RADHA--a new male germ line-specific chromosomal protein of Drosophila.

A new chromosomal protein - RADHA - of Drosophila is described that is specific for the male germ line. It is encoded by a single-copy gene, located in the region 96C-D of D. melanogaster polytene chromosomes. Transcription of the radha gene is restricted to the primary spermatocyte stage. The protein initially accumulates in some of the Y-chromosomal lampbrush loops. After meiosis it is found in the nuclei of spermatids and might be involved in chromatin rearrangement processes in the male germ line. RADHA is a basic protein with a C-terminal leucine zipper region and several segments capable of forming coiled-coil structures.

Heterochromatin.

The properties of heterochromatin are reconsidered in the context of our present understanding of gene silencing, telomeric and centromeric properties, position-effect variegation and X-chromosome inactivation. It is proposed that the chromatin in heterochromatic chromosomal regions is generally similar in its molecular composition to that in silenced chromosomal regions. Heterochromatic appearance hence reflects not a particular quality of the respective chromosomal regions but only a specific kind of chromatin packaging comparable to that required for the inactivation of genes. This packaging may be initiated by particular signals in the DNA but can be propagated over more extended chromosomal regions by the formation of multiprotein complexes that interact with histones and possibly cell-specific additional components (RNA or proteins) that determine the status of the chromosome in a particular cell type.

Drosophila melanogaster histone H2B retropseudogene is inserted into a region rich in transposable elements.

We have isolated and characterized the genomic sequence of a Drosophila melanogaster histone H2B pseudogene that is localized outside of the cluster of the replication-dependent histone genes and has all the properties of a retropseudogene. It is highly homologous to the transcribed region of the D. melanogaster histone H2B gene, but not to its flanking regions, and is surrounded by short direct repeats. The pseudogene contains several point mutations that preclude its translation. The sequence of the 3' region of this pseudogene is compatible with the hypothesis that the 3' terminal stem-loop structure of the histone H2B mRNA has served as a primer for the reverse transcription event from which this pseudogene originated. Analysis of the regions flanking the histone H2B pseudogene revealed the presence of three different types of transposable elements, suggesting that this chromosomal locus represents a hotspot for transposition.

Naturally occurring testis-specific histone H3 antisense transcripts in Drosophila.

While analysing the transcription of the cluster of cell-cycle regulated histone genes in Drosophila hydei, we have found transcripts spanned both histone H3 and H4 genes and were antisense for histone H3. As the two histone genes are in opposite orientation, these transcripts contained the sense strand of the histone H4 gene. Such transcripts were present in both poly(A)+ and poly(A)- RNA fractions. The polyadenylated molecules contained a poly(A) tail at the 3' end of the stem-loop structure, which is characteristic for cell-cycle regulated histone mRNAs. The antisense RNA of histone H3 is synthesized exclusively in testes. By developing an improved protocol of in situ hybridization to Drosophila testis squashes, we could demonstrate that the antisense transcripts are localized in the nuclei of primary spermatocytes. Possible functions of this RNA are discussed.

Two types of polyadenated mRNAs are synthesized from Drosophila replication-dependent histone genes.

The polyadenylation of replication-dependent histone H2B, H3 and H4 mRNAs in Drosophila melanogaster was analysed. Two types of mRNAs, containing a poly(A) tail, can be detected in addition to non-polyadenylated messengers, which represent the majority of replication-dependent histone mRNAs. Firstly, conventional polyadenylation signals, localized downstream from the stem-loop region, are used to produce polyadenylated mRNAs. The messengers of this type, generated from the D. melanogaster H2B gene, are preferentially synthesized in the testis of the fly. Secondly, a distinct type of polyadenylated histone mRNA has been identified. This mRNA, which is present in many different tissues and constitutes a minor part of the total histone mRNA pool, contains a short poly(A) tail, added to the end of the 3' terminal stem-loop structure, which is in most cases lacking several nucleotides from its 3' end. The sites of polyadenylation within the stem-loop are not preceded by a normal polyadenylation signal. The possible functions of the polyadenylated histone transcripts are discussed.

The localization of histone H3.3 in germ line chromatin of Drosophila males as established with a histone H3.3-specific antiserum.

A rabbit antiserum, specific for the histone H3.3 replacement variant, was raised with the aid of a histone H3.3-specific peptide. Immuno blot experiments demonstrated the specificity of this polyclonal antiserum. In addition, we showed on immuno blots that two monoclonal antibodies isolated from mice with systemic lupus erythematosus (SLE) display strong reactivity with the H3.3 histone, but not with its replication-dependent counterparts. Our observations indicate that histone H3.3 might play a role as autoantigen in SLE. We used the histone H3.3-specific antiserum to characterize the germ line chromatin in cytological preparations of Drosophila testes, because our previous studies had shown that a histone H3.3-encoding gene is strongly expressed in the germ line of Drosophila males. The antiserum reacted with some of the lampbrush loops in spermatocytes and with chromatin of the postmeiotic germ cells of males. Our data indicate that histone H3.3 is not evenly distributed throughout the chromatin of germ cells, but is concentrated in distinct regions. Histone H3.3 disappears from the spermatid nuclei, along with the other core histones, during the late stages of spermatogenesis. In Drosophila polytene chromosomes, however, a rather uniform distribution of the histone H3.3 was observed. The possible role of histone H3.3 is discussed.

