Denaturation of deoxyribonucleic acid in situ effect of formaldehyde.

In situ denaturation of nuclear deoxyribonucleic acid (DNA) is studied by use of acridine orange to differentially stain native versus denatured DNA, and a flow-through cytofluorometer for measurements of cell fluorescence. Thermal- or acid-induced DNA denaturation is markedly influenced by formaldehyde. Two mechanisms of the formaldehyde action are distinguished. If cells are exposed to the agent during heating, DNA denaturation is facilitated, most likely by the direct action of formaldehyde as a "passive" denaturing agent on DNA. If cells are pretreated with formaldehyde which is then removed, DNA resistance to denaturation increases, presumably due to chromatin cross-linking. It is believed that both effects occur simultaneously in conventional techniques employing formaldehyde to study DNA in situ, and that the extent of each varies with the temperature and cell type (chromatin condensation). Thus, profiles of DNA denaturation of cells heated with formaldehyde do not represent characteristics of DNA denaturation in situ; DNA denaturation under these conditions is modulated by the reactivity of chromatin components with formaldehyde rather than by DNA interactions with the macromolecules of nuclear mileu.     

In situ factors affecting stability of the DNA helix in interphase nuclei and metaphase chromosomes

The data from earlier cytochemical studies, in which the metachromatic fluorochrome acridine orange (AO) was used to differentially stain single vs double-stranded DNA, suggested that DNA in situ in intact metaphase chromosomes or in condensed chromatin of G0 cells is more sensitive to denaturation, induced by heat or acid, than DNA in decondensed chromatin of interphase nuclei. Present studies show that, indeed, DNA in permeabilized metaphase cells, in contrast to cells in interphase, when exposed to buffers of low pH (1.5-2.8) becomes digestible with the single-strand-specific S1 or mung bean nucleases. A variety of extraction procedures and enzymatic treatments provided evidence that the presence of histones, HMG proteins, and S-S bonds in chromatin, as well as phosphorylation or poly(ADP)ribosylation of chromatin proteins, can be excluded as a factor responsible for the differential sensitivity of metaphase vs interphase DNA to denaturation. Cell treatment with NaCl at a concentration of 1.2 N and above abolished the difference between interphase and mitotic cells, rendering DNA in mitotic cells less sensitive to denaturation; such treatment also resulted in decondensation of chromatin visible by microscopy. The present data indicate that structural proteins extractable with greater than or equal to 1.2 N NaCl may be involved in anchoring DNA to the nuclear matrix or chromosome scaffold and may be responsible for maintaining a high degree of chromatin compaction in situ, such as that observed in metaphase chromosomes or in G0 cells. Following dissociation of histones, the high spatial density of the charged DNA polymer may induce topological strain on the double helix, thus decreasing its local stability; this can be detected by metachromatic staining of DNA with AO or digestion with single-strand-specific nucleases.     

DNA content by in situ fluorescence imaging and S-phase detection, with chromatin structure preserved

OBJECTIVE: To test the feasibility of in situ DNA quantitation of adherent cells' nuclei by fluorescence imaging, preserving chromatin structure and to follow-up S phase, in relation to DNA content, in order to assess the precision of DNA measurements. STUDY DESIGN: Double labeling experiments involved total DNA staining with Hoechst 33342 and BrdU immunostaining (after either Br photolysis and DNA strand break labeling by terminal transferase or acid denaturation) to detect replicating DNA. An epifluorescence microscope was used, images captured with a CCD camera and quantitative total DNA measurements done in 12 bits with IPLab software. BrdU results were related to DNA content on an individual cell basis. Cell cycle analyses were run with Imastat software (developed in the laboratory) on Hoechst-stained cells and on double labeled cells. RESULTS: In cells progressing through the cycle, as assessed by BrdU, a corresponding increase in DNA content was measured. Early S differed from G1 (P < .05). Imastat analyses gave a CV for GI peak of 6-7%. CONCLUSION: Quantitative fluorescence imaging allows a sensitive determination of DNA content for adherent-cell nuclei in situ. Topologic analyses of nuclear components will be possible in relation to DNA content.     

