BIOLOGICAL OVERVIEW

The Drosophila gene boule is a homolog of the human gene Deleted in Azoospermia (DAZ), which, when mutated, causes azoospermia (no sperm production) by blocking meiotic cell divisions. Deletions of portions of the human Y chromosome long arm occur in a fraction of men with azoospermia. These deletions define a Y chromosomal region likely to contain one or more genes required for spermatogenesis (the Azoospermia Factor, AZF). Deletion of the AZF region is associated with highly variable testicular defects, ranging from complete absence of germ cells to spermatogenic arrest with occasional production of condensed spermatids. The AZF region contains the DAZ gene, which is transcribed in the adult testis and encodes an RNA binding protein (Reizo, 1995). In fact, multiple copies of DAZ (>99% identical in DNA sequence) are clustered in the AZF region and there exists a functional DAZ homolog (DAZH) on human chromosome 3 (Saxena, 1996).

boule and DAZ encode closely related proteins that contain a predicted RNA-binding motif; both loci are expressed exclusively in the testis (Eberhart, 1996). boule expression is limited to males and the boule transcript is absent from flies lacking a germ line, indicating that expression is testis specific. The appearance of boule transcript during the life cycle coincides with the onset of testis development and spermatogenesis. Spermatocytes are formed in boule mutants, but fail to undergo meiotic divisions (Castrillon; 1993, Eberhart, 1996 and Santel, 1997). Comparison of the localization of Cyclin A in boule mutant and in wild-type germ cells supports the conclusion that the meiotic prophase is normal in boule mutants. Although the meiotic prophase appears wild type in boule mutant germ cells. subsequent stages are aberrant. Cyclin A, which is exclusively cytoplasmic in the extended premeiotic G2, and enters the nucleus at the transition between G2 and M phases only to be degraded rapidly after nuclear translocation, persists in boule mutants (Eberhart, 1996).

What then is the target of Boule during meiosis? An important clue lies in the similarity of boule and twine. Twine (Twe) is required for meiotic entry in males: spermatocytes lacking twine function fail at the G2/M transition and, as a result, carry out incomplete differentiation of tetraploid spermatids. This phenotype is also produced by conditional alleles encoding the cyclin-dependent kinase Cdc2 (Alphey, 1992; Courtot,1992 and Sigrist, 1995), as well as by mutations in two other genes, boule and pelota, that encode proteins predicted to interact with RNA. Twine/cdc25 is a phosphatase that activates cdc2 kinase activity during meiotic divisions. Like the zygotic protein String, Twine activates cell cycle progression by removing phosphate groups from cdc2, the cyclin dependent kinase that forms heterodimers with Cyclin A and Cyclin B. Confirming this notion, ectopic expression of String, which is known to activate the cdc2 kinase before mitosis, results in a partial rescue of meiotic divisions in twine mutant testis (Sigrist, 1995).

Boule is required for Twine protein expression. By assaying the progress of spermatogenesis in the twine partial loss of function background, boule is a candidate in vivo regulator of twine. Spermatocytes from males with partial loss of function for twine and with only a single wild-type copy of boule fail to enter meiosis. Moreover, these cells do not carry out the G2/M transition; unlike wild-type or twine mutant spermatocytes, these double mutant cells fail either to relocalize cyclin A from the cytoplasm to the nucleus or to form bipolar spindles (Maines, 1999).

To explore the mechanism by which boule acts as an enhancer of twe mutants, Twine expression was examined in genetic backgrounds deficient for boule, using a reporter construct. Expression of the Twine reporter protein is dramatically reduced in a boule mutant compared with wild type. This reduction occurs at the level of protein, and not RNA, accumulation, since the amounts of twine–lacZ and endogenous twine RNA present are not reduced but are instead increased in the absence of boule. Given that, in wild-type flies, Twine RNA accumulates well before the onset of meiosis, these data indicate that (1) Twine is translationally regulated and (2) efficient Twine translation requires Boule (Maines, 1999).

Support for this hypothesis is provided by a study of the patterns of Boule localization and Twine expression in developing spermatocytes. Boule protein translocates from the nucleus to the cytoplasm just before the first meiotic division, and is required in the cytoplasm for meiotic entry (Cheng, 1998). Boule relocalization is coincident with the first detectable expression of the Twine reporter, consistent with cytoplasmic Boule being essential in the translation of Twine protein (Maines, 1999).

