Expression, regulation, and immunolocalization of putative homeodomain transcription factor 1 (PHTF1) in rodent epididymis: evidence for a novel form resulting from proteolytic cleavage.

PHTF1 is an 84-86-kDa membrane protein found in the endoplasmic reticulum of male germ cells in rodents. There are no evident signs of PHTF1 in the spermatozoa released into the lumen of the seminiferous tubules but PHTF1 is present in the epididymal epithelium. Characterization of the epididymal Phtf1 messenger by Northern blot and reverse transcription-PCR identified a 3-kilobase transcript in the epididymis, similar to that previously reported in the testis. The transcript is present in the proximal part of the epididymis and it appears when the rats reach 4 wk of age. Through immunofluorescence analysis, PHTF1 was localized in the principal cells of the initial segment and the caput epididymis. Colocalization with different markers indicated PHTF1 is in the endoplasmic reticulum saccules applied to the trans face of the Golgi system. Western blot analyses revealed a shorter form of the protein--about 56-kDa versus the 84-kDa form found in the testis. Using the canine epididymal cell line CIM 20, transfected by N- and C-terminal myc-tagged PHTF1, we demonstrated that the 56-kDa epididymal form could result from proteolytical processing.

Putative homeodomain transcription factor 1 interacts with the feminization factor homolog fem1b in male germ cells.

The Phtf1 gene encodes a membrane protein abundantly expressed in male germinal cells. Using a two-hybrid screening procedure we have identified FEM1B, an ortholog of the C. elegans feminization factor 1 (FEM-1), as a binding partner for PHTF1. We studied FEM1B expression in the rodent testis and found that Fem1b mRNA is present at high levels during meiosis and after, during spermiogenesis, in a similar manner to Phtf1 mRNA. Accordingly, Western blot and immunofluorescence revealed the presence of PHTF1 and FEM1B in the same cell types, and by coimmunoprecipitation we demonstrated the association between these proteins. We characterized some aspects of this interaction and showed that the ANK domain of FEM1B is necessary for the interaction with the amino extremity of PHTF1. Next, we found that FEM1B can bind several intracellular organelles and demonstrated that PHTF1 would recruit FEM1B to the endoplasmic reticulum membrane. Previous in vitro experiments had suggested that the human FEM1B was involved in apoptosis. After comparing expression profiles of FEM1B and PHTF1 with apoptotic events occurring in the normal seminiferous tubules, we suggest that neither FEM1B nor PHTF1 are directly implicated in apoptosis in this tissue.

Phtf1 is an integral membrane protein localized in an endoplasmic reticulum domain in maturing male germ cells.

Phtf1 is a gene evolutionarily conserved from Drosophila to human that is abundantly expressed in testis. In adult rat, transcripts were abundant in germinal meiotic and postmeiotic cells. Phtf1-specific antibodies revealed weak activity in a juxtanuclear region of early pachytene spermatocytes. Labeling progressively extended to the entire cytoplasm of step 2-3 spermatids, became intense from step 4, and persisted until the end of spermiogenesis, when it was eliminated in the residual bodies. Phtf1 displayed the properties of an integral membrane protein. In transfected cells and haploid cells of rat seminiferous epithelium, it colocalized with ER markers (calnexin and calmegin, respectively). By using both ER and Golgi markers (TGN-38, p58), we were able to show that, in pachytene spermatocytes and in Golgi phase spermatids, phtf1 labeled a region neighboring the cis-Golgi that probably corresponded to the peripheral Golgi region. Phtf1 staining was not related to beta-COP, AP1, or AP2 aptamers, indicating that it was not transported between Golgi saccules or between the Golgi complex and plasma membrane. However, aptamer labeling showed that chlatrin vesicles could be engaged in a new traffic route, raising the possibility of a meiotic proacrosomal vesicle origin. Colocalization between phtf1 and calmegin decreased during the acrosomal phase. During the maturation phase, phtf1 was able to identify different ER domains, as described previously for the peripheral Golgi region. Phtf1 provides a potential new marker for Golgi modifications as well as for many of the obscure transformations undergone by the endoplasmic reticulum. It could help to elucidate the morphogenic events connected with the transformation of spermatogenic cells.

