| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Testis |
Specifically Expressed in Human Testis1
Institute of Physiology3
Clinic of Anaesthesiology,4 University of Lübeck, D-23538 Lübeck, Germany
Cell Physiology Group,5 Medical Faculty, Martin-Luther-University Halle, D-06112 Halle, Germany
Institute of Vegetative Physiology,6 D-50931 Cologne, Germany
Institute of Physiology,7 University of Zürich, CH-8057 Zürich, Switzerland
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
in the mouse, termed mHIF-1I.1. Here, we demonstrate that expression of mHIF-1I.1 increases during puberty, further demonstrating its gene induction in postmeiotic germ cells. Using 5'-rapid amplification of cDNA ends, we identified a novel HIF-1 isoform in the human testis, called hHIF-1Te. Like mHIF-1I.1, hHIF-1Te mRNA is derived from an alternative promoter-first exon combination, but with a different genomic organization and a different nucleotide sequence. Reverse transcription-polymerase chain reaction analysis confirmed that hHIF-1Te is exclusively expressed in the testis. As determined by immunofluorescence of ejaculated sperm cells, HIF-1 protein is mainly localized in the postacrosomal head and in the midpiece of spermatozoa. Though overlapping with mitochondrial localization in human and mouse spermatozoa, neither hHIF-1Te nor hHIF-1 associated with mitochondria. In contrast with the ubiquitously expressed HIF-1 protein and the mouse testis-specific mHIF-1I.1 isoform, the hHIF-1Te mRNA sequence predicts a protein with an N-terminal truncation of the DNA-binding domain. As shown by yeast two-hybrid assays, hHIF-1Te still formed heterodimeric complexes with HIF-1ß. However, hHIF-1Te was incapable of forming a DNA-binding HIF-1 complex. Overexpression of exogenous hHIF-1Te resulted in the inhibition of the endogenous HIF-1 transcriptional activity, demonstrating that the testis-specific hHIF-1Te isoform is a dominant-negative regulator of normal HIF-1 activity.
environment, gene regulation, puberty, spermatogenesis, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1ß1 heterodimer specifically recognizing the HIF-binding site within cis-regulatory hypoxia response elements. Under normoxic conditions, the von Hippel-Lindau tumor-suppressor protein (pVHL) targets the HIF-1 subunit for rapid ubiquitination and proteasomal degradation [3]. Binding of the pVHL tumor-suppressor protein requires modification of HIF-1 by a family of low-affinity, oxygen-dependent prolyl-4-hydroxylases [4–6].
We previously cloned the mouse HIF-1 gene (Hif1a) and found that its expression is driven by two different promoters located 5' to two alternative first exons, designated mouse Hif1a exon I.1 and exon I.2 [7–9]. While the upstream exon I.1 promoter exhibits tissue-specific features, the downstream exon I.2 promoter is a typical housekeeping-type promoter driving ubiquitous transcription [8, 9]. The mouse mHIF-1I.1 mRNA isoform encodes for a predicted protein product that is 12 amino acids shorter than the predicted mHIF-1I.2 protein. The mHIF-1I.1 mRNA is detectable by the sensitive reverse transcription-polymerase chain reaction (RT-PCR) and RNase protection assays in a wide variety of mouse cell lines and organs [7– 9]. However, by using the less sensitive method of in situ hybridization, we found a ubiquitous expression of mHIF-1I.2 and a highly testis-specific expression of mHIF-1I.1 [10]. In testis, the putative mHIF-1I.1 protein was localized to the midpiece of mature spermatozoa. Besides testis, upregulation of mHIF-1I.1 expression has also been reported in activated T-lymphocytes [11], explaining the widespread expression pattern of mHIF-1I.1 when determined by RT-PCR [9].
