Introduction to Flow Cytometry

0.25 0.5 0.75 Mole(dABrdU)

40

(0

1.0

2

o 3

20 V

N.

400

450

500 550 Wavelength, nm

600

Figure 11.28. Quenching of Hoechst 33258 fluorescence by increasing molar ratios of poly(dA.BrdU). Ethidium bromide was also included in the sample and there was a slight increase in EB fluorescence with increasing proportion of poly(dA.BrdU). The inset shows the ratio of fluorescence at 462 nm (Hoechst) to that at 5 77 nm (EB) versus poly(dA.BrdU) concentration which represents poly(dA.BrdU) content. Redrawn from Latt (1977).

CO

24 hrs of BrdU incorporation

O 3

Control

X 2

5 o

O 100

200

100

"o3

1:4 mixture of BrdU - substituted and control cells

= 1 0

O

0

200

100

200

Figure 11.29. Quenching of Hoechst 33258 fluorescence in tissue culture cells. Top left, cells exposed to BrdU for 24 hours and, top right, control cells. The bottom panel shows a 1:4 mixture of BrdU treated and untreated cells.

CELL CYCLE KINETICS

237

(0

o

Ho-33342 fluorescence

Ho-33342 fluorescence

Figure 11.30. Flash labelling of tissue culture cells with BrdU for 2 hours where panels A and B show the control and treated cells respectively. Some quenching of Hoechst 33342fluorescencein the S-phase cells can be seen in panel B. The quenching 'moves' these cells to the left and they then tend to overlie each other and some overlapping with Gl cells, which have not incorporated, occurred. fluorescence in cell kinetic studies by following time courses after labelling. However, the method has resolution problems as quenching is directly proportional to the quantity of BrdU incorporated. It can be difficult to obtain sufficient uptake of BrdU in flash labelling experiments to monitor the S-phase fraction independently of Gl and G2 + M as quenched S-phase DNA overlaps the Gl component. This is illustrated in figure 11.30 where 90° light scatter is plotted on the ordinate versus Hoechst 33342 fluorescence on the abscissa as dot-plots for control and BrdU-treated cells (5 |ig ml" 1 BrdU for 2 hours), panels A and B respectively. The fluorescence quenching of the S-phase fraction in the BrdUtreated cells is clearly apparent and many of these would overlap the Gl peak, see panel B. The separation we obtained was not regarded as being sufficient to enable us to use this method for pulse BrdU exposure experiments where the S-phase fraction was to be followed with time. However, this method has been used for continuous labelling experiments. The problems we encountered with BrdU quenching have been overcome by the second method using this analogue which is due to Gratzner et al. (1975) who raised a polyclonal antiserum to the DNA/BrdU complex in rabbits. Some of the original data obtained after BrdU incorporation and immunoperoxidase staining are reproduced in figure 11.31 which shows patchy staining in a nucleus and two chromosomes, one of which exhibits sister chromatid exchange. The first use of this technique in flow cytometry was reported by Gratzner and Leif (1981) where dual parameter data of light scatter signals versus BrdU fluorescence were able to identify the labelled cell fraction after a 60 minute exposure to 10|iM BrdU. Gratzner (1982) produced monoclonal antibodies to both BrdU and IdU as a means of detecting DNA replication and Dolbeare et al (1983) extended this

238

NUCLEIC ACID ANALYSIS

Figure 11.31. Reproductions of some of Gratzner's original immunoperoxidase data (Gratzner et al., 1975) showing incorporation of BrdU into a nucleus and two chromosomes one of which (right) exhibits sister chromatid exchange.

method to measure total cellular DNA simultaneously with BrdU incorporation. Cells were pulse labelled with the analogue, treated with a nuclear isolation buffer and fixed then subjected to an hydrolysis step to denature double-stranded DNA to single stands as the anti-BrdU antibody only recognizes BrdU in single stands. The nuclei were then incubated with the antibody which was subsequently tagged with fluorescein (green) and total DNA was stained with propidium iodide (red). The redrawn data from Dolbeare et al. (1983) are shown in figure 11.32 where the BrdU associated fluorescence (green) is plotted on the ordinate versus DNA fluorescence (red) on the abscissa as a contour map where panels A and B respectively show the control and the BrdU incorporation data after flash labelling. The 'horse-shoe' pattern in panel B is characteristic with Gl and G2 + M cells showing no BrdU associated fluorescence and the S-phase cells exhibiting differing degrees of incorporation which are related to the rates of uptake at different stages of DNA synthesis. Dean et al. (1984) have shown how the rate of DNA synthesis during S-phase can be calculated from data such as those in figure 11.32. Similar data sets are shown in figure 11.33 where an in vivo mouse tumor was analysed both before and after hyperbaric oxygen treatment (Wilson et al., 1985). In both panels there is a component with S-phase DNA content which has not been stained by the anti-DNA/BrdU antibody indicating that some cells with an Sphase DNA content are not synthesizing DNA. However, after hyperbaric oxygen this component is decreased, suggesting that some tumour cells are nutrient limited. A time course experiment using this technique in a mouse tumour is shown in figure 11.34 (Begg et al., 1985). Immediately after the BrdU pulse the data are qualitatively similar to those in figures 11.32 and 11.33. However, with

CELL CYCLE KINETICS

239

o o •o

DNA content

DNA content

Figure 11.32. Bivariate analysis of BrdU incorporation (green fluorescein fluorescence) on the ordinate versus total DNA (red propidium iodide fluorescence) on the abscissa for tissue culture cells. Controls in panel A, BrdU incorporation in panel B. Redrawn from Dolbeare et al (1983).

increasing time between BrdU administration and sampling of the population there is movement of the labelled cohort through S-phase then G2 and division to appear in the G l region of the two-dimensional map with a halving of their green fluorescence intensity. It can also be seen that cells with a Gl DNA content at the time BrdU was administered track through the S-phase domain with no green fluorescence. It is obvious from the data shown in figure 11.34 that we can now track the behaviour of the two different subsets with time, namely cells which were synthesizing DNA when BrdU was administered and those that were not. Calculation of the proportions in each cell cycle phase can now be carried out and these data can be used to estimate the cell cycle time and intermitotic phase times as described in section 11.5.3. The BrdU—antibody technique is potentially of very great importance in cell biology proliferation regulation studies as will be shown in section 12.3. However, it can be a little tricky and a considerable amount of work on the stoichiometry and sensitivity (Dolbeare el al., 1985) and the role of the denaturation step (Moran el al., 1985) has been carried out. Beisker, Dolbeare and Gray (1987) have also improved the immunocytochemical procedures to obtain increased detection sensitivity. This is not a technique which should be undertaken lightly and each cell type should be investigated in pilot studies before embarking on 'full-blow7 experiments. If you are really serious about using this method you should acquire a copy of the November 1985 issue of Cytometry (volume 6) which was entirely devoted to this technique and do some specialized reading. This issue has also been published as a book, edited by Gray and Mayall (1985).

240

NUCLEIC ACID ANALYSIS i

100

IN VIVO 80-

60

-

40

20-

1 -Li

1

1

i

20

r1—i

1

40

1

60

1

1

80

1

1 100

DNA

100-j

HYPERBARIC

80-

60•o 40

-

20-

20

40

60

80

100

DNA Figure 11.33. An in vivo tumour showing BrdU incorporation after flash labelling, top panel, and following hyperbaric oxygen treatment, bottom panel. Both data sets show some cells with S-phase DNA content which have not incorporated BrdU. However, this component is reduced in the tumour treated with hyperbaric oxygen suggesting that these S-phase cells are nutrient limited. Redrawn from Wilson et al, 1985.

CELL CYCLE KINETICS Oh

241

2h

4h

6h

•o CQ

DNA

DNA

DNA

DNA

Figure 11.34. BrdU pulse-chase experiment where the in vivo tumour was followed at 2-hourly intervals up to 6 hours. Redrawn from Begg et ah, 1985.

11.5.6 Biotinolated nucleotides The use of biotinolated nucleotides is similar in principle to the BrdU technique except that the biotinolated precursor is secondarily probed with labelled streptavidin (see section 73.3). This method has been used by Blow and Watson (1987a,b) to study in vitro DNA replication of Xenopus sperm nuclei in the

Figure 11.35. Panels A and B respectively show photomicrographs of DNA and biotin associated fluorescence. Nuclei show different degrees of biotin incorporation in panel B. There is also some variation in DNA staining intensity in panel A. The flow cytometric data in panel C show that biotin fluorescence (ordinate) is both directly and quantitatively related to total DNA (abscissa). This figure is reproduced by permission of the editor of the EMBO journal.

242

NUCLEIC ACID ANALYSIS

cell-free system described by Blow and Laskey (1986). Nascent DNA was biotin labelled by adding biotinolated 11-dUTP to the egg extract growth medium. The system was then sampled at intervals and total DNA was stained with propidium iodide and fluorescenated streptavidin was used to probe for the newly replicated DNA. An example is shown in figure 11.35 where panels A and B respectively show photomicrographs of DNA and biotin associated fluorescence from the same population as was analysed flow cytofluorimetrically in panel C. The fluorescence microscope data show qualitatively that there are different degrees of biotin incorporation per nucleus and possibly some intensity variation of DNA fluorescence. The flow system data (panel C) show that the two are both directly and quantitatively related. This technique was used to show that nuclei act as independent and integrated units of DNA replication (Blow and Watson, 1987a; Blow et al, 1988).

11.6 Tloidy' I debated for some time the wisdom of including a specific section devoted to this topic, which has probably created more problems than any other single area in flow cytometry. Obviously, I decided to 'have-a-go' and doubtless many others will in the future. The first point to make is that 'ploidy', which is derived from cytogenetics terminology, is a bad term to use in flow cytometry. Euploid means that a cell has the correct number of chromosomes. Aneuploid means that a cell has an incorrect number of chromosomes, and applies if it has too many or too few. In flow cytometry we obtain a measure of the total quantity of DNA and although the number of chromosomes is usually related directly to the DNA content of the nucleus this is not always so. Kraemer et al. (1972) have reported DNA constancy despite variability in chromosome number. It is also possible for a tumour cell, for instance, to have more chromosomes than it should have, and hence it would be classed as aneuploid, but less DNA than expected. The reverse is also possible; however, it is generally accepted that fluorescence intensity is proportional to DNA content (Kerker et al, 1982) but this may not always correlate directly with chromosome number. These considerations prompted Hidderman et al. (1984) to coin the phrase DNA index to describe the quantity of DNA assayed flow cytofluorimetrically where DNA index is the ratio of the Gl peak position of test cells to the Gl peak position of a known diploid standard. Hence if the diploid control Gl peak is recorded in channel 200 and the Gl peak of the test cells is scored in channel 300 the DNA index would be 1.5. Not infrequently one is asked by a clinician to do 'some DNA studies' or even more vaguely 'some flow work7 on this or that type of tumour or tissue. Depressingly, it is often more senior people, but I suppose this is not surprising. The first thing that has to be established is the question that is to be answered. Always say yes but with the very firm qualification that the objectives must be clearly defined. Rule 1, don't start till the questions and objectives have been defined and

TLOIDY'

243

preferably committed to paper. Some (perhaps this should read many) clinicians seem to think that if a 'fancy7 machine is available all they need to do is chop a chunk out, dump it on the bench and somehow, perhaps by magic, a result will appear which is meaningful. This type of attitude is not uncommon and stems from established behaviour patterns. Clinicians will order an investigation, be this a chest X-ray, serum electrolytes or whatever, someone else takes the picture or assays the serum and written reports, usually prepared by yet another person, then appear in the patient's notes. If the results are within normal limits the primary data, i.e. the chest X-ray, may not even be looked at by the requesting clinician. I'm referring here to routine investigations with an established place and sometimes usefulness. Tlow studies', particularly 'ploidy', are tending to be viewed in the same category but these are not yet routine investigations and the value of the majority of assays have yet to be evaluated fully in the routine clinical context. Clinicians, particularly the surgical variety, tend to be 'action-men' (that's how they are trained) and expect a result even if they don't know what to do with it when they get it. After the objectives have been established it is the duty of the scientist in charge of the flow system to alert the prospective user to the potential problems. This applies to all assays but trouble most often seems to arise with 'ploidy' and I'll consider the problem areas under headings of stoichiometry, binding site modulation, standards and logistics. Some of these have been considered elsewhere in the book but I hope it helps to have them all lumped together in one place. 11.6.1 Stoichiometry It must be quite clear from the considerations in section 7.3.7 that measuring 'ploidy' is somewhat more complex than simply chopping out a chunk of tissue and dumping it on the bench. Even if you subsequently manage to produce a perfect single nuclear suspension from the disaggregation a number of variables have to be considered, some of which can be controlled and some of which can not. One variable which is controllable is the dye concentration. If dye is in excess then from figure 7.6 we can see that all the DNA binding sites will be filled and CX will be equal to the initial number of available binding sites in the DNA. I could have done some fancy mathematics here with differential equations to prove this formally, but as it is intuitively self-evident I haven't bothered. 11.6.2 Binding site modulation The most important variable, which may not be controllable, is the initial number of free binding sites in the DNA and the 'free' has been stressed in both figure 7.6 and in section 7.3.7 as this variable can be modulated by many factors. Chromatin and DNA associated proteins may mask binding sites. This applies particularly to acridine orange (Fredericq, 1971), and the techniques using this dye have been developed to remove histones to increase the number of binding sites available (Darzynkiewicz et a\., 1975, 1977a; Traganos et al.f 1977). Chromatin

244

NUCLEIC ACID ANALYSIS

associated proteins may also reduce the accessibility of the phenanthridinium dyes (PI and EB) to DNA (Angerer and Moudrianakis, 1972; Brodie, Giron and Latt, 1975). This can occur not only due to masking but also to conformation changes and by the same token any conformation modulating agents, including DNA binding drugs, may alter dye accessibility or dye binding capacity. The state of chromatin condensation and associated DNA denaturability can make profound differences to the number of accessible binding sites and this has been exploited by Darzynkiewicz el al. {1977b,c) for many years to distinguish between G2 and mitotic cells. In tissues which have been fixed there may be permanent cross linking in DNA and its associated proteins due to the fixation process which cannot be 'unfixed7 and this too can mask potential binding sites. Prolonged formalin exposure, particularly in non-buffered saline, and mercury-containing fixatives can give rise to problems of this nature. The time interval between interruption of the blood supply to the surgical specimen and either fixation or processing can also modify the available binding sites. Cells contain endonucleases and proteolytic enzymes which digest DNA and nucleoproteins respectively. Normally, any intracellular proteolytic enzymes are contained in subcellular compartments but with cell death these enzymes are released and generally 'chewup' the cells from the inside. This inevitably degrades DNA, reduces the number of binding sites and degrades the quality of the data obtained. Hedley et al. (1984) have stressed the importance of adequate fixation of paraffin embedded material. Clearly, if fixation is not adequate there will be tissue degradation both before and during the embedding process. In the latter the temperatures reach above the DNA melting point and if tissue is inadequately fixed there will be temperature dependent DNA degradation. Feichter and Goerttler (1986) have identified inappropriate storage conditions of paraffin embedded biopsies as a cause of artefact. Prolonged (years, not specified) central heating temperatures were sufficient to degrade tissue contained in wax blocks. These authors have also identified tissue damage before fixation as a source of artefact. Biopsies which had been stored frozen at temperatures between — 40 °C and — 60 °C which were thawed, fixed and subsequently embedded gave uninterpretable DNA histograms. 11.6.3 Standards Some method of standardization and specific calibration is needed for every instrument. In more conventional assay systems, e.g. a spectrofluorimeter, it is possible to make up solutions containing known concentrations of the fluorochrome to produce a standard calibration curve. This is also normal practice for spectroscopic techniques using light absorption. The test sample is then compared with the standard calibration curve to determine the concentration and 'quantity' within it. This type of procedure is used in flow cytometry for making pH measurements (see sections 8.4.1 and 14.4) but this is simply not possible with flow cytometric DNA measurements. A 'nuclear-sized' package of DNA containing 'nuclear-quantities' of DNA can only be made by a cell. Try dissolving purified calf thymus DNA in water as I tried some time ago. You end up with a

'PLOIDY'

245

saturated colloidal 'sludge' well before you reach nuclear-like concentrations. Moreover, it is completely impossible to run this sludge through the instrument. However, some help is at hand. Biology has provided a variety of cells with different DNA content and two particularly useful examples are trout and chicken red blood cells which, unlike their mammalian counterpart, retain their nuclei. Vindelov, Christensen and Nissen (1982) and Vindelov et al. (1982) have advocated spiking the test sample with both trout and chicken red blood cells treated in exactly the same way as the test cells. Hence, the peak position of the latter can be determined relative to both internal standards and examples are shown in figure 11.36. Jacobsen (1983) used a combination of trout erythrocytes and human lymphocytes as a standardization procedure and Iversen and Laerum (1987) have used trout and salmon erythrocytes as well as human leukocytes as internal standards for ploidy control. However, it should be noted that DNA fluorescence from the compacted chromatin from lymphocytes can be less than that from cells with less compacted chromatin, which is a dye accessibility problem. These systems are fine for fresh tissue where it can be guaranteed that the test sample and the internal standards have been treated identically. But, it is not always satisfactory for archival material where the fixation and embedding history may not be known, which makes it impossible to treat the standardizing cells identically to the test cells. This problem is only partly resolved with physical standards in the form of microbeads which are now available with different fluorochromes in a range of concentrations which can be used to set up the instrument identically from day to day. 3000

2000 o c


O* 0)

1000

50

100

150

200

Channel No. (DNA)

Figure 11.36. A sample of peripheral blood monocytic cells (MO) 'spiked' with both chicken and trout red blood cells (CRBC and TRBC respectively) which act as internal standards. Redrawn from Vindelov et a\., 1982.

246

NUCLEIC ACID ANALYSIS

Our efforts in this area have been to use a known diploid standard of normal colonic mucosa from a single specimen as well as microbeads to set up the instrument. We originally obtained 30 cm of a single colonic specimen and embedded this in 40 different paraffin blocks and we've used three blocks in four years, so it will keep us going for some time to come. We then rely upon finding normal cells in any aneuploid specimen to calculate the DNA index if this is what is wanted. In my experience it is always possible to find cells with diploid DNA content in an aneuploid tumour even if it means cutting sections from a second or third block. If there is no aneuploid spike in a tumour specimen it is easy to check that it does in fact contain tumour by looking at a 4 |im histological specimen cut adjacent to the thick section used for flow cytometry. Furthermore, it is often possible to distinguish between multiple subsets, including between normal and malignant, on forward and 90° light scatter characteristics (Watson etal., 1987a) even if the DNA content is identical. It has been our practice to score aneuploidy if, and only if, we can see two distinct spikes corresponding to normal diploid and aneuploid components respectively. Using these criteria may mean that some of our studies have tended to underestimate the proportion of aneuploidy in tumour groups. A minority of tumours with a 'shoulder' on the right of the diploid spike, which may have been due to aneuploid components, were scored as diploid but at least we can be sure that those scored as aneuploid were in fact aneuploid. In short, we use presumed normal diploid elements within the tumour as the internal control.

11.6A Logistics When your prospective new user has assimilated this little lot and come back for more you know that he/she is serious. It is now time to establish who is going to prepare and then extract the information from the raw data after the samples have been run through the instrument. Remember that the time taken to 'rail-road' the samples through the instrument is totally insignificant compared with the time for preparation, staining, data extraction and interpretation. In a multi-user system the flow cytometry staff, which seldom number more than two, simply do not have the time to prepare everyone's cells as well as running the samples and keeping the instrument in good running order. If prospective users either cannot, or occasionally will not, provide the necessary support to prepare the samples and extract the information from the raw data you have no option but to refuse to collaborate. People who are serious will understand and find a way of providing the necessary support. Those who do not don't matter, so nothing is lost by refusing.

11.7

RNAandDNA

We have already seen examples of the potential power of multiparameter cell cycle analysis in section 11.5, namely the capacity to distinguish between newly replicated and total DNA using BrdU and biotinolated nucleo-

RNA AND DNA

247

tides. In this section we consider the simultaneous measurement of RNA, DNA and size using acridine orange and Hoechst/pyronin-Y. 11.7.1 Acridine orange A number of staining techniques have been developed to quantitate RNA and DNA simultaneously using the metachromatic property of acridine orange but the pioneering work in this area of flow cytometry has been carried out by Darzynkiewicz and colleagues at the Sloan Kettering in New York. Our group has used the two-step technique of Traganos et al. (1977) for a number of years with only minor modifications and it is worth some space to describe the technique in detail. Unfixed cells suspended in whole medium are initially treated with a solution containing the non-ionic detergent triton X-100 and 0.1 N HC1 in isotonic saline. The triton permeabilizes the external cell membrane which remains stable at the low pH and allows the highly polar dye to enter the cell. The low pH also removes basic histones from chromatin thus making available some dye binding sites which would otherwise be inaccessible. After 60 to 90 seconds the second solution is added. This is a phosphate/citrate buffer containing acridine orange and EDTA. The latter chelates divalent cations which destabilizes double-stranded nucleic acids with RNA being more susceptible than DNA. The final pH is in the range from 4.5 to 4.7 which gives the optimum red/green discrimination. Dissociation of double-stranded RNA to single-stranded species which fluoresce red takes about 5 minutes after which the green and red fluorescence signals remain stable for 20 to 30 minutes depending on cell type. This method depends critically on the selective denaturation of double-stranded RNA as any DNA denaturation will be scored as red fluorescence with concomitant loss of green, and a number of warnings have to be sounded about using this technique. Firstly, some DNA denaturation will occur if cells are over exposed to the first solution containing triton X-100 at pH 1. Secondly, DNA is inherently less stable after histone extraction and denaturation will occur with prolonged exposure to EDTA (divalent cations have been chelated) plus AO in the second step. Every cell type should, therefore, be tested for optimum exposure times in both steps. Thirdly, the DNA phosphate-to-dye molar ratio is fairly critical (Nicolini et al, 1979). For mammalian cells the final cell concentration should be about 10 5 ml" * with the AO concentration between 4 and 5 jigml" 1 . In practice 0.4 ml of solution 1 is added to 0.2 ml of medium containing cells which should be at a concentration of about 106 ml ~ \ Then 1.2 ml of solution 2 are added containing AO at a concentration of 6 jag ml" \ giving final concentrations of about 105 cells and 4 |ig ml~ *. Fourthly, the fluorescence emission spectrum is proton concentration dependent and the final pH should be in the range 4.5-4.7 for the best red/green discrimination. Fifth, the analysing filter combination in the cytometer is critical as the 'green' and 'red' fluorescence emissions do not occupy discrete bands in the spectrum. There is a green DNA 'tail' into the red region and a red RNA 'tail' into the green region. The latter is less critical than the former as DNA is present in much larger quantities than RNA and the contribution of the red RNA tail into the green region is

100

100

A

•

t

<50

< 50

z o

Q

100

B

c

t

" ' ;. ft •

< 50

v

tv .

z o

•

100

100

Figure 11.37. Acridine orange staining of stimulated T-cells for DNA (ordinate) and RNA (abscissa) simultaneously. Panel A shows the control non-stimulated cells. Stimulated cultures are shown in panel B where cells increase their RNA content between Go and Gl. The latter phase is subdivided into two namely G l A and GlB on an arbitrary RNA level above which cells are seen to enter S-phase. Panel C shows data after treating stimulated cells with vincristine.

100

RNA AND DNA

249

insignificant compared with the magnitude of the green intensity from DNA. The long-wavelength red tail from the DNA emission spectrum could, however, constitute a problem particularly for cells with low RNA levels where the DNA content is many orders of magnitude greater than the RNA content. In this case a significant quantity of the signal in the red photomultiplier could be due to DNA fluorescence. These problems can be largely overcome by the correct optical filtration with the use of a 580 nm dichroic mirror, a band-pass filter from 515 to 560 nm for the green photomultiplier (PMT) and a 630 nm long-pass filter for the red PMT. The latter minimizes the green DNA tail contribution to the red RNA signal, and the magnitude of this effect should be checked by treating a test sample with ribonuclease. In our instrument we have found that only between 10 and 15% of the red signal is due to DNA fluorescence in lymphocytes and EMT6 mouse tumour cells. However, this may be different in different cell types depending on the RNA to DNA ratio and the stability of the DNA. Finally, it seems to be more difficult to perfect this technique with instruments which interrogate cells in the jet as opposed to those that interrogate in the flow chamber and a number of explanations have been proposed. These include stripping of cytoplasm by the shearing forces in the nozzle and disturbance of the electrostatic stacking of dye molecules along single-stranded molecules either by mechanical forces or by induced electrical currents within the nozzle. Darzynkiewicz et al. (1976) have used acridine orange to show that G o thymocytes can be distinguished from Gl after PHA stimulation and these data are reproduced in figure 11.37. Panels A and B show the control unstimulated population and cells stimulated with phytohaemaglutanin (PHA) respectively. The Go 'resting' cells (panel A) increased their RNA content as they progressed into Gl before they entered DNA synthesis. Panel C shows stimulated cells treated with vincristine to block the population in G2 + M and it is apparent that there is emptying of the G l component with the higher RNA levels. It is also clear from these data that a 'critical' RNA level had to be reached before Gl cells could enter S-phase. Data such as these have enabled Darzynkiewicz et al. (1976, 1977a) to discriminate between early and late Gl, termed GlA and GlB respectively, where GlB has been defined as the RNA content of Gl cells above which cells can be observed to enter S-phase. Subsequently Darzynkiewicz et al. (1977b) demonstrated that interphase and metaphase chromatin exhibited different sensitivities to thermal denaturation and on this basis they were able to discriminate between G2 and mitotic cells (Darzynkiewicz et al, 1977c). Fixed cells were treated with ribonuclease to remove RNA then heated at various temperatures to partially denature DNA. Thus, on subsequent staining with acridine orange any single-stranded DNA would fluoresce red. Some of these data are shown in figure 11.38 where green fluorescence (double-stranded DNA) is plotted on the ordinate versus red fluorescence (single-stranded DNA) on the abscissa after heating to 50,85 and 95 °C, panels A, B and C respectively. The discrete cluster which became increasingly separated from the main body of the data with increasing temperature was

NUCLEIC ACID ANALYSIS

250

5 0 ' 'c

A O

O

c
o

w

85°C

B

95'

• HI •

I I I

InB• ••m •

M

s

,v

1[

FTT-

Gi

1 1

50

1 1 1

50

iii ei i i •

50

Red fluorescence Figure 11.38. Temperature dependent DNA denaturation in ribonuclease treated fixed cells assayed with acridine orange. With increasing temperature there is increasing DNA denaturation manifest by a decrease in green fluorescence and a concomitant increase in red fluorescence. Mitotic cells exhibited greater thermal denaturability as shown by the cluster which becomes increasingly separated from G l and S-phase cells.

composed of mitotic cells which was confirmed with colcemid treatment. This result was a little surprising as it showed that the condensed chromatin in mitotic cells was more susceptible to thermal denaturation than chromatin in interphase. It was also found that BrdU incorporation into DNA suppressed acridine orange DNA staining (Darzynkiewicz et al, 1978). As this analogue is not incorporated into RNA it was possible to identify cells 'in transit7 from Go to GlA (GlT) on their RNA content where their DNA fluorescence was not suppressed. These data are shown in figure 11.39 with DNA plotted on the ordinate versus RNA on the abscissa where panel A shows G l Q (Go) non-stimulated lymphocytes. In panel B the PHA stimulated cells had been exposed to BrdU for approximately one cell cycle, thus all cells within the division cycle had incorporated the analogue. These cells are included in the dashed polygon and the GlA and GlB compartments clearly have reduced green (DNA) fluorescence compared with non-stimulated cells labelled G l Q . The GlT cells, in transit from G o to GlA, are those with nonsuppressed green fluorescence but with RNA levels comparable with those in GlA. These cells are stimulated as their RNA levels have increased above the G o cells but they had not passed through DNA synthesis as their green fluorescence was not reduced. Studies by Traganos et al (1979) with DMSO induced senescence in Friend leukaemia cells revealed non-cycling cells in S and G2 as well as in Gl and these various studies have enabled Darzynkiewicz et al. (1980) to identify 12 different 'compartments' in the cell cycle. Similar studied have been carried out with EMT6 cells synchronized by mitotic selection (Watson and Chambers, 1978a). The RNA levels increased in Gl before

RNA AND DNA

251

80

80

B *

G1Q

GIQ

GIT

• V#

\ o CO

f.

m F

0* •

50 F

rt

•• *#

•

#

: >

GIB 100

>600

(RNA)

Figure 11.39. BrdU quenching of acridine orange green fluorescence (DNA) in stimulated lymphocytes where green (DNA) fluorescence is scored on the ordinate with red (RNA) fluorescence on the abscissa. Panel A shows data from the control non stimulated cells (GlQ). Stimulated cells which have encorporated BrdU are enclosed in the dashed polygon in panel B. Cells in transit to G l (GlT) have Gl RNA levels but have not entered S-phase and have not incorporated BrdU hence, their green fluorescence is not quenched.

S-phase commenced. This is a similar finding to that of Darzynkiewicz et al. (1976) in stimulated T-cells. However, during early and mid S-phase the levels remained virtually constant but in late S and G2 there was a further rapid increase in RNA then a decrease in mitosis (Watson and Chambers, 1978a). The latter was confirmed by parallel studies with tritiated uridine uptake (Watson and Chambers, 1978b). Actinomycin-D (ACT-D) is a cytotoxic agent which blocks RNA transcription and the pharmacological dose range in man, assuming an even whole body distribution, is within the range 1 —10 ng ml~ 1 when translated to tissue culture terms. When synchronized EMT cells were treated with ACT-D at 6 ng ml" 1 the RNA level decreased exponentially with a half-time of about 32 hours which is consistent with ribosomal-S RNA degradation. At a lower dose of 1 ng ml" 1 the RNA level remained constant following exposure indicating that the production rate was equal to the rate of degradation. We were very surprised to find that subsequent mitosis was only delayed and not inhibited if ACT-D was added to synchronized cells after mid Gl. Mitosis took place in a proportion of cells even though the RNA level was decreasing at the higher dosage. However, mitosis was inhibited if ACT-D (at either dose) was added just before mitotic selection. This observation suggested that these cells were programmed for subsequent mitosis soon after the last mitosis, and it is possible that this occurs in GlA. Multi-parameter analysis of lipopolysaccharide-stimulated B-lymphocytes using acridine orange to give measurements of RNA and DNA, with both 90° and

252

NUCLEIC ACID ANALYSIS DAY T' COL -VF. CONTROL 11-MAY-£

•E*'-.

? DAY LP3 22H COL 4-. 3 iASH 1 1 -MAY -84-

F/^wr^ 11.40. Acridine orange staining for RNA (630RF AREA) versus DNA (520RF AREA) in 5 day stimulated spleen B-cells. The top panel shows the control data. The middle and bottom panels respectively show the results of colcemid exposure and its withdrawal.

RNA AND DNA

253

forward light scatter has shown that the post-mitotic, pre-DNA synthesis phase (Gl) can now be subdivided into four distinct subphases namely, Go, Gl', GlA and GIB (Kenter et al, 1986). Figure 11.40 shows three data sets where DNA (520RF AREA) is plotted against RNA (630RF AREA) for day 5 LPS-stimulated B-cells. The data in the top, middle and bottom panels respectively were from control and colcemid-arrested cells and from cells 4.5 hours after colcemid release and each panel shows the characteristic GlA, GlB, S-phase and G2 + M pattern. Our initial interpretation of the data in figure 11.40 was that with colcemid (middle panel) GlA emptied into GlB, the latter emptied into S-phase which subsequently emptied into G2 + M where the population was blocked and that there was a residual non-cycling 'Go' population represented by the cells with lowest RNA content. This interpretation seemed to be substantiated by the data in the bottom panel of figure 11.40 obtained 4.5 hours after colcemid was washed off. There was a relative increase in the proportion of GlA and GlB cells which were presumed to have arisen from a resumption of progression through the cell cycle with division at mitosis refeeding the GlA compartment. This interpretation, however, was not entirely satisfactory as there did not seem to be progression through S and G2 + M. The data were then re-examined as the DNA (520RF AREA) versus 90° light scatter (4'80RS WDTH) data space which is shown in figure 11.41. The left and right columns give perspective 3-D views and conventional contour plots respectively where the top, middle and bottom panels correspond to those in figure 11.40. The control data (top panels) show two peaks in Gl distinguished on their high and low 90° light scatter with small proportions in S and G2 + M. The same overall pattern was seen after colcemid (middle panels) with two exceptions. First, there was the expected accumulation of cells with G2 + M DNA content. Secondly, and unexpectedly, there was an increase in the proportion of cells with G l DNA content which exhibit low 90° scatter signals. After release from colcemid (bottom panels) the proportion of cells in S and G2 + M remained about the same but there was a decrease in the Gl DNA content population with low 90° scatter. The three data sets were gated as shown in the right column of figure 11.41 to give the proportions in G l ' (region 1, defined as Gl DNA content with low 90° scatter) GlA plus GlB (region 2), G2 + M (region 3) and S (region 4). The region 2 data were then analysed in the RNA/DNA data space to obtain the proportions in GlA and GlB and a summary of these data are shown in figure WAI. With colcemid there was a relative increase of cells in G l ' due to emptying of GlA and GlB with progression into S-phase with G2 + M accumulation. After colcemid release the proportions in S and G2-f M remained constant but there was a redistribution within Gl from G l ' to GlA and GlB. Thus, the large 'spikes' in the three panels of figure 11.40 were not Go cells, but G l ' cells plus some in GlA. Furthermore, the three ill-defined but distinct RNA distributions in the top panel of figure 11.40 represent cells in Gl', GlA and GlB respectively. This was confirmed by two further pieces of evidence. Firstly, unstimulated cells die and lyse with a half-time of about 18 hours, and after 5 days in culture there were no detectable unstimulated G o cells. Secondly, if the G l ' cells had been Go cells then the

254

NUCLEIC ACID ANALYSIS FILE PA

Tieee

n

A

F I L E Tteee5-*ae! D ^ T E I I —M«* — * S S * i QAV S U^3 . tS« COL

Figure UAL DNA (52ORF AREA) versus 90°light scatter (488RS WDTH) for figure 11.40 data. The right column shows conventional contour plots with four gated regions used for data analysis.

stimulated cells within the cell cycle would have had to double in 4.9 hours after release from colcemid in order to attain the proportions seen in the various phases 4.5 hours after colcemid was washed off. The doubling time of the population, from cell counts, was about 36 hours. Thus, this new G l ' compartment, which may be similar or identical with the GlT compartment of Darzynkiewicy et al. (1980), is within the cell cycle and is blocked by colcemid. Kenter and Watson (1987) also used the acridine orange technique in conjunction with Southern blotting (Southern, 1975) to study the S(i to C|i immunoglobulin heavy-chain switching which occurs when stimulated spleen Bcells convert from being IgM expressors to IgG secretors. The proportions of the

RNA AND DNA

255

Cell cycle redistribution of B-cells with COLCEMID

+ 300-

G2 + M

50 -|

40-

o G1A

+ 100
o.

4) (0 (0 JC

a o "o o % O

c a> o a> a a>

30

0) (0 JC

• G1 f

• G2+M

20

a o

"o >» u =

0 -

(0

a>

O)

c

(0

G1B

10

G1B

control colcemid wash

-100 control colcemid wash

Figure 11.42. Summary of colcemid induced changes in stimulated spleen B-cells showing the unexpected transit block between Gl' and GlA and its rapid reversal following colcemid removal. The expected G2 + M block is also apparent but this was not reversed within the 4.5 hours after colcemid removal. population in Go, Gl', GlA, GlB and S + G2 + M were determined at 24 hour intervals for 4 days after stimulation with LPS and dextran sulphate. It was then possible to estimate the various phase times with their standard deviations from a computer cell cycle model (Watson and Taylor, 1977). The experimental data represented by the points, and computer predicted results represented by the curves, are shown in figure 11 A3. The conversion from IgM expression to IgG secretion is accompanied by rearrangements within the immunoglobulin heavychain locus which are manifest by a loss of the S|i DNA signal from restriction enzyme digest patterns, and the loss of this signal was analysed at intervals after stimulation by Southern blotting. Computer simulations were then carried out

256

NUCLEIC ACID ANALYSIS

1001

4>

o 0) CO *-• 0)

G n +G1P

o

5 CL

DAYS 11.43. Computer model predicted data, curves, versus experimentally determined proportions of B-cells in G0 + Gl' (OX GlA (#), GlB (D) and S + G2 + M ( • ) at the times in days shown on the abscissa after LPS and dextran sulphate stimulation. with the results from the parallel cell cycle kinetic analysis to predict the proportion of S|i DNA signal remaining after stimulation. These results predicted that the SjJ, to Cjl heavy-chain switch would occur at approximately 85% into the cell cycle which placed it in mid S-phase on the first pass through the cell cycle after stimulation. This is an interesting result as it corresponds to the time in S-phase at which the immunoglobulin locus is replicated and the results are shown in figure 11.44. The predictions were confirmed experimentally by treating stimulated cells with aphidocholine which blocks cells in early S-phase. This also blocked the loss of the S(i signal but the switch was manifest by loss of the signal some 6 hours after release from aphidocholine.

11.7.2 Hoechst/pyronin-Y The staining procedures developed for RNA and DNA using acridine orange are somewhat 'harsh7 and the permeabilization step using acid—triton effectively excludes the addition for monoclonal antibody probing simultaneously with RNA and DNA. Furthermore, acridine orange staining is an equilibrium

RNA AND DNA

257

100 0.5tc

80-

3

4

5

6

7

Time, days

Figure 11.44. Computer predicted DNA rearrangement of Sji signal, curves, versus experimental data, points, obtained from Southern analysis of S|I signal loss. The curves were simulated with switch probabilities of 0.25, 0.5, 0.75 and 1.0 at times corresponding to early Gl(0.25tc), mid G l (0.5tc), Gl/S interface (0.75tc) and mid S-phase (0.85tc). The best correspondence between experimental and simulated data occurred at 0.85 tc with unit switch probability.

technique where the dye equilibrates with DNA, RNA, core and sheath and some groups have found it difficult to obtain consistent results. These considerations prompted Shapiro (1981) to investigate the possibility of using the combination of Hoechst 33342 plus pyronin-Y for RNA and DNA using dual wavelength excitation (UV and blue). CCRF-CEM cells and lymphocytes were used, adjusted to a concentration of 106 ml ~ l in medium and initially stained with Hoechst 33342

25S

NUCLEIC ACID ANALYSIS

Ho:DNA-

AO:DNA-

Figure 11.45. Hoechst 33342/pyronin-Y estimation of RNA and DNA simultaneously, from Shapiro (1981).

at concentration of either 5 |im or 10|iM. After 45 minutes incubation at 37°C pyronin-Y was added to give a final staining concentration of 5 |iM and the cells were analysed after a further 45 minutes incubation. Results from control and concanavalin-A stimulated lymphocytes were comparable with those from acridine orange as were those from CCRF-CEM cells. Results from the latter are given in figure 11.45 where the left panel shows pyronin-Y RNA fluorescence on the ordinate versus Hoechst 33342 DNA fluorescence on the abscissa. Comparable acridine orange staining with cells from the same culture are shown in the right panel. The DNA histograms are very similar but, the RNA resolution in late S-phase and G2 + M is not as good with pyronin-Y as with acridine orange. Shapiro (1981) also demonstrated that only about 60% of the 'RNA' signal from pyronin-Y could be abolished with ribonuclease in ethanol fixed CCRF-CEM cells. Therefore, there was a considerable background, due to non-RNA poly-anion binding, above which the specific RNA staining had to be scored. This non-specific background is likely to vary considerably between different cell types and is possibly one of the reasons for the inferior RNA resolution in late S-phase and G 2 + M compared with acridine orange. Nevertheless, Shapiro's results demonstrated that RNA and DNA could be measured simultaneously with this technique in 'intact' cells. The initial results from Shapiro (1981) prompted a comprehensive physicochemical study of pyronin-Y binding with synthetic polynucleotides by Kapuscinsky and Darzynkiewicz (1987) and a summary of their conclusions is as follows. Pyronin-Y is not RNA specific and only double-stranded nucleic acids form fluorescence complexes with the dye. Fluorescence is quenched by guanine

EMISSION SPECTRUM ANALYSIS

259

residues in both RNA and DNA. At high dye: phosphate ratios, when binding sites are saturated, about 80% of the fluorescence emission is from sites which do not contain guanine. Condensed complexes of pyronin-Y with nucleic acids have considerably reduced fluorescence compared with free dye and at concentrations above 5 x 10 ~5 M the dye preferentially condenses RNA (Darzynkiewicz et al, 1986). Thus, at high dye concentrations with mixtures of nucleic acids the fluorescence will tend to be from DNA which is more resistant to condensation than RNA (Portela and Stockert, 1979). At first sight these conclusions would seem to exclude the use of pyronin-Y as a stain for RNA. However, Darzynkiewicz et al. (1987) have shown that the presence of 0.8 jiM Hoechst 33342 in a 6.6 [iM solution of pyronin-Y increases the specificity of pyronin-Y for RNA and under strictly controlled conditions it can be used very effectively as a probe for RNA in fixed cells. This dye also enters viable cells, but it is taken up into mitochondria (Cowden and Curtis, 1983; Darzynkiewicz et al, 1986) and, therefore, cannot be used as a vital RNA stain (Darzynkiewicz et al, 1987).

11.8

Emission spectrum analysis

It was shown in figure 8.9 that multiple subsets are apparent in mouse bone marrow when Hoechst 33342 stained DNA fluorescence is analysed on two wavelengths simultaneously. Figure 11.46 shows time course results obtained with chicken thymocytes when stained at concentration of 5 |iM. The various panels were recorded at 15, 30, 45 and 125 minutes after the stain was added, top left, top right, bottom left and bottom right respectively (Watson et al., 1985a). Each panel shows the dot-plot of violet (440RF AREA) versus green (560RF AREA) fluorescence in the lower right-hand corner with the associated histograms, and each division on both the X- and Y-axes respresents 100 digitization channels with a maximum of 1024. In each panel there is a single peak on the violet analysis channel (Y-axis) but there are two peaks on the green channel (X-axis) at 15,45 and 125 minutes after staining. At 30 minutes, top right, there is a single peak on each axis. The dot-plots at 15 and 125 minutes show two major subsets very clearly. Figure 11.47 displays data comparable to that in figure 11.46 but with the dye concentration reduced to 1 (iM and only the 15 and 30 minute data sets are shown. Again, two major subsets are apparent. It appeared that the subset labelled 1 at 15, 45 and 125 minutes in figure 11.46 exhibited a 'tighter' distribution compared with that labelled 2. Furthermore, it also appeared that the former remained stationary with time but that the latter showed increasing green fluorescence. This observation is substantiated by the data shown in figure 11.47 (1 (iM Hoechst 33342), where the subset labelled 1 maintains its position at 15 and 30 minutes, but that labelled 2 shows an increase in both violet (Y-axis) and green (X-axis) with time. Moreover, subset 2 shows a greater initial increase in violet fluorescence compared with green, its position tends to be towards the upper left domain of the dot-plot. The proportions of cells in these two major subsets were counted by applying

260

NUCLEIC ACID ANALYSIS 15 min

30 min

560RF AREA

560RF AREA

1+2

45 min

125 min

560RF AREA

560RF AREA

Figure 11.46. Dot plots of violet fluorescence on the ordinate (440RF AREA, where 440, RF and AREA respectively denote analysis wavelength, right angle fluorescence and area under each pulse) versus green on the abscissa (with 560 denoting analysis wavelength) for chicken thymocytes stained with 5 uM Hoechst 33342 at 15, 30, 45 and 125 minutes after staining (respectively top left, top right, lower left and lower right). The associated histograms are shown against the relevant axes.

the elliptically gated regions as shown on the dot-plots. The mean percentages of cells within the three gated regions of figures 11.46 and 11.47 were scored as follows. Subsets 1 and 2 respectively comprised 33.2%+ 0.99% and 41.7%+1.53% of the population, where the limits represent 2 standard errors. Subset 3 contained 12.0%+ 0.55% and subsets 1 + 2 combined constituted 75.0%+ 1.09% of the whole population. About 10% of each data set was not included within any of the regions. It is clear from these data that the subsets designated 1 and 2 are discrete entities. It is also clear that the behaviour of these subsets with Hoechst 33342 staining is different and figure 11.48 shows the green and violet coordinates of the

EMISSION SPECTRUM ANALYSIS

261

15 min

560RF AREA

—-1

< <

< ^

•

U-. CC

o

j ^•' ^>

/ / Mr

<

30 min

560RF AREA

Figure 11.47. Similar display to that in figure 11.46 but with 1 uM Hoechst 33342 at 15 and 30 minutes after staining, top and bottom respectively. centers of the gated regions for subsets 1 and 2 of figure 11.46 plotted against time. Neither the green nor violet coordinates change significantly with time in subset 1, although there seems to be a tendency for both to fall slightly, and the violet coordinate does not change in subset 2. The significant feature, however, is the rise and then the plateauing of the green coordinate of subset 2. These data demonstrated that two different chicken thymocyte subsets could be distinguished by analysing the emission spectrum time course of the Hoechst 33342-DNA complex at two different wavelengths simultaneously. In one subset {33% of the

262

NUCLEIC ACID ANALYSIS 700

Region 1 600

violet > 500

green

c

I

o

400

0

700 T

30

60

90

Region 2 . o green


2 600 /

500

400

120

!_ 9

30

;

•

• violet

60

90 Minutes

120

Figure 11.48. Green and violet fluorescence intensities for subsets 1 and 2, regions 1 and 2 respectively, plotted against time after staining with 5 uM Hoechst 33342 (data from figure 11.46).

total, designated 1) the development of both violet and green fluorescence was very rapid at a stain concentration of 5 |iM with maximal expression occurring in 15 minutes and no change thereafter. In the second subset (42% of the total, designated 2) the violet fluorescence was expressed equally rapidly, and to the same intensity (channel 500), but the green fluorescence increased throughout the 2 hour time course from an initial value of 410 at 15 minutes towards a plateau level of about 625 at 125 minutes. It was also interesting to note that decreasing the dye concentration to 1 |iM did not decrease the violet emission intensity of subset 1 (about channel 500, see figures 11.46 and 11.47), but it did decrease the green emission intensity by about 25%, from channel 500 to approximately 400. Again, however, neither the green nor the violet emission intensity changed in subset 1 at the lower dye concentration during the observation interval. In contrast, subset 2 at the lower dye concentration showed increasing fluorescence on both the green and the violet analysis bands with the latter increasing more rapidly than the former. The quantity of light emitted from a DNA—fluorochrome complex at a given time after staining with a Vital' dye will depend on a number of factors. These may be divided into three basic categories which will be considered under the headings transport, binding and microenvironment. The transport factors include the concentration of the dye, the rate at which it can traverse not only the external cell

EMISSION SPECTRUM ANALYSIS

263

membrane, but also the cytoplasm and possibly nuclear membrane, the rate at which the free fluorochrome is metabolized (if at all) and the rate at which the fluorochrome is 'excreted' by the cell. The binding factors include the ratio of the number of accessible binding sites for the dye to the intra-cellular dye concentration in the vicinity of the DNA and the rate constants for association and dissociation of the fluorochrome with DNA. Once the fluorochrome is bound to DNA the energy of the fluorescence emission, as opposed to the quantity, will depend on the type of binding of fluorochrome to DNA and the microenvironment (see section 3.8.4). Considering these various factors we could make some generalizations about the behaviour of the two major subsets observed in these experiments. It was apparent that the violet fluorescence from subset 1 was not dependent on the dye concentration within the range used (albeit relatively limited), and it was also temporally independent within the observed interval. Therefore, the factors included under both the 'transportation' and 'binding' headings had reached a saturated steady state for violet fluorescence which was neither concentration nor rate limited. This, however, was not true for the green emission from this same subset with the peak being at about channel 400 at 1 pM Hoechst 33342, and at about channel 500 at 5jiM Hoechst 33342 with no change with time at either concentration. Thus, the expression of green fluorescence in subset 1 was dye concentration dependent but time independent. The time independence shows that the expression of green fluorescence from subset 1 had reached a steady state by 15 minutes and the concentration dependence shows that saturation had not been attained at the lower Hoechst 33342 concentration and it may also not have been attained at the higher concentration, but at both concentrations a steady state had been reached. Turning now to population 2 we saw a very different pattern. The violet fluorescence remained constant from 15 to 125 minutes at the higher dye concentration, but there was a considerable increase between 15 and 30 minutes at the lower concentration (see the Y-axis histograms in figure 11.47). The green fluorescence intensity of this subset also showed increasing fluorescence with time not only at the lower (see X-axis histograms figure 11.47), but also at the higher dye concentration (see figure 11.48). Thus, the expression of both the violet and green fluorescence were dye concentration and time dependent in subset 2 over the concentration and time ranges used in these experiments. Moreover, it is clear from figure 11.47 that violet fluorescence in subset 2 was expressed more rapidly than green during the first 30 minutes. This was also apparent at the higher dye concentration (see figure 11.48) where at 15 minutes the green fluorescence intensity was relatively less than that of the violet, but beyond 30 minutes the green exceeds the violet. This is illustrated in figure 11.49 which shows the green: violet ratios for both subsets at the higher dye concentration plotted against time. Subset 2 exhibited an increasing ratio throughout the time course with a tendency towards a plateau level and subset 1 showed a near constant level, or possibly a small decrease towards a plateau value. Thus, the 'shape' of the Hoechst 33342 emission spectrum of cells in subset 2 changes with time with an

264

NUCLEIC ACID ANALYSIS 1.25

i Region 2

. ^o

o

OC +•« 4) O

> c

1.00

d) 0)

Region 1

k.

O

0.75 30

60

90

120

Minutes

Figure 11.49. Ratio of green to violet fluorescence for subsets 1 and 2, regions 1 and 2 respectively, plotted against time after staining with 5 uM Hoechst 33342.

increase in green emission, but that of subset 1 did not. It was this difference which made it possible to distinguish the subsets. These various observations strongly suggested that two different binding sites, or two types of binding at the same site might exist; one emitting violet fluorescence preferentially, the other emitting green preferentially indicating different binding energy states. It is also pertinent to observe that the total fluorescence emission from subset 2 was greater than that from subset 1 at the higher dye concentration at 125 minutes. This can be seen from the data presented in figure 11.48 where the sum of the green and violet intensities is 1035 for subset 1 and 1150 for subset 2, an increase of 10%. It is highly unlikely (though not impossible) that subset 2 contains 10% more DNA than subset 1. It is much more likely that the differences observed are due to different dye binding properties in the two subsets with both containing approximately the same number of available sites which give rise to violet fluorescence preferentially and with subset 2 containing more binding sites which give rise to green fluorescence preferentially. From the evidence available in these experiments we proposed that two DNA binding sites, or two binding states, exist for Hoechst 33342. As the violet fluorescence was expressed more rapidly than the green we considered this to be the primary, 'V-site', and the green, 'G-site', to be secondary. In further more extensive studies (Smith et a\., 1985) it was shown that the Hoechst 33342 emission from DNA is highly pH dependent, which can have profound consequences for the green: violet ratio. In summary it is possible to make the following statements.

EMISSION SPECTRUM ANALYSIS

265

(1) The bisbenzimidazoles bind preferentially to repetitive A —T sequences (see review by Latt, 1979) with dye molecules occupying the minor groove of the helix at low dye: DNA-phosphate ratios. (2) The fluorescence emission primary mode binding tends to be violet biassed. (3) Increasing the dye: DNA-phosphate ratio leads to further binding where the dye molecules align more closely with the base planes (Bontemps et al.f 1975; Latt and Stetten, 1976). This secondary binding leads to fluorescence quenching (Bontemps et al, 1975) and being cooperative suggests conformational changes in DNA. (4) A manifestation of the quenching process is a shift of the fluorescence emission to longer wavelengths (lower energy).

12 Nucleic acids and protein

The central dogma of biology, a term coined by Francis Crick, is that DNA makes RNA and RNA makes protein. It is self-evident, therefore, that the capacity simultaneously to measure DNA, RNA and protein at the individual cell level is of fundamental importance. We have seen in the previous chapter that DNA and RNA can be measured simultaneously. However, those measurements were of total DNA and RNA and ideally we would also wish to be able to measure specific gene copy number, messenger RNA (mRNA) and the protein product at the single cell level. The last of these can be quantitated with ease by flow cytometry using monoclonal antibodies and developments are taking place for the detection and measurement of specific genes with in situ hybridization in whole nuclei (Trask et al, 1985,1988). Furthermore, some recent work has enabled ribosomal RNA to be detected in suspended cells using hybridization (Bauman and Bentvelzen, 1988) and Dunne, Thomas and Lee (1989) have sorted small numbers of tells directly onto nitrocellulose filters then probed for interleukin-1 mRNA. However, specific mRNA and copy number cannot yet be quantitated, but I suspect that this will be possible within the next five years or so with improvements in hybridization techniques, the use of nick-translation with biotinolated nucleotides, streptavidin fluorochrome amplification and improvements in instrumentation. The simultaneous staining of DNA with propidium iodide and protein with fluorescein, either directly (FITC) or indirectly with antibodies, has been mentioned in sections 3.S3, 10.3 and 11.1.2 and application of this particular combination will be considered further in sections 12.3, 12.4 and chapter 15. The major problem with propidium iodide is that cells must be permeabilized to allow entry of the highly polar dye. This can only be overcome for viable cell analysis using the Hoechst dyes.

12.1

Viable cells

The combination of DNA staining with Hoechst 33342 and surface immunofluorescence with fluorescein on viable lymphocytes and myeloma cells was introduced by Loken (1980). This combination of fluorochromes requires two excitation wavelengths, UV for DNA/Hoechst and 488 nm for fluorescein. Loken (1980) excited both dyes with a single argon laser containing a specially

267

VIABLE CELLS

constructed set of multi-line mirrors which allowed both the UV and blue lines to be emitted by the laser. Initially, some difficulty was encountered due to the brightness of the DNA/Hoechst fluorescence emission in comparison with that from the fluorescein (see section 3.8.3). However, this was overcome by using biotinolated antibodies with avidin-coupled fluorescein as the second layer to give a degree of immunofluorescence amplification, reducing the output power of the UV compared with the blue laser lines by altering the current and magnetic field applied to the laser tube and by electronic compensation for the spectral overlap (see section 4.2.3). Some of these manoeuvres are highly specialized and it is preferable to use two lasers (if you have them) for three reasons. Firstly, the light outputs can be independently controlled with complete precision. Secondly, both beams can be focussed to the same point or offset in the vertical plane for sequential illumination. Finally, sequential excitation overcomes potential spectral overlap problems. Loken (1980) probed lymph node lymphocytes with an anti-Thy-1.2 antibody then stained with Hoechst 33342 at a concentration of 1.0(ig ml" 1 . Tlymphocytes stain more dimly with this concentration of Hoechst 33342 and he was able to show that these cells were Thy-1.2 positive and obtained a complete discrimination from the B-cells. Exponentially growing S107 myeloma cells, which express large quantities of cell surface IgA, were similarly double stained with Hoechst 33342 and a rabbit anti-IgA antibody and these results are reproduced in figure 12.1. Panel A shows IgA-associated fluorescence on the ordinate versus

cr

o

(0

DNA Fluorescence

Frequency

Figure 12.1. Panel A shows the dot-plot of specific IgA versus specific DNA fluorescence, from fluorescein and Hoechst 33342 respectively, after electronic compensation. Panels B and C respectively show the associated DNA and IgA monodimensional histograms. Redrawn from Loken, 1980.

268

NUCLEIC ACIDS AND PROTEIN

DNA/Hoechst fluorescence on the abscissa with the respective mono-dimensional histograms in panels B and C. Hollander and Loken (1988) extended this approach to the analysis of human bone marrow using sequential excitation with two lasers emitting 488 nm and UV light respectively. The 488 nm line excited either fluorescein or phycoerythrin tagged CD4 or CD34 and the UV was used for Hoechst/DNA. The progenitor cells defined as CD34 positive had the same G 0 /Gl DNA peak as other subsets and these were relatively quiescent with an overall DNA content proliferative index of about 6%. Almost all the cells with S, G2 + M DNA content were within the lymphoid, erythroid and myeloid series after CD34 expression was lost during

NUCLEAR-ASSOCIATED ANTIGENS

269

involved in DNA packing and together with DNA they form chromatin. There are five histones. Four of these, in pairs making an octomeric unit, constitute the cylindrical nucleosome core around which is wrapped two turns of doublestranded DNA. The 'input' and 'output' DNA strands to the nucleosome are tacked into place by the fifth histone. DNA between each nucleosome is termed linker DNA and the whole structure resembles 'beads on a string'. The nuclear enzymes include DNA polymerases, restriction enzymes, ligases, damage recognition and repair enzymes, transcriptases, DNAases and topoisomerases. Polymerases, as their name implies, are responsible for linking together the bases and deoxyribose sugars during DNA synthesis. Restriction enzymes cut DNA at very specific base sequences. They may be involved in physiological geonomic rearrangements of the types which occur at the immunoglobulin locus with the C|i to S|i heavy chain switch as spleen B-cells convert from IgM expressions to IgG secretors. The ligases join DNA strands together at specific sites and their function is the converse of restriction enzymes. DNA damage recognition/repair enzymes are not clearly understood. Thymine dymers, induced by UV irradiation, are repaired in normal cells and there may well be a component where damage has to be recognized before it can be repaired. The repair process involves DNAases which digest out a stretch of bases either side of the damaged site before reconstruction takes place on the complementary strand. It is, however, contentious as to whether specific enzymes exist which recognize the damage before it can be repaired. The transcriptases are responsible for transcribing the genetic code contained within the DNA into messenger RNA which is subsequently translated into protein in the ribosomes. Topoisomerases are DNA strand passing enzymes and are of two types, I and II, which respectively pass single- and double-stranded DNA. They are responsible for the topological integrity of DNA and are involved in winding and unwinding super-helical DNA by strand breakage and rejoining.