Identification and characterization of the Drosophila histone H4 replacement gene.

Replacement variant genes for different histones have been described in most higher eukaryotes. However, so far no such gene has been found for histone H4. We have isolated from both Drosophila melanogaster and D. hydei a novel histone H4 encoding gene, H4r, which displays all the properties of a histone replacement variant gene: it contains introns, generates polyadenylated mRNA, represents the predominant H4 transcript in non-dividing tissues and is present in the genome as a single copy. The encoded polypeptide is identical to the Drosophila cell-cycle regulated histone H4. The fact that it is a single copy gene makes it prone to genetic analysis and it might be a useful tool for studying nucleosome structure and function.

Transcription of gypsy elements in a Y-chromosome male fertility gene of Drosophila hydei.

We have found that defective gypsy retrotransposons are a major constituent of the lampbrush loop pair Nooses in the short arm of the Y chromosome of Drosophila hydei. The loop pair is formed by male fertility gene Q during the primary spermatocyte stage of spermatogenesis, each loop being a single transcription unit with an estimated length of 260 kb. Using fluorescent in situ hybridization, we show that throughout the loop transcripts gypsy elements are interspersed with blocks of a tandemly repetitive Y-specific DNA sequence, ay1. Nooses transcripts containing both sequence types show a wide size range on Northern blots, do not migrate to the cytoplasm, and are degraded just before the first meiotic division. Only one strand of ay1 and only the coding strand of gypsy can be detected in the loop transcripts. However, as cloned genomic DNA fragments also display opposite orientations of ay1 and gypsy, such DNA sections cannot be part of the Nooses. Hence, they are most likely derived from the flanking heterochromatin. The direction of transcription of ay1 and gypsy thus appears to be of a functional significance.

Spermatogenesis in Drosophila.

A short summary on the present knowledge on spermatogenesis in Drosophila is given which also points out particular questions of interest in the context of this morphogenetic process. Such points of interest are the formation of lampbrush loops in primary spermatocytes, the chromosomal events during meiosis, the occurrence of chromatin rearrangements and the regulation of gene activities at the posttranscriptional level. The activities and some major conclusions from my laboratory are subsequently described. They include studies of the expression of histone variants, the structure and function of lampbrush loops and the expression of genes participating in sperm morphogenesis.

Tracking heterochromatin.

The Second International Workshop on Drosophila Heterochromatin, held in Honolulu from January 4-7, 1995, brought together about 70 scientists from the US, Canada, Germany, Italy, Russia, and the Netherlands. After the first of these international meetings, five years ago, Mary Lou Pardue and Wolfgang Hennig, in these columns, commented on its proceedings, and on heterochromatin in general. Although the questions that they raised cannot yet be answered exhaustively, important and sometimes surprising new observations have been made, some previously tentative answers have been firmed up, and some theoretical views underwent significant shifts. We wish to reflect here a few of the data presented at the second workshop, and express some thoughts suggested to us by these recent findings.

Structure and expression of histone H3.3 genes in Drosophila melanogaster and Drosophila hydei.

We demonstrate that in Drosophila melanogaster the histone H3.3 replacement variant is encoded by two genes, H3.3A and H3.3B. We have isolated cDNA clones for H3.3A and cDNA and genomic clones for H3.3B. The genes encode exactly the same protein but are widely divergent in their untranslated regions (UTR). Both genes are expressed in embryos and adults; they are expressed in the gonads as well as in somatic tissues of the flies. However, only one of them, H3.3A, shows strong testes expression. The 3' UTR of the H3.3A gene is relatively short (approximately 250 nucleotides (nt)). H3.3B transcripts can be processed at several polyadenylation sites, the longest with a 3' UTR of more than 1500 nt. The 3' processing sites, preferentially used in the gonads and somatic tissues, are different. We have also isolated the Drosophila hydei homologues of the two H3.3 genes. They are quite similar to the D. melanogaster genes in their expression patterns. However, in contrast to their vertebrate counterparts, which are highly conserved in their noncoding regions, the Drosophila genes display only limited sequence similarity in these regions.

Minor-myosin, a novel myosin isoform synthesized preferentially in Drosophila testis is encoded by the muscle myosin heavy chain gene.

