Fine specificities of autoantibodies directed against the Ro, La, Sm, RNP, and Jo-1 proteins defined by two-dimensional gel electrophoresis and immunoblotting

Although useful for specific purposes, immunofluorescence, precipitation in agarose gels, and the m.w. estimation of RNA or proteins immunoprecipitated from transformed cells often provide partial or ambiguous definition of autoantibody specificity. We have analyzed organ and cell extracts by one-and two-dimensional electrophoresis together with Western blotting to define the fine specificities of antibodies to the ribonucleoprotein (RNP) antigens Ro, La, Sm, RNP and Jo-1. One-dimensional analysis identified the Ro protein as a 57 kilodalton (kd) protein, although many anti-Ro sera also react with a 50 kd protein. La antisera react with 50 and 43 kd proteins. The 50 kd La protein readily breaks down into 43, 25, and smaller immunoreactive cleavage products. Partial proteolysis of Ro and La proteins in human spleen extracts produces similar immunoreactive products, providing evidence for a common structure. The major immunoreactive Sm antigens defined by human polyclonal antisera and a mouse monoclonal antiserum were doublets of 25/26 and 16/18 kd, whereas anti-RNP sera reacted with a protein of 68 kd. Most Sm-RNP antisera contained antibodies reactive with additional proteins, especially when whole cell extracts were used as a source of antigens. Two-dimensional analysis provided characteristic maps of the antigens. Ro and La were acidic, and La showed a unique set of acidic charge isomers at 50 and 43 kd. Anti-Sm antibodies reacted with discrete dots corresponding to both the acidic and basic regions of the first-dimension (charge) gels, whereas the RNP antigen showed a series of basic charge isomers of 68 kd. Many anti-Sm-RNP sera reacted with other closely spaced proteins of a similar charge and size to the Sm and RNP antigens, suggesting antibody cross-reactivity or reactivity with closely related functional proteins. Although Jo-1 had the same m.w. as the undegraded La antigen, the fingerprints were quite distinctive on two-dimensional electrophoresis. The results of this study indicate how the source and preparation of antigen extracts, as well as protein degradation, influence the m.w. determinations of soluble protein antigens. With these factors taken into account, two-dimensional fractionation with immunoblotting provides a highly discriminating, sensitive, and reproducible method of analysis of autoantibody specificity. This technique can be used to standardize reference antisera and to study protein antigens in normal and abnormal cell and tissue extracts, and could lead to new or more precise correlations with clinical disease.     

Identification and specificity of a cDNA encoding the 70 kd mitochondrial antigen recognized in primary biliary cirrhosis

Mitochondrial autoantibodies are characteristic of the disease primary biliary cirrhosis (PBC), but the immunoreactive mitochondrial antigens have not been defined. We used a rat liver cDNA library in lambda gt 11-Amp3 to clone a 1370-base pair insert that coded for a polypeptide reactive with PBC sera. This insert was subcloned for expression into pBTA224, a plasmid vector in the same reading frame as lambda-Amp3. A positive clone, designated pRMIT, that expressed a fused polypeptide of 160 kd, was recognized by 25 of 25 sera from patients with PBC and none of 96 sera from normal persons or patients with systemic lupus erythematosus, rheumatoid arthritis, or chronic active hepatitis. This fused polypeptide was shown to correspond with the 70 kd mitochondrial autoantigen by several experiments. First, lysates of pRMIT in J101 absorbed out the 70 kd reactivity of PBC sera when probed against fractionated placental mitochondria. Second, affinity-purified antisera reactive with the fused polypeptide also reacted with the 70 kd mitochondrial antigen. Third, such affinity-purified antisera produced the characteristic anti-mitochondrial pattern of immunofluorescence on tissue sections. Finally, immunization of BALB/c mice with the fused polypeptide elicited antibodies to mitochondria. These murine antibodies reacted with the 70 kd mitochondrial protein and also produced typical mitochondrial immunofluorescence on tissue sections. The nucleotide and amino acid sequence of the recombinant protein, which encodes for approximately a 48 kd protein, showed no significant homologies with known proteins, and there were no homologies with mitochondrial genomic DNA. The availability of a recombinant form of the 70 kd mitochondrial autoantigen will allow several definitive questions to be addressed in PBC, including identification of B cell epitopes, T cell recognition, and a model of PBC in mice.     