12.3.1 Quantitation with antibodies One of the first reports of nuclear-associated antigen quantitation is due to Gershey (1980) who used an antibody directed to SV-40 large T-antigen which binds to chromosomes in infected cells (D'Alisa and Gershey, 1978; D'Alisa, Korf and Gershey, 1979). A fluorescenated probe was used for large T with propidium iodide for DNA content. Since then a number of antibodies have been produced which recognize nuclear antigens and some are now commercially available. They include antibodies to p53 (Harlow et al, 1981), p62c~myc (Evan et al, 1985), p55c~f°s (Evan et al, 1985) and topoisomerase II (Liu, 1983). Some antibodies recognize nuclear antigens in stimulated cells, 6-B1012/N (Reeve, personal communication) and Ki-67 (Gerdes et al, 1983) and others recognize preferentially cell cycle dependent nuclear antigens. These include specific regions in condensed chromatin (MAB 244-7, mW 34 Kd), interchromatin granules (MAB 780-3, mW 105 Kd and 41 Kd) and euchromatin (MAB 58-15, mW 36 Kd) (Epstein and

270

NUCLEIC ACIDS AND PROTEIN

Clevenger, 1985; Clevenger et al, 1985). This list is by no means exhaustive and the number of antibodies which recognize nuclear components is increasing very rapidly. The precise functions of most of the proteins cited in the previous paragraph are not understood but possibilities include both reception and/or transduction of gene regulatory signals from the cytoplasm. Likely candidates for regulatory functions are the various steroid receptors and the protein products of the c-myc, c-fos and c-myb genes; p 5 5 c ~ / o s appears to be a factor necessary for transcription (Distel et al, 1987; Lech, Anderson and Brent, 1988) but the function of the c-myc and related proteins of the myc family (encoded by the N-myc, L-myc and v-myc genes) are not known although the evidence suggests that the c-myc product is involved in cell proliferation regulation. Kelly et al (1983, 1984) have shown that c-myc messenger RNA increases rapidly after mitotic stimulation of lymphoid cells. Similar results with hepatocytes after partial hepatectomy (Makino, Hayashi and Sugimura, 1984) and with growth factor stimulation of quiescent 3T3 cells (Greenberg and Ziff, 1984) have been obtained. It may also play a part in differentiation as mRNA copy number shows a peak at 4—5 weeks in developing placenta (Pfeiffer-Ohlsson et al, 1984) and a peak during spermatogenesis with stem cells and mature sperm showing very low levels (Stewart, Bellve and Leder, 1984). The protein products of the mouse c-myc and human c-myc, N-mcy and L-myc genes share a common amino acid sequence motif, the 'leucine zipper7, in conserved regions (Landschultz, Johnson and McKnight, 1988). This motif is also found in the yeast DNA regulatory protein GCN4 and in the proteins encoded by the v-fos and jun oncogenes which have transcriptional activity (Vogt, Bos and Doolittle, 1987; Landschultz et al, 1988). Evan and Hancock (1985) have shown that p62 c ~ myc is one of a discrete set of non-histone and non-nuclear matrix proteins which elute from the nucleus at salt concentrations below 200 mM. This evidence taken together with the structure of the conserved region suggests a DNA binding function which can be modulated rapidly by ionic changes within the physiological concentration range. Furthermore, both the protein and its mRNA are turned-over with half-times of between 20 and 40 minutes in stimulated cells (Hann, Thompson and Eisenman, 1985; Rabbitts et al, 1985) with no cell cycle phase dependency (Thompson et al, 1985; Rabbitts et al, 1985). The mRNA and protein were synthesized de novo during each phase of the cell cycle and a time course experiment showed that an increase in p62c~myc level in Gl preceeded entry into S-phase. With subsequent development of quiescence the level in Gl decreased before the S-phase fraction decrease (Rabbitts et al, 1985; and see later in figure 12.3). Most of the above information was obtained with blotting techniques which can only give the grand average for the whole population. However, the work of Rabbitts et al (1985) also contained flow cytometric data. The left panel of figure 12.2 shows the results obtained from serum-stimulated exponentially growing 3T3 cells stained for DNA (ordinate) versus p62c~myc associated immunofluorescence (abscissa) using propidium iodide (red) and fluorescein (green) respectively

NUCLEAR-ASSOCIATED ANTIGENS

271

Figure 12.2. The left panel shows DNA (ordinate) versus specific fluorescence on the abscissa for stimulated 3T3 cells. The right panel shows that the p62 c~myc associated fluorescence is abolished by pre-incubating the anti-p62 c ~ myc antibody with the peptide used as the immunogen.

after freeze-thaw premeabilization. The monoclonal antibody used to probe ( M Y C J.5E2Q) w a s S y n thetic peptide induced (Evan et al, 1985) and the right panel shows that preincubation of the antibody with the peptide used as the immunogen before staining abolished the green fluorescence signal. These data demonstrated that the p62 c ~ myc levels do not exhibit cell cycle dependent changes which confirmed the Western blotting data of Thompson et al (1985). Figure 12.3 shows the time course experiment mentioned above. Total DNA content and p62c~myc levels were determined simultaneously at daily intervals in 3T3 cells after stimulating quiescent cells by splitting the monolayer and reseeding at low density in fresh medium. The solid curve and squares show the p62c~myc changes in Gl (left ordinate) versus time (abscissa) and the open triangles and dashed curve depict the proportions in S-phase (right ordinate). The potential value of flow cytometric techniques is also demonstrated by work from Clevenger et al (1985), R. J. Epstein (1988) and R. J. Epstein et al (1988). The former group of authors used a monoclonal antibody directed towards interchromatin granules (MAB 780). These data, obtained with the paraformaldehyde/triton technique and the usual fluorescein (antibody) and propidium iodide (DNA) staining combination, are shown in figure 12.4. The log of the MAB 780 associated fluorescence is plotted on the ordinate versus DNA on the abscissa. The arrowed population is intensely stained and represents mitotic cells, but these would not have been apparent with a 'bulk' analysis procedure such as Western blotting which cannot reveal minority subsets in heterogeneous samples. p62 c-myc

NUCLEIC ACIDS AND PROTEIN

272 2500-1

p 6 2 c-myc

-60

CO

o

-40

S-phase

1500-

CO JZ

a i

CO

CNJ

CD

a

D A

N

^

"A

A.

500-

A A

4

1

T

20

10

6 Time, Days

Figure 12.3. Time course for changes in p62 c myc levels in G l (squares and left ordinate) and percentages of cells in S-phase (triangles and right ordinate) after stimulating quiescent 3T3 cells with fresh medium (Rabbitts et al, 1985).

MAB 780 240
c o

^

I!

*" - 120 c 2 o

120

240

DNA content (channel number) Figure 12.4. The log of the MAB 780 associated fluorescence plotted on the ordinate versus DNA on the abscissa. The intensely stained arrowed population represents mitotic cells. Redrawn from Clevenger et al. (1985).

NUCLEAR-ASSOCIATED ANTIGENS

273

Epstein (1988) and Epstein et al. (1989) have studied topoisomerase II levels simultaneously with DNA using fluorescein immunofluorescence and propidium iodide in T-4 7D breast cancer cells during oestrogen stimulation and some of these results are shown in figure 12.5. The origin in each panel is on the left with DNA displayed obliquely from top left to bottom right and topoisomerase II associated fluorescence displayed from bottom left to top right. Frequency, as usual, is scored on the vertical (Y) axis. Control cells stained with the fluorescenated second antibody only are shown in the top left panel. The bottom two panels demonstrate the effects of stimulating cells with oestrogen for 4 and 16 hours, shown in the left and right panel respectively. With increasing time of stimulation there are increasing number of cells exhibiting positive staining with the anti-topoisomerase II antibody. However, the bottom right panel(16 hours stimulation) shows that a small subset of cells in Gl do not exhibit a marked increase in topoisomerase II. The top right panel shows the effects of treating the 16 hour stimulated cells with a high salt buffer which partially dissociates topoisomerase II from the nucleus thus reducing the fluorescence signal. These data demonstrate that this type of flow cytometric technique is capable of determining changes within subsets in heterogeneous populations that could not be made with bulk sample methods. 12.3.2 Turnover measurements Turnover studies of p62c~myc were also conducted by Rabbitts et al (1985) using flow cytometry and these data are shown in figure 12.6. Exponentially growing 3T3 cells at 48 hours after serum stimulation, when both the p62 c ~ myc levels and proportion in S-phase were at a maximum (see figure 12.3), were treated with cyclohexamide which blocks protein synthesis. The MYC 1—6E10 antibody was used to probe for the protein at intervals thereafter and the top, middle and bottom panels respectively show the p62c~myc changes in Gl, S and G2 + M. There was an initial small rise in p62 c ~ myc levels in all cell cycle phases followed by a rapid fall compatible with a half-time of 30 minutes. Western blot data from parallel experiments using actinomycin-D to block mRNA then subsequently the protein, gave directly comparable qualitative results, see figure 12.7. However, these data were only meaningful as the protein levels and turnover were very similar throughout the cell cycle. If there had been marked cell cycle dependent absolute levels and/or turnover the Western blot data would have been meaningless. Furthermore, this would not have been apparent without either similar studies on synchronized cells or flow cytometric analysis. 12.3.3 Cell cycle modulation The data discussed in sections 12.3.1 and 12.3.2 concerning the role of p62 c ~ myc point to some form of control mechanism which is acting at the G o to Gl transition with p62c~myc elevation being required not only to facilitate entry of cells into the cell cycle but also to maintain cells in a potentially active proliferating state. Some evidence suggests that p62c~myc participates in DNA synthesis (Studzinski et al, 1986). However, cells are extraordinarily efficient and logical in

#2 FLU CON

Figure 12.5. DNA (propidium iodide) versus topoisomerase II content (fluorescein immunofluorescence) in T-47D cells after oestrogen stimulation (Epstein, 1988, and Epstein et al., 1989). The origins are on the left with DNA displayed from top left to bottom right and topoisomerase II displayed from bottom left to top right. Fluorescence control cells are shown in the top left panel. Oestrogen stimulation for 4 and 16 hours are shown bottom left and bottom right respectively. The data after treatment with high salt buffer partially to dissociate topoisomerase II are shown top right.

NUCLEAR-ASSOCIATED ANTIGENS

lib

G1

f

I\t

*

20

40 Minutes

G2+M

Figure 12.6. Turnover of p62 c myc. Exponentially growing 3T3 cells were treated with cyclohexamide which blocks protein synthesis. The p62 c ~ myc levels were then followed versus time. There was an initial increase then a rapid decrease in all cell cycle phases (top, middle and bottom panels represent Gl, S and G2 + M respectively) compatible with a half-time of 30-40 minutes (Rabbitts et al., 1985).

276

NUCLEIC ACIDS AND PROTEIN

Act ino. D 30

60

120

240 mins.

Figure 12.7. Pulse-chase Western blot data for cells treated for 1 hour with [3H]lysine in the presence of 5 jig ml" 1 actinomycin-D then assayed at 30, 60, 120 and 240 minutes. There was a rapid decrease in the signal which was completely undetectable at 120 minutes (Rabbitts et a\., 1985). performing their various functions and there is no reason for the c-myc gene to be activated early in Gl to produce mRNA for p62 c ~ myc with both being turned-over at such rapid rates if the protein is only required in S-phase. Because of these considerations Jon Karn of the Laboratory of Molecular Biology, Cambridge, constructed a series of retroviral vectors containing the c-myc gene in various configurations to investigate p62c~myc cell cycle dependency using the BrdU-antibody technique. These retroviral constructs were composed of SV-40 promoter regions (SV0), c-myc long terminal repeat sequences (LTR), a c-myc 'mini-gene' containing the protein coding exon sequences without the introns and the neomycin resistance gene (NEO) for selection purposes. A number of retroviruses were constructed and the structures of three of these, VSN-2, NSM-4 and MSN-7, are shown in figure 12.8. All three viruses contained SV0 (SV-40

NUCLEAR-ASSOCIATED ANTIGENS

277

VSN-2

c

M

svol

MSV LTR

Mo LTR

NEO

I

R

NSM- 7 c MSV LTR

SK

M

^1 C

C

I

MSV LTR 11

SK

||

PR

CC

*

11

svoi

NEO

M !

* | 1

1

PH

MSN-4

i i

XSK

R.B

M

c svo

1

I

P E

M

i

NEO i

1

Cl

R.S.B

XSK

Cl

1

MYC

Mo LTR

\

P E

HP

*

i

t

H.P

Mo LTR Cl

XSK

Figure 12.8. Structures of the retroviral vectors carrying the human c-myc gene with introns sliced out (stippled) drawn in the form of proviral DNA constructs cloned in £ coli plasmids.

promoter), the NEO gene and myc LTRs, but only NSM-4 and MSN-7 contained the myc gene. In NSM-4 the myc gene was under the control of SVO, but in MSN-7 it was under control of its own LTR. After cells were infected with these viruses there was no enhanced production of p62c~myc in the VSN-2 cell line, some enhancement in the NSM-4 cells and considerable enhancement in the MSN-7 cells which was the expected result (Karn et al, 1989). The growth rates of the three cell lines showed progressive increases with p62c~myc amplification, with V S N < N S M < M S N , which I'll abbreviate to V-, N- and M-cells. The cell cycle kinetics were determined by following the changes in BrdU signals versus total DNA at hourly intervals after flash labelling and selected data sets are shown in figure 12.9 for the three cell lines. The left, middle and right columns show the data from the V-, N- and M-cells respectively and the rows, from top to bottom, show the data obtained at 0, 4, 8 and 12 hours after labelling. The origins in these three-dimensional displays are on the left of each panel where BrdU-associated green fluorescence (520RF AREA) is scored obliquely from bottom left to top right and total DNA (630RF AREA) is scored from top left to bottom right. Frequency is scored on the vertical axis. The large 'spike' in each of the top three rows, the time zero data for each of the three cell lines, represents unlabelled Gl cells and is centered at channel 300 on the DNA axis. The S-phase fractions are represented by the characteristic U-shaped patterns and the G2 + M cells, which are also unlabelled, are represented by the small 'spikes' centered at

27Z

NUCLEIC ACIDS AND PROTEIN

:

•

*

:

2.9. Bivariate perspective histograms of DNA versus BrdU for V-, N- and M-cells in the columns at 0,4,8 and 12 hours after pulse labelling in the rows. The origins are on the left with DNA displayed from top left to bottom right and BrdU incorporation displayed from bottom left to top right.

channel 600 on the DNA axis. The second row shows the data at 4 hours and in each panel the most obvious change is that some cells which were in late S-phase when the BrdU was added have progressed through G2 + M, split into two and in so doing have halved their BrdU content and now appear with Gl DNA content at channel 300 on the DNA axis (520RF AREA). These are the Gl cells with the positive green signals which appear to 'stream out7 from the large Gl spikes from bottom left to top right. The next row down shows the data obtained at 8 hours and a number of further changes have taken place. Firstly, most of the cells which were initially in S-phase (BrdU positive) have passed through G2 + M, halved their BrdU content and are now scored with Gl DNA content and positive green (BrdU) signals. Secondly, a relatively large fraction of the BrdU labelled V-cells are

NUCLEAR-ASSOCIATED ANTIGENS

279

still in G2 + M compared with the M-cells (row 3, left and right panels respectively). The N-cells, middle panel, show an intermediate pattern. Thirdly, it is now quite obvious that there has been significant progression of the initially unlabelled Gl cells which are BrdU negative. These have moved out of Gl along the DNA axis (630RF AREA) in the direction top left to bottom right and are progressing towards G2 + M. Finally, it is equally clear that the M-cells have progressed further towards G2 + M than have the V-cells, and again the N-cells occupy an intermediate position. By 12 hours the M-cells, right panel bottom row, are exhibiting a pattern qualitatively similar to their time zero counterpart in the right panel of the top row, but where the BrdU labelled cells have half their original green fluorescence signals. In contrast the V-cells at 12 hours, left panel bottom row, still show some cells in G2 + M with positive BrdU signals (these are situated on the rear right wall of the display as the spike) but the majority of these have a G l DNA content. Also, some of the initially unlabelled Gl cells are still in S-phase and G2 + M, where the latter are clearly seen as a spike at channel 600 on the DNA axis. Again the N-cells display an intermediate pattern between the V- and M-cells. The data in figure 12.9, together with the remaining majority of data sets in the experiment, were analysed by placing gates around both the labelled and unlabelled Gl, S and G2 + M regions to obtain the proportions in each of these compartments. Figure 12.10 illustrates this procedure where panel A shows the control time zero data for the V-cells and panel B shows the zones used in the

B

o a

o o k_

OQ

DNA fluorescence Figure 12.10. Panel A shows BrdU uptake into V-cells at time zero where the characteristic incorporation horse-shoe is readily evident. Panel B shows the gates used to assess the unlabelled proportions in Gl, S and G2 + M (zones 1, 3 and 2 respectively) and the labelled proportions in Gl, S and G2 + M (zones 4, 6 and 5 respectively).

28>0

NUCLEIC ACIDS AND PROTEIN Zone 1 (G1)

50-.

Zone 3 (S)

Zone 2 (G2:M)

VSN2

12

16

VSN2

0

4

8

12

16

12

16

Time, hours Figure 12.11. The proportions defined in zones 1, 3 and 2 of figure 12.10, unlabelled cells, plotted as the points on the ordinates versus time on the abscissae. The cell cycle phases, Gl, S and G2 + M, are arranged in the columns for V-, N- and M-cells in the rows. The curves were calculated from the 'Hartmann-Pederson' type computer model (Watson and Taylor, 1977; Kenter and Watson, 1987) described in section 11.5.3.

gating analysis procedure where 1, 3 and 2 correspond to unlabelled Gl, S and G2 + M cells respectively and where 4, 6 and 5 correspond to labelled Gl, S and G2 + M cells. The data for the unlabelled population are shown in figure 12.11 where the various proportions are plotted as the points on the ordinates versus time on the abscissae. The cell cycle phases, Gl, S and G2 + M are arranged in the columns for V-, N- and M-cells in the rows. Figure 12.12 shows the comparable data for the labelled cells, i.e. those initially in S-phase. The curves were calculated from the 'Hartmann-Pederson' type computer model (Watson and Taylor, 1977; Kenter and Watson, 1987) described in section 11.5.3. The durations of Gl, S and G2 + M with their 95% confidence limits are shown in figure 12.13 where it can be seen that the durations of S-phase and G2 + M were not significantly different in the three cell types. However, there was a progressive reduction in the duration of G l from V- to N- to M-cells. Further experiments were carried out where the three cell types were BrdU pulse-chased into serum free medium. Again, the durations of S-phase and of G2 + M were not altered but there was a progressive increase in the length of Gl which was greater in V- than in N- than in M-cells. These data not only add support to the hypothesis that p62 c ~myc is involved in cell cycle regulation but also suggest very strongly that this control is exerted in Gl. Furthermore, the control mechanisms for c-myc mRNA and protein are themselves typical of fast response regulatory systems. Blanchard et al. (1985) have

281

NUCLEAR-ASSOCIATED ANTIGENS Zone 4 ( G 1 )

Zone 5 (G2:M)

Zone 6 (S)

50-t

VSN2

8

12

16

Figure 12.12. Similar display to figure 12.11 for zones 4, 6 and 5 (Gl, S and G2 + M respectively) of figure 12.10 which represent the labelled cells which initially were in S-phase.

7-

6-

5-

.§ 4-

G2:M G1

I

3-

NM

NM

NM

Figure 12.13. Durations of Gl, S and G2 + M with their 95% confidence limits. The durations of S-phase and G2 + M were not significantly different in the three cell types. However, there was a progressive reduction in the duration of G l from V- to N- to M-cells.

282

NUCLEIC ACIDS AND PROTEIN

200-

i

°oi 1 0 0 a

30

60

90

120

Minutes

Figure 12.14. Turnover of p62c ~ myc in quiescent 3T3 cells, a similar experiment to that infigure12.6, but these data show that the protein is turned over with a very much longer half-time.

shown that the c-myc gene is transcribed at an equal rate in both quiescent and stimulated Chinese hamster fibroblasts. However, the message was not detectable in the former which means that degradation must have been taking place at the same rate as production. Following stimulation the mRNA level increased which had to be due to a decrease in the degradation rate as there was no demonstrable increase in transcription. This, therefore, represents part of a negative servo control loop which is inhibited by serum stimulation causing a rise in mRNA transcripts and a subsequent rise in protein content. The reverse control process seems to operate at the protein level. In quiescent cells the half-life of the protein is relatively long (250—350 minutes, see figure 12.14) but the absolute content is low; in stimulated cells the half-life is shorter but the level is higher. Thus, the p62 c ~myc content is controlled by two different processes, predominantly at the mRNA degradation level in quiescent cells and predominantly at the protein level after stimulation. The combination of these two control processes is capable of giving rise to very tight regulation of the absolute protein content with the possibility for very rapid modulation.

12.3.4 Dual antigens plus DNA This type of assay can be carried out using two antibodies and a DNA stain as mentioned in section 7.3.2 but the method requires dual wavelength excitation. Our approach (Watson et al.f 1987b) was to probe the c-myc protein with a mouse monoclonal anti-p62 c " m};c antibody then secondarily stain this with a sheep anti-mouse immunoglobulin (IgG) coupled to the UV fluorochrome amino-methyl coumarin acetic acid (AMCA). The c-fos protein was then probed with a synthetic peptide induced rabbit anti-serum which was then stained with a fluorescenated swine anti-rabbit antibody (FITC-SaR) and DNA was stained with propidium iodide. Figure 12.15 shows the emission spectra of the three fluorochromes together with the excitation wavelengths used and the band passes

CYTOPLASMIC ANTIGENS

80-

'LOW BLUE (AMCA)

7n-

283

60-

LU IZ

50-

f\

40-

> 30HI CC

I

2010/ 400

(PI)

I

I

I

I

1 1

I

I

I

I

I

I

I

1 1 1

\ \ \ \ \

1

/ f\ i ' / ^ i

i ' / ^i V;; 500

337 356

488

KRYPTON

ARGON

1

' '

4:

\i

/

3( ) 0

RED

GREEN (FLU)

"£<£-:\

1 600

tv •Xv!v!v!vXv!vN 700

WAVELENGTH

Figure 12.15. Emission spectra of AMCA, fluorescein and DNA/propidium iodide with the band passes through which the spectra were analysed indicated by the stippling. The krypton UV lines and the argon 488 nm line are indicated by the arrows.

through which the emissions were quantitated. The 90° scatter versus DNA data space for nuclei extracted from a paraffin wax embedded biopsy of ovarian carcinoma is shown in figure 12.16 which exhibits an aneuploid (Anu) as well as a diploid (Dip) component. These two populations were then gated and the p55c~fos versus p62c~myc associated fluorescence data (520 and 420 nm respectively on the ordinate and abscissa) for both populations are shown in figure 12.17.

12.4 Cytoplasmic antigens 12A.I

Immunoglobulin Zeile (1980) was perhaps the first investigator to analyse intracytoplasmic immunofluorescence simultaneously with DNA in a report which appeared in the same issue of Cytometry as Gershey's paper (1980, and see section 12.3.1) where SV-40 large T antigen was assayed. Zeile (1980) studied bone marrow from 12 patients with multiple myeloma which was disaggregated with collagenase, treated with ice-cold saline containing 0.01% triton X-100 then stained with fluorescein-conjugated anti-human immunoglobulin and propidium iodide.

284

NUCLEIC ACIDS AND PROTEIN

DNA

5

Anu

Figure 12.16. The 90° light scatter versus DNA for an aneuploid ovarian cancer biopsy extracted from paraffin wax. The diploid (Dip) and aneuploid (Anu) components are arrowed.

420 nm (MYC)

c/> O U-

o CM m

Figure 12.17. The green fluorescence (520 nm, p55 c fos versus low blue fluorescence (420 nm, p62c~myc) data space for the data of figure 12.16 in which the diploid (Dip) and aneuploid (Anu) components are completely separated.

CYTOPLASMIC ANTIGENS

285

Characteristic patterns of normal and pathological immunoglobulin-producing cells were obtained. Hayden et al (1988) extended this approach to quantitate cytoplasmic and surface antigens simultaneously but they did not also measure DNA although this could be carried out with some modifications to their technique. Their method involved initial cell surface antigen staining with a directly fluorescein-conjugated antibody followed by permeabilization and probing cytoplasmic antigens with biotinolated antibodies which were subsequently stained with avidin— phycoerythrin or streptavidin—texas red. A number of permeabilization steps were investigated including 0.1% gluteraldehyde plus 1% saponin (Newell, Hannam-Harris and Smith, 1983), 0.01% triton X-100 (Zeile, 1980), lysolethicin (Schroff et al, 1894) and 50% ethanol in either PBS or RPMI (Walker et al, 1985) and they finally chose the last of these. This method could be used to add DNA staining by probing for the latter on the red channel with propidium iodide, using the green fluorescein channel for the 'brightest' antigen and probing the second antigen with a biotinolated antibody and staining this with streptavidin-AMCA. This combination of fluorochromes was used in the previous section.

12.4.2 Cytoskeleton Mammalian cells contain filamentous protein networks which make up the cytoskeleton (Ramaekers et al, 1983). These can be classified into four types based on their diameters assessed by electron microscopy (Schliwa and van Blerkom, 1981). Microtrabecular filaments are the smallest measuring 2—3 nm in diameter. Microfilaments measure 5—7 nm and microtubules are 22—25 nm in diameter. The last group comprises the intermediate filaments which measure 7-11 nm in diameter. These different groups are further characterized by the various proteins which contribute to their structure, i.e. actin, tropomyosin, tubulin and the intermediate filament (IF) proteins. The latter are subdivided into five different groups which are tissue type specific namely, keratins (epithelia), vimentin (mesenchyme), desmin (muscle), neurofilament proteins (nerve cells) and glial fibrillary ascidic protein (GFAP, astrocytes) (Anderton, 1981; Franke et al, 1978, 1981; Holtzer et al, 1981; Lazarides, 1980, 1981, 1982; and Osborne et al, 1981). Moreover, neoplastic transformation does not alter tissue specific expression of intermediate filament proteins and antibodies to these have been used in the differential diagnosis of 'difficult' biopsies using immunofluorescence and immunoperoxidase techniques (Battifora et al, 1980; Ramaekers et al, 19S2a,b; Sieinski, Dorsett and Ioachim, 1981). These techniques have also been adapted for flow cytometry and used to distinguish between (Ramaekers et al, 1983) and sort cells (Oud et al, 1985) on their IF-protein content and to study tubulin expression in the cell cycle (Wang et al, 1988). The latter is potentially of importance in proliferation regulation. Wang and Rozengurt (1983) have shown that there appears to be an interaction between microtubules and cyclic AMP in modulating initiation of DNA synthesis in 3T3 cells and Ball, Albrecht and Carrey (1982) report that microtubules are involved in

NUCLEIC ACIDS AND PROTEIN

286

T24

T24

B

PI

Molt-4

Molt-4

4) O O (0 w4>

PI

|

Z T24+Molt-4

T24+Molt-4

T24 gate r

PI

|f^ DNA content

1

Cytokeratin, FITC

Figure 12.18. Dual parameter analysis of the T24 bladder cancer cell line and the lymphoid cell line Molt-4 using DNA and cytokeratin content. Panels A, C and E respectively show the DNA histograms for T24, Molt-4 and a mixture of the two. Panels B, D and F show the bivariate dot-plots of DNA (PI staining) versus FITC immunofluorescence probed cytokeratin. T24 was positively labelled (B, top right) but Molt-4 was negative (D, middle right). Panel F (bottom right) shows the mixture and the T24 gate' was used to assess the DNA histograms associated with each cell type in the mixture.

CYTOPLASMIC ANTIGENS

287

cytomegalovirus initiation of DNA synthesis. Moreover, Crossin and Carney (1981a,b) have shown that tubulin depolymerization early in the cell cycle is sufficient to initiate DNA synthesis (1981a) and that taxol stabilization of tubulin inhibits initiation of DNA synthesis after thrombin and epidermal growth factor stimulation (1981b). One of the first examples of the use of intermediate filament antibodies in flow cytometry to distinguish between different cell types is due to Ramaekers et al. (1984) and their results are reproduced in figure 12.18. The T24 bladder cancer cell line and Molt-4 lymphoid cells were assayed simultaneously for cytokeratin and DNA both separately and in a mixture of the two. The left column from top to bottom shows the DNA histograms (propidium iodide staining) for T24, Molt-4 and the mixture. The right column shows the bivariate cytograms of DNA on the ordinate versus cytokeratin associated immunofluorescence on the abscissa. T24 cells exhibited positive staining for cytokeratin (top panel) but Molt-4 showed no staining (middle panel). The mixture of the two cell types shown in the bottom panel on the right suggested that the two cell types could be completely resolved in the DNA/cytokeratin data space. This was confirmed by placing a gate to include the T24 cells as shown and re-analysing the DNA histogram data of each component of the mixed population. The results were essentially identical to those from the individual populations. It is also worth noting that a chicken RBC internal standard was used in these experiments and that the T24 and Molt-4 DNA histograms are very similar to the hypothetical histograms used as illustrations in figure 11.10.

13 Chromosomes

Conventional chromosome analysis and karyotyping is time consuming, laborious and to some extent subjective as it relies upon interpretation of banding patterns and morphological features. These can vary with staining conditions and preparative methods and results are usually presented for very few metaphase spreads. This inevitably introduces a selection bias as only those preparations which are interpretable can be presented. However, these may not always be representative of the population under study. These considerations apply to any image analysis system including the best of them all, the human eye and mind, but these will always be limited in statistical precision due to the relatively few numbers of objects analysed. A new approach in which many thousands of chromosomes can be analysed using flow cytometry, termed flow cytogenetics, was submitted to the National Academy of Sciences, USA, by Joe Gray and colleagues in December 1974 (Gray et al, 1975b). Chinese hamster cells (M3-1, clone 650A) were used and chromosomes were prepared according to the method of Wray and Stubblefield (1970). The chromosomes were stained with ethidium bromide and the results are reproduced in figure 13.1 A. Later in 1975 Gray et al. (1975a) produced the first flow karyotype derived from a human cell line which is shown in figure 13.IB. By today's standards the human data set shown in this figure is not very impressive; however, this heralded the start of a new area of research and since those initial reports there have been incredible technological and preparative advances. The major technical developments have been made at the Lawrence Livermore and Los Alamos biomedical laboratories but other groups have made very significant contributions. It is now possible with flow cytometry to assay not only for total DNA but also for base composition, centromeric index, banding patterns and chromosome associated proteins.

13.1

Harvesting mitotic cells

Chromosomes are easiest to harvest from tissue culture adapted cells growing as a monolayer. The culture is treated with colcemid for about one cell cycle time to enrich the metaphase fraction and these are harvested by mitotic selection (Teresima and Tolmach, 1963) as described in section 11.5.2. More

HARVESTING MITOTIC CELLS

289

9000 -

I 4>

4500 -

1.0

Fluorescence intensity

Figure 13.1. Panel A shows the flow karyotype of M3-1 Chinese hamster cells clone 650A and panel B shows the karyotype derived from a human cell line redrawn from Gray et al. 1975b (A), 1975a (B). difficulty is encountered with suspension cultures as it is not possible to positively select for the mitotic fraction as with monolayers. However, even ficol-paque enriched peripheral blood lymphocytes, stimulated to divide by PHA (T-cells) or LPS (B-cells) and treated with colcemid can give a very workable chromosome: intact nuclear yield. Following cell lysis to release chromosomes (see below) any debris, clumps and intact nuclei can be removed to some extent by sedimentation through a 10—50% glycerol linear concentration gradient (Gooderham and Jeppeson, 1983; Fantes et al., 1989) or a 20-40% sucrose gradient (Wray, 1973).

290

CHROMOSOMES

13.2

Chromosome preparation

The ideal preparation method would produce chromosomes in fluid suspension with no chromosomal fragmentation or clumping and no intact nuclei. In practice, even with the best procedures, there is frequently some fragmentation, occasional clumping and some intact nuclei. Chromosomes are liberated from the metaphase enriched population by combinations of hypotonic solutions, detergent treatment and mechanical shearing. The latter has included syringing through fine-gauge needles (Gray et al, 1975a,b; Yu et al, 1981), the use of mechanical homogenizers (Sillar and Young, 1981), vigorous vortexing (Young et al, 1981) and sonication. These procedures can also damage chromosomes but this can be minimized by addition of polyvalant cations or reducing agents (Blumenthal et al, 1979; Sillar and Young, 1981). The shearing vigour required to disrupt the external cell membrane depends to some extent on cell type. Fibroblasts which have 'tougher' membranes than lymphoblastoid cell lines and lymphocytes require the more vigorous shearing (Langlois et al, 19SO; Stohr et al, 1982; Stewart et al, 1985) but, this has to be evaluated for each cell type (Gray and Langlois, 1986). 13.2.1 Hexylene glycol This was the first chromosome isolation technique to be used in flow cytometry and was based on that of Wray and Stubblefield (1970). The harvested mitotic cells were first suspended in hypotonic KC1 (75 mM) at 4 °Cfor 30 minutes then resuspended and lysed by mechanical shearing in 1.0 M hexylene glycol (2-methyl-2,4-pentanediol) and 1 mM CaCl2 in PIPES buffer, pH 6.7. The latter was composed of 10 mM piperazine-N,N"-bis(2-ethane sulphonic acid) monosodium monohydrate plus 10 mM PMSF (phenylmethylsulphonylfluoride). The presence of the divalent cations (Ca2 + ) stabilizes the chromosomes and analysis of both histone and non-histone proteins shows minimal alterations following this isolation method (Wray and Wray, 1979, 1980) which also yields high molecular weight DNA (Yu et al, 1981). A variation on this theme has been used by Sillar and Young (1981) where cells were resuspended in a buffer containing 0.75 M hexylene glycol, 25 mM trisma base, 0.5 mM CaCl2 and 1.0 mM MgCl2, pH 7.5, for 10 minutes before disruption with an MSE homogenizer. Chromosomes from a number of cell types including fibroblasts, lymphoblastoid cells and various tumour cell lines have been obtained by this method. It also has the advantage that the chromosomes are not contracted and they are suitable for banding pattern analysis following staining and sorting (Gray et al, 1979).

13.2.2 Polyamine This method involves treating the mitotic cells with hypotonic buffer containing 15 mM Tris-HCl, 0.2 mM spermine, 0.5 mM spermidine, 2 mM EDTA, 0.5 mM EGTA, SO mM KC1, 20 mM NaCl and 14 mM p-mercaptoethanol, pH 7.2. After washing and resuspending in the same buffer the detergent digitonin is added to a final concentration of 0.1% and the suspension is then subjected to

CHROMOSOME PREPARATION

291

vigorous vortexing (Sillar and Young, 1981, modified from Blumenthal et al, 1979). This method can be used with most DNA fluorochromes if divalent cations are added during the staining procedure (Lalande et al, 1985) and again high molecular weight DNA is obtained (Blumenthal et al, 1979). However, it suffers from the disadvantage with ethidium bromide staining that the chromosomes are very contracted and are not suitable for subsequent banding (Davies et al, 1981; Buys, Koerts and Aten, 1982).

13.2.3 Hypotonic PI detergent The double-helical structure of DNA is normally in an 'extended7 form as the negatively charged phosphate residues of the sugar—phosphate backbones on the outsides of the helix exert mutually repulsive electrostatic forces as the phosphates are in relatively close apposition. The phenenthridinium dyes, ethidium bromide and propidium iodide, intercalate with the double helix such that their linked tricyclic ring structures (see figure 11.4) 'slot-in' between, and parallel to, the bases positioned in the center of the helix. The positive charges of the dyes then neutralize the negative charges of the phosphates and the helix compacts along its length. Propidium iodide is more efficient than ethidium bromide in this respect as it has two positive charges as opposed to just one. This property has been exploited by Aten, Kipp and Barendsen (1980), Buys et al. (1982), Bijman (1983) and Kooi et al (1984) to stabilize the DNA of chromosomes during hypotonic detergent lysis. Metaphase cells were first treated with hypotonic KCl containing propidium iodide then triton-X 100 was added and the cells were syringed. The chromosomes so produced gave high molecular weight DNA (Yu et al, 1981) and were suitable for subsequent banding. There is also the advantage that the stabilizing agent, propidium iodide, is the DNA fluorochrome which is fine if you want a measure of total DNA per chromosome. If, however, you want to use a different stain then non-fluorescent stabilizing agents, e.g. psoralen derivatives (Yu et al, 1981), have to be used.

13.2.4 Ohnuki buffer This buffer enables chromosomes to be prepared in elongated form whilst preserving their morphology (Ohnuki, 1965, 1968). It has been used by Bartholdi et al (1989) to produce banding analysis by slit-scanning (see section 13.5.2). Metaphase cells are suspended in the hypotonic buffer comprised of an equimolar solution of 55mM KCl, NaNO 3 and NaC 2 H 3 O 2 in a ratio of 10:5:2 (Ohnuki, 1965). Following swelling the cells are treated with the polyamine isolation buffer described in section 13.2.2.

13.2.5 Magnesium sulphate An isolation technique which is suitable for many DNA stains and chromosomes from various cell types, including hybridomas, human amniotic cells and chorionic villus biopsies (Trask et al, 1984) has been developed by van den Engh et al (1984, 1985). The mitotic cells were resuspended in hypotonic MgSO 4 containing dithioerythritol and triton X-100 prior to syringing.

292

CHROMOSOMES

13.3

Staining

All of the DNA specific stains and the phenanthridinium dyes have been used for total DNA staining of chromosomes. The former group has the potential disadvantage that UV excitation is required but this constitutes no problem for mercury arc lamp based systems or those with a laser tunable to UV lines.

13.3.1 Total DNA Comparative studies of the UV excited ligands, DAPI, DIPI and Hoechst stains have been carried out by Otto and Tsou (1985) and CVs between 2.2% and 2.9% were obtained. Overall the best results were obtained with DAPI with the lowest CV and a brightness only 3% less than DAPI. Hoechst 33342 gave the worst results in these experiments with a CV of 2.9% and a brightness 14% less than DAPI. However, all these stains gave the same number of chromosome peaks with chinese hamster chromosomes. Chromomycin A 3 has been used as a single DNA stain in slit-scan studies (see section 13.5) but there is considerable overlapping of the human chromosomes with only eight well-defined peaks being obtained. These are shown in figure 13.2 which is redrawn from Bartholdi et al. (1989). The polyamine/ethidium bromide method developed by Sillar and Young (1981) would seem to have a number of advantages as the dye/DNA complex is 100

X,8-12

Chromomycin fluorescence

1024

Figure 13.2. Human karyotype stained with chromomycin A 3 in which only eight major peaks can be defined.

STAINING

293

exited by the most frequently used 488 nm argon line, many of the chromosomes are resolved uniquely, ethidium bromide is relatively inexpensive, the method is very reproducible, it can be performed readily on a regular cell sorter (Young et al, 1981) and chromosomal DNA suitable for cloning and library generation can be obtained (Davies et al, 1981, and see figure 6.6). Figure 13.3 shows two karyotypes from peripheral lymphocytes of the same individual assayed six months apart (redrawn from Young et al, 1981). However, as can be seen from figures 6.6 and 13.3 not all chromosomes can be resolved with this method. There is complete overlap of the 9—12 group, considerable overlap of the 13—15 group and frequently pairs of the smaller chromosomes cannot uniquely be resolved.

13.3.2 A-T:G-C composition A dual staining technique was developed at Lawrence Livermore Laboratories using a dual-beam system (Dean and Pinkel, 1978) to excite chromosomes stained with chromomycin A 3 and Hoechst 33258 (Carrano et al, 1979; Langlois et al, 1982). These dyes bind independently to DNA with chromomycin A 3 showing G—C specificity and Hoechst 33258 exhibiting A—T specificity (see section 11.1). The Hoechst dyes are excited by UV light and chromomycin A 3 is excited by mid blue light (458 nm argon line) but not by UV. The UV excited emission from Hoechst 33258 overlaps the absorption spectrum of chromomycin A 3 and the fluorescence energy from the Hoechst 33258/DNA complex undergoes resonant energy transfer to the chromomycin A 3 (Langlois et al, 1980). Thus, by arranging for the double-stained chromosomes to pass first through the UV beam and then through the 458 nm beam we get two flashes of fluorescent green light. The first is due to UV energy absorbed by the Hoechst 33258 bound to A—T rich sequences which is then energy transferred to chromomycin A 3 and the latter emits green fluorescence in direct proportion to A-T content. When the chromosome subsequently passes through the 458 nm excitation beam the energy is absorbed directly by chromomycin A 3 bound preferentially to G-C rich sequences and the second flash of green fluorescence is directly proportional to G-C content. The processes involved in this technique are shown diagramatically in figure 13.4 and figure 13.5 shows a bivariate karyotype produced by this method and redrawn from Langlois et al. (1982) where Hoechst 33258 dependent fluorescence is plotted on the ordinate versus chromomycin A 3 on the abscissa. The left panel shows the whole distribution and the right panel is a 'magnified' view of the smaller chromosomes. The 9—12 group still remains a single cluster but the remainder have largely been resolved and figure 13.6 shows a statistical summary of the positions of the peaks of the chromosome distributions from a study of 10 normal individuals.

13.3.3 Bromodeoxyuridine We saw earlier (section 11.5.5) that fluorescence from Hoechst dyes is quenched by bromodeoxyuridine incorporation into DNA (Latt, 1977). However, fluorescence from chromomycin A 3 is slightly enhanced (Swartzendruber, 1977)

Fluorescence intensity Figure 13.3. Human flow karyotype prepared by the polyamine/ethidium bromide method from Young et al., 1981. The two panels were obtained from the same individual with an intervening interval of six months. There are three peaks on the far right where usually there is only one with this staining method (compare with figure 6.6). The peaks labelled IA and IB represent heteromorphism of the centric heterochromatin in the chromosome.

STAINING

295

458nm beam detector Collecting lens

UV beam detector Double stained chromosomes

458nm beam Figure 13.4. Sequential excitation of Hoechst 33258; chromomycin A 3 stained chromosomes with first the UV beam and second the mid-blue beam. The Hoechst 33258: DNA complex is excited on passing through the UV beam and the absorbed energy is transferred to chromomycin A 3 which emits fluorescence in proportion to A—T content. The chromomycin A 3 is excited directly on passing through the mid-blue beam and this same dye then emits fluorescence in proportion to G—C content.

oo

LO CM CO CO

4 O

z

Chromomycin Figure 13.5. Bivariate karyotype produced by dual laser excitation and double staining with Hoechst 33258 (ordinate) and chromomycin A 3 (abscissa). The whole karyotype is shown on the left and 'magnified' view of the smaller chromosomes on the right. Note that almost all the chromosomes are completely resolved apart from the 9—12 group. Redrawn from Langlois et al, 1982.

CHROMOSOMES

296

1

I

I

I

2 4

4 -

°

O3

O5 O6

3 08

o 0) o

13

0

2 -

1 -

09-12

Q 14.15

18 o

Y0

016

20

O17

21° 0 °i y° 22

I

I

1

2

I

I

3

4

5

Chromomycin 13.6. Statistical summary of the chromosome peak positions from ten individuals with the Hoechst: chromomycin technique. Small ellipses (chromosomes 2, 7, 8, 9—12,18 and 20) indicate little variation. In contrast chromosomes 14 and 15 exhibit fairly large variability in position from individual to individual.

just as with ethidium bromide. These properties can be used to study the times in S-phase at which different chromosomes are replicated (Cremer and Gray, 1982, 1983) and two approaches can be employed. Firstly, monolayer cells can be synchronized by mitotic selection. Different aliquots of cells would be flash labelled with BrdU for one hour at different times during the cell cycle and treated with colcemid to harvest metaphase cells on the first pass through mitosis after taking up the analogue during S-phase. Chromosomes which were replicated during the interval that BrdU was present would then exhibit a partial loss of Hoechst related fluorescence and a slight gain in chromomycin A 3 fluorescence. It was stated in section 11.5.5 that flash labelling with BrdU produced insufficient quenching of Hoechst 33342 fluorescence to make a reliable distinction between cells which had replicated their DNA during the labelling interval and those that had not (see figure 11.30). However, the position is somewhat different for individual chromosome analysis. In the former case an attempt is being made to assay for Hoechst quenching in whole nuclei within a background where the majority of DNA has not incorporated BrdU. In the latter case the majority of replicated DNA within a given chromosome will have incorporated the analogue if that chromosome was being replicated when the label was present. The second approach would use asynchronously growing cultures and BrdU would be added for intervals of one hour at increasing times before addition of

STAINING

297

colcemid and harvesting the metaphase cells. Thus, if the interval between BrdU labelling and addition of colcemid and harvesting mitotic cells was short we would be observing chromosomes replicated late in S-phase. If the interval between labelling and colcemid was successively increased we would be observing chromosomes replicated earlier and earlier in S-phase.

13.3.4 Partial sequence specificity A number of non-fluorescent minor groove ligands (Baguely, 1982; Zimmer and Wahnert, 1986; Smith, Debenham and Watson, 1989) with A-T specificity, including distamycin and netropsin, can compete for Hoechst dye binding sites (Meyne et al, 1984; Lalande et al, 1985). However, even though they have A—T specificity they also have specific sequence specificity (Lane, Dabrowiak and Vournakis, 1983; Martin and Holmes, 1983; Scamrov and Beabealashvilli, 1983; Van Dyke and Dervan, 1983a,b). Thus, the non-fluorescent minor groove ligands will compete for Hoechst dye binding sites and reduce its fluorescence in preferred A-T regions where the flanking sequences are most conducive to their binding. Such non-fluorescent ligand/Hoechst dye combinations have been used by Sahar and Latt (1978, 1980) and Latt et al. (19&0a,b) to identify polymorphic regions in human metaphase spread chromosomes which contain highly repetitive DNA sequences. This method has been applied in flow cytometry to identify chromosomes 9 and 15 which have long stretches of DNA which bind Hoechst dyes but do not bind netropsin or distamycin (Meyne et al, 1984; Lalande et al, 1985). Thus, these two chromosomes have very little reduction in their Hoechst fluorescence but the remainder exhibit considerable reduction which enables the distinctions to be made.

13.3.5 Chromosome-associated proteins A very brief summary of some of the nuclear proteins was given in section 12.3. The non-histone nuclear structural proteins also contribute to the chromosome skeleton in metaphase with loops of chromatin 'pegged' to the central scaffold, a concept originally proposed by Adolph, Cheng and Laemmli (1977), Paulson and Laemmli (1977) and Laemmli et al. (1978). More recently Earnshaw and colleagues have shown that the kinetochore (centromere) is part of the metaphase chromosome scaffold (Earnshaw et al., 1984). Moreover, Earnshaw and Heck (1985) and Earnshaw et al (1985) have shown that topoisomerase II is a structural component of the scaffold and have demonstrated its central location in mitotic chromosomes using immunofluorescence. As yet, little work has been carried out in flow cytometry to assay for metaphase chromosome proteins as opposed to nuclear-associated proteins. However, extensive studies have been carried out using immunocytochemical techniques to study nucleosome heterogeneity, visualize the presence of histones in transcriptionally active chromatin and isolate DNA sequences associated with specific chromosomal proteins (Bustin, 1987) and considerable potential exists in this area for both flow and image cytometry. Our group have used similar techniques to those of Earnshaw and Heck

298

CHROMOSOMES

Hoechst fluorescence Figure 13.7. A mixture of human chromosomes (Hi, H2, H3 and H4) and Chinese hamster chromosomes (Cl, C2, C3 and C4) double stained for DNA with Hoechst 33258 and human histone H2B with FITC immunofluorescence. The human chromosomes stained with greater intensity on the immunofluorescence

to assay for topoisomerase II in whole nuclei (see figure 12.5 and Epstein, 1988; Epstein et al, 1989) and there seems no reason why this could not also be extended to chromosome analysis. Flow cytometric methods have been developed to assay for histone proteins using immunofluorescence (fluorescein) simultaneously with DNA Hoechst fluorescence and an example from Trask et al. (1984) is shown in figure 13.7. A mixture of Chinese hamster and human chromosomes was used which were probed with an antibody raised against human histone H2B. Fluorescein immunofluorescence is plotted on the ordinate versus DNA on the abscissa and the human chromosomes labelled with greater intensity compared with the hamster chromosomes. Autoantibodies which recognize centromeric structures have been found in the sera of patients with scleroderma (Moroi et al, 1980) and Earnshaw and Rothfield (1985) have identified a family of human centromeric proteins using autoimmune sera from patients with scleroderma. Recently, the gene encoding the major human centromeric autoantigen has been cloned (Earnshaw et al, 1987). Dicentric chromosomes have two centromeres and Fantes et al (1989) have been developing methods to assay for such chromosomes following radiation damage using immunofluorescence flow cytometric methods with sera from patients with proven autoimmune disease.

FLOW KARYOTYPE ANALYSIS

299

13.3.6 In situ hybridization All the chromosome staining techniques described above are assaying small, or relatively small, differences between chromosomes to effect a discrimination and very high precision instrumentation, preparation and staining is required. In spite of the sophistication developed over the last 15 years the 9—12 group still defies unique identification. However, all chromosomes have unique DNA sequences which could be identified with in situ hybridization using cDNA probes. Considerable work has been carried out in this area with chromosomes and nuclei attached to slides and Nederlof et al. (1989) have been able to perform triple colour fluorescence in situ hybridization to detect multiple nucleic acid sequences. However, these methods are not yet applicable to chromosomes for flow analysis though Trask et al. (1985) have performed in situ hybridization in whole nuclei in suspension. In situ hybridization techniques for flow analysis are going to be difficult to perfect, but I believe they will be perfected in the not too far distant future and judging by the speed of events in this field it would be a brave (or foolhardy) person who believes otherwise.

13.4

Flow karyotype analysis

Analysis of flow karyotypes involves calculating the relative area under each peak in univariate histograms and the relative volumes of the peaks in bivariate data sets. For example the normal flow karyotype should have 4.348% (^-) of the whole distribution under each unique autosomal peak, the same value under the X chromosome peak in the female but 2.174% under both the X and Y chromosomes in the male. The composite 9—12 peak should comprise 17.392% of the whole distribution. 13A.I

Univariate Analysis of monodimensional histograms of the types shown in figure 6.6 and 13.3 is carried out by fitting multiple Gaussian distributions to the various peaks in the histogram. The equation used in the fitting procedure (Gray and Langlois, 1986) is shown below. g(x)=Yj(0.3989xAi/(ji)

x exp(-0.5 x { ( j - /

i

This may look a little fearsome to the majority of biologists but really, it's fairly simple. However, if you do find it unintelligible it's only because you are not familiar with the nomenclature, and I'll try to explain. If you insist that you have a maths phobia and can't face this you had better skip the next section as well and go on to slit-scanning. If you know all about these things you shouldn't be reading either this or the next section so you too should go on to slit-scanning. So, you're going to give it a try, take a deep breath. There is a total of i separate distributions in the experimental data, each one corresponding to a particular

300

CHROMOSOMES

chromosome. If all chromosomes were uniquely resolved i would equal 23 in the female and 24 in the male. In practice, due to overlapping, e.g. 9—12, the value of / is less than the theoretical maxima. Aif ot and ft are the area, standard deviation and mean respectively of the /th peak. The expression exp{ — 0.5 x [(x — ft)/ad 2} is the curve which describes the normal distribution and 0.3989 x A-JCi is a constant which normalizes the /th distribution to unity with A{ being the scaling factor for the /th peak. The numerical value 0.3989 is equal to 1/^/(2 x TC). The x is the channel number on the X-axis (fluorescence intensity) of the whole distribution and the symbol Z with its / subscript represents the summation of the contributions of all the / theoretical distributions to that particular channel x. This is denoted by g(x). The expression d(x) represents a continuum which is added to the whole of the theoretical function to approximate any background noise underlying the peaks in the experimental data. The value of / can usually be fixed by inspection of the experimental data and the / values for A, a and \i are varied until the best fit between the theoretical data, g{x), and its corresponding experimental data, f(x), in each channel is found. 13.4.2 Bivariate Analysis of bivariate data in principle is no more complicated than univariate. The mathematical notation is just extended to the two dimensions simultaneously assuming that Gaussian distributions can be fitted in both dimensions and the notation is as follows,

C{x,y)= £ where N is the number of peaks, C(x,y) is the theoretical frequency at the location (x,y) and where the function Gi(x,y) is dependent on the correlation coefficient, p, between x and y, and the means and standard deviations in both the x and the y direction and the volume under each peak. I'm not giving the full expansion of this function here, if you want it you will find it in Cytometry by Dean, Kolla and van Dilla (1989) but it is interesting to note that it was used during the last war. The probability of hitting an aeroplane with a shell is Gaussian distributed in both the x and y directions and a 'correlation coefficient' dependent on the angle of the aeroplane's vector with respect to the position of the gun. The above mathematical function was used to calculate the number of shells required to hit the target at a given height and velocity. It turned out that the theory predicted far fewer shells to obtain a hit than was observed in practice. Then some bright spark realized that this was a three-dimensional problem as the altitude of the plane could not be calculated exactly and the function was extended to three dimensions with the Z-axis (altitude) being added. Theory now predicted that more shells would be required but it was still an underestimate of the reality as just hitting the plane was not always sufficient to bring it down. Some hits were 'non-lethal'.

SLIT-SCANNING

13.5

301

Slit-scanning

The principles involved in slit-scanning were described in chapter 8 and figure 8.2 showed an example of chromosome slit-scanning from Lucas and Gray (1987). 13.5.1 Centromeric indices The centromeric index is the relative position of the centromere along the length of a given chromosome and flow cytometric calculation of this parameter was introduced by Lucas et a\. (1983). Inspection of the slit-scans in figure 8.2 reveals that the centromeric index can be calculated with confidence for only four of the profiles, namely those labelled 2, 4, 5 and 8. In these studies total DNA was measured by propidium iodide fluorescence from human chromosomes after isolation with the hypotonic PI technique and these were then slit-scanned in the object plane with a 1.3 |im 488 nm argon beam. The fluorescence intensity profile

0.90

0.80

_ Y,21,22 -

x


0.70 -

E o
o

0.60 -

0.50 -

2

3

4

DNA fluorescence Figure 13.8. Slit-scan centromeric index (ordinate) versus total DNA (abscissa). The contours of the smaller chromosomes to the left of the dashed line were drawn at different levels compared with those on the right. The squares represent the centromeric indices derived from positively identified chromosomes in human/rodent hybrid cell lines.

302

CHROMOSOMES

of each chromosome was digitized at 20 ns intervals with a model 8100 Biomation waveform recorder as it passed through the exciting light. The data were transferred to computer and the centromeric indices were calculated and plotted on the ordinate versus total DNA content on the abscissa as shown in figure 13.8, which again is redrawn from Lucas and Gray (1987). The peaks from the smaller chromosomes to the left of the dashed line were drawn with different contouring levels than those on the right. The squares within the contours were the data from positively identified human chromosomes in human/rodent hybrid cell lines. The feature of these data is the encouraging separation of the 9-12 group which is not resolved with PI alone or with the Hoechst/chromomycin dual-staining technique. Moreover, the distinct possibility exists that even better resolution will be obtained with the dual-staining method in conjunction with slit-scan centromeric index determination. 13.5.2 Banding Similar types of techniques are being developed at the Los Alamos laboratories to identify human chromosomes by banding patterns (Bartholdi et al., 1989) using a cytometer to record simultaneous waveforms (Johnston et al., 1985). This uses image plane slit-scanning, as opposed to object plane, but it also uses a model 8100 Biomation waveform recorder digitizing at 20 ns intervals. Human chromosomes were prepared from metaphase cells treated in Ohnuki's buffer then released using the polyamine method and stained with chromomycin A3. This stain produces low-contrast R-banding (Sahar and Latt, 1978; Schweitzer, 1976, 1981; van de Sande et al, 1977) similar to that of acridine orange (ISCN, 1985). A typical chromomycin flow karyotype from Bartholdi et al. (1989) was shown in figure 13.2 in which eight major peaks could be resolved on total chromomycin fluorescence. Chromosomes 1-3 were each identified as a single peak and their slitscan profiles were relatively easy to identify and correlated well with the acridine orange R-banding. None of the remaining chromosomes could be resolved uniquely on total chromomycin fluorescence and no clear waveform profiles could be identified which corresponded to R-banding of the smaller chromosomes, 13—22. However, profiles were found in the 4—7 and X,8—12 peaks which corresponded reasonably well with the R-banding patterns of these chromosomes. At present it is not possible to sort chromosomes for positive identification on either centromeric index value or banding pattern. Centromeric index calculation is the simpler of the two to express as a single number and requires between 50 and 100 ms computing time (Lucas and Gray, 1987). However, sorting would require this to be reduced by two orders of magnitude and the current generation of computers, fast as they are, are simply no match for this problem which is further compounded for banding pattern analysis. In the latter there are multiple peaks and troughs in the waveform profile as opposed to just one for centromeric index calculation. It is not yet clear how the data contained in the waveform profile are going to be reduced to the information of a unique single number which would be required in the current generation of sort decision electronics.

APPLICATIONS

$03

13.6 High-speed sorting No chapter on flow cytometric chromosome analysis can fail to mention the high-speed sorter developed initially at Lawrence Livermore (Peters et al, 1985; Gray et al, 1987) which is a phenomenal example of modern bio-technological instrumentation. Attempts are beind made to clone, map and understand the whole human genome which is composed of about 109 DNA bases encoding 0.5-1.0 X 106 genes. This 'human genome project' (Caskey, 1986) is a simply monumental task which will be facilitated considerably by the availability of DNA purified from individual chromosomes. Chromosome sorting can be carried out with ease using conventional instruments for identification purposes, but the time taken to obtain workable microgram quantities of sorted DNA is prohibitive. As an example let us take the Y chromosome and suppose that 1 jig of DNA is needed. This chromosome contains approximately 10 ~ 13 g of DNA and the total throughput rate of a standard sorter is about 1000 chromosomes s~ l for the highresolution analysis that is needed. However, only about 2% of the total chromosome complement is Y, thus 20 chromosomes will be sorted per second which is 2 pg and we need 1 |ig. It's not too difficult to appreciate that this would take half a million seconds which is about 140 hours and this assumes that everything (instrument, chromosome preparation and all the rest of the paraphernalia) is working perfectly all the time. The LLNL-HiSS (LLNL stands for Lawrence Livermore National Laboratory and HiSS stands for high-speed sorter) was developed to reduce sorting times by about one order of magnitude. It operates with a jet velocity of 50 m s ~ l compared with 1 0 m s " 1 in regular instruments, a sample pressure of 200 psi compared with 15-20 psi and a droplet production frequency of 215 000 per second compared with 30 000—40 000 whilst maintaining the analytical resolution required, an amazing achievement.

13.7 Applications The uses of flow cytogenetics are being developed for diagnosis, gene mapping, cDNA library production and radiation dosimetry. Arguably, the most exciting developments are in the areas where there is an absolute requirement for flow cytometry, namely the generation of chromosome specific libraries.

13.7.1 Diagnosis Many diseases are genetic in origin. Perhaps the best understood examples are the haemoglobinopathies thalassemia (Orkin et al, 1979; Little et al., 1980) and sickle cell anaemia (Orkin et al, 1978) where restriction site polymorphism analyses of the globin loci (Jefferys, 1979) have formed the basis for diagnosis of variants. Further well recognized genetic anomalies with gross chromosomal abnormalities include duplications, deletions and translocations. Kleinfelter's syndrome (XXY), the XYY mosaic associated with psychopathic

304

CHROMOSOMES

Figure 13.9. Isometric hidden line elimination plots of hoechst versus chromomycinfluorescenceversus frequency for a normal female, panel A, and a female with trisomy 21 (Downs syndrome), panel B. Note the increased volume of peak 21 in panel B which is clearly apparent in spite of the somewhat higher background which partly underlies peak 21 in panel A. disorders and Patau's, Edward's and Downs syndromes, respectively associated with trisomies of 13, 18 and 21 are all examples of chromosomal duplications. Turner's syndrome (X0) represents a Y deletion and Burkitt's lymphoma and chronic myeloid leukaemia are associated with t(8:14) and t(9:22) translocations respectively. Conventional chromosome banding analysis is carried out in the diagnosis of suspected cases and this is being extended to small amniotic samples using flow cytometric methods (Gray et al, 1988). An example from Gray and Langlois (1986) is given in figure 13.9 which shows trisomy-21. Techniques such as these can be readily adapted to genetic anomaly screening programs where the specific abnormality is known. However, it is also important to determine the degree of variation within the normal population. An example of a normal variation was shown in figure 13.3 where there are three peaks, namely 1A, IB and 2, on the far right. Usually, chromosomes 1 and 2 occupy the same peak with this particular staining method but this idividual exhibited heteromorphism of the centric heterochromation of chromosome 1 which was confirmed by G- and C-banding (Young et al, 1981).

