Helicobacter pylori Protocols

Introduction:

Medical Significance of H. pylori

Martin J. Blaser Until the discovery of Helicobacterpylori in 1982, the normal human stomach was generally considered to be sterile, or transiently populated by oropharyngeal bacteria carried there by peristalsis. However, we now know that from one-third to one-half of the human population carries H pylori, and that once infected, most persons remain infected for decades, if not for life (I). H. pylori infection is almost invariably associated with gastric inflammation and induces the most common inflammatory lesion, known as chronic superficial gastritis (2). This lesion is so common, especially among the elderly, that it was thought to represent the aging stomach, but we now know that eradication of H. pylon from the stomach results in clearance of this pathology (3). However, based on present knowledge, those persons with H. pylori-induced chronic superficial gastritis do not have any symptoms. Until the discovery of H pyZori, chronic superficial gastritis was essentially a medical curiosity, appreciated only by pathologists who could not agree about the nosology of gastritis.

H. pylori Association with Gastric Diseases Now it is clear that H. py2ori infection is highly associatedwith a number of the most important diseases of the upper gastrointestinal tract. These include duodenal and gastric ulcers, gastric adenocarcinoma, and non-Hodgkin’s lymphomas of the stomach (Table 1). Although H pylori infection IS not present in all cases of these disorders (Fig. l), there IS a substantial body of evidence that it is involved in the causation of each of these (4-12). Conversely, there are other diseasesof the upper gastrointestinal tract with which H pylori infection is not associated, including reflux esophagitis, Barrett’s esophagus, bile reflux gastritis, pernicious anemia, and the gastropathy associated with use of nonsteroidal anti-inflammatory drugs (NSAIDS). Thus, H. pylori is Ed&d

From Methods 1r-1Molecular Mehne, Helicobacter pylon Protocols by. C. L. Clayton and H L T. Mobley Humana Press Inc , Totowa,

I

NJ

2

Blaser Table 1 Association of H. py/ori Infection with Important Inflammatory and Neoplastic Conditions of the Upper Gastrointestinal Tract Estimated Condition odds ratio Refs. Peptrculcer disease 4-12 43 Adenocarcinomaof the stomach 8 6-9 6 10,rr Non-Hodgkin’s lymphoma of the stomach Barrett’s esophagus

1

12

Uninfected

Fig. 1. H pylori infection, ulcer disease,atrophic gastritis, and gastric neoplasra. Associationof H. pylon infection andmedical condttronsof the upper gastrointestinal tract as shown m a Venn diagram. DU, duodenal ulcer; GU, gastric ulcer; CAG, chronic atrophic gastntls; GC, gastrx cancer (adenocarcinoma of the stomach), L,

Non-Hodgkin’s lymphoma of the stomach. related to risk of certain diseases of the upper gastrointestinal tract, but 1snot nonspecifically associated with them all. Not All H. pylorblnfected Persons Become Ill A seeming paradox is that most infected persons never develop any consequences of colonization with H pyZori during the decades-long course of the infection; this is not unusual for other host-adapted agents of chronic mfectious diseases. For example, Mycobacterium tuberculosis infects about one-third of the world’s population, as well, yet rt has become clear that less than 10% of normal hosts will develop any clinical consequences (most commonly, pulmonary tuberculosis) of the infection. The other 90% have no clim-

Medical Significance of H. pylori H. pylon

3

Mection

Weeks-Months 0

Fig. 2. Natural history of H pylorr infection. Within months after H pylon is acquired, infected persons develop chronic superfictal gastrrtis. This persists, essentially for life, in most infected persons. However, one subgroup of infected persons develops peptic ulceration, another develops chrome (multifocal) atrophic gastritis, which may lead to adenocarcinoma of the stomach, and a third group develops lymphoproliferative disorders, such as MALT-type lymphomas. Reprinted with permission from ref. 23.

tally apparent consequences during their lifetime. Examining the clinical data another way, we find that among persons with chronic destructive pulmonary inflammation, with cavities, the vast majority are caused by A4. tuberculosis, and a few result from other disorders, such as Wegener’s granulomatosis and chronic fungal infections. This paradigm nearly parallels that known for H. pylori and peptic ulceration. Although 95% of those with idiopathic (non-NSAID or Zollinger Ellison-induced) ulcers (23). Similarly, even in parts of the world where gastric cancer is hyperendemic, less than 5% of H. pylori-infected persons develop these neoplasms. However, in parallel with the ulcer data, attributable risk analysis suggests that H. pylori is responsible for 60430% of these tumors (9). Natural History of H. pylori-Infection From studies conducted by hundreds of investigators all over the world, it has become clear that H pylori infection has a natural htstory (Fig. 2). Within

4

Blaser

weeks after the organism is acquired, chronic superficial gastritis develops (14,15). In most persons, this lesion persists, without any consequences, for the duration of the infection (2). One subgroup of patients develops peptic ulcer disease, another develops a variety of lymphoproliferative disorders, and in others, chronic superficial gastritis progresses to atrophic gastritis. The latter lesion is important because it is the major recognized risk factor for the development of gastric cancer. The development of cancer probably takes 3-6 decades of infection to manifest. An important aspect in understanding H. pylori infection is to consider why this diversity of outcomes develops. The determinants for specific consequences of infection may be divided into four categories. First, there may be differences among H. pylori strains m their virulence potential. Second, there may be important differences among infected hosts in their genetic characteristics. Third, exposure of hosts to environmental cofactors (such as smoking or particular dietary substances) differs and may relate to outcome of H. pylori infection. Finally, as with other pathogenic agents infecting humans, the age at which H. pylorz is acquired may affect outcome. In fact, there now is evidence for each of these possibilities (1619). The determinants of outcome are multifactorial and complex. Management of H. py/ori Infection Over the last decade, important advances have permitted the ready and accurate diagnosis of H pylori mfection (20). Similarly, antimicrobial therapy has now improved to the point that safe and well-tolerated relatively short term (7-10 d) regimens are more than 90% effective (22,22). It now is clear that H. pylori-infected persons who develop peptic ulceration or (MALT) lymphomas should be treated, and perhaps an argument can be made for those with family or exposure histories that point to high risk of developing gastric cancer. However, at this time it is premature to consider eradicating H. pylori infection in everyone. Defimtion of risk factors for the development of specific diseases will facilitate this quest. Similarly, investigators also should focus on whether this well-adapted and persistent bacterium may have any beneficial effects for humans. If such properties are discovered, or if there are commensal H pylori, there may be individuals in whom infection should be maintained, Conclusions In the 14 yr since the discovery of H. pylori, there has been remarkable scientific progress. We have learned about the role of these organisms m important human diseases, methods of diagnosis are readily available, and effective treatments have been found. An improved understanding of the pathogenesis of infection will facilitate further medical progress.

Medical Significance of H. pylori

5

References 1. Pounder, R. E. and Ng, D. (1995) The prevalence of Helicobacterpylori infection m different countries. Aliment Pharmacol. Ther 9, S33-S39. 2. Dooley, C P., Fitzgibbons, P. L., Cohen, H , Appleman, M. D., Perez-Perez, G I., and Blaser, M J. (1989) Prevalence of Helicobacter pylori infection and histologic gastritis in asymptomatic persons. N. Engl. J Med 321, 1562-l 566 3. Blaser, M J. (1990) Helicobacterpylori and the pathogenesis of gastroduodenal inflammation. J. Infect. Dzs. 161,626-633 4. Stpponen, P., Helske, T., Jarvmen, P., Hyvarinen, H., Seppala, K., and Smrala, M. (1994) Fall in the prevalence of chronic gastritis over 15 years: analysts of outpatient series in Finland from 1977, 1985, and 1992. Gut 35, 1167-l 171. 5. Nomura, A., Stemmermann, G. N., Chyou, P., Perez-Perez, G. I , and Blaser, M. J. (1994) Helicobacterpylori infection and the risk for duodenal and gastric ulceration Ann Intern Med. 120,977-981. 6. Nomura, A., Stemmermann, G. N., Chyou, P., Kato, I., Perez-Perez, G I , and Blaser, M. J. (199 1) Helicobacterpylori infection and gastric carcmoma m a population of Japanese-Americans in Hawaii N. Engl J, Med 325, 1132-l 136 7 Parsonnet, J., Friedman, G. D., Vandersteen, D. P , et al. (1991) Helzcobacter pylon infection and the risk of gastric carcinoma. N Engl. J Med 325, 1127-l 13 1 8. Forman, D., Newell, D. G., Fullerton, F., Yarnell, J. W., Stacey, A. R., Wald, N., and Sitas, F. (1991) Association between infection with Helicobacterpylori and risk of gastric cancer: evidence from a prospective investigation. BMJ 302, 1302-1305. 9. Forman, D. (1995) The prevalence ofHelicobacterpylorz infection in gastric cancer. Aliment. Pharmacol Ther 9, S71-S76. 10. Parsonnet, J., Hansen, S., Rodriguez, L., Gelb, A. B , Warnke, R. A., Jellum, E., Orentretch, N., Vogelman, J. H., and Friedman, G. D. (1994) Helxobacterpylori infection and gastric lymphoma. N Engl J Med 330, 1267-1271. 11. Wotherspoon, A. C., Doglioni, C., Diss, T. C., et al. (1993) Regression of primary low-grade B-cell lymphoma of mucosa-associated lymphoid tissue type after eradication of Helicobacterpylori. Lancet 342,575-577. 12. Blaser, M. J., Perez-Perez, G. I., Lindenbaum, J., Schneidman, D., VanDeventer, G., Marin-Sorenson, M., and Weinstein, W. M. (1991) Association of mfection due to Helwobacter pylori with specific upper gastrointestinal pathology. Rev Injkt.

Dis 13, S704-S708.

13. Kuipers, E. J., Thijs, J. C., and Festen, H. P. M. (1995) The prevalence of Helicobacterpylori in peptic ulcer disease. Aliment. Pharmacol Ther 9, S59-S69. 14. Morris, A. and Nicholson, G. (1987) Ingestion of Campylobacterpylondrs causes gastritis and raised fasting gastric pH. Am. J, Gastroenterol. 82, 192-l 99. 15. Morris, A. J., Ali, M. R , Nicholson, G. I., Perez-Perez, G I., and Blaser, M. J. (199 1) Long term follow-up of voluntary ingestion of Helicobacter pylon. Ann. Intern.Med

114,662,663.

16 Figura, N., Guglielmetti, P., Rossolini, A., Barberi, A , Cusi, G., Musmanno, R. A., Russt, M., and Quarantas, S. (1989) Cytotoxin production by Campylobacter

Blaser

6

strains isolated from patients with peptic ulcers and from patients with gastritis only. J, Clin Mlcroblol. 27,225,226. T. L., Dooley, C. P., and Blaser, M. J. (1990) Characterization of and serologrc response to proteins m Helrcobacter pylorl broth culture superwith vacuolizmg cytotoxm activity. Infect. Immun. 58,603-6 10. H. M., Engstrand, L., Pedersen, N L., and Graham, D. Y (1994) Helicobacter pylorl infection: genetic and environmental influences A study of twins. Ann Intern Med 120,982-986. Blaser, M. J., Perez-Perez, G. I., Kleanthous, H., Cover, T. L., Peek, R. M , Chyou, P. H., Stemmerman, G. N., and Nomura, A. (1995) Infection with Hekobacter pylon strains possessmg cagA associated with an increased risk of developing adenocarcinoma of the stomach. CancerRes. 55,2 11 l-2 115. Feldman, R. A. and Evans, S. J. W (1995) Accuracy of diagnostic methods used for epidemrologmal studies of Helrcobacterpylorl. Aliment Pharmacol. Ther. 9, S21S31. de Boer, W., Driessen, W., Jansz, A., and Tytgat, G (1995) Effect of acid suppression on efficacy of treatment for Hellcobacter pylorz infection. Lancet 345, 817-820. Lind, T., vanzanten, S. J. O., Unge, P., et al. (1995) The Mach I study: optimal one-week treatment for H pylorl defined? Gut 37, A4. Blaser, M. J. and Parsonnet, J (1994) Parasitism by the “slow” bacterium Helicobacter pylon leads to altered gastric homeostasis and neoplasia. J. Clan pylori

chronic 17 Cover, human natants 18. Malaty,

19.

20.

21.

22. 23.

Invest

94,443

Detection of H. pyhi Infection by Biopsy Urease, Histology, and Culture Stewart Goodwin 1. Introduction Detection of Helicobacter pylori in specimens obtained by endoscopy requires the gastroenterologist to select suitable patients for endoscopy, and to take an adequate number of biopsy specimens; there must also be correct cleaning of the biopsy forceps. 7.7. Selection

of Patients for Endoscopy

Any antibiotic active against H. pylori will cause a reduction in the numbers of bacteria in the stomach. If the patient has been treated with such an antibiotic or has ingested an over-the-counter preparation containing bismuth such as Pepto-Bismol, in the preceding 4 wk before the endoscopy, then endoscopy must be delayed until 4 wk after the last antibiotic treatment (1). Also, if the patient has received a drug that reduces the acid in the stomach and raises the pH, this will affect the area of the stomach to be biopsied. Hz-receptor antagonists (ranitidine and cimetidine) raise the gastric pH, but proton pump inhibitors, such as omeprazole and lanzoprazole, raise the gastric pH to a higher level. When the gastric pH has been raised, H. pylon is not so common in the antrum of the stomach, but is found in the corpus more frequently than in patients who have not received such drugs (2); m such patients, at least one biopsy specimen must be taken from the corpus and the antrum. Various chemical agents used during the endoscopy procedure can be antibacterial to H. pylori. Benzocaine has a minimum inhibitory concentration for From* Methods /n Molecular Me&he, Hellcobacter pylorr Protocols Edlted by* C. L Clayton and H. L T. Mobley Humana Press Inc., Totowa,

7

NJ

8

Goodwin

H. pylori of 0.14-0.7 mg/L, and simethicone is also inhibitory. Lidocame is not inhibitory to H. py2ori. 1.2. Cleaning of Endoscopy Forceps An efficient cleaning and disinfecting method for the biopsy forceps must always be employed between each patient. If disinfection is inadequate, then a subsequent patient(s) may become infected with H. pylon’ from a previous patient. However, rarely will a positive culture be obtained from such inadequately disinfected forceps. Another danger is that a strong disinfectant, such as glutaraldehyde, may remain on the biopsy forceps and kill the bacteria, rendering the culture falsely negative. Most modern endoscopes are purchased with a machine that fully disinfects and cleans the instrument between patients. 1.3. The Number of Biopsy Specimens to be Obtained and Their Sites The number of specimens to be taken is determined first by the range of tests for H. pylori that will be requested. It is recommended that one specimen be taken from the antrum for the rapid biopsy urease test. Histology is always performed on one or more specimens, but some physicians avoid specimens for microbiological processing because these are an extra expense. If several biopsies are taken for histology and one specimen for the biopsy urease test, detection of H. pylori may not need a specimen for microbiological processing. However, an important opportunity with endoscopy is to determine the antibiotic sensitivity of the isolate to guide correct antibiotic treatment. Because H. pylori is not evenly distributed around the stomach surface, but has a patchy distribution, the more biopsy specimens that are taken, the less likely it is that the presence of H. pylori will be missed, and these extra specimens can be processed microbiologically. At least one specimen for histology should be taken from the antrum and one from the corpus. At least one specimen from the antrum should be sent for microbiological processing and, to be sure of detecting the organism, probably two specimens are required, so that one can be processed for microscopy and the other for culture. Second, if the patient has received a proton pump inhibitor, at least one biopsy specimen must be taken from the corpus (2). All specimens should be kept at 4OCbefore and after transport to the laboratory. Detection of H. pylori involves a biopsy urease test, a histological process that ensures detection of spiral bacteria, an efficient transport medium for microbiology, correct microbiological microscopy and inoculum preparation for culture on solid media with the correct incubation atmosphere, and length of incubation. These are described in the followmg sections.

Detection by Biopsy, Histology, Culture

9

2. Materials

2.1. Tests in the Endoscopy Unit Biopsy Urease Tests. These testsdetect the presence of preformed urease produced by H pylon in the biopsy specimen; growth of the organisms is not required. The urease of H. pylori hydrolyzes urea in the medium, with the production of ammonium ions, which raise the pH and change the color of the indicator. 2.1.1.

2.1.1 .~.COMMERCIAL CL0 TEST

Delta-West Ltd., Bentley, Western Australia. This 2% urea unbuffered gel preparation has proved its stability and reliability, and has been found to be very accurate, when read after 3 h, with a sensitivity of 90% and a specificity of 100% (3). However, a minimal amount of 10,000 cm/ml is necessary to obtain positivity, and the sensitivity of all urease tests drops to 50% m patients after antimicrobial treatment (4). 2.1 1 2. LIQUID UREA MEDIUM

This can be prepared in the microbiology laboratory and is usually cheaper than the CL0 test. The concentration of urea has been varied by many workers, but a medium with 6% urea is probably the most rapid and reliable. In theory, higher concentrations may result in inhibition of the urease of H. pylori. The ingredients for the liquid urea medium are: 60 g/L urea; 0.012 g/L phenol red; 2 g/L KH,PO,; 1 g/L peptone; 5 g/L NaCl; 10 g/L glucose, in distilled water. Three-milliliter aliquots are dispensed into small containers with screw caps. This solution, without buffer, produces the most rapid result, but is unstable and should be made up fresh at least once a week. 2.1.2. Simple Carboifuchsin Staining Carbolfuchsin solution prepared as follows: 0.4 g basic fuchsin, 2 g phenol crystals, 4 mL absolute alcohol (made up to 100 mL in distilled water); absolute alcohol, glass slides. 2.2.

Histology (see Note I)

2.2.1. Imprint Cytology Requirements: a glass microscopy slide, sterile needle, absolute alcohol, 2% Giemsa solution (Giemsa concentrate, Difco, Detroit, MI) in distilled water. 2.2.2. Modified Giemsa Method Requirements: 2% (v/v) Giemsa (see Section 2.2.1.) solution m distilled water, absolute alcohol.

Goodwin

10 2.2.3. Simple Carbolfuchsin Staining

Requirements: 4% (v/v) neutral formaldehyde, distilled water, and carbolfuchsin solution prepared as in Section 2.1.2. 2.2.4. lmmunohistochemical

Staining

Requirements: Antibodies MABOOl and MAB002 (Bioproducts for Science, Indianapolis, IN). 2.3. Microbiological Processing of Specimens 2.3.1. Transport Media to the Laboratory (see Note 1) 1. H. pylon rapidly losesviability at room temperaturein 0.9% (w/v) NaCl, so this should not be used unless the specimen can be received m the laboratory within 1 h. It should be dispensed in 0 5-mL ahquots. 2 Brucella broth (Oxoid CM169 [Basmgstoke, UK]; BBL 11088 [Becton Dickinson, Cockeysville, MD]; Difco 0495). These should be made up followmg the mstructions of the maker, with the addition of 0 5% (w/v) bovine serum albumin. Three-milhllter aliquots should be dispensed into small containers with screw caps

2.3.2. Preparation of lnoculum for Solid Media Requirements: sterile forceps, normal saline, a ground-glass grinder or an electric

tissue grinder

(Potter-Elvehjem

homogenizer

[Citenco,

Boreham

Wood, Hertsfordshire, UK]) or a mechanical Teflon homogenizer, glutaraldehyde or alcohol (to clean the blades and parts of the grinder), and carbolfuchsin solution (see Section 2.2.3.). 2.3.3. Phase Contrast Microscopy A phase contrast microscope is required. 2.3.4. Culture on Solid Media (see Note 3) 1. Most authorities recommend maculation of one selective medium with antibiotics, and also a medium without antibiotics. Freshly prepared moist plates are essential to obtain the maximal isolation rate of H pylori, although if the fresh plates are wrapped immediately in a plastic bag and stored at 4’C, they can be used for 2 wk. Whole blood or blood lysed by saponm (7%) is incorporated m all

media, and must be thoroughly mixed with the liquid agar to ensure an even distribution. Blood lysed by freezing or thawing is less successful for the growth of H. pylori.

2. Blood agar bases: Brain-heart mfusion (BHI) agar (Oxoid CM375B; BBL 99069 [Cockeysville, MD]; Difco 0418); Brucella agar (Oxoid CM169; BBL 11088; Difco 0964); Columbia agar (Oxoid CM33 1; BBL 11125 or Difco 0792).

Detection by Biopsy, Histology, Culture

I7

3. Antibrotic supplements for selective media (see Note 4): 10 mg/L vancomycm, 10 mg/L amphotericin, and 5 mg/L cefsulodin. If cefsulodin is not available, it may be replaced by. 5 mg/L trimethroprim, although, theoretically, this is not useful, because the rich medium has a high thymidme concentration; however, trimethroprim may be synergistic with other antibiotics in the combination. Some workers add 5 mg/L trimethropim to the other three agents. 4. Commercially poured plates of selective medium: Biomerieux 43263 Pylori medium (Marcy-L’Etoile, France) does not contain blood, but is good for primary isolation of H. pylori. Although commercial media often become too dried to be used successfully, Biomerieux’s Pylori medium has been effective in the United Arab Emirates, one of the driest and most remote spots for growth of H. pylon. 5 Other supplements: Many workers recommend 1% IsoVitalex (BBL 11875) for optimum isolation. Triphenyl-tetrazolium chloride (40 mg/L) is added to produce golden-colored colonies of H. pylon, which are easier to detect tf the plate has many commensals on it.

2.3.5. Incubation Jars, Atmosphere, and Temperature (see Note 5) 1. The correct atmosphere contains 5-7% 02, and 5-10% COz. An anaerobic jar (without catalyst) should be evacuated to 220 mm Hg and tilled with an anaerobic gas mixture containing 10% CO*, 10% Hz, and 80% Nz, which gives an atmosphere of 5% 02, 7% COz, 8% Hz, and 18% Nz. This atmosphere gives good growth, and, occastonally, better growth than a standard campylobacter gas mixture (see item 2). 2. Gas generating kits for the campylobacter system (Oxoid BR60 and BR56; BBL Campypak 7 1034). 3. In the absence of items 1 or 2, an anaerobic gas-generating kit (BBL 6065 1; Oxoid AN025A or AN035A) may be used and the catalyst palladium pellets removed; however, this method has not been directly compared with either item. 4. It is important to maintain a humid atmosphere in the jars by placing motst filter paper or moist cotton wool at the bottom of the jar. However, the moist atmosphere allows growth of fungi and other contaminants, and so, between each use, the jars should be well cleaned and wiped around with a disinfectant, such as 70% propyl alcohol. 5. The temperature of incubation should be 37’C.

2.3.6. Identification of H. pylori in the Clinical Laboratory Standard laboratory tests can be used for oxidase, catalase, and urease. Requirements: oxidase reagent (1% N,N,NZ,NZ, tetramethyl-p-phenylenediamine dihydrochloride); catalase reagent (3% [v/v] hydrogen peroxide); urease reagent (2 g urea, 0.5% [w/v] phenol red, 20 mg sodium azide) made up to 100 mL with O.OlM phosphate buffer and adjusted to pH 5.5; or urea test broth (BBL 11797; Oxoid SRO2OK).

12

Goodwin

2.3.7. Growth in Liquid Media Requirements: BHI broth (Oxoid CM225B; BBL 99070); or Brucella broth (Oxoid CM169; BBL 11088), with 5% v/v fetal calf or horse serum (some people find 1% satisfactory). A container with a screw-top lid, such as a medmm-sized glass bottle or a urine specimen container, or a conical flask with a cotton wool plug, may be put in an anaerobic jar with a micro aerobic atmosphere and shaken in an incubator. Alternatively, a conical flask may be shaken in a gyratory water bath shaker at 150 rpm (New Brunswick Scientific, Edison, NJ) fitted with a gassing hood connected to a gas mixture of 10% COZ, 5% 02, and 5% HZ. This will produce maxtmum growth after 72 h. Growth without shaking can be achieved with a tissue culture flask, such as a tissue culture flask (Falcon, Los Angeles, CA). Such a tissue culture flask can be incubated in a variable atmosphere incubator containing 2% HZ, 5% 02, 7% C02, and 86% N2, with 90% relative humidity. The flask should contain 20 mL of BHI Broth (see above) supplemented with 1% (v/v) horse serum. Alternatively, a lOO-mL tissue culture flask 1sused with 10 mL BHI broth. Both flasks can be incubated at 37OC in a 10% CO2 incubator with 98% relative humidity (see Note 5). 3. Methods and Identification 3.1. Tests in the Endoscopy Unit 3.1.1. Biopsy &ease Tests 3.1.1.1.

COMMERCIAL CL0 TEST

Peel back the label covering the well and insert the biopsy into the agar by means of a sterile needle. Replace the label and insert the plastic container into a trouser pocket so that the warmth of the human body will promote the most rapid result. Examine the test for up to 1 h. The color of the medium changes from yellow to red or purple, tf H. pyhri urease 1spresent. This test is extremely valuable to an endoscoptst, because tt can give a posmve result for H pylon before the patient leaves the clinic. The sensitivity of this test is at least 75%. However, a negative test after 1 h may result from the fact that the area biopsted contams fewer H. pyhi; either histology or microbiology will prove posittve. Some people keep the CL0 test for 24 h, but at that time a false postttve biopsy urease test may be caused by the presence of other rapid ureasepositive organisms, such as streptococci and staphylococci. This underlines the need to confirm a positive biopsy urease test by either histology or microbiology or a urea-breath test. However, a positive result within 1 h of sampling is highly specific for H. pylori.

Detection by Biopsy, Histology, Culture

13

3.1 .I .2. LIQUID UREA MEDIUM Insert the biopsy specimen into the liquid urea medium and observe in the endoscopy unit, preferably in an incubator at 37OC.The container is kept until the patient leaves the endoscopy unit, and then the container can be sent to the microbiology laboratory to be incubated at 37°C usually overnight. If the medium changes from yellow to red, the presence of H pylon is indicated, but the specificity of the test is higher after 4 h than after 24 h for the same reasons that cause a false positive biopsy urease test as mentioned in Section 3-l. 1.1, 3.1.2. Biopsy Carbolfuschsin

Stain in the Endoscopy Unit

In the endoscopy unit, smear a biopsy specimen on a glass slide, fix with absolute alcohol, and stain with carbolfuchsm for 3-5 min; rinse slide in tap water and examine under the microscope. In experiences hands, this 1sa sensltive technique that gives the investigator an immediate result. If the smear is made with sterile forceps and slide, the biopsy can be subsequently submitted for culture, or fixed in formalin for histology. Gastric mucosal brushings may be stained m a similar fashion. 3.2. Histology (see Note 6) 3.2.1. Imprint Cytology Put the biopsy specimen on a glass microscopy slide and lightly press with a sterile needle. Fix the imprint in absolute alcohol and stain by the modified Giemsa method. Put the 2% (w/v) Giemsa solution on the slide for 20 mm at room temperature; rinse the slide m tap water and examine under the microscope. This method has been found to have a sensitivity and specificity equal to that of conventional histology. 3.2.2. Histology-Modified

Giemsa Method

Place the biopsy specimen in 4% (v/v) neutral formaldehyde. Paraffin embedded sections are routmely dewaxed and taken to water, and then incubated in 2% Giemsa solution m distilled water for 30 mm at room temperature. After rinsing in tap water, quickly dehydrate the section through ethanol solutions, clear with xylene, mount in dixylol-phthalate mountant (DPX), and examine under an oil immersion lens (x 1000). H pylori appears as blue, curved, or loosely spiral organisms. 3.2.3. Histology4imple

Carbolfuchsin Staining

Place the biopsy specimen m 4% (v/v) neutral formaldehyde. Transfer dewaxed tissue sections by stages into distilled water and stain for 5 min in carbolfuchsin solution. After rinsing in tap water, briefly decolorize the sec-

14

Goodwin

tions with acetone and examme under an oil nnmersion lens (x 1000). The stain shows I-J pylori as dark red, curved, or loosely spiral organisms against a reddish background. 3.2.4. lmmunohistochemical

Staining

Antibody MABOOl reactswith H pylori in formahn-fixed biopsy specimens. Another section is exposed to MAB002, which is specific for Cumpylobacter coli and should be negative in the biopsy specimen. If the antibody MABOOl produces a positive result and MAB002 produces a negative result, this 1shighly suggestive of H pylori in the section. However if a pathologist is well-expertenced m detecting H pylorz with standard stains m Sections 3.2.2. or 3.2 3., this method provides no additional information.

3.3. Microbiological

Processing

3.3.1. Microscopy Remove the biopsy specimen(s) from the transport medium with sterile forceps and place in 0.5 mL of saline or the transporting fluid. The sample used for microscopy must not be ground in a grinder first, because detection by microscopy 1s impossible after such a procedure. Cut off a small corner of the biopsy for microscopy and smear over a slide. If two specimens are sent for microbiologrcal processmg, a small portion of each should be smeared on the slide. 3.3.1 .l. MODIFIED GRAM STAIN

Standard Gram stain constituents are used, except that carbolfuchsm 1sused in place of safranin. Inspect the slide under oil mnnerston for the presence of typical curved bacteria. Provided the smear has been well spread over the slide, this test produces a very high rate of successm detecting H pylori (see Note 5). Gram stain of H. pylorz colonies shows a range of shapes from short rods to curved rods and occasionally S-shaped spirals. 3.3.1.2.

PHASE CONTRAST MICROSCOPY

Place a finely minced biopsy homogenate in a few drops of saline, cover with a cover slip, and examine with x400 magnification. If pylori 1sidentified as characteristic black curved rods. 3.3.2.

Culture (see Note 7)

3.3 2.1. PREPARATION OF INOCULUM AND INOCULATION OF AGAR PLATES 1.

Remove the biopsy specimensfrom the transport medium with sterile forceps and place m 0.5 mL of saline or the transporting fluid.

Detection by Biopsy, Histology, Culture

15

2. Grind m one of the grinders detailed in Section 2.3.2 -in an electric tissue grinder for not more than 10-20 s at 10,000 rpm, or in the Teflon homogenizer for 1 min; rf necessary, add 0.5-l mL of nutrient broth to suspend the ground material. Grinding is essential to achieve the largest number of colonies. An alternative to grinding is to macerate the specimen between two broken-tipped Pasteur pipets. 3. Place two drops of homogenate on one agar plate of selective medium and on one plate of a nonselective medium. Spread over the whole surface of the plate, either with wire loop, or with a sterile glass rod with a rounded bent end (“hockeysttck”), or a bent Pasteur pipet with a similar end. 4. With glutaraldehyde or alcohol, clean and disinfect the blades and parts of the grinder that would have been exposed to biopsy material. Ground glass grinders must be cleaned and autoclaved.

3.3.2.2. ATMOSPHERES FOR INCUBATION Put the plates in an anaerobic jar filled with the atmosphere described in Section 2.3.4., or place a gas generating kit in the jar and add the required water to the sachet.

3.3.2.3. LENGTH OF INCUBATION AND REGENERATION OF THE MICROAEROBIC ATMOSPHERE

1. The use of gas generatmg kits lacks complete control of the atmosphere, and requires replacement of the gas pack sachet every 2 d and whenever the jar is opened. Up to 10% of commercial gas generating envelopes may be deficient; therefore many workers simultaneously use two envelopes instead of one 2. Incubate the plates at 37°C for up to 7 d for primary isolation, but if culture is used to determine success of cure of H. pylori infection, then the plates must be incubated for up to 10 d Examine the plates after 2 or 3 d and then daily (see Note 8).

3.3.3. identification of H. pylori (see Note 9) Colonies of H. pylori are small (2 mm) and translucent. Gram stain shows H. pylori as Gram-negative curved rods and H. pylori is oxidase, catalase, and urease positive: 1. To test for oxidase, transfer a colony with a platinum loop onto filter paper soaked with oxidase reagent. A positive reaction is indicated by a deep blue/black color. 2. Catalase activity is detected by putting a loopful of a colony mto a drop of 3% (v/v) hydrogen peroxide on a glass slide, and a positive reaction is seen by bubbles of oxygen. Samples from blood agar may be weakly false-posmve because of blood.

3. Ureaseactivity ISdetectedby inoculating a loopful of the organism into 0.25 mL of urea broth and incubating at 37°C m a water bath. Alternattvely, a microtiter tray with urea reagent can be used A positive red color occurs in a few minutes.

16

Goodwin

3.3.4. Growth in Liquid Media (see Note 10) Incubate the containers with screw cap lids loosened to allow the microaerobic atmosphere to enter the containers. Put the containers into ajar with a mlcroaerobic atmosphere (see Section 2.3.5.). Place the jar on a reciprocating shaker in an incubator at 37°C. A heavy growth 1sobtained after 18 h mcubatlon; 0.1 mL of this broth can be used as the inoculum for other liquid media. When cultures are first put into the liquid medium, not all the organisms show motility; two more transfers at daily mtervals into liquid medium achieve the highest percentage of motile and presumably viable organisms. 3.3.5. Subculture of Single Colonies of H. pylori Subculture of single colonies of H. pylori 1sdifficult, but can be achieved if one colony is subcultured into an area of only 1 cm diameter. Incubate the plate for 2-3 d, and then spread the growth over a wide area. If a single colony 1s spread over a wide area on a plate, growth may not occur. Subculture of strains in the laboratory is possible in a candle jar with a source of humidity (see Section 2.3.2.5.), which may save the expense of gas-generating kits (5). 3.3.6. Preservation of H. pylori Isolates 1. Lyophilization: An isolatefreshly grown for 2-3 d is required. The superiority of the liquid state (L drying) techmque over freeze drying has been discussedm detail by Goodwin et al. (6). 2 In liquid nitrogen: BHI broth (Oxold CM225B; BBL 11060; Difco 0037) with 20% (v/v) glycerol will preserve culturesfor more than 6 yr. 3. At -70°C: 1 mL of 1% peptonewater (Oxoid CM509; BBL 99095; Dlfco 1807) with 25% glycerol is inoculatedwith a fresh culture of H pylon at -7OOC.This will preserveH pylon for at least 1 yr. 4 In semi-sohdagarat 37’C: BHI agar0.17% (w/v) (Oxoid CM375B; BBL 99069, Difco 0148), with 10%horse serumand 0.25%yeastextract,is inoculated with a freshly grown strain of H pylori. This semisolidmedium is left m a CO* mcubator at 37°C with 98% humidity Cultures can be subcultured monthly into the samemedium 4. Notes 1. Many stains have been used to detect H. pylori in histological sections. The Brown-Hopps stain uses Gallego’s differentiating solution, a 1.5% tartrazme solution, and basic fuchsm as part of the counter-staining technique. This stain shows the mucus as yellow, and H pylon as purple or red curved bacteria, often within the mucus layer; a microscope without a daylight blue filter should be used. The Warthin-Starry stain is excellent at showing H. pylori, but it is tlmeconsuming and costly and requires an experienced techmclan. Also, the silver

Detection

2.

3.

4.

5.

6. 7

8. 9

10.

by Biopsy,

Histology,

Culture

impregnation gives H. pylori a granular appearance, which can lead to difficulty in differentiating it from silver precipitate in the mucus layer, consequently, with a false-positive biopsy result. Alternative transport media: a. 20% glucose solution: 20 g glucose is dissolved in 100 mL sterile distilled water and 3-mL aliquots are dispensed into small, screw-cap containers b. Semisolid Stuart’s Transport Medium (Oxoid CM1 11; BBL 11743; Dtfco 062 1) can also be used successfully for transport. H pylon will survive for 6 h in this medium at temperatures up to 15°C and for up to 48 h if kept at 4°C. c. BHI broth (see Section 3.3.6.) with 10% (v/v) glycerol. H. pylon will remain viable for several months, if biopsy specimens are frozen in this solution at -7OY!. However, some workers have found that H. pylon does not remain viable m such a frozen sample. Alternative supplements have been used that enable growth of H. pylori in the absence of blood: 0.2% (w/v) charcoal, 1% (w/v) corn starch, 10% (v/v) egg yolk emulsion Oxoid SR47, and 5-10% (v/v) horse, bovine, or fetal calf serum. But, for primary rsolatron, a blood-containing medium is probably the most reliable. Nalidixic acid is contraindicated, because 14% strains are sensitive to it. Colistin and polymyxin mhrbrt 5% of strains and should not be used. Sknrow’s supplement contains polymyxm, so there will be mhrbrtion growth of some strains. The supplement used for Cumpylobacter isolation in connectton with Skit-row’s medium is FBP, but this should not be used for H. pylori, because sodium metabisulfite is inhibitory to H. pylon. Incubation of plates in a CO2 incubator does not produce as high a primary isolation rate as incubation m aJar with a controlled atmosphere. If a COZ incubator IS used for static liquid culture or maintenance of strains in semisolid medium, the humidity must be maintained at 95%, but this also allows fungi to grow, and the incubator must be cleaned regularly with an antifungal agent. The hematoxylin and eosin stain falls to reveal H pylon m a significant proportion of specimens. The abtlity of a laboratory to detect H pylon by culture usually increases with increasing skill of the workers. The most skilled laboratories are able to culture H. pylori m 10% of specimens in which spiral bacteria were not seen on Gram stain. Because H. pylon is so frequently isolated from endoscopy specimens, most laboratories are lulled mto a false sense of efficiency; they should determine how frequently they can culture H pylori when it is not seen on Gram stain. Isolates of H. pylori on agar must not be left in air for longer than 45 mm or they may die. The plates must be returned to the microaerobic atmosphere before 45 mm If there is any doubt about the identification of H pylori, then further tests may be performed; y-glutamyl ammopeptidase and alkaline phosphatase enzymatic activity (Rosco Diatabs) will show a positive test for H. pylori. Many workers add 0 25% (w/v) yeast extract as a supplement to achieve maximum growth.

18

Goodwin

References 1. Marshall,

D J (1993) Treatment strategies for Helzcobacter pylori mfection. Clm. North Am 22, 183-198. 2. Stolte, M. and Bethke, B. (1990) Elimination of Helicobacter pylorz under treatment with omeprazole. Zekwhrift Fur Gastroenterologze 28,27 l-274. 3. Morris, A., McIntyre, D., Rose, T., and Nicholson, G. (1986) Rapid diagnosis of Campylobacterpyloridis infection. Lancet i, 149. 4. Deltenre, M., Glupczynski, Y., De Prez, C., Nyst, J. F., Burette, A., Labbe, M., Jonas, C., and Dekoster, E. (1989) The reliability of urease tests, histology and culture in the diagnosis of Camphylobacterpylorr infection. Stand J Gastroenterol Gastroenterol

24(Suppl.

160), 19-24.

5. Holloway, Y., Schiphuis, J., Weites, L., and Snijder, J. A. M. (1994) Luxuriant growth of Helrcobacterpylorz and Campylobacter species in candle jars after primary isolation. Eur. J. Urn. Microbtol. Infect Du. 13,831. 6. Goodwin, C. S., McCullough, C., and Boehm, J. (1989) Successful lyophilization of Campylobacter pylori and spiral organisms from the stomachs of animals. Pathology 21,227-229.

3 Serology and Urea Breath Test in the Diagnosis of H. pylori Infection Peter R. Hawtin 1. Introduction The selectivity of Helicobacterpylori for the antral mucosa of the stomach and gastric metaplasia in the duodenum combined with the inaccessibility of those sites has hindered the investigation of this infection. Consequently, the study of the epidemiology and treatment of II. pylori has relied heavily on indirect methods of detecting infection. Therefore, the accuracy of findings from such studies, particularly those with small populations, was highly dependent on the sensitivity and specificity of the detection systemsemployed. The calculation of sensitivity, specificity, and positive and negative values of a test requires a “gold standard.” The choice of gold standard, against which to validate the test, is important as it will have its own margins of error. Ideally, the gold standard and its margins of error should be identified when quoting sensitivity and specificity rates in validation of any new diagnostic test. For example, if a test has a sensitivity of 95% against a gold standard that has a true sensitivity of 85%, then the true sensitivity of the first test for its target would be 0.95 x 0.85 = 0.8, i.e., 80 and not 95%. In the case of H pylori, the most commonly used gold standard in early work on indirect diagnostic methods was culture and/or histology. These two methods possibly have relatively the lowest sensitivity of any test for the presence of H pylori. This is owing to the potential sampling error in removing biopsy material from an uninfected site. This, combined with variable culture efficiency and identification in histologtcal sections, will lead to false negatives (I). This would be a particular problem in patients with any changes to the gastric mucosal epithelium from typical antral epithelium, such as intestinal metaplasia, because the changes would From Methods m Molecular Medicine, Hehcobacter pylori Protocols Edited by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

19

NJ

20

Ha wtin

Table 1 Calculation of Test Sensitivity, Specificity and Predictive Values Gold standard Gold standard positive negative Total Test posrtive A B A+B C D C+D Test negative Total A+C B+D A+B+C+D Sensitivity = A/(A + C) x 100, speclficlty = D/(B + D) x 100 PosItwe predlctwe value = A/(A+B) x 100, negative predlctlve value = D/(C + D) x 100

remove the niche availabihty to H. pylori and reduce the infective load. The problems can be munmtzed, for example, by staining hrstologrcal sectrons with the modified Giemsa stain, making the recognition of H pylori easier, even when the organism IS present in small numbers. The combination of culture and histology is preferable to using either of the two independently, and certainly preferable to using endoscopic appearance alone, to estimate the sensitivity and specificity of a test. Many workers only count a case as H. pylon positive when two out of three indirect tests are posrttve. Calculation of sensitivity, specificity, positive predictive, and negative predictive value IS shown in Table 1. Indirect methods such as serology and urea breath test (UBT) have distinct advantages over detection methods that require biopsy material. The prmcrpal advantage is these two methods, which, in effect, sample the whole stomach and duodenum, although indirectly, thus avoiding the pitfalls of sampling error and subsequent culture or histology. The pitfall of sampling error may be reflected in the apparent lower specificity rates of the 13Carbon UBT against “gold standards” such as CLO-test@,histology, and culture (2), suggesting that the UBT produced false positives However, it is possible that the three gold standards produced some false negatives by missing the infection. 7.7. Serology Many formats have been developed for the detection of specific antibodies to H pylori. These include latex agglutination, capillary, and enzyme-linked immunosorbent assays(ELISAs). Many of these use undefined antigens that are relatively easy to prepare but can include crossreacting antigens that reduce specificity. The impact of crossreacting antigens and nonspecific antibodies to produce false posrtrves can be muumized by determination of threshold concentration of antigen and dilution of serum m the test system using a checker board titration. Calculation of the sensitivity and specificity of each concentration of antigen and dilution of serum against a gold standard will reveal the

Serology and Urea Breath Test

21

optimal combination, However, the assay should then be validated by screening a significant number of samples from a variety of patients covering all ages and disease states. This will ensure that the assay characteristics are maintained in routine usage. The optimization of the test system may be different for each diagnostic purpose. For example, if the test is to be used as a screen for endoscopy (where only positives are referred for endoscopy), then it is important that no H. pylon infected individuals are missed, i.e., no false negatives are produced by the test. Under this scheme, a number of false-positive results could be tolerated as those individuals would then be referred for endoscopy. The antigen concentration and serum dilution would be adjusted to ensure 100% negative predictive value against a set of validated sera. Alternatively, the cutoff value for positivity could be adjusted. Assay optimization may also be necessary for the population it is intended to screen. Adjustment to the cutoff may be required in some countries or areas of high immigration to prevent false positivity owing to crossreacting antibodies from frequent campylobacter infections. Similarly, serodiagnosis in a pediatric population from a developed country might require a lower cut-off value to increase sensitivity. Assay performance can be accurately determined by the construction of a receiver operating characteristic, which describes the overall performance independently of predefined cutoff values (3). Serology has achieved extremely high rates of sensitivity and specificity against all gold standards such as urea breath test, culture, and histology. These standards are all direct tests for detecting the presence of H. pylori. This means that, in the great majority of cases, positive serology indicates the presence of Z-I. pylori and not past infection; this suggests that once infected, always infected, unless the gastric mucosa is altered to displace the preferred epithelium. A variety of H. pylori antigens are recognized by each host as demonstrated on Western blotting; this has restricted the manufacture of defined capture antigens for general serodiagnostic assays that require high sensitivity rates (4). However, as more specific proteins and genotypes of H. pylori are identified associated with disease entities, it is possible that serological assayswill be developed for the diagnosis of those pathologies (5). Control of interbatch reproducibility in the manufacture of antigens is essential to maintain an assay’s characteristics. This is easier for defined antigens, especially those produced as recombinant proteins, than for undefined antigens. Variables such as ill-defined growth media and atmosphere can influence each batch of H. pylori antigen production. It is essential that quality control of each batch of antigen be done before release for use in the test proper. Quality control might include comparison of batches of antigen by protein pro-

Ha wth

22 Table 2 Examples

of Capture

Antigens

Antigen Acid-extract” Sonicate High molecular weight cellassociatedproteins Purified 120~kDaprotem

Used in

H. py/ori Serology

Reported sensitivity, % 95.6 100 98.7

Reported specificity,% 83 60.7 100

Gold standard Histology Histology UBT

Ref. 7 22 9

96

92

Histology

22

aLater reports have Improved the sensltwty and specific@ of assays based on acid-extracted antigens (23)

filing using gel electrophoresls and by assessing the antigen’s performance against a standard set of sera. Clearly, the more complex the antigen, the more difficult it is to maintain interbatch comparability. Capture antigens extracted from multiple strains of H. pylori are relatively more.difIicult to produce with high interbatch reproducibility. Antigens extracted from a single strain of H. pylori have been shown to be sufkent even when screening sera of subjects from different countries (6). Not many routine laboratories are equipped for protein separation and purification. An alternative to the sophisticated approach has been shown to be successful and in wide use (7). This assay used an acid extracted protein of a single strain H. pylori as the capture antigen, but other antigens, such as a simple whole-cell sonicate, could be used in the same system but possibly with poorer performance (8). Table 2 describes a few antigen preparations that have been used. H. pylori capture antigens may not trap antibodies specific to the closely related Helicobacter heilmanni, although samples from patients infected with this organism will be rare. Patients with a clear clinical history of upper gastrointestinal symptoms who are seronegative should be investigated further for such infections. Prozone effects should also be excluded as a reason for seronegatlvity in these patients. However, it has been recognized that a small minority of H. pylori culture positive patients fail to seroconvert at all (9).

1.2. Urea Breath Test H. pylori produces a powerful urease that is unusual among important medical bacteria (10). A simple demonstration of the power of the urease IS to inoculate a Christensen’s urea slope with a small moculum of H. pylori. A red color appears almost immediately owing to the rise m pH on liberation of ammonia from urea.

Serology and Urea Breath Test

23

The reaction is shown in outline below: urease NH*-C-NH2 +

II

CO2+ NH3

0 This ureolytic activity of H. pylori can be used for microbiological identification purposes, However, it is the CO2 liberated during the hydrolysis of urea by urease that is used in diagnostic presence/absence tests. The carbon molecule in urea can be substituted with either the stable isotope Cl3 or the radtoactive isotope C 14.Each isotope liberated from urea as either 13C02 or 14C02 can then be distinguished from the predominant atmospheric Cl2 by mass or radioactivity, respecttvely. The use of either isotope suffers from one major disadvantage each. The use of Cl4 is limited to where suitable radioactivity facilities are available and Cl3 can only be detected by expensive mass spectrometry. These limitations aside, the Urea Breath Test (UBT) has become an extremely useful specific and sensitive tool m the detection of H. pylori mfection, both in humans and in animals. The basis of the UBT 1sthat when a subject ingests urea labeled with either isotope, if H pylori is present then the urease will hydrolyze the urea liberating labeled CO2 that is absorbed into the blood stream, exchanged m the lungs, and exhaled. Detection of labeled CO2 in a breath sample thus reveals the presence of H. pylori infection in the stomach. Essentially, Cl4 and Cl3 UBTs are very similar m application, but it has been debated that, as Cl4 does not occur naturally, then it could be a more sensitive test when used with a lower cutoff value In contrast, the Cl3 UBT involves the measurement of the enrichment of Cl2 with C13,expressed as a ratio, as Cl3 does occur naturally. Important influences on the accuracy of a UBT include testing a subject too quickly after any treatment that has an effect on H pylori, including selfadministered antacids, which are now available as over-the-counter pharmaceuticals. False negatives could occur if the infection IS being suppressed (21-13). Therefore, the subject should have completed any course of treatment I that may influence the UBT for between 14 d and 1 mo before testmg. Other factors such as gastric surgery can produce false-negative UBTs thought to be owing to rapid transit of urea through the remnant stomach (14). Protocols for each form of the UBT vary and should be optimized by each operator when being used for the first time. There is considerable debate, that remains unresolved, on the necessity for a test meal in the protocol. It was thought that the test meal was necessary to delay gastric emptying, thus prolongmg the contact time between enzyme and substrate. This was useful when the mterpretation

Ha wtin

24

of the test was made by using an area under the curve (AUC) calculation. AUC required several collections of breath over an extended period to provide sufficient points to plot an excretion curve of the labeled C02. Many workers now take just one or two post-dose samples of breath for analysis. Urea has been classed as a pharmaceutical and as such now requires a hcence. This means that UBTs are only available on a named patient basis. 2. Materials 2.1. Serology 1. Antigen. Acid-extracted antigen preparation sufficient to coat required number of microtiter plates 2 Posltlve and negative control sera. 3. Coating buffer: 0. IA4 carbonate buffer, pH 9.6 4 Wash buffer. 145 mMNaCl,O.O5% v/v Tween-20. 5. Dlluent: 0.05M Tris-acetate buffer, pH 7.6, containing 1% w/v bovine serum albumin (globulin free, fraction V), 145 mMNaC1, 0 05% v/v Tween-20. 6. Conjugate. Antihuman IgG-peroxidase (see Note 1). 7 Substrate* 58 mg tetramethylbenzidme in 10 mL DMSO, 100 pL of this diluted with 10 mL 0 1M sodium acetate buffer, pH 6.0; add 1 pL of hydrogen peroxide 8. 2M H2S04. 9 0 2M glycine-HCl buffer, pH 2.2 10. 1MNaOH. 11. Dialysis tubing: bolled before use and rinsed with distllled water. 12. Sterile distilled water 13. Growth media: blood agar plates.

2.2. Urea Breath Test 1 2. 3. 4. 5

Carbon-labeled urea (see Notes 3 and 4) Distilled water. Test meal (see Note 2). Breath collectlon contamers. Anhydrous CaCl, (Cl4 UBT only)

6. Hyamine hydroxide (Cl4 UBT only). 7 Ethanol (Cl4 UBT only). 8. Phenolphthalein (Cl4 UBT only).

3. Methods 3.1. Serology 3.1.1. Antigen Preparation 1 Inoculate about 20 blood agar plates with a type strain ofH. pylon, e.g., NCTC 11638, to produce a confluent lawn cultnre. Incubate for 48 h at 37°C under rmcroaeroblc conditions (5-10% CO+ 5% 02* balance N, or HZ) in a humtdified atmosphere

Serology and Urea Breath Test

25

2. Harvest all growth mto sterile distilled water and wash the bacteria twice by suspension and centrimgation. 3. Deposit the bacteria by centrifugation and resuspend (4% w/v) in 0.2M glycineHCl buffer, pH 2.2 Incubate for 15 min at 20°C. 4. Deposit the bacteria by centrifugation and decant the supematant into a sterile container. Neutralize the supematant pH with 1M sodium hydroxide and dialyze against distilled water ovemtght at 4°C. 5. Estimate the protein concentration of the dialysate, aliquot, and store at -20°C

3.1.2. Assay Procedure 1. Couple the antigen to a microtiter tray well at 5 pg/mL in O.lM carbonate buffer pH 9.6 (100 $/well). Incubate the tray either overnight at 20°C or for 1 h at 37°C in a humidified atmosphere. 2. Remove excess antigen with three washes usmg 200 pL wash buffer. 3. Dilute serum samples 1:200 in diluent. Add 100 pL of each diluted sample to a designated well but leave one well as a control blank. Include a positive and negative control serum. Incubate the tray for 2 h at 37’C in a humidified atmosphere. 4. Wash the tray as described in step 2. 5. Add 100 pL of diluted conjugate in same buffer used to dilute the sera Incubate the tray for 30 min at 37°C. 6. Wash the tray as described in step 2 7. Add 100 pL of substrate and incubate for 10 min at 20°C. 8. Stop the reaction by the addition of 50 pL of 2MH2S04 and measure the optical density (OD) of each well at 450 nm.

3.1.3. Assay Interpretation Interpretation can be either qualitative or quantitative. A qualitative result is obtained by comparison against a standard serum of which the ODd5c correlates with the established cutoff value (established in the validation of the assay to calculate the sensitivity and specificity). It is indicative of the presence of antlH pylori antibodies if the sample ODd5c exceeds the standard serum value. Values close to the cutoff are equivocal, loosely termed the gray-zone. The definition of a grey-zone is variable but can be identified by simple statistics on a population of serum samples validated by other gold standard tests, Quantitative interpretation is obtained when the ODdSo is converted to a measurement of mass of specific H pylori antibody, expressed as either mass of antibody (IS) or ELISA units (16) by extrapolation from a standard curve.

3.2. Cl4 Urea Breath Test Since the introduction of the Cl4 UBT, the dose of Cl4 urea radiation has decreased from 0.4 to 0.1 Mbq (17). The lower amount of Cl4 urea was calculated to result in an equivalent dose of between 4 and 8 l.,tSv (18). Not suitable for children or in pregnancy.

Hawtin

26

1 Subject should fast overnight before test. 2. Time zero’ the subject drinks a liquid meal and then the test dose of 0.1 MBq C14urea in 20 mL of water. Optionally, subjects are requested to he on each side for 2 min each to aid distribution of the label, although evidence to support the necessity of this IS lacking. 3. Time zero + 40 min: take a measured breath sample from the subject by exhaling normally through anhydrous CaCl, mto a vial containing 2 mm01 of hyamme hydroxide in 2 mL of ethanol with phenolphthalein. The subject contmues to breathe through the system until decolorization occurs, indicating that sufficient CO, has been trapped. 4. Add 10 mL of scmtlllant to the vial and measure Cl4 activity m a hquid scmtlllatlon counter. Express the results as counts per minute (cpm)

3.3 . Cl3 Urea Breath Test Preferred test for children and in pregnancy. 1. 2 3 4.

Subject should fast overnight before the test. Time zero: SubJect drmks test meal. Time zero + 5 mm: collect predose breath samples (see Notes 5-7). Time zero + 10 min: subject drinks urea solution containing between 75 and 100 mg C13-urea in 50 mL of water. Optionally, subjects are requested to he on each side for 2 min each to aid distribution of the label, although evidence to support the

necessityof this 1slacking. 5 Time zero + 40 mm* collect postdose breath samples.

3.4. Cl3 UBT Interpretation Extensive valldatlon of the Cl3 UBT has been done and generally the calculated upper limit for excess 613C02 excretion in H pylori negative SubJects was 4.9 per mil(19). It 1sessential that each UBT have a cutoff value re-established during its development in a new center or population. It is useful to understand how mass spectrometrlsts express their results. The volume of exhaled breath collected is not critical to the Cl3 UBT. This 1s because it is the comparison of the ratios of C12:C13,measured by isotope ratio mass spectrometry, m the individual samples that is calculated. The ratios are expressed as parts per thousand or per mil (O/00) and are relative to an international standard, PDB calcium carbonate. However, PDB is no longer available and secondary standards have been calibrated. PDB IS derived from Belemnitella

americana,

extracted from the Pee Dee formation

of South Caro-

lina. The ratio of C12:C13m this standard is close to that occurring naturally and 1s expressed as a carbon isotope ratio of 0 per mil. Consequently, as baseline breath samples are depleted m Cl3 isotope, they can have negative per ml1 values. This 1scompensated in H. pylori positive subjects by isotope ratios greater than that of PDB and therefore give a positive value per mil. By taking

27

Serology and Urea Breath Test

the difference, or 6 (change), between pre- and postdose ratios, a measurement is made of the enrichment, or excessof Cl3 isotope. It is the extent of the Cl3 excess, usually a minimum of about 65 per mil between pre- and postdose sample, that determines the presence or absence of H! pylori. It is calculated as below: 6c13= &mp~e - bmdard) 1Rstandard x 1000; where R = Ci3/Cr2. 4. Notes 1. Specific coqugates for both mununoglobulm 2.

3.

4. 5.

6. 7.

class and subclass and for species, e.g., mouse or human, can be substituted according to test requirements Some protocols for UBTs include a mixture of Calogen and Ensure as a test meal (2). Calogen is a suspension of 50% arachis oil (peanut oil) in water to which some patients may show a severe allergic reaction (20) Other drinks have been used successfully, e.g., orange juice, but a test meal must not contribute to 13COZ excretion by its own metabolism as is the case with carbohydrates. Although the actual dose of radiation to the subject is estimated to be equivalent to one day’s background exposure, radiological containment facihties are essential for b-emitting Ci4-urea. Furthermore, the adsorbed dose to the body should be minimized by reducing the bladder voiding Interval to 2 h. This will reduce the dose to the urinary bladder wall significantly (22) The voiding interval can be easily reduced by encouraging the subject to drink immediately after the breath sample has been taken and void the bladder as soon as possible The carbon labeled urea of choice is used, either Cl4 or Ci3-urea. There are two techniques for collecting breath samples in the Cl3 UBT. One involves the subject breathing normally through a straw with the tip in the base of a vacutamer tube. The subject continues to breathe until beads of condensation appear on the glass around the tip of the straw upon which the straw is slowly withdrawn from the tube. The tube is then sealed with a rubber bung. It is prudent to prepare at least 3 tubes in this manner at each stage of the UBT. The second method uses special collection bags (Quintron, Milwaulkee, WI) to ensure the trapping of a suitable quantity of breath. This method IS better tolerated by children. It is important to handle reagents and equipment carefully to prevent introduction of contaminating urea. It IS essential that a pre-Cl3 urea dose breath sample is taken as this baseline is required for comparison with postdose samples to establish changes m Ci2:Ci3 ratio.

References 1. Glupczynski, Y. (1994) The diagnosis of Helzcobacter pylori microbiologist’s perspective. Rev. Med. Mcroblol. 5, 199-208.

mfection*

a

Ha wtin

28

2. Logan, R., Dill, S., Walker, M., Bauer, F., Gummet, P A., Hirschl, A , Rathbone, B., Johnson, P , Baron, J H., and Misiewicz, J. J (1990) The detection of H pylorl using a European standard t3Carbon urea breath test. Rev Esp Enfermedades Digestwas 78(Suppl. l), 14. 3. Trautmann, M., Moldryzk, M., Vogt, K., Held, T., and Marre, R. (1994) Use of a receiver operating characteristic in the evaluation of two commercial enzyme immunoassays for detection of Hellcobacter pylorl mfection. Eur. J Clin Mwoblol Infect Du 13,8 12-8 19. 4 Stacey, A R., Hawtin, P. R., and Newell, D. G. (1990) Antigemcity of fractions of Helicobacterpylorz prepared by fast protein liquid chromatography and urease captured by’monoclonal antibodies. Eur J Clin. Microblol. Infect. Du 9,732-737. 5 Cover, T L , Glupczynski, Y., Lage, A. P., Burette, A., Tummuru, M. K. R., Perez-Perez, G. I., and Blaser, M. J. (1995) Serologic detection of infection with cagA+ Helrcobacterpylorz strains. J. Clw. Mlcroblol 33, 1496-1500. 6. The Eurogast Study Group (1993) An international association between Helacobacter pylori infection and gastric cancer. Lancet 341, 1359-1362. 7. Newell, D. G., Johnston, B. J , Ah, M. H., and Reed, P. I (1988) an enzymelinked mmmnosorbent assay for the serodiagnosts of Campylobacterpylorr associated gastritis. Scan J. Gastroenterol. 23(Suppl. 142), 53-57 8. Booth, L., Holdstock, G., MacBrtde, H., Hawtm, P., Gibson, J , Ireland, A , Bamforth, J., DuBoulay, C. E., Lloyd, R. S., and Pearson, A. D. (1986) Clinical importance of Campylobacter pylorldis and associated serum IgG and IgA antibody responses in patients undergoing upper gastrointestinal endoscopy J Clrn Pathol. 39,215-219.

9. Evans, D J., Evans, D. G., Graham, D. Y , and Klein, P. D (1989) A sensitive and specific serologic test for the detection of Campylobacter pylorl mfection. Gastroenterology 96, 1004-1008. 10. Hawtin, P. R., Stacey, A. R , and Newell, D. G (1990) Investigation of the structure and localisation of the urease of Heltcobacterpylorz using monoclonal antibodies. J Gen. Microbzol 136, 1995-2000. 11. Graham, D. Y., Klein, P. D., Evans, D. J., Graham, D Y., Klein, P D., Evans, D. J , Alpert, L. C., Opekum, A. R., and Boutton, T W. (1987) Campylobacterpylori detected non-invasively by the 13C-urea breath test Lancet i, 1443,1444 12 Bell, G. D , Well, J., Harrison, G., Bell, G. D., Weil, J , Harrison, G., Morden, A , Jones, P. H., Gant, P. W., Trowell, J. E., Young, A. K., Dane&mend, T. K , and Logan, R. F. (1987) 14C-urea breath analysis, a non-invasive test for Campylobacter pylon in the stomach. Lancet, i, 1367,1368. 13 Weil, J., Bell, G. D., Powell, K., Weil, J , Bell, G. D., Powell, K., Morden, A., Harrison, G., Gant, P W., Jones, P. H , and Trowell, J. E (1991) Omeprazole and Helicobacterpylori temporary suppression rather than true eradication. Alzment. Pharmacol

Therap. 5,303-3 13.

14 Weil, J. and Bell, G D. (1989) Detection of Campylobacter pylori by the t4Cbreath test, in: Campylobacter pylori and Gastroduodenal disease (Rathbone, B. J. and Heatley, V R., eds.), Blackwell Scientific, Oxford, pp. 74-87.

29

Serology and Urea Breath Test

15 Steer, H. W., Hawtm, P. R., and Newell, D. G. (1987) An ELISA technique for the serodtagnosis of Campylobacter pyloridzs mfection m patients with gastrttis and benign duodenal ulceration. Serodiagnosis Immunother 1,253-259. 16. von Wulffen, H. (1992) An assessment of serological tests for the detection of Hellcobacterpylorr.

Eur J Clm. Mcroblol

Infect Du

l&577-582

17. Rauws, E. A. J., Royen, E. A. V., Langenberg, W., Rauws, E. A. J., Royen, E. A. V., Langenberg, W., Woensel, J. V., Vrij, A. A., and Tytgat, G. N. (1989) t4Curea breath test m C. pylori gastritis. Gut 30,798-803 18. Bell, G D. (1993) Urea breath tests, in Helicobacter pylori znfectzon (Northtield, T. C., Mendall, M., and Goggin, P. M., eds.), Kluwer Academic, Dordrecht, pp 127-138.

19. Logan, R. P. H. (1992) Detection of Helicobacter pylon by the 13C-urea breath test, in Hehcobacter pylori and Gastroduodenal Disease (Rathbone, B J and Heatley, R. V., eds ), Blackwell Scientific, Oxford, pp 88-106. 20. British National Formulary No. 30 (September 1995) 596. Pubs British Medical Association & Royal Pharmaceutical Society of Great &tam. 2 1. Stubbs, J. B. and Marshall, B. J. (1993) Radiation dose estimates for the carbon 14-labelled urea breath test. J NucE Med 34, 821-825 22 Gerstenecker, B., Eschweiler, B., Vogele, H., Koch, H. K., Hellerich, U., and Ktst, M (1992) Serodiagnosis of Hekobacter pylorz infections with an enzyme immunoassay using the chromatographically purified 120 kilodalton protein. Eur J Clin. Mcroblol

Infect Du

11, 595-601.

23 Wilcox, M. H., Dent, T H S., Hunter, J. O., Gray, J. J , Brown, D. F J., Wight, D G. D., and Wratght, E P (1996) Accuracy of serology for the diagnosis of Helicobacter pylori Infection--a comparison of eight kits J Clin Path01 49, 373-376.

The Detection of If. py/ori by the Polymerase Chain Reaction Nicholas

P. Mapstone

1. Introduction A number of methods are currently available for the detection of Helicobacterpylori, including serology, culture, histology, and isotope breath tests. All have relative advantages and disadvantages of sensitivity, specificity, convenience, expense, and immediacy. The polymerase chain reaction (PCR) was an obvious choice of test once the technique became available to many laboratories. PCR tests for various bacteria, including H. pylori, are available on a wide range of specimens, including cultures and direct clinical samples. The ideal PCR test for H pylori would be highly sensitive and specttic, cheap if performed in bulk, and could produce results much faster than would be possible with culture. PCR tests for H. pylori have been described for a number of genomic targets, such as the urease gene (I), and genes encoding cytotoxtc factors, such as cagA (2). We chose the 16s ribosomal RNA gene of H. pylori as our target for the PCR (3), because sequence information was readily available, and this gene formed part of the basisof the classification of H. pylori as a separate organism. There have been varying conclusions drawn about the ability to detect H. pylori in fecal specimens. The presence of the bacteria in a small proportton of Meckel’s diverticulae (4) and metaplastic rectal mucosa (5), and the culture of the organism from children (6) and adults (7), has raised the possibility of fecal oral spread of infectron. We were able to use PCR on fecal specimens to indrcate the presence of H. pylori DNA in feces in a large proportion of other patients (8). However, others using similar, but not identical, techniques (9) were unable to detect H. pylori DNA in this situation. From Methods m Molecular Me&me, Helicobacter pylon Protocols EdIted by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

31

NJ

32

Mapstone

2. Materials 1 Colony scrapes from cultured H. pylori, the DNA of which will serve as a positive control in future reactions 2 Gastric biopsies, taken from sterilized and scrubbed (see Note 1) gastric biopsy forceps directly into test tubes, frozen, and stored at -7O’C. 3. Gastric aspirates taken via sterilized endoscopy scope channels and frozen. 4 Tooth scrapings, taken using sterile wooden sticks, broken mto test tubes, and frozen. 5. Saliva specimens taken by the patient spitting into a universal container, the saliva then being transferred to a test tube and snap frozen. 6. Fecal samples collected by patients into standard microbiology fecal specimen containers. Immediately followmg receipt, a portion of the sample is placed in a test tube and frozen for future sampling (see Note 2) 7 Proteinase K. 8 Digestion buffers I. 100 mA4 NaCl, 10 mM Tris-HCl, 25 mM EDTA. 9. Digestion buffers II: 100 mMNaC1, 10 mMTris-HCl, 25 mA4EDTA, 1% SDS. 10. 5MNaCl 11 CTAB (hexadecyltrtmethyl ammonium bromide)/NaCl mixture* 10% CTAB m 0.7M NaCl 12 Chloroform/lsoamylalcohol, 24: 1 measured by vol. 13 Phenol/chloroform/isoamyl alcohol, 25~2411 measured by vol. 14 Ethanol 15 3M Sodium acetate, pH 5.2. 16 Primers HP1 5’-CTG GAG AGA CTA AGC CCT CC-3’, HP2 5’ ATT ACT GAC GCT GAT TGT GC- 3’, which result in a 1 lo-bp-sized PCR product and HP3 5’-AGG ATG AAG GTT TAA GGA TT-3’. 17. Molecular biology grade water 18. Agarose. 19. TAE electrophoresis buffer: 40 mM Tris-acetate, 1 mM EDTA, pH 8.0. 20. 10 mg/mL Ethidium bromide stored at 4°C in a bottle wrapped in foil. 2 1 Gel tray (Hoefer, San Francisco, CA). 22 Gel comb with 15 teeth, producing 40-pL capacity wells 23. Gel unit (MAX submarine agarose gel unit [Hdefer, model HE99]). 24 Gel running buffer 1000 mL TAE with 40 pL ethidium bromide solution 25 TE buffer, 10 mM Tris-HCl, 1 mA4EDTA, ph 7.5. 26. DNA size ladder (see Note 3). 27. DNA sample loading buffer (10X): 0.25% bromophenol blue, 0.25% xylene cyanol, 25% Flcoll (type 400) m water. 28. 10X PCR buffer: 500 mM KCl, 100 rruV Tris-HCl, 0.1% (w/v) gelatm, pH 8.3, stored at -20°C. 29. dNTPs, made up from concentrated stock soluttons of individual nucleotides (dATP, dGTP, dCTP, and dTTP), to produce a final concentration of 1.25 & for each individual nucleotide and 8 mM total nucleotide. 30. Taq DNA polymerase.

33

Detection by PCR 3 I. Thermal cycler. 32. 10 &Magnesium 33. Mineral oil

chloride.

3. Methods

3.1. DNA Extraction: Saliva/Dental Plaque/Gastric Biopsies 1. Place the specimen (either 100 pL of saliva or the broken end of a toothpick with scraped dental plaque, or a gastric biopsy) in a test tube (see Note 4). 2. To each test tube, add 0.5 mg of proteinase K dissolved in 250 J.& of digestion buffer II. A further 250 pL of digestion buffer I is added to each tube. 3. Place the tubes m rotation in an incubating oven for 24 h. 4. To each tube add 100 pL of 5M NaCl and 70 pL of CTAB/NaCl, m that order (NaCl first, followed by CTAB/NaCl). 5. Incubate the tubes in a water bath at 65’C for 10 min. 6. To each tube add an equal volume of 24: 1 chloroform/isoamyl alcohol and then spin at low speed (6OOOg) for 3 min. Complexes of CTAB-protein and polysaccharide appear as a white interface after centrifugation. Remove the supernatant from each tube and place in a clean test tube. Discard the old tube and contents. 7. To each tube add an equal volume of 25:24: 1 phenol/chloroform/isoamyl alcohol and then spin at low speed (6000g) for 3 mm. Remove the supernatant from each tube and place in a clean test tube. Discard the old tube and contents. 8. To each tube add an equal volume of chloroform!isoamyl alcohol and then spm at low speed (6000g) for 3 min. Remove the supematant from each tube and place in a clean test tube. Discard the old tube and contents. 9. To each tube add twice the volume of ice-cold ethanol and one-tenth the volume of 3M sodium acetate. Leave the tubes m a fridge at -20°C for 24 h 10. Spin each tube at 12,000g for 5 min and elute the ethanol 11. Wash the precipitate in ice cold 70% ethanol; centrifuge again at high speed for 5 min and remove the ethanol. Dry the ensuing sample in a vacuum desiccator for 1 h, and dissolve m 50 & of molecular biology grade water.

3.2. DNA Extraction: Fecal Samples 1. Defrost frozen fecal samples and add an equal volume of water to the sample Agitate to produce a suspension of the sample. Transfer 20 pL (see Note 5) of this mixture to a separate test tube (see Note 6). 2. Treat the sample in the same way as the saliva specimens noted m Section 3.1.

3.3. Single-Step PCR 1. Collect samples to be analyzed. For each reaction, the specimens and a positive control (ZY.pylon DNA) and a negative control (water) should be included 2. Allocate one test tube for each specimen, including the controls. To each test tube add 50 pmol of primer HP1 and 50 pmol of HP2 3. Centrifuge each tube for 5 s to ensure that the pnmer 1sat the bottom of the test tube.

34

Mapstone

4. Place the test tubes in a vacuum chamber and evaporate the ammonium m which the primers are received. 5. Once the primers are dry, add to each test tube 0 2 pL of 7’uqpolymerase, 8 & dNTP, 5 pL of PCR buffer 10 $ (see Note 7) of magnesium chloride, and 26 pL of water for an ultimate volume of 50 & 6. Over the top of this rntxture, add 50 pL of mineral oil to prevent reagent evaporation 7. To each test tube add 1 pL of the sample to be tested. 8. Place each test tube m the PCR machine. 9. Thermal cycle the test tube contents with the following temperature profile. 1 cycle of 3 min at 95”C, 1 mm at 6O’C; 38 cycles of 30 s at 95’C and 30 s at 60°C; then 1 cycle of 3 mm at 72°C 10. Remove the test tubes and store at 4°C until gel visualization. 3.4. Nested

PCR (see Note

8)

Perform steps 1-9 using primers HP1 and HP3 as in Section 3.3., but with 30 cycles instead of the 38 noted in step 9. Using 1 pL of the ensuing PCR products as the sample, perform another PCR using primers HP1 and HP2, but again with 30 cycles in step 9.

3.5. Gel Visualization 1 Dissolve 2.4 g of agarose in 120 mL of TAE and heat for 2 mm on high power m a microwave oven. Add 10 p,L of ethidium bromide and cool with constant stlrring to 55°C. Add the ensuing mixture to a gel tray with gel comb and leave to cool. Place the gel tray in an electrophoresis bath contaming 1 L of TAE and ethidium bromide. 2. Remove the PCR products from storage and prepare a mixture of 12 K of PCR product, 12 pL TE buffer, and 6 p.L loading dye. 3. Place 30 pL of the mixture in each comb slot and run at 150 V for 90 min. 4. Take a picture. The PCR product should be 110 bp long. 4. Notes 1. Most endoscoplc samples are taken by forceps that have been sterilized with glutaraldehyde. Though this will kill many bacteria, it will not necessarily render Its DNA insusceptible to the PCR. By using forceps on meat samples, we found evidence of contammatlon of biopsy forceps if they had not been scrubbed as well as sterilized (IO). Scrubbing of biopsy forceps has since been instituted as routme m this department for other reasons. 2. Length of storage of feces was found to be important in the sensltlvtty of detection of H. pylon’ DNA. Initial work on fecal samples that had been kept m the freezer for 6 mo-1 yr showed fewer positive results than when feces were tested within a month of collection. 3. A lOO-bp ladder 1s most appropriate, considering the size of PCR product with the mam 2 primers (110 bp).

Detection by PCR

35

4. Each batch of specimens to be DNA extracted should include one test tube with no sample but an equal volume of water. This test tube undergoes all of the purification steps together with the biopsy specimens, and controls against contamination of the specimens during DNA extraction. This is at least as important as using negative controls during the actual PCR 5. Feces contain chemicals that will inhibit PCR, and some of these chemicals will persist in the sample despite DNA extraction. A number of purification techniques for such samples have been suggested, but if only a small amount of feces is used (in this example, 10 uL), the concentration of inhibitors appears to be so low as not to interfere with the PCR in most samples. We did not need to use bead purification or other techniques to prevent PCR inhibition in most of the fecal specimens. 6. Unsuccessful attempts to perform PCR upon feces have centrifuged the fecal samples before DNA extraction, using the supernatant This may be appropriate when the technique is attempting to demonstrate DNA from organisms that are living mamly m the large intestine. However, H pylorr DNA will originate mainly from the stomach and the DNA may well be from dead organisms, mixed in with particulate matter from the stomach. We have found DNA extraction from whole fecal specimens, missmg any centrifugation step, to be more successful m demonstrating H. pylon DNA in feces. 7. When the assay is being developed, a PCR should be run to confirm the opttmum magnesium concentrations for that particular reaction. A PCR is set up with six test tubes each, with H. pylori DNA and contaming magnesium concentrations varying from 0.5 to 3 mM. Depending on the quantity of PCR product (crudely estimated by the intensity of the band seen on agarose gel electorphoresis), this will determme the amount of magnesium (and water) to be used m subsequent reactions 8. We found ordinary PCR to be sensitive enough to detect H. pylori DNA in gastric biopsies, aspirates, and fecal samples, but not in oral samples. To detect the very low copy numbers of bacteria in this context, we used nested PCR in which an initial 30 cycles of PCR is repeated, with internal primers and the previous PCR product as the template for the reaction. This mcreases sensitivity approx 1OO-fold.

