Tetraploid cycle in ageing solid tumours

When the mouse mammary adenocarcinoma 755 (Ca-755) reaches the plateau phase of growth, non-cycling cells with a G2-DNA content can be observed. They may belong to the diploid cell cycle but they could also be blocked in G0 or G1 of a tetraploid cycle. This hypothesis was tested in three ways: (1) non-cycling G2 nuclei were stained with a combination of Feulgen and naphthol yellow which revealed two populations, one with a low protein content and the other with a high protein content--the latter may represent nuclei ready to begin a new phase of DNA synthesis; (2) Feulgen staining and autoradiography were performed after tritiated thymidine had been administered to mice continuously: this showed that there were cells synthesizing DNA with a DNA index above 2; and (3) cells having 80 chromosomes, corresponding to the tetraploid cycle, were found almost exclusively in the plateau phase tumours. On the other hand, the use of texture and DNA parameters of the Feulgen stained nuclei showed that they were concentrated in a diploid cycle for tumours in the exponential phase of growth and were divided between a diploid and tetraploid cycle for 'plateau' cells. Neither the cause for, nor the role played by, polyploid cells is known.     

Assessment of cell viability by flow cytometric analysis using DNase exclusion

A new method was developed for selective measurement of DNA distributions in viable cell populations. The method is based on the fact that non-viable cells lose membrane integrity and treatment of such cells with DNase should remove their DNA. The DNase-treated cells were stained with DNA fluorochrome 4′-6-diamidino-2-phenylindole (DAPI) in the presence of Triton X-100. DNA distribution was measured by flow cytometry prior to and after treatment with DNase. Percentage of cells stained after DNase treatment was considered as an index of cell viability. Optimal conditions for DNase treatment and application of DNase exclusion test for the analysis of spontaneous cell death, selective death of cells arrested in S/G2 phases, instant cell disintegration induced by cytotoxic compounds and cell death induced by hyperthermia are described.     

Nuclear non-histone protein content of G0/G1 cells as related to growth fraction in mouse mammary tumors.

In eight mouse mammary tumors with varying growth fractions DNA and non-histone nuclear protein (NHNP) were determined by absorption cytophotometry of Feulgen-Naphthol Yellow S stained, isolated cells. It was found that: 1. The mean NHNP content of cells with postmitotic DNA content (G0 + G1) increased with increasing growth fraction. 2. The mean NHNP content of S and G2 cells in the eight tumors did not vary significantly with growth fraction. 3. The frequency distributions of NHNP in G0/G1 cells were unimodal and right-skewed. The results are interpreted as follows: A) G0 cells differ from G1 cells by their lower content of NHNP. B). If it is assumed that the G0 and G1 compartments are arranged in series, the cells in the transition from G0 to late G1 may account for the unimodality and skewedness of the NHNP frequency distributions of postmitotic cells.     

Flow cytometric analysis of live cell proliferation and phenotype in populations with low viability.

BACKGROUND: Combined analysis of DNA content and immunofluorescence on single cells by flow cytometry provides information on the proliferative response of subpopulations to stimuli in mixed cell preparations; however, in low-viability cell preparations, dead cells interfere with accurate flow cytometric data analysis because of nonspecific binding of antibodies and altered DNA-staining profiles. Light scatter differences between nonviable and viable cells are unreliable, particularly after the cell permeabilization step that is necessary for DNA staining. We developed a method for identification of nonviable cells by fluorescence in cell preparations that are stained simultaneously for cell surface or intracellular immunofluorescence and DNA content. MATERIALS AND METHODS: Nonviable cells that have lost membrane integrity are identified by uptake of 7-amino-actinomycin D (7-AAD). Transfer of 7-AAD from stained nonviable cells to unstained viable cells after permeabilization is prevented by blocking DNA binding with nonfluorescent actinomycin D (AD). Pyronin Y(G) (PY) is used for DNA staining because the orange spectral emission of PY can be separated from the green fluorescein isothiocyanate (FITC) emission and the red emission of 7-AAD, respectively. RESULTS: Application of the method to the analysis of the T-cell leukemia cell line Molt-4f and of cultured human peripheral blood mononuclear cells is presented. In both cell preparations, 7-AAD staining permitted reliable dead cell exclusion. Live, 7-AAD-negative Molt-4f cells showed higher expression levels of cell surface CD4 and of intracellular CD3, showed a higher proportion of cells in the G1 phase of the cell cycle, and showed a lower coefficient of variation of the G1 peak compared with data obtained from all the cells in the preparation. Live, CD8+ lymphocytes from OKT3-stimulated cultures of human peripheral blood mononuclear cells showed a specific proliferative response as measured by DNA content analysis. CONCLUSIONS: The results show that cells stained with FITC-labeled antibodies can be analyzed by single-laser flow cytometry for DNA content combined with dead cell discrimination. Furthermore, they emphasize the need for exclusion of dead cells from the analysis of cell preparations with low viability to obtain reliable data on immunofluorescence and cell-cycle distributions.     