Molecular characterization of a novel gene family (PHTF) conserved from Drosophila to mammals.

PHTF1 (putative homeodomain transcriptional factor; HGMW-approved symbol PHTF1) is a putative homeobox gene located at band 1p11-p13 of the human genome. We report here the cloning and sequencing of its mouse and Drosophila orthologs. The conservation between mouse and human proteins extends over the entire protein and is localized at the putative homeodomain and at the N- and C-terminal regions of Drosophila protein sequence. Blast searches allowed us to identify another member of the PHTF family, PHTF2, located at 7q11.23-q21 of the human genome. The strongest homologies between human PHTF1 and PHTF2 are localized to the domains that we already described in Drosophila, i.e., the putative homeodomain and the N- and C-terminal regions. The human and mouse genes display 98% similarity to one another, 56% similarity with the Drosophila gene, and 67% similarity with PHTF2, suggesting that phtf might define a novel gene family of highly divergent homeobox genes. Finally, Northern blot analysis showed that while PHTF1 is expressed mainly in testis, PHTF2 is predominantly expressed in muscle, suggesting that these two genes may have acquired different functions after their duplication and divergence.

Binary transgenic mouse model for studying the trans control of globin gene switching: evidence that GATA-1 is an in vivo repressor of human epsilon gene expression.

To test whether human GATA-1 (hGATA-1) is involved in the transcriptional control of globin gene switching, we produced transgenic mice overexpressing hGATA-1, crossed them with mice carrying a human beta-globin locus yeast artificial chromosome (beta YAC), and analyzed globin gene expression in their progeny. Mice carrying both the hGATA-1 and the beta YAC transgenes had normal levels of gamma- and beta-globin mRNA and no distortion in the rate or in the timing of gamma-to-beta switch, indicating that hGATA-1 is not involved in the developmental control of gamma- and beta-globin genes. In contrast, mice carrying the hGATA-1 and the beta YAC transgenes had 5- to 6-fold lower expression of the human epsilon globin gene compared with beta YAC mice lacking the hGATA-1 transgene. These results provide direct in vivo evidence that hGATA-1 is a specific repressor of human epsilon gene expression. Our findings also suggest that binary transgenic mouse systems based on overexpression of transcriptional factors can be used to investigate the trans control of human globin gene switching. Systems as the one we describe here should be useful in the study of any developmentally controlled human gene for which transgenic mice are available.

GATA1 and YY1 are developmental repressors of the human epsilon-globin gene.

The human epsilon-globin gene is transcribed in erythroid cells only during the embryonic stages of development. Expression of epsilon-globin gene, however, can be maintained in adult transgenic mice following removal of DNA positioned between -467 and -182 bp upstream of the epsilon-globin cap site. We have identified three protein binding regions within this silencer; a CCACC motif around -379, two overlapping motifs for YY1 and GATA around -269 and a GATA motif around -208 and we have analyzed their function during development by studying several mutants in transgenic mice. Mutation of the -208 GATA motif allows high epsilon-globin transgene expression in the adult suggesting that, in addition to its positive effects on transcription, GATA-1 also plays a negative role in the regulation of globin gene expression during development. Repression of epsilon gene expression in the adult also requires a functional YY1 binding site at position -269. Finally, mutation of the -379 CCACC site results in a small but detectable level of epsilon expression in adult erythroid cells. Thus, multiple proteins, including GATA-1, participate in the formation of the epsilon gene repressor complex that may disrupt the interaction between the proximal epsilon-promoter and the locus control region (LCR) in definitive erythroid cells.

Erythroid regulatory elements.

