Escape from gene silencing in ICF syndrome: evidence for advanced replication time as a major determinant

Chromosomal abnormalities associated with hypomethylation of classical satellite regions are characteristic for the ICF immunodeficiency syndrome. We, as well as others, have found that these effects derive from mutations in the DNMT3B DNA methyltransferase gene. Here we examine further the molecular phenotype of ICF cells and report several examples of extensive hypomethylation that are associated with advanced replication time, nuclease hypersensitivity and a variable escape from silencing for genes on the inactive X and Y chromosomes. Our analysis suggests that all genes on the inactive X chromosome may be extremely hypomethylated at their 5' CpG islands. Our studies of G6PD in one ICF female and SYBL1 in another ICF female provide the first examples of abnormal escape from X chromosome inactivation in untransformed human fibroblasts. XIST RNA localization is normal in these cells, arguing against an independent silencing role for this RNA in somatic cells. SYBL1 silencing is also disrupted on the Y chromosome in ICF male cells. Increased chromatin sensitivity to nuclease was found at all hypomethylated promoters examined, including those of silenced genes. The persistence of inactivation in these latter cases appears to depend critically on delayed replication of DNA because escape from silencing was only seen when replication was advanced to an active X-like pattern.

Enrichment for submitotic cell populations using flow cytometry

BACKGROUND

One of the most dramatic events during the course of the mammalian cell cycle is mitosis, when chromosomes condense and segregate, the nuclear envelope breaks down, and the cell divides into two daughter cells. Although cells undergoing mitosis are cytologically distinguishable from nonmitotic cells, few molecular markers are available to specifically identify mitotic cells, especially cells within different stages of mitosis.

METHODS

We applied the flow cytometric method of Juan et al. (Cytometry 32:71-77, 1998) to obtain cells with various levels of the molecular markers cyclin B1 and phosphorylated histone H3; fluorescence microscopy was then used to identify sorted cells in different stages of mitosis.

RESULTS

We observed the substantial enrichment of submitotic cell populations.

CONCLUSIONS

This method represents an effective approach to obtain an enriched population of submitotic cells without the use of drug treatments or prior synchronization.

Very late DNA replication in the human cell cycle

G2 was defined originally as the temporal gap between the termination of DNA replication and the beginning of mitosis. In human cells, the G2 period was estimated to be 3-4 h. However, the absence of replicative DNA synthesis during this period designated G2 has never been shown conclusively. In this report, we show that, at some autosomal and X linked loci, programmed DNA replication continues within 90 min of mitosis. Furthermore, the major accumulation of cyclin B1, a cell-cycle marker that is usually ascribed to G2, overlaps extensively with very late DNA replication. We conclude that the G2 period is much shorter than previously thought and may, in some cells, be nonexistent.

Fragile-X syndrome and myotonic dystrophy: parallels and paradoxes

Fragile-X syndrome and myotonic dystrophy are caused by triplet repeat expansions embedded in CpG islands in the transcribed non-coding regions of the FMR1 and the DMPK genes, respectively. Although initial reports emphasized differences in the mechanisms by which the expanded triplet repeats caused these diseases, results published in the past year highlight remarkable parallels in the likely molecular etiologies. At both loci, expansion is associated with altered chromatin, aberrant methylation, and suppressed expression of the adjacent FMR1 and DMAHP genes, implicating epigenetic mediation of these genetic diseases.

Sequence analysis of long FMR1 arrays in the Japanese population: insights into the generation of long (CGG)n tracts

The human fragile-X syndrome is associated with expansions of a (CGG)n triplet repeat within the FMR1 gene. Whilst normal FMR1 arrays consist of variable numbers of (CGG)7-13 blocks punctuated with single AGG triplets, unstable arrays contain longer blocks of uninterrupted (CGG)n. The degree of instability, and subsequent risk of expansion to the fragile-X mutation, is dependent upon the length of this uninterrupted repeat. Detailed analyses of normal FMR1 array structures suggest that longer uninterrupted blocks of repeat could arise either through a process of gradual slippage or a more dramatic loss of an intervening AGG triplet. Up to 15% of Japanese and Chinese individuals have FMR1 triplet arrays centred on 36 repeats in length, a modal group not found in Caucasians. As longer FMR1 arrays have been associated with high-risk fragile-X haplotypes in some populations, we investigated the nature of these larger arrays. Sequence analysis revealed that the unusual length is due to the presence of a novel (CGG)6 block within the array. Several haplotypically related arrays contain blocks of (CGG)16 or (CGG)15, consistent with the fusion of adjacent (CGG)9 and (CGG)6 blocks after loss of the intervening AGG triplet. This is compatible with inferences from the Caucasian population that AGG loss is a mechanism by which long blocks of identical repeats are generated.