Expression of proliferation associated antigens in the cell cycle of synchronized mammalian cells.

Flow cytometric bivariate analysis was used to investigate the expression of PCNA, p120 and p145 during the cell cycle of a mammalian cell line (CHO-K1). Initially, aliquots of cells in exponential and plateau (G0) phase were analyzed for proliferation associated antigen expression. Expression of PCNA and p145 during G0 was markedly depressed (less than 12% positive) while 54% of the G0 cells stained positive for p120. The fluorescent intensity (mean channel fluorescence) of these G0 positive p120 cells, however, was only slightly above the mean channel fluorescence (MCF) of cells stained with a negative isotype control. In asynchronous cultures, all three antigens were expressed in greater than 70% of the cells, with PCNA staining being greater than 95%. Cells were then synchronized using mitotic selection (mitotic index of 97%) and antigen levels were measured as cells progressed synchronously through the cell cycle. From DNA analysis histograms, it appeared that the degree of synchrony was approximately 90% throughout the remainder of the cell cycle. The bivariate DNA/PCNA, DNA/p120, and DNA/p145 histograms for mitotic cells indicated that both p120 and p145 expression were elevated (percent positive and MCF) while PCNA levels were near controls (MCF). In early G1, all three markers were depressed (less than 12% positive); however PCNA levels rose precipitously in mid-G1 (greater than 50% positive). In late G1 to early S, p145 levels increased concomitantly with increases in p120. All three antigens were elevated throughout S phase and began to decline as cells moved from G2/M to G1 of the next cell cycle with p145 expression decreasing first. This report indicates that all three proliferation associated antigens studied are differentially expressed in the cell cycle and therefore may be useful in detecting and assessing the proliferation state.     

Caffeine dissociates complexes between DNA and intercalating dyes: application for bleaching fluorochrome-stained cells for their subsequent restaining and analysis by laser scanning cytometry

BACKGROUND: Removal of the nucleic acid-bound fluorochrome is desirable when stained cells have to be reanalyzed using other fluorochromes. It is also often desirable to remove DNA-bound antitumor drugs from drug-treated cells, to improve cell staining. We have previously observed that in aqueous solutions, the methylxanthine caffeine (CFN) decreases interactions between planar aromatic molecules such as intercalating dyes or antitumor drugs and nucleic acids. The aim of this study was to explore whether this property of CFN can be utilized to remove the DNA-bound intercalating dyes propidium iodide (PI) or 7-aminoactinomycin D (7-AAD) from the cells and whether the bleached cells can be restained and reanalyzed. METHODS: HL-60 cells were fixed in 70% ethanol and their DNA was stained with PI or 7-AAD. The cells were then rinsed with a 0.05 M solution of CFN in phosphate-buffered saline (PBS) or with PBS alone. The decrease in intensity of cell fluorescence during rinsing was measured by laser scanning cytometry (LSC) to obtain the bleaching kinetics of individual cells. The bleached cells were then restained with PI, 7-AAD, or the protein-specific fluorochrome sulforhodamine 101(S101). Their fluorescence was measured again by LSC. In addition, free DNA was subjected to gel electrophoresis, DNA bands in the gels were stained with ethidium bromide (EB), and the gels were rinsed with a solution of CFN or PBS to bleach the DNA band's fluorescence. RESULTS: Rinsing the PI or 7-AAD-stained cells with solutions of CFN led to nearly complete removal of PI and a more than 75% decrease in 7-AAD fluorescence after 10 min. The rinse with PBS decreased the PI cell fluorescence intensity by less than 30% and the 7-AAD fluorescence by about 50%. The differences in kinetics of PI or 7-AAD removal by CFN from G2/M versus G1 cells suggest that these intercalators bind more strongly to DNA in chromatin of G2/M than G1 cells. The CFN-bleached cells were then successfully stained with S101 and again with PI or 7-AAD. The bivariate analysis of the LSC merged files of the cells sequentially stained with PI and S101 revealed typical DNA/protein distributions. The fluorescence of EB-stained DNA bands in gels was also nearly completely removed by rinsing gels in 0.05 M CFN; PBS alone had a distinctly lesser effect. CONCLUSION: Solutions of CFN can dissociate the DNA-bound PI, 7-AAD, EB, and possibly other intercalating fluorochromes. The bleached cells can be restained and reanalyzed by LSC. This approach can also be used to remove such fluorochromes from nucleic acids immobilized in gels and perhaps in other solid matrices. Analysis of the kinetics of fluorochrome removal from cells can possibly be used to study their binding affinities to nucleic acids in situ.     

