|
Gene names- achintya and vismay
Synonyms - TGIF Cytological map position- 49A12--13 Functions- transcription factors Keywords- spermatogenesis |
Symbol- achi and vis FlyBase IDs: FBgn0033749 and FBgn0033748 Genetic map position- Classification- homeodomain proteins Cellular location- nuclear |
Two tightly linked and nearly identical homeobox genes of the TGIF (TG-interacting factor) subclass called vismay and achintya (often referred to as TGIF) are essential for spermatogenesis in Drosophila. 'achintya' is a Sanskrit word meaning 'that which is beyond thought and contemplation', and relates to initial difficulties in interpreting the mutant analysis; 'vismay' is a Hindi word meaning 'surprise', which described the reaction when the genome sequence revealed the tandem duplication (Ayyar, 2001). In flies deficient for both genes, spermatogenesis is blocked prior to any spermatid differentiation and before the first meiotic division. This suggests that vismay and achintya function at the same step as two previously characterized meiotic arrest genes, always early and cookie monster. Consistent with this idea, both always early and cookie monster are still expressed in flies deficient in vismay and achintya. Conversely, Vismay and Achintya proteins are present in always early mutant testes. Co-immunoprecipitation experiments further suggest that Vismay and Achintya proteins exist in a complex with Always early and Cookie monster proteins. Because Vismay and Achintya are likely to be sequence-specific DNA binding factors, these results suggest that they help to specify the spermatogenesis program by recruiting or stabilizing Always early and Cookie monster to specific target genes that need to be transcriptionally regulated during testes development (Wang 2003; Ayyar, 2003).
TGIF is a transcription factor of the TALE homeodomain class (Burglin, 1996) that has been implicated in a number of distinct pathways. TALE-superclass homeodomains are characterized by the presence of an additional three amino acids (Three Amino-acid Loop Extension) between helices 1 and 2. They comprise an ancient family, with representation from yeast to humans, and they act as transcription factors, often in collaboration with other homeodomain proteins. Members of the PBC and Meis classes of TALE proteins function as cofactors of Hox homeodomain proteins. TGIF was first identified as a competitor of the retinoic acid receptor for binding to retinoic acid response elements (Bertolino, 1995). Subsequently TGIF was demonstrated to interact with Smads, and a role has been proposed for it as a negative regulator of TGFß signalling based on in vitro and cell culture experiments (Wotton, 1999a). The findings that TGIF binds transcriptional repression proteins such as HDAC, mSin3A and CtBP and that TGIF can displace the CBP/p300 co-activator from Smad complexes suggest that it acts to build a repression complex on Smad target gene promoters (Wotton, 1999b; Wotton, 2001a; Wotton, 2001b). TGIF acting as a Smad co-repressor has been proposed to impose a response ceiling on transcription from TGFß response genes. TGIF has also been suggested to act as a competitive inhibitor of the TALE-class homeodomain protein Meis2 (Yang, 2000) in neuronal cell lines (Ayyar, 2003 and references therein).
Consistent with an in vivo role in TGFß/BMP signalling, TGIF has been identified as one of a small group of genes implicated in the human developmental disorder holoprosencephaly (HPE). This failure of forebrain formation is a relatively common developmental disorder affecting 1 in 250 conceptuses and 1 in 16000 live births. Loss-of-function mutations in TGFß family members in the mouse and zebrafish exhibit holoprosencephaly phenotypes. Four regions in the human genome (HPE 1-4) have been correlated with HPE and HPE 4 has been mapped to a 6 Mb region on chromosome 18 at p11.3, which includes the TGIF locus. Collections of HPE families have revealed TGIF alleles with mutations that affect protein function, which provide a plausible case for the relevance of TGIF to HPE but surprisingly these mutations do not appear to be more prevalent in the HPE group. Another potentially relevant gene, twisted gastrulation has also been recently found to be located at 18p11.3 (Ayyar, 2003 and references therein).