13.7.2 Genomic libraries Until 1981 most chromosome-specific genomic libraries had been produced from human/rodent hybrid cell lines (Gusella et al, 1980; Olsen, McBride and Otey, 1980; Schmeckpeper et al, 1979). In 1981 Davies et al produced the first library from a flow cytometry sorted chromosome and they derived 50 000 recombinants representative of the human X chromosome and it

APPLICATIONS

305

was their data which were used as the illustration in figure 6.6. This was followed by similar work from Krumlauf, Jeanpierre and Young (1982) who produced libraries from chromosomes 21 and 22. In the next two years libraries were constructed from chromosomes 1 (Kanda et al, 1983), 13 (Lalande et al, 1984a) and X (Lalande et al, 1984b) then the flood gates opened. By the end of 1986 Scambler et al. (1986) and Donlon et al. (1986) had produced libraries derived from chromosomes 7 and 15 respectively and the National Laboratory Gene Library Project of the USA, a joint venture by Lawrence Livermore and Los Alamos Laboratories had produced libraries from all 24 human chromosomes (Deaven et al, 1986; van Dilla et al, 1986; Fuscoe, Clark and van Dilla, 1986).

13.7.3 Gene mapping Unique sequence DNA in the form of a cDNA probe with incorporated radioactivity or a fluorescent tag can be hybridized to chromosomes to locate not only the chromosome containing that sequence but also the hybridization site on the chromosome. Hence, a probe (gene) map in relation to chromosome bands can be built up. In the study by Krumlauf et al. (1982), in which sorted chromosomes were used to generate probes, it was found that 90% of the single-copy clones derived from chromosome 22 hybridized to chromosome 22 and 10% to chromosome 21. In view of the proximity of chromosomes 21 and 22 on the histogram (see figures 13.3 and 6.6) and the early stage in the development of the staining technique used to sort these chromosomes (Sillar and Young, 1981) this must have been a gratifying result. One of the first reports of flow cytometry being used to assist in gene mapping is due to Lebo et al (1979) where the p-, y- and 5globin genes were assigned to the short arm of chromosome 11 by sorting coupled with restriction analysis. Lebo et al. (1984) have mapped McArdle's syndrome, a rare genetic defect resulting in myopathy due to enzymic deficiency of muscle glycogen breakdown, to chromosome 11 and the insulin gene to the short arm of this chromosome (Lebo et al, 1982). A very direct and interesting approach to gene mapping is afforded by cell lines containing chromosomal break points and translocation in the vicinity of the gene being mapped (Bernheim et al, 1982; Lebo et al, 1984; Collard et al, 1985; Harris et al, 1986). This operates as follows. The c-myb oncogene had been localized to chromosome 6 between bands q22 and q24 (6q22-q24) by in situ hybridization (Harper et al, 1983). Collard et al (1985) then used a primary teratocarcinoma cell line in which there were translocations involving chromosomes 6 and 11. There were two break points on the number 6 chromosomes, at 6q21 on the first of these and at 6q23 on the second and one break point on one of the number 11 chromosomes, at I l q l 3 . The shorter chromosome 6 fragment (6q23—qterm) was translocated to the chromosome 11 break point. The longer chromosome 6 fragment (6q21—qterm) was translocated to the 6q23 break point of the second number 6 chromosome which now had 6q21 to 6q23 duplicated. Finally, the (ql3-qterm) fragment of chromosome 11 was translocated to the 6q21 break point site of the first number 6 chromosome. A diagramatic summary of these

306

CHROMOSOMES Chromosome 6

Chromosome 11

q13"

q21

q23-

Figure 13.10. Chromosome banding analysis of translocations in chromosomes 6 and 11 of a teratocarcinoma cell line. The top panel shows the normal chromosomes with the break points in the teratocarcinoma cell lines indicated by the arrows. Both chromosomes 6 were cleaved at q21 and q23 respectively and one chromosome 11 was cleaved at ql3. The middle panel shows the rearrangements with the 6q21—qterm fragment being translocated to the q23 break point on the second chromosome 6. The 6q23—qterm fragment from the latter translocated to the ql3 break point of chromosome 11 and the

APPLICATIONS

307

13-16+11/6

9-12

o 4) 0)

21

DNA Fluorescence Figure 13.11. Flow karyotype from the teratocarcinoma where both of the chromosome 6 translocations show a net gain in DNA. The tit translocation appears with chromosome 3, the 6/11 translocation with chromosomes 4 and 5 and the 11/6 translocation, which has lost DNA, has moved out of the 9-12 peak and appears in the 13—It group.

Figure 13.10 (cont) Ilql3-qterm fragment from 11 translocated to the q21 break point of the first number 6 chromosome. These translocations, 6/11, tit, 11/1 are depicted in the bottom panel together with the normal chromosome 11. The identical repeat segments of the tit translocation are indicated by the arrows. This figure was abstracted from the work of Collard et al.r 1985.

308

CHROMOSOMES

translocations is shown in figure 13.10. Hence, if c-myb was located in the (6q23—qtrem) fragment one copy of the gene would appear on chromosome 11, the 11/6 translocation, and the second copy would appear on the second chromosome 6, the 6/6 translocation. If, however, c-myb was located between 6q21 and 6q23 the gene would be deleted from the first number 6 chromosome and both copies would appear on the second number 6 and it would not be present on 11. The flow karyotype of this cell line which is shown in figure 13.11 revealed a reduced area under the 9-12 group as the 11/6 translocation, had considerably less DNA (see figure 13.11) and this now appeared in the region of the 13-16 chromosome group. Similarly, both chromosomes 6 had more DNA than normal. The 6/6 translocation appeared in the same peak as chromosome 3 and the 6/11 appeared with chromosomes 4 and 5. A total of 60 000 chromosomes from each of these peaks were sorted directly onto nitrocellulose filters and were hybridized with the c-myb probe and only chromosomes from the 3 + 6/6 peak gave a positive signal. Thus the gene had to be located in the q21—q23 repeated region of chromosome 6/6. The previous study had mapped this gene to (6q22—q24) and combination of these two results assigns c-myb to 6q22—q23.

13.7.4 Radiation bio-dosimetry Radiation induces a large number of chromosomal anomalies which are dose dependent. However, not all cells will exhibit detectable changes, which applies particularly at low doses and, furthermore, the changes in affected cells will not be uniform. Flow karyotype analysis is an attractive approach to detecting low-dose environmental radiation exposure as large numbers of chromosomes can be analysed rapidly but even with flow we run into the problems of rare-event analysis. One approach is to assess radiation-induced alterations in the karyotype which include broadening of the peaks and an increase in the underlying background continuum (Otto and Oldiges, 1980; Nusse and Kramer, 1984; Welleweerd et al, 1984). However, these types of end-point can be difficult to assess quantitatively. A second approach is the detection of specific chromosome anomalies, e.g. dicentrics, the frequency of which is dose dependent. Two methods are currently being evaluated for dicentric frequency assessment, namely the use of centromeric indices (Lucas and Gray, 1987), and immunocytochemical assessment of centromeric protein content which should be doubled in dicentric chromosomes (Fantes et al, 1989).

14 Dynamic cellular events

Time-dependent biological events monitored by flow cytometry have been studied almost since the technology was invented and examples of DNA histogram changes in populations of cells released from a synchronizing event giving cell cycle kinetic data were described in section 11.5.3. In these types of study the time-dependent biological reference frame is long and sampling the population at intervals of hours gives the required resolution and information. In this chapter the use of flow technology to study dynamic events within the temporal reference frame of a few seconds to minutes will be considered. The majority of such studies have been conducted in the general area of enzyme reaction kinetics, termed flow cytoenzymology (FCE) (Dolbeare and Smith, 1979; Watson, 1980a,b; Dolbeare, 1981; Watson, 1984), where the aim is to determine both the quantities and activities of enzymes in populations of single intact cells. However, similar techniques can also be used to determine rates of influx or efflux of molecules across the cell membrane, binding of ligands to intracellular constituents, and changes in intracellular calcium, pH and membrane potential in response to various stimuli. Total quantities of enzymes can be determined using antibodies (Coons et al., 1941) or fluorescently labelled tight-binding inhibitors (Gapski et al, 1975) and activities can be determined using various fluorogenic and non-fluorogenic substrates which yield fluorescent and light-absorbing products respectively. With increasing time after mixing cells with substrate greater quantities of either fluorescent or light-absorbing product are released within or on the cells of the population, and these can be monitored quantitatively with time. Flow techniques have important advantages over conventional biochemical and histochemical assays of enzyme activities, particularly as cells can be assayed under near physiological conditions. Artefactual results may be generated by the disruption of cellular and subcellular permeability barriers in the preparation of the cell-free extracts for conventional biochemical methods (Youdim and Woods, 1975). Quantitative data can be obtained with conventional methods using bulk samples, but heterogeneity within samples is completely masked by average values. Useful qualitative information regarding heterogeneous cell populations can be obtained by traditional histochemistry; however, the capacity for quantitative accuracy is extremely limited. Furthermore, histochemical procedures

310

DYNAMIC CELLULAR EVENTS

are slow and certainly not ideal for collection of detailed dynamic biochemical data. With flow technology the enzyme activity of each individual cell in a sample is recorded by the cytometer. This results in the very important capacity to obtain distributional information for identification and quantitation of cell subpopulations differing in enzyme activities in a heterogeneous cell sample, including minority populations. The techniques are rapid, highly sensitive and good statistical precision is obtained. Another important advantage to flow cytometry over conventional methods is the capacity for multi-parametric analysis. Potential exists for measurement of multiple enzyme activities, multiple substrates (Watson, 1980a; Malin-Berdel and Valet, 1980), and enzyme activities simultaneously with other biologically and pharmacologically relevant cellular parameters such as DNA content, glutathione and oncoproteins.

14.1

Incorporation of time

Clearly, in order to measure the dynamics of reactions it is necessary to include time in the data base, and we will start with this, the most important single parameter for such studies and there are four methods available. 14.1.1 Discontinuous sequential sampling This is appropriate whenever the dynamic event under study is varying within a time frame of a few tens of minutes to hours. The population is sampled discretely at defined intervals after the initiating event and the measurements are made at those defined time intervals. Examples include mixing cells with substrate for enzyme kinetics of 'slow7 reactions and DNA histogram analyses after a synchronizing event to give cell cycle data. 14.1.2 Continuous interrupted sampling With this technique cells are mixed with substrate and the sample is introduced into the instrument as fast as possible. 'As fast as possible' in this context usually means a 'dead' time of about 20—30 seconds but if everything goes very well this can be reduced to a minimum of 15 seconds. The sample runs through the instrument continuously and the data acquisition computer is instructed to record for a defined interval, say 5 seconds, then to wait for a further defined interval, say 10 seconds, before recommencing data acquisition. The median of the distribution obtained within each 5 second recording interval can be printed out during the period when data are not being collected. The median values are then plotted against time to generate a reaction progress curve. This technique was introduced in flow cytometry over a decade ago to study the hydrolysis of fluorescein diacetate (FDA) by EMT6 cells (Watson et al., 1977) and some of these data are reproduced in figure 14.1.

INCORPORATION OF TIME

311

60-1

LU

u z

096

UJ

o UJ

O

•—-•

048

ID

°'24 TIME, MINUTES Figure 14.1. Hydrolysis of FDA (mM concentrations shown against each curve) by EMT6 mouse mammary tumour cells. (Watson et al., 1977).

14.1.3 Continuous time recording The continuous interrupted sampling technique introduced by Watson et al. (1977, 197S) had one major deficiency. The fluorescence from substrate hydrolysis was increasing during the recording interval; hence the distributions obtained during the recording interval were skewed to the right and the median values were overestimated. This was not a problem for 'slower7 reactions but it could be a severe limitation for 'fast' reactions. This was overcome very elegantly by Martin and Swartzendruber (1980) by incorporating time directly into the data base. Their initial method used a voltage ramp generator where the potential increased linearly with time and a recording of the voltage was made from the ramp generator each time a cell was analysed in the flow chamber. This method had one disadvantage, namely that time could only be recorded over an interval of about 15 minutes. However, immediately following the introduction of continuous time recording by Martin and Swarzendruber (1980) we incorporated the time stamp from the data acquisition computer directly into the data base. Hence, events were timed automatically with 50 ms resolution by the computer clock and this time, in the form of the number of clock 'ticks' (one tick per 50 ms), was recorded for each event analysed. Thus, unlimited time could now be recorded and an example of mouse bone marrow where cells were incubated with FDA is shown in figure 14.2 and four subsets labelled A, B, C and D can be defined.

312

DYNAMIC CELLULAR EVENTS

Time—-• Figure 14.2. Contour plots of fluorescence from fluorescein diacetate hydrolysis versus computer clock time for mouse bone marrow where four subsets, labelled A, B, C and D are apparent.

14.1.4 'Stop—flow' cytometry in flow Each of the various methods described above suffers from a considerable disadvantage, which is the finite time taken between mixing cells with substrate or ligand and recording the first event. This 'dead time' is due to mixing, location of the containing vessel in the instrument, surge pumping the reaction mixture through the sample feed tube into the flow chamber, and restoration of stable flow. In our instrument the absolute minimum dead time is 15 seconds if everything proceeds smoothly, but more often this was about 20 seconds. There was a clear need to reduce this temporal delay, so as to facilitate the analysis of biochemical events taking place over the time period of a few seconds. Such a development would be valuable not only in flow cytoenzymology (e.g. for substrate diffusion kinetics, very rapid reaction and membrane enzyme analysis), but also for a variety of other biological applications involving very short time scales (e.g. drug uptake and ion flux analysis). Kachal, Glossner and Schneider (1982) described a flow chamber in which timeresolved measurements could be made within 1 second but this was not variable. Recently, we have developed a system which markedly reduces the temporal reference frame so that this can be varied within the interval 1 to 20 seconds (Watson et ah, 1988a, 1991a). The technique employs computer-controlled precision drive syringe pumps, one for substrate the other for cells, together with variable length tubing between a mixing chamber and analysis point, selected by an array of zero dead-space pinch valves. The different tube lengths at a given

INCORPORATION OF TIME

313

FCN

Figure 14.3. Schematic of 'stop-flow' device which allows kinetic measurements to be made during the first few seconds of a dynamic reaction. Pi and P2 = pumps 1 and 2; J = 4-way junction; PVA = pinch value assembly; FCN = flow chamber needle. pump flow rate give different times between mixing and analysis. A schematic of the system is shown in figure 14.3. Calibration was effected using two sets of microbeads with different fluorescence intensities (Polysciences Inc. Warrington, PA, USA). Pumps 1 and 2 were filled with beads of the higher and lower intensity respectively. The concentrations were adjusted to give a flow rate of 250 beads per second with both pumps running at 100 |il min" 1 . The concentration of beads in pump 2 (lower intensity) was about 1.4 times greater than in pump 1. Both pumps were activated before data collection in order to fill the input pipes to the mixing chamber, and they were then stopped. The chamber was then flushed through and pump 1 was restarted before data collection. Pump 2, containing the lower intensity beads, was started during the run after about 1500 of the 10 000 requested events had been recorded. This procedure was repeated for each tube length at various flow rates. Figure 14.4 shows fluorescence versus time data as frequency contour plots at the 3, 6 and 12 event levels with the fluorescence and flow rate histograms adjacent to their respective axes. The saw-tooth pattern of the flow rate histogram is caused by interruption of data collection due to time sharing on the PDP 11/40 computer which dumps the data to disk, displaying data on the screen between each buffer dump and checking the pump flow rates before recommencing data collection. These data were recorded with flow rates of 100 ul min" 1 from each pump and

314

DYNAMIC CELLULAR EVENTS

TIME Figure 14.4. Fluorescence versus time data as frequency contour plots at the 3, 6 and 12 event levels with the fluorescence and flow rate histograms adjacent to their respective axes.

INCORPORATION OF TIME

-30

-20

-10

0

315

20 5 10 Tube Length, cm

Figure 14.5. Plots of time from mixing chamber to analysis point versus tube length for pump flow rates of 100, 150, 200 and 250 ul min" 1 .

panels A, B, C and D respectively were obtained for tube lengths of 5, 10, 20 and 40 cm. Full scale on the abscissa is 40 seconds, and each division on the ordinate represents 100 digitization steps. The long vertical lines drawn through each panel represent the time which is flagged in the data base when pump 2 was activated. The immediate flow rate increase is apparent and the sudden surge is also manifest by a widening of the distribution of the higher intensity beads. The short vertical lines show the time at which the beads from pump 2 first appeared in the data record. The interval between these lines represents the time for the lower intensity beads from pump 2 to travel from the mixing chamber to the analysis point. Plots of time from mixing chamber to analysis point versus tube length are shown in figure 14.5 for pump flow rates of 100, 150, 200 and 250 JLXI min~ \ The regression lines for these pump flow rates extrapolated almost to a common point and all the intersections are included in the region defined by the open circle shown in figure 14.5. Therefore, the four equations defined by the regression analyses were solved simultaneously to give a common intersection point with X and Y coordinates of —24.73 cm and 0.6 s respectively. This value of 24.73 cm represents the volume of the 'non-variable' (dead-space) section of the system which comprises an individual pair of arms of the four-way junctions, the flow chamber injection needle, the mixing chamber and its outflow. The volume of unit length of the tubing contained in the variable section was measured using

316

DYNAMIC CELLULAR EVENTS

0.25

\L°9e

SORT

LJJ O _l

0.15

CO

0.05-

4.6

5.0

5.5

Flow rate

10 SQRT Flow rate

15

Figure 14.6. The slopes of the four regression lines plotted against both the square root and the log of the respective flow rates. Regression analyses gave correlation coefficients of 0.9999 and 0.991 respectively. Hamilton syringe injection of eight different lengths. The average of the eight readings showed that 1.0 cm was equivalent to 1.046 pi. The volume of the dead-space, excluding the mixing chamber (18.1 pi), was measured (again by Hamilton syringe injection) to be 8.1 pi. Thus the total dead-space volume was 26.2 pi, which corresponds well with the theoretical value of 25.8 pi (24.7 cm X 1.046 pi cm" 1 ) represented by the intersection point in figure 14.5. The vertical displacement of the intersection point, 0.6 s, constituted a puzzle until we discovered that the assembler language routines controlling the pump flow rates, which had been developed for a different application, not only test, but also correct for, any 'backlash' before commencing delivery. This involves a feedback hysteresis loop to the computer and the time taken for backlash correction is inversely proportional to the requested flow rate. At 100 pi min" 1 the backlash

ENZYME KINETICS

317

assessment and correction time is between 500 and 600 ms, which accounts for the observed vertical displacement. This minor problem was only encountered when the second pump was started during data collection when backlash had to be tested for and corrected. Figure 14.6 shows the slopes of the four regression lines plotted against both the square root and the log of the respective flow rates. Regression analyses gave correlation coefficients of 0.9999 and 0.991 respectively. From this we assumed that the square root transformation was more appropriate, which gave a slope of — 0.0283 and intercept of 0.528. From these various data it was possible to express time in relation to flow rate and length of tubing between mixing and analysis points. This is given by the following equation, T= ((0.528 - (0.0283 x SQRT(F'))) x (L + 24.73)) + 0.6 where T is time, F is the average flow rate of the two pumps and L is tube length. The expression 0.528 —(0.0283 x SQRT(F)), derived from figure 14.6, is equivalent to the slope, m, in the equation Y= mX+ C, and L + 24.73 is equivalent to X, derived from figure 14.5. The constant C, 0.6 s, can be dropped when both pumps are started before data collection.

14.2

Enzyme kinetics

The first non-kinetic description of enzyme measurements by flow cytometry (Hulett et ah, 1969) exploited the phenomenon of fluorochromasia (Rotman and Papermaster, 1966) in which the lipophilic, membrane-permeable fluorogenic substrate fluorescein diacetate (FDA) was converted to the comparatively polar fluorescent product fluorescein by intracellular esterases. Since then there have been considerable advances which will continue with the further developments of probes.

14.2.1 Substrates A schematic for the conversion of substrate to product by enzyme action in whole cells is shown in figure 14.7. Substrate must first cross the external membrane, then interact with the enzyme generating the reaction product which can then be measured if it remains inside or associated with the cell. Unlike classical biochemical assays only product which is cell-associated is 'seen' by the instrument. The ideal substrate for flow enzymology is one which is lipophilic, non-toxic and either colourless or non-fluorescent. During the reaction this is converted to a highly polar product which is either light absorbing (coloured) or fluorescent. Substrate lipophilicity generally is required for rapid transport across the external cell membrane and high polarity of the product is desirable in order to 'trap' the product within the cell. However, no substrate meets all these criteria and a number of 'tricks' have been employed to simulate the behaviour of the 'ideal' substrate. Many 'classical' histochemical enzyme-staining procedures deposit insoluble

318

DYNAMIC CELLULAR EVENTS INTRACELLULAR

ENZYMES

Figure 14.7. Schematic for the conversion of substrate (Se) to product (P) by enzyme action in whole cells. Substrate must first cross the external membrane, then interact with the enzyme generating the product which can then be measured if it remains inside or associated with the cell. light-absorbing products which enables cells with the enzyme activity being studied to be identified. Some of these techniques have been adapted to flow technology. Peroxidase activity in white blood cells (Kaplow and Eisenberg, 1975) has been assayed based on supravital benzidine dihydrochloride staining where zinc chloride was used as a stabilizer for the blue reaction product (Kaplow, 1975). oc-Naphthol acetate has been used as a substrate for esterases (Kaplow, Dauber and Lerner, 1976) where the reaction product was coupled to fast blue salt BB to give a final product which was grey—black. The methods developed by Kaplow and collaborators are excellent for cellular identification in the systems for which the techniques were developed, but relatively 'static' patterns were obtained. Furthermore, these light-absorbing methods are relatively insensitive and not suitable for 'real-time' studies. This is particularly true where enzyme progress curves are required in order to obtain the kinetic parameters of enzyme action, and fluorogenic substrates are more suited for these types of studies. There is now a considerable variety of such substrates based on a number of different fluorophores. These include fluorescein, rhodamine, naphthol, naphthylamine, quinolines, methylumbelliferone and monochlorobimane for studying esterases, lipases, sulphatases, phosphatases, glucuronidases,

ENZYME KINETICS

319

transferases, peroxidases, peptidases, transpeptidases, galactosidases, arylmidases, glucosidases and glutathione-S-transferases to mention just a few. In general, the fluorescein and rhodamine based substrates and monochlorobimane are either weakly or non-fluorescent which is ideal. In contrast, the 4-methylumbelliferone conjugates are generally fluorescent but both their excitation and emission wavelengths are much shorter than those of the released product 4-methylumbelliferone and with correct optical design any 'overlap' can be minimized or eliminated. Techniques using the naphthol derivatives involve 'trapping' the released product within the cell by coupling with 5-nitrosalicylaldehyde which forms an insoluble fluorescent complex with an emission spectrum that is shifted into the red (Dolbeare and Smith, 1977). A comprehensive list of substrates and the reactions to release the associated fluorophores is not in order here and the reader is referred to the Molecular Probes Catalogue (Molecular Probes Inc. Eugene, Oregon, USA) and an excellent volume Applications of Fluorescence in the Biomedical Sciences (Lansing-Taylor et ah, 1986).

14.2.2 Light absorption quantitation Most light-absorbing reaction products are blue, blue—grey or black and a helium—neon (HeNe) laser emitting red light is ideal for such assays for two

CELL STREAM, SINGLE FILE PHOTOSENSORS

Figure 14.8. Absorbed light quantitation with two sets of light-sensitive detectors.

320

DYNAMIC CELLULAR EVENTS

reasons. Firstly, red light is well absorbed by the products and secondly, HeNe lasers are very stable. The absorbed light can be quantitated with two sets of lightsensitive detectors and a typical system is depicted in figure 14.8. Cells in flow intersect the focus and as each cell is illuminated there is a decrease in intensity on the detector A. This decrease in intensity is proportional to the cross-sectional area of the cell and is largely independent of the optical density. In contrast, the amount of light scattered into the S detector is inversely proportional to optical density. The instrument is set up initially with an unstained sample so that the amplitudes of the pulses from the S and A detectors are adjusted to be approximately equal. When a sample containing cells stained with light-absorbing material is subsequently analysed there is a decrease in the amplitude of the S detector (but not of the A) whenever a stained cell passes through the light beam. Thus, by comparison of the S-pulse with the A-pulse for each cell it is possible to define clusters which absorb light and hence to identify the stained cells in the population. 14.2.3 Fluorescence quantitation Fluorescence measurement is simpler, quicker and more sensitive than light absorption for a number of reasons. Firstly only one detector is needed for quantitation and secondly, the reaction is followed as it is taking place. Finally, there is no 'upper' detection limit in contrast to absorption methods where the lower' detection limit due to loss of signal represents greater activity. This problem can be partly overcome by having the capacity to vary the gain in the preamplifier circuits by defined quantities (Cox et a\., 1987). For absorption work the gain would be increased in the S detector and with fluorescence the gain would be reduced. However, this does have limitations. If very high fluorescnce activities are recorded during a run a neutral density filter can be placed in the light path to the detector. This attenuates the signal by a further defined quantity depending on the optical density of the filter. But no comparable manoeuvre can be applied in light absorption methods. 14.2.4 Cytoplasmic enzymes Kaplow and Eisenberg (1975) identified lymphocytes, monocytes, neutrophils and eosinophils in peripheral blood using peroxidase activity. These data are reproduced in figure 14.9, in which scattered light (ordinate) is scored against axial light loss (abscissa). Each cell in the sample is represented by a single dot on the storage oscilloscope with x and y coordinates that are respectively proportional to axial light loss and scattered light intensity. This technique has also been applied to patients with various haematological disorders and characteristic patterns of eosinophilia, lymphocytic leukaemia and chronic granulocytic were obtained (Kaplow, 1979). Monocytes contain large quantities of esterases and this enzyme activity is demonstrated in figure 14.10 which shows two samples from peripheral blood in which the monocyte clusters exhibited low (panel a) and high (panel b) esterase activity (Kaplow et al, 1976). Kaplow (1979) has also investigated the influence of incubation time on monocyte esterase activity and

ENZYME KINETICS

321

Figure 14.9. Identification of lymphocytes, monocytes, neutrophils and eosinophils in peripheral blood using peroxidase activity in which scattered light (ordinate) is scored against axial light loss (abscissa) from Kaplow and Eisenberg (1975).

Figure 14.10. Monocyte esterase activity in two samples from peripheral blood in which the monocyte clusters exhibited low (panel a) and high (panel b) esterase activity (Kaplow et al., 1976).

322

DYNAMIC CELLULAR EVENTS

has observed a progressive increase in axial light loss, but accurate quantitation of the reaction velocity was not possible. Dolbeare and Smith (1977) studied peptidases with a fluorescence technique which required product trapping. This type of technique was subsequently used by Pallavichini et al. (1977) not only to distinguish between intestinal crypt and villus cells, but also to sort these subpopulations based on leucine—amino peptidase activity. The differentiated villus cells contain considerably higher activity of this enzyme than the non-differentiated cells of the crypt. The continuous interrupted sampling technique (Watson et al., 1977) with FDA as substrate was used to show that cells in late plateau phase EMT6 cultures, in which more than 95% of the population was arrested in a Gl/G 0 state, exhibited considerably higher esterase activity than exponentially growing cells (Watson et al., 1978). The progress curves are shown in figure 14.11, with the data from the plateau phase and exponentially growing population in panels A and B respectively. The micromolar concentrations of substrate are shown against each progress curve. Figure 14.12 shows the substrate-dependent velocity plots associated with these data, which should theoretically follow the Michaelis-Menten rectangular hyperbola (Michaelis and Menten, 1913). There was a highly abnormal 'double-sigmoid' pattern for the plateau-phase cells but the exponentially growing cells exhibit less abnormal kinetic behaviour. Krisch (1971) has reported abnormal kinetic behaviour in certain esterase preparations and it has been suggested that the reaction mechanism may need two or more interacting catalytic sites. Substrate activation, where the reaction proceeds most rapidly when more than one substrate molecule is bound to a single enzyme molecule, may also be involved (Adler and Kistiakowsky, 1961). Furthermore, it is highly probable that more than one esterase is involved with the hydrolysis of FDA, which is a non-specific substrate. Guibault and Kramer (1966) have shown that FDA hydrolysis is catalysed by a number of enzymes including oc- and ychymotrypsin and lipase. However, further factors must also be considered for the abnormal kinetic behaviour shown in figure 14.12. Firstly, cellular and sub-cellular permeability barriers are likely to exist in the intact cell which could limit the availability of substrate at the enzyme site. Secondly, there could be active transport mechanisms for the substrate with intracellular accumulation. Each of these factors would result in a difference between the substrate concentration external to the cell and at the site of enzyme reaction and all of these various factors could contribute to the upward concavity (lag kinetics') seen for plateau phase cells in figure 14.12. Substituted naphthol was used as a substrate which did not require a trapping agent for the fluorescence product by Dolbeare and Phares (1979). This enabled them to produce progress curves for (3-glucuronidase and acid phosphatase. The reaction was allowed to proceed outside the flow cytometer and was stopped at intervals before flow analysis in a discontinuous assay. However, because of the long incubation time required to develop fluorescence, product diffusion became a limiting factor in these experiments. Dolbeare and Smith (1979) used a

ENZYME KINETICS

323

30-

144

TIME, Minutes Figure 14.11. Esterase activity in log and plateau phase EMT6 cultures using FDA as substrate, panels A and B respectively. The micromolar concentrations of substrate are shown against each progress curve.

DYNAMIC CELLULAR EVENTS

324

1-On

LOG

Substrate Concentration

Figure 14.12. Substrate-dependent velocity plots associated with the data in figure 14.11 which should theoretically follow the Michaelis-Menten rectangular hyperbola. combination of forward light scatter and (3-glucuronidase activity to distinguish lymphocytes, monocytes and macrophages in washings from the peritoneal cavity of the rat. These data are reproduced in figure 14.13 which shows the twodimensional histogram of frequency on the vertical axis versus light scatter and fluorescence on the two horizontal axes. Vanderlaan, Cutter and Dolbeare (1979) have shown that y-glutamyl transpeptidase activity is considerably higher in hepatocytes following carcinogen exposure than in control populations. However, Vanderlaan et al. (1979) have also shown that in mixed populations it can be difficult to identify positively the transformed cells in the absence of product trapping as the reaction product can diffuse from cells with high enzyme activity and subsequently enter those with little or no activity. However, this problem can be largely overcome by short incubation times and multiparameter analysis. Alkaline phosphatase and an arylamidase can be used as markers for normal and transformed Wi-38 cells (Dolbeare et al., 1980). These results demonstrated that the enzymes in the transformed cells differed by not only a qualitative but also a quantitative change in alkaline phosphatase, as well as a quantitative loss of arylamidase.

ENZYME KINETICS

325

Figure 14.13. Forward scatter versus (3-glucuronidase activity versus frequency which distinguished between lymphocytes, monocytes and macrophages in washings from the peritoneal cavity of the rat (Dolbeare and Smith, 1979).

Glutathione (GSH) metabolism is a critical determinant in control of cellular response to anticancer drugs and radiation. Depletion of protective GSH results in sensitization. Kosower et al. (1979) were the first to use fluorescent bimanes to study intracellular thiols and Rice et al. (1986a) using flow cytometry have shown that monochlorobimane (mClB) can act as a reporter molecule in GSH metabolism. As an extension to this we developed a sensitive multi-parametric flow cytoenzymological assay using continuous time measurements to determine reaction kinetics for conjugation of mClB with GSH in populations of intact viable cells catalysed by the enzyme family, the glutathione-S-transferases (Workman and Watson, 1987). Monochlorobimane is a non-fluorescent probe which after conjugation with GSH results in a relatively insoluble fluorescent complex excited by UV light with emission in the blue (460—510 nm). The reaction is illustrated in figure 14.14 and the reaction rate is a function of added mClB, intracellular GSH content and the activity of the transferase(s). The progress curves generated should reach an asymptope when all intracellular GSH is converted. Figure 14.15 shows the results from a FicoU—Paque lymphocyte-enriched preparation obtained

326

DYNAMIC CELLULAR EVENTS

CH 3 +

V-Glu-Cys(SH)-Gly

CH2CI

GSH

rnCIB ( non-fluorescent, penetrates cell)

GSH - S - transferase

mClB - GSH conjugate ( fluorescent, trapped in cell)

Figure 14.14. Reaction of monochlorobimane with the tripeptide glutathione mediated by transferases to produce the insoluble fluorescent conjugate.

from heparinized whole blood of a normal subject which was mixed with 120 pM mClB plus propidium iodide at a concentration of 50 |ig ml ~ l (Workman, Cox and Watson, 1988). Forward and 90° scatter signals were collected together with blue and red fluorescence. The latter is also excited by UV from the DNA of dead cells which do not exclude propidium iodide. Panel A shows forward versus 90° scatter in which three regions can be defined. Regions 1 and 2 correspond to lymphocytes and monocytes respectively. Region 3 represents debris and propidium iodide positive dead cells. Panels B and C show the accumulation of blue fluorescence with time as frequency contour plots where it can be seen that monocytes (panel C) exhibited greater activity than lymphocytes (panel B) consistent with larger quantities of GSH and/or GSH-S-transferase. A more conventional display, derived from the data in panels B and C, shows medians of the distributions at

ENZYME KINETICS

327

SCAT TEI

A

|

o

A //n

REGION 2 REGION 1

B

REGION 3

f

U-

9O° SCATTER

REGION 2 MONOCYTES

REGION 1 LYMPHOCYTES

figure 14.15. Ficoll—Paque lymphocyte-enriched preparation obtained from heparinized whole blood of a normal subject which was mixed with 120 JiM mClB plus propidium iodide at a concentration of 50 |lg m l " l (Workman et al., 1988). Forward and 90° scatter signals were collected together with blue and red fluorescence. The latter is also excited by UV from the DNA of dead cells which do not exclude propidium iodide. Panel A shows forward versus 90° scatter in which three regions can be defined. Regions 1 and 2 correspond to lymphocytes and monocytes respectively. Region 3 represents debris and propidium iodide positive dead cells. Panels B and C show the accumulation of blue fluorescence with time as frequency contour plots where it can be seen that monocytes (panel C) exhibited greater activity than lymphocytes (panel B) consistent with larger quantities of GSH and/or GSH-S-transferase. A more conventional display, derived from the data in panels B and C, shows medians of the distributions at discrete time intervals plotted against time in panel D. The dead cells and debris in region 3 exhibited no activity.

328

DYNAMIC CELLULAR EVENTS

discrete time intervals plotted against time in panel D. The dead cells and debris in region 3 exhibited no activity. The data in figure 14.15 represent a fivedimensional set as two scatter and two fluorescence measurements were recorded with the fifth parameter, time. Further flow cytoenzymological studies using single substrates have included the measurement of elevated y-glutamyl transpeptidase levels in aflatoxin-induced rat hepatoma cells with implications in toxicology (Manson et al, 1981), the detection of acid-P-galactosidase activity in human fibroblasts (Jongkind, Verkerk and Sernetz, 1986) and the determination of cell cycle phase-specific changes in cellular phosphatase and glycosidase activities in stimulated lymphocytes and cultured cell lines (Britten and Dyson, 1987).

14.2.5 Membrane enzymes An automated fluorimetric method for assaying alkaline phosphatase using 3-0-methylfluorescein phosphate was developed by Hill, Sumner and Waters (1968). An adaption of this technique has been applied to intact cells using flow cytometry by Watson, Workman and Chambers (1979). Again, the continuous interrupted method was used. The results for EMT6 cells are shown in figure 14.16. These are obviously very different from those for intracellular esterases shown in figure 14.11 and they presented a considerable interpretative problem which was not resolved until a sample was viewed under the fluorescence microscope. It was then seen that the fluorescence was concentrated in 'halos' at the external cell membrane with no fluorescence from the interior of the cells. Subsequent incubation of cells with product, 3-0-methylfluorescein, failed to demonstrate entry into intact cells. The viability of EMT6 cells decreases considerably after 3 hours incubation with protein-free phosphate-buffered saline (PBS). In a sample of cells so treated it was found that the fluorescence from 2>-omethylfluorescein was no longer located at the cell surface but was being emitted from granular structures surrounding the nucleus. This fluorescence was considerably more intense than that emitted from the cell surface, so much so that the latter could no longer be seen. These various data suggested that substrate hydrolysis in intact viable cells takes place at or within the cell by phosphatases located in the plasma membrane and that the product was lost from the immediate vicinity of the cells by diffusion and consequently was not 'seen' by the instrument. Thus a steady state would be reached in which the rate of production of fluorescent product would equal the rate of loss, giving rise to asymptotic fluorescence responses from the population at each substrate concentration. On the assumption that this hypothesis was correct, Watson et al (1979) were able to show that the progress curve at a given concentration, S, should be described by the equation, x(1.0-exp(-fcxf))

(14.1)

where P(t) is the fluorescence response at time t, P^ is the theoretical maximum asymptotic response at infinite substrate concentration, Km is the Michaelis

ENZYME KINETICS

329

Time, minutes

Figure 14.16. Progress curves for hydrolysis of 3-O-methylfluorescein phosphate. The micromolar concentrations of substrate are shown adjacent to each curve.

constant and k is the rate constant.for product loss assuming a first order kinetic process. When a steady state exists (i.e. beyond 2.5—3 minutes in figure 14.16) the term exp( — kxt) in equation 14.1 will tend to zero and the asymptotic fluorescence response will be given by the expression P^ x S/(S + Km) and will vary with substrate concentration. P^ can be eliminated from this expression by taking ratios, thus, P (t ) Pltfoo)

S

(*

_

(14.2)

where P^oo) and P2(to0) are the asymptotic fluorescence values associated with substrate concentrations SY and S2 and where R12 is the ratio of P2(^oo) to P^oo)Equation 14.2 can be rearranged to give,

330

DYNAMIC CELLULAR EVENTS

where for N substrate concentrations / varies from 1 to (w — 1) and j varies from (/ +1) to N, to give a triangular matrix for the ratios containing jN(N— 1) values. Thus, by plotting Rtj against [(1/St) — (R tj/SJ)] a line with slope Km is obtained which intersects the ordinate at unity. In three separate analyses values of 110.0 + 31 |lM, 122.3 + 31 uM and 120.5 + 24 |lM were obtained where the limits were calculated at two standard errors and in all cases the ordinate intercept did not differ from unity, p > 0.1. It can be seen from equation 14.2 that if P^ is the asymptotic fluorescence response at infinite substrate concentration, Soo, then, P

S

_oo__^o_

(I

_-

where Pn is the asymptotic response at Sn. The term Soo^Soo+Km) is unity; therefore, Pn x (Sn + Km) = P^ x Sn. By plotting Pn x (SB + Km) against Sn a line of slope PQO is obtained which intersects the origin. Thus P^ can be obtained after a value for Km is found. It was also shown previously (Watson et al, 1979) that the maximum reaction velocity, Vmax, is equal to the product of P^ X k. As equation 14.1 is an inverted exponential an approximate value of k can be found from the time taken for the curves to reach half their maximum height and a more accurate estimate can be obtained by curve fitting. The latter gave a rate constant of 1.98 min~ \ thus Vmax can be defined in the arbitrary units of channels per minute. The three analyses gave maximum reaction velocities of 107.0 110.0 and 109.2 channels per minute. The progress curves in figures 14.11 and 14.16 are very different and two major differences between the fluorogenic substrates should be considered. Firstly, 3-omethylfluorescein, the reaction product of 3-o-methylfluorescein phosphate, is considerably less polar than fluorescein, the reaction product of FDA. Thus, if 3-omethylfluorescein enters the cell we would expect the rate constant for leakage from the cell to be greater than that for fluorescein. Previous studies indicated that fluorescein leaks out of EMT6 cells with a half-time of 7-8 minutes (Watson et al, 1979; Watson, 1980a), thus the leakage rate constant is about 20 times greater than that for fluorescein. Secondly, fluorescein diacetate is lipophilic and will penetrate the cell without prior hydrolysis by any membrane esterases. The direct observations made with the fluorescence microscope suggest that the product of 3-0-methylfluorescein phosphate hydrolysis either does not enter the cell or it diffuses out of the cell very rapidly although some remains associated with the external membrane. Both possibilities are compatible with the magnitude of the loss rate constant whatever the mechanism. These results are interesting biologically as they suggest that dephosphorylation at the cell membrane does not necessarily lead to appreciable uptake or accumulation in the cytoplasm.

14.2.6 Dual substrate analysis The substrate-dependent initial velocity plots shown in figure 14.12 suggested the possibility that two or more enzymes were hydrolyzing fluorescein diacetate in EMT6 cells. As this is a non-specific esterase substrate it was not

ENZYME KINETICS

331

unreasonable to expect that a second non-specific esterase substrate with a different molecular structure might show different reaction rate characteristics. Furthermore, it was also not unreasonable to postulate that two different cell types would contain either different esterases or if the same enzymes were present they were unlikely to be present at the same concentration or exhibit the same activity. Thus, by assaying with two different non-specific esterase substrates (fluorescein diacetate and 4-methylumbelliferyl acetate) simultaneously using dual laser excitation (argon 488 nm and krypton UV) in a mixture of two different cell types it should be possible to distinguish between the populations. The results of one such assay are shown in figure 14.17, where the green (fluorescein diacetate hydrolysis) fluorescence for each cell is scored on the ordinate against the violet (4methylumbelliferyl acetate hydrolysis) fluorescence on the abscissa. Two distinct populations are apparent. The cluster to the upper left represents the EMT6 mouse mammary tumour and that to the lower right represents the human colonic adenocarcinoma HT29. These two cell lines were chosen for the initial analysis for two main reasons. Firstly, the cell types were sufficiently different to stand a reasonable chance of obtaining a discrimination and, secondly, the plating efficiencies and culture characteristics were well documented and understood. Apart from the primary objective of distinguishing between different cell types these assays were being developed with a view to maintaining cell viability. This

0) O

c 0)

o 0)

o 3

0)

Violet fluorescence Figure 14.17. Dual non-specific esterase substrate assay where green fluorescence (fluorescein diacetate hydrolysis) for each cell is scored on the ordinate against violet fluorescence (4-methylumbelliferyl acetate hydrolysis) on the abscissa.

332

DYNAMIC CELLULAR EVENTS

has largely been achieved with both cell types maintaining a plating efficiency of about 90% even after a 6 hour exposure to maximum concentrations of the substrate mixture. It has also been reported that a discrimination can be made between fibroblasts, a lung tumour and a second colon carcinoma, all of human origin, using this combined esterase assay (Chambers and Watson, 1980; Watson, 1980b). These results were sufficiently encouraging to attempt to discriminate between the different subsets to be found in the much more complex biological system of bone marrow. Using a combination of six simultaneous measurements on each cell (UV and blue light both scattered at 90° and at narrow forward angles, coupled with the two non-specific esterase substrates, fluorescein diacetate and 4-methylumbelliferyl acetate) it has been possible to identify between 10 and 16 different subsets in mouse bone marrow (Watson and Chambers, 1980; Watson, 1980b). The same combination of fluorochromes has been used by Malin-Berdel and Valet (1980) to characterize mouse spleen and bone marrow cell subpopulations based on differing esterase and phosphatase activities. Some examples of this type of multiparameter analysis, in which only one substrate was used, were shown in chapter 5 (figures 5.5 and 5.8) as an illustration of multi-parameter data handling and display. A combination of esterase and y-glutamyl transpeptidase activity using fluorescein diacetate and y-glutamyl-7-amino-4-methyl-coumarin simultaneously as substrates has been used to discriminate between normal and malignant liver cell lines. This technique also requires a twin-laser instrument to excite the reaction product from y-glutamyl transpeptidase activity (4-methylumbelliferone, krypton UV excitation) and esterase activity (fluorescein, 488 nm argon excitation). Esterase and y-glutamyl transpeptidase activities were both greater in the aflatoxin-transformed malignant cells which enabled an excellent discrimination to be made between the two cell types.

14.2.7 Inhibition kinetics Chloroethylnitrosoureas (Cnus) are an important class of antitumour agent. Their exact mode of action is not known but chloroethylation of DNA and subsequent crosslinking has been implicated (Gibson, Mattes and Hartley, 1985). However, protein carbamoylation may also be involved as organic isocyanates as well as alkylating species are formed when Cnus decompose in aqueous environments. Carbamoylating agents have been shown clearly to inactivate a number of enzymes, and inactivation of glutathione reductase has been used to determine the protein carbamoylation potential of Cnus (Babson and Reed, 1978). Because of the susceptibility of serine hydrolases to isocyanate inactivation (Brown and Wold, 1973) Paul Workman of our laboratories proposed that flow cytometric measurement of esterase activity might form a basis for the determination of intracellular carbamoylation by Cnu-derived isocyanates (Dive et al., 1987b,c). This proved to be the case and an example of esterase inhibition induced by BCNU-derived isocyanate is shown in figure 14.18. By using a number of different drug doses as in figure 14.19 it is possible to obtain an estimate of the

333

ENZYME KINETICS

B 50

I

40

UJ

O

UJ

30

I

20

11.

10

U

o

0

100

200

300

TIME ( SEC )

Figure 14.18. Inhibition of esterase activity by BCNU-derived isocyanate (Dive et al, 1987b). The top two diagrams (panel A) show size versus fluorescence on the X- and Z-horizontal axes with time (which increases upwards on these displays) on the Y-axis. Contour plots of size versus fluorescence for ten equal 'time-slices' are shown within these three-dimensional data spaces. Top left is the control and top right shows the inhibition due to BCNU. Panel B shows the medians of the fluorescence distributions of each 'time-slice' plotted against time.

drug concentration which causes a 50% reduction in FDA hydrolysis activity, defined as the I50 value. Using the I50 it is possible to compare directly the inhibitory effects of a number of Cnus and related cytotoxic agents and the potential utility of this is discussed in section 15.3.2.

14.2.8 Short time scale kinetics Kinetic measurements in the temporal reference frame of a few seconds after biochemical perturbation are potentially of great importance for studying cell activation and a new technique for this type of measurement was described in

334

DYNAMIC CELLULAR EVENTS

100

10'6

10'5

10"4 BCNU CONCENTRATION (M)

10'3

10'2

Figure 14.19. Enzyme activity for multiple concentrations of BCNU. The dose which produces 50% inhibition of enzyme activity, I50, can be estimated from data such as these. section 14.1.4. The method was tested with EMT6 mouse mammary tumour cells in exponential growth adjusted to a concentration of 2 x 105 cells ml" 1 and introduced into pump 1 (see figure 14.3). FDA was made up at a concentration of 2 |oM and loaded into pump 2. Both pumps were started and the high tension voltage of the green detector was adjusted to record the fluorescein emission histogram with a mean in about channel 600 at pump flow rates of 100|j] min~ l for a tube length of 40 cm. Recordings were then made at each of the four tube lengths at the same flow rate. From the data shown in figure 14.20 it can be seen that the fluorescence distribution is progressively shifted to higher values with increasing tube length. It should also be noted that the distribution remains constant with time as the period taken for cells to flow from the mixing chamber to analysis point is constant. This overcomes one of the potential problems associated with continuous time recording of 'fast' reactions where, because of the reaction velocity, only relatively few cells can be recorded with a given fluorescence intensity. A number of further recordings were made with various tube lengths and pump flow rates. The medians of the green fluorescence distributions so obtained are plotted against time in figure 14.21 which shows that the accumulation of fluorescence was biphasic over the first 16 seconds of the reaction. A regression analysis (solid line) was carried out for the data beyond 5

MEMBRANE POTENTIAL

335

5 cm

10 cm

o c 0)

20 cm

3

& 0)

40 cm

Fluorescence intensity Figure 14.20. FDA hydrolysis fluorescence distributions for four tube lengths at the same flow rate from EMT6 cells. seconds and this gave a correlation coefficient of 0.99 which is compatible with a linear increase in fluorescein with time between 5 and 16 seconds. However, the dotted curve drawn by eye, is more compatible with the biological reality and the lag probably represents the time taken for substrate to diffuse across the external cell membrane and cytoplasm to the site(s) of enzyme action.

14.3

Membrane potential

Fluorescent membrane potential probes are charged symmetrical, 'mirrorimage ring-structure molecules with hydrocarbon side chains. The molecule retains lipophilicity as the charge is not localized and the relative lipophilicity can be engineered by changing the length of the side chains. The cationic cyanin dyes 7

DYNAMIC CELLULAR EVENTS

336

6OO1 ft

I

* 400 o o o

1 200 LL

10

15

Time, seconds Figure 14.21. Biphasic FDA hydrolysis over the first 16 seconds of the reaction shown by the dashed curve. The data from 5 to 16 seconds can also befittedwith a straight line (non-interrupted) but this is less likely to represent the biological reality. were the first such molecules to be recognized as membrane-potential probes (Sims et al, 197A). They partition in lipids to an extent that varies partly with side chain length and partly with potential across the cell membrane. A second class of dyes, the oxanols, are functionally similar to the cyanins differing only in that they are anionic (Rink et al, 1980). Cyanins are excluded as cells undergo depolarization whereas oxanols are excluded on re- and hyper-polarization. The cyanins, however, suffer from the disadvantage that they are taken up extensively by mitochondria which can reduce the 'signal-to-noise' ratio when assessing external membrane potential. This problem, however, is overcome by using the oxanols. These probes have been used in studies of granulocyte and lymphocyte function and in mitogenic stimulation of T-cells (Tsien, Pozzan and Rink, 1984).

14.4

Calcium

Ligand activation of cell surface receptors linked to the phosphatidyl inositol pathway generates the second-messenger molecules inositol triphosphate and diacylglycerol (Berridge and Irvine, 1984). Inositol triphosphate mobilizes calcium leading to an increase in free calcium ions (Ca2 + ) in the cytoplasm. As stated in section 8.4.2, Indo-1 is the preferred calcium probe as it exhibits a shift in emission spectrum on binding calcium. These spectra are shown in figure 14.22 where it is apparent that as the Ca2 + concentration rises there is a progressive shift of the emission to shorter wavelengths. In the absence of Ca 2+ the emission peak is at

CALCIUM

337

0.1-1 mM

X 0)

4

4) O 0) O (0

o =2 2

500

400 Wavelength, nm

Figure 14.22. Changes in the emission spectrum of Indo-1 with increasing Ca2 + concentration which are given in nanomoles against each curve except for the maximum which is micromolar. These data were redrawn from Grynkiewicz et al. (1985) who pointed out that the 'notch' at 465 nm was an instrument artefact.

480-485 nm and at saturating concentrations this has shifted to 400-405 nm. Thus, the ratio of the emissions at 400 nm and 480 nm gives a relative measure of calcium chelated by the probe. Rabinovitch et al. (1986) and more recently Keij et al. (1989) have produced a method for expressing the fluorescence ratio as a concentration of Ca 2+ by the following equation, nMCa2+=Kdx

f

b2

The meanings of these various symbols are as follows. Kd is the Indo-1/Ca2 + dissociation constant which is numerically equal to 250 nM (Grynkiewicz et al., 1985). R is the 400:480 nM fluorescence ratio and Rmax and Rmin are the 400:480 nM fluorescence ratios for Indo-1 loaded cells in the presence of saturating concentrations of Ca2 + and in the absence of Ca 2+ respectively. A value for Rmax can be obtained in intact cells by treating with the ionophore ionomycin (Rabinovitch et al, 1986) which allows equilibration of internal Ca2 + with the very much higher external concentration to take place. Rm\n is obtained by treating the cells with ionomycin + EGTA. The latter chelates Ca 2 + external to

DYNAMIC CELLULAR EVENTS

10

2 3 Time, minutes

2 3 Time, minutes

Figure 14.23. T-cell activation showing [Ca2 + ] versus time where panels A and B give dot-plots for each cell and population means respectively.

the cells and hence 'drains' Ca2 + from the cells. S b2 and Sf2 are the fluorescence values at 480 nm for Ca2 + -bound Indo-1 and Ca2 + -free Indo-1 respectively and these too are obtained from cells treated respectively with ionomycin and ionomycin + EGTA. A slight variation on this theme was used by Keij et al. (1989) where saponin + EGTA was used instead of ionomycin + EGTA. Keij et al. (1989) have used the formula given above to present results in the form of Ca 2+ concentrations in studies involving both T- and B-cell activation with PHA and pneumococcal polysaccharide respectively. The T-cell results showing [Ca2 + ] versus time are reproduced in figure 14.23 where panels A and B give dot-plots for each cell and mean population values respectively. Time was not incorporated into the list-mode data base directly for these studies but it was 'inferred' from the computer recorded duration of the run by dividing sequentially the list-mode data into 255 equal channels. Thus, a time-related factor was produced which was used to plot [Ca2 + ] versus time. This is not the ideal method for incorporating time into the data base as extreme care must be taken to keep the flow rate constant. However, it is the only compromise possible for commercial instruments which do not include the facility to incorporate time in the data base from the computer clock. However, Griffioen et al. (1989) have used this method to study Ca 2+ kinetic changes in lymphocytes. Indo-1 has also been used to study Ca 2+ kinetic changes in lymphocytes under various stimuli by Chused et al. (1986,1987) who were able to distinguish between T- and B-lymphocytes from mouse spleen cultures where the mixed population was subjected to B-cell mitogenic stimuli. Within 20 seconds the B-cell component had responded as indicated by their increase in Ca2 + . It has also been used to study Ca 2+ changes in polymorphonuclear leukocytes (Lopez, Olive and Mannoni,

DRUG TRANSPORT

339

1989) and in platelets subjected to thrombin stimulation (Davies et al, 1988). In the latter study there was an initial increase in Ca 2+ which was maximal by about 15 seconds followed by a slow decline. The use of flow cytometry enabled two subsets to be defined where one responded fully but the other responded partially or not at all. Rabinovitch et al. (1986) studied not only Ca 2+ changes during lymphocyte activation but also cell surface determinants simultaneously with Ca 2+ using antibodies labelled with phycoerythrin in a multi-parameter assay. Heterogeneous Ca 2 + responses were observed which to some extent were related to immunophenotype. Only examples of Ca2 + responses using Indo-1 have been given in this section as this is the current probe of choice; however, the uses of fura-II and quin-II together with Indo-1 have been reviewed by Ransom, DiGiusto and Cambier (1987).

14.5

Mitochondrial function

Rhodamine-123 is a class of cationic dye which partitions into electronegative environments and has been described as a mitochondrial specific dye (Johnson et al, 1981; Weiss and Chen, 1984). This description was based on fluorescence microscopy observations which revealed intense mitochondrial staining and is not strictly correct as rhodamine-123 will partition into any electronegative compartment. However, the interior of mitochondria is very electronegative if the proton pump is functioning normally and staining intensity gives some indication of the integrity of normal mitochondrial function. Rhodamine-123 has been studied in some in vitro tumour cell lines where it is reported to be preferentially retained and selectively cytotoxic (Lampidis et al, 1982). A decrease in its uptake has been used as an indicator of loss of mitochondrial function possibly associated with decreased respiratory activity in L1210 cells treated with anticancer drugs (Bernal, Shapiro and Chen, 1982). Martinez, Vigil and Vila (1986) have also used this probe as an indicator of respiratory activity and its uptake is increased in some small cell lung cancer cell lines after interferon-y treatment (Jabbar, Twentyman and Watson, 1989).

14.6

Drug transport

Transport of any agent across the external cell membrane from without occurs by at least three mechanisms; see review by Goldenberg and Begleiter (1984): (1) passive diffusion, where rate of increase of the intracellular concentration is directly proportional to the concentration gradient; (2) facilitated diffusion, where influx involves specific interactions between transport molecules and drug which, unlike passive diffusion, is temperature sensitive and exhibits saturation kinetics but the intracellular concentration cannot exceed that externally; (3) active transport, where the agent can be concentrated within the cell

340

DYNAMIC CELLULAR EVENTS

against a gradient using an energy-dependent mechanism. Clearly, these factors also apply to transport from inside the cell to outside, and this has important implications in multi-drug resistance. Flow cytometric quantitation depends on fluorescence and any direct measurements of drug uptake must rely on the drug itself being fluorescent, quenching of a reporter fluorochrome by the drug or the availability of a fluorescent analogue. Two classes of agents fall into these categories. First, the commonly used anthracyclines, daunorubicin and adriamycin, are both fluorescent and can induce quenching of propidium iodide/DNA emission. Second, fluorescent analogues of the anti-folates, aminopterine and methotrexate, have been synthesized (Gapski et al, 1975; Henderson, Russell and Whiteley, 1980; Rosowsky et al, 1982; Kumar et al,

1983a,b).