Epigenetic variation illustrated by DNA methylation patterns of the fragile-X gene FMR1

Genomic methylation patterns of mammals can vary among individuals and are subject to dynamic changes during development. In order to gain a better understanding of this variation, we have analyzed patterns of cytosine methylation within a 200 bp region at the CpG island of the human FMR1 gene from leukocyte DNA. FMR1 is normally methylated during inactivation of the X chromosome in females and it is also methylated and inactivated upon expansion of CGG repeats in fragile-X syndrome. Patterns of methylation (epigenotypes) were determined by the sequencing of bisulfite-treated alleles from normal males and females and alleles from a family of five brothers who are methylation mosaics and are affected to various degrees by the fragile-X syndrome. Our data indicate that: (i) methylation of individual CpG cytosines is strikingly variable in hypermethylated epigenotypes obtained from a single individual, suggesting that maintenance of cytosine methylation is a dynamic process; (ii) methylation of non-CpG cytosines in the region studied may occur but is rare; (iii) mosaicism of methylation in the analyzed fragile-X males is remarkably similar to that found for the active X and inactive X alleles in normal females, suggesting that the methylation mosaicism of some fragile-X males reflects similar on and off states of FMR1 expression that exist in normal females; (iv) hypermethylation is slightly more pronounced on fragile-X alleles than on normal inactive X alleles of females; (v) the general dichotomy of hypo- and hypermethylated alleles persisted over the 5 year period that separated samplings of the fragile-X males; (vi) methylation variability was most pronounced at a consensus binding sequence for the alpha-PAL transcription factor, a sequence that may play a role in regulating expression of FMR1.

A variable domain of delayed replication in FRAXA fragile X chromosomes: X inactivation-like spread of late replication

The timing of DNA replication in the Xq27 portion of the human X chromosome was studied in cells derived from normal and fragile X males to further characterize the replication delay on fragile X chromosomes. By examining a number of sequence-tagged sites (STSs) that span several megabases of Xq27, we found this portion of the normal active X chromosome to be composed of two large zones with different replication times in fibroblasts, lymphocytes, and lymphoblastoid cells. The centromere-proximal zone replicates very late in S, whereas the distal zone normally replicates somewhat earlier and contains FMR1, the gene responsible for fragile X syndrome when mutated. Our analysis of the region of delayed replication in fragile X cells indicates that it extends at least 400 kb 5' of FMR1 and appears to merge with the normal zone of very late replication in proximal Xq27. The distal border of delayed replication varies among different fragile X males, thereby defining three replicon-sized domains that can be affected in fragile X syndrome. The distal boundary of the largest region of delayed replication is located between 350 and 600 kb 3' of FMR1. This example of variable spreading of late replication into multiple replicons in fragile X provides a model for the spread of inactivation associated with position-effect variegation or X chromosome inactivation.

Separation of cells at different times within G2 and mitosis by cyclin B1 flow cytometry

Multivariate flow cytometry using specific cyclin proteins and DNA content can identify cell populations at different points within the cell cycle. Quantification of cyclin B1 and DNA content reveals that cells with high levels of cyclin B1 predominantly have a 4C DNA content and are therefore in G2 or mitosis. We have examined whether separation of cells by levels of cyclin B1 could be used to discriminate cells at discrete times within these phases. Post-replicative cells progressively enter into fractions with higher levels of cyclin B1, indicating that this protein can be used as a marker of time in G2. Furthermore, cells in particular phases of mitosis can be greatly enriched by separation based on cyclin B1 levels. This method can thus be used to isolate cells at specific times within G2 and mitosis, periods of the cell cycle that have been difficult to study by cell fractionation.