A comparison of low-power helium-cadmium and argon ultraviolet lasers in commercial flow cytometers

The feasibility of installing a low power ultraviolet (UV) laser in a commercial flow cytometer was evaluated by testing an Ortho Cytofluorograf 50HH and a Coulter Epics V. Both instruments were equipped with two argon ion lasers, one emitting at 488 nm and the other in the UV region and were tested by measuring the DNA content of cells stained with Hoechst 33342 or DAPI. The coefficient of variation (CV) of the G0/G1 peak of the DNA histograms produced by each instrument did not deteriorate markedly when results obtained at 100-125 mW were compared to those obtained at 10 mW. These tests indicated that a helium-cadmium laser (He-Cd) which can produce 10 mW at 325 nm should work well as a UV laser in these instruments. An Ortho Cytofluorograf IIs was purchased with a 10 mW He-Cd laser installed in the forward position. Studies of DNA content have confirmed that this low power UV laser can produce CVs of 2.2% with DAPI stained fixed cells and 3.6% with Hoechst 33342 stained viable lymphocytes. Thus, the He-Cd laser should provide a reasonable alternative as a UV source for flow cytometers.     

Normal and perturbed Chinese hamster ovary cells: correlation of DNA, RNA, and protein content by flow cytometry

Quantitative, correlated determinations of DNA, RNA, and protein, as well as RNA to DNA and RNA to protein ratios, were performed on three-color stained cells using a multiwavelength-excitation flow cytometer. DNA-bound Hoechst 33342 (blue), protein-fluorescein isothiocyanate (green), and RNA-bound pyronin Y (red) fluorescence measurements were correlated as each stained cell intersected three spatially separated laser beams. The analytical scheme provided sensitive and accurate fluorescence determinations by minimizing the effects of overlap in the spectral characteristics of the three dyes. Computer analysis was used to generate two-parameter contour density profiles as well as to obtain numerical data for subpopulations delineated on the basis of cellular DNA content. Such determinations allowed for analysis of RNA to DNA and RNA to protein ratios for cells within particular regions of the cell cycle. The technique was used to study the interrelationship of DNA, RNA, and protein contents in exponentially growing Chinese hamster ovary cells as well as in cell populations progressing the cell cycle after release from arrest in G1 phase. The sensitivity of the method for early detection of conditions of unbalanced growth is demonstrated in the comparison of the differential effects of the cycle-perturbing agent, adriamycin, on cells treated either during exponential growth or while reversibly arrested in G1 phase.     

Simultaneous detection of cyclin B1, p105, and DNA content provides complete cell cycle phase fraction analysis of cells that endoreduplicate