During spermatogenesis in Drosophila, stem cells located at the tip of the testes divide asymmetrically to give a stem cell daughter and a primary spermatogonial cell. The primary spermatogonial cells divide four times mitotically to give a cluster of 16 spermatogonia encased by two somatic cyst cells. After the fourth mitotic division, the cells undergo DNA replication and, now called primary spermatocytes, they enter a long (approximately 3.5 days) G2 phase which is a period of extensive transcription preceding the first meiotic metaphase. During this period the cells enlarge 25-fold. Upon entry into meiosis, transcription is shut down. The spermiogenesis genes, required to build functional sperm, tend to be transcribed during the long G2 phase and then held under translational control for later protein production after meiosis (Ayyar, 2003).
The control pathway underlying spermatogenesis is, as yet, poorly defined but a few 'meiotic-arrest' mutants have been identified. All the meiotic arrest mutants have a similar phenotype -- mature primary spermatocytes arrest development, and fail to enter either the meiotic divisions or spermatid differentiation. The currently identified meiotic arrest genes have been subdivided into two classes. The aly-class genes [always early (aly) and cookie monster (comr)] appear to be higher in the control hierarchy and regulate transcription of some genes involved in entry into meiosis (boule, twine, Cyclin B) and also of many spermiogenesis genes (e.g. fuzzy onions, janus B, don juan, gonadal) required for the differentiation of functional sperm. In contrast, can-class meiotic arrest genes (including cannonball, meiosis 1 arrest (mia) and spermatocyte arrest) do not affect transcription of the meiosis cell-cycle genes but are required for spermiogenesis gene transcriptional activation (Ayyar, 2003 and references therein).
To place achi/vis within this scheme the expression of a set of meiosis-related genes and a selected set of spermiogenesis genes were examined in Df(2R)achi1 homozygous mutant testes by RT-PCR analysis, and in homozygous mutant males by in situ hybridization. Both the set of spermiogenesis genes tested (fuzzy onions, janus B, don juan, gonadal) and the meiosis-related cell-cycle genes (boule, twine, Cyclin B) showed strongly reduced expression in the mutant, placing achi/vis in the aly class of meiotic arrest genes. Transcription of other genes (RP49, polo and Cyclin A) was not affected in the mutants. To determine whether Drosophila achi/vis is required upstream in the pathway for transcription of other meiotic arrest genes, the expression of aly and comr was tested in achi/vis mutant testes. In situ hybridization on achiZ3922 visZ3922 mutant testes revealed aly and comr transcripts at levels similar to wild type, and RT-PCR analysis on Df(2R)achi1 demonstrated robust expression of aly and can transcripts. In the RT-PCR analysis the levels of aly and can actually appear somewhat higher than wild type. This result is not interpreted, however, as indicative of a regulatory interaction but rather as a reflection of the altered cellular composition of the mutant testes. Similarly, aly and comr are not required for the expression of achi/vis because normal levels of achi/vis transcripts were found, by RT-PCR, in aly and comr homozygous mutant testes (Ayyar, 2003).
Neither of the two previously described aly-class meiotic arrest genes contain a predicted DNA binding domain, yet they are both chromatin associated, and are clearly required for transcriptional activation. A simple model would be that the gene products, Aly, Comr and Achi/Vis all act together as components of a single mechanism required for gene activation in spermatogenesis. If this were true it would predict that the phenotype of aly and comr mutations might be indistinguishable from the achi/vis loss-of-function phenotype. To test this, the phenotypes were examined in detail. As noted above the Df(2R)achi1 phenotype includes an expansion of early primary spermatocytes, indicative of an early role for achi/vis in the primary spermatocyte stage. A similar cellular defect has not previously been described for aly but, aly mutants also display expansion of the early primary spermatocyte population presumably due to a defect in progression through the primary spermatocyte differentiation program. This phenotype is not common to all meiotic arrest mutants and progression through the primary spermatocyte stages in mia mutants appears similar to wild type. In both aly and achi/vis mutants the primary spermatocytes do exhibit some spermatocyte differentiation; they increase in size and chromosomal reorganization occurs, giving clear chromatin clumps, as visualized by either DAPI or anti-histone labelling. However, the chromatin fails to organize as tightly in the aly mutant as in the wild type and the cells arrest with peripheral chromatin clumps with a fuzzy appearance. The chromatin morphology of comr is identical to that of aly. In achi/vis mutant testes the chromatin appears to follow a wild-type program up to the generation of mature primary spermatocytes with peripheral chromatin clumps, however, the cells arrest with rounded chromatin clumps that are not apposed to the nuclear periphery and that resemble the chromatin configuration in meiotic stages. It is concluded that the achi/vis phenotype is similar but not identical to that of aly and comr (Ayyar, 2003).