References 1. Clayton, C., Kleanthous, K., and Tabaqchali, S. (1991) Detection and identitication of Helzcobacterpylorz by the polymerase chain reaction. J Clin Pathol. 44, 515-516. 2. Owen, R. J., Hurtado, A., Banatvala, N., Abdi, Y., Davies, G. R , Feldman, R., and Hardie, J. M. (1994) Conservation of the cytotoxin-associated (cagA) gent of Hehcobacter pylori and investigation of association with vacuolating-cytotoxin activity and gastroduodenal disease. Ferns Immunol Med Microbial 9,307-3 15. 3. Ho, S. A., Hoyle, J. A., Lewis, F. A., Seeker, A. D., Cross, D., Mapstone, N. P., Dixon, M. F., Wyatt, J. I., Tompkins, D. S., Taylor, G. R., and Qutrke, P. (1991)

Mapstone

36

4.

5.

6 7

8.

9.

Direct polymerase chain reaction test for detection of Hellcobacter pylori m humans and animals. J. Ciin. Mcrobiol. 29,2543-2549. Bemelman, W. A., Bosma, A., Wiersma, P. H , Rauws, E. A., and Brummelkamp, W. H. (1993) Role of Helicobacterpylori in the pathogenesis of complicattons of Meckel’s diverticula. Eur. J. Surg. 159, 171-175. Kestemberg, A., Marino, G., DeLima, E., Garcia, F. T., Carrascal, E., and Arredondo, J. L. (1993) Gastric heterotopic mucosa m the rectum with Hehcobacter pylon-like organisms: a rare cause of rectal bleedmg. Int J Colorectal Disease 8,9-12 Thomas, J. E., Gibson, G. R., Darboe, M. K., Dale, A., and Weaver, L. T. (1992) Isolation of Helicobacterpylori from human faeces. Lancet 340, 1194,1195 Kelly, S M., Pitcher, M. C., Farmery, S. M., and Gibson, G. R. (1994) Isolation of Helicobacterpylori from feces of patients with dyspepsia m the United Kingdom. Gastroenterology 107, 167 1-1674. Mapstone, N. P., Lynch, D. A., Lewis, F. A., Axon, A. T., Tompkins, D. S , Dixon, M. F., and Quirke, P. (1993) PCR identification of Helicobacterpylorl m faeces from gastntis patients . Lancet 341,447 van Zwet, A., Thqs, J. C., Kooistra, S. A., Schnm, J , and Snijder, J A (1994) Use of PCR with feces for detection of Hehcobacterpylorz mfections m patients. J Clin Pathol 32, 1346-1348.

10. Mapstone, N. P., Lynch, D. A., Lewis, F. A., Axon, A. T., Tompkins, D. S., Dixon, M. F., and Quirke, P (1993) Identification of Helicobacter pylori DNA in the mouths and stomachs of patients with gastritis using PCR. J. Clin Path01 46, 540-543

5 PCR for the Detection of H. pylori in Gastric Juice Aspirates and Environmental Water Samples Stephen B. Fritz and T. Ulf Westblom 1. Introduction Helicobacterpylori is a curved Gram-negative bacillus that infects the human gastric mucosa. Most methods used to diagnose H pylori require the collection of gastric samples by biopsy during gastroduodenoscopy. A positive culture or visualization of the organism by histologic examination of biopsy material can give a definitive diagnosis (I). To make a diagnosis without gastroduodenoscopy, indirect tests, such as ELISA or urea breath tests, must be relied on (2). The polymerase chain reaction (PCR) can be used to make a highly specific diagnosis from gastric juice aspirates or environmental samples. PCR is a highly sensitive technique that can detect very small amounts of DNA. The DNA molecule is chemically stable and can survive m the environment for long periods (3). PCR may therefore be useful in detecting the presence of H. pylori, even when the organism is in a nonculturable state or no longer alive. Described here are two methods for the detection of H. pylori using PCR. One protocol that can be utilized for patient diagnosis requires gastric juice aspirates; the other method, which may be used for epidemiological research, utilizes environmental water samples, such as raw sewage. The two protocols are presented m a unified format with the salient differences mentioned when needed.

2. Materials 1. Reactionbuffer (10X): 100mMTris-HCl, pH 8.3,500 mMKC1, 15 mMMgC&, 0.01% (w/v) gelatin. 2. 5 U/$ AmpliTaq DNA polymerase. 3. 2.5 mA4Combined deoxynucleotide triphosphates. Edited

From. Methods m Molecular Medzlne, by C L Clayton and H L T Mobley

37

Hehcobacter Humana

pylon Protocols Press Inc , Totowa,

NJ

38 4. Primers* We use degenerate ohgonucleotide primers that are 23- and 24-mer corresponding to amino acids of the small subunit of the H pylon ureA gene, as described previously (4). Stock concentration 40 pmol/pL 5. Template DNA: Thts protocol assumes that the samples bemg tested have been prepared prior to amplification (see Notes 1 and 2). 6. Nusieve 3: 1 agarose (FMC Bioproducts, Rockland, ME). 7. Mineral 011. 8 Ethidium bromide. 9. TBE Buffer (1 OX): 890 mM Tris-borate, 20 miI4 EDTA, pH 8.3 10 Positive control template DNA: Prepared from pure H. pylon culture and heat inactivated for 10 min at 98’C.

3. Methods (see also Note 3) 1. Make a master mix in a 1.5 mL mtcrocentnfuge tube using the following quantities of reagents/50 pL reaction (see Note 4): 30.5 pL stenle distilled water, 5.0 pL 10X reaction buffer, 4.0 pL combined dNTPs, 2.5 pL primer I,2 5 pL primer 2. Keep the mixture cold by placing it on ice. Make certain the mixture is homogeneous by prpetmg up and down or by vortexing the sample. 2. Add an appropriate amount of AmphTaq polymerase to master mrx. Use 0 5 @.,I 50 $, reaction (see Note 5) 3. Ahquot 45 pL of master mix mto microcentrifuge tubes of appropriate size for the PCR machine. 4. Overlay each aliquot with 100 $ of mineral oil (see Note 6) 5 Add 5 pL of each target sample to be tested through the mineral oil, changing the pipet trp between each sample. Also add 5 uL of positive control to one tube and 5 pL of distilled HZ0 to another tube for a negative control. 6. Perform 20 cycles of the form. 95°C for 1 mm; 55’C for 2 min, 72°C for 2 min, followed by 29 cycles of 89°C for 1 min; 55’C for 2 mm; 72°C for 2 min, and complete the reaction with 1 cycle of 89°C for 1 min; 55’C for 2 mm; 72’C for 10 min (see Notes 7 and 8). 7 Visualize the PCR products in a 24% TBE gel stained with ethtdmm bromide (see Note 9).

4. Notes 1 Gastric juice aspirate samples are prepared as follows. A 5-mL vol of gastric juice IS aspirated and buffered to a neutral pH with 5 mL of Tris (0 67M, pH 7 4) Each sample is then concentrated by centrifugation at 10,OOOg for 20 mm. The supernatants are removed and the pellets are resuspended m 25 pL of stertle distilled water. Each concentrated target sample 1sthen heat-macttvated at 98°C for 10 mm before amplification. 2. Environmental water samples are prepared as follows: 10 mL of each sample IS concentrated by centrtfugatton at 10,OOOgfor 20 mm. The supematants are removed and the pellets are resuspended in 50 pL of sterile distilled water. Since the pellets may contain much nonbiologic matenal that could interfere with the amplificatton,

PCR Detection in Aspirate and Water

3.

4. 5.

6. 7.

8. 9.

39

the sample DNA is then extracted with a commercial kit, such as G nome (Bio 10 1, La Jolla, CA) and purtfied by a Geneclean II protocol (Bio 101). Because of the extreme sensitivity of the PCR assay, it is important to minimize the risk of contamination, We recommend the following precautions: a. Separate physically into three distinct rooms the sample preparation, preamplification, and amplification/gel loading areas. Provide each room with separate supplies and equipment. b. Perform the preamplification step in a laminar flow hood equipped with ultraviolet lights, which should be turned on when the hood IS not in use. c Prepare and handle reagents using positive-displacement pipets equtpped with aerosol-resistant disposable tips and plungers. d. Change gloves upon begmnmg work in the preamphfication area and frequently during experiments. e. When preparing reagents, always use new glassware, plasttcware, and pipets that have not been exposed to DNA in use in the laboratory. f. Always use a negative control that contains all components of the PCR reaction, except template DNA Master mtxes can be made up m advance, stored at -20°C, and thawed for use in PCR reactions. We have also used Vent (exo-) polymerase when testing gastric juice aspirate samples and environmental water samples. The Vent polymerase is more heatstable; therefore, a greater number of cycles can be run yielding greater sensitivity. We used 2 pL of 1 U/pL Vent stock per 50-pL reaction. Mineral oil acts to prevent evaporation during the PCR run The exact condmons will depend on the make of the temperature cychng machine and should be varied accordingly. We do a final extension cycle of 72°C for 10 min to ensure that all the amplifying DNA strands are fully extended at the end of the reaction. We mn 30 out of the 50 cycles at a reduced melting temperature to prevent the need for adding additional Ampliraq polymerase midway through the reaction, The lower melting temperature will still melt our amplification product, but prevent melting of strands larger than 500 bp. We have also used Vent polymerase for more than 50 cycles under these conditions: 95’C for 1 min, 60°C for 2 min, 72°C for 2 min. Always load the samples into the PCR cycler once it has preheated to the 95°C melting temperature. This helps prevent primer-dimer formation. Occasionally, when running high cycle numbers on environmental water samples, there may be some nonspecific amplification. If this occurs, determination of H pylon DNA will require Southern blot transfer and hybridization with a radiolabeled H. pylori probe (5) We have used a probe made from pure cultures of H. pylori using our PCR protocol.

References 1. Westblom, T. U (199 1) Laboratory diagnosis and handling of Helzcobacterpylort, in Helicobacter pylori in Peptic Ulceration and Gastritis (Marshall, B. J., McCallum, R. W., and Guerrant R. L., eds.), Blackwell, Boston, MA, pp. 81-91

Fritz and Westblom

40

2. Westblom, T U. (1993) The comparative value of different diagnostic tests for Helzcobacter pylorz, in Helmobacter pylorr: Bzology and Clznical Practice (Goodwin, C. S. and Worsley, B., eds.), CRC, Boca Raton, FL, pp. 329-342. 3 Doran, G H., Dickel, D. N., Ballmger, W E., Jr., Agee, 0. F., Latpis, P J , and Hauswirth, W W. (1986) Anatomical, cellular and molecular analysts of 8,000-yr-old human brain tissue from the Wmdover archaeological site Nature 323,803-806 4

Westblom,

T. U., Phadms, S., Yang, P., and Czmn, S. J. (1993) Diagnosis of

Helzcobacter pylorz infection by means of a polymerase chain reaction assay for gastrrc Juice aspirates. Clin Infect. Dis 16,367-37 1. 5 Sambrook, J., Fntsch, E. F., and Mamans, T., (1989) Molecular Clonzng A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York

In vitro Susceptibility

Testing of H. pylori

Alistair McLaren 1. Introduction Since first isolated and cultured, there have been numerous reports stating the in vitro susceptibilities of Helicobacterpylori to a wide range of antibacterial agents (1,2), bismuth compounds (3), and compounds mhibiting acid secretion (4). When analyzing minimum inhibitory concentration (MIC) data, it is apparent that H. pylori is extremely susceptible in vitro to a wide range of agents, and as such, should not be a major problem to eradicate. However, this is not the case, with the only effective regimens being combinations of agents possessing different mechanisms of activity. As a result, it would be easy to dismiss in vitro activity as having any bearing on in vivo efficacy, since there is little correlation between MIC and clmical success.For example, H. pylori is extremely susceptible to amoxycillin in vitro, MIC ~0.1 pg/mL, but this agent is not effective as monotherapy (see Table 1). The basic activity of an agent must always tie into the equation of efficacy to some degree, but in the case of H. pylori eradication, it is only one of many important factors to be taken into consideration. When assessingthe suitability of an agent to eradicate H. pylori, speed of kill, i.e., the bactericidal activity of the agent, must be of paramount importance to agents that only remain in the stomach for such a short period of time (5,). For example, clarithromycin has a comparable MIC to amoxycillin, but has a rapidly tidal mode of action, making it more effective as a monotherapy. Other important factors for a drug being used against H. pylori include the ability of the agent to penetrate the mucus layer of the stomach, acid stability, agent bioavailability, emergence of resistant variants, and activity against nondividing or dormant cells. From Mefhods m Molecular Medrcme, Hellcobacter pylon Protocols Edlted by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

41

NJ

42

McLaren Table 1 Geometric Mean MIC of Agents Against 14 Clinical isolates of H. py/ori

Compound Amoxycillin Ampicillin Benzylpenicillm Cefuroxime Chlortetracycline Clarithromycm Gentamicin Clindamycine

GeometricmeanMIC, pg/mL 0.02 003 0.06 0.18 0 19 0.06 063

1.1

When analyzing drug interactions, it is apparent that certain combinations of antibiotics become more rapidly tidal than single agents and synergy, or at least an additive effect can be seen. Synergy as a concept can be hard to define (6), and any in vitro observations should always be verified by in vivo analysts. Agents that by themselves are bacteristatlc, in combmation can become bactericidal. For example, tetracycline by itself is bacteristatic, but in combination with bismuth salts shows a synergistic interaction with respect to tidal activity. The procedures described in this chapter are for carrying out MIC analysis, rate-of-kill studies, and interactions of agents in combination. Basic MICs are performed in agar by serially diluting compounds, inoculatmg with H. pylori, and incubation at 37OCfor 72 h. Cidal testing is performed in broth; analysis of viability can be analyzed by using Miles-Misra technique or, more rapidly, by flow cytometry. Drug combinations can be analyzed by creating a 96-well microtiter checkerboard and sampling at various time points. Combinattons can be further analyzed at sub-MIC concentration vta time kill studies mcorporating Miles-Misra or flow cytometry. Flow cytometric analysis of H. pylori gives a rapid method of detecting viability when compared to culture techniques. Actual numerical viability cannot be achieved. Results are quoted as percentage viability of the control cultures. Bacterial suspensions at approx 105-lo7 CFU/mL can be readily analyzed. Flow cytometry sorts bacteria according to their forward and side scatter, which is based on density, shape, and particle size. It is necessary to identify the region of resultant scatter that corresponds to the profile of the bacteria, so background noise must be limited as much as possible. This can be achieved by optimizing the quality of the bacterial cultures used, and by clearing the broth media of any debris. Broth cultures are prepared m small tissue culture flasks containing 5 mL of liquid growth medium. For all tidal testing, rt

43

In vitro Susceptibility Testing

is recommended to subculture strains at least twice in broth after 24 h growth. After initially suspending plate cultures in broth, H. pylon is incubated for 24 h microaerobically. Flasks are laid flat to increase surface area to volume ratio, Gentle agitation on a rocking table optimizes growth. After 24 h, 1 mL of H, pylon broth can be added to 4 mL of fresh media and reincubated. Subculturing the organism will clear the media of dead bacteria that ~111initially be picked up from the plate culture. It will also put the strains into a synchronous state after 2 subcultures. For flow cytometric analysis, it is essential to filter broth containing serum to remove particles that will interfere with scatter. A cleaner broth containing less background noise can be prepared by growing H pylori in S-cyclodextrin supplemented broth. Why combinations of drugs work better than single agents against H. pylorz can be hard to define. A positive interaction cannot always be measured in vitro as synergy or an additive effect, even though there is a clear advantage in vivo. For example, bismuth-containing compounds have been shown clinically to lower the emergence of resistance in H. pylori (7), thereby enhancing activity. Another example

is the increase in effectiveness

of amoxycillin

when in

combination with acid-blocking agents: an increase m potency is attained by placing the organism

in an environment

it cannot tolerate. Some combmations

do provide a clearly advantageous effect in vitro. These should always be confirmed with studies in small animals. When analyzing data, a mean fractional bactericidal concentration (FBC) index of CO.5is indicative of synergy, as is a drop in viability of >2 loglo compared to single agents in tidal studies. 2. Materials 1. Disposable plastics: 90-mm Petri dishes triple vent; 150 x 20 mm Petri dishes (Stemlm) 6-well cluster trays; 96-well tissue culture cluster trays (Costar); 10 pL

inoculating loops (NUNC). 2. 3. 4. 5. 6. 7. 8. 9.

H pylori strains for analysis

Columbia agar base (Oxold, Bassingstoke, UK). Defibrinated horse blood (Tissue Culture Services) Fetal bovme serum. Fildes and yeast extracts. Made up from a 1: 1 ratio of fildes and 20% yeast extracts Beta-cyclodextrin (Sigma). Brain heart infusion (BHI) broth (Oxoid). Growth media, solid: Heated blood (chocolate) Columbia agar containing 5% deflbrinated horse blood. Heat to 80°C until agar turns from red to brown Do not overheat as blood will separate from the agar. 10. Growth media, liquid: BHI broth supplemented with 5% fetal bovme serum and 0.5% fildes and yeasts extracts, subsequently filter sterilized through 0.45 l.an pore filter; or 0.5% (w/v) @cyclodextrm. 11. Phosphate buffered saline (PBS).

McLaren

44

12. Multipoint inoculator and 25well inoculating trays (Denley). 13 Microaerobic work station; variable atmosphere mcubator or gas Jars. H. pylon will optimally grow at oxygen concentration of 5-l 0% and carbon dioxide concentration of lO-20%. Any incubator producing these conditions or anaerobic jars combined with mlcroaeroblc gas generatmg envelope system will be sultable for supportmg H. pylori growth.

14 15. 16 17. 18 19

FACscan (Becton Dickinson, Cockysville, MD) Luckham 4RT rocking table. Sigma 2- 15 bench centrifuge. Filters. 0.2, 0.45, and 5 pm pore size Fluorescein diacetate (FDA). 96-prong moculator for 96-well microtiter trays (Denley).

3. Methods

3.1. Minimum Inhibitory

Concentration

(M/C)

1. Prepare cultures by subculturmg H pylon onto freshly poured chocolate Columbia agar plates and incubate for 48 h mtcroaerobically (see Note 1) 2. Prepare doubling dilutions of test compounds and add 1 mL to triple-vented agar plates Take into account l/20 dilution after addition of agar 3. Add 19-mL chocolate Columbia agar to each plate and ensure the test compounds are suitably mixed mto the agar. 4. When set, briefly dry agar plates 5. Harvest H. pylon cultures and suspend the moculum m 1 mL PBS. Dilute the bacteria to 1O* bacteria/ml. 6 Transfer 0.6 mL of H. pylon suspensions to a sterile moculatmg tray 7. Inoculate plates, incubate 72 h microaerobtcally. 8 The MIC is quoted as the lowest concentratton at which no growth is seen.

3.2. Rate-of-Kill Analysis 3.21. Analysis Using Viable Counting via Miles-Misra Methodology 1 Once the MIC has been established, a rate-of-kill study can be carried out. Ideal test concentrations are two and five times the MIC value. 2. Prepare a starter moculum by harvesting a 48 hour plate culture (see Note 2) and seeding a 5 mL broth culture Incubate the broth culture for 24 h microaerobically Subculture again, adding 1 mL of moculum to 4 mL broth and remcubate for 24 h mtcroaerobically (see Note 3) 3. To a 6-well tissue culture cluster tray, add 2.7 mL of growth medium to the first well, which wrll be used as the no-drug growth control

4. Add 2.6 mL of growth medium to compound test wells. 5. Prepare test compounds to a final concentration of two and five times the MIC m growth medium. Take mto account a 30-fold dilution m the experiment and prepare compounds to 60- and 150-times the MIC. 6 Add 100 $ of test compounds to the wells.

In vitro Susceptibility Testing

45

-8tdalthrcxnydn -+&uradme 4 metrcnlcJ2zde

01 0

6 Time

24

30

(hours)

Fig. 1. The bactericidal activity of arnoxycillin, clarithromycm, cefuroxtme, and metronidazole using 2X MIC concentration against H. pylori 3236E 7. Preparation of moculum. The concentration of the 24-h broth should be in the region of log CFU/mL. The broth should be diluted sufftciently to a level of 10’ CFU/mL. Add 300 IL to both the control and test wells, giving a starting inoculum of 106CFU/mL. 8. Dispense 180~pI., aliquots of PBS into 96-well mtcrotiter trays, leaving the first column blank. 9 Take lOO-pl., samples from the control and test wells and place m the first columns 10. Take up 20 @Lfrom the first columns and add and mix in with the second column, giving a lo-fold dilution. 11. Repeat this procedure down to lo-6 dilution, replacing the plastic after each dilution step to prevent excess carry over. 12. Pick up 50 pL from each well, starting from the lowest dilution, and place 5 x 10 pL drops onto freshly prepared chocolate Columbia agar 13. This is the initial time zero reading; repeat the procedure at specific time-points, e.g , 3,6,9, l&24,48, and 72 h. 14. Incubate plates microaerobically for 6 d. The incubation period is extended as H. pylori can become dormant and the colonies take longer to grow If the plates are read too early, a false representation of the mode of action can be obtained. 15. Read colonies on the highest possible dilution, i.e., where the growth is at the optimum, but there are still single colonies rather than confluent growth; and taking into account the dilution factor, calculate the viable count and consequent rate of ktll. 16. Figure 1 is an example of a typical rate-of-kill study.

3.2.2. Rate-of-Kill Analysis (Flow Cytometric Method) 1. The experiment carried out m this procedure is essentially the same as that performed m the previous rate-of-kill analysis methodology. It differs by having a

46

2

3

4 5.

6. 7. 8. 9. 10. 11.

12 13.

McLaren higher mttial moculum (there 1sno mittal dilution of the starter moculum) and by using either growth medium twice filtered through a 0.2~pm milhpore filters or using broth supplemented with 0.5% @cyclodextrm and subsequently filtered twice through 0.2~pm filters. Prepare inoculum by harvesting a plate culture of H pylon and suspending this in 5 mL filtered growth medium m a small tissue culture flask. Incubate at 37°C microaerobically. Subculture after 24 h by transferring 1 mL of culture mto 4 mL of fresh medium and reincubatmg for 24 h. Set up an experiment as in previous methodology, m 6-well cluster trays using two and five times MIC concentrations of the compounds under test and filtered growth medium A no-drug control must also be prepared in the same medium. Incubate microaerobically for 24 h at 37°C. After 24 h, dilute the samples and no drug growth control 1: 100 in filtered broth containing FDA (see Note 4) The final concentration of FDA should be prepared at 10 pg/mL Add 10 clr, experimental to 890 $ broth and add 100 l.iL FDA Incubate samples for 13 mm at 37°C. Split the sample into two equal 500~pL aliquots. Add 0.1 mL phosphate buffered saline (PBS) beads (5 x lO”mL-I). Centrifuge samples at 10,000 rpm for 6 min. Gently resuspend pellets in the same volume of filtered PBS and keep on ice until FACS analysts. Measure viability by setting the threshold at F 1.1. The region taken to represent the border between live and dead bacteria is defined by the pomt of inflection from the scatter of the negative control. It is necessary to identify the region of the resultant scatter plot that corresponds to the profile expected from H. pylon. When the region has been defined, the FACscan analyzes the first 1000 particles in that region and sorts them according to parttcle fluorescence (see Note 5). Calculate vtabthty ratio = H. pylori events/Bead events (a) % = a/H pylon events (control)/Bead events (control) X dilution factor for control. Repeat sampling and analysis at 48 h (see Figs. 2 and 3; Table 2).

3.3. In vitro Analysis of Agents in Combination Against H. pylori 3.3.1. Analysis Using a Two-Dimensional Checkerboard in 96-Well Cluster Trays 1. Prepare moculum by growing organism as previous broth cultures, usmg 0.5% P-cyclodextrm as the growth medium. This is the medium used for all dilutions m this study. 2. The MIC for both agents should fall approximately halfway along the dilution range. Prepare the drug with the lowest MIC, agent A (most active against H. pylon) to 40 times the final concentration required in broth supplemented with 0.5% P-cyclodextrin. Prepare the other drug, agent B (least active against H. pylon), to 2 ttmes the required final concentration m the same media. Ensure at least 4 mL final volume of B

In vitro Susceptibility Testing

47

1

6

Fig. 2. Example of flow cytometric analysis of standard agents after 24 h incubation with H pylori 3236E at twice MIC value. 1, growth control; 2, gentamicin; 3, erythromycm; 4, amoxycillin; 5, clarithromycin; 6, azithromycin. Relative viabilities were calculated with respect to the growth controls. The gates were set to include the area that the peak of the growth control lies within. By overlaying control graphs onto test compound graphs, It can be seen that a tidal effect IS characterized by a shift of the graphs to the left. 3. Double-dilute agent B 2 + 2 mL 11 times in growth medium. 4. Agent B is then added to the cluster trays, leaving the first row completely empty: 190 & of each dilution is added to column 2, 100 & of each dilution to columns 1 and 3-12. 5. 100 $ of growth medra (contaming no drug) is added to the first row, except column 2, where 190 & is added. 6. Add 10 mL agent A to column 2.

McLaren

48

+ + -w-I+ + +I-

0

12

24

36

Control Gentamicin Erlthromycin Amoxyclllin Clariihromycir Azthromycin

48

Time (h)

Fig. 3. Cldal activity of the same experiment using viable counting.

Table 2 Calculated Percentage Viability after Flow Cytometric Analysis of Standard Agents. (Analysis of Cidal Activity by FACscan)B Treatment pg.g/mL Growth control Gentamycm (0.5) Erythromycin (0.12) Amoxyclllm (0.25) Clarithromycin (0.12) Azithromycin (1) No FDA control

24-h incubation H. pylori events

Bead number

Ratio

Viability, %

8180

847

9.92

100

6765

866

7.81

78

5314

1034

5.13

51

4590

1641

2.8

28

338

2208

0 15

1.5

304

2371

0 13

13

207

1888

0 11

11

OAll vlablhtles calculated with respect to growth control

7. Double-dilute agent A in the wells: Take up 100 mL from all the second column wells and add and mix to the next column down. Repeat until all the wells are diluted.

In vitro Susceptibility Testing

49

8. The checkerboard is now complete and ready for inoculation. The no drug control being at the top of the first column, the first row being drug A only, and the first column being drug B only. 9. Dilute the H. pylon broth culture 1:lO and add 100 pL to each well. Incubate 24 h microaerobically. 10. Prepare chocolate Columbia agar plates for sampling Ensure plates are at least 120 x 90 mm. 11. Sterilize a 96-prong multipomt inoculator. Sample the experimental microtiter tray and inoculate the agar plates. 12. Incubate the inoculated plates for 5 d microaerobically 13. Repeat after 48 h. 14. The plates can be read and scored by assessing the amount of growth on each spot. No growth = 0; l-5 colonies = 1; 6-10 colomes = 2; 11-15 colonies = 3; 16-20 colonies = 4; ~20 colonies = 5; confluent growth = 6 15. Data analysis. Results can be analyzed by calculating the FBCs and consequent FBC index and mean FBC index. Synergy = mean FBC index of ~0.5; an additive effect = mean FBC index of 0.5-i; and antagonism = mean FBC index of > 1. FIC index = FIC A + FIC B, where FIC A = MBC A in presence of B/MBC A and FIC B = MBC B in presence of A/MBC B The mean FIC index is the arithmetic mean of the FIC index. 16 An example of an mteraction of two agents in combination is given in Fig. 4

3.3.2. Rate-of-Kill Analysis of Drugs in Combination Using Miles-Misra Technique 1 The methodology is identical to that employed in Section 3 2 1 except for a few modifications. 2. Since two drugs are being used instead of one, add 100 IL less growth medium. 3. Add the agents at MIC, one-half and one-quarter times MIC. 4. A 2 log drop in viabthty in the combmation with respect to the agents singly IS indicative of synergy.

3.3.3. Rate-of-Kill Analysis of Drugs in Combination Using Flow Cytometry 1. The methodology is identical to Section 3.2.2. The modificattons are as for Section 3 3 2. 2. An advantageous interaction can be determined by a decrease m percentage viability with respect to single agents.

4. Notes 1. When considering strains to be analyzed it is best to use fresh hospital isolates together with at least one standard strain, e.g., H pylon NCTC 11637. It may

McLaren

50 GR1223 11X/CLAR

.

24h

.

00

Fig. 4. A 3D representation of the interaction of 2 drugs in combination against iY pylon 3236E The 2 drugs under evaluation are clarithromycm and GR1223 11X, Ranitidine bismuth citrate. At the concentrations under test, neither drug by itself is able to be totally effective agamst the organism. In combmatton, the 2 agents are much more tidal, totally eradicating H pyfori after 24 h.

2.

3.

4.

5. 6.

also be useful to include nitroimidazole and macrolide translation inhibrtor-resistant strains. For rate-of-kill studies, the age of the culture and consequent growth phase may affect the rate of kill. It may be useful to test agamst a variety of dtfferently aged H pylon broth cultures to obtain a full picture of the rate of kill of specific agents. The age of the cultures in the method for Section 3.2.1. is a 24h culture Older cultures, such as 72-h broths, will still attam high enough viability to assess the rate of kill. A large percentage of H pylori will now have become coccoid forms. As H. pylori is a fairly fastidious microaerobe, tt 1s important not to dilute the starter broths too heavily before addition to test wells. Over-dilution will result m H pylori not growing in the no-drug controls or any of the test wells. Keep the startmg moculum above lo5 CFU/mL for all of the studies. FDA is not normally used for analysis of Gram-negative organisms by flow cytometry, but is excellent for the analysis of H pylori viabihty. Propidium iodide was not taken up by H. pylon and no subsequent fluorescence was observed. When analyzing flow cytometry traces, a shift of the peak to the left 1s mdtcative of increase in tidal activity against H. pylori. In all of the studies, it IS useful to add a control test drug to momtor the performance of the test and to ensure reliability of the data generated.

In vitro Susceptibility Testing

51

References 1. McNulty,

C. A. and Dent, J. C. (1988) Susceptibility of clinical Isolates of agents. Eur. J. Clan Mcrobiol

Campylobacterpylori to twenty-one antimicrobial Infect. Dis. 7,%X-569

2. Lambert, T , Megraud, F , Gerbaud, G., and Courvalm, P. (1986) Susceptrbrlrty of Campylobacterpylorrdzs to 20 antimtcrobial agents. Antlmlcrob. Agents Chemother. 30,510,511.

3. Goodwin, C. S., Blake, P., and Blincow, E. (1986) The minimum inhibitory and bactericidal concentrations of antibiotics and anti-ulcer agents against Campylobacterpyloridis. J. Antimicrob. Chemother. 17,309-3 14. 4. Suerbaum, S., Leying, H., Klemm, K., and Opferkuch, W. (1991) Antrbactenal activity of pantoprazole and omeprazole against Helicobacterpylon. Eur. J Clm Microblol. Infect. DES. 10,92,93.

5. Millar, R., M., and Pike, J (1992) Bactericidal actrvtty of antimicrobial agents against slowly growing Helicobacterpylori. Antimicrob. Agents and Chemother. 36(Nl), 185-187. 6. Moellering, R. C. (1979) Antimicrobial synergism-An elusive conept. J Infect Dis 140(N4), 639-641.

7. Goodwin, C S , Marshall, B. J., Blincow, E D , Wilson, D. H , Blackbourn, S , and Phrhps, M. (1988) Prevention of mtroimidazole resrstance in Campylobacter pylorz by coadmmistration of collordal bismuth subcitrate: clinical and in vitro studies. J. Clm Path01 41,207-2 10

Defined Media for H. pylori David J. Reynolds 1. Introduction Helicobacter pylori has been widely studied since its discovery in 1982 by Marshall and Warren (I), but many aspects of its structure, metabolism, and physiology, including its specific growth requirements, are still largely unknown, The organism is generally grown in complex media containing tissue extracts (e.g., of brain or heart) or proteolytic enzyme digests of meat or casein supplemented with blood or serum, which has made the metabolic pathways utilized by the bacterium difficult to determine. For example, although early studies based on acid formation from sugars and detection of preformed enzymes found no evidence of saccharide fermentative pathways (2,3), more recent evidence mdicates that H. pylori does indeed catabolize sugars. In a series of studies, Mendz, Hazell, and colleagues found clear evidence for the pentose phosphate pathway (4), glucokinase activity (5), and the fermentation of glucose to lactate (6). Unfortunately, the approach used by these workers (‘H and t3C nuclear magnetic resonance spectroscopy) is expensive to set up and use, which restricts its wider availability to researchers studying H. pylori metabolism, Here, a simple defined medium is described that is capable of supportmg growth comparable to that achieved in complex media. Each of the individual constituents can be selectively omitted or further components added and the effect on the growth and morphology studied. The medium was developed using tissue-culture medium (RPMI, Life Technologies, UK) as a basic chemically defined medium, and a series of components was added until the medium would support growth. These components were chosen based on the published analysis of Isosensitest broth (Umpath, Basingstoke, UK), a medium used in this laboratory to grow H. pylori. The resulting supplemented medium was then prepared de ylovo from reagent grade From Methods m Molecular Medicme, Hehcobacter pylon Profocols Edlted by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

53

NJ

Reynolds chemicals and then simplified by sequentially removmg each of the key ingredients and noting the effect on growth. Initial studies during the development of the defined medium highhghted lipoic acid, FeS04, adenme, bovine serum albumm (BSA), and nonessential ammo acids as key ingredients. Some of these are common additions to media used to grow other bacteria. However, the requirement for BSA, a protein known to act as a carrier for long chain fatty acids (LCFA), may indicate either that LCFAs are required by H pylon, if they are present m the BSA used in this study, or that LCFAs are toxic and it is their removal by BSA that enhances growth. Alternattvely, other required growth factors may be either carried by BSA or present as an impurity. The pattern of growth in the defined medium is similar to that of complex media with logarithmic growth to approx 1 x log CFU/mL followed by a short stationary phase and a rapid death phase. This is typical of complex media studied by other workers, e.g., Mueller-Hinton broth (7) and Brucella broth (8,9). However, a major difference is that, m the defined medium, the rate of death m the death phase was lower and the conversion to the coccoid form reduced from 99 to 50% (10). This IS a potentially important observation, as it shows that conversion to the coccoid form can be regulated by changes induced by altered nutrition. Environmental factors are known to modulate the expression of vuulence determinants (I I), and similar processes might also be involved m regulating the coccoid transformation. 2. Materials 1. The strainsthat have been grown m the defined medium are: NCTC 11637 (type strain), Roberts (clinical isolate from Manchester, UK), and 8 isolates from stomach biopsies collected during 199 l-l 992 m Birmingham, UK Two other clmical isolates failed to grow m the medium, but these also grew poorly in standard media. All strains were mamtamed on Isosensitest agar (Unipath) containmg 2% heat-mactivated newborn calf serum (Life Technologies) 2 Inorganic salts solution: (This is a 4X workmg concentration aqueous solution, at 5X concentration, precipitation can occur.) 0.4 g/L Ca(NO& * 4Hz0, 1 6 g/L KCl, 0.4 g/L MgS04 7Hz0, 24 g/L NaCl, 3.2 g/L Na2HP04 anhydrous Stenhze by membrane filtration (0.2~pm membrane bottle filter [Costar, Cambridge, MA or Sartorius, Goettingen, Germany], 0.2~pm syrmge filters [Gelman Sciences, Northampton, UK or Sartonus, depending on volume]), and store at 4°C for up to 6 mo (see Note 1). 3 7 5% NaHC03 solution’ This IS a 37 6X workmg concentration solution Either bought from Life Technologies or made up m lOO-mL lots of 7.5% (w/v) m water and sterilized by membrane filtration. Store at 4°C for up to 6 mo. 4 FeS04 solution: This is a 1000X working concentration solution 200 mg m 100 mL of water acidified with 100 pL concentrated HCl, sterilization is by membrane filtration Store at 4°C for up to 6 mo.

Defined

55

Media

5. 10% Glucose solution: This is a 50X working concentration solution. 10 g glucose m 100 mL water. Sterilize by membrane filtration, store at 4°C for up to 6 mo 6. 0.05% Phenol red: This 1sa 100X working concentration solution. 25 mg phenol red in 50 mL of a 20% (v/v) solution of propan-2-01 in water. Sterilize by 0.2 pm membrane filtration, store at 4OC for up to 6 mo (see Note 2). 7. 14% Bovine serum albumin solution: This is a 28X working concentration solutton. 14 g of BSA is dissolved in 80 mL of water by stirring overnight at 4°C then the volume made up to 100 mL. The solution is prefiltered through a Millipore GF/C filter (Millipore, Watford, UK) prior to sterilization by membrane filtration. Store at 4°C for up to 1 mo (see Note 3). 8. Adenine: For large volumes of medium, this can be added directly at 50 mg/L and the medium sterilized by membrane filtration Alternatively, a stock of 6.5 mg/L can be made (130 mg m 20 mL, requires heating to dissolve) and filter sterilized, This stock can be stored at -20°C for up to 12 mo. It is best to make several ahquots of 4 mL (3.85 mL needed for 500 mL medium) as crystallization occurs at low temperatures and heating is necessary to dissolve the crystals. 9. Lipoic acid (thioctic acid): This is a 3333X working concentratton solutton 20 mg in 2 mL 100% ethanol, sterilization of the solution is not necessary Store at -20°C (to avoid evaporation of the solvent) for up to 2 mo. 10. Ammo acids solution: These are 100X working concentration solutions made from tissue-culture grade reagents (Sigma, Poole, UK). All of the amino actds are soluble in water at the concentrations specified in Table 1, but some may need persuading by adding small amounts (e.g., lo-pL lots to a 20-mL vol) of HCl and gentle heating (5O’C). Store at -20°C for up to 6 mo; it is not necessary to sterilize each individual amino acid if the final medium is to be sterilized (see Section 3.) Avoid freeze-thaw cycles with glutamine by freezing a reasonable number of aliquots initially. 11. Vitamins solution. This is a 100X workmg concentration solution prepared from tissue-culture grade reagents (Sigma), as listed in Table 2. Most of the vitamins are readily soluble in water, but the folic acid should be dissolved in 0.2MNaOH before addmg; riboflavin needs heating, and P-ammobenzoic acid should either be warmed or predissolved in 100% ethanol. Steriltze by membrane filtration and store at -20°C in the dark (cyanocobalamin is light-sensitive) for up to 6 mo

3. Methods 3.1. Preparation

of the Defined Medium

1. For 500 mL of medium, mix the ingredients in the following order (see Notes 4 and 5): 125 mL of 4X Inorganic salts, 10 mL of 50X glucose, 3.85 mL of adenine solution, 0.15 mL of 3333X lipoic acid, 5 mL of 100X vitamins, 200 mL of water, 5 mL of each of the 20 amino acid solutions (100X), 5 mL of 100X phenol red (or 5 mL water, if indicator is not required), 17.85 mL of 28X BSA, 0 5 mL of 1000X FeS04, 13.3 mL of NaHC03; adjust the pH to 7.4. Make up to 500 mL with water (this will take approx 19.5 mL) and recheck the pH

Reynolds

56 Table 1 Stock Solutions of Individual Amino Ammo acid Alanine HCI Arginine Asparagine Aspartic acid Cystine Glutamic acrd Glutamme Glycine Histldme Isoleucine Leucme Lysme Methionme Phenylalanine Proline Serine Threonine Tryptophan Tyrosine Valine

acids (100x) Concentration, mgl20 mL 89 1264 150 133 240 147 600 75 220 525 524 725 151 330 115 105 476 102 360 468

2. Sterilize by filtration through a 0.2~pm membrane filter. The easiest means of doing this is to use a 500-mL bottle filter connected to a vacuum pump. Check a sample of the finished medium for sterihty by incubation for 48 h at 37°C and then store at 4°C for up to 2 mo. 3. Where media that are deficient in a single component (e.g., an amino acid) are required, stock concentrates containing all of the components except amino acids can be made up and the individual amino actds added, substttutmg water for the ammo acids omitted. This results m a large number of media (20 in the case of amino acids) and, when volumes are small (up to, e g., 30 mL), syringe filters are useful for sterilization.

3.2. inoculation

of the Medium

1. Grow the culture overnight on agar plates (see Section 2, step 1) and then scrape the culture from the agar surface, using a sterile plastic loop (Nunc, Roskllde, Denmark). Suspend the culture m l-2 mL of Hanks balanced salts solution by vortexing in a sterile mlcrofuge tube, centrifuge at 12,000g (MSE

57

Defined Media Table 2 Stock solution of vitamins (100x) Vitamin

Weight, mg

D-Biotin Choline chloride Folic acid Myo-inositol Niacinamide p-aminobenzoic acid n-pantothenic acid Pyridoxine hydrochloride Riboflavin Thiamine hydrochloride Vitamin B r2 (cyanocobalamin) Water

2 30 10 350 10 10 12.5 10 2 10 0.05 to 100 mL

Microcentaur, Crawley, UK) for 1 mm, and resuspend in the same solution. Standardize to approx 1 x lo* cells per mL by counting in a Helber chamber (Weber Scientific Lancing, UK) and diluting by the appropriate factor. 2. Prewarm the medium and allow to equilibrate m a microaerobic environment. Inoculate at 1 in 100 dilution, i.e., for a lOO-mL conical flask with 15 mL of medium, add 150 pL of suspension; for a 24-well tray with 1 mL of medium per well, add 10 $,/well. 3. Manipulation of the culture and maculation should be carried out in a sterile au environment, preferably a Class II safety cabinet. 4. Immediately after inoculation, transfer the cultures back to the mtcroaerobic environment, preferably a variable atmosphere incubator set at 5% CO*, 5% O2 balance Nz (VAIN, Don Whitley Scientific, Shipton, UK). Flasks should be shaken gently (1 . 18g) (IKA Vibrax VXR, Janke & Kunkel, Germany); 24-well trays (lidded trays from Nunc) should be shaken more vigorously (3.46g).

3.3. Growth in the Medium 1. Growth over time can be assessed by the standard techtuques of measurement of absorbance at 600 nm (OD6sc,),using l-n& samples m semimicro plastic cuvets, by serial dilution viable count on Isosensitest agar (Unipath) plates, or by phase contrast microscopy, using a Helber chamber (Weber Scientific). Microscopic examination also allows the proportions of coccoid and rod-shaped cells to be assessed (see Notes 6-14) 2. When 24-well trays are being used, volumes will generally be too small to allow use of semimicro cuvets. It is convenient to transfer lOO-200~pL vol to flexible plastic 96-well trays (Falcon, Los Angeles, CA) and to assay the OD600 on a Tltertek Multtskan plate reader (Flow Laboratories, Rickmansworth, UK) or equivalent.

58

Reynolds

4. Notes 1 All the morgamc chemicals and glucose can be obtained from BDH, Poole, UK. The amino acids (tissue-culture grade), vitamins, and other organic chemicals are from Sigma. Water was deionized-distilled water (Millipore) 2 The phenol red indicator IS parttcularly useful if the medium has been made aseptically from sterile ingredients, as tt is possible for the experienced eye to determme the pH with reasonable accuracy from the color of the medium. This saves time if many media with small differences m composition are being made, e.g , omission of single amino acids 3. The bovine serum albumin (BSA) can be obtained either from Sigma (Fraction V, catalogue no. A-7906 ) or Advanced Protein Products, Brierley Hill, W. Midlands, UK (catalogue no. PF-201-47). 4. Water is added early in the preparation of the medium and the bicarbonate added last to avoid problems with precipitation. The precipitate does redissolve, but this slows the procedure considerably. 5. It is best to prepare the medium in a lammar flow hood, or other sterile air environment, to avoid contamination of sterile stock solutions. 6. The medium in the wells of a 24-well tray will evaporate during mcubation. For optimal shaking, 0.5 mL is ideal, but this will only permit 24 h of mcubation before volume loss becomes a problem. Using 1 mL/well allows incubation for 48 h, but shaking and thus gas exchange are less effective. Edge effects, i.e., disproportionate evaporation from wells at the edge of the plate, can be pronounced. 7 Liquid cultures must be shaken; growth in static cultures is negligible. 8 For all cultures, but particularly those grown in 24-well trays, where contammanon is more likely, it is important to confirm growth as H pylori by comparison of their protein profile, using sodium dodecyl sulfate polyacrylamide gel electrophoresis (12) analysis on minigels, with that of the moculum 9 As soon as the culture begins to go coccoid, the ODhoOwill drop This is because the shape change alters the scattering of the incident light. The drop in ODbOalags behind the drop m CFU count by several hours, suggesting that the loss of cultivability precedes the change m shape. 10. In the defined medium, there is a tendency for some strains to clump. This is more severe at higher BSA concentrations. 11 There is no discernible change m the pH of the medium during growth, as evidenced by a lack of change in the color of the phenol red Indicator 12. Inclusion of xanthme, guanine, or uracil m the medium was detrimental to growth Addition of hemoglobin to the medium had no effect on growth. 13. H pylon strains Roberts and 11637 were both sequentially passaged m the medium five times and attained similar growth at each passage. A large moculum was necessary to achieve growth during serial passage Typically, this was 0.75 mL of a 24-h culture (about 2.5 x lo7 bacteria) mto 15 mL of medium, and growth gave a 20-fold increase in numbers of bacteria at each passage of 24 h 14. Growth m the defined medium is comparable to that in a complex medium (Isosensitest broth plus 2% heat-inactivated newborn calf serum) m terms of

Defined Media

59

50

m I I 100 150 200 Time (h post inoculation)

Fig. 1. Comparison of the growth of H pylorz in a complex medium and the defined medium. Growth was assessed as CFU, q , and total count, 0, for strain Roberts grown m (A) Isosensitest broth +2% heat-inactivated newborn calf serum, and (B) the defined medmm. Representative of three experiments. maximum growth rate (Fig. 1, doubling time during exponential growth phase for defined medium = 2 h 33 min, for ISB + FCS = 2 h 37 min). The maximum total counts and CFU counts were also similar at approx 1 x IO9 bacteria per mL. There was a difference m the death rate: the mean halving time (time for the number of CFU to halve) of the complex medium was lower than that for the defined medium (1 h 51 min compared to 2 h 4 min, respectively); i.e., in the defined medium, bacteria became noncultivable less rapidly The slower death of the culture in the defined medium was reflected in the rate of conversion to the

Reynolds

60

20 26 44 51

75 96 114 141147 169

u

44 64 71 138 Time (h post inoculation) Fig. 2. Coccoid development in a complex medium and the defined medium. Growth was assessed as CFU, Cl, and differential mtcroscopical counts of the rod (hatched columns) and coccoid (clear columns) forms of the organism Representative of two experiments

nonculttvable coccotd form of the organism (Fig. 2). In the defined medium, conversion was slower and only about 50% of the culture became coccotd even after prolonged mcubatton. In contrast, m the complex medium, conversion was rapid and over 99% of the culture became coccoid.

References 1. Warren, J. R. and Marshall, B. J. (1983) Unidentified curved bacilli on gastric eptthelmm m active chronic gastritis Lancet 1, 1273-1275

Defined Media

61

2. Marshall, B. J. and Warren, J. R. (1984) Unidentified curved bacilli m the stomach of patients with gastntts and peptic ulceration. Lancet 1, 13 11-13 15 3. McNulty, C. A. and Dent, J. C (1987) Rapid identification of Cumpylobacter pylori (C. pylondls) by preformed enzymes. J. Clan. Mlcrobiol. 25, 1683-1686. 4. Mendz, G L. and Hazel& S. L. (1991) Evidence for a pentose phosphate pathway m Hellcobacterpylorl. FEMS Mxroblol. Lett &I,33 l-336 5 Mendz, G. L. and Hazell, S L. (1993) Glucose phosphorylation in Helzcobacter pylon. Arch Blochem. Bzophys. 300,522-525

6. Mendz, G. L , Hazell, S. L., and Bums, B. P. (1993) Glucose utilization and lactate productton by Helicobacterpylori. J. Gen. Mlcroblol. 139,3023-3028. 7. Sorberg, M., Nilsson, M., and Nilsson, L. E. (1993) Morphologic conversion of Helicobacter pylorl from bacdlary form to coccoid form evaluated by biolumtnescence, microscopy and viable count. Acta Gastro-Enterol Belg. Suppl 56,52. 8 Catrenich, C. E and Makin, K. M (1991) Charactertzatlon of the morphologic conversion of Helicobacter pylorl from bacillary to coccotd form. &and. J Gastroenterol

26(Suppl.

Ml),

58-64.

9. Cellini, L., Allocati, N., Di Camph, E., Masulli, M., and Dainelli, B. (1993) Morphological forms in Helzcobacterpylorl Acta Gastro-Enter01 Belg Suppl 56, 108. 10 Reynolds, D J and Penn, C. W. (1994) Characteristics of Hellcobacter pylorl growth in a defined medium and determination of its ammo acid requirements. J. Gen. Mlcroblol.

140,2649-2656.

11. Mekalanos, J. J. (1992) Environmental signals controllmg expression of virulence determinants in bacteria J Bacterrol. 174, l-7 12. Laemmli, U. K (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680485.

8 Studies of Respiration

in If. pylorl

Roger J. Miles and Huai-Tzu Chang 1. Introduction In the procedures described here, respiratory metabolism is estimated by direct determination of oxygen uptake, measured as a reduction in the dissolved oxygen tension (DOT) of cell suspensionsor extracts. Continuous monitoring of DOT is achieved using a Clark-type oxygen electrode linked to a chart recorder. The volume of cell suspension required 1sof the order of 1 mL (0.5-l mg cell protein mL) and the method is sufficiently sensitive to detect the consumption of a few nmol oxygen per mL. This sensitivity and the rapidity with which measurements may be made makes the technique particularly useful for the study of organisms, such as Helicobacter pylon, which give relatively poor growth yields and which, being microaerobic, are likely to be more suscepttble to prolonged experimental procedures at high oxygen tensions. To date, only a small number of Helicobacter studies have directly measured oxygen uptake. However, it has been shown that H. pylorz cells growing in broth media do consume oxygen (I) and that both whole cells and cell membrane fractions, suspended m buffer, oxidize a number of organic substrates (1,2). In these studies, oxygen uptake was monitored by DOT measurement. Thus, the technique may be applicable for monitoring of Helicobacteer-inhibitor interactions, determination of btochemlcal diversity, and fundamental investigations of metabolism and its regulation. Also, since the technique provides kinetic data, it may enable assessmentof the likely significance of substratemetabolism, at the substrate concentrations found m the organism’s natural habitat. Polarographic measurement of DOT to monitor oxygen uptake has been widely used m biological systemsand the general principles of the method and the derivation of rates and quantities of oxygen uptake have been described previously (e.g., ref. 3). The methodology that follows only briefly describes From Methods m Molecular Medrone, Hehcobacter pylorl Protoco/s Edited by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

63

NJ

64

Miles and Chang

the setting up of the apparatus used, and is principally a description of cell preparation and other experimental conditions that have been successfully used to monitor I-Z. pylori respiration. Many of the variables m the protocols have not been systematically studied. However, cell background has been demon-

strated to be particularly important m studies of substrate oxidation by washed cells suspended m buffer (I). In the absence of added substrate, cells derived from stirred broth cultures or from agar plates gave high and variable rates of oxygen uptake, which could be sustained for several hours. In contrast, under similar conditions, oxygen consumption by cells from statically incubated cul-

tures was transient or undetectable. Though in general terms, it would appear preferable to use cells from homogeneous s&red/shaken cultures, m practice this may not be the case, perhaps because shear forces damage cells, causing

leakage of metabohzable substances.Cells harvested from agar plates may also contain a proportion of nongrowing and lysing cells, A further important feature in the preparation of cell suspensionsis the total time required. The procedure

descrtbed aims to complete centrifugation, washmg, and resuspenstonprocedures rapidly, within about 15 min, so as to retam maximum metabolic activity. 2. Materials 1. Bacterial inoculum and suitable liquid and agar growth media (see Note 1). 2. Variable atmosphere incubator (e g., VAIN cabinet, Don Whitley Sctentific, Shipton, UK) or anaerobic jar plus gas generating lut (5-10% [v/v] O2 and 10% [v/v] COz). 3. PR buffer: Sterile quarter-strength Ringer solution (Oxoid, Basingstoke, UK) containing 50 mM PIPES buffer, pH 6.5 (see Note 2). 4. TRIS-phenylmethylsulfonyl fluoride (PMSF) buffer: 20 mM Tris-HCl, pH 7 0, contannng 100 uA4PMSF (for preparation of cell extracts only). 5. Mtcrocentrtfuge (MSE Micro Centaur or equivalent) and sterile microcentrifuge tubes. 6. Ultracentrifuge (for preparation of cell extracts only) 7 Water-jacketed, magnetically-stirred, flat-bottomed reaction vessel(s) for use with oxygen electrode (see Notes 3 and 4) 8. Oxygen electrode (Notes 3 and 4), Teflon membrane, 3M KC1 solution, and sodium dithionite. 9 Thermocnculator 10 Chart recorder (see Note 4) 11 Metabolic substrates (filter sterilized); e g., sugars, organic acids, alcohols (5-500 pko 12. Microsyrmge(s) for accurate delivery of 5-50 pL substrate solution

3. Methods 3.1. Cell Growth and Preparation

of Who/e Cell Suspensions

1. Inoculate agar plates (see Note 1) and incubate at 37°C for 2-3 d in a suitable atmosphere For most purposes, 5-l 0% O2 and 10% CO2 m nitrogen is appropri-

Respiration in H. pylori

65

ate and may be obtained using an anaerobe jar with a gas generating kit (e.g., Oxoid, BR56) or a variable atmosphere incubator. 2. Dispense hqutd medium (50 mL; see Note 1) in 500-mL flat disposable tissue culture flasks (Nunc or similar) and inoculate with a loopful of a fresh plategrown culture (see Note 5). Incubate flasks statically at 37°C in an appropriate atmosphere and monitor OD,,,. 3. For optimal yield of metabolically active cells, harvest cultures toward the end of the exponential growth phase, i.e., after approx 72 h for cells grown m bramheart infusion (BHI) broth supplemented with 10% (v/v) horse serum 4. Dispense culture in 1.5-mL microcentrifuge tubes and centrifuge m a microcentrifuge at 13,000g for 3 mm. Carefully remove the supernatant and resuspend the cell pellet in PR buffer using a Pasteur pipet. Using a similar procedure, wash the cells twice and finally resuspend in PR buffer at approx 1.5 mg cell protein/ml. The volume of cell suspension required is l-2 mL per oxygen electrode system, plus a sufficient sample for determination of cell protein. The total time for the preparation of suspensions should be kept to a minimum (approx 15 mm). Cell suspensions should be used fresh as activity declines during storage (see Note 6).

3.2. Preparation

of Cell Extracts

1. Membrane fractions with high specific rates of substrate (NADPH, succinate) oxidation have been prepared from cells grown on chocolate agar plates contaming 5% (v/v) horse blood. Inoculate plates with 0.1 mL of a broth culture and incubate at 37°C in a suitable atmosphere. When growth has become confluent (48 h), wash cells off plates using TRIS-PMSF buffer; cell suspensions may be stored frozen (-20°C) at this stage. 2. Thaw cell suspensions on ice; sonicate cells using the MSE Soniprep 150 at maximum amplitude for 8 x I-min periods, separated by l-mm intervals Centrifuge at 700g for 10 mm to remove cell debris, then centrifuge the supernatant at 100,OOOg for 40 min at 4°C to obtain the cell membrane fraction. Wash membranes three times (4V) and resuspend in PR buffer. Store as aliquots at -70°C

3.3. Preparation

of Oxygen Electrode and Reaction

Vessel

1. The detailed procedure for assembling the oxygen electrode system will vary according to the design of the apparatus; manufacturer’s mstructions should be consulted. In the Rank Brothers (Cambridge, UK) instrument, the electrodes (a platinum working electrode and an Ag-AgCl reference electrode) are mounted beneath the cylindrical reaction vessel (maximum volume, 5 mL; diameter, 15 mm) and are separated from the contents of the vessel by a semtpermeable Teflon membrane. Contact between the electrodes is maintained by fine tissue (e.g., lens tissue) saturated with 3M El; the membrane is held in place using a rubber washer. The tissue has a small central hole, which is placed over the central platinum electrode to enable direct contact with the membrane In mstruments of this type, briefly rinse and dry the electrodes, cut and fit a new piece of tissue, add

66

Miles and Chang

sufficient KC1 (3M) to wet the electrodes, and cover with a new piece of membrane Thoroughly clean and wash the reaction vessel with sterile distilled water and assemble the system. 2. Connect the thermoctrculator to the water Jacket of the reaction vessel and the output of the oxygen electrode to a chart recorder. Set the polarizing voltage of the platinum electrode (relative to the reference electrode) to -0.6 volts; at this voltage, current is directly proportional to DOT m the reaction vessel 3. Calibrate the electrode with air saturated water (210 nmol dissolved O*/mL at 37°C; for other temperatures, see ref 4) obtained by simply adding stertle distilled water (that has not been shaken, thus avoiding super-saturation) to the reaction vessel and magnetically stn-ring. The position of the baseline (zero current) may be confirmed by adding a few crystals of sodmm dithionite, which will reduce DOT to zero within a few seconds. It is convenient to set up the chart recorder so that an-saturated water gives a 90% deflection. 4. Remove water from the reaction vessel usmg a vacuum lme or Pasteur pipet, taking care not to damage the membrane. Wash the vessel if sodium dithionite has been used.

3.4. Measurement of Oxygen Uptake I. Add 1-2 mL of cell suspension or extract or culture to the oxygen electrode vessel. Set the magnetic stirrer speed to give adequate mixing with a high but stable speed. Do not alter the speed during the course of DOT measurement. Allow up to approx 10 min for temperature equilibration and for the cell suspension to become saturated with oxygen, and then isolate from air In the Rank Brothers instrument, thts is achieved using a cylmdrical plug with a fine central pore (1 mm diameter). Adjust the level of the plug so that the meniscus of the cell suspension or extract IS within the central pore, essentially ehmmatmg oxygen transfer to the suspension. 2 Allow sufficient time to enable the initial rate of oxygen uptake to be accurately determined (see Note 7) and add test substrates (see Note 8) or mhibltors (2-10 & of appropriate concentrations), via the central pore m the plug, usmg a microsyrmge. In experiments with washed cell suspensions, when there 1s complete substrate utilization, the metabolism of a number of substrates can be followed sequenttally. The contents of the electrode vessel may be reoxygenated as required by raising the plug and allowing air to enter. The Km value for oxygen appears low (I), substrate metabolism would be expected to be mdependent of DOT m the range of 5100% of saturation. In PR buffer, there is little reduction in the viability of H. pylori (strain NCTC 11637) cell suspensions incubated at 37°C for up to approx 1 h, but thereafter viability declines rapidly and 1s reduced by more than 50% at 3 h (see Note 6). In experiments with cells suspended in growth medium, any oxygen consumption by the unmoculated medium should be determined BHI medium, containing serum, consumed oxygen at an appreciable rate, which was highest in freshly prepared medium and at high DOT (I).

67

Respiration in H. pylori 3.5. Calculation

of Kinetic Parameters

of Substrate

Utilization

1. Provided metabolic intermediates do not accumulate during substrate metabolism, the rate of change in DOT will be proportional to the rate of substrate utthzation, and reaction stoichiometries can be determined from the total reduction m DOT (3). If tt is assumed that the relationship between the rate of substrate metabolism (v) and substrate concentration (s) follows Michaehs kinetics, the saturation constant (K, value) and maximum rate of substrate metabolism (I’,,,,,) may also be estimated from plots of l/s against l/v, in whtch the intercepts on the ordinate axis and abscissa are l/V,,,,, and -l/K,,,, respectively. Where K,,, values are high, the rate of metabolism (proportional to rate of change in DOT) may be determined at different initial values of s. Alternatively, where Km values are low, kinetic data may be estimated from the analysis of smgle curves representing the complete utilization of substrate. It is necessary to assume that change m DOT at any time (t) after substrate addition IS proportional to the substrate used Thus, substrate concentration at time t is: (initial substrate concentration) x (a - b)la where a is the total change in DOT following substrate addition and b the change at time t. 2 The metabolic activtty of cell suspensions or extracts may be related to viable count or cell protein. However, the spectfic activities of washed cell suspensions prepared on different days will vary, unless growth condittons, inoculum preparation, and cell harvesting procedures are rigorously controlled One way of overcoming this is to inmate all individual experiments by the addition of a low concentration of a rapidly metabolized control substrate, e g , pyruvate (I), and to express rates of metabolism of substrates added subsequently, relative to the rate for this substrate. The control substrate may also be added at the end of experiments to confirm that metabolic activity has not declined

4. Notes 1. Chocolate blood agar (prepared by adding 10% [v/v] horse blood to Columbta agar base and heating at 70°C for 20 min) is a suitable solid medium. BHI broth, supplemented with 10% (v/v) horse serum, may be used as a liquid medmm. However, it may be preferable to grow organisms in recently described defined media (X6). Also, Deshpande et al. (7) describe a simple procedure for modifymg the gas phase of cultures grown in complex media, which leads to increased growth rate and yield. 2. The activity of cell suspensions was increased when catalase (160 U/mL) was included in PR buffer used for washing and suspending cells. 3. A suitable oxygen electrode system is available from Rank Brothers that incorporates a water-jacketed glass or perspex reaction vessel. 4. The most suitable experimental arrangement is to have two oxygen electrode systems linked to a dual channel chart recorder; this allows experiments to be conducted simultaneously with an appropriate control and simplifies comparison of recorder traces.

Miles and Chang

68

5 To improve reproducibility of cell preparattons, medmm may be maculated with liquid cultures (0.5 mL per 50 mL medium); these should be at a well-defined stage of growth, e.g., exponential phase cells that have reached a particular cell density. The use of inocula stored at -70°C or in liquid nitrogen (--196’(Z) might also be considered. Homogeneous cell suspensions (usually with a cryoprotectant, e.g., 1.5M glycerol) are dispensed in ampuls and Frozenand thawed at precisely controlled rates, to ensure reproducibility. The use of low-temperature storage to provide standardized inocula has been successfully used in metabolic studies with many organisms 6. Storing cell suspenstons for 4 h at room temperature or overnight at 4°C or -20°C reduced oxygen uptake rates in BHI broth by 30,90, and 96%, respectively 7. For growing cells (i.e., suspended in growth media), the rate of oxygen uptake prior to makmg additions is essentially constant. Ideally, washed whole cell suspenstons in the absence of added substrate should consume oxygen only at low rates. Membrane preparations do not consume oxygen without added substrate. 8. The substrate concentratton necessary to give detectable metabolism will vary with the affinity constant (K, value) toward the selected substrate. Values of K, for metabolism of, for example, pyruvate and succinate by the type-strain of H pylorz were low (K, < 10 @4), indicating a high afftmty. For these substrates, initial concentrations of 50 fl will give close to maximal metabolic rates. However, substantially higher concentrations are needed when K, values are high, as, for example, with ethanol (K,,, > 5 mM). In general, it is advantageous to use low substrate concentrations, since, followmg substrate exhaustion, additions of further different substrates may be made, increasing the mformation available from each experimental run.

References 1. Chang, H. T., Marcelli, S W., Davison, A. A , Chalk, P. A., Poole, R. K , and Miles, R. J. (1995) Kinetics of substrate oxidation by whole cells and cell membranes of Helxobacter pylori. FEMS Mlcroblol Lett 129,33-38. 2 Baer, W., Koopman, H , and Wagner, S. (1993) Effects of substances inhibiting or uncoupling respiratory-chain phosphorylation of Hellcobacterpylorl Zbl Bakt 280,253-258. 3. Estabrook, R. W. (1967) Mnochondrial respiratory control and the polarographtc measurement of ADP:O ratios, in Methods in Enzymology, vol. 10 (Eastabrook, R. W. and Pullman, M. E., eds.), Academic, New York, pp. 41-47. 4. Truesdale, G. A., Downing, A. L., and Lowden, G. F. (1955) The solubility of oxygen in pure water and sea water. J Appl Chem. 5,53-62 5. Reynolds, D. J and Penn, C. W. (1994) Characterlsttcs of Helzcobacter pyZori growth in a defined medium and determination of tts ammo acid requirements. Mlcrobrology

140,2649-2656.

6. Nedenskov, P. (1994) Nutritional Appl Envwon

Mxroblol.

requirements for growth of Helrcobacterpylorl

60,3450-3453.

7. Deshpande, M., Calenoff, E., and Daniels, L. (1995) Raptd large-scale growth of Helxobacter pylorl m flasks and fermentors Appl Environ. Mlcrobiol 61, 243 l-2435.

9 Use of NMR to Study H. pylori Metabolism Peter A. Chalk, Andrew D. Roberts, Alistair A. Davison, David J. Kelly, and Peter J. White 1. Introduction Until recently, the culture of Helicobacter pylori in vitro has only been achievable with complex undefined media supplemented with blood products, such as Brain-heart infusion (BHI) medium, blood agar, chocolate agar, or lysed blood agar (I). Substrate utilization and metabolic activity during growth on these media IS difficult to define; this has led to the use of techniques such as nuclear magnetic resonance (NMR) spectroscopy to investigate the metabolic pathways active in H. pylori (e.g., refs. 2-5). Defined media for the growth of H. pylori have now been described (6,7), however, and may facihtate future metabolic studies in this organism by more conventional techniques. NMR spectroscopy has many features that make it a powerful technique for metabolic studies. It is noninvasive and does not harm the sample under study. There is no need for the chemical isolation or purification of species before analysis, and serial measurements can be carried out with the same sample. Studies can be carried out with suspensions of intact viable cells. This allows the study of enzyme processes m their natural environment with normal control elements intact and without the necessity of destroying the cell environment and supplementing extracts with relevant cofactors (not all of which may be known to the investigator). Another advantage of NMR spectroscopy is the exceptional chemical resolution it provides, allowing, for example, from a single spectrum, the discrimination of the fate of individual carbon atoms within the same molecule and the analysis of metabolic flux simultaneously into several divergent pathways. For these reasons, NMR lends itself to the study of the influence of cell environment on metabolic traffic. From Methods UI Molecular Medcme, Hehcobacter pylon Pfotoco/s Edited by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

69

NJ

70

Chalk et al.

In many instances, it may be feasible to carry out metabolic analyses by NMR spectroscopy m real time. The limitation here is that this can tie up the NMR spectrometer (an expensive piece of equipment!) for extended periods. Also, NMR measurements are generally carried out m narrow sample tubes under nitrogen; this limits real-time studies to anaerobic conditions, which are not a preferred environment for many organisms, including H. pylori. Although devices for sterile aeration and mixing m NMR cell metabolism studies have been described (a), and their application to H. pylori research would be an interesting future development, most laboratories are likely to be limited to carrying out aerobic studies by time-point analysis. The major limitation of NMR spectroscopy for metabolic tracking is the inherent insenstttvity of the technique, especially compared to radioactive tracer studies, such aswhen using 14Cor 32Plabels. Partrcularly with 13CNMR, it can take a significant amount of time to accumulate data; consequently, shortlived metabohtes and compounds at low concentration are likely to be missed. Also, because of the low inherent sensitivity of NMR, it 1sgenerally necessary to use very dense cell suspensions during experiments, which may be a cause of unnatural cell physiology. A further constramt, if using 13C-enriched precursors, is that these compounds can be difficult and expensive to source. Despite these considerations, NMR spectroscopy can be ideally suited to mapping major metabolic processesand investigating the effects of different physiological conditions. NMR detects the behavior in a strong magnetic field of those atomic nuclei that have a nuclear spin and so not all atomic species are available for analysis. The most commonly used isotopes are ‘H, i3C, and 31P,but m principle other isotopes could be used, such as 15N, i70, 33S, and lgF. These isotopes vary greatly in their sensitivity of detection (Table 1). Another important factor in the utility of a particular isotope is its natural abundance. The r3C nucleus is a rather insensitive isotope for NMR studies; however, its natural abundance is only 1.l%, and this can be exploited for metabolic studies by utilizing substrates 13C-enrichedon specific carbon atoms, with minimal background interference from endogenous metabolites. On the other hand, for compounds accumulatmg in sufficiently high levels, it may be possible to obtain useful information by investigating the 13Cnatural abundance signals using normal nonenriched substrates and/or endogenous signals from samples under different physiological conditions. In contrast to 13C,the ‘H atomic nucleus is particularly sensitive for NMR studies, but its natural abundance is over 99%. A complex biological sample will contain many thousands of NMR-visible compounds that may potentially mask the signals of interest. Indeed, the large ‘H signal from the water m the sample can obscure the smaller signals for metabolites. There are a number of techniques for sup-

71

NMR Study of Metabolism Table 1 Commonly

Available

Relative sensitivity Isotope (13C = 1)

Isotopes

of Interest

Natural abundance, Absolute % sensitiwty

‘H *H 3H

63 0.6 76

99.98 0.02 0

5675 0.008 0

13C

1

1.1

I

14N

0 06

15N ‘70

0 06 1.8

‘9F 23Na 3’P 33P

52.4 58 4.2 0.1

99.6

0.4 0 04 100 100 100 0.76

5.7

0.02 0.06 4730 525 376 0.1

for Metabolic

Studies

Chemical shift range, Comments wm 10 10 Requires labeled compounds 10 Requires labeled compounds The most sensitive NMR nucleus 250 Mostly requires labeled compunds Very broad lines unless in 900 a symmetrical chemical environment 900 Requires labeled compounds Broad signals, very low 1700 sensitivity 600 Easily observed 30 Easily observed 150 Easily observed Very broad lines unless 600 in a symmetrxal chemical environment

pressing this water signal, for example, by presaturatlon of the water resonance, but they will, of necessity, obliterate metabolite signals in the region of the water. NMR spectroscopy distinguishes between atoms of the same type in different chemical environments because of the effect of surrounding electron density in shielding the applied magnetic field. This changes the resultant chemical shift of a resonating nucleus. It is the key element of NMR (together with integration and analysis of spin/spin coupling) that enables its use in elucidating molecular structure. However, NMR visible isotopes vary greatly in the degree of resolution that they provide. For example, the range of chemical shift values seen commonly for 13Cmolecules is about 200 ppm but, for ‘H and 31P,it 1s only about 10 and 30 ppm, respectively. The chemical shift range and hence the ease of discrimination between different metabolites, may be a crucial factor in deciding the best approach for a particular study. NMR is now used frequently for metabolic studies and several reviews have been written on the subject (9,ZO). Sophisticated developments of the tech-

Chalk et al.

72

nique, such as isotoper analysis, have been described (10). Clearly, the potential applications of NMR analysis to the study of H pylori metabohsm are many and varied. The methods required for each particular study vary considerably, depending on the application. Nevertheless, the basic techniques that use NMR are simple and straightforward. In this chapter, we will describe some methods that have given useful results in our laboratories; they may provide a starting point for someone setting out with the technique. 2. Materials 1 H pylon type stram NCTC 11637. As with other areas of H pylon study, results should be confirmed with more than one strain of orgamsm. Matertals for the growth and maintenance of organisms are readily available from general microbiology supply houses. We mamtam strains by subculturmg twice-weekly on chocolate Columbia agar (Umpath [Bedford, UK], prepared according to manufacturers mstructtons, using 5% sterile defibrmated horse blood), supplemented wrth 10 mg/L each of amphotericin B, polymyxin B, and vancomycm (Sigma [Poole, UK], prepare at 10,20, and 20 g/L, respectively, m water, filter sterthze, and add to media immediately before pouring plates). BHI media (Unipath) is prepared according to manufacturer’s mstructions and supplemented with antibiotics as above and 5% fetal calf serum (Sigma). 2 13C enriched substrates can be difficult to find. We have obtained compounds from Cambridge Isotope Laboratories (Andover, MA), MSD isotopes (Montreal, Canada), Euriso-top Groupe CED (Gif-sur-Yvette, France), and Sigma/Aldrich

Substratesare madeup in DzO,adjustedto pH 7.0, and filter sterilized. 3. Phosphate-buffered saline (PBS). 136 mM NaCl, 2 mM KCI, 8 mA4 Na,HPO,, 1.5 mM KHZ P04, pH 7.2. After sterilization, we supplement with antibiotics, as above

3. Methods 3.1. Real-Time Analysis of Anaerobic Metabolism of 13C Labeled Substrate - [W3C] Fumarate Metabolism by H. pylori (Fig. I) 1 Inoculate 10 mL of BHI media in a 25 cm* vented tissue culture flask with a loopful of H pylon cells from an agar plate. Incubate overmght (preferably on a rocking tray) in a microaerobic cabmet [12% CO*, 5% 02, 5% HZ, balance N2, 37”C]. 2. Seed approx 40 chocolate Columbia agar spread plates with 0 1 mL each of culture Incubate m a microaerobic cabinet for 2 d. 3 Harvest cells into 25 mL of ice cold PBS (see Note 1). Vortex to resuspend cells; centrifuge for 20 min at 4°C at 3500g. Discard supernatant. Resuspend cells in 25 mL PBS. Centrifuge for 20 mm at 4°C at 3500g. Discard supernatant and resuspendcells with an equal volume of PBS (this should give approx 4-6 mL of cell suspension)

NMR Study of Metabolism

Incubation

Time (Mln)

Fig. 1. Anaerobic metabolism of [U-13C] fumarate by H pylon. (A) Spectrum after 5 min (B) Spectrum after 60 mm (C) changes of metabolite mtegrals with time (normahzed to 1 carbon per molecule) (1, fumarate; 2, malate; 3, succmate, 4, brcarbonate; 5, CO*; 6, acetate). Complex peak patterns are caused by carbon couplmgs because of urnversally labeled substrate. See text for explanation.

Chalk et al.