So far, no alternative promoter for the human HIF1A gene has been reported and no homology could be detected between the mouse Hif1a alternative exonI.1 and the human HIF1A gene 5' region. To investigate whether a human counterpart to mouse mHIF-1I.1 exists, we performed 5'-rapid amplification of cDNA ends (RACE) using human testis RNA. A novel human testis-specific HIF-1 isoform, termed hHIF-1Te, was detected. Unlike mouse mHIF-1I.1, human hHIF-1Te mRNA was exclusively detectable in testis, even by RT-PCR, and showed a different genomic organization. Surprisingly, characterization of the DNA-binding function and transcriptional activity revealed that hHIF-1Te acts as a dominant negative regulator of normal HIF-1 function.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The pcDNAhHIF1- was constructed by ligation of a blunted EcoRV-BamHI (MBI Fermentas, Vilnius, Lithuania) fragment derived from pBSKhHIF-1T7 [12] into the blunted HindIII site of pcDNA3 (Invitrogen, Karlsruhe, Germany). Full-length hHIF-1Te was obtained by RT-PCR using human testis mRNA (Clontech, Heidelberg, Germany) and the forward and reverse primers, 5'-CGGGATCCCGATGTGCAAGACACTGCATTCTTAG-3' and 5'-GCTCTAGAGCGGGGATCAAAAATTGAACTAACC-3', respectively. The product was cloned into the BamHI and XbaI sites of pcDNA3, yielding pcDNAhHIF-1Te.
Animal Experimentation
Male C57BL/6 mice were killed by cervical dislocation and the testes were excised and rapidly frozen in liquid nitrogen for later RNA extraction. The experimental protocols were performed according to the German Animal Protection Law (Ministry for Nature and Environment of Schleswig-Holstein application number V21/A20/02). Histological pictures were obtained from the formaldehyde-fixed contralateral testis as described previously [10].
HIF-1 mRNA Isoform-Specific RT-PCR
Total RNA isolation from mouse tissue and RT-PCR encompassing mouse mHIF-1 exon I.1 or I.2, respectively, to exon III was performed as described previously [10]. RT-PCR analysis of total mHIF-1 mRNA was performed using primers encompassing mouse exons V (5'-TCAAGTCAGCAACGTGGAAG-3') to exon VI (5'-TATCGAGGCTGTGTCGACTG-3'). For the ß-actin control PCR reaction, amplification was performed with 27 cycles of 95°C for 60 sec, 58°C for 60 sec, and 72°C for 60 sec, using an exon III forward primer (5'-TGTTACCAACTGGGACGACA-3') and an exon IV reverse primer (5'-TCTCAGCTGTGGTGGTGAAG-3').
Human isoform-specific RT-PCR was performed with a commercially available cDNA panel (Origene, Rockville, MD). An aliquot (1 ng) of each cDNA was subjected to PCR amplification using each 10 pmol of the forward primers hHIFexI.1 (5'-CACCTCTGGACTTGCCTTTCCTTC-3') or hHIFexI.2 (5'-ATGTGCAAGACACTGCATTCTTAG-3') and the common reverse primer hHIF3.1 (5'-CACCAGCATCCAGAAGTTTCCTCAC-3'), located on exons II and III, in 25 µl 1x PCR buffer (Ambion, Huntingdon, U.K.), 0.25 mM deoxy-NTPs, and 0.5 U SuperTaq DNA polymerase (Ambion). After 40 cycles of 95°C for 60 sec, 56°C for 60 sec, and 72°C for 90 sec, the PCR products were analyzed by agarose gel electrophoresis and ethidium bromide staining.