A cytochemical study of the nonhistone protein content of condensed and extended chromatin

Cytochemical techniques have been used to study the distribution of nonhistone proteins in sections of interphase nuclei and mitotic chromosomes. Condensed chromatin, including the heterochromatin of interphase nuclei from frog liver, and mitotic metaphase and anaphase chromosomes from bovine kidney, show little or no staining for nonhistone protein. Regions of frog liver nuclei which contain extended chromatin (euchromatin) stain intensely for nonhistone protein. These differences in nonhistone staining of condensed and extended chromatin support the suggestion that regions of condensed chromatin contain considerably less nonhistone protein than regions of extended chromatin. The results suggest further that there may be considerably less nonhistone protein associated with chromosomes and interphase heterochromatin than has been reported in most previous analyses of isolated chromatin and chromosome preparations.     

Identification of specific and cross-reactive antigens of Leishmania donovani chagasi by human infection sera

Cloned Leishmania donovani chagasi (Ldc) promastigotes were analyzed by SDS-PAGE separation and immunoblotting with human infection sera. The patterns of antigen reactivity were compared by using sera from individuals with Ldc, Leishmania mexicana amazonensis (Lma), Trypanosoma cruzi, Mycobacterium tuberculosis, or Mycobacterium leprae infections. Sera from individuals with these infections recognized Ldc antigens in several m.w. ranges. Reactivity was due to recognition of Ldc molecules and not to Ldc culture medium components, as shown by comparing Ldc promastigotes grown in the presence or absence of fetal bovine serum (FBS), by immunoblotting of FBS, and by [35S]methionine labeling. The major findings of the study were as follows. Immunoblots with Ldc promastigotes could be used to distinguish individuals with Ldc infections from those with Lma infections. Persons with Ldc infections had antibodies to a Ldc antigen of approximately 32 to 35 kd not recognized by persons with Lma infections. Individuals cured of acute Ldc infection did not develop antibodies that differed in specificity to those present during their acute phase of infection. Ldc antigens in the 62 to 66 kd region were recognized by all individuals with Ldc or Lma infections but were not recognized by individuals in the other disease groups or by control sera. This region was found to contain at least four distinct bands, one of which appeared to be glycosylated as indicated by periodic acid-Schiff staining and concanavalin A labeling; an apparently nonglycosylated protein of 62 to 63 kd was eluted from SDS-PAGE gels and was used to diagnose Ldc infection by the ELISA. Whereas crude Ldc antigen gave false positive results with T. cruzi and mycobacteria infection sera, the eluted 62 to 63 kd protein was 100% specific and sensitive in the diagnosis of Ldc infection.     

Evidence that the G1-S and G2-M transitions are controlled by different cdc2 proteins in higher eukaryotes.

Xenopus eggs contain two distinct cdc2 homologs of 34 and 32 kd. We show that the 32 kd cdc2 protein, like the 34 kd protein, is a kinase. However, unlike the 34 kd homolog, the 32 kd cdc2 kinase activity does not decrease dramatically at the end of mitosis. The 32 kd protein does not associate with mitotic cyclins B1 and B2 but does associate with cyclin A and a novel doublet of proteins of 54 kd that may regulate its activity. We also show that depletion of the 32 kd cdc2 homolog from a Xenopus extract blocks DNA replication, but does not inhibit entry into mitosis. By contrast, depletion of the 34 kd cdc2 homolog or absence of mitotic cyclins from an extract does not inhibit replication, but does block entry into mitosis. Our results indicate that in higher eukaryotes, DNA replication (G1-S) and mitosis (G2-M) may be controlled by distinctly different cdc2 proteins.     

Identification of centromere proteins in different mammalian cells

The characterization of centromeric proteins is facilitated using anti-centromere antibodies present in the sera of patients with the CREST variant of scleroderma. We have employed these sera to determine whether or not those proteins are present in different mammalian species, as well as to study their tissue distribution. Here, we describe the immunofluorescent pattern and the proteins recognized by CREST sera in dividing and resting cells from mouse, rat, swine, hamster, rabbit, and man. In nuclear preparations from cultured cells, thymocytes and spermatozoa from these species, the antigens recognized by CREST sera are proteins of 18 to 20 kDa in all species tested, except in rat. Additionally, two peptides of 80 and 140 kDa were observed in human preparations. In contrast, a 50 kDa peptide is the primary protein detected by the sera in rat nuclei.     