In vitro expansion of GGC:GCC repeats: identification of the preferred strand of expansion

The human fragile-X syndrome, a major cause of inherited mental retardation, is associated with expansion of the trinucleotide repeat GGC:GCC. Repetitive sequences in DNA are subject to slippage during catalysis by DNA polymerases. We characterized the extent of slippage of synthetic GGC:GCC repeats by various DNA polymerases: Taq DNA polymerase, Klenow fragment of DNA polymerase I, DNA Sequence, DNA polymerase-alpha and polymerase-beta, as well as HIV reverse transcriptase. All of these enzymes were found to expand GGC:GCC repeats, with the most extensive expansion exhibited by Taq DNA polymerase. Starting with a template and primer, each 15 nucleotides (nt) in length, the product of one round of synthesis by Taq polymerase is as long as 250 nt. Sequence analysis of cloned DNA fragments expanded by Taq polymerase indicates that expansion involves multiple triplet additions and that it is asymmetric. The asymmetric distribution of terminal nucleotides in the expanded product is consistent with active expansion of the GCC strand and passive additions onto the GGC strand. The preferential elongation and expansion of the GCC strand was confirmed in studies utilizing longer repeats within a single-stranded M-13 template.

Association of fragile X syndrome with delayed replication of the FMR1 gene

The fragile X syndrome is commonly associated with mutant alleles of the FMR1 gene that are hypermethylated and have large expansions of CGG repeats. We present data here on the replication timing of FMR1 that confirm predictions of delayed replication of alleles from affected males. The normal FMR1 allele replicates late in S phase, while alleles from affected males replicate later, the major peak of replication occurring in the flow cytometry fraction usually referred to as G2/M. The delayed timing of replication is not the direct result of a single replication fork stalling at the expanded CGG repeat, because delayed replication was observed for regions on both sides of the repeat. The domain of altered replication timing includes sites at least 150 kb 5' and 34 kb 3' of the repeat, indicating that genes in addition to FMR1 may be affected.

Methylation analysis of CGG sites in the CpG island of the human FMR1 gene

The fragile-X syndrome of mental retardation is associated with an expansion in the number of CGG repeats present in the FMR1 gene. The repeat region is within sequences characteristic of a CpG island. Methylation of CpG dinucleotides that are 5' to the CGG repeat has been shown to occur on the inactive X chromosome of normal females and on the X chromosome of affected fragile-X males, and is correlated with silencing of the FMR1 gene. The methylation status of CpG sites 3' to the repeat and within the repeat itself has not previously been reported. We have used two methylation-sensitive restriction enzymes, AciI and Fnu4HI, to further characterize the methylation pattern of the FMR1 CpG island in normal individuals and in those carrying fragile-X mutations. Our results indicate that: (i) CpG dinucleotides on the 3' side of the CGG repeat are part of the CpG island that is methylated during inactivation of a normal X chromosome in females; (ii) the CGG repeats are also part of the CpG island and are extensively methylated as a result of normal X-chromosome inactivation; (iii) similar to normal males, unaffected fragile-X males with small CGG expansions are unmethylated in the CpG island; for affected males, the patterns of methylation are similar to those of a normal, inactive X chromosome; (iv) in contrast to the partial methylation observed for certain sites in lymphocyte DNA, complete methylation was observed in DNA from cell lines containing either a normal inactive X chromosome or a fragile-X chromosome from an affected male.(ABSTRACT TRUNCATED AT 250 WORDS)

Epigene conversion: a proposal with implications for gene mapping in humans

Epigenetic modification of DNA is now recognized as a potentially important factor in the inheritance and expression of some mutations; its ability to complicate human genetic analysis is concurrently becoming apparent. One unusual form of epigenetic modification, dominant position-effect variegation (PEV), has been used as a model for Huntington disease. In dominant PEV, a fully dominant mutant phenotype results from stable epigenetic inactivation of an allele adjacent to the structural alteration (cis-inactivation) combined with a complementary inactivation of the homologous normal allele (trans-inactivation). We now propose that trans-inactivation of the normal allele may occasionally persist through meiosis. Such "epigene conversion" occurring at the Huntington disease locus in a few percent of meioses would largely account for the published anomalies in that region's genetic map. This concept could also explain anomalous linkage map data for other disease-causing alleles in humans.

Estimating the stability of the proposed imprinted state of the fragile-X mutation when transmitted by females