Different sensitivity of DNA in situ in interphase and metaphase chromatin to heat denaturation

Heat denaturation of DNA in situ, in unbroken cells, was studied in relation to the cell cycle. DNA in metaphase cells denatured at lower temperatures (8 degrees-10 degrees C lower) than DNA in interphase cells. Among interphase cells, small differences between G1, S, and G2 cells were observed at temperatures above 90 degrees C. The difference between metaphase and interphase cells increased after short pretreatment with formaldehyde, decreased when cells were heated in the presence of 1 mM MgCl2, and was abolished by cell pretreatment with 0.5 N HCl. The results suggest that acid-soluble constituents of chromatin confer local stability to DNA and that the degree of stabilization is lower in metaphase chromosomes than in interphase nuclei. These in situ results remain in contrast to the published data showing no difference in DNA denaturation in chromatin isolated from interphase and metaphase cells. It is likely that factors exist which influence the stability of DNA in situ are associated with the super-structural organization of chromatin in intact nuclei and which are lost during chromatin isolation and solubilization. Since DNA denaturation is assayed after cell cooling, there is also a possibility that the extent of denatured DNA may be influenced by some factors that control strand separation and DNA reassociation. The different stainability of interphase vs. metaphase cells, based on the difference in stability of DNA, offers a method for determining mitotic indices by flow cytofluorometry, and a possible new parameter for sorting cells in metaphase.     

Ultrastructure of meristematic cells of dormant and released buds in Tradescantia paludosa

The lateral bud meristems of Tradescantia paludosa show a characteristic cytohistological zonation during dormancy. The cells comprising this so called ‘zone of inhibition’, which is located at the extreme tip of the bud apex, rarely synthesize nuclear DNA or undergo mitotic division. These nuclei are as large as prophase nuclei, yet contain only telophase (2C) amounts of DNA and significantly lower amounts of histone as compared to the 2C nuclei of the actively dividing cells.Ultrastructural observations of the nuclei in the ‘zone of inhibition’ show that a large proportion of the chromatin is organized as less condensed, diffuse, euchromatin fibrils; however, the chromatin of the actively dividing nuclei of the cells outside the ‘zone of inhibition’ or in the released bud meristems is organized to a greater extent as condensed clumps of heterochromatin. When the dormancy is released, the nuclei in the ‘zone of inhibition’ synthesize DNA and histone and undergo cell division in approx. 4 days. Striking changes in the organization of chromatin fibrils take place during this transition period. The diffuse chromatin fibrils of the nuclei in the ‘zone of inhibition’ progressively become more and more condensed as the cell prepares to undergo the first mitotic division after the release of dormancy. This change which is coupled with the synthesis of histones in the nuclei of the ‘zone of inhibition’ suggests a prominent structural role of these basic proteins in the organization of the chromatin. The large volume of 2C nuclei of the ‘zone of inhibition’ seems, therefore, to result not from a great nuclear mass, but probably from a relatively small degree of condensation of chromatin.     

Detection of 5-bromodeoxyuridine (BrdUrd) incorporation by monoclonal antibodies: role of the DNA denaturation step

Immunochemical detection of cells that incorporate 5-bromodeoxyuridine (BrdUrd) requires prior denaturation of DNA in situ to make BrdUrd binding sites accessible to the antibodies. A technique is described in which the DNA denaturation step is facilitated by a) prior dissociation of histones from DNA and b) the use of low ionic strength buffer in which the cells are suspended during heating. Dissociation of histones is achieved by cell treatment with 0.08N HCl at 0 degree C, which a) increases accessibility of DNA to propidium iodide (and following the denaturation to the antibodies); b) lowers stability of DNA to thermal denaturation; c) decreases differences between various cell types due to variability in chromatin structure; and d) ensures more complete DNA denaturation. Cell heating (80-95 degrees C) at low ionic strength (1 mM Na+) eliminates the need for formamide and results in extensive and rapid DNA denaturation. The method was applied in Friend leukemia, L1210 and HL-60 cell lines, and to bone marrow, experimental animal tumor and primary human tumor cells.     