The quenching of propidium iodide DNA fluorescence as a means of assessing nuclear adriamycin content was demonstrated by Krishan and Ganapathi (1979) and these investigators also demonstrated the feasability of measuring intracellular anthracycline fluorescence directly (Krishan and Ganapathi, 1980). Durand and Olive (1981) used similar direct fluorescence techniques to quantitate cellular adriamycin and additionally they showed time courses for uptake and compared nuclei with whole cells. Some of these data are reproduced in figure 14.24. Moreover, Durand (1981) demonstrated a considerable adriamycin concentration gradient across multi-cell spheroids. Lower adriamycin fluorescence was found in cells in the center which had greater cell survival probability than those at the periphery where fluorescence was greater. Multi-parameter analysis of adriamycin and daunomycin uptake in mouse bone marrow and acute myeloid leukaemia cells has been carried out by Sonneveld and van den Engh (1981) using forward and 90° light scatter properties to identify different subsets (Visser et al, 1980). Fluorescence was related directly to cytotoxicity and myeloid progenitor cells were an order of magnitude more sensitive to daunomycin compared with adriamycin, a result which correlated with fluorescence. In contrast, the fluorescence from acute myeloid leukaemia cells was greater for adriamycin than for daunomycin. Further studies of daunomycin fluorescence uptake by rat bone marrow were carried out by Nooter, van de Engh and Sonneveld (1983) where dead cells, lymphocytes, blast cells, granulocytes and red cells were separately identified on light scatter. At higher concentrations the lymphocyte fraction exhibited greater fluorescence than granulocytes arid at given concentrations the dead cells exhibited considerably greater fluorescence than viable cells. These various effects were attributed to quantitative differences in active transport of these agents into different cell types. Absolute calibration of drug content per cell was also effected in these studies using [3H]-daunomycin and, in conjunction with cell sorting, a detection limit below 10~ 18 M cell" 1 was obtained. Anthracyclines are actively transported both into and out of cells, but by different pump mechanisms. One of the efflux pumps is the 170 kda membrane associated glycoprotein (pg170) encoded by the Multi-Drug-Resistance, MDR-1,

DRUG TRANSPORT

341

1000o | 500-

50

100 150 Fluorescence

200

250

0)

o

O

3-

o

30

60 90 Time (minutes)

120

Figure 14.24. Panel A shows adriamycin fluorescence distributions for 2 hour exposures at the concentrations, in Jig ml" 1 , indicated adjacent to each histogram. Time courses for the increase in adriamycin fluorescence for whole cells (O) and nuclei ( # ) exposed to 1.0 |lg ml" 1 are shown in panel B. Redrawn from Durand and Olive (1981).

gene (see section 15.3). Extensive evidence now exists which shows that multidrug-resistant cells have increased numbers of membrane associated gp 170 molecules and as a result MDR cells exhibit lower fluorescence on exposure to anthracyclines compared with sensitive cells as these agents are actively transported out of MDR cells with greater efficiency. Ross et al (1989) have used this property to cell sort and isolate a very low frequency but highly MDR population from within a sensitive P388 cell population on duanorubican fluorescence. The correct functioning of the efflux pump mechanisms can be modulated by Ca 2+ channel blockers (chloropromazine, trifluoperazine and

342

DYNAMIC CELLULAR EVENTS

verapamil; Tsuro et al, 1982). A number of studies on the modulation of anthracycline uptake by Ca2 + channel blockers have been carried out using flow cytometry in resistant and sensitive cells in both animal and human systems (Krishan et al, 1986, 1987; Ross, Joneckis and Schiffer, 1986; Nooter et al, 1989; Morgan et al, 1989; Watson, Morgan and Dive, 1989). The 'bottom line' in all these studies is that the lower accumulation of anthracyclines in multi-drugresistant cells can be partially reversed by exposure to Ca 2+ channel blockers which interfere with the efflux pump mechanism. This is easily monitored by flow cytometry to give a predictive assay for MDR which will be described in section 15.3. Similar types of studies have been carried out with cells sensitive and resistant to the antimetabolite methotrexate using fluorescent analogues of the latter. However, methotrexate resistance is different from that in MDR and two of the three mechanisms identified are attributable to Fisher (1961, 1962). The enzyme dihydrofolate reductase, DHFR, converts folic acid to folinic acid which is required in one of the pathways to DNA synthesis. Methotrexate binds to DHFR with an affinity about 40-fold greater than that of folic acid and blocks its action. Fisher (1961) demonstrated that methotrexate resistance could be caused by elevation of dihydrofolate reductase levels. Cells grown continuously in medium containing methotrexate exhibit amplification of the DHFR gene (Hanggi and Littlefield, 1976; Alt, Kellems and Schimke, 1976) where the multiple copies are contained in 'double minute' chromosomes and are manifest in such cells at metaphase. Stable mutants are obtained when these double minutes become integrated into the normal chromosomes and the amplified gene can then be transmitted predictably from one generation to the next which confers considerable methotrexate resistance on the population. Fisher (1962) also demonstrated that the rate of influx of methotrexate (MTX) was about 15-fold lower in resistant compared with sensitive cells. At equilibrium the intracellular MTX concentration was between 10- and 30-fold lower in the resistant cells. However, this is not due to metabolic changes as dihydrofolate reductase levels were equal in both the resistant and sensitive cells. Changes in methotrexate membrane transport contributing to resistance (Schimke, 1984) have also been demonstrated by Sirontak et al. (1981) and Dembo, Sirotnak and Moccia (1984) have presented evidence that the cellular influx and efflux pathways for methotrexate are different. The third mechanism of methotrexate resistance is due to an alteration of the affinity of methotrexate binding to DHFR (Flintoff, Davidson and Siminovitch, 1976; Haber et al, 1981). To date relatively limited work has been carried out on methotrexate-resistant cells using flow cytometry. This has been in part due to the difficulty of producing fluorescent analogues with low background as pointed out by Gaudray, Trotter and Wahl (1986). However, these authors overcame this problem with a fluorescenated derivative of the ligand and presented data which suggested that staining heterogeneity with the analogue was partly due to transport variability within the population. Assaraf and Schimke (1987) loaded cells with fluorescent methotrexate then examined the ability of hydrophilic and lipophilic antifolates to displace the probe. Methotrexate-resistant cells incubated with methotrexate

CONCLUDING REMARKS

343

200-

4

6

8

Time (hours) Figure 14.25. Uptake kinetics for fluorescent methotrexate in a parental cell line (AA8, # ) and a dihydrofolate reductase diffident cell line (DG-44, O) from Assaraf et al (1989).

(hydrophilic) exhibited no reduction in fluorescence compared with sensitive cells however, incubation with lipophilic analogues produced reductions in fluorescence in both sensitive and resistant cells. This is very compelling evidence for a membrane active-transport mechanism for methotrexate and hydrophilic analogues which is deficient in resistant cells. The lipophilic analogues enter cells by passive diffusion without the need for an active-transport mechanism, hence these displaced the fluorescenated probe bound to DHFR in both the sensitive and resistant cell lines. Similar types of studies have been carried out by Rosowsky et al (1986) and Wright et al (1987). Figure 14.25 shows the uptake of fluorescent methotrexate versus time in a parental Chinese hamster cell line, AA8, and in a DHFR-deficient cell line, DG-44. The former exhibits very considerably greater fluorescence which is attained at a more rapid rate. The effect of a temperature reduction, which slows the uptake rates of active-transport processes but not of passive diffusion, is shown in figure 14.26. There was a profound decrease in fluorescent methotrexate uptake in parental CHO when the temperature was decreased from 37°C to 4°C, an effect which was easily observed in the initial stages of uptake with a temperature decrease to 27°C. Both of these last two data sets were redrawn from Assaraf, Seamer and Schimke (1989).

14.7

Concluding remarks

Measurement of dynamic cellular events in intact cells under physiological or near physiological conditions, irrespective of the assay system, is arguably one of the most powerful techniques in cell biology. Physiological and

344

DYNAMIC CELLULAR EVENTS

25

Time (minutes) Figure 14.26. Temperature-dependent uptake of fluorescent methotrexate in parental CHO cells. The insert shows that cooling from 37°C to 4°C effectively abolishes uptake. The main graph shows that a reduction in uptake can be measured with a temperature decrease to only 27°C, from Assaraf etal. (1989).

pathological processes are not static but are continuously varying. Observing and measuring an event in isolation gives us little or no insight into what gave rise to the event or what the future consequences might be. The measurement of time structures our perception of the biological process as it places sequential interdependent events in the correct temporal relation to each other. It is becoming increasingly obvious that many of the initial events involved in cell proliferation and differentiation control take place within a time frame of seconds, but the effects of those initiating events (e.g. growth factor stimulation, membrane calcium flux and pH changes and carcinogen interaction) may only become apparent some days, weeks or even years later. Clearly, an understanding of the dynamics of inappropriate or abnormal initiating events will be of crucial importance in attempts to modulate pathological states. Incorporation of time into flow cytometry data bases, particularly for short time scale processes, undoubtedly will make contributions to that understanding.

15 Applications in oncology

This chapter originally started life as applications in clinical medicine and as such it would have been relatively easy to write five years ago. However, with the amazing proliferation of flow cytometry applications in medicine during the past few years this has become such a daunting prospect that I 'ducked-out' and decided to contract it to applications in clinical oncology. The latter discipline in the UK means all aspects of cancer patient management including diagnosis, pathology, surgery, radiation therapy and drug treatment. Writing in my capacity as a clinical oncologist I think I can cope with this, but I would like to stress the 'think', as this is still a formidable undertaking. At this point I would advise people who have no interest in cancer and whose area has not been covered to move smartly on to the epilogue where you may find just a brief mention of your field which has been omitted from the volume. However, I suppose that won't help much so you might as well stop here. After further deliberation I decided not to embark on a comprehensive account of all that has been done in cancer with flow cytometry although many of the references are included. I thought it might be more constructive to try to concentrate on what needs to be done and how flow cytometry might be able to help. In order to effect this approach it is necessary to ask a number of questions although I'm not trying to pretend there are any 'real' answers at present. As a physician confronted by a patient with cancer I would like to know a number of things. Firstly, the exact diagnosis, which is usually straight forward, but not always. For instance, it is sometimes difficult to be precise about tumours in lymph nodes in the neck. Even the very best histopathologists can have difficulty discriminating between lymphomas and undifferentiated carcinomas at this site particularly if the latter is arising in the nasopharynx. The metastasis can look exactly like a lymphoma and the primary may be very small, even microscopic, and have been missed on ENT examination. This doesn't happen very often, but it does occur. Similar problems can arise with primary melanoma in the maxillary antrum. Again, this is rare but does happen. These are diagnostic problems; can flow cytometry help here? I'll illustrate some further problems and very serious dilemmas by a story of two patients (which could be entitled a Tale of Two Titties' and is an equally sad story) who presented with apparently identical disease. The human mind has an

346

APPLICATIONS IN ONCOLOGY

extraordinary capacity to recall anecdotal rare coincidental events, unlike the computer which is hopeless in this respect, but as certain aspects of the saga were so harrowing it's not surprising they made such an indelible impression and are still so readily recalled. It started in 1984 shortly after the SAC meeting in Asilomar. Two young women in the 30—35 age group presented on the same day having both had wide biopsy excisions of small lumps ( < 1.5 cm) in the upper outer quadrants of their right breasts. Both had had two children and both had been taking oral contraceptives intermittently during the previous 10—15 years. In neither case had there been skin involvement or deep tethering and neither had palpable lymph nodes in the axilla. In both cases there had been complete excision, as far as it was possible to tell, and the histological diagnoses were identical, moderately differentiated infiltrating duct carcinomas. Both patients were treated identically with radiation to the breast and gland areas and during treatment they became mutually supportive good friends which tends to occur with considerable intensity when people are under severe stress. One was dead within 10 months with generalized metastatic disease and the other is alive and well over 5 years later though she still suffers psychological trauma due to her friend's death and the realization that this could still happen to her. Obviously, these two cases were very different in spite of the superficial similarities, but were the tumours different or were the patients surrounding the tumours different? It was almost certainly a combination of the two; however, both factors are difficult to investigate although it has long since been a part of empirical oncological knowledge that faster growing tumours tend to be associated with a reduced chance of a successful outcome. Was the tumour growing faster in the patient who died? These are questions which relate to prognostic factors; can flow cytometry help us here? These two cases also illustrate a number of further classes of problems, one of which is therapy selection. With the benefit of the retrospectroscope we may ask if it was necessary to treat with post-operative radiotherapy the patient who survived. After all, the excision of the small lesion appeared to have been complete and there were no palpable lymph nodes which all added up to a favourable prognosis. The patient who died also fell into this category and the decision to treat her was correct but, clearly, radiotherapy was the wrong choice. This raises further points. Did she have microscopic disease in axillary lymph nodes which was not clinically apparent? If so, were the cells resistant to radiation and consequently survived to metastasize or had the disease undergone generalized metastasis before diagnosis? If this were the case then local radiotherapy was totally inappropriate and, in the present state of our knowledge, chemotherapy should have been employed. However, would this have made any difference and can flow cytometry help us with these types of problems? We can attempt to gain answers to some of these questions using clinical trials where one treatment modality or combination is compared with another. The problems with this approach are multi-fold. Firstly, many anti-cancer clinical trials are based purely on empiricism and there is insufficient basic knowledge available to make more than a random guess as to what treatment modalities or

APPLICATIONS IN ONCOLOGY

347

combinations are worth comparing. One is reminded here that if you give enough chimpanzees a type writer each and leave them for long enough they will type out, purely by chance, the whole of the works of William Shakespeare and probably the Bible as well. The clinical trial om'cionados may well, at this point, cry 'what about lymphomas, leukaemias, chorion carcinoma and teratoma'. And, indeed these are some of the few tumours which are curable by chemotherapy, but it is worth noting that the initial major advances in these tumours did not come about from clinical trials but by inspired guesses. Moreover, some of the other guesses were not well reported as they didn't work. The lack of real knowledge is the major contributor to the overwhelming vast excess of essentially negative results obtained from such multi-center clinical trial enterprises over the past quarter of a century. Moreover, most trials now seem to accept within the statistical design specification that there is likely to be a very small difference in results between comparative arms of such studies. This means that large numbers of patients, up to 25 000 in some examples, need to be 'recruited' (a highly used euphemism) to obtain statistically significant results. This is great for keeping cancer research statisticians employed and maintaining high turnover in the paper industry but bad for trees and most of the time it doesn't do the patients much good either. Those that fare particularly badly are the poor devils recruited to phase I toxicity studies where the aim is merely to determine the maximum dose of a new agent that can be achieved with no attempt or hope of obtaining any benefit for the patients in the trial. Obtaining fully informed consent from these patients, who are usually pre-terminal, is not usually sought with rigour as nobody would agree to enter such an experiment (current euphemism is 'study') if they were told the whole truth in straight forward, blunt and simple language which might run something like this. 'Mr Smith, you have known for sometime now that you have a cancer and that this is not likely to be curable. We have tried all the currently available treatment and the disease is still progressing. There is a new drug which has never been tried in human beings although we know the dose which kills 50% of rats, mice and some other animals. We don't know if it is likely to have any beneficial effects but we are certain it has side effects which could be very severe. What we want to do with you is give you increasingly large doses until we find out how bad the side effects are and how much humans can tolerate.' Would you say yes to this proposition if it had been spelt out as clearly as this? Moreover, I would submit that the realization that such large numbers are required for some trials is tantamount to admission of failure at the very outset. Even if a positive result from a clinical trial is obtained, and by positive I mean that it is shown that one form of treatment confers a survival advantage, the physician still has a problem. The results of the trial will say, for example, that at a given level of statistical significance (say 0.95) 20% of the patients will live 25% longer if treated according to protocol A compared with B. This isn't a great deal of help as we do not know which 20% of the total will benefit or even if the extra 25% of time is worth having from the patients' point of view, notwithstanding the fact that most of this 'gained' time will have been spent undergoing intensive and frequently very

348

APPLICATIONS IN ONCOLOGY

toxic therapy. Perhaps the most telling statistic is that only 19% of Ontario cancer specialists would have consented to enter a chemotherapy trial for metastatic lung cancer (Mackillop et al, 1988). The major problem in the therapeutic management of cancer is that we do not know when we should intervene at the individual patient level. The population statistics obtained from clinical trial results cannot directly be applied to the patient who is consulting you and is in front of you at this very minute. As a further example let us suppose that a series of independent prognostic factors are available which tell us that 80% of patients with a given tumour do not require further intervention following surgical removal. We have a probability of 0.8 of being correct if we manage the whole of the group by observation alone and the vast majority of cancer therapists would be quite pleased with themselves if they had 80% five-year survival figures for the majority of the tumours they manage even if they didn't prescribe any treatment. But, what about the remaining 20% who are doomed to die with this philosophy? Equally, we may have prognostic factors which tell us that 60% of patients will survive if they are treated with adjuvant radiotherapy plus a particular chemotherapy regime. Again, in order to obtain fiveyear survival figures of 60% all patients within the group would have to be treated which means that 40% would be subjected to toxic intervention with no benefit. I don't want to give the impression that I am a disbeliever in clinical trials, in fact I believe in them absolutely as this is the only true and objective means of testing the efficacy of treatment. The problem lies in our depressingly considerable lack of real understanding of the processes involved in the diseases and the modulation of both abnormal and normal cell behaviour, let alone that of organ systems, by the treatments we employ. However, there is no need to get too depressed as very considerable technological strides have been made in the past 15 years which are enabling us to gain a better understanding of the biology. And here, I would like to draw your attention to the quotation from Whitehead at the beginning of the book,'... it is not the imagination of man that improves but his capacity to measure which increases ...'. Flow cytometric methods combined with those of biochemistry and of molecular and cellular biology are beginning to lighten the darkness by enabling us to measure, and some assay systems are emerging which may allow us to grasp the 'Holy Grail' of cancer therapy, individualized treatment specifically designed for and aimed at a given tumour within an individual patient. 15.1

Diagnosis

Flow cytometry has a very limited role in primary diagnostic decision making. A biopsy will be taken and subjected to standard histological procedures and interpretation and in the majority of cases the diagnosis can be made. Even where there is potential difficulty, e.g. neck lymph nodes where the tumour could be anaplastic carcinoma or lymphoma, flow cytometry still has little or no use. Antibodies to cytoskeletal elements can assist in making discriminations such as these using classical immunoperoxidase or immunofluorescence staining

DIAGNOSIS

349

(Ramaekers et al, 1982b, 1983) and these assays have been adapted for flow cytometry (Ramaekers et al, 1984, and see figure 12.18). The latter was completely convincing and produced very pretty data, but we may ask if flow cytometry was really necessary? The answer must be that it isn't as the simpler classical methods can cope with this perfectly adequately. If you happen to have a flow cytometer and time to spare then certainly do this type of assay but you wouldn't buy a cytometer just to do this particular job. 15.1.1 Leukocyte classification Flow cytometry has been used extensively in leukocyte and bone marrow immunophenotyping (Nicola et al, 1981; Hoang et al., 1983; Loken and Lanier, 1984; Watt et al, 1984; Loken, 1986; and see in leukocyte typing, 1984) and can be useful in subclassification within the lymphomas and leukaemias which is carried out after the primary diagnosis has been established. A large number of antibodies to various leukocyte cell surface differentiation determinants are now available and the list is increasing all the time. Immunological typing of lymphomas has been carried out for many years (Lukes and Collins, 1974). However, following the advent of monoclonal antibody technology (Kohler and Milstein, 1975) and the recognition of discrete differentiation stages in both T-cells (Reingertz et al, 1980) and B-cells (Nadler et al, 1982,1983) which can be identified with antibodies there has been an explosion in this field. Antibodies to these various differentiation antigens have been used extensively in the classification of lymphomas and leukaemias (Griffin et al, 1981; Anderson et al, 1984; Foon and Todd, 1986) and the role of flow cytometry, which can encompass multi-dimensional analyses (Civin and Loken, 1987; Terstappen and Loken, 1988), has recently been reviewed by Parker (1988). I'm not going any further with this as it is a whole book's-worth in its own right.

15.1.2 Cytological prescreening The technology can be of potential assistance in primary diagnostic procedures in screening where large number of cells, or large numbers of samples have to be analysed. Any automated system which is reliable could have a very valuable place in cytological prescreening. The problems with manual methods are considerable. They are labour intensive, skilful, tedious, repetitive and require a high degree of sustained concentration. Some of these factors are mutually exclusive for many individuals. One of the major problems, almost by definition, is that the vast majority of screening specimens are entirely normal. Assuming there are no problems with specimen collection and preparation (that's a big 'if') perhaps only one sample in every hundred is likely to be questionable let alone obviously malignant. Whenever the expectation of observing an event is low it is very difficult for the human mind to maintain concentration so that when eventually that event does present it is not missed. The objective in any automated instrumental system is to screen out the normal samples to leave the questionable and abnormal specimens for humans to make the decisions. If at the outset 99% of

APPLICATIONS IN ONCOLOGY

350

t
o

(0

o

z

•

•

•

•

'

.

'

-

•

•

<

'

Nucl


D e c i s i o n •!::•:••! ^/r-^^rtx^- :•"::: L l y ^ ^ S : boundary, j : ^ ^ : ^ : : : : : : : : : : : : : : : ^ ^ : : ^ :

|

.

:

:

•

I,

Nuclear-to-cell-diameter ratio ^

Figure 15.1. Nuclearfluorescenceplotted against nuclear-to-cytoplasmic ratio calculated from the slit-scan profiles. The horizontal line represents the abnormal cell decision boundary.

specimens are normal and a machine can decide that 90% are normal then only 10% have to be looked at by humans. This has two advantages. Firstly, the expectation of seeing an abnormal specimen is increased 10-fold which helps to maintain concentration and increases performance and, secondly, it decreases the number of skilled 'cerebral units' required to perform a repetitive and boring task. Slit-scan flow cytometry which was described in section 8.1 has been used by Wheeless et al to analyse both gynaecological (1984) and bladder specimens (1986a) in an extensive experimental prescreening programme. The specimens were stained with acridine orange and total nuclear fluorescence was plotted on an individual cell basis against nuclear-to-cytoplasmic ratio calculated from the slitscan profiles of each cell. A typical example is shown in figure 15.1 where the horizontal line represents the abnormal cell decision boundary. No screening system (including humans) will be 100% effective all the time. The most important fundamental requirement of automated prescreening systems is that the number of false negative results, defined as a positive which the instrument fails to recognize, should be acceptably low. A second but less critical requirement is that the false positive rate, defined as being scored positive by the instrument when in fact it was negative, should not be unacceptably high. Exactly what constitutes acceptable low or not unacceptably high is debatable; however, if the false positive rate is very high there is little point in using an automated system as it will not be screening out a worthwhile proportion of the total. Wheeless et al. (1986b) have developed a statistical analysis system to evaluate their multi-parameter data and in 251 gynaecological specimens have achieved a

PROGNOSIS

351

1.2% false negative error rate at the expense of a 24% false positive error rate. This false negative error rate of 1.2% compares very favourably with human performance. The 24% false positive error rate means the 75% of normal specimens were correctly classified, which would cut down the manual screening needed to 25%. It would seem, therefore, that this is well worth pursuing. In the feasability programme for bladder prescreening a total of 153 irrigation specimens have been analysed (Wheeless et al, 1986a). Of the 38 normal specimens only four were classified as abnormal, a false positive error of 10.5%. Sixty nine specimens were derived from patients with transitional cell carcinoma grades 1—2 and 66 were correctly classified as abnormal, a false negative rate of 4.3%. A total of 17 specimens were derived from grade 2—3 lesions and all were classified correctly. Only eight cases of dysplasia were analysed and the instrument scored seven of these as abnormal. These results are also extremely encouraging particularly as the chance of obtaining a false negative twice from the same patient must be extremely small and, undoubtedly, we should look forward to having such instrumentation available within the next decade or so to assist in routine screening programs.

15.2 Prognosis The most extensively studied parameter in attempts to produce prognostic information using both image analysis and flow cytometric systems has been total nuclear DNA content. This gives encouraging results in a number of tumours and the advent of antibodies to oncogene-encoded proteins has added another dimension to this field and both of these are discussed.

15.2.1 DNA index DNA index (ploidy) has been studied extensively (Tribukait, 1984; Friedlander, Hedley and Taylor, 1984) and the spectrum of studies seem to fall into three categories. Those in common or relatively common tumours where a large number of different studies have been performed, those in common tumours where few studies have been performed and those in rare tumours. The first category includes tumours of the breast, ovary, colo-rectum and non-Hodgkin's lymphoma. This almost certainly reflects the feeling that we could be doing better with these tumours if we had better prognostic factors, an important point I will return to later. The second group includes lung cancer (one archival study up to the end of 1988) where the general impression is that there is little point in obtaining better prognostic factors as they won't make any difference. The last group comprises rare tumours which are of considerable academic interest. However, it was only with the advent of nuclear extraction from archival biopsies (Hedley et al, 1983) that such studies became practicable as the assay could be carried out in reasonable numbers of samples where the outcome is known. I'm not going to embark on an exhaustive review of this topic, and the following references are included to complement those in the remainder of this section: Alanen et al, 1989; Auer et al, 1984; Barlogie et al, 1978; Ewers et al, 1984;

352

APPLICATIONS IN ONCOLOGY

Fossa et al, 1984; Hedley, Friedlander and Taylor, 1985; Horsfall et al, 1986; Jakobsen et al, 1984; Lakhanpal et al, 1987; McGuire and Dressier, 1985; Merkel, Dressier and McGuire, 1987; Van den Ingh, Griffioen and Cornelisse, 1985; Volm et al, 1985; Wolley et al, 1982. The best overall agreement has occurred in ovarian cancer where a number of studies have reported that DNA index is an important prognostic determinant (Blumerfeld et al, 1987; Rodenberg et al, 1987; Friedlander et al, 1988; Kallioniemi et al, 1988b). Moreover, Kallioniemi et al {1988a) have shown that DNA index relates to histological features and have indicated that the proportion of cells with S-phase DNA content (referred to subsequently as S-phase fraction) may also be an important determinant (Kallioniemi et al, 1988b). Rodenberg et al. (1988) have carried out morphometric analyses with DNA index and found agreement and good prognostic significance. Trebeck et al (1987) have studied Brenner ovarian tumours and found elevated DNA indices in higher grade tumoups and some evidence suggests that the rare borderline malignancy tumours with elevated DNA indices are the more aggressive (Friedlander et al, 1984). Our group has carried out DNA index determinations in archival biopsies from 225 patients with ovarian cancer of epithelial origin, 6 patients with borderline low potential malignancy and 12 normal ovaries. All normal ovaries had diploid DNA content as did 5 out of 6 cases of 'borderline7 malignancy. The majority of cases of carcinoma, 170 out of 225, were aneuploid. There was a statistically significant difference in the distribution of aneuploidy, p < 0.001, between invasive carcinoma and normal ovaries, and between invasive carcinoma and those classified as borderline low potential malignancy, p < 0.025. There was no difference between normal ovaries and those of borderline malignancy, p>0.1. The five-year survival of patients with diploid tumours was 70% which compared with 22% for patients with aneuploid tumours, p < 0.001, and these data are shown in figure 15.2. The group of patients with well differentiated diploid tumours (n = 39) had the best prognosis with a five-year survival of 82%. In contrast, patients with poorly differentiated aneuploid tumours (n=119) had a very poor prognosis with a five-year survival of only 10%. The 50% survival probability of the group of patients with moderately and poorly differentiated diploid tumours {n = 16) was comparable to the value of 47% in the group with well differentiated aneuploid lesions (n = 51), see figure 15.3. These results are fairly typical for ovarian carcinoma. The position in breast cancer seems to be less clear. Cornelisse et al (1987), in a large series of over 560 patients, found that DNA index was a weak but significant prognostic factor. In contrast, in another large series of 354 patients there was no correlation between prognosis and DNA index (Dowle et al, 1987). Furthermore, Hedley et al (1987) with 490 patients and Kallioniemi et al (1987) with 93 patients reported that S-phase fraction was a more significant prognostic variable than DNA index. In advanced disease both hormone responsiveness (Stuart-Harris et al, 1984) and chemosensitivity (Masters et al, 1987) correlated with DNA index. In general, the breast cancer data suggest that proliferation is more important prognostically than the DNA index.

PROGNOSIS

353

0.2

0.0

1

2

3 Time (years)

4

Figure 15.2. Survival curves for ovarian cancer patients with normal DNA index (Dip) and elevated DNA index (Anu).

The overall results in colo-rectal carcinoma are somewhat similar to those in breast cancer. Armitage et al. (1985), Kokal et al. (1986) and Scott et al. (1987) have reported DNA index as an independent prognostic variable. Bauer et al. (1987) report that proliferation is the more important factor compared with DNA index in early stage disease, Quirke et al. (1987) suggest that both DNA index and S-phase fraction have prognostic implications in rectal cancer, Chang, Enker and Melamed (1987) in a small study of 30 patients report that the DNA index predicts for recurrence in rectal tumours and Blijham et al. (1985) found that both DNA index and S-phase fraction were prognostically significant in Duke's stage C disease. However, Finan et al. (1986) have shown that the prognostic significance of DNA index is lost in patients with widespread hepatic metastasis, a result which is not altogether surprising. DNA index (Banner et al., 1987) and both DNA index and Sphase fraction (Walloch, Solans and Herman, 1987) are reported to correlate with development of adenomas. These results were not unexpected as the proliferative activity of adenometa is increased compared with normal mucosa.

APPLICATIONS IN ONCOLOGY

354

n=39

2. "5 3

O 0.4

0.2

0.0

1

2

3 Time (years)

4

5

Figure 15.3. Survival curves for ovarian cancer patients according to histological grade and DNA index. Dip and Anu respectively denote normal and elevated DNA index. W and P signify well and poorly + moderately differentiated tumours respectively.

Non-Hodgkin's lymphoma, albeit relatively uncommon, has received fairly considerable attention. Bauer et al. (1986b) and Young et al. (1987) both report that S-phase fraction predicts for a bad prognosis and Macartney et al. (1986) found a greater risk of developing more clinically aggressive disease in the higher S-phase fraction tumours. In non-endemic Burkett's lymphoma the high DNA index tumours carried the worst prognosis (Lehtinen et al., 1987). Lung cancer, one of the commonest and most depressing tumours to treat has received relatively little attention, possibly for the reasons given in the introduction to this section. In the one archival study up to the end of 1988 which comprises 100 non-small cell tumours it was found that DNA index was a powerful prognostic factor (Zimmerman et al., 1987). DNA index studies have been carried out in a number of other tumours (recently reviewed by Hedley, 1989) including those arising in the ENT region

PROGNOSIS

355

(Kaplan et al, 1986; Tytor, Franzen and Olofsson, 1987; Sanders et al, 1988; Grace et al, 1988), malignant melanoma (Meecham and Char, 1986; Coons et al, 1987) and adrenal tumours (Bowlby, DeBault and Abraham, 1986; Amberson et al, 1987; Hosaka et al, 1988) where the results were generally consistent. DNA index and/or S-phase fraction correlated with prognosis and histology where the more poorly differentiated neoplasms showed elevated DNA indices. In carcinomas of the urinary system there have been some unexpected results though insufficient studies have been carried out to date. In bladder cancer, Murphy, Chandler and Trafford (1986) and Blomjous et al (1988) found correlation between DNA index, tumour grade and prognosis; however, Jacobsen et al (1987) have reported that a normal DNA index carried a poorer prognosis in advanced tumours. In renal adenocarcinoma Rainwater et al {1987b) in a series of over 200 cases found that DNA index was prognostically significant in well differentiated tumours, but Ekfors et al (1987) found that DNA index did not relate to survival in 96 cases. Some interesting results have arisen in studies of the blastomas. Tomita et al (1988) have found that elevated DNA index is a good prognostic factor in medulloblastoma. In neuroblastoma, Gansler et al (1986) have reported that elevated DNA index and low S-phase fraction are prognostically favourable. Results in nephroblastoma have shown that tetraploid tumours (DNA index = 2.0) carry a worse prognosis than diploid and aneuploid tumours (Rainwater et al, 1987a) and Schmidt et al (1986) correlated S-phase fraction and DNA index with aggressive behaviour. As a general summing up we can state that DNA index and/or S-phase fraction are important prognostic factors in most cancers. However, it should not go unmentioned that this was strongly suggested by the pioneering work of N. B. Atkin at the Gray Laboratory, Northwood, who measured DNA content in populations of individual nuclei with manual light absorption microscope techniques two decades ago (Atkin, 1972) which was well before flow cytometric techniques became generally available. Atkin and Kay (1979) were also the first to introduce the concept of multi-variate analysis for such studies as they had recognized that DNA index alone (modal DNA as they called it) was insufficient to make an absolute distinction between normal and malignant and between good and bad prognostic groups of patients. Recently, van der Linden et al (1989) have reported that multi-variate prognostic index analysis in breast cancer including nodal status, tumour size, mitotic activity plus DNA index gave better prognostic information than DNA index alone. The two major advances in this field in the past 10 years have been in automation with its more general availability and nuclear extraction from archival biopsies enabling retrospective studies to be carried out in even very rare tumours. However, although more assays can now be carried out more quickly than in Atkin's time we are still confronted with a number of technical and interpretive problems. Firstly, with the increased number of cells that can be analysed it is possible to obtain estimates of the fraction of cells with S-phase DNA content. I

356

APPLICATIONS IN ONCOLOGY

would again, however, draw your attention to the distinction between percentage of cells with S-phase DNA content and percentage of cells in S-phase (see section 11.5.5). Perhaps more striking correlates would be obtained in prospective studies in which patients were to be given BrdU before biopsy as this is the only way to make this distinction. However, this would entail considerably more work than wax block extraction and would take very much longer. Secondly, the S-phase fraction, particularly if this is 'low', can be difficult to estimate reliably even with the 'best' computer software. This is due to a number of problems including overlapping of mixed diploid and aneuploid populations which is compounded further in archival material where thick sections have been cut as some nuclei will have been sliced. Furthermore, some of the commercial computer software packages are really not sufficiently sophisticated to cope with these types of data sets. I suspect that published reports in this field which have only addressed DNA index and not S-phase fraction were limited by software capability. Nevertheless, if the S-phase fraction can be estimated reliably and if it gives meaningful correlations which can be used clinically it may not be necessary to make the distinction between fraction of cells with S-phase DNA content and percentage of S-phase cells.

15.2.2 Oncogenes It has been my opinion for some time that DNA index and/or S-phase fraction will not be sufficiently reliable to make good enough distinctions between good and bad prognostic groups to influence patient management. Because of this, which I freely admit is a prejudice and not based on scientific considerations, our group embarked on studies to see if measurement of oncogene-encoded proteins could give additional information. In order to 'set the scene' I'll begin with a brief review of this field. The human genome contains a number of genes which may be involved in the aetiology of cancer (Bishop, 1985; Varmus, 1984; Slamon, 1987). These segments of DNA, the proto-oncogenes, came to light from the pioneering work of Peyton Rous (1911) who discovered that chicken sarcomas could be transmitted by cellfree extracts. In 1951, Gross showed that a retroviral agent was responsible for a murine leukaemia and since then a number of transmittable tumours in higher vertebrates have been shown to be due to retroviruses including Rous sarcoma. The retrovirus genome is composed of RNA and is small with only three genes. These are called gag, pol and env and they code for core protein, a reverse transcriptase (Baltimore, 1970; Temin and Mizutani, 1970) and the viral envelope respectively. When a cell is infected, the viral RNA is handled in exactly the same way as the cell's normal RNA and it is translated to protein. The reverse transcriptase product of the pol gene can now splice a new segment of DNA into the host cell genome which is complementary to the viral RNA. Many rounds of transcription are induced and viral RNA identical to that initially infecting the cell is produced. Core protein and envelope are manufactured simultaneously and amplification of the virus takes place which is released on rupture of the cell.

PROGNOSIS

357

Not all retroviruses are oncogenic, but most that are contain an extra gene. These 'one-genes have three-letter acronyms, for example sre, erb and myc stand for Rous sarcoma, chicken erythroblastosis and chicken rat/elocytosis respectively. The retroviruses tended to be regarded as curiosities with little or no relevance to human cancer until the key discovery was made by Stehlin et al. in 1976 where it was shown than normal cells contain segments of DNA which are structurally very similar to the retroviral v-src gene. After this progress was rapid and many of the retroviral oncogenes were found to have cellular counterparts which were very similar though not identical (Bishop and Varmus, 1982). Furthermore, not all oncogenes have retroviral homologues. This had lead Coffin et al. (1981) to propose that the identifiers of the retroviral oncogenes should be prefixed with V-' to distinguish them from cellular oncogenes to be prefixed with 'c-'. We now know that one class of oncogenes encode proteins which are intimately involved in growth regulation at the levels of extracellular signal transmission (c-sis, platelet-derived growth factor, Doolittle et al., 1983; Waterfield et al, 1983); signal reception and transduction across tne external membrane (c-erbB, epidermal growth factor receptor, Downward et al, 1984, and c-fms, receptor for colony stimulating factor 1, Roussel et al, 1984; Scherr et al, 1985); intracellular signal transmitters (N-ras, Wakelman et al, 1986, product related to G-proteins, Hurley et al, 1984) and nuclear signal receivers and/or transducers (v-erb-A, thyroid hormone receptor, Sap et al, 1986; Weinberger et al, 1986, which has extensive homology with oestrogen receptor, Green et al, 1986). Further likely candidates for nuclear transducers are the proteins encoded by myc, fos and myb, which have DNA binding capacity and p55c~fos has been shown to be involved with transcription (Distel et al, 1987; Lech et al, 1988). A second class of 'oncogenes' has now been identified as tumour-suppressor genes where loss of normal function is associated with tumour formation. This possibility, which in engineering terms would be described as the failure of a negative feedback servo control loop mechanism, was considered by Comings (1973) and has now been identified in a number of hereditary human tumours (Knudson, 1983, 1985). The first suspicion that hereditary factors might be involved in cancer was voiced by Norris (1820) based on observations of what was probably malignant melanoma in successive generations of the same family. Familial polyposis coli (McKusic, 1962; Ashley, 1969) carries a well defined risk of carcinoma development which, at a given age, is about three orders of magnitude greater than that in the remainder of the population. Another good example is retinoblastoma (Knudson, 1971, 1978) where a visible deletion involving band 14 on the long arm of chromosome 13 (13ql4) is found in a small proportion of cases (Knudson et al, 1976; Yunis and Ramsay, 1978). It is now established that the genetic abnormalities in retinoblastoma are not random (Benedict et al, 1983a) and that predisposition is a recessive inheritance via one affected and one normal chromosome 13 with the gene located at 13ql4 (Benedict et al, 1983b; Cavenee et al, 1983). Development of the neoplastic phenotype is a consequence of somatic mutation at 13ql4 in the normal chromosome and hence complete loss of the

358

APPLICATIONS IN ONCOLOGY

retinoblastoma suppressor gene function. Increasing evidence suggests that p53, discovered a decade ago (Lane and Crawford, 1979; Linzer and Levine, 1979) also has tumour-supressor function (Wolf and Rotter, 1985; Masuda et al, 1987; Munroe et al, 1988; Finlay, Hinds and Levine, 1989; Hinds, Finlay and Levine, 1989; Banker et al, 1989) where mutation of the gene or conformation changes in the normal protein contribute to transformation. The fundamental pathological diad of cancer is disordered proliferation and the capacity of malignant cells to metastasize. The genes discussed so far are obviously concerned with proliferation control and probably also with control of differentiation. Further discoveries have potentially linked the metastasis phenomenon with disordered functioning of oncogenes encoding cytoskeleton elements. The v-fgr gene encodes a hybrid protein containing a portion of the actin molecule (Naharro, Robins and Reddy, 1984) and onc-D codes for a non-muscle tropomyocin (MartinZanca, Hughes and Barbacid, 1986). This would seem to be highly significant as it might signify a third category of 'oncogenes7, those concerned with sub-cellular architecture and hence possibly with metastatic potential, to complement those involved with proliferation. Most studies involving the oncogenes carried out to date in both tissue culture and in human cancer have relied upon hybridization techniques. These methods, particularly for mRNA, require fresh tissue which is not always obtainable in sufficient quantities from cancer patients. Also, it may take a considerable time to accumulate sufficient material and information to make meaningful clinical correlates with fresh tissue. This applies particularly to the rare tumours. Moreover, neither the gene nor its message is the effector molecule; this is the province of the protein and monoclonal antibodies directed to specific oncoproteins have been developed. Examples include p53 (Harlow et al, 1981), p21 c ~ras (Furth et al, 1982) and p62c~myc (Evan et al, 1985). Herein lie possibilities for clinical applications using protein probes for not only body fluids but also biopsy material. Studies with the former indicate the potential for oncoproteins to become a new generation of tumour markers. Crawford, Pirn and Bulbrook (1982) have detected antibodies against the cellular protein p53 in sera from patients with breast cancer. Antipeptide antibodies have detected and compared oncogene-related proteins in urine in normal subjects, pregnancy and in cancer patients (Niman et al, 1985). Similar studies have been conducted using an antibody to the c-myc protein. A 40 000 da breakdown product of p62c~myc was detected in cancer patients and in pregnancy (Chan et al, 1987). Such antibodies have been used not only for immunoprecipitation and Western blotting, but also for immunocytochemical localization in biopsies of both frozen and formalin-fixed normal and tumour tissue. They have been used to define differential ras gene expression in malignant and benign colonic disease (Thor et al, 1984). Slamon et al (1989) studied HER-2/neu expression in breast and ovarian cancers and found good correlation between immunocytochemical staining, Western blotting and both DNA and mRNA hybridization signals. However, these investigators stress that it is important to identify which antibodies can be

PROGNOSIS

359

1.0

1 copy 2-5 copies >5 copies

0.8 0.6 CD

_

0.4 0.2

:

CO

1

•

I

1

2000

4000

6000

CO

1.0

J5

D

o

—

\

—

0.8 \]

Negative i + 2+ >2+

0.6 •

0.4

:

I

i

L

"-

0.2 :

0.0 0

i

2000

4000

6000

Time (days) Figure 15.4. HER-2/neu expression and survival probability in ovarian cancer where the top panel is for gene copy number and the bottom panel is for the protein product assayed immunocytochemically. used in formalin-fixed biopsies, and frozen sections had to be used for their immunocytochemistry as completely negative staining was observed after formalin fixation. Patient survival correlated inversely with HER-2/neu gene amplification and its protein expression in ovarian cancer and these data are reproduced in figure 15.4. Immunocytochemical studies with an anti-p62 c ~ mK monoclonal antibody, Myc 1-6E10, have shown that normal testis expressed only small amounts of the protein. Seminomas exhibited increased nuclear and cytoplasmic staining. Undifferentiated teratomas showed barely detectable staining whereas well differentiated epithelial structures and yolk-sac elements exhibited intense staining (Sikora et al, 1985). Normal colonic mucosa exhibited maximal expression in the maturation zones of the crypts of Lieberkuhn where there was mixed nuclear and cytoplasmic staining (Stewart et al, 1986). However, p62c~myc is known to be nuclear associated (Evan and Hancock, 1985) and in further more extensive studies it was shown that this protein is redistributed from a nuclear to a cytoplasmic location with increasing maturation of the normal colonic mucosa (Sunderesan et al, 1987). This has also been seen in mucosa freshly

APPLICATIONS IN ONCOLOGY

360

fixed within seconds of biopsy (Forgacs et al, 1986) and active exclusion of the protein from the nucleus may be part of the normal control mechanism for regulating proliferation and differentiation. In familial polyposis coli, which inexorably progresses to the malignant invasive phenotype, nuclear staining persisted to the surface of the crypts and in carcinomas the staining was predominantly nuclear (Sunderesan et al, 1987). Immunocytochemical techniques have the advantage that the architecture is preserved but they suffer from the disadvantage of poor quantitation; however, the careful work of Slamon et al. (1989), see figure 15.4, was able to effect a discrimination into low', 'medium' and 'high' HER-2/neu protein expression which had meaningful clinical correlates. Fluorescence quantitation of p62 c~myc and DNA simultaneously was developed in our laboratories to study a number of tumours using flow cytometry in an attempt to overcome some of these problems (Watson et al, 1985b). The technique was adapted from that of Hedley et al (1983) and uses pepsin to release nuclei by cytoplasmic digestion after dewaxing and rehydration (see section 7.1.3). As the c-myc protein is nuclear associated (Evan and Hancock, 1985) it can be stained with the Myc 1-6E10 antibody after nuclear 1000H

800

o

O

o o

c


O

o

—o—

600-

CM

o

a

o

(0

—o—

NOR

SEM

MTU c myc

MTI

MT+YS

Figure 15.5. Levels of p62 ~ in normal testis (NOR) and testicular cancer. Both semonomas (SEM) and teratomas exhibited elevated levels. In the latter the p62c~myc increased with increasing differentiation from MTU (malignant teratoma undifferentiated) to MTI (malignant teratoma intermediate) to MT + YS (malignant teratoma with yolk sack elements).

PROGNOSIS

361

extraction. However, very great care must be exercised with this method for the reasons given in section 7.3.4 and if you've forgotten what they are please read that bit again. Low p62c~myc levels were found in normal testicular tissue and significantly elevated levels were found in both teratoma (p < 0.001) and seminoma (p < 0.001). However, the oncoprotein level increased significantly with increasing differentiation in teratoma (p<0.01). These data are shown in figure 15.5. Patients who developed recurrence and died within three years of diagnosis had lower levels than those who were disease free since their initial treatment, p < 0.001 (Watson et ah, 1986) and these results are shown in figure 15.6. Somewhat similar results were obtained in colon neoplasia (Watson et ah, 1987c). Normal mucosa and polyps from non-malignant specimens exhibited lower p62c~myc levels than carcinomas (/?<0.02) but there was a wide spread in the latter as can be seen in figure 15.7.

1000n

800
600-

o CM CD

a

400-

200-

A/W

R/D

Figure 15.6. Comparison of p62c~myc levels in testicular cancer patients who developed recurrence and died (R/D) within three years of diagnosis with those who were disease free (A/W) at three years. Seminomas and teratomas are represented by the open and closed symbols respectively.

362

APPLICATIONS IN ONCOLOGY uuu-

•

i

800o

E

600-

o CM CO

a

•

•

>

i

400-

1 200-

T

f

|

N HISTOLOGY Figure 15.7. Increasing expression of p62c~myc mucosa to polyps then to carcinomas.

levels from normal colonic

Villous adenometa (pre-malignant) exhibited elevated levels as did morphologically normal mucosa from specimens containing a carcinoma. There was also a tendancy for p62c~myc levels to fall as the tumours became more poorly differentiated, a finding which was corroborated in parallel studies with Northern blotting for mRNA (Sikora et al, 1987). Comparison of normal with dysplastic mucosa and carcinomas developing in patients with ulcerative colitis have shown that the nuclear p62c~myc content increased with the transition from 'mild' to 'severe7 dysplasia (Forgacs et al, 1986), results which are shown in figure 15.8. Moreover, the protein content was raised in morphologically normal colonic mucosa derived from malignant compared with non-malignant specimens (Watson et al, 1991b). We have also studied some gynaecological malignancies using these methods and aneuploidy with a well defined Gl peak was not observed in a total of 127 biopsies of uterine cervix neoplasia (Hendy-Ibbs et al, 1987). Normal biopsies exhibited higher p62c~myc levels than carcinomas, p< 0.000 01, and there was a progressive decrease in oncoprotein level with progression from cervical intraepithelial neoplasia, CIN I to CIN III, p<0.05 as shown in figure 15.9. Furthermore, the maximum fluorescence signal in the normal tissues occurred at a lower antibody concentration compared with tumour tissue, p<0.000 01. There

PROGNOSIS

363

70006000o o § 5000-

o

(0
2 400030002000H 1000-

NORMAL CHRONIC COLON U.C

MILD SEVERE DYSPLASIA

CARCINOMA

Figure 15.8. Comparisons of normal colonic mucosa, quiescent mucosa in ulcerative colitis, dysplastic mucosa in the latter mild and severe and invasive carcinomas developing in patients with ulcerative colitis.

was no correlation with histological grade, stage, age or prognosis in patients with invasion. Serious papillary ovarian carcinoma expressed significantly higher p62 c " m y c levels compared with normal ovary, p < 0.000 03 (Watson et al., 1987a). Biopsies classified as 'borderline low potential malignancy' exhibited levels between normal ovary and carcinoma. The difference between normal and borderline was significant at p < 0.003, but no difference between borderline and frankly invasive biopsies was observed, p = .149. There was no difference between the histological grades of carcinomas. All normal ovaries had diploid DNA content as did 5 out of 6 cases of 'borderline' malignancy. The majority of cases of carcinoma, 28 out of 36, were aneuploid. There was a statistically significant difference in the distribution of aneuploidy, p < 0.005, between invasive carcinoma and those classified as borderline low potential malignancy. These results are summarized in figure 15.10 which shows p62c~myc levels for normal, borderline and invasive carcinoma. It would seem that elevation of p62c~myc preceded the development of aneuploidy in the evolution of the malignant phenotype in ovarian cancer; however, there is no evidence to suggest that these are causally related. The findings in colonic neoplasia, seminoma and ovarian cancer were partially anticipated but those in both carcinoma of the cervix and in the histological breakdown of teratoma were completely contrary to initial expectation. However, an increase in either the gene or mRNA copy number (or both), which should give

APPLICATIONS IN ONCOLOGY

364

2520-

10 50 10 (0

c

I & o •o

I]

5

0 10 5

-P.

Figure 15.9. Decrease in p62c ~ myc levels with 'progression' from normal cervical mucosa (panel A) to invasive carcinoma (panel E). Panels B, C and D were obtained from patients with intraepithelial neoplasia, CIN I, CIN II and CIN III, respectively. rise to an increased protein production rate, need not necessarily be reflected in a marked increase in the total protein content for two main reasons. Firstly, inappropriately increased message may result in rate limitation at the protein synthesis level. Secondly, an increase in protein degradation may offset an increased production rate. The latter is most likely to occur with a protein which has a short half-life and, clearly, this is a distinct possibility for p62c~myc with a half-life of 20—30 minutes in rapidly cycling and stimulated cells (Greenberg and Ziff, 1984; Rabbitts et al., 1985). Hence, the lower absolute levels in carcinoma of the cervix compared with normal and undifferentiated compared with the better differentiated teratomas, may reflect an increased protein turnover and an increased cell production rate in the former. Further possibilities include posttranslational protein modification in the more malignant teratomas and in cervical

PROGNOSIS

365

7000H

6000-

g 5000 c
o

•

• I

0)

O 4000-

IP

o E 3000-

_

<0

t t

a D

2000-

° I

8

D O

o

B

°

O

2 1000-

N

B

C

Figure 15.10. Levels of p62c ~ myc in normal ovaries (N), borderline tumours (B) and invasive carcinoma (C). The closed circles represent lesions with elevated DNA indices. The normal ovary square symbols represent the oncoprotein levels in elements predicted to be normal cells in tumour specimens where this component constituted more than 25% of the whole specimen. carcinoma giving rise to an alteration or partial occlusion of the epitope recognized by the antibody and a possible increase in the susceptibility of the protein to proteolysis in neoplastic cells in the preparation for the assay. There is a small shred of evidence that post-translational protein modification may occur in cervical carcinoma. Maximum binding was observed at different antibody concentrations in the normal and malignant cells which might indicate a change in binding constant. An interesting feature of the cervix study was the apparent absence of aneuploidy. However, 30% of neoplastic specimens had a positive skew to the diploid spike which could have been due to a small proportion of cells with a 5 to 10% increase in DNA content although in no specimen was there a well defined second peak. These may have represented aneuploid components which the method was not able to resolve in these particular specimens.

366

APPLICATIONS IN ONCOLOGY

We have seen in section 15.2.1 that DNA index can be a bad prognostic factor in a number of tumour types but not all patients with diploid tumours have a good prognosis. Part of our work to quantitate oncoproteins simultaneously with DNA was attempting to help with this distinction and indeed highly significant differences in p62c~myc levels were observed between good and bad prognostic groups of testicular cancer patients. However, inspection of figure 15.6 reveals considerable overlap and it would not be possible to assign patients with certainty to good and bad prognostic groups with these data even though 'p was less than 0.001. As can be seen the problem lies with the low values which completely overlap in the good and bad prognostic group. The results obtained in ulcerative colitis (figure 15.8) are more encouraging where a considerable increase in p62c~myc levels occurred with progression from mild to severe dysplasia which from a clinical standpoint is exactly where the physician would like to be alerted that something is going wrong. However, we have to ask if the p62 c ~ myc levels are telling us anything we don't already know. I suspect that they are not as the histological distinction between mild and severe dysplasia is relatively easy to make histologically. Perhaps we should look in detail at the mild dysplasias with high p62 c ~myc levels in prospective studies to see if this gives us an indication that the mucosa is in the process of turning to severe dysplasia. Nevertheless, these results were encouraging. The ovarian cancer results of Slamon et al. (1989) with immunoperoxidase staining for HER-2/neu are also encouraging. However, there are many more parameters which could be measured. These include a whole battery of nuclear-associated oncogene-encoded proteins, quite apart from p62 c ~mycf and cell surface receptor status (EGFr and oestrogen receptor; Sainsbury et al, 1985) which may provide additional data in a number of tumour types to make more reliable distinctions between good and bad prognostic groups of patients.

15.3 Therapy selection It is quite clear from the types of studies and results carried out to date, and outlined in the previous two sections, that multi-parametric information with multi-factorial analysis will be required to effect a sufficiently reliable predictive discrimination between good and bad prognostic groups to influence clinical therapeutic judgements. I do not believe it is justifiable at present, or that it will be in the future, to base therapy judgements merely on univariate analysis of DNA index data. Moreover, even when we can distinguish absolutely between good and bad prognostic groups, and I believe we will be able to do this, it's not entirely clear what we are going to do with the patients in the bad prognostic groups as, apart from some rare exceptions, we tend to use our most aggressive therapy at the outset and treat to established normal tissue tolerance. If assays, or assay systems, are developed which can tell us that a particular 'biochemical profile' of a tumour with a particular therapy regime is going to give us close to 100% success then there is no need to change the treatment. However, this is probably not a

THERAPY SELECTION

367

realistic possibility for the majority of human cancers with the currently available treatment strategies. It seems to me to be a relatively pointless exercise in becoming better and better at predicting the good prognostic groups of patients whose numbers will become smaller and smaller if we cannot improve therapy for the bad prognostic groups whose numbers will become bigger and bigger as a consequence of the improved classification. Reclassification does not, in itself, improve results as, by definition, it just places patients in different categories. This may make us feel more comfortable when we look at the results from the good prognosis groups but less comfortable with the results from the bad prognostic groups which alerts us to the problem. One good example is the redefinition of the lymphomas into Hodgkin's and non-Hodgkin's categories and the substratifications into low and high grades. This gave rise to more aggressive chemotherapy for the high grade non-Hodgkin's tumours and somewhat less aggressive therapy for the low grade tumours in each major subdivision. If we are to make comparable advances in the more difficult 'solid' tumours, the big three Bs, breast, bronchus and bowel, we must think in terms of quantitative comparative subcellular biochemical measurements in normal and malignant cell not only for developing new drugs and therapy combinations but also in attempts to minimize administration of ineffective therapy. The latter would at least spare a large number of patients very considerable toxicity which many, usually the relatives, would consider a positive gain in a very negative environment. The assay systems described in the next three sections are possible ways in which these approaches could be addressed. 15.3.1 GSH metabolism Oxygen can be dangerous stuff when handled incorrectly. This also applies in aerobic biological systems as normal oxygen metabolism produces some peroxides and superoxides during energy releasing steps. These highly reactive short lived species can be very damaging to a number of cruicial targets, including DNA, and aerobic organisms have developed ways and means of coping with these potential threats. One of these mechanism is the tripeptide glutathione containing a sulphydryl group (—SH—) in the cysteine residue which 'mops-up' reactive radicles. Highly metabolically active cells can contain up to 15% of their total molar content as glutathione. Reduction of intracellular glutathione can increase sensitivity to not only radiation (Bump, Yu and Brown, 1982; Malaise, 1983; Clark et al, 1984; Schrieve, Denekamp and Mitchinton, 1985) but also cytotoxic agents (Babson, Abell and Reed, 1981; Roizin-Rowle etal, 1982; Ono and Schieve, 1986). These observations have potentially important clinical implications as exemplified by the very limited work we have carried out to date. The normal range for GSH estimation was established for lymphocytes from a number of normal individuals in our laboratories using mClB as described in section 14.2.4 and two patients with leukaemic central nervous system involvement have also been studied. These results are given in figure 15.11 where the peripheral lymphocytes from the two

APPLICATIONS IN ONCOLOGY

368

10

100

200

300

400

500

Time (seconds) Figure 15.11. Bluefluorescencedevelopment for the conjugation of GSH with monochlorobimane in normal lymphocytes (closed symbols with 2 standard error limits) and in peripheral lymphocytes from two patients with CNS leukaemic involvement refractory to irradiation. patients are compared with lymphocytes from the normal subjects. Both patients (open symbols) were well above the two standard error limits of the normal range. Both had had intensive chemotherapy previously and both were totally refractory to CNS radiation and died without exhibiting any response. These results are nothing to shout about yet but lymphatic leukaemia lymphocytes are usually sensitive to radiation. This begs the question: had the genes which encode the enzymes responsible to GSH production (y-glutamic acid transferase and glutathione synthatase) undergone adaptive amplification in response to the previous chemotherapy and produced increased intracellular levels of GSH and hence induced radiation resistance? This is a distinct possibility as gene amplification can occur very rapidly under some circumstances. Rice, Hoy and Schimke (1986b) have reported that even transient hypoxia can cause amplification of the dihydrofolate reductase gene. Perhaps in future it would be advisable to assay the lymphocytes of such patients for GSH at intervals during chemotherapy to monitor any changes that might occur and to assay for gene amplification if changes are noted. If this did occur and subsequent CNS involvement developed which required radiation then thiol depleting agents possibly could be employed.