Fragile-X syndrome is a major cause of mental retardation in humans. The X-inactivation imprinting model accounts for the unusual pattern of inheritance and expression of this syndrome. According to this model, the fragile-X mutation creates a local block to the attempted reactivation of the mutant X chromosome prior to oogenesis. This local block results in an "imprinted" fragile-X chromosome that is deleterious in males and in females for whom this chromosome is predominantly the active X chromosome. The imprinted state of the fragile-X mutation is inferred to be stable when transmitted by an imprinted female because the penetrance of the syndrome in sons of affected females is estimated to be 1.0. To provide a more precise estimate of the stability of the proposed fragile-X imprint, we have analyzed published pedigrees that include restriction fragment length polymorphism and cytogenetic data from sibships with mothers who are interpreted as having an imprinted fragile-X allele. We conclude that the fragile-X imprint was stable in 46 out of 48 female meioses. This analysis leads to a preliminary estimate of about 96% for the stability of the imprint through female meiosis. Two imprinted females had progeny who appeared to be carriers of a nonimprinted fragile-X allele. If this interpretation is correct, then reversion from the imprinted to the nonimprinted state, or "erasure," can occasionally occur when the mutant fragile-X allele is transmitted by an imprinted female. We discuss the genetic and epigenetic significance of possible female erasure. We request DNA and cytogenetic information from unpublished pedigrees to quantify further the stability, during female meiosis, of the proposed imprinted state of the mutant fragile-X allele.

Hormonal control of gene expression: interactions between two trans-acting regulators in Drosophila

The steroid hormone 20-hydroxyecdysone (20HE) and the Broad-Complex locus (BRC) are involved in regulating developmental changes in gene expression around the time of metamorphosis in Drosophila. We have investigated the regulatory interactions between 20HE, BRC, and a set of genes expressed in the fat body of third-instar Drosophila larvae. RNA levels for two hormone-inducible genes, Larval Serum Protein-2 and P1, accumulate to normal levels in BRC-mutant larvae. In contrast, RNA levels for the P6 gene were affected by mutations at BRC. On the basis of the results of experiments in which hormone concentrations were varied in BRC-mutant or wild-type larvae, we conclude that 20HE can both increase and decrease P6 RNA levels in the absence of BRC product(s). BRC appears to be a trans-acting modulator of the response of P6 to the hormone. We propose that BRC attenuates the repressive effect of the hormone, expanding the range of hormone concentrations that induce the gene, thus allowing P6 RNA to reach high levels during the third larval instar. The results are discussed in relation to other genes that are regulated by the same two trans-acting factors. A model is presented that refines the model of Ashburner et al. (1974, Cold Spring Harbor Symp. Quant. Biol. 38: 655-662) for the hormonal regulation of gene activity.

Possible erasure of the imprint on a fragile X chromosome when transmitted by a male

Although most males with the fragile-X [fra(X)] syndrome do not reproduce, there are 2 published pedigrees that include affected males who have daughters and who thus appear to have transmitted the fragile-X chromosome to their progeny. In addition, one published fra(X) pedigree includes an apparently normal male who expresses cytogenetically the fra(X) site at high frequency and who has 3 daughters. In the 6 daughters of these 3 males, there is little or no cytogenetic expression of the fra(X). I interpret these pedigrees within the context of my X-inactivation imprinting model of the fra(X) syndrome (Genetics 117:587-599): the cytogenetic manifestation of the imprinted state of the mutant fra(X) chromosome [high percentage of cytogenetic expression] is no longer present in daughters of imprinted males. I propose that the imprinted state is erased when an imprinted fragile-X chromosome is passed through a male. Such erasure in the gender opposite to the gender that established the imprint is in accord with other examples of chromosome imprinting in mammals. Additional data from unpublished fra(X) pedigrees are requested.

Proposed genetic basis of Huntington's disease

I propose that Huntington's disease (HD) is caused by dominant position-effect variegation, a phenomenon for which new information is available in Drosophila melanogaster. The essential features of this proposal are that (1) the HD mutation is the result of a chromosome alteration that inactivates transcription of a nearby structural gene or genes (cis-inactivation); the combination of this proposed chromosome alteration and the structural gene(s) is termed the HD allele; (2) there is pairing in some somatic cells between the HD and HD+ alleles on homologous chromosomes; (3) as a result of this somatic pairing, the HD mutation also inactivates transcription of the HD+ structural gene on the normal homologue (trans-inactivation), resulting in complete dominance of the mutation; (4) polymorphism for an X-linked recessive modifier of position-effect variegation means that the age of onset of symptoms of HD will depend on which parent the HD mutation was inherited from. The fully dominant nature of HD and the parental-source effect on the age of onset are thus both understandable within the genetic and epigenetic paradigm of position-effect variegation.