Linker DNA destabilizes condensed chromatin.

The contribution of the linker region to maintenance of condensed chromatin was examined in two model systems, namely sea urchin sperm nuclei and chicken red blood cell nuclei. Linkerless nuclei, prepared by extensive digestion with micrococcal nuclease, were compared with Native nuclei using several assays, including microscopic appearance, nuclear turbidity, salt stability, and trypsin resistance. Chromatin in the Linkerless nuclei was highly condensed, resembling pyknotic chromatin in apoptotic cells. Linkerless nuclei were more stable in low ionic strength buffers and more resistant to trypsin than Native nuclei. Analysis of histones from the trypsinized nuclei by polyacrylamide gel electrophoresis showed that specific histone H1, H2B, and H3 tail regions stabilized linker DNA in condensed nuclei. Thermal denaturation of soluble chromatin preparations from differentially trypsinized sperm nuclei demonstrated that the N-terminal regions of histones Sp H1, Sp H2B, and H3 bind tightly to linker DNA, causing it to denature at a high temperature. We conclude that linker DNA exerts a disruptive force on condensed chromatin structure which is counteracted by binding of specific histone tail regions to the linker DNA. The inherent instability of the linker region may be significant in all eukaryotic chromatins and may promote gene activation in living cells.     

Unstainable DNA in cell nuclei. Comparison of ten different fluorochromes

Ten fluorochromes with specificity for DNA were used to compare the stainability of nuclei of exponentially growing, nondifferentiated Friend leukemia (FL) cells with that of dimethylsulfoxide-induced, fully differentiated FL cell nuclei. Decreased accessibility of DNA to several dyes, particularly pronounced in the case of some intercalators, was observed in differentiated cells. Dye binding was also compared for both sets of nuclei following extraction of nuclear proteins, mostly histones, with 0.1-N HCl. Acid extraction of nuclear proteins increased the accessibility of DNA to varying degrees, depending upon the fluorochrome. In most cases, the differences in fluorescence between differentiated and nondifferentiated nuclei stained with most intercalating dyes was abolished by acid treatment. The results are discussed in terms of the mode of interaction between DNA and the various fluorochromes and the factors associated with chromatin structure, which may affect or be associated with different degrees of proliferative activity.     

Thermal denaturation of nucleohistones--effects of formaldehyde reaction

Thermal denaturation of native or partially dehistonized nucleohistones shows two melting bands at 66 and 81° in 2.5 × 10^?4 M EDTA, pH 8.0. These correspond to the melting of DNA segments bound by the less basic and the more basic half-molecules of histones, respectively. These two melting bands combine into a broad melting band from around 70 to 85° when these nucleohistones are pre-treated with formaldehyde. A formaldehyde reaction which fixes histones on DNA by covalent bonds account for the effect. Formaldehyde fixation also increases the melting temperature of some free DNA segments from around 42 to around 55°. This is interpreted as a result of closed or rigid boundaries between free DNA and formaldehyde-reacted histone-bound DNA segments. MgCl₂ dissociates histones from DNA more effectively and leaves longer free DNA segments than does NaCl. Thermal denaturation of a formaldehyde-reacted nucleoprotein thus provides an effective tool for comparing the relative size of free DNA regions on nucleoproteins. The effect of reversible binding of ligands on helix-coil transition of DNA is descussed and found not adequate for thermal denaturation of nucleohistones.     