74

-‘nyrm,m~, 120

200

Y.rT~-m-,,rr-rr,--m-~-mrm 180 ‘60 140

120

‘T---r--~r--T---prrrrrrrrTn100 ID (10

4u

20

wm

Fig. 2. Compartson of anaerobic (A) and aerobic (B) metabolism of [2-13C] pyruvate by H pylon Spectra taken after 180 min mcubatton at 37°C by which time pyruvate (202 ppm) is totally depleted (1, acetate [CH3 ‘3C02H], 2, lactate [CH3 13CHOH CO*H], 3, ethanol [CH3 ‘3CH,0H]; 4, alanme [CH, 13CHNH2 C02H])

4. For many metabolites it wtll be necessary to use the above suspension directly For fumarate, dilute 10 times m PBS 5 Mix 100 pL of D,O (see Note 2), and 800 $ of cell suspension. Add to sterile 5-mm diameter NMR tube. Gas the head space of tube with a stream of nitrogen. Add 100 pL of OSM [U-t3C] sodium fumarate, pH 7.0, immediately vortex, and place into NMR spectrometer. 6 At 5-mm intervals, acquire proton decoupled carbon spectra (see Notes 3-5) over about 250 scans with a repetition time of 1 s and 35’C pulses. Use 37’C and nitrogen gas for probe services

3.2. Time-Point Analysis of Aerobic and Anaerobic Metabolism of 13C Labeled Substrates - [2-13C] Pyruvate Metabolism by H. pylori (Fig. 2) 1 Steps l-3 as in Section 3 1 2. To sterile 1 5-mL Eppendorf tubes, add 100 pL D20 and 850 & of cell suspenston. Add 50 & of 1M [2-13C] sodmm pyruvate, pH 7.0 3 Gas the head space of the anaerobic sample with a stream of nitrogen Seal with Nescotilm (Osaka, Japan). Incubate the tube on its side at 37°C on a rocking tray. 4 Mtx aerobic samples Transfer to 6-well tissue culture trays. Incubate at 37°C on a rocking tray under aerobic conditions (We use 10% CO*, 20% 02, balance N2 ) 5 Anaerobic samples, At time intervals, transfer tube to a microfuge. Centrifuge at 13,000g for 5 mm. Transfer supernatant to a 5-mm NMR tube

NMR Study of Metabolism

0

20

40

60

75

80

lncubatlon

100

Time

120

140

En

lea

(Mln)

Fig. 3. Analysis of succinate respiration to acetate using proton NMR. concentration, 0, succinate concentration.

n , Acetate

Aerobic samples. At time intervals, transfer the samples to a microfuge and centrifuge as above. 6. Acquire proton decoupled carbon spectra using a repetition time of 1 s and 35°C pulses until a suitable signal-to-noise ratio is obtained

3.3. Time-Point Analysis by Proton NMR: Aerobic Succinate Metabolism by H. pylori (Fig. 3) 1. Inoculate 2 x 20 mL of BHI broth media with a loopful of a 2-d-old plate culture. Incubate at 37°C overnight and use the entire cultures as inocula for 2 x 500 mL

Chalk et al.

76

2. 3.

4

5. 6.

BHI broth in 2-L shake flasks. Incubate overnight m a microaerobtc atmosphere at 37°C with shaking. Harvest cells at 35OOg, 4°C for 20 mm, wash once m fully oxygenated 20 mA4 potassium phosphate buffer, and resuspend m 5 mL of the same buffer. Add 0.06 mL of 100 rnA4sodmm succinate to 2.94 mL of cell suspension to give a final succinate concentration of 2 mM Incubate at 37% A 0 2-mL sample is immediately removed and the cells centrifuged out at 13,OOOgm a microfuge for 2 mm. The supematant is stored frozen at -20°C until used for NMR analysis. At time intervals, further samples are removed and treated in the same way. For NMR analysis, add to 0.2 mL of the supematant 0.3 mL of D,O and 5 pL of a 10 mMstock solution of Trimethyl silyl propionate (TSP) m D,O to give a final concentration of 0.1 mM. Acquire proton spectra using 60°C pulses, 4 s acquisition time, and 2 s nradiation at water frequency. The integral from the known concentration of the trimethyl signal of TSP IS used as an internal standard for quantification. It has the advantage that the signals from its trimethyl fimction (at 0 ppm) are well-separated from water and most common chemicals

3.4. ldenfificafion

of Mefabolifes

1 A number of techniques are available for determmation of metabohte identity The most straightforward is comparison of chemical shift data against that of putative products (see Table 2). Care should be taken that the chemical shift data is compared at similar pH values, since the shift of some carbon atoms can be sensitive to this, particularly carboxylic acids and carbons adJacent to nitrogen Having identified a candidate product, a few milligrams of authentic standard can be spiked mto samples to check whether the natural abundance peaks of the putative products superimpose over those of the sample (Fig 4) 2 To help in identifying products, Dept-135 spectra (distortionless enhancement by polarization transfer) may be obtamed. A Dept-135 spectrum will show methine and methyl carbons positive, methylene carbons negative, and quatemary carbons absent (Fig. 5). If further clarification of CH, and CH above chemical shift is required, a Dept-90 may be obtained, which will only show methine carbons 3 The spin of neighboring atoms affects the chemical shift of a resonating nucleus, because, depending on the orientation of a neighbormg atom in the field, it will either enforce or diminish the applied magnetic field. In practice, signals for both of these possibilmes are seen. This feature of NMR, called spin-spin couphng, gives rise to the multiple patterns m some spectra that are mdicative of the number and kinds of atoms adjacent to the nucleus being observed, The analysis of spin-spin coupling can be a useful technique for metabolic studies (e.g., ref. 4) and forms the basis of isotoper analysis (I 0). If the products have originated from a fully 13C labeled substrate, then it is likely that the products will be labeled likewise and will exhibit spin-spin couplings between adjacent i3C nuclei. This can be seen m Fig. 1, where the use of universally labeled fumarate has given rise

Table 2 13Carbon Chemical Shifts of Commonly Seen Products Compound Acetate

Chemical shift CH3

I C02H Pyruvate

I co

202

CH3

I

CHOH

I

COzH Alanine

21 179 26

CO,H

CH3

of H. pykwi Compound

CH3

I Lactate

of Metabolism

170 20

Malate

I

COIH Ethanol Bicarbonate Carbon dioxide Formate Urea Succmate

CH3

CH~OH HC032CO2 HCO,H NH,CONH, C02H

I CH2 CH2

Fumarate

C02H C02H

I

CH

CHOH

I CH2

Glycerol

Aspartate

C02H C02H I CHNH;!

I CH2

CO,H

C02H CH,OH

Glutamate

CH2

Sorbitol

C02H CH20H I CHOH

I

CHOH

I

CHOH

179

56 28 34 181 63 71 68 71

I

CHOH

69

I

Gluconate

CH,OH C02H

64 176

I

CHOH

72

I

CHOH

I

CHOH

69 70

I I

CH20H 38

65 176

I

CHOH 53

74

I

31

172 176

CH,OH C02H CHNH2 I CH,

31

132

175 65

I

175 16

179 172

39

I

48

55 159 124 170 163 179

68

I CHOH

182 17

177

I

66

132

I

COzH

I

I

CHNH*

Chemical shift

69 64

78

Chalk et al.

A

Fig. 4. Identification of metabohtes. Sptkmg of final mix of products from anaerobic [2-13C] pyruvate metabolism of H pylorz (A) with authentic alanme standard (B). (1, acetate [CH, “CO,H]; 2, lactate [CH, 13CHOH CO,H]; 3, ethanol [CH, 13CH,0H]; 4, alanine [CH, CI-INH2 C02H]). Note the appearance of peaks from all the alanme carbons, as well as enhancement of the previously observed peak in the upper spiked spectrum to complex peaks m some products due to internal spin-spin couplings between labeled carbons For example, the [U-13C] malate i3CH, at 67 5 ppm, gives a doublet of doublets because of spin-spin couplmgs from adjacent 13CH2 and i3C02H In contrast, the signals resulting from bicarbonate and carbon dioxide are single peaks, because there is no coupling 4. Two-dtmensional NMR techniques can be used to help map adJacent carbons. 2D technrques can also be very useful rn interpreting proton spectra and may allow the correlation of protons to their adJacent carbons.

4. Notes 1, The choice of buffer will depend on the nuclear species being observed For r3C NMR, carbon-based buffers, such as Tris and MES, are best avoided because they give rise to significant natural abundance peaks that can complicate the interpretation of spectra PBS is a good choice For 31P NMR, the converse is true Growth media can give good results, but the presence of competmg substrates may influence the pattern of products obtained. 2. 5-10% D20 can be included m the final analysis sample as a convenient means of referencing values of chemical shift. For proton NMR, TSP can be used as an internal standard for quantifying integrals.

79

NMR Study of Metabolism 2

B

III,,

~,,‘-‘-,-“,,“‘~~~~,,~mr’r-jmrT1 2x) an

lb)

I(D

7 1, ““P-n~‘-rrrrrr, 14)

mm tm

!rn

77”

,171 80

‘7-rr-T17~ 83

r, ‘TTln 43

rr

/, , -, , /, , ?, 20 Fm

Fig. 5. Identification of metabohtes: Dept- 135 spectrum of metabolic products from anaerobic [2-13C] pyruvate metabolism by Z-Lpylori. Comparison of normal spectrum (A) with Dept-135 spectrum (B). Peak asstgnments as for Fig. 2.

3. All modem spectrometer systemsare capable of observing 13Ccarbon while decoupling protons. However, the major limitation is sensitivity, requiring dense cell suspensions and/or long accumulation times. Use of a higher field strength will increase the intrmstc sensitivity. However acquiring i3C at higher field strength requires a larger frequency range to be decoupled for protons. Care needs to be taken in order to avoid excessive heating of the sample, particularly if it is htghly ionic. Composite pulse decoupling sequences are preferable, as they allow a lower decoupling power to be used. Probe size also affects sensitivity. Standard liquid probes tend to be 5 mm outside diameter A probe that can accept lo-mm sample tubes will give an increase in sensitivity of 34-fold. This translates into a 9- to 16-fold reduction of the time required to accumulate spectra because, as data IS accumulated, genuine signals increase in proportion to the number of scans, but random noise Increases in proportton to the square root of the number of scans. 4. It is difficult to build a probe capable of observing all nuclei without compromising sensitivity; choice of probe design can be an important factor m these expertments. If only one nucleus, e.g., 13C, 1sto be observed, then tt IS worth considering a dedicated probe to achieve the best possible signal-to-notse ratio. Alternatively, probes commonly come m a high frequency range (covermg t3C to 31P) and a low frequency range (covering the remainder). The proton coils on most probes are also capable of observing lgF and 3H. 5 Care must be exercised when using integrals as a means of quantlfymg components. The relaxation rate of the atoms concerned must be borne m mind. With

Chalk et al.

80

13C NMR, comparison of integrals of carbon in different chemical species may not accurately reflect their relative amounts. If protonated carbons only are to be measured, then a faster repetition rate may be used than if quaternary carbons are to be measured

References 1. Goodwin, C. S and Armstrong, J. A (1990) Microbiological aspects of Helzcobacterpylorz. Eur. J, Clzn Mzcrobzol. Infect Dzs 9(l), 1-13. 2 Mendz, G L and Hazell, S. L (1993) Glucose phosphorylation in Helzcobacter pylon. Archives Bzochem. Bzophys 300(l), 522-525.

3. Mendz, G. L., Hazell, S L , and Bums, B. P. (1993) Glucose utilisatton and lactate production by Helzcobacterpylori. J. Gen. Mzcrobzol 139,3023-3028. 4 Chalk, P. A., Roberts, A D , and Blows, W. M (1994) Metabolism of pyruvate and glucose by intact cells of Helicobacter pylorz studied by i3C NMR spectroscopy. Mzcrobioiogy 140,2085-2092. 5 Mendz, G. L , Hazell, S. L., and Bums, P. (1994) The Entner-Doudoroff pathway m Helzcobacterpylorz. Arch Bzochem. Biophys. 312,349-356. 6 Reynolds, D. J. and Penn, C W. (1994) Characteristics of Helzcobacter pylorz growth in a defined medium and determination of its ammo acid requnements Microbiology 140,2649-2656. 7. Nedenskov, P (1994) Nutritional requirements for growth of Helzcobacterpylorz. Appl Envzron Mzcrobzol

60(9), 345&3453.

8 Williams, H. J., Gao, Y., and Scott, I. (1991) Simple devices for sterile aeration and mixing m NMR cell metabolism studies. J Magn Reson. 94,405407 9 Lundberg, P., Harmsen, E., Ho, C., and Vogel, H. J. (1990) Nuclear magnetic resonance studies of cellular metabolism. Anal Bzochem 191, 193-222. 10. Jeffrey, F. M. H., Rajagopal, A., Malloy, C. R., and Sherry, A. D. (1991) 13CNMR* a simple yet comprehensive method for analysis of intermediary metabolism. TIBS 16,5-10

Isolation of H. py/oriGenomic and Restriction Analysis

DNA

Robert J. Owen and Jane Bickley 1. Introduction Since Helicobacter pylori was first described in 1983 (I), the study of genomic DNA has been central to the development of its microbiology and molecular genetics. For instance, DNA base composition estimation (mol% G+C) was crucial in demonstratmg affinities of the microorganism to the genus Campylobacter (2). Likewise, DNA-DNA hybridization assaysrevealed a high degree of base sequencehomology between different isolates of H. pylon, yet a low relatedness to Campylobacterfetus and other speciesof Campylobacter (3). In 1987, rRNA-DNA hybridization and hybrid thermal stability analyses were used to show that H. pylorz was phylogenetically distinct from Campylobacter sense strict0 and that the species merited classification in a new genus (4). The most significant application of DNA analysis has been in showing the diversity between genomes of different strains within H. pylori. The first indication of such genome diversity was from restriction endonuclease digest analysis of genomic DNA (6) and was subsequently confirmed by ribosomal RNA gene analysis and polymerase chain reaction (PCR)-based analysis of urease and other genes (7). This article describes methods for extraction and endonuclease digest analysis of genomic DNA from H pylori. Such analyses provide a sensitive means of directly detecting minor genomic differences between strains. The basic principle of the method is to treat high-mol-wt, purified genomic DNA with a high-frequency cutting restriction endonuclease, such as Hind111 (6-bp cutter) or Had11 (4-bp cutter) (see Note 9). The DNA is cleaved into fragments of different lengths, depending on the number and position of the particular recFrom Methods 111Molecular Medrcme, Hehcobacter pylon Protocols Edited by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

81

NJ

82

Owen and Bickley

ognition sequences. The band fragments generated by restriction enzyme digestion are then separated according to size by electrophoresis m an agarose gel to give a pattern of bands composed of fragments of l-20 kb. Differences in the DNA fingerprints are caused by band sizepolymorphisms, which reflect inter-strain sequence divergence (see Notes 10-12). This article describes the isolation of DNA from H pylori by the cetyltrimethylammomum bromide (CTAB)

procedure (8) and details the procedures for restriction

analysis.

2. Materials 1. Freshly cultured cells of H. pylon. 2. TE buffer 10 mA4 Tris-HCl, 1 n&f EDTA, pH 8 0. Store at 4°C. 3 10% Sodmm dodecyl sulfate (SDS) (BDH Pharmaceuticals, London, UK). Store at room temperature. 4. 20 mg/mL Protemase K (Stgma, Poole, UK). Store m ahquots at -10°C. 5. 5MNaCl. Store at room temperature. 6. CTAB/NaCl: Dissolve 4.1 g of NaCl, dissolved in 80 mL of water; 10 g CTAB IS slowly added while heating and stlrring. The final volume is adjusted to 100 mL Store at room temperature. 7 Chloroform/isoamyl alcohol (24: 1). Store at 4°C 8 Phenol (BDH)/chloroform/isoamyl alcohol (25:24*1) Store at 4°C 9. Isopropanol. Store at 4°C. 10 70% (v/v) ethanol. Store at 4“C 11 Buffers for restriction endonuclease digestion: The appropriate buffers are obtained from the manufacturer and are generally supplied as 10X stock solutions. These are stored at 4°C for periods of l-2 wk, or at -20°C indefinitely. 12. Spermidine (Sigma): Prepared to a concentration of 0. IA4 in 0 IA4 Tns, pH 7.0, and allquoted into small volumes, e.g., 100 &. Store mdefimtely at -20°C and defrost as necessary. 13. Restriction endonucleases. Available commercially from several manufacturers. Enzymes are stored at -20°C. 14. Stop mix: 2.5 mL 1% bromophenol blue (Sigma), 2.5 g Flcoll (Sigma), 1.0 mL 0.5M EDTA, pH 8.0, 10.0 mL water (to gave final volume). 15. Agarose: For electrophoresis of restriction enzyme digests, high-purity agarose is required, e.g., Ultrapure agarose, electrophoresis grade (Life Technologies, Paisley, UK). 16. Electrophoresls buffer: 1X Tris-borate buffer (TBE) 53.9 g Tris, 27.5 g boric acid, 3.7 g EDTA, pH 8.3. Make up to 5 L with water and store as stock solution at room temperature. 17. Mol-wt markers: Blotinylated h Hind111 fragments (Life Technologies), supplied as 10 pg/lOO pL: 10 pL h fragments, 26 pL sterile water, 4 pL HlndIII core buffer. Heat at 60°C for 5 mm, cool on Ice, and add 10 & stop mix Store at 4°C. The fragment sizes (bp) are 23,130; 9416; 6557; 4361; 2322; 2027; 564; and 125 18. Electrophoresis tank: Select size, depending on the number of samples to be run, e.g , Life Technologies Horizontal Gel Electrophoresis Systems, models

Isolation of Genomic DNA

19 20. 2 1.

22 23.

83

H4 and H5. The H4 tank holds up to 60 samples and the H5 tank, a maximum of 14 samples Power pack: A standard power pack is reqmred to supply up to 100 V Heating apparatus: A heatmg block or a microwave oven. Ethtdium bromide (Stgma): Prepare a stock solution of 10 mg/mL in water, store in a light-proof container. Use diluted to a concentration of 0 5 %I$ for staming gels. UV Transilluminator. Camera and film.

3. Methods

3.1. Cell Growth and DNA Extraction 1 To extract DNA from bacterial cultures, make a suspension of H. pylon cells (1 x lo* CFU) in 900 clr, of 1X TE buffer 2 Spin the sample for 5 min in a microcentnfuge to form a compact pellet. 3. Discard the supernatant and resuspend the pellet m 567 pL of TE buffer using a disposable spatula. 4. To the resuspended pellet, add 30 p,L of 10% SDS and 3 pL of 20 mg/mL protemase K, to give final concentrations of 100 pL/mL proteinase K in 0 5% SDS. The sample is mixed thoroughly and incubated for 1 h at 37°C 5. After mcubation, add 100 & of 5M NaCl and mix well, followed by 80 pL of CTAB/NaCl solution. Incubate for 10 mm at 65°C. 6. Add an approximately equal volume of chloroform/isoamyl alcohol, mix thoroughly, and spm for 5 mm 7 Remove the aqueous, viscous supernatant to a clean tube, leavmg the interface 8. Add an equal volume of phenol/chloroform/isoamyl alcohol to the aqueous phase, mix well, and again spin for 5 min Gloves must be worn and care taken when using phenol 9. Transfer the supernatant to a clean tube, and to it add 0.6 vol isopropanol to precipitate the nucleic acid Shake the tube vigorously and then mix gently until a stringy white DNA precipitate appears. 10. Pipet off the isopropanol solution, taking care not to suck any of the DNA into the pipet. 11. Wash the DNA with 70% ethanol to remove residual CTAB. 12. Briefly spin the tube to pellet the DNA, before removing the ethanol with a pipet. 13. Finally dry the DNA under vacuum and then redissolve in 100 p.L TE buffer. 14. Determine the concentration of the DNA solution spectrophotometrically 15. A flowchart of the procedure is given in Fig. 1.

3.2. DNA Restriction Digestion 1 Chromosomal DNA samples: The concentrations of the DNA samples are calculated from the optical density readings at 260 nm (A260). Results are used to calculate the volume of sample required to give 5 pg of DNA in the digestion reaction (total volume = 20 pL). An Az6c of 1.O is equivalent to 47 pg/mL DNA.

Owen and Bickley

84

I

I

Spin 5 mins

_

1

1

Dry under

vacuum

Fig. 1. Preparation of genomic DNA from

2. 3. 4 5. 6 7.

H

pylori by the CTAB method

A drlution of l/50 IS used to measure the optical densny (i.e., 10 pL DNA sample m 0 49 mL water) The concentratron of DNA IS determmed from. [DNA] = Az6s x 47 x 50 (pg/mL) To obtain the DNA concentration in pg/pL, the above value is drvided by 1000. The volume of sample required to give 5 pg of DNA m the digestion reactron IS 5 x [DNA] -’ mL. To a sterile 0.5 mL microcentrtfuge tube, add a sample volume contaming 5 pg of DNA, and make up to 16 $ wrth sterrle distilled water and cool on me. Add 1 pL of spermidme and 2 p.L of the approprtate 10X digestton buffer and mix well. Add 5-10 U of the appropriate restrrctton enzyme (1.e , in most cases, that amounts to 1 pL of enzyme) (see Notes l-3), mix well, and centrrfuge, briefly to ensure that all reagents are at the bottom of the tube. Leave on ice for 5 mm. Incubate at the optrmum temperature (usually 37°C) for 3 or 4 h. Stop the reaction by adding 4 p.L of stop mix; cool on ice. Analyze the samples mnnedtately, or store for a few days at 4°C or at -20°C for longer periods untrl requrred

3.3. Agarose

Gel Electrophoresis

1. Clean and dry the gel tray. Seal the open edges with two layers of autoclave tape, which should overlap one another. Locate the reqmred comb (or combs, If a

Isolation of Genomic DNA

2.

3.

4.

5.

6.

7.

8. 9. 10. 11.

85

double rank gel is to be used) in positton and place the tray on a level surface. To ensure that the tray is completely level, a spirit-level is used, as it is important that the whole gel is of even thickness. Prepare a 0.8% agarose gel by adding a measured volume of electrophoresis buffer to the correct amount of powdered agarose, depending on the size of gel to be used. For the H5 gel tank, use 65 mL TBE buffer and 0.52 g agarose. For the H4 gel tank, 200 mL TBE buffer and 1.6 g agarose. Mix these in a conical flask and cover the mouth of the flask (but do not seal completely) to prevent evaporation during heating. Heat the mixture in a mtcrowave (or other heat source) on a medium setting, removing the flask periodically to ensure good mixing. Continue heating until the agarose has completely melted and the mixture is boihng. Remove the agarose from the heat and cool to approx 50°C by holding the flask under cold, running water. Pour the molten agarose into the gel tray and remove any bubbles with a Pasteur pipet. Leave the gel to completely set for 30-45 min at room temperature. Once the gel has set, carefully remove the comb (or combs) and autoclave tape and place the gel tray in the electrophoresis tank. Add sufficient electrophoresis buffer to cover the gel to a depth of about 1 mm. Prepare DNA digests for agarose gel electrophoresis as described in Section 3.2 Remove samples from the fridge or freezer immediately before loading onto the gel, heat at 65°C for 5 min, then cool on ice. Mix the samples and briefly centrifuge, and pipet only from the upper part of the sample for gel loading. Carefully load 10 pL of each sample and 7 $ of mol-wt markers (see Note 4) mto the wells. Withdraw the pipet slowly without disturbing the loaded samples, and load the gel as quickly as possible to mimmtze diffusion of samples out of the wells. Electrophorese at 30 V for 16 h (overnight). Switch off the power supply and remove the gel from the electrophoresis tank. Place the gel, still supported by the gel tray, in ethidmm bromide stam for 30-60 min (see Notes 5 and 6). View the gel on a UV transilluminator and photograph to record the result (see Notes 7 and 8). Features of typical digest profiles of H. pylori DNA are shown in Fig. 2

4. Notes 1. Approximately l-5 U of restriction enzyme should be used m a 20-pL reaction, In general, digestion for longer periods of time, or with excess enzyme, does not cause problems, unless there is contamination with DNase or exonuclease. 2. Reaction volumes should be kept to a mimmum by reducing the volume of water, as much as possible. The restriction enzyme should contribute less than 10% (by volume) of the final reaction mixture. Otherwise, the enzyme activity may be inhibited by glycerol. 3. When digesting large numbers of DNA samples with the same enzyme, the spermidine, buffer, and restriction enzyme should be premixed in the proportions 1:2: 1, and 4 pL added to each sample.

Owen and Bickley

86

23.19.46.6 4.4-

Fig. 2. Agarose gel electrophoresis of Hue111 restriction digest fragments of chromosomal DNA from different strains of H. pylori. Isolates were from a series of four patients before (B) and after (A) treatment (11). 4. Two to three wells of mol-wt markers should be included on each gel and placed at both ends and in the center of the gels. These help to indicate any distortion in the gel, since all the markers should run symmetrically. 5. Ethidium bromide is a powerful mutagen and gloves should always be worn when handling gels or solutions containing the dye. Ethidium bromide must be neutralized before disposal, as follows (in 1 L vol): 250 pL 2.5MHCl + 250 pL 0.5M KMn04 (left for 2 h); 250 pL 2.5M NaOH (left for 30 min). The neutralized solution must be disposed of with copious quantities of running water. Similarly, gels containing ethidium bromide must be disposed of with care, and sealed in plastic bags before being discarded for incineration. 6. Destaining of gels is not usually necessary. However, detection ofvery small amounts of DNA may be easier if the background fluorescence caused by unbound ethidium bromide is reduced by soaking the stained gel in 1 mM MgSO, for 30-60 min. 7. An eye and face shield must be worn when viewing and photographing gels on the UV transilluminator. 8. Polaroid 667 film provides an instant record of a gel. It is essential to obtain a good quality photographic negative for future prints, enlargements, or for densitometric scanning of restriction patterns. Tri-X Pan Professional film (Kodak, Rochester, NY) can also be used for this purpose. 9. Various restriction endonucleases have been used to digest DNA from H. pylori (Table 1). In practice, most analyses have been based on the use of HaeIII or HindIII, because they give distinctive patterns of bands. Nevertheless, about 40% of strains of H. pylori have DNA that is resistant to Hue111 activity (6).

Isolation of Genomic DNA

87

Table 1 Restriction Digest Analysis of DNA from H. py/ori NCTC 11638 by Various Restriction endonuclease HueIII, HindIIIa BamHI, BstEII, EcoRI

Endonucleases Comment

Clear and dtstmct band patterns DNA cut to varying degrees

MspI, PVUII, SUCI, StuI

ACCI, AZuI, A@, BscI, @I, PstI, RsaI, &XII, SalI, SphI, Tag1

DNA not cut

“It should be noted that about 40% of other strains are not cut by Hue111 Most strams are cut by HzndIII

10. A significant disadvantage of digest analyses performed usmg high frequency cutting enzymes and conventional electrophoresis, is that the resultant band profiles are complex, including more than 25 bands. Such profiles are difficult to analyze and to interpret, because individual bands can be identified usually only in the high-mol-wt region; the rest of the profile includes areas of light and dark intenstty (Fig. 2). 11. Comparative analysis of such DNA digests is usually by subjective visual comparison. More objective approaches include the computation of similarities from high molecular weight band signatures (9), or by densitometric scanning of the complete profile with computer-based analysis for determmation of strain semilarities (ZO). Both of the latter methods are time-consuming and difficult when gel-to-gel comparisons are necessary. 12. Restrictton digest analysis is best used to compare pairs of strains, e.g , before and after treatment isolates, which should be run m adjacent lanes to facilitate direct visual comparison. Such analyses are also an important stage in preparing Southern blots, as required for ribotyping.

References 1 Marshall, B. (1983) Unidentified curved bacilli on gastric epithelmm in active chronic gastritis. Lancet i, 1273-1275 2. Marshall, B. J., Joyce, H., Anwar, D. I., Goodwin, C. S., Pearmans, J. W , Warren, J. R , and Armstrong, J A. (1984) Original isolation of Campylobacterpylondzs from human gastric mucosa. Muzrobtos Lett 25,83-88. 3. Bukholm, G., Nedenskov-Sorensen, P , and Bovre, K. (1989) DNA-DNA hybridization incompatibility of Campylobacter pylorl with other Campylobacter and Wolinella species. APMIS 97,472-474. 4. Romamuk, P. J., Zoltouska, B., Trust, T. J., Lane, D. J., Olsen, G. J., Pace, N. R , and Stahl, D. A. (1987) Cumpylobacterpylorz, a spiral bacterium associated with human gastritis, is not a true Campylobacter sp. J Bacterlol 169,2 137-2 141

Owen and Bickiey

88

5. Goodwin, C S , Armstrong, J. A , Chilvers, T., Peters, M., Collins, M. D., Sly, L , McConnell, W., and Harper, W. E. S. (1989) Transfer of Campylobacter pylorz and Campylobacter mustelae to Helrcobacter gen. nov. and Heltcobacter pylori comb. nov., as Helxobacter mustelae comb. nov., respectively. Int J Systematic Bacterial. 39, 397-405.

6. Owen, R J , Bickley, J , Costas, M., and Morgan, D. R. (1991) Genomic vartation m Hellcobacter pylori. application to identification of strains. Stand J Gastroenterol. 26(Suppl. Ml), 43-50. 7 Owen, R. J., Bickely, J., Hurtado, A., Fraser, A., and Pounder, R. E. (1994) Comparison of PCR-based restriction length polymorphism analysis of urease genes with rRNA gene profiling for monitoring Hehcobacterpylori infections m patients on triple therapy. J Clan Mlcroblol 32, 1203-1210. 8. Wilson, K. (1987) Preparatton of genomic DNA from bacteria, m Current Protocols zn Molecular Biology (Ausubel, F M , Brent, R , Kingston, R E , Moore, D D , Smith, J. A, Seidman, J. G., and Struhl, K., eds.), Wiley, New York, pp 2.4.1,2.4 2 9. Owen, R. J., Fraser, J., Costas, M., Morgan, D., and Morgan, D. R. (1990) Signature patterns of DNA restriction fragments of Helzcobacterpylort before and after treatment J Clin Pathol. 43,646-649. 10. Nwokolo, C. U., Bickley, J., Attard, A. R., Owen, R. J., Costas, M , and Fraser, I A. (1992) Evidence of clonal variants of Helicobacter pylorl in three generations of a duodenal ulcer disease family. Gut 33, 1323-l 327. 11. Fraser, A. G., Bickley, J., Owen, R J., and Pounder, R. E. (1992) DNA fingerprints of Hellcobacterpylorl before and after treatment wtth omeprazole. J. Clan Path01 45, 1062-1065

11 Ribosomal RNA Gene Restriction (Ribotyping) of H. pylori

Pattern Analysis

Wee Tee 1. Introduction Ribotyping, a method used to type strains of bacteria by analyzing the restrtctton enzyme digestion patterns of the ribosomal RNA (rRNA) genes, was first developed and proposed as a taxonomical tool for the identttication and differentiation of bacteria by Grimont and Grimont, m 1986 (1). Since its introduction, ribotyping has found wide application in molecular epidemiology as a marker for a variety of bacterial genera, and has been successfully applied to differentiate many bacterial species (l-6,), including Helicobacter pylorz (7-14). For epidemiological study, this typing method can be used to trace the source of infection or outbreak, determine the mode(s) of transmission, or monitor the distribution, movement, or occurrence of a particular strain of bacteria. For clinical study, it can be used to assessthe role of an organism in clinical disease, such as the monitormg of treatment regimens to determine whether clinical relapse is caused by reinfection with a new resistant isolate, or recrudescence caused by the original colonizing strain that has developed resistance to the antibiotic used (13). It can also determine whether an infection is caused by multiple strains or by a single strain (13). This chapter describes the use of the rRNA gene restriction patterns as a typing scheme for H. pylori. Ribotyping relies on the use of an rRNA probe that hybridizes to a selected portion of the bacterial genome, namely, the DNA fragments carrymg the rRNA genes, to produce simple strain-dependent patterns of bands. The probe is usually derived from E. coli 16s rRNA and is capable of hybridization to the DNA of most bacterial genera, because rRNA genes are highly conserved among all bacteria. Detection of strain differences From Methods VI Molecular Me&me, Hehcobacter pylon Protocols Edlted by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

89

NJ

90

Tee

both within and between spectes becomes possible because the rRNA gene sequences are often present in multiple copies in a variety of genetic arrangements in bacterial chromosomes. They are sufficiently heterogeneous to make a suitable target for comparison of different bacterial species and frequently show differences between strains within a bacterial species. To maximize the detection of DNA restriction site heterogeneity, a variety of restriction enzymes were assessed to determine which restriction endonuclease(s) produced DNA restriction fragment patterns that provide the most discrtmmation among epidemiologically distinct strains. Six restriction enzymes, namely HindIII, HaeIII, DruI, BarnHI, PstI, and EcoRI were evaluated for their ability to type H. pylorz. Only HlndIII, HaeIII, and DraI yielded DNA fragments that produced well-resolved bands by ribotypmg, and Hue111 appeared to be the most discrtmmatory, producing patterns with 4-12 distinct bands. Unfortunately, 25-30% of H pylori strains were not digested by this enzyme, which would result m a large proportion of untypable strains if only Hue111was used for digestion. In contrast, both DruI and Hind111digested DNA from all of more than 200 H. pylon strains tested. DruI yielded digest patterns consisting of 2-5 bands ranging from 1 to 6 kilobase (kb) pairs in size, but it appears to have the least discriminatory power compared with Hue111 or HindIII. Ribotyping patterns obtained from Hind111 digests consisted of 2-8 bands, ranging m size from 1 to 10 kb pairs. From these observations, it appears that Hind111 is the most appropriate restriction endonuclease, digesting DNA from the majority of H. pylori strains and providing a good degree of discrimination. A variety of probes have been used successfully for ribotyping H. pylon These include a probe prepared from a recombinant probe in which rRNA genes were cloned into plasmid pBR322 to obtain an 11.9-kb plasmid pKK3535 (‘161, and cDNA probes obtained by reverse transcription of either E. coli 16s and 23s rRNA (7) or NTCC type strain 11638 H. pylori 16s and 235 rRNA (9-12). Some of these probes are commercially available, e.g., E. cob rRNA probe (Bethesda Reseach Laboratories, Gaithersburg, MD), 16s and 23s rRNA from E colz’(Boehringer Mannheim, Mannheim, Germany, or Pharmacia, Uppsala, Sweden). The probe used for the ribotyping protocol in this chapter, pKK3535, was kindly provided by Martin Altwegg and colleagues from the Department of Medical Microbiology, University of Zurich, Switzerland (2) This recombinant plasmid carries the 7.5 kb BumHI fragment insert of rrnB operon encoding 5S RNA, 16s RNA, 23s RNA, and tRNAzGlu genes. The detailed physical map of pKK3535 has been published (16). This plasmid can be propagated m E. coli and then labeled for ribotyping of H. pylori; however, the yield of pKK3535 m E. coli is low, about 0.2 mg/L of culture (16). To overcome this, the insert containing the rRNA genes has been subcloned into high copy num-

Ribotyping of H. pylori

97

ber plasmid vectors and transformed into host E. coli, resulting in recombmant plasmid yields of >400 ng/ mL of culture (Z 7). There are several ways of labeling probes for ribotypmg using radioactive (2,14) or nonradioactive methods (7-15). Radiorsotope labeling using 32Plabeled nucleotides involves incorporating the radiolabeled nucleotide mto the DNA enzymatically by nick translation or by random priming. After hybridization, the presence of the labeled probe is detected by autoradiography. Radioisotopic probes are generally found to be more sensitive than nonradioisotopic probes, with greater band intensity and less background. Direct detection by autoradiography provides a permanent record of the results and eliminates the time-consuming steps of washing involved in nonradioisotopic detection. However, because of radioactive hazards associated with 32P-Iabeled nucleotides, this method requires trained personnel and special facilities for the handling and disposal of radioactive waste. In addition, radiolabeled nucleotides have a relatively short half-life, requiring frequent reordering of tsotopes and limited opportunity for probe reuse. As a result, a variety of nonradioactive systemshave been developed for labeling and detection of products of DNA hybrtdization. The most popular methods are based on biotin- (7), or digoxigenin- (8,12) labeled nucleotides. These techniques rely on an enzyme reaction for detection, using a signal-generating substrate, such as color reactions or chemiluminescence. For example, digoxigenin molecules covalently linked to nucleotide bases can be incorporated into DNA, using standard DNA labeling procedures with a DNA polymerase. The presence of the digoxigenin-labeled DNA is then detected, using an antibody to digoxigenin, which is conjugated to the enzyme alkaline phosphatase. The hybrtds are subsequently detected by adding the substrate for the enzyme, leading to the appearance of a colored spot on the filter. A number of kits are commercially available for a range of nonradioactive labeling and detection systems. Briefly, ribotyping consistsof the following steps: labeling the rRNA probe, using the digoxigenin system (Nonradioactive DNA Labeling and Detection Kit, Boehringer Mannheim); restriction endonuclease digestion of whole cell H. pylon DNA; separation of restriction fragments by agarose gel electrophoresis; transfer of restriction fragments to a fixed membrane; probing with a labeled ribosomal RNA; and detection of hybridized probe. The outline of the procedure is summarized diagrammatically in Fig. 1. Various modifications of the stepsused in the method have been published (7-13) with a range of different probes used and a number of detection methods (radioactive and nonradtoactive) applied. The list of techniques and the choice of methods will depend on the resources of individual laboratories. The protocol provided m this chapter uses recombinant plasmrd pKK3535 as the source of the probe and a nonradioactive digoxigenin system for labeling and detection,

92

Tee

Chromosomal

DNA

Digestlon with restriction endonuclease

+

(Hind Ill)

Separation of DNA fragments by gel electrophoresls,

55V for 17 hrs

South& transfer (Transfer onto nylon membrane. Remove and bake at 120°C 30 Ins.) T

I I

Prehybrldlzatlon

Hybrldlzatlon

for 2 hrs at 37OC Plasmld pKK3535 probe labelled wlth dlgoxlgenln dUTP

at 37°C overnight twice wlth SXSSC/O.l% SDS

+ Wash 4

twice wlth O.lXSSC/O.i%

SDS

t lmmunodetectlon of dlgoxlgenln-labelled probe (Non-radioactlve DNA labelllng and detectlon klt, Boehrlnger Mannhelm, Germany)

Fig. 1. Diagrammatic representationof ribotyping procedure. 2. Materials 2. I. Equipment Gel electrophoresis tank (20 cm wide, 24 cm long, International Biotechnologies, New Haven, CT); power pack; vacuum transfer apparatus (Hybaid, Teddington, UK); nylon membranes, positively charged (30 cm x 3 m roll, Boehringer Mannheim); Whatman 3MM Chromatography paper (Whatman,

93

Ribotyping of H. pylori

Maidstone, UK); 6-L large square storer or lunch box (-23 cm2), plastic bags (22 x 26 cm), plastic bag sealer, and shaking water bath. 2.2. Stock Solutions and Buffers Distilled water was used for preparation of buffers and solutions. Sterilize solutions by autoclaving at 121OC for 15 min. Store autoclaved reagents at room temperature unless otherwise indicated. 1 2. 3. 4. 5. 6. 7. 8. 9. 10

IMTris-HCl, pH 7.5. lMTris-HCI, pH 9.5. 2MTris-HCl, pH 7.4. 0.5M EDTA, pH 8.0. 5MNaCl. 10% SDS. 20X SSC: 3MNaC1,0.3M Nas citrate. 10% Casein. 1M MgCl*. 10X TBE (Tris-borate-EDTA) electrophoresis buffer stock solution* 108 g Trisbase, 55 g boric acid, and 40 mL 0.5MEDTA m 1 L of dHzO (final molarrtres. 0.89M Tris-base, 0 89M boric acid, and 20 mA4 EDTA).

2.3. DNA Probe Labeling 1. Tris-EDTA (TE) buffer: 10 nnI4 Tris-HCl, pH 7.4, 1 nnI4 EDTA, pH 8.0. 2. DNA labeling and detection kit, nonradioactrve (Boehringer Mannhelm), stored at -20°C: vial 5, hexanucleotide mixture; vial 6, dNTP labeling mixture (dATP, 1 mmol/L, dGTP, 1 mmol/L, dTTP, 0.65 mmol/L, Drg-dUTP, 0 35 mmol/L, pH 6.5), Klenow enzyme (2 U/pL)

2.4. Restriction Endonuclease Digestion of Chromosomal DNA HirtdIII restriction endonuclease and buffer (Boehringer Mannheim). 2.5. Agarose Gel Electrophoresis 1. 1X TBE: dilute l/10 of stock solution for working solution Made fresh pnor to use 2. 0.7% (w/v) agarose m 1X TBE buffer (make up 200 mL for a large gel).

2.6. Southern

Transfer

1. 0.25MHCl. 2. Transfer buffer: O.SMNaOH, to 2 L wrth dHzO.

1.SMNaCl. Mix 40 g NaOH, 176 g NaCl Make up

2.7. Hybridization 1. Herring sperm DNA (10 mg/mL): 200 mg herring sperm DNA m 20 mL dHzO. Heat to boiling until all DNA is dissolved. Allow to cool. Shear DNA by passing the suspension through a syringe and 23-gage needle. Dispense in 1 mL aliquot

Tee

94

m microcentrifuge tubes and stored at -20°C. Before use, boil for 5 min and cool rapidly on ice to denature. 2 Prehybridization’hybridization solution: 40 mL 10% casein; 200 mL Formamide; 6.6 mL Sarkosyl; 40 mL 10% SDS; 100 mL 20X SSC; 4 mL 10 mg/mL herrmg sperm DNA (denatured first) (Boehringer Mannhelm), 9 mL H,O 3. 2X SSC, 0 1% SDS MIX 200 mL 20X SSC, 20 mL 10% SDS, 1780 mL dH,O 4 0.1X SSC, 0 1% SDS (10 mL 20X SSC, 20 mL 10% SDS, 1970 mL dH,O)

2.8. Detection

Using Digoxigenin

System

1 Buffer 1. O.lM Tris-HCl, pH 7 5,0.15MNaCl.

Mix 200 mL 0. IA4 Tris-HCl,

pH

7 5 + 60 mL 5M NaCl + 1740 mL dHzO.

2 Buffer 2. Buffer 1 with 1% casein 3 Buffer 3: 0 IA4 Tris-HCl, pH 9 5, O.lMNaCl, 50 mM MgCl, Mix 200 mL 1M Tris-HCl, pH 9.5 + 40 mL 5MNaCl+ 100 mL lMMgCl* + 1660 mL dHzO 4 DNA labeling and detection kit, nonradioactive (Boehrmger Mannheim), stored at -20°C: vial 8, anti-digoxigenin Fab fragment conjugated to alkaline phosphatase; vial 9,75 mg/mL mtroblue tetrazolium salt and 70% (v/v) dimethylformamide, vial 10, 50 mg/mL 5-bromo-4-chloro-3-mdolyl phosphate in dimethylformamide. 5. DNA mol-wt marker II (digoxigenm labeled HzndIII digested h phage DNA). 6. Color solution: 30 mL buffer 3, 135 pL vial 9, 105 pL vial 10 for a large gel (make fresh just before use) 3. Method

3.1. DNA Probe Labeling

Using Digoxigenin

System

Approximately 2 ltg of plasmtd pKK3535 are dtgested with 1 $ BamHI (10 U/&, Boehrmger Mannheim) for 3 h at 37°C. The digest is purified by using phenol-chloroform extraction followed by chloroform extraction and then adjusted to 0.2M NaCl before precipitation with 2 vol of ethanol. After centrifugation, the pellet is washed wtth 70% ethanol. The plasmtd DNA is labeled with digoxigenin using the nonradioactive DNA labeling and detection kit of Boehringer Gmb H, according to the manufacturer’s instructions. This procedure 1s summarized as follows: 1. Dilute approx 2 pg of lmeanzed plasmid DNA to 26 pL with distilled water 2 Denature the DNA by boiling for 5 min, then chill quickly on ice 3. Add 8 pL of 5X DIG labeling buffer containing random primers (vial 5 supplied with the kit) and 4 pL of dNTP mix (vial 6, supplied with the kit) and 1 pL of Klenow enzyme to the sample and incubate overnight at 37Y. 4 Add 1 pL of glycogen and 160 pL of TE to make a total volume of 200 pL 5. Precipitate the DNA probe by adding 8 pL of 5MNaCl(4%) and 2 vol(400 pL) of ethanol, then centrifuge for 10 min at 12,OOOg

6. Dtscard supernatant and wash the pellet with 70% ethanol. 7. Dry the pellet.

95

Ribotyping of H. pylori

8 Suspend the DNA probe in 100 p.L TE buffer. 9. Denature to single strand by boiling for 5 mm and then chill on me. 10 Add the single-stranded DNA probe to 30 mL of hybridization solution and store at 4°C. Dilute the probe 1:3 in hybridization solution for use The diluted probe is able to be reused five times before there is a significant loss of sensitivity.

3.2. Restriction

Digestion

of Chromosomal

DNA of H. pylori

2 ~18 of each DNA sample 1s digested with 30 U of Hind111 restrmtion enzyme at 37°C overnight under the conditions specified by the manufacturer (Boehringer Mannhelm) (see Notes 1 and 2).

3.3. Separation of Digested DNA Fragments by Gel Electrophoresis DNA fragments are separated by electrophoresis through 0.7% (w/v) agarose gel (20-cm-long, 24-cm-wide gel tank [International Biotechnologies, New Haven]) at 55 V for 17 h rn 1X TBE (see Note 3).

3.4. Southern Blot Hybridization 3.4.1. Preparing the Gel for Transfer 1. Acid treat the gel containing the separated fragments by soaking in 0.25M HCl for 20 min. This step is incorporated because the preliminary results of rRNA gene restriction patterns showed that the DNA fragments yielded were ranging from 1 to 10 kb in molecular size Acid treatment is recommended to improve the transfer of long molecules >5 kb (18). 2. Wash three times with distilled water for 5 min each wash. 3. Neutralize with OSMNaOH-1.5MNaCl solution (transfer buffer).

3.4.2. Southern Transfer The DNA is transferred to nylon membrane (nylon membranes, positively charged, Boehringer Mannheim) using a vacuum transfer apparatus (Hybaid) at 40 cm HZ0 vacuum for 1 h with 0.5M NaOH-1.5M NaCl (transfer buffer) solution as described (see also Notes 4 and 5). 1. Cut one piece of Whatman filter paper (23 x 26 cm), prewet m water Place the filter paper centrally on the white porous screen, making sure no air bubbles are trapped under the paper. 2. Cut rubber mask with an opening in the center on all four sides, lust 5-10 mm smaller than the size of the gel (e.g., 19.5 cm wade by 20.8 cm long). Place this black rubber mask over the Whatman paper, mark the four corners of the aperture on the Whatman paper. Remove the rubber mask 3. Cut one piece of nylon membrane (21 x 22 cm), prewet in water. Place and cover completely the marked corners on the Whatman paper. Again, make sure no air bubbles are trapped between the paper and the nylon membrane (see Note 4)

96

Tee

4. Replace the rubber mask over the nylon membrane, leavmg an opening just smaller than the size of the gel 5. Secure the ltd frame over the mask. 6. Fill the central aperture of the mask with 100 mL of transfer buffer. 7 Carefully position the gel over the opening in contact with the nylon membrane by slowly sliding the gel off the gel tray to cover the central aperture of the rubber mask Make sure no air bubbles are trapped under the gel (see Note 4). 8 Turn on the vacuum (40 cm HZ0 pressure or vacuum regulator set between 3 and 4) and ensure gel 1s under vacuum. 9 Add transfer buffer until gel IS covered. 10. After 1 h of transfer, remove the nylon membrane. 11 Bake the nylon membrane at 12O’C for 30 min.

3.5. Hybridization 3.5. I. Prehybridization 1. Place the nylon membrane m a plastic bag. Add 10 mL prehybridizatron Seal the bag. 2 Incubate the bag containing the nylon membrane at 37°C for 2 h.

solution.

3.5.2. Hybridization with Labeled Probe 1. The prevtously labeled DNA probe stored m hybrtdtzatton solution is denatured by boiling for 10 min and cooled rapidly on ice 2. Add this to the bag containing the nylon membrane and incubate at 37’C overnight.

3.5.3. Washing 1 After hybridizatton, wash the membrane twice for 5 mm at room temperature with 2X SSC, 0 1% SDS. 2. Wash twice for 30 min at 42°C with 0.1X SSC, 0 1% SDS.

3.6. Detection (Digoxigenin System) The presence of digoxtgenin m the labeled DNA probe is detected with an alkaline phosphate conjugated antibody as described in the nonradioactive detection kit (Boehringer Mannhelm). The only modification from the method described IS that 1% casem in 0. lMTris-HCl, pH 7.5- 0.15MNaCl was used in the blocking step and as a conjugate dlluent. This is described below: 1. Wash the membrane briefly in buffer 1. 2. Incubate 30 min shaking m buffer 2. 3. Incubate 30 mm in a plastic bag, shaking m 1:5000 dilutton of vial 8 (anti-DIG/ AP conjugate) in buffer 2. (For a large gel, use 3 mL 10% [w/v] casein, 30 mL buffer 1, 6 pL conjugate.) 4 Wash twice for 15 min each in buffer 1. 5. Wash for 2 min m buffer 3. 6. Seal membrane in a plastic bag with freshly prepared 30 mL of color solution.

Ribotyping of H. pylori

97

7. Leave m the dark for at least 2 h (the intensity of bands depend on the sensitivity of the probe and/or the concentration of DNA used) (see Notes 2 and 6). 8 Stop reaction by washing membrane in TE or distilled water and air-dry m dark.

4. Notes 1. It is essential that the H. pylori DNA extracted should be of high quality and sufficiently pure for efficient digestion by restrictton endonucleases. Partial digestion with high numbers of faint uneven bands or smears of bands may result if DNA is insufficiently pure. The DNA may require another extraction with phenol/chloroform and/or ethanol precipitation. The DNA concentration of approx 2 @sample should be sufficient to produce well-resolved bands of adequate intensity. 2. Weak or faint band patterns suggest either the DNA concentration is too low and needs to be concentrated, or, alternatively, that the probe has lost activity. The development with substrate can be left for a longer period (overmght, if needed) instead of 2 h. However, longer periods of incubation may result in higher background coloring. It is preferable to use fresh probe and relabel if loss of sensitivity is the cause. 3. 1X TBE buffer for gel electrophoresis should be made fresh just before use. Old TBE buffer can precipitate if left over a few days and spots may result from particles of precipitation. 4. Spots or blotches may also appear because of incomplete or uneven transfer of digest fragments from gel to the membrane, because of air bubbles trapped between the gel and the membrane, or between the membrane and the Whatman paper. 5. Always handle nylon membranes with blunt forceps and gloved hands. Fingerprints, marks, or smears may be left on the membrane during the washing or the detection steps. 6. High background colormg may be reduced by decreasing the amount of labeled DNA in hybridization; increasing the volume of prehybridizatton solution, to allow the membrane filter m the bag to float freely; and raising the concentration of blocking reagents for detection.

References 1. Grimont, F. and Grtmont, P. A. D. (1986) Ribosomal ribonucleic acid gene restriction patterns as potential taxonomic tools. Ann. Inst Pasteur Mzcrobiol 137B, 165-l 75. 2. Altwegg, M., Hickman-Brenner, F. W., and Farmer, J. J., III (1989) Ribosomal RNA gene restriction patterns provide increased sensitivities for typing Salmonella typhi strains. J. Infect. Du. 160, 145-149. 3. Bialkowska-Hobrzanska, H., Harry, V., Jaskot, D., and Hammerberg, 0. (1990) Typing of coagulase negative staphylococci by southern hybridization of chromosomal DNA fingerprints using a ribosomal RNA probe. Eur. J. C&n. Microbial Infect Dls. 9,588-594.

4. Tee, W., Mayall, B., Lucas, C. R., Rayner, R., Pearson, S., and Dwyer, B. (1992) Use of ribosomal RNA gene restriction patterns to investigate two outbreaks of Campylobacter enteritis in Melbourne, Australia. J Med. Microblol 37,385--388.

98

Tee

5. Owen, R. J., Beck, A., Dayal, P. A , and Dawson, C. (1988) Detection of genomic variation in Providencza stuartiz clinical isolates by analysis of DNA restrlctlon fragment length polymorphism containing rRNA clstrons. J Clm Microbial 26, 2161-2166. 6. Saunders, N. A., Hanson, T G., Haththotuwa, A , and Taylor, A G. (1991) A comparison of probes for restriction fragment length polymorphisms (RFLP) typing of Legionella pneumophlla serogroup/strains. J Med Microbial 35, 152-158. 7. Morgan, D. D. and Owen, R. J. (1990) Use of DNA restriction endonuclease digest and ribosomal RNA gene probe patterns to fingerprint Helzcobacter pylorl and Hellcobacter mustelae isolated from humans and animal hosts. Mol Cell Probe 4,321-334. 8. Tee, W., Lambert, J., Smallwood, R , Schembn, M , Ross, B , and Dwyer, B. (1992) Ribotyping of Hellcobacter pylorr from clinical specimens. J Clan Microblol. 30, 1562-1567.

9. Fraser, A G., Buckley, J., Owen, R. J., and Pounder, R. E (1992) DNA fingerprints of Hellcobacterpylorz before and after treatment with omeprazole. J. Clan Path 45,1062-1065.

10. Owen, R. J., Bickley, J., Lastovlca, A., Dunn, J. P., Borman, P., and Hunton, C (1992) Rlbosomal RNA gene patterns of Hellcobacter pylorz from surgical patients with healed and recurrent peptic ulcers. Epldemlol Infect 108,39-50 11. Prewett, E. J., Buckley, J., Owen, R. J., and Pounder, R. E. (1992) DNA patterns ofHelicobacterpylor1 Isolated from gastric antrum, body and duodenum. Gastroenterology 102,829-833.

12. Owen, R J , Bell, G D., Desai, M., Mueno, M., Gant, P. W , Jones, P. H., and Lmton, D. (1993) Biotype and molecular fingerprints of metromdazole resistant strains of Helicobacterpylori from antral gastric mucosa. J Med Microbzoi 38,6-12. 13 Rautehn, H., Tee, W., Seppala, K , and Kosunen, T U. (1994) Ribotyping

patresistance in paired clinical samples of Hebcobacter pylori J. Clin. Mlcrobrol. 32, 1079-1082 Stull, T. L., LiPuma, J J., and Edlind, T D. (1988) A broad-sprectrum probe for molecular epidemiology of bacteria: rlbosomal RNA. J Infect Dzs. 157,280-285. Gustaferro, C. A. and Persing, D. H. (1992) Chemllummescent umversal probe for bacterial ribotyping. J Clm Mzcrobiol. 30, 1039-1041. Brosms, J., Ullnch, A., Raker, M. A., Gray, A , Dull, T J., Gutell, R R., and Noller, H. F. (198 1) Construction and fine mapping of recombinant plasmlds containing the rm B ribosomal RNA operon of E cob Plasmzd 6, 112-l 18. Cravloto, A., Beltran, P , Delgado, G., Navarro, A., Eslava C., Le6n, S., and Gonzblez, A. R. (1994) A subcloned ribosomal RNA probe for bacterial rlbotype analysis. J Infect. Dis. 169, 1413,1414. Brown, T. (1989) Analysis of DNA sequences by blotting and hybridization, m Current Protocols zn Molecular Biology (Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl, K , eds.), John Wiley & Sons, New York, pp. 2.9.1-2.9.10. terns and emergence of metronidazole

14. 15. 16.

17.

18.

12 Differentiation of H. py/cwlStrains Using PCR RFLP Harry L. T. Mobley and Paul A. Foxall 1. Introduction Helicobacterpylori strains have been shown to display considerable heterogeneity with respect to DNA sequence. Diverse restriction fragment length polymorphism (RFLP) patterns are generated among strains by restriction endonuclease digestion of whole chromosomal DNA (I-.?), digestion of specific polymerase chain reaction (PCR) products (4), or arbitrary primer-PCR and random amplified polymorphic DNA-PCR (5-7). These techniques demonstrate numerous distinct and reproducible patterns that can be used to dlfferentiate strains. Sequence analyses of specific genes have confirmed that there are frequent nucleotide substitutions among strains, usually in the third position of codons, which may not result in amino acid substitution within the corresponding polypeptides (8). This type of heterogeneity is found in all H. pylori strains, whether they are considered virulent or nonvirulent, To take advantage of this microdlversity, PCR products of specific genes can be PCR-amplified from H. pylori chromosomal DNA. When digested with appropriate restriction endonucleases, discrete digestion patterns are observed in agarose gels for each distinct strain. While the PCR-RFLP patterns vary among strains, these patterns are nevertheless quite stable within each mdividual strain. This method IS based on the work of Foxall et al. (4) in our laboratory. In this chapter, we outline the application of PCR amplification to the analysis of the restrictlon digest patterns of the urease structural genes, ureA and ureB, amplified from clinical isolates of H. pylori. The substrate for PCR amplification is DNA extracted from H. pyZori by alkali lysis and phenol-chloroform extraction. The 2.4 kb PCR products, amplifiable from all strains, are

From Methods /n Molecular Medmw, Hekobacter pylon Protocols Edlted by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

99

NJ

Mobley and Foxall

100

subjected to digestion with HaeIII, a restriction enzyme that recognizes the four-base sequence GGCC. Strains can be differentiated based on the sizes of the resulting restrtction fragments. This technique, or simple modifications of it, offer a sensitive epidemiological tool for differentiating or confirming the identities of clinical isolates and has been used in numerous studies (945). The restriction patterns generated by Hue111digestion of PCR amplified ureA and ureB can be used to differentiate strains on the basis of patterns on agarose gels, to identify the presence of multiple strains in a single biopsy specimen, to check for reinfection with the same strain following eradicatron therapy, and to identify identical (or different) strains among gastric biopsies obtained from family members. 2. Materials 1. H pylorz strains can be cultured directly from gastric biopsies taken by upper endoscopyof an infected patient by a gastroenterologist,or can be cultured from a frozen stockculture of a previously Isolated strain. 2. Growth medium: Brucella agar plates containing 5% sheep blood with and without Skirrow’s supplement (Product Diagnostics, Austin, TX). 3. Anaerobic jars. 4. CampyPaks (BBL, Cockeysville, MD) are envelopes used to generate the microaerobic atmosphere necessary for cultivation of H. pylon. 5. Lysis buffer A. 2 miI4 Tris-HCl, pH 8.0,25 mM glucose, 1 rmI4 EDTA, 100 pg/mL lysozyme. 6. Lysis buffer B: 50 n&I sodium chlorrde, 1% (w/v) sodium dodecyl sulfate, 2.5 mg/mL proteinase K 7. Sterile distilled water 8. Phenol-chloroform-uoamyl alcohol (1: 1:32).

9. Oligonucleotides with the specified nucleotide sequencecan be synthesizedby 10 11 12. 13. 14 15. 16. 17

any specialty facility that carries out this service. S’AGGAGAATGAGATGA-3’ (bp 308-322) (4.10) and SACTTTATTGGCTGGT-3’ (bp 27 18-2703) (4. IO). Taq polymerase PCR amplification buffer, 200 @dNTPs (both from Perkin-Elmer, Norwalk, CT). Agarose. NuSieve GTG agarose (FMC Btoproducts, Rockland, ME). TE buffer. 10 miI4 Trts-HCl, pH 8.0, 1 n&I EDTA. 0.5 Clg/mL Ethtdmm bromide DNA standards, 1-kb ladder (Gtbco-BRL, Gaithersburg, MD). UV light box equipped with camera and Polaroid film

3. Methods

3.1. Cell Growth 1 Gastric btopsles, obtained by upper endoscopy, are used to inoculate Brucella agar plates containing 5% sheep blood and Skurow’s supplement. Alternatively,

PCR-RFLP Differentiation

101

H. pylon strains that have been previously isolated can be used to maculate the agar plates (see Note 1). 2. Incubate plates for 3-4 d at 37°C under microaerobic condmons (sealed anaerobic jar with activated CampyPak). H pylori is identified as small pinpoint colonies that are positive for urease, oxidase, and catalase. The resulting bacterial growth can be used directly, or suspensions of H pylori can be stored for later use in TE (pH 8.0), 5 mM glucose, 50% glycerol at -70°C.

3.2. Preparation

of DNA for PCR

1. To purify chromosomal DNA from H. pylon, resuspend two 10-pL loopfuls of confluent growth m 400 pL of Lysis buffer A (see Note 2) and incubate the mixture for 30 min at 37°C. 2. Add 300 pL of Lysis buffer B (see Note 3) and incubate for 60 min. 3. Extract the sample with 500 pL of phenol-chloroform (see Note 4) and centrifuge at room temperature in a microcentrifuge for 6 min. 4. Remove 500 clr, the aqueous phase, ethanol precipitate the DNA, and resuspend the pellet in 50 pL of sterile distilled water. Store at 4°C (see Note 5). Typically, 5 pL vol will be used in the subsequent PCR amplification reactions

3.3. PCR Amplification To amplify the 2.4-kb fragment containing the ureA and ureB genes, two ohgonucleotide primers are used with sequences:SAGGAGAATGAGATGA-3’ (bp 308-322) (410) and 5’ACTTTATTGGCTGGT-3’ (bp 2718-2703) (4,10). In a thin-walled PCR tube (Perkin Elmer Cetus), mix on ice: water, PCR amplification buffer, dNTPs, primers, and template (see Note 6). Add Taq polymerase just prior to starting the reactions. After an imtial denaturation step of 94OC for 3 min, a 30-cycle amplification is used with these primers and consists of annealing at 50°C for 2 min, extension at 72°C for 4 min, and denaturation at 94°C for 1 min (see Notes 7-9). 3.4. Restriction Enzyme Digestion and Agarose Gel Electrophoresis 1. To determine whether the correct-sized PCR product was obtamed, run approx

0.2 pg of DNA on a 0.7% agarosegel. Stain the gel with ethidmm bromide and photographthe gel on the UV light box. PCRproductsof 2 4 kb should be amplified from the DNA isolated from each strain (Fig. 1). 2. Prior to restriction digest analysis, ethanol precipitate the amplified DNA and wash the DNA pellets twice with 70% ethanol. Resuspend the pellet in 20 I.~L of water.

3. Incubate the DNA with 5-10 U of HaeIII restriction endonucleaseat 37’C for 3 h in the buffer recommended or supplied by the enzyme manufacturer.

4. Electrophoresethe PCR products and DNA standardson a 3.5% w/v Nusieve GTG agarosegel (see Note 10).

Mobley and Foxall

102

L 1

2

3

4

5

6

7

8

9 10 11 12

L

Fig. 1. PCR amplification of DNA fragments containing the completeureA-ureB ureasestructuralsubunit genesfrom clinical isolatesof H. pylori. DNA was amplified from chromosomalDNA and electrophoresedin a 0.7% agarosegel. L: h Hind111standard; Lanes 1-12: 2.4-kb PCRproductsderived from 12 clinical isolatesof H. pylori.

5. Stain the gel with ethidium bromide, and photograph the gel on a UV light box (Fig. 2). 3.5. Analysis of Patterns on Agarose

Gels

Prior to HaeIII digestion, PCR-amplification should yield a 2.4-kb fragment as determined by agarose gel electrophoresis. Following digestion of the 2.4-kb product with liaeII1, the patterns observed on the stained agarose gel will contain an average of four distinct bands, but can range from one (no HaeIII sites) to six or seven bands (there may be more actual bands, but very small bands will not have been resolved on the gel). If the DNA template was derived from one strain, the sizes of the HaeIII-digested PCR product will add up in size to approximately 2.4 kb (however, some small fragments may not have been resolved). If the DNA template was derived from more than one strain (as when DNA was isolated directly from a gastric biopsy from a patient infected with more than one strain, or when DNA was isolated from a mixed culture), then the fragments derived from the HaeIII-digested PCR products will add up in size to significantly greater than 2.4, and, theoretically, as high as 4.8 kb. A third strain would obviously increase the number of bands and the total sizes of the fragments, and so on.

PCR-RFL P Differentiation

103

ABCDEFGHIJS

$8. -1.0 -0.52 - 0.39 - 0.34 -0.30

Fig. 2. Hue111 restriction digest patterns of the amplified urease genes on a 2.4-kb PCR product from representative H. pylori clinical isolates. Amplified DNA was digested with Hue111 and electrophoresed on a 3.5% Nusieve GTG agarose gel. Lanes A-J, representative Ii pylori clinical isolates; lane S. 1-kb ladder DNA standard. Strains can be considered different if one or more bands differ in size. For accurate analysis, strains that are being compared should be run in adjacent lanes. To confirm differences, the 2.4 kb products can be digested with different 4-bp cutting restriction endonucleases and reanalyzed on an agarose gel (see Note 11).

4. Notes 1. For successful primary isolation and subculture of H. pylori from gastric biopsy material, once inoculated, keep exposure time of the plates to an oxygen environment as short as possible. If subculturing from a known pure stock, better growth is often observed on Brucella agar plus 5% sheep blood plates when Skirrow’s supplement is not added. 2. Lysis buffer A without lysozyme can be stored at 4°C. Prewarm to 37X’ and add lysozyme just prior to use. Gently resuspend bacterial growth by gyration of the inoculating loop in the buffer. Do not vortex; this may cause shearing of the template chromosomal DNA. 3. Make up lysis buffer B fresh for each procedure. Once added, mix by gentle inversion of the microcentrifuge tube, again to avoid excessive shearing of template DNA. 4. Extraction with phenol chloroform effectively inactivates and removes the proteinase K prior to PCR amplification, and any other proteinaceous amplification inhibitors. The organic and aqueous phases should be mixed by gentle inversion over a number of minutes until an emulsion is produced. It is not necessary to do more than one extraction and no chloroform extraction is required; both of these steps would lead to loss of template DNA during manipulations.

104

Mobley and Foxall

5. H pylorz template DNA for PCR reactions can be stored at 4°C for 4-6 months

and still produce amplified product 6. When possible, try to make up a master mix of the amplification reaction wtthout template or Taq polymerase added. This minimizes pipeting errors, and ensures

7

8.

9.

10

11

that all reactions occur in the same reaction matrix. After mixing, aliquot the master mix at the required volume into the reaction tubes. Always include a negative control without template DNA in each run. All pipetmg manipulations should be carried out using aerosol-resistant barrier tips, even the preparation of template DNA prior to amplification. It is desirable to have a set of pipets solely dedicated to transfer of amphfied reactions, using another set of pipets for template production, PCR reagent dilutions, and PCR amplification set-up. All reagents, buffers, primers, and template DNA solutions should be aliquoted mto low repeat-use volumes to reduce the possrbrlrty of contammatmg such stocks with amplified products. If contammation is detected, discard m-use ahquots, and use fresh, unopened ahquots of all reagents. When examining the ethidium bromide-stained agarose gel for the 2 4-kb PCR product, differences m the intensity of amplified product can be observed. This IS often caused by varying concentrations of starting template DNA m different samples, as no effort is made to quantify the template DNA prior to amplification. Either more template DNA can be added to the PCR reaction (up to 20 pL), or the volume that the ethanol-precipitated PCR product is resuspended in can be reduced accordingly To detect the full range of restriction fragment lengths, the digest can be split into 40%/60% aliquotes of the final volume. The 40% aliquot of the complete set of samples is loaded onto the gel, run for one-thud of the time, and then the 60% sample aliquots are loaded and the gel run to completion The second set of samples allows detection of the smaller fragments (less than 150 bp) that may be generated in the digestion; the larger volume helps m detection of such small fragments upon staining. Double restriction digests can also be used to drfferentiate between strains in the event of one pattern bemg predominant in a population of isolates. If the enzymes require different buffer conditions, sequential digestions should be carried out with the enzyme utilizing the lower concentrations m the first reaction, adJustmg buffer conditions with concentrated stocks prior to addmon of the second restrrction enzyme.

References 1. Langenberg, W., Rauws, E. A. J., WrdjoJokusumo, A., Tytgat, G. N. J., and Zanen, H. C. (1986) Identification of Campylobacter pylorzdis isolates by restriction endonuclease DNA analysis. .J. Clin. Microbial. 24,414417. 2. MaJewski, S. I. H and Goodwin, C. S. (1988) Restriction endonuclease analysis of the genome of Campylobacterpylori with a rapid extraction method: evidence for considerable genomtc variation. J, Infect Du 157,465-47 1.

PCR-RFL P Differentiation

105

3. Oudbier, J. H., Langenberg, W., Rauws, E. A. J., and Bruin-Mosch, C. (1990) Genotypical variation of Campylobacter pylorl from gastric mucosa. J. Clan Microbzol. 28,5.59-565.

4. Foxall, P. A., Hu, L.-T., and Mobley. H. L. T. (1992) Use of polymerase chain reaction-amphfied Hellcobacter pylori urease structural genes for differentiation of isolates. J Clin. Mtcroblol 30,739-74 1. 5. Akopyanz, N., Bukanov, N. O., Westblum, T. U., Kresovich, S , and Berg, D. E (1992) DNA diversity among clinical isolates of Hehcobacterpylori detected by PCR-based RAPD fingerprintmg. Nuclew Aczds Res. 20,5 137-5 142. 6 Akopyants, N. S., Eaton, K. A., and Berg, D. E (1995) Adaptive mutation and co-colonizatton during Helzcobacterpylori infection of gnotobtotic piglets Infect Immun 63, 116-121.

7 Taylor, N. S., Fox, J. G., Akopyants, N. S., Berg, D. E., Thompson, N., Shames, B., Yan, L., Fontham, E., Janney, F., Hunter, F. M., and Correa, P. (1995) Long term colonization with single and multiple strams ofHelzcobacter Pylori assessed by DNA fingerprinting. J. Clin Mlcrobrol 33,9 18-923. 8. Courcoux, P., Freuland, C , Piemont, Y., Fauchere, J. L., and Labigne, A. (1990) Polymerase cham reaction and direct DNA sequencing as a method for distinguishing between different strains of Helicobacter pylon. Rev Esp. Enf Dzgest 78(Suppl.

l), 29,30.

9. Bamford, K B , Bickley, J , Collins, J. S., Johnston, B. T., Potts, S , Boston, V., Owen, R J , and Sloan, J. M. (1993) Helicobacter pylon: comparison of DNA fingerprints provides evidence for mtrafamilial infection. Gut 34, 1348-1350 10 Clayton, C. L , Kleanthous, H., Coates, P. J., Morgan, D. D., and Tabaqchali, S (1992) Sensitive detection of Hehcobacter pylorr by using polymerase chain reaction. J. Clin. Microbial. 30, 192-200. 11 Desai, M., Lmton, D , Owen, R. J., and Stanley, J. (1994) Molecular typing of Hellcobacter pylori isolates from asymptomatic, ulcer and gastritis patients by urease gene polymorphism. Epidemlol. Infect. 112, 151-160. 12 Hurtado, A. and Owen, R. J. (1994) Identification of mixed genotypes m Helzcobacter pylori from gastric biopsy tissue by analysis of urease gene polymorphisms. FEMS Immun. Med. Mlcrobiol. 8,307-3 13. 13. Lopez, C. R., Owen, R. J., and Desai, M. (1993) Differentiation between isolates of Helicobacteripylorl by PCR-RFLP analysis of urease A and B genes and comparison with ribosomal RNA gene patterns FEMSMicrobiol Lett 110,37-44 14. Moore, R. A., Kureishi, A., Wong, S., and Bryan, L. E. (1993) Categorization of clinical isolates of Hehcobacter pylori on the basis of restriction digest analysis of PCR-amplified ureC genes. J. Clin Mzcrobiol 31, 1334,1335 15. Owen, R. J., Bickley, J., Hurtado, A., Fraser, A., and Pounder, R. E. (1994) Comparison of PCR-based restriction length polymorphism analysis of urease genes with rRNA gene profiling for monitoring Helicobacterpylorl infections in patients on triple therapy J. Clin Mzcrobiol 32, 1203-1210.

13 PCR-DNA Sequence Typing Agnes Labigne 1. Introduction There is a great need to develop molecular techniques for the typing of Helicobacter pylori isolates since classical bacteriological assays, such as serotyping or lysotyping, are lacking. Such techniques would provide the microbiologist with the tools necessary to differentiate one strain from another one (essential for epidemiological studies), to determine whether individuals might be colonized by several isolates, and to elucidate whether ulcer relapse following treatment is associated with the colonization of the gastric mucosa by the same isolate or is caused by reinfection. Different molecular techniques have been developed and used for the typing of H. pylon isolates. The most effective ones are based on the high degree of genomic variability that exists among isolates belonging to the H. pylori species. This was first demonstrated by restriction fragment length polymorphism (RFLP) of the whole chromosome of H. pylori restricted with Hind111or Hue111 (I). Later, several authors used gene amplification followed by restriction of the polymerase chain reaction (PCR) product to demonstrate this polymorphism (2-4). Both approaches are able to discriminate between isolates; however, the former is time-consuming (because of the requirement for the culturing of isolates, followed by DNA extraction, as well as endonuclease restriction), and the latter often requires the use of several restriction enzymes to be highly discriminative. More recently, the arbitrary-primer PCR, also designated PCR-based random amplified polymorphic DNA (RAPD) tingerprinting (5), was introduced as a way to differentiate isolates. This technique is very powerful; however, to achieve good reproducibility, it is necessary to prepare well-purified and well-calibrated DNA. These molecular typing approaches are described in other chapters m this volume. From Methods WI Mol8Cular Mechcne, Hehcobacter pylori Protocols Edrted by C. L. Clayton and H L. T Mobley Humana Press Inc , Totowa,

107

NJ

108

Labigne

The PCR-DNA sequence typing of H. pylori isolates is a procedure that, in a very short period of time (treatment of 12-24 isolates within 24 h), allows one to differentiate isolates with a very high degree of discrimination without bacterial culture or DNA purification. Its limitations are based on rts use of radioisotopes (35S). The procedure is based on the gene amplification of a highly polymorphic genomic region known to be present in all H pylori isolates (6). This PCR step is followed by the determination of the nucleotide sequence of the amplified DNA fragment (PCR product) (7,8). The reading of the nucleotide sequence over a 150-bp region, and the comparison of the sequence with that obtained from another isolate, indicates whether the two compared isolates are identical (same nucleotide sequence reading) or different isolates. Using this molecular-typing approach, we have shown that, of 80 mdependent isolates tested, none of them had the same nucleotrde sequence. In contrast, over a 1-yr follow-up, it was demonstrated that one patient was infected by a unique strain of H. pylon and that, over that period, the sequence remained unchanged. Moreover, the determination of the nucleotide sequence from strains isolated before and after treatment from the same patients allowed one to conclude that the observed relapses were associated with the colonization of the gastric mucosa by the same isolates, demonstrating, therefore, that relapse was caused by treatment failure or by reinfection from the same source. The general procedure includes three major steps.The first one concerns the amplification of a DNA fragment of the H. pylorz genome. As mentioned, the two ohgonucleotides used for the amplification have to target a gene that is present in all H. pylon isolates. In addition, the amplified DNA fragment should demonstrate high polymorphism over a short stretch of DNA. The HP1 and HP2 oligonucleotide primers depicted in this procedure target the urease regton of the H. pylon isolates (internal to the so-called ureC open reading frame [9, IO]) and allow the amphfication of a 294-bp fragment (7,8). It is crucial that the PCR be carried out under conditions when only a single amplification product is produced, allowing the direct sequencmg of the PCR product without purification steps. During the second step of the procedure, the sequencing of the double-stranded PCR fragment generated by amplification is accomplished followmg a treatment of the PCR product by exonuclease I and shrimp alkaline phosphatase; these enzymes remove, respectively, the excess of primers used for the PCR amplification and the excess of deoxynucleotides (dNTPs). Sequencing is then performed by the dideoxynucleotide technique of Sanger (II), using one of the two primers, HP1 or HP2 in the annealing reaction, followed by a labeling step, and a termination step. The four reactions (T, C, G, A) ortgmatmg from a PCR product are then run in parallel on denaturing

PCR-DIVA Sequence Typing acrylamide gels and the nucleotide by reading of the autoradiograms.

109 sequences of the PCR products determined

2. Materials and Reagents 2.1. Materials 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

DNA Thermal Cycler. Microcentrifuge. Mini-horizontal gel electrophoresis apparatus. Water or dry bath at 37’C. High-voltage power supply (V > 2000). Boiling water bath. A sequencing apparatus composed of a set of two plates, two spacers, and a shark comb; two drawing clamps. A 50-mL syringe, adhesive tape (at least 3 cm wade), 100-mL graduated cylinder, beaker (200 mL), and Erlenmeyer flask (250 mL). Regular mtcrocentrifuge tubes (1.5 and 0.5 mL) Colored microcentrifuge tubes (1.5 mL blue, yellow, red, and green). Radiological cassette (35 x 43 cm). Gel dryer. UV light (wavelength 300-360 run).