Two progenitor cells for human oogonia inferred from pedigree data and the X-inactivation imprinting model of the fragile-X syndrome

Laird has proposed that the human fragile-X syndrome is caused by abnormal chromosome imprinting. The analysis presented here supports and extends this proposal. Using published pedigrees that include DNA polymorphism (RFLP) data, we establish that the states of the fragile-X mutation termed "imprinted" and "nonimprinted" usually can be distinguished by the level of cytogenetic expression of the fragile-X chromosome. This information is then used to assess the state of the fragile-X allele in carrier progeny of individual women who inherited a nonimprinted fragile-X chromosome. From this assessment, an estimate is made of the frequency, in individual women, of primary oocytes with an imprinted fragile-X chromosome. The results of this analysis provide additional support for the specific model in which chromosome imprinting occurs in a female in, on average, half of her primary oocytes. This is the expected frequency if X-chromosome inactivation is the initial step in the imprinting of the mutant fragile-X allele. Moreover, this analysis suggests a biological explanation for peculiarities of fragile-X inheritance described by others as "clustering" and the "Sherman paradox." We interpret these peculiarities as consequences of a very small number of oogonial progenitor cells. Two progenitor cells for oogonia is the best integer estimate of the number of such cells at the time of the initial event that leads to chromosome imprinting.

Population genetic consequences of the fragile-X syndrome, based on the X-inactivation imprinting model

We have examined the population genetic consequences of the model of Laird (Genetics 117:587-599, 1987) in which the fragile-X syndrome is caused by "imprinting" of a mutant chromosome. The imprinting event in this model results from a block to reactivation of an inactive X chromosome prior to oogenesis. If it is assumed that males carrying the imprinted chromosome never reproduce, the frequencies of females and males carrying the imprinted chromosome are expected to be equal. When a mutation-selection balance is established, there are expected to be somewhat more than twice as many females carrying the nonimprinted fragile X as carry the imprinted fragile-X chromosome, the excess depending on the fertility of fragile-X females. Nonpenetrant (transmitting) males, i.e., those with the nonimprinted fragile-X chromosome, are expected to be present at about the same frequency as are males with the syndrome. More than one-third of the nonimprinted chromosomes in the population are expected to be newly arisen in each generation. We have considered possible alternatives to the model of a mutation-selection balance. Nonimprinted carrier females would need to have 100% fertility excess to avoid postulating a high mutation rate to account for the very high prevalence of the syndrome.

A Drosophila rRNA gene located in euchromatin is active in transcription and nucleolus formation

P-element transformants of a single rRNA gene (rDNA) were used to investigate the relationship between the organization of the nucleolus organizer (NO) and rDNA function in Drosophila melanogaster. In situ hybridization to rRNA in polytene nuclei of salivary glands demonstrated that an rRNA gene can be transcribed at a high rate when inserted into chromosomal sites other than the NO. Structures that resemble morphologically the endogenous nucleoli ('mininucleoli') were associated with four different euchromatic sites of rDNA insertion. Molecular analyses revealed that these mininucleoli contained both rRNA and an antigen specific to nucleoli. Phenotypes resulting from rDNA deficiencies were rescued partially by the presence of the transformed rDNA, indicating that the transcripts and mininucleoli associated with the rDNA insertion sites were functional. Thus, two conserved features of rDNA organization in eukaryotes, namely tandem repetition and heterochromatic localization, are not required for rRNA gene function. We conclude that 'nucleolar organizing activity' is an intrinsic property of the rDNA or its RNA products.

Intercalary heterochromatin of Drosophila as a potential model for human fragile sites

We summarize our proposal that "intercalary heterochromatin" of Drosophila is a useful model for human fragile sites. Comparison with Drosophila site 11A suggests that the normal allele of fragile site Xq27 is a meiotic pairing site.

Fragile-X mutation proposed to block complete reactivation in females of an inactive X chromosome

I discuss two aspects of my proposal that fra(X) chromosomes exist in two states, imprinted and non-imprinted: why do males not imprint the fra(X); does the "Sherman paradox" rule out my proposal?

Proposed mechanism of inheritance and expression of the human fragile-X syndrome of mental retardation

A mechanism is proposed for the inheritance and expression of the fragile-X-linked syndrome of mental retardation in humans. Two independent events are required for expression of the syndrome: the fragile-X mutation, and X chromosome inactivation in pre-oogonial cells. The fragile-X mutation at site Xq27 has little or no effect until the chromosome is inactivated in a female as part of the process of dosage compensation. At a stage where the inactivated X chromosome would normally be reactivated in preparation for oogenesis, the mutation results in a local block to the reactivation process. This block to reactivation leads to mental retardation in progeny by reducing the level of products from the unreactivated Xq27 region in male cells, and, for a heterozygous female, in somatic cells in which the normal X chromosome has been inactivated. Published data relevant to this proposed mechanism are discussed.