Fluorescent staining of L cell centromeres and chromocenters with 1-dimethylaminonaphthalene-5-sulfonyl chloride and G-bandings

Fluorescent staining patterns of L cell chromosomes with 1-dimethylaminonaphthalene-5-sulfonyl chloride (dansyl chloride) were studied. Ordinary air-dried L cell metaphase chromosomes exhibited relatively uniform and bright yellowish green fluorescence by dansyl-staining under the fluorescence microscope. However, after the chromosome preparations were treated with 10 mM NaCl for 24 h at 4 °C, which produced distinctive G-bands with Giemsa-staining, the centromeric regions and several interstitial regions of some particular chromosomes were clearly fluorescent but other regions showed only dull fluorescence. After the treatment of chromosome slides with cupric sulfite reagent, which converts sulfhydryls and disulfides to thiosulfates chromosomes showed clear G-bands which were indistinguishable from those after 10 mM NaCl treatment. By dansyl-staining, however, the cupric sulfite-treated chromosomes exhibited very faint fluorescence on their contour alone, and neither centromeric regions nor some interstitial regions of marker chromosomes had distinctly bright fluorescence.Although Giemsa-staining disclosed dark chromocenters in approx. 75% of interphase nuclei irrespective of pretreatments, dansyl-staining revealed bright chromocenters in approx. 60% of interphase nuclei in control slides, in about 40% of nuclei in 10 mM NaCl-treated slides, and in only about 30% of nuclei in cupric sulfite-treated preparations.These observations indicated that in the air-dried chromosome preparations, the distribution of protein over the metaphase chromosome is relatively uniform along its length, and that G-bands in the chromosome and Giemsa-staining of chromocenters in interphase nuclei are not significantly affected by apparent loss of protein from the preparations. It was also suggested that particular protein may be associated with the centromeric regions of L cell chromosomes. Some technical details of dansyl fluorochroming and the significance of the observations were discussed.     

Leishmania mexicana pifanoi: analysis of the antigenic relationships between promastigotes and amastigotes by gel diffusion, immunoelectrophoresis, and immunoprecipitation

The antigenic relationships between Leishmania mexicana pifanoi promastigotes, axenically grown amastigotes, and amastigotes isolated from the footpads of infected hamsters or from a J774 macrophage cell line were studied by three serologic methods. Amastigote and promastigote antigens were disrupted by freeze-thawing of intact cells in a lysis buffer. Antisera were prepared in rabbits by repeated subcutaneous inoculations of the parasite antigens in complete Freund's adjuvant and were tested against the homologous and heterologous antigens in a series of gel diffusion experiments. Negative results were obtained in all control experiments. In each instance, the homologous antigen-antiserum reactions yielded the largest numbers of precipitin lines. A pattern of cross-reactivity was also observed in the heterologous systems. Results indicated that the amastigote and promastigote forms had unique and common antigens. The two parasite antigen-antiserum systems were also examined by immunoelectrophoresis. Qualitative and quantitative differences between the promastigote and amastigote antigens were readily demonstrable by this technique. Results indicated that each parasite form had specific and many common antigens. In the homologous system, major proteins, with molecular weights (MW) of 23, 52, and 68 kd, were demonstrated in the promastigotes by immunoprecipitation of lactoperoxidase-catalyzed radioiodinated cells. In a similar (homologous) system, axenically grown amastigotes were found to contain three proteins with MW of 38, 70, and 74 kd and, therefore, different from those demonstrated for the promastigotes. All the results suggested that the three amastigote stages of different origins are antigenically similar to one another, but differ from the promastigote forms.     