15.3.2 Drug resistance The development of anti-cancer drug resistance was first described 40 years ago by Burchenal et al. (1950). Mechanisms in drug resistance include decreased cellular influx, increased efflux, increased catabolism, increases in competitive inhibitors, changes in the partitioning of drug within intracellular compartments, alteration of intracellular targets, reduction of DNA damage and gene amplification. Some of the transport and resistance mechanisms involving

THERAPY SELECTION

369

anthracyclines and methotrexate have been mentioned or considered in section 14.6 and mechanisms involving some of the other anti-cancer drugs are considered here. Goldenberg et al. (1970) demonstrated that transport of nitrogen mustard (HN2) into cells was mainly an active process mediated by a carrier molecule and their data suggested that HN2-resistant mutants were deficient in the carrier. It was then shown in HN2-sensitive and resistant cells that this was the carrier for choline (Goldenberg, Vanstone and Bihler, 1971). Study of cells sensitive and resistant to chlorambucil have revealed a resistance mechanism which is not related to transport as the kinetics of uptake were identical in the two cell types (Harrap and Hill, 1970). Drug uptake in both resistant and sensitive cell lines was temperature independent, directly proportional to concentration and unaffected by ouabane and metabolic inhibitors all of which strongly support passive diffusion as the means of transport (Hill, 1972). However, alkylating activity declined at over double the rate in the resistant cells with decreased binding to DNA, RNA and proteins (Hill, 1972) which are all indications of increased intracellular metabolism of the drug in resistant cells. Passive diffusion is also the means by which the nitrosoureas, CCNU, BCNU (Begleiter, Lam and Goldenberg, 1977) and chlorozotocin (Lam, Talgoy and Goldenberg, 1980) gain access to the cell interior as the uptake of each is unsaturable, independent of temperature and unaltered by metabolic inhibitors. The esterase inhibition kinetic assay described in section 14.2.7 is a potential means of assessing not only transport of nitrosoureas and related isocyanates across the external membrane but also the damage they do having gained access. Some cells are resistant to nitrosoureas but as yet we have scant knowledge of the mechanisms involved. Schabel et al (1978) have studied resistance to alkylating agents and cross-resistance with nitrosoureas. A cell line totally resistant to cyclophosphamide retained sensitivity to BCNU, CCNU and methyl-CCNU. BCNU-resistant cells exhibited no cross-resistance with cyclophosphamide or melphalan, but a melphalan-resistant line showed partial cross-resistance with chlorozotocin and cyclophosphamide, complete cross-resistance with ris-platin but no resistance to BCNU, CCNU or methyl-CCNU. The uptake of phenylalanine mustard (PAM, melphalan), a derivative of HN2, is an active process but it does not share the choline pathway with HN2. PAM cytotoxicity can be reduced by co-incubation with a number of amino acids, an effect first reported by Vistica, Toal and Rabinowitz (1976) with L-leucine and L-glutamine. This, and further evidence (Vistica, Toal and Rabinowitz, 1978) suggested that PAM and leucine share a common carrier molecule. Melphalan resistance has been noted in a colchicine-resistant clone of Chinese hamster ovary (CHO) cells, CHRC5, initially isolated by Ling and Thompson (1974). These cells were also cross-resistant with nitrogen mustard, chlorambucil, adriamycin and puromycin and Ling, Gerlach and Kartner (1984) used the term multi-drug resistance (MDR) to describe this phenotype. During the past 15 years cross-resistance or multi-drug resistance, which is

370

APPLICATIONS IN ONCOLOGY

thought to be a major cause of failure in cancer chemotherapy, has been recognized increasingly. The phenomenon is characterized clinically by good responses to primary cytotoxic chemotherapy but poor responses on relapse not only to the drugs used initially, but also to others not used in the primary therapeutic schedule. The development of 'classical' multi-drug resistance (MDR) as described by Ling et al (1984) occurred during in vitro selection and a number of mechanisms have been proposed. These include changes in the intracellular drug distribution (Supino et al, 1986), modified membrane transport (Skovsgaard and Nissen, 1982), reduced DNA cleavage (Glisson et al, 1986; Capranico, Dasdia and Zunino, 1986; Supino et al., 1988), altered drug target sites (Potmesil et al., 1988) and the capacity to secrete drugs using energy dependent active efflux mechanisms. The MDR phenotype described by Ling et al. (1984) correlated with the presence of a high molecular weight membrane-associated glycoprotein (Juliano and Ling, 1976; Carlsen, Till and Ling, 1976). Initially, it was postulated that resistance was conferred by reduced drug accumulation due to decreased influx (Ling and Thompson, 1974; See et al, 197A). However, the drugs involved in the MDR phenotype are very different and have different influx pathways all of which would have to be involved. Moreover, enhanced efflux of actinomycin-D, vincristine and vinblastine occurs in adriamycin-resistant P388 cells (Inaba and Sakurai, 1979). These various considerations suggested a common efflux pathway and that MDR may be due to increased efflux mediated via an energy-dependent carrier-mediated transport pathway involving the 170 kda membrane located glycoprotein, gp 170 . Anthracycline secretion has been studied extensively in both sensitive and resistant cells (Dano, 1973; Skovsgaard, 197Sa,b,c; Di Marco, 1978; Inaba et al, 1979) and gp 170 involvement has been confirmed in some (Riordan and Ling, 1985), but not all (Slovak et al, 1988; Mirski, Gerlach and Cole, 1987) cell lines exhibiting MDR. Energy-dependent active efflux is supported by studies with the calcium channel blocking agent verapamil which can partially overcome some resistance and cross-resistance (Tsuro et al, 1982; Twentyman, Fox and Bleehen, 1986a). In the latter studies it was shown that the degree of resistance to adriamycin was indirectly proportional to verapamil concentration. However, there may well be many drug-resistance pathways (Warr, Anderson and Fergusson, 1988) and Beck et al. (1986) have shown that vinka alkaloid resistance but not multi-drug resistance is reversed by verapamil in human leukaemic cells. The gp 170 is encoded by the mdr-1 gene and its expression varies considerably in different normal tissues (Fojo et al, 1987), perhaps reflecting variation in tissue requirements for excretion of toxic substances (Nelson, 1988). Amplification of this gene has been described in a multi-drug-resistant small-cell lung cancer cell line, H69/LX4, (Reeve, Rabbitts and Twentyman, 1989) which was developed in our laboratories from the parent NCIH69/P (Twentyman et al, 1986b). Hybridization techniques (Southern, 1975; Thomas, 1980) and Western blotting can be used to study gp 170 at the gene, message and protein levels; however, these methods have two major disadvantages. Firstly, they cannot assay

THERAPY SELECTION

371

for functional activity. Secondly, they cannot identify minority subsets in heterogeneous populations as only the grand average for the whole population can be determined. Chemotherapy may select for pre-existing multi-drug resistance in a minority subset within a tumour, a view supported by the work of Ross et al (1988) where highly multi-drug-resistant P388 cells were isolated from a drug-sensitive population by flow cytometric cell sorting. It is very unlikely that this minority subset would be apparent using blotting techniques. For example, if there was mdr-1 gene amplification of 10-fold in a minority subset comprising 1% of a tumour, which is not an unlikely scenario, we would obtain a 9% increase in hybridization signal in the tumour specimen compared with a control. Southern hybridization techniques could not hope to make this distinction. The same arguments apply to similar analysis of mRNA. Furthermore, message copy number need not relate quantitatively to the number of gp 170 molecules in the membrane as translation to protein could be rate limited particularly in the unfavourable environments in which tumour cells frequently find themselves. The need to develop techniques to overcome these problems is self-evident. However, apart from the requirement to measure the functional activity of gp 170 in minority subsets of heterogeneous samples it would also be desirable to be able to perform the assays on relatively small biopsy samples in viable cells under near physiological conditions. Two flow cytometric kinetic methods which meet these requirements are now available. One is 'indirect' and due to Morgan et al (1987, 1989) and the other is 'direct' and due to Herweijer, van den Engh and Nooter (1989). The 'indirect' assay involves the use of the bisbenzimidazole DNA dye Hoechst 33342 whose cellular uptake parallels that of a number of cytotoxic agents involved in multi-drug resistance and hence correlates with sensitivity to those agents (Lalande, Ling and Miller, 1981). Furthermore, Hoechst 33342 uptake is modified by verapamil (Krishan, 1987) and it is possible that the dye is secreted by cells via the gp 170 pathway. We saw in sections 8.4.3 and 11.8 that the fluorescence emission spectrum from the DNA/Hoechst 33342 complex varies in different cell types (Watson et al, 1985a; Smith et al, 1985). It is dependent on the dye: DNA—phosphate ratio and pH (Smith et al, 1985) and it has been used as a means of sorting viable cells on DNA content (Arndt-Jovin and Jovin, 1977; Hamori et al, 1980; Lydon et al, 1980). These considerations prompted an investigation of the possibility of using Hoechst 33342 as a reporter probe for MDR in viable cells by monitoring the fluorescence emission spectrum on an individual cell basis. It was anticipated that this analysis might be more demanding in MDR cells than in bone marrow or thymocytes (which it was, see sections 8.4.3 and 11.8) and the instrument was modified to give greater resolution. Data were collected from a minimum of three photomultipliers each guarded by 10 nm narrow band-pass filters for simultaneous analysis in the violet (395-405 nm), blue (495-505 nm) and red (595—605 nm) regions of the spectrum. Some studies also included the addition of the indigo (445—455 nm) and green (545—555 nm) channels and measurement

372

APPLICATIONS IN ONCOLOGY

on such narrow band-passes required the high-efficiency light collection flow chamber (Watson, 1985). Forward and 90° light scatter signals were also collected as a means of helping to identify debris and clumps of cells (Watson et al., 1985b). The cell throughput rates were reduced to between 250 and 350 cells per second and the data were collected list-mode on disc which were later processed with a VAX S600 computer. The photomultiplier voltages, amplifier gains and thresholds of the instruments were set using control samples stained for 1 hour so that the easily defined Gl peak was placed at the same scale position on each photomultiplier analysis channel. Subsequent fluorescence from the test samples could then be compared with the control on each analysis channel to give a measure of relative changes in the emission spectrum. A total of 10 4 cells was analysed for each data set. A number of multi-drug-resistant cell lines have been developed in our laboratories and these were used in the investigation. Single-cell suspensions were obtained by treating wth 0.4% trypsin plus 0.02% versene and the concentrations were adjusted to 2 x 105 cells ml""1 in growth medium. A stock solution of Hoechst 33342 was dissolved in distilled water at a concentration of l m M . Aliquots of cells were stained by adding 10 jil of the stock solution to 1 ml of the cell suspensions to give a final stain concentration of 10 |iM. The cells were then incubated at 37 °Cin the dark. Verapamil, when this was used, was added to give a final concentration of 3.3 |iM. Time courses for the development of violet (400 nm, left panel) and red (600 nm, right panel) flourescence in the H69/P and LX4 cell lines are shown in figure 15.12 where the closed and open symbols represent the results from the sensitive (H69/P) and resistant (LX4) cell lines respectively. Note there is very much faster 400 nm

120

80

40

20

40 60 Time (minutes)

0

600 nm

r Y 20

40 60 Time (minutes)

Figure 15.12. Time courses for the development of violet (400 nm) and red (600 nm) fluorescence in the H69/P and LX4 cells lins. The closed and open symbols represent the results from the parent (H69/P) and resistant (LX4) cell lines respectively. The error bars were drawn at the 95% confidence limits.

THERAPY SELECTION

2>73

1-2r O O

0.8

o (0

o 0.4 o

2o LL

0

20 40 Time (minutes)

60

Figure 15.13. Red: violetfluorescenceratios calculated from the data in figure 15.12 where the closed and open symbols again represent the sensitive (H69/P) and resistant (LX4) cells. The lower ratio in LX4 is a manifestation of the violet bias, and hence lower dye: DNA-phosphate ratio, in these cells. development of fluorescence in the parent and a considerable violet (400 nm) bias in LX4 compared with its red {600 nm) fluorescence. The instrument was set up with the parent cell line as the control and any differences in total fluorescence intensity arising from differences in cellular DNA content can be 'normalized' by taking the red: violet ratio. This is shown in figure 15.13 with the closed and open circles again representing the sensitive (H69/P) and resistant (LX4) cells. The lower ratio in the latter clearly demonstrates the considerable violet bias to the emission. Figure 15.14 shows a time course for the development of fluorescence at three wavelengths (600 nm, circles; 500 nm, triangles; and 400 nm, squares) for H69/P and LX4 in left and right panels respectively where verapamil was added to both after 25 minutes exposure to Hoechst 33342. The solid symbols represent the results following verapamil treatment. There was no difference in the sensitive cell line (left panel) but there was a pronounced change in the LX4 multi-drug-resistant cell line (right panel) with considerable increase in fluorescence at all wavelengths. The red: violet ratios calculated from the data of figure 15.14 are shown in figure 15.15 where the red shift after verapamil addition in the LX4 cells (triangles) is very apparent; however, there was no change in the sensitive parent (circles). This technique for assessing gp 170 MDR depends on only two basic factors, namely the violet bias to the fluorescence emission from the DNA/Hoechst 33342 complex at low dye: DNA-phosphate ratios and the postulate that the dye is pumped out of cells by gp 170 . Cells with increased numbers of gp 170 molecules in the membrane should, therefore, maintain a lower intracellular Hoechst 33342

APPLICATIONS IN ONCOLOGY

374

LX4

H69

Verapamil 140

Verapamil 140

100

20

40 60 Time (minutes)

20

40 60 Time (minutes)

Figure 15.14. The time course for the development of fluorescence at three wavelengths (600 nm, circles; 500 nm, triangles; and 400 nm, squares) for H69/P and LX4 in the left and right panels respectively. Verapamil was added after 25 minutes exposure to Hoechst 33342. The solid symbols represent the results following verapamil exposure.

c

o o §0.8

c 0.4 0) o 0

20 40 Time (minutes)

60

Figure 15.15. Red: violet ratios from the data in figure 15.14. The red shift after verapamil addition in the LX4 cells (triangles), due to an increase in the dye: DNA-phosphate ratio, is very apparent; however, there was no change in the sensitive parent (circles).

THERAPY SELECTION

375

concentration for a given extracellular concentration. Consequently, the fluorescence emission will be violet biassed compared with cells containing lower numbers of membrane gp 170 molecules. Verapamil, which blocks calcium channels and reduces the activity of gp 170 , will decrease the efflux of dye in cells with elevated gp 170 levels. This results in a higher intracellular dye concentration, greater binding to DNA and a relative red shift to the emission spectrum. Cells with very low numbers of gp 170 molecules will show little or no emission spectrum shift with verapamil. We have not shown directly that Hoechst 33342 is actively transported out of cells by the gp 170 pathway, but the evidence pointing to this conclusion is compelling. Firstly, the cellular uptake of Hoechst 33342 parallels that of cytotoxic agents involved in MDR (Lalande, Ling and Miller, 1981). Secondly, the intracellular content of this agent is modified by verapamil (Krishan, 1987, and see above). Thirdly, the predicted red shift in the emission spectrum due to verapamil only occurs in gp 170 elevated MDR cells. Finally, of three MDR cell lines investigated, the phenomenon was only seen in LX4 cells in which the mdr-1 gene was found to be amplified (Reeve et al, 1989). It was not found in MDR cells without mdr-1 gene amplification. The technique has been applied in a very limited number of patients with smallcell lung cancer to date, but this has already demonstrated the feasability of the approach. The illustrating example from a small sample of a biopsy specimen was disaggregated with trypsin/versene, stained and assayed both with and without verapamil. In each of the panels in figure 15.16 violet fluorescence (400RF AREA) is plotted on the X-axis obliquely from left to right with red fluorescence (600RF AREA) on the Z-axis from right to left. The origins are furthest from the eye with frequency on the Y-axis in these three-dimensional data spaces. The top two panels show the results obtained from a disaggregated small-cell lung cancer biopsy at 2 and 5 minutes after staining with Hoechst 33342 (left and right respectively) where a number of subsets can be seen and there was a considerable violet bias to the emission of the major population at these short incubation times. The bottom left panel shows the results obtained within the same experiment from the verapamil pre-treated drug-sensitive cell line H69/P analysed 2 minutes after staining which was identical to the non-verapamil treated sample as was expected from the data in figure 15.15. This panel represents a reference standard where cells in the other three panels which are angled further to the right than the data in this panel are classified as violet biassed. The bottom right panel shows the results from the biopsy after 5 minutes staining where the cells were pretreated with verapamil. Note the better defined peaks compared with the non-verapamil treated cells (top right panel) where the major spike (checked with ethidium bromide staining) represents aneuploid tumour cells. There has been both an increase in total fluorescence and a relative shift towards the red (600RF AREA) fluorescence axis. The degree of shift in the aneuploid component is comparable to that shown in figure 15.15 and indicated the possibility of MDR.

BIOPSY z

mn%

8I0PSV 5

u

z

BIOPSY

L• Figure 15.16. Small-cell lung cancer patient biopsy assayed on the violet (400RF AREA) and red (600RF AREA) channels at short times after staining with Hoechst 33342. The origins are furthest from the eye with violet fluorescence scored from top left to bottom right and red fluorescence scored from top right to bottom left. Results from verapamil treated H69 cells at 2 minutes after staining, which act as a reference standard, are shown in the bottom left panel. The two top panels show the biopsy results at 2 and 5 minutes respectively after staining with Hoechst 33342. In both panels the major population is angled to the right compared with the H69 data indicating a violet bias. The data in the bottom right panel were obtained from biopsy cells pretreated with verapamil and stained for 5 minutes. These show a dramatic shift in the major spike compared with the top right panel. The latter has increased in fluorescence intensity on both the red and violet channels but the former is relatively greater than the latter indicating a red shift due to verapamil.

THERAPY SELECTION

377

The 'direct' method of Herweijer et al (1989) is very similar except that they used daunorubicin as the reporter probe for MDR with verapamil and cyclosporin-A as the efflux modifiers in sensitive and resistant human ovarian cancer cell lines. The resistant line, A2780/R, was maintained continuously in medium containing daunorubicin at a concentration of 2 uM but the cells were incubated in drug-free medium for 48 hours prior to analysis. Time courses for the uptake of daunorubicin were then performed for the sensitive (A2780/S) and resistant lines and the efflux modifiers were added to the resistant cells after 30 minutes. The results with verapamil are shown in figure 15.17 where three different concentrations of the efflux modifier were used. It is clearly apparent that there was a verapamil concentration dependent increase in daunorubicin fluorescence in the resistant cells. Hepes-buffered Hanks' salt solution was added to the sensitive cells at 60 minutes to act as a control. These results, I'm sure, represent the processes being observed in the biology as they exhibit the same effect as seen in figure 15.15. This, of course, assumes that we were correct (Morgan et al, 1989) as the control in figure 15.17 from Herweijer et al (1989) is not adequate. Verapamil should have been added to both cell types at 30 minutes to exclude the possibility of similar changes in the sensitive cells. Comparable results using cyclosporin-A as the efflux modifier are shown in figure 15.18. Again varying doses of cyclosporin-A were added at 30 minutes, and on a molar basis this agent was considerably more efficient in modifying efflux than verapamil, compare with figure 15.17. However, again I would challenge the control as no cyclosporin-A was administered to the sensitive cells at 30 minutes.

4500n

A2780/S O C 4) O 0)


30

60 Time (minutes)

90

Figure 15.17. Daunorubicin time course uptake in sensitive (A2780/S) and resistant (A2780/R) cell lines. Verapamil increased the uptake of daunorubicin in the resistant cell line.

378

APPLICATIONS IN ONCOLOGY

4500n

A2780/S 0)

o
0(iM

30

60 Time (minutes)

90

Figure 15.18. Similar experiment to that in figure 15.17 where cyclosporin was used as the daunorubicin efflux modifier. Both this and the previous figure were redrawn from Herweijer et al. (1989).

A non-kinetic method for improving the detection of MDR in daunorubicintreated cells has been proposed by Ross et al. (1989). This involves measuring daunorubicin fluorescence and cell volume, which enable intracellular concentration to be calculated which is lower in MDR cells. Antibodies are available to both the internal and external domains of gp 170 and these can be used to estimate the number of molecules per cell. An example obtained with H69 and LX4 is shown in figure 15.19 where there is greater fluorescence associated with the resistant LX4 cell line. This type of assay may be useful in some circumstances; however, it is measuring total quantity of gp 170 and this may not correlate with its functional activity which is the important parameter to quantitate in any attempt to assay for MDR. Nevertheless, combinations of these types of techniques can, and have, been applied directly to patient biopsies (Morgan et al, 1987, 1989; Watson, Morgan and Smith, 1989) and with further refinements they should be able to detect a 1:1000 drug-resistant minority subset within the sample. This may be very useful as it could tell us if we need to administer efflux modifiers in combination with drugs which are known to either induce or select for multi-drug resistance. A further avenue for exploitation is in the general area of drug transport into cells. It was shown in section 14.2.7 that esterase inhibition by carbamoylation with nitrosoureas and related isocyanates can be assayed flow cytometrically and comparisons between agents can be made using the I50 value which is the dose required to induce 50% inhibition of activity. An example of BCNU inhibition was given in that section and similar assays were performed with a number of drugs using both flow cytometry for intact cells and conventional spectrofluorimetry

THERAPY SELECTION

379

Figure 15.19. Two cell lines, H69 parent (middle panel) and MDR LX4 (bottom panel) stained for DNA (630RF AREA, top left to bottom right) simultaneously with immunofluorescence for gp 170 (520RF AREA, bottom left to top right). The top panel is the fluorescence control data set. A few H69 cells exhibit some gp 170 fluorescence but the LX4 cells which have MDR-1 gene amplification show considerable fluorescence.

3S0

APPLICATIONS IN ONCOLOGY

Table 15.1. Iso values (M) Drug CHI CCNU BCNU CEI TCNU ACNU CHLOZ GANU

Intact cells

Sonicates

Ratio

1.1 x 10"' 3.8x10"' 5.0x10"' 1.0 x 10"' 1.8 xlO" 4 7.3 x l O " 3 >1.5xlO"2 > 1.5 xlO" 2

2.6x10"' 8.3x10"' 9.2x10"' 1.6x10"' 9.9x10"' 3.2 x l O " 3 2.4 x l O " 3 2.4 x l O " 3

0.423 0.458 0.543 0.625 1.818 2.281 > 6.250 > 6.250

with sonicated preparations and the I50 values for the two methods are compared in table 15.1 together with the ratio of results from intact cells to those in sonicates. The results in this table are interesting. The I 50 ratios for CHI, CCNU, BCNU and CEI indicate that only about half the concentration of drug is required to obtain 50% esterase activity inhibition in whole cells compared with sonicates. With TCNU and ACNU this is reversed with double the drug concentration required in whole cells compared with sonicates; however, with chlorozotozin and GANU this has increased to greater than six-fold. These results almost exactly parallel the hydrophilicity of the agents. CHI, CCNU, BCNU and CEI are lipophilic, but chlorozotozin and GANU are hydrophilic, readily soluble in water at physiological pH and temperature and would not be expected to cross the external cell membrane without an active transport mechanism. We have seen earlier that nitrosoureas enter cells by passive diffusion, thus we would not expect chlorozotozin and GANU to enter cells. This conclusion would seem to be substantiated by the data in table 15.1 and the method seems capable of assessing not only the ability of such agents to cross the external cell membrane, but also their capabilty of inducing damage having entered the cell. This technique, like those described above for MDR, could also readily be applied to small biopsies taken directly from a tumour. 15.3.3 Tumour growth rate Surgery and radiotherapy were the mainstays of cancer therapy until the late 1950s when chemotherapy was introduced. A number of landmarks can be identified in radiotherapy, the first of which was that it was ever used at all for cancer therapy, and the second was that it was used so soon after the discovery of radium (Curie and Curie, 1899). It is also worth remarking that hormone manipulation for breast cancer by oophorectomy was introduced three years earlier by Beatson (1896). The next major 'contribution' to radiotherapy was the introduction of external beam fractionated radiation in which multiple small doses were delivered on a daily basis over a number of weeks. In fact, this was entirely serendipitous. The early X-ray machines had a very low radiation output and

THERAPY SELECTION

381

frequently overheated and had to be switched off intermittently or they broke down completely. Thus, it was logistically easier for staff and patients to treat a number of patients each for a short time and let the machines cool off between treatments rather than have one patient there all day and wait between treatments. However, it was not until the late 1950s and early 1960s that it was shown why this approach can actually work. A further landmark was the introduction of high energy radiation (super-voltage) in the 1950s where the maximum absorbed dose occurs at some depth into the tissues thus sparing the skin from damage. The recognition that cells are relatively more resistant to radiation under anaerobic conditions (Crabtree and Cremer, 1933; Mottram, 1935) and that tumours contain hypoxic cells which are resistant to radiation at critical distances from blood vessels was another important step (Thomlinson and Gray, 1955). This lead to radiation treatment in hyperbaric oxygen (van den Brenk, Madigan and Kerr, 1968) and the introduction of electron-afBnic chemical sensitizing agents (Adams and Cooke, 1969). One of the greatest problems with all non-surgical cancer treatments is the complete non-specificity and all tumours of a given type tend to be treated identically and in radiotherapy this involves irradiation on a three or five fraction per week schedule with no therapy at the weekend. Unfortunately, tumours don't know that they should stop growing at the weekend when nobody is around to treat them and the same applies to chemotherapy schedules with extended intervals (2-4 weeks) between courses. The net rate constant for growth of a tumour (K) is dependent on the excess of the rate constant for cell production (KP) compared with the rate constant for cell loss (KL) where K= KP — KL. If a tumour has a very high rate of cell production which is only partially arrested by each week of fractionated radiotherapy or by each course of chemotherapy there is a distinct possibility that it could grow back to its original size or beyond during the treatment intervals. However, some breaks in treatment are necessary to allow the normal tissues to recover as these too are damaged by the therapy. A fractionated course of cancer therapy, which could be either radio- or chemotherapy, is depicted in figure 15.20. This may look a little complicated at first sight but it's really very simple and a considerable underestimate of the complexity of the reality. Five therapy fractions are depicted by the arrows. Just before fraction 1 both the tumour and the treatment target volume containing the tumour have an initial 'surviving fraction' of unity. If chemotherapy is being used the target volume is the whole patient. For the sake of simplicity we will assume that the dose given at each fraction kills 50% of the tumour cells. It is highly probable that this same dose will kill relatively more of the critical normal cells within the treatment volume and, again for simplicity, I've chosen this as 60%. Thus, after the first fraction the tumour will be reduced to 50% of its initial cell content (solid triangle) and the normal cell component will be reduced to 40% (solid circle) and at this point we have killed relatively more of the patient than the tumour. What happens next is critical to understanding why fractionated therapy works. Normal tissues have the capacity to recover rapidly by repairing damage at the cellular level, regenerating within the target volume and by repopulating from without the

382

APPLICATIONS IN ONCOLOGY

1

2 -•9

\

i

3

4

5

.?

•?

**<

!

!

/I

O

>

Jo.50,4

Time—• Figure 15.20. Representation of the response of a tumour to fractionated therapy. target volume if the latter is localized (i.e. radiotherapy). This is depicted by the dotted line from the solid circle at 0.4 at fraction 1 to the open circle at 1.0 at fraction 2. The normal tissues will not recover to above 1.0 due to the normal homeostatic mechanisms. The tumour does not recover so rapidly but it does regrow, from the solid triangle at 0.5 at fraction 1 to the open triangle at 0.6 at fraction 2. When the second therapy fraction is delivered the normal cells are again reduced to a surviving fraction of 0.4 (solid circle at fraction 2) and the tumour is reduced to 0.5 of 0.6, i.e. 0.3 (solid triangle at fraction 2). This is the point at which the first net gain is achieved. The normal tissues again recover to 1.0 (dotted line) and the tumour again regrows. Although the proportional increase in the latter is the same as previously (exponential growth) the absolute increase is less than between fractions 1 and 2 and it regrows to less than 0.4, the open triangle at fraction 3. This process continues with increasing numbers of therapy fractions and the tumour volume declines exponentially as shown by the dot-dash-dot curve as long as the fractions are spaced as shown. However, if fractions 2,3 and 4 had been omitted the tumour would have regrown, represented by the dashed curve, to its original size (solid square) just before fraction 5 (which would now be fraction 2) and such a spacing of therapy fractions could not hope to contain tumour growth. Thus, the optimum interval between treatment fractions is critical and primarily dependent on the rate of tumour growth which, in turn, is dependent on the rate of cell production by the tumour. There was no realistic means of assessing tumour cell production rate until Begg et al. (1985) produced a very elegant method which I would place along side, or

THERAPY SELECTION

383

even above, the oxygen effect and sensitizers in terms of potential importance to radiotherapy. That importance lies in the simple fact that for the very first time we are able to measure something in a tumour which has direct relevance to the speed at which the tumour is growing and prescribe treatment specifically for that tumour. Because of its potential importance I'll describe it in some detail though the reader should be encouraged to read the original paper (Begg et al, 1985) and the follow-up correspondence (White and Meistrich, 1986). The potential doubling time, T pot, of a tumour is given by the expression,

where t$ and LI are the duration of S-phase and labelling index respectively and where 1 is a correction factor for the position of S-phase in the cell cycle of exponentially growing populations (Steel, 1968). The value for k is always likely to be between 0.9 and 1.0 in human tumours although it could have a wider range for cells growing in tissue culture, hence, to a reasonable approximation, a value for the potential doubling time, T pot, can be obtained by knowing the S-phase duration and labelling index (proportion in S-phase). The importance of T pot is shown by the relationship, KP = log e 2/T pot where KP is the rate constant for cell production. Hence, by measuring the S-phase duration and LI simultaneously we can obtain the rate constant for cell production in a tumour. Begg et al. (1985) solved this problem by suggesting that a single dose of bromodeoxyuridine should be administered to patients, then at a defined interval later (about 4 hours) a single biopsy should be taken from the tumour. Part of this would be used for the normal histological diagnosis and part for disaggregation and double staining for total DNA (propidium iodide) and BrdU (fluorescein) for flow cytometric analysis as described in section 11.5.5. We have already seen in figure 12.10 that such data sets can be gated to obtain the proportion of cells which have taken up BrdU (the labelling index). Moreover, we can also determine the proportion of cells initially labelled with BrdU which have halved their BrdU content at mitosis and have divided to appear in Gl at an interval after BrdU was administered. Let us suppose for simplicity that 50% of cells have halved their BrdU content and entered Gl 4 hours after administration. It's not too difficult to appreciate that the remaining 50% will also take about 4 hours to halve their BrdU content and enter Gl giving a total S-phase duration of about 8 hours. Similarly, if after 4 hours only 25% of labelled cells have halved their BrdU content and entered Gl then there will be three more cohorts of 25% of the total each of which will require about 4 hours to complete DNA synthesis. Thus, to a first approximation the S-phase duration will be about 16 hours. In principle it really is as simple and as elegant as that and as a radiotherapy and flow cytometry person I wish I'd thought of it and I bet many other people do too! This technique is now being used in clinical trials at Mount Vernon Hospital, Northwood, to assign patients with high rate constants for cell production, KP

384

APPLICATIONS IN ONCOLOGY

calculated from Tpot/ to hyper-fractionated radiotherapy schedules (multiple fractions per day) and those with low values of KP to conventional single daily fraction schedules. The difference between this clinical trial and all the others is that the two arms of the study are designed around a highly relevant biological parameter, the cell production rate, which is actually being measured within the tumour and flow cytometry is the only means of doing this.

15.4

Future prospects

It is always foolhardy to attempt to predict the future, except perhaps in nuclear physics where some spectactular predictions were later found to be true, and I'm not going to fall into this trap in oncology. However, at the beginning of this chapter I asked a number of questions and was somewhat scathing (I hope) about some of the basic premises on which anti-cancer clinical trials have been carried out. There is never any justification for destructive negative criticism about anything and if we are to be critical we should only be so in a constructive and positive sense. It is clear from the examples in the chapter that flow cytometry could help us in these matters. A number of assays are now available which give prognostic information (DNA index, S-phase fraction, oncoprotein quantitation) but, as I have pointed out, this really doesn't help us very much in the design of therapy strategies. However, help is also at hand here and, in an ideal world, every patient who is about to undergo either attempted curative or diagnostic surgery for cancer should have a prior injection of BrdU so that an estimate of the rate constant for cell production can be obtained. This might enable us to develop fractionated therapy regimes for chemotherapy similar to those studies now being carried out at Mount Vernon Hospital for radiation which are designed for a particular tumour. Furthermore, whenever therapy with radiation and drugs is to be contemplated the cells from the tumour biopsy should be assayed for GSH levels, methotrexate uptake and MDR using both functional assays described. This would be particularly important for breast cancer where many adjunctive chemotherapy regimes using adriamycin and methotrexate are frequently employed. By carrying out prior biochemical assays on the tumour cells under near-physiological conditions we should be able to predict if there is likely to be a subpopulation of cells which is drug or radiation resistant and employ suitable modifying agents where necessary. It goes without saying that all of these suggestions should be carried out under carefully controlled clinical trials where the basic premises of those trials are based on sound biochemical measurements to stratify patients into subgroups which need special attention as opposed to the overall empiricism applied to all patients with a particular tumour type which is the general rule at present.

16 Epilogue

If you have read the whole of this book I salute you. If you are not primarily involved in this field, have read the whole book and have understood everything, then I salute both of us. The intention was an introduction and, indeed, that is all that it is as, on the applications front, I have omitted very considerably more than IVe included. The problem with writing about flow cytometry is that it covers the whole of biological science and it is not possible to include everything, thus a considerable degree of selection must be employed. However, I hope that the first 10 chapters, which were intended to show how these instruments work, how they can be used and some of the associated problems, were not only intelligible but also prove to be useful. The last five chapters were intended to show where they can be used in a number of fields in cell biology including oncology. However, the biggest applications omission is the whole of immunology which, if it had been included, would have more than doubled the size of the book. The technology is also being used increasingly in microbiology, plant biology and in the aquatic sciences and none of these have even been mentioned up till now. The last of these is assuming increasing importance and in recognition of this the September 1989 issue of Cytometry, containing 21 papers and over 600 references, was devoted to this topic. There has been a simply phenomenal revolution in the biological sciences over the past 40 years which, I believe, represents a Renaissance equivalent to those in art and literature in the fifteenth and sixteenth centuries, mathematics, physics and chemistry in the sixteenth and seventeenth centuries and physics and astronomy in each of the decades either side of the turn of this century. Chemistry and physics respectively did not really begin until the chemical balance and measurement systems (length and time) were invented. Biology is considerably more complex than chemistry and physics as it includes both of these disciplines. The current revolution in biology is due, in part, to our capacity to recognize and measure cellular and subcellular constituents and to relate these to function of both individual cells and complex interactions between cells in multi-cellular cooperative organ systems. The methods by which those measurements are made is the means by which the revolution has taken place and flow cytometry has been part of that revolution for the past 20 years. It would seem to me that these instruments together with those of image analysis coupled with techniques in biochemistry and

386

EPILOGUE

molecular biology have given us an unprecedented capacity to quantitate in all aspects of cell biology. We are no longer inhibited in our capacity to measure in the biological sciences and for those who have the courage to use their imagination there would seem to be no limit to what can be achieved. That is the only real message I would like this book to convey.

References

Adams, G. E. and Cooke, M. S. (1969) Electron-affinic sensitization. 1. A structural basis for chemical radiosensitizers in bacteria. Int. ]. Radial. Biol. 15; 457-71. Adler, A. J. and Kistiakowsky, G. B. (1961) Isolation and properties of pig liver esterases. /. Biol. Chem. 236; 3240-5. Adolph, K. W., Cheng, S.M. and Laemmli, U.K. (1977) Role of non-histone proteins in metaphase chromosome structure. Cell 12; S05-16. Airy, G. B. (1838) On the intensity of light in the neighbourhood of a caustic. Trans. Cambridge Phil. Soc. 6; 378-402. Airy, G. B. (1848) Supplement to a paper, On the intensity of light in the neighbourhood of a caustic. Trans. Cambridge Phil Soc. 8; 593-9. Alanen, K. A., Klemi, P.J., Joensuu, H., Kujari, H. and Pekkala, E. (1989) Comparison of fresh, ethanol-preserved and paraffin-embedded samples in DNA flow cytometry. Cytometry 10; 81-5. Alt, F.W., Kellems, R.E. and Schimke, R.T. (1976) Synthesis and degradation of folate reductase in sensitive and methotrexate resistant lines of S-180 cells. /. Biol. Chem. 251; 3063-74. Amberson, J. B., Vaughan, E. D., Gray, G. F. and Naus, G. J. (1987) Flow cytometric analysis of nuclear DNA from adrenocortical neoplasms. A retrospective study using paraffinembedded tissue. Cancer 59; 2091-5. Anderson, K. C, Bates, M. P., Slaughenhoupt, B. L, Pinkus, G. S., Schlossman, S. F. and Nadler, L. M. (1984) Expression of human B cell-associated antigens on leukaemias and lymphomas: A model of human B cell differentiation. Blood 63; 1424—33. Anderton, B. H. (1981) Intermediate filaments: a family of homologous structures. /. Muse. Res. Cell Motility 2; 141-6. Angerer, L. M. and Moudrianakis, E. N. (1972) Intercalation of ethidium bromide with whole and selectively deprotinized deoxynucleoproteins from calf thymus. /. Mol. Biol. 63; 505-21. Archer, J. S. and Wust, G. J. (1973) In vitro incorporation of orotic acid by spleen and liver cells in rats. Proc. Soc. Exptl. Biol. Med. 142; 262-5. Armitage, N. C, Robins, R. A., Evans, D. F., Turner, D. R., Baldwin, R. W. and Hardcastle, J. D. (1985) The influence of tumour DNA abnormalities on survival in colorectal cancer. Brit. ]. Surg. 72; 828-30. Armstrong, J. A. (1956) Histochemical differentiation of nucleic acids by means of induced fluorescence. Exptl. Cell Res. 11; 640-3. Arndt-Jovin, D.J. (1979) Cell differentiation. In Flow Cytometry and Sorting (eds M. R. Melamed, P. F. Mullaney and M. L. Mendelsohn) chapter 24, p. 453. John Wiley and Sons, New York, Chichester, Brisbane, Toronto.

388

REFERENCES

Arndt-Jovin, D. J., Grimwade, B. G. and Jovin, T. M. (1980) A dual laser flow sorter utilizing a CW pumped dye laser. Cytometry 1; 127-31. Arndt-Jovin, D. J. and Jovin, T. M. (1977) Analysis and sorting of living cells according to deoxyribonucleic acid content. /. Histochem. Cytochem. 25; 585—9. Ashley, D.J. B. (1969) Colonic cancer arising in polyposis coli. /. Med. Genet. 6; 376—8. Assaraf, Y. G., Seamer, L. C. and Schimke, R. T. (1989) Characterization by flow cytometry of fluorescein-methotrexate transport in Chinese hamster ovary cells. Cytomerry 10; 50-5. Assaraf, Y. G. and Schimke, R. T. (1987) Identification of methotrexate transport deficiency in mammalian cells using fluoresceinated methotrexate and flow cytometry. Proc. Natnl Acad. Sci. USA. 84; 7154-8. Aten, J. A., Kipp, A. and Barendsen, G. A. (1980) Flow-cytometric determination of damage to chromosomes from x-irradiated Chinese hamster cells. In Flow Cytometry IV (eds O. D. Laerum, T. Lindmo and E. Thorud) pp. 287-92. Universitetsforlaget, Oslo. Atkin, N. B. (1972) Modal deoxyribonucleic acid value and survival in carcinoma of the breast. Brit. Med. ]. 1; 271-2. Atkin, N. B. and Kay, R. (1979) Prognostic significance of modal DNA value and other factors in malignant tumours based on 1465 cases. Brit. ]. Cancer 40; 210-21. Auer, G., Ericksson, E., Azavedo, E., Casperson, T. and Wallgren, A. (1984) Prognostic significance of nuclear DNA content in mammary adenocarcinomas in humans. Cancer Res. 44; 394-6. Babson, J. E., Abell, N. and Reed, D. J. (1981) The protective role of glutathione redox cycle against adriamycin mediated toxity in isolated hepatocytes. Biochem. Pharmacol 31; 2299-304. Babson, J. R. and Reed, D. J. (1978) Inactivation of glutathione reductase by 2-chloroethyl nitrosourea-derived isocyanates. Biochem. Biophys. Res. Comm. 83; 754—62. Baker, S. J., Fearin, E. R., Nigro, J. M., Hamilton, S. R., Preisinger, A. C, Milburn-Jessup, J., van Tuinen, P., Ledbetter, D. R , Barker, D. F., Nakamura, Y., White, R. and Vogelstein, B. (1989) Chromosome 17 deletions and p53 gene mutations in colorectal carcinoma. Science 244; 217-21. Baguely, B. C. (1982) Nonintercalative DNA-binding anti-tumour compounds. Mol. Cell. Biochem. 43; 167-81. Baisch, H. and Beck, H.-P. (1978) Comparison of cell kinetic parameters obtained by flow cytometry and autoradiography. In Biomathematics and Cell Kinetics (eds. A.-J. Valleron and P. D. M. Macdonald) p. 411. Elsevier/North Holland, Amsterdam. Baisch, H., Beck, H.-P., Christensen, I. J., Hartmann, N. R., Fried, J., Dean, P. N., Gray, J. W., Jett, J. H., Johnston, D. A., White, R. A., Nicolini, C, Zeitz, S. and Watson, J. V. (1982) A comparison of mathematical methods for the analysis of DNA histograms obtained by flow cytometry. Cell and Tissue Kinet. 15; 235—49. Baisch, H., Gohde, W. and Linden, W. A. (1975) Analysis of PCP-data to determine the fraction of cells in the various phases of the cell cycle. Rad. Environ. Biophys. 12; 31—9. Ball, R. L, Albrecht, T. B. and Carrey, D. H. (1982) Microtubule involvement in initiation of cell DNA synthesis by human cytomegalovirus. /. Cell Biol 95; 355a. Baltimore, D. (1970) RNA-dependent DNA polymerase in virons of RNA tumour viruses. Nature 226; 1209-11. Banner, B.F., Chacho, M.S., Roseman, D.L. and Coons, J.S. (1987) Multiparameter flow cytometric analysis of colon polyps. Am. ]. Clin. Pathol 87; 313-18. Barlogie, B., Gohde, W., Johnson, D. A., Smallwood, L, Schumann, J., Drewinko, B. and

REFERENCES

389

Freireich, E. J. (1978) Determination of ploidy and proliferative characteristics of human solid tumours by pulse cytophotometry. Cancer Res. 38; 3333-9. Bartholdi, M. F., Meyne, J., Johnson, R. G. and Scott-Cram, L. (1989) Chromosome banding analysis by slit-scan flow cytometry. Cytometry 10; 124-33. Battifora, R , Sun, T. T., Bahu, R. M. and Rap, S. (1980) The use of anti-keratin antiserum as a diagnostic tool: thymoma versus lymphoma. Human Pathol. 11; 635-41. Bauer, K. D., Clevenger, C. V., Endow, R. K., Murad, T., Epstein, A. L. and Scarpelli, D. G. (1986a) Simultaneous nuclear antigen and DNA content quantitation using param'nembedded colonic tissue and multiparameter flow cytometry. Cancer Res. 46; 2428-34. Bauer, K. D., Lincoln, S. T., Vera-Roman, J. M., Wallemark, G. B., Chmeil, J. S., Madurski, M. L, Murad, T. and Scarpelli, D. G. (1987) Prognostic implications of proliferative activity and DNA aneuploidy in colonic adenocarcinomas. Lab. Invest. 57; 329-35. Bauer, K.D., Merkel, D.E., Winter, J.N., Marder, R.J., Hauck, W.W., Wallermark, C.B., Williams, T.J. and Variakojis, D. (1986b) Prognostic implications of ploidy and proliferative activity in diffuse large cell lymphoma. Cancer Res. 46; 3173-8. Bauman, J. G. J. and Bentvelzen, P. (1988) Flow cytometric detection of ribosomal RNA in suspended cells by fluorescent in situ hybridization. Cytometry 9; 517-24. Beatson, G. T. (1896) On the treatment of inoperable cases of carcinoma of the mamma: suggestions for a new method of treatment, with illustrative cases. Lancet ii; 104-7. Beck, H.-P. (1978) A new analytical method for determining duration of phases, rate of DNA-synthesis and degree of synchronization from flow cytometric data on synchronized cell populations. Cell and Tissue Kinet. 11; 139-48. Beck, H.-P. (1981) Proliferation kinetics of perturbed cell populations determined by the BUdR-33258 technique; radiotoxic effects of 3H-thymidine. Cytometry 2; 170-4. Beck, W T., Cirtain, M. C, Look, A. T. and Ashmun, R. A. (1986) Reversal of vinca alkaloid resistance but not multi-drug resistance in human leukaemia cells by verapamil. Cancer

Res. 46; 778-84. Begg, A. C, McNally, N. J., Schrieve, D. C. and Karcher, H. (1985) A method to measure the duration of DNA synthesis and the potential doubling time from a single sample. Cytometry 6; 620-6. Begleiter, A., Lam, H.Y.P. and Goldenberg, G.J. (1977) Mechanism of uptake of nitrosoureas by L5178Y lymphoblasts in vitro. Cancer Res. 37; 1022-7. Beisker, W., Dolbeare, F. and Gray, J.W. (1987) An improved immunocytochemical procedure for high-sensitivity detection of incorporated bromodeoxyuridine. Cytometry 8; 235-9. Benedict, W.F., Banerjee, A., Mark, C. and Murphee, A. L. (1983a) Non-random chromosomal changes in untreated retinoblastoma. Cancer Genet. Cytogenet. 10; 311—33. Benedict, W. F., Murphee, A. L, Banerjee, A., Spina, C. A., Sparkes, M. C. and Sparkes, R. S. (1983fe) Patient with 13 chromosome deletion: evidence that the retinoblastoma gene is a recessive cancer gene. Science 219; 973-5. Berridge, M. J. and Irvine, R. F. (1984) Inositole triphosphate, a novel second messenger in intracellular signal transduction. Nature 312; 315—21. Bernal, S. D., Shapiro, H. M. and Chen, L. B. (1982) Monitoring the effects of anticancer drugs on L1210 cells by a mitochondrial probe; rhodamine 123. Int. ]. Cancer 30; 219-24. Bernheim, A., Metessau, P., Geuellaen, G., Fellous, M., Goldberg, M. E. and Berger, R. (1982) Direct hybridization of sorted human chromosomes: localization of the Y chromosomes on the flow karyotype. Proc. Natnl. Acad. Sci. USA 80; 7571-5.

390

REFERENCES

Bernoulli, D. (1738) Hydrodynamica. Argentorati. Bijman, J. T. (1983) Optimization of mammalian chromosome suspension preparations employed in a flow cytometric analysis. Cytometry 3; 354-8. Bishop, J.M. (1985) Viral oncogenes. Cell 32; 23-36. Bishop, J. M. and Varmus, H. (1982) Functions and origins of retro viral transforming genes. In Molecular Biology of Tumour Viruses, Part III, RNA Tumour Viruses (eds R. Weiss, N. Teich, H. Varmus and J. Coffin), pp. 999-1108. Cold Spring Harbor Press, New York. Blanchard, J.-M., Piechaczyk, M., Dani, G, Chambard, J.-C, Franchi, A., Pouyssegur, J. and Jeanteur, P. (1985) c-myc gene is transcribed in G0-arrested fibroblasts and is posttranscriptionally regulated in response to growth factors. Nature 317; 443—5. Blijham, G., Schuette, B., Reynders, M., Wiggers, Th, Arends, J., Volovics, L. and Bosman, F. (1985) Flow cytometric determination of ploidy levels and life cycle analysis on 297 paraffin embedded colorectal carcinoma specimens. Proc. Am. Soc. Clin. Oncol 4; 22a. Blomjous, C.E.M., Schipper, N. W., Baak, J. P. A., van Galen, E. M., de Voogt, H.J. and Meyer, G J. L. M. (1988) Retrospective study of prognostic importance of DNA flow cytometry in urinary bladder carcinoma. /. Clin. Pathol. 41; 21-5. Blow, J. J. and Laskey, R. A. (1986) Initiation of DNA replication in nuclei and naked DNA by a cell-free extract of Xenopus eggs. Cell 47; 577-87. Blow, J.J., Sheehan, M.A., Watson, J.V. and Lasky, R.A. (1988) Regulated DNA replication in cell-free extracts of Xenopus eggs. Cancer Cells 6; 297—303. Blow, J.J. and Watson, J.V. (1987fl) Nuclei act as independent and integrated replication units in a xenopus cell-free DNA replication system. Embo ]. 6; 1997—2002. Blow, J. J. and Watson, J. V. {1987b) A probe for nascent DNA in intact nuclei. Trends Genet.

3; 233. Blumenthal, A.B., Dieden, J.D., Kapp, L.N. and Sedat, J.W. (1979) Rapid isolation of metaphase chromosomes containing high molecular weight DNA. /. Cell Biol 81; 255-9. Blumerfeld, D., Braly, P. S., Ben-Ezra, J. and Kleverz, R. R. (1987) Tumor DNA content as a prognostic feature in advanced epithelial ovarian carcinoma. Gynaecol. Oncol. 37; 389-98. Bohmer, R. M. (1979) Flow cytometric cell cycle analysis using the quenching of 33258 Hoechst fluorescence by bromodeoxyuridine incorporation. Cell and Tissue Kinet. 12; 101-10. Bohr, N. (1909) Determination of the surface tension of water by the method of jet vibration. Phil Trans. R. Soc. Lond. A209; 281-317. Bonnard, G (1984) Streptavidin-biotin bridge technique: application in light and electron microscope immunocytochemistry. In Immunolabelling for Electron Microscopy (eds J. M. Polak and I.M. Varndell) chapter 8, pp 95-111. Elsevir Scientific Publishers, Amsterdam. Bonner, W.A., Hulett, H.R., Sweet, R.G. and Herzenberg, L.A. (1972) Fluorescence activated cell sorting. Rev. Sci. Instr. 43; 404-9. Bontemps, J., Houssier, G and Fredericq, E. (1975) Physico-chemical study of the complexes of '33258 Hoechst' with DNA and nucleohistone. Nucleic Acid Res. 2; 971-84. Bowlby, L. S., DeBault, L. E. and Abraham, S. R. (1986) Flow cytometric analysis of adrenal cortical tumour DNA. Relationship between cellular DNA and histopathologic classification, Cancer 58; 1499-1505.

REFERENCES

391

Bowman, P. and Mclaren, A. (1970a) Viability and growth of mouse embryos after in vitro culture and fusion. /. Embryol. Exptl. Morphol. 23; 693-704. Bowman, P. and Mclaren, A. (1970b) The reaction of mouse blastocyst and its zona pellicida to enzymes in vitro. ]. Embryol. Exptl. Morphol. 24; 331—4. Brachet, J. (1940a) La detection histochimique des acides pentose nucleiques. Compt. Rend. Soc. Biol. 133; 88-90. Brachet, J. (1940b) La localisation des acides pentoenucleiques pendant la developement des Amphibiens. Compt. Rend. Soc. Biol. 133; 91-4. Bradley, D. F. and Wolf, M. K. (1959) Aggregation of dyes bound to polyanions. Proc. Natnl. Acad. Sci. USA 45; 944-52. Britten, R. A. and Dyson, J. E. (1987) Flow cytometric determination of cell cycle phasespecific changes in cellular phosphatase and glycosidase activities. Cell Biochem. Fund. 5; 17-26. Brodie, S., Giron, J. and Latt, S. A. (1975) Estimation of accessibility of DNA in chromatin from fluorescence measurements of electronic excitation energy transfer. Nature 253; 470-1. Brown, W. E. and Wold, F. (1973) Alkyl isocyanates as active-site-specific reagents for serine proteases. Identification of the active-site serine as the site of reaction. Biochem. 12; 835-40. Brunsting, A. and Mullaney, P. F. (1972) A light scattering photometer using photographic film. Rev. Sci. Instr. 45; 1514-19. Bump, E. A., Yu, N. Y. and Brown, J. M. (1982) Radiosensitization of hypoxic tumor cells by depletion of intracellular glutathione. Science 217; 544-5. Burchenal, J. R , Biedler, J. L. Nutting, J. and Stobbe, G. D. (1950) Effects of 4-amino-N10methyl-pteroyglutamic acid on leukocytes of normal and leukaemic mouse. Blood 5;

167-76. Bustin, M. (1987) Immunochemical analysis of the structure and function of chromosomal proteins. Cytometry 8; 251-9. Buys, C H . C M . , Koerts, T. and Aten, J.A. (1982) Well-defined human chromosomes isolated from mitotic fibroblasts by a new method. Hum. Genet. 61; 157-9. Cambier, M.A., Christie, W.J., Wheeless, L.L. and Frank, I.N. (1976) Slit-scan cytofluorometry: basis for automated urinary tract cytology./. Histochem. Cytochem. 24; 305-7. Cambier, J. L. and Wheeless, L. L. (1975) The binucleate cell: implications for automated cytopathology. Acta Cytol. (Baltimore) 19; 281-5. Campbell, A. K. (1983) Intracellular Calcium. John Wiley and Sons, New York. Capranico, G., Dasdia, T. and Zunino, F. (1986) Comparison of doxorubicin-induced DNA damage in doxorubicin-sensitive and -resistant P338 murine leukaemia cells. Brit. ]. Cancer 37; 227-31. Carlsen, S. A., Till, J. E. and Ling, V. (1976) Modulation of membrane drug permeability in Chinese hamster ovary cells. Biochim. Biophys. Acta 455; 900—12. Carrano, A. V., Gray, J. W., Langlois, R. G., Burkhart-Schultz, K. and van Dilla, M. A. (1979) Measurement and purification of human chromosomes by flow cytometry. Proc. Natnl. Acad. Sci. USA 76; 1382-4. Caskey, C. T. (1986) A milestone in human genetics. Cold Spring Harbor Symp. Quant. Biol. 51; 1115-19. Caspersson, T. O. and Schultz, J. (1938) Nucleic acid metabolism of the chromosomes in relation to gene reproduction. Nature 142; 294—5.

392

REFERENCES

Cavenee, W.K., Dryja, T.P., Phillips, R.A., Benedict, W.F., Godbout, R., Gallie, B.L., Murphee, A. L, Strong, L. C. and White, R. L. (1983) Expression of recessive alleles by chromosomal mechanisms in retainoblastoma. Nature 305; 779-84. Chaiet, L. and Wolf, F. J. (1964) The properties of streptavidin, a biotin-binding protein produced by streptomycetes. Arch. Biochem. Biophys. 106; 1-5. Chambers, S. H. and Watson, J. V. (1980) Identification of different cell types using two enzyme reactions measured simultaneously in single cells with a multi-parameter flow cytometer. Brit. ]. Cancer 42; 174. Chambers Twentieth Century Dictionary (1976) (ed A. M. Macdonald), p. 1330. W. and R. Chambers Ltd., Edinburgh. Chan, S., Gabra, H., Hill, F., Evan, G. I. and Sikora, K. (1987) A novel tumour marker related to the c-myc oncogene product. Mol. Cell Probes 1; 73-82. Chang, K.-J., Enker, W. E. and Melamed, M (1987) Influence of tumor cell DNA ploidy on the natural history of rectal cancer. Am. J. Surg. Pathol. 4; 184-8. Christensen, L, Hartmann, N. R., Keiding, N., Larsen, J K., Noer, H. and Vindelov, L. (1978) Statistical analysis of DNA distributions from cell populations with partial synchrony. In Third International Symposium on Pulse-cytophometry (ed. D. Lutz) pp. 71-8. European Press, Ghent. Churchman, J. W. (1927) The structure of B. anthracis and reversal of the Gram reaction. /. Exptl. Mel 46; 1009-20. Chused, T.M., Wilson, A.H., Greenblatt, D., Ishida, Y., Edison, L.J., Tsien, R.Y. and Finelman, F. D. (1987) Flow cytometric analysis of murine splenic B lymphocyte cytosolic free calcium response to anti-IgM and anti-IgD. Cytometry 8; 396-404. Chused, T. M., Wilson, A. H., Seligman, B. E. and Tsien, R. Y. (1986) Probes for use in the study of leukocyte physiology by flow cytometry. In Applications of Fluorescence in the Biomedical Sciences (eds. D. Lansing-Taylor, A. S. Waggoner, R. F. Murphy, F. Lanni and R.R. Birge) chapter 26, pp. 531-44. Alan R. Liss, New York. Civin, C. I. and Loken, M. R. (1987) Cell surface antigens on bone marrow cells: dissection of haematopoetic development using monoclonal antibodies and multiparameter flow cytometry. Int. ]. Cell Cloning 5; 267-88. Clark, E.O., Epp, E.R., Biaglow, J.E., Morse-Gaudio, M. and Zachago, E. (1984) Glutathione depletion, radiosensitization, misonidazole potentiation in hypoxic Chinese hamster ovary cells by buthione sulfoximine. Radiat. Res. 98; 370-80. Clevenger, C. V., Bauer, K. D. and Epstein, A. L. (1985) A method for simultaneous nuclear immunofluorescence and DNA content quantitation using monoclonal antibodies and flow cytometry. Cytometry 6; 208-14. Coffin, J. H., Varmus, H. E., Bishop, J. M., Essex, M., Hardy, W. D., Martin, S., Rosenberg, N. E., Scolnick, E. M., Weinberg, R. A. and Vogt, P. K. (1981) Proposal for naming host cell derived inserts in retrovirus genomes. /. Virol. 40; 953-7. Collard, J. G., de Boer, P. A. J., Janssen, J. W. G., Schijven, J. F. and de Jong, B. (1985) Gene mapping by chromosome spot hybridization. Cytometry 6; 179-85. Comings, D. E. (1973) A general theory of carcinogenesis. Proc. Natnl. Acad. Sci. USA 70;

3324-8. Coons, J. S., Bines, S., Kheir, S. and Soong, S.-J. (1987) DNA flow cytometry in stage I cutaneous melanoma. Cytometry Suppl. 1; 55 a. Coons, A. H., Creech, H. J. and Jones, P. N. (1941) Immunological properties of an antibody containing a fluorescent group. Proc. Soc. Exptl. Biol. Med. 47; 200—2. Coons, A. H. and Kaplan, M. H. (1950) Localization of antigen in tissue cells. II.

REFERENCES

393

Improvements in a method for the detection of antigen by means of fluorescent antibody. /. Exptl. Med. 91; 1-13. Cornelisse, C. J., van de Velde, C. J. H., Caspers, R. J. C, Moolenaar, A J. and Hermans, J. (1987) DNA ploidy and survival in breast cancer patients. Cytometry 8; 225-34. Corsetti, J.P., Cox, C, Leary, J.F., Cox, M.T., Blumberg, N. and Doherty, R.A. (1987) Comparison of quantitative acid-elution technique and flow cytometry for detecting fetomaternal hemorrhage. Analyt. Clin. Lab. Sci. 17; 197-206. Coulter, W. H. (1956) High speed automatic blood cell counter and cell size analyzer. Proc. Natnl. Electronics Conf. 12; 1034-40. Cowden, R. R. and Curtis, S. K. (1983) Supravital experiments with pyronin-Y, a fluorochrome of mitochondria and nucleic acids. Histochemistry 77; 535—42. Cox, H., Munn, C. F. and Watson, J. V. (1987) A dynamic range of 7 log-orders in a flow cytometer using linear amplifiers. Cytometry Suppl. 1; 102a. Crabtree, H. G. and Cremer, W. (1933) The action of radium on cancer cells. I. Effects of hydrocyanic acid, iodo-acetic acid and sodium fluoride on the metabolism and transplantability of cancer cells. Proc. R. Soc. Lond. B113; 226-8. Crawford, L. V., Pirn, D. C. and Bulbrook, R. D. (1982) Detection of antibodies against the cellular protein p53 in sera from patients with breast cancer. Int. ]. Cancer 30; 403-8. Cremer, C. and Gray, J. W. (1982) Application of the BrdU/thymidine method to flow cytogenetics: differential quenching/enhancement of Hoechst 33258 fluorescence of late-replicating chromosomes. Somatic Cell Genet. 8; 319-27. Cremer, C. and Gray, J. W. (1983) Replication kinetics of Chinese hamster chromosomes as revealed by bivariate karyotyping. Cytometry 3; 282-6. Crissman, H. A., Darzynkiewicz, Z., Tobey, R. A. and Steinkamp, J. A. (1985a) Correlated measurements of DNA, RNA and protein in individual cells by flow cytometry. Science 228; 1321-4. Crissman, H. A., Darzynkiewicz, Z., Tobey, R. A. and Steinkamp, J. A. (1985 b) Normal and perturbed Chinese hamster ovary cells: correlation of DNA, RNA and protein content by flow cytometry. /. Cell Biol. 101; 141-7. Crissman, H. A., Oka, M. S. and Steinkamp, J. A. (1976) Rapid staining methods for analysis of deoxyribonucleic acid and proteins in mammalian cells. /. Histochem. Cytochem. 24; 64-71. Crissman, H. A. and Steinkamp, J. A. (1973) Rapid, simultaneous measurement of DNA, protein and cell volume in single cells from large mammalian cell populations. /. Cell Biol.

59; 766-71. Crissman, H. A., Stevenson, A. P., Kissane, R. J. and Tobey, R. A. (1979) Techniques for quantitative staining of cellular DNA for flow cytometric analysis. In Flow Cytometry and Sorting (eds M. R. Melamed, P. F. Mullaney and M. L. Mendelsohn) chapter 14, pp. 234-62. John Wiley and Sons, New York, Chichester, Brisbane, Toronto. Crissman, H. A. and Tobey, R. A. (1974) Cell cycle analysis in 20 minutes. Science 184; 1297-8. Crosland-Taylor, P. (1953) A device for counting small particles suspended in fluid through a tube. Nature 171; 37-8. Crossin, K. L. and Carney, D. H. (1981a) Evidence that microtubule depolymerization early in the cell cycle is sufficient to initiate DNA synthesis. Cell 23; 61—7. Crossin, K. L. and Carney, D.H. (1981b) Microtubule stabilization by taxol inhibits initiation of DNA synthesis by thrombin and by epidermal growth factor. Cell 27; 341-50.