Three euchromatic DNA sequences under-replicated in polytene chromosomes of Drosophila are localized in constrictions and ectopic fibers

We examined three regions of under-represented euchromatic DNA sequences (histone, Ubx, and 11 A), for their possible correlation with euchromatic constrictions in polytene chromosomes of Drosophila melanogaster. Cloned sequences were hybridized to filters and to chromosomes prepared for light microscopy. Under-represented sequences hybridized to DNA within constrictions and in ectopic fibers. In contrast, adjacent sequences that were fully endoreplicated in the Ubx and 11 A regions in polytene cells hybridized to sites just adjacent to their respective constrictions. For one region (Ubx), sequences under-represented in salivary gland cells were fully endoreplicated in fat body cells. For this particular region, the morphology of the polytene chromosomes differs between these two cell types in that the specific constriction is absent at this region in fat body polytene chromosomes, thus strengthening the correlation between under-representation and chromosome constrictions. Although all three sequences are in regions that have been classified by others as "intercalary heterochromatin," we detect no common functional or sequence organizational feature for these examples of under-represented DNA. We suggest that the lower efficiencies of the replication origins, or special regions of termination at these sites, are the primary cause of the under-replication, and that this under-replication is sufficient to confer the properties of intercalary heterochromatin.

Chromosome structure and DNA replication in nurse and follicle cells of Drosophila melanogaster

In the nurse cells of Drosophila, nuclear DNA is replicated many times without nuclear division. Nurse cells differ from salivary gland cells, another type of endoreplicated Drosophila cell, in that banded polytene chromosomes are not seen in large nurse cells. Cytophotometry of Feulgen stained nurse cell nuclei that have also been labeled with 3H-thymidine shows that the DNA contents between S-phases are not doublings of the diploid value. In situ hybridization of cloned probes for 28S + 18S ribosomal RNA, 5S RNA, and histone genes, and for satellite, copia, and telomere sequences shows that satellite and histone sequences replicate only partially during nurse cell growth, while 5S sequences fully replicate. However, during the last nurse cell endoreplication cycle, all sequences including the previously under-replicated satellite sequences replicate fully. In situ hybridization experiments also demonstrate that the loci for the multiple copies of histone and 5S RNA genes are clustered into a small number of sites. In contrast, 28S + 18S rRNA genes are dispersed. We discuss the implications of the observed distribution of sequences within nurse cell nuclei for interphase nuclear organization. In the ovarian follicle cells, which undergo only two or three endoreplication cycles, satellite, histone and ribosomal DNA sequences are also found by in situ hybridization to be underrepresented; satellite sequences may not replicate beyond their level in 2C cells. Hence the pathways of endoreplication in three cell types, salivary gland, nurse, and follicle cells, share basic features of DNA replication, and differ primarily in the extent of association of the duplicated chromatids.

Control of DNA replication and spatial distribution of defined DNA sequences in salivary gland cells of Drosophila melanogaster

In dividing cells, each sequence replicates exactly once in each S-phase, but in cells with polytene chromosomes, some sequences may replicate more than once or fail to replicate during S-phase. Because of this differential replication, the control of replication in polytene cells must have some unusual features. Dennhöfer (1982a) has recently concluded that the total DNA content of the polytene cells of Drosophila salivary glands exactly doubles in each S-phase. This observation, along with previous studies demonstrating satellite underreplication in salivary gland cells, led us to consider the hypothesis that there is a "doubling of DNA" mechanism for the control of DNA replication in polytene cells. With this mechanism, a doubling of DNA content, rather than the replication of each sequence, would signal the end of a cycle of DNA replication. To test this hypothesis, we have reinvestigated the replication of several sequences (satellite, ribosomal, histone and telomere) in salivary gland cells using quantitative in situ hybridization. We find that underreplication of some sequences does occur. In addition we have repeated Dennhöfer's cytophotometric and labeling studies. In contrast to Dennhöfer, we find that the total DNA contents of nonreplicating nuclei do reflect this partial replication, in accord with Rudkin's (1969) result. We conclude that DNA replication in polytene cells is controlled by modifications of the mechanism operating in dividing cells, where control is sequence autonomous, and not by a "doubling of DNA" mechanism. In situ hybridization to unbroken salivary gland nuclei reveals the distribution of specific sequences. As expected, satellite, histone and 5S sequences are usually in a single cluster.(ABSTRACT TRUNCATED AT 250 WORDS)