DNA denaturation in situ. Effect of divalent cations and alcohols

Heat denaturation profiles of rat thymus DNA, in intact cells, reveal the presence of two main DNA fractions differing in sensitivities to heat. The thermosensitive DNA fraction shows certain properties similar to those of free DNA: its stability to heat is decreased by alcohols and is increased in the presence of the divalent cations Ca2+, Mn2+, or Mg2+ at concentrations of 0.1-1.0 mM. Unlike free DNA, however, this fraction denatures over a wide range of temperature, and is heterogeneous, consisting of at least two subfractions with different melting points. The thermoresistant DNA fraction shows lowered stability to heat in the presence of Ca2+, Mn2+, or Mg2+ and increased stability in the presence of alcohols. It denatures within a relatively narrow range of temperature, consists of at least three subfractions, and, most likely, represents DNA masked by histones. The effect of Ca2+, Mn2+, or Mg2+ in lowering the melting point of the thermoresistant DNA fraction is seen at cation concentrations comparable to those required to maintain gross chromatin structure in cell nuclei or to support superhelical DNA conformation in isolated chromatin (0.5-1.0 mM). It is probable that factors involved in the maintenance of gross chromatin organization in situ and/or related to DNA superhelicity also have a role in modulating DNA-histone interactions, and that DNA-protein interactions as revealed by conventional methods using isolated chromatin may be different from those revealed when gross chromatin morphology remains intact.     

Isolation and characterization of the nuclear matrix in Friend erythroleukemia cells: chromatin and hnRNA interactions with the nuclear matrix.

Nuclear matrices from undifferentiated and differentiated Friend erythroleukemia cells have been obtained by a method which removes DNA in a physiological buffer. These matrices preserved the characteristic topographical distribution of condensed and diffuse "chromatin" regions, as do nuclei in situ or isolated nuclei. Histone H1 was released from the nuclear matrix of undifferentiated cells by 0.3 M KCl; inner core histones were released by 1 M KCl. Nuclear matrix from differentiated cells did not maintain H1, and histone cores were fully released in 0.7 M KCl. KCl removed the core histones as an octameric structure with no evidence of preferential release of any single histone. Electron microscopy of KCl-treated matrix revealed no condensed regions but rather a network of fibrils in the whole DNA-depleted nuclei. When nuclear matrices from both types of cell were exposed to conditions of very low ionic strength, inner core histones and condensed regions remained. These observations support the contention that inner core histones are bound to matrix through natural ionic bonds or saline-labile elements, and that these interactions are implicated in chromatin condensation. hnRNA remained undegraded and tenaciously associated to the matrix fibrils, and was released only by chemical means which, by breaking hydrophobic and hydrogen bonds, produced matrix lysis. Very few nonhistone proteins were released upon complete digestion of DNA from either type of nuclei. The remaining nonhistone proteins represent a large number of species of which the majority may be matrix components. The molecular architecture in both condensed and diffuse regions of interphase nuclei appears to be constructed of two distinct kinds of fibers; the thicker chromatin fibers are interwoven with the thinner matrix fibers. The latter are formed by a heteropolymer of many different proteins.     

Metakaryotic stem cell lineages in organogenesis of humans and other metazoans

A non-eukaryotic, metakaryotic cell with large, open mouthed, bell shaped nuclei represents an important stem cell lineage in fetal/juvenile organogenesis in humans and rodents. each human bell shaped nucleus contains the diploid human DNA genome as tested by quantitative Feulgen DNA cytometry and fluorescent in situ hybridization with human pan-telomeric, pan-centromeric and chromosome specific probes. From weeks ∼5–12 of human gestation the bell shaped nuclei are found in organ anlagen enclosed in sarcomeric tubular syncytia. Within syncytia bell shaped nuclear number increases binomially up to 16 or 32 nuclei; clusters of syncytia are regularly dispersed in organ anlagen. Syncytial bell shaped nuclei demonstrate two forms of symmetrical amitoses, facing or “kissing” bells and “stacking” bells resembling separation of two paper cups. Remarkably, DNA increase and nuclear fission occur coordinately. Importantly, syncytial bell shaped nuclei undergo asymmetrical amitoses creating organ specific ensembles of up to eight distinct closed nuclear forms, a characteristic required of a stem cell lineage. Closed nuclei emerging from bell shaped nuclei are eukaryotic as demonstrated by their subsequent increases by extra-syncytial mitoses populating the parenchyma of growing anlagen. From 9–14 weeks syncytia fragment forming single cells with bell shaped nuclei that continue to display both symmetrical and asymmetrical amitoses. These forms persist in the juvenile period and are specifically observed in bases of colonic crypts. Metakaryotic forms are found in organogenesis of humans, rats, mice and the plant Arabidopsis indicating an evolutionary origin prior to the divergence of plants and animals.     