2.2. Reagents 1. 1OX PCR: 100 mA4 Tris-HCl, pH 8.3,25 mM MgClz, 500 nnI4 KC1 2. Nucleotrdes: 2 mM each dATP, dCTP, dGTP, dTTP. 3. Ohgonucleotides (20 pmol/pL): HP 1, 5’-AAA GCT TTT AGG GGT GTT AGG GGT TT-3’ and HP2,5’-AAG CTT ACT TTC TAA CAC TAA CGC-3’. 4. Taq DNA polymerase (5 U/pL). 5. Paraffin oil. 6. 0.5MEDTA, pH 8.0: 186 g dissolved in 1000 mL of distilled water. EDTA wrll only dissolve under neutral pH. Adjust the pH to 8.0 with 1ONNaOH solutron, while stirring to permit complete dissolution. 7. 50X TAE: Trizma-base 242 g, glacial acetic acid 57.1 mL, 0.5MEDTA 100 mL: drssolve and adjust to pH 8.0. Can be kept at room temperature (see Note 1). 8. Ethidium bromide: Stock solution 10 mg/mL (Caution: use gloves for handle this reagent). 9. Agarose gel: Dissolve 1 g of normal agarose and 1 g of low melting pomt agarose in 100 mL of 1X TAE buffer; add ethidium bromide to a final concentration of 400 pg/mL, cool at 55°C before pouring in the casting boxes. 10. 5X Blue loading buffer: For a final volume of 10 mL, mix 5 mL of glycerol, 5 mL of (Tris [O. 1 mA4, pH 8.01 EDTA [-O.Ol n-&I, pH 8.01) and 7 mg of bromophenol blue. 11. Sequenase PCR product sequencing kit (Amersham product number US70 170, Amersham, France). 12. [a-35S]-dATP, with specific activity ranging from 1000 to 1500 Ci/mmol. 13. Filter paper and Whatman 3MM paper.

110

Labigne

14. Tergazyme(or any glasslaboratory detergent). 15 Autoradrograpby film (Kodak XAR-5 or Hype&m Amersham [35 x 43cm]) 16 20X TTE buffer 216 g Trts-base, 72 g taurine, 20 mL 0.5M EDTA; stir and adjust the final volume to 1000 mL with distilled water. 17 Stock acrylamide, 40%: dissolve 38 g of acrylamtde, and 2 g of bu-acrylamide in distilled water, and adjust the volume to 100 mL Add 5 g of Amberhte MB 1 and stir with a magnetic bar for 30 mm to delomze the acrylamlde solution. Pretilter on a regular filter paper to remove the beads, then filter through a nitrocellulose

filter (0.45 pm). Storem dark at 4°C Alternatively, ready to use40% acrylamtde solution from Bioprobe (ref. BIAC02), or another brand can be used. 18. TEMED (N&VW-tetra-methyl-ethylenediamme). 19. 10% Ammonium persulfate (APS) Dissolve 1 g m 10 mL distilled water, fractionate and store at -20°C. Samples should only be thawed once.

3. Methods 3.7. Polymerase Chain Reaction 3.1.1. Preparation of the Bacterial Lysate 1. Resuspend H pylori cells from a plate in 100-500 pL of sterile disttlled water (dependmg on the density of the growth on the plate); the suspension should correspond to approx lo8 bacteria/ml, equivalent to an ODbOs of 0.4-0.5 (see Notes 2 and 3). 2. Boil the suspension for 5 mm in a water bath in order to lyse the bacteria and release the H pylon genomic DNA (see Note 4). Cool the suspension on ice for 5 mm, and spin the suspension at 12,000g for 5 min in a microcentrifuge 3 Collect the supematant and store it on ice for immediate use (see Note 5)

3. I .2. Amplification Reaction 1. Set up the amplification reaction by adding (in the given order) the followmg reagents to a steriie 0.5 mL microcentrifuge tube 53.5 @, distilled and sterile water, 10 $, 10X PCR buffer, 4 I.& each of the 4 nucleotides (dATP, dCTP, dGTP, dTTP), 5 pL oligonuclotide HPl, 5 pL oligonucleotide HP2, 0.5 $ Taq polymerase, 10 p,L boiled, and 10 & ice-cooled H pylon supernatant. Vortex, spin 5 s, and add 3 drops of paraffin oil. 2 Amplification is performed m a thermal cycler by performing 25 cycles including the three steps: 2 mm at 94°C for denaturation, 2 mm at 56°C for anneabng, and 2 mm at 72°C for elongation (see Note 6).

3.1.3. Controlling the Quality and the Yield of the PCR Product Prepare a 2% agarose mini-gel as indicated m Section 2.2., and mix independently 1 pL and 20 pL with agarose 5X Blue loading buffer (BLB) to achieve a final concentration of 1X BLB; load these two samples u-rwells (see Note 7), as well as markers for fragment length. Migrate, and visualize under UV. The sample conststing of 1 pL out of the 100 & PCR product should be

PCR-DNA Sequence Typing

111

wsible to guarantee a good sequencing reaction (good yield) and the 2O+L sample should demonstrate the presence of a single band migrating as a 294-bp fragment (see Note 8). 3.2. Sequencing Reaction 3.2.1. Preparation of the Sequencing Gel 1. Weigh 42 g of urea. Place urea in a 100-mL cylmder; add 15 mL of stock acrylamide (40%), 5 mL of 20X TTE taurine buffer (see Note 9), and adJust the volume to 100 mL with disttlled water; mix by turning upside down after sealing the cylinder with parafilm. When the urea is completely dissolved, adjust the final volume to 100 mL with water. Pour the acrylamide solution into a 200-mL beaker. 2 During the dissolution of urea, carefully wash the large glass plates (part of the sequencing apparatus) with Tergazyme; rinse these extensively with distilled water, and then with 90°C ethanol. Wipe away any trace of alcohol with good quality absorbent disposable towels. 3 Position the two lateral spacers on the sides of one of the two plates, and superpose the second plate on the top of the spacers. Maintain the two plates with one or two drawing clamps. 4 Tape the glass plates to form a U, being careful to eliminate all air bubbles trapped underneath the tape. Envelop the bottom edge of the two plates with tape, ensuring that the tape at the comers is air tight. Plates are now ready for use, and the acrylamide solution can be poured. 5 Just before pouring the gel, add 25 pJ.,of TEMED and 1 mL of APS (10%) freshly prepared or thawed. Mix well with a stirring bar. 6. Take up the acrylamide mto a 50-mL syringe and inject the acrylamide solution slowly, without introducmg air bubbles. Initiate the pouring by inclining the two plates at a 45” angle; lower them slowly to a horizontal position while pouring. 7. Insert the comb (shark teeth side opposite to that of gel) and maintain the comb with a drawing clamp (the clamp should press on the comb and not on the gel) Let the acrylamide solidify in the horizontal position, resting the plates on two Petri plate covers (comb side up). When the gel has solidified, cover the top with plastic wrap, and let the gel rest a minimum of 2 h and a maximum of 20 h.

3.2.2. Cleaning the PCR Product from Primers and Nucleotides 1. This step should be achieved according to the mstructions of the manufacturer when using the Sequenase PCR product sequencing kit (Amersham). 2. Briefly, 5 pL of the PCR product should be treated simultaneously with 1 pL of exonuclease I (10.0 U/pL) to digest all the HP1 and HP2 ohgonucleotides (primers) that were not extended during the PCR, and 1 pL of shrimp alkaline phosphatase (2.0 U/pL) to dephosphorylate all the unincorporated nucleotides (dNTPs). Mix and incubate the 7 pL at 37°C for 15 mm. Both enzymes are then inactivated by incubating the reaction at 80°C for 15 min.

112

Labigne

3.2.3. Sequencing Reaction 1. Annealing of the HP1 oligonucleotide (pnmer) with the amplified treated template DNA. To the 7 & corresponding to the treated PCR product DNA (see Section 3.2 2 ), add 1 pL of HP1 (1 e., 20 pmol) and 2 pL of distilled water Vortex and spm briefly the 10 & (mixture 1) Denature the amplified DNA by heatmg for 3 mm at 100°C in a boiling water bath, then immediately plunge the tube mto Ice water for 5 mm CentrlfUge briefly and keep mixture 1 on ice while preparing the other reagents. 2 Prepare four tubes for each DNA template. one blue (2’-dideoxythymidme 5’ triphosphate, ddT), one yellow (2’-dideoxycytldine 5’ triphosphate, ddC), one red (2’-dideoxyguanosme 5’ triphosphate, ddG), and one green (2’-dldeoxyadenosine 5’ triphosphate, ddA). Introduce 2 5 pL of each ddNTP mixture m the bottom of each tube, corresponding to reactions T, C, G, and A, respectively (2.5 & of ddT m the blue tubes, 2.5 $ of ddC in the yellow tubes, 2.5 pL of ddG m the red tubes, and 2.5 pL of ddA in the green tubes). Close the hds and incubate the tubes at 37’C 3. Dilute the mixture of deoxynucleotldes (dNTPs), designated in the Amersham kit as dGTP, 1.5 m water and keep the dilution on ice. 4 Set up the following reaction (mixture 2) by mixing, then vortexing and spmnmg: 2 pL of reaction buffer, 1 pL of O.lM DTT (dithiothreitol), 2 pL of the diluted (1:5) deoxynucleotldes (dGTP), 0.5 & of 35S-dATP, 2 & of Sequenase (optional, see Note 10). Mixture 2 must be kept on ice until it 1sused (see Note 11). 5. Labeling reaction: Place the tube containing mixture 1 (primer annealed to the template) m a microcentrifuge. Spin for 2 s. Leave the tube in the microcentrifuge, and open the lid. Spot on the side of the tube (without touching mixture 1) 7.5 & of mixture 2. Close the tube gently and centrifuge for 2 s. The labeling reaction (mixture 3) begins at this point. Incubate for 5 min. During this time, place the four colored tubes (ddT, ddC, ddG, ddA) preincubated at 37’C in the microcentrifuge, with lids open, and distribute on the top side of each of the four tubes 3.5 pI., of mixture 3. Close the lids and wait for the end of the 5-min labeling reaction. 6. Elongation and termination of the polymerization. At the end of the 5-min mcubation, centrifuge for 2 s. This allows the beginning of the polymerization of the nucleotides, as well as the dideoxynucleotides, that will determine the termination of the elongation. Transfer the tubes at 37°C and incubate for 10 min. 7. Stopping the reaction: When the IO-mm incubation at 37°C is over, transfer the four colored tubes in the microcentrifuge, lids open, and add 4 mL of the blue stop solution on the top side of each of the four tubes. Close the tubes, and centritige for 2 s to stop all the reactions at the same time (mixture 4). Place the tubes on ice or, alternatively, store them at -20°C if the loading on the gel IS postponed

3.2.4. Migration of the Samples 1 Remove the tapes from the plates, and wet them around the comb with distilled or deionized water to facilitate the removal of the comb. Remove the comb; wash and wipe it. Using a squeeze bottle of distilled water, abundantly wash the large

113

PCR-DNA Sequence Typing

indentation of the comb to remove the urea, as well as any pieces of polymertzed acrylamide. Absorb all the water with filter paper and introduce the shark’s teeth of the comb in the acrylamide to form loading wells. The points of the comb should be embedded at least 2 mm to prevent leakage from one well to another. Mount the plates on the apparatus, and fill the two tanks with 1X TTE buffer. Flush each of the wells with buffer to clean again the wells with 1X TTE and to eliminate any air bubbles. Turn on the current to preheat the gel at 35 W and constant power for 15 min. Loading the samples: Just before loading the samples, denature the four reactions corresponding to one template by incubating the four colored tubes in a water bath at 90°C for 2 min. During this denaturation step, turn off the current of the apparatus sequencing gel and clean the four wells that will be loaded by flushmg buffer with a Pasteur pipet. Using a pipetman, load 4 pL of each of the ddT, ddC, ddG, and ddA colored tubes in this order (see Note 10). Turn on the current and let the samples migrate until the bromophenol blue (dark blue) elutes out of the gel (generally, 1.5-2 h at 55 W). If a longer readmg 1s required, let the samples migrate until the cyan01 xylene (green-blue) reaches the bottom and count 1 h before turning off the current.

3.2.5. Gel Fixation and Autoradiography

of the Gel

1. When the current is off, remove the buffer from the bottom tank and store with the radioactive waste. Empty and discard into the sink the buffer from the upper tank. Detach one of the two plates, leaving the acrylamide gel restmg on one of the plates. Place a sheet of Whatman 3MM paper on the surface of the gel, and lift it progressively, starting from the bottom of the gel to the top The gel should remain stuck on the Whatman paper. Cut off the excess paper, and cover the gel with plastic wrap. Dry the gel for at least 60 min at 80°C under vacuum. 2. Remove the plastic wrap, and expose the gel in a radiographic cassette in contact with a film at room temperature for 18 h, or more if required.

3.3. Reading of the Nucleotide Sequence 1. Reading: An illustration of the autoradiogram that can be obtained followmg 24 h of exposure is shown in Fig. 1, where the gel is read from the bottom to the top, as follows; the first line corresponds to the reading of the nucleotide sequence of the stram No. 1, and the second line to that of the strain No. 2. TGTCGCTACGAGCATGAGCAATTTAGCCCTTAAAGAATACCTGAAATCCCAAGATTTAGAATTGAAGCATTG.. , TGTCGCCACAAACATGAGCAATTTAGCCCTTAAAGAATACTTAAAATCCCAAGATTTAGAATTGAAGCATTG... The two nucleotide sequences differ from each other by 5 nucleotides (nucleotides shown as bold A, T, G, and C); it can be concluded that the two isolates are not the same. 2. If following the reading of the sequence from the HP1 oligonucleotide, the first 150 oligonucleotides appeared to be the same, the identity of the strains

Labigne

174 TCGA

Isolate No1

TCGA

Isolate No2

Fig. 1. Autoradrograph of a sequencing gel illustrating the nucleotide sequences of two independent isolates. The arrows pomt out the differences between the two nucleottde sequences. might be confirmed by a second sequencmg reaction in which the HP2 oligonucleotrde will be used in the annealing step (see Section 3.2 3.) instead of the HP 1 olrgonucleotlde 3. A reading of the 132 nucleottdes using the HP1 ohgonucleotides has usually been dtscriminative enough to classify isolates as identical or different rsolates. Below is the 132 nucleotide-prototype sequence that can be read from HP 1; nucleotrdes represented as bold A, T, C, or G letters are those for which, so far, we have observed some variabthty. These nucleotides correspond to hot spots.

PCR-DNA Sequence Typing

115

TCTTCTCAAGCAATTGTCGCTACAAACATGAGCAATTTAGCC CTT AAA GAA TAC TTA AAA TCC CAA GAT TTA GAA TTC AAG CAT TGC GCG ATT GGG GAT AAG TTT GTG AGC GAA TGC ATG CGA TTG AAC AAA.

4. Notes 1. When preparmg an electrophoresis buffer, always use Tris-base and glacial acetic acid to adjust the pH Never use Tris-HCI or Tris-base and hydrochloric acid since the presence of chloride ions m the buffer will cause major problems in the migration of the samples (including terrible white smears) 2. It is very important to record the origin of the lysate and to know whether it was obtained from a single colony of H pylori, or whether it might originate from several colonies and, possibly, from different isolates. 3 The amplification might also be performed directly from a biopsy specimen For successful typing, the amplified product will be adequate for sequencing if it fulfills the following criteria, i.e., a single 294-bp band visible from a 20-a sample of the PCR product (total volume 100 pL) and still visible when 1 pL of the PCR product is loaded on a mini-gel. 4. In order to prevent cross-contamination between the samples, it is important to have a boiling lysate in a volume below or equal to 400 pL in a 1.5-mL microcentrifuge tube 5. If the supernatant is not used directly for PCR, it can be stored at -20°C. In this case, the supematant must be reboiled for 3 min immediately prior to initiating the PCR. 6. The time (2 min) for each step can be shortened to 1 min for each of them, if the DNA thermal cycler used is very effective in controllmg the temperature. If the sequencing reaction is postponed, the PCR product has to be stored at -20°C 7. The wells of the mini-gel should be 3-5 mm in width to allow an adequate appreciation of the PCR product yield. 8. The HP1 and HP2 oligonucleotides used in the PCR reaction should give a single band. If this is not the case, the temperature of annealing might have to be increased by 1 or 2 degrees to reach this goal, i.e., a single product of amplification. 9. Taurme is used instead of the more classical boric acid because of its compatrbtlity with glycerol-contammg samples. Note that the sequencing reaction (see Section 3.2.3.) is performed by using 2 pL of undiluted sequenase provided m glycerol buffer, but that, in regular sequencing protocols, the sequenase is usually diluted to 1:8 before use. 10. If the first nucleotrdes of the sequence cannot be read clearly, then the reading of the gel might be improved by adding 1 pL 0. 1M MnCl, to the 7.5 p.L of mixture 2 (as given in Section 3.2.3.). 11 Mixture 2 can be prepared for several sequencing reactions run in parallel (up to three strains, if a 12-tube microcentrifuge is available, or six, if a 24-tube microcentrifuge can be used). In these cases, mix, spin, and store on ice, respectively, four times or seven times, the volumes of all the reagents of mixture 2, and

Labigne

116 spot 7.5 pL on the upper part of the tube containing mixture labeling step of three or six templates at a time.

1, allowing the

References 1 Langenberg, W , Rauws, E., Widjojokusumo, A., Tytgat, G N. J , and Zanen, H C. (1986) Identification of Campylobacter pyloridis isolates by restriction endonuclease DNA analysis. J Clin. Mzcrobiol 24,414-4 17. 2. Clayton, C., Kleanthous, H., Morgan, D. D , Puckey, L , and Tabaqchali, S (1993) Rapid fingerprinting of Helzcobacter pylorz by polymerase chain reaction and restriction fragment length polymorphism analysis. J. Clan. Mzcro. 31, 1420-1425 3. Moore, R. A., Kuretshi, A., Wong, S., and Bryan, L. E. (1993) Categorization of cluucal isolates of Helicobacterpylori on the basis of restriction digest analyses of polymerase chain reaction-amplified ureC genes. J Clin Mzcrobzol. 31,1334-1335 4. FuJlmoto, S., Marshall, B., and Blaser, M. (1994) PCR-based restriction fragment length polymorphism typing ofHelzcobacterpylori. J Clan. Mzcrobiol 32,33 l-334 5. Akopyanz, N., Bukanov., N.O., Westblom, T. U., Kresovich, S., and Berg, D. E (1992) DNA diversity among clinical isolates of Helicobacter pylorr detected by PCR-based RAPD fingerprinting. Nuclezc Acid Res 20,5 137-5 142 6. Labtgne-Roussel, A , Courcoux, P., and Moyen, E (1989) Development of gene probes for the detection and characterizatton of Campylobacter pylon, in Gastroduodenal Pathology and Campylobacterpytori (Megraud, F. and Lamouliatte, H., eds.), Excerpta Medica, Amsterdam, pp. 123-125. 7 Ferrero, R. L. and Labigne, A. (1983) Orgamzatton and expression of the H. pylon urease gene cluster, m Helzcobacter pylorz Bzology and Clznzcal Practice (Goodwin, C. S. and Worsley, B., eds.), CRC, Boca Raton, FL, pp. 171-190 8. Courcoux, P., Freuland, C., Ptemont, U., Fauchere, J. L., and Labigne, A. (1990) Polymerase chain reaction and direct DNA sequencing as a method for distmguishmg between different strains of Helicobacter pylori, in Gastroduodenal Pathology and Helzcobacter pylon, Enfermedades Digestwas, No. 12, Suppl I, Abstract P-49. 9. Labigne, A., Cussac, V., and Courcoux, P. (1990) Shuttle cloning and nucleotide sequences of Helicobacter pylorz genes responsible for urease activity J Bacterial. 173, 1920-l 93 1 10. Kansau, I., Raymond, J., Bingen, E., Courcoux, P., Kalach, N., Bergeret, M., Braimi, N., DuPont, C., and Labigne, A. (1996) Genotypmg of Helzcobacterpylorz isolates by sequencing of PCR products and comparison with the RAPD technique. 11. Sanger, F , Nicklen, S., and Coulson, R. (1977) DNA sequencing with chain-terminating inhibitors. Proc Nat1 Acad Scz USA 74,5463-5467.

H. pylori DNA Fingerprinting

Using the Arbitrarily Primed PCR (AP-PCR) or Random Amplified Polymorphic DNA (RAPD) Method Douglas E. Berg, Janaki Lelwala-Guruge, Engin T. Incecik, Kalpana Srivastava, and Natalia S. Akopyants 1. Introduction An ability to distinguish individual strains of Helicobacterpylori with sensitivity and efficiency IS valuable for studies of the epidemiology, populatton genetic structure, and evolution of this gastric pathogen. The arbitrarily primed polymerase chain reaction (AP-PCR), or random amplified polymorphic DNA (RAPD) method (I-#.), provides one of the most sensitive and efficient means for distinguishing individual strains, and has been particularly useful for H. pylori (5-7). In overview, the method entails PCR amphtication with an ohgonucleotide primer of arbitrarily chosen sequence and no known match to sequences in the target genome. This allows initiation of DNA synthesis from genomic sites to which the primer 1s fortuitously, and usually only partially, matched (Fig. 1). The method detects DNA sequence diversity throughout the genome, rather than just at individual loci; less DNA is needed than m most other DNA fingerprinting methods; the DNA need not be very large nor be double-stranded; and no DNA labeling or hybridization, nor information about target DNA sequences,is needed. There are two principal variants of the AP-PCR protocol, one using oligonucleotide primers of about 10 nucleotides (nt) (3,4), and a second using longer primers, which often may have been constructed for other purposes, such as conventional PCR or DNA sequencing (1,2). A number of primers that work well for H. pylori were identified empirttally as those yielding arrays of some 3-20 prominent fragments that are strain-specific and reproducible (e.g., primers 1281 and 1254, m Fig. 2). Other From Methods m Molecular Mechne, Hehcobacter pylorl Profocols Edlted by C. L Clayton and H L T Mobley Humana Press Inc , Totowa,

117

NJ

Berg et al.

AP-PCR

Product

Profile

Templalc DNA

DETAIL A

.

I

1. law strbqency annrallng of arbilrary primen lo pdally matched slles

/ -.

Arbitrary Primer

2, PCR amplllldon RAPD pmducl

Fig. 1. Strategy for DNA fingerprinting by the arbitrarily primed (AP) PCR or random amplified polymorphrc DNA (RAPD) method In the top left are diagrammed the genomes of related but genetically distinct strains of H. pylon that may have diverged from a common ancestor by mutation and/or gene transfer from other strains Pairs of thick half-arrows indicate primer annealing to pairs of sites that result in AP-PCR products; thin half-arrows indicate the same primer annealing to individual sites that are not near enough to other potential primer binding sites in opposite orientation to yield AP-PCR products. The annealing of primers to pairs of incompletely matched sites, which 1spostulated to be responsible for many AP-PCR bands from prokaryotic genomes, IS diagrammed in the DETAIL section (lower left), and the array of products that would be generated from the two strains compared here IS diagrammed at the right. primers did not give such good amplification (e.g., primer 1292, in Fig. 2), and were not studied further. There are several explanations for dtfferences in these profiles among strams, including: rearrangements, such as deletions or insertions in regions from whtch a given fragment is amplified; base substitution differences m potential primer binding sites; and even sequence differences between binding sites that affect the ease of DNA synthesis in vitro.

We (S-7) and others (8,9) are using the AP-PCR fingerprinting method m several types of studies, mcludmg testing for relatedness among epidemiologitally linked H. pylori isolates, such as those from members of the same family; analyzing recrudescence after apparent curing of infection;

detecting divergence

among closely related isolates;and testing for co-colonization and also horizontal gene transfer during experimental

or naturally

occurring mixed infections.

A P- PC I? Fingerprinting

119

Primer strein

1281 1

2

3

4

5

1

2

1254 3

1292 4

5

1

2

3

4

5

Fig. 2. Representative tests to identify primers useful for AP-PCR analysis of H. pylori. The sequence of primer 1292, which gave very little amplification, is 5’CCCGTCAGCA. The sequences of primers that gave informative profiles are listed in the text. The H. pylori strains listed here as 1,2,3,4, and 5 are NCTC 11638, WV229, NCTC 11637, P466, and MO 19, respectively.

2. Materials 1. Freezing medium: Tryptic soy broth (Difco, Detroit, MI) supplemented with 20% glycerol (TSB glycerol). 2. Culture medium: As solid medium, brain-heart infusion (BHI) agar (Difco), supplemented with 7% saponin-lysed horse blood (Colorado Serum, Denver, CO) and antibiotics (6 mg vancomycin, 5 mg trimethoprim, and 0.8 mg amphotericin B per 1 L, all from Sigma, St. Louis, MO). As liquid medium, BHI broth (Difco) supplemented with 7% horse serum, instead of blood, and, sometimes also, with 1% isovitalex (Becton Dickinson, Cockeysville, MD). 3. Urease, catalase, and oxidase test solutions for H. pylori identification: urease test solution, 10% urea in phosphate buffered saline, pH 7.2, and 1% phenol red; catalase test solution, 30% H202. Oxidase test solution, 1% N,N,N’,N’-tetramethyl-p-phenylenediamine dihydrochloride in water, or a commercial oxidase kit (Difco). 4. GTEL buffer: 50 mMglucose, 50 mMTris-HCl, pH 8.0,50 mMEDTA, 10 mg/mL lysozyme (Boehringer Mannheim, Indianapolis, IN). Store this solution without lysozyme at 4°C and add lysozyme from 10 mg/mL stock just before use. 5. TESK lysis buffer (solution B): 50 mM Tris-HCl, pH 8.0, 50 mM EDTA, 1% SDS, 50 pg/mL proteinase K, 20 pg/mL RNase. Store this solution without proteinase K and RNase at room temperature. Just before use, add proteinase K and RNase (both from Sigma) from 20 mg/mL stocks that were stored at -2O’C. 6. Tris-saturated phenol: Melt solid phenol crystals in a water bath at -80°C. Add an equal volume of 1M Tris-HCl, pH 8.0, and mix well to generate a white emulsion. Allow aqueous and organic phases to separate by standing for 20 min at room temperature, and discard the upper aqueous phase; the dissolved water increases the volume of the phenol phase by at least 30%. Add an equal volume of lMTris-HCl, pH 8.0, allow the phases to separate, and discard the upper aqueous phase. Add one-fifth vol of 100 mA4 Tris-HCl, pH 8.0, and store the solution

Berg et al.

120

7.

8. 9. 10 11. 12. 13. 14. 15 16

17. 18. 19 20. 2 1. 22. 23

24 25. 26. 27.

at 4°C m a dark bottle to prevent oxidation (detected by pink color). In general, Tris-saturated phenol can be stored for up to 2 wk before use. Pheno/chloroform/isoamyl alcohol. Mix 25 vol Tris-saturated phenol, 24 vol chloroform, and 1 vol isoamyl alcohol. This results in a two-phase emulsion with a thin aqueous layer overlying a uniform organic phase. Use the bottom organic phase. 70 and 95% ethanol. TE: 10 mA4 Tris-HCl, pH 8.0, 1 mMEDTA 10% Sodium dodecyl sulfate (SDS) 20 mg/mL Proteinase K (Boehmger Mannheim) (stored in small ahquots, -20°C) 5MNaCl 10% Cetyltrtmethyl-ammonium bromide (CTAB) (Sigma) m 5MNaCl. 24: 1 Chloroform/isoamyl alcohol Isopropanol AP-PCR Primers* 5’-AAGAGCCCGT 1247 5’-CCGCAGCCAA 1254 5’-AACGCGCAAC 1281 1283 5’-GCGATCCCCA 5’-GTGGATGCGA 1290 D14216 5’-NNNAACAGCTATGACCATG 5’-AGTGAATTCGCGGTGAGATGCCA D11344 D8635 5’-GAGCGGCCAAAGGGAGCAGAC D9355 5’-CCGGATCCGTGATGCGGTGCG D14307 5’-GGTTGGGTGAGAATTGCACG These primers are synthesized commercially (e g., by DNAgency, Malvern, PA). 10X PCR buffer. 100 miM Tris-HCl, pH 8.3, 500 mM KC1 MgCl, PCR stock solution. 25 mMMgC12 dCTP, dGTP, dATP, and dTTP (lOOmA stock solutions) (Boehringer Mannhelm) AmpliTaq DNA polymerase (Perkin-Elmer Cetus, Norwalk, CT). Perkin-Elmer TC480 thermal cycler Agarose: SeaKem LE agarose (FMC Bio Products, Rockland, ME). 50X stock Trts acetate runmng buffer (TAE) for agarose gels: 242 g Tris-Base, 57 mL glacial acetic acid, 100 mL 0 5M EDTA, pH 8.0 per 1 L Store at room temperature. Size marker: I-kb DNA ladder (Gibco-BRL, Gaithersburg, MD). Ethtdmm bromide. 10 mg/mL water stock solution. Use 6 pL/150 mL gel, Clarrthromycm stock solutrons: 2 mg/mL in methanol. Store at 4°C for up to 2 wk before use Metromdazole stock solutrons 2.4 mg/mL m distilled water.

3. Methods 3.7. Culture and Maintenance

of H. pylori

1. To preserve bacteria m fresh biopsies before culturmg them, prepare plastrc freezer tubes containing 0.5 mL TSB glycerol; while in the endoscopy surte, put

AP-PCR Fingerprinting

727

mdividual biopsies in different tubes and hold them temporarily at -20°C or 4°C; and move them to a -70°C freezer for long term storage as soon as possible. Culture H. pylon cells from these TSB glycerol-preserved biopsy specimens, as described in step 2, below. To store established H pylon strains, such as those isolated from biopsies, suspend cultures that have grown for several days on solid medium in TSB glycerol, m general at concentrations of at least log cells/ml, distribute 0 5 mL of suspension to each of a number of freezer tubes, and store at -70°C. 2 When culturing H pylon from frozen stocks, we normally grow the bacteria on solid medium, inittally, and then inoculate broth medmm with suspensions of the plate-grown bacteria. To begin, thaw the frozen stock quickly under tap water, pipet at least 10 pL onto the surface of agar medium m a Petrt plate, and streak or spread the cells, using a sterile loop. Incubate plates m a microaerobic atmosphere (5% O,, 10% CO,, and 85% N2) and 100% humidity, for 2-3 d for established stocks, or up to 6 d for initial culture of nonadapted cells from biopsies (strains vary m growth rate). Scrape the bacterial growth from the agar with a sterile loop, and suspend it in 5-10 mL broth m a 100 x 15 mm Petri plate. Shake the plates slowly m the incubator for 3-5 d to obtain stationary phase cultures To check for possible contammation by other, less fasttdtous, faster growing bacteria, spread a 50-pL ahquot on agar and incubate overnight

3.2. /den tifica tion H pylori are identified by their distinctive colony morphology and slow growth rate, and by positive urease, catalase, and oxidase tests and/or by the spiral shape of the bacteria with or without Gram stammg. The biochemical tests are performed on a glass slide or on a piece of filter paper. Pick one or a few colonies with a sterrle loop and suspend them in a drop of each test solution. A positive urease test 1sindicated by rapid change (within 1 mm) from orange to pink; a positive catalase test is Indicated by rapid bubbling of the H20, solution; and a positive oxidase test is indicated by rapid development of a dark blue color. If these reactions seem slow, it can be useful to examine the bacterial suspension microscopically, and look for the spiral-shaped bacterial cells that are characteristic of 23 pylori.

3.3. Selection of Antibiotic-Resistant Fractions of Mixed Resistant/Susceptible Bacterial Populations from Individual Biopsies Grow pools of bacteria m liquid medium to titers of about 2 x 1Ogcells/ml. Transfer 100 uL of each culture into a well in a sterile ELISA microtiter plate, and dilute each serially in IO-fold increments in phosphate buffered salme to densities of about 1 x lo4 cells/ml. Spot IO-pL aliquots of each dilution on solid medmm containing metromdazole (16 clg/mL) or claruhromycm (8 l&r&).

122

Berg et al.

Grow pooled bacteria from spots having at least 10 colonies for DNA extraction and fingerprinting of resistant fractions. 3.4. Standard Phenol DNA Preparation for AP-PCR Analysis (5) The following method was used originally for AP-PCR analysis of H pylon (5). We sometrmes use an alternative CTAB method (see Section 3.5.), especially when larger DNA fragments may be needed later for additional expertments, such as long distance (>I 0 kb) PCR. 1. Collect cells from 3-5 mL of stationary phase culture by centrifugation, and discard the supernatant. 2 Suspend the pellet m 200 Ilz, of GTEL buffer. Incubate for 30 mm at 37°C. 3. Add 500 pL of TESK lysls buffer and mix gently, shaking l&20 times by hand Avoid vortexing. Incubate at 55°C for 2 h or overnight. 4. Add 500 & of phenol/chloroform/isoamyl alcohol, vortex, and spin in a microcentrrfuge for 3 mm. Carefully collect the upper phase and place m a new microcentrifuge tube 5 Add 500 pL of chloroform, mix to get a white-colored emulsion, centrifuge for 3 min, and collect the upper phase. 6 Add one-tenth volume of 3M sodium acetate and mix, add 2 vol of 95% ethanol and mix, and incubate for at least l-2 h or overnight at -20°C. 7. Centrrfbge wrth tubes in a fixed orientation to facilnate detection of the DNA pellet. 8 Remove the supematant completely with Pasteur pipet-vacuum siphon. Wash with 70% ethanol. 9. Dry the DNA pellet briefly (3-5 mm) on the bench top, until it changes from white-colored to invisible. Avoid overdrying, because this makes the DNA difticult to resuspend. 10. Add 50 pL of water, and dissolve DNA thoroughly before use, by standing at 4’C overnight. 11. To assess DNA yield and mtegnty, mix 2 & of DNA solution with 8 pL Hz0 and 2 6 of gel-loading buffer (0.1% bromophenol blue, 50% glycerol), and electrophorese in a 0.8% agarose gel containing 0.5 pg/mL ethidmm-bromide in 1X Tris acetate running buffer Amounts of DNA above 20 ng are judged by the intensity of fluorescence relative to a h DNA marker of known concentration (New England Biolabs, Beverly, MA), and the intactness of the DNA is assessed by the compactness of the band.

Store the DNA at 4°C (for several months) or at -20°C. 3.5. Alternative CTAB DNA Preparation for A P-PCR Analysis (adapted from ref. 10) 1. Collect cells from 3-5 mL of stationary phase culture by centrifugation in a 15-mL disposable polypropylene tube, and resuspend pellet in 500 JJL TE buffer 2 Add 30 pL of 10% SDS and 3 & of 20 mg/mL proteinase K, and mix thoroughly. Incubate for 2 h at 55OC or overnight at room temperature.

AP- PC R Fingerprinting

123

3. Add 100 pL of 5MNaCl and mix thoroughly 4. Add 100 pL of CTAB/NaCl solution. Mix thoroughly but gently and incubate for 10 min at 65°C. 6. Add an approximately equal volume (approx 0.7 mL) of chloroform/isoamyl alcohol, mix thoroughly, and centrifuge to separate phases. 7. Transfer the aqueous, viscous supernatant with a wide-mouth pipet tip (to minimize DNA breakage) to a microcentrifuge tube, leaving the interface behind Add an equal volume of phenol/chloroform/isoamyl alcohol, extract thoroughly, and spin in a microcentrifuge. Repeat this phenol/chloroform/isoamyl alcohol extraction if a large white precipitate is present at the interface. 8. Transfer the supernatant to a fresh microcentrifuge tube, usmg a wide-mouth pipet tip. Add 0.6 vol uopropanol, and mix thoroughly but gently, until a strmgy, sticky, white DNA aggregate becomes visible. Catch the aggregate with the hooked end of a glass micropipet or with a plastic pipet tip. Alternatively, precipitate the DNA by brief centrifugation at room temperature. 9. Briefly immerse the hook or pipet tip contaming the DNA aggregate m 1 mL of 70% ethanol to remove residual CTAB, and then remove the excess ethanol by air drymg the DNA for a few minutes at room temperature. Avoid overdrying, because this makes the DNA difficult to resuspend. 10. Dissolve the pellet in 50 pL of H,O, to estimate DNA amount and intactness as outlined above (see Section 3 4.), and store until needed.

3.6. Af-P CR Fingerprinting

(5)

Dilute the DNA in drstllled water to a concentration of 10-20 ng/pL. Carry out AP-PCR in 25 pL+ containing 20 ng of H. pylori genomic DNA (see Notes 8 and 9), 3 mM MgC12, 20 pmol of primer, 1.0-l .2 U of AmphTaq DNA polymerase, 250 pA4 each of dCTP, dGTP, clATP, and dTTP in 10 mM Tris-HCl, pH

8.3,50 mMKC1, and 0.001% gelatin, under a drop of mineral oil (see Notes 4-6). Use the following cycling program for 10 nt primers: 40 cycles of 94”C, 1 min;

36°C 1 min; and 72OC,2 mm. Use the following cycling program for longer primers: 4 cyclesof 94”C, 5 min; 40°C, 5 min; and 72”C, 5 min (low stringency amphfication); 30 cyclesof 94°C 1 min; 55°C 1min; and 72°C 2 min (high stringency amplification); and then a final extension at 72°C for 10 min (see Note 7). After PCR amplification,

mix 8 l,tL of the product with 2 & of gel-loading

buffer (0.1% bromophenol blue, 50% glycerol) and load samples on a 2% agarose gel containing 0.5 mg/mL ethtdium bromide m 1X TAE buffer. Include a 1-kb DNA ladder (Gibco-BRL) or equivalent size standard in one or more lanes of the gel, and photograph under UV light. 4. Notes

4.1. Applications 1. H. pylon genome diversity: In initial studies of H. pylori using the AP-PCR method, each of 60 independent isolates from patients at one hospital was distin-

Berg et al. gutshed with Just a single arbitrarily chosen primer (5). This result 1s m accord with other DNA fingerprinting results indicating that H pylori is an extremely diverse species, and epidemtologic evidence that H pylon mfections tend to be commumty-acquired, not nosocomial. In contrast, other pathogens, such as Vzbrzo cholera 01 or E colz 0157:H7 tend to have highly clonal population structures, reflecting a tendency to proliferate in outbreaks or epidemics (1143); with V cholera and E colz, separate but epidemiologically related Isolates are often indtstmguishable from one another by AP-PCR, or by standard DNA fingerprmtmg methods (Berg, Gtlman, and Barrett, unpublished). 2 H pylon genome stability: Usually isolates recovered from mtttal and follow-up biopsies from U S and Western European residents have not been dtstingutshable by AP-PCR or other DNA fingerprinting methods (.5,7-9,14) This 1sconsistent with suggestions that many of the H. pylori infected persons in these soctettes carry Just one strain, and also that the genome of any individual strain tends to be quote stable, despite the great diversity among isolates from different persons. With a few patients, however, at least two markedly different tingerprmts have been obtamed from sequentral biopsy isolates, suggesting that these persons harbor more than one strain. Separate evidence that H. pyZorr genomes are quite stable emerged from the fingerprmting of a strain that initially had grown only weakly m gnotobtotrc ptglets and of a spontaneous mutant derivative of it that had been selected for more abundant growth by a dozen sequential passages in piglets. No differences were seen between the AP-PCR fingerprints of the starting strain and Its well-adapted derivative (6). Stmtlarly, no changes were seen m the AP-PCR fingerprints of a standard antibiotic-sensitive laboratory strain and anttbiotic-resrstant derivatives of it that had been generated by transformatton with DNAs from unrelated climcal isolates (Incecrk and Berg, unpublished). This matchmg of AP-PCR profiles fits with the idea that transformation usually involves relatively short DNA segments, and that any given array of AP-PCR fragments constitutes less than 1% of the H pylon genome (1700 kb), and, thus, would not be much affected by DNA sequence differences m small regions 3. Cocolomzation and the topology of bacterial growth. Tests of indivtdual smgle colony isolates have occasionally identified patients likely to carry more than one H. pylon strain (e.g., see ref. 7), but such tests can be quite laborious. Therefore, to more effictently detect at least some cases of cocolonization, we have begun fingerprintmg pools of bacteria from bropstes, rather than single colony isolates. Reproducible differences in AP-PCR profiles from different biopsy sites have been observed wtth a few of the patients tested to date, indicating that different strams predominate at those sites. Some of the bacteria are resistant to metronidazole or to clarithromycin, and comparisons of AP-PCR fingerprints of pooled resistant subpopulations and corresponding unselected populations have revealed additional instances m which two strains are present, one sensitive and one resistant to the clmically useful antibiottc (Lelwafa Guruge, Gilman, and Berg, unpublished).

A P- PCR Fingerprinting

125

Restriction analysis of specific PCR-amplified DNA fragments (PCR-RFLP, described m Chapter 11) can also yield evidence of cocolomzation and nonrandom stram distributrons (14); it has been very useful m studies of experimental mixed infection with two previously characterized strains (6) We prefer the AP-PCR method for studies of natural human infection, however, because it tests for sequence divergence throughout the genome. This would be important if distinguishable coresident strains sometimes derive from the same ancestor by mutation or interstram DNA transfer, because such events would leave most loci unchanged. Such pairs of strains would tend to be missed in PCR-RFLP tests focused on just one or a few loci.

4.2. Optimization Success in AP-PCR

and Reproducibility tests depends on having profiles that are rich in detail

(multiple bands) and also highly reproducible. Several reaction parameters have been found to markedly affect the patterns obtained, including levels of MgCl,, primer, and polymerase, and the quality of template DNA. 4. MgCl** No AP-PCR products were obtained with 1 mM MgClz, weak amphtication was obtained with 2 mA4MgClz, and reproducibly good arrays of AP-PCR fragments were obtained with 3-4 mM MgC12 (15). A report that focused on possible artifacts in AP-PCR tests (16) was based primarily on use of MgCl, concentrations that seem suboptimal for microbial genomes. It is noteworthy that most protocols for standard PCR (with primers that perfectly match particular template sequences) recommend lower MgC l2 levels (1.5 mM) than are optimal for AP-PCR; low Mg2+ concentrations suppress the type of misprimmg on which AP-PCR depends. The original description of the AP-PCR method used 2 mA4 with higher plant genomes (3) or 4 mi’~4 for bacterial genomes (Z). The apparent need for more MgCl* for good AP-PCR results from bacteria, as compared to plants, may reflect the much lower complexity of bacterial genomes, and a greater dependence on pairs of sites that are mcompletely matched to the primer sequence. 5. Primer concentration: The arrays of products obtained with 20-50 pmol of primer per 25 & of reaction were more informative than those obtained with 5 pmol or 100 pmol, with each of several primers tested (15). 6. DNA polymerase: In control experiments with E. colr K-12 strains, richer profiles were obtained with 1 O-l.2 U of AmphTaq DNA polymerase per 25-pL reaction than with 0.2 or 0.5 U, levels that are often sufficient for standard PCR with primers that perfectly match defined bacterial target sequences Use of lower levels of AmpliTaq for AP-PCR resulted in less amplification overall, and disproportionate effects on the yields of certain DNA fragments. Overall reproducibility also seemed to be lower when suboptimal levels of AmpltiTaq polymerase were used (I 7). Although we usually used AmpliTaq (which is produced in E. colr from the cloned gene; Perkin Elmer Cetus, Norwalk, CT), equally good profiles have also been obtained with Taq polymerase (directly from Thermus

Berg et al. aquaticus). Informative, although different, profiles were obtained with a DNA polymerase from T fravus (4), and with Stoffel fragment (18), an engineered derivative of the Tug polymerase. It should be noted that more products are generated, and the size distributron of products is smaller with Stoffel fragment than with Taq or AmpliTaq. Hence, Stoffel fragment may be better suited for new versions of the AP-PCR method that exploit dye labeled primers and polyacrylamide gels m automated sequencing instruments, and computer-assisted data acquisition and analysis (18, Applied Biosystems, Foster City, CA, personal communication) 7. Cycling parameters: A number of different annealmg temperatures and cycling times can be used, although the exact parameters can affect the profiles obtained. For example, m a study of E COZZstrains with 10 nucleotide (nt) primers, good profiles were obtained with annealing temperatures of 38’C instead of 36°C and an annealing time of 5, instead of 1, mm (19). In the case of longer (>I6 nt) arbitrary primers, the original protocol (I) used 40°C as the nonstrmgent annealing temperature for the first two cycles, and then a higher temperature (60°C) m subsequent cycles. Equally good arrays of products were generated using constant low stringency annealing temperatures of 40°C (2) or even 36’C (17) throughout the entire (4w5 cycle) reaction, The parameters cannot be varied without bound, however. For example, m studies of E. colr strains with 10 nt primers, less amplification and less informative profiles were obtained with annealing at 40°C or at 25 or 30°C, instead of at 36-38°C. The AP-PCR patterns were weaker after 25 or 30 cycles than after 40 cycles under our standard conditions, but extendmg the amplification from 40 to 45 cycles did not have much effect (Kersulyte, Brikun, and Berg, unpublished) Thus, overall, the flexibility in cycling parameters implies that useful results should be obtainable with any thermal cycler exhibiting good mterwell reproducibility. This is true even if rates of heating and coolmg, and actual mtrawell temperatures, vary among mstruments and complicate mterlaboratory comparisons (e.g., ref 21) 8. Reproducibility. Reproducibility is critically important, if differences m AP-PCR profiles from independent isolates are to be interpreted correctly. In general, the arrays of AP-PCR products obtained are highly reproducible, given well-optimized and uniform amplification conditions. Sporadic reports of lack of reproducibility are often attributable to suboptimal or nonconstant amplrtication conditions (parameters such as MgClz concentration, polymerase activity, DNA quantity, inhibitors, and inappropriate temperature during the reaction, reviewed in ref. 16). To test for reproducibility, we recommend parallel AP-PCR tests using different amounts of DNA from the same preparation (e.g., 5 and 20 ng) or DNA preparations from duplicate cultures. With H. pylon, as with most species, the arrays of fragments obtained using the protocols presented here are highly reproducible (see, e.g , refs 5%7,19,21-24). Results indicating an mhibitor in one of three parallel DNA preparations from a strain, evident only with the higher levels of DNA extract, are presented m Fig. 3A (prep. 2). Such inhibitors can usually be removed

AP-PCR Fingerprinting

127 Prep

mm-1

2

[DNA1 1 h I b

Isolate -- A

B

-- A

B

-- A

3

I II m

kb

B

Fig. 3. Reproducibility of AP-PCR DNA fingerprinting. (A) Effects of inhibitor remaining in DNA preparation on AP-PCR profiles. Four ng (1) and 20 ng (h) of DNA from each of three DNA preparations of the same H. pylori strain were fingerprinted with primer 128 1, as detailed in Methods. With preparation 2, the inefficient amplification of larger DNA fragments when 20 ng, but not when 4 ng, of template DNA was used, indicates a trace of inhibitor remaining in the DNA preparation. No amplification is obtained, at least at the higher DNA concentrations, when more inhibitor remains. The nature of this inhibitor has not been determined, but it can usually be removed by further washing (e.g., addition of 300 pL of distilled water, precipitation with 0.6 vol, and resuspension in 25 & distilled water). (B) Two H. pylori isolates recovered from the same infected individual can be closely related, but distinguishable. The identity of profiles of a given isolate obtained with 4 ng and with 20 ng of template DNA, the differences in profiles from the two isolates (A and B) with each of several primers (128 1, 1283, and 1290, respectively), and equivalent differences in profiles among these isolates seen with other preparations of their DNAs, indicate that these two isolates are indeed genetically distinct. That they are also closely related can be inferred by comparison of these profiles with those obtained from epidemiologitally unrelated strains (e.g., Fig. 2). by further washing (see Fig. 3). Given this ability to recognize when differences in profile are artifactual, and when they are not, pairs of strains that are closely related but truly divergent (Fig. 3B) can be identified with confidence. It should be noted that, in general, although profiles obtained in the same week are usually highly reproducible, the profiles obtained several months apart some-

Berg et al. times seem subtly different. Such variation might reflect differences among batches of thermostable DNA polymerase, or subtle changes in the activity of an enzyme preparation during storage. 9 Input DNA quality: Recent tests have shown that mformative and reproducible AP-PCR profiles can be obtained from E colz simply by boiling stationary phase broth cultures, diluting them lo- to 1OO-fold to avoid Inhibitors, and then using 5 pL of these diluted samples in 25-pL AP-PCR tests (25) A simple boilmg method has also been used for template preparation from H pylon (9).

4.3. Caiibra tion The AP-PCR method was calibrated. by using E. coli strains of known ancestry, and by comparmg AP-PCR test results with results from other frequently used methods. Although most such tests involved other bacterial species, the outcomes should also contribute to the evaluation of I-I pylori AP-PCR results. Several such calibration experiments are described below. 10. Cahbration using E. co11 strains of known ancestry. Two sets of E coli K- 12 strains were used to help calibrate the AP-PCR method: recently constructed tsogenic strains that differed by about 11% m genome size, because of differences m plasmid and chromosomal DNA content; and historic derivatives of the origmal E. coEzK-12 laboratory stram that were chosen to represent divergent branches of the K-12 pedigree These strains are known to have evolved m research laboratortes during the last half century solely by mutation and selecnon, without horizontal gene transfer (I 7) In studies of the strains differing in genome size, about 12% of bands obtamed from the strain with the largest genome were not obtained from its isogemc derivative with the smallest genome. Further tests showed that plasmids and deletable chromosomal segments each contributed to the AP-PCR profiles in approximate proportion to then sizes. Only a small number of polymorphisms for AP-PCR bands were found among historic strains from divergent branches of the E. colz K- 12 pedigree (5% of bands scored), despite half a century of subculturing, mutagenesis, and intense selection in laboratories throughout the world. This outcome implies that AP-PCR profiles of strains related by descent from a recent common ancestor are quite

constant.The few changesobserveddid not correlate with the mutationsthat had been specifically selected when each new strain was generated. Certain polymorphic bands that were not generated from a strain from one step in the pedigree reappeared in profiles from strains at subsequent steps m the same pedigree, further tests showed that several such polymorphrsms were causedby deletions, which could not posstbly undergo reversion. These results suggested a model m

which many of the changesleadmg to AP-PCR differences had accumulatedduring storage m room temperature stab cultures durmg the early years of E co11 genetics (use of ultracold freezers did not become commonplace until the 197Os), and mutagenic and selection pressures that differed markedly from those operatmg during more rapid growth (I 7). Although H. pylon strains are not stored in

the same way, there IS also hkely to be selection pressure for new mutations during adaptation of fresh clinical Isolates to laboratory media, or conversely, during infection of a fresh host for which the strain may not be well adapted, and such mutations might affect some AP-PCR profiles. 11. Comparison with MLEE typing* The multilocus enzyme electrophoresls (MLEE) typing method has been widely used by population geneticists and evolutromsts for several decades (25), and 1sbased on differences among strains in the electrophoretic mobility of each of a number of diagnostrc enzymes. Differences in mobility of a given enzyme among different isolates of a species is Interpreted to reflect differences in charge (amino acid sequence) of the protein, and thus drvergence of the underlying gene. To compare AP-PCR and MLEE data, a set of 75 independent isolates of diarrheagenic E. cok, which by MLEE typing represented 15 major clonal groups, were characterized by AP-PCR tests (19). Several distinct strains were drstmguished in most MLEE groups with each arbitrary primer tested (Strains wrthin any grven group were not distinguished by MLEE tests wtth any of 20 diagnosttc enzymes). Overall, these experiments indicated that AP-PCR typing IS consrderably more sensitive than the more laborious MLEE typing. 12. Comparrson with PFGE Pulsed field gel electrophoresis (PFGE) of large chromosomal DNA fragments, generated by a restriction endonuclease that cuts target genomes, infrequently has become a favorite tool for many research groups interested in epidemiology and genome evolution because it is highly discrrminatory (see ref. 27). However, PFGE analysis also requires costly equipment and reagents, long DNA preparation and electrophoresrs times, and a high level of expertise Alternatrvely, more efficient and lower cost typing methods of near-equivalent resolution are needed. Several recent studies indicated that AP-PCR typing is nearly as sensitive as PFGE typing (15,23,27), but a PFGE analysis of closely related but divergent H. pylon strains identified by AP-PCR would also be of great interest.

4.4. Perspective Just five years have passed since the first publication of the AP-PCR or RAPD method of fingerprinting microbial genomes (I-4). During this pertod, the method has become immensely useful for epidemiology and infection control, and for population and molecular genetrcs and evolution. There is reason to anticipate that AP-PCR will continue to be a valued method for studies ofH. pylori for years to come. Further improvements in H. pylori DNA preparation method, increases in reproducibility, and general standardization of protocols may make the method accessible to service as well as research laboratories. Given the complextty of AP-PCR profiles, in terms of band intensity as well as presence or absence of

bands, the wider application and further development of user friendly software, such as those marketed by Applied

Biosystems,

Applied Maths (KortriJk,

130

Berg et al.

Belgium), and Solltech (Iowa Ctty, IA) for recording Images and groupmg isolates according to common patterns, should greatly facilitate future analyses of H. pylori populations. Analysis of the genomic sequence dtfferences that underlie individual polymorphic bands should give new insights into genome evolution. In particular, when the differences in AP-PCR profiles between independent bacterial isolates are slight but reproducible, analyses of the corresponding genomic sequencesmay help us understand the relative contributions of interstrain DNA transfer vs mutation in H. pylori evolution (6,28,29). More generally, the anticipated technical advances in AP-PCR methodology and the processing of information gleaned from it should greatly enhance our understanding of how H. pylori is transmitted among sensitive hosts, and adapts and evolves during long-term chronic infection. Acknowledgments We thank R. H. Gilman and A. Dubois for stimulating discussions and for some of the H. pylorr-infected biopsy material characterized here. This work was supported by grants DK48029 and AI38 166 from the National Institutes of Health and VM-121 from the American Cancer Society to DEB. ET1 is the recipient of an Integriertes Auslandstudium stipend for exchange between the University of Regensburg and Washington University, funded by the Deutscher Akademischer Austauschdienst. References 1. Welsh, J. and McClelland, M (1990) Fingerprinting genomesusing PCR with arbitrary primers. Nucleic Acids Res l&72 13-72 18. 2 Welsh,J and McClelland, M. (1993) Characterizationof pathogenicmicroorganisms by genomm fingerprinting using arbitrarily primed PCR, m Dzagnostzc Molecular Mzcrobzology- Principles and Appllcatlons (Persing, D. H., Smith, T. F., Tenover, F C., and White, T. J., eds), American Society for Microbiology, Washington,DC, pp. 595-602. 3. Williams, J. G. K., Kubelik, A. R., Llvak, K. J., Rafalski, J. A., and Tingey, S.V (1990) DNA polymorphismsamplified by arbitrary primers are useful asgenetic markers.NuclezcAcids Res. 18,653 l-6535. 4. Williams, J. G. K , Hanafey, M. K., Rafalski, J. A., and Tingey, S. V. (1993) Genetic analysisusing random amplified polymorphic DNA markers. Methods Enzym. 218,704-740.

5. Akopyanz,N , Bukanov, N. O., Westblom,T. U , Kresovlch, S , andBerg, D E (1992) DNA diversity amongclinical isolatesof Hekobacter pylorl detectedby PCR-basedRAPD fingerprinting. Nucleic Acids Res 20,5 137-5142. 6 Akopyants, N. S., Eaton, K. A., and Berg, D. E. (1995) Adaptive mutation and co-colonizationduring Hellcobacterpylori mfection of gnotobiotic piglets. Infect Immun. 63, 116-12 1.

A P- PC I? Fingerprinting

131

7. Taylor, N. S., Fox, J. G., Akopyants, N. S , Berg, D. E., Thompson, N., Shames, B., Yan, L., Fontham, E , Janney, F., Hunter, F. M., and Correa, P. (1995) Long term colonization with single and multiple strains ofHelicobacterpylorr assessed by DNA fingerprinting. J. Clan. Microbtal 33, 9 1S-923. 8. Schutze, K., Hentschel, E., Dragosics, B., and Hirschl, A. M. (1995) Helzcobacter pylori reinfection with identical organisms: transmission by the patient’s spouses Gut 36,83 l-833. 9. Xia, H. X., Windle, H J., Marshall, D. G., Smyth, C. J , Keane, C. T , and O’Morain, C. A. (1995) Recrudescence of Helicobacter pylort infection after apparently successful eradication: application of randomly amplified polymorphic DNA fingerprinting. Gut 37,30-34 10 Ausubel, F M., Brent, R., Kingston, R. E., Moore, D D., Seidman, J G , Smith, J. A., and Struhl, K. (1994) Current Protocols in Molecular Biology, vol. I, Greene, Wiley, New York, unit 2.4. 11. Whittam, T. S., Wachsmuth, I. K., and Wilson, R. A. (1988) Genetic evidence of clonal descent of Escherichia coli 0157:H7 associated with hemorrhagic colitis and hemolytic uremic syndrome. .I. Infect. Dis. 157, 1124-l 133. 12. MacDonald, K L. and Osterholm, M. T. (1993) The emergence of Escherzchza coli 0157:H7 infection in the United States: the changing epidemiology of foodborne disease. J Am Med Assoc 269,2264-2266. 13. Karaolis, D. K. R , Lan, R., and Reeves, P. R. (1995) The sixth and seventh cholera pandemics are due to independent clones separately derived from envuonmental, non-01 Vzbrzo cholerae. J. Bactenol 177,3 19 l-3 198. 14. Akopyanz, N., Bukanov, N 0 , Westblom, T. U , and Berg, D. E (1992) PCR-based RFLP analysis of DNA sequence diversity m the gastric pathogen Heltcobacter pylori. Nucleic Acids Res 20,622 l-6225. 15 Berg, D. E., Akopyants, N., and Kersulyte, D. (1994) Fingerprinting microbial genomes using the RAPD or AP-PCR method. Methods Mel Cell Btol 5,13-24 16 Ellsworth, D. L., Rittenhouse, K. D., and Honeycutt, R. L. (1993) Artifactual variation in randomly amplified polymorphic DNA banding patterns. Biotechnzques 14,2 14-2 17.

17 Brikun, I., Suziedelis, K., and Berg, D. E. (1994) DNA sequence diversity among historic strains of Escherzchia coli K-12 detected by arbitrary primer PCR (random amplified polymorphic DNA) fingerprinting. J. Bacterial 176, 1673-1682. 18. Sobral, B. W. S. and Honeycutt, R. J. (1993) High throughput genetic mapping of polyploids using PCR-generated markers. Theoret Appl. Genet. 86, 105-l 12. 19. Wang, G., Whittam, T. S , Berg, C. M., and Berg, D. E (1993) RAPD (arbitrary primer) PCR is more sensitive than multilocus enzyme electrophoresis for dlstinguishmg related bacterial strains. Nucleic Acids Res 21,5930-5933 20. Van Belkum, A., Kluytmans, J., van Leeuwen, W., Bax, R., Quint, W., Peters, E , Fluit, A., Vandenbroucke-Grauhs, C., van den Brule, A., Koeleman, H., Melchers, W., Meis, J., Elaichouni, A., Vaneechoutte, M., Moonens, F., Maes, N , Struelens, M , Tenover, F., and Verbrugh, H. (1995) Multicenter evaluation of arbitrarily pnmed PCR for typing of Staphylococcus aureus strains. J Clan Mtcrobiol 33, 1537-1547

132

Berg et al.

21 Kersulyte, D., Woods, J. P., Keath, E J., Goldman, W. E., and Berg, D. E. (1992) Diversity among clinical isolates of Htstoplasma capsulatum detected by polymerase chain reaction with arbitrary primers. J, Bacterrol 174, 7075-7079. 22. Woods, J P., Kersulyte, D., Tolan, R. W., Berg, C. M., and Berg, D E. (1994) Use of arbitrarily primed polymerase chain reaction analysts to type disease and carrter strains of Nezsserra menmgltldis isolated during a umversity outbreak J Infect Du 169, 1384-1389. 23. Kersulyte, D , Struelens, M J., Deplano, A., and Berg, D E. (1995) Compartson of arbitrarily primed PCR and macrorestrtction (pulsed field gel electrophoresis) typing of Pseudomonasaerugznosastrains from cystic fibrosis patients. J Clin Microbrol 33,22 16-22 19. 24. Madico, G., Akopyants, N. S., and Berg, D. E. (1995) Arbitrarily primed PCR DNA fingerprmtmg of Escherzchza coli 0157:H7 strains using templates from boiled cultures. J Clm Microblol 33, 1534-1536. 25 Selander, R. K. and Musser, J. M. (1990) Populatron genetics of bacterial pathogenesis, m The Bacteria, vol XL (Iglewski, B. H. and Clark, V. L., eds.), Academic, New York, NY, pp 1 l-36. 26. Taylor, D. E., Eaton, M., Chang, N., and Salama, S. (1992) Constructton of a Hellcobacter pylon genome map and demonstration of diversity at the genome level. J. Bacterzol 174,2332-2337. 27. van Belkum, A. (1994) DNA fingerprinting of medically important mtcroorgantsms by use of PCR. Clin. Mlcrobiol. Rev. 7, 174-184. 28 Cohan, F. M. (1994) Genetic exchange and evolutionary drvergence m prokaryotes. Trends Ecol Evol. 9, 175-181. 29. Lenski, R. E. (1993) Assessing the genetic structure of mtcrobial populations. Proc Natl. Acad SCLUSA g&4334--4336.

15 Molecular Methods for Detecting Ulcerogenic

Strains of H. pyhi

John C. Atherton 1. Introduction Certain nonconserved genotypic and phenotypic characteristics of H. pylori are associated with increased risk of peptic ulceration in the human host. These characteristics can be divided into two groups: first, those relating to vacuolatmg cytotoxin activity (1,2), and differences in the gene encoding the cytotoxin, vacA (3); second, those relating to the cytotoxin-associated gene, cagA (4,5). About 50% of the US strains so far examined produce vacuolating cytotoxm activity in vitro, and patients harboring such strains are more likely to have past or present peptic ulcers than patients harboring noncytotoxlc strains (1,2,3). The gene encoding the cytotoxin, vacA, is present in virtually all strams. However, it has a mosaic structure unusual for a bacterial toxin, consistmg of regions that are relatively well-conserved among all strains and regions that are conserved only among a proportion of strains. The two best-characterized of these latter regions are the signal sequence, which has three allelic types (sla, slb, and s2), and the mid-region, which has two allelic types (ml and m2). vacA alleles with all possible combinatrons of signal sequence and mid-region types have been described, with the exceptton of s2/m 1. Both slgnal sequence and mid-region types are independent predictors of the in vitro cytotoxm activity of a strain. However, signal sequence type appears to be a better marker of a strain’s ulcerogenic potential than either mid-region type or cytotoxin phenotype (3). The second nonconserved characteristic of H. pylori shown to be a marker for increased ulcer risk is the presence of a cytotoxm-associated gene, cagA, and its high molecular weight product (4,5). cagA is lmked to a number of From Methods 111Molecular Medrcme, Hehcobacter pylon Protocols Edited by C L Clayton and H L T Mobley Humana Press Inc, Totowa,

733

NJ

134

A therton

neighboring genes including picA and B (permits induction of cytokines genes A and B). The presence or absence ofpicA and B are predicted with complete accuracy (to date) by the presence or absence of cagA. Like cagA, these genes have unknown functions, although in vitro studies suggest that they may be important for the enhanced host epithelial cell interleukm-8 response to

cagA-positive strains (6). Although cagA is distant from vacA on the chromosome, its presence is closely (but not invariably) associated with vacA types s1a and s1b. It 1snot yet clear whether these vacA subtypes or cagA are better markers of a strain’s ulcerogenic potential, but preliminary evidence favors the vacA subtypes: type sla strains are associated with a higher risk of duodenal ulcer in the host than are type s 1b strains, despite similar

associations

of these

vad types with cagA status (7). Several words of caution are necessary when working with vacA and cagA typing systems. First, the vacA work, in particular, is based on H. pylon isolates from US patrents, and preliminary

work with strains isolated from patients

from other countries shows that these may not fit neatly into the current vacA typing system. Second, m non-US populations, age of acquisition of infection may be different, and this, together with different host factors, may change the ulcer risk associated with infection by certain strain types. Finally, cagA and vacA genotypes have been described merely as being associated with

ulcer-disease m the host; the genotypic characteristics directly involved in disease pathogenesis remain to be identified. 2. Materials 2.1. Preparation

of Strains for Typing

1. Tryptlcase soy agar with 5% sheep blood (BBL, Cockeysvllle, MD). 2 Sulfite-free Brucella broth (8): 10 g tryptone, 10 g peptamm, 1 g glucose, 2 g yeast extract, 5 g sodium chloride, m 1 L deionized water. Autoclave. When cool, add 5% fetal bovine serum. 3. Disposable hydrogen and carbon dioxide generator envelope (CampyPak Plus, BBL) 4 Airtight gas jar 5 Carbon dioxide incubator set at 5.5% CO*, 100% relative humidity, 37°C.

2.2. Genomic DNA Extraction 1. Molecular biology grade dlstllled water. 2. TE buffer, pH 8.0: 10 MTris-HCl, pH 8.0, 1 MEDTA, pH 8.0. 3, GES reagent: Add guanadlnum thlocyanate, 60 g, OSM EDTA 20 mL, water 20 mL, heat at 65’C untii dissolved; add 5 mL 10% Sarkosyl; add water to 100 mL 4. 7.5M ammonium acetate 5. Chloroform (molecular biology grade) 6, Isopropanol(2-propanol). 7. 70% ethanol.

Detecting Ulcerogenic Strains

135

2.3. Polymerase Chain Reaction 1 PCR kit including 10X PCR buffer, dATP, dCTP, dGTP, dTTP, each at 10 mM, magnesium chloride at 50 mM, Tuq DNA polymerase at 5 U/pL. 2. Molecular biology grade distilled water. 3. Primers, each at 25 pA4. 4 Extracted DNA from H. pyZorz strains, including control strains. 5. 1 5- and 0.5-mL polypropylene tubes (the latter to fit thermal cycler) 6 Mineral oil 7. Thermal cycler. 8. Equipment and reagents for agarose gel electrophoresis.

2.4. Colony Hybrid&a tion 1. 2. 3 4. 5. 6. 7 8.

9

10. 11. 12

Nltrocellulose transfer and immobilization membranes. 10% SDS (sodium dodecyl sulfate). Denaturatlon solution: 0.5M sodium hydroxide plus 1.5M sodium chloride. Neutralization solution: 1.5M sodium chloride, 0.5h4 Tris-HCl, pH 7 4. 2X SSC: 0.3M sodium chloride, 0.03M sodium citrate. Ultraviolet crosslmker or vacuum oven. Heat-sealable polyester pouches. Prehybridization solution (made by adding 58 mL distilled water, 5 mL 100X Denhardt’s solution, 5 mL 10% SDS, 30 mL 20X SSC, 1.3 mL salmon sperm DNA at 10 mg/mL. 100X Denhardt’s solution is 2% Ficoll, 2% polyvmylpyrolidone, and 2% bovine serum albumin. 20X SSC 1s 3M sodium chloride and 0.3Msodmm citrate Salmon sperm DNA should be denatured at 95°C for 5 min and cooled on ice before use). Random primed DNA labeling kit (Boehringer Mannheim, Indianapolis, IN). This should include dCTP, dGTP, dTTP (each at 0.5 rmJ4); reaction mixture containing hexanucleotide mixture and 10X buffer; and Klenow enzyme at 2 U/& a3*P-dATP (spec. act, 650 Ci/mmol). 0 lXSSCplusO.l%SDS X-ray film, cassette with intensifying screens, and developer.

3. Methods

3. I. Strategy for Typing H. pylori 3.7.1. vacA Mid-Region Typing Mid-region typing is based on nucleotide differences m a 0.75-kb region of vacA. Two simple techniques are described that give identical results: PCR and DNA/DNA hybridization. The simplest form of hybridization, and the one described here, is colony hybridization, which avoids the need for pre-extraction of DNA. The positions of primers and probes used in vacA mid-region typing are shown schematically in Fig. 1, and primers are listed in Table 1. The probe pCTB4 corresponds to bp 2212-2650 in tax+ strain 60190 (9) and VA6 corre-

Atherton

136 vacAORF Slgnal sequence I

N-termmal

region

I

/

/

/

/ /-

/iid-region

from stram 60190

\

\

hybndlzatton

region

3 9 kb \

\

\

(type

VA3-F

A), V&3-R

pCTB4 mid-region

from stram 87203

PCR primers Probe for colony

I I\,

/I

PCR primers Probe for colony

C-termmal

mid-region I

(type

m2)

-

-

VA4-F

VA4-R

hybrldrzation VA6

Fig. 1. Strategy for vacA mid-region typing by PCR and hybridization Primer sequencesand size of ampltfied products are given m Table 1. Exact posmons and sizesof probes are given m the text

sponds to the same region of tax- strain 87-203, corresponding to bp 536-843 of the published partial vacA sequence (9). 3.1.2. vacA Signal Sequence Typing Signal sequence typing is based on nucleotide differences in the second half of the region encoding the vacA signal sequenceand m the region encoding the first few N-terminal ammo acids of the mature toxin The region involved is less than 50 bp, making DNA/DNA hybridization technically difficult, and so two separate PCR strategies have been used. The first strategy uses conserved primers and differentiates products on the basis of size: all strains give products, but type sla and slb vacA grve smaller products than type s2. The second strategy uses forward primers specrfic for each subtype (sla, s1b, and s2) and a common reverse primer; three reactions are performed for each strain, of which one is positive and the other two negative. The positions of primers used m these strategies are shown schematically in Ftg. 2 and primers are listed in Table 1. 3.1.3. cagA Typing cagA typing is more straightforward than vacA typing, because tt depends purely on the presence or absence of the gene. Both PCR and colony hybrrdization can be used; suitable PCR primers are listed m Table 1 and large PCR-generated probes from any part of the gene are suitable for use in colony hybridization (I 0).

Table 1 Oligonucleotide Gene and region amplified vacA nudregion

Primers

Used for vacA and cagA Typing

Genotype identified ml m2

2; u

vacA signal sequence

cagA

slls24 sla slb s2 cagA+

Primer designation VA3-F VA3-R VA4F VA4R VAl-F VAl-R SSl-Fb SS3-Fb SS2-Fb Fl” BIC

Primer sequence S’GGTCAAAATGCGGTCATGG3’ S’CCATTGGTACCTGTAGAC3’ SGGAGCCCCAGGAAACATTG3’ SCATAACTAGCGCCTTGCAC3 SATGGAAATACAACAAACACAC3’ S’CTGCTTGAATGCGCCAAAC3’ S’GTCAGCATCACACCGCAAC3’ S’AGCGCCATACCGCAAGAG3 SGCTAACACGCCAAATGATCC3’ S’GATAACAGGCAAGCTTTTGAGG3’ S’CTGCAAAAGATTGTTTGGCAGA3

%acA types sl and s2 are differentiated on the basis of differences m size of the PCR product %sed with reverse pnmer VAl-R These prnners are taken from ref 10.

Size of PCR product, bp 290 352 259/286” 190 187 199 349

Atherton

138 &nal

sequence

1 sequence re_einn VAl-F

ssl-F

w

ss3-F

I I

Mature orotern rem VAl-R

sla

sib

Ir-1171111 VAl-F

s2

i VAl-R

I 7-Y

ss2-F

--=

VAI-R

--___

Fig. 2. Strategy for PCR typing of vacA signal sequence regions. Primers VAl-F and VA1 -R give a product for all strains, but the product from type s 1a and s 1b strains IS smaller than that from type s2 strains. Primers s&F, ss3-F, and ss2-F, each used with primer VA 1-R, grve products of the approprrate size only from types s 1a, s 1b, and s2 strains, respectively. Shaded areas show posmons of nucleotide differences to type sla; hortzontal arrows indicate PCR prtmers; vertical tilled arrows indicate experimentally determined signal sequence cleavage site; vertrcal hollow arrow indrcates presumed signal sequence cleavage site

3.2. Preparation of Strains for Typing Both plate- and broth-cultured bacteria are suitable for strain typing by PCR or colony hybridization. 1 Plate cultures should be seeded from single colonies to avoid starting with more than one strain (see Notes l-3) One plate is enough for genomtc DNA extraction for PCR or for preparing membranes for colony hybridization. 2 Broth cultures should be seeded from a plate passed from a single colony. Though commercral Brucella broth supplemented with FBS is often satisfactory, growth IS more consistent and faster m sulfite-free Brucella broth (8). A 48-h culture m 20 mL broth yields sufficient bacteria for typing by PCR or colony hybridrzatton

3.3. Typing by Polymerase Chain Reaction As with all PCR-based techniques, strmgent precautions should be taken at all stagesto avoid possible contammatron, both with extrinsic DNA (especially that derived from previous PCRs) and with PCR inhibitors (II). For vacA typmg of either mid-region or signal sequence, when specific rather than conserved primers are used, contaminatton or inhibition can be deduced, as more than one reaction is performed. For example, a vacA mid-region type ml strain IS expected to yield a positive reaction with ml primers, and a negative reaction with m2 primers; positive reactions with both indicate contamination. For cagA typing, on the other hand, there is no check, other than the controls, for

139

Detecting L&erogenic Strains

falsely positive reactions caused by contamination. Thus for cagA typing colony hybrtdization 1s the preferred method. Two quick methods of chromosomal DNA extraction for PCR are described: a simple boiling method and a modified GES method. The latter takes a little longer, but yields cleaner preparations, which give more conststent results on freeze/thawing, and the resultant DNA can also be used for purposes other than PCR.

3.3.1. Boiling Method of Chromosomal DNA Extraction 1. 2. 3 4. 5. 6.

Prepare 48-h plate culture of H. pylori. Suspend bacterial cells from one-fourth of a plate in 500 pL sterile distilled water. Boil for 10 mm. Centrrfuge at top speed in a microcentrifuge for 10 min Remove supernatant to a fresh tube. Use immediately, or store at -20°C.

3.3.2. Modified GES Method of Chromosomal DNA Extraction This method microcentrifuge 1 2 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

is relatively tubes.

quick as it is performed

throughout

m 1.5-mL

Prepare 48-h plate culture of H. pylon. Suspend bacterial cells from one-fourth of a plate in 100 pL TE buffer. Add 500 $ GES reagent, and mix by inversion. Incubate at room temperature for 5 min. Add 250 & 7.5M ammonium acetate, precooled on ice, mix by inversion, and incubate on ice for 5 min. Add 850 pL chloroform and mix well by inversion. Centrifuge at top speed in a microcentrifuge for 5 min, then carefully remove the aqueous layer (see Note 4). Repeat steps 6-8. Add 0.54 vol of isopropanol and mix well by inversion. Centrifuge at top speed in a microcentrifuge for 2 min, then remove supematant. A small pellet may or may not be visible. Wash twice with 70% ethanol. Remove all traces of ethanol and air dry for 10 min. Dissolve in 100 & TE buffer, pH 8.0 (see Note 5). Use immediately or store at -20°C.