Erythroid differentiation leads to the production of red blood cells that contain a high level of hemoglobin. This level is mainly regulated by globin gene transcription during development and differentiation. Although numerous cis-acting sequences are involved in transcriptional activity of globin genes, combinations of three motifs, CCACC, SP1 and GATA represent the core elements of their regulatory sequences. These combinations are also found in promoters and/or enhancers of non-globin genes specifically expressed in the late stages of erythroid differentiation. The CCACC and SP1 sequences bind proteins that do not display erythrocytic specificity, and the GATA sequences bind a family of transacting factors recently cloned. The GATA family members are distinctive for a highly homologous DNA binding domain that exists in two zinc fingers reminiscent of those of the glucocorticoid receptor. None of the GATA family members displays only erythroid specificity, but gene disruption followed by rescue indicates that GATA-1 is necessary for terminal erythroid differentiation throughout development. The GATA/SP1 and GATA/CCACC associations are present in positive, negative or inducible regulatory sequences suggesting that other elements control the fine tuning of erythroid gene expression. NF-E2, which is a major transcriptional activator, members of the ets family which are implicated in the early stages of erythropoiesis and finally c-erbA which directly regulates a set of erythroid-specific genes are proteins that bind these latter regulatory motifs.

Demonstration of a human epsilon-globin gene silencer with studies in transgenic mice.

The human epsilon-globin gene displays normal developmental regulation in transgenic mice; it is expressed only in embryonic and in definitive erythroid cells. We show here that deletion of a negative element located between -182 and -467 bp upstream of the epsilon-globin gene cap site results in continuation of epsilon gene expression in the definitive erythroblasts of the fetal liver and in the red blood cells of adult transgenic mice. These data provide direct in vivo evidence that cis acting silencing elements are involved in the developmental control of the epsilon-globin gene.

Autonomous developmental control of human embryonic globin gene switching in transgenic mice.

The mechanisms by which expression of the beta-like globin genes are developmentally regulated are under intense investigation. The temporal control of human embryonic (epsilon) globin expression was analyzed. A 3.7-kilobase (kb) fragment that contained the entire human epsilon-globin gene was linked to a 2.5-kb cassette of the locus control region (LCR), and the developmental time of expression of this construct was studied in transgenic mice. The human epsilon-globin transgene was expressed in yolk sac-derived primitive erythroid cells, but not in fetal liver or bone marrow-derived definitive erythroid cells. The absence of epsilon gene expression in definitive erythroid cells suggests that the developmental regulation of the epsilon-globin gene depends only on the presence of the LCR and the epsilon-globin gene itself (that is, an autonomous negative control mechanism). The autonomy of epsilon-globin gene developmental control distinguishes it from the competitive mechanism of regulation of gamma and beta-globin genes, and therefore, suggests that at least two distinct mechanisms function in human hemoglobin switching.

Developmental regulation of human fetal-to-adult globin gene switching in transgenic mice.

Transgenic mice containing a human fetal (gamma-) or adult (beta-) globin gene linked to the beta-globin gene locus activation region (LAR) express the gene throughout development. By contrast, transgenic mice containing LAR linked to both a fetal and an adult globin gene display the normal developmental switch from fetal to adult gene expression. This suggests that the human fetal-to-adult globin gene switch is controlled through a mutually exclusive interaction between LAR and either the gamma- or beta-globin gene, resulting in the expression of only one gene at any given moment.

Butyrate induces expression of transfected human fetal and endogenous mouse embryonic globin genes in GM 979 erythroleukemia cells.

We have analyzed the expression of endogenous murine genes and of transfected human fetal A gamma globin gene in GM 979, a mouse erythroleukemia line which produces adult as well as embryonic globins. Optimal induction of the endogenous murine adult globin genes was obtained with DMSO or HMBA while the epsilon y and beta h1 embryonic genes were preferentially induced by butyrate. Similarly, the transferred human A gamma-globin gene was preferentially induced by butyrate. These results as well as previous observations in vivo or in erythroid cell cultures suggest that butyrate preferentially induces the expression of fetal globin genes.

Cis- and trans-acting elements involved in the regulation of the erythroid promoter of the human porphobilinogen deaminase gene.