BACKGROUND: DNA analysis of endoreduplicating cells is difficult because of the overlap between stem-line G2 + M cells and 4C G1 cells. Simultaneous flow cytometry of DNA and cyclin B1 analytically separates these populations. The objective here was to develop simultaneous flow cytometry of DNA, cyclin B1, and p105 (highly expressed in mitosis) for improved, complete cell cycle phase fraction analysis of endoreduplicating cell populations. METHODS: Monoclonal antibody, GNS-1, reactive with human cyclin B1, was conjugated with fluorescein at three different fluorochrome-to-protein (F/P) ratios and tested for optimal sensitivity in a flow cytometric assay. A formaldehyde-methanol fixation procedure was optimized for retention of p105 within mitotic cells by analytic titration of formaldehyde. p105 was stained indirectly with Cy5-conjugated secondary antibody, followed by GNS-1, and DNA was stained with Hoechst 33342. The specificity of p105 in this assay was tested by comparison of manual and flow cytometric mitotic indices and by sorting and microscopic inspection. RESULTS: F/P 4.1 provided optimal fluorescein labeling of GNS-1. Formaldehyde (0.5%), followed by methanol permeabilization, fixed cells sufficiently to quantify stem-line and endoreduplicated G1, S, G2, and M phase fractions. Kinetic measurements of these fractions for both populations were demonstrated. CONCLUSIONS: The fluorochrome-to-protein ratio is important and can be optimized objectively for these assays. A permeabilization-sensitive antigen (p105), previously requiring formaldehyde/detergent-fixed cell preparations, was shown to work equally well with formaldehyde/ methanol fixation. Three-laser, two-parameter intracellular antigen analysis can be successfully coupled with DNA content analysis. Cell cycle kinetic analysis of endoreduplicating populations should be improved.     

Expression of proliferating cell nuclear antigen (PCNA)/cyclin during the cell cycle

The expression of proliferating cell nuclear antigen (PCNA), also called cyclin, was quantified in the cell lines SP2/0 and MOLT-4 and in mouse splenocytes induced to proliferate in vitro with mitogens. Autoantibody from a patient with systemic lupus erythematosus was used to label PCNA in cell suspensions after the cells had been fixed and permeabilized. In the same cells DNA was stained by propidium iodide. The cells were then analysed by flow cytometry for PCNA and DNA content. The PCNA profiles in proliferating spleen cells and the cell lines were similar. Most G0-G1 cells did not express significant amount of PCNA. A dramatic increase in PCNA immunofluorescence was observed in late G1 cells, and further increases were observed in S-phase cells. G2-M cells showed a reduced level of PCNA immunofluorescence relative to S-phase cells but were still elevated relative to G0-G1 cells. Proliferating cells arrested at the G1-S boundary by exposure to cytosine arabinoside showed an increased PCNA immunofluorescence as compared to unstimulated cells.     

Basal-cell subpopulations and cell-cycle kinetics in human epidermal explant cultures

Cultured human epidermal cells were studied by cell sorting and autoradiography after different 3H-thymidine (3H-dThd)-labelling procedures and after labelling with DNA precursors that are incorporated via salvage or de novo pathways. It was shown that 3H-dThd incorporation was the best measure of the rate of DNA replication. Dose-response experiments with pulse and continuous labelling revealed that all S- and G2-phase cells were cycling, whereas some 20% of the cells stayed in G1-phase for long periods of time. Most, if not all of these cells were probably non-proliferating differentiated keratinocytes. At least two subpopulations of S-phase cells could be discriminated on the basis of the rate of incorporation of DNA precursors. The difference in precursor incorporation did not seem to be caused by differences in nucleotide metabolism but rather to reflect true differences in the rate of DNA replication. Continuous labelling experiments showed that these subpopulations also were apparent in the G1- and G2-phases. Studies of the grain-count distribution revealed that cells that appeared to move rapidly through the S-phase moved slowly through the G2-phase, and vice versa. Cells stained with acridine orange were subjected to a two-parameter analysis in the cell sorter by simultaneous measurement of the DNA and RNA fluorescence. Autoradiography of sorted cells revealed that, on average, cells with low RNA contents incorporated 3H-dThd at a higher rate than cells with high RNA contents.     

Flow cytometric cell cycle analysis using the quenching of 33258 Hoechst fluorescence by bromodeoxyuridine incorporation.

Cultures of L cells were grown in medium containing 2.0 mg/l bromodeoxyuridine (BUdR) and stained with the fluorescent dye 33258 Hoechst for flow cytometric analysis. During exposure to BUdR, The cells replace thymidine by BUdR in the newly synthesized DNA. The new DNA is not stainable with 33258 Hoechst, which is highly specific for thymidine. The temporal development of the fluorescence distributions after addition of BUdR to the growth medium has been investigated in the flow cytometer, and the data were used to calculate the mean durations of the phases G1, S and G2 + M in exponentially growing cultures as well as the cycle transit times in synchronized cultures. The percentage of non-cycling cells was determined in each experiment.     