Leishmania mexicana pifanoi: Analysis of the Antigenic Relationships Between Promastigotes and Amastigotes by Gel Diffusion,Immunoelectrophoresis, and Immunoprecipitation1

ABSTRACT. The antigenic relationships between Leishmania mexicana pifanoi promastigotes, axenically grown amastigotes, and amastigotes isolated from the footpads of infected hamsters or from a J774 macrophage cell line were studied by three serologic methods. Amastigote and promastigote antigens were disrupted by freeze-thawing of intact cells in a lysis buffer. Antisera were prepared in rabbits by repeated subcutaneous inoculations of the parasite antigens in complete Freund's adjuvant and were tested against the homologous and heterologous antigens in a series of gel diffusion experiments. Negative results were obtained in all control experiments. In each instance, the homologous antigen-antiserum reactions yielded the largest numbers of precipitin lines. A pattern of cross-reactivity was also observed in the heterologous systems. Results indicated that the amastigote and promastigote forms had unique and common antigens. The two parasite antigen-antiserum systems were also examined by immunoelectrophoresis. Qualitative and quantitative differences between the promastigote and amastigote antigens were readily demonstrable by this technique. Results indicated that each parasite form had specific and many common antigens. In the homologous system, major proteins, with molecular weights (MW) of 23, 52, and 68 kd, were demonstrated in the promastigotes by immunoprecipitation of lactoperoxidase-catalyzed radioiodinated cells. In a similar (homologous) system, axenically grown amastigotes were found to contain three proteins with MW of 38, 70, and 74 kd and, therefore, different from those demonstrated for the promastigotes. All the results suggested that the three amastigote stages of different origins are antigenically similar to one another, but differ from the promastigote forms.     

Adenovirus E1b-58kd tumor antigen and SV40 large tumor antigen are physically associated with the same 54 kd cellular protein in transformed cells

The adenovirus E1b-58kd tumor antigen has been detected in a physical association with a 54 kilodalton cellular protein in adenovirus-transformed mouse cells. Antibody specific for the E1b-58kd protein coimmunoprecipitates a 54 kd protein from transformed, but not from productively infected, cells. Monoclonal antibody specific for the cellular 54 kd protein coimmunoprecipitates the adenovirus E1b-58kd protein from transformed cell extracts. The same or closely related cellular 54 kd protein, associated with the adenovirus E1b-58kd protein, was present in the SV40 large T antigen-54 kd complex previously detected in SV40-transformed mouse cells. The identity of the 54 kd protein is based on the immunological specificities of the anti-54 kd monoclonal antibodies and partial peptide maps of the 54 kd protein associated with the adenovirus and SV40 tumor antigens. The adenovirus E1b-58kd-54 kd complex, like the SV40 large T antigen-54 kd complex, is heterogeneous in size or mass. While all of the cellular 54 kd protein in the adenovirus-transformed cell extract is found in a complex with the E1b-58kd protein, some of the viral 58 kd antigen is detected in a form not associated with the 54 kd protein. The fact that the adenovirus and Sv40 tumor antigens, both required for transformation, can be found in physical association with the same cellular protein in a transformed cell is a good indication that these two diverse viral proteins share some common mechanisms or functions.     

The maternally expressed Drosophila gene encoding the chromatin-binding protein BJ1 is a homolog of the vertebrate gene Regulator of Chromatin Condensation, RCC1.

Using monoclonal antibodies I have identified a nuclear protein of Drosophila, BJ1 (Mr approximately 68 kd), and isolated its gene. Biochemical analysis demonstrates that the BJ1 protein is associated with nucleosomes and is released from chromatin by agents which intercalate into DNA, as previously shown for the high mobility group proteins (HMGs). On polytene chromosomes the protein is localized in all bands, with no preference for particular loci. Both the BJ1 protein and in particular the BJ1 mRNA are strongly expressed maternally. In early embryos all nuclei contain equal amounts of BJ1. During neuroblast formation, BJ1 mRNA becomes restricted to cells of the central nervous system, and higher protein levels are found in the nuclei of this tissue. In late embryonic stages, the mRNA almost completely disappears, but significant amounts of BJ1 protein persist until morphogenesis. The BJ1 gene encodes a 547 amino acid polypeptide featuring two different types of internal repeats. The sequence from amino acids 46 to 417 containing seven repeats of the first type has been highly conserved in evolution. 45% of the amino acids in this region are conserved in seven similar tandem repeats of the human gene Regulator of Chromatin Condensation, RCC1. The phenotype of a cell line carrying a mutation of RCC1 suggested a main function for this gene in cell cycle control. A yeast gene, SRM1/PRP20, also contains these repeats and shows 30% amino acid identity to BJ1 in this region. Mutations in this gene perturb mRNA metabolism, disrupt nuclear structure and alter the signal transduction pathway for the mating pheromone. Complementation experiments argue for a common function of these genes in the different species. I propose that their gene products bind to the chromatin to establish or maintain a proper higher order structure as a prerequisite for a regulated gene expression. Disruption of this structure could cause both mis-expression and default repression of genes, which might explain the pleiotropic phenotypes of the mutants.     