394

REFERENCES

Cupp, J. E., Leary, J. F., Cernichiori, E., Wood, J. C. and Doherty, R. A. (1984) Rare-event analysis methods for detection of fetal red blood cells in maternal blood. Cytometry 5; 138-44. Curie, P. and Curie, M. (1899) Sur la radioactivite provoquee par les rayons de Becquerel. Compt. Rend. Acad. Sci. (Paris) 129; 714-16. D'Alisa, R. M. and Gershey, E. C. (1978) Simian virus 40 T antigen binds to host cell chromosomes. Nature 274; 164-6. D'Alisa, R. M., Korf, B. R. and Gershey, E. C. (1979) T-antigen bonding on chromosomes of simian virus 40 infected muntjac cells. Cytogenet. Cell Genet. 24; 27-36. Dann, O., Bergen, G v Demant, E. and Vogl, G. (1971) Trypanocide diamine des 2-phenylbenzoforan, 2-phenylindoles. Leibigs Ann. Chem. 749; 68—72. Dano, K. (1973) Active outward transport of daunomycin in resistant Ehrlich ascites tumor cells. Biochim. Biophys. Ada 323; 466-83. Darzynkiewicz, Z., Andreeff, M., Traganos, F., Sharpless, T. and Melamed, M. R. (1978) Discrimination between cycling and noncycling lymphocytes by BUdR-suppressed acridine orange fluorescence in a flow cytometric system. Exptl Cell Res. 115; 31—5. Darzynkiewicz, Z., Kapuscinski, J., Carter, S. P., Schmid, F. A. and Melamed, M. R. (1986) Cytostatic and cytotoxic properties of pyronin-Y. Relation to mitochondrial localization to the dye and its interaction with RNA. Cancer Res. 46; 5760-6. Darzynkiewicz, Z., Kapuscinski, J., Traganos, F. and Crissman, H. A. (1987) Application of pyronin Y(G) in cytochemistry of nucleic acids. Cytometry 8; 138-45. Darzynkiewicz, Z., Traganos, F. and Melamed, M. (1980) New cell cycle compartments identified by multiparameter flow cytometry. Cytometry 1; 98—108. Darzynkiewicz, Z., Traganos, F., Sharpless, T. and Melamed, M. R. (1975) Conformation of RNA in situ as studied by acridine orange staining and automated cytophotometry. Exptl. Cell Res. 95; 143-53. Darzynkiewicz, Z., Traganos, F., Sharpless, T. and Melamed, M. R. (1976) Lymphocyte stimulation: A rapid multiparameter analysis. Proc. Natnl Acad. Sci. USA 73; 2881-6. Darzynkiewicz, Z., Traganos, F., Sharpless, T. and Melamed, M. R. (1977a) Cell cycle related changes in nuclear chromatin of stimulated lymphocytes as measured by flow cytometry. Cancer Res. 37; 4635-40. Darzynkiewicz, Z., Traganos, F., Sharpless, T. and Melamed, M. R. (1977b) Different sensitivity of DNA in situ in interphase and metaphase chromatin to heat denaturation. /. Cell Biol 73; 12S-3S. Darzynkiewicz, Z., Traganos, F., Sharpless, T. and Melamed, M. R. (1977c) Recognition of cells in mitosis by flow cytometry. ]. Histochem. Cytochem. 25; &75-&0. Davies, K. E., Young, B. D v Elles, R. G., Hill, M. E. and Williamson, R. (1981) Cloning of a representative genomic library of the human X chromosome after sorting by flow cytometry. Nature 293; 374-6. Davies, T.A., Drotts, D., Weil, G.K. and Simons, E.R. (1988) Flow cytometric measurements of cytoplasmic calcium changes in human platelets. Cytometry 9; 138-42. Dean, P. N., Dolbeare, F., Gratzner, H. G., Rice, G. C. and Gray, J. W. (1984) Cell cycle analysis using a monoclonal antibody to BrdUrd. Cell and Tissue Kinet. 17; 427-36. Dean, P. N. and Jett, J. H. (1974) Mathematical analysis of DNA distributions derived from flow microfluorometry. /. Cell Biol. 60; 523-7. Dean, P. N., Kolla, S. and van Dilla, M. A. (1989) Analysis of bivariate flow karyotypes. Cytometry 10; 109-23.

REFERENCES

395

Dean, P. N. and Pinkel, D. (1978) High resolution dual laser flow cytometry. ]. Histochem. Cytochem. 26; 622-7. Deaven, L. L, van Dilla, M. L, Bartholdi, M. F., Carrano, A. V., Cram, L. S., Fuscoe, J. C, Gray, J.W., Hildebrand, C.E., Moyzis, R.K. and Perlman, J. (1986) Construction of human chromosome-specific DNA libraries from flow-sorted chromosomes. Cold Spring Harbor Symp. Quant. Biol. 51, 159-67. Dembo, M., Sirotnak, F. M. and Moccia, D. M. (1984) Effects of metabolic depravation on methotrexate transport in L1210 leukaemia cells: further evidence for separate influx and efflux systems and different energetic requirements. /. Membrane Biol. 78; 9—17. Dent, G. A., Leglise, M. C, Pryzwansky, K. B. and Ross, D. W. (1989) Simultaneous paired analysis by flow cytometry of surface markers, cytoplasmic antigens, or oncogene expression with DNA content. Cytometry 10; 192—8. Descartes, R. (1637) 'Dioptrice copvt secundum de refractione' reprinted from La Diopterique in Philosophie sev dissertio de methodo, Diopterice et meteora, 1656, pp. 56—64.

Apud Ludovicum and Danielem Elzevirios, Amstelodami. Royal Society Library, London. Di Marco, A. (1978) Mechanism of action and mechanism of resistance to antineoplastic agents that bind to DNA. Antibiot. Chemother. 23; 216-27. Distel, R.J., Ro, H.-S., Rosen, B.S., Groves, D.L and Spiegelman, B.M. (1987) Nucleoprotein complexes that regulate gene expression in adipocyte differentiation: direct participation of fos. Cell 49; 835—44. Dittrich, W. and Gohde, W. (1969) Impulsfluorometrie dei einzelzellen in suspensionen. Z. Naturforsch 24b; 360-1. Dive, C. (1988) Flow cytoenzymology with special reference to cancer chemotherapy. Ph.D. Thesis, Council for National Academic Awards. Dive, C, Workman, P. and Watson, J. V. (1987a) Improved methodology for intracellular enzyme reaction and inhibition kinetics by flow cytometry. Cytometry 8; 552—61. Dive, C, Workman, P. and Watson, J. V. (1987/;) Novel dynamic flow cytoenzymological determination of intracellular esterase inhibition by BCNU and related isocyanates. Biochem. Pharmacol 36; 3731-8. Dive, C, Workman, P. and Watson, J. V. (1987c) Flow cytoenzymological analysis of intracellular carbamoylation by chloroethylnitrosoureas. Cytometry Suppl. 1; 343a. Dive, C, Workman, P. and Watson, J. V. (1988) Inhibition of intracellular esterases by antitumour chloroethylnitrosoureas: Measurement by flow cytometry and correlation with molecular carbamoylation activity. Biochem. Pharmacol 37; 3987—93. Dive, C, Workman, P. and Watson, J. V. (1990) Modulation of tumour cell membrane permeability by the developmental antitumor ether lipid SRI 62—834: analysis by multiparametric flow cytometry. Cancer Res. (In press.) Dolbeare, F. A. (1981) Fluorogenic quantification of specific chemical species in single cells. In Modern Fluorescence Spectroscopy (ed. E. L. Wehry) pp. 251—93. Plenum Press, New York. Dolbeare, F., Beisker, W., Pallavini, M.G., Vanderlaan, M. and Gray, J.W. (1985) Cytochemistry of bromodeoxyuridine/DNA analysis: Stoichiometry and sensitivity. Cytometry 6; 521-30. Dolbeare, F., Gratzner, H.G., Pallavicini, M.G. and Gray, J.W. (1983) Flow cytometric measurement of total DNA content and incorporated bromodeoxyuridine. Proc. Natnl Acad. Sci. USA 80; 5573-7.

396

REFERENCES

Dolbeare, F. A. and Phares, W. A. (1979) Naphthol AS-BI (7-Bromo-3hydroxy-2-napthoo-anisidine) phosphatase and naphthol AS-BI (3-D-glucuronidase in Chinese hamster ovary cells: biochemical and flow cytometric studies./. Histochem. Cytochem. 27; 120—4. Dolbeare, F. A. and Smith, R. A. (1977) Flow cytometric measurement of peptidases with the use of 5-nitrosalicylaldehyde and 4-methoxy-(J-naphthylamine derivatives. /. Clin. Chem. 23; 1485-91. Dolbeare, F. A. and Smith, R. E. (1979) Flow cytoenzymology: Rapid enzyme analysis of single cells. In Flow Cytometry and Sorting (eds M. R. Melamed, P. F. Mullaney and M. L. Mendelsohn) chapter 17, pp. 317-33. John Wiley, New York. Dolbeare, F. A., Vanderlaan, M. and Phares, W. (1980) Alkaline phosphatase and an acid arylamidase as marker enzymes for normal and transformed W l - 3 8 cells. /. Histochem. Cytochem. 28; 419-26. Donlon, T. A., Lalande, M , Wyman, A., Bruns, G. and Latt, S.A. (1986) Isolation of molecular probes associated with chromosome 15 instability in Prader-Willi syndrome. Proc. Natnl Acad. Sci. USA 83; 4408-12. Doolittle, R. F., Hunkerpiller, M. W., Hood, L. E., Devare, S. G., Robbins, K. C, Aaronson, S. A. and Antoniades, H. W. (1983) Simian sarcoma virus one gene, v-sis, is derived from the gene (or genes) encoding a platelet derived growth factor. Science 211; 275-6. Dovichi, N. J., Martin, J. C, Jett, J. H. and Keller, R. A. (1983) Attogram detection limit for aqueous dye samples by laser-induced fluorescence. Science 219; 845-7. Dovichi, N.J., Martin, J.C., Jett, J.H., Trkula, M. and Keller, R.A. (1984) Laser induced fluorescence of flowing samples as an approach to single molecule detection in liquids. Analyt. Chemistry 56; 348-54. Dowle, C. S., Owainati, A., Robins, A., Burns, K., Ellis, I. O., Elston, C. W. and Blarney, R. W. (1987) Prognostic significance of the DNA content of human breast cancer. Brit. ].

Surg. 74; 133-6. Downward, J., Yardem, Y., Mayes, E., Scrace, G., Totty, N., Stockwell, P., Ullrich, A., Schlessinger, J. and Waterfield, M. D. (1984) Close similarities of epidermal growth factor receptor and v-erb B oncogene protein sequences. Nature 307; 521-7. Dunne, J.F., Thomas, J. and Lee, S. (1989) Detection of mRNA in flow sorted cells. Cytometry 10; 199-204. Durand, R. (1981) Flow cytometric studies of intracellular adriamycin fluorescence in multicell spheroids in vitro. Cancer Res. 41; 3495-8. Durand, R. and Olive, P. (1981) Flow cytometric studies of intracellular adriamycin in single cells in vitro. Cancer Res. 41; 3489—94. Earnshaw, W.C., Halligan, B., Cooke, C.A., Heck, M.M.S. and Liu, L.F. (1985) Topoisomerase II is a structural component of mitotic chromosome scaffolds. /. Cell Biol. 100; 1706-15. Earnshaw, W. C, Halligan, B., Cooke, C. A. and Rothfield, N. (1984) The kinetochore is part of the metaphase chromosome scaffold. /. Cell Biol 98; 352-7. Earnshaw, W. C. and Heck, M. M. S. (1985) Localization of topoisomerase II in mitotic chromosomes. /. Cell Biol 100; 1716-25. Earnshaw, W. C. and Rothfield, N. (1985) Identification of a family of human centromere proteins using autoimmune sera from patients with scleroderma. Chromosoma 91; 313-21. Earnshaw, W.C., Sullivan, K.F., Machlin, P.S., Cooke, C.A., Kaiser, D. A., Pollard, T.D., Rothfield, N. F. and Cleveland, D. W. (1987) Molecular cloning of cDNA for CENP-B, the major human centromere autoantigen. /. Cell Biol. 104; 817-29.

REFERENCES

397

Ehrlich, P., Krause, R., Mosse, M., Rosin, H. and Weigert, C. (1903) In Enzyklopodie der mikroskopischen Technik. Urban and Schwarzenberg, Berlin. Ehrlich, P. and Lazarus, A. (1898) Die anaemia. In Nothnagel's Specielle Pathologie und Therapie, Bd. 8, Vienna. Eisert, W. G. (1981) High resolution optics combined with high spatial reproducibility in flow. Cytometry 1; 254-9. Eisert, W. G. and Dennenloehr, M. (1981) Nozzle design for the generation of plane liquid surfaces. Cytometry 1; 249-53. Ekfors, T. O., Lipastl, J., Nurmi, M. J., and Eerrola, E. (1987) Flow cytometric analysis of the DNA profile of renal cell carcinoma. Pathol Res. Pract. 182; 58-62. Elias-Jones, J., Hendy-Ibbs, P., Cox, H. Evan, G. I. and Watson, J. V. (1986) Cervical brush biopsy specimens suitable for DNA and oncoprotein analysis using flow cytometry. /. Clin. Path. 39; 577-81. Epstein, A. L. and Clevenger, C. V. (1985) Indentification of nuclear antigens in human cells by immunofluorescence, immunoelectron microscopy, and immunocytochemical methods using monoclonal antibodies. In Recent Advances in Non-histone Chromosomal Protein Research, Vol. 1, (ed. I. Bekhor) pp. 117-37. CRC Press, Boca Raton Press, FL. Epstein, R.J. (1988) Hormone modulation of drug-induced DNA damage and repair in human tumour cells. Ph. D. Thesis, Cambridge University. Epstein, R.J., Smith, P.J. and Watson, J.V., Waters, C. and Bleehen, N.M. (1989) Oestrogen potentiates topoisomerase-II-mediated cytotoxicity in activated subpopulation of human breast cancer cells: implications for cytotoxic drug resistance in solid tumours. Int. ]. Cancer 44; 501-5. Epstein, R. J., Watson, J. V. and Smith, P. J. (1988) Subpopulation analysis of drug-induced cell-cycle delay in human tumours using 90° light scatter. Cytometry 9; 349-58. Euler, L. (1755) Principes generaux du mouvement des fluides. Hist, de l'Acad. de Berlin. Euler, L. (1759) De principiis motus fluidorum. Novi Comm. Acad. Petrop. 15; 1-57. Evan, G. I. and Hancock, D. C. (1985) Studies on the interaction of the human c-myc protein with cell nuclei: p62 c ~ m K as a member of a discrete subset of nuclear proteins. Cell 43; 253-61. Evan, G. I., Lewis, G. K., Ramsay, G. and Bishop, J. M. (1985) Isolation of monoclonal antibodies specific for human and mouse proto-oncogene products. Mol. Cell Biol. 5; 3610-16. Ewers, S. B., Langstrom, E., Baldetorp, B. and Killander, D. (1984) Flow cytometric DNA analysis in primary breast carcinomas and clinicopathological correlations. Cytometry 5; 408-19. Fantes, J.A., Green, D.K., Malloy, P. and Summer, A.T. (1989) Flow cytometry measurements of human chromosome kinetochore labeling. Cytometry 10; 134-42. Feichter, G. E. and Goerttler, K. (1986) Pitfalls in the preparation of nuclear suspension from paraffin-embedded tissue for flow cytometry. Cytometry 7; 616. Fellner-Feldegg, H. (1985) Dual laser flow cytometry: Focal length compensation when focussing through a single lens. Cytometry 6; 286-9. Finan, P.J., Quirke, P., Dixon, M.R., Dyson, J.E.D., Giles, G.P. and Bird, C.C. (1986) Is DNA aneuploidy a good prognostic indicator in patients with advanced colorectal cancer. Brit. ]. Cancer 54; 327-30. Finlay, C. A., Hinds, P.W. and Levine, A.J. (1989) The p53 proto-oncogene can act as a suppressor of transformation. Cell 57; 1083-93.

398

REFERENCES

Fisher, G. A. (1961) Increased levels of folk acid reductase as a mechanism of resistance to amethopterin in leukaemic cells. Biochem. Pharmacol. 7; 75-80. Fisher, G. A. (1962) Defective transport of amethopterin (methotrexate) as a mechanism of resistance to the antimetabolite in L5178Y leukaemia cells. Biochem. Pharmacol 11; 1233-4. Fleisher, T. A., Marti, G. E. and Hegengruber, C. (1988) Two-color flow cytometric analysis of monocyte depleted human blood lymphocyte subsets. Cytometry 9; 309-15. Fleming, A. (1922) On a remarkable bacteriolytic element found in tissues and secretions. Proc. R. Soc. London. B93; 306-17. Flintoff, W. E., Davidson, S. V. and Siminovitch, L. (1976) Isolation and partial characterization of three methotrexate-resistant phenotypes from Chinese hamster ovary cells. Somatic Cell Genet. 2; 245-61. Fojo, A. T., Ueda, K., Slamon, D. J., Poplack, D. G., Gottesman, M. M. and Pastan, I. (1987) Expression of a multidrug-resistant gene in human tumours and tissues. Proc. Natnl. Acad. Sci. USA 84, 265-9. Foon, K.A. and Todd, R. F. (1986) Immunologic classification of leukaemias and lymphoma. Blood 68; 1-31. Forgacs, I. C, Sunderesan, V., Evan, G., Wight, D. G. D., Neale, G., Hunter, J. O. and Watson, J. V. (1986) Abnormal expression of c-myc oncogene product in dysplasia and neoplasia associated with ulcerative colitis. Gut 27; 1285a. Fossa, S. D., Thorud, E., Shoaib, M. C, Petterson, E. O., Hois, J. and Scott Knudsen, O. (1984) DNA flow cytometry in primary breast carcinoma. Ada Path. Microbiol. Immunol. Scand. Sect. A, 92; 475-80. Fox, M. H. (1980) A model for the computer analysis of synchronous DNA distributions obtained by flow cytometry. Cytometry 1; 71-7. Fox, M. H. and Coulter, J. R. (1980) Enhanced light collection in a flow cytometry. Cytometry 1; 21—5.

Franke, W.W., Schmid, F., Osborne, M. and Webber, K. (1978) Different intermediatesized filaments distinguished by immunofluorescence microscopy. Proc. Natnl. Acad. Sci. USA 75; 5034-8. Franke, W.W., Schmid, E., Schiller, D.L., Winter, S., Jarasch, E.D., Moll, R., Denk, H., Jackson, D. W. and Illmensee, K. (1981) Differentiation related patterns of expression of proteins of intermediate-sized filaments in tissues and cultured cells. Cold Spring Harb. Symp. Quant. Biol. 46, 431-53. Fredericq, E. (1971) The chemical and physical properties of nucleohistones. In Histones (ed. D.M.P. Phillips) p. 135. Plenum Press, New York. Fried, J. (1976) Method for the quantitative evaluation of data from flow cytofluorometry. Comp. Biomed. Res. 9; 263-76. Fried, J. (1977) Analysis of deoxyribonucleic acid histograms from flow cytometry. J. Histochem. Cytochem. 25; 942-51. Fried, J. and Mandel, M. (1979) Multi-user system for analysis of data from flow cytometry. Comput. Programs Biomed. 10; 218-30. Fried, J., Peres, A. G. and Clarkson, B. D. (1976) Flow cytofluorimetric analysis of cell cycle distributions using propidium iodide. /. Cell Biol. 71; 172-81. Fried, J., Peres, A. G. and Clarkson, B.D. (1978) Rapid hypotonic method for flow cytofluorimetry of monolayer cell cultures. Some pitfalls in staining and data analysis. /. Histochem. Cytochem. 26; 921-33.

REFERENCES

399

Friedlander, M.L., Hedley, D.W., Swanson, C. and Russell, P. (1988) Prediction of longterm survival by flow cytometric analysis of cellular DNA content in patients with advanced ovarian cancer. /. Clin. Oncol 6; 282-90. Friedlander, M L , Hedley, D.W. and Taylor, I.W. (1984) Clinical and biological significance of aneuploidy in human tumours - a review. /. Clin. Path. 37; 961-74. Friedlander, M.L., Russell, P., Taylor, I.W. and Hedley, D.W. (1984) Flow cytometry analysis of cellular DNA content as an adjunct to the diagnosis of ovarian tumours of borderline malignancy. Pathology 16; 301-6. Fulwyler, M.J. (1965) Electronic separation of biological cells by volume. Science 150; 910-11 Fulwyler, M.J. (1977) Hydrodynamic orientation of cells. ]. Histochem. Cytochem. 25; 781-3. Furth, M. E., Davis, L. J., Fleurdelys, B. and Scolnick,- E. (1982) Monoclonal antibodies to the p21 product of the transforming gene of Harvey murine sarcoma virus and of the cellular ras gene family. /. Virol 43; 294-304. Fuscoe, J. C, Clark, L. M. and van Dilla, M.A. (1986) Construction of fifteen human chromosome-specific DNA libraries from flow-purified chromosomes. Cytogenet. Cell Genet. 43; 79-86. Ganesan, A. K., Smith, C. A. and van Zeeland, A. A. (1981) In DNA Repair: A Laboratory Manual of Research Techniques Vol. 1 (eds. E. C. Friedberg and P. C. Hanawalt) p. 89. Marcel Dekker, New York. Gansler, T., Chatten, J., Varello, M., Bunin, G. R. and Atkinson, B. (1986) flow cytometry DNA analysis of neuroblastoma. Correlation with histology and clinical outcome. Cancer 58; 2453-8. Gapski, G. R., Whiteley, J. M., Rader, J. I., Cramer, P. L. Henderson, G. B., Neef, V. and Huennekens, F. M. (1975) Synthesis of a fluorescent derivative of amethopterin. J. Med. Chem. 18; 526-8. Gatti, G., Houssier, C. and Fredericq, E. (1975) Binding of ethidium bromide to ribosomal RNA. Biochim. Biophys. Ada 407; 308-19. Gaudray, P., Trotter, J. and Wahl, G. M. (1986) Fluorescent methotrexate labelling and flow cytometric analysis of cells containing low levels of dihydrofolate reductase. ]. Biol.

Chem. 261; 6285-92. Gautschi, W. (1964) Error function and Fresnel intergrals. In Handbook of Mathematical Functions (eds M. Abramowitz and I. A. Stegun) chapter 7, p. 297. Applied Mathematics series 55, National Bureau of Standards, USA. Gerdes, J. Schwab, U., Lemke, H. and Stein, H. (1983) Production of a mouse monoclonal antibody reactive with a human nuclear antigen associated with cell proliferation. Int. ]. Cancer 31; 13-20. Gershey, E. L. (1980) SV-40 infected muntjac cells: cell kinetics, ploidy and T antigen concentration. Cytometry 1; 37-41. Gibson, N.W., Mattes, W.B. and Hartley, J.A. (1985) Identification of specific DNA lesions of chlorethylating agents: Chloroethylnitrosoureas, chloroethylmethanesulphonates and chloroethylimidazotetrazines. Pharma. Ther. 31; 153-63. Glazer, A.N. and Shyer, L. (1984) Phycoflour probes. Trends Biochem. Sri. 9; 423-7. Glisson, B. S., Gupta, R., Smallwood-Kentro, J. and Ross, W. W. (1986) Characterisation of aquired resistance in a Chinese hamster ovary cell line: loss of drug-stimulated DNA cleavage activity. Cancer Res. 46; 1934—8.

400

REFERENCES

Goad, C. A. (1978) A clustering algorithm for mixtures of monotone densities. Los Alamos Sci. Lab. Rep. LA-7120-MS (Jan). Goldenberg, G. J. and Begleiter, A. (1984) Alterations of drug transport. In Antitumour Drug Resistance (eds B. W. Fox and M. Fox) chapter 10, pp. 241-98, Handb. Exp. Pharm. 72. Springer-Verlag, Berlin, Heidelberg, New York, Tokyo. Goldenberg, G. J., Vanstone, C. L. and Bihler, I. (1971) Transport of nitrogen mustard on the transport carrier for choline in L5178Y lymphoblasts. Science 172; 1148—9. Goldenberg, G. J., Vanstone, C. L. Israels, L. G., Use, D. and Bihler, I. (1970) Evidence for a transport carrier for nitrogen mustard in nitrogen mustard-sensitive and -resistant L5178Y lymphoblasts. Cancer Res. 30; 2285-91. Golding, J. W. (1978) The use of staphylococcal protein-A as an immunological reagent. /. Immumol Meth. 20; 241-53. Goldsmith, H. L. and Mason, S. G. (1961) Axial migration of particles in Poiseuille flow. Nature 190; 1095-6. Gooderham, K. and Jeppeson, P. (1983) Chinese hamster metaphase chromosomes isolated under physiological conditions. A partial characterization of associated nonhistone proteins and protein cores. Exp. Cell Res. 144; 1-14. Govi, G. (1885) L'Ottica di Tolomeo de Eugenia. Modern translation of Ptolomy's Optics. Turin. Grace, A., Patterson, B., Gullane, P., Cummings, M. B. and van Nostrand, P. (1988) Flow cytometry of adenoid cystic carcinoma predicts tumour recurrence and survival. Cytometry Suppl. 2; lla. Gratzner, H. G. (1982) Monoclonal antibody to 5-bromo- and 5-iododeoxyuridine: A new reagent for detecting DNA replication. Science 218; 474-5. Gratzner, H. G. and Leif, R. C. (1981) An immunofluorescence method for monitoring DNA synthesis by flow cytometry. Cytometry 1; 385-9. Gratzner, H. G., Leif, R. C, Ingram, D. J. and Castro, A. (1975) The use of antibody specific for bromodeoxyuridine for the immunofluorescent determination of DNA replication in single cells and chromosomes. Exptl. Cell Res. 95; 88-94. Gray, J. W. (1974) Cell cycle analysis from computer synthesis of deoxyribonucleic acid histograms. /. Histochem. Cytochem. 22; 6A2-50. Gray, J. W. (1976) Cell-cycle analysis of perturbed cell populations; Computer simulations of sequential DNA distribution. Cell and Tissue Kinet. 9; 499-510. Gray, J. W. (1980) Flow cytometry and cell kinetics: relation to cancer therapy. In Flow Cytometry IV (eds O. Laerum, T. Lindmo and E. Thorud) p. 485. Universitetforlaget, Oslo, Norway. Gray, J.W., Carrano, V.A., Moore, D.H., Steinmetz, L.L., Minkler, J., Mayall, B.H., Mendelsohn, M. L. and van Dilla, M. A. (1975a) High-speed quantitative karyotyping by flow microfluorometry. Clin. Chem. 21; 1258-62. Gray, J. W., Carrano, V. A., Steinmetz, L. L, van Dilla, M. A., Moore, D. H., Mayall, B. H. and Mendelsohn, M. L. (1975b) Chromosome measurement and sorting by flow systems. Proc. Natnl Acad. Sci. USA 72; 1231-4. Gray, J. W., Dean, P. N., Fuscoe, J. C, Peters, D. C, Trask, B. J. van den Engh, G. J. and van Dilla, M. A. (1987) High-speed chromosome sorting. Science 238; 323-9. Gray, J. W., Dean, P. N. and Mendelsohn, M. L. (1979) Quantitative cell-cycle analysis. In Flow Cytometry and Sorting (eds M. Melamed, P. Mullaney and M. L. Mendelsohn) p. 383. John Wiley and Sons, New York, Chichester, Brisbane, Toronto.

REFERENCES

401

Gray, J. W. and Langlois, R. G. (1986) Chromosome classification and purification using flow cytometry and sorting. Ann. Rev. Biophys. Chem. 15; 195-235. Gray, J. W., Langlois, R. G., Carrano, A. V., Burkhart-Schultz, K. and van Dilla, M. A. (1979) High resolution chromosome analysis: one and two parameter flow cytometry. Chromosoma 73; 9-27. Gray, J. W. and Mayall, B. H. (1985) Monoclonal Antibodies Against Bromodeoxyuridine. Alan R. Liss, New York. Gray, J. W., Trask, B. J., van der Engh, G., Silva, A., Lozes, G., Grell, S., Schonberg, S., Yu, LC. and Golbus, M. S. (1988) Application of flow karyotyping in prenatal detection of chromosomal aberrations. Am. ]. Hum. Genet. 42; 49-59. Green, N. M. and Joynson, M. A. (1970) A preliminary crystallographic investigation of avidin. Biochem. ]. 118; 71-2. Green, N. M. and Toms, E. J. (1970) Purification and crystallization of avidin. Biochem. J. 118; 67-70. Green, S., Walter, P., Kurman, V., Krust, A., Bornert, J.-M., Argos, P. and Chambon, P. (1986) Human oestogen receptor cDNA: sequence, expression and homology with v-erb-A. Nature 320; 134-9. Greenberg, M. E. and Ziff, E. B. (1984) Stimulation of 3T3 cells induces transcription of the c-fos proto-oncogene. Nature 311; 433-8. Griffin, J. D., Ritz, J., Nadler, L. M. and Schlossman, S. F. (1981) Expression of myeloid differentiation antigens on normal and malignant myeloid cells. /. Clin. Invest. 68; 932—9. Griffioen, A.W., Rijkers, G.T., Keij, J. and Zegers, B.J.M. (1990) Measurement of cytoplasmic calcium in lymphocytes using flow cytometry. Kinetical studies and single cell analysis. /. Immumol. Methods 120; 23—7. Gross, L. (1951) Spontaneous Leukaemia developing in C3H mice following innoculation, in infancy, with AK-leukaemia extracts, or AK-embryos. Proc. Soc. Exp. Biol, Med. 76; 27-32. Grynkiewicz, G., Poenie, M. and Tsien, R. Y. (1985) A new generation of Ca 2+ indicators with greatly improved fluorescence properties. /. Biol. Chem. 260; 3440-50. Guesdon, J.-L., Ternynck, T. and Avrameas, S. (1979) The use of avidin-biotin interaction in immunoenzymatic techniques. /. Histochem. Cytochem. 27; 1131-9. Guibault, G. G. and Kramer, D. N. (1966) Lipolysis of fluorescein and eosin esters. Kinetics of hydrolysis. Analyt. Biochem. 14; 28—32. Gusella, J. F., Keys, C, Varsanyi-Breiner, A., Kao, F., Jones, C, Puck, T. T. and Housman, D. (1980) Isolation and characterization of DNA segments from specific human chromosomes. Proc. Natnl Acad. Sci. USA 77; 2829-33. Haber, D. A., Beverly, S. M., Kiely, M L. and Schimke, R. T. (1981) Properties of an altered dihydrofolate reductase encoded by amplified genes in cultured mouse fibroblasts. /. Cell Biol 256; 9501-10. Haeuptle, M.-T., Aubert, M. L, Djione, J. and Kraehenhubl, J.-P. (1983) Binding sites for lactogenic and somatogenic hormones from rabbit mammary gland and liver. /. Biol. Chem. 258; 305-14. Hamming, R. W. (1973) Numerical Methods for Scientists and Engineers. McGraw-Hill, New York. Hamori, E., Arndt-Jovin, D. J., Grimwade, B. G. and Jovin, T. M. (1980) Selection of viable cells with known DNA content. Cytometry 1; 132-5. Hanggi, U. J. and Littlefield, J. W. (1976) Altered gene regulation of the rate of synthesis of

402

REFERENCES

dihydrofolate reductase in methotraxate-resistant hamster cells. /. Biol. Chem. 251;

3075-82. Hann, S. Rv Thompson, C. B. and Eisenman, R. N. (1985) c-myc oncogene protein is independent of the cell cycle in human and avian cells. Nature 314; 366-9. Harlow, E., Crawford, L.V., Pimm, D. C. and Williamson, N.M. (1981) Monoclonal antibodies specific for Simian Virus 40 tumour antigen. /. Virol. 39; 861-9. Harmon, P., Maus, R. and Morrissey, W. (1988) Expert Systems - Tools and Applications. John Wiley, New York. Harper, M. E., Franchini, G., Loue, J., Simon, M. L, Gallo, R. C. and Wong-Stall, F. (1983) Chromosomal sublocalization of human c-myb and c-fes cellular one genes. Nature 304; 169-71. Harrap, K. R. and Hill, B. T. (1970) The selectivity of action of alkylating agents and drug resistance. 3) The uptake and degradation of alkylating drugs by Yoshida sarcoma cells in vitro. Biochem. Pharmacol 19 209-17. Harris, P., Morton, C. C, Guglielmi, P., Li, F., Kelly, K. and Latt, S. A. (1986) Mapping by chromosome sorting of several gene probes, including c-myc, to the derivative chromosomes of a 3:8 translocation associated with familial renal cancer. Cytometry 7; 589-94. Harrison, M. F. (1953) Percentage of binucleate cells in livers of adult rats. Nature 171; 611. Hartmann, N. R. and Pederson, T. (1970) Analysis of the kinetics of granulosa cell populations in the mouse ovary. Cell and Tissue Kinet. 3 1—11. Hayden, G. E., Walker, K. Z., Miller, J. F. A. P., Wotherspoon, J. S. and Raison, R. L. (1988) Simultaneous cytometric analysis for the expression of cytoplasmic and surface antigens in activated T-cells. Cytometry 9; 44—51. Hecht, R. M., Schomer, D. F., Oro, J. A., Bartel, A. H. and Hungerford, E. V. (1981) Simple adaptations to extend the range of flow cytometry five orders of magnitude for the DNA analysis of uni- and multicellular systems. /. Histochem. Cytochem. 29; 771-4. Hedley, D.W. (1989) Flow cytometry using paraffin-embedded tissue: five years on. Cytometry 10; 229-41. Hedley, D.W., Friedlander, M. L and Taylor, I.W. (1985) Application of DNA flow cytometry to paraffin-embedded archival material for the study of aneuploidy and its clinical significance. Cytometry 6; 327-33. Hedley, D. W., Friedlander, M.I., Taylor, I.W., Rugg, C.A. and Musgrove, E.A. (1983) Method for analysis of cellular DNA content of paraffin-embedded pathological material using flow cytometry. /. Histochem. and Cytochem. 31; 1333—5. Hedley, D. W., Rugg, C. A. and Gelber, R. D. (1987) Association of S-phase fraction with prognosis of node positive early breast cancer. Cancer Res. 47; 4729—35. Hedley, D. W., Rugg, C. A., Ng., A. B. P. and Taylor, I. W. (1984) Influence of cellular DNA content on disease-free survival in stage II breast cancer patients. Cancer Res. 4t4t;

5395-8. Henderson, G. B., Russell, A. and Whiteley, J. M. (1980) A fluorescent derivative of methotrexate as an intracellular marker of dihydrofolate reductase in L1210 cells. Arch. Biochem. Biophys. 202; 29-34. Hendy-Ibbs, P., Cox, H., Evan, G. I. and Watson, J. V. (1987) Flow cytometric quantitation of DNA and c-myc oncoprotein in archival biopsies of uterine cervix neoplasia. Brit. ]. Cancer 55; 275-82. Herweijer, H., van den Engh, C. J. and Nooter, K. (1989) A rapid and sensitive flow cytometric method for the detection of multidrug-resistant cells. Cytometry 10; 463-8.

REFERENCES

403

Herzog, A. and Schutze, H. R. (1969) Fluorochromierung von saugetierchromosomen mit einem basisch substituierte bis-benzimidazol derivat (Praparten 33258 'Hoechst'). Deut. Tierarztl Wschr. 75; 467-76. Herzberger, M. (1966) Optics from Euclid to Huygens. Applied Optics 9; 1383-93. Herzenberg, L. A., Sweet, R. G. and Herzenberg, L. A. (1976) Fluorescence activated cell sorting. Sci. American 243; 108-12. Hidderman, W., Schumman, J., Andreefr", M., Barlogie, Bv Herman, C. J., Leif, R. C, Mayall, B. H., Murphy, R,. F. and Sandberg, A. A. (1984) Convention on nomenclature for DNA cytometry. Cytometry 5; 445—6. Hill, B. T. (1972) Studies on the transport and cellular distribution of chlorambucil in the Yoshida ascites sarcoma. Biochem. Pharmacol. 21; 495-502. Hill, H. D., Sumner, G. K. and Waters, M. D. (1968) An automated fluorimetric assay for alkaline phosphatase using 3-O-methylfluoresein phosphate. Analyt. Biochem. 24; 9-14. Hilwig, I. (1970) On the influence of a benzimidazole derivative (fluorochrome) on cell lines in tissue culture. Z. Zellforsch. 104; 127-37. Hilwig, I. and Gropp, A. (1975) pH dependent fluorescence of DNA and RNA in cytologic staining with 33258 Hoechst. Exptl Cell Res. 91; 457-60. Hinds, P., Finlay, C. and Levine, A. J. (1989) Mutation is required to activate the p53 gene for cooperation with the ras oncogene and transformation. /. Virol. 63; 739—46. Hoang, T., Gilmore, D., Metcalf, D., Cobbold, S., Watt, S. M., Clark, M. R. and Waldmann, H. (1983) Separation of hemopoetic cells from adult mouse marrow by use of monoclonal antibodies. Blood 61; 580—8. Hogg, N., Ross, J. D., Jones, D. B., Slusarenko, M., Walport, M. J. and Lachmann, P. J. (1984) Identification of an anti-monocyte monoclonal antibody that is specific for membrane complement receptor type 1 (CRl). Eur. ]. Immunol. 14; 236-43. Hollander, Z. and Loken, M. R. (1988) Simultaneous analysis of DNA content and surface antigens in human bone marrow. Cytometry 9; 485-90. Holtzer, H., Bennet, G. S., Tapscott, S. J., Croop, J. M. and Toyama, Y. (1981) Intermediatesized filaments: Changes in syntheses and distribution in cells of the myogenic and neurogenic lineages. Cold Spring Harb. Symp. Quant. Biol. 46; 317—29. Hooke, R. (1665) Obfers IX. Of the colours obfervable in Mufcovy Glafs and other thin bodies. In Micrographia: on some phyfiological descriptions of minute bodies made by magnifying glasses with observations and inquiries thereupon, pp. 47—67. Jo. Montyn and Ja. Allefetry, London MDCLXV. Royal Society Library, London. Hopwood, D. (1985) Cell tissue fixation, 1972-1982. Histochem. J. 17; 389-442. Horan, P. K., Slezak, S. E. and Poste, G. (1986) Improved flow cytometric analysis of leukocyte subsets: Simultaneous identification of 5 cell subsets using two-color immunofluorescence. Proc. Natnl. Acad. Sci. USA 83; 8361-5. Horsfall, D.J., Tiley, W.D., Orell, S.R., Marshal, V.R. and McCant, E.L. (1986) Relationship between ploidy and steroid hormone receptors in primary invasive breast cancer. Brit. ]. Cancer 53; 23-8. Hosaka, Y., Rainwater, L. M., Grant, C. S., Young, W. F., Farrow, G. M., van Heerden, J. A. and Lieber, M. M. (1988) Adrenocortical carcinoma: nuclear deoxyribonucleic acid ploidy studied by flow cytometry. Surgery 102; 1027-31. Howard, A. and Pelc, S. R. (1951) Nuclear incorporation of 32P as demonstrated by autoradiographs. Exptl Cell Res. 2; 178-87. Hudson, B., Upholt, W. B., Devinny, J. and Vinograd, J. (1969) The use of an ethidium analogue in dye-buoyant density procedure for the isolation of closed circular DNA; the

404

REFERENCES

variation of the superhelix density of mitochondrial DNA. Proc. Natnl. Acad. Sci. USA 62; 813-20. Hulett, H.R., Bonner, W.A., Barrett, J. and Herzenberg, L.A. (1969) Cell sorting: Automated separation of Mammalian cells as a function of intracellular fluorescence. Science 166; 747-9. Hulett, W. R., Bonner, W. A., Sweet, R. G. and Herzenberg, L. A. (1973) Development and application of a rapid cell sorter. Clin. Chem. 19, 813-16. Hunt, S. P., Pini, A. and Evan, G. (1987) Induction of c-/bs-like protein in spinal cord neurones following sensory stimulation. Nature 328; 632—4. Hurley, J.B., Simon, ML, Teplow, D.B., Robinshaw, J.D. and Gilman, A.G. (1984) Homologies between signal transducing G proteins and ras gene products. Science 226;

860-3. Huygens, C. (1690) Traite de la Lumiere. Dover Publications, New York; reprinted 1962 as Treatise on Light.

Imoto T., Johnson, L.N., North, A.C.T., Phillips, D.C. and Rupley, D. (1972) Vertebrate lysosyme. In The Enzymes Vol. 7, (ed. P. D. Boyer) chapter 21, pp. 606-868. Academic Press, New York, London. Inaba, M. and Sakurai, Y. (1979) Enhanced efflux of actinomycin-D, vincristine and vinblastine in adriamycin-resistant subline of P388 leukaemia. Cancer Lett. 8; 111-15. Inaba, M., Kobayashi, H., Sakurai, Y. and Johnson, R.LC. (1979) Active efflux of daunorubicin and adriamycin in sensitive and resistance sublines of P388 leukaemia. Cancer Res. 39; 2200-3. ISCN (1985) An international system for human cytogenetic nomenclature. In Cytogenetics and Cell Genetics (eds D. G. Harnden and H. P. Klinger) p. 11, Karger, A. G., Basle. Iversen, O.E. and Laerum, O.D. (1987) Trout and salmon erythrocytes and human leukocytes as internal standards for ploidy control in flow cytometry. Cytometry 8; 190-202. Jabbar, S. A. B., Twentyman, P. R. and Watson, J. V. (1989) The MTT assay underestimates the growth inhibitory effects of interferons. Brit. ]. Cancer 60; 523-8. Jacobberger, J. W., Fogleman, D. and Lehman, J. M. (1986) Analysis of intracellular antigens by flow cytometry. Cytometry 7; 356-64. Jacobsen, A.-B., Fossa, S.D., Lunde, S., Melvik, J.E. and Petersen, E.D. (1987) Flow cytometric DNA measurements in paraffin-embedded bladder carcinoma tissue before and after pre-cystectomy radiotherapy. Radiother. Oncol. 10; 149-53. Jacobsen, A. (1983) The use of trout erythrocytes and human lymphocytes for standardization in flow cytometry. Cytometry 4; 161—6. Jacobsen, A., Skovgaard-Poulsen, H., Lindegaard Madsen, E., Ellebaek Petersen, S. and Sommer Hansen, H. (1984) Ploidy level of human breast carcinoma. Acta Radiol. Oncol. 23; 103-7. Jefferys, A. J. (1979) DNA sequence variants in the G y-, A y-, 5- and (3-globin genes of man. Cell 18, 1-10. Jett, J. H. (1978) Mathematical analysis of DNA histograms from asynchronous and synchronous cell populations. In Third International Symposium on Pulse-cytophotometry (ed. D. Lutz) p. 93. European Press, Ghent. Johnson, L.V., Walsh, M. L, Bockus, B.J. and Chen, L. B. (1981) Monitoring of relative mitochondrial membrane potential in living cells by fluorescence microscopy. /. Cell Biol

88; 526-35.

REFERENCES

405

Johnston, D.A., White, R.A. and Barlogie, B. (1978) Automatic processing and interpretation of DNA distributions: comparison of several techniques. Comp. Biomed. Res. 11; 393-404. Johnston, R. G., Bartholdi, M. F., Hiebert, R. D., Parson, J. D. and Cram, L. S. (1985) Slit-scan flow cytometer for recording simultaneous waveforms. Rev. Sci. Instrum. 56; 691—5. Jongkind, J. F., Verkerk, A. and Sernetz, M. (1986) Detection of acid-^-galactosidase activity in viable human fibroblasts by flow cytometry. Cytometry 7; 463-6. Jovin, T. M. (1979) Fluorescence polarization and energy transfer: theory and applications. In Flow Cytometry and Sorting (eds M. R. Melamed, P. F. Mullaney and M. L. Mendelsohn) pp. 137-65. John Wiley and Sons, New York, Chichester, Brisbane, Toronto. Jovin, T. M., Morris, S. J., Striker, G., Schultens, H. A., Digweed, M. and Arndt-Jovin, D. J. (1976) Automated sizing and separation of particles by ratios of light scatter intensities. /. Histochem. Cytochem. 24; 269-83. Juliano, R. L. and Ling, V. (1976) A surface glycoprotein modulating drug permeability in Chinese hamster ovary cell mutants. Biochim. Biophys. Ada 455; 152—62. Kachal, V., Glossner, E. and Schneider, H. (1982) A new flow cytometric transducer for fast sample throughput and time resolved kinetic studies of biological cells and other particles. Cytometry 3; 202—12. Kachal, V., Kordwig, E. and Glossner, E. (1977) Uniform lateral orientation of flat particles in flow through systems caused by flow forces. /. Histochem. Cytochem 25; 774-80. Kachal, V. and Menke, E. (1979) Hydrodynamic properties of flow instruments. In Flow Cytometry and Sorting (eds M. R. Melamed, P. F. Mullaney and M. L. Mendelsohn) chapter 3, pp. 41-59. John Wiley and Sons, New York, Chichester, Brisbane, Toronto. Kallioniemi, O.-P., Hietonen, T., Mattila, J., Lehtinen, M., Lauslahti, K. and Koivula, T. (1987) Aneuploid DNA content and high S-phase fraction of tumour cells are related to poor prognosis in patients with primary breast cancer. Europ. ]. Cancer Clin. Oncol. 23;

277-82. Kallioniemi, O.-P., Mattila, J., Punnonen, R. and Koivula, T. (1988a) DNA ploidy levels and cell cycle distribution in ovarian cancer: relationship to histopathological features of the tumour. Int. ]. Gynaecol. Pathol. 7; 1-11. Kallioniemi, O.-P., Punnonen, R., Mattila, J., Lehtinen, M. and Koivula, T. (19881?) Prognostic significance of DNA index, multiploidy, and S-phase fraction in ovarian cancer. Cancer 61; 334-9. Kamentsky, L. A. and Melamed, M. A. (1967) Spectrophotometric cell sorter. Science 156; 1364-5. Kamentsky, L. A., Melamed, M. R. and Derman, H. (1965) Spectrophotometer: a new instrument for ultrarapid cell analysis. Science 150; 630—1. Kanda, N., Schrek, R., Alt, F., Bruns, G., Baltimore, D. and Latt, S. A. (1983) Isolation of amplified DNA sequences from IMR-32 human neuroblastoma cells: facilitation by fluorescence-activated cell sorting of metaphase chromosomes. Proc. Natnl. Acad. Sci. USA 80; 4069-73. Kaplan, A. S., Caldarelli, D. D., Chacho, M. S., Bruce, D. R., Hutchinson, J., Conway, S. and Coons, J. S. (1986) Retrospective DNA analysis of head and neck squamous cell carcinoma. Arch. Otolaryngol. 112; 1159-62. Kaplow, L. S. (1975) Substitute for benzidine in myeloperoxidase stains. Am. J. Clin. Pathol 63; 451-4.

406

REFERENCES

Kaplow, L. S. (1979) Leukocyte peroxidase and non-specific esterase. In Flow Cytometry and Sorting (eds M. R. Melamed, P. F. Mulaney and M. L. Mendelsohn) chapter 29, pp. 531—45. John Wiley and Sons, New York, Chichester, Brisbane, Toronto. Kaplow, L. S., Dauber, H. and Lerner, E. (1976) Assessment of monocyte esterase activity by flow cytophotometry. /. Histochem. Cytochem. 24; 363—72. Kaplow, L. S. and Eisenberg, M. (1975) Leukocyte differentiation and enumeration by cytochemical-cytographic analysis. In First International Symposium on Pulsecytophotometry (eds C. A. M. Haanen, H. F. P. Hellen and J. H. C. Wessels) pp. 262-74. European Press Medicon, Ghent. Kapuscinsky, J. and Darzynkiewicz, Z. (1987) Interaction of pyronin Y(G) with nucleic acids. Cytometry 8; 129-37. Kapuscinsky, J., Darzynkiewicz, Z. and Melamed, M. R. (1982) Luminescence of the solid complexes of acridine orange with RNA. Cytometry 2; 201-11. Karn, J., Watson, J. V., Lowe, A. D., Green, S. M. and Vedeckis, W. (1989) Regulation of cell cycle duration by c-myc levels. Oncogene 4; 773—87. Keij, J. F., Griffioen, A. W., The, T. H. and Rijkers, G. T. (1989) INCA: software for Consort 30 analysis of flow cytometric calcium determinations. Cytometry 10; 814—17. Kelly, K., Cochran, B.H., Stiles, C D . and Leder, P. (1983) Cell specific regulation of the c-myc gene by lymphocyte mitogens and platelet derived growth factor. Cell 35; 603-10. Kelly, K., Cochran, B. H., Stiles, C. D. and Leder, P. (1984) The regulation of c-myc by growth signals. Curr. Topics Microbiol Immun. 113; 117—26. Kenter, Amy L. and Watson, J. V. (1987) Cell cycle kinetics model of LPS-stimulated spleen cells correlates switch region rearrangements with S-phase. ]. Immunol Meth. 97; 111-17. Kenter, Amy L, Watson, J. V., Azim, Tasmin and Rabbitts, T. H. (1986) Colcemid inhibits growth during early G l in normal but not in tumorigenic lymphocytes. Exp. Cell Res. 167; 241-51. Kerker, M., van Dilla, M. A., Brunsting, A. Kratohvil, J. P., Hsu, P., Wang, D.-S., Gray, J. W. and Langlois, R. G. (1982) Is the central dogma of flow cytometry true: That fluorescence intensity is proportional to cellular DNA content? Cytometry 3; 71-8. Kersten, W., Kersten, H. and Szybalski, N. (1966) Physicochemical properties of complexes between deoxyribonucleic acid and antibiotics which affect ribonucleic acid synthesis (actinomycin, daunomycin, cinerubin, nogalamycin, chromomycin, mithramycin and olivomycin). Biochemistry 5; 236-44. Khalfan, H., Abuknesha, R., Rand-Weaver, M., Price, R.G. and Robinson, D. (1986) Aminomethyl coumarin acetic acid: a new fluorescent labelling agent for proteins. Histochem. J. 18, 497-9. Kim, M. and Perry, S. (1977) Mathematical methods for determining cell DNA synthesis rate and age distribution utilizing flow micro-fluorometry. /. Theor. Biol. 68; 27—42. Knudson, A. G. (1971) Mutation and cancer: statistical study of retinoblastoma. Proc. Natnl Acad. Sci. USA 68; 820-3. Knudson, A. G. (1978) Retinoblastoma: a prototypic hereditary neoplasm. Semin. Oncol. 5;

57-60. Knudson, A. G. (1983) Model hereditary cancers of man. Prog. Nucleic Acid Res. Mol. Biol. 29; 17-25. Knudson, A. G. (1985) Hereditary cancer, oncogenes and antioncogenes. Cancer Res. 45; 1437-43.

REFERENCES

407

Knudson, A. G., Meadows, A. T v Nichols, W. W. and Hill, R. (1976) Chromosomal deletion and retinoblastoma. N. Engl. ]. Med. 295; 1120-3. Kohler, G. and Milstein, C. (1975) Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256; 495-6. Kokal, W., Sheibani, K., Terz, J. and Harada, R. (1986) Tumor DNA content in prognosis of colorectal carcinoma. /. Am. Med. Ass. 255; 3123-7. Kooi, M. W., Aten, J. A., Stap, J., Kipp, J. B. A. and Barendsen, G. A. (1984) Preparation of chromosome suspensions from cells of a solid experimental tumour for measurement by flow cytometry. Cytometry 5; 547—9. Kosower, N.S., Kosower, E.M, Newton, G.L. and Ranney, A.M. (1979) Bimane fluorescent labels: labeling of normal human red blood cells under physiological conditions. Proc. Natnl. Acad. Sci. USA 75; 33S2-6. Kraemer, P., Deaven, L, Crissman, H. and Dilla, M. (1972) DNA constancy despite variability in chromosome number. In Advances in Cell and Molecular Biology Vol. 2 (ed. E.J. Dupraw) pp. 47-108. Academic Press, New York. Krisch, K. (1971) Carboxylic ester hydrolyases. In The Enzymes Vol. 5 (ed. P. D. Boyer) pp. 43-69. Academic Press, New York. Krishan, A. (1975) Rapid flow cytofluorimetric analysis of mammalian cell cycle by propidium iodide staining. /. Cell Biol 66; 188-93. Krishan, A. (1987) Effects of drug efflux blockers on vital staining of cellular DNA with Hoechst 33342. Cytometry 8; 642-5. Krishan, A. and Ganapathi, R. (1979) Laser flow cytometry and cancer chemotherapy: detection of intracellular anthracyclines by flow cytometry. /. Histochem. cytochem. 2 7;

1655-6. Krishan, A. and Ganapathi, R. (1980) Laser flow cytometric studies on intracellular fluorescence of anthracyclines. Cancer Res. 40; 3895-900. Krishan, A., Sauerteig, A., Gordon, K. and Swinkin, C. (1986) Flow cytometric monitoring of cellular anthracycline accumulation in murine leukaemia cells. Cancer Res. 46;

1768-73. Krishan, A., Sridhar, K.S., Davila, E., Vogel, C. and Sternheim, W. (1987) Patterns of anthracycline retention modulation in human tumour cells. Cytometry 8; 306-14. Krumlauf, R., Jeanpierre, M. and Young, B. (1982) Construction and characterization of genomic libraries from specific human chromosomes. Proc. Natnl. Acad. Sci. USA 79; 2971-5. Kumar, A. A., Freisheim, J. H., Kempton, R. J., Anstead, G. M., Black, A. M. and Judge, L. (1983a) Synthesis and characterization of a fluorescent analogue of methotrexate. /. Med. Chem. 36; 111-13. Kumar, A. A., Kempton, R.J., Anstead, G.M., Proce, E.M. and Freisheim, J.H. (1983b) High-performance liquid chromatography of methotrexate analogues containing terminal lysine or ornithine and their dansyl derivates. Anal. Biochem. 128; 191—5. Kurnich, M. B. (1955) Pyronin-Y in the methyl green-pyronin Histological stain. Stain Technol 30; 213-30. Kurtz, I. and Balaban, R. S. (1985) Fluorescence emission spectroscopy of 1,4-dihydroxyphtholonitrile. Biophys. ]. 48; 499-508. Laemmli, U. K., Cheng, S. M., Adolph, K. W., Paulson, J. R., Brown, J. A. and Baumbach, W. R. (1978) Metaphase chromosome structure: the role of non-histone proteins, Cold Spring Harbor Symp. Quant. Biol. 42; 351—60.

408

REFERENCES

Lakhanpal, S., Gonchroff, N.J., Katzmann, J.A. and Habdwerger, B.S. (1987) A flow cytofluorometric double staining technique for simultaneous determination of human mononuclear cell surface phenotype and cell cycle phase. /. Immunol. Methods 96; 35-40. Lalande, M , Dryja, T. P., Schreck, R. R., Shipley, J., Flint, A. and Latt, S. A. (1984a) Isolation of human chromosome 13-specific DNA sequences cloned from flow sorted chromosomes and potentially linked to the retinoblastoma locus. Cancer Genet. Cytogenet. 13; 283-95. Lalande, M , Kunkel, L.M., Flint, A. and Latt, S.A. (19841?) Development and use of metaphase chromosome flow sorting methodology to obtain recombinant phage libraries enriched for parts of the human X chromosome. Cytometry 5; 101—7. Lalande, M., Ling, V. and Miller, R. G. (1981) Hoechst 33342 dye uptake as a probe of membrane permeability changes in mammalian cells. Proc. Natnl Acad. Sci. USA 78;

363-7. Lalande, M., Schrek, R.R., Hoffman, R. and Latt, S.A. (1985) Identification of inverted duplicated 15 chromosomes using bivariate flow cytometric analysis. Cytometry 6; 1—6. Lam, H.Y. P., Talgoy, M. M. and Goldenberg, G.J. (1980) Uptake and decomposition of chlorozotocin in L5178Y lymphoblasts in vitro. Cancer Res. 40; 3950—5. Lammler, G. and Schutze, H. R. (1969) Vital-fluorochromierung tierischen zellkerne mit einen neuen fluorochrom. Naturwiss. 56; 286. Lampidis, T. J., Bernal, S. D., Summerhays, I. C. and Chen, L. B. (1982) Rhodamine 123 is selectively toxic and preferentially retained in carcinoma cells in vitro. Annals N. Y. Acad. Sci. 397; 299-302. Landschultz, W. H., Johnson, P. F. and McKnight, S. L. (1988) The Leucine Zipper: A hypothetical structure common to a new class of DNA binding proteins. Science 240; 1759-65. Lane, D.P. and Crawford, L.V. (1979) T-antigen is bound to a host protein in SV40transformed cells. Nature 278; 261-3. Lane, M. J., Dabrowiak, J. C. and Vournakis, J. N. (1983) Sequence specificity of actinomycin D and netropsin binding to pBR322 DNA analysed by protection from DNase I. Proc. Natnl Acad. Sci. USA 80, 3260-4. Langlois, R. G., Carrano, A. V., Gray, J. W. and van Dilla, M. A. (1980) Cytometrial studies of metaphase chromosomes by flow cytometry. Chromosoma 77; 229—51. Langlois, R. G., Yu, L. C, Gray, J. W. and Carrano, A. V. (1982) Quantitative karyotyping of human chromosomes by dual laser beam flow cytometry. Proc. Natnl. Acad. Sci. USA

79; 7876-80. Lanier, L. L. and Loken, M. R. (1984) Human lymphocyte subpopulations identified by using three-color immunofluorescence and flow cytometry analysis. Correlation of Leu2, Leu-3, Leu-7, Leu-8 and Leu-11 cell surface antigen expression. /. Immunol 132; 151-6. Lansing-Taylor, D., Waggoner, A. S., Murphy, R. F., Lanni, F. and Birge, R. R. (1986) Applications

of Fluorescence in the Biomedical Sciences (eds D. Lansing-Taylor, A. S.