The cell cycle associated change of the Ki-67 reactive nuclear antigen expression

The change of human nuclear antigen expression in proliferating cells recognized by a monoclonal antibody, Ki-67, during the cell cycle was investigated in HeLa S3 cells using a bivariate-flow-cytometric analysis. The antigen was immunocytochemically stained with FITC, and DNA was stained with propidium iodide (Pl). The expression of the antigen increased with cell-cycle progression, especially in the latter half of S-phase and reached a maximum at G2M-phase, although its content varied greatly from cell to cell. The cells in which DNA synthesis was inhibited by treatment with hydroxyurea increased markedly in the antigen expression (as compared to untreated cells). Treatment with adriamycin also elevated the antigen content. After digestion with DNase I, but not after RNase treatment, FITC fluorescence from the antigen disappeared. These results suggest that the Ki-67 antigen is bound to DNA and its expression does not depend on DNA replication. Although the biological implications of the antigen remain unresolved, the antigen may be considered to be essential for maintaining the proliferating state of cells.     

Cytometry of chromatin bound Mcm6 and PCNA identifies two states in G1 that are separated functionally by the G1 restriction point<Superscript>1</Superscript>

Background  

Cytometric measurements of DNA content and chromatin-bound Mcm2 have demonstrated bimodal patterns of expression in G1. These patterns, the replication licensing function of Mcm proteins, and a correlation between Mcm loading and cell cycle commitment for cells re-entering the cell cycle, led us to test the idea that cells expressing a defined high level of chromatin-bound Mcm6 in G1 are committed - i.e., past the G1 restriction point. We developed a cell-based assay for tightly-bound PCNA (PCNA*) and Mcm6 (Mcm6*), DNA content, and a mitotic marker to clearly define G1, S, G2, and M phases of the cell cycle. hTERT-BJ1, hTERT-RPE-1, and Molt4 cells were extracted with Triton X-100 followed by methanol fixation, stained with antibodies and DAPI, then measured by cytometry.     

Different instability of nuclear DNA at acid hydrolysis in cancerous and noncancerous cells as revealed by fluorescent staining with acridine orange

Ehrlich cancer cells and inflammatory cells in mouse ascitic fluid were hydrolyzed and stained with acridine orange (AO). The AO hydrolysis curves for G1/G2 + M phase cancer cells and inflammatory cells were differentially determined using flow cytometry by monitoring the metachromatic red-shifted fluorescence of the fluorochrome bound to the single-stranded DNA produced by acid hydrolysis. By computer fitting of the Bateman function to the hydrolysis curves, the kinetic parameters k1 (rate constant for the production of single-stranded DNA), k2 (rate constant for the degradation of the produced single-stranded DNA), and y0 (theoretical value of the single-stranded DNA present initially) were determined. It was found that the k2 value, which reflects the degree of DNA instability, was much higher for cancer cells in both the G1 and G2 + M phases than for inflammatory cells. This finding led us to develop a method for the differential AO staining of cancer cells and non-cancerous cells utilizing the different degree of DNA instability at acid hydrolysis. AO staining after hydrolysis with 2N HCl at 30 degrees C for 8.5 min was found to be the optimal method. In the 60 cases of human malignant epithelial and nonepithelial tumors tested, all of the malignant tumor cells emitted metachromatic red fluorescence, while all of the nonmalignant tumor cells (5 cases of benign tumor) and normal cells emitted orthochromatic green fluorescence when observed with a violet excitation light under a fluorescence microscope. This new technique can be a useful tool for the screening of malignancy in exfoliative cytology and also for basic cancer research.     

Sequential estimation of nuclear DNA and silver staining of nucleoli in plant cells

A method for sequential estimation of nuclear DNA and silver staining of nucleoli in plant cells is described. Feulgen staining is done first and nuclear DNA estimated by absorption cytophotometry. Following this, the slides are stained with AgNO3. The method has been used to study the process of nucleolar fusion in garlic (Allium sativum L.) meristem root tip cells. It was found that during interphase nucleoli rarely fused, thus most fusion must have occurred before the G1 phase of the cell cycle.     