The normal chromatin association of Aly and Comr proteins is essential for their function, and the localization of these two proteins is mutually dependent, i.e., in an aly mutant Comr protein remains cytoplasmic, and vice versa. In contrast, both Comr and Aly proteins localize to chromatin in testes mutant for the downstream, can-class, genes. To determine whether achi/vis plays a role in the production or localization of the other aly-class proteins, the levels and localisation of Aly and Comr proteins were examined in achiZ3922 visZ3922 mutant testes. Both Aly and Comr proteins were detected by Western blotting in achiZ3922 visZ3922 mutant testes. Immunofluorescent staining revealed that Aly and Comr proteins are nuclear in achiZ3922 visZ3922 testes. This places achi/vis downstream of, or parallel to, comr and aly (Ayyar, 2003).
Many of the phenotypes observed in the achi/vis deficiency are also observed in aly and comr mutants. In addition, immunolocalization studies suggest that Vis, Achi, Aly and Comr proteins are co-expressed in the nuclei of primary spermatocytes. These observations prompted a test to see if these proteins may be present as a complex in wild-type testes. This was by carrying out immunoprecipitation (IP) experiments with the anti-Achi antibody and determining if either Aly or Comr is co-immunoprecipitated. Interestingly, both Aly and Comr can be co-immunoprecipitated with Vis/Achi from wild type, but not from pingpong testes. These results suggest that Vis and Achi proteins are present in a complex with Aly and Comr during wild-type testes development (Wang, 2003).
It is concluded that Drosophila TGIF is required for a specific transcriptional program in Drosophila spermatogenesis. The extended G2 phase that primary spermatoctyes undergo prior to entry into meiotic division is an important period in spermatogenesis. It lasts for about 3.5 days and during this time the cells increase in volume 25-fold, execute the transcription program required to produce all the transcripts necessary for sperm differentiation and undergo a striking sequence of chromatin reorganization. The switch in transcriptional activity that occurs in primary spermatocytes is one of the most dramatic of any differentiation pathway, with a great many genes being expressed exclusively in this cell type. At the end of this stage virtually all transcription is switched off; many of the transcripts produced during this period are held under translational inhibition to be released after meiosis in a coordinated program of protein production that mediates sperm differentiation. Most male sterile mutations affect these later stages of sperm manufacture and typically result in a block in spermiogenesis at the very late stage of sperm individualization. A relatively small set of mutations have been characterized that block spermiogenesis during the extended primary spermatocyte G2 phase. In these mutants, spermatocytes fail to enter meiotic division and fail to initiate spermatid differentiation (Lin, 1996). These meiotic arrest genes are required to initiate the primary spermatocyte-specific transcriptional activation program. Previous genetic and biochemical analyses have suggested that the aly-class genes act before the can-class genes, at the top of a regulatory hierarchy (White-Cooper, 1998). Therefore the aly-class meiotic arrest genes provide an entry point into the mechanisms that initiate and orchestrate the transcriptional program of spermiogenesis, that control spermatocyte differentiation and that regulate the entry into the meiotic divisions (Ayyar, 2003).
The relatively small number of these meiotic arrest genes currently identified presents the beguiling prospect that the transcriptional program of spermiogenesis may be controlled by an ancient simple mechanism. Interestingly, two of the can-class meiotic arrest genes (cannonball and no hitter) encode testis-specific components of the general transcription factor TFIID, suggesting that the genes activated during the primary spermatocyte stage may share a distinct core promoter type (Aoyagi, 2000; Aoyagi, 2001; Hiller, 2001). The homology of aly to a C. elegans gene implicated in a pathway leading to chromatin remodelling factors suggests that aly may have a role in chromatin reorganization to allow access for specific transcription factors and the testis specific TFIID to target promoters (Beitel, 2000; White-Cooper, 2000). The characterization of Drosophila TGIF represents the first description of a sequence-specific transcription factor implicated in this pathway. Combined mutations in achi and vis or deletion of both genes lead to a recessive male sterile meiotic arrest phenotype. A markedly decreased expression was found for both the spermiogenesis genes and also for CycB and twine, required for entry into meiosis, placing achi/vis into the aly-class of meiotic arrest mutants. Drosophila TGIF does not appear to be required for the transcriptional activation of other meiotic arrest genes since aly, comr and cannonball are all expressed in achi/vis mutants. Similarly, function of other meiotic arrest genes is not required for transcription of achi/vis (Ayyar, 2003).