Searching for structural proteins involved in spermatogenesis of Drosophila, we found a novel myosin isoform in the testis of Drosophila hydei and D. melanogaster. The transcript encoding this isoform, which we called 'minor-myosin', initiates within the intron between exons 12 and 13 of the muscle myosin heavy chain (mMHC) gene. Minor-myosin contains a common myosin tail but no ordinary myosin head domain. Instead, it has a short N-terminal domain which displays similarity with the N-termini of certain myosin light chain proteins. Western blots with male germ line mutants showed that the novel mMHC isoform is synthesized in the male germ cells, mainly postmeiotically. However, minor-myosin is not testis-specific, as it is expressed at a low level in the fly carcasses. The possible functions of the myosin isoform in the male germ line are discussed.

Degenerating gypsy retrotransposons in a male fertility gene on the Y chromosome of Drosophila hydei.

During the evolution of the Y chromosome of Drosophila hydei, retrotransposons became incorporated into the lampbrush loop pairs formed by several of the male fertility genes on this chromosome. Although insertions of retrotransposons are involved in many spontaneous mutations, they do not affect the functions of these genes. We have sequenced gypsy elements that are expressed as constituents of male fertility gene Q in the lampbrush loop pair Nooses. We find that these gypsy elements are all truncated and specifically lost those sequences that may interfere with the continuity of lampbrush loop transcription. Only defective coding regions are found within the loop. Gypsy is not transcribed in loops of many other Drosophila species harboring the family. These results suggest that any contribution of gypsy to the function of male fertility gene Q does not depend on a conserved DNA sequence.

Interspecific sequence comparison of the muscle-myosin heavy-chain genes from Drosophila hydei and Drosophila melanogaster.

The muscle-myosin heavy-chain (mMHC) gene of Drosophila hydei has been sequenced completely (size 23.3 kb). The sequence comparison with the D. melanogaster mMHC gene revealed that the exon-intron pattern is identical. The protein coding regions show a high degree of conservation (97%). The alternatively spliced exons (3a-b, 7a-d, 9a-c, 11a-e, and 15a-b) display more variations in the number of nonsynonymous and synonymous substitutions than the common exons (2, 4, 5, 6, 8, 10, 12, 13, 14, 16, 17, and 19). The base composition at synonymous sites of fourfold degenerate codons (third position) is not biased in the alternative exons. In the common exons there exists a bias for C and against A. These findings imply that the alternative exons of the Drosophila mMHC gene evolve at a different, in several cases higher, rate than the common ones. The 5' splice junctions and 5' and 3' untranslated regions show a high level of similarity, indicating a functional constraint on these sequences. The intron regions vary considerably in length within one species, but the corresponding introns are very similar in length between the two species and all contain stretches of sequence similarity. A particular example is the first intron, which contains multiple regions of similarity. In the conserved regions of intron 12 (head-tail border) sequences were found which have the potential to direct another smaller mMHC transcript.

Transcription of repetitive DNA sequences in the lampbrush loop pair Nooses formed by sterile alleles of fertility gene Q on the Y chromosome of Drosophila hydei.

The Y chromosomal lampbrush loop-forming male fertility genes of Drosophila consist mainly of repetitive DNA sequences that do not code for proteins. We investigated whether differences in the transcription of these sequences can be detected in male-sterile alleles of male fertility gene Q, which forms the loop pair Nooses. The loop consists, for approximately two-thirds, of repeats of the Y-specific ay1 family of repetitive DNA sequences. Of the remaining one-third, at least one-half is represented by defective retrotransposons of the gypsy family. Both sequence types are interspersed throughout the loop. Using both ay1 and gypsy sequences as probes for transcript in situ hybridization, we show that, at the level of the light microscope, transcription of neither sequence is detectably affected in the loops formed by a male-sterile allele of gene Q. We conclude that the transcription of ay1 and gypsy is required, but not sufficient for the function of gene Q.

Discrimination of related transcribed and non-transcribed repetitive DNA sequences from the Y chromosomes of Drosophila hydei and Drosophila eohydei.

The short arm of the Y chromosome of Drosophila hydei carries a single male fertility gene, gene Q, which forms the lampbrush loop pair Nooses. Conflicting observations have been reported concerning the identity of the repetitive DNA sequences that are transcribed in this loop pair. It has been claimed by other investigators that the loop transcripts contain repeats of two distinct, but related families of Y-specific repetitive DNA sequences, ay1 and YsI. We reinvestigated this issue, using as probes single ay1 and YsI repeats which, under stringent conditions, hybridize only to members of their own family. Under non-stringent conditions, both repeats hybridize in situ to Nooses transcripts. However, if hybridization conditions are stringent, only the ay1 probe hybridizes to loop transcripts. Hybridizations to Northern blots of testis RNA confirm these results. Further, YsI repeats are not found the closely related species D. eohydei. We conclude that the YsI repeats are not relevant for the function of fertility gene Q.