Heterologous Twine expression rescues the boule meiotic-entry defect. If boule mutants fail in meiotic entry because of inadequate accumulation of Twine protein, heterologous expression of Twine should drive meiotic entry in a boule mutant. To test this prediction, Twine was expressed in a boule-independent manner by placing twine under the control of the spermatocyte-specific beta2-tubulin gene promoter and the 5' and 3' untranslated sequences of the beta2-tubulin gene. Expression of twine in this context is sufficient to restore fertility to twine mutant males (Maines, 1999).

Introduction of the beta2-twine construct into a boule mutant drives meiotic entry, with some cells completing at least one division in all individuals studied. Meiotic spindles, which are absent in boule mutant males lacking the beta2-twine transgene, are easily apparent in transformed lines. Although the observed rescue could result from Twine mRNA overexpression, this seems unlikely for two reasons: (1) amounts of endogenous Twine mRNA are already quite high in boule mutants; (2) the rescue is not dose dependent and the presence of either one or two copies of the -beta2-twine transgene has no deleterious effect in the wild type. Furthermore, the activity of the beta2-twine construct in a boule mutant does not reflect a general ability to drive meiotic entry. For example, although the male sterile phenotype produced by mutation of the pelota gene closely resembles that produced in twine or boule mutants, the beta2-twine construct does not restore meiotic entry in a pelota male. Thus, the beta2-twine transgene appears to specifically compensate for the defect in endogenous Twine translation in a boule mutant, thereby restoring meiotic entry (Maines, 1999).

Boule cannot be required to stabilize the Twine mRNA, because the Twine message is abundant in a genetic background lacking boule activity. Instead, Boule is likely to play a part in Twine translation. Given that Boule contains an RNA-recognition motif (RRM), it is thought that Boule could influence Twine expression through direct binding to the Twine mRNA. Twine is not, however, the sole target of Boule activity, since boule mutants show defects in spermatid differentiation that are absent in twine mutant males (Eberhart, 1996). Taken together, these data indicate that Boule may be required for both meiosis and spermatid differentiation, and may have a role in coordinating these two events (Maines, 1999).

The boule and twine gene products act downstream of a second set of genes required for meiotic entry. The products of these genes, which include spermatocyte arrest (sa) and meiosis I arrest (mia), are also required for the expression of Twine protein but not of twine mRNA (White-Cooper, 1998). Mutations in these genes result in a failure to accumulate Boule protein. This inability of sa and mia mutants to express Boule and therefore Twine, presumably contributes to their failure to initiate meiotic divisions (Maines, 1999).

Although proteins of the Dazl family are required for fertility in several organisms, little is known about their biochemical function. Given the similarities in sequence and in expression patterns among Boule and other Dazl-family members, and also in the phenotypes induced by mutation of these proteins, it is suggested that a role in translation or translational control will be a general property of such proteins. Moreover, given that inactivation of the murine Dazl protein results in a proliferation defect, it is possible that transcripts encoding Cdc25 proteins will prove to be a general target for Daz-related proteins (Maines, 1999 and references therein).

GENE STRUCTURE

cDNA clone length - 2506

Bases in 5' UTR - 769

Bases in 3' UTR - 1047

PROTEIN STRUCTURE

Amino Acids - 228

Structural Domains

The Drosophila gene boule is a homolog of human DAZ. The two genes encode closely related proteins that contain a predicted RNA-binding motif. The RNA binding domain of the Boule protein, in the N-terminal portion of the protein, is 42% identical to the human DAZ protein and a closely related mouse protein, Dazla. Boule also has considerable sequence similarity (33% identity) to a second region of DAZ and Dazla. This region, in the C-terminal portion of the proteins, contains a motif, termed a DAZ repeat, that is present once in the mouse sequence and is repeated seven times in the human protein. Strikingly, the position of the RNP domain and the first DAZ repeat are conserved among the fly, mouse and human proteins. Whereas the human DAZ gene maps to a 500-kb region of the Yq that is frequently deleted in azoospermic men, the mouse gene, like boule, is autosomal (Eberhart, 1996).

date revised: 3 March 2000

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.