Waggoner, R. F. Murphy, F. Lanni and R. R. Birge). Alan R. Liss, New York. Latt, S. A. (1977) Fluorimetric detection of deoxyribonucleic acid synthesis; possibilities for interfacing bromodeoxyuridine dye techniques with flow cytometry. /. Histochem. Cytochem. 25; 913-26. Latt, S.A. (1979) Fluorescent probes for DNA microstructure and synthesis. In Flow

REFERENCES

409

Cytometry and Sorting (eds M. R. Melamed, P.F. Mullaney and M L. Mendelsohn) pp. 263-84. John Wiley and Sons, New York, Chichester, Brisbane, Toronto. Latt, S. A v George, Y. S. and Gray, J. W. (1977) Flow cytometric analysis of bromodeoxyuridine substituted cells stained with 3325S Hoechst. /. Histochem. Cytochem. 25; 927-34. Latt, S. A., Juergens, L. A., Mathews, D. J., Gustashaw, K. M. and Sahar, E. (1980a) Energy transfer-enhanced chromosome banding. An overview. Cancer Genet. Cytogenet. 1; 187-96. Latt, S. A., Sahar, E., Eisenhard, M. E. and Juergens, L. A. (1980b) Interactions between pairs of DNA-binding dyes: Results and implications for chromosome analysis. Cytometry 1; 2-12. Latt, S.A. and Stetten, G. (1976) Spectral studies on 33258 Hoechst and related bisbenzimidazole dyes useful for fluorescent detection of deoxyribonucleic acid synthesis. /. Histochem. Cytochem. 24; 24—33. Latt, S. A. and Wohlleb, J. (1975) Optical studies of the interaction of Hoechst with DNA, chromatin and metaphase chromosomes. Chromosoma 52; 297-316. Lazarides, E. (1980) Intermediate filaments as mechanical integrators of cellular space. Nature 283; 249-56. Lazarides, E. (1981) Intermediate filaments: Chemical heterogeneity in differentiation. Cell 23; 649-50. Lazarides, E. (1982) Intermediate filaments: a chemically heterogeneous developmentally regulated class of proteins. Ann. Rev. Biochem. 51; 219-50. Lebo, R. V., Bruce, B. D., Dazin, P. F. and Payan, D. G. (1987) Design and application of a versatile triple-laser cell and chromosome sorter. Cytometry 8; 71—82. Lebo, R. V., Carrano, A. V., Buckhardt-Schultz, K., Dozy, A. M , Yu, L.-C. and Kan, Y. W. (1979) Assignment of the (}-, y- and 5-globin genes to the short arm of chromosome 11 by chromosome sorting and DNA restriction analysis. Proc. Natnl. Acad. Sci. USA 76;

5804-8. Lebo, R. V., Gorin, F., Fletterich, R. J., Rao, F. T., Cheung, M C, Bruce, B. D. and Kan, Y. H. (1984) High-resolution chromosome sorting and DNA spot blot analysis assign McArdle's syndrome to chromosome 11. Science 225; 57—60. Lebo, R. V., Kan, Y. W., Cheung, M.C., Carrano, A.V., Yu, L.-C, Chang, J.C, Cordell, B. and Goodman, H. M. (1982) Assigning the pleomorphic insulin gene to the short arm of chromosome 11 by chromosome sorting. Human Genet. 60; 10—15. Le Bouteiller, P. P., Mishal, Z., Lemonnier, A. and Kourilsky, F. M. (1983) Quantification by flow cytofluorimetry of HLA class I molecules at the surface of murine cells transformed by cloned HLA genes. /. Immunol. Meth. 61; 301—15. Lech, K., Anderson, K. and Brent, R. (1988) DNA bound fos proteins activate transcription in yeast. Cell 52; 179-84. Lee, L. G., Chen, C.-H. and Chiu, L. A. (1986) Thiazole orange: A new dye for reticulocyte analysis. Cytometry 7; 508-17. Lehtinen, T., Lehtinen, M , Aine, R., Kallioniemi, O.-P., Leino, T., Hakala, T., Leinikki, P. and Alavaikko, M. (1987) Nuclear DNA content of non-endemic Burkitt's lymphoma./. Clin. Pathol 40; 1201-5. Le-Perq, J. B. and Paoletti, C. (1967) A fluorescent complex between ethidium bromide and nucleic acids. /. Mol. Biol 27; 87-106. Leserman, L. D., Barbet, J., Kourilsky, F. and Weinstein, J.N. (1980) Targeting to cells of

410

REFERENCES

fluorescent liposomes covalently coupled with monoclonal antibody or protein A. Nature 288; 602-4. Leukocyte typing (1984) (eds A. Bernard, L. Boumsell, J. Dausset, C. Milstein and S. F. Schlossman). Springer-Verlag, Berlin, Heidelberg, New York and Tokyo. Levitt, D. and King, M. (1987) Methanol fixation permits flow cytometric analysis of immunofluorescent stained intracellular antigens. /. Immunol. Meth. 96; 233—7. Ling, V. and Thompson, L. H. (1974) Reduced permeability in CHO cells as a mechanism of resistance to colchicine. /. Cell Physiol. 83; 103-16. Ling, V., Gerlach, J. and Kartner, N. (1984) Multidrug resistance. Breast Cancer Res. Treat. 4; 89-94. Linzer, D.I.H. and Levine, A.J. (1979) Characterization of a 54K Dalton cellular SV40 tumour antigen present in SV40-transformed cells and uninfected embryonal carcinoma cells. Cell 17; 43-52. Little, P. F. R., Annison, C, Darling, S., Williamson, Camba, L. and Modell, B. (1980) Model for antenatal diagnosis of beta-thalassaemia and other monogenic disorders by molecular analysis of linked DNA polymorphisms. Nature 285; 144-7. Liu, L.F. (1983) DNA topoisomerases - enzymes that can catalyse the breaking and rejoining of DNA. CRC Crit. Rev. Biochem. 15; 1-24. Loewe, H. and Urbanietz, J. (1974) Baisisch substituerte 2,6-bis-benzimidazole derivate, eine neve chemotherapeutisch active korperklasse. Arzneim. Forsch. 24; 1927—33. Loken, M. R. (1980) Simultaneous quantitation of Hoechst 333A2 and Immunofluorescence on viable cells using fluorescence activated cell sorter. Cytometry 1; 136—42. Loken, M. R. (1986) Cell surface antigen and morphological characterization of leukocyte populations by flow cytometry. In Monoclonal Antibodies (ed. P. C. L. Beverly). Meth. Haematol 13; 132-44. Loken, M. R. and Herzenberg, L. A. (1975) Analysis of cell populations with a fluorescence activated cell sorter. Annals N. Y. Acad. Sci. 254; 163-71. Loken, M R. and Lanier, L. L. (1984) Three-color immunofluorescence analysis of Leu antigens on human peripheral blood using two lasers on a fluorescence activated cell sorter. Cytometry 5; 151-8. Loken, M. R., Parks, D. R. and Herzenberg, L. L. (1977) Identification of cell asymetry and orientation by light scattering. /. Histochem. Cytochem. 35; 790-5. Loken, M R., Sweet, R. G. and Herzenberg, L. A. (1976) Cell discrimination by multiangle light scatter. /. Histochem. Cytochem. 24; 284-91. Lopez, M., Olive, D. and Mannoni, P. (1989) Analysis of cytosolic ionized calcium variation in polymorphonuclear leukocytes using flow cytometry and Indo-1 AM. Cytometry 10; 165-73. Lucas, J. N. and Gray, J. W. (1987) Centromeric index verus DNA content flow karyotypes of human chromosomes measured by means of slit-scan flow cytometry. Cytometry 8; 273-9. Lucas, J.N., Gray, J.W., Peters, D.C. and van Dilla, M.A. (1983) Centromeric index measurement by slit-scan flow cytometry. Cytometry 4; 109—16. Ludlam, T. and Slansky, R. (1977) Search and classification of reaction mechanisms in multiparameter data. Phys. Rev. D16; 100-13. Lukes, R. J. and Collins, R. D. (1974) Immunologic characterization of human malignant lymphomas. Cancer 34; 1488-1503. Lydon, M.J., Keeler, K.D. and Thomas, D.B. (1980) Vital DNA staining and sorting by flow microfluorometry. /. Cell Physiol. 102; 175-81.

REFERENCES

411

Macartney, J. C, Camplejohn, R. S., Adler, J., Stone, M. G., Powell, G. (1986) Prognostic importance of DNA flow cytometry in non-Hodgkin's lymphoma. ]. Clin. Pathol. 39; 542-6. MacDonald, H. R. and Zaech, P. (1982) Light scatter analysis and sorting of cells activated in mixed lymphocyte culture. Cytometry 3; 55-8. MacDonald, P. D.M. (1975) The mathematical theory of exponentially growing cell populations. In Mathematical Models and Cell Kinetics (ed. A.-J. Valleron) p. 15. European Press, Ghent. Mackillop, W. J., Palmer, M. J., O'Sullivan, B., Ward, G. K., Steele, R. and Dotsikas, G. (1988) Clinical trials in cancer: the role of surrogate patients in defining what constitutes an ethically acceptable clinical trial. Brit. ]. Cancer 59; 388—95. Makino, Rv Hayashi, K. A. and Sugimura, T. (1984) c-myc is induced in rat liver at a very early stage of regeneration or by cycloheximide treatment. Nature 310; 697—8. Malaise, E. P. (1983) Reduced oxygen enhancement of the radiosensitivity of glutathionedeficient fibroblasts. Radiat. Res. 95; 486-94. Male, D. K. (1986) Immunology, an Illustrated Guide. Gower Medical Publishing, London, New York. Malin-Berdel, J. and Valet, G. (1980) Flow cytometric determination of esterase and phosphatase activities and kinetics in haemopoietic cells with fluorogenic substrates. Cytometry 1; 222-8. Mann, R. C. (1987) On multiparameter data analysis in flow cytometry. Cytometry 8; 184-9. Manson, M.M., Legg, R.F., Watson, J.V., Green, J.A. and Neale, G.E. (1981) An examination of the relative resistance to aflatoxin Bj and susceptibilities to y-glutamyl p-phenylene daimine mustard of y-glutamyl transferase negative and positive cell lines. Carcinogenesis 2; 661-70. Martin, J. C. and Swartzendruber, D.E. (1980) Time: a new parameter for kinetic measurement in flow cytometry. Science 207; 199-200. Martin, M. M. and Linquist, L. (1975) The pH dependence of fluorescein fluorescence. /. Luminescence 10; 381-90. Martin, R.F. and Holmes, N. (1983) Use of 125I-labelled DNA ligand to probe DNA structure. Nature 302; 452-4. Martinez, A. O., Vigil, A. and Vila, J. C. (1986) Flow cytometric analysis of mitochondrial associated fluorescence in young and old human fibroblasts. Exp. Cell. Res. 164; 551—5. Martin-Zanca, D v Hughes, S. H. and Barbacid, M. (1986) A human oncogene formed by the fusion of truncated tropomyosin and protein tyrosine kinase sequences. Nature 319; 743-8. Masters, J.R.W., Camplejohn, R.S., Millis, R.R. and Rubens, R.D. (1987) Histological grade, elastosis, DNA poloidy and response to chemotherapy of breast cancer. Brit. ]. Cancer 55; 455-7. Masuda, H v Miller, C, Koeffler, H. P., Battifora, H. and Cline, M. J. (1987) Rearrangement of the p53 gene in human osteogenic sarcoma. Proc. Natnl. Acad. Sci. USA 84; 7716-19. Maxwell, J. C. (1891) Treatise on Electricity and Magnetism. Oxford University Press (3rd edition). McDonald, J. K. and Barrett, A. J. (1986^) Mammalian Proteases: a Glossary and Bibliography Volume 1: Endopeptidases. Academic Press, London, Orlando, San Diego, New York, Austin, Montreal, Sydney, Tokyo, Toronto. McDonald, J. K. and Barrett, A. J. (1986b) Mammalian Proteases: a Glossary and Bibliography

412

REFERENCES

Volume 2: Exopeptidases. Academic Press, London, Orlando, San Diego, New York, Austin, Montreal, Sydney, Tokyo, Toronto. McGuire, W. L. and Dressier, L. G. (1985) Emerging impact of flow cytometry in predicting recurrence and survival in breast cancer patients. /. Natnl. Cancer Inst. 75; 405-10. McKusic, V. A. (1962) Genetic factors in intestinal polyposis. /. Am. Med. Ass. 182; 271-7. McNeil, P.L., Kennedy, A.L., Waggoner, A.S., Taylor, D.L. and Murphy, R.F. (1985) Light-scattering changes during chemotactic stimulation of human neutrophils: Kinetics followed by flow cytometry. Cytometry 6; 7—12. Meccham, W. J. and Char, D. H. (1986) DNA content abnormalities and prognosis in uveal melanomas. Arch. Ophthamol 104; 1626-9. Meissel, M. N. (1951) Luminescence-microscope analysis of the functional condition of living matter (Russian text). Izvest. Akad. Nauk. S.S.S.R., Ser. Fiz. 15; 788. (Quoted from Chem. Abstr. (1952) 46; 6693.) Mendelsohn, M. L. (1962) Autoradiographic analysis of cell proliferation in spontaneous breast cancer of C3H mouse III. The growth fraction. /. Natnl. Cancer Inst. 28; 1015-29. Merkel, D.E., Dressier, L.G. and McGuire, W.L. (1987) Flow cytometry, cellular DNA content, and prognosis in human malignancy. /. Clin. Oncol 5; 1690-1703. Metcalf, R. L. and Patton, R. L. (1944) Fluorescence microscopy applied to entomology and allied fields. Stain Technol. 19; 11-27. Meyer, K. (1971) Hyaluronidases. In The Enzymes Vol.5 (ed. P.D. Boyer) chapter 11, pp. 307-20. Academic Press, New York, London. Meyer, R.A., Haase, S.F., Poduslo, S.E. and McKhann, G.M. (1974) Light scattering patterns of isolated oligodendroglia. /. Histochem. Cytochem. 22; 594—7. Meyne, J., Bartholdi, M. F., Travis, G. and Cram, L. S. (1984) Counterstaining human chromosomes for flow karyology. Cytometry 5; 580—3. Michaelis, L. and Menten, M. L. (1913) Die kinetic der invertinwirkung. Biochem. Z. 49;

333-69. Mie, G. (1908) Beitrage zur optik truber medien, speziell holloidaler metallosungen. Ann. Physik. 15; 377-445. Miller, M. R., Castellot, J. J. and Pardee, A. B. (1979) A general method for permeabilizing monolayer and suspension cultured animal cells. Exptl. Cell Res. 120; 421—5. Mirski, S. E. L., Gerlach, J. H. and Cole, S. P. C. (1987) Multidrug resistance in a human small cell lung cancer cell line selected in adriamycin. Cancer Res. 47; 2594—8. Moran, R., Darzynkiewicz, Z., Staiano-Coico, L. and Melamed, M. R. (1985) Detection of 5-bromodeoxyuridine (BrdUdr) incorporation by monoclonal antibodies: Role of the DNA denaturation step. /. Histochem. Cytochem. 33; 821-7. Morgan, S.A., Smith, P.J., Twentyman, P.R. and Watson, J.V. (1987) Hoechst33342: DNA spectral shift analysis of multi-drug resistant human tumour cells. Cytometry Suppl. 1; 344a. Morgan, S. A., Watson, J.V., Twentyman, P.R. and Smith, P.J. (1989) Flow cytometric analysis of Hoechst 33342 uptake as an indicator of multi-drug resistance in human lung cancer. Brit. ]. Cancer 60; 282-7. Moroi, Y., Peebles, C, Fitzler, M. J., Steigerwald, F. and Tan, E. M. (1980) Autoantibody to centromere (kinetochore) in scleroderma sera. Proc. Natnl. Acad. Sci. USA 77; 1672-31. Moroney, M. J. (1951) Facts from Figures. Penguin, London. Mottram, J. C. (1935) On the alteration in the sensitivity of cells towards radiation produced by cold and anaerobiosis. Brit. J. Radiol. 8; 32—9.

REFERENCES

413

Mueller, W. and Crothers, D. (1975) Interaction of heteroaromatic compounds with nucleic acids. 1. The influence of heteroatoms and polarizability on the base specificity of interacting ligands. Eur. ]. Biochem. 54; 267-78. Mueller, W., Bunemann, H. and Dattagupta, N. (1975) Interaction of heteroaromatic compounds with nucleic acids. 2. The influence of substituents on the base and sequence specificity of intercalating ligands. Eur. ]. Biochem. 54; 279-91. Mueller, W. and Gautier, F. (1975) Interaction of heteroaromatic compounds with nucleic acids. A-T-specific non-intercalating ligands. Eur. ]. Biochem. 54; 385-94. Mullaney, P. F., Crowell, J. M., Salzman, G. C, Martin, J. C., Hiebert, R. D. and Goad, C. A. (1976) Pulse-height light-scatter distributions using flow-system instrumentation. /. Histochem. Cytochem. 24; 298-304. Munroe, D. G., Rovinski, R., Bernstein, A. and Benchimol, S. (1988) Loss of highly conserved domain on p53 as a result of gene deletion during Friend virus-induced erythroleukaemia. Oncogene 2; 621—4. Murphy, R. F. (1985) Automated identification of subpopulations in flow cytometric list mode data using cluster analysis. Cytometry 6; 302—9. Murphy, W.M., Chandler, R.W. and TrafTord, R.M. (1986) Flow cytometry in deparaffinised nuclei compared with histological grading for the pathological evaluation of transitional cell carcinoma. /. Urol. 135; 694-8. Musgrove, E., Rugg, C. and Hedley, D. (1986) Flow cytometric measurement of cytoplasmic pH: A critical evaluation of probes. Cytometry 7; 347—55. Nadler, L. M., Anderson, K. C, Marti, G., Bates, M., Park, E., Daley, J. F. and Schlossman, S. P. (1983) B4, a human B lymphocyte associated antigen expressed on normal, mitogen-activated, and malignant B lymphocytes. /. Immunol. 131; 244-50. Nadler, L. M., Stashenko, P., Reinhertz, E. L, Ritz, J., Hardy, R. and Schlossman, S. F. (1982) Expression of normal differentiation antigens on human leukaemia and lymphoma cells. In Malignant Lymphoma: Etiology, Immunology, Pathology, Treatment (eds S. A. Rozengurtz and H. S. Kaplan) pp. 107—20. Academic Press, New York. Naharro, G., Robins, K. and Reddy, E. P. (1984) Gene product of v-fgr one: hybrid protein contains a portion of actin and a tyrosine-specific protein kinase. Science 223; 63-6. Nederlof, P. M., Robinson, D., Abuknesha, R., Wiegunt, J. Hopmann, A. H. N., Tanke, H. J. and Rapp, A. K. (1989) Three-color fluorescence in situ hybridization for the simultaneous detection of multiple nucleic acid sequences. Cytometry 10; 20-7. Nelson, A.J. (1988) A physiological function for multidrug-resistant membrane glycoproteins: a hypothesis regarding the renal organic cation-secretory system. Cancer Chemother. Pharmacol. 22; 92-3. Newell, D. G., Hannam-Harris, A. and Smith, J. L. (1983) The ultrastructural localization of immunoglobulin in chronic lymphocytic lymphoma cells: Changes in light and heavy chain distribution induced by mitogen stimulation. Blood 61; 511-19. Newton, I. (1672a) A letter of Mr Isaac Newton, Professor of the Mathematicks in the University of Cambridge; containing his new theory about light and colours: sent by the author to the publisher from Cambridge, 6 th February 1671/2; in order to be communicated to the Royal Society. Phil. Trans. R. Soc. Lond. 6; 3075-87. Newton, I. (1672b) Answers to some considerations upon His doctrine of light and colours. Phil. Trans. R. Soc. Lond. 7; 5084-103. Newton, I. (1687) Philosophiae Naturalis Principia Mathematica. Original latin manuscript Royal Society of London, facsimile reprint by William Dawson Ltd, London.

414

REFERENCES

Newton, I. (1704) The light of the Sun confifts of Rays of differently refrangible. In Opticks: on, a Treatise of the Reflexions, Refractions, Inflexions and Colours of Light, Prop. II, Theor. II,

Exper. 3. pp. 18-22. Sam. Smith and Benj. Walford, London MDCCIV, in the Wren Library bench mark NQ.16.198, Trinity College, Cambridge. Nicola, N. A., Metcalf, D., von Melcher, H. and Burgess, A. W. (1981) Isolation of murine foetal hemopeotic progenitor cells and selective fractionation of various erythroid precursors. Blood 58; 376-86. Nicolini, C, Belmont, A., Parodi, S., Lessin, S. and Abraham, S. (1979) Mass action and acridine orange: Static and flow cytofluorometry. /. Histochem. Cytochem. 27; 102-13. Nielsen, O., Larsen, J. K.; Christensen, I. J. and Lernmark, A. (1982) Flow sorting of mouse pancreatic B cells by forward and orthogonal light scatter. Cytometry 3; 177-81. Niman, H. L, Houghten, R. A., Walker, L. E., Reisfeld, R. A., Wilson, I. A., Hogle, J. M. and Lerner, R. A. (1983) Generation of protein-reactive antibodies by short peptides in an event of high frequency: Implications for the structural basis of immune recognition. Proc. Natnl. Acad. Sci. USA 80; 4949-53. Niman, H. L., Thompson, A. M. H., Yu, A., Markman, M , Willems, J.J., Herwig, K. R., Habib, N.A., Wood, C.B., Houghton, R.A. and Lerner, R.A. (1985) Anti-peptide antibodies detect oncogene-related proteins in urine. Proc. Natnl Acad. Sci. USA 82; 7724-8. Nooter, K., Oostrum, R., Jonker, R. R., van Dekken, H., Stokdijk, W. and van de Engh, C. J. (1989) Effects of cyclosporin A on daunorubicin accumulation in multidrug-resistant P388 leukaemia cells measured by real-time flow cytometry. Cancer Chemother. Pharmacol. 23; 296-300. Nooter, K., van den Engh, C.J. and Sonneveld, P. (1983) Quantitative flow cytometric determination of anthracycline content of rat bone marrow cells. Cancer Res. 43; 5126-30. Norris, W. (1820) Case of fungoidal disease. Edinburgh Med. Surg. ]. 16; 562-5. Nusse, M. and Kramer, J. (1984) Flow cytometric analysis of micronuclei found in cells after irradiation. Cytometry 5; 20—5. Ohnuki, Y. (1965) Demonstration of the spiral structures of human chromosomes. Nature 208; 916-17. Ohnuki, Y. (1968) Structure of chromosomes, morphological studies of the spiral structure of human somatic chromosomes. Chromosoma 25; 402—28. Olsen, A. S., McBride, O. W. and Otey, M. C. (1980) Isolation of unique sequence human X chromosome deoxyribonucleic acid. Biochemistry 19; 2419-28. Ono, K. and Schrieve, D. C. (1986) Enhancement of EMT6/SF tumour cell killing by mitomycin C and by cyclophosphamide following in vivo administration of buthione sulfoximine. Int. ]. Radiat. Oncol Biol Phys. 12; 1175-8. Orkin, S., Alter, B., Atlay, C, Mahoney, M , Lazarus, H., Hobbins, J. and Nathan, D. (1978) Application of endonuclease mapping to the analysis of prenatal diagnosis of thalassaemia caused by globin-gene deletion. N. Engl ]. Med. 299; 166—72. Orkin, S., Old, J., Wetherall, D. and Nathan, D. (1979) Partial deletion of beta-globin gene DNA in certain patients with beta O-thalassaemia. Proc. Natnl Acad. Sci. USA 76; 2400-4. Ormerod, M. G., Payne, A.W. R. and Watson, J.V. (1987) Improved program for the analysis of DNA histograms. Cytometry 8; 637-41. Osborn, ML, Geisler, N., Shaw, G., Sharp, G. and Weber, K. (1981) Intermediate filaments. Cold Spring Harb. Symp. Quant. Biol 46; 2490-4.

REFERENCES

415

Otten, G.R. and Loken, M. R. (1982) Two color light scattering identifies physical differences between lymphocyte subpopulations. Cytometry 3; 182-7. Otto, F. J. and Oldiges, H. (1980) Flow cytogenetic studies in chromosomes and whole cells for the detection of clastogenetic effects. Cytometry 1; 13—17. Otto, F. and Tsou, K. C. (1985) A comparative study of DAPI, DIPI and Hoechst 33258 and 33342 as chromosomal DNA stains. Stain Technol 60; 7-11. Oud, P. S., Hendrik, J. B. J., Beck, H. L. M., Veldhuitzen, J. A. M , Vooijs, G. P., Herman, C. J. and Ramaekers, F. C. S. (1985) Flow cytometric analysis and sorting human endometrial cells after immunocytochemical labelling for cytokeratin using a monoclonal antibody. Cytometry 6; 159-64. Pallavichini, M. G., Dolbeare, F. A., Gray, J. W., Cohen, A. M. and Dethlefsen, L. A. (1977) Separation of crypt and villus cell populations by flow cytometry. 3rd International Symposium of Pulse Cytophotometry, Vienna. Parker, J. W. (1988) Flow cytometry in the diagnosis of lymphomas. Cytometry Suppl. 3; 38-43. Parks, D. R., Hardy, R. R. and Herzenberg, L. L. (1984) Three-color immunofluorescence analysis of mouse B-lymphocyte subpopulations. Cytometry 5; 159—68. Paulson, J. R. and Laemmli, U. K. (1977) The structure of histone depleted metaphase chromosomes. Cell 12; 817-25. Perry, S. and Reynolds, J. (1956) Methyl green—pyronin as a differential nucleic acid stain for peripheral blood smears. Blood 11; 1132-9. Peters, D., Branscomb, E., Dean, P., Merrill, T., Pinkel, D., van Dilla, M. and Gray, J. W. (1985) The LLNL high-speed sorter: Design features, operational characteristics and biological utility. Cytometry 6; 290-301. Pfeiffer-Ohlsson, S., Goustin, A. S., Rydnert, J., Wahlstrom, T., Bjersing, L, Stehelin, D. and Ohlsson, R. (1984) Spatial and temporal pattern of cellular myc oncogene expression in developing human placenta; Implications for embryonic cell proliferation. Cell 38;

585-96. Plato (c. 370 BC) Republic. Book X, verse 602. Portela, R. A. and Stockert, J. C. (1979) Chromatin fluorescence by pyronin staining. Experimentia 35; 1663—5. Potmesil, M., Hsiang, K.-H., Liu, L. F., Bank, B., Grossberg, H., Kirschenbaum, S., Forlenzar, T. J., Penzinen, A., Kangonis, D., Knowles, D., Traganos, F. and Siber, R. (1988) Resistance of human leukaemic and normal lymphocytes to drug-induced DNA cleavage and low levels of DNA topoisomerase II. Cancer Res. 48; 3537—A3. Pretlow, T. G., Wein, E. E. and Zettergen, J. G. (1975) Problems connected with separation of different kinds of cells. In International Review of Experimental Pathology, Vol. 14 (eds. G.W. Richter, and M. A. Epstein) p. 91. Academic Press, New York. Price, B. J., Kollman, V. H. and Salzman, G. C. (1978) Light scatter analysis of microalgae. Correlations of scatter patterns from pure and mixed asynchronous cultures. Biophys. ]. 22; 29-36. Price, B. J., Salzman, G. C, Crowell, J. M., Ingram, M., Hansen, K. M. and Mullaney, P. F. (1976) Biologic cell classification by multi-angle light scattering flow system. Biophys. }. 16; 64A. Ptolomy (c. AD 150) Opticks. See Govi, G. (1885). Alexandria. Quastler, H. and Sherman, F. G. (1959) Cell population kinetics in the intestinal epithelium of the mouse. Exptl. Cell Res. 17; 420-38. Quirke, P., Dixon, M. R., Clayden, A. D., Durdey, P., Dyson, J. E. D., Williams, N. S. W. and

416

REFERENCES

Bird, C. C. (1987) Prognostic significance of DNA aneuploidy and cell proliferation in rectal adenocarcinomas. /. Pathol. 151; 285-91. Rabbitts, P. H v Watson, J. V., Lamond, A., Fischer, W., Forester, A., Stinton, M. A., Evan, G. I., Atherton, E., Sheppard, R. C. and Rabbitts, T. H. (1985) Metabolism of c-myc gene products: c-myc mRNA and protein expression in the cell cycle. Embo ]. 4; 2009—15. Rabinovitch, P. S., June, C. H., Grossman, A. and Ledbetter, J. A. (1986) Heterogeneity among T-cells in intracellular free calcium response after mitogen stimulation with PHA or anti-CD-3. Simultaneous use of indo-1 and immunofluorescence with flow cytometry. /. Immunol. 137; 952-61. Rainwater, L. M., Hosaka, Y., Farrow, G. M., Kramer, S. A., Kelalis, P. P. and Lieber, M. M. (1987a) Wilms' tumors: relationship of deoxyribonucleic acid ploidy to patient survival. /. Urol. 138; 974-7. Rainwater, L. M., Hosaka, Y., Farrow, G. M. and Licber, M. M. (1987b) Well differentiated clear cell carcinoma: significance of nuclear deoxyribonucleic acid patterns studied by flow cytometry. /. Urol. 137, 15-19. Ramaekers, F. C. S., Beck, H., Vooijs, P. and Herman, C. J. (1984) Flow-cytometric analysis of mixed cell populations using intermediate filament antibodies. Exp. Cell. Res. 153; 249-53. Ramaekers, F.C.S., Puts, J.J.G., Kant, A., Moesker, O., Jap, P.H.K. and Vooijs, G.P. (1982a) Antibodies to intermediate filaments in the characterization of human tumours. Cell. Biol. Int. Rep. 6; 625-32. Ramaekers, F.C.S., Puts, J.J.G., Kant, A., Moesker, O., Jap, P.H.K. and Vooijs, G.P. (1982b) Differential diagnosis of human carcinomas, sarcomas and their metastases using antibodies to intermediate filaments. Eur. ]. Cancer Clin. Oncol. 18; 1251-7. Ramaekers, F. C. S., Puts, J. J. G., Moesker, O., Kant, A., Huysmans, A., Haag, D., Jap, P.H.K. and Herman, C.J. (1983) Antibodies to intermediate filament proteins in the immunohistochemical identification of human tumours: an overview. Histochem. ]. 15; 691-713. Ransom, J. T., DiGiusto, D. L. and Cambier, J. (1987) Flow cytometric analysis of intracellular calcium mobilization. Methods Enzymol. 141; 53-63. Rayleigh, J.W.S. (1871) On the scattering of light by small particles. Phil. Mag. 41; 447-54. Reeve, J. G., Rabbitts, P. H. and Twentyman, P. R. (1989) Amplification and expression of mdr-1 gene in a multi-drug variant of a small cell lung cancer cell line NCI-H69. Brit. ]. Cancer 60; 339-42. Reingertz, E. L, Kung, P. C., Goldstein, G. and Levey, R. H. (1980) Discrete stages of human intrathymic differentiation: Analysis of normal thymocytes and leukaemic lymphoblasts of T-cell lineage. Proc. Natnl. Acad. Sci. USA 77; 1588-92. Reynolds, O. (1883) An experimental investigation of the circumstances which determine whether the motion of water shall be direct or sinuous, and the law of resistance in channels. Phil. Trans. R. Soc. Eondon. 174; 935-83. Rice, G.C., Bump, E.A., Shrieve, D.C., Lee, W. and Kovacs, M. (1986a) Quantitative analysis of cellular glutathione in Chinese hamster ovary cells by flow cytometry utilising monochlorobimane: some applications to radiation and drug resistance in vitro and in vivo. Cancer Res. 46; 6105-10. Rice, G. C, Hoy, C. A. and Schimke, R. T. (1986b) Transient hypoxia induces amplification of the dihydrofolate reductase gene. Proc. Natnl. Acad. Sci. USA 83; 5978-82.

REFERENCES

417

Rinaldini, L. M. J. (1958) The isolation of living cells from animal tissues. Int. Rev. Cytol 7;

587-647. Rink, T. J., Montecucco, C, Hesketh, T. R. and Tsien, R. Y. (1980) Lymphocyte membrane potential assessed with fluorescent probes. Biochim. Biophys. Acta 595; 15-30. Riordan, J. R. and Ling, V. (1985) Genetic and biochemical characterization of multidrug resistance. Pharmac. Ther. 28; 51-75. Riozin-Towle, L, Hall, E.J., Flynn, M., Baglow, J.E. and Varnes, M.E. (1982) Enhanced cytotoxicity of melphalan by prolonged exposure to nitroimidazoles: the role of endogenous thiols. Int. J. Radiat. Oncol. Biol. Phys. 8; 757-60. Rodenberg, G J., Cornelisse, G J., Heintz, P. A., Hermans, J. and Fleuren, G. J. (1987) Tumor ploidy as a major prognostic feature in advanced ovarian cancer. Cancer 59; 317-23. Rodenberg, C.J., Cornelisse, C.J., Hermans, J. and Fleuren, G.J. (1988) DNA flow cytometry and morphometry as prognostic indicators in advanced ovarian cancer: a step forward in predicting the clinical outcome. Gynaecol. Oncol. 29; 176—87. Rockwell, S. C, Kallman, R. F. and Fajardo, L.F. (1972) Characteristics of a serially transplanted mouse mammary tumour and its tissue culture adapted derivative. /. Natnl. Cancer Inst. 49; 735-47. Rosowsky, A., Wright, J. E., Cucchi, C.A., Boeheim, K. and Frei, E. (1986) Transport of a fluorescent antifolate by methotrexate-sensitive and methotrexate-resistant human leukaemic lymphoblasts. Biochem. Pharmacol. 35; 356-60. Rosowsky, A., Wright, J.E., Shapiro, H., Beardsley, P. and Lazarus, H. (1982) A new fluorescent dihydrofolate reductase probe for studies of methotrexate resistance. /. Biol

Chem. 157; 14162-7. Ross, D. D., Joneckis, C G and Schiffer, G A. (1986) Effects of verapamil on the in vitro intracellular accumulation and retention of daunorubicin in blast cells from patients with acute non lymphatic leukaemia. Blood 68; 83—8. Ross, D.D., Ordonez, J.V., Joneckis, C.C., Testa, J.R. and Thompson, B.W. (1988) Isolation of highly multidrug resistant P388 cells from drug sensitive P388/S cells by flow cytometric cell sorting. Cytometry 9; 359—67. Ross, D.D., Thompson, B.W., Ordonez, J. V. and Joneckis, C.C (1989) Improvement of flow cytometric detection of multidrug-resistant cells by cell volume normalization of intracellular daunorubicin content. Cytometry 10; 185—91. Rotman, B. and Papermaster, B. W. (1966) Membrane properties of living mammalian cells as studied by enzymatic hydrolysis of fluorogenic esters. Proc. Natnl Acad. Sci. USA 55; 134-41. Rous, P. (1911) Transmission of a malignant new growth by means of a cell-free filtrate. /. Am. Med. Assoc. (Chicago) 6; 198-202. Roussel, M.F., Rettenmier, C.W., Look, A.T. and Scherr, C.J. (1984) Cell surface expression of v-/ms-coded glycoproteins is required for transformation. Mol Cell Biol 4; 1999-2009. Ryan, D. H., Mitchell, S. J., Hennessy, L. A., Bauer, K. D., Horan, P. K. and Cohen, H. J. (1984) Improved detection of rare CALLA-positive cells in peripheral blood using multiparameter flow cytometry. /. Immunol Meth. 74; 115—28. Sahar, E. and Latt, S. A. (1978) Enhancement of banding patterns in human chromosomes by energy transfer. Proc. Natnl Acad. Sci. USA 75; 5650-4. Sahar, E. and Latt, S. A. (1980) Energy transfer and binding competition between dyes to enhance staining differentiation in metaphase chromosomes. Chromosoma 79; 1—28.

418

REFERENCES

Sainsbury, J. R. C, Farndon, J. Rv Sherbet, G. V. and Harris, A. L. (1985) Epidermal growth factor receptors and oestrogen receptors in human breast cancer. Lancet i; 364—6. Salzman, G.C, Crowell, J.M., Goad, C A., Hansen, K.H., Heibert, R.D., LaBauve, P.M., Martin, J. C. and Mulaney, P. F. (1975a) A flow-system multiangle light-scattering instrument for cell characterization. Clin. Chem. 21; 1297-1304. Salzman, G.C., Crowell, J.M., Hanse, K.M., Ingram, M. and Mullaney, P. F. (1976) Gynaecologic specimen analysis by multiangle light scatter in a flow cytometer. /. Histochem. Cytochem. 24; 308-14. Salzman, G. C, Crowell, J. M., Martin, J. C, Truijillo, T. T., Romero, A., Mullaney, P. F. and LaBauve, P. M. {1975b) Cell classification by laser light scattering: identification and separation of unstained leukocytes. Ada Cytol. (Baltimore) 19; 374—7. Salzman, G. C, Mullaney, P. F. and Price, B. J. (1979) Light-scattering approaches to cell characterization. In Flow Cytometry and Sorting (eds M. R. Melamed, P. F. Mullaney and M. L. Mendelsohn) chapter 5, pp. 105—24. John Wiley and Sons, New York, Chichester, Brisbane, Toronto. Sanders, M., Maners, A., Snyderman, N., Owens, R. and Milligan, E. (1988) The prognostic importance of clinical and flow cytometric parameters in head and neck squamous carcinomas: Prospective and retrospective studies. Cytometry Suppl. 2; 44a. Sap, J., Munoz, A., Damm, K., Goldberg, Y., Ghysdal, J., Leutz, A., Beug, H. and Vennstrom, B. (1986) The c-erb-A protein is a high-affinity receptor for thyroid hormone. Nature 324; 635-40. Scambler, P. J., Wainwright, B. J., Watson, E., Bates, G., Bell, G., Williamson, R. and Farrall, M. (1986) Isolation of further anomalous informative DNA sequence from chromosome seven closely linked to cystic fibrosis. Nucleic Acid Res. 14; 1951-6. Scamrov, A.V. and Beabealashvilli, R.S. (1983) Binding of actinomycin D to DNA revealed by DNase I footprinting. Febs Lett. 164; 97-101. Schabel, F. M., Trader, M. W., Laster, W. R., Wheeler, G. P., Witt, M. H. (1978) Patterns of resistance and therapeutic synergism among alkylating agents. Antibiot. Chemother. 23; 200-15. Scherr, C. J., Rettenmeir, C. W., Sacca, R., Roussel, M. F., Look, A. T. and Stanley, E. R. (1985) The c-fms proto-oncogene product is related to the receptor for the mononuclear phagocytic growth factor, CSF 1. Cell 41; 665-76. Schimke, R. T. (1984) Gene amplification, drug resistance and cancer. Cancer Res. 44; 1735-42. Schliwa, M. and van Blerkom, J. (1981) Structural interaction of cytoskeletal components. /. Cell Biol. 90; 222-35. Schmeckpeper, B.J., Smith, K.D., Dorman, B.P., Ruddle, F.H. and Talbot, C.C. (1979) Partial purification and characterization of DNA from the human X chromosome. Proc. Natnl. Acad. Sci. USA 76; 6525-8. Schmidt, D., Wiedermann, B., Keil, W., Sprenger, E. and Harms, D. (1986) Flow cytometric analysis of nephroblastomas and related neoplasms. Cancer 58; 2494-500. Schrieve, D. C, Denekamp, J. and Mitchinton, A. I. (1985) Effects of glutathione depletion by buthione sulfoximine on radiosensitization by oxygen and misonidazole in vivo. Radiat. Res. 102; 283-94. Schroff, R.W., Bucana, C D . , Klein, R.A., Farrell, M.M. and Morgan, A.C. (1984) Detection of intracytoplasmic antigens by flow cytometry. /. Immunol. Meth. 70,

167-77. Schuette, W.H., Shackney, S.E., MacCollum, M.A. and Smith, C.A. (1983) High

REFERENCES

419

resolution method for the analysis of DNA histograms that is suitable for the detection of multiple aneuploid G l peaks in clinical samples. Cytometry 3; 376-86. Schummelfeder, N., Ebschner, K. R. and Krogh, E. (1957) Die grundelage der diferenten fluorochromierung von ribo- und deoxyribonuklrinsaure mit akridinorange. Naturwissenschafler 44; 467-8. Schweitzer, D. (1976) Reverse fluorescent chromosome banding with chromomycin and DAPI. Chromosoma 58; 307-24. Schweitzer, D. (1981) Counterstain-enhanced chromosome banding. Hum. Genet. 57; 1-14. Scott, N. A., Welland, H. S., Moertel, C. G., Cha, S. S., Beart, R. W. and Lieber, M. M. (1987) Colorectal cancer, Duke's stage, tumour site, preoperative CEA level, and patient prognosis related to tumour DNA ploidy pattern. Arch. Surg. 122; 1375-9. Sculthorpe, H.M. (1978) Metachromasia. Mel Lab. Sci. 35; 365-70. See, Y. P., Carlsen, S.A., Till, J.E. and Ling, V. (1974) Increased drug permeability in Chinese hamster ovary cells in the presence of cyanide. Biochim. Biophys. Ada 373; 242-52. Shapiro, H. M. (1981) Flow cytometric estimation of DNA and RNA content in intact cells stained with Hoechst 33342 and pyronin-Y. Cytometry 2; 143—50. Shapiro, H. M., Schildkraut, R., Curbelo, R., Turner, R. B., Webb, R. H., Brown, D. C. and Block, M. J. (1977) Cytomat-R: A computer-controlled multiple laser source multiparameter flow cytophotometer system. /. Histochem. Cytochem. 25; 836-44. Sharpless, T.K., Bartholdi, M. and Melamed, M. R. (1977) Size and refractive index dependence of simple forward angle scattering measurements in a flow system using sharply-focussed illumination. /. Histochem. Cytochem. 25; 845-56. Shortman, K. (1972) Physical procedures for the separation of animal cells. Annu. Rev. Biophys. Bioeng. 1; 93-130. Sieinski, W., Dorsett, B. and Ioachim, H. L. (1981) Identification of prekeratin by immunofluorescence in the differential diagnosis of tumours. Human Pathol. 12; 452-8. Sikora, K., Chan, S. Y. T., Evan, G. I., Gabra, R , Markham, N., Stewart, J. and Watson, J. V. (1987) c-myc oncogene expression in colorectal cancer. Cancer 59; 1289-95. Sikora, K., Evan, G., Stewart, J. and Watson, J. V. (1985) Detection of the c-myc oncogene product in testicular cancer. Brit. ]. Cancer 52; 171—6. Sillar, R. and Young, B. D. (1981) A new method for the preparation of metaphase chromosomes for flow analysis. /. Histochem. Cytochem. 29; 74—8. Sims, P.J., Waggoner, A.S., Wang, C.H. and Hoffman, J.S. (1974) Studies on the mechanism by which cyanin dyes measure membrane potential in red blood cells and phosphatidyl cholinic vesicles. Biochemistry 13; 3315—30. Sirontak, F.M., Moccio, D.M., Kelleher, L.E. and Goutas, L.J. (1981) Relative frequency and kinetic properties of transport-defective phenotypes among methotrexate-resistant L1210 clonal cell lines derived in vitro. Cancer Res. 41; 4447—52. Skogen-Hagenson, M.J., Salzman, G.C., Mullaney, P.F. and Brockman, W.H. (1977) A high efficiency flow cytometer. /. Histochem. Cytochem. 25; 784-9. Skovsgaard, T. (1978a) Carrier-mediated transport of daunorubicin, adriamycin and rubidazone in Ehrlich ascites tumor cells. Biochem. Pharmacol. 27; 1221-7. Skovsgaard, T. (197Sb) Mechanism of resistance to daunorubicin in Ehrlich ascites tumor cells. Cancer Res. 38; 1785-91. Skovsgaard, T. (1978c) Mechanism of cross-resistance between vincristine and daunorubicin in Ehrlich ascited tumor cells. Cancer Res. 38; 4722-7.

420

REFERENCES

Skovsgaard, T. and Nissen, N. I. (1982) Membrane transport of anthracyclines. Pharmacol. Ther. 18; 293-311. Slamon, D.J. (1987) Proto-oncogenes and human cancer. N. Engl. ]. Med. 317; 955-7. Slamon, D. J., Godolphin, W., Jones, L. A., Holt, J. A., Wong, S. G., Keith, D. E., Levin, W. J., Stuart, S.G., Udove, J., Ullrich, A. and Press, M.F. (1989) Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 244; 707-12. Slovak, M.L., Hoeltge, G.A., Dalton, W.S. and Trent, J.M. (1988) Pharmacological and biological evidence for differing mechanisms of doxorubicin resistance in two human tumour cell lines. Cancer Res. 48; 2793-7. Smith, P. J., Debenham, P. G. and Watson, J. V. (1989) A role of DNA topoisomerases in the active dissociation of DNA minor groove-ligand complexes. Mutation Res. 217; 163-72. Smith, P. J., Nakeff, A. and Watson, J. V. (1985) Flow-cytometric detection of changes in the fluorescence emission spectrum of a vital DNA-specific dye in human tumour cells. Exptl. Cell Res. 159; 37-46. Sonneveld, P. and van den Engh, C. J. (1981) Differences in uptake of adriamycin and daunomycin by normal BM and acute leukaemia cells determined by flow cytometry. Leukaemia Res. 5; 251-7. Southern, E. M. (1975) Detection of specific sequences among DNA fragments by gel electrophoresis. ]. Mol Biol. 98; 503-17. Spielman, L. and Goren, S. L (1968) Improved resolution in Coulter counting by hydrodynamic focussing. /. Colloid Interface Sci. 26; 175-S2. Steel, G. G. (1968) Cell loss from experimental tumours. Cell and Tissue Kinet. 1; 193-207. Steen, H. B. and Lindmo, T. (1979) Flow cytometry: a high resolution instrument for everyone. Science 204; 403—4. Stehlin, D., Varmus, H.E., Bishop, J.M. and Vogt, P.K. (1976) DNA related to transforming genes of avian sarcoma virus is present in normal avian DNA. Nature 260; 170-3. Steinkamp, J. A. (1983) A differential amplifier circuit for reducing noise in axial light loss measurements. Cytometry 4; 83—7. Steinkamp, J. A., Hansen, K. M. and Crissman, H. A. (1976) Flow microfluorometric and light scatter measurements of nuclear and cytoplasmic size in mammalian cells. /. Histochem. Cytochem. 24; 292-7. Steinkamp, J.A., Stewart, C. C. and Crissman, H.A. (1982) Three-color fluorescence measurements on single cells excited at three laser wavelengths. Cytometry 2; 226-31. Stewart, G., Harris, P., Gait, J. and Ferguson-Smith, M. (1985) Cloned DNA probes regionally mapped to human chromosome 21 and their use in determining the origin of nondisjunction. Nucleic Acids Res. 13; 4125-32. Stewart, J., Evan, G.I., Watson, J.V. and Sikora, K.E. (1986) Detection of the c-myc oncogene product in colonic polyps and carcinomas. Brit. ]. Cancer 53; 1—6. Stewart, T. A., Bellve, A. R. and Leder, P. (1984) Transcription and promoter usage of the c-myc gene in normal somatic and spermatogenic cells. Science 226; 707-10. Stohr, M., Eipel, H., Goerttler, K. and Vogt-Schaden, M. (1977) Extended application of flow microfluorometry by means of dual excitation. Histochemistry 51; 305—13. Stohr, M., Hutter, K. J., Frank, M. and Goerttler, K. (1982) A reliable preparation of monodispersed chromosome suspensions for flow cytometry. Histochemistry 74; 57-61. Stovel, R. T. (1977) The influence of particles on jet breakoff. /. Histochem. Cytochem. 25; 813-20.

REFERENCES

421

Strugger, S. (1948) Fluorescence microscope in examination of bacteria in soil. Canad. ]. Res. C26; 188-93. Stryer, L. (1978) Fluorescence energy transfer as a spectroscopic ruler. Ann. Rev. Biochem. 4t7; 819-46. Stuart-Harris, R., Hedley, D. W., Taylor, I. W., Levene, A. L. and Smith, I. E. (1985) Tumour ploidy and survival in patients receiving endocrine therapy for advanced breast cancer. Brit. ]. Cancer 51, 573-6. Studzinski, G. P., Brelvi, Z. S., Feldman, S. C. and Watt, R. A. (1986) Participation of c-myc in DNA synthesis of human cells. Science 234; 467-70. Sturzbecher, H. W., Addison, C. and Jenkins, J. R. (1988) Characterization of mutant p53hsp72/73 protein-protein complexes by transient expression in monkey COS cells. Molec. Cell Biol 8; 3740-7. Sunderesan, V., Forgacs, I. C, Wight, D. G. D., Wilson, B., Evan, G. I. and Watson, J. V. (1987) Abnormal distribution of the c-myc oncogene product in familial polyposis coli. /. Clin. Path. 40; 1274-81. Supino, R., Mariani, M , Capranico, G., Colombo, A. and Parmiani, G. (1988) Doxorubicin cellular pharmacokinetics and DNA breakage in a multi-drug resistant B16 melanoma cell line. Brit. J. Cancer 57; 142-6. Supino, R., Prosperi, E., Formelli, F., Mariani, M. and Parmiani, G. (1986) Characterization of a doxorubicin-resistant murine melanoma line: studies on cross-resistance and its circumvention. Brit. J. Cancer 54; 33-42. Swartzendruber, D. E. (1977) Microfluorometric analysis of cellular DNA following incorporation of bromodeoxyuridine. /. Cell Physiol. 90; 445-54. Sweet, R. G. (1965) High frequency recording with electrostatically deflected ink jets. Rev. Sci. Instrum. 36; 131-6. Szoka, F. and Papahadjopoulos, D. (1978) Procedure for the preparation of liposomes with large internal aqueous space and high capture by reverse-phase evaporation. Proc. Natnl. Acad. Sci. USA 75; 4194-8. Taft, E. B. (1951) The specificity of the methyl green-pyronin stain for nucleic acide. Exptl. Cell Res. 2; 312-26. Takahashi, M. (1966) Theoretical basis for cell cycle analysis, I. Labelled mitosis wave method. /. Theoret. Biol. 13; 202-11. Takahashi, M. (1968) Theoretical basis.for cell cycle analysis. II. Further studies on labelled mitosis method. /. Theoret. Biol 18; 195-209. Takahashi, M., Hogg, J. and Mendelsohn, M. L. (1971) The automated analysis of FLM curves. Cell and Tissue Kinet. 4; 505-18. Tanke, H. J., Niewenhuis, I. A. B., Koper, G. J. M., Slats, J. C. M. and Ploem, J. S. (1981) Flow cytometry of human reticulocytes based on RNA fluorescence. Cytometry 1; 313-20. Tate, E., Wilder, M. E., Scott-Cram, L. and Wharton, W. (1983) A method for staining 3T3 cell nuclei with propidium iodide in hypotonic solution. Cytometry 4; 211—15. Taylor, I.W. (1980) A rapid single step staining technique for DNA analysis by flow microfluorimetry. /. Histochem. Cytochem. 28; 1021-4. Taylor, I.W. and Milthorpe, BK. (1980) An evaluation of DNA fluorochromes, staining techniques and analysis by flow cytometry. /. Histochem. Cytochem. 28; 1224-32. Temin, H. and Mizutani, S. (1970) RNA-dependent DNA polymerase in virons of Rous sarcoma virus. Nature 226; 1211-13. Tengerdy, R. P. (1965) Properties of fluorescein-labeled human gamma globulin used in quantitative immunofluorescence tests. Analyt. Biochem. 11; 272-80.

422

REFERENCES

Teresima, T. and Tolmach, L. J. (1963) Growth and nucleic acid synthesis in synchronously dividing populations of HeLa cells. Exptl. Cell Res. 30; 344-62. Terstappen, L. W. M. M., de Grooth, B. G., Nolten, G. M. J., ten Napel, C. H. K, van Berkel, W. and Greve, J. (1986) Physical discrimination between human T-lymphocyte subpopulations by means of light scattering, revealing two populations of T8-positive cells. Cytometry 7; 178-83. Terstappen, L.W. M. M. and Loken, M. R. (1988) Five-dimensional flow cytometry as a new approach for blood and bone marrow differentials. Cytometry 9; 548-56. Thomas, J.A., Buschbaum, R.N., Zimmick, A. and Racker, E. (1979) Intracellular pH measurements in Ehrlich ascites tumour cells utilizing spectroscopic probes generated in situ. Biochemisrry 18, 2210-18. Thomas, P.S. (1980) Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. Proc. Natnl Acad. Sci. USA 77; 5201-5. Thomlinson, R. H. and Gray, L. H. (1955) The histological structure of some human lung cancers and the possible implications for radiotherapy. Brit. ]. Cancer 9; 539—49. Thompson, J. M., Gralow, J. R., Levy, R. and Miller, R. A. (1985) The optimal application of forward and ninety-degree light scatter in flow cytometry for the gating of mononuclear cells. Cytometry 6; 401—6. Thompson, S. C., Bowen, K. M. and Burton, R. C. (1986) Sequential monitoring of peripheral blood lymphocyte subsets in rats. Cytometry 7; 184—93. Thor, A., Horan Hand, P., Wunderlich, D., Caruso, A., Muraro, R. and Schlom, J. (1984) Monoclonal antibodies define differential ras gene expression in malignant and benign colonic disease. Nature 311; 562-4. Thornthwaite, J. T., Sugarbaker, E. V. and Temple, W.J. (1980) Preparation of tissues for DNA flow cytometric analysis. Cytometry 1; 229-37. Titus, J.A., Sharrow, S.O., Connoly, J.M. and Segal, D.M. (1981) Fc (IgG) receptor distributions in homogeneous and heterogeneous cell populations by flow microfluorimetry. Proc. Natnl. Acad. Sci. USA 78; 519-23. Tobey, R. A. and Crissman, H. A. (1975) Unique techniques for cell cycle analysis utilizing mithramycin and flow microfluorometry. Exptl. Cell Res. 93; 235—9. Tomita, T., Yasue, M., Englehard, H. M., McLone, D. G., Gonzalez-Crussi, F. and Bauer, K. D. (1988) Flow cytometric DNA analysis of medulloblastomas. Prognostic implications of aneuploidy. Cancer 61; 744—9. Traganos, F., Darzynkiewicz, Z., Sharpless, T. and Melamed, M. R. (1977) Simultaneous staining of ribonucleic and deoxyribonucleic acid in unfixed cells using acridine orange in a flow cytofluorimetric system. /. Histochem. Cytochem. 25; 46—56. Traganos, F., Darzynkiewicz, Z., Sharpless, T. and Melamed, M. R. (1979) Erythroid differentiation of Friend leukaemia cells as studied by acridine orange staining and flow cytometry. /. Histochem. Cytochem. 27; 382—9. Trask, B., van den Engh, Gray, J. W., Vahderlaan, M. and Turner, B. (1984) Immunofluorescence detection of histone 2B on metaphase chromosomes using flow cytometry. Chromosoma 90; 295—302. Trask, B., van den Engh, G., Landegent, J., Jansen in de Wai, N. and van der Ploeg, M. (1985) Detection of DNA sequences in nuclei in suspension by in situ hybridization and dual beam flow cytometry. Science 230; 1401-3. Trask, B. van den Engh, G., Pinkel, D., Mulliken, J., van Dekken, H. and Gray, J. W. (1988) Fluorescence in situ hybridization to interphase cell nuclei in suspension allows flow

REFERENCES

423

cytometric analysis of chromosome content and microscopic analysis of nuclear organization. Human Genet. 78; 251—9. Trebeck, C. Ev Friedlander, M. L, Russell, P. and Baird, P. J. (1987) Brenner tumours of the ovary: A study of the histology, immunohistochemistry, and cellular DNA content in benign, borderline and malignant ovarian tumours. Pathology 19; 241-6. Tribukait, B. (1984) Clinical DNA flow cytometry. Med. Oncol. Tumour Pharmacother. 1; 211-18. Truneh, A. and Machy, P. (1987) Detection of very low receptor numbers on cells by flow cytometry using a sensitive staining method. Cytometry 8; 562-7. Truneh, A., Machy, P. and Horan, P. K. (1987) Antibody-bearing liposomes as multicolor immunofluorescence markers for flow cytometry and imaging. /. Immunol Meth. 100; 59-71. Tsien, R. Y. (1980) New calcium indicators and buffers with high selectivity against magnesium and protons: Design, synthesis and properties of prototype structures. Biochemistry 19; 2396-404. Tsien, R.Y., Pozzan, T. and Rink, T.J. (1984) T-cell mitogens cause early changes in cytoplasmic free Ca 2 + and membrane potential in lymphocytes. Nature 295; 6S-70. Tsuro, T., Iida, H., Tsuragoshi, S. and Sakurai, Y. (1982) Increased accumulation of vincristine and adriamycin in drug-resistant P388 tumour cells following incubation with calcium antagonists and calmodulin inhibitors. Cancer Res. 42; 4730-3. Twentyman, P. R., Fox, N. E. and Bleehen, N. M. (1986a) Drug resistance in human lung cancer cell lines: cross-resistance studies and effects of the calcium transport blocker, verapamil. Int. ]. Rad. Oncol Biol Phys. 12; 1355-8. Twentyman, P.R., Fox, N.E., Wright, K.A. and Bleehen, N.M. (1986b) Derivation and preliminary characterization of adriamycin resistant lines of human lung cells. Brit. ]. Cancer 53; 529-37. Tytor, M., Franzen, G. and Olofsson, J. (1987) DNA patterns in oral cavity carcinomas in relation to clinical stage and histological grading. Pathol Res. Pract. 182; 202-6. Valet, G., Raffael, A., Moroden, L, Wursch, E. and Ruhenstroth-Bauer, G. (1981) Fast intracellular pH determination in single cells by flow cytometry. Naturweiss 68; 265—6. van Bockstaele, D. R., Berneman, Z.N. and Peetermans, M. E. (1986) Flow cytometric analysis of hairy cell leukaemia using right-angle light scatter. Cytometry 7; 217—20. van den Brenck, H. A. S., Madigan, J. P. and Kerr, R. C. (1968) Experience in Melbourne with the use of hyperbaric oxygen combined with megavoltage radiation in 614 cases of advanced malignant disease. In Frontiers of Radiation and Oncology, pp. 162—74. Kager, Basel and New York, van den Engh, G.J., Trask, B.J., Cram, S. and Bartholdi, M. (1984) Preparation of chromosome suspensions for flow cytometry. Cytometry 5; 108—17. van den Engh, G. J., Trask, B. J., Gray, J. W., Langlois, R. G. and Yu, L.-C. (1985) Preparation and bivariate analysis of suspensions of human chromosomes. Cytometry 6; 92—100. Van den Ingh, H. F., Griffioen, G. and Cornelisse, C. J. (1985) Flow cytometric detection of aneuploidy in colorectal carcinomas. Cancer Res. 45; 3392-7. Vanderlaan, M., Cutter, C and Dolbeare, F. (1979) Flow microfluorimetric identification of liver cells with elevated y-glutamyl transpeptidase activity after carcinogen exposure. /. Histochem. Cytochem. 27; 114-19. van der Linden, J. C, Linderman, J., Baak, J. P. A., Meijer, C. J. L. M. and Herman, C. J. (1989) The multivariate prognostic index and nuclear DNA content are independent prognostic factors in primary breast cancer patients. Cytometry 10; 56-61.

424

REFERENCES

van de Sande, J. H., Lin, C. C. and Jorgenson, K. F. (1977) Reverse banding on chromosomes produced by a guanosine-cytosine specific DNA binding antibiotic: olivomycin. Science 195; 400-2. van Dilla, M A., Deaven, L. L, Albright, K. L, Allen, N. A., Aubuchon, M. R., Bartholdi, M. F., Brown, N. C, Campbell, E. W., Carrano, A. V., Clark, L. M , Cram, L. S., Crawford, B. D., Fuscoe, J. G, Gray, J. W., Hildebrand, G E., Jackson, P. J., Jett, J. H., Longmire, J. L., Lozes, C. R., Luedemann, M. L, Martin, J. C, McNinch, J. S., Meinche, L. J., Mendelsohn, M. L, Meyne, J., Mayzis, R. K., Munk, A. C, Perlman, J., Peters, D. C, Silva, A. J. and Trask, B. J. (1986) Construction and availability of chromosome specific DNA libraries. Biotechnology 4; 537-52. van Dilla, M. A., Fulwyler, M. J. and Boone, I. U. (1967) Volume distribution and separation of normal human leukocytes. Proc. Soc. Exp. Biol. Med. 125; 367—70. van Dilla, M. A., Trujillo, T. T., Mullaney, P. F. and Coulter, J. R. (1969) Cell microfluorometry: a method for rapid fluorescence measurement. Science 163; 1213-14. Van Dyke, M. W. and Dervan, P. B. (1983a) Methidiumpropyl-EDTA. Fe(II) and DNase I footprinting report different small molecule binding site sizes on DNA. Nucleic Acid Res.

11; 5555-67. Van Dyke, M. W. and Dervan, P. B. (1983b) Chromomycin, mithramycin and olivomycin binding sites on heterogeneous deoxyribonucleic acid. Footprinting with (methidiumpropyl-EDTA)iron(II). Biochemistry 22; 2373-7. van Ewijk, W., van Soest, P. L, Verkerk, A. and Jonkind, J. F. (1984) Loss of antibody binding to prefixed cells: fixation parameters for immunocytochemistry. Histochem. ]. 16; 179-93. Van, D. C and Kettman, J. R. (1972) In vitro cooperation of cells of bone marrow and thymus origins in the generation of antibody-forming cells. /. Immunol 108; 73-80. Varmus, H. E. (1984) The molecular genetics of cellular oncogenes. Annu. Rev. Genet. 18; 553-612. Vindelov, L. L., Christensen, I. J., Jensen, G. and Nissen, N. I. (1982) Limits of detection of nuclear DNA abnormalities by flow cytometric DNA analysis. Results obtained by a set of methods for sample-storage, staining and internal standardization. Cytometry 3; 332-9. Vindelov, L. L., Christensen, I. J. and Nissen, N. I. (1982) Standardization of high-resolution flow cytometric DNA analysis by the simultaneous use of chicken and trout red blood cells as internal reference standards. Cytometry 3; 328-31. Virchow, R. (1847) Die pathologische Pigmente. Virchows Arch. Path. Anat. 1; 397-486. Visser, J.W. M., Haaijman, J. and Trask, B. (1978) In Immunofluorescence and Related Techniques (eds W. Knapp, K. Holubar and G. Wick) p. 147. Elsevir/North Holland Biomedical Press, Amsterdam. Visser, J.W.M., van den Engh, G.J. and van Bekkum, D.M. (1980) Light scattering properties of murine haemopoetic stem cells. Blood Cells 6; 391-407. Vistica, D. T., Toal, J. N. and Rabinowitz, M. (1976) Amino acid-conferred resistance to melphalan. 1) Structure-activity relationship in cultured L1210 cells. Cancer Treat. Rep.