Although the gross transcriptional consequences of loss of either achi/vis or aly appear similar, the mutant phenotypes are distinct. Both show effects on early primary spermatocytes with an expansion of this cell type presumably due to a slowing of the progress of differentiation through this stage. All the aly-class mutants (aly, comr and achi/vis) exhibit defects in chromatin organization but whereas the aly and comr mutant spermatocytes arrest with 'fuzzy' chromatin condensation (Lin, 1996), in achi/vis mutants the chromatin rounds up in condensed 'blobs' which are reminiscent of meiotic pro-metaphase. This difference in the phenotype is consistent with the finding that Drosophila TGIF is not required for the normal localization of Aly and Comr proteins, and suggests that TGIF is also not required for the chromatin remodelling mediated by Aly and Comr. This raises the question of whether the aly-class genes all act together as components of a simple transcription activation switch or whether they may be a somewhat heterogeneous collection with more diverse roles within the spermatogenesis transcriptional program (Ayyar, 2003).
To identify Drosophila TALE class homeodomain genes BLAST searches of Drosophila genomic sequence databases were performed. This analysis revealed an uncharacterized TALE sequence, which showed strong sequence similarity to the TGIF family. ESTs homologous to this sequence were also found in cDNA libraries from several tissues and developmental stages. The complete Drosophila genome sequence revealed that these EST sequences are in fact the products of two closely juxtaposed tandemly-repeated transcription units, achintya and vismay. The two genes are highly similar (93% at nucleotide level and 97% at protein level) (Ayyar, 2003).
A comparison of the protein encoded by this locus with vertebrate TGIF sequences reveals that the sequence similarity exists both within the homeodomain (including RYN as the TALE amino acids characteristic of the TGIF family) and for about 30 amino acids on the carboxy-terminal side of the homeodomain. Such a C-terminal domain is also found in the yeast TALE protein MATalpha2 and members of the PBC family, where these residues fold into additional helices required for the proper functioning of the homeodomain (Phillips, 1994; Sprules, 2000). The complete genome sequence of the mosquito Anopheles gambiae has recently been published and also reveals the presence of a TGIF homolog (AgCG54405). This gene encodes a protein (AgCP3385) with 63% similarity to Achi/Vis. In addition to the sequence conservation in the homeodomain and C domain there are additional blocks of homology between these invertebrate sequences (Ayyar, 2003).
vis (CG8821) and achi (CG8819) are both predicted to encode TALE homeodomain proteins. Analysis of genomic and cDNA sequences suggests that both genes have the potential to encode two protein products, which are referred to as long (L) and short (S) isoforms. The long and short forms of both Vis and Achi, that have been directly observed by immunoblot analysis, differ in length because of the presence or absence of an alternatively spliced exon. Blast and ClustalW analyses indicate that Vis and Achi are almost identical. For example, AchiS and VisS are predicted to be identical in 406 of 424 residues (96% identity). In addition, the alternatively spliced exons in Vis and Achi are 129 residues in length and are predicted to be 100% identical to each other. The high degree of identity between these proteins is unusual for homeobox genes (Wang, 2003).
Blast and ClustalW analyses also reveal that the mammalian proteins most similar to Vis and Achi are human TGIF and TGIF2. The similarity to TGIF and TGIF2 is most obvious in the homeodomain (TGIF is ~78% identical to Achi), but the identity extends to residues C-terminal to the homeodomain. A protein predicted from the genome sequence of the mosquito, Anopheles gambiae, shows several additional regions with similarity to Vis and Achi. Other members of the TGIF subgroup have homology to Vis and Achi primarily within the homeodomain (Wang, 2003).
date revised: 12 June 2003
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.