Electrophoretic Mobility Shift Assay
Mouse and human HIF-1 isoforms as well as human HIF-1ß/aryl hydrocarbon receptor nuclear translocator (hARNT) were synthesized by in vitro transcription and translation (IVTT) in a rabbit reticulocyte lysate system (Promega, Madison, WI). Electrophoretic mobility shift assay (EMSA) was performed as described before [13, 14]. Briefly, oligonucleotides containing a HIF-1 DNA-binding site derived from the erythropoietin gene were gel purified on 10% polyacrylamide gels before 5' end labeling of the sense strand with (
-32P)ATP (Amersham Biosciences, Freiburg, Germany.). Unincorporated nucleotides were removed by gel filtration over Bio-Gel P60 columns (Bio-Rad, München, Germany). Labeled sense strands were annealed to a 2-fold molar excess of unlabeled antisense strands. DNA-protein binding reactions were carried out for 4 h at 4°C in a total volume of 20 µl of 10 mM Tris-HCl (pH 7.5), 50 mM KCl, 50 mM NaCl, 1 mM MgCl2, 1 mM EDTA, 5 mM dithiothreitol, 5% glycerol, containing in vitro-translated proteins, 100 ng of sonicated calf thymus DNA (Sigma, Buchs, Switzerland), and 2 x 104 cpm of oligonucleotide probe. Samples were run through 4% nondenaturing polyacrylamide gels at 200 V in TBE buffer (89 mM Tris, 89 mM boric acid, 5 mM EDTA). The gels were dried and radioactive signals were recorded by phosphorimaging. Supershift analysis was performed using the monoclonal anti-HIF-1 antibody mgc3 (Transduction Laboratories, Heidelberg, Germany) as described before [15].
5'-Rapid Amplification of cDNA Ends
For 5'-rapid amplification of cDNA ends (5'-RACE), human testis Marathon-Ready double-stranded cDNA was purchased from Clontech. Amplification was performed by nested PCR with the AP1 and AP2 primers (Clontech) and the gene-specific primers hHIF3.1 (5'-CACCAGCATCCAGAAGTTTCCTCAC-3') and hHIF3.2 (5'-CCTCACACGCAAATAGCTGATGGTAAGC-3'), respectively, according to the manufacturer's instructions. The PCR products were digested with XhoI and NotI and subcloned into the pBluescript vector (Stratagene, Amsterdam, The Netherlands). Following sequence determination (SeqLab, Göttingen, Germany), the positions of the inserts were located by comparison with the human HIF1A genomic sequence located on chromosome 14 (GenBank accession no. AL137129.4).
HIF-1 Immunofluorescence
HIF-1 immunofluorescence was performed as described previously [14]. Ejaculated sperm cells were washed in PBS and fixed with 3.5% formaldehyde, pelleted, resuspended in 80% ethanol, spread on glass slides, and air dried. After rehydration, unspecific binding sites were blocked with PBS containing 10% fetal calf serum. Subsequently, the sperm cells were incubated with a monoclonal anti-HIF-1 antibody (Transduction Laboratories) or 3% BSA in PBS alone. Fluorescein isothiocyanate-labeled rabbit anti-mouse IgG (DAKO Corp., Copenhagen, Denmark) secondary antibodies were used for the detection by fluorescence microscopy (Axioplan 2000; Carl Zeiss Vision GmbH, Mannheim, Germany).
Mitochondria Pull Down
Radioactively labeled hHIF-1, hHIF-1Te, and transcription factor A mitochondrial (TFAM) proteins were synthesized by IVTT (Promega, Mannheim, Germany) in the presence of (-35S)methionine. Rat liver mitochondria were prepared as described in detail previously [16]. Mitochondrial import was studied in 100 µl of the in organello transcription buffer system containing ATP (1 mM) in the presence of mitochondrial protein (2 mg/ml) and rabbit reticulocyte lysate (10% v/v) containing the radiolabeled proteins. After incubation at 30°C for 1 h, mitochondria were sedimented by centrifugation, dissolved in SDS sample buffer, and analyzed by SDS-PAGE. Gels were treated for fluorography, dried, and exposed to x-ray film.