Two cis-acting sequences, recognized by two erythroid-specific trans-acting factors, are involved in the regulation of the erythroid promoter of the human gene coding for porphobilinogen deaminase (PBGD). The first region, located at -70, binds the erythroid factor NF-E1, and point mutations within this region abolish the induction of transcription of this promoter during murine erythroleukemia (MEL) cell differentiation. The second region, located at -160, binds the erythroid-specific factor NF-E2 and the ubiquitous factor AP1. Using UV cross-linking, we show that NF-E2 has a higher molecular weight than AP1, demonstrating that NF-E2 is not an erythroid-specific degradation product of AP1. By point mutagenesis of the NF-E2/AP1 binding site, we define mutations that abolish binding of either NF-E2 alone or AP1 and NF-E2 together. Regulation of transcription of the PBGD erythroid promoter is abolished by those mutations, suggesting that NF-E2 but not AP1 is necessary for correct regulation of this promoter in erythroid cells.

Regulated expression of the overlapping ubiquitous and erythroid transcription units of the human porphobilinogen deaminase (PBG-D) gene introduced into non-erythroid and erythroid cells.

The human gene coding for porphobilinogen deaminase (PBG-D) is transcribed into two distinct transcription units giving two mRNAs. These units originate from two adjacent promoters distant of 3 kilobase pairs. The upstream promoter is active in all cell types, whereas the downstream promoter is active only in erythroid cells. We have studied the expression of this gene either after introduction of the corresponding human chromosome into murine erythroid cells using somatic hybrids or after transfection into both erythroid and non-erythroid cells. Using somatic hybrids, we showed that activation of the erythroid-specific promoter of the PBG-D gene did not reduce the rate of initiation of the ubiquitous promoter. Transfection experiments in erythroid cells showed that the PBG-D erythroid transcription unit, controlled by the PBG-D erythroid promoter, was correctly transcribed and regulated. Furthermore, we found that the PBG-D erythroid promoter alone was sufficient for correct expression and regulation of a reporter gene during erythroid differentiation. When the human PBG-D gene was transfected into non-erythroid cells, only the ubiquitous promoter was active. Deletion of the ubiquitous promoter did not lead to any activation of the erythroid promoter, suggesting that its inactivity in non-erythroid cells was not due to promoter occlusion but to a strict erythroid specificity.

Alternative transcription and splicing of the human porphobilinogen deaminase gene result either in tissue-specific or in housekeeping expression.

Porphobilinogen deaminase [PBGD; porphobilinogen ammonia-lyase (polymerizing), EC 4.3.1.8] is a cytosolic enzyme involved in the heme biosynthetic pathway. Two isoforms of PBGD, encoded by two mRNAs differing solely in their 5' end, are known: one is found in all cells and the other is present only in erythroid cells. We have previously shown that the human PBGD is encoded by a single gene and have now cloned and characterized this gene, which is split into 15 exons spread over 10 kilobases of DNA. We demonstrate that the two mRNAs arise from two overlapping transcription units. The first one (upstream) is active in all tissues and its promoter has some of the structural features of a housekeeping promoter; the second, located 3 kilobases downstream, is active only in erythroid cells and its promoter displays structural homologies with the beta-globin gene promoters.

Molecular cloning and complete primary sequence of human erythrocyte porphobilinogen deaminase.

We have cloned and sequenced a cDNA clone coding for human erythrocyte porphobilinogen deaminase. It encompasses the translated region, part of the 5' and the 3' untranslated regions. The deduced 344 amino acid sequence is consistent with the molecular weight and the partial amino-acid sequence of the NH2 terminal of the purified erythrocyte enzyme. Southern analysis of human genomic DNA shows that its gene is present as a single copy in the human genome and Northern analysis demonstrates the presence of a single size species of mRNA in erythroid and non-erythroid tissues and in several cultured cell lines. Quantitative determinations indicate that the amount of PBG-D mRNA is modulated both by the erythroid nature of the tissue and by cell proliferation, probably at the transcriptional level.

Assignment of human uroporphyrinogen decarboxylase (URO-D) to the p34 band of chromosome 1.

A cDNA probe corresponding to mRNA encoding human uroporphyrinogen decarboxylase (URO-D) was used to determine the chromosomal localization of the URO-D gene in the human genome. In agreement with previous studies, we have found that the locus for URO-D is located on chromosome 1 in hybrid cell mapping panels. The use of in situ hybridization allowed us to map the URO-D locus to band 1p34.