The molecular cloning and identification of a gene product specifically required for nuclear movement in Aspergillus nidulans

A temperature-sensitive mutation in the nudC gene (nudC3) of Aspergillus nidulans specifically prevents the microtubule-based movement of nuclei in this organism at the restrictive temperature. The mutation does not affect short term growth, nuclear division, or the movement of other subcellular organelles. Immunofluorescence analysis of cells blocked at the restrictive temperature, using antitubulin antibodies, shows that the inability of nuclei to move under these conditions is not related to an inability of a particular class of microtubule to form. The inability to move nuclei in this mutant is also shown to be independent of both mitosis and the number of nuclei in the cell as a double mutant carrying both nudC3 and a cell cycle-specific mutation blocks with a single immotile nucleus at the restrictive temperature. The molecular cloning of the nudC gene and sequence analysis reveal that it encodes a previously unidentified protein of 22 kd. Affinity-purified antisera reactive to the nudC protein cross reacts to a single protein of 22 kD in Aspergillus protein extracts. This purified sera failed to reveal a subcellular location for the nudC protein at the level of indirect immunofluorescence. The data presented suggest that the 22-kD nudC gene product functions by interacting between microtubules and nuclei and/or is involved in the generation of force used to move nuclei during interphase.     

CENP-E, a novel human centromere-associated protein required for progression from metaphase to anaphase.

We have identified a novel human centromere-associated protein by preparing monoclonal antibodies against a fraction of HeLa chromosome scaffold proteins enriched for centromere/kinetochore components. One monoclonal antibody (mAb177) specifically stains the centromere region of mitotic human chromosomes and binds to a novel, approximately 250-300 kd chromosome scaffold associated protein named CENP-E. In cells progressing through different parts of the cell cycle, the localization of CENP-E differed markedly from that observed for the previously identified centromere proteins CENP-A, CENP-B, CENP-C and CENP-D. In contrast to these antigens, no mAb177 staining is detected during interphase, and staining first appears at the centromere region of chromosomes during prometaphase. This association with chromosomes remains throughout metaphase but is redistributed to the midplate at or just after the onset of anaphase. By telophase, the staining is localized exclusively to the midbody. Microinjection of the mAb177 into metaphase cells blocks or significantly delays progression into anaphase, although the morphology of the spindle and the configuration of the metaphase chromosomes appear normal in these metaphase arrested cells. This demonstrates that CENP-E function is required for the transition from metaphase to anaphase.     

Simultaneous Purification of Murine Mammary Tumor Virus Structural Proteins: Analysis of Antigenic Reactivities of Native gp34 by Radioimmunocompetition Assays

All the structural proteins (gp47, gp34, p27, p23, p16, and p12) of the murine mammary tumor virus (MuMTV) were simultaneously purified utilizing alkylagarose chromatography as the initial fractionation step. Least-hydrophobic MuMTV polypeptides (p23, p16) and the slightly hydrophobic p27 were separated from moderately hydrophobic proteins gp47 and p12 by passage through octylimino (C(8))-agarose; the gp47 and p12 could be removed from the matrix by elution with ethylene glycol, whereas the most hydrophobic MuMTV protein, gp34, was eluted using nonionic detergent together with ethylene glycol. Subsequent purification steps involved ion-exchange or gel filtration chromatography. The resulting protein preparations appeared near-homogeneous on analysis by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. Recoveries of MuMTV proteins, based on their approximate individual contribution to total virus protein, ranged from about 20% for gp47 to greater than 100% for the minor structural component p23, the major phosphoprotein of MuMTV. Antiserum against purified C3H MuMTV gp34, together with purified, radioiodinated gp34, was used to develop a radioimmunoassay which showed that from 13 to 14% of total MuMTV protein by weight is gp34. Using this assay system, the group-specific antigenic reactivity of gp34 was also demonstrated. When solubilized preparations of C3H, RIII, and GR MuMTV's were used as competing antigens in gp34 radioimmunoassays with anti-C3H MuMTV serum, both group- and type-specific differences in antigenic reactivity were found.     