Transient Transfection and Reporter Gene Assays
HepG2 cells (American Type Culture Collection, Manassas, VA) were cultured in Dulbecco modified Eagle medium high glucose containing 10% fetal calf serum, 50 IU/ml penicillin, and 50 µg/ml streptomycin (Invitrogen). HepG2 cells were seeded into six-well plates and transiently transfected with the pH3SVL firefly luciferase reporter gene plasmid that contains six HIF DNA-binding sites derived from the transferrin gene as described earlier [17]. In brief, each well was transfected with 13 µl Fugene (Roche, Basel, Switzerland) and 2 µg pH3SVL, 2 µg pcDNAhHIF-1 or pcDNAhHIF-1Te, and 0.4 µg pRLSV40 renilla luciferase reporter gene plasmid. After incubation under hypoxic (1% O2) or normoxic (20% O2) conditions, the cells were lysed and firefly and renilla luciferase activities were measured according to the manufacturer's recommendations (Promega).
Yeast Two-Hybrid Protein-Protein Interaction Assays
Yeast two-hybrid analysis was performed using the Gateway-compatible Proquest system (Invitrogen). Yeast cultures, transformation, and screening were performed according to the manufacturer's instructions (Invitrogen). Constructs were generated by fusing the cDNAs of full length or parts of hHIF-1, hHIF-1Te, or hARNT to the GAL4 activation domain (AD) or the GAL4 DNA-binding domain (DB) employing the vectors pExp-AD502 and pDBLeu, respectively. Plasmid pairs were cotransfected into yeast strain MaV203 and plated on minimal yeast synthetic complete medium lacking tryptophan and leucine (Sc-Leu-Trp-). Up to five colonies of each interaction were plated onto Sc-Leu-Trp-His supplemented with 10 mM 3-amino-1,2,4-triazole to examine activation of a HIS3 reporter. Expression of the lacZ reporter gene was evaluated by determining ß-galactosidase activity with a filter assay. For analyzing the uracil reporter (which is the most stringent selection method), five independent colonies were plated onto Sc-Leu-Trp-Ura. No self-activity of the used constructs could be observed after cotransfection of each pDBLeu and pExp-AD502 plasmid containing an insert with either the pExp-AD502 or pDBLeu vector lacking an insert, respectively.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Isoforms During Testis Development
We have previously demonstrated that the mouse testis-specific mHIF-1I.1 mRNA and protein isoform is expressed in the center of distinct seminiferous tubuli, whereas the ubiquitously expressed mHIF-1I.2 isoform is located to the basal layers of all tubuli [10]. To determine the onset of mHIF-1I.1 mRNA expression, we analyzed the developmental expression pattern of the mHIF-1 isoforms in the testis during mouse puberty (Fig. 1). The first round of spermatogenesis in the mouse starts at birth and is completed in about 34 up to 40 days postnatally. By determining the expression of mHIF-1I.1 in mice of different ages, it can be deduced at which stage of spermiogenesis the gene is expressed. At Day 12, the first meiotic spermatocytes appear. From Day 18 until Day 25, round spermatids increase in number and progressively differentiate to condensing spermatids up to Day 40 [18]. While we could not detect elongated spermatids at Postnatal Days 15 and 25 (Fig. 1B), even not at higher magnification (data not shown), fully developed elongated spermatids could be observed after Day 40 (Fig. 1B). As determined by RT-PCR with primers spanning exons V–VI, total mHIF-1 mRNA levels do not change during mouse testis development (Fig. 1A). However, a switch could be observed between Days 25 and 40 in the expression of the two mHIF-1 isoforms. While up to Day 25 the ubiquitous mHIF-1I.2 was the primary mRNA isoform, beginning with Day 40, the testis-specific mHIF-1I.1 became the primary mRNA isoform detectable in mouse testis (Fig. 1A).
|
DNA-Binding Properties of the Mouse Ubiquitous and Testis-Specific mHIF-1 Isoforms
The predicted protein product derived from the testis-specific mHIF-1I.1 mRNA is 12 amino acids shorter than the ubiquitous mHIF-1I.2 protein. Because this deletion is close to the basic domain of mHIF-1, we analyzed whether a functionally DNA-binding HIF-1 complex still could be formed. As shown by EMSA in Figure 2, in vitro transcribed/translated mouse mHIF-1I.1, mHIF-1I.2 as well as human hHIF-1 formed a DNA-binding complex together with hARNT, which could be supershifted by the addition of an anti-HIF-1 antibody. Thus, the N-terminal deletion of the mouse testis-specific HIF-1 isoform apparently does not affect its heterodimerization and DNA-binding properties.