3.3.3. Polymerase Chain Reaction 1. PrelabelO.5~pL tubes (for thermal cycler). This should include tubes for samples and appropriate positive and negative controls (see Note 6). 2. Prepare master mix on ice (see Note 7). The reagent concentrations shown m Table 2 are effective for all the primers listed in Table 1.

A therton

140 Table 2 PCR Constituents Constituent Water 10X PCR buffer Mix of 4 nucleotides, 2 5 mM each Prtmer 1, 25 w Primer 2, 25 pA4 MgCl*, 50 mA4 Tuq DNA polymerase, 5 U/mL Total Template DNA

Volume per tube@. 36 5 4 1 1 1.5 0.5 49

Final concentratron 1x 0.2 mA4 each 0.5 p44 0.5 pM 1.5 mM 2 5 u/50 pL

1

3 Add 49 & of master mix to each reactton tube (on ice). 4. Add 1 pL of template DNA solution to each tube and 1 $ of water or TE to the negative control tube (see Note 8). 5. Overlay with mmeral or1 (see Note 9). 6 Centrifuge briefly to remove any bubbles. 7. Put tubes in thermal cycler. The following cycles work well, but cycle length could undoubtedly be shortened without compromtsing effictency. 94°C 1 min 35 cycles 52°C 1 min (55°C for cugA primers) 72°C 1 min (2 min for cagA) > 6 mm 72°C soak 4°C 8 Visualize products on a 2% agarose gel with ethidmm bromide, using suitable DNA size markers

3.4. Typing by Colony Hytwdizafion Colony hybridization, in contrast to other DNA/DNA hybridizatron techniques, avoids the need for pre-extraction of chromosomal DNA. This makes it quick and simple, and means that it can be used to check results obtained usmg PCR without using the same DNA sample or re-extractmg DNA. Other DNA/ DNA hybridrzatron techniques are equally suitable. 3.4.1. Preparation of Membrane 1 SpotH pylori cells onto anitrocellulose membraneusing agradpattern to record positron of cells from each stram. Cells should form a clearly visible spot of 3-5 mm. Several methods work. a. Transfer directly from a plate culture using a cotton swab (be careful not to disrupt the mtrocellulose membrane by vigorous transfer).

Detecting &erogenic

Strains

141

b. Make a dense cell suspension in saline from plate or broth cultured cells and spot 5 pL onto membrane (this method IS more reproducible and gives more uniform loading). 2. Soak each membrane sequenttally in the following solutions. Between soaks remove excess solution by placing membrane on a paper towel. Soak membrane from the side without cells. a. 10% SDS for 3 min. b. Denaturation solution for 5 mm. c. Neutralization solution for 5 min. d. 2X SSC for 1 min. 3. Bmd DNA to nitrocellulose by ultraviolet cross-linking or baking m a vacuum oven. 4. Use membrane immedtately, or store at room temperature until use.

3.4.2. Hybridization Hybridization using 32P-labeled probes is described. Using nonradioactive probes should be equally effective. 1 Put nitrocellulose filter m a heat-sealable polyester pouch. 2. Preheat prehybridization solution to 68’C and add 4 mL per filter to bag. Seal without bubbles and incubate at 68°C for 90 mm. 3 Label probe, The method described is random-primed DNA labeling with 32P-dATP (11,12), but other methods could also be used. a. Denature 100 ng DNA in 10 pL water by heating to 95°C for 5 mm and cooling on ice. b. Add 1 pL of dGTP, dCTP, dTTP, and 2 p.L reaction mixture containing hexanucleotides and 10X reaction buffer c. Add 5 pL 32P-dATP (see Note 11). d. Add 1 pL Klenow enzyme. e. Incubate at 37°C for 30 min. f. Heat to 95’C for 10 min and cool on ice 4. Purify probe (see Note 11) 5. Add purified probe to bag (add volume containing IO-20 mg labeled probe per membrane). 6. Incubate at 68“C for at least 16 h. 7. Wash twice for 30 min with 0.1X SSC plus 0.1% SDS at 68°C (high stringency). Considerably shorter washes can be used tf the probe is purified 8. Visualize hybridization by exposure to X-ray film.

4. Notes 1. Strains should be passaged as few times as possible before typing. This is because genomic stability of H. pylori, during m vitro passage, has not been formally assessed 2. When growing H. pylon from smgle colonies or freezer stock, more consistent and luxuriant growth is obtained by growing under microaerobic conditions (gen-

Atherton

3

4

5. 6.

7.

8

9. 10. 11

erated using a CampyPak Plus m a gas Jar) than by growing under aerobic condttions in a carbon dioxide incubator. Because of sparse growth when subculturmg from a single colony, it may be necessary to perform a subsequent subculture of multiple colonies to get enough bacteria to set up a series of membranes for colony hybrtdizatton. One way to remove the aqueous layer, without contaminatton with cell debris at the aqueous/chloroform interface or with chloroform, IS to use pipet tips that have been cut off to Increase the bore (e.g , blue tips with 5 mm cut off). If tips are cut, tt is vital to use sctssors and gloves never exposed to H pylorz This may have to be diluted tenfold before PCR, depending on DNA concentratton Minimum controls for each primer pan should include a positive control, a DNA negative control, and a no-DNA negattve control. For example, for VUCAtype ml primers, controls should include DNA from a known ml strain, a known m2 sham, and a distdled water or TE buffer control prepared in parallel with the DNA Depending on the number of strains to be tested, tt is often convenient to prepare two or even three master mixes at once and then test strains with two or three primer pairs in parallel PCR can accommodate a wide range of template DNA concentrations and tolerate many contaminants, so tt 1susually unnecessary to check DNA concentratton before PCR If the PCR is not working, the problem is usually too much rather than too little template That sufficient DNA template has been obtained can be checked, at least following extraction by the GES method, by visualtzation on a 0.7% agarose gel stained with ethtdium bromide. Be careful not to contaminate mineral oil with DNA or crosscontaminate tubes by splashing (ideally each tube should be closed before the next IS opened). Mineral 011is the most common source of contammatton Full radiation precautions should be taken when using 32P. Purification can be omitted, but greatly improves the quality of results One method is to add the probe mix to 0.3 mL of a loading mix of dextran blue 2 mg/mL in water with 30% glycerol, run through a disposable chromatography column (e g., EconoPak lODG, Bio-Rad, Hercules, CA), and collect the blue fraction

Acknowledgment John Atherton is funded by a Clmtcian cal Research Council.

Scienttst Fellowshtp

from the Medi-

References 1. Figura, N , Gughelmettr, P , Rossohm, A , Barberi, A , Cuss, G , Musmanno, R A , Russi, M , and Quaranta, S (1989) Cytotoxm production by Campylobacter pylorz strains isolated from patients with peptic ulcers and from patients with chronic gastritis only. J Clw Microblol. 27, 225-226 2 Rautelin, H., Blomberg, B., Jarnerot, G , and Damelsson, D (1994) Nonopsomc activation of neutrophtls and cytotoxin production by Helicobacter pylonulcerogemc markers. Sand J Gastroenterol. 29, 128-132.

Detecting

Ulcerogenic

Strains

143

3. Atherton, J C , Cao, P , Peek, R. M., Jr., Tummuru, M. K. R., Blaser, M J., and Cover, T. L. (1995) Mosaicism in vacuolating cytotoxin alleles of Helicobacter pylori: association of specific vacA types with cytotoxin production and peptic ulceration. J. Biol Chem 270, 17,77 l-l 7,777 4 Cover, T L., Dooley, C P., and Blaser, M J. (1990) Characterization of and human serologic response to proteins in Hehcobacter pylori broth culture supernatants with vacuolizmg cytotoxin activity Infect. Immun 58,603-6 10 5 Crabtree, J. E , Taylor, J. D., Wyatt, J. I., Heatley, R V., Shallcross, T M , Tompkins, D. S., and Rathbone, B. J. (1991) Mucosal IgA recogmtion of Hekobacter pylorr 120 kDa protem, peptic ulceration, and gastric pathology. Lancet 338,332-335

6. Tummuru, M. K. R., Sharma, S. A., and Blaser, M. J. (1995) Helzcobacterpylorl picB, a homolog of the Bordtella pertussts toxin secretion protein, 1srequired for induction of IL-8 in gastric epithelial cells. A402 Mzcrobiol. 18,867-876 7. Atherton, J. C., Peek, R. M., Tham, K. T., Blaser, M. J., and Cover, T L. (1996) Clinical and pathological importance of heterogeneity in vacA. The vacuolating cytotoxin gene of Hehcobacterpylori. Gastroenterol, in press. 8. Hawrylik, S. J., Wasilko, D. J., Haskell, S. L., Gootz, T. D., and Lee, S E. (1994) Bisulfite or sulfite inhlblts growth of Helzcobacter pylorl. J. Clwz. Mcrobiol. 32, 790-792

9. Cover, T. L., Tummuru, M. K. R., Cao, P., Thompson, S. A., and Blaser, M J. (1994) Divergence of genetic sequences for the vacuolating cytotoxm among Helicobacterpylorr strains. J Biol. Chem. 269, 10,566-10,573 10. Tummuru, M. K. R., Cover, T. L., and Blaser, M. J. (1993) Cloning and expression of a high molecular weight major antigen of Helicobacterpylon: evidence of linkage to cytotoxin production. Infect Immun 61, 1799-1809. 11. Kwok, S. (1990) Procedures to mmimlze PCR-product carry-over, and Orrego, C. (1990) Organizmg a laboratory for PCR work, in PCR Protocols. A Gurde to Methods and Applications (Innis, M. A., Gelfand, D. H., Sninsky, J. J , and White, T. J , eds.), Academic, San Diego, CA, pp. 142-145 and 447-454. 12. Feinberg, A. P. and Vogelstein, B (1983) A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity Anal Biochem. 132, 6-13. 13 Feinberg, A P. and Vogelstein, B. (1984) A technique for radlolabeling DNA restriction endonuclease fragments to high specific activity. Addendum. Anal. Biochem

137, 266,267.

H. pylori DNA Transformation by Natural Competence and Electroporation Zhongming Ge and Diane E. Taylor 1.

Introduction

Helicobacterpylori is an important etiological pathogen of human stomach diseases, such as gastritis, peptic ulcer, and gastric carcinoma (1). In the past few years, great progress has been made in the cloning and characterization of H. pylori genes. Successof these studies stems in part from the finding that chromosomal and recombinant plasmid DNA are able to be efficiently transformed into H pylori cells by natural competence (24, and electroporation ($5). Such techniques allow the transfer of cloned H. pylon genes, manipulated in vitro, which can then shed light on the structural and functional relationships of the genes of interest. In this chapter, we describe the protocols for the isolation of H. pylori chromosomal and plasmid DNA, natural transformation, and electroporation. Natural transformation of bacteria involves the uptake of DNA molecules from the environment by a cell, and the heritable integration of the mtroduced DNA into its chromosome. During the course of evolutton, bacteria have developed different mechanisms to accomplish this process; these are beyond the scope of this chapter and have been reviewed by Lorenz and Wackernagel (6). There are a limited number of bacteria for which natural transformation has been demonstrated; transformation efficiencies vary significantly (6). Natural transformation of plasmid DNA is rare or occurs at much lower frequencies than those of bacterial chromosomal DNA (7). However, the frequencies of plasmid transformation can be increased by using a homologous plasmid that is resident m the recipient cells (3,8,9). H. pylori is a Gramnegative bacterium, and its natural transformation was first described by From’ Methods m Molecular Medicme, Hellcobacter pylor~ Protocols Edlted by C L Clayton and H L T Mobley Humana Press Inc, Totowa,

145

NJ

146

Ge and Taylor

Nedenskov-Sorensen et al. (2) and studied m detail by Wang et al. (3). Using rifampicin, streptomycin, and metronidazole resistance as genetic markers, Wang et al. (3) demonstrated that almost all H. pylori strains are naturally competent and transformable, A plasmid isolated from II. pylon was also capable of being naturally transformed into the plasmid-harboring strains at a fairly efficient rate, although the plasmid-free strains took up the plasmids poorly (3). Several mutant H pylon genes, such asJlaA, gvrA, and copA carried on plasmids, have been introduced into their parental cells by means of natural tranformation (4, IO, I I). Electroporation (electrotransformation) provides an alternative for the introduction of bacterial genes of interest into recipient cells (12-14). This technique has been widely applied m various organisms, including bacteria, animal, plant, fungal, and protozoan cells (15). The basic principle of electroporation is that a brief high-voltage pulse is used to increase the transient membrane permeabrlity of the recipient cells, which can then take up DNA molecules more efficiently. Efficiencies of electrotransformatron vary between bacterial strains and species. In the case of H. pylori, electrotransformation is capable of giving efficiencies m the range of approx 1.0 transformantjpg to 2.0 x lo3 transformants/clg plasmrd DNA, depending on the H. pylori strains used (3). In some circumstances, it appears to be more efficient than natural transformation. In one such example, the plasmid pUOA26 was able to be introduced into plasmid-free I?, pylori strains by electroporation, but not by natural tranformation (3). Successful delivery of the target genes mto H. pylori cells by this method has been achieved (.5,16,17), making this valuable technique available for genetic studies of H. pylori. 2. Materials All chemicals are purchased from BDH (Alberta Canada) unless specified. 1. H pylori strains for transformation* Almost all H pylon strains are naturally competentandtransformable (24). 2 Microaerobic conditions:a supplyof 5% (v/v) CO,, 5% (v/v) H, andbalancedN2 3. H pylori chromosomalandplasmid DNA. 4 Isolation buffer* O.l5MNaCl,O lMEDTA, pH 8 0 5. TE: 10mA4Tris-HCl, pH 8 0, and 1 mMEDTA. 6 TES: 10ml4 Tris-HCl, pH 8.0, 1 mMEDTA, 100rmZPNaC1. 7. Solution I 40 mMTris-HCl, pH 8.0,2.5 mM EDTA, 15% (w/v) sucrose. 8. Solution II: 0.2MNaOH, 1% SDS.Preparefreshly prior to use. 9 RNasedigestion buffer: 50 mMTris-HCl, pH 8.0, 150mMNaC1. 10 Brain heart infusion (BHI) (Umpath, Hampshire, UK)-YE media 37 g/L BHI, 5 g/L yeast extract (Difco, Detroit, MI), 50 mL/L horse serum, 10 mg/L vancomycm (Sigma-Aldxh

Canada, Ontano, Canada), 10 mg/L amphotericm B

(Sigma), and, for the preparationof agarplates,16g/L agar.Generally,BHI, yeast

147

DNA Transformation

extract, and agar are prepared separately and autoclaved. Antibtottcs amphotericin B and vancomycin are filter-sterilized. Serum and antibiotics are added Just prior to inoculation or preparation of agar plates 11. Sodium acetate solution: 3M sodium acetate, pH 4.8. The solution should be stored at -20°C until used 12. Glycerol solution: 10% glycerol sterthzed by autoclaving. The solution should be ice-cold until used. 13. Isopropanol: 100% (v/v).

3. Methods 3.1. H. pylori Chromosomal

DNA lsolafion

The procedures for the isolation of H. pylon chromosomal DNA vary, based on personal preference. The protocol described below, which is modified from the method of Ezaki et al. (IB), is simple, rapid, and efficient. The DNA preparation can be subjected to restriction endonuclease digestion, polymerase chain reaction (PCR), hybridization, natural transformation, and electroporation (see Note 1). 1 Inoculate a plate culture of 2-3 d either into 100 mL BHI-YE broth in a 500-mL flask or onto four BHI-YE agar plates. Keep the plates or the flask in anaerobic jars containing 5% (v/v) COZ, 5% (v/v) HZ, 90% (v/v) Nz. Grow the bacterium at 37°C for 2-3 d. For broth culture, moderate shaking is required. 2. Scrape off cells from plate cultures and suspend them in 2 mL of isolation buffer, Harvest cells from broth culture by centrifugation at 302Og in a Beckman JA-20 rotor (Fullerton, CA) and resuspend in 2 mL of isolation buffer. 3. Pellet cells at 4°C at 302Og for 10 min. 4. Remove as much of the supernatant as possible Resuspend the cells in 2 mL isolation buffer. 5 Add 200 pL 20% (w/v) SDS and 100 pL RNase (10 mg/mL). 6. Incubate cells at 65°C for 5 min. The lysate should become clear and viscous, 7. Extract the lysate once with 2.5 mL of phenol*chloroform.rso-amyl alcohol (25:24: 1). Separate the aqueous phase from the organic phase by centrifugatlon m the Beckman JA-20 rotor at room temperature at 302Og for 5 min 8. Transfer the aqueous phase (-2 mL) into a new tube and wash the organic layer once with 1 mL of isolation buffer. 9. Combine the original and the washing solutions (-3 mL). Extract once wrth 3 mL of chloroform. 10. Centrifuge as above and collect the aqueous phase. 11. To the aqueous phase, add 7.5 mL of cold 95% ethanol and incubate at room temperature for 10 min (see Note 2). 12. Pellet DNA by centrifugation at 18,500g at 4’C in a Beckman JA-20.1 rotor. 13. The pellet 1s washed twice with 1 mL of cold 70% ethanol. 14. Dry and redissolve DNA pellet m 1 mL TE. The chromosomal DNA should be stored at 4°C or -20°C.

148

Ge and Taylor

3.2. Preparation of H. pylori PIasmids Most procedures for the preparation of bacterial plasmids are derivatives of the method described by Birnboim and Doly (19), although modtficatrons are often made to optimize conditrons for respective plasmrds. The protocol described below IS adapted from the method of Birnbotm and Doly wtth a few modificatrons (see Note 3). 1 Grow H pylon cells as describedin Section3.1 (step 1) 2. Collect cells by centrtfugation at 4°C at 4360g for 10 mm in a Beckman JA-20 rotor, followed by washing once with 2 mL TES buffer. 3. Resuspend cells m 2 mL ice-cold Solution I and place the sample on ice for 5-10 mm. 4. Add 3 mL Solution II into ice-cold cells and mix well with inversion several times (no vortexmg). Place cells on ice for 10 min. 5. Add 3 mL of sodium acetate (pH 4.8) mto the lysed cells and keep on me for 5-10 mm. 6. Centrifuge the cells at 4°C at 17,400g for 10 mm m a Beckman JA-20 rotor. 7 Transfer the supematant (about 7 mL) into a new tube and discard the pellet. 8 Add 4.2 mL of cold isopropanol (kept at -20°C) mto the tube and mcubate the mixture on ice for 10 min. 9 Spin down DNA at 4°C at 17,400g for 15 min. 10 Discard supematant and dissolve the pellet in 2 mL of RNase digestion buffer Add 50 PI., of DNase-free RNase (10 mg/mL) and incubate for 60 min at 37’C. 11 Extract once with 2 mL pheno1:chlorofor-m (1 *l), followed by centrifugation at room temperature at 12,900g for 5 min 12. Collect supematant and mix with 2 5-fold of cold 95% ethanol. Precipitation is performed at -80°C for 10 min or 60 min at -20°C. 13. Pellet DNA by centrtfugation at 4°C at 18,500g for 15 mm. 14 The pellet 1swashed twice with 1 mL of cold 70% ethanol. Dry and dissolve the DNA m 500 & of TE.

3.3. Natural Transformation Efficiencies of natural transformation vary greatly between H. pylori strains (3,4). Factors that could affect the transformatton efficiency are discussed in Note 4. The transformation can be performed either on an agar surface or in broth. Based on our experience, the agar surface gives higher transformatron efkencies and IS simpler than the latter. 3.3.7. On Agar Surface 1. Inoculate 100 pL of recipient H. pylon cells from its stock at -80°C onto a cold BHI-YE agar plate (see Note 5) and incubate for 2 d under the same conditions as described m Section 3.1.) step 1. 2. Scrape off a heavy loop of cells from the above plate and spread H pylon cells on a cold plate in a diameter of 8-10 mm. Continue incubation for 5 h.

DNA Transformation

149

3 Spot 10-20 pL of DNA (0.5-2 pg) in TE directly onto the inoculated area (bacterial lawn) (see Note 6). 4. Incubate the plate for 16-24 h under microaerobic conditions 5. Streak the transformed cells onto BHI-YE plates containing selecttve anttbtotics. 6. Grow the tranformants for 3-4 d.

3.3.2. In BHI-YE Broth The protocol described below is modified from Nedenskov-Sorensen et al. (2) and has been demonstrated to be reliable in our laboratory. In addttion, an alternative of this protocol can also be found in ref. 4. 1. One hundred microliters of an H pylon stock at -80°C are inoculated mto 2 mL BHI-YE broth and mcubated for 48 h at 36°C wtth vigorous shaking under microaerobic conditions. 2. DNA (10-20 pg) is added to a broth culture of the reciptent cells and grown for 18-20 h to a density of 1O8cm/ml under microaerobic conditions. 3. The mixture is incubated in a 5% CO2 atmosphere at 36°C for 20 min 4. The transformed cells (200 pL) are spread onto BHI-YE plates containing selective antibiotics

3.4. Electrotransformation Efficiencies of electrotransformation vary among bacterial species, different strains, and even the different preparations of the same strain (15). With several E. coli strains including LE329, DHSa, MC 106 1, and WM 1100, Dower et al. (14) obtained 10g-lO1o transformants/ltg of plasmid DNA using electroporation. However, the efficiencies of electrotransformation with the H. pylon genes appear to be lower than those for E. coli strains. The protocol described below, adapted from Wang et al. (3), is capable of giving lo4 transformants& plasmid DNA, when the strain used as the recipient cell contains a homologous plasmid. 3.4.1. Preparation of Recipient H. pylori Cells 1. Inoculate 100 & of H pylori stock culture at -80°C on BHI-YE plates.

2. After 18-24 h incubation,cells are scrapedoff the plate and suspendedin 30 mL cold double-disttlled water 3. Harvest cells by centrifugation at 4360g at 4°C for 5 mm and the pellet 1s resuspended in 20 mL of cold 10% glycerol. 4. Spm down the cell as before and resuspend the pellet m 2 mL ice-cold 10% glycerol. The cells are aliquoted as 0.2 mL vol and are then ready for electroporatton. For long-term storage, competent cells in 10% glycerol are frozen m dry ice and

stored at -80°C until needed.Frozencells can be thawedon ice andused rmmedtately for electroporation.

Ge and Taylor 3.42. Electroporation

of H. pylori Cells

1 Freshly prepared cells on ice can be used immediately; frozen cells are thawed on ice for 15 mitt Plasmid DNA (0.5-2 clg in 5-l 0 clr, TE buffer) is added to a microcentrifnge tube on ice and mixed with 0.2 mL cell suspension The sample 1schrlled on ice for l-2 min. 2. The Bio-Rad gene pulser is set up at 2 5 kV (12 5 kV/cm) with a 25-l,rF capacitor and a resistance of 600 ohm in parallel (see Note 7). 3. The bactenal/plasmid DNA mixture is added slowly to a prechilled (kept at -20°C) electroporation cuvet (0.2-cm gap) (Bio-Rad, Hercules, CA). Bubbles that form m the sample within the cuvet are removed and the sample is then evenly spread on the bottom of the chamber of the cuvet 4. The sample is subjected to a smgle-pulse electroporatton with a time constant of 12.5 ms. 5. After the pulse, an aliquot of 10 pL of the sample 1sunmedtately diluted in 90 pL of ice-cold BHI-YE broth and plated onto BHI-YE agar to determine the number of cells that survived electroporation 6. The remammg suspension IS transferred onto a cold plate and incubated for 12 h at 36°C. The cells are then inoculated onto selecttve media, followed by mcubatton for 5 d to allow the growth of transformants under microaerobic conditions.

4. Notes 1. DNA prepared using this method is not suitable for the construction of physical and genetic maps of H. pylori, because the DNA 1srandomly sheared during the rsolation process, Readers can find protocols for the isolation of complete genomtc DNA m Chapter 18 of this volume. 2. At this step, precipitated DNA can be spooled out with a mtcroptpet tip and redissolved in 500 pL of TE. The DNA isolated m this way is purer, since the contamination of SDS, salt, proteins, and organic reagents in the preparation is minimized. 3. Both broth and plate cultures of H. pylon can be used for the tsolation of plasmids. However, we noticed that relatively higher plasmid yields were obtained from plate cultures than from broth cultures. This observatton may result from the fact that a small amount of growth medium still remains in the cell pellet. Some components of the medium may protect cells from lysis. 4 Several factors, such as a time to collect H pylon cells, concentration of DNA, and restriction-modification systems of the recipient cells potentially affect the efficiencies of natural transformation. It has been observed that under the growth condition described m the protocol, recipient H pyiorr cells, after initial inoculation, give higher transformation frequencies at 5 or 10 h (3). However, the timepoint might be changed for optimal transformanon in mdrvidual laboratories when the growth conditions of H. pylori are changed. The transformanon frequencres ofH. pylon are lmearly related to donor DNA concentrations (3). In the case of H. pylori UA83 1, the number of Rrf transformants obtained with 5 clg DNA is 12-fold higher than that obtained using 0 5 pg DNA. Generally, 0.5-2 pg DNA is sufficient for transformation.

151

DNA Transformation

H pylon shows significant diversity at the genome level among Its strains (20). Also, their restrlctlon-modification systems appears to be different (3). These genetic traits might constitute the barriers for gene (particularly plasmid) transfer between strains. 5 Fresh cold plates used for inoculatmg recipient cells yield a higher transformation efficiency. The reason for this observation is unknown at present. 6. While performing the transformation, we observed that the cells, which have grown beyond the Initial inoculated area when the donor DNA solution IS applied, contain more transformants than the cells grown on the original spot We suggest, therefore, that 15 pL is an optimal volume for both covering a bacterial lawn of g-10 mm and producing freshly grown cells 7. Efficiencies of electrotransformation depend on a number of factors. Some of them have been discussed in detail by Miller (15). In addition, H. pylori produces large amounts of DNase, which may play a role in natural transformation, perhaps by digesting incoming DNA before uptake (3). Settings of the parameters of the electroporation apparatus significantly influence transformation efficiencies. In the protocol described in Section 3.4 , a reslstance and a time constant are adjusted from 200 ohm and 5 ms to 600 ohm and 12.5 ms, respectively. Such modifications give better transformation frequencies However, these specifications are only optimal for a Blo-Rad Gene Pulser Conditions for other instruments need to be set up according to the manufacturer’s instructions.

Acknowledgments We thank the following

agencies for supporting

our work on H pylon*

Canadian Bacterial DiseasesNetwork (Centres of Excellence Program), GlaxoWellcome Canada, the Medical Research Council of Canada and Alberta ATRT. DET is a Scientist with the Alberta Heritage foundation for Medical Research. References 1 Lee, A., Fox, J , and Hazell, S. (1993) Pathogenicity of Helzcobacter pylon. a perspective. hfect Immun 61, 1601-1610. 2. Nedenskov-Sorensen, P., Bukholm, G., and Bnvre, K. (1990) Natural competence

for genetic transformation in Cumpylobacterpylori.

J Infect Du

161,365-366.

3. Wang, Y., Roos, K. P., and Taylor, D E. (1993) Transformation of Helzcobacter pylori by chromosomal metronidazole resistance and by a plasmld with a selectable chloramphenicol resistance marker. J Gen Mlcroblol. 139,2485-2493 4. Haas, R., Meyer, T. F., and van Putten, J. P. M. (1993) Aflagellated mutants of Helicobacter pylon generated by genetic transformation of naturally competent strains using transposon shuttle mutagenesis Mol Microblol 8,753-760. 5. Ferrero, R , Cussac, V., Courcoux, P , and Labigne, A. (1992) Construction of lsogenic urease-negative mutants of Hekcobacter pylon by allellc exchange J Bactenol.

174,4212-4217.

152

Ge and Taylor

6. Lorenz, M. G. and Wackernagel, W. (1994) Bacterial gene transfer by natural genetic transformatron in the environment, Microbzol Rev. 58,563-602 7. Stewart, G. J and Carlson, C. A. (1986) The btology of natural transformation Annu Rev Microbzol 40,21 l-235. 8. Chang, S and Cohen, S. N. (1979) High frequency transformatton of Bacrllus subtilzs protoplasts by plasmid DNA. Mol. Gen Genet. 168, 11 l-l 15. 9. Wang, Y. and Taylor, D T. (1990) Natural transformatron in Campylobacter species. J Bacterzol 172, 949-955. 10. Ge, Z , Hiratsuka, K., and Taylor, D. E (1995) Nucleotide sequence and mutational analysis indrcate that two Helicobacter pylorl genes encode a P-type ATPase and a cation-bmding protein associated wtth copper transport. Mol Mlcroblol.

15,97-105.

11 Moore, R. A , Beckthold, B., Wong, S., Kurershl, A., and Bryan, L. E. (1995) Nucleottde sequence of the gyrA gene and characterization of ctprofloxacmresistant mutants of Helicobacter pylon. Antrmwob. Agents Chemother. 39, 107-111. 12 Chassy, B. M and Fhckmger, J L. (1987) Transformatton of Lactobacdlus case1 FEMS Mlcroblol

Letts 44, 173-I 77.

13. Miller, J. F , Dower, W. J., and Tompkins, L. S. (1988) High-voltage electroporation of bacteria. genetic transformatton for Campylobacter jejuni with plasmid DNA. Proc Nat/. Acad. Scl USA 85,856-860 14 Dower, W. J., Miller, J F., and Ragsdale, C. W (1988) High efficiency transformation of E colr by high voltage electroporatron. Nucleic Acids Res 16,6 127-6 145. I5 Miller, J. F (1994) Bacterial transformation by electroporation Methods Enzym 235,375-385

16. Suerbaum, S., Josenhans, C., and Labigne, A (1993) Cloning and genetic charactenzatron of the Hehcobacter pylori and Helxobacter mustelae jlaB flagellm genes and construction of H pylorl jlaA- and jlaB-negative mutants by electroporation-mediated allehc exchange J Bacterzol 175,3278-3288 17. Cover, T. L , Tummuru, K R., Cao, P., Thompson, S. A , and Blaser, M (1994) Divergence of genetic sequences for the vacuolatmg cytotoxm among Helzcobacterpylori strains. J Biol Chem. 269, 10,566--10,573 18 Ezaki, T , Takeuchi, N., Liu, S., Kat, A., Yamamoto, H., and Yabuuchi, E. (1988) Small-scale DNA preparation for rapid genetic identification of Campylobacter spectes wrthout radtoisotope Mzcrobiol Immunol. 32, 14 l-1 50 19 Birnbotm, J. C and Doly, J. (1979) Rapid alkaline extractron procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7, 15 13-l 523 20 Taylor, D E , Eaton, M., Chang, N , and Salama, S (1992) Constructton of a Helicobacter pylori genome map and demonstration of diversity at the genome level J Bacterzol 174,6800-6806

17 Random Mutagenesis of the H. pylori Genome Agnes Labigne 1. Introduction Transposable elements are well-known genetic tools that enable the geneticist to generate mutations by disrupting the linear continuity of a specific gene and, consequently, affect its expression. This approach, in addition to providing an efficient way to create mutants, which are easily detectable because of the acquisition of drug resistance phenotype (when a drug-resistance transposable element is used), also provides an easy way to recover and identify the mutated gene. Several genetic systems have been developed m the past for many bacterial species and have permitted random and generalized mutagenesis to be performed on given genomes by introducing, via conjugatton, suicide plasmids harboring transposableelements.Unfortunately, suchtransposondelivery systems have not yet been found to be functional in Helicobacterpylori. A procedure is described here that overcomes this handicap by usmg shuttle mutagenesis; a method that involves the introduction of a transposon mto cloned genes in Escherichia co& and then the transfer of the transposon msertion to the H. pylori genome by allelic exchange via homologous recombination. The procedure has been adapted from the shuttle mutagenesis approach described by Seifert (I), in which the MiniTn3 transposable element has been modified by the replacement of the P-lactamase-encoding gene with a kanamycin gene originating from Campylobacter coli (2). This kanamycin gene has been shown to be well-expressed in H. pylori (3). The Tn3 transposon was chosen as the basis for a transposable element for H. pylori, because it is known to preferentially insert in AT-rich DNA sequences, and so would be suited to the H. pylori genome, which has a GC of 35-36%. The MiniTnKm (Fig. l), in accordance with the original MiniTn3, has several features (I). First, it retains the 38-bp inverted repeated sequences of the EdWl

From Methods in Molecular Medrcme, Hehcobacter pylon Protoco/s by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

153

NJ

Labigne

154

MiniTn3 38bp

20.19’ 3%

1.2 kb

b B

5 G

MiniTn3-Km sbp

1.8 kb

1oxP

=bp

C/a I Hmd /?I Bgl II Pa I

PILL570

Badf PSI I

I

5.3 kb Fig. 1. Restriction map of MiniTn3. According to Seifert et al (I), the derivative MimTn3-Km also designated MiniTnKm, as well as the coqugatlve vector used for the cloning of Sau3A H pylorl fragements.

original Tn3, but lacks the Tn3 transposase and resolvase genes. Therefore, to obtain the transposition of the deficient transposon, the strain harboring the mmiTnKm has to encode a functional Tn3 transposase in trans. Second, m order to obtain the resolution of the cointegrate, an obligatory intermediate during the Tn3 transposition process, the lox site (a specific site for recombmation originating

from the Pl phage), which has been Inserted between the two

3%bp inverted repeats, has to be recognized by the cre gene product provided zn trans. A major advantage of using such trans-dependent elements 1s that when finally inserted into the genome of H. pylori following allelic recombination, the transposon cannot retranspose; the mutatron created by the insertion of the MimTnKm

is therefore a stable mutation.

The random mutagenesis of H pylori genome mvolves three major steps. First, the generation of small chromosomal fragments (34 kb in length) by partial Sau3A restriction digests of the chromosome followed by their ligation into the BgZII linearized and dephosphorylated spectinomycin-resistant pILL570 vector. This vector (2) is a pBR322 derivative in which the DNA

Random Mutagenesis of Genome

155

sequencesresponsible for the Tn3 immunity have been deleted (2). The recombinant plasmids are transformed into a strain containing the pTCA plasmid, a pACY 184 derivative plasmid, compatible with the pILL570 derivative plasmid. pTCA encodes the Tn3 transposase and has been made immune to Tn3. The bank of H. pylori genome DNA is stored as individual clones in 96-well microtitration plates at -80°C. The second step consists of the random mutagenesis of each cloned fragment in E. coli. Ninety-six independent recombinant plasmids can be mutagemzed in parallel without extensive DNA manipulations. The random mutagenesis is achieved by two consecutive mating steps. During the first mating, the IncF derivative plasmid, pILL553 (4/ (i.e., pOX38::TnKm, naturally derepressed for transfer) is introduced into HB 101 containing the pTCA plasmid, expressing constitutively the Tn3 transposase (TnpA) and the pILL570 derivative plasmid, containing the cloned H. pylori fragment to be mutagenized. The kanamycin HB 101 transconjugants are selected on a medium contaming tetracycline (resistance conferred by pTCA), kanamycin (resistance conferred by pILL553),

and spectinomycin

(resistance conferred by pILL570

denvatlves).

These are grown at 30°C for 48 h, in order to allow the transposition event to occur, and the bacteria are mated with a plasmid-free recipient, the NS2114 (Rif) recipient strain, in order to select for the transfer of the cointegrates, i.e., to select for the transfer of PILL570 derivative plasmids in which transposition took place. The NS2 114 strain harbors the cre gene encoding a product able to resolve the cointegrate by specific recombination, givmg rise to two replicons, one consistmg of the original donor of transposon (pILL553), and the other consisting of PILL570 derivative hybrid plasmid m which MiniTnKm has been Inserted. Positive selection of resolved forms of the cointegrates is obtained by selecting the kanamycin NS2 114 transconjugants on medium containing high antibiotic concentrations (Km, 400 pg/mL and Sp, 400 pg/mL), a selective media on which only high copy number plasmid will be able to confer such levels of resistance, but the unresolved comtegrate-harboring strains will not grow. The third step in the random mutagenesis procedure consists of a cesium chloride extraction of the pILL570-derivative::TnKm plasmids from pooled NS2114 transconjugants, and their electrotransformation into an H. pylori strain competent for transformation (3). This is the only step where DNA manipulation is required. Randomly mutagenized H. pylori are then selected on a kanamycin containing H. pylori selective medium.

2. Materials 2.1. Material 1 Microcentrifuge. 2. Horizontal gel electrophoresis apparatus.

156

Labigne

Table 1 Plasmids Required for the Random Mutagenesis Procedurea Plasmid

Phenotyptc characteristics

Size

Refs.

pILL570 pTCA pILL553

ReppBR322, mob, Sp ReppACYC 184, Tc, TnpA, ImmTn3 pox38 :TnKm

5.3 kb ND >50 kb

4 I 4

aReppBR322 and ReppACYC184. rephcative fonctrons from pBR322 or pACYC184, mob. coqugative plasmid resulting from the presence of OrlT from the IncP plasmids; Sp, Tc, Km* resrstance to spectmomycm, tetracycline, and kanamycm, respectively; TnpA* consrtrtutrve expression of the Tn3 transposase, IrnmT113nnmun~ty to Tn3, ND not deternuned 3 37OC mcubator (with and wtthout shaking)

4. Anaerobic Jars (BBL, Cockeysville, MD, or other brand) with a carbon dioxide generator envelope with no catalyst. 5. Ultracentrifuge equipped with a 60Ti (Beckman, Fullerton, CA) angular rotor, a vertical VT165 rotor (Beckman), and a swinging SW27 rotor, as well as the polyallomer tubes that tit the rotors (quick-seal tubes for Vtt50 and Vti65, and regular polyallomer disposable tubes for SW27) 6 Voltage generator

7. Boiling water bath. 8. Regular microcentrifuge 9 Elutip-D

tubes (1 5 and 0 5 mL)

columns NA 010/2 (Schletcher

and Schuell ref. 462 617) (D-37586

Dassel, Germany). 10. UV light (wavelength 300-360 nm) 11 Gene Pulser apparatus (Bra-Rad) and 0.2-cm electroporatron cuvets (Bio-Rad).

12. 96-well moculator designed to deliver lo-&

ahquots.

2.2. Reagents 1. The plasmids required for the procedure are listed in Table 1. 2. Bacterial strains: HB 101 (5) (hsd R hsdA4 recA supE44 lacZ4 EeuB6 proA thl-l Sm); NS2 114Sm (recA, SmlOO, contains a h-cre prophage), (1) and its rrfampicm spontanous mutant (4), DHl (6) (F-supE44 recA1 endA gyrA96 tht-I hsdRI 7qr,-rnk+] relA1). H pylorz strain N6 1sused for the preparation of chromosomal DNA and as the recipient strain for the electroporation experiments, because of its favorable transformability (3). 3. E co12 strams are grown m L broth without glucose (10 g of tryptone, 5 g of yeast extract, 5 g of NaCl per liter), or on L agar plates (containing 1 5% agar) AntibiE. colz strains are as follows: 20 or 500 mg/L kanamycm, 8 mg/L tetracycline, 100 or 500 mg/L

otic concentrations for selection of plasmtd-harboring

spectmomycin, 50 mg/L nahdrxic acid, 100 mg/L rtfamptcm. 4 Solid media for H pylori growth: horse blood agar supplemented with vancomycm (10 mg/L), polymtxin B (2500 U/L), trtmethoprtm (5 mg/L), and amphotericin B (4 mg/L). For the selectton of transformants (mserttonal

Random Mutagenesis of Genome

5. 6. 7. 8. 9. 10. 11. 12.

13. 14.

157

mutants), and their growth, kanamycm is added at the final concentration of 20 mg/L. Protemase K: 20 mg/mL (store as aliquot at -20°C). EDTA-Sarkosyl: to 3.25 mL OSM EDTA, pH 8.0, add 1.25 mL 10% sarkosyl (Ciba-Geigy NL97, Basel, Switzerland). TES: O.O5MNaCl,5 mMEDTA, 30 mMTrn+HCl, pH 8.0. TES-CsCl-aprotinin solution: to 99 mL TES add 1 mL of aprotmm (Sigma, St Louis, MO) and 126 g of cesmm chloride, Calf intestinal mucosa alkaline phosphatase (Pharmacia, ref 27-0620). 50X TAE: 242 g Trizma-base, 57.1 mL glacial acetic acid, 100 mL 0.5MEDTA: dissolve and adJust the pH at 8.0. Can be kept at room temperature (see Note 1). Ethidium bromide: stock solution 10 mg/mL (Caution: Use gloves to handle this reagent) Agarose gel: Dissolve 1 g of normal agarose in 100 mL of 1X TAE buffer; add ethidium bromide to the final concentration of 400 pg/mL, cool at 55°C before pourmg m the casting box. 5X Blue loading buffer: for a lo-mL final volume, mix 5 mL of glycerol, 5 mL of Tris-EDTA, pH 8 0, buffer (10 mA4-1 mM, respectively) and 7 mg of bromophenol blue. Sucrose solutions: 10 % sucrose m 1M NaCl, 20 mM Tris-HCl, pH 8 0, 5 mM EDTA; 40 % sucrose in 1M NaCl, 20 mM Tris-HCl, pH 8 0,5mM EDTA.

3. Methods 3.1. Construction of a Genomic Bank in the plL L570 Vector 3.1.1. Preparation of Chromosomal DNA 1. Inoculate 15 plates of blood agar H. pylori selective medium with a culture of the 2 3. 4. 5. 6 7.

N6 strain, and incubate the plates in a jar under microaerobic conditions at 37OC for 48 h (see Note 1). Harvest the cells with a swaband resuspendthem in a 25-mL 15% glycerol, 9% sucrose sterile solution, and pellet the cells by spinning at 4500g for 30 min (see Note 2). Resuspend the cells in 2 mL 25% sucrose, 50 mM Tris (pH 8.0) containmg 10 mg/mL of lysozyme and transfer the suspension directly into a quick-seal polyallomer ultracentrifuge tube (30 mL) (see Note 3). Add by injecting in the quick-seal tube 10 pL of proteinase K (20 mg/mL) followed by 650 pL of the EDTA-Sarkosyl solution. Incubate for a few minutes in a water bath at 65°C. As soon as the lysate starts to clear, fill up the tube with the cesium chloride-aprotinm solution. Spin in the angular rotor for 48 h at 40,000 r-pm at 18°C. Collect the DNA by punching the polyallomer tube with an 18-gage needle in the upper part of the tube to allow air to enter; the needle should be left inserted in the tube so the operator can control the flow out of the tube. If required, a 20-mL syringe may be connected to the needle to hasten the flow. A second 18-gage needle will be punched in the bottom part of the tube. Collect the drops m a

158

Labigne

tube and save, m a separate Eppendorf tube, the material that will flow out of the tube bearing a viscous appearance. These viscous fractions correspond to the chromosomal DNA; the total volume of DNA/CsCl should be approx 1.5-2 mL 8 Introduce the sample by pouring (again without pipeting the DNA, to prevent any damage) m a dialysis bag, and dialyze the DNA solution three times against 1000 mL TE, buffer pH 7.5, over a 24-h period. 9. Pour the DNA solution m a sterile tube, and store at 4°C. This procedure provrdes a highly purified and concentrated high molecular weight chromosomal DNA m a final volume of 1.5-2 mL, without the reqmrement for phenol extraction

3.1.2. Partjal Sau3A Restriction of Chromosomal DNA 1. Chromosomal DNA from H pylori N6 is partially cleaved with the restrmtion enzyme endonuclease Suu3A m order to generate a malonty of fragments wrth a size ranging between 3 and 5 kb. This might require prekinetic experiments on a small DNA volume (25 pL) in order to determine the amount (dilution) of enzyme to use and the duration of the incubation at 37°C. Once defined, reproduce the digestion conditions m the same final volume (30 pL) for a series of 24 such reactions. To synchronize the beginning and the end of the 24 reactions, spot 2 pL of the diluted Sau3A enzyme on the top of the Eppendorf tube containing 25 l.tL of DNA and 3 pL of 10X Suu3A buffer. Spin to start the reaction. Spot 2 pL of EDTA 0 5M again on the side of the tube and spm to stop the reaction. 2. Pool the 24 samples (final volume 24 x 32 pL = 768 fi) and inactivate the enzyme by heating 5 min at 65’C. 3 Load half of the sample (380 6) on the top of two 36 rnL 10-40% sucrose gradients and spin m a swinging rotor (SW27) at 18°C for 22 h at 27K (see Note 4). 4 Collect the two gradients in order to obtain about 50 fractions from each gradient (15 drops/fraction). 5 Load on a 1% 1X TAE-ethidnnn bromide agarose gel, 20 p,L of the fractions corresponding to the upper part of the gradient, as well as a molecular weight (Ladder). 6. Pool the fractions containing DNA fragments, ranging from 3 to 5 kb; measure the final volume of the pooled fractions, and increase the volume of the DNA fragments with TE buffer to lower the sucrose concentration to 8% (dilutton to I :5). 7 Add 3M acetate sodium to achieve a final concentration of 0.3M, mix, and add two volumes of cold ethanol to precipitate the DNA. Mix well and store at -20°C for a few hours before spinning at 15,000 rpm to collect the DNA. 8. Resuspend the lmeartzed 3-5 kb DNA fragment in 50 pL.. of distilled and sterile water, and load 2 @ on the gel to estimate the final concentration (see Note 5) 9 Store this stock of Sau3A-partially restricted fragments at -20°C

3.7.3. Ligation of the Genomic Fragments into the plLL570 Vector, and Storage of the HBlOl (pTCA) Genomic Bank 1 Following plasmid purification of the pILL570 vector by cesmm chlortdeethidmm bromide ultracentrifugation, according to classical protocols, lin-

i

Random Mutagenesis of Genome

2

3.

4

5.

6.

7.

8.

9. 10. 11.

12. 13.

earize 3 pg of purified pILL570 with the endonuclease BglII in a buffer compatible with the dephosphorylation reaction (“One for all” 10X buffer from Pharmacra, for example). Followmg 1 h of restriction, load 100 ng of DNA to check that the plasmtd has been completely lmearized (presence of a umque DNA band migrating as a 5.3-kb fragment). Dephosphorylate the linearized plasmid by adding 0.1 U of calf mtestmal alkaline phosphatase and incubate the mix for 30 min at 37°C . Inacttvate the enzymes at 85°C for 15 mm. Load the sample on a preparative 1% agarose gel prepared m 1X TAE-ethtdmm bromide, and run under constant voltage. Electroelute the lmearized fragment (5.3 kb in size), and purify the eluted fraction by passing the DNA through an Elutip column according to the recommendattons of the manufacturer Precipitate the DNA from the 400~pL equate by adding 800 pL of cold ethanol, spin for 15 min at 4“C at 10,OOOg. Dry the pellet and resuspend the DNA in 20 p.L of distilled water. The concentration of the lmearrzed vector should be 150 ng/pL. Set up a first ligation reaction by addmg 300 ng of lmeartzed vector, 50 ng of chromosomal Suu3A fragments (final volume of 10 pL), 1X ligation buffer. Add 100 U of lrgase and incubate the ligation reaction at 14°C for 18 h. Prepare 100 mL of HB 10 1 harboring the pTCA plasmid-competent cells, following a standard protocol ( 50 mM CaC12 treatment of the cells), ending wtth a 5mL suspension of competent cells. Transform 200 l.tL of competent cells with 10 pL of the ligation reaction and select transformants on L-agar plates containing tetracycline (8 pg/mL) and spectinomycin ( 100 pg/mL). Check by mini-plasmid extraction and plasmid restrictton from 12 tetracyclmespectinomycm resistant transformants that the plasmids do contain single inserts and that the proportion of plasmid without insert is very low (see Note 6). If the amount of clones harboring hybrid plasmtds is too low, readapt the conditions of ligatton and transformation in order to obtain more than 100-200 clones from each ligation mix. When the correct conditions are defined, multiply, as much as possible, the ligation and transformation steps and save the independent transformants by storing 96 clones at once m a 96-well mrcrottter plate. Fill each well wtth 186 p.L of L-broth supplemented with the two antibiotics, tetracycline and spectinomycin. Inoculate each well with a single transformant clone with a tooth pick, incubate the 96-well plate at 37’C. When growth 1svisible, add 7% of DMSO (i.e., 14 PL of DMSO to the 186 & of culture) and store the plate at -80°C. 20-25 of these 96-well microtitration plates will be representative of the whole genome of H pylori, with hybrid plasmids havmg inserts of 3-5 kb. The mutagenesis will then be performed in E colz by manipulating from I-4 96-well plates in parallel

Labigne

160 3.2. Mutagenesis

of 96 Independent

DNA Inserts

3.2.1. Transposition of MiniTnKm on the plLL570 Derivative Hybrid Plasmids 1. Thaw one stored 96-well plate (plate 0 = frozen stock) Prepare a new 96-well plate and distrtbute m each well 100 pL of L-broth without antibiotic (plate 1) 2. Inoculate each well of plate 1 with 10 & of each well of plate 0 using a 96-well inoculator designed to dehver a lo-@, liquid sample. Plunge the inoculator into the wells of plate 0 and then plunge it mto the wells of plate 1, 3, Incubate plate 1 at 37°C with shaking for a few hours, whtle storing again plate 0 at -80°C. 4. In parallel, inoculate 50 pL of an overnight liquid culture of strain DHI-harbormg the pILL553 plasmid (the donor of MmiTnKm) into 5 mL of L-broth without antibtottcs Shake gently at 37°C 5. When both the cultures in the 96-well plate and the DHl (pILL553) culture are m the early exponential growth phase, add 100 I.~Lof the DHl(pILL553) culture to each well of the microtitration plate 1 using a repetitive distributor Incubate the plate at 37°C for another 4 h (see Note 7). 6. Select HBlOl transconlugants harboring the pILL553, m addition to the pTCA and the pILL570 dertvatlve plasmtds, by spotting 10 pL of the mating mixture (DHl [pILL553] x HB 101 [pTCA, pILL5701) from the 96 wells of microtitration plate 1 on a square L-medium agar plate contammg 50 ug/mL kanamycin, 100 clg/mL. spectinomycm, and 8 pg/mL tetracycline, using the 96-well inoculator. Incubate the plates at 37°C for 24 h (see Note 8). At this step, one should obtain confluent spots as a result of the high frequency of transfer of the pILL553 plasmid 7. Using toothpicks, maculate each of the 96 growing spots obtained on the square L-medium agar plate, into 200 pL of L-broth containing the three antibiotics (kanamycm, spectinomycin, and tetracycline) distributed in the 96 wells of a new microtitration plate (plate 2). Incubate with gentle shaking at 30°C for 36-48 h. During that time, transposition should occur and lead to the formation of cointegrates between the pILL553 MimTnKm donor and the pILL570 derivative plasmid.

3.2.2. Positive Selection of the Clones in Which MiniTnKm Transposition has Occurred by a Second Mating Step 1 Grow the E colz NS2114 cells overnight m 5 mL of L-medium broth free of antibiotics 2. Distribute m a new microtitration plate (plate 3) 200 pL of L-broth free of antibiotic. Inoculate twice each well of plate 3 with 10 ~.ILof each well of plate 2, using a 96-well inoculator (volume of inoculatton 2 x 10 pL). 3 Incubate plate 3 at 37°C with shaking for a few hours until growth becomes visible. 4 In parallel, inoculate 50 pL, of the overnight liquid culture of strain NS2 114 mto 5 mL of L-broth without antibiotic. Shake gently at 37°C.

Random Mutagenesis of Genome

761

5. When both the cultures in the 96-well plate (plate 3) and the NS2 114 culture are in the early exponential growth phase, add 100 pL of the NS2 114 culture to each well of the microtltratron plate 3 using a repetitive distributor, and incubate the plate at 37 “C with gentle shaking for another 6 h. 6 Inoculate with 10 pL from each well of plate 3 a square L-medium agar plate containing 500 clg/mL spectmomycm, 500 pg/mL kanamycin, and 100 ClgimL rifamptcm using the 96-well inoculator. Incubate the plates for 3648 h at 37°C 7. The growing colonies should be E co/i NS2114 cells harboring the pILL553 original plasmid and the pILL570 derivative, in which MiniTnKm has been inserted. Because of the structure of pILL570, very few msertions of MmrTnKm will occur in the vector. In contrast, insertions of the transposon should occur m the AT-rich regions of the cloned fragment. 8. At this stage, a series of mini-plasmid preparations from independent clones should be made before the extraction of the pooled disrupted hybrtd plasmtds m cesium chloride, in order to check that the process of transposition did work 9. First check that the NS2114 E. cok cells are tetracyclme-sensitive, then compare the restriction digestion patterns of several pILL570 derivatives grown from plate 1 and from the respective plasmids from plate 3.

3.3. Transformation of a Naturally Competent H. pylori /so/ate by Electroporation 3.3.1. Purification of a Pool of Disrupted Hybrid Plasmids 1, Pool from the Km(500) Sp(500) Rif- plates all the growing colonies and inoculate a lOO-mL L-broth medium containing Km(50), Sp(100) with the colonies without any amplificatton step, in order to avoid the selection of faster-growing clones. This pool corresponds to 96 independent cloned fragments, each randomly disrupted by the MiniTnKm transposon. Incubate with shaking at 37°C for 18 h. 2. Proceed to the extraction of the pool of disrupted plasmids by an alkaline lysis procedure (any protocol of plasmid purification is suitable), and proceed to the purification of the 96 disrupted plasmids by recovering the pool of supercoiled DNA plasmids from a cesium chloride-ethidium bromide gradient 3. Extract the ethtdmm bromide with isopropanol saturated with NaCl, and precipitate the plasmid with cold isopropanol Resuspend the pellet of the pool of plasmids m distilled and sterile water to achieve a final plasmid DNA concentration of 1 pg/pL (see Note 5). Store the pooled plasmid DNA solutions at 4°C

3.32. Electrotransformation and Selection of Randomly Mutagenized H. pylori The following procedure refers to one electrotransformatton. It must be repeated several times to increase the number of independent mutants that can

beobtainedfrom onepool of disruptedplasmids (Le., from a setof 96 mdependent cbromosomal

inserts).

162

Labigne

1. Grow H pylori N6 on blood agar plates for 48 h under microaerobic conditions at 37’C One plate with confluent growth of H pylori will be sufficient to perform one electrotransformation 2 Harvest the cells from the plate with a swab, and resuspend them in 20 mL of 15% glycerol, 9% sucrose sterile solution Pellet the cells for 30 min at 4500g and resuspend the pellet m 50 pL of glycerol-sucrose at 4’C 3 Add 2 pL of the supercoiled purified plasmids (2 pg of DNA), and mix gently with a tip by pipetmg. 4. Transfer the cells and the DNA to a prechilled (kept at -20°C) 0.2 cm electroporation cuvet and place in a Gene Pulser apparatus. Achieve a pulse by setting the apparatus as follows: 25 F, 2 5 kV, and 200 ohms. 5. Resuspend the cells m 100 pL of SOC buffer and spread the suspension on blood agar plates without kanamycm. Incubate the plate at 37’C under microaerobic conditions for 36 h. 6. Harvest the bacterial cells on a swab onto a blood agar plate containing 20 pg/mL of kanamycin, as well as a medium that allows the direct positive screening of a precise phenotype resultmg from the msertional mactivation of a gene. 7. Kanamycm-resistant H. pylon transformants can be detected following mcubation at 37°C under microaerobic conditions for 60-l 20 h

4. Notes 1 Because of the genomic variability that exists among clinical isolates, it is a prerequisite to extract the DNA from the H. pylon strain that will be mutagenized at the end. The only consideration that has to be taken into account is that the stram used be transformable. For convenience, such a strain has been designated the N6 strain m this protocol. 2. A long centrifugation time is required because of the presence of glycerol, which will slow down the sedimentation of the bacteria. 3. The lysis of the bacteria must be carried out followmg the transfer of the cell suspension into the ultracentrifuge tube in order to prevent damage of the DNA through pipeting steps 4. All the solutions used have to be autoclaved before their use for the preparation of the sucrose gradient 5. The smallest DNA concentration that will be visible by the naked eye under UV trradiatton followmg ethidium bromide stammg can be estimated as being 40 ng of DNA for a 4-mm-wide well 6. At this stage, the host stram harbors two plasmids: the pILL570 derivative plasmid, plus the pTCA piasmtd. 7. The pILL553 plasmid is a ~0x38 derivatrve plasmid, derepressed for autotransfer. As a consequence, it will spread very efficiently within 1 h mto the recipient host 8 The use of 100 Clg/mL of spectmomycin is imperative because of the low level resistance of HB 101 to spectinomycin.

Rancfom Mutagenesis of Genome

763

References 1. Selfert, H S., So, M., and Heffron, F. (1985) Shuttle mutagenests. a method of transposon mutagenesis for Saccharomyces cerevwae. Proc Natl. Acad. Scz USA 83,735-739. 2. Labigne-Roussel,

3.

4.

5.

6.

A., Courcoux, P., and Tompkins, L (1988) Gene dtsruptton and replacement as a feasible approach for mutagenesis of Campylobacterjejunr. J Bacterzol 170, 1704-1708 Ferrero, R. L., Cussac, V., Courcoux, P., and Labigne, A. (1992) Constructton of isogenic urease-negative mutants of Hehcobacter pylorl by allelic exchange J Bacterial. 174,42 12-42 17. Labigne, A., Courcoux, P., and Tompkins, L. (1992) Cloning of Cumpylobacter Jejune genes required for leucine btosynthesis, and constructton of leu-negattve mutant of C. jejuni by shuttle mutagenesis. Res Microbzol 143, 15-26. Boyer, H. W. and Roulland-Dussoix, D. (1969) A complementatton analysis of the restriction and modification of DNA in Escherzchra toll. J Mol. Biol 41, 459-472. Bachmann, B. J. (1996) Derivations and genotypes of some mutant derivatives of Eschericlua cob, in E co11and Salmonella typhlmurium, Cellular and Molecular Bzology, 2nd ed (Neidhardt, F. C., Curtiss, III, R., Ingraham, J. L., Lm, E. C C., Low, K. B., Magasanik, B., Regnikoff, W. S., Riley, M., Schaechter, M., and Umbarger, H. E., eds.), ASM, pp. 2460-2488.

Construction of a Genomic Map of If. pylari by Pulsed-Field Gel Electrophoresis (PFGE) Nicholas Chang, Qin Jiang, and Diane E. Taylor 1. Introduction The inability of conventional gel electrophoresis to separate DNA molecules exceeding 50 kb in size led to the development of pulsed-field gel electrophoresis (PFGE) by Schwartz et al, (I) in 1982. He introduced the concept of applying two alternating electric fields (i.e., pulsed-field) to separate DNA molecules greater than 50 kb embedded in an agarose gel matrtx. Since then, many instruments based on this principle have been developed. For a discussion of various pulsed-field applications, see review articles by La1 et al. (2) and Crete et al. (3). It was shown that, under the influence of an electric field, a DNA molecule embedded in a gel matrix undergoes reorientation, elongation, and migration along the field toward the anode. When a second field is applied in an alternate direction, the DNA molecule must reorientate, elongate, and migrate along the direction of the new field. Larger DNA molecules will take longer to reorientate than smaller molecules; therefore the larger ones spend less ttme migrating down the gel than the smaller per pulse time. Consequently, larger DNA molecules will appear near the origin while the smaller molecules will migrate furthest. This principle becomes important when dlfferent fragment sizes of DNA are to be separated. PFGE has become an important tool for determining genome sizes, physical mapping of the chromosome, and localization of genes on the chromosome of prokaryotic microorganisms. The PFGE method described below, for genome sizing and gene mapping of H. pylori genome, employs the hexagonal electrode array system designed by LKE3 Instruments (Pharmacia, Uppsala, Sweden) and it is a modification of the

Ed&d

From Methods m Molecular Medrcme, Hehcobacter pylorl Protocols by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

165

NJ

Chang, Jiang, and Taylor

166

contour-clamped homogenous electric field (CHEF) system produced by BioRad (Hercules, CA) and other manufacturers. For a detailed explanation of this system, see the review by Chu (4). Two other PFGE methods employed to a lesser extent for H. pylori are transverse alternating field electrophoresls (TAFE) and field inversion gel electrophoresls (FIGE). These will be discussed where relevant. Genomic mapping and sizing of H. pylori DNA begins first with the isolation of the DNA from the bacterial cell embedded within the agarose gel matrix. A typical regime for studying the bacterial DNA is as follows: 1 Growing the bacterial cells to optimum concentration, usually wlthm 24-48 h. 2 Embedding an appropriate concentration of cells in l-1.5% low melting point agarose gel 3 Lysis of bacterial cells by treatment with sarcosyl and protemase K, thereby freeing the bacterial DNA from the cell intact 4. The DNA within the agarose IS then subjected to restriction digestion with specific restriction enzymes to produce DNA fragments of varying sizes 5. The digested DNA embedded in agarose is then subjected to PFGE to separate out the DNA fragments A DNA restriction profile is now available for sizing, physical mapping, and genomic mapping. 6. Each DNA fragment can now be labeled with a suitable radlotsotope, such as 32P, and, with cross-hybridization, one can determine the fragment order on the chromosome, resulting in a physical map 7. Each DNA fragment can also be hybridized to any number of radlolabeled gene probes to determine the location of the genes on the chromosome, resulting in a genetic map 8. Analysis of DNA fragments from double digestion helps confirm fragment order of the physical map 9 The reader should be aware that the methods described below provide a basis for preliminary bacterial genome mapping. Other mapping strategies may be required to further refine the map obtained by PFGE Additional information can be obtained from the review by Fonstein and Haselkorn (5) on strategies for refining a genetic map.

2. Materials 2.1. Growth Media 1 2. 3. 4. 5. 6 7.

Brain-heart infusion (BHI) agar. Bovine serum albumin or fetal calf serum. Amphotericin B. Vancomycin Petri plates. Anaerobic Jar Supply of gas tank containing 5% C02, 5% H2, and balance of N2.

Genomic Mapping by PFGE

167

2.2. Cell-L MP Agarose Mixture 1 McFarland turbidity standards or a spectrophotometer to measure cell density at OD600 nm. 2. Low-melting-point agarose (LMP). 3. Lauroyl sarcosine. 4. Proteinase K. 5. Sterile clear plastic tubes (14 mL/17 x 100 mm) with caps. 6 Plastic molds (100~pL capacity) for making 100~@, sized-agarose blocks 7. Sterile clear glass cover slips for slicing DNA-agarose blocks. 8. Mlcrocentrifuge tubes (1-mL capacity). 9. Agarose (electrophoresis grade). 10. PFGE apparatus.

2.3. Incubation

Temperatures

Water baths are required at 37,45, 50, and 65°C.

2.4. Materials for Cell Lysis and PFGE 1. ESP (EDTA, N-lauroylsarcosine, proteinase K) solution 1X: 0.25M EDTA (pH g&9.5), autoclaved, add IV-lauroylsarcosine to 1% (w/v) concentration, shake vigorously to get into solution, then add 5 mg/mL proteinase K. Freeze in 4-5-mL aliquots/plastic tube. 2. TE (Tris-EDTA) solution (10X): 10 miI4 Tris-HCl, pH 7.4, 100 mM EDTA, dilute 1: 10 with sterile distilled water for use; LMP: Prepare 1.5% LMP m 100 mL TE solution, store in 60°C incubation, PMSF (phenylmethylsulfonylfluoride) solution. 3. Prepare O.lM PMSF by dissolvmg 17.5 mg in 1 mL of lsopropanol. Store m 500 $ or 1-mL aliquots (microcentrifuge tube) at -2O’C!, or the solution can be prepared fresh before use. 4. Restriction buffers: According to manufacturer’s supply or specification. 5. Electrophoresis running buffer: 0.5M Tris-Boric acid-EDTA (TBE) stock solution: 12 1.Og Tris-base, 61 .OBoric acid, 7.4 g EDTA, pH 8.8 Dilute 1: 10 for use. When using a system not designed by LKB, readers should consult the manufacturer’s protocol for buffer concentration.

3. Method 3.1. Growth of H. pylori H. pylori is a Gram-negative microaerophilic bacterium, an organism that IS difficult to grow, whether in liquid or solid media. For the purpose of convenience and ease of handling, growth on solid media should be the method of choice for most DNA isolation experiments. On solid media, the bacterium requires complex media supplemented with 5-l 0% blood or serum. A variety of growth media are described in the literature (6-8). Successful cultivation of

168

Chang, Jiang, and Taylor

H. pylori can be achieved by any of the published methods and quite often depends on personal preferences, the availability of media, and the ability to provide a microaerobic environment m the laboratory. The protocol outlined below has been used with success, parttcularly in the Isolation of the organism from gastric biopsies, usually within 2-4 d. 1. Prepare culture plates of BHI agar contaming 5% BSA, 10 pg/mL of vancomycm and amphotencin B, and 1 5% yeast extract 2 Streak the bacteria onto the plate or spread with a glass spreader. 3. Incubate the plates either m a microaerobic chamber or anaerobic jar containmg 5% CO,, 5% HZ, and 90% N, 4. When an anaerobic jar is used, the air mixture is allowed into the jar vra a PFTE IN-LINE filter (0 2 pm, Nalgene, Rochester, NY) This eliminates contammation by antibiotrc resistant bacteria from the gas supply. A combmatron of antrbrotic regimen can also be used, e.g., 10 ug/mL of vancomycin, 5 pg/mL of polymyxm B, 5 pg/mL of trimethroprin, and 10 ug/mL of amphotericin B 5 Incubate the plates for 48 h at 37°C. 6. Check the condition of growth and for any contammation by observrng a small sample of suspension (m PBS) under a dark field microscope 7 Alternatively, prepare a Gram stam and observe under a light microscope 8. Do not allow bacterial growth to reach the coccoid stage, as determined by morphologrcal observatron. DNA IS usually degraded in the cocci 9. Appropriate bacteria should appear as slightly curved rods.

3.2. Preparation

of LMP Agarose Blocks

1 Resuspend the H pylorz cells in 1 mL of TE buffer to obtain a turbidity level equivalent to McFarland turbidity tube No 5 (which has an approx ODs5c of 0 25 or 2-5 x lo9 CFU/mL). 2. Vortex cells into an even suspension 3. Transfer 100 & of the cell suspension into 1 mL of 1.5% molten LMP agarose m TE buffer and mrx. 4 Dispense immediately in 100~pL aliquots of the cell-agarose mixture into 100~p,L capacity plastic molds. 5 Allow the agarose mixture to solidify at 4°C.

3.3. Ceil Lysis Correct preparation of the ESP solutton is a very important step because completion of cell lysis must be achieved for further DNA work. 1. Transfer the solidified agarose-cell mixture (agarose blocks) very carefully mto 5 mL of ESP solution with an L-shaped thin glass rod that tits into the mold to a maximum of 12 agarose blocks. 2. Incubate the tube in a 50°C water bath for 48 h 3 When agarose blocks appear almost transparent, cell lysis has occurred, otherwise mcubate again with fresh ESP solution for another 24 h

Genomic Mapping by PFGE

169

4. Store the agarose blocks at 4°C until used (do not remove blocks from ESP solutron). The DNA m the agarose is stable for up to 6 mo storage.

3.4. Restriction

Enzyme Digestion

Check for undigested DNA by conventional mini-gel electrophoresrs. If the integrity of the DNA in the agarose is uncertain, simply run an undigested DNA agarose sample by conventional gel electrophoresis, and observe for any DNA breakdown products in the form of a smear; intact DNA will remain at the top. The DNA-agarose blocks can now be removed from the ESP tube for restriction enzyme digestion. 1 Each DNA-agarose block contains enough DNA material for several PFGE runs. The DNA-agarose block can be cut into at least SIX thin insert pieces with a glass coverslip. The thinner the insert, the thinner the DNA band will appear in the PFGE gel. 2. Transfer each Insert mto a 1-mL microcentrifuge tube contammg 1 mL of TE buffer plus PMSF (if only one insert is to be run, cut out only one insert from the agarose block). Use a clean sterile spatula. 3. Wash the agarose insert three times with fresh TE-PMSF (with 15-min intervals between washes). 4 Replace the TE-PMSF solution with 1 mL TE buffer and wash the agarose-Insert with three changes of TE buffer. Remove TE buffer and add 100 pL of restriction enzyme buffer according to manufacturer’s directions. 5 Allow 1 h incubation at room temperature. 6. Replace with 100 pL fresh restriction buffer 7. Add l&20 U of enzyme (see Note 1); mix well 8. We generally favor incubation at 37°C or as directed by manufacturer for 8 h or more as required. 9. Wash the DNA agarose Insert with TE buffer. 10. Load mto PFGE agarose wells.

3.5. PFGE 1. Prepare 15&200 mL of a 1% agarose m TBE buffer (diluted 1: 10 from stock). Bring to a boil and let it cool to about 56°C. 2. Pour the agarose gel to no more than 4 mm thick on a gel plate. 3. Make a smgle long well in the agarose gel with an aluminum or plastic plate 140 mm long, 30 mm high, and 3 mm wide. 4. Allow gel to solidify. Although combs containing any number of wells are usually available for the individual user, a single long well is easier to work with and air bubbles can be avoided during insert placement 5. The 3-mm width should be sufficient to allow the DNA agarose sample to sit m the gel, and the DNA-agarose insert (or plug) can be cut to a minimum wtdth oflmm.

Chang, Jiang, and Taylor

170

Table 1 Molecular Sizes and PFGE Pulse limes for Separation Fragment sizes (kb) separation range

DNA markers

Pulse, s/run time, h

30-200 50-500 100-1000

h Concatamers h Concatamers Yeast DNA

200-l 500

Yeast DNA

WI, 10/4,20/16 5f4, 15/4,25/16 70115, 130/l 1 or 10/4, 15/4, 45/16 7017, lOOf7, 145/10

6. Samples of the DNA-agarose plugs are posttioned m the well m any order desired, makmg sure that matenal for handlmg DNA-mserts are clean (a metal spatula can be dipped m alcohol or can be flamed before use). Normally, the well allows up to 16 DNA-agarose mserts, including DNA size standards, which should be positioned in the first and last lanes. 7. With a Pasteur pipet, dispense 1% melted agarose over the well until all the DNAagarose inserts are covered, allowing the agarose to solidify. 8. Remove rubber mold from gel plate 9 Transfer the gel plate mto the PFGE gel box containing 2 L of 0.05M TBE running buffer 10 The running buffer level should be just above the gel surface; do not add any more than required to cover the PFGE-gel. 11. Begm pulsed-field run as required, using Table 1 as a general guide for setting the pulse times; feel free to modify run times or follow manufacturers’ specifications (The temperature of the running buffer should be set at lO-15°C during electrophoresis.) For fragment and chromosome size determination, at least three separate PFGE runs are required for accurate determmation: one run to determine the size range between 30 and 300 kb; the second run to determine the range between 50 and 400 kb, and a third run to estimate the size range between 200 and 1000 kb, or more. In the last case, yeast DNA markers are needed alongside h DNA size markers. 12. At the completion of PFGE, the gel is transferred mto a solution of ethidmm bromtde for at least 15 min (see Note 2). 13. Replace the EtBr stain with water and allow at least an hour for destaining. 14. Observe gel under UV light (see Notes 3-8). 3.6. Southern

Blot

Hybridization

Different PFGE patterns can be observed after DNA is digested separately with different enzymes. This is because each restriction enzyme recognizes a unique cutting sequence on the same piece of DNA. Therefore, one fragment from one digestion may overlap fragments from another digestion. Based on this principle, a labeled fragment from one digestion will hybridize with the

Genomic Mapping by PFGE

171

fragment(s) from the second digestion on the same region of the chromosome because of DNA homology. This phenomenon can be detected by exposmg the radioactive blot to autoradiography.

3.6.1. Transferring

DNA onto Membranes

Capillary transfer of DNA from gel to membranes is described by Southern (‘lo). Protocols developed from this idea are similar, and can be chosen by personal preference. The following method 1s a modified procedure from the molecular clonmg manual by Sambrook et al. (II/. 1. Mark the position of the standards and the origin of each DNA sample by makmg Incisions on both stdes of the gel. 2. Soak the gel in an acidic solution (0.25MHCl) for 10 min. An alternative method to break the larger fragments of DNA is to expose the gel to UV light (260 nm) for 20-30 min. This step IS recommended for efficient blottmg of DNA larger than 100 kb. 3. Transfer the predenaturing treated DNA-gel directly into denaturing solutton (0.5MNaOH, 0.5MNaCl) and leave tt for 30 min. 4. Transfer the gel into neutralizing solution (IMTris-HCl, pH 7.4,SMNaCl) for at least 30 min. 5 Cut out one prece of nitrocellulose or nylon membrane and four pieces of Whatman 3MM paper to the same size as the gel. Finally, place the gel along wrth the precut membrane and 2 pieces of Whatman paper in 10X SSC for 2-3 min before they are positroned for transfer. 6. The setup starts from the bottom layer. The gel is placed on a piece of long Whatman filter paper that is latd over the glass bridge on a container of 20X SSC, and should be above the two overhanging ends. Cover the gel with nitrocellulose membrane. Mark the posittons of standards and orrgm described in step 1 onto the membrane with a pencil using the incisions made prevtously m the gel. On top of the membrane are laid two pieces of wet Whatman filter paper, and then two pieces of dry Whatman paper Add a pile of paper towels and, lastly, place a weight on top.

3.6.2. Hybridization 1. Set three water baths at 65,45, and 37’C. 2. Cut out the fragments from the LMP PFGE gel, store the fragments separately in Microcentrifuge tubes Two inserts of each fragment should be used for one labeling. 3. Put the tube in the 65°C water bath for about 5 mm or until the agarose 1s completely melted. An alternative way of melting agarose is to put the tube m boiling water for 2-3 min. 4. Cool the tube down in a 45°C water bath for 5 min. Add agarose buffer (25X, 0.75M Brs-Tris, 0.25M EDTA, 4 pL/lOO pL of melted agarose), and agarose (lU/lOO mg agarose), mix well.

172

Chang, Jiang, and Taylor

5. Incubate the mixture at 45°C for 2-3 h Gently mtx several times during the Incubation period. When digestion is completed, the mixture should not solidify even at a cool temperature 6. Biol the DNA mixture for 5 mm, dip on ice m-mediately, then add dATP, dGTP, dTTP, random primers, labeling buffer, Klenow, H,O and 32PdCTP according to mstructions from the Random Primers DNA Labeling System Kit (GIBCO-BRL, Gaithersberg, MD), (see Note 11). 7 Add 250 pL of 10% herring sperm DNA to the tube containing the probe, and boil the tube for 10 mm. 8 Add the labeled probe into the container that contams the blot and hybridization buffer (50% formamide, 25% 20X SSC, 2% [w/v] SDS, 10% 10X PM buffer [0 2% ficoll, 0.2% BSA, 0.2% polyvinyl pyrrolidone], 0.4% 0.25MEDTA, 9.6% H,O). 9. Incubate the blot at 37’C in an oven overnight The hybrrdizatron can be shortened to 2 h. 10. Wash the blot with washing solution (20X SSC 250 mL, 20% SDS 10 mL, 0.25M EDTA 4 mL, H,O 730 mL) at room temperature for 15 mm, then transfer the blot mto a fresh solutton and wash at 65°C for 30 min. 11. An optional step used to get rid of nonspecific background hybridization is to wash the blot with another liter of high stringency buffer (20X SSC, 10 mL; 20% SDS, 10 mL; 0.25M EDTA, 4 mL; H,O, 976 mL) for 30 min to 2 h 12. Wrap the blot in plastic wrap and expose to X-ray film.