Transformation of ram spermatid chromatin

In order to investigate the sequence of changes in nuclear basic proteins throughout ram spermiogenesis, we have used different techniques to obtain populations of spermatid nuclei in specific stages of maturation. Sedimentation of testis cells at 1 gravity followed by treatment with Triton X-100 resulted in one population of round spermatid nuclei (steps 1–a), one of non-round spermatid nuclei (steps 8b-15), and one of elongated spermatid nuclei (steps 12–15). Populations of non-round spermatid nuclei were obtained by treatment with EDTA (steps 9–15), by sonication (steps 12–15) and digestion by DNase (steps 13–15). Nuclear proteins, extracted either directly with dilute acid or following a reducing treatment with 2-mercaptoethanol were characterized by polyacrylamide gel electrophoresis.The most striking alterations in protein composition occur during the elongation phase (steps 8–12). The five histones are displaced from chromatin at the same rate. When they are freed of histones (step 12), the nuclei start to accumulate the sperm-specific protein (BNSP) which is then partly extractable by dilute acid without a thiol that is required for its extraction from more mature nuclei. This stepwise replacement process is accompanied by a reduction of the basic protein amount bound to DNA. As soon as histones begin to disappear, eight spermatidal protein fractions are present in the nuclei until the BNSP synthesis reaches its maximum rate. Except for one, they all contain cysteine and are partially intermolecularly cross-linked in the chromatin. After in vivo and in vitro labelling experiments, they are synthesized in elongating spermatids (steps 8–11). None are degradation products of histones.Correlations of the times of onset of EDTA, sonication and DNase resistances with changes in the basic nuclear proteins point out that stabilization and condensation of spermatid chromatin is promoted through a progressive increase in disulfide bridges.     

Denaturation, renaturation, and loss of DNA during in situ hybridization procedures

With the aim of optimizing in situ hybridization methods, alkaline, acid, and thermal denaturation procedures have been studied for their ability to separate the DNA strands of nuclear DNA and for the DNA losses they induce. Isolated methanol/acetic acid-fixed mouse liver nuclei have been used as a biological object. The results, obtained with acridine orange staining and microfluorometry, show that all denaturations studied lead to almost complete strand separation. Quantitative DNA staining and cytometry indicated that with heat and alkaline denaturation about 40% of the DNA is lost. Acid denaturation led to about 20% DNA loss. For the alkaline denaturation, the DNA retention could be improved to a 20% DNA loss by adding 70% ethanol to the denaturation medium. During hybridization, another 20% DNA loss occurs. When denatured nuclei are brought under annealing conditions, a rapid renaturation of a considerable fraction of the remaining DNA occurs. The extent of renaturation was dependent on the type of denaturation used. For the ethanolic alkaline denaturation, it was estimated to be 35%. Quantitative nonautoradiographic in situ hybridization experiments with acetylaminofluorene-modified mouse satellite DNA showed that alkaline denaturation procedures are superior to the heat and acid denaturation. As proven by acridine orange fluorescence measurements, hybridization conditions can be designed that permit DNA.RNA hybridization under in situ DNA.DNA denaturing conditions. These conditions should be very useful, especially for in situ hybridization with single-stranded RNA probes.     

Circular dichroism of histone-bound regions in chromatin.