|
Identification of a Novel Human HIF-1 Isoform by 5'-RACE
Although several splice variants of human hHIF-1 have been published, so far, no distinct hHIF-1 isoform has been reported that would be specific to the testis. Thus, we performed 5'-RACE using human testis cDNA and primers located on HIF1A exons II and III. Apart from the ubiquitously expressed hHIF-1 isoform, a novel cDNA species was cloned with a different N-terminal sequence (Fig. 3A), which was termed hHIF-1Te. Sequence comparison with the known human chromosome 14 genomic HIF1A sequence revealed 100% similarity with a sequence region located downstream of the ubiquitous HIF1A exon I (Fig. 3B). Thus, hHIF-1Te mRNA is the result of a transcription from an alternative promoter-first exon combination, as we found in the mouse before. However, the genomic order is inverted when compared with the corresponding region on mouse chromosome 12 (Fig. 3B).
|
In contrast with the mouse mHIF-1I.1 testis isoform, there is no ATG translational start codon at the corresponding site in hHIF-1Te. An in-frame ATG start codon in the 5' untranslated region is followed by a TAG stop codon six triplets downstream (Fig. 3A). The next ATG is located further downstream in the second helix of the basic-helix-loop-helix (bHLH) domain located in exon II (Fig. 3). The resulting reading frame encodes for a predicted protein of 767 amino acids with a calculated molecular mass of 86 kDa compared with 826 amino acids and a calculated molecular mass of 93 kDa of the ubiquitous hHIF-1 isoform.
Human hHIF-1Te mRNA Is Exclusively Expressed in the Testis
To determine the tissue expression pattern of the novel hHIF-1Te mRNA isoform, isoform-specific RT-PCR was performed using two human tissue cDNA panels, forward primers located on the specific first exons, and a common reverse primer located on exon II (Fig. 4). Strikingly, while hHIF-1 was expressed ubiquitously, hHIF-1Te expression was detected exclusively in testis (Fig. 4A). The hHIF-1Te mRNA could not be detected in human epididymidis, vesicula seminalis, or penis, further confirming its highly tissue-specific expression pattern (Fig. 4B).
|
HIF-1 Protein Expression in Human Spermatozoa
By immunofluorescence analysis using a monoclonal anti-HIF-1 antibody, HIF-1 protein was located mainly to the postacrosomal head and the midpiece of spermatozoa (Fig. 5A). Because the testis-specific HIF-1 protein isoform colocalized with mitochondria in the midpieces of both human (Fig. 5A) and mouse [10] sperm cells, the question arose whether HIF-1 might associate with mitochondria or even be incorporated into mitochondria. To answer this question, rat liver mitochondria were incubated with radioactively labeled in vitro-translated hHIF-1, hHIF-1Te, or the mitochondrial TFAM; sedimented; lysed; and subjected to SDS-PAGE and fluorography. Whereas TFAM was sedimented and incorporated and processed by the mitochondria, no similar effect could be observed with either hHIF-1 or hHIF-1Te (Fig. 5B).
|
hHIF-1Te Is a Dominant-Negative Regulator of Normal HIF-1 Function
The predicted protein sequence deduced from the primary hHIF-1Te DNA sequence lacks part of the N-terminal bHLH domain. Because the N-terminal regions of both HIF-1 and ARNT have been shown to be involved in heterodimerization and DNA-binding activity of HIF-1 [19, 20], we examined the impact of the testis-specific deletion on HIF-1 function.