Flow cytometric determination of cell cycle parameters of V79 cells by continuous labeling with bromodeoxyuridine

A simple method with which to determine the cell cycle parameters, TG1, TS and TG2M (the durations of the G1, S and G2 + M phases) is described. V79 Chinese hamster lung cells were used to evaluate the method. After continuous labeling with bromodeoxyuridine (BrdU), V79 cells were stained with anti BrdU-monoclonal antibody with FITC (fluorescein isothiocyanate) and with PI (propidium iodide). The individual cells were checked by flow cytometry for green and red fluorescences whose signal intensities corresponded to the BrdU and cellular DNA contents. The durations of G1, S and G2 + M phases of V79 cells were determined by measuring the cell fractions containing the nonlabeled G1, labeled S and nonlabeled G2 + M phases. The reliability of this method is discussed.     

A new method for rapid and sensitive detection of bromodeoxyuridine in DNA-replicating cells

A new flow cytometric technique, involving differential fluorescence analysis of two DNA-binding fluorochromes, was used to quantify cellular incorporation of the base analog, bromodeoxyuridine (BrdU), into DNA over short time periods. During analysis of stained cells, the blue fluorescence signal of Hoechst 33342, which is quenched by BrdU-substituted DNA, was subtracted, on a cell by cell basis, from the green-yellow fluorescence signal of mithramycin, which remained stoichiometric to cellular DNA content. Bivariate contour profiles obtained for CHO cells pulse-labeled for 30 min showed that fluorescence quenching of Hoechst 33342 in BrdU-labeled, S phase cells produced fluorescence difference signals that were significantly greater than the difference signals from G1 and G2 + M phase cells. Analysis of L1210 cells demonstrated that the amount of BrdU detected was proportional to the length of the labeling period. The novel technique is simple, rapid, and mild; it produces minimal cell loss and does not significantly affect cellular moieties such as DNA, chromatin, or RNA.     

DNA synthesis in the second and third cell cycles of mouse preimplantation development. A cytophotometric study

Female Swiss mice were sacrificed at 2 h intervals between 16–30 and 40–56 h after insemination. One-, 2- and 4-cell embryos were stained by the Feulgen method and cytophotometric measurement of their nuclear DNA content was carried out. The cells with 2C and 4C DNA content were assumed to be in G1 and G2 phase and those with intermediate DNA content in S phase of the cell cycle. The fractions of cells which had passed a given phase of the cell cycle were calculated for various times after insemination and utilized for measurements of the second and third cell cycle timing. Results of measurements for the second cell cycle: G1 phase 1.3 h, S phase 6.1 h, G2 phase 15.4 h, whereas for the third cell cycle: G1 phase 1.6 h, S phase 7.4 h, G2 phase 0.5 h. The first cleavage division was calculated as 1.6 h, the second as 1.3 h and the third as 1.2 h. Complete intra-embryonic synchronization of the DNA-synthesizing nuclei was preserved during the entire synthesis phase of 2-cell embryos, while in 4-cell embryos they were slightly asynchronized. Among mitotic cells of the first cleavage division and G1 cells of 2-cell embryos a slight interembryonic asynchronization was found which deepened during subsequent cell cycle phases.     

Immunoreactivity to antinucleoside antibodies in X-irradiated HeLa cells

Synchronized HeLa cells were stained with antibodies to purine and pyrimidine nucleosides by immunofluorescent and immunoperoxidase techniques. These antibodies react only with denatured or single-stranded regions of DNA. Nuclear attachment of antibody was seen only during the period of DNA synthesis as determined by ³H-thymidine incorporation. Positive nuclear immunoreactivity was seen in approx. 15% of cells obtained by mitotic selection at a time corresponding to the G 1 phase. After exposure to ionizing radiation, 80% of the G 1 cells were reactive. Induction of immunoreactivity was dose dependent over the range of 100 to 1 000 rads. Treatment of irradiated G 1 cells with deoxyribonuclease completely eliminated the positive nuclear reaction. Exposure to ribonuclease had no effect. Incubation of the G 1 cells for 90 min at 37 or 0–4 °C after the administration of 1 000 rads resulted in a prompt decrease of immunoreactivity to control levels. However, in the presence of 0.04 μg/ml actinomycin D, positive nuclear staining remained at high levels. No such effect could be observed as a result of exposure to cytosine arabinoside or hydroxyurea. It is concluded that X-irradiation of G 1 HeLa cells produces single-stranded regions in nuclear DNA that can be detected by anti-pyrimidine and anti-purine antibodies.