3.6.3. Construction of PFGE Map The next step is to analyze the radroautograph patterns of each hybrrdization fragment (see Note 10). Using H. pylori NCTC 11637 as an example, the alignment of two Not1 and one NruI fragments is shown in Fig. 1. Here, fragment No. 5 of NCTC 11637 from iVru1 digestion 1s labeled with 32P. This fragment was found to hybridize to Not1 restriction fragments 5 and 9. Therefore, fragments 5 and 9 of the Not1 digestion are adjacent to one another on the chromosome. In this way, all other fragments can be aligned m the appropriate order by fragment crosshybridization (13). After completion of the physical map, the genetic map can be constructed, using a similar method, but with additional use of known gene probes to produce a preliminary physical and genetic map, as shown m Fig. 2.

4. Notes 1 Several restrtctton enzymes are available that cut H pylon DNA mto fragment sizes that can be separated by PFGE. For a restriction enzyme to be of any use, it should cut the DNA into at least five fragments and no more than 15, and these fragments should be spread out over a range of 30-500 kb. Such enzymes should have a 6-8 bp recognition sequence cut site. In this protocol, the two enzymes used to cut H. pylon DNA are NruI and Not1 (9). The enzyme NruI has a 6-bp recognition sequence conststmg of 5’ TCG/CGA 3’, while Not1 has an 8-bp rec-

I

Genomic Mapping by PFGE

173

NmI chgeslm

Not1 dlgesilon

of NCTC 11637 1

of NCTC 11637 PFGE

fragment

*t 1

5

labelled v/Oh “P by N~ckiranslabon cm8s hybridize to the PFGE Southern blot of Not1 digestion of NCTC 11637

PFGE

I

fragment

alignment

NruI

Fig. 1. Strategy for mappmg H pylon The above drawing illustrates the experimental sequence for the ahgnment of restriction fragments generated by Not1 and NruI digestion of H. pylori DNA. *represents 32P label and * -* represents 32P labeled fragment. ognition sequence made up of 5’ GC/GGCCGC 3’. The reader should feel free to try out new restriction enzymes, but should bear in mind that the type of base pair recognition sequence has little bearing on the ability to cut the DNA, based on the G+C% of the DNA. A restriction enzyme with only GC recognition sequence does not necessarily indicate that it will have a greater tendency to cut DNA with a high G+C% content. 2. Ethidium bromide intercalates between the two DNA strands and illuminates brightly when exposed to ultraviolet (UV) light. The brightness (or intensity) of the DNA bands and their location in the gel are the first indication of the success of the PFGE run. When the gel is scanned from the top down, much useful information can be derived. The gel must be destained to give a clear background; otherwise, there will be masking of faint bands, particularly bands less than 50 kb in size. DNA break-

Chang, Jiang, and Taylor

174

23s’

26KDa

Fig. 2. Genomic map of the ff. pylon NCTCI 1637. The map was constructed by hybridrzation of 32P labeled DNA fragments as described in Section 6.2. DNA probes used for mapping are described m ref. 12.

3.

4. 5.

6.

down products can be a problem because they often appear as a smear m the 30-l 00-kb region and light up under UV light, masking the presence of any DNA bands in the area. In this case, uncut chromosomal DNA should be prerun m a conventional mmr-gel for l-2 h to remove degraded DNA DNA contamination from other bacteria may be deduced by the presence of intense DNA in the gel well at the top. In this case, the digested DNA appears as famt bands m the gel. This may suggest that the contammatmg bacteria were not susceptible to ESP lysis and the DNA was not accessible or susceptible to the restriction enzyme used. Overloaded DNA may also remam in the well. Uncut DNA will enter the gel, but often remains as a broad band at the top. Plasmids have not been a problem in PFGE. If the presence of a plasmrd 1ssuspected, isolate it and determine whether it can be cut by the restriction enzyme to be used. DNA that are completely cut by the restriction enzyme at its cut site always appear as intensely bright bands. Partially cut DNA fragments often appear dull under UV light. A DNA band that IS intensely bright under UV light, and wider than other bands, indicates the presence of fragments banding together. This is particularly so when smaller size bands appear brighter than larger size fragments. There are indirect methods of estimating the number of fragments having the same size, such as double-digestion with another restriction enzyme, or comparison of the total size obtained by digestion with two or three restriction enzymes. It is highly unusual that more than three fragments will have the same size. Partial fragments may also add to the intensity of the DNA band.

Genomic Mapping by PFGE

175

7. The postttons of the h-size markers, or any makers at the left and right periphery of the gel, will mdicate the consistency of current flows in the gel. The mdtvidual marker bands from the left end and the right end should run parallel, in a straight line, and hortzontally, although occasionally slanting does occur If slanting occurs, it must be taken into account when measuring the fragment stzes. A slanting pattern indicates that the current flow is not consistent from left to right. The wiring in the hexagonal electrode should be checked and replaced, if necessary, or the pulse control unit needs to be examined for defects. If the slant is not extreme, measurement of fragment migration can still be made, as long as h-size markers or other markers are present at the gel pertphery and at the center. 8. Each DNA fragment size should be determined from an average of at least three separate PFGE runs, and, if posstble, from more than one set of pulse times. Measurement of fragment size 1s gained by using the mobilittes of the h DNA ladder to construct a calibration curve; the size of the unknown fragment can then be determined by the distance it has migrated, i e., plot a graph with h DNA ladder vs distance migrated. For determination of standard deviation, more than three runs are required for statistical accuracy. The standard deviation can be calculated by first obtammg the arithmetic mean X of the fragment, using the formula. Arithmetic mean = X = CXIN where X = size of the fragment obtained from one PFGE run, CX = the sum of the sizes obtained from any number of the PFGE runs for a particular fragment, and N = the number of PFGE runs performed to determine the size of the fragment. Example: XI, X2, X3, X4 represent the sizes obtained from four runs. x=zx/N=x1+x2+x3+x&JN The standard deviation is then calculated from the formula. s = 1-m

= 1(X,

-X)2

+

(X,

-X)2

+ (X, -Q2

+

(X4

-Q2/4

9. Alternative methods of PFGE, such as TAFE or FIGE, are used only tf the system described in this section is not available, or if a comparison of data for different systems is required. The FIGE system can separate out fragments from 10-I 50 kb very well. The user should follow the manufacturers’ protocols, since the running buffer and pulse time conditions are different for each system. 10. Of the few ways to construct a physical map, fragment crosshybridization gives the most direct evidence and accurate results. One fragment from one enzyme digestion is used as a probe to hybridize against the DNA fragments from a Southern blot of the other enzyme digestion. The physical map then can be constructed by analyzing the hybridizatton patterns from each fragment. 11. To label the probes, [a32P]-dCTP is used because of labelmg consistency and cost-effectiveness when compared with nonradiolabeling (chemilummescence). However, precautions are needed to handle 32P, and the safety guidelines should always be carefully followed.

176

Chang, Jiang, and Taylor

Acknowledgments We thank the following agencies for supporting our work on H. pylori: Canadian Bacterial DiseasesNetwork (Centres of Excellence Program), GlaxoWellcome Canada, the Medical Research Council of Canada, and Alberta ATRT. DET is a Scientist with the Alberta Heritage Foundation for Medical Research. References 1. Schwartz, D. C., Saffran, W., Welsh, J., Haas, R., Golderberg, M., and Cantor, C. R. (1982) New techniques for purifying large DNAs and studymg their properties and packaging. Cold Sprmg Harbor Symp Quant Btol 47, 189-l 95. 2 Lai, E , Birren, B. W., Clark, S M , Simon, M. I., and Hood, L (1989) Pulsed field gel electrophoresis. BloTechmques 7,34-42 3. Crete, N , Delabar, J. M., Sinet, P. M., and Creau-Goldberg, N. (1991) Accurate evaluation of the sizes of DNA fragments (from 30 to 4700 kb) m pulsed field gel electrophorests BioTechniques 11,7 1 l-7 18. 4. Chu, G (1989) Pulsed-field electrophoresis in contour-clamped homogeneous electrtc fields for resolution of DNA by size or topology. Electrophoresls 10, 290-295. 5 Fonstein, M. and Haselkorn, R. (1995) Physical mapping of bactertal genomes J Bacterlol 177,3361-3369. 6. Morshed, M. G., Karita, M., Konisht, H., Oktta, K., and Nakazawa, T. (1994) Growth medium containmg cyclodextrin and low concentration of horse serum for cultivation of Helicobacterpylori. Microblol Immunol 38(11), 897-900. 7. Xta, H. X., Keare, C. T., and O’Morain, C A. (1994) Culture of Hellcobacter pylori under aerobic condttion on solid media Euro. J Clm. Mtcroblol Ifect. Dis 13,4OHO9 8 Nedenskov, P (1994) Nutritional requirement for growth of Helicobacterpylorz. Appl Environ Mlcrobiol 60,3450-3453. 9 Taylor, D. E., Eaton, M., Chang, N., and Salama, S M (1992) Constructton of a Hellcobacter pylorl genome map and demonstration of diversity at the genome 1evel.J Bact. 174,6800-6806. 10. Southern, E. M. (1975) Detection of specific sequences among DNA fragments separated by gel electrophorests. J. Mol Blol 98, 503 11. Sambrook, J., Frttsch, E. F., and Maniatis, T. (1989) Molecular Cloning. A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 12. Jrang, Q. (1994) Genomtc analysis of Campylobacter and Helicobacter. M SC thesis, Umverstty of Alberta, Edmonton, Alberta 13. Jiang, Q,, Hiratsuka, K., and Taylor, D. E. (1996) Variabtltty of gene order m different Hellcobacterpylon strains contributes to genome diversity. Mol Micro 20(4),833-842

Molecular Characterization of #f. pylori Surface Antigens Peter Doig, Paul W. O’Toole, and Trevor J. Trust 1. Introduction Despite its clinical significance, relatively little is known about the components of Helicobacter pylori that allow it to colonize, persist, and elicit an inflammatory response within the host. Bacterial surface components frequently influence colonization and persistence of a pathogen, as well as the disease process (Z). In the case of H. pylon, one macromolecular assembled protein component that is unequivocally located on the surface of the bacterium is the sheathed flagellum. Two other proteins that can be isolated m abundance from suspensions of H. pylori cells, but that are normally regarded as intracellular in other bacterial species, are urease and a GroEL analog, Hp60K (2-6). Whether these large macromolecular assemblies are surface protems on H. pylori or are releasedas a result of cellular lysis is uncertain. However, numerous other intracellular proteins can be readily isolated from culture supernatants after mild shearing and extraction procedures, and these released proteins have initially been incorrectly interpreted as representing surface proteins (7-9). To unambiguously identify surface-exposed proteins of H. pylon, we have used as probes polyclonal and monoclonal antibodies directed against the outer membrane components (ZO). Many surface-exposed proteins are constituents of the outer membrane, and are frequently implicated in pathogenesis. Using these antibodies, we have identified surface-localized outer membrane proteins, including porin species, and we have been able to clone various surfacelocated proteins (Z&I 3). Here, we describe some of the techmques we have employed to identify surface components and some of the molecular methods used to clone and study such components. We also describe techniques we From Methods /n Molecular Mechne, Helicobacter pylon Protocols E&led by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

177

NJ

Doig, O’Toole, and Trust

178

have employed to produce mutants defective in various surface components, which then can be used to study the component’s function and/or role m pathogenesis. Many basic techniques, such as sodium dodecyl sulfate polyacrylamrde gel electrophoresrs (SDS-PAGE), agarose gels, using restriction enzymes, and so on, are not described, since details of these techniques have been extensively described elsewhere (14-l 6). 2. Materials 2.7. General 1. Chocolate blood agar (CBA). Brain-heart infusion (BHI) agar plus an additional 3 g of agar per liter. After autoclaving, while the agar is still at approx 80°C supplement it with 5% (v/v) defibrmated horse blood. 2. Equipment and solutions for SDS-PAGE and Western blotting (this is described m detail in vol 32 of the Methods w Molecular Biology series). 3. General equipment and supplies for molecular biology, such as water baths, agarose gel equipment and solutions, transilluminator, and restriction enzymes

2.2. Preparation

of Outer Membranes

for Use as an Antigen

1. 2. 3. 4

10 mM Tris-HCl, pH 7.5 French pressure cell or sonicator. DNase and RNase (Boehrmger Mannheim, Mannheim, Germany) Sucrose solutions of 20, 52, 58, 64, 70% (given as w/v) all made up m 10 mM Tris-HCl, pH 7.5. 5 2% (v/v) Triton X-l 14, 20 mM Tris-HCl, pH 7 5

2.3. Immunization

Protocol

1 Stain 0 5% Coomassie brilliant blue R-250 dissolved in 20% ethanol, 10% acetic acid and filtered through Whatman 3MM paper prior to use. 2. Destain: 20% (v/v) ethanol, 10% acetic acid (v/v) in water. 3 10 mM phosphate buffered salme (PBS), pH 7 5. 4. A tissue grmder (preferably Teflon tipped with a steel shaft). 5. Freund’s complete adjuvant and mcomplete adjuvant (Gibco, Grand Island, NY). 6. New Zealand White Rabbit. 7 Various gage needles and 5- and IO-mL syringes.

2.4. Confirmation

of Surface Localization

by Electron Microscopy

1. Copper grids coated with formvar. 2. 10 mA4phosphate buffered salme, pH 7.5, containing 1% (w/v) bovme serum albumin. 3 15 nm colloidal gold particles coated with protein A (Amersham Canada, Oakville, ON, Canada). 4 1% (w/v) ammonium molybdate, pH 7.5.

179

Characterization

of Surface Antigens

2.5. Preparation

of Genomic DNA from H. pylori

1. 2. 3 4 5 6.

TE buffer. 10 mA4 Tris-HCl, pH 8.0, 1 mM EDTA. GES solution: 5Mguanidinmm thiocyanate, 100 mMEDTA, 0.5% (w/v) Sarkosyl. A sterile, hooked glass capillary Ice-cold ethanol (95%) 70% ethanol. Proteinase K (Boehrmger Mannhelm) stock solution (20 mg/mL) in water.

2.6. Preparation

of a Genomic Library

1. Restriction enzyme BarnHI. 2. Arms of the h insertion vector hZAP Express, predigested with BumHI and treated with calf intestinal phosphatase, are purchased from Stratagene (La Jolla, CA) 3. Gigapack Gold (Stratagene). 4. Various growth media as described by Stratagene m their hZAP and Gtgapack manuals.

2.7. lmmunoscreening 1. 2. 3. 4. 5. 6.

2.8. Construction 1. 2. 3. 4. 5 6. 7. 8.

9.

of the ;1ZAP Genomic Library

LB plates. Nitrocellose filters (0.45~pm pore size). Tris-buffered saline (TBS): 10 mMTris, 150 mMNaC1, pH 7.5. 2% skim milk m TBS Alkaline phosphatase-conjugated antirabbit secondary antibody. Color development reagent: 0.165 mg/mL 5-bromo-4-chloro-3-indolyl phosphate (Boehringer GmbH, Mannhelm, Germany) and 0.33 mg/mL nitro blue tetrazolmm (Sigma, St Louis, MO) dissolved m 100 mM Tris-HCI, 150 mM NaCl, 5 mM MgC&, pH 9.5.

of a Vector Containing

an Inactivated

Gene

Purified plasmid DNA (see Note 8). Restriction enzyme to cut plasmid at desired site. Antibiotic resistance cassette. E. coli XL-1 blue. LB agar with and without the antibiotics necessary for selection of the plasmid and the cassette. PCR oligonucleotide primers. Plasmtd pUC18. PCR reaction buffer. 10 mM Tris-HCl, pH 8.3, 50 mM KCl, including 2.5 mA4 MgC&, 200 @all four dNTPs, 200 nA4each primer, and 0 5 U Tuq polymerase (AmpliTaq, Perkm-Elmer Cetus, Norwalk, CT). A thermal cycler.

2.9. Hectroporation

of H. pylori

1. Hams F12 cell culture medmm (Gibco) supplemented with 10% (v/v) fetal calf serum, 4 pg/m.L amphotericin B, 10 pg/mL vancomycin, and 5 pg/mL tnmethopnm.

Doig, O’Toole, and Trust

180

2. 15% (v/v) glycerol, 9% (w/v) sucrose. 3. CBA plates containing the appropriate antibtottc for selection of the antibtotic resistance cassette. 4. Electroporator

2.10. Natural Transformation 1. BHI broth containing 1% fetal calf serum. 2 BHI broth. 3 CBA plates containing the appropriate annbiotic for selection of the antibiotic resistance cassette.

3. Methods 3.1. Preparation

of Antisera

In order to clone the genes encoding a surface-localized component, the protein must first be either purified or enriched. This preparation may then be used to produce an antigen for the development of a specrfic antiserum, which is used to detect clones expressing the antigen. Below, we describe a method for purifying the outer membrane relatively free of inner membrane contamination (see Note 6).

3.1.1. Preparation of Outer Membranes for Use as an Antigen 1 Grow bacteria as lawns for 2 d on 40 chocolate blood agar (CBA) plates (10 cm diameter) and harvest by suspension in 10 mA4 Tris-HCl, pH 7 5 2. Collect cells by centrifugation (12,OOOg, 4°C 20 mm) and wash three times with 10 mA4 Tris-HCl, pH 7.5 3 Resuspend cells m 10 mM Tns-HCl, pH 7.5, and lyse by three passages through a French pressure cell (15,000 psi) (see Note 1) 4 DNase (0 1 mg) and RNase (0.5 mg) are added to the cell suspension and the mixture incubated at room temperature for 30 min. 5. Remove unlysed cells by centrifugation (4OOOg, 4°C 15 min) 6. Collect total membranes by centrtfugation (4O,OOOg, 30 mm, 4’C) and suspend m 10 mMTris-HCl20% (w/v) sucrose, pH 7.5. 7. Layer this suspension on top of a 52/58/64/70% (w/v) sucrose step gradient (listed top to bottom) 8 Subject the gradient to ultracentrifugation in a swinging bucket rotor (150,OOOg for 18 hat 1O’C) 9. Remove the bands from the tube, wash three times with 10 mA4Tris-HCl, pH 7 5 (150,OOOg for 30 mm), and assay for protein, succmate dehydrogenase (SDH) activity, and ketodeoxyoctonate (KDO) content (see Note 2). The fraction containing the least SDH activity, and highest KDO content, is the crude outer membrane preparation 10. Remove most of the remaining inner membrane contamination, as measured by SDH activity, by extraction with 2% (v/v) Triton X-l 14,20 mMTns-HCl, pH 7.5.

Characterization of Surface Antigens

181

Suspend membranes m the Tnton solution to a final concentration of 0.4 mg protem/mL and incubate at 4°C for 30 min. Warm the preparatton to 15’C and subject to ultracentrifugation (150,OOOg for 1 h at 15’C). Resuspend the pellet in 10 mMTris-HCl, pH 7.5 (precooled to 4’C), and wash three times by centrifugation to remove any remaining Triton.

3.1.2. Immunization Protocol 1. SDS-PAGE is performed by the method of Laemmli (I 7) in a mint-slab apparatus using 15% acrylamlde gels (see vol. 32 of this series for details of SDSPAGE). Prepare the sample by adding outer membrane (2 mg/mL) to an equal volume of SDS-PAGE sample buffer and heat at 100°C for 15 min. Add 20 pg of protein to each well. 2. Visualize protein bands after stammg with 0.5% Coomassie brilliant blue R-250 and destammg with 25% (vol/vol) ethanol and 10% acetic acid (v/v) in water 3. Excise bands containing the protein of interest from the gel with a scalpel or razor blade. Neutralize the gel slices by several changes of PBS. 4. Place the gel slices m a manual tissue grinder cooled on ice, add PBS, and grind to a tine paste. Dilute this paste to form a slurry and mtx 1.1 with Freund’s complete adjuvant. 5 Inject this mixture intramuscularly (0.5 mL) and subcutaneously (4 sites 0.25 ml/site) into a New Zealand White Rabbit. 6 Give mtramuscular and subcutaneous booster mjections in Freund’s mcomplete adJuvant on d 14 and 35. 7 On d 45, exsanguinate the rabbit and collect the serum; store at -20°C 8. Assess the antiserum for specificity to H. pylon by Western immunoblottmg, using the following antigens: whole cells of H pylorz, outer membranes of H. pylori, and whole cells of all E co11strams to be used in subsequent molecular techniques. All samples should be boiled in SDS-PAGE sample buffer for 15 mm prior to loading on the gel.

3.1.3. Confirmation of Surface Localization by Electron Microscopy 1. For immunoelectron mtcroscopy, float a copper grid coated with formvar on a drop of a turbid suspension of H. pylon’ m PBS for 1 mm, remove, and float the grid on a drop of PBS containing 1% (w/v) bovine serum albumin for approx 30 min 2 Incubate the grid on a drop of PBS containing the diluted antibody for 1 h (see Note 10). 3. After incubation, remove the grid and remove nonspecifically bound immunoglobulin by floatmg the grid on three successive drops of PBS (10 min/drop) 4. Float the grid on a drop of PBS containing a 1:50 (v/v) dilution of 15 nm colloidal gold particles coated with protein A. 5. After incubation for 30 min, remove the nonspecifically bound colloidal gold particles by floating the grid on three drops of PBS, as described in Section 3 1 3 , step 3. 6. Negatively stain the grids by floating the grid on a drop of 1% (w/v) ammomum molybdate for one mm. Draw off the excess stain from the grid with a piece of

182

Doig, O’Toole, and Trust Whatman filter paper, and then examme by transmission electron microscopy to detect surface localized gold particles

3.2. Molecular Cloning Many of the basic techniques are described in detail elsewhere (16). Described below are the techniques that can be employed to isolate clones expressing a particular surface antigen. We have attempted to give all relevant details of any procedure that differs from a standard technique. 3.2.7. Preparation of Genomic DNA from H. pylori DNA is isolated from members of the genus Helicobacter by a modtfication of the guanidine lysis protocol of Pitcher and coworkers (18), as described below. 1, Harvest cells from a single CBA plate (24-48 h growth) by a stertle cotton swab, resuspend m 1 mL TE buffer, and pellet by centrtfugation for 20 s m a mtcrofuge 2. Resuspend the pellet in 100 & TE buffer and add 500 pL GES solution. 3. Mix the sample by repeated gentle inversion; lysis usually ensues to completton with 15 min 4. Precipitate the crude DNA by addition of 1 mL ice-cold ethanol (95%), and remove on a sterile hooked glass capillary 5 Followmg bnef immersion of the DNA coated glass capillary in 70% ethanol, resuspend the DNA gently m 500 pL TE for 30 mm on tee 6 Add proteinase K (Boehringer Mannheim) to a final concentration of 20 pg/mL from a stock solution and incubate the sample at 50°C for 60 min 7. Recover the DNA by precipitation with ethanol as described above, and resuspend overnight at 4°C in 100 pL of TE

3.2.2. Preparation of a Genomic Library Standard techniques are used to produce a phage library containing H. pylon DNA. These techniques have been described extensively elsewhere and will not be described in detail here (see Note 7). Commercially available kits are used, and the methods recommended by the manufacturer are followed. Details not supplied by the manufacturers of the various kits are supplied below. Cut genomrc DNA with the restriction enzyme BamHI to yield lO-20-kb fragments. This IS achieved by removmg aliquots from the drgest tube at varrous time intervals (0,5, 15,30, and 60 mm). Stop the digestion of the DNA by immediately placing the aliquot at 80°C for 20 min to denature the enzyme. Analyze aliquots are analyzed by agarose electrophoresis and pool fractions containmg the appropriate sized fragments for use m producing the library. 3.2.3. lmmunoscreening of the itZAP Genomic Library (see Note 7) 1 Overlay phageon LB platesto produce 50-100 plaques/plate. 2 Place rntrocellose filters on top of the plate and allow to stand for 5 min.

183

Characterization of Surface Antigens

3. Remove the filters and wash wtth TBS extensively to remove any debris. 4. Block the filters m a solutron of 2% skim milk m TBS for 1 h, wash with TBS three times (10 minIwash), and add primary antibody at an appropriate dilutton 5. After 2 h at room temperature with gentle shaking, wash the filters three times with TBS, and add an alkaline phosphatase conJugated antirabbit secondary antibody. 6. After 2 h at room temperature with gentle shaking, wash the filters three times with TBS, and develop, using the color development reagent. 7. Retsolate positive plaques and amplify by standard techniques. 8. Make confluent lawns of the positive plaques; some of the bacterial cell debris can then be analyzed by SDS-PAGE and Western immunoblot to determine the identity of the peptide cloned. 9. Further mampulatton and characterization of the phage clone, such as bluescript excision, are performed as described by the manufacturer 10 Mapping the insert, subclonmg and sequencing the gene, encoding the H pylon peptide can then be performed (see Note 11).

3.3. lnsertional

Mutagenesis

The followmg section describes how to produce an isogenlc mutant m H pylori for a given gene. These mutants can be useful in assigning a function to the gene or gene product and discerning its possible role in pathogenesis

through the use of animal models. However, it must be noted that one may not be able to inactivate a gene that is essential for cell vtabtlity using these methods.

3.3.1. Construction of a Vector Containing an Inactivated Gene Using an Existing Restriction Site Without a Deletion The cat gene from PRY 109, derived originally

from C. coli, or the kanamy-

cm resistance cassette of pILL600 (19) can be cloned mto the mtragemc restriction site of the gene being studied. To employ this method, there must be a unique restriction site within the gene, preferably downstream from the start site, but not much farther than the center of the codmg region. Insertions near the end of the coding region should be avoided, since, in such cases, an active or partially active gene product might still be produced. 1. Cut purified plasmtd DNA containing the cloned gene and the antibrotlc reststance cassette with the appropriate restriction enzyme 2. MIX the DNAs at a ratio of 1: 1 and ligate by standard techniques 3. Transform E coEi strain TB 1 with ligation mixture Select transformed cells on a mednun containing the appropnate antibiotics for both the cassette and the plasmtd. 4. Examine cells containing the plasmtd by Western immunoblot to confirm that inactivation has occurred (see Note 3) 5. Propagate the plasmid and purify the DNA by standard methods

Doig, O’Toole, and Trust

184

DYNE

1

PCR 1

,--

cDYNE

1

PCR 2

Hindlll

1

Hhdlll, cbnlng Info pUC18Mlndfll

Hlndlll

<

jm Hlndlll

Restrldlon slle. Kmr-fragment lfgaflon

Hlndlll -L

Fig. 1. Schematic representation of the strategy for deletion-msertronal mactrvanon of an H pylon gene See text for complete details. Vectors are indicated by the thrcker black lmes, and cloned H pylon DNA by the thinner black lmes The locatron of the gene is indicated by the box m the first panel and the blackened area represents the region to be deleted. RS is the restrrctron site to be mtroduced into the sequence by PCR. The varrous constructs indicated are not drawn to the same scale

3.3.2. Construction of a Vector Containing an Inactivated Gene Using PCR to Generate a Deletion A two-step PCR-based procedure can be employed to mutate a particular gene locus (Fig. 1). Here we describe a method for generating a deletron in a gene inserted mto the clonmg site of pUCl8. Prrmer design is of the utmost importance. The internal oligonucleotide primers are designed to have comple-

Characterization

of Surface Antigens

185

mentary sequences at their 5’-ends, resultmg in PCR products with a 20- to 30nucleotide complementary overhang. The primers should also be designed to introduce a novel restriction site into the gene by the addition of the site into the complementary primer ends. Primers should also be far enough apart to introduce a significant deletion, but must allow sufficient complementary DNA at both sides of the deletion to allow homologous recombmation. The other primers (DYNA and DYNB [see Note 51) are complementary with sequences in the pUCl8 lad gene flanking the multiple cloning site. 1. Perform two separate PCRs with a plasmid contammg the wild-type allele as template (for simplification it is shown as one step in Fig. 1) 2. Purify the two PCR products by Magic mimpreps (Promega, Madison, WI), mixed at a ratio of l-1, and use as template in a second PCR reaction, using prtmers DYNA and DYNB only. The resulting PCR product will result m a deletion m the allele and an additional restriction site. 3 Purity the product as described above and cleave the ends with Hind111 Reclone the fragment into pUC 18 4 Clone the kanamycin resistance cassette into the unique restriction site generated by the PCR reaction as describedin Section3 3.1. 5. Propagate and purify the plasmid.

3.3.3. Construction of a Vector Containing an Inactivated Gene Using the Polymerase Chain Reaction 1. Perform PCR reactions in a standard reaction buffer, including 2.5 rruV MgCl,, 200 p.A4 all four dNTP’s, 200 ti each primer, and 0.5 U Tuq polymerase (AmphTaq, Perkm-Elmer Cetus), using a Perkin-Elmer model 9600 Thermal Cycler. 2. After a preamplification denaturation step of 95’C!, 2 mm, incubate reactions for 30 cycles of 94”C, 1 min; T,, 1 mm; 72’C, 1 min. The annealing temperature T, for the reaction is 24°C below the lower melting temperature of the primers 3.4. Transformation

of H. pylori

3.4.1. Electroporation

of H. pylori

After construction of the vector, H. pylori may be transformed by either electroporation or natural transformation. Many strains of IL pylorz have been shown to be naturally competent for uptake of DNA at a frequency of approx 10m6/pgDNA. However, in some cases,natural transformation is not successful, and electroporation may be tried. Both methods are described. 3.4.1.1. PREPARATION OF ELECTROCOMPETENTCELLS 1. Inoculate bacterial cells from CBA plates into Hams F 12 cell culture medium (Gibco, Burlington, Ontario, Canada) supplemented with 10% (v/v) fetal calf serum, 4 pg/mL amphotericm B, 10 pg/mL vancomycm, and 5 Clg/mL trimethoprim.

Doig, O’Toole, and Trust

186

Grow cultures m 100 mL medium m 250-mL flasks m microaerobic Jars, at 37°C

with shakingat 100‘pm. 2. After 24-48 h, when the cell density reaches lo* cfu/mL as measured by viable counting of test cultures, collect cells by centnfugatlon. 3 Wash the cells twice m 15% (v/v) glycerol, 9% (w/v) sucrose, and resuspend m l/5000 vol of the origmal culture m the glycerol-sucrose solution, 4. Store these electrocompetent cells at -70°C until use. 3.4.1.2. ELECTROPORATION 1 Thaw electrocompetent cells on ice and mix with l-2 pg DNA. 2 Transfer this suspension to a 0.1 -cm cuvet and subject to one pulse (25 RF, 200 Q and 1.5 kV, giving a time-constant of 4.5-4.8 ms) m a Blo-Rad electroporator 3 Resuspend cells immediately m 10 mL BHI broth supplemented with 5% fetal calf serum, and incubate for 16 h at 37”C, with shaking m a mlcroaeroblc envlronment 4 Collect bacterial cells by centrifugation and plate for single colony lsolatlon on CBA plates containing 20 pg/mL kanamycin

3.4.2. Natural Transformation 1 Harvest cells from plates after growth for 48 h, and resuspended m BHI broth contammg 1% fetal calf serum to an OD600nmof 0.6. 2. Add l-10 pg of plasmid DNA to a 0 5-pL ahquot of the cell suspension in a sterile mlcrocentrifuge tube and mamtam in a CO2 Incubator at 37’C for 3 h, with occasional gentle mlxmg. 3. Harvest the cells by centrlfugatlon at 13,000g for 30 s, resuspend m 100 pL BHI broth, and spread on the surface of a CBA piate. 4. After incubation for 24 h, harvest the visible layer of bacterial growth and transfer, with serial dilution, to a CBA plate containing kanamycin at 20 ccg/mL 5. Transformants typically grow to form visible colonies after 36 h. Purify by several rounds of single colony Isolation.

3.4.3. Confirmation of integration To confirm that double crossover rather than plasmtd integration has occurred, purified lineages of each mutant should be characterized by PCR and Southern hybridization. 1 Using primers flanking the insertion sites m the respective genes, carry out PCR on DNA purified from the H pylon transformant. The mutants should have a PCR product that has increased its size by the size of the cassette minus any deletion Introduced mto the gene 2 Use Southern hybridization with probes based on the gene, the kanamycin resistance gene, and the parent plasmld, to confirm that a single insertion of the marker into the expected chromosomal fragment has occurred without integration of the

plasmld itself,

Characterization of Surface Antigens 4.

187

Notes

1 Cells may also be disrupted by sonication. We have found that 6-8 30-s pulses with l-2 mm coolmg on ice between pulses will effecttvely lyse most of the bacteria 2. The assays used to quantitate the outer membrane content of a fraction m relation to inner membrane contaminatron are described elsewhere. Sources of the assay procedures are as follows: The succmate dehydrogenase (SDH) assay descrrbed by Osborn et al. (20) is used to assess inner membrane content The ketodeoxyoctanate assay described by Osbom et al (20) IS used to quantify outer membrane content The Bradford assay was used to measure protein, employing bovine serum albumin (BSA) as a standard (21). 3. E. coli cells harbormg this plasmid may produce no proteins reacting with the specific anttbody. However, a truncated protein mtght be expected to occur in some cases. 4. Chloramphemcol-resistant transformants were isolated at a frequency of approx IO-Qlg. 5. DYNA, 5’-AAAGGGGGATGTGCTGCAAGGCG-3; and DYNB, 5-GCTTCCGGCTCGTATGTTGTGTG-3’

(22).

6. It has been shown that the outer membrane of H pylon IS difficult to completely separate from the inner membrane, simrlar to the case with Campyiobacter sp. (23). Since this preparation ~111ultimately be used as the antigen source for identification of surface localized components, great care must be taken to assure that the outer membrane preparation 1s as free of inner membrane contammatton as possible. Significant inner membrane contammation will increase the probability of isolating nonsurface components. 7. For a more extensive explanation of this technique, see ref. 16 8. Plasmid DNA is isolated by an alkaline lysrs procedure usmg Magic mmiprepsTM as described by the manufacturer in all cases 9. The E. coEistram containing the plasmid with the mactivated gene can be examined by Western immunoblot to assure inactivation has occurred, if previous analysis has shown the gene expresses in E cob. The same is true for the transformed H pylon strains. 10. The dilution of the antrserum that gives clear surface labeling without a high background must be determined empirically. Typically, an antrserum that gives a strong reaction by Western rmmunoblot at a dilution of 1: 1000 is diluted to 1: 100 for immunoelectron mtcroscopy. Excessrvely concentrated antibody may result in nonspecific binding to the surface of H. pylori and result in a false assignment of the antigen to the surface. 11. Frequently, we have observed rearrangements and/or deletions occurring in cloned DNA encoding H. pylori outer membrane components When characterizing the clone, care should be taken that this has not occurred. It IS useful to have a complete restriction map of the parental clone, so that it may be compared to any clone resulting from subclonmg or other genetic manipulation.

188

Doig, O’Toole, and Trust

Acknowledgments The development of the methods described in this chapter was supported in part by grant Al29927 from the National Institutes of Health. The contmumg support of Astra Pharma is also acknowledged. References 1. Blaser, M. J. (1993) ffelzcobacter pylon: mlcroblology of a ‘slow’ bacterial mfectlon. Trends Mlcroblol 1,225-260 2. Dunn, B. E , Roop, R. M , Sung, C. C., Sharma, S A , Perez-Perez, G I., and Blaser, M. J. (1992) Identlficatlon and purification of a cpn60 heat shock protem homolog from Helrcobacterpylori. Infect. Immun. 60, 1946-l 95 1. 3. Austin, J. W., Doig, P., Stewart, M., and Trust, T. J (1992) Structural comparison of urease and a GroEL analog from Helicobacter pylorz J. Bacterlol 174, 7470-7473. 4 Dunn, B E., Campbell, G. P., Perez-Perez, G. I, and Blaser, M J. (1990) Purification and characterization of urease from Helicobacter pylon. J Blol Chem 265,94&l-9469. 5 Hu, L. T. and Mobley, H. L. T. (1990) Purification and N-terminal analysis of urease from Hebcobacterpyion Infect Immun S&992-998. 6. Mobley, H. L T , Cortesia, M. J , Rosenthal, L. E., and Jones, B. D (1988) Charactenzatlonof ureasefrom Campylobacter pylon. J Clw Mlcroblol 26,83 l-836. 7. Dolg, P , Austin, J. W., Kostrzynska, M , and Trust, T J (1992) Production of a conserved adhesion by the human gastroduodenal pathogen Hekobacterpylori. J Bactenol.

174,2539-2547

8 Doig, P., Austin, J W., and Trust, T. J. (1993) The Helrcobacter pyZorz 19.6 kilodalton protein IS an iron conlaming protein resembling femtm. J Bacterial 175,557-560.

9. O’Toole, P., Logan, S. M., Kostrzynska, M , WadstrBm, T., and Trust, T. J (1991) Isolation, biochemical characterization and molecular analysis of a species speclfic protein antigen produced by the gastric pathogen Hellcobacter pylon. J Bactenol.

173, 505-513.

10. Doig, P. and Trust, T. J. (1994) Identification of surface-exposed outer membrane antigens of Helxobacterpylorz Infect. Immun 62,4526-4533 11. O’Toole, P. W., Kostrzynska, M., and Trust, T. J. (1994) Non-motile mutants of Helicobacterpylorl and Hellcobacter mustelae defective m hook production Moi Mxrobiol.

14,691-703.

12. O’Toole, P W., Austin, J. W., and Trust, T. J. (1994) Identification and molecular characterization of a maJor ring-forming surface protein from the gastric pathogen Hebcobacter mustelae Mol Mxroblol. l&349-36 1. 13. Exner, M. M., Doig, P., Trust, T. J., and Hancock, R E. W. (1995) Isolation and characterizationof a family of porm proteins from Helzcobacter pylorz. Infect Immun. 63, 1567-1572. 14 Walker, J. M. (1984) Nucleic acids Methods ln Molecular Biology, vol. 2 Humana, Clifton, NJ

Characterization of Surface Antigens

189

15. Walker, J M. (1984) Protezns Methods zn Molecular Bzologv, vol. 1 Humana, Chfton, NJ. 16 Sambrook, J., Frrtsch, E. F , and Maniatis, T. (1989) Molecular Clonzng A Laboratory Manual Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 17. Laemmli, U. K (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227,680-685. 18. Pitcher, D. G., Saunders, N. A., and Owen, R. J. (1989) Rapid extraction of bacterial genomrc DNA with guanidium thiocyanate. Lett Appl Mzcrobzol. 8, 15 l-l 56. 19. Labigne, A., Courcoux, P., and Tompkins, L. (1992) Cloning of Campylobacter Jejune genes required for leucme biosynthesrs, and constructron of a leu-negative mutant by shuttle transposon mutagenesis. Res Mcroblol. Inst Pasteur 143, 15-26. 20. Osborn, M J., Gander, J. E., Parisi, E., and Carson, J. (1972) Mechanisms of assembly of the outer membrane of Salmonella typhzmurzum: tsolatton and characterization of cytoplasmic and outer membrane. J. Blol. Chem. 247,3962-3972 2 1 Bradford, M. M. (1976) A raped and sensitive method for the quantrtation of mlcrogram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72,248-254.

22 Hultman, T., Bergh, S., Moks, T., and Uhlen, M (1991) Bidirectional sohd phase sequencing of zn vitro amplified plasmid DNA. Biotechniques 10, 84-93 23. Page, W. J. and Taylor, D E. (1988) Comparison of methods used to separate the inner and outer membranes of cell envelopes of Campylobacter species. J Gen. Microblol

134,2925-2932

Isolation and Characterization of Porins from Ii. pylori Robert E. W. Hancock and Maurice Exner 1. Introduction The outer membranes of Gram-negative bacteria represent selective, permeability barriers to environmental molecules. This function is accomplished in two ways. First, outer membranes exclude many larger hydrophilic molecules, including enzymes and other proteins, and most hydrophobic ones, by virtue of their unique composition. It has been demonstrated m Escherichia coli that outer membranes are asymmetric bilayers containing a unique species of glycolipid, lipopolysaccharide (LPS), in its outer leaflet, and phospholipids in its inner leaflet. The highly negatively charged LPS is stabilized by divalent cations. The observations that Helicobacter outer membrane proteins are Triton X-l 00 insoluble in the absence of EDTA or NaCl(1,2), that Helicobacter contains substantial amounts of LPS (3), and that Helicobacter is resistant to highly hydrophobic antibiotics, such as trimethoprin and nalidixic acid (4), lead one to believe that this organism is thematically similar. Second, selective permeation of small hydrophilic molecules through the outer membrane occurs through the water-filled channels of protems, termed porins. Such proteins have now been identified in a wide variety of bacteria, including Helicobacter pylori. The term, porin, has been somewhat misused to describe proteins with the biochemical features of known porins, including heatmodifiable behavior on sodium dodecyl sulfate-polyacrylamide gel electroporesis (SDS-PAGE). However, we use the term here to denote only those proteins with demonstrated ability to reconstitute channels in bilayer membranes. Porins are perhaps the best-characterized outer membrane proteins. Their major function is in passage of small molecules through the outer membrane, From Methods m Molecular Medune, Helioobacter pylort Protocok Edlted by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

191

NJ

192

Hancock and Exner

although they fall into two main categories, the nonselective and so-called specific porms. These two classesof proteins differ by the demonstrated existence in the latter of a specific substrate binding site, which, at low concentrations, substantially accelerates passageof this substrate through the outer membrane. However, the distinction between these two classes has become blurred by some recent observations: (1) the Rhodobacter porm, previously termed nonspecific, was cocrystallized with an unidentified substrate in its channel (.5,6); (2) the sucrose-specific porm of E. coli could also function as a general porm (7); and (3) the lmipenem/basic amino acid porin OprD of Pseudomonas aerugiaosa demonstrated homology to the nonspecific porin superfamily (8). Porms have been demonstrated to have a variety of roles in pathogenesis. These include a general (9) or specific (10) role m the passage of antibiotics mto cells and a correspondmg role in antibiotic resistance, as demonstrated when porins are deleted by mutation. They also possess nnmunomodulatory properties, including the ability to mitogenically activate B-cells, modulate host cell functions, and induce cytokine release (11-13). In addition, porms have been demonstrated to have potential as vaccine components. Thus the study of porms m Helzcobacter is a valid pursuit. Our own studies were initiated in an attempt to clarify the known difficulties in treating Helicobacter infections m VIVO,despite its apparent high degree of in vivo susceptibility. This discrepancy has been variously explained for subsets of antibiotic classesas being caused by reduction of antibiotic activity at low pH (I#), influence of slow growth in vivo on antibiotic susceptibility (15), low bioavailability of certain antibiotics m the upper gastrointestmal tract (16), development of antibiotic resistance (I 7,18), recrudescenceof infection (19), or poor patient compliance because of the requirement for long periods of treatment (20,22). However, few basic studies have been performed on the mechanisms of antibiotic uptake and action m Helzcobacter; without such studies, the fundamental underlying reasons for reduction in antibiotic effectiveness m vlvo cannot be properly investigated. The following methods describe procedures for porin isolation. Our laboratory has had years of experience in porin isolation and characterization from numerous species. Thus, it is of some mterest that, in our hands, methods that permitted the successful purification of crystallizable porin using starting materials that contained at least three other contaminating porms (22), were unsuccessful for H. pylori. The described method is generally applicable for almost any porin and for most outer membrane proteins. It relies on the fact that porms form highly stable P-barrel structures, containing, in the known structures, 16 or 18 transmembrane P-strands (23-25). Such structures will not easily renature once disassembled. However, they are highly resistant to attack by even the most powerful detergents (i.e., sodium dodecyl sulfate [SDS]),

Isolation/Characterization of Porins Table 1 Porins of tielicobacter

193 pylori

Molecular mass of monomer, kDaa HopA HopB

48 49 50 67 31

HopC HopD HopE

Conductance, nsb 0 36 0.36 0.30 0.25 1.5

aMolecular mass refers to the apparent molecular mass observed when heat-denaturedsamples are run on SDS-PAGE bAverage single channel conductance of the monomeric pro&m m 1 OMKCl

since detergents cannot penetrate the ordered structure of porins. This gives rise to the property of heat-modifiability, by which porins migrate in a folded configuration on SDS-PAGE after solubihzatlon at low-to-moderate temperatures m SDS, binding SDS only on the periphery of the porin molecule. If the nonheated porin runs as a native trimer (the usual oligomeric state of porms), it will appear at a higher apparent molecular mass or lower mobility than the monomeric, fully dlssoclated porin that has been preboiled in SDS. For some porins the primers can be disassociated into monomers after treatment in SDS at room temperature without affecting the P-barrel structure of the monomers. Alternatively, they might actually exist as monomers m the outer membrane. Such monomer porins run at a lower apparent molecular weight than when fully dissociated because of their more compact structure, and these proteins require more extensive heating to unfold them (i.e., 95°C for 20 min) (2). The above properties permit one to utilize detergents and SDS-PAGE as purification procedures. The following procedures have permitted purification of milhgram levels of five porins (HopA, HopB, HopC, HopD [I], and HopE [26]) from H. pylorz (Table l), and nearly half a gram of OprP porin from Pseudomonas aeruginosa. 2. Materials 2.1. Pork Isolation

Procedures

2.1.1. Isolation of Membrane Fractions 1 Brain-heart infusion (BHI) agar (Accumedia, Baltimore, MD). 2. 1% hemoglobin powder (Accumedia). 3 20% (w/v) sucrose, 10 mMTns-HCl, pH 8.0, containing 50 pg/mL of deoxynbo-

nuclease1 (Sigma, St.Louis, MO). 4. 70% (w/v) sucrose, 10 mM Tris-HCl, pH 8.0.

194 5 6 7. 8 9. 2.12.

Hancock and Exner 18% (w/v) sucrose, 10 rnA4 Tris-HCl, pH 8.0. Reagent A: 2% Na+ZO,, 0 02% NaK Tartrate, 0. lMNaOH, 1% SDS (added last). Reagent B: 0.5% CuS04. Reagent C: Mix 25 mL of reagent A with 1 mL of reagent B. Make fresh dally Folm Ciocalteu Reagent (BDH, Toronto, Ontario).

ldenfification

of Heat Modifiable

Proteins

1. 11% running gel (10 mL): 4.2 mL H20, 2.75 mL 40% (w/v) acrylamide (29.1 acrylamide/bis) (Bio-Rad, Hercules, California), 2 5 mL 1.5MTris-HCl, pH 8 8, 0 18 mL 10% (w/v) SDS, 0.16 mL 5MNaC1,O 07 mL 200 mM EDTA, 0 1 mL 10% ammonium persulfate, 0 04 mL TEMED. 2. 4% Stacking gel: (10 mL) 5 36 mL H20, 1.0 mL 40% acrylamlde, 2 5 mL 0.5M Tris-HCl, pH 6 8, 1 0 mL 10% SDS, 0 1 mL 10% ammomum persulfate, 0 04 mL TEMED. 3 10X SDS-PAGE running buffer: 30 g Trts-HCl, 10 g SDS, 144 g glycme. Make up to 1000 mL with distilled water. 4 2X Solubihzatlon mix. 2% SDS, 12 mMTris-HCl, pH 8.0, 10% glycerol, 50 mM EDTA/lOO mg bromophenol blue. 5 0 375MTris-HCl, pH 8 6 6 Coomassie blue stain: 90 mL methanol, 20 mL acetic acid, 90 mL water, 200 mg Coomassre brillrant blue R250 7. Destain solution 290 mL distilled water, 80 mL methanol, 30 mL acetic acid

2.1.3. Selective Solubilization and Column Chromatography

of Porins

1. 10 mMTris-HCI, pH 8.0,l .O% Triton X- 100 (Sigma, St. LOUIS, MO), 2 mMMgCl* 2. 10 MTris-HCl, pH 8.0, 0.5% sodium lauryl sarcosmate (Sigma). 3. 10 mM Tris-HCl, pH 8.0, 3% octylpolyoxyethylene (OPOE) (Bachem Bioscience, Philadelphia, PA). 4. 10 mA4 Tris-HCl, pH 8.0, 0.08% iV,N-dtmethyldodecylamme-N-oxide (LDAO) (Fluka Chemika, Ronkonkoma, NY) 5. 1.OM NaCI.

2.1.4. Gel Purification 1. SDS-PAGE reagents (see Section 2 1.2.). 2 Coomassie blue stain. 3. 10 mM Tris-HCI, 0.08% A’,&dtmethyldodecylamine-N-oxide, 150 mMNaC1

1 rnJ4 EDTA,

3. Methods 3.7. Porin kfenfificefion and Isolation Procedures 3.1.7. Isolation of Membrane Fractions 1 Grow H. pylon for 3 d at 37°C m an atmosphere of 10% CO, on BHI agar that is supplemented with 1% hemoglobin after autoclaving. In our hands, this supported

Isolation/Characterization of Porins

2

3 4. 5

6

7 8 9. 10.

11.

195

growth as well as, or better than, traditional growth procedures including nncroaerobic growth with a CampyPak system (Becton Dickmson, Cockeysville, MD) or growth on BHI agar with 5% sheep blood, or on chocolate agar (see Note 1). To gain large quantities of starting materials, two hundred 100 x 15 mm plates are each streaked out from plates grown for 2 d under the above conditions (see Note 1) Harvest cells by scraping off the plates with a cotton swab, and resuspend m 20% (w/v) sucrose, 10 mM Tris-HCI, pH 8.0, containing 50 pg/mL of deoxyribonuclease I. The volume of sucrose should be approx 20 mL, and this should be adjusted so that the final cell resuspension 1sthick and VISCOUS. Disrupt the pooled cell suspension using a French pressure cell by two passages at 15,000 psi (see Note 1). Remove unbroken cells by centrifugation at IOOOg for 10 min. Layer the supernatant in IO-mL quantities onto a sucrose gradient comprrsmg (from bottom) 10 mL of 70% (w/v) sucrose and 20 mL of 18% (w/v) sucrose in a 40-mL centrlfugatton tube, and centrifuge at 65,000g for 4 h, or overnight Collect the membrane fraction, which appears as a whitish opaque band at the interface of the 18 and 70% gradients, by piercing the bottom of the tube with a needle and collecting the opaque fractions. Add water to dilute out the sucrose, approx 4- to 5-fold. Centrifuge the diluted fractions at 250,OOOg for 1 h at 4’C 1n a Beckman 60T1 rotor or eqmvalent. Resuspend the pellet after centrifugation at about 20 mg protem/mL in 10 mA4 Tris-HCI, pH 8.0 (see Note 3) Protein yields are determined by a modified Lowry assay that contains detergent (27), and, rn our hands, gives more accurate assays of membrane protein concentrations. Starting with the cells harvested from 200 plates, isolate 850 mg of protein

3.1.1.1.

PROTEIN ASSAY

1. Add 5 pL and 10 pL of a I-10 dilution of each protein sample to a 13-mrn borosilicate glass tube, and prepare a standard curve using additions of 0, 5, 10, 15, 20, and 25 pL of 0.1% BSA (in water). 2 Add 1 mL of reagent C, and let tubes stand at room temperature for 15 mm. 3 Add 100 pL Folin Ciocaloteu reagent, vortex immediately, and let stand at room temperature for 30 mm. 4. Read absorbance at 650 nm against a reagent blank

3.1.2. /den tifica tion of Heat Modifiable Proteins As drscussed in the introduction, porins generally comprise heat modifiable proteins. To examine heat modifiable proteins 1n H. pylori, two-dimensional (unheated vs heated) SDS-PAGE is utilized. It was necessary to do this, since, generally speaking, there are no freely available assays to follow por1n punfication. Thus, SDS-PAGE mobility and heat modifiable behavior are general criteria that can be monitored during purification.

196

Hancock and Exner

94 67 43 30 20

Fig. 1. Two dimensional SDS-polyacrylamide gel identifying heat modifiable proteins. Spots appearing left of the diagonal represent dissociated aggregates; spots appearing to the right of the diagonal are proteins that have become denatured only after heating. Spots 1,2,3, and 4 represent HopA, B, C, and D, respectively. Molecular masses (kDa) are indicated on the right. Reproduced with permission from ref. I.

1. Solubilize samples containing 30 pg of membrane proteins at room temperature (23°C) in SDS-containing solubilization-reduction mix and load into 6-mm wells on a 10 x 150 x 1.25 mm 4% SDS-polyacrylamide stacking gel with a 100 x 150 x 1.25 mm 11% running gel (see Note 2). The gel is run at 100 V for 4-5 h, using a conventional discontinuous buffer system (28). 2. Excise a vertical gel strip comprising a single lane, equilibrate in 0.375M Tris-HCl, pH 8.6, for 1 min, and then wrap in cellophane and heat to 95°C for 20 min in the same buffer (see Note 3). 3. Remove the strips from the cellophane and overlay on a second SDS-PAGE gel that has been poured so that it contains only an 11% separating (running) gel and not a stacking gel (see Note 3); initiate electrophoresis by applying a constant voltage of 100 V. 4. Place gels in a clean tray and stain with Coomassie blue stain for 10 min. Following staining, place the gel in destain solution and agitate. Frequent changes of the destaining solution allows for more rapid visualization of the bands. 5. Proteins that do not demonstrate heat-modifiable behavior run identically in both dimensions and appear on a diagonal running from the top left-hand to bottom right-hand corner. Off-diagonal spots to the left of the diagonal are oligomer-forming proteins. Off-diagonal spots to the right of the diagonal are proteins that retain at least part of their folded structure in the presence of SDS at low temperature. A typical result for H. pylori is shown in Fig. 1,

3.1.3. Selective Solubilization and Column Chromatography

of Porins

1. Suspend samples containing 100 mg of membrane proteins from Section 3.1. I. to a final volume of 10 mL in 10 nnI4 Tris-HCI, pH 8.0, containing 1% (v/v) Triton

Isolation/Characterization

2.

3. 4. 5. 6. 7. 8.

197

of Porins

X-100 and 2 rmI4 MgCl, at room temperature. Use an 18-22-gage needle and syringe to make resuspension easier. To ensure complete suspension, sonicate the sample briefly (15 s) with a narrow probe sonicator (Fisher dismembranator [Pittsburgh, PA] or equivalent at a power giving use to cavitation-formation of small bubble in the suspension). Centrifuge the suspension at 180,OOOg for 1 h on a Beckman 60Ti rotor (Fullerton, CA) or equivalent. Remove the supernatant and resuspend the pellet, as above, in 10 mL of 10 nnI4 HCl, pH 8.0,0.5% (w/v) sodium lauryl sarcosinate. Centrifuge samples as per step 3. At this stage, the pellet contains largely outer membrane protems. Resuspend the pellet as above in 5 mL of 10 mA4Tris-HCl, pH 8.0,3% octylpolyoxyethylene (OPOE), and centrifuge at 180,OOOg for 1 h. Save the supernatant; repeat step 6 and compare the supernatant from thts step with the previous OPOE-soluble protein preparation. Separate the OPOE-soluble samples by fast protein liquid chromatography (FPLC) on a Mono Q HR 5/5 anion exchange column (Pharmacia, Toronto,

Ontario), with elution using a salt gradient of 0-1MNaCl m a column buffer of 10 mM Tris-HCl, pH 8.0,0.08% N, N dimethyldodecylamme-N-oxtde

(LDAO)

(seeNote 4). Run portions of all fractions collected on SDS-PAGE and collect fractions containing

the heat-modifiable

proteins

of interest and pool as

appropriate. 3.1.4. Gel Purification 1. Solubilize FPLC fractions in solubilization mix at room temperature for 10 mm. 2. Load a solubilized sample containing up to 500 pg of protein mto a single preparative well (see Note 4) and run on an 11% SDS-PAGE gel. 3. Excise thin, vertical gel slices from the left and right sides of the gel and stain

with Coomassieblue. Use thesestainedstrips as a guide to precisely excise,by use of a razor blade, bands corresponding to the proteins of interest from the

unstainedgel (seeNote 5). 4. Crush gel slices are crushed and soak individually over mght at 4°C in 0.5 mL of a buffer consistmg of 10 mA4 Tris HCl, pH 8.0, 1 mM ethylene diamme tetraacetate (EDTA), 150 nnI4 NaCl, 0.08% LDAO (see Note 4). This can be done using a standard test tube on a tube roller, but agitation IS not absolutely necessary Aspirate the liquid with a pipet and filter to remove any remaining acrylamrde. The extraction process can be repeated rf the mmal extraction gives

a low yield of protein Store the eluted proteins are storedby freezing at -70°C.

3.2. Assay of Pore-Forming Activity A full description of the assay procedures utilized to assay porins is not within the scope of this chapter, since such assaysare highly technical and, for the system with the greatest utility (planar lipid bilayers), specialized apparatus is required. Instead, a general description is presented here.

198

Hancock and Exner

3.2.1. Planar Lipid Bilayer Methods This method has two major advantages and two potential disadvantages, compared to the liposome methods described in Section 3.22. One advantage is that virtually any porin can be assayed, regardless of its channel size. In addition, the method has single molecule sensitivity and requires extremely modest amounts of porin (ng) to yield a substantive amount of information about the channel formed by the porm. One potential disadvantage is that only about 1 in 10,000 molecules is actually assayed (i.e., enters the planar bilayer and forms channels). The porin must be extremely pure so that one has reasonable certainty that it is the purified protein species, rather than a contaminant, that is forming transbilayer channels. In practice, if a protein is 99% or more pure and forms channels at a concentration of 500 pg/mL, experience has taught us that it is this protein that almost certainly is forming the channels. A second concern is created by the fact that ion movement (i.e., current) is used to assess channel formation. It can be difficult to determine the exact size of biological molecules capable of passing through the channel (i.e., the exclusion limit of the porin), except by extrapolation to other channels, where both kinds of studies have been performed. The planar lipid bilayer method requires specialized apparatus (outlined below) and some training in its use. However, it is a very productive, rapid procedure for determining channel propertres. The apparatus is diagrammed in Fig. 2. The central part comprises a chamber that is machined from a 5 x 2.5 x 3 cm block to create two equal compartments separated by a 1-mm Teflon divider. One of these compartments contains a viewing window, and the Teflon divider is perforated by a 0.1-2 mm2 hole. The hole is anointed at its edges with a lipid solution (typically 1.5% oxidized cholesterol or 1.5% diphytanoyl phosphatidyl choline in n-decane), and dried under ajet of hot air to provide a surface to which a membrane can adhere. Each compartment is then filled with 6 mL of a salt solution (e.g., 1M KCl). Subsequently, the hole is wiped with a Teflon rod onto which 5 p,L of one of the above lipids has been pipeted. The lipid will cover the hole (assessed by measuring a high resistance when a voltage is passed across two electrodes dipping into the two compartments). Within a short time the lipid thins out until it forms a bilayer. This can be observed using a short focal length telescope in incident light coming from a suitable focused light source, such as a microscope light, since the observed lipid changes from multicolored to black, because of the optical properties of lipid globules and lipid bilayers, respectively. This gives the methodology its alternate name, black lipid bilayers. To the electrode dipping into the solution m one compartment, a direct current voltage source is attached. To the electrode in the other, a current amplifier (Ketthly 427, Cleveland, OH), oscilloscope to monrtor the amplified signal (Tektronix 5 11 A, Beaverton, OR), and rapid response chart recorder (Houston

199

Isolation/Characterization of Porins Faraday chamber 0.C voltage

source / n

storage 0scilloscope ctwt recorder Fig. 2. Schematic representation of planar lipid bilayer model membrane system. Electrodes connected to the voltage source and current amplifier are placed in an aqueous salt solution within the Teflon chamber, and a lipid bilayer is painted across the hole separating the chamber compartments. Conductance increases caused by the msertlon of porins into the bilayer are monitored on the oscilloscope and recorded on the chart recorder Reproduced with permrssion from ref 37.

Instruments 45 12, Austin, TX) are attached. With the naked membrane, appbcation of a voltage (typically

1O-50 mV) results in a very small current (approx

2 PA), smce bilayers have httle permeability to ions. If a porin in detergent solution 1s added to the compartment on one side of the membrane, it will become diluted to a very low detergent concentration, and because of its avrdity for membranous (lipid) environments, it will spontaneously incorporate into the membrane, forming a conduit for the movement of ions through the channel. This can be observed, after signal amplificatron, as a stepwise increase in conductance in both real time on the chart recorder and at a faster time resolution on the oscilloscope.

Some of the channel properties that can be easily checked by varying experimental parameters are whether the channel is water-filled, whether there is a strong selectrvity for cations over anions (or vice versa), the influence of lipid composition (usually none), whether the channel aggregates m the membrane, the effect of voltage on channel properties, whether the channel permits only unidirectional flux of ions, and the variability in the sizes of individual channels. Refer to the literature for exactly how these properties are studied

200

Hancock and Exner

(29-31). The major readout, however, is average single channel conductance of the channel m given salts, which is itself proportional to the volume of the channel and its geometry. The single channel conductance can range from 10,000-l 1,000 pS for mdividual porin channels (see Table 1 for Helzcobacter porm properties). With a slightly different setup in which only a voltage source and a multimeter (Keithly 610, Cleveland, OH) are connected to the electrodes, one can measure macroscopic conductance, and determine such properties as voltage dependence, selectivity for one ion over another, and whether the channel contams a specific binding site for a gtven substrate (see ref. 30 for procedures). 3.2.2. Liposome Procedures One can also utilize hposomal procedures to measure channel properties. These are useful for large channels like the OmpF porin of E. COZZ,but are technically challenging with many possible experimental pitfalls, as evidenced by disparities between two groups studying Pseudomonas OprF porin by one of these procedures, liposome swellmg (32,33). In addition, they are exceptionally difficult to perform with charged substrates, require high concentrations of soluble substrates, and cannot be utilized for small pores or specific porins, unless the substrate is known Nevertheless, m the hands of skilled practitioners, such methods do yield pore exclusion limits equivalent to those measured m VIVO.It must be emphasized that both these and the above procedures are models and must eventually be confirmed by in vtvo assessments. For details of how these procedures are performed, see refs. 31,34, and 35. 4. Notes 1 Other growth media work quite adequately,and it appearsthat the expressionof porin proteins is not dramatically changedby using different media (2) In addition, our preliminary data suggestthat the nongrowing coccoidforms yield porin preparations that are virtually indistinguishable from those obtained from the growing spiral forms One important factor is that the plates must not be dry or the cells will not grow It is useful to store agar plates at 4“C shortly after pouring

them so that condensationwill form, keeping the plates moist. Media supplemented with hemoglobin can be stored for 3 mo or longer at 4°C without showing decreased cell growth, although other media, such as that supplemented wtth P-cyclodextrin, do not appear to support cell growth if the media are a month or more old.

When growing large amountsof cells, tt may be preferable to use larger culture plates (150 x 15 mm or larger). To achieve sufficient growth on a plate, it is necessary to streak out a large number of cells; streaking out single colonies will not produce sign&ant growth. The cells from a single plate are usually only sufficient to streak out 1O-l 5 additional plates.

Isolation/Characterization of Porins 2. When disrupting ceils for membrane preparations, a somcator, or glass bead agitator, can be utilized instead of a French pressure cell although the efficiency of cell breakage is decreased. The efficiency may be increased by freezing the cells at -2O“C prior to disruption. Separate batches of cells can be frozen and then pooled together prior to disruption. 3. After cell membranes have been collected, they may be directly resuspended in a detergent solution if a solubihzation procedure using the given detergent is to follow. Larger gels (100 x 150 x 1.25 mm) tend to give better results, but mimgels (80 x 60 x 0 75 mm) may also be used. Larger gels facilitate the handling of gel strips, and it is easier to place the larger gel strip on the second separating gel without damaging it. When heatmg a gel strip, it is wrapped m cellophane, but precautions to make it completely water tight are not necessary, because small leaks will not cause a significant loss of protein from the gel. Heating must be carried out at 95°C for 20 min; trimeric porins may dissociate after heating to 65°C for 10 min, but heat modifiable monomers require further heating to ensure that they fully unfold. It should be noted that at least one heat modifiable porin, OprF of Pseudomonas aeruginosa, is quite stable to boilmg in the presence of SDS, and extensive heating is required to unfold this protein (36). Placement of the gel strip on a secondary separating gel can be achieved by sliding the gel strip between the glass plates, or by separating the glass plates, inserting the gel strip, and then repositioning the glass plate. The latter method has proven to be quite successful, since tt minimizes damage to the gel strip, and it enables a tighter contact between the gel strip and the separating gel. A stacking gel may be used for the second dimensional gel run, but experience has shown that this produces smears and streaks, instead of compact spots. 4 In our hands, OPOE is the detergent of choice for porin purification, since it yields porin of crystallizable quality (Egh and Hancock, unpublished results) It is quite expensive, however, and we replace it with NJ-dimethyldodecylamine&oxide (LDAO) for column chromatography, since this latter detergent has similar properties but a lower critical micellar concentration, such that a fourfold lower concentration can be utilized in column buffers to maintain proteins m solution. An important factor in obtaining reproducible solubihzation results IS the concentration of protein m the sample. The starting concentration m detergent should be 10 mg/mL. Changing this concentration will vary the amount of protein extracted in each solubilization step. A concentration of 10 mg/mL of detergent will enable the extraction of HopE in the Triton X-100~soluble fraction, while HopA, B, C, and D will be extracted in the OPOE-soluble fraction When fractions are resuspended during solubilization, a large amount of granular material will not resuspend, as the fractions contain insoluble material (possibly peptidoglycan fragments). However, this granular material must be meticulously agitated to release any soluble proteins bound to them. If this is not done, a considerable portion of the porin proteins may be found m the final pellet, but an extraction using 3% OPOE with 0.4M NaCl will usually solubilize the remaining porin.

202

HancockandExner

After detergent solubllizatlon, the samples can be further purified usmg FPLC, or they can be directly cut out of a preparative gel Samples for FPLC are normally dialyzed m the starting buffer, which contams 0.08% LDAO However, occaslonally, this results in the entire sample fluting in the void volume. This can be solved by leaving the sample m 3% OPOE, as this ensures that the porms will remain soluble and will enhance binding to the column. TypIcally, the porins cannot be purified after one column run, but if partially purified fractions are pooled and lyophlhzed, separation can be achieved This, however, may result m proteins that are not active m pore-forming assays. 5. As with the 2D gels, better results are usually achieved using larger gels, which permit better protein separation. When cutting unstained bands from a gel, it is often difficult to obtain a pure species, even when stamed sectlons are present as a gmde. It should be noted that after staining the guide slices m Coomassie blue, they may have to be placed m water to swell them to the onginal size, smce staining may dehydrate and shrink the gel. To improve the chances of obtaining a pure band, a number of very thin slices can be excised from the region where the band of interest is, as this decreasesthe chances of including other proteins. Similarly, changmg the acrylamide concentration can ald in separatmg proteins with snmlar molecular masses. When elutmg a protein from a gel slice, a greater yield 1sachieved if the gel IS crushed into small fragments. The volume of liquid used is dependent on the srze of the gel slice, but 0.5-l mL is usually sufficient. Followmg elutlon, the proteins may be concentrated by ultrafiltration or by preclpltatlon, and then they should be frozen at-7O’C. Any pore-forming assaysshould be done as soon aspossible after gel elution, as some gel purified proteins rapidly lose their pore-formmg ability. Freezing at -70°C usually does not affect ponn activity, and can ensure longer retention of activity

References 1. Exner, M M., Dolg, P. Trust, T J., and Hancock, R. E. W (1995) Isolation and characterization of a family of porm proteins from Helicobacter pylon. Infect Immun 63, 1567-1572. 2. Exner, M. M (1996) Ph.D thesis, University of British Columbia. 3 Geis, G , Leymg, H., Suerbaum, S , and Opferkuch, W. (1990) Unusual fatty acid substitution in lipids and llpopolysaccharides of Hellcobacter pylorl. J Clan Mlcrobiol

28, 930-932.

4 Goodwin, C. S., Blake, P , and Blincow, E. (1986) The mimmum inhibitory and bactericidal concentrations of antibiotics and anti-ulcer agents against Campylobacterpylondrs. J Antimxrob Chemother. 17,309-314. 5. Weiss, M. S. and Schulz, G. E (1993) Porin conformatlon m the absence of calcium. Refined structure at 2.5 A resolution. J Mel Biol 231, 817-824 6. Weiss, M. S. and Schulz, G. E. (1992). Structure of porm refined at I .8 A resolution. J Mel Biol 227,493-509. 7. Schulem, K., Schmid, K , and Benz, R. (1991) The sugar-specific outer membrane channel ScrY contains functional characteristics of general diffusion pores and substrate-specific porms. Mel Mlcroblol 5,2233-224 1.

Isolation/Characterization of Porins

203

8. Huang, H., Jeanteur, D., Pattus, F., and Hancock, R. E W. (1995) Membrane topology and site-specific mutagenesis of Pseudomonas aerigznosa porm OprD. Molec. Microbial. 16,93 1-941. 9. Hancock, R. E. W. (1987) Role of porins in outer membrane permeability. J. Bacterial 169,929-933. 10. Trias, J. and Nikaido, H. (1990) Outer membrane protein D2 catalyzes facilitated diffusion of carbapenem and penem through the outer membrane ofPseudomonas aeruginosa. Antimicrob. Agents Chemother. 34,52-57 11 Tufano, M. A., Rossano, F., Catalanotti, P., Liguori, G , Capasso, C., Ceccarelli, M. T., and Marinelli. P. (1994) Immunobiological activities of Hebcobacterpylori porms. Infect lmmun 62, 1392-1399 12. Sharma, P., Sharma, B K., Sharma, S., Rawal, I. J., Saxena, S. N., Pamgrahr, D., and Ganguly, N K. (1990) Mechanism of protection provided by active unmunization with porms in mice challenged with Salmonella typhi. Jpn J Exp. Med 60,247-252.

13 Tufano, M. A., Ianniello, R., Galdiero, M., Demartino, L., and Galdiero, F. (1989) Effect of Salmonella typhimurium porins on biological activities of human polymorphonuclear leukocytes Mtcrobiol. Path 7, 337-346 14. Sabath, L D., Lorian, V., Gerstein, D., Loder, P B , and Finland, M (1968) Enhancing effect of alkalinization of the medium on the activity of erythromycm against gram-negative bacteria. Appl Microbtol 16, 1288-1292. 15 Millar, M. R. and Pike, J (1991) Bactericidal activity of antimicrobial agents agamst slowly growing Heltcobacterpylori. Antimtcrob Agents Chemother 36, 185-187. 16. Veldhuyzen van Zanten, S. J O., Goldte, . J , Hollingsworth, J , Sillettr, C., Richardson, H., and Hunt, R H (1992) Secretion of intravenously admnustered antibiotics in gastric juice* implications for management of Helicobacter pylon. J Clin. Path01 45,225-227. 17. Seppala, K., Farkkila, M., Nuutinen, H , Kakala, K., Vaananen, H., Rautelin, H , and Kosunen, T. U. (1992) Triple therapy of Helzcobacterpylon infection in peptic ulcer A 12-month follow-up study of 93 patients. Stand J Gastroenterol. 27,973-976 18. Haas, C. E., Nix, D. E., and Schentag, J. J. (1990) In vitro selection of resistant Helicobacterpylori. Antimtcrob Agents Chemother 34, 1637-164 1. 19. Marshall, B. J., Goodwin, C S., Warren, J. R., Murray, R., Blincow, E. D , Blackbourn, S. J , Phillips, M., Waters, T. E., and Sanderson, C. R. (1988) Prospective double-blind trial of duodenal ulcer relapse after eradication of Campylobacter pylori. Lancet ii, 7-l 0. 20. Glupczynski, Y. and Burette, A. (1990) Drug therapy for Heltcobacter pylort infection: problems and pitfalls. Am J. Gastroenterol 85, 1545-1551. 21. Glupczynski, Y., Labbe, M., Van der linden, M. P., Nyst, J F., and Burette, A. (1989) Lack of antibiotic compliance m patients treated for Campylobacter pylon-assocrated gastritis. Am. J. Gastroenterol. 84, 1126 22. Worobec, E. A., Martin, N. L., McCubbin, W. D., Kay, C M , Brayer, G D , and Hancock, R E. W. (1988) Large-scale purification and biochemical characterrza-

HancockandExner

204 tion of crystallization-grade

porin protein P from Pseudomonas

aerugtnosa.

Btochtm. Brophys Acta 939,366374.

23. Weiss, M. S., Abele, U., Weckesser, J., Welte, W., Schiltz, E., and Schulz, G E. (1991) Molecular architecture and electrostatic properties of a bacterial porin Science 254,1627-1630.

24. Weiss, M. S., Kreusch, A, Schiltz, E , Nestel, U., Welte, W., Weckesser, J , and Schulz, G E. (1991) The structure of porin from Rhodobacter capsulatus at 1.8 A resolutton. FEBS Letts 280,379-382. 25. Schnmer, T., Keller, T. A., Wang, Y F , and Rosenbusch, J. P. (1995) Structural basis for sugar translocation through maltoporin channels at 3.1 A resolution Science 267,5 12-5 14 26. Doig, P., Exner, M. M., Hancock, R. E. W., and Trust, T. J (1995) Isolatton and charactertzation of a conserved porin protein from Heltcobacter pylort J Bactertol

177,5447-5452

27 Sandermann, H., Jr. and Strominger, J. L. (1972) Purification and properties of C 55 tsoprenoid alcohol phosphokmase from Staphylococcus aureus. J. Btol Chem 247,5123-5131. 28 Laemmh, U. K. (1970) Cleavage of structural proteins durmg the assembly of the head of bacteriophage T4 Nature (London) 227,680-X%5.

29. Benz, R. and Hancock, R. E. W. (1981) Properties of the large ion-permeable pores formed from protein F of Pseudomonas aeruginosa in lipid bilayer membranes. Btochtm Btophys. Acta 646,298-308. 30. Benz, R., Schmid, A, and Hancock, R. E. W (1985) Ion selectivity of gramnegative bacterial porms J Bacterzol 162,722-727. 31. Hancock, R. E. W (1986) model membrane studies of porin function, m Bacterial Outer Membrane as Model Systems (Inouye, M., ed.), Wiley, New York, pp. 187-225. 32. Yoshthara, E and Nakae, T (1989) Identificatton of porins in the outer membrane of Pseudomonas aerugznosa that form small diffusion pores. J Btol Chem 264,6297-630 1. 33 Yoshimura, F., Zalman, L., and Nikaido, H. (1983) Puriticatton and properties of Pseudomonas aeruginosa porin. J. Biol. Chem. 258,2308-23 14. 34 Nikaido, H and Rosenberg, E. Y. (198 1) Effect on solute size on diffusion rates through the transmembrane pores of the outer membrane of Escherrchza coli. Gen. Physiol 77, 121-135. 35 Nikatdo, H and Rosenberg, E Y. (1983) Porm channels m Escherzchta cob. studies with liposomes reconstituted from purified proteins J Bactertol. 153,241-252. 36. Hancock, R. E. W. and Carey, A. M (1979) Outer membrane of Pseudomonas aerugmosa: heat-2-mercaptoethanol-modifiable proteins. J Bactertol 140,902-9 10 37. Hancock, R. E. W. and Brown, M. H. (1992) Bacterial porins. Today’s Life Scz. 4, 24-32.