Morphological analyses of active genes and chromatin

Chromatin with nascent ribonuclear protein (RNP) fibers representing transcription of ribosomal DNA (top) and nonribosomal DNA (bottom). These two types of transcription can be distinguished on the basis of the size of transcription units, chromatin morphology and the inferred ratio of DNA packing, the frequency of RNP fibers (number of fibers per micrometer of chromatin), and the solitary v tandem repeat occurrence of fiber arrays. In some cases, the structure of RNP fibers is also distinctive for different transcription units. The micrographs presented herein, from Laird and Chooi, were taken of chromatin samples prepared similarly to the method described by Miller and Bakken. Cells and nuclei from Drosophila were ruptured in low ionic strength buffer; nuclear and cytoplasmic constituents were recovered on electron microscopic grids under conditions that facilitated spreading of chromatin and nascent ribonuclear protein fibers. Bar represents 1 micron.

Structural paradox of polytene chromosomes

The observation of thick chromatin fibers in interbands of Dipteran polytene chromosomes suggests that there should be 5 to 10 times more mass and DNA in interbands than is commonly thought to be present. To resolve this paradox, the chromatin content of interbands was estimated, using whole-mounted polytene chromosomes from Drosophila melanogster. Densitometry of high voltage electron microscopic negatives provides an estimate of less than 4:1 for the average ratio of cross-sectional dry mass (or mass per unit chromosome length) of bands relative to interbands. This ratio, combined with an estimate of the length of chromosome composed of interbands, indicates that at least 26% of chromosome mass is contributed by interband chromatin. Since DNA comprises a similar proportion of chromatin mass in bands and interbands (Laird et al., 1980b), these data imply that DNA sequences in interbands represent at least 26% of the euchromatic genome of D. melanogaster. This result calls for reinterpretation of some of the genetic and molecular data from Diptera. The discrepancy between this higher estimate of interband mass and DNA, and previous estimates of 3-5%, is discussed. One possibility is that previous measurements were made on prominent interbands, which are proposed here to be in regions that are delayed in DNA replication. Such interbands would be reduced in polyteny and DNA content compared with the average interband region. The concept of local variations in polyteny is also used here to explain major differences in the cross-sectional mass of bands. This leads to a revised model of polytene chromosomes in which at least three levels of polyteny, rather than one or two levels, can be present within one euchromatic region.

Morphology of transcription units in Drosophila melanogaster

We have used an electron microscopic analysis to define and to characterize active transcription units of Drosophila melanogaster. The lengths and spacings of nascent ribonuclear protein (RNP) fibers were determined on embryonic chromatin that was spread using techniques introduced by Miller and Beatty (1969). The data are consistent with the occurrence of specific sites of transcription initiation and termination. We apply the term transcription unit (TU) to a chromatin region bounded by these control sites. Two classes of TUs are active in Drosophila melanogaster embryonic cells--those synthesizing ribosomal RNA and those synthesizing non-ribosomal RNA. The classes can usually be distinguished on the basis of TU size, chromatin morphology and inferred DNA packing ratio, frequency of RNP fibers (number of fibers per mum of chromatin), and the solitary vs. tandem repeat occurrence of fiber arrays. The results indicate that non-ribosomal transcription units have lengths in accord with the expectation that DNA of each chromomere is transcribed as a unit. Some nascent fiber arrays in D. melanogaster have more complex patterns of RNP fiber lengths. We suggest that these are a consequence of cleavage of RNP fibers at specific sites during transcription. These sites of transcriptional control and the amounts of DNA between them provide a basis for further relating units of transcription to units of gene function.