Native, NaCl-treated, trypsin-treated, and polylysine-bound nucleohistones were studied in 2.5 × 10^?4 M EDTA, pH 8.0, using circular dichroism (CD) and thermal denaturation. Removal of histone I by 0.6 M NaCl has a much smaller effect on both Δε₂₂₀ and Δε₂₇₈ than the removal of other histones. This indicates that histone I has less helical content and less conformational effect on the DNA in nucleohistone. By extrapolating to 100% binding by histones other than I, the positive CD band near 275 nm is close to zero. Comparison is also made between the effects of binding by the more basic and the less basic halves of histones by trypsin-digestion and polylysine-binding experiments. Trypsin digestion of nucleohistone reduces melting band IV at 82°C much more than melting band III at 72°C. However, the CD changes of Δε₂₇₈ and Δε₂₂₀ induced by trypsin digestion are small, unless melting band III is also reduced by the use of a higher trypsin level. This implies that the less basic halves of histones, which stabilize DNA to 72°C (melting band III), have more helical structure and are more responsible for conformational change in DNA than are the more basic halves, which stabilize DNA to 82°C (melting band IV). Polylysine binding to nucleohistone diminishes melting band III but has no effect on melting band IV. This binding affects only slightly the Δε₂₂₀ of nucleohistone, indicating that polylysine interferes very little with the structure of the less basic halves of bound histones. The implications of these studies with respect to chromatin structure are discussed.     

Methods of denaturation and renaturation of DNA in interphasic chromatin: cytochemical quantitative analysis by Methyl Green staining

Almost diploid nuclei (as judged from the microdensitometric evaluation of the Feulgen positive material) of granular and Purkinje cells of the rat cerebellar cortex, were submitted to in situ DNA denaturation and renaturation experiments. We assessed the double-strandedness of DNA, by Methyl Green staining according to Scott (1967). Under these conditions a stoichiometric ratio between bound dye and DNA exists, suitable for quantitative microdensitometric measurements. Our data show that DNA in the interphasic chromatin is never completely denatured after the treatments we used. Furthermore, the renaturation takes place in a different way in the two cell types. Owing to the unlike chromatin packing of granular and Purkinje nuclei, we suggest that nuclear proteins must interfere differently on the in situ denaturation and renaturation processes.     

NaCl-aided Hoechst 33258 staining method for DNA quantification and its application

Summary We investigated the effect of salt on the fluorescence staining procedure for quantification of the amount of DNA in cell nuclei in situ. For this, NaCl was added at various concentrations to the Hoechst 33258 fluorochrome (Hoe) medium for staining DNA. The fluorescence intensity of free DNA-Hoe solution was not changed by the addition of NaCl, but that of the nuclei-Hoe complex in situ increased 4-fold on increasing the NaCl concentration up to 1 M. SDS polyacrylamide gel electrophoresis showed that histones H1, H2A, and H2B dissociated from cell nuclei in the presence of 1 M NaCl, resulting in increasing accessibility of DNA to the fluorochrome.The applicability of the NaCl-aided fluorescence staining method was evaluated by measuring the ploidy classes of various cells. The amount of DNA in spermatozoa is half that in 2n hepatocytes, but by the conventional Hoe staining procedure the fluorescence intensity of spermatozoa is higher than that of 2n hepatocytes, due to differences in accessibility of the dye to DNA. In contrast, by the NaCl-aided procedure, the fluorescence intensity of 2n hepatocytes was twice that of spermatozoa. The effectiveness of the NaCl-aided Hoe staining method was checked using cultivated human gingival cells and hepatocytes of LEC rats with hereditary hepatitis. In all cases, reasonable proportionality between the fluorescence intensity and the amount of DNA was observed.     