21 Methods for the Identification of tf. pylori Host Receptors Thomas Bon%, Torkel Wadstriim, Staffan Normark, Jeffrey I. Gordon, and Per G. Falk 1. Introduction Bacterial attachment to host receptors is a prerequisite for colonization of epithelial cell surfaces, in particular, continuously renewing mucosal surfaces, such as the gastrointestinal tract. Microbes express adhesion molecules for interactions with eukaryotic cell surface proteins or glycoconjugates, such as glycoprotems and glycoliptds (1). The combination of high receptor speciticity (2) and restricted receptor distribution will target bacteria to spectfic tissues, i.e., cell populations. This is referred to as tissue troptsm and partly determines the niche a bacterium is able to occupy. In addition, competition between bacterial species for space and nutrients selects for bacteria able to colomze specific niches. Bacteria unable to adhere to the epithelial cells and mucus lining will be exposed to the local nonspecific host defense mechanisms (such as peristalsis and turnover of the epithehal cell populations and the mucus layer) and eventually removed. The biological relevance of adherence as an initial step in the infectious process has focused interest to the structures involved in these processes. Bacterial adhesins and host receptors are both potential targets for novel antimicrobial drug design (3). Antimicrobial agents could be chemically coupled to soluble high-affinity receptor analogs and kill pathogens, such as H pylori, once they are targeted by the complex. Soluble receptor analogs would competitively interfere with bacterial attachment, utilizing the same mechanism as naturally occurrmg scavenger molecules m human secretions, such as milk and saliva. Receptor analogs could be developed for high-affinity mteractions and would thereby be efficient inhibitors at From Methods m Molecular Medrone, Hehcobacter pylon Protocols Edlted by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

205

NJ

206

Borhn et al.

low concentrations. Both drug targeting and competitive adhesion mhibition receptor analogs could exhibit a higher specificity for the pathogenic microbes, circumventmg the negative effects of broad spectrum antibrotics. It is currently unclear how the human- and primate-specific gastric pathogen H pylori functions, either as a primary causative agent or as a cofactor m the pathogenesis of chronic active gastritis, gastric and duodenal ulcers, and gastric adenocarcinoma. This reflects a lack of information about the nature and expression of the organism’s virulence factors and the gene products involved in the bacteria’s homing to gut epithelmm. In order to understand the molecular cross-talk between H pylori and the gastric epithelium, it will probably be necessary to recapitulate the chronic course of infection that is seen in humans. The use of genetically engineered mice for creating a model for studymg the impact of attachment on tropism and microbial pathogenesis has been elegantly demonstrated for poliomyelitis (4). For H. pylori, an animal model will have to be engineered m which the conditions that favor the human stomach as a natural site for bacterial colonization are mimicked as closely as possible. This involves, among other things, making sure that the optimal/ functional microbial receptors are available in the gastric epithelium (5). 7.7. H. pylori Receptor Characterization by Hemagglutination (HA) Characterization of potential microbial host receptors often requires a reliable supply of cells for binding experiments. Red blood cells (RISC) have been used for initial characterization of many microbial host receptors, based on aggregation (hemagglutinauon) reactions, since erythrocytes and epithelial cells exhibit similarities in their glycosylation patterns. In addition, since the cells are easy to handle, enzymatic treatments, such as protease and glycosidase treatments of the potential receptor structures, are feasible (see ref. 6 and references therein). RBC from individuals of different blood groups, as well as from different animal species,exhibit differences in their receptor architecture that are useful for dissection of fine-tuned differences in microbial receptor specificity. To further characterize potential receptor structures, hemagglutination inhibition assays(HAI) can be performed (7). This involves comparmg the results from the positive hemagglutination reaction with the results from assaysm which the bacterial cells have been preincubated with inhibitors. For instance, for inhibition of sialic acid-dependent hemagglutmation reactions, slalic acid-rich proteins such as fetuin and transferrin, could be used to block the bacterial adhesins. 7.2. H. pylori Receptor Characterization by Particle Agglutination Assay (PAA) When the results from the initial RBC experiments and the in sztuassaysare summarized, a putative receptor structure can often be defined. The particle

Identification of Host Receptors

207

agglutination assay (PAA) can then be used for the evaluation of such a receptor structure. The method has also been applied for rapid screening of surface adhesins of bacteria. In particular, bacterial binding to particles coated with receptor structures, such as fibrinogen, fibronectin, and immunoglobulins, have been monitored (15). This technique also has a clinical application, using particle-immobilized carbohydrate epitopes for the screening of uropathogenic Escherichia coli, and it was recently used to screen for surface adhesins of H. pylori (16). The principle for agglutination of latex particles and bacteria is the same as for hemagglutination. The latex particles coated with bacterial receptor structures will adhere to the bacterial adhesins and cause agglutination, During the coating process of the latex particles, the receptor molecules will passively absorb to the surface of the particles and the orientation of the molecules will be at random. Carboxylate-modified latex particles are commercially available for the linkmg of the receptor molecules covalently to the particle. The covalent binding will ensure an increased coating efficiency of substancesunable to absorb passively. In addition, the orientation of the receptor molecules can, to a certain degree, be standardized and predicted.

7.3. Screening of Host Receptor Distribution In Situ In situ screening of the microbial colonization patterns in target tissue is often guided by the set of clinical symptoms that characterize the bacterial pathogen. These provide clues to the distribution of potential receptors that mediate microbial adherence. Histochemical and immunohistochemical techniques, such as in situ hybridization and electron microscopy, can be used for the detection and localization of microbes in tissues.An alternative approach ISbased on the direct screening of potential receptor structures,using the microbes as probes. This was first demonstrated by the binding of fluorochrome-ConJugated E. colz to frozen sections of human kidney (8). An adaptation of this technique (9) utilizes the bacteria as probes to study the distribution of receptor epitopes in fixed histological sections of potential target tissues. The advantage of this model is that it allows for a direct evaluation of the cellular distribution of receptor molecules, It can also be used to characterize receptors directly in the tissue, by performmg double labelings (Fig. l), biochemical modifications, and inhibition assays(IO), as outlined in Section 1.4. This approach has been used to identify fucosylated blood group antigens as functional target tissue receptors for H pylori (Fig. 1 and ref. II), and the membrane cofactor protein (CD 46) as a keratinocyte receptor for Streptococcuspyogenes, causing infections in human skin (22).

1.4. Receptor Characterization

and IdenfMcafion

In Situ

Once a tissue-specific microbial binding pattern is established in situ, the receptor identity remains to be revealed (6). As an initial approach, periodate

208

Fig. 1. Immunocytochemical

Borh

et al.

stainings showing the coexpression of receptors for

H. pylori and fucosylated epitopes in surface mucous cells of the human gastric mucosa. A section of human stomach was overlaid with FITC conjugated H. pylori staining green

(top panel). In addition, a monoclonal antibody directed against the blood group H antigen in combination with TRITC conjugated rabbit antimouse antibodies staining red (middle panel) was used. Inspection of the section in fluorescence microscope through an FITC filter reveals the bound bacteria in green; the TRITC filter shows the staining of

Identification of Host Receptors

209

oxidation (13) can be used to determine whether the bacteria bmd to protein- or carbohydrate-based receptor epitopes. This treatment can be performed directly on cells or tissues and will cleave carbon-carbon linkages carrying vicinal hydroxyl groups, as are found in the ring structure of many sugar residues, while leaving peptides intact. Furthermore, by adjusting the conditions, it is possible to discriminate between sialylated and nonsralylated glycoconjugate receptors (Fig. 2 and ref. 10). Protease pretreatment of the tissue sections will remove most protein-based receptor structures, including glycoprotein receptors, and can be used as a further step to discriminate between protein vs carbohydrate receptors or glycoprotein vs glycolipid receptors, since the glycoliplds will be retained after protease digestion (6,10,24). Inhibition experiments, in situ, are performed by analogy with the HAI, to help in the characterization of the molecular structure of the binding epitope. This is most conveniently done by bacterial inhibition experiments, i.e., preincubatlon of bacteria with defined soluble proteins or glycoconjugates prior to overlay on tissue sections or cells, to prevent attachment in a competitive fashion. Like restriction enzymes, glycosidases are very substrate specific and will remove only the carbohydrates corresponding to their specificity. Some glycosidases are, in addition, specific for the glycosidic linkage, such as fucosidases with restricted specificity for a 1-3/4-linked fucose (Takara Biochemical, Berkeley, CA), while fucosidase from beef kidney (Boehringer Mannheim, Mannheim, Germany, or Sigma, St. Louis, MO) exhibit a broad specificity for a-linked fucose. Loss of binding activity, after fucosidase treatment of a receptor-active glycoconjugate, indicates a vital role for fucose residue(s) in the interaction. Neuramimdases, with a spectficity for sialic acid residues, also exhibit some differences m linkage-specitictty, although neuramimdase from Vibrio cholerae recognizes o2,3-, a2,6-, and a2,8-linked terminal sialic acid. In addition to traditional immunohistochemical analyses, for which several protocols for single and multi-labeling of histological sections have been published (Z 7), detection of carbohydrate antigens can be made on intact gastrointestinal tract tissue, i.e., wholemounts (18) This technique will provide valuable information about the spatial and cellular dlstribution of different glycoconjugates, such as microbial receptor structures.

1.5. Characterization of Microbial Receptor Structures by Solid-Phase Analyses In vitro analyses of the glycolipids and glycoproteins, prepared from whole intestinal tissues or epithelial scrapings, make it possible to determine the the fucosylatedblood group antigensin red. Double exposure (lower panel) shows that surfacemucouscells coexpressthe bacterial cell receptorsandfbcosylatedblood group antigens (yellow because of green-stammg bacteria supenmposmg red stained antibodies).

210

Bon% et al. Dog stomach Human stomach Ulexeuropaeustype1 (UEAl) Sambucusnigra(SNA) Helicobacterpylori

periodate,pH5.5, 10min.,t4“C

periodate,pH4.5,l hr. RT

bufferctrl.

periodate,pH4.5,I hr, RT

/den tifica tion of Host Receptors

211

detailed distribution of specific oligosaccharrdes among different lipid and protein caners. These preparations also serve as sources of material to be used for detailed structural analyses of compounds identified through the initial immunological and biochemical studies (6). Glycoconjugates with potential receptor epitopes are most conveniently screened for bacterial binding m solidphase assay systems.Solid-phase analyses are performed by bacterial overlays to immobilized proteins/glycoproteins on Western blots, and to glycoliptds separated on silica gel high performance thin layer chromatograms (HPTLC). 2. Materials 2.1, Materials

for Hemagglu tinin Assays

1, Phosphate buffered saline (PBS): 150 mMNaC1, 10 mMsodmm/potassmm phosphate, pH 7.6. 2. H. pylori incubated on blood agar plates, supplemented with Isovitalex (BBL Microbiology Systems, Cockeysville, MD) and antibiotics, under mtcroaerobic conditions (CO2 lo%, O2 5%) for 2 d (see Chapter 2). 3. Erythrocyte suspensions of bovine, chicken, guinea pig, horse, monkey, mouse, pig, rabbit, and sheep origins were used, in addition to human erythrocytes of blood groups A, B, and 0. 4. Protease inhibitors; 1 pg/mL leupeptin, 1 pg/mL aprotinin, 1 pg/mL pepstatm, 100 pg/mL L- 1-chloro-3-(4-tosylamido)-4-phenyl-2-butanone (TPCK), 1 mM phenyl methyl sulfonyl fluoride (PMSF), 5 mM EDTA (all protease mhtbitors are available from Boehringer Mannheim). 5 TBS: 5 mMTris-HCl, pH 7.4/8-O, 150 mMNaC1. 6. Clostndwm perfringens neuraminidase (Sigma).

Fig. 2. (opposite page) Immunocytochemical stainmgs summarizing the effects of periodate oxidation under various condittons on sialylated and fucosylated sugar epitopes (as illustrated by lectin staining), and H. pylon attachment. a-l-fucose-specific UEA 1 staining to dog gastric mucosa is resistant to NaI04 oxidation when this is performed at pH 5.5, for 10 min on ice (5rst lane, middle panel), but sensitive to the harsher oxidation condition, pH 4.5, 1 h at room temperature (first lane, lower panel) (see Note 12). The sialic acid-specific SNA staining of dog stomach is abolished by the milder oxidation (second lane, middle panel). The control (second lane, lower panel) shows that lectin staming is unaffected by the procedure in the absence of NaIO+ The binding of H. pylort to human gastric mucosa (third lane) is resistant to the siahc acid-specific oxidation conditions (middle panel), but 1s eliminated under the harsher conditions that will also remove UEAl staining (lower panel) However, periodate oxidation wtth 10 mM NaIO, in 0. IM NaAc, pH 4.5, for 1 h at room temperature, as described in Section 3.4.1.) destroys the carbohydrate-based receptor

epitopes for lectins in surface mucouscells, but leavesthe pepttde epitopes, such as the intrinsic factor protein, intact (10).

212

Bor& et al.

2.2. Materials 1 2. 3. 4

for Particle Agglutination

Assays (PAA)

Latex beads: 0.8 1 @4 diameter (Difco Laboratories, Detroit, MI). Glycine buffer: 0.17M glycine-NaOH, pH 8 2. Carboxylate modttied latex (CML)* 0.85 wdiameter (Seradyn, Indianapolis, IN). l-ethyl-3 (3-dimethyl ammo propyl) carbodimide (EDPC) (Sigma).

2.3. Materials for The In Situ Adhesion

Assay

1. Labeling buffer. 0.2M carbonate buffer, pH 9.2 (10% Na&Os, 90% NaHCO& 2. Labeling tags: 10 mg/mL fluorescein isothtocyanate (FITC), tetramethylrhodamme tsothrocyanate (TRITC) (both from Sigma), or N-hydroxysuccimmide ester activated dtgoxigenin (Dig-NHS) (Boehrmger Mannheim), or biotin (Biotin-XX-NHS) (Calbtochem, La Jolla, CA), freshly prepared m dimethylsulfoxtde (DMSO). 3. Blockmg buffer: PBS, pH 7.6,O 05% Tween-20, 1% bovine serum albumin (BSA). 4. Xylene or Hemo-De (Fisher Scientific, Pittsburgh, PA). Hemo-De is a watersoluble orange oil-based alternative to the volatile solvent xylene 5. Isopropanol (Merck, Darmstadt, Germany). 6. PAP-Pen: (Research Products International, Mt. Prospect, IL) 7 Protease mhibttor mix; 1 mM PMSF, 1 clg/mL leupeptm, 1 mg/mL pepstatin (all three from Boehrmger Mannhelm), 5 mM EDTA, and 10 mA4 benzamidmeHCl (Sigma) 8. Fluorochrome conjugated antrdigoxigenm Fab-antibodies (Boehringer Mannhelm) diluted 1: 100, or streptavidin (Jackson ImmunoResearch Laboratories, West Grove, PA), diluted 1: 1000 m blocking buffer. 9 Microscopy cover glass VWR Scientific Corp , San Franctsco, CA.

2.4. Materials for Periodate Oxidation, Protease Treatment, and Preparation of Intestinal Wholemounts 1 2 3. 4. 5.

Buffer 1. O.lMNaAc, pH 4.5. Buffer 2. 10 tiperiodate, O.lMNaAc, pH 4.5 Buffer 3: 50 mMNaBH4 in PBS, pH 7.6. Blocking buffer: PBS, pH 7.6, 0.05% Tween-20, 1% BSA. Protease inhibitors: 1 pg/mL leupeptin, 1 pg/mL aprotmm, 1 pg/mL pepstatin, 100 pg/mL L-I-chloro-3-(4-tosylamido)-4-phenyl-2-butanone (TPCK), 1 mM phenyl methyl sulfonyl fluoride (PMSF), 5 r&f EDTA (all protease mhibitors from Boehringer Mannhelm, diluted according to the manufacturer) 6. PBS: 150 mMNaC1, 10 mM sodmmpotassmm phosphate, pH 7.6. 7. Reducing solution: 20 tidithiothreitol (DTT), 20% ethanol, 120 mMTris-HCl, pH 8.2. 8. Blockmg buffer for wholemounts: PBS contaming 2% BSA and 0.3% Triton X- 100.

2.5. Materials for Glycopro tein/Glycolipid

Receptor

Preparations

1. Laemmh buffer containing protease inhibitors: 1% SDS, 2% P-mercapto-ethanol, 5% glycerol, 40 mM Tris, pH 6.8, 1 mg/mL leupeptm, 1 mg/mL aprotmm,

213

identification of Host Receptors

1 mg/mL pepstatin A, 100 mg/mL TPCK, 100 mg/mL PMSF, 1 mA4 EDTA (see Section 2.1.) Item 4). 2. Blocking buffer: Tris buffered saline, pH 7.4 (TBS; see Section 2-l., item 5), 1% Blocking reagent (Boehringer Mannheim).

3. Methods 3.1. Receptor Characterization 3.1.1. Hemagglutination Assay

by Hemagglutination

I. Wash titrated blood twice m PBS, pH 7.6. Dilute RBC from the pellet to 0.75% (v/v) in PBS. 2. Wash H. pylorz in PBS and suspend to a final concentration of 10” cells/ml (see Note 1). 3. Use twofold dilution series of the bacterial suspension for estimating the hemagglutinations, and for comparison of different erythrocytes and different bacterral strains. 4. Mix RBC and equal volumes of bacterial suspensions (25 pL) m wells of microtrter plates (U-bottom shaped, Nunc A/S, Roskilde, Denmark) and allow to settle at room temperature for 1-2 h. Use PBS as negative control. Estimate the titer of hemagglutination as one hemagglutination unit (1 HAU), 1 e , the smallest amount of bacterial cells that causes complete hemagglutination under the conditions described Hemagglutinations are estimated by the naked eye

3.1.2. Hemagglutination

Inhibition Assay (HAI)

1. Dilute bactertal cell suspensions with PBS to give 4 HAU Perform mhtbitlon tests by mixing 25 pL of bacterial cell suspension with 15 pL of a serial dilution of inhibitor in PBS, starting with 20 ug (fetuin, asialofetuin, glycophorm, asialoglycophorin, transferrin, bovine submaxillary mucin, and orosomucoid, all from Sigma), for 1 h at room temperature. 2. Then, add 25 pL of RBC 0.75% (see Section 3.1.1.) step 1) suspension to each well and allow to settle at room temperature for 1-2 h. The control wells contain PBS alone, to assure that inhibitors do not cause nonspecific agglutination (see Note 2).

3.1.3. Hemagglutination Inhibition by Glycosidase/Protease Treatments of Erythrocytes 1. Remove sialic acid by incubation of washed RBC wtth Clostridium perfrzngens neuraminidase at a concentration of 50 mU/mL for 20 min at 37°C in TBS. Wash RBC as in step 4 (see Note 3). 2. Perform trypsin and chymotrypsin treatments of erythrocytes essentially as described previously (14). Wash RBC in TBS, pH 8.0, and resuspend to 20% (v/v) The higher pH was used to activate the proteases. Add the proteases in concentrattons of 0, 0.05,0.1,0.2, 1.0, and 5.0 mg/mL. 3. After incubation at 28°C for 2 h, terminate the digestions by adding protease inhibitors (see Section 2.1 , item 4) and incubate for 15 mm (see Note 4).

Both

214

et al.

4. Finally, gently wash the RBC five times, with ice-cold TBS, 1 mMEDTA, 1 mM PMSF, pH 7.4, and use the same day. Process a control erythrocyte suspension as above, but without enzyme (see Note 5).

3.2. Receptor Characterization by PAAs 3.2.7. Coating Unmodified Latex Particles 1. Mix 1 mL of latex parttcles (Section 2.2., item 1) with 3 mL of glycme buffer (Section 2.2., item 2). 2 Centrifuge at 4500g for 5 mm. 3 Resuspend the pellet m 3 mL glycme buffer 4. Add 100 pg of protein, such as fetuin, to study the sialic acid specificity of H pylon, and incubate at 30°C for 12 h on a horizontal shaker. 5 Centrifuge at 9200g for 5 min 6 Remove the supernatant 7. Resuspend the pellet in 2 mL glycme buffer contaming 0.01% ovalbumin (Sigma) and 0.0 1% merthiolate (Merck) and keep at 4°C for 12 h before use.

3.2.2. Coating Carboxylate Modified Latex (CML) Particles 1. Wash 100 pL particles (see Section 2 2., item 3) in 1 mL O.lMsodmm phosphate buffer, pH 8.1, by centrifugation at 12,OOOgfor 10 mm. 2. Resuspend the pellet in O.lM sodium phosphate buffer, pH 8.1, containmg 2 mg/mL EDPC. 3. Incubate overnight at 4°C on an orbital shaker. 4. Wash twice in 0. 1M sodium phosphate buffer. 5 Suspend the particles in 1 mL sodium phosphate buffer containing 1 mg protein. 6. Incubate overnight at 4°C on an orbital shaker 7. Wash twice in 0. 1M sodium phosphate buffer. 8. Store at 4°C (see Note 6)

3.2.3. Particle Agglutination Assay (PAA) 1. Place 20 pL coated particles on a glass slide and add 20 p.L bacterial suspension (1 x log bacteria/ml) 2. Mix gently and read the agglutination reaction after 2 mm (see Note 7).

3.3. Screening of Host Receptor Distribution 3.3.7. H. pylori Labeling Procedure

In Situ

1. Wash freshly harvested bacteria twice (1 OOOgm an Eppendorf table top centrifuge, for 3 min) in 1 mL labeling buffer (see Section 2.3., item 1) and resuspend in 1 mL of the same buffer (see Note 8). 2. Add 100 ~18of the labeling tag (see Section 2 3 , item 2) in 10 pL DMSO, to the bacterial suspenston. Samples are incubated for 1 h at room temperature in the dark on a slow-rocker table (Nutator, Becton Dickinson, Sparks, MD) (see Note 9).

lden tifica tion of Has t Receptors

215

3. Wash bacteria twice m 1 mL blockmg buffer (see Section 2 3 , item 3), and dilute the bacterial suspension to a density of 1.O OD6c0, in blocking buffer, and freeze in aliquots (-20’(Z), m the same buffer.

3.3.2. Tissue Section Overlay Assay 1. Fix tissues in 10% formalm and paraffin embed; cut in 5-pm-thick sections, and affix to microscope glass slides, according to standard procedure (19). Series of consecutive sections should be used for binding analyses to minimize variations in tissue quality that could affect adhesion patterns. 2 Deparaffimze histological sections and rehydrate in a series of washes, startmg with 100% xylene or Hemo-De for 10 min, followed by another 3 mm separate rinse in this solvent, 3 x 5 min in isopropanol, 5 min m slowly running distilled water, and 3 x 5 min in PBS. Incubate the rehydrated sections in blockmg buffer (see Section 2 3., item 3) for a minimum of 1 h, or overnight m 4°C to reduce nonspecific binding. 3 In order to confine the bacterial suspension to the tissue, wipe the glass dry around the tissue and encircle the tissue with wax (PAP-Pen). It is crucial that the tissue never dries during the overlay procedures, since that could induce arttfactual binding patterns. 4 Dilute the labeled bacterial suspensions to 0.1-0.2 OD,,, in blocking buffer with protease inhibitor mix (see Section 2.3., item 7) and apply 200~pL aliquots to the prepared tissue sections. Overlay bacteria onto tissue sections for 1 h at room temperature, followed by six washes for 5 mm each m PBS/Tween, on a rotating table (see Note 10). 5. Detect dig/biotin-labeled bacteria with fluorochrome conjugated anttdigoxigenm antibodtes/streptavidm (see Section 2.3., item 8) that are incubated on the sections for 30-60 min at room temperature, followed by washing 3 x 10 min in PBS/Tween. Incubations with fluorochromes should be performed in the dark (under aluminum foil), m order to prevent fading of fluorescens. Protect the tissue by a microscopy cover glass mounted on the section with PBS/glycerol (1: 1, v/v) 6. Gentle handling is important, since touching the cover glass once posmoned on the tissue section could disrupt the bacterial binding 7. Assess distribution of bacteria adherent to cells using fluorescence microscopy. The fluorescence is stable for at least a month at 4°C in the dark (see Note 11).

3.4. Receptor Charecterizetion and identification In Situ 3.4.7. Periodate Oxidation of H. pylori Receptors In Situ 1. Wash deparaffinized and rehydrated tissue sections (see Section 3.3.2.) twice for 5 min in buffer 1 (see Section 2.4., item I), followed by incubation m buffer 2 (see Section 2.4., item 2), for 1 h at room temperature, in the dark. 2 Wash sections once for 5 mm, in buffer 1, and twice m PBS, pH 7.6. Reduce the tissue by immersion m buffer 3 (see Section 2.4., item 3), for 30 mm at room temperature, followed by 2 x 5 min washes in PBS. 3, Block and overlay sections with bacteria, as described in Section 3.3.2 (see Note 12)

Borkn et al.

216 3.42. Protease Treatment of H. pylori Receptors In Situ

1 Incubate deparaffnized and rehydrated tissue sections (see Section 3.3.2.) with 0.2 U of protemase K from Zkchrrutum albus (Boehrmger Mannheim) for 2 h at 37”C, in PBS. 2 Wash sections three times m PBS, and treat with blockmg buffer (see Section 2 4., item 4) with protease inhibitors (see Section 2 4 item 5), before overlay with H. pylorz (see Note 13).

3.4.3. inhibition of H. pylori Adherence In Situ by Soluble Receptor Structures 1. Premcubate bacterial suspensions with dilution series of l-100 pg/mL of soluble receptor structures for 1 h at room temperature. 2 Wash the bacterial suspension once in PBS, pH 7 6, to remove excess receptors, and add to sections as described m Section 3.3 2 3. Direct counting of the numbers of adherent bacteria at 20&400x magmfication is a convenient approach to evaluate reduction m bacterial binding compared to a nonmhibtted control (11) (see Note 14)

3.4.4. Fucosidase and Neuraminidase of Soluble Receptor Structures

Treatment

1 Incubate 100 ug of glycoprotem with 100 mU of bovine kidney a-L-fucosidase or Vibrlo cholerue neurammidase (both from Boehrmger Mannheim) for 2 h at 37T, in PBS, pH 7 6. 2. Inactivate glycosidases by incubation at 85°C Include untreated glycoprotein, incubated at 85”C, as control. 3. Use the glycoprotem preparations in inhibition experiments as described in Section 3.4.3. (see Note 15).

3.4.5. Preparation of Intestinal Wholemounts for Screening of Receptor Distribution 1 Remove the entire intestine en bloc and flush with cold PBS, followed by 10% phosphate buffered formyl salme (Fisher Scientific GP, San Francisco, CA) 2 Open the small intestine and colon by a longitudmal incision along the mesenteric side of the cephalocaudal axis of the gut. For the stomach, prepare strips from the major curvature to give a representative sample of the cellular composition along the entire length of the gastric eptthelium Mount the opened tissues on a wax or plastic support with pms. 3. Fix tissues in phosphate buffered formyl salme for 1 h at room temperature, followed by incubation m reducing (see Section 2 4., item 7) solution for 1 h at room temperature, to remove mucus. The tissue should be mtermtttently and gently flushed using a Pasteur pipet during this procedure. 4. If the carbohydrate marker (lectms, antibodies) is detected with horseradish peroxidase, incubate the tissues for 15 mm at room temperature m 1% H202 m block-

Identification of Host Receptors

5. 6.

7. 8.

217

ing buffer to reduce endogenous peroxidase activity. If alkaline phosphatase is used for detection, a Tris-based buffer must be used instead of PBS. Block tissues in blocking buffer (see Section 2.4 , item 8) for 30 min at room temperature. Overlay lectin/primary antibodies on tissues overnight at 4°C. Dilute reagents in blocking buffer. For lectins, a concentration of 5 pg/mL usually provides a good signal If a secondary antibody is used, add thts after the tissues have been washed five times, for 5 min each, in PBS or TBS, and incubate for l-2 h at room temperature. After staining with commercially available substrates for peroxidase (DAB) or alkaline phosphatase (e.g., BCIP/NBT), the tissues can be stored for several months m formalm.

3.5. Biochemical Preparation and Characterization of Microbial Receptor Structures by Solid-Phase Analyses 3.5. I. Glycopro tein Preparation 1, Sacrifice animals and immediately dissect and snap-freeze trssuesin liquid mtrogen Mince mesenchymal organs, such as liver, into smaller fragments prior to freezing. Flush intestinal segments with ice-cold PBS to remove lummal contents. If the goal is to recover epithehal material, the epithelium can be scraped of the serosal support using a razor blade. This will enrich for molecular entitles specific for epnheha1 cells. The composition of total fractions of proteins/glycoproteins and lipids/ glycolipids will differ distinctly if they are prepared from total tissues or eptthehal scrapings. Perform dissection and all handling of tissues on ice prior to freezing 2. Lyophihze deep frozen tissue. 3. For protein/glycoprotein preparation, homogenize the freeze-dried tissues to a powder with a glass rod. Dissolve the powder m a Laemmh buffer containmg protease inhibitors (see Section 2.5., item 1) to a concentration of 100-200 mg powder/ml buffer, sonrcate for 10 min, and leave on ice for 1 h Avoid boiling to prevent aggregation of proteins. 4. Spin suspensions and recover the supematant and determine protein concentrations. 5 For Western blot analyses, apply l-25 pg of protein to SDS-polyacrylamide gels (good separation is achieved by using precast gradient gels, 415% or &20%, e.g., from Bro-Rad Laboratories, Hercules, CA), and electroblot onto PVDFor nitrocellulose-membrane 6 Blotted proteins can be analyzed with antibodies, lectins, or bacteria (IO) Several ways for detecting bound lectin, antibodies, bacteria are available, such as direct enzymatic detection (horseradish peroxidase, alkaline phosphatase), silver staining, or chemtluminescence based on alkaline phosphatase (Tropix, Bedford, MA) or horseradish peroxidase (ECL, Amersham, Buckmghamshire, UK).

3.5.2. Glycolipid Preparation 1. For glycolipid preparation, extract the lyophrhzed powders with methanol (-20 mg dry weight/ml), somcate the solution for 10 min, and incubate at 65°C for 1 h.

Borkn et al.

218

2.

3.

4.

5.

6.

7.

Followmg centrifugation, the supernatants are transferred to new tubes The pellets are re-extracted, using the same procedure, with methanol, chloroform: methanol, 1:2 (v/v), and chloroformmethanol 2: 1 (v/v), respectively Evaporate the pooled supernatants at 65°C under N, and subject them to mild alkaline hydrolysis m 3 mL of 0 2M potassium hydroxide in methanol at room temperature for 3-12 h. Neutrahze the alkaline solution through the addition of 50 pL of acetic acid and reconstttute m 300 pL of chloroform and 2650 pL of water to yield a final solution of 6 mL chloroformmethanokwater, 1: 10:9 (v/v/v). Perform desaltmg reversed phase chromatography on 5 g Bond-Elut C- 18 columns (Vartan Sample Preparation Products, Harbor City, CA), preconditioned in 40 mL of chloroform:methanol,water, 1: 10:9 (v/v/v). Apply samples to the column four times before washing the column with 20 mL of 1: 10.9 and elute to complete dryness. Elute alkali-stable lipids with 30 mL of chlorofotmmethanol, 2: 1 (v/v), evaporate to dryness under N2, and vacuum-desiccate for 1 h. Perform acetylation of lipids in 1 mL each of chloroform, pyridine, and anhydrous acetic acid overnight at room temperature in the dark; stop by the addition of 5 vol of methanol and evaporation, followed by 5 repeated evaporattons m small volumes of methanol. Remove the sphingomyelin fraction on a 5 g silica gel Bond-Elut column preconditioned in 20 mL each of chloroformmethanol, 85.15 (v/v), and chloroform. Samples are applied to the columns in 3 mL of chloroformmethanol, 98.2 (v/v), and elute lipids free of sphingomyelin with 30 mL of chlorofornumethanol, 90: 10 (v/v) Evaporate these samples under N2 and deacetylate in 0.5 mL each of methanol and toluene and 1 mL of 0.2h4 potassium hydroxide m methanol, at room temperature overnight Neutralize the solutton in 50 pL acetic acid and desalt as above. The composition of the glycosylated components of the partly purified alkah stable lipid preparations can be analyzed after separation on plastic or alumina backed high performance thm layer chromatography (HPTLC; Merck) plates, eluted with chloroformmethanolwater, 60:35.8 (v/v/v) (ref. 20 summarizes different solvents and ways of glycohpid detection) In addition, immunological detection can be done using lectins, monoclonal antibodies, or whole bacteria, as reviewed m ref. 10.

3.5.3. Detection of Bacterial Protein/Glycoprotein by Western Blot Overlay

Receptors

1, Separate proteins (0. l-l .Omg) by gel-electrophoresrs, either under native conditions, to preserve conformattonal integrity, or by SDS-PAGE, for enhanced presentatton of peptide epitopes; transfer to a mtrocellulose membrane. 2. Incubate membranes with blocking buffer (see Sectton 2.5 , item 2) overnight, and wash 2 x 10 mm in TBS. 3. Add a Dig-labeled bacterial suspenston of 0 1 OD,,, (see Section 3.3.1 ) and incubate for 8 h at room temperature, followed by 6 x 5 min washes in TBS.

/den tifica tion of Uos t Receptors

219

4 Add AP-conjugated anti-Dig antibody (Boehringer Mannhelm) diluted 1 2000 m blocking buffer, and Incubate for 1 h, and wash 5 x 10 min in TBS; bound antibody is detected with BCIP/NBT.

3.5.4. Detection of Bacterial Glycolipid Receptors by HPTLC Overlay 1. Dilute glycolipids in a proper solvent (e.g., chloroform:methanol.water, 65:35*8 [v/v/v]) and spot in bands (0.5-l cm wide) about 1 cm from the bottom of an alumina-backed HPTLC plate (Si-60, Merck), leaving 1 cm on each edge, using a Hamilton syringe. Between 1 ng and 1 ug of pure glycolipid is required for immunologtcal detection (depending on the specificity and affinity of the probe) Approximately 1 pg of glycolipid m a band is required for chemical detection using anisaldehyde, which will stain all glycohpids green, or resorcmol, which stains gangliosides blue to purple (20); this could be used as a control to verify the purity and mobility of the glycolipids. After the samples have drted for a few minutes, place the plate in a developing tank, which has been equilibrated with eluent (e.g , chloroform:methanol:water, 65:35*8 [v/v/v]; see ref. 20 for examples of proper solvents for different glycolipid fractions). Allow the solvent front to migrate 4-5 cm and air-dry the plate for 1 to 2 h. 2 Fix the plate for 30 s in 0.25% poly-isobutylmetacrylate plastic (P28, Aldrich, Steinhetm, Germany) dissolved in diethyl ether, to preserve the silica gel, and air-dry overnight or for 30 mm in a vacuum desiccator Plastic fixative should be freshly prepared the day before to ensure that it is completely dissolved. If the solution is too concentrated or too old, the specificity of the overlay may be altered significantly. 3. Block the fixed plates in 5% BSA in PBS for 1 h at room temperature. Air bubbles must be removed, since they raise the background staining m combinatton with ECL-detection In addition, it is essential that the plates do not dry, because this will markedly increase background. 4. Aspirate blocking off the plate, then overlay it with a suspension of labeled bacteria (107-lO*/mL) in 5% BSA-PBS. Bacteria may either be metabolically radiolabeled with 35S-methionine, or Dig-labeled for tmmunologic detection, as described in Section 3.3.1. After incubation for 2 h at room temperature, wash the plates 3 x 5 mm with PBS 5. If radiolabeled bacteria are utilized, au-dry the plate and expose tt to film If Dtglabeled bacteria or purified bacterial adhesin is used as probe, cover the plate with an antibody suspension (horseradish peroxidase conmgated sheep anti-Dig [Boehringer Mannheim], or anti-adhesin antibodies diluted 1:2500 to 1:5000 m 5% BSA-PBS). Incubate the plates for 1 h at room temperature, followed by washing 3-5 times in PBS. Secondary antibodies should be HRP-conjugated for subsequent ECL detection. Dilute them 1:5000 to 1: 10,000 in 5% BSA-PBS and incubate on the plates for 1 h. 6. Wash the plate six times in PBS. We have adapted enhanced chemilummescence (ECL) for use in HPTLC overlays. Mix the two components of ECL substrate (Detection reagents 1 and 2; Amersham) 1: 1 and apply on top of the HPTLC

t3or6n et al. plate. Leave the solution m place for 1 mm. A thin film of substrate solution should always remam on top of the plate. If too little substrate remains, it ~111be absorbed during film exposure and lead to increased background. Wrap the plate m cellophane and place m a film cassette. Place autoradlographlc film on top of the plate for 1 min and then develop. Pre-flashed film should be used to increase the linearity m the exposure, allowing for quantitatlon of bmdmg intensity Several exposures are usually required to optimize the signal-to-noise ratio. It should be noted that the signal intensity decreases markedly with time If a colorlmetrlc substrate 1s to be used, DAB (e.g., Sigmafast [Sigma]) appears to provide the highest sensitivity. However, colorlmetric detection requires a higher concentration of secondary antibodies, i.e., in the range of 1:500 (see Note 16).

4. Notes 1 The H. pylon strain CCUG 17874 could be used as a positive control, smce It hemagglutinates human erythrocytes efficiently (21). Growth conditions are important for the expression of efficient hemagglutmation properties, since these are restricted to plate-grown bacterial cultures (22). 2. For interpretation of the HA1 assay, it is necessary to consider that the reaction of the bacterial cells with erythrocytes can be medlated by several structurally dlfferent glycoconjugates, which can make interpretation of the results complicated Bacterial overlay onto electrophoresis-separated erythrocyte membrane proteins, immobilized on Western blot membranes, has been used to identify glycophorm A as the erythrocyte receptor for S-fimbriated E co11 (23) Methods for dlssection of immobilized glycoprotein and glycolipid receptors are further discussed in Sections 3 5.3. and 3 5.4 3. In order to verify the effect of neurammldase treatment of RBC, the siallc acid dependent hemagglutination reaction of S-fimbnated Escherichla coli could be titrated (23) 4 S-fimbriated E co/z could also be used to assessthe efficiency of protease treatment, since S-fimbnated strains preferentially interact with slalylated glycoproteins (6,14,23) The integrity of the glycollpids could be assessed usmg P-fimbnated E coli, since they are restricted in receptor specificity to the carbohydrate structure Galcl I-4Ga1, presented in glycohpids (2). 5. RBC can usually be stored for up to l-2 wk at 4°C without loss of hemagglutmatlon properties However, RBC from some animals, such as chickens, are less stable, and might lyse after a week. Protease-treated RBC should preferably be used the same day, since they will usually rapidly lyse. Make sure that the RBC do not autoagglutinate. 6. Make sure that the particles do not autoagglutinate after the coating procedure (mild sonicatlon can be used to redisperse the particles). Use both positive and negative controls to verify the coatmg of the particles Particles should be stored at 4°C and are stable for 1 mo. Particles should not be frozen 7. Perform the agglutination assay against a dark background, in order to read the results more accurately.

/den tifica tion of Host Recep tot-s 8. Bacteria can be surface-labeled with fluorochromes or with marker molecules, such as drgoxigenin (Dig) or biotin, that can be efficiently detected with a wide range of conjugated antibodies or streptavidin, respectively. Labeling with marker molecules has proven very useful, since the same batch of labeled bacterta can be applied in a number of adhesion assays, in addition to binding in s&z, for instance, and microbial Western blot- and HPTLC-overlays (see Sections 3.5.3.-3.5.4.). Furthermore, labeled bacteria can be stored frozen for extended periods of time, thus circumventing problems with batch-to-batch variability. 9. When labeling with biotin, it is important to remember that the streptavidin molecule is sensitive for the steric presentation of the biotm epitope. By using a spacer arm coupled to biotin, the availability of the biotin tag will be much improved for streptavidin interactions. Activated biotin with spacer arms of different lengths are commercially available, such as Biotin-X-NHS and BiotmXX-NHS (Calbiochem). Introduction of labeling markers such as fluorochromes, Dig, or biotin, will modify the bacterial surface proteins and could possibly interfere with the binding properties of the adhesins. This could potentially result in the lack or reduction of expected binding, or by the introduction of artifactual binding properties. Alternative detection methods could be used for the imtial zn situ experiments, such as unmunodetection using antibodies or serum directed against whole bacteria (available from DAK0 A/S, Glostrup, Denmark). Alternatively, direct staining of adherent bacteria, using fluorochromes, which are taken up by the bacteria and become fluorescent following DNA mtercalation (e.g., acridine orange), could be used for comparison Acridine orange (Sigma) stammg is performed by washing the slide with tissues or cells containing adherent bacteria in a solution of acridine orange (10 mg/mL, 50 mM sodium acetate buffer, pH 4.0), for 10 min in the dark, followed by a brief rinsing in PBS (12). 10. Bacterial binding in sztu is much improved by the addition of the protease inhibitor mix. 11. The in situ adherence assay was primarily designed as a method for identifying potential bacterial receptors in host target tissues. Bacterial adherence to cells or tissues in sztu demonstrates the presence of potential receptor molecules. However, it does not necessarily indicate that the bacteria will have a natural habitat at this location in vivo, since a combination of environmental factors will determine whether or not a particular site will form an appropriate bacterial niche. 12. In order to verify the specificity in Inactivation of carbohydrate-based receptor epitopes, an antibody against a peptide antigen could be used to probe the tissue section after periodate oxidation, to monitor the integrity of protein-based antigen structures. Carbohydrate-specific antibodies or lectins could be similarly used to probe for macttvation of carbohydrates (6,1#) By performing the periodate reaction under milder conditions (pH 5.5, 10 min, on ice), a selective cleavage of sialic acid can be achieved. This was titrated using tissue sections of dog gastric mucosa, by the loss of binding of the NeuAca2,6Gal/GalNAc recognizing Sumbucus nzgra (SNA) lectin. No effect on the ability of Fucal,2Gal+peclfic Ulex europaeus type 1 (UEAl) to bmd surface

Bor& et al.

13

14.

15.

16

mucous cells was found under these conditions. Using this approach, we could demonstrate (Fig. 2) the carbohydrate-specific, sialic acid-independent binding of H pylorz to human gastric surface mucous cells (9, IO). Although this method dtscrimmates very effictently between protein and carbohydrate structures (receptors), it is important to remember that there are carbohydrate structures resistant to periodate oxidation, like 0-acetylated sialic acid. In addition, there are protein structures that may become inactivated and lose there bmding or enzymatic properties. Variability m bacterial binding to the tissue secttons could be caused by heterogeneity m tissue quality, technical procedures (fixatton and embedding), and/or phenotyptc (e.g., localized inflammatory processes or various stages of neoplasttc transformatton) factors. Using consecutively cut tissue sections will reduce the variability in the inhibition experiments, especially since the receptor density and presentation will vary between different parts of the biopsies and between mdivtduals. All mhibrttons should be done, as far as possible, in the same set of experiments, in order to make accurate comparisons of inhibitory activity Common glycoconjugates used for characterization of siahc acid binding bacteria are the highly sialyiated glycoproteins fetuin and glycophorin A (both available from Sigma). Mucm fractions, like bovine submaxillary mucms (Sigma), are well-characterized and rich sources of complex, both sialylated and fucosylated, carbohydrate epitopes, as are protein fractions from human milk, like human secretory IgA and tc-casein. Both milk proteins have proven to be very useful m initial inhibition experiments (6,~11,24). The detailed characterization of carbohydrate-based microbial receptor epitopes by inhibition assays requires access to purified and structurally characterized carbohydrate chains, preferably presented in defined glycoproteins, 1.e , either naturally purified glycoprotems, such as blood group-specific milk proteins or synthetic glycoprotems. Conjugation of many (25-50) chains of the same oligosaccharrde to a nonglycosylated carrier protein like human serum albumin, creates a highly defined neoglycoprotem. They also provide multivalent attachment sites for the bacterial adhesms, which often potentiate bacterial binding to the compound and increase the mhibitory activity (II, 14). The effect of the glycosidase treatments can be analyzed by lectin overlays of SDS gel electrophoresis-separated, Western blot-muuobihzed glycoprotems that have been digested with glycosidases. Ulex europaeus type 1 (UEA 1) lectm, detectmg terminal fucose, Sambucus nzgra (SNA) lectin, detectmg a2,6-linked sialic acid, and Maackta amurensls agglutimn (MAA), binding to cL2.3-linked sialic acid, are suitable lectins for the verification of the efficiency of fucosidase and neuraminidase digestions (all available as Dig-labeled compounds from Boehringer Mannheim). The Western blot and HPTLC overlays allow for the detection of receptor-active structures isolated from target tissue and associated secrettons and solutions, such as gastric juice and serum. In addition, the receptor activity for structurally characterized and defined glycoproteins, such as neoglycoproteins, can be evaluated

223 in solid phase (IO,Il). The Western blot and HPTLC overlay assays are, however, only qualitative; for a comparison of affinity to different membrane components, an enzyme linked nnmunosorbent assay (ELISA) may provide a way for titration of bacterial binding to various receptor candidates (I 0).

Acknowledgments T. B. has been supported by the Swedish Medical Research Council (Project no. 11218), the Swedish Society for Medical Sciences, the Swedish Society of Medicine and Lion’s Cancer ResearchFoundation, University of UmeA, Sweden. P. F. has been supported by the Swedish Medical Research Council (project no. 11595), the Swedish Cancer Society (3523-B95-02XBB), and Pharmacia Research Funds.

References 1. Sharon, N. (1986) in The lectms; properties, functions, and applications in biology and medicme (Liener, I. E , Sharon, N., and Goldstein, I. J., eds.), Academic, New York, pp. 494-526. 2. Stromberg, N., Nyholm, P.-G., Pascher, I., and Normark, S (1991) Saccharlde orientation at the cell surface affects glycolipid receptor function. Proc Nat1 Acad. Scz. USA 88,9340-9344.

3 BorBn, T and Falk, P (1994) Helicobacterpylori* binds to a blood group antigen Scz Am Sci Med. Sept/Oct., 28-37. 4 Ren, R. B., Costantini, F., Gorgacz, E. J., Lee, J. J., and Racaniello, V R. (1990) Transgenic mice expressing a human poliovirus receptor* a new model for pohomyelitis. Cell 63,353-362. 5. Falk, P. G., Bry, L., Holgersson, J., and Gordon, J, I. (1995) ExpressIon of a human a- 1,3/4-fucosyltransferase in the pit cell lineage of FVB/N mouse stomach results m production of Leb-containing glycoconjugates: a potential transgenic mouse model for studying Helicobacterpylon infection. Proc. Nat1 Acad Scz USA 92,15 15-l 5 19 6. Falk, P., BorCn, T., and Normark, S. (1994) Strategies for characterization of microbial host receptors. Meth. Enzymol. 236,353-374. 7. Lelwala-Guruge, J., Ljungh, A., and Wadstrom, T (1992) Haemagglutinatlon patterns of Helzcobacter pylori-frequency of sialic acid-specific and non-sialic acid-specific haemagglutinms. APMIS 100,908-913. 8. Korhonen, T. K., Parkkmen, J., Hacker, J., Finne, J., Pere, A, Rhen, M., and Holthbfer, H. (1986) Binding of Escherzchia coli S fimbrlae to human kidney epithelium. Infect. Immun. 54, 322-327. 9. Falk, P., Roth, K. A , Bortn, T., Westblom, T. U., Gordon, J I., and Normark, S. (1993) An in vitro adherence assay reveals that Helzcobacter pylorz exhibits cell lineage-specific tropism in the human gastric epithelium. Proc NatZ. Acad Sci. USA 90,2035-2039.

10. Falk, P., BorCn, T., Haslam, D., and Caparon, M. G. (1994) Bacterial adhesion and tissue colonization assays. Methods Cell Bzol. 45, 16 l-1 87

224

Borbn et al.

11 Boren, T , Falk, P., Roth, K. A., Larson, G., and Normark, S. (1993) Attachment of Helicobacter pylort to gastric epithelium mediated by blood group antigens Sctence 262, 1892-l 895 12. Okada, N., Lrszewski, M. K., Atkinson, J. P., and Caparon, M. (1995) Membrane cofactor protein (CD46) is a keratinocyte receptor for the M protein of the group A streptococcus. Proc Nat1 Acad Set USA 92,2489-2493. 13. Woodward, M. P , Young, W. W , Jr., and Bloodgood, R. A. (1985) Detection of monoclonal antrbodies specific for carbohydrate eprtopes using periodate oxrdatron J Immunol Meth 78, 143-153. 14. Stromberg, N. and Boren, T. (1992) Actinomyces tissue spectticity may depend on differences m receptor specificity for GalNAcp-containmg glycoconjugates Infect Immun 60,326&63217.

15. Paulsson, M., LJung, A., and Wadstrom, T. (1992) Rapid identiticatron of fibronectm, vttronectm, lammm and collagen cell surface binding proteins on coagulase negative staphylococci by particle agglutination assays. J Clan Mtcrobiol.

24,2006-2012.

16. Lelwala-Guruge, J., Ascencro, F., Kreger, A. S., LJungh, A., and Wadstrom, T. (1993) Isolatron of a stahc acid-specific surface haemagglutmins of Heltcobacter pyiort strain NCTC 11637 Zbl Bakt-Int. J. Med Mtcrobiol. 280,93-l 06. 17 Falk, P., Roth, K. A., and Gordon, J. I. (1994) Lectms are sensitive tools for detining the differentiation programs of mouse gut eprthelial cell lineages. Am 3 Physiol. 266, G987-G 1003.

18. Herrmston, M L., Green, R. B., and Gordon, J. I (1993) Chrmerrc-transgenic mice represent a powerful tool for studying how the proliferation and differenttatron programs of intestmal epithelial cell lineages are regulated Proc. Natl. Acad Scz USA 90,886&8870

19. Luna, L. G., ed. (1968) Manual of Histologic Starning Methods of the Armed Forces Institute of Pathology, 3rd ed., McGraw-Hill, New York. 20. Stahl, E. (1967) Diatnschichts Chromatographte-em Laboratortums-handbuch, Springer Verlag, Berlin. 21 Hirmo, S., Kelm, S., Schauer, R., Nilsson, B., and Wadstrom, T. (1995) Recogmtron of different stalic acid linkages by Helzcobacter pylort strains. Mol Mtcrobtol., m press. 22 Evans, D. G., Evans, D. J., Jr., Moulds, J. J., and Graham, D. Y (1988) N-acetylneurammyllactose-binding fibrillar hemagglutinm of Campylobacter pylort. a putative colonization factor antigen. Infect Immun. 61,2896-2906. 23. Parkkmen, J , Rogers, G. N., Korhonen, T K , Dahr, W , and Finne, J (1986) Identification of O-linked sialyloligosaccharides of glycophorin A as the erythrocyte receptors for S-timbriated Escherichia coli Infect Immun 54,37-42 24. Stromqvrst, M., Falk, P., Bergstrom, S., Hansson, L., Lonnerdal, B., Normark, S., and Hernell, 0. (1995) Human milk K-casein and inhibition of Helicobacterpylon adhesion to human gastric mucosa. J Pedtatr Gastroenterol Nutr 21,288-296

22 Host Responses

to H. pylori

Molecular Analysis of Cytokine Gene Expression Susan M. Farmery and Jean E. Crabtree 1. Introduction The interactions between Helicobacterpylori and the human gastric epithehum have been modeled in an in vitro cell culture system. The model permits investigation of the interplay between the bacteria and eprthehal cells m a controled environment. Following coculture, the cells, bacteria, and culture supernatants are available for analysis. There are several established human gastric cancer cell lines that can be used as model systemsfor examining bacterial-epithelial interactions. The protocol described herein employs epithelial cells grown as a monolayer, or m suspension, and cultured with specific H. pylori strains. The cultured cells are harvested, and changes in gene expression determined by a semi-quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) assay(1). The use of RT-PCR technology permits the analysis of gene expression in small numbers of cells; experiments designed in 24- or 96-well plates become feasible. Using the methods described in this chapter, it has been possrble to design experiments involving combinations of bacterial strains and pharmacological agents (cytokines, metabolic inhibitors, blocking antibodies, and so on), at different concentrations and time-points, that would be impractical by other methodologies. The protocol relies on the extraction of total RNA reverse transcription of the mRNA into complementary DNA (cDNA), PCR amphticatron of the cDNA using gene-specrfic primers, and the quantification of PCR product. The amount of mRNA present in the original sample can then be extrapolated. Extraction of pure nondegraded RNA is critical to the successof the technique. Steps should be taken to ensure processes are undertaken in a nucleaseEd&d

From: Methods m Molecular Medrc~ne, Hellcobacter pylon Protocols by C L Clayton and H L T. Mobley Humana Press Inc , Totowa,

225

NJ

226 free environment (see Note 1). Various RNA extraction methods have been evaluated in our laboratory, and protocols for the two preferred techniques are described in Section 3.3. The first, based on the method of Chomczynskt and Sacchi (2), uses RNAzolTMB (Biotecx, Houston, TX). Cells are lysed in the presence of the RNAzol reagent, which promotes the formation of complexes of RNA with guanidinium and water while inactivating ribonucleases. The addition of chloroform causes the separation of aqueous and organic phases with the DNA and proteins discarded in the organic phase. RNA can then be precipitated from the aqueous phase by the addition of alcohol. The second method uses Catrimox-14 (Iowa Biotechnology, Coralville, IA) (3). This cationic surfactant solution lyses cells, inhibits ribonucleases, and precipitates RNA and DNA as electrically neutral complexes. The RNA IS then extracted from the complexes by the addmon of lithium chloride, which preferentially solubilizes contaminating DNA and proteins. Followmg either technique, the RNA is resuspended in water containing ribonuclease mhibitor, and can be stored at -70°C. The presence of contammatmg genomtc DNA m the RNA is undesirable. If it is not possible to design gene-specific primers that span introns, genomic DNA will co-amplify with cDNA. Even tf primers that should anneal exclusively to cDNA are selected, the presence of variable amounts of contaminating genomtc template will affect the efficiency of the PCR m an unpredictable manner, possibly generating misleading results. For this reason, the RNA is checked before reverse transcription for the presence of genomic DNA by PCR with genomic DNA primers. With the RNA extraction techniques described, DNA contammation is rarely a problem; however, it is possible to DNase-treat RNA before proceeding to the reverse transcriptron, If necessary. The reverse transcription step synthesizessingle-stranded cDNA from RNA. The choice of reverse transcriptase (RT) may depend on a combination of factors, but both avian myoblastoma virus RT and Maloney murme leukemia virus RT (MMLV-RT) have been used successfully. A protocol for the use of MMLV-RT is described. The choice of RT primer is also a constderatlon; oligo(dT)r5 will prime the reverse transcription of all polyA-tailed mRNAs and is used routmely in our laboratory, although random hexamer primers have also been employed where the amplification product is far removed from the 3’ end of the gene. The product of the RT reaction should constitute a pool of cDNA species represented in proportion to the amounts of mRNA in the origtnal cell population. The cDNA is stable and can be stored at 4OC. The presence of mRNA for any given gene can be determined by amplification of the corresponding cDNA by PCR, using gene specific primers (Table 1) The protocol described here expresses changes m gene expression proportionally to expression of a so-called housekeeping gene, glyceraldehyde 3-phosphate dehydrogenase (G3PDH), which acts as an mternal standard. All such

Host Response: Cytokine Gene Expression Table 1 Primers Used for Cytokine

Assays

G3PDH- 1 G3PDH-2

GAG TCA ACG GAT TTG GTC GT

lkactin- 1 ll-actinHSP70A- I

ATC CAT CTA GTT AAG GAT

HSP70A-2 IL-8- 1 IL-8-2

227

GCT GCC ATG GAA TTT GCC AT ATG TTT GAG ACC TTC AA CTC TTG CTC GAA GTC CA GCC TGA GGA GCT GCT GCG ACA CCC TGC TCT CTG TCG GCT CGG GAA CCA TCT CAC TG TCT TGG ATA CCA CAG AG

techniques rely on the expression of the internal standard gene remaining unchanged. p-actin has often been used as an internal standard, but its expression is regulated in a variety of circumstances, and changes have been observed following bacterial-epithelial interactions; G3PDH appearsto be more constantly expressed (4). G3PDH and the gene of interest are co-amphfied in a duplex PCR. The relative generation of each PCR product will reflect the ratios of their respective mRNAs in the original cells. A limitation of the technique 1sthat the absolute ratio of PCR product for G3PDH to the gene under investigation will be strongly influenced by the respective efficiencies of their amplification reactions. However, in experiments that compare control and test condttions, changes in product ratios can be assumed to reflect changes in mRNA copy number. Different methods of quantifying PCR product have been employed; radiolabeled nucleotides can be mcorporated into the products and quantified by scmtillation counting. Alternatively, computerized image analysis can ascribe a quantitative value to the band corresponding to a given product on an ethidium bromide-stained gel. Protocols for both techniques are described m Section 3.9. The described techniques have been used to demonstrate changes in IL-8 gene expression in gastric cancer cell lines followmg coculture with different H. pylori clinical isolates (5). 2. Materials 1 Water: Wherever possible, use nuclease-free molecular biology grade water (BDH) 2. KATO-3 American Type Culture Collection (ATCC) HTB 102; European Collection of Animal Cell Cultures (ECACC) 8609004. 3 AGS (ATCC CRL 1739, ECACC 89090402). 4. Media: RPMI-1640 (ICN), supplemented with 10% heat inactivated fetal calf serum (FCS) wtth or without 40 pg/mL gentamicm 5. RNAzol B: Available from Biotecx.

6. Catrimox: Available from Iowa Biotechnology. 7. 2MLiCl:

Prepare 500 mL 2MLiCl

solution, store at room temperature.

228

Farmery and Crabtree

Table 2 2X RT Mix

Regent

Mg’A

RT buffer (Promega) dNTP mtx oligo(dT), s rRNasin MMLV RT Water

Stock 25 mA4

10x 20 mM

0.5 mg/mL 40,000 U/mL 200,000 U/mL

2X Workmg solution 6mA4 2x 2mh4 50 pghrlL 2000 u/mL

12,000 U/mL to volume

Final 3mM

1x lmIl4 25 pg/mL 1000 U/mL 6000 U/mL

8. EDTA: Dissolve disodmm ethylenedtammetetraacetate in water, adJust to pH 8 0 with NaOH, make up to 0.5M (6). 9 TAE, O.O4MTns-acetate, 1 mA4EDTA. 10 TE: 10 m/t4 Tris-HCl, pH 7.9,O. 1 mM EDTA. 11 100 rm’t4DTT: Prepare 100 tidithiothreitol solution, store at -20°C m aliquots. 12. Recombinant ribonuclease inhibttor rRNasm, Promega 13 DNase I: RNase free, supplied by Life Technologies. 14. 2.5MAmmonium acetate: Store at room temperature. 15. RNA solubilization solution: Immediately prior to use, prepare a workmg stock of water containing 1 U/pL rtbonuclease inhibitor (rRNAsin, Promega, Madison, WI) and 1 m&f dithiothreitol. 16. dNTP solution: Using Pharmacia dNTP set (Pharmacta Biotech, Uppsala, Sweden), dilute with water to yield a mix that has each nucleottde at 20 mA4, store at -20°C in aliquots. 17. 25 mM MgC12: supplied by Promega with PCR buffer, or prepared as aqueous solution, stored frozen m aliquots. 18. 2X RT mix: Prepare a working stock solution immediately prior to use, keep on ice as the enzyme 1s labile (Table 2). 19. PCR primers: If the oligonucleotrdes are in ammonia solution, the solute should be dried off by heating (pL amount) or under vacuum centrifugation (larger quantities) and redissolved m water. The relative concentration of different primer pairs in multiplex PCRs needs to be determined empirtcally. Working solutions of primer mix are prepared so that the primer can be added to the PCR tube m a convenient volume (2-10 pL). Sequences for primers used for the analysts of IL-8 and G3PDH (7,8) are shown on Table 1. 20. 2X PCR mix* Prepare a working solution of 10 pL per PCR reaction immediately prior to use; keep on ice (Table 3). If radiolabeled PCR product IS required, 33P-dATP can be added to the mix at a concentration of 50 pCi/mL 21 Size Marker Ladder. Commercial preparations, 100 bp ladder size range (Life Technologtes, Paisley, UK), are used.

229

Host Re,sponse: Cytokine Gene Expression Table 3 2X PCR Mix Stock

Regent PCR buffer (Promega) M&l;? dNTP mix Taq polymerase Water

2X Working solution

Final

10x 25 mA4

2x 3mM

1x 1.5 mM

20 mA4 5000 u/mL

0.4 mM

0.2 mM

100 U/mL to volume

50 U/mL

22. Sample loading buffer: 0.25% bromophenol blue, 0.05% xylene cyan01 FF, and 15% Ficoll (Type 400; Pharmacia) in water, store at room temperature (see Note 5) 23 Agarose (Genetrc Technology Grade, ICN Biomedicals, Aurora, OH): Prepare a 2% solution of agarose in 1X TAE buffer. Heat until botlmg, cool to 55“C, add ethidium bromtde (50 pg/lOO mL), pour. Leave to set. 24. Ethrdlum bromide: 10 mg/mL solution, stored in dark bottle at room temperature.

3. Metholds 3.1. Cell Culture 1. Rout1 nely maintain cells in RPMI- 1640 with 10% FCS and gentamicm. 2. Seed 1 mL of AGS cells at a density of 1 x 105/mL mto 24-well tissue culture plates (ICN) and culture for 2-3 d to confluency (approx 5 x 105/mL). KATO-III cells, which are generally nonadherent, can be resuspended in antibiotic-free culture medmm and used directly (see Section 3.2.).

3.2. Bacferial

Epithelial

Coculture

1. Harvest H pylon strains grown on blood agar base Number 2 (Oxoid, Basingstoke, UK) containing 7% fresh horse blood into antibiotic-free culture medium. Brothcultured H pylori can also be used successfully Resuspend bacteria at a concentration of 2.5 x 107/mL to give an eventual bacteria:cell ratio of approx 100.1 and use mmediately (see Note 2). 2. Carefully remove the antibiotic-containing culture media from the confluent monolayers of eptthelial cells. Add 1 mL of bacterial suspension and incubate together with control cell cultures at 37°C in a 95% sir/5% CO2 incubator 3. At varying time-points postculture, cell-free supernatants can be harvested for measurement of released protein and cells for extraction of RNA (see Section 3.3.).

3.3. RNA Extraction 3.3.1. Using RNAzol B 1. Add RNAzol B accordingto manufacturer’sinstructionto cellsznsitu, lysecellsby passing through pipet several times, and transfer to autoclaved mtcrocentnmge tubes

230

Farmery and Crabtree

2. 3. 4. 5. 6 7 8.

Add chloroform (0 1 mL/l mL lysate), vortex for 15 s, leave on ice for 15 min. Centrifige at 13,OOOgat 4’C for 15 min Transfer upper (aqueous) phase to clean microcentrifUge tube Add an equal volume of isopropanol, and store for 15 mm at 4°C to precipitate RNA. Samples centrifuged at 13,OOOgat 4°C for 25 min to pellet RNA Wash pellets twice with 2 vol of 70% ethanol Dry pellets briefly under vacuum.

3.3.2. Using Ca trimox 1. Catrimox is added to the cell suspensions or to adherent cells. A volume no less than the volume in which the cells are suspended should be used Cells pelleted in microcentrlfuge tubes are resuspended in a small volume and extracted 1n 1 mL of Catnmox. 2, The nucleic acid-surfactant complex 1spelleted by centnfugation at either 13,OOOg (CO.5 million cells) or IOOOg (OS-10 million cells) for 5 min at room temperature. 3 Discard supernatant by gentle aspiration 4. Add 1 mL 2M L1C1, vortex vigorously, centrifuge for 5 m1n at 13,OOOg, and aspirate supematant Repeat L1Cl wash. 5. Wash pellet with 70% ethanol, aspirate, and dry pellet briefly under vacuum.

3.4. Solubilization

of RNA and Check for DNA Contamination

1 RNA is dissolved 1n 20 pL of RNA solubilizatlon solution. 2. 1 & of the RNA is checked for the presence of DNA by PCR using primers for heat shock protein (see Section 3.6.). 3 RNA that 1snot immediately reverse-transcribed 1sstored at -7O*C 4. Samples with DNA contamination can be treated with DNase. RNA incubated at 37°C for 30 mm with 1000 U/mL DNase I 1n the presence of 200 U/mL nbonuclease inhibitor 1n TE buffer containing 10 mA4MgC1, and 1 mA4dlthlothreltol. 5 RNA 1sprecipitated in the presence of 0 5 vol of 5Mammon1um acetate and 2 vol ice-cold ethanol at -2O’C for 30 m1n The precipitate 1s pelleted at 13,000g at 4”C, washed with 70% ethanol, and solub111zed as before.

3.5. Reverse

Transcription

1. Dispense 10 @ of 2X RT workmg m1x into autoclaved 0.7-mL m1crocentrlfuge tubes 2 Add 10 pL of the RNA solution to yield a 20-pL final reaction volume. 3. Incubate at 42°C for 1 h, then heat to 95’C for 5 min to heat-inactivate the reverse transcriptase. 4. Check the presence of cDNA by setting up a PCR for a constitutively expressed gene, such as G3PDH (see Note 3).

3.6. PCR 1. Dispense working solution of primer m1x (1 $4 each), and make the volume up to 10 pL with water.

Host Response: Cytokine Gene Expression 2. 3. 4. 5.

231

Add a drop of light mineral oil. Add 1 @, template cDNA (or RNA for genomtc DNA check). Move tubes to thermocycler and start PCR program to heat PCR block to 95“C. Once tubes have equilibrated at 95’C, add 10 pL of 2X PCR mix and allow the cycling program to complete.

3.7. Thermocycler Programming The optimal cycling parameters of the PCR should be determined for each set of primer combinations (see Note 4). The program outlined below is often satisfactory and works well with the primers given in Table 1. Step 1 95°C (5 min) initial denaturation Step 2 95OC (1 min) denaturation Step 3 55°C (1 min) annealmg Step 4 72°C (1 min) extensron From step 4 return to step 2 and cycle 29 more times 72°C (5 min) final extension Step 5 Step 6 4°C (hold)

3.8. Agarose

Gel Electrophoresis

1. 2. 3. 4 5.

Prepare gel tray: tape ends, adjust comb. Pour 2% agarose solution at 55°C to required depth and leave to set Remave comb (wrap gel in cling film and store at 4”C, if necessary). Place gel on its tray into the tank, and just submerge beneath TAE running buffer MIX samples, and size-marker ladder, with loading buffer (approx 1 pL buffer/l0 uL sample) and pipet into wells (see Note 5). 6. Replace lid on tank, connect leads, and set voltage to run at 10 V/cm as measured between the electrodes, run gel for 30-60 min as necessary to separate the product adequately. 7. Visualize DNA by UV light source, and photograph (see Fig. 1).

3.9. Quantification of PCR Product 3.9.7. Sy Band Excision and Scintillation Counting I. Carefully excise bands containing relevant PCR products, place in counting vial. 2. Add 0 3 mL sodium hypochlorite (15% available chlorine), and incubate at 45°C for 4 h. 3. Add 3 mL Packard Hionic Scintillant. Count in p-counter.

3.9.2. By Densitometry 1. The gel is run with a track containing loading buffer diluted in water to act as a background track for subtraction during image analysis. 2. The image is digitized using a UVP Gel documentation system (UVP, San Gabriel, CA) (GDS 5000) with the Imagestore 5000. The image is electronically captured, and digitized.

232

Farmery and Crabtree

12345L

@ IL-8 @ G3PDH Fig. 1. PCR analysis of mRNA using primers specific for IL-8 (369 bp) and G3PDH (157 bp) in KATO-III gastric epithelial cells after 3 h culture with H. pylori. Lane 1 cell only control; lanes 2 and 3, type 2 strains of H. pylori (G25, G50); lanes 4 and 5, type 1 strains (G32 and NCTC 11637). L, 100 bp ladder. Reproduced with permission from ref. 5. 3. Peak height and area measurements are determined for each band on the track using GelBase, a software package for the GDS 5000 (see Fig. 2).

3.9.3. Expression of Results 1. All determinations are performed in triplicate. 2. Results are expressed as a mean of the ratio of PCR product for the gene of interest to the internal standard gene.

4. Notes 1. Establishing an RNase-free environment: RNases are very active enzymes and do not require cofactors. They are present in the cells under investigation, as well as in the environment, particularly on the skin, but also in commercial chemicals, plasticware, and glassware. Different degrees of stringency with regard to RNase elimination have been described, but precautions have to be practical, based on the space where the experiments have to be undertaken. The wearing of gloves is an essential precaution. Increasingly, commercial, pre-prepared, guaranteed RNase-free reagents are employed, reducing the risk of contaminating reactions with RNases. As long as cells are intact, endogenous RNases are compartmentalized and the mRNA protected. Once the cells are in lysing medium, RNases will be inactivated, and the RNA is safe. If cells are frozen prior to extraction, they should be snap frozen, stored at -70°C and only just thawed prior to addition of the lysing solution. The RNA is again susceptible to degradation once it is dis-

Host Response: Cytokine Gene Expression

233

Fig. 2. Example of a track analyzed using the Gelbase software. Peak 1 is IL-8 (peak height 72, peak area 606). Peak 2 is G3PDH (peak height 135, area 1582). Peak 3 is prime: dimer. solved in aqueous solution. For that reason, RNase inhibitor IS added at that stage Again the RNA should be stored at -70°C, or, preferably, reverse-transcribed immediately. The cDNA is stable and can be stored for several months at 4°C withcut apparent degradation. 2 Bacteria cell ratio Ideally, with each epnheltal cell hne used for functional studies, the culture system should be optimized to use a bacteria:cell ratio that allows optimal secretion of the protein of interest (e g., IL-8) The use of high bactena:cell ratios, which results m poor eplthelial cell viability, should be avoided Recent studies have shown ratios of 100: 1 to be optimal. Cell viability should be momtored by trypan blue exclusion at the end of the experimental period 3 Optmiizing RT: Although not routmely performed in our laboratory, it is possible to determme the efficiency of the RT reaction by setting up a parallel reaction using radiolabeled dCTP, allowing the products to be visualized after alkalme gel electrophoresis. A protocol IS given in ref. 9. 4 Optimizing PCR. A comprehensive discussion of the theoretical and practical aspecls of optimizing PCR are described in ref. 10. Briefly, the mam considerations include primer design, cycle profile, choice of buffer and enzyme, and the use of hot-start PCR Primers should be designed to avoid self-annealing and the formation of both heterodlmers and homodlmers. All the primers to be used in the reaction should have similar dlssoclation temperatures. If oligo(dT) IS used, primers should be within approx 1 kb of the 3’ terminus. The most critical step of the temperature cycle profile is the annealing temperature, which relates to the dissociation temperature of the primers; the higher the annealing temperature, the mclre stringent the hybridization between pnmer and template The technique of Touchdown PCR involves starting with a high annealing temperature, so the early PCR cycles are very specific, but lowering the temperature gradually with each cycle, as the ratio of specific to nonspecific target sequence increases. The most frequently used thermostable polymerase is Tuq polymerase; however, newer enzymes, such as Pfu or Vent, could be substituted. Commerctal reaction buffer is usually sold with the enzyme; however, the Mg2+ concentration may need to be optimized, usually m combination with annealing temperature. The

234

Farmery and Crabtree

PCR protocol described in this chapter is a hot-start PCR; the DNA is denatured in the absence of reaction mix and the PCR reaction only commenced once this high temperature has been obtained This technique precludes the formation of nonspecific hybridization, which may occur as the reaction ramps up from room temperature. 5. Optimization of electrophoresis conditions. If gels are to be analyzed by computerized densitometry, it is important not to use Ioadmg buffer with dye that is too intense. For these gels, the usual sample buffer is diluted 1.100 with 15% Ficoll, and used as before.

References 1. Dallman, M J., Larsen, C. P., and Morris, P J. (1991) Cytokine gene transcription m vascularised organ grafts: analysis by semi-quantitative polymerase chain reaction. J Exp Med. 174,493-496. 2. Chomczynski, P and Sacchi, N. (1987) Single-step method of RNA isolation by acid guamdinmm thiocyanate-phenol-chloroform extraction Anal Blochem 162, 156-159. 3 Macfarlane, D E. and Dahle, C. E (1993) Isolatmg RNA from whole blood-the dawn of RNA-based diagnosis? Nature 362, 186-188. 4. Hedges, S , Agate, W., Scensson, M., Sjogren, A-C , Ceska, M., and Svanborg, C. (1994) Uroepithelial cells are part of a mucosal cytokine network. Infect Immun. 62,23 15-232 1 5 Crabtree, J. E , Farmery, S. M., Lindley, I. J , Figura, N., Peichl, P., and Tompkins, D. S. (1994) CagA/cytotoxic strains of Helrcobacter pylon and interleukm-8 m gastric epithelial cell lines. J Clrn Pathof 47, 945-950. 6 Sambrook, J , Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning, A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 7 Crabtree, J. E., Wyatt, J. I., Trejdosiewicz, L. K., Peichl, P , Nichols, P. N , Ramsay, N , Primrose, J N., and Lindley, I. J. D. (1994) Interleukin-8 expression in Hellcobacterpylori infected, normal and neoplastic gastroduodenal mucosa J Clin. Path01 47,61-66 8. Paludan, K. and Thestrup-Pedersen, K. (199 1) Use of the polymerase cham reaction m quantification of interleukm 8 mRNA in minute epidermal samples. J Invest Dermatol. 99, 830-835. 9 Gurr, S. J. and McPherson, M. J (1991) in PCR* A Practical Approach (McPherson, M. J , Quirke, P., and Taylor, G. R , eds.), IRL, Oxford, UK, pp. 147-170. 10 Kidd, K. K. and Ruano, G. (1995) Optimizing PCR, m PCR 2, A Practzcai Approach (McPherson, M. J., Hames, B. D., and Taylor G. R., eds.), IRL, Oxford, UK, pp l-22.