Analysis of chromatin-associated fiber arrays

Electron microscopic examination of chromatin from embryonic nuclei of Oncopeltus fasciatus and Drosophila melanogaster reveals arrays of chromatin associated fibers. The lengths and spacings of these fibers were analyzed to provide a basis for defining and interpreting regions of transcriptionally active chromatin. The results of the analysis are consistent with the interpretation of some fibers as nascent RNA with associated protein (RNP). The chromatin segments underlying these fiber arrays were classified as ribosomal or non-ribosomal transcription units according to definitions and criteria described by Foe et al. (1976). Nascent fibers on active ribosomal transcription units were analyzed and compared for Drosophila melanogaster, Triturus viridescens, and Oncopeltus fasciatus. A common feature of the fiber patterns on ribosomal TUs is that origin-distal fibers exhibit greater length variability and a lower slope relative to proximal fibers. The region of increased variability in fiber lengths is correlated with the expected location of 28S ribosomal RNA sequences in the distal half of each ribosomal transcription unit. Because 28S ribosomal RNA appears to contain more extensive regions of base sequence complementarity, we suggest that the length of ribosomal RNP fibers is influenced under our spreading conditions by the secondary structure of the nascent RNA. In order to calculate the RNA content of RNP fibers, chromatin morphology was used to estimate lengths of transcribed DNA. The packing ratio of DNA in chromatin, which we express as the length of B-structure DNA divided by length of chromatin, is 1.1-1.2 and 1.6 for the DNA in active ribosomal and non-ribosomal chromatins, respectively. These DNA packing ratios are used to determine the extent to which nascent RNP fibers are shorter than the transcribed DNA (expressed as DNA/RNP length ratio). For non-ribosomal transcription units and for proximal fibers of ribosomal transcription units. DNA/RNP length ratios are relatively constant within each array. However, considerable variability in this ratio (4-23) is observed for different arrays of fibers. Possible sources of this variability are considered by comparing ratios derived from the presumably identical ribosomal transcription units. Further analysis of the morphology of nascent fibers may elucidate the contributions of proteins and successive RNA sequences to RNP structure.

Comparative organization of active transcription units in Oncopeltus fasciatus

We have analyzed electron micrographs of chromatin-associated fiber arrays from embryos of the milkweed bug, Oncopeltus fasciatus. The analysis has revealed that the arrays have highly ordered patterns of fiber spacings and lengths. These patterns support the interpretation that the fibers are nascent RNA with associated proteins (RNP fibers) which have resulted from transcription of the DNA in the underlying chromatin segment. In particular, the patterns indicate that the chromatin underlying each array is delimited by specific sites for initiation and termination of transcription. We apply the term transcription unit to a chromatin segment thus bounded. The analysis has further revealed that transcription units can be grouped into two principal classes--ribosomal and nonribosomal. Active transcription units of these two classes differ in DNA content, in their proximity to other active transcription units, and in their chromatin morphology. For certain developmental stages, fiber frequencies (that is, the nubmers of fibers per mum of chromatin) are also useful in distinguishing ribosomal from nonribosomal arrays. The most definitive of the above classification criteria is chromatin morphology as observed under our preparative conditions. We propose that term rho chromatin for the unbeaded or smooth chromatin that underlies nascent ribosomal RNP fibers. DNA in rho chromatin has a calculated packing ratio of approximately 1.2 mum of B structure DNA per mum of chromatin. Nu chromatin is used to designate the beaded chromatin for which we calculate a DNA packing ratio of 1.6-2.3 in our preparations. This calculation for nu chromatin is based on the inference that the beads are nucleosomes (nu bodies, PS particles, unit particles). The beaded morphology is observed between fibers of nonribosomal transcription unit as well as for most fiber-free chromatin. The detection of specific sites of transcriptional initiation and termination and the classification of transcription units can provide a basis for further analysis of transcriptional control.

The size of poly(A)-containing RNAs in Drosophila melanogaster embryos

The size range of poly(A)-containing RNA from Drosophila melanogaster embryos has been estimated by hybridization with 3H-labeled poly(U) and subsequent fractionation on sucrose gradients. The median size of nuclear poly(A)-containing RNA is about 30 S (6000 nucleotides), and the median size of cytoplasmic poly(A)-containing RNA is about 17 S (1800 nucleotides). The relationship of these sizes to messenger RNA needed to code for protein and to the length of DNA contained in a chromomere is discussed.

Transcription of nonrepeated DNA in mouse brain

Under normal conditions of DNA renaturation, about 60 percent of mouse DNA fragments renature at a rate consistent with their being present only once per sperm. These nonrepeated sequences (also called single-copy or unique) may be used in RNA-DNA hybridization experiments to provide quantitative estimates of RNA diversity. About 10 percent of the mouse single-copy sequences are transcribed in mouse brain tissue. Estimates of about 3 percent were obtained for mouse liver and kidney RNA's. If only one of the complementary DNA strands is transcribed, this hybridization value implies that the equivalent of at least 300,000 different sequences of 1000 nucleotides are expressed in mouse brain tissue. It is suggested that the large amount of DNA in mammals is functionally important, and that a substantial proportion of the genome is expressed in the brain.