60; 1363-7. Vistica, D. T., Toal, J. N. and Rabinowitz, M. (1978) Amino acid-conferred protection against melphalan: characterization of melphalan transport and correlation of uptake with cytotoxicity in cultured L1210 murine leukaemia cells. Biochem. Pharmacol. 27;

2865-70.

REFERENCES

425

Vogt, P.K., Bos, T.J. and Doolittle, R.F. (1987) Homology between the DNA-binding domain of the GCN4 regulatory protein of yeast and the carboxyl-terminal region of a protein coded for by the oncogene jun. Proc. Natnl Acad. Sci. USA 84; 3316—19. Volm, M , Mattern, J., Sonka, J., Vogt-Schaden, M. and Wayss, K. (1985) DNA distribution in non-small cell lung carcinomas and its relationship to clinical behaviour. Cytometry 6;

348-56. von Bertalanffy, L. and Bickis, I. (1956) Identification of cytoplasmic basophilia (ribonucleic acid) by fluorescence microscopy. /. Histochem. Cytochem. 4; 481-93. Wakelman, M.J.O., Davies, S. A., Houslay, M.D., McKay, L, Marshall, C.J. and Hall, A. (1986) Normal p 2 1 c ~ r a s couples bombesin and other growth factors to inositole phosphate production. Nature 323; 173-6. Waldeyer, K. (1863) Untersuchungen uber den Usprung and den Verlauf des Axencylinders bei Wirbellosen und Wirbelthieren sowie uber dessen Endverhalten in der quergestreiften Muskelfaser. Henle Pfeifer. Z. rat. Med. (3 Reihe) 20, 193-256. Walker, K. Z., Hayden, G. E., Goodnow, C. C, Adams, E., Bastin, A. and Raison, R. L. (1985) Co-expression of an epitope on human free K-light chains and on a cytoplasmic component of activated T-cells. /. Immunol. 134; 1059—64. Walloch, J., Solans, C. and Herman, C. J. (1987) Comparison of DNA content and histologic grade of colon adenomas. Cytometry Suppl. 1; l l a . Walport, M. J., Ross, G. D., Mackworth-Young, C, Watson, J. V., Hogg, N. and Lachmann, P. J. (1985) Family studies of erythrocyte complement receptor type 1 levels: reduced levels in patients with SLE are acquired, not inherited. Clin. Exptl. Immunol. 59; 547—54. Wang, D., Li, S. Zeng, C, Cheng, R. and Xue, S. (1988) Microtubule and microfilament distribution and tubulin content in the cell cycle of Indian Muntjac cells. Cytometry 9;

368-73. Wang, Z. and Rozengurt, E. (1983) Interplay of cyclic AMP and microtubules in modulating the initiation of DNA synthesis in 3T3 cells. /. Cell Biol. 96; 1743-50. Ward, D. C, Reich, E. and Goldberg, I. H. (1965) Base specificity of the interaction of polynucleotides with antibiotic drugs. Science 149; 1259—63. Waring, M. J. (1965) Complex formation between ethidium bromide and nucleic acids. /. Mol Biol. 13; 269-82. Waring, M. J. (1970) Variation of the supercoils in closed circular DNA by binding of antibiotics and drugs: Evidence for. molecular models involving intercalation. /. Mol Biol. 54; 247-79. Warr, J. R., Anderson, M. and Fergusson, J. (1988) Properties of verapamil-hypersensitive multidrug-resistant Chinese hamster ovary cells. Cancer Res. 48; 4477-83. Waterfield, M. D., Scrace, G. T., Whittle, N., Stroobant, P., Johnson, A., Wasteson, A., Westermark, B., Huang, J. and Deuel, T. F. (1983) Platelet derived growth factor is structurally related to the putative transforming protein p28SIS of simian sarcoma virus. Nature 304; 35-9. Waterman, D. (1986) A Guide to Expert Systems. Addison-Wesley, Reading, MA. Watson, J. D. and Crick, F. H. C. (1953) Molecular structure of nucleic acids. A structure for deoxyribonucleic acid. Nature 171; 737-8. Watson, J.V. (1977a) The application of age distribution theory in the analysis of cytofluorimetric DNA histogram data. Cell and Tissue Kinet. 10; 157-69. Watson, J.V. (1977b) Fluorescence calibration in flow cytofluorimetry. Brit. ]. Cancer 36; 396.

426

REFERENCES

Watson, J. V. (1980«) Enzyme kinetic studies in cell populations using fluorogenic substrates and flow cytometric techniques. Cytometry 1; 143-51. Watson, J. V. (1980b) Identification of different cell types and their separation using flow cytometric techniques. In Cancer, Assessment and Monitoring (eds T. Symington, A. E. Williams and J. G. McVie) pp. 77-95. Churchill Livingstone, Edinburgh, London, New York. Watson, J. V. (1981) Dual laser beam focussing for flow cytometry through a single crossed cylindrical lens pair. Cytometry 2; 14-19. Watson, J. V. (1984) Flow cytometric methods for studying enzyme activity in populations of individual cells. In Antitumour Drug Resistance (eds B. W. Fox and M. Fox) chapter 8, pp. 187—203. Springer-Verlag, Berlin, Heidelberg, New York, Tokyo. Watson, J. V. (1985) A method for improving light collection by 600% from square cross section flow cytometry chambers. Brit. ]. Cancer 51; 433—5. Watson, J. V. (1987) Time, a quality control parameter in flow cytometry. Cytometry 8;

646-9. Watson, J. V. (1989) A flow cytometry chamber with An light collection suitable for epifluorescence microscopes. Cytometry 10; 681—8. Watson, J. V. and Chambers, S.H. (1978a) Nucleic acid profile of the EMT6 cell cycle in vitro. Cell and Tissue Kinet. 11; 415-22. Watson, J.V. and Chambers, S.H. (1978b) Inferences on RNA synthesis rates during mitosis from flow cytofluorimetric RNA histograms. /. Histochem. Cytochem. 26; 691—5. Watson, J. V. and Chambers, S. H. (1980) Physico-chemical properties of subpopulations in mouse bone marrow. Brit. ]. Cancer 42; 184a. Watson, J. V., Chambers, S. H. and Smith, P. J. (1987) A pragmatic approach to the analysis of DNA histograms with a definable G l peak. Cytometry 8; 1-8. Watson, J. V., Chambers, S. H., Workman, P. and Horsnell, T. S. (1977) A flow cytometric method for measuring enzyme reaction kinetics in populations of single cells. Febs Lett. 81; 179-92. Watson, J. V., Cox, H., Hellon, C, Workman, P. and Dive, C. (1988a) Time-resolved flow cytometric measurement of dynamic events in the interval 1 to 20 seconds in whole cells. Cytometry Suppl. 2; 93a. Watson, J. V., Cox, H., Hellon, C, Workman, P. and Dive, C. (1991a) Stop-flow cytometry in flow: Measurement of 'fast' dynamic events in the interval 1 to 20 seconds in whole cells. (In submission.) Watson, J.V., Curling, O.M., Munn, C.F. and Hudson, C.N. (1987a) Oncogene expression in ovarian cancer: A pilot study of c-myc oncoprotein in serous papillary ovarian cancer. Gynaecol. Oncology 28; 137—50. Watson, J. V., Horsnell, T. S. and Smith, P. J. (1988b) Data compression: 8-dimensional flow cytometric data processing with 28K addressable computer memory. /. Immunol. Meth. 113; 205-14. Watson, J.V., Morgan, S.A. and Dive, C D . (1989a) Drug resistance analysis in populations of individual cells using flow cytometry. In Ovarian Cancer; Biological and Therapeutic Challenges (eds F. Sharp, W. P. Mason and R. E. Leake) chapter 14. Chapman and Hall Medical, London. Watson, J.V., Morgan, S.A. and Smith, P.J. (1989b) Analysis of the Hoechst-33342 fluorescence emission spectrum in individual cells with reference to multi-drug resistance. Protides Biol. Fluids 36; 359-72.

REFERENCES

427

Watson, J. V., Munn, C. F., Cox, H. and Robinson, D. (1987b) a simultaneous assay for c-myc and c-fos oncoproteins plus DNA in human archival cancer biopsies. Cytometry Suppl. 1; 2a. Watson, J. V., Nakeff, A., Chambers, S.H. and Smith, P.J. (1985a) Fluorescence emission spectrum analysis of chicken thymocytes using Hoechst-33342. Cytometry 6; 310—15. Watson, J. V., Sikora, E. K. and Evan, G. I. (1985 b) A simultaneous flow cytometric assay for c-myc oncoprotein and cellular DNA in nuclei from paraffin embedded material.

/. Immunol Meth. 83; 179-92. Watson, J.V., Stewart, J., Cox, H., Sikora, K.E. and Evan, G.I. (1987c) Flow cytometric quantitation of the c-myc oncoprotein in archival neoplastic biopsies of the colon. Mol Cell Probes 1; 151-7. Watson, J. V., Stewart, J., Cox, H., Sikora, K. E. and Evan, G. I. (1991b) c-myc oncoprotein levels are raised in morphologically normal colonic mucosa derived from malignant compared with non-malignant specimens. (In submission.) Watson, J. V., Stewart, J., Evan, G., Ritson, A. and Sikora, K. (1986) The clinical significance of flow cytometric c-myc oncoprotein quantitation in testicular cancer. Brit. ]. Cancer 53; 331-7. Watson, J. V. and Taylor, I. W. (1977) Cell cycle analysis in vitro using flow cytofluorimetry after synchronization. Brit. J. Cancer 36; 281-7. Watson, J.V. and Walport, M.J. (1985) How does flow cytometry interpret Gaussian distributed information. ]. Immunol Meth. 77; 321-30. Watson, J. V. and Walport, M. J. (1986) Molecular calibration in flow cytomety with subattogram detection limit. J. Immunol Meth. 93; 171-5. Watson, J.V., Workman, P. and Chambers, S.H. (1978) Differences in esterase activity between exponential and plateau growth phases of EMT6/M/CC cells monitored by flow cytofluorimetry. Brit. ]. Cancer 37; 397-402. Watson, J. V., Workman, P. and Chambers, S. H. (1979) An assay for plasma membrane phosphatase activity in populations of individual cells. Biochem. Pharmacol 28; 821-7. Watt, S.M., Gilmore, D.J., Clark, M.R., Davis, J.M., Swirsky, D.M. and Waldmann, H. (1984) Haemopoetic progenitor cell heterogeneity revealed by a single monoclonal antibody, YW 13.1.1. Mol. Biol. Med. 2; 351-68. Waymouth, C. (1974) To disaggregate or not to disaggregate. Injury and disaggregation, transient or permanent? In Vitro 19; 97-103. Weinberger, C, Thompson, C. G, Ong, E. S., Lebo, R., Gruol, D. J. and Evans, R. M. (1986) The c-erb-A gene encodes a thyroid hormone receptor. Nature 324; 641-6. Weisblum, B. and Hanessler, E. (1974) Fluorimetric properties of the bisbenzimidazole derivative Hoechst 331518, a fluorescent probe specific for AT concentration in chromosomal DNA. Chromosoma 46; 255—60. Weiss, M.J. and Chen, L.B. (1984) Rhodamine 123: A lipophilic mitochondrial-specific vital dye. Kodak lab. Chem. Bull. 55; 1-4. Welleweerd, J., Wilder, M.E., Carpenter, S.G. and Raju, M.R. (1984) Flow cytometric determination of radiation induced chromosome damage and its correlation with cell survival. Radiat. Res. 99; 44-51. Weltzien, H. U. (1979) Haemolytic and membrane perturbing properties of lysophosphatidylcholine. Biochim. Biophys. Acta 559; 259-87. Wheeless, L.L., Berkan, T.K., Patten, S.F., Reeder, R.E., Robinson, R.D., Eldidi, M.M., Hulbert, W. G and Frank. I. N. (1986a) Multidimensional slit-scan detection of bladder cancer preliminary clinical results. Cytometry 7; 212—16.

428

REFERENCES

Wheeless, L L, Hardy, J.A. and Balasubramanian, N. (1975) Slit-scan flow system for automated cytopathology. Ada Cytol. (Baltimore) 19; 45-52. Wheeless, L. L. and Patten, S. F. (1973a) Slit-scan cytofluorometry. Ada Cytol (Baltimore) 17; 333-9. Wheeless, L. L. and Patten, S. F. (1973fc) Slit-scan cytofluorometry: basis for automated cytopathology pre-screening. Ada Cytol (Baltimore) 17; 391-4. Wheeless, L. L, Patten, S. F., Berkan, T. K., Brookes, C. L, Gorman, K. M , Lesh, S. R., Lopez, P. A. and Wood, J. C. (1984) Multidimensional slit-scan prescreening system: preliminary results of a single blind clinical study. Cytometry 5; 1-8. Wheeless, L. L, Robinson, R. D., Cox, C, Berkan, T. K. and Reeder, J. E. (19861?) A statistical analysis of prescreening alarms in a population of normal and abnormal gynaecologic specimens. Cytometry 7; 205—11. White, R. A. and Meistrich, M. L. (1986) A comment on: A method to measure the duration of DNA synthesis and the potential doubling time from a single sample. Cytometry 7; 486-90. Wilson, G. D., McNally, N. J., Dunphy, E., Karcher, H. and Pfragner, R. (1985) The labelling index of human and mouse tumours assessed by bromodeoxyuridine staining in vitro and in vivo staining and flow cytometry. Cytometry 6; 641—7. Wing, M. G., Montgomery, A. M. P., Songsivilai, S. and Watson, J. V. (1990) An improved method for the detection of cell surface antigens in samples of low viability using flow cytometry. /. Immunol Meth. 126; 21-7. Wolf, D. and Rotter, V. (1985) Major deletions in the gene encoding the p53 tumor antigen cause lack of p53 expression in HL-60 cells. Proc. Natnl Acad. Sci. USA 82; 790-4. Wolley, R.C., Schreiber, K., Koss, L.G., Karas, M. and Sherman, A. (1982) DNA distribution in human colonic carcinomas and its relationship to clinical behaviour. /. Natnl Cancer Inst. 69; 15-22. Workman, P., Cox, H. and Watson, J. V. (1988) Glutathione metabolism in human lymphocytes by flow cytometry. Cytometry Suppl. 2; 47a. Workman, P. and Watson, J. V. (1987) Flow cytoenzymology of glutathione metabolism. Cytometry Suppl. 1; 48a. Wray, W. (1973) Isolation of metaphase chromosomes, mitotic apparatus and nuclei. In Methods in Cell Biology IV (ed. D. Prescot) pp. 283-306. Academic Press, New York. Wray, W. and Stubblefield, E. (1970) A new method for rapid isolation of chromosomes, mitotic apparatus, or nuclei from mammalian fibroblasts at near neutral pH. Exp. Cell Res.

59; 469-78. Wray, W. and Wray, V. P. (1979) Isolated metaphase chromosomes II: proteins of Chinese hamster chromosomes. /. Histochem. Cytochem. 27; 454-7. Wray, W. and Wray, V. P. (1980) Proteins from metaphase chromosomes treated with fluorochromes. Cytometry 1; 18-20. Wright, J. E., Rosowsky, A., Boeheim, K., Cucchi, C. A. and Frei, E. (1987) Flow cytometric studies of methotrexate resistance in human squamous carcinoma cell cultures. Biochem. Pharmacol 36; 1561-4. Yablonka-Reuvein, Z. (1988) Discrimination of myogenic and nonmyogenic cells from embryonic skeletal muscle by 90° light scattering. Cytometry 9; 121-5. Youdim, M. B. E. and Woods, H. F. (1975) The influence of tissue environment on the rates of metabolic processes and the properties of enzymes. Biochem. Pharmacol 24; 317-23. Young, B., Ferguson-Smith, M., Sillar, R. and Boyd, E. (1981) High-resolution analysis of

REFERENCES

429

human peripheral lymphocyte chromosomes by flow cytometry. Proc. Natnl. Acad. Sci. USA 78; 7727-31. Young, G.A.R., Hedley, D.W., Rugg, C.A. and Hand, H.J. (1987) The prognostic significance of proliferative activity in poor histology non-Hodgkin's lymphoma: A flow cytometry study using archival material. Eur. ]. Cancer Clin. Oncol. 23; 1497-1504. Young, T. (1802) On the theory of light and colours. Phil. Trans. R. Soc. Lond. 92; 12-48. Young, T. (1804) Experiments and calculations relative to optical physics. Phil Trans. R. Soc. London. 9 4 , 1-16.

Yu, L.-C, Aten, J., Gray, J.W. and Carrano, A.V. (1981) Human chromosomes isolated from short-term lymphocyte culture for flow cytometry. Nature 293; 154-5. Yunis, J. J. and Ramsay, N. (1978) Retinoblastoma and sub-band deletion of chromosome 13. Am. J. Dis. Child 132; 161-3. Zeile, G. (1980) Intracytoplasmic immunofluorescence in multiple myeloma. Cytometry 1; 37-41. Zeitz, S. and Nicolini, C. (1978) Flow microfluorometry and cell kinetics: a review. In Biomathematics and Cell Kinetics (eds A.-J. Valleron and P. D.M. Macdonald) p. 357. Elsevier/North Holland, Amsterdam. Zimmer, C. and Wahnert, U. (1986) Nonintercalating DNA-binding ligands: specificity of the interaction and their use as tools in biophysical, biochemical and biological investigation of the genetic material. Prog. Biophys. Mol. Biol. 47; 31-112. Zimmerman, P.V., Hawson, G.A.T., Bint, M.H. and Parsons, P.G. (1987) Ploidy as a prognostic determinant in surgically treated lung cancer. Lancet ii; 533-9. Zucker, R. M , Elstein, K. H., Easterling, R. E. and Massaro, E.J. (1988) Flow cytometric discrimination of mitotic nuclei by right-angle light scatter. Cytometry 9; 226-31. Zucker-Franklin, D. (1974) The percentage of monocytes among 'mononuclear' cell fractions obtained from normal human blood. /. Immunol. 112; 234—40.

Index

aberrations, 20-1, 23-7, 29 astigmatic, 29 chromatic, 20-1, 25-7 spherical, 23-4 absorption, 34-6, 49, 51, 160-5, 168, 175, 185 filters, 34-6, 51, 160-5, 168, 175 fluorochromes to plastic tubes, 185 spectra, 49 accuracy, see precision acetoxy-methyl esters, 133, 146-7 acid denaturation of DNA, 135-6 acridine orange, 50, 185, 204-7, 243, 247-56, 283-5, 350 BrdU quenching, 250 cell cycle subsets, 250-6 metachromatic fluorescence, 50, 204-7 multi-parameter analysis with, 283—5 precipitation with, 206 screening with, 350 staining with, 50, 206, 247-51 structure of, 205 actinomycin-D, 251, 276, 370 ADC, see analogue-to-digital conversion adjunctive chemotherapy, 384 age distribution theory, 211—13 agglutinating fixatives, 122-3 air line filtration, 155 Airy disk, 33 alkaline phosphatase, 324 allophycocyanin, 50 amino-methyl coumarin acetic acid (AMCA), 50, 125, 128, 282-3 amplification, 68-70, 129-31, 154, 159-60, 169, 171-2, 241 electronic, 66-70, 154, 159-60, 171-2 differential, 68-70 linear, 66, 171-2 logarithmic, 66-7, 154, 159-60, 171-2 fluorochrome, 129-33, 169, 241 biotin: streptavidin, 130-1, 169, 241 liposomes, 130-1, 169 polyclonal antibodies, 129

analogue-to-digital conversion, 72—3, 76, 17 A, 176

offset, 72, 173, 176 resolution of, 72 analysis of data, see data analysis aneuploidy, see DNA, index angle of incidence, 18-19, 26-31 antibodies, 125-33, 237-8, 269-73, 282-3 bromodeoxyuridine, 23 7-8 combination staining, 127-9, 282-3 labels for, 125 monoclonal, 128 polyclonal, 129 staining with, 126-33, 269-73 structure of, 127-9 fragments, 128 heavy chains, 127 light chains, 127 recognition site, 127 variable regions, 127 synthetic peptide induced, 128, 132, 271, 282-3 anomalous diffraction, 62 antigens, 123, 131-2, 247, 268-87, 297-8, 358-6, 364 cell surface, 123, 285 cytoplasmic, 131-2, 283-7 nuclear antigens, 132, 247, 268-83, 297-8, 358-6, 364 histones, 247, 268-9 interchromatin granules, 269 Ki-67, 269 p53, 269, 358 p55c~fos, 269-70, 282, 283 p62c~myc, 269-82, 358-6 p62c~myc, staining, 270-1 problems, 132 staining and quantitation, 269—3 topoisomerases, 269, 274, 297-8 turnover measurements, 273, 364 AO, see acridine orange aperture, numerical, 40 applications in oncology, 345-84

432

INDEX

applications in oncology (cont.) adjunctive chemotherapy, 384 bladder cancer, 350, 355 breast cancer, 351-2, 384 Brenner tumour, 352 bromodeoxyuridine, 356, 383 Burkitt's lymphoma, 354 cancer diagnosis, 348—51 carbamoylation, 379 cell loss, 381 cell production rate, 381-2 cervical cancer, 362 chlorambucil, 369 chloroethylnitrosoureas, 369, 380 CIN, 362 clinical trials, 347 colonic adenomata, 353, 357, 362 colorectal cancer, 351, 353, 361 cyclosporin-A, 377 cytological prescreening, 349-51 acridine orange, 350 bladder, 350 gynaecological, 350—1 nuclear-to-cytoplasmic ratio, 350 slit-scanning, 350 cytoskeleton, 358 daunorubicin, 377 diagnosis, 348-51 leukocyte classification, 349 drug resistance, 340-2, 368-80, 384 active efflux, 370 adriamycin resistance, 370 cross resistance, 369 drug resistance pathways, 370 glutathione, 325-8, 367-8, 384 gp 170 , 370-1, 373, 375, 379 Hoechst 33342, 371, 373 MDR-1 gene, 369, 371 mechanisms, 368-9 multi-drug resistance, 369-78, 384 DNA index, 242-4, 351-6 efflux modifiers, 377 electron affinic sensitisers, 381 ENT tumours, 354 esterase inhibition, 378 fractionated radiotherapy, 380-1 future prospects in oncology, 384 glutathione, 325-8, 367-8, 384 hereditary tumours, 357 HER2/neu, 358-60, 366 homeostatic mechanisms, 382 hyperbaric oxygen, 381 hyperfractionation, 384 hypoxic cells, 381 leukaemia, 367 lung cancer, 354 malignant melanoma, 355, 357

medulloblastoma, 355 methotrexate, 384 monochlorobimane, 325, 367 multi-variate analysis, 355 neuroblastoma, 355 nitrogen mustard, 369 non-Hodgkin's lymphoma, 351, 354 oncogenes, 356-66 oophorectomy, 380 ovarian borderline malignancy, 352 ovarian cancer, 352, 363 phenylalanine mustard, 369 potential doubling time, 383 prognosis, 351-66 archival biopsies, 351 DNA index, 351-6 oncogenes, 356-66 p21c~ras, 358 p53, 358 p62c~myc, 358-66 p62c~myc turnover, 364 receptor status, 366 renal adenocarcinoma, 355 retinoblastoma suppressor gene, 358 retro viruses, 356 reverse transcriptase, 356 S-phase fraction, 352, 354-5 Testicular cancer, 360-1 therapy selection, 366-84 tumour growth rate, 356, 380—4 tumour suppressor genes, 357 verapamil, 372, 375, 377 archival biopsies, 120-2, 132, 351-6,

358-66 dewaxing 121 digestion, 121 nuclear extraction, 121—2 problems, 132 arc lamps, deuterium, mercury and xenon, 56 argon lasers, 57, 148 arrays, see data arrays astigmatism, 29 AT specific dyes, 50, 51-2, 133, 148-9, 185, 201-2, 256-66 AT:GC composition, 293 autofluorescence, 152, 158 back flushing, 185 background compensation, 100—1 background current, 155 background fluorescence, 151-2 background subtraction, 68-70, 92, 100-1 base-pair composition, 293 band-pass filters, 37, 51, 148, 161, 175, 282-3, 371 base-line DC current, 155

INDEX beam geometry, 58-60 beam splitter, 168 Bernoulli, 5-6 binding site modulation, 134—5, 243—4 biotin, 130-1, 169, 241 bridge, 130-1 streptavidin, 130-1, 169, 241 biotinolated antibodies, 130-1 biotinolated nucleotides, 241 binomial statistics, 14, 113-14 bisbenzimidazoles, 50, 52-1, 133, 148-9, 185, 201-3, 256-66, 292, 371 structure of, 203 bits, bytes and binary, 72-6 bit-mapping, 86-92 2-dimensional, 86 multi-dimensional, 88-92 bivariate (two-dimensional) data, 81, 195-8, 283-5 bladder cancer, 350, 355 bleach, 185 bleaching, 49, 53, 167, 173 blocking bar, 41 bone marrow analysis, 148-9 BrdU, BrdUrd see bromodeoxyuridine breast cancer, 351-2, 384 Brenner tumour, 352 brightness, 55 bromodeoxyuridine, 235-41, 250, 276-82, 293, 356, 383 acridine orange quenching by, 250 antibodies to, 237-8 growth kinetics with, 238-9, 279-82 incorporation, 238-41 quenching with, 235, 250 stoichiometry, 239 tumour growth rate, 239-41, 356, 380-4 boundary layer, 7-8 building vibration, 154 Burkitt's lymphoma, 354 calcium, 50, 133, 147, 336-9 measurement of, 337-8 probes for, 50, 147 calibration, 146-7, 158-9, 176, 244-6, 337-8 biological standards, 244-6 calcium 337-8 chicken, red cells, 244-6 coincidence in, 176 fluorescent microbeads, 159 labelled antibodies, 158-9 pH, 146-7 trout red cells, 244-6 cancer, see application in oncology cancer diagnosis, 348-51 capillary bore, 156

433

carbamoylation, 379 carboxyfluorescein, 131, 146 Casperson, 2 cell cycle, 207-8, 223-41, 250-6, 273-82, 356, 383 after Howard and Pelc, 208 kinetics, 223-41, 273-82 bromodeoxyuridine, 235-41, 250, 279-82, 356, 383 FPI analysis, 233-5 mitotic selection, 225-7 modelling population kinetics, 228-41 PLM curve, 207-8 stathmokinetic analysis, 224-5 modulation, 273-82 subsets, 250-6 cell loss, 381 cell production rate, 381-2 cell separation, 106-16 electrostatic sorting 107-16, see cell sorting iron particles, 106 magnetic beads, 106 sedimentation, 106 cell size, 64, 186-91 cell sorting, 106-16 droplet charging, 107-9 droplet deflection, 108-10 droplet generation, 107 efficiency, 115-16 electrostatic sorting, 107-16 high speed, 303 ink-jet writing, 106 one-droplet sort, 112 pre-sorting enrichment, 114 problems, 116, 134 purity, 110-14 statistics, 112-14 sorting yield, 110-14 three droplet sort, 112 times, 110 centromere, 297, 301 centromeric indices, 139, 301 cervical cancer, 362 chamber design, seeflowchambers chicken red cells, 244-6 chicken thymocytes, 259-62 chlorambucil, 369 chloroethylnitrosoureas, 332-4, 369, 380 chromatic aberration, 20-1, 25-7 chromatin, 268-9 charged couples devices, 66, 140 chromatic compensation focussing, 26-31 chromomycin A3, 50, 201-2, 292-3 structure of, 202 chromosome analysis, 115, 139, 283-5, 288-308

434

INDEX

chromosome analysis {cont.) applications, 303-8 diagnosis, 302-4 genomic libraries, 304-5 gene mapping, 305-8 radiation bio-dosimetry, 308 banding, 302 bivariate karyotype, 295-6 bromodeoxyuridine, 293 centromeric indices, 139, 301 chromosome-associated proteins, 297-8 dual-beam excitation, 283-5, 293 harvesting mitotic cells, 225-7, 288-9 high-speed sorting, 303 karyotype analysis, 299-300 one dimensional (monovariate), 299 two-dimensional (bivariate), 300 in situ hybridization, 299, 305 preparation, 290—1 hexylene glycol, 290 hypotonic PI detergent, 291 magnesium sulphate, 291 Ohnuki buffer, 291 polyamine, 290 slit-scanning, 301 staining, 292-3 base composition, 293 partial sequence specificity, 297 total DNA, 292 univariate karyotype, 115, 295-6 CIN, 362 clinical trials, 347 clumping, 136, 175 coaxial streaming, 6 coefficient of variation, 72, 143-4, 165, 170-1 coincidence, 13-17, 176-9 correction for, 177-9 probability of, 16 use of, 176 coincident focussing, 28-31 colcemid, 224, 283, 289 coordinate coding, 7 7-SO collagenase, 119 colonic adenomata, 353, 357, 362 colorectal cancer, 351, 353, 361 combination staining, 51-3, 128, 266, 282-3 AMCA/FITC/PI, 128, 282-3 antibodies, 282-3 Hoechst/FITC, 266 spectral considerations, 51-3 compensation, 26-31, 68-70, 99-101 chromatic, 26-31 data, 99-101 electronic, 68-70 computing, 72-105

conic section, 23 conjugate distances, 23—4 conjugate foci, 45—6 contamination, 154-5, 185 air lines, 155 fluorochromes in feed tubes, 185 sheath, 154-5 continuous interrupted sampling, 310 continuous time, 182-5, 311-12 contour maps, 81-2, 85, 89-90, 195-8, 284-5 core, 11-14 diameter, 11, 13-14 position, 12 stability, 12 coumarins, 50, 125, 128, 282-3, 330-2 counting within gates, 92-4, 277-81 Crosland-Taylor, 10, 107 crossed cylindrical lenses, 27-31, 153, 179, 186 cuvette flow chamber, 41, 42-3 light collection from, 42-3 cyanin dyes, 51, 146—7 cyclosporin-A, 377 cylindrical lenses, 25 cytological prescreening, 349-51 cytokeratins, 285-6 cytoplasmic antigens, 283-7 cytoplasmic enzymes, 310, 320-8 cytoskeleton, 285-7, 358 dansyl chloride, 50 DAPI, 51, 201-2, 292 dark current, 155 data analysis, 81-2, 85-105, 195-8, 213-23, 283-5 background compensation, 100—1 deconvolution of distributions, 96-8, 213-23 distribution assessment, 94-100 distribution shape analysis, 97-100 gating, 81-2, 85-94, 195-8, 283-5 Gaussian distribution, 97—100 'skewed-normal' distribution, 98, 100 data acquisition see data capture data arrays, 76 one-dimensional, 76 multi-dimensional, 76 two-dimensional, 76 data capture, 73, 178 buffering, 73 dedicated memory, 73 FIFO, 73, 178 list-mode, 73 data display, 80-90, 94, 195-8, 283-5 contour maps, 81-2, 85, 89-90, 195-8, 283-5

INDEX dot-plots, 81 hidden surface elimination, 83-4 multi-parameter, 85-6 one-dimensional, 81, 94 stereo-perspective graphics, 84 three-dimensional, 84 two-dimensional, 81, 195-8, 283-5 data processing, 76-92, 96-8, 102-5, 195-8, 283-5 arrays, 76 bit-mapping, 86-92 bit-shifting, 79, 102-5 coordinate coding, 77-9, 88 decoding coded data, 79 distribution deconvolution, 96-7 integer arithmetic, 75-6 multi-parameter, 76-80, 85-92, 195-8, 283-5 ranking coded data, 79 daunorubicin 377 denaturation, 122-3, 135-6, 249-50 DNA, 135-6, 249-50 acid, 135-6 heat, 249-50 proteins, 122-3 agglutinating fixatives, 122-3 cross-linking fixatives, 122—3 dedicated memory, 73 detection limit, 169-70 detectors see photodetectors de waxing, 121 dichroic mirrors, 37-9, 168 dichroic combinations, 38-9 serial, 38 serial/parallel, 39 dicyanodihydroxybenzene (DCDHB), 50, 146-7 dielectric interface, 18 differential amplifiers, 68-70 diffraction, 31-4, 62 diphenyl hexatriene, 125, 133 DIPI, 50, 201, 293 dirty optics, 156 disaggregation, 117-21 enzymatic, 118-22 cleavage sites, 118-20 collagenase, 119 elastase, 119 factors influencing, 118 glycosaminoglycans, 118 hyaluronidase, 119 lysosyme, 119 pepsin, 118 pronase, 120 protease, 120 trypsin, 118 mechanical, 117

435

chopping, 117 syringing, 117—18 wax embedded material, 120-1 disc of least confusion, 24 discrimination between populations, 170-2 distribution assessment, 94-100 mean, 94 median, 95 mode, 95 shape, 97-100 distribution deconvolution, 96—8 discrimination, 96-7, 144-5, 171-2 between distributions, 144-5 labelled and unlabelled, 96-8 rare events, 144—5 with log amplifiers, 171—2 double sheath, 12 double threshold, 177-9 DNA, 50-3, 68-9, 134-6, 148-9, 160, 201-65, 276-82, 293, 352-6, 371, 373, 375, 380-4 binding site modulation, 134—5, 243—4 bromodeoxyuridine, 235-41, 250, 276-82, 293, 356, 383 antibodies to, 237-8 growth kinetics with, 238-9, 279-82, 356, 380-4 incorporation, 238—41 stoichiometry, 239, 243 tumour growth rate, 239-41, 356, 380-4 denaturation, 135-7, 239 acid, 135-6 heat, 135, 249-50 chromosomes, see chromosome analysis dyes, see nucleic acid stains emission spectrum analysis, 148-9, 259-65, 371, 373, 375 histogram, 76, 159, 208-11 histogram analysis, 211-13 age distribution theory, 211-13 rectilinear integration, 213 multiple Gaussian, 214-15 polynomial, 215-16 single Gaussian, 216-23 TCW analysis, 223 index, 241-4, 351-6 stains, see nucleic acid stains standards, 244—6 stoichiometry, 243 drug resistance, 340-2, 368-80, 384 drug transport, 339-43 adriamycin, 340-1 CNUs, 332-3 methotrexate, 342—3 dual-beam focussing, 26—31 dyes, see fluorochromes

436

INDEX

dynamic cellular events, 309-44 alkaline phosphatase, 324 calcium, 336-9 continuous interrupted sampling, 310 continuous time, 311—12 cytoplasmic enzymes, 310, 320-8 discontinuous sequential sampling, 310 drug transport, 339-43 adriamycin, 340-1 CNUs, 332-3 methotrexate, 342-3 dual substrate analysis, 330—2 esterases, 310, 312-17, 320-4 enzyme kinetics, 317, 335 esterase inhibition, 322-33 fluorescence quantitation, 320-1 p-galatosidase, 328 P-glucuronidase, 322, 325 y-glutamyl transpeptidase, 324, 328 glutathione, 325-8f 367-8, 384 glutathione S-transferase, 326 incorporation of time, 310-17 inhibition kinetics, 332-3 leucine amino-peptidase, 322 light absorption quantitation, 319 membrane enzymes, 328-30 membrane potential, 336-7 Michaelis-Menten hyperbola, 322 mitochondrial function, 339 peroxidases, 320—1 progress curves, 323, 327, 329, 333 short time scale kinetics, 312-17, 333-5 substrates, 317-25, 328, 330-2 fluorescein diacetate, 180, 185, 322-4, 330-2 methylumbelliferyl acetate, 330-2 monochlorobimane, 50, 325, 367 3-o-methyl fluorescein phosphate, 328 verapamil, 341, 372, 375 dynamic range, 66-8, 72-3, 95, 154, 159-65 A-to-D converters, 72-3 log amplifiers, 66-8, 154, 159-60 neutral density filters, 160 variable gain, 160-5 dynode chain, 65

ADC offset, 73, 173, 176 resolution, 72 noise, 151—4 photodetectors, 65-6, 140 photomultipliers, 65 sequential illumination triggering, 71 signal processing, 66-70 solid-state devices, 66 triggering, 70-1, 177 voltage thresholds, 71, 177-9 electron affinic sensitizers, 381 electronic amplification, 66-70, 154, 159-60, 171-2 electrostatic sorting, see cell sorting emission spectrum, 49, 145—9, 259—65 emission spectrum analysis, 145-9, 259-65, 371, 373, 375 pH, 146-7 calcium, 147 DNA, 148-9, 259-65, 371, 373, 375 ENT tumours, 354 enzymatic disaggregation, 118—22 factors influencing, 118 enzyme cleavage sites, 118-20 enzyme progress curves, 323, 327, 329, 333 eosin, 1 esterases, 310, 312-17, 320-4 esterase inhibition, 322-33, 378 ethidium bromide, 50, 175, 203-4, 288, 290, 292 structure of, 204 Euler, 5-6 excitation, 54-60 conventional sources, 56 brightness, 56 deuterium arc, 56 focussing geometry, 54-5 mercury arc, 56 xenon arc, 56 size, 54 lasers, 57-60 beam geometry, 58—60 focussing, 22-31, 58-9 pulse shape, 60 excited state, 134-5

eddy currents, 7 efflux modifiers, 377 elastase, 119 electronics, 65-73, 77, 140, 151-4, 159-60, 171-2, 176-9 amplification, 66-70, 154, 159-60, 171-2 differential, 68-70 linear, 66, 171-3 logarithmic, 66-8, 154, 159-60 analogue-to-digital conversion, 72-3, 76, 17 A

filtration, 34-9, 51, 136, 148, 155, 160-5, 167-8, 175, 267-8, 282-3 optical, 34-9, 51, 148, 160-8, 282-3 absorption, 34-6, 160-5, 168, 175 band-pass, 37, 51, 148, 161, 175, 282-3, 371 dichroic mirrors, 37, 168, 281-3 dichroic combinations, 38—9, 282—3 for AMCA/FITC/PI, 282-3 for Hoechst/FITC, 267-8 long-pass, 34-6, 51, 168, 175

437

INDEX neutral density, 35, 160-5 short-pass, 36, 168 sample, 136 sheath, 155 air lines, 155 filter breakthrough, 51-4 fixation, 122-3 acetone, 122 ethanol, 122 formaldehyde (formalin), 122-3 gluteraldehyde, 122-3 methanol, 122 paraformaldehyde, 122-3 epitope modulation by, 123 flow chambers, 41-7, 107, 372 cuvette, 41 light collection efficiency of, 43, 372 diverging refraction from, 42 modified cuvette, 43-4 spherico-ellipsoidal, 45—7 stream-in-air, 41-3, 107 flow karyotyping, 295-6, 299-300 flow rates, 12-17 calculation of, 12-14 Poisson statistics, 15-17 fluid dynamics, 5—17 Bernoulli, 5 boundary layer, 7 coaxial streaming, 6-10 Crosland-Taylor, 10 Euler, 5-6 flow rates, 12-17 hydrodynamic focussing, 8-12 laminarflow,6—8 pressure profile, 6 Reynolds number, 6 turbulent flow, 7-8 fluorochrome amplification, 129-31, 169, 241 biotin:streptavidin, 130-1, 169, 241 liposomes, 130-1, 169 fluorochrome combinations, 51—3, 57 fluorochromes, 50-3, 123, 125, 128-9, 131, 133, 146-9, 169, 175, 179-80, 185, 191, 196-7, 201-7, 243, 247-66, 269, 282-3, 288, 290-3, 322-5, 328, 330-2, 367, 371, 373 absorbance to tubing, 185 acetoxy-methyl ester derivatives, 133, 146-7 acridine orange, 50, 185, 204-6, 243, 247-56 allophycocyanin, 50 amino-methyl coumarin acetic acid (AMCA), 50, 125, 128, 282-3 bisbenzimidazoles, 50, 51-2, 133, 149-9, 185, 201-3, 256-66, 292, 371

carboxyfluorescein, 132, 146 chromomycin A3, 50, 201-2, 292-3 combinations of, 51 coumarins, 50, 125, 128, 282-3, 330-2 cyanins, 50, 146-7 dansyl chloride, 50 DAPI, 50, 201-2, 292 dicyanodihydroxybenzene (DCDHB), 50, 146-7 diphenyl hexatriene, 125, 133 DIPI, 50, 201, 293 ethidium bromide, 50, 175, 203-4, 288, 290, 292

fluorescein, 50-3, 125, 128-9 fluorescein diacetate, (FDA), 180, 185, 322-4, 330-2 fluorescein isocyanate (FITC), 50, 125, 133, 169 fura-II, 50, 148 Hoechst 33258, 50, 201-3, 292 Hoechst 33342, 50, 52-3, 148-9, 185, 256-66, 371, 373 Indo-1, 50, 147 3-o-methylfluoresceinphosphate (MFP), 338 4-methyl umbellferyl acetate (MUA), 330-2 mithramycin, 50, 191, 201-2 monochlorobimane, 50, 325, 367 naphthol derivatives, 50 olivomycin, 201-2 oxanoles, 50, 205 phenanthridinium, 49-52, 133, 175, 179, 196-7, 203-4, 269, 288, 290-2 phycoerythrin, 50, 125, 129 polyanion, 204-7 propidium iodide, 49-52, 123, 179, 196-7, 203-4, 269, 291-2 pyronine-Y, 50, 205, 256-9 quin-II, 50, 147 resorufin, 50 rhodamine, 50, 125, 169 rhodamine-123, 133 SITS, 50 Texas red, 50 thiazole orange, 50, 207 thioflavine T, 50, 207 tricyclic antibiotics, 50, 201-2 tricyclic heteroaromatic, 204-7 fluorescence, 49-60, 145-9, 169, 259-65, 371, 373, 375

absorption spectra, 49 auto, 152, 158 breakthrough, 51-3 emission spectra, 49 excitation, 54-60 liposomes, 169

438

INDEX

fluorescence (cont.) nature of, 47 quenching of, 49, 54-5, 134-5, 167, 235, 250 resonant energy transfer, 54 spectrum analysis, 145-9, 259-65, 371, 373, 375 fluorescein, 50-3, 125, 128-9 fluorescein diacetate (FDA), 180, 185, 322-4, 330-2 fluorescein isocyanate (FITC), 50, 125, 133, 169 fluorescent microbeads, 159 focussing, 8--12, 22-31, 54, 153, 179, 186 chromatic compensation in, 26-31 conjugate distances, 22-4 crossed cylindrical lenses, 25, 27—31, 153, 179, 187 focal length, 22-4 hydrodynamic, 8-12 multiple beams, 26-31 spherical lens, 22-7, 29-30, 54 single beam, 25-6 forward light scatter, 61-3, 186-91 fractionated radiotherapy, 380-1 fura-II, 50, 147 (3-galatosidase, 32S gating, 81-2, 85-94, 195-8, 283-5 one-dimensional, 94 multi-dimensional, 86-92, 283-5 two-dimensional, 82, 85, 89, 195-8, 283-5 Gaussian, 57-60, 97-100, 143, 170-1, 231 distribution, 97-100, 231 profile, 51-60 statistics, 98-100, 143, 170-1 gene constructs, 276-7 gene mapping, 305-8 gene switching, 255-6 genomic libraries, 304-5 p-glucuronidase, 322, 325 y-glutamyl transpeptidase, 324, 328 glutathione, 325-8, 367-8, 384 glutathione S-transferase, 326 glycosaminoglycans, 118 haematoxylin, 1 helium-neon (HeNe) lasers, 57 helium-cadmium (HeCd) lasers, 57 hereditary tumours, 357 HER2/neu, 358-60, 366 homeostatic mechanisms, 382 Hooke, Robert, 32 high-speed sorting, 303 high-tension voltage, 151, 161-4, 173-5 hidden surface elimination, 83-4

histogram, 60, 76, 81-3, 97-100, 208-23 analysis, 211-13 deconvolution, 97-9, 213-23 DNA 76, 60, 208-11 mono-dimensional, 81 two-dimensional, 82-3 histones, 247, 268-9 Hoechst 33258, 50, 201-2, 292 Hoechst 33342, 50, 51-3, 148-9, 185, 256-66, 371, 373 Huygens principle, 32, 61 hyaluronidase, 119 hydrodynamic focussing, 8-12 double sheath, 12 factors involved in, 9—10 Reynolds number, 6 single sheath, 11 stability of, 11 hydrostatic pressure, 5 hyperbaric oxygen, 239-40, 381 hyperfractionation, 384 hypoxic cells, 381 illumination variation, 58-60 immunofluorescence staining, 50, 126-32 amplification, 129-31 cell surface, 126-7 combination, 127-9 intracellular antigens, 131-2 dyes, 50 incandescent light sources, see arc lamps indo-1, 50, 147 ink-jet writing, 106 information extraction, 3, 76, 80, 92 instrument hygiene, 185 instrument performance, 150-85 integer arithmetic, 75-6 interchromatin granules, 269 intermediate filaments, 285-7 intermitotic phase times, 227 interference, 31-4, 36-7 filters, 36-7 nature of, 32-3 thin films, 33 Young's experiment, 32 interference filters, 33, 36-7, 51, 148, 161, 175, 282-3, 371 construction of, 37 principles of, 33 uses of, 37, 51, 148, 161, 175, 282-3, 371 intracellular antigens, 268-87 cytoplasmic, 283-7 nuclear, 268-83 inverse square law, 54 ionophore, 146-7

INDEX jet-in-air, 41-3, 107 light collection from, 42-3 Ki-67, 269

krypton lasers, 57, 148 laminar flow, 6-8 lasers, 58-60, 148, 156 argon, 57, 148 beam geometry, 58-60 dye, 57 helium-neon (HeNe), 57 helium-cadmium (HeCd), 57 krypton, 57 lasing lines, 57 noise 156 stabilization, 57, 156 current, 156 light, 156 TEM output modes, 58 laws of, 5, 54 conservation of energy, 5 inverse square, 54 motion, 5 lens formula, 22 lenses, 22-31, 54, 153, 179, 186 crossed cylindrical, 27-31, 153, 179, 186 cylindrical, 25 spherical, 22-4, 26-7 thick, 29-30 thin, 23-5, 54 leucine amino-peptidase, 322 leucine zipper, 270 leukaemia, 367 leukocyte classification, 349 light, 18-66, 134-5, 140, 167, 175, 235, 250, 372 absorption, 34, 47-9 bleaching, 49, 53-4, 167 collection, 39-47 collection efficiency, 40-7, 167, 175, 372 colour, 20 detectors, 65-6, 140 diffraction, 31-4, 62 energy transfer, 49, 53-4 filtration, 34-9 flux, 23, 54, 167 fluorescence, 47—54 focussing, 23—31 interference, 31—4 luminescence, 47 phosphorescence, 47 quantum phenomena, 49 quenching, 49, 54-5, 134-5, 167, 235, 250 reflection, 18, 62 refraction, 18-19, 62

439

resonant energy transfer, 54 stabilization, 57 thin films, 33 wave nature of, 32 light flux, 23, 54, 167, 169 light collection, 39-47, 372 cone -|-angle, 40-1 efficiency, 40-7 flow chamber design in, 41—7 cuvette, 41 collection efficiency of, 43, 372 diverging refraction from, 41 modified cuvette, 43-4 spherico-ellipsoidal, 45-7 stream-in-air, 41—3 numerical aperture, 40 light scatter, 18-19, 31-4, 60-4, 141-2, 151-2, 167, 183-200 anomalous diffraction, 62 applications, 183-200 dual-angle scatter, 191-4 forward, 186-91 multi-angle scatter, 198-200 viability determination, 194-8 white cell discrimination, 183 diffraction, 31-4, 62 forward, 61-3, 186-91 Maxwell, 61 Mie scattering, 61 multi-angle, 141-2 Rayleigh scattering, 62-3 reflection, 18, 62 refraction, 18-19, 62 right angle, 64, 151-2, 167, 191-4 sweep scanning, 142 linear amplification, 66, 171-2 linearity, 173-6 A-to-D converters, 173 amplifiers, 173 measurement of, 174-5 photomultipliers, 174—5 liposomes, 131 list-mode data, 73 logarithmic amplification, 66-8, 154, 159-60, 171-2 log-normal distributions, 49, 53, 229 long-pass filters, 51, 168, 175 luminescence, 47 lung cancer, 354 lysosyme, 119 malignant melanoma, 355, 357 mean, 94 measurement range, 66-8, 72-3, 155, 159-65 A-to-D converters, 72-3 log amplifiers, 66-8, 154, 159-60

440

INDEX

measurement range (cont.) neutral density filters, 160 variable gain, 160—5 mechanical disaggregation, 117-18 mechanical vibration, 154 building, 154 laser cooling water, 154 optical mounts, 154 median, 95 medulloblastome, 355 methotrexate, 384 3-o-methyl fluorescein phosphate, 328 4-methyl umbelliferyl acetate (MUA), 330-2 membrane enzymes, 328-30 membrane potential dyes, 50, 133, 146-7 metachromasia, 51, 204—6 Michaelis-Menten hyperbola, 322 microbeads, 159 Micrographia, 33

mithramycin, 50, 191, 201-2 structure of, 202 mitochondrial dyes, 133 mitotic selection, 225-7, 288-9 mode, 95 modified cuvette, 41-4 monochlorobimane, 50, 325, 367 mono-dimensional (univariate) data, 81 multi-angle scatter, 141-2, 198-9 multi-beam focussing, 26—31 multi-detectors, 39, 142, 198 multi-drug resistance, 340-2, 369-78, 384 multi-parameter data and analysis, 76-80, 85-92, 195-8, 283-5 multi-parameter display, 81-6 multi-variate analysis, 355 Muscovy glass, 33 naphthol derivatives, 50 neuroblastoma, 355 neutral density filters, 35, 160-5 Newton, 3, 20-1, 33 nigericin, 146-7 nitrogen mustard, 369 noise, 150-8 biological, 158 dirty optics, 156 electronic, 151-4 fluidic, 154 light sources, 156 mechanical, 154 oscilloscope traces in, 151-2 preparative, 157 stray light, 155-6 non-Hodgkin's lymphoma, 351, 354 non-linear response, 174-6 nuclear antigens, 132, 247, 268-83, 297-8, 358, 364

antibody staining and quantitation, 269-73 histones, 247, 268-9 interchromatin granules, 269 Ki-67, 269 myc constructs, 276-7 p53, 269, 358 p55c~fos, 269-70, 282-3 p62c~myc, 269-82 p62 c-mK staining, 270-1 problems, 132 topoisomerases, 269, 274, 297-8 turnover measurements, 273, 364 nucleic acids and protein, 266—87 non-viable cells, 268-87 viable cells, 266-8 nucleic acid stains, 49-53, 133, 148-9, 175, 179, 185, 196-7, 201-7, 243, 227-66, 269, 283-5, 288, 290-2, 350, 371 DNA specific, 50, 52-3, 133, 148-9, 185, 201-3, 256-66, 292, 371 bisbenzimidazoles, 52-3, 133, 148-9, 185, 201-3, 256-66, 292, 371 phenylindoles, 201-2 tricyclic antibiotic, 50, 201-2 non-specific poly-anion, 50, 185, 204—7, 243, 247-59, 283-5, 350 acridine orange, 50, 185, 204-6, 243, 247-56, 283-5, 350 tricyclic heteroaromatic dyes, 204-7 nucleic acid specific, 49-52, 133, 175, 179, 196-7, 203-4, 269, 288, 290-2 phenanthridinium, 49-52, 133, 175, 179, 196-7, 203-4, 269, 288, 290-2 RNA part-specific', 50, 185, 204-6, 243, 247-59, 283-5, 350 acridine orange, 50, 185, 204-7, 243, 247-56, 283-5, 350 pyronine-Y, 259-60 thioflavine T, 50 nuclear-to-cytoplasmic ratio, 138, 350 numerical aperture, 40 obscruation bar, 43 off-axis aberration, 30 off-scale data, 96 oligonucleotide hybridization, 2, 299, 305 olivomycin, 201-2 oncogenes, 356-66 oophorectomy, 380 opticalfiltration,34-9, 51, 148, 160-5, 168, 175, 282-3 absorption, 34-6, 168, 175 band-pass, 37, 51, 148, 161, 175, 281-3 coloured glass, see absorption dichroic mirrors, 37, 168

INDEX dichroic combinations, 38-9, 282-3 interference, 36 long-pass, see absorption neutral density, 35, 160-5 short-pass, see interference oscilloscope traces, 38 ovarian borderline malignancy, 352 ovarian cancer, 352, 363 oxanoles, 51, 204 p21 c ~ ras , 358 p53, 271, 358 p62c~myc, 358-66 p62c~myc, turnover, 364 paraxial lens formula, 22 pepsin, 120-2, 351 archival material, 120-2, 351 specificity, 118 peroxidases, 320-1 permeabilization, 122-5 detergent, 123-4 fixation, 122—4 freeze—thaw, 124—5 hypotonic lysis, 123 lysolecithin, 125 pH, 133, 146-7 measuremnt of, 146-7 probes for, 146-7 phenanthridinium dyes, 49-52, 133, 175, 179, 196-7, 203-4, 269, 288, 290-2 phenylalanine mustard, 369 phenylindole dyes, 201—2 phosphorescence, 47 photocathode, 65 photodetectors, 65-6, 140 photolysis, 53 photomultiplier noise, 151-3 photomultipliers, 65-6, 151-2, 161-3, 167, 173-5 cathode, 65 high-tension voltage, 65, 151-2, 161-2, 167, 173-5 linearity, 173-5 noise, 151-3 sensitivity, 65 signal-to-noise ratio, 150, 167 spectral response, 65-6 phycoerythrin, 50, 125, 129 PLM curve, 228-31 ploidy, see DNA, index polyanion stains, 204—7 Poisson statistics, 15-17, 110-14, 143, 177 potential doubling time, 383 precision, 13-17, 72, 173-6 ADC offsets, 72, 173, 176 ADC resolution, 72 non-linear response, 174-6

441

coincidence, 13-17, 176 preparation of samples, 117-35, 206, 243-4, 246-59, 269-73, 282-3 enzymatic disaggregation, 118-22 mechanical disaggregation, 117-18 permeabilization, 122-5 staining, 125-35, 206, 243-4, 246-59, 269-73, 282-3 wax embedded material, 120-1 pressure profile, 6 prognosis in cancer, see applications in oncology pronase, 120 propidium iodide, 49-52, 123, 179, 196-7, 203-4, 269, 291-2 structure of, 204 prospects in cancer, see applications in oncology protease, 120 protein A, 131 protein turnover measurement, 273, 364 primary data space, 84—92 Ptolemy, 18-19 pulse shape analysis, 179-80 pyronine-Y, 50, 204, 256-9 quantum phenomena, 49 quality control, 176-85 coincidence correction, 177-9 inspection, 177 pulse shape analysis, 179-80 time, 182-5 quenching, 49, 54, 134-5, 167, 235, 250 quin-II, 50, 147 radio-immuno assay, 158-9 rate-event analysis, 143-5 discrimination in, 144-5 statistics, 143 resorufin, 50 Rayleigh scattering, 62-3 receptor status, 366 reflection, 18, 62 refraction, 18-19, 22-4, 62 converging, 22-4 diverging, 22-4 measurement of, 19-20 refractive index, 18-21 renal adenocarcinoma, 355 resolution, 170-2 resonance cavity, 36—7 resonant energy transfer, 54 retinoblastoma suppressor gene, 358 retro viruses, 356 reverse transcriptase, 356 Reynolds number, 6 rhodamine, 50, 125, 169

442

INDEX

rhodamine-123, 133 ribonuclease, 50 right angle scatter, 151-2, 168, 179 RNA and DNA staining, 246-59 S-phase fraction, 352, 354-5 sample contamination, 116, 154—5 sample filtration, 136 scatter, see light scatter screening, see applications in oncology secondary data space, 84-92 sensitivity, 34-9, 40-7, 166-70, 372 bleaching, 167 excitation intensity, 166 exposure time, 166 fluorochrome amplification, 169 light collection efficiency, 40-7, 167, 372 measurement of, 160-70 optical filtration, 34-9, 167-9 sequential illumination, 30 sequential illumination triggering, 71-2 sheath contamination, 154-5 sheath filtration, 155 short-pass filter, 36, 168 'skewed-normal' distribution, 98, 100 side scatter, see right angle scatter signal processing, 66-71 signal-to-noise ratio, 150, 166, 336 SITS, 50 slit-scanning, 137-41, 291, 301, 350 chromosome analysis, 138-9, 291, 301 cytological prescreening, 137—8, 350 image plane, 140—1 object plane, 137-40 Snell's Law, 18 solid-state devices, 66, 107, 140, 142 sorting efficiency, 115-16 sorting purity, 110—14 sorting times, 110 sorting yield, 110-14 sources of variation, 165 spherical aberation, 23-4 S-phase probability distribution, 218-19, 233 spherical lenses, 22—4, 26—7 spherico-ellipsoidal flow chamber, 45—7 spectral, 49, 145-9, 259-65 absorption, 49 emission, 49, 145-9, 259-65 staining, 125-36, 243-4, 246-59, 262-4, 269-73, 282-3 antibody combination, 127-9, 282-3 cytoplasmic antigens, 131—2 DNA, see nucleic acid stains factors influencing, 134-5, 243-4, 262-4 fluorochrome amplification, 129-31 interactive stains, 132-3

liposomes, 131 non-interactive stains, 136 nuclear antigens, 132-3, 269-73 RNA, 246-59 surface antigens, 126-7 standards, 159, 244-6 statistics, 14-17, 97-100, 110-14, 143, 170-1 binomial, 14, 113-14 Gaussian, 97-100 Poisson, 15-17, 110-14 rare event, 143 Student's-T, 97, 170-1 stoichiometry, 134—5, 239 stereo-perspective graphics, 84 stray light, 155-6 stream-in-air, see jet-in-air streptavidin, 130-1, 169, 241 sweep-scanning, 142 SV-40 large T, 269, 283 synthetic peptide antibodies, 132 tertiary data space, 84-92 testicular cancer, 360-1 Texas red, 50 therapy selection in oncology, see applications in oncology thiazole orange, 50, 207 thick lenses, 29-30 thinfilms,33 thin lenses, 23-5 thioflavine T, 50, 207 three-dimensional (trivariate) data, 84-6 time, 182-5, 310-17, 320-35 in quality control, 182-5 incorporation into data base, 310-17 continuous, 311-13 continuous interrupted sampling, 310 computer clock, 181, 311 discontinuous sequential sampling, 310 short time scales, 312-17, 333-5 kinetics with, 310, 312-17, 320-35 topoisomerases, 269, 274, 297-8 transverse emission mode (TEM), 57 tricyclic antibiotics, 51, 201—3 tricyclic heteroaromatic dyes, 204—7 triggering, 177-9, 191 trivariate (three-dimensional) data, 84-6 trout red cells, 244-6 trypsin, 118 tryptophan, 132 tubulin, 283, 287 tumour growth rate, 239-41, 356, 380-4 tumour suppressor genes, 357 turbulent flow, 7-8 two-dimensional (bivariate) data, 81, 195-8, 283-5

INDEX

443

tyrosine, 132

vortex, 10

variable gain, 160-5 •1-,^-f ,«., , we . ,„« verapamil, 341, 372, 375, 377 viability, 194-8 vibration, 154 voltage thresholds, 177-9

L J J J L • I ^™ n ^C-T ^ wax embedded material, 120-2, 351-6, 358

"6

Young's experiment, 32