HIF-1 DNA-binding activity was assessed by EMSA using in vitro-translated proteins and a radioactively labeled oligonucleotide derived from the erythropoietin gene 3' hypoxia response element. As shown in Figure 6A, hHIF-1 formed a DNA-binding complex together with hARNT, which could be supershifted by addition of a monoclonal anti-HIF-1 antibody. In contrast, hHIF-1Te was incapable of forming a functional DNA-binding HIF-1 complex.
|
One possible explanation for the lack of DNA-binding activity could be the failure of hHIF-1Te to heterodimerize with hARNT. To examine heterodimerization, yeast two-hybrid assays were performed. Auto-transactivation by the HIF-1 isoforms was prevented by using transactivation domain-depleted constructs. As shown in Table 1, both the ubiquitous hHIF-14–400 and the testis-specific hHIF-1Te1–341 isoforms were able to interact with hARNT. Under the most stringent selection conditions (uracil–), only the hHIF-14–400 interaction with hARNT conferred yeast growth. Apart from the bHLH-PAS A domain [21], also the lack of the HIF-1 Per-Arnt-Sim (PAS) B domain has been shown to affect the DNA-binding activity of HIF-1 [19]. Thus, we tested the role of the PAS B domain common to hHIF-1 and hHIF-1Te for the interaction with ARNT. Indeed, the PAS B domain (hHIF-1230–345) interacted with full-length hARNT in the yeast two-hybrid assays, conferring yeast growth even under highest stringency (Table 1). On the other hand, the PAS B domain of hARNT (hARNT349–467) was also sufficient for the interaction with both hHIF-1 isoforms (Table 1). No interaction occurred between hHIF-1 and hHIF-1Te. These results suggest that the failure of DNA-binding activity of hHIF-1Te is due to the lack of the protein domain interacting with DNA rather than due to the lack of heterocomplex formation with ARNT.
|
Because hHIF-1Te does not bind DNA but is still able to recruit ARNT, it might act as a negative regulator of normal HIF-1 function. To investigate this hypothesis, empty expression vectors or expression vectors containing either hHIF-1 or hHIF-1Te were transiently cotransfected together with a HIF-1-dependent firefly luciferase reporter gene and a constitutive renilla luciferase control expression vector into HepG2 cells. Following incubation under normoxic (20% O2) or hypoxic (1% O2) conditions for 24 h, luciferase activities were determined and firefly reporter gene activity was normalized to the renilla control activity. In cells cotransfected with the empty expression vector, hypoxia led to a 23-fold stimulation of reporter gene activity due to induction of the endogenous HIF-1 protein (Fig. 6B). Overexpression of hHIF-1Te significantly inhibited hypoxic reporter gene induction by 50% (P < 0.05, n = 6 independent transfections, paired t-test), suggesting that hHIF-1Te sequestrates endogenous hARNT and hence negatively regulates normal HIF-1 function. In contrast, overexpression of hHIF-1 did not further stimulate reporter gene induction, probably because all endogenous hARNT is engaged in the HIF-1 complex.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In this paper, we describe a novel HIF-1 isoform, termed hHIF-1Te, which is exclusively expressed in human testis. The predicted coding region lacks 59 amino acids of the N-terminal bHLH domain, resulting in a dominant-negative function of hHIF-1Te. We previously reported a testis-specific mHIF-1I.1 mRNA isoform in the mouse, which encodes for a predicted protein product that is shortened by only 12 amino acids [10]. At least following in vitro expression, this N-terminal deletion did not affect DNA-binding efficiency, despite its vicinity to the bHLH DNA-binding domain. However, it should be emphasized that the actual N-termini of the human or mouse ubiquitous or testis-specific protein isoforms in vivo never have been reported. Moreover, antibodies specific for the different N-termini of HIF-1 are not available. Thus, it might well be that in mouse testis in vivo, a downstream ATG is actually employed, resulting in a dominant-negative HIF-1 protein isoform similar to that shown here for the human hHIF-1Te isoform. Due to the lack of suitable antibodies, it is also impossible to determine the exact onset of HIF-1 protein translation from the testis-specific mRNA isoform during spermatogenesis.