Disruption of CENP antigen function perturbs dynein anchoring to the mitotic kinetochore

Injection of purified autoantibodies against human centromeric proteins into HeLa cells during interphase disrupts the organization of the kinetochore and interferes with chromosomal movements during the subsequent mitosis even though the chromosomes retain the ability to bind microtubules. We have investigated the hypothesis that this phenotype arises from effects on cytoplasmic dynein, the microtubule motor protein. In previous experiments we found that introduction of anticentromere antibodies into cell nuclei during the G₁- or S-phases causes a prometaphase-like arrest, while injections during G₂-phase cause a metaphase arrest. We show here that, in both cases, the level of detectable cytoplasmic dynein at kinetochores is significantly decreased. In contrast, when injected cells were permitted to enter mitosis in the absence of microtubules (conditions where trilaminar kinetochores could be detected by electron microscopy), the intensity of dynein labeling on the kinetochores was identical to that seen in uninjected control cells exposed to colcemid. Therefore, the loss of dynein label on mitotic kinetochores was correlated both with the injection of anticentromere antibodies and with the presence of intact spindle microtubules. We suggest that the injection of anticentromere antibodies somehow weakens the association of dynein with the kinetochore, so that when microtubules are present, these motor molecules are pulled away from the kinetochores as they generate force. This model offers an explanation for the failure of chromosomes of injected cells to move normally in mitosis even though they have attached microtubules.     

Centromere proteins

Human anti-centromere sera from scleroderma patients were used to detect centromere antigens of mouse fibroblast cells. An M_r=59000 centromere protein was localized exclusively on mitotic chromosomes. The association of this protein with the mitotic chromosomes proved to be DNase I sensitive. In interphase nuclei, this centromere antigen was not detectable by immunoblot techniques. The results suggest that the M_r=59000 mitosis specific protein may be necessary for the structural stability of kinetochores during mitosis.     

The 34 kd pp60src substrate is located at the inner face of the plasma membrane

The subcellular localization of the 34 kd protein substrate of the pp60src kinase was investigated by immunofluorescence microscopy and subcellular fractionation. When permeabilized fibroblasts were stained with a monoclonal anti-34 kd protein antibody, a diffuse reticular pattern was observed. The 34 kd protein was not exposed on the outside surface of the cell. Double immunofluorescence staining experiments established that the 34 kd protein distribution was similar to that of the membrane-associated protein alpha-spectrin. The 34 kd protein was found in cell sections to be concentrated along the cell edges. Taken together, these results suggested that the 34 kd pp60src substrate was associated with the inside surface of the plasma membrane. This conclusion was supported by subcellular fractionation experiments in which the 34 kd protein was observed to fractionate with the plasma membrane. These localization studies support further the hypothesis that many of the primary effects of the pp60src kinase occur at the plasma membrane.     

The chemical composition of nuclei and chromosomes isolated by the Langmuir trough technique

The pattern of staining for DNA, histone, and nonhistone protein has been studied in whole cells and in nuclei and chromosomes isolated by surface spreading. In whole interphase cells from bovine kidney tissue culture, nuclear staining for DNA and histones reveals numerous small, intensely stained clumps, surrounded by more diffusely stained material. Nuclei in whole cells stained for nonhistone proteins also contain intensely stained regions surrounded by diffuse stain. These intensely stained regions also stain for RNA, indicating that the regions contain nucleolar material. Electron microscopy of kidney cells confirms that multiple nucleoli are present. Kidney nuclei isolated by surface spreading show an even distribution of stain for DNA, histones, and nonhistone proteins, indicating that the surface forces disperse both condensed chromatin and nucleoli. DNA and protein staining was also studied in metaphase chromosomes from testes of the milkweed bug, Oncopeltus fasciatus. Staining for DNA and histones in metaphase chromosomes is essentially the same in sections of fixed and embedded testes as in preparations isolated by surface spreading. However, striking differences are noted in the distribution of nonhistone proteins. In sections, nonhistone stain is concentrated in extrachromosomal areas; metaphase chromosomes do not stain for nonhistone proteins. Chromosomes isolated by surface spreading, however, stain intensely for nonhistone proteins. This suggests that nonhistone proteins are bound to the chromosomes as a contaminant during the isolation procedure. The relationship of these findings to current work with chromosomes isolated for electron microscopy is discussed.