Accurate determination of DNA content in single cell nuclei stained with Hoechst 33258 fluorochrome at high salt concentration

Summary In an attempt to achieve accurate quantification of DNA levels in cell nuclie, we studied the influence of salt concentration on the fluorescence of cell nuclei complexed with Hoechst-33258 (Hoe) fluorochrome. The fluorescence of cell nuclei was compared with that of extracted DNA as well as that of nucleosome core. Conformational changes in these complexes were examined by measuring both fluorescence anisotropy and fluorescence lifetime in the nanosecond region. The results showed that the fluorescence of DNA-Hoe was quenched by the nucleosomal structure, there being an associated increase in anisotropy and a decrease in the fluorescence lifetime; however, the fluorescence was restored to the orginal level by the addition of a high concentration of NaCl, CsCl, or LiCl. The reduction in fluorescence may have been due to loss of fluorescence energy caused by collision of the fluorophore with histones in the nucleosome. The addition of 1 M NaCl to the medium used for staining with Hoe greatly stabilized the fluorescence of DNA in cell nuclei. The DNA content of individual cell nuclei was determined by comparing the fluorescence of these nuclei with that of a standard DNA solution. For lymphocytes and liver ploidy cells, reasonably accurate values were obtained by applying the present method.     

HAMLET interacts with histones and chromatin in tumor cell nuclei

HAMLET is a folding variant of human alpha-lactalbumin in an active complex with oleic acid. HAMLET selectively enters tumor cells, accumulates in their nuclei and induces apoptosis-like cell death. This study examined the interactions of HAMLET with nuclear constituents and identified histones as targets. HAMLET was found to bind histone H3 strongly and to lesser extent histones H4 and H2B. The specificity of these interactions was confirmed using BIAcore technology and chromatin assembly assays. In vivo in tumor cells, HAMLET co-localized with histones and perturbed the chromatin structure; HAMLET was found associated with chromatin in an insoluble nuclear fraction resistant to salt extraction. In vitro, HAMLET bound strongly to histones and impaired their deposition on DNA. We conclude that HAMLET interacts with histones and chromatin in tumor cell nuclei and propose that this interaction locks the cells into the death pathway by irreversibly disrupting chromatin organization.     

Denaturation and condensation of DNA in situ induced by acridine orange in relation to chromatin changes during growth and differentiation of Friend erythroleukemia cells

DNA in situ is progressively denatured when the cells or nuclei are treated with increasing concentration of acridine orange (AO). This transition can be monitored by flow cytometry as a decrease in green fluorescence. The complexes of denatured DNA and AO undergo immediate condensation and aggregation; this step is manifested by appearance of red luminescence and formation of precipitates that can be detected by electron microscopy. The precipitates form preferentially in heterochromatin as well as in ribosomes and polysomes. Their formation and further aggregation affects cellular light scatter properties in both the forward and right-angle direction. The AO-induced DNA denaturation and condensation was studied in nuclei of Friend erythroleukemia cells from exponentially growing, differentiated or quiescent cells. The DNA in nuclei of quiescent cells, from plateau-phase cultures, was the most sensitive to denaturation; it denatured (measured by changes in luminescence) at an AO concentration between 50 and 80 microM with the midpoint of the transition (Cd) at 70 microM. DNA in nuclei of differentiated cells (dimethyl-sulfoxide-induced erythroid differentiation) was more resistant (Cd = 77-83 microM), whereas DNA in exponentially growing cells was the most resistant (Cd = 86 microM). Extraction of proteins with 0.1 M HCl at 0 degree C abolished the differences between the cells and shifted the transition to a lower AO concentration (Cd = 46 microM). For comparison, the midpoint transitions representing condensation of free, nucleic acids measured as light scatter changes occurred at 13, 22, 31 and 53 microM of AO, for rRNA, tRNA, and denatured and native-calf thymus DNA, respectively. Denaturation and condensation of DNA, which can be induced by AO either in isolated nuclei or viable permeabilized or fixed cells provides a new approach to discriminate cell subpopulations with different chromatin structure by flow cytometry. The molecular mechanisms of this phenomenon are discussed.