23Nonprimate Animal Models of H. py/ori Infection Andrew A. McColm 1. Introduction Ever since the realization that Hekobacterpyhi was intimately assoctated with the development of gastritis and peptic ulcer disease in humans, there has been a need for a simple animal model in which modes of pathogenicity, transmission, immunization, and chemotherapeutic intervention can be evaluated. Whereas small animals such as mice and rats are parttcularly well suited as experimental hosts for many infections, early studies suggested that H. pylori had a very narrow host range that did not extend to these species. Although many attempts to infect small laboratory animals with H. pylon’ were apparently made, these proved generally unsuccessful (1,2) and the view became established rapidly that “H. pylori will not colonize many of the usual laboratory animal species,including conventionally reared rats, mice, rabbits, guinea pigs, speci fit-pathogen-free pigs, colostrum-deprived piglets, and gnotobiotic rats and mice” (3). An apparent exception was the claim that H. pylori would colonize Mongolian gerbils particularly after gastric lesions were produced by indomethacin (4); however, this work has never been substantiated nor followed up. Instead, most attention was paid subsequently to the use of naturally occurring Helicobacter mustelae infections of ferrets (5,6), experimental challenge with H pylori in gnotobtotic piglets (7) and the important development of the Helkobacter felis model in mice and rats (8,9). Although these models, in particular H. felis, have many advantages, they are still only substitutes for the “target” organism, i.e., H. pylon. Therefore, efforts continued to try to produce effective H. pylon’ colonization in a suitable small animal host. Proof that gastric colonization with H. pylon’ could occur in mice was first reported by Karita et al. (IO), who showed that Balb/c nude mice could support long-term (>20 wk) colonization with associated gastritis; howFrom Methods II) Molecular Medrcne, Hehcobacter pylon Protocols Edited by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

235

NJ

236

McColm

ever, their euthymic counterparts demonstrated only transient infection (about 2 wk). More recently, an improved model has been described in which specific-pathogen-free (SPF) CD1 mice or conventional (CV) Balb/c mice showed good colonization and developed gastric pathology resembling human disease after oral challenge with mouse adapted fresh clinical isolates of H pylon (2 I), In addition, our group at Glaxo has been using a simple mouse model of gastric H. pylorz colonization over the last few years in which we have examined some of the tmportant determinants of successful colonization (22). These are listed later in thuschapter. By nature of their size, availability, and ease of mampulation, mouse models of disease (infectious or otherwise) tend to be utilized mostly as simple, primary tests m which mittal compound selection, for example, can be performed. After this mittal selection process, more advanced secondary and tertiary models better able to emulate the clmical condition may be necessary for subsequent studies. Nonprimate models of H pylori that fall into this category include the established gnotobiotic pig model (7,13-25) as well as the more recent dog and cat models (3,16-18). Apart from some species of nonhuman primates (see Chapter 24), the cat is the only other animal to have been positively shown to naturally acquire H. pylori infections and demonstrate associated pathology. Consequently, although information IS still preliminary, this model is an important advance offering considerable scope for examination of mechanisms of transmission, pathogenesis, and perhaps factors important in the long-term progression to atrophic gastritis and gastric carcmoma. Although some reference will be made to the pig, dog, and cat models of H. pylorz in the Notes section, this chapter will concentrate primarily on the salient methodologies of the newer H pylori mouse models and thetr potential use in selection of therapeutic agents, vaccines, and so on. 2. Materials 2.1. Animals In the work of Karita et al. (IO), 6-wk-old male athymic nude Balb/c mice and their euthymic counterparts were evaluated. The nude animals were maintamed constantly m lamina flow racks and were fed autoclaved food and water, whereas the euthymic animals were maintained in conventional conditions. Marchetti et al. (I I) used conventional (CV) CD 1 and Balb/c mice, mamtained to a 12-h light-dark schedule, These were fasted for 24 h before any mampulation (e.g., infecting, killing, and so on). In our model (22), female HSD/ICR mice (approx 20 g body wt) are used and housed in conventional conditions and also maintained to a 12-h light-dark schedule. No fasting is performed before infecting, sampling, and so on.

Nonprimate Models of H. pylori Infection

237

2.2. Bacteria Fresh clinical isolates obtained from biopsies of patients with gastritis or peptic ulcer disease are preferable to lab strains.

2.3. Solid Culture Media Clinical isolates of H. pylori and samples isolated from mice can be mamtained on Columbia agar containing 5% horse blood, 0.2% cyclodextrin, and Dent’s or Skirrow’s antibiotic supplement (II) (Oxoid, Basingstoke, UK). However, at Glaxo, we have developed a novel antibiotic supplement that inhibits all contaminating organisms liable to be isolated from the mouse gastrointestil?al tract. The composition of this medium is as follows: Columbia chocolate agar with 5% horse blood containing: Amphotericm B 50 pg/mL Vancomycin 100 pg/mL Polymyxin B 3.3 pg/mL Bacitracm 200 pg/mL Nalidlxlc

acid

10.7 pg/mL

Note that it is important to make up agar plates no more than 24 h in advance of when required.

2.4. Liquid Culture Media Brucell,abroth cultures containing 2% fetal bovine serum (FBS) (10) or tryptose soya broth containing 5% FBS (12) are normally used for challenge mocula. 3. Methods 3.1. Karifa Model (10) 3.1.1. Challenge lnoculum Athymic nude Balb/c mice (6 wk old) and their euthymlc counterparts were challenged with broth grown cultures of H. pylon (Brucella broth containing 2% FBS) grown m shake flasks (120 rpm) maintained at 37°C for 48 h in a gas mixture of 80% Nz, 15% COz, and 5% Oz. with gas replacement at 24 h. For optimal results, a single 2-mL challenge volume of broth containing approx 1O8 organisms/‘nL was administered orally to each mouse after 1 d of fasting. This extremely large volume would significantly dilate the stomach, thus permitting longer and more effective contact between the bacteria and the mucosal surface. 3.1.2. Evaluation of Infection Prior to sampling for gastric colonization, groups of 5 mice were fasted for 48 h. Samples (5 x 5 mm) were then cut from the posterior wall of the fundic and pyloric: regions of each stomach as well as the duodenum. The tissue pieces

238

McColm

were washed with 1 mL phosphate buffered saline (PBS) before being homogenized m 0.5 mL PBS. Quantitative evaluation of the numbers of H. pylon/ “unit tissue” was performed by inoculation of 0. 1-mL aliquots of the homogenate onto blood agar containing vancomycin (10 mg/L), polymyxm B (2500 II-I/L), and trimethoprim (5 mg/L) and incubation for 4 d at 37OC in the same gas phase as in Section 3.1.1. Colonies were identified as H pylori by then characteristic morphology and by standard criteria such as Gram stammg, production of urease, and so on. In addition to tissue samples for mtcrobiology, which were taken from the postertor wall of the stomach, samples similarly obtained from the anterior wall were fixed with 10% formalin and assessedfor histopathologtcal changes using standard hematoxylin and eosin (H&E). 3.1.3.

Characteristics

of the Infection

No colonization occurred with the type strain NCTCI 1637; however, all 3 clinical H pylori isolates evaluated showed colonization of the fundic and pyloric regions of all nude mice for at least 20 wk. Progressive increases in bacteria numbers (to approx 103/umt tissue) occurred up to wk 4 but, thereafter, numbers stabilized. No bacteria were seen m the duodenum of nude mice. Conversely, H pylori was seen in all three locations in the euthymic animals but numbers were significantly lower (Cl 02) and only persisted for 2 wk. Therefore, it is evident that, in these conventional mice, only transient low grade colonization occurs and it is posstble that the presence of H pylori in the duodenum reflects the gradual elimination of the bacteria. Evtdence for a progressive histological gastritis with mucosal erosions, capillary dilatation, and infiltration of monocytes was reported in this model. A major disadvantage of thts model, however, is that in order to achieve a level and longevity of infection suitable for therapeutic evaluation studies, athymic nude animals are required. 3.2. Marchetti Model (11) This model represents a significant advance, as colonization was demonstratedfor at least 8 wk m both SPFCD I mice aswell asCV CD 1and Balb/c mace. 3.2.1. Challenge lnoculum After fasting for 24 h, mice were given 0.25 mL of 0.2M NaHCOs solutron to neutralize acidity immediately prior to challenge with either freshly isolated or lab strains of H. pylori. The bacteria were either Type I strains expressing vacuolating cytotoxin (VacA) and cytotoxin-associated antigen (CagA) or Type II strains expressing neither. The challenge protocol differed from that of Karita (10) in three main ways: the volume was considerably less (oral admmtstration of lo9 colony forming units [CFU] in 0.15 mL sterile sahne); the frequency of challenge was greater (d 0, 3, and 5) and, instead of using broth

Nonprimate Models of H. pylori Infection

239

grown cultures, the bacteria were grown on Columbia agar containing 5% horse blood, 0.:!5% cyclodextrin, and Dent’s or Skirrow’s antibiotic supplement.

3.2.2, Evaluation of Infection Animals were fasted for 24 h before sampling. After removal of the stomach and opening along the lesser curvature, the nomnucosal, squamous forestomach was discarded before gentle streaking of the entire remaining mucous surface onto an agar plate. Although some contaminating Gram-negative bacteria were seen,posilive identification of H. pylori could be made after 3-5 d culture usmg Gram stain, urease, and catalase testsand morphological criteria. As in Karita’s model, gastric pathology was evaluated by standard H&E methods on small samples of the mucosa.

3.2.3. Characteristics of the Infection As in the Karita model (IO), the established laboratory strain (NCTC11637) failed to colonize mice, whereas fresh clinical isolates of both Type I and Type II strains showed detectable colonization in some ammals by 1 wk postchallenge and in all animals by 4-8 wk after challenge. Bacteria recovered from 2-wk infections were more efficient than clinical isolates in ensuring colonization with all animals exhibiting detectable infections within 1 wk. Bacterial numbers ranged from 5 x lo3 to 2 x IO4 CFU/lOO mg gastric mucus. Infection with Type I strains provoked gastric pathology similar in some, but not all, respects to human gastritis. This was present in some mice by 2 wk and in all by 8 wk and was characterized by loss of gastric gland architecture, production of epithelial erosions, and inflammatory cell infiltration into the lamina propria. Conversely, Type II strains only produced a mild histopathology with slight inflammatory infiltration; however, there was no indication that the extent of H. pylori colonization was any less (II). In addition, serum antibodies could be detected from 4 wk onward. The principal antigens recognized appeared to be a 55-kDa protein (possibly analagous to the heat shock protein, Hsp60, recognized in human H. pylori infections) and a 26-kDa band that was probably the UreA subunit of urease; however, the 128-kDa CagA antigen, normally detected in human infections, was not recognized. 3.2.4.

Use of the Model

Despite possible differences in H. pylori antigen recognition by the mouse immune response compared to in humans, this model reproduces many of the characteristics of human H. pylon’ infections and offers considerable potential for comparative treatment studies. Marchetti et al. (11) have evaluated whether oral vaccination with bacterial lysates or purified antigens (urease or VacA) in the presence of heat-labile enterotoxigenic toxin (LT) of Escherz’chia coli could

240

McColm

Table 1 Host and Bacterial/Culture Factors Examined in Developing the Mouse H. py/ori Model

Host factors Mouse stram Status(e.g.,age;CV, germ-free [GF], gut decontaminatedmice) Immunosuppression Achlorhydrta

Bacterial/culturefactors H. pylorl stram (fresh clinical isolatesor lab strains) Solid agar grown vs broth grown cultures Age of challenge moculum (e.g.,24,48 h, and so on) Effect of repeatedin vitro subculture Challengesize,frequency,numberof bacteria, and so on Selective antibiotics in sampling media

prevent colonizatton. Their procedure, based on prior studies performed with H. felis (19,20), was as follows: on d 0, 7, and 14, groups of SPF CD1 mice

received an oral inoculum of 0.5 mL saline contammg 10 ug LT with or without antigens (etther bacterial lysate equivalent to lo9 bacteria, 100 pg urease, or 100 ug VacA). On d 2 1,23, and 25, all animals were challenged with 1O9CFU H. pylori and colonization assessedon d 35. Approximately 80-l 00% protection was observed in the antigetiadjuvant groups; however, LT by itself also elicited an overall nonspecific protection rate of 17.8%. Purified VacA obtamed from Type I bacteria only protected against Type I challenge and not Type II. 3.3. Giexo Model (12) The fact that fresh clinical isolates or mouse adapted bacteria were necessary for colonization m previously reported studies (10, II) suggeststhat “quality” of the challenge inoculum is vital for success.Similarly, the observation that athymlc animals kept m a sterile envtronment become colonized to a greater extent than conventionally housed euthymic ammals (10) mdicates that the status of the host mouse 1sImportant also. For example, the conventional (CV) mouse stomach already harbors a background population of several indigenous bacterial and fimgal species along with a huge range of lower gut organisms introduced via coprophagy. This not only presents a physrcal barrier to gastric colonization with H pylori after oral challenge, but also may lead to severe contamination of the cultures used to quantify challenge success,Contamination can be &nimlzed by use of the selective supplement described in Section 2.3. We have embarked on a program of studies to examme a range of host and bactertal/cultural factors liable to influence colonization of mice with H pylon (Table 1). The conclusions obtamed are listed m the Notes section; however, the procedures for establishing the basic model are detailed m Sections 3.3.1. and 3.3.2.

Nonprimate Models of H. pylori Infection

241

3.3.1. Challenge inoculum Plate grown cultures of a clinical isolate of H pylori (designated stram 4 187E), obtained from a patient with duodenal ulcer, were used to Inoculate 5-mL volumes of tryptose soya broth in 25-mL capacity flat-bottomed tissue culture flasks. The cultures were incubated for 24 h without shaking in a humidified microaerobic workstation (Don Whitley, Shipley, Yorks, UK). They were then bulked together and subcultured by addition of 5 mL culture to 20 mL fresh broth; incubation was continued in 75-mL capacity flasks. After a further 24 h, the resultant cultures were then used to challenge mice and to initiate further 25-mL volume cultures as described 3 lines above for a second challenge. In order to provide enough challenge moculum to infect 180 mice with 1 mL of broth administered on 2 successive days, it is necessary to establish initially 12 parallel 5-mL cultures. These are then subcultured into 10 x 25-mL cultures, which are in turn subcultured into 8 x 25-mL cultures for the second mouse challenge. Viable counts of bacteria are performed on each of the 2 challenge inocula. Normally 5 x 107-5 x lo* bacteria are administered in each challenge. 3.3.2. Evaluation of infection Although quantitative counts of H. pylori can be made by standard homogenization and serial dilution techniques of either whole stomachs or circular “punch biopsies,” this can be time consuming for large numbers of animals. For therapy screening purposes, for example, it is desirable to use as simple and rapid a procedure as possible: after rapid killing of the mice by COz, the stomachs are excised and opened from the pylorus upward along the greater curvature. After crude removal of the contents by forceps, the stomachs are washed briefly in 2 changes of tap water, blotted dry on tissue, and the entire mucosal surface is then spread evenly for about 10 s over the surface of a Columbia chocolate agar plate containing selective antibiotics. The cultures are then immediately placed in a variable atmosphere incubator (VAIN, Don Whitley, Shipley, Yorks, UK) and incubated in a gas mix of 5% 02, 85% Nz, and 10% CO2 at 37°C for 7 d. Two criteria are used for assessing colonization: 1. Percentmice colonized (i.e., numberof H. pylori-positive culture plates) 2. Percentbacterial score(a semiquantltativeanalysisof the numbers of H. pylorz within the stomach--seeTable 2). 3.3.3. Characteristics of the Infection Over the last 5 yr we have challenged several thousand mice with H. pylori and have achieved a colonization success rate of approx 98%. Whereas the

242

McColm Table 2 Scoring System for Visual Assessment No. of Colonies Present 0 l-20 >20450 >50-<100 >1004200 >200

of Cultures Scorea 0 1 2 3 4 5

OFor a group of 10 mice, the maximum score is 50. The scores of all treatment and control groups are expressed as a percentage of this value so that comparisons of compound efticacy, as well as the relative severity of the test, can be made. We have found that this evaluation method is quicker, simpler, and as accurate as performing whole stomach viable counts of H. pylori. In an ideal test, the high-dose clarithromycin (50 mg/kg) control treatment will eradicate most or all infections, while the low dose (5 mg/kg) should have little or no effect (see Table 7).

majority of our work has concentrated on the use of HSD/ICR mice challenged with the 4187E strain of 2X pylori, many other strains of mouse and bacteria have been investigated. Most mouse strains examined will support at least some colonization, whereas most fresh clinical isolates of bacteria are similarly able to colonize (12). Like previous workers (ZO,lZ), we have been unable to show colonization with the established laboratory type strain, NCTCl1637. Bacterial numbers rarely exceed 1O5CFU/stomach irrespective of mouse or bacterial isolate; however, once established, the bacteria may remain for several months. Very little significant gastritis was observed in our model-again irrespective of host or challenge strain. In some cases,mild cellular infiltrates and lymphoid aggregates were seen; however, in general, there was little difference from control, unchallenged animals even after several months’ colonization. Despite the absence of a meaningful inflammatory response, serum antibodies (IgG and IgA) are detectable in colonized mice from approx 4 wk onward. Histological evidence suggested that the bacteria were mostly located on the mucosal surface in or under the mucus layer. Although all regions of the stomach (and the duodenum in a very few animals) support H. pylori colonization, quantitative viable counts indicated that there seemed to be a marked predilection for the cardiac antrum area (12).

Nonprimate Models of H. pylori Infection Table 3 Screening and Initial

Procedure for Identification Selection of Novel Anti-Helicobacfer

243

Agents

Day

Procedure

051

Challenge of female HSD/ICR mice (20 g body wt) p-0. with a I-mL vol of Tryptose Soya broth containing 107-lo8 CFU of H. pylori (4187E). This results in successful long-term gastric colonization of 98-l 00% of the mice. Treatment of mice b.d., p.o. (i.e., 8 doses in total). Novel compounds are tested initially at 5 and 50 mglkg and compared with the standard compound, clarithromycin (given at the same doses). All treatments and formulation/water controls are administered at 0.2 ml/mouse/dose. 10 mice/group are used. Mice sampled for H. pylori. The stomachs are removed, opened, washed briefly in tap water and the entire mucosal surface spread evenly for about 10 sec. over the surface of a Columbia chocolate agar plate containing selective antibiotics? Cultures are incubated at 37°C microarobically for 7 d. Culture plates evaluated for H. pylori growth and results expressed as % mice colonized and % bacterial score.

5-8 (inclusive)

12

19

aAmphotericin B (50 &mL), vancomycin (100 pg/mL), polymyxin tin (200 &mL), and nalidixic acid (10.7 I.lg/mL). b.d., twide daily; p.o., orally.

B (3.3 pg/mL), bacitra-

3.3.4. Use of the Model in Compound Screening The simplicity in establishing our model, coupled with the reproducibility of response to challenge, enabled us to devise a rapid screening strategy for identification and initial selection of novel anti-Helicobacter agents (21). The

standard screening procedure is in Table 3. 4. Notes

4.1. Factors Affecting Colonization of Mice with H. pylori The model as described provides a reproducible and simple method for ensuring uniform

colonization

of animals suitable for evaluation

of therapies;

however, various host, bacterial, or cultural factors may influence outcome. The main conclusions are summarized below: 1. Early studies showed that germ-free (GF) mice or mice treated with antibiotics to partially decontaminate the gut were easier to colonize than conventional (CV) animals (see Table 4). However, for practical purposes, use of CV mice is easier

and less labor intensive and will give 100% colonizationrates provided that the conditions stipulated in Sections 3.3.1. and 4.2. are adhered to.

McColm

244

2. No apparent colomzatton improvement occurred m achlorhydrtc mice-therefore there is no requirement for pretreating with H, antagonists. 3. Immunosuppression with cyclosporm or use of SCID or nude mice did not improve outcome. 4. Most strains of mouse tested (21 m number) supported at least some colomzation; however, the readily obtainable outbred strain, HSD/ICR, consistently showed the highest colonization rates. 5. Most clmical isolates of H. pylorz tested were able to colonize at least some mice. No attempt was made to differentiate these into Type I or Type II strains. Strain 4187E was chosen as it consistently demonstrated the best colonization rates. Note that lab strains such as NCTCI 1637 do not readily colonize. 6. No correlation appeared to exist between percent colonization and the level of challenge moculum, as some of the H pylon strams, which grew best in vitro, performed less well at infecting mice (see Table 5)-this confirmed that factors such as bacterial strain and challenge quality are more important than absolute numbers, e g* a. Broth grown bacteria > plate grown bacteria. b Mimmum challenge inoculum: lo6 CFU/mouse c Media Ttyptose soya broth (TSB) > brain heart infusion broth (BHI) > brucella broth. d Age of challenge culture: 24 h broth > 48 h > 72 h. e Stattc cultures > shaken cultures. f. Volume of challenge:- 2 or 1 mL > 0.5 mL > 0.25 mL. g. Frequency of challenge. 2 or 3 challenge doses on successive days > 1 dose (>, better than). Of these parameters, age of culture is of paramount importance. Despite the progressive increase in bacterial counts m broth cultures after 24,48, and 72 h, there is a corresponding decrease in ability to colonize mice. In addition, shaken cultures were less successful in colonizing than static cultures, despite supporting higher bacterial numbers. These observations provide further evidence that culture quality rather than bacterial numbers is critical to success. Note that for optimum results, it 1s important to mrmmrze the time between harvesting cultures and challenging mice. Challenge volume did, however, seem to be important as equivalent numbers of bacterra administered as a large volume colonized better than if given as a smaller volume. This indicates that the greater gastric distension and likely decrease in gastric emptying rate accompanying a large volume permit a more effective contact of bacteria with the mucosa. 4.2. The Importance

of Urease

Despite adhering to the above conditions, rt was noticed that H pylori subcultured on agar in vitro for periods greater than 4-6 wk often displayed a

Nonprimate Models of H. pylori lnfection

245

Table 4 Effect of Mouse G.I. Tract Flora on Gastric Colonization with /f. py/ori Mice colomzed with H pylori”, O/O

Mouse type/strain GF (CD- 1) CV (HSDIICR) CV (Antibiotic

treatedjb

95 61 82

9 mo postchallenge (mice given approx 5 x lo7 CFU of a mixture of 27 clmical H. &orr Isolates on d -2 to 0). bBacitracm. 4 mg b.d./mouse p o Polymyxm B* 0 5 mg b.d./mousep.o All antibodiesgwen d -8 to -4 prechallenge Cefirlaxone

0.01% in drinking water. Table 5 Percent Colonization of HSCVICR Mice with Individual If. pyloristrains

H pylori strain Flegarty Din (4 187E) Cl27 C264 reduced ability

Challenge (log 10 CFU/mouse)

Mice colonized atd21,%

8.3 7.5 9.7 6.0

30 80

to colonize mice when subsequently

10

60 grown in broth (Table 6).

Infectivity could be restored, however, by exposure of the bacteria to acid conditions in the presence of physiological urea (5 mA4)as follows: 0.1 mL 150 mM urea solution, 2.6 mL citrate phosphate buffer, pH 2, and 0.3 mL H. pylori suspension (plate grown, suspended in saline, - 109/mL). Incubation at 37°C is carried out for 30 min maximum followed by culture of surviving bacteria on agar as normal; note that the H pylon suspension must be added last to the incubation mix. This enhances the actlvlty of urease, an important factor m colonization (Z5), and selects those bacteria showing the highest urease posltivity. However, boosting urease activity alone is not the sole determinant for restoring full infectivity of cultured H. pylon (Table 6) as other factors, e.g., motility, may be relevant. Therefore, in order to maximize the chances of infectivity, it 1s best to preserve replicate ampoules of the original stock in liquid nitrogen, remove one as required, and acid treat and maintain in vitro as a source of challenge mocula for 1 mo maximum.

McColm

246 Table 6 Restoration of H. pyIori4167E Infectivity for Mice Using Acid/Urea Treatment and Use of a Fresh Isolate0

(Number [%] mice colonizedby different challenge mocula) Old N 1821370 49 2%

Old (AT) 1:lO 241120 20%

N 85/106 80.2%

1:lO 10140 25%

New N 138/144 95.8%

1:lO 63t82 76.8%

l*lOO 16139 41%

%ummation of several expertments Old, bacteria mamtamed for at least 1 mo in vrtro by repeated subculture. Old (AT), as above but exposed for 30 mm to pH 2.0 citrate/phosphate buffer contammg 5 mA4urea. New, fresh Isolate from hqurd nitrogen of ongmal stock, acid treated and mamtamed m vitro for 1 mo maximum N, neat challenge 1 10, 1O-fold drlutton of neat challenge

As a result of the above model development studies, the following standard protocol for reproducibly achieving 100% colonization rates is vigorously adhered to: HSD/ICR mice are challenged p.o. on 2 successive days with 1 mL of a 24-h tryptose soya broth (TSB) culture of H. pylon (4187E) containing 107-1 OS CFU/mL. Bacteria passaged for no more than 1 mo in vitro (chocolate Columbia agar subcultures) and, exposed to urea/acid treatment to stimulate urease activity, will colonize better than older cultures, 4.3. Selective Medium Note that before using the selective antibiotic supplement described in Section 2.3., it is very important to confirm the tolerance of the H. pylori isolatesbemg used to nalidixic acid. Some isolatesare sensitive to concentrationsaslow as 10.7 pg/mL, and it may be necessaryto reduce this slightly and increase polymyxin B levels, 4.4. Use of the Mouse

H. pylori

Model

in Compound

Screening

The aim of this test is to identify compounds that will eradicate H. pylori as rapidly aspossible; therefore, a short (4 d) regimen was chosen. Clarithromycm is used as the positive control standard since it shows the best eradication figures when administered clmically as monotherapy. When tested at 5 and 50 mg/kg x 8, the pattern of response is remarkably consistent, showing 100% of mice remaining colonized at the low dose and 0% colonized at the top dose. Any deviation from this would indicate a less or more severe test, which should be kept m mmd when evaluating the activity of novel entities.

Nonprimate Models of H. pylori Infection

247

Table 7 Activity of Standard Agents In H. pylori Infected Mice Compound Clarithromycin

Azithromycin

Water control

Dose, mgfkg x 8

Mice colonized, %

Percent score

5 25 50 5 25 50

100 30 0 100 60 0 100

84 14 0 76 26 0 94

0.2 m.L x 8

As this test does not have the facility to examine the phenomena of suppression, recrudescence, and so on, it is regarded as an “all or nothing test” in which eradication is the end point. Since isolation of just one H. pylori colony from a mouse stomach is sufficient to assign a positive result, it is important to have a secondary readout of therapeutic response. The percent bacterial score criterion is useful in this regard for helping to discern more subtle differences in compound action. In order to assessand validate our mouse model of H. pylori, we have investigated its ability to rank standard antibiotics, detect synergistic drug combinations, compare route of compound delivery, and evaluate various factors likely to affect therapeutic outcome, e.g., achlorhydria, use of gastric emptying inhibitors, and so on (21). As these studies have confirmed the suitabihty of the model for compound evaluation, we have used it so far to screen several hundred novel structures with proven in vitro anti-He&obacter activity. Table 7 shows that clarithromycin is slightly superior to azithromycin in this model, thus reflecting in vitro and clinical action. 4.5. Other H. pylori Animal Models 4.5.1.

Pig Model

The first reports of experimental H. pylori infection of pigs used piglets deltvered by Caesarian section into sterile isolation units (7,13). These gnotobiotic animals could then be challenged with H pylon and maintained for up to 2 mo in germ-free conditions, during which time the extents of colonization and gastritis were monitored. Animals are generally challenged within the first 4-5 d of life with broth grown bacteria after temporary mduction of achlorhydria (e.g., cimetidine, 60 mg/kg once daily for 2 d) and fasting for 12 h. The infection was characterized by a progressive increase m luminal gastric mucus from 2 wk onward, some mtld transient diarrhea 2 d postchallenge, and the

McColm development of prominent submucosal and mucosal lymphoid follicles (nodules) by 4 wk (7). Histopathological chronic active gastritis occurred m all piglets and increased in intensity with time. This consisted mitially of neutrophilic aggregates and some monocytes present mostly in the nonglandular regions of the cardia. However, from 2 wk onward, mononuclear cells predominated, formmg large lymphoid follicles in the submucosa and lamina propria. This was accompanied by the identification of H. pylori, mostly m the superficial mucus secreting area of the gastric epithelmm. Whereas gnotobtotic pigs reproduce many of the features of diseases assocrated wtth H. pylon’ in humans, they have only llmtted usefulness owing to the difficulties m mamtaimng them within isolators for long-term studies (maximum of 45-60 d). A significant advance was the use of barrier born pigs, which, although free of specific pathogens at the time of H. pylon challenge (e.g., 8 wk of age or more), could subsequently be kept under normal experimental conditions for significantly longer periods (more than 6 mo), thus permittmg long-term evaluation of the infection (1415). A characteristic of the pig model is the relatively patchy distribution of I-I. pylon over the gastric mucosa (15), which may result in false negative results for cultured endoscopic biopsies. A more accurate method is to carry out precise topographic mapping of urease distribution on the gastric mucosa. This necessitates killing the animal and marking the stomach into large numbers of equivalent sections using a gridded template (IS). From each section (2.1 cm2 area), 2-mm diameter punch biopsies are cut out and placed mto 1 mL of 2% urea solution containing 0.001% phenol red. Change m color from yellow to pink is recorded at 559 nm in a spectrophotometer and is scored as follows: if absorbance reaches 0.600 within 60 min, the urease reaction is graded as “strong;” after a period of 3 h, any pmk samples with absorbance lower than 0.600 are classed as “weak” reactions. Although urease mapping results are not always in agreement with those of parallel cultures, the sections showing strong urease reactions tended to also reflect viable counts. A drawback of urease mapping is the need to kill the pig before evaluation; therefore, longitudinal studies of the infection course over time are not possible in the same animal. However, the development of a 13C-urea breath test in barrier born pigs (22) has countered this by allowing accurate long-term evaluation of infections m conscious animals without recourse to anesthesia and invasive endoscopy. The method is described as follows (22): Pigs infected with H. pylori are fasted overnight before breath testing. Initial experiments were done m anesthetized animals that had been incubated endotracheally; however, by use of a modified anesthetic mask covering the snout, samples can easily be collected from conscious pigs. Exhaled breath is collected over a 1-min period into apolythene bag (Scantec,Sweden) both before

Nonprimate Mode/s of H. pylori Infection

249

and after the administration of 13C-urea (2 mg/kg dissolved in water and admmistered by stomach tube). Breath 1susually collected at 20,40, and 60 min after urea administration, and samples are removed from the bag by syringe mto 20-mL vol vacutainers for analysis by mass spectrometry. The ratio of 13C/12Cin exhaled COZ is expressedas ppm excessof 13Cover the natural existing baseline value. A result exceeding 5 ppm is considered positive. In control pigs, the median excessof 13C02in expired air was found to be 2.2 (range O--12),whereas H pylori infected pigs showed significantly higher values (median 23, range lU3) when examined on different occasions 4-l 0 wk after infection (22). 4.5.2. Dog and Cat Models Gnotobiotic beagle pups have been successfully colonized with brucella broth grown H pylori and maintained in sterile isolation units for up to 1 mo (3). The method used a human isolate known to cause gastritis in gnotobiotic piglets and consisted of a single 2-mL oral challenge of 2 x lo* bacteria from a 24-h, 250-mL shake flask culture. The bacteria were washed and suspended m peptone water prior to challenging the animals when they were 7 d old. Although this model displays both lower overall gastic H. pylori densities and a different pattern of bacterial distribution (a predilection for the fundic mucosa) compared to m humans or gnotoblotic piglets, dogs can be maintained in gnotobiotic conditions for several years, thus allowing more prolonged studies to be performed. The model has been used to examine contact transmission from infected to naive dogs (3). More recently, the domestic cat has been identified as a natural host for H. pylori infections (16,Z7). A preliminary commumcatlon (28) has also reported experimental colonization of specific pathogen-free cats with broth grown cultures of H. pylori (3 x 3-mL challenges administered on alternate days to cimetidine-treated animals). The animals developed multifocal gastritis, comprising lymphoid aggregates and nodules mostly in the antral mucosa with H pylori organisms visible within the glandular crypts. Cats are the only other species except for certain species of nonhuman primates to have been positively shown to naturally acquire H. pylori infections. Consequently, this model is an important advance offering considerable scope for examination of mechanisms of transmission, pathogenesis, and perhaps factors important in the long-term progression to atrophic gastritis and gastric carcinoma. In addition, it may well function as a suitable secondary stage therapy model for further evaluation of compounds selected in a primary mouse test. References 1. Cantorna, M. and Balish, E (1990) Inability of human clinical strains of Helicobacterpylori to colonisethe alimentary tract of germ-free rodents.Can J Microbial.

36,237-241.

McColm

250

2 Ehlers, S., Warrelmann, H., and Hahn, H. (1988) In search for an ammal model for experimental Campylobacter pylon infection. admmtstration of Campylobatter pylorz to rodents. Zentralb Bakteriol Mzkrobzol. Hyg 268,341-346 3. Radm, M. J., Eaton, K. A , Krakowka, S., Morgan, D. R , Lee, A , Otto, G., and Fox, J (1990) Helicobacter pylorz gastric infection m gnotobiotic beagle dogs. Infect Immunol

58,2606--26 12.

4 Yokota, K., Kurebayashi, Y , Takayama, Y., Hayashi, S., Isogai, H , lsogat, E , Imai, K., Yabana, T , Yachi, A , and Oguma, K (199 1) Colonisatton of Helzcobatter pylori in the gastric mucosa of Mongolian gerbils Microbzol Immunol 35,475-480. 5. Fox, J. G., Cabot, E. B., Taylor, N. S., and Laraway, R. (1988) Gastric colomsation by Campylobacterpylori subsp. mustelae m ferrets. Infect Immun. 56,2994-2996.

6. Fox, J. G., Correa, P., Taylor, N. S., Lee, A., Otto, G., Murphy, J. C., and Rose, R. (1990) Helzcobacter mustelue-associated gastritis in ferrets. An animal model of Helicobacterpylorz gastritis m humans. Gastroenterology 99,352-361. 7 Krakowka, S., Morgan, D. R., Kraft, W. G., and Leunk, R. D. (1987) Estabhshment of gastric Campylobacterpylori infection m the neonatal gnotobiotic piglet Infect Immun 55,2789-2796

8 Lee, A , Fox, J. G., Otto, G., and Murphy, J (1990) A small animal model of human HeZicobacter pylon active chronic gastritis. Gastroenterology 99, 1315-1323. 9. Fox, J. G., Lee, A., Otto, G., Taylor, N. S., and Murphy, J. C. (1991) Helicobacter fells gastritis in gnotobiotic rats: an animal model of Helicobacterpylori gastritis Infect Zmmun. 59,785-791.

10 Karita, M., Kouchiyama, T , Oktta, K , and Nakazawa, T. (1991) New small animal model for human gastric Helicobacterpylori mfection: success in both nude and euthymic mice. Am J Gastroenterol 86, 1596-1603 11. Marchetti, M., Artco, B., Burrom, D., Figura, N., Rappuoh, R., and Ghiara, P (1995) Development of a mouse model of Helzcobacterpylorz infection that mimics human disease Sczence 267,1655-1658 12. McColm, A. A., Bagshaw, J. A., O’Malley, C., and McLaren, A. (1995) Development of a mouse model of gastric colomsation with Helzcobacter pylon. Gut 37(Suppl. l), 198. 13 Lambert, J. R (1987) Colonisatton of gnotobtottc piglets with Campylobacter pyloridis-an animal model? J Infect Dzs 155, 1344. 14. Engstrand, L. S., Gustavsson, A., Jorgensen, A, Schwan, A., and Scheynius, A (1990) Inoculation of barrier-born pigs with Helzcobacterpylorz a useful animal model for gastritis type B Infect Immun. 58, 1763-1768 15. Engstrand, L. S , Rosberg, K , Hubmette, R., Berglmdh, T., Rolfsen, W , and Gustavsson, S. (1992) Topographic mapping of Helicobacter pylori colomsatton in long-term infected ptgs Infect. Immun 60, 653-656. 16. Handt, L K., Fox, J. G , Dewhirst, F E., Fraser, G J , Paster, B. J , Yan, L L., Rozmiarek, H., Rufo, R , and Stalis, I H. (1994) Helicobacter pylorz isolated from the domestic cat public health imphcations. Infect Immun 62,2367-2374.

Nonprimate Models of H. pylori Infection

251

17. Handt, L. K., Fox, J. G., Stalls, I. H., Rufo, R., Lee, G , Lmn, J., Lt, X., and Kleanthous, H. (1995) Characterisation of feline Helzcobacter pylori strains and associated gastritis m a colony of domestic cats. J. Clm. Mlcroblol. 33,2280-2289. 18. Fox, J. G., Batchelder, M., Mark, R. P., Yan, L., Li, X., Shames, B., Hayward, A., Campbell, J., and Murphy, J. C. (1995) Hehcobacterpylorl induced gastritis m the domestic cat. Gut 37(Suppl. l), 206 19. Chen, M , Lee, A., and Hazel& S. L. (1992) Immumsation against gastric Helicobacter infection m a mouselHelicobacter felis model Lancet 339, 1120,112l. 20. Ferrero, R. L., Thiberge, J. M., Huerre, M., and Labigne, A. (1994) Recombinant antigens prepared from the urease subunits of Helicobacter spp. : evidence of protection in a mouse model of gastric mfectton. Infect Immun. 62,498 1-4989. 2 1. McColm, A. A., Bagshaw, J. A., Wallis, J., and McLaren, A. (1995) Screening of anti-Hebcobacter therapies in mice colonised with H. pylon. Gut 37(Suppl. l), 367 22. Meyer-Rosberg, S. and Gustavsson, S. (1993) 13C-urea breath test for diagnosis of experimental Hellcobacterpylori in barrier born pigs. Gut 34,594-598.

24 The Nonhuman

Primate Model for H. pylori Infection

Andre Dubois and Douglas E. Berg 1. Introduction Helicobacter pylori establishes chronic mfections in the human gastric mucosa that can last for decades, and that are a major cause of gastritis and peptic ulcer disease (PUD) and a risk factor for gastric cancer. The importance of H. pylon as a human pathogen has led to major efforts to understand its interactions with its human host, including the development and use of various animal models. Several heterologous models have been developed, Including:

1. Gnotoblotlc newborn piglets, which are easily infected by H pylon of human origin (1,2), but are best for short term studies lastmg no more than about a month, 2. Mice and ferrets, which can be colomzed for months and years, respectively, although most easily by Hellcobacter species other than H pylori (3,4), 3. Particular strains of mice, which can be colonized by selected H pylon strams (56); and 4. Domestic cats, which can also carry H. pylon (7), but which exhibit an mflammatory response different from that seen in humans (mononuclear mfiltratlon deep near the lamina propna, not throughout the mucosa) and which also differ m diet and gastric anatomy Although each of these models has been quite valuable for specific studies, each is also limited by major differences from humans m gastric anatomy, physiology, diet, receptors for H pylori adherence, immune and inflammatory responses, and/or inability to repeatedly endoscope the same animal. It IS in this context that nonhuman primates are of particular interest. Natural infections by bacteria that are morphologically similar to H pylori from humans have been reported in a number of nonhuman primates (8-l 3). Usually, these bacteria were not tested for their match to H. pylori from humans. In one case, however, isolates from naturally infected pigtailed macaques From Methods m Molecular Me&me, Hehcobacter pylori Protocols Edlted by C L Clayton and H L T Mobley Humana Press Inc , Totowa,

253

NJ

254

Dubois and Berg

(Macaca nemestrina) turned out to belong to another species, Helicobacter nemestrinae (14). In contrast, Helicobacter strains isolated from rhesus monkeys match H. pylori isolated from humans by 16s rRNA sequencing, as well

as antlgemcally and by morphology (1 I, 12,15). Rhesus monkeys harbormg at a given time remam infected for years. In addition, they have gastritis and elevated specific IgG, which resolve if the bacteria are eradicated by antimicrobial therapy. Experimental transmission of H pylori seemsto have been successful in chimpanzees(Pan troglodytes) (I 3), cynomolgus monkeys (Maraca fascicularzs) (17), Japanese monkeys (Macaca fuscata) (Id), and rhesus monkeys (Macaca mulatta) (2 7). However, some of these primates, such as chimpanzees, are not widely available. More importantly, in some monkey infections, recovered strains were found to not match the input strains or the match was not tested. Rhesus monkeys seem to be among the best suned of nonhuman primates as a model for human H. pylori infection. 1 Like humans,they areomnivora, eating small animalsas well asplant materials when living in the wild (18), 2. They provide a goodmodel for diseasesof the human stomachbecausethe physiology of their gastrrc secretron and emptying IS srmilar to that in humans H. pylori

(11,12,19,20);

3 Ulceratrons of the gastroduodenal mucosa have been observed in 2-5% of necropsies of rhesus monkeys (Parker et al. 1211 and unpublished results); 4. Gastroduodenal ulcers were also induced experimentally m them in early experrments (22), and resident H. pylori infection probably contributed to the forma-

tion of theselessons,although this was not specifically testedat that time; 5. Cases of gastric adenocarcinoma and lymphoma have been reported sporadically, especially in older rhesus monkeys (23); 6 Monkeys can be examined repeatedly for months or years using videogas-

troscopy,biopsy, mtubatron, and scmtigraphymethodsthat were developed for human gastromtestmal studies (12,19,24), 7. More specific immunological reagents are available for rhesus monkeys than for any other nonhuman prtmate, and they are widely used for many types of btomedical studies; and 8. We recently demonstrated that rhesus monkeys can, m fact, be experimentally Infected with H pylon strains isolated from humans or from rhesus monkeys (25)

The gastric mucosa of rhesus monkeys, like that of humans, can also be colonized (and sometimes cocolonized) by a different Helicobacter orgamsm that closely resembles “Gastrospirillum hominis ” (also called H. heilmannii) (I I, 12). G. hominis is generally considered nonculturable, although the culture of one such G hominzs has just been reported (26). Analysis of DNA and rRNAs of rhesus monkey and human tissues heavily colonized by G. hominis indicates that there are probably many bacterial species included in this

Nonhuman Primate Model for H. pylori

255

denomination, and that tt is premature to propose an official name (27,28). The name G. hominis will be used to represent these various species m this chapter. Because the rhesus monkey model of H pylori infection appears to be the most advanced and the most promising nonhuman primate at this time, this chapter will focus exclusively on this animal model.

2. Materials 1. Ammals Domestic, colony-reared, male rhesus monkeys (age 2-5 yr; weight 35 kg) were obtained from Laboratory Animals Breeders and Servtces, Yemassee, SC They had been bred, reared and socially housed either m indoor gang cages, in outdoor corrals, or in a large free-ranging colony on a sea island. H. pylori infection IS enzootic in that colony, with prevalence, mcidence, and age distribution similar to those observed among humans in third-world countries (‘29). 2. Videogastroscopies and gastric mucosal biopsies: We used an EG2700 Pentax (Orangeburg, NY) video-gastroscope with an outer diameter of 9.0 mm for the msertlon tube, and of 9.3 mm for the rigid distal portion. The depth of field is 5-100 mm, the angle of view is loo”, and the tip deflections are 2 l-l 20’ up-down and 12&120° right-left The diameter of the biopsy channel is 2 2 mm, and the insertion tube total length is 1051 mm. The videogastroscope is connected to an EPM-330 Pentax videoprocessor, and a Sony monitor, thus allowmg excellent visualization of the entire stomach as well as of the proximal duodenum Biopsy forceps 1.8mm cup drameter, 150 cm long (KW1815S, Pentax, Orangeburg, NY). 3. Freezing medium: TSB-glycerol, tryptic soy broth (Difco, Detroit, MI) supplemented with 20% glycerol 4. Culture Medium: As solid medium, Brain Heart Infusion agar (Difco), supple-

mented with 7% saponin-lysedhorse blood (Colorado Serum,Denver, CO) and antibiotics (6 mg Vancomycin, 5 mg Trimethoprim, 0.8 mg Amphotericm B per 1 L; all from Sigma, St. Louis, MO). As liquid medium, Brain Heart Infusion broth (Difco) supplemented with either 7% horse serum or 4% calf serum mstead of blood, and sometimes also with 1% isovitalex (Becton Dickmson, Cockeysville, MD). 5. Urease, catalase, and oxidase test solutions for H. pylorz tdentification: urease test solution, 10% urea in phosphate buffered salme, pH 7.2, and 1% phenol red; catalase test solution, 30% HZO,. Oxidase test solution, l%N,N,NZ,NI-tetramethyl-pphenylenediamme dihydrochloride m water, or a commercial oxidase kit (Difco) 6. Multiple H. pylori strains have been tested for ability to colonize monkeys (30,31). Before inoculation, primary isolates are stored at -70°C m saline or Brucella broth with 20% glycerol. Two days before inoculation of monkeys, the strains are transferred to flasks containing 25 mL of brain heart infusion broth + 4% fetal calf serum, which are then incubated with shaking for 2 d in an atmosphere of 90% NZ, 5% 02, and 5% COZ. On the day of inoculation, the cultures are centrifuged at 4’C and resuspended in Brucella broth at 10s-log H pylori CFU/mL, and mixed when indicated.

256

Dubois and Berg

3. Methods 1, Upper gastrointestinal videogastroscoptes. Perform gastroduodenal videogastroscopic examination under general anesthesia. Between each endoscopy, care must be taken to disinfect the vrdeogastroscope by soaking sequentially for 10 mm m a proteolytic enzyme solution (metrizyme, metrex) and then m an activated dialdehyde solution of 2% glutaraldehyde so&on (Cidex@, Johnson & Johnson Medical, Arlington, TX); rinse the instrument sequentially with sterile water before and after each solution, and then air dry The macroscoptc appearance of corpus and antral mucosae should be assessed qualuatively but, as previously reported (11,12), has not been stgmficantly related to any of the other features of infection. 2 Mucosal biopsies. Obtain pmch biopsies of the gastric mucosa from the corpus and from the antrum for histologrcal scoring of infection and of gastritis (12). These btopsies measure approx 1 mm in diameter, weigh 4-5 mg and should include the muscularts propria They can be oriented under low magmfication mrcroscope on pickled cucumber slices covered with a thin layer of albumin, and are fixed m neutral 10% buffered formalm and embedded in paraffin Five-micrometer thick sections are stained with Hematoxylm and Eosm (H&E) and viewed under 100-1000x magnification For culture, mnnedtately place m 0.1 mL of sterile 0.9% NaCl on me, code, and homogenize within 3 h with a sterile ground-glass cone-shaped pestle fitting a tapered 1.5-mL Eppendorf tube. Streak an aliquot (l-2 &) on Campylobacter chocolate agar plates supplemented with trimethoprrm, vancomycin, amphotericm B, and polymixm B (Remel, Lenexa, KS), and incubate at 37°C in an atmosphere of 90% NS, 5% Ox, and 5% CO* H pylon is identified: as forming pinhead-sized colonies that grow wtthm 7-10 d, as having urease, oxidase (Becton-Dickmson, Cockeysville, MD), and catalase activities; and by microscopy as Gram-negative and curved or gull-wing rods. Isolates are characterized biochemmally, antigenically, and genetically (see steps 4 and 5) 3. Determination of plasma IgG. Determine plasma IgG levels blindly usmg a previously described enzyme-linked immunosorbent assay (ELISA) with >95% sensitivity and specificity for human infection (12,32). 4 Rtbosomal RNA gene sequence of H. pylon isolated from rhesus monkeys (IS). The sequence of the cloned 16s rRNA gene can be determined and compared with 16s rRNA sequences reported for human isolates of H. pylon and for other Helzcobacter species. 5 DNA fingerprintmg To efficiently drstmgursh among H. pylon strams, the “arbitrarily primed PCR” or “random amplified polymorphtc DNA” (“RAP,,‘) fingerprmting method is used exactly as described (25, see Chapter 14) The PCR RFLP fingerprintmg method (15; see Chapter 12) can also be used (see Chapter 14). 6. Treatments to cure natural H pylon and “G hommrs ” infection A 10-d triple therapy of amoxlcillin + metronidazole + brsmuth subsalicylate (7,7, and 10 mg/kg, respectively) diluted in 5 mL of sterile water and admmistered mtragastrically twice a day for 10 d was partially successful We have recently found a 10-d

Nonhuman Primate Model for H. pylori

257

quadruple therapy wrth omeprazole, clarithromycin, amoxicillin, and peptobismol (0.4, 10, 7, and 10 mg/kg, respectively) given orally twice a day to be more successful as a suspension in water. 7. Follow-up examinations. Repeat endoscopies and plasma IgG determinations at regular intervals in animals naturally infected to verify persistence of infection. We have also endoscoped animals at 1 wk, and 1,2,3,5, and 6 mo after the end of treatment, or after inoculation. 8. Inoculation protocol (25). Treat monkeys with famotidine (Pepcid@, Merck, West Point, PA, 2 mg/kg, given i.m. 14 and 1 h before inoculation) to suppress acid output. After an overnight fast, the animals are endoscoped as described, phenol red is sprayed to estimate the pH of the gastric mucosa (generally a pH of 2-7), and 5 mL of 0.25M NaHCO, is introduced onto the antral mucosa to further neutralize gastric acid. Spray a suspension of 10slog CFU of H pylorl (1 mL suspension of each strain) onto the gastric antrum. Reendoscope the monkeys 57 d after inoculation, and generally at 3-5-wk intervals thereafter to harvest biopsies, at which time plasma is also obtained for IgG determmations (see Notes l-7 for analysis of data).

4. Notes 1 Natural infection of the rhesus monkey stomach. Colony-reared rhesus monkeys are often naturally infected with strams that match H. pylon, the easily distmguishable G homints (Figs. 1 and 2), or both, by usual taxonomlc criteria (15) As with humans, H pylori infection is often patchy, and natural or experimental infection 1s sometimes detected histologically, but not by culture tests, or vice versa. At least for H. pylori, the frequency of monkeys in the Yemassee colony that are infected increases with age: about 60% of monkeys 2 yr of age are H pylori infected, whereas some 90% of animals age 1 l-22 yr are Infected. Although many of the young H. pylori-free animals are G. hommrs, Infected, many older ammals carry both H. pylorl and G. hominu, suggestmg that colomzation by one pathogen does not seriously interfere with colonization by the other. In diagnosmg natural infection, it is important to note that G homuus and H. pylori tend to occupy different niches in the monkey gastric mucosa. As m humans, H. pylorr are generally present only m the mucus layer immediately adjacent to the superficial epnhelial cells, to which they can be attached through pedestal formation (Fig. 3A) Rarely, a few bacteria can be found inside superticial epithelial cells (Fig 3B). G. hominis are also visible in the mucus layer but do not attach to the superficial epithelial cells (Fig. 4), and they are most abundant in the acid-producing body of the stomach, sometimes within parietal cells As in humans, G. hominis is generally associated with less gastritis than H pylori, and it is often difficult to distinguish the presence of H. pylorz by histology in the presence of heavy colonization by G hominu. Therefore, mixed mfection can be missed if no culture is performed. Many authors have tried to culture G. homznzs from the stomach of humans and animals, but they were uniformly unsuccessful with one recent exception (26).

258

Dubois and Berg

Fig. 1. Photomicrograph of the gastric gland of an H. pylori-infected rhesus monkey (H&E; original magnification, x1000). Note the presence of 2-5 pm, l-3 loose spiral bacteria, associated to mucus depletion of superficial epithelial cells, and intense mononuclear and neutrophilic infiltration. 2. Ribosomal RNA gene sequence ofH. pylori isolated from rhesus monkeys (15). Analysis of the 127 nucleotide sequence (corresponding with residues 1240-l 366 of E. coli 16s rRNA gene) indicates that the H. pylori strains from rhesus monkey 85D08 and from humans are closely related, with 99.2-100% homology. In contrast, H. pylori 85D08 is less closely related to all other Helicobacter species examined, with 83.5-96.9% homology in this region, as expected (Table I). 3. Relationship between gastric inflammation and infection with gastric organisms. For all 26 animals tested (whether infected or uninfected), inflammation scores of the corpus and antrum were significantly correlated to each other (r = 0.69; p < 0.01). Therefore, the average of the two scores can be used as an overall index of gastric inflammation. As shown in Fig. 5, 13 rhesus monkeys infected with H. pylori that we studied, including three animals also infected with G. hominis, had mean scores 22.0, whereas three of four apparently uninfected rhesus monkeys that we tested had scores CO.5 (p = 0.005). Among the nine monkeys infected with G. hominis alone, eight animals had scores 11.0 (Fig. 5) and five of these nine rhesus monkeys had a score of 0 (Fig. 5), whereas all 11 rhesus monkeys infected with H. pylori alone had scores 22.0 @ < 0.001). Analyzing the data in another way, all 13 rhesus monkeys with mean gastritis scores 22.0 had detect-

Nonhuman Primate Model for H. pylori

Fig. 2. Gastric gland of a G. hominis-infected rhesus monkey (H&E; bar = 10 pm). These bacteria are much longer than H. pylori (5-10 pm), and have a markedly different shape (5-10 tight spirals). Note the absence of neutrophilic infiltration and the discrete mononuclear infiltration. able H. pylori infection compared to none of 13 with lower scores @ < 0.001; FET, 2-tailed). Figure 5 shows that a mean gastritis score 22.0 is 100% sensitive and 100% specific for H. pylori infection, and that the gastritis score of H. pylori-infected monkeys was significantly greater than that of uninfected animals or animals infected with G. hominis alone @ c 0.05). 4. Assessment of the serologic response to H. pylori. When using antimonkey antibody conjugates, H. pylori-specific plasma IgG levels are significantly @ c 0.05) higher in the H. pylori-infected animals compared to either the animals infected with G. hominis or the uninfected rhesus monkeys (Fig. 6). In addition, regression analysis showed that anti-H. pyEori plasma IgG levels are significantly correlated with gastritis scores for the entire group of 26 animals (r = 0.52; p < 0.01). Of rhesus monkeys without detectable H. pylori infection, 11 (85%) of 13 had IgG optical density ratios of ~0.5 compared with only 1 (8%) of 13 with detectable H. pyZori (p c 0.00 1, Mantel-Haenszel corrected chi square). The specificity of the ELISA is therefore 92%, and its sensitivity is 85%. In one of the two animals with IgG ratio >0.5 and no detectable H. pylori or infection with G. hominis, the gastritis score was 1.O in the annum and 2.0 in the corpus. The other rhesus monkey had a marked infection with G. hominis in the corpus (score 3.0) but not in the antrum and a gastritis score of 3.0 in the antrum and 0.0 in the

260

Dubois and Berg

Fig. 3. Transmission electron micrographs of H. pylori in rhesus monkeys. (A) Note the virtual disappearance of microvilli, the formation of pedestals (arrows), and the close association between H. pylori and these pedestals (original magnification, x 12,000). (B) Intact microvilli and intracellular organelles on two superficial epithelial cells, one of which has two H. pylori within its cytoplasm (original magnification, x10,000).

corpus. The high level of inflammation and the H. pylori-specific antibodies suggest that these two animals were infected with H. pylori as well, but that the organism had not been detected. In any event, these studies demonstrate that H. pylori infection in rhesus monkeys induces a specific immune response that can be accurately detected serologically using H. pylori antigens isolated from humans. Similar results are observed with antihuman antibody conjugates (r = 0.80; p < 0.001) although the specificity (83%) and sensitivity (69%) of the ELBA are slightly less than when using the antimonkey conjugates. 5. Natural history of infection with H. pylori. In rhesus monkeys initially found to be infected with H. pylori or G. hominis alone, values for infection, gastritis, and plasma IgG remained constant during the subsequent 7-l 5 mo tested (Table 2). Thus, as in humans, chronic infections with these organisms have relatively stable characteristics. 6. Effect of treatment on infection, inflammation, and serologic response. An initial 1O-d triple therapy of 10 rhesus monkeys using metronidazole, peptobismol, and amoxicillin failed to cure G. hominis infection and was successful in curing H. pylori from only 60% of the animals tested because, in the remaining 40%, the bacteria

Nonhuman Primate Model for H. pylori

261

Fig. 4. Transmission electron micrograph of G. hominis in rhesus monkeys (original magnification, x18,000). Note that the bacteria are not closely associated with epithelial cell surface, and that microvilli are intact.

emerged resistant to metronidazole (Fig. 7). In contrast, a 10-d quadruple therapy with omeprazole, clarithromycin, amoxicillin, and peptobismol eradicated H. pylori from each of five animals tested, including three of the animals that had not been cured by prior triple therapy. Each such cured animal has remained H. pylori-free for the 5 mo since therapy was finished (based on histology and culture at regular intervals). The gastritis scores and anti-H. pylori IgG levels decreased in each animal during this period to levels characteristic of uninfected animals. This successful eradication of H. pylori from naturally infected monkeys allows studies of reinfection of animals known to be highly susceptible to H. pylori and it should also facilitate establishment of specific-pathogen [H. pylori]-free monkey colonies.

Dubois and Berg

262 Table 1 Aligned sequence Data for Helicobacfer sp. 16s Gene (127 nt) (E. co/i 1240-l 366) H H H H H H H H H H ” ” H

py1ori 95DW pylcxi 5443' py1ori NTCT 1163W pylori NTCT 11637E acfnonvx ATCC 29263' fell8 ATCC 49179' fells DS3' fennelliee Ccw 19920~ sp CC"0 29259' m"Sfe1.w ATCC 43771' sp 91-169-al= mur~darium ATCC 49282 cinaedf CCC4 199lV

TQQQQMCAC

AAAQAcIAR(IC

?+P.*AcrucQA

AQTQQAQCCA ATCTTcAnn?.

cP.cmcTc*Q

0

c

TTCWATTQT R

c c c

H ” H H H x :: ” :: M

::

A

Vee ref. IS. bSeeref ISa ‘GenBank 3or is 0

25'

8

5:

20.

itI+

co F

1.5.

bz

lo-

d

05.

O”

0

. I~-Il_i

v

0

.

UNINFECTED (N=4)

H. pyhri + (N=13)

GHLO (N=9)

+

Fig. 5. Indivtdual values and means (+ SEM) of gastritis scores m uninfected, H pylon-infected, and G hominu-infected rhesus monkeys as determined by histology and culture. Gastritis score was sigmficantly @ < 0 05) greater in H pylon-infected rhesus monkeys compared to both uninfected and G hominzs-infected animals. 7 Experimental infection with H. pylori (25). Several tests have been carried out that establtsh that monkeys can be infected experimentally with H pylorz strains of human as well as of monkey origin. In the first set of tests, naive 2-3-yr old, colony raised monkeys that had somehow evaded natural H. pylort infection, but happened to be colonized with G homuus, were inoculated with mixtures of H. pylorz strains; bacteria recovered from colonized animals were typed by arbitrarily primed PCR (see Chapter 14), and host inflammatory and immunologic responses were followed Among five H pylon-free animals inoculated with a mixture of two human plus one monkey

Nonhuman Primate Model for H. pylori

263

. Q 9

1

0

0 5

0

LJL

0

’

UNINFECTED (N=4)

u. pyhw~ + (N =13)

A-bGHLO

+

(N=9)

Fig. 6. Individual values and means (It: SEM) of plasma IgG ratios in uninfected, H pylori-infected, and G. hominu-infected rhesus monkeys as determined by histology and culture. Plasma IgG was significantly 0, c 0.05) greater in H pylon-infected rhesus monkeys compared to both uninfected and G hominis-infected ammals strain, one monkey became persistently infected, one became only transiently infected, and the other three animals remained H pylori free. The bacteria recovered from the first two animals matched the monkey input strain m DNA fingerprint. A parallel inoculation of an animal with a natural low grade mfectron (H pylorz not detectable in most biopsy specimens) also led to the establtshment of the input monkey strain. However, new, possibly recombinant, strains (drstinguishable from input and resident strains by DNA fingerprinting) began to emerge about 4.5 mo after infection, and one of these was found to be the predominant strain when this monkey was biopsred about 14 mo later. A subsequent trial used two new human isolates and the three animals that had resisted colonization by the monkey strain, plus the one that had spontaneously cleared his infection. The trial resulted in persistent infection in two animals, and transient infection in two others. DNA fingerprmtmg indicated that just one of the two human input strains was the successful colonizer in each of these four cases. These results show that rhesus monkeys can be infected experimentally with H. pylon and that individuals differ in susceptibility to particular bacterial strains We also found that the transient and persistent infections differed in histologic and serologic responses, and that mixed infections as described here may permit the emergence of recombinants with higher fitness We propose that these findings and inferences also apply to human infection. A second set of experimental infections has been undertaken this time using four young colony-reared monkeys that were likely to be highly susceptible to H. pylon infection, by the criteria that they had become naturally infected m their colony. These monkeys were cured of then natural H. pylori (and also G homims) infections by antibiotic treatment (see Notes 5 and 6), and held for 5 mo to allow healing of H pylon-induced mucosal lesions, and dissipation of histologic and immunologic indicators of their prior infection. They were then inoculated with a mixture of seven new human strains. Culture and histologic tests showed that each

Table 2 Natural History

Time, mo

H. pylori-infected (n = 6) 0 7-15 G. homznis-infected (n = 5) 0 7-l 1

of H. py/ori or G. Hominis

Infections

H pylon or G hornmu mfection scoreb (light rmcroscopy)

in Untreated

Animals

over Timea

H. pylorl infection scorec (culture)

H pylori I@ ratio”

Gastritis scored

C

A

C

A

C

A

1.s+o5 1.2 + 0.2

2.3 + 0 4 2.7 + 0 1

0 7 + 0.2 0.8 + 0 1

1.o + 0.0 1 0 + 0.0

2.7 + 0.3 2.2 + 0 1

3.0 + 0.0 28+0.1

1 15 + 0.28 166+025

2.8 + 0.2 3.0 + 0.0

2.2 + 0.5 2 4 + 0.4

0 0

0 0

0.4 + 0.2 0

0.2 + 0.2 0

0.41 + 0.20 0 15+006

‘Values are mean + SEM for the corpus (C) and antrum (A) %oded H&E-stained slides were scored for intensity of mfectron at 1000x using a scale of O-3 (index 0, no bacteria, 1, colomes seen m 1 to 200 fields of view, 2, colomes seen m 3 to 8/10 fields; 3, colonies seen m 9 to 1000 fields) CHomogenlzed biopsies cultured m mtcroaerobtc envnonment and a score of 1 was given rf colonies growmg w&m 7 d were identified as H. pylon based on (1) pmheadstzed “water-spray” colonies posihve for urease acttvtty; (2) Gram-neganve curved or “gull-wing” rods therefrom; and (3) m at least one culture from each rhesus monkey a kmetrc assay demonstratmg high urease spectlic actrvrty, otherwise, a score of 0 was given %astrttrs rated independently from mfectton scoring on coded H&E slides using a scale of O-3 (0, intact mucosal lmmg and essentrally no mfiltratron of the lamina propna with lymphocytes and plasma cells, 1, mild increase of mononuclear mflltratron, localized m upper half of the mucosa; 2, marked mononuclear mflltratton extending from the surface to the musculans mucosae, 3, presence of polymorphonuclear leukocytes m glands, which was always associated with marked mononuclear mfiltrahon and surface erostons) ‘Plasma IgG levels were determined blindly using a modrficatton of a prevrously described ELBA using anttmonkey IgG conlugates (12)

265

Nonhuman Primate Model for t-l. pylori TRIPLE

THERAPY

-+-

UNERADICATED (N -

TIME 3-

FROM TRIPLE

-1 TIME 3-

0 FROM

-1

OF TREATMENT

THERAPY

TRIPLE

TIME

START

1 START

-+-

2

3

0 FROM

1 START

(months)

UNERADICATED (N - 2) ERADICATED

4

5

OF TREATMENT

THERAPY

-c

2

3

2)

6 (months)

UNERADICATED (N - 2) ERADICATED

4

OF TREATMENT

5

6 (months)

Fig. 7. Effect of 10 d of intragastnc administration of amoxlcillin + metromdazole + bismuth subsalicylate (7,7, and 10 mg/kg, respectively, diluted in distilled water; b id.) on mean (* SEM) H pylon infection scores, on gastritis scores, and on plasma IgG in 2 groups of H pylori-infected rhesus monkeys. There were 4 rhesus monkeys whose infections were cured at 6 mo (o), and 2 in whom infections relapsed before 6 mo (+)

of these animals became infected by 1 wk, and remained infected for the followmg 4 mo tested to date. DNA fingerprinting of strains recovered at 1 wk indicated that Just one of the seven input strains predominated in each monkey; isolates from later biopsies are currently being tested to learn how stable these populations are. Thus, these experiments illustrate that prior colonization of a monkey (or human) by H. pylori does not protect against further infection, and suggest that the selection and curing of young naturally infected monkeys provides a good means of ensuring that the animals used will be htghly susceptible to experimental mfectton.

266

Dubois and Berg

5. Perspective Our studies to date have shown that (1) domestic colony-raised rhesus monkeys are often naturally and persistently Infected by strains of H. pylori that match human pathogenic strains by usual taxonomic criteria, and (2) the risk of such infection increases with the age of the monkey. The immunologic, inflammatory, and immune consequencesof such natural infectlons, and some of the mucosal ulcerations, resemble those seen in humans Infected with H. pylon. The high risk of infection seen with these animals matches that m developing countries, where easy transmission of pathogens by fecal-oral or oral-oral routes (and possibly through contaminated water or food) puts humans at high risk of infection beginning m early infancy, and continuing through life. The rhesus monkey is also proving to be a robust and mformatlve experimental infection model. Our experiments to date have indicated differences among individual monkeys m terms of susceptibility to a given strain, a corresponding diversity among strams in terms of ability to colonize a given host or hosts, and transient as well as persistent infection. We propose that each of these findings mirrors aspectsof natural human infection by H pylon, and that the rhesus monkey model reflects the diversity among humans in traits Important for H. pylori growth, as well as mutation and selection pressures on H pylori strains during decades of carriage m a given human host (2,251. Although the rhesus monkey model described here IScostlier than other more heterologous animal models, we propose that this model ~111be invaluable for certain analyses of particular relevance to the human condition, such as: 1. Factorsthat underlie mdlvidual pathogen-host specificity; 2. Factors that determine whether infection will be persistent or transient, 3. The contributions of putative virulence determinants that are polymorphic among H pylorz isolates, such as vacuolatmg cytotoxm production or the cytotoxm as-

sociated“tag” region of the chromosometo colomzationand disease; 4. Physiologiceventsdurmgthe acutephaseof infection that have beenso difficult to studyin humans,aswell aslaterduring themuchbetterstudiedchronicinfection;and 5. The developmentof vaccinesthat alone, or possibly m combmattonwith antibiotic therapy, hold the great promise of permanent cure of infectlon, even in high risk populations in developing countries and in the underclass of the prosperous societies m North America and Western Europe.

These are among the considerations that help motivate and guide our further testing and development of rhesus monkeys as a most human-like model for H pylori infection. Acknowledgments We thank N. S. Akopyants, M. J. Blaser, E. S. Drazek, N. Fiala, R, H. Gilman, C. Hadlck, L. M. Heman-Ackah, C. K. Lee, P. T. Mehlman, E. T.

Nonhuman Primate Model for H. pylori

267

Incecik, G. I. Perez-Perez, and R. I. Walker, for stimulating discussions and/or collaboration m the experiments leading to the development of the rhesus monkey model for H. pylori infection described here. We were supported by research grants A138 166 and DK48029 from the National Institutes of Health, VM- 12 1 from the American Cancer Society, RO-83GM from the Uniformed Services University of the Health Sciences, and Gl-83HE from OraVax, Inc. The data presented in the text as well as those illustrated in the figures and tables came from US Government studies either yet unpublished or published in Gastroenterology ([1991] 100,884-891 and [1994] 106, 14051417) and in J. Clin. Microbial. [ 1994132, 1799-l 804). The experiments reported herem were conducted according to the principles set forth in the “Guide for the Care and Use of Laboratory Ammals,” Instttute of Laboratory Animal Resources, National Research Council, HHS/NIH Publ. No. 8523. The opinions and assertions contained herein are the prrvate ones of the authors and are not to be construed as official or reflecting the views of the Department

of Defense, the Uniformed

Services University

of the Health Sci-

ences, or the Defense Nuclear Agency. References 1. Krakowka, S., Morgan, D. R., Kraft, W. G., and Leunk, R D (1987) Estabhshment of gastric Campylobacterpylori infection in the neonatal gnotobiotic piglet. Infect. Immun. 55,2789-2796.

2. Akopyants, N. S., Eaton, K. A., and Berg, D. E. (1995) Adaptive mutation and cocolonization during Hehcobacterpylori infection of gnotobiotic piglets. Infect. Immun. 63, 116-121. 3. Lee A., Fox, J. G., Otto, G., and Murphy, J. (1990) A small animal model of human Helxobacterpylon active chronic gastritis. Gastroenterologv 99, 13 15-1323 4. Fox, J. G., Otto, G., Taylor, N. S., Rosenblad, W., and Murphy, J. C (1991) Helicobacter mustelae-induced gastritis and elevated gastric pH m the ferret (Mustela putorius furo). Infect Immun. 59, 1875-l 880 5. Marchetti, M., Arico, B., Burroni, D., Figura, N., Rappuoh, R., and Ghiara, P. (1995) Development of a mouse model of Helicobacterpylorz infection that mimics human disease. Science 267, 1655-1658. 6. McColm, A. A., Bagshaw, J. O’Malley C., and McLaren, A (1995) Development of a mousemodel of gastric colonisation with Helicobacterpylon. Gut 37(Suppl l), 198(abstract). 7. Handt, L. K., Fox, J. O., Dewhirst, F. E., Fraser, G. J., Paster, B. J., Yan, L. L., Rozmiarek, H., Rufo, R., and Stalis, I. H. (1994) Helzcobacter pylori isolated from the domestic cat: public health implications. Infect. Immun 62,2367-2374. 8. Newell, D. G., Hudson, M J., and Baskerville, A. (1987) Naturally occurrmg gastritis associated with Campylobacter pylon infection m the Rhesus monkey. Lancet ii, 1338.

Dubois and Berg

268

9. Curry, A., Jones, D. M., and Eldridge, J (1987) Spiral organisms in the baboon stomach Lancet ii, 634,635. 10 Bronsdon, M. A. and Schoenknecht, F D (1988) Campylobacter pylort isolated from the stomach of the monkey Macaca nemestrtna J. Clan Microblol 26, 1725-I 728. 11. Dubois, A., Tarnawskl, A., Newell, D. G., Fiala, N., Dabros, W., Stachura, J , Krivan, H., and Heman-Ackah, L M (1991) Gastric iqury and invasion of parieta1 cells by spiral bacteria in rhesus monkeys. Are gastritis and hyperchlorhydrla Infectious diseases? Gastroenterology 100, 884-891 12 Dubois, A., Fiala, N., Heman-Ackah, L. M., Drazek, E S., Tarnawskl, A , Fishbein, W. N , Perez-Perez, G I, and Blaser, M. J (1994) Natural gastric mfection with Helxobacterpylon m monkeys. A model for human infection with splral bacteria. Gastroenterology 106, 1405-1417. 13 Hazell, S. L , Elchberg, J. W., Lee, D. R., Alpert, L., Evans, D G., Evans, D. J., Jr., and Graham, D. Y. (1992) Selection of the chimpanzee over the baboon as a model for Helzcobacterpylorl infection Gastroenterology 103, 848-854. 14 Bronsdon, M. A., Goodwin, C. S., Sly, L. I., Chllvers, T , and Schoenknecht, F D. (199 1) Helzcobacter nemestrznae sp. nav., a spiral bacterium found m the stomach of a pigtailed macaque (Macaca nemestrma). Int. J. Syst. Bacterzol. 41,148-153. 15. Drazek, E. S , Dubois, A., and Holmes, R. K. (1994) Characterization and presumptive identification of Helicobacter pylorl isolates from rhesus monkeys J Clm. Microbial 32, 1799-l 804. 15a.Romaniuk, P. J., Zoltowska, B , Trust, T. J , Lane, D. J., Olsen, G J., Pace, N R., and Stahl, D. A. (1987) Campylobacter pylon, the spiral bacterium associated with human gastritis, 1snot a true Campylobacter sp J Bacterlol 169,2 137-2 14 1 16. Shuto, R., Fujloka, T., Kubota, T , and Nasu, M. (1993) Experimental gastritis induced by Helicobacterpylon in Japanese monkeys Infect. Immun. 61,933-939 17 Euler, A. R., Zurenko, G. E., Moe, J. B., Ulrich, R. G., and Yagl, Y. (1990) Evaluation of two monkey species (Macaca mulatta and Macaca fasclcularis) as possible model for human Hellcobacter pylorrdlsease J Clin Mlcroblol 28, 2285-2290.

18. Napier, J. R. and Napier, P H. (1967) A Handbook ofLwzng Przmates. Morphology, Ecology and Behavior of Non-Human Primates. Academic, New York, NY 19. Dubols, A., Nompleggl, D., and Castell, D 0. (1988) Histamine H2 receptor stlmulation increases gastric emptying m monkeys. Am J Physzol. 255, G767-G77 1, 20. Srinivasan, V. and Dubols, A (1995) Use of non-human primates for studies of gastrointestinal diseases, in Essentials of Experimental Surgery- Gastroenterology (Moody, F., ed.), Harwood Academic, Reading, UK, 1995; Chapter 27, pp. 1-19. 21. Parker, G. A., Gilmore, C. J , and Dubois, A. (1981) Spontaneous gastric ulcer in a rhesus Monkey Bram Res Bull 6,445-447. 22 Natelson, B. J., Dubois, A , and Sodetz, F. J. (1977) Effect of multiple stress procedures on monkey gastroduodenal mucosa, serum gastrin and hydrogen ion kinetics. Am J Dig DES 22, 888-897.

Nonhuman Primate Model for H. pylori

269

23. O’Gara, R. W. and Adamson, R. H. (1972) Spontaneous and induced neoplasms in nonhuman primates, in Pathology of Human Prrmates Part I (Fiennes, R. L., ed.), Basel, Karger, pp. 190-238. 24. Bruley des Varannes, S., Mizrahi, M., and Dubois, A. (1991) Relation between postprandial gastric emptying and cutaneous electrogastrogram in primates. Am J Physzol 261(Gastrointestinal

Liver Physiol. 24), 248-255

25. Dubois, A., Berg, D. E., Incecik, E T., Fiala, N., Heman-Ackah, L. M., Perez-Perez, G. I., and Blaser, M. J (1996) Transient and persistent experimental infection of non-human primates with Helzcobacter pylon. implications for human disease Zr$ Immunol. 64,2885-2891.

26. Andersen, L. P., Norgaard, A., Holck, S., Blom, J., and Elsborg, L (1996) Isolation of a “Hehcobacter heilmannii”-like organism from the human stomach. Eur J. Clin. Mlcroblol.

If: Dis 15,95,96.

27. Solnick, J. V., O’Rourke, J., Lee, A., and Tompkins, L. S. (1993) Molecular analysis of urease genes from a newly identified uncultured species of Helrcobacter J If:

DLT. 168,379-383.

28. Mendes, E. N., Querroz, D. M., Dewhirst, F. E., Paster, B J., Carvahlo, S. D , Magnano, A. G. P., Moura, S. B., Rocha, G. A , and Ohvera, A M R. (1995) Reservoir hosts for Gastrospzrillum mfection. Gut 37(Suppl l), A26 (abstract 104) 29. Dubois, A., Fiala, N., Weichbrod, R. H., Ward, G. S., Nix, M., Mehlman, P. T., Taub, D M., Perez-Perez, G I., and Blaser, M. J. (1995) Seroepizootiology ofH pylori gastric infection m socially-housed rhesus monkeys. J. Clan. Mlcroblol. 33,1492-1495.

30. Cover, T. L. and Blaser, M. J. (1994) Purification and characterization of the vacuolating toxin from Hellcobacterpylorl J. Bzol. Chem 267, 10,570-10,575. 3 1 Tummuru, M. K., Cover, T. L., and Blaser, M. J. (1993) Cloning and expression of a high-molecular-mass major antigen of Helicobacterpylori: evidence of linkage to cytotoxin production Znfect. Immun 61, 799-809. 32. Perez-Perez, G. I., Dworkin, B. M., Chodos, J., E., and Blaser, M J. (1988) Campylobacterpylori antibodies in humans Ann Int. Med 109, 11-17