What could be the function of a negative regulator of HIF-1 in the testis? Clearly, it is difficult to study transcriptional regulation (or silencing) in postmeiotic cells. Regarding oxygen biology, two major features of the seminiferous tubuli should be noted: they are highly hypoxic and completely avascular. Vessels are located between the tubuli and oxygen reaches the site of spermiogenesis exclusively by diffusion. Because of the diffusion distance and the high oxygen consumption of the proliferating cells, oxygen partial pressures (pO2) are likely to be very low. Indeed, pO2 values as low as 2 mm Hg have been reported, which are among the lowest values found in the body and otherwise occur only in the vicinity of mitochondria [26].
Due to the hypoxic conditions, there is a rather high level of constitutive expression of HIF-1 in premeiotic cells of the mouse testis [10]. Following ischemia, HIF-1 protein levels have been reported to be further upregulated in rat testis but not in the epididymidis [27, 28]. However, in postmeiotic cells, gene transcription is generally downregulated [23]. The testis-specific expression of a dominant-negative regulator of normal HIF-1 function might hence be involved in the silencing of the many genes that are under the transcriptional control of HIF-1, despite its strong activation by the hypoxic environment. This mechanism overlaps with the decrease of transcripts for the ubiquitous HIF-1 isoform in postmeiotic regions of the testis [10]. Of note, testis-specific variants of many of the glycolytic enzymes are expressed exclusively during spermatogenesis [29–31]. A block in normal HIF-1 function might contribute to the decrease in expression of the ubiquitous isoforms of the glycolytic enzymes, which are all under the transcriptional control of HIF-1.
Downregulation of transcriptional activity by dominant-negative regulation is a common mechanism in testis. The Id proteins are, to date, the best characterized negative regulators of bHLH transcription factors [32]. These short proteins, which contain a HLH but no basic domain, inhibit transcriptional activity by heterodimerization with certain transcription factors. Notably, Id 4 is specifically expressed in testis [32]. The hHIF-1Te might function through a similar mechanism. However, it is currently unknown whether the Id proteins interfere also with HIF-1 function.
Another avascular hypoxic tissue is the corneal epithelium of the eye. Interestingly, a hypoxia-inducible splice variant of the HIF-3 gene, termed inhibitory PAS protein (IPAS), has been reported to be highly expressed in the corneal epithelium [33, 34]. IPAS antisense oligonucleotides applied to the mouse cornea induced angiogenesis due to derepression of HIF-1-mediated vascular endothelial growth factor (VEGF) gene expression in hypoxic corneal cells, suggesting a novel mechanism for the maintenance of an avascular phenotype [33]. The hHIF-1Te might have a similar function during spermatogenesis, though this might not be directly related to the angiogenesis function of VEGF. Consistent with this speculation, transgenic mice overexpressing VEGF in the testis are infertile and show spermatogenic arrest [35]. However, in contrast with IPAS, hHIF-1Te does not interact with hHIF-1 (Table 1) and the critical target gene(s) to be repressed in the testis is (are) currently unknown.
The significance of stringent regulation of HIF-1 in the testis is further supported by recent findings in male VHLf/day Cre mice. These animals show oligospermia, reduction in testicular weight, and infertility, suggesting that impaired regulation of HIF-1 results in defects in spermatogenesis [36]. Because HIF-1 knockout mice die during embryogenesis [37], the role of HIF-1 in testis cannot be investigated in this animal model. However, a recent report on HIF-2 knockout mice showed that these animals suffer, among other things, from azoospermia [38]. Therefore, it will be important to determine the spatiotemporal expression pattern of HIF-2, HIF-3, and its splice variant IPAS during spermatogenesis. Testis-specific gene targeting experiments will be required to elucidate the specific function of each HIF- family member during spermatogenesis.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Roland H. Wenger, Institute of Physiology, University of Zürich-Irchel, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland. FAX: 41 0 1 63 56814; roland.wenger{at}access.unizh.ch
Received: 26 January 2004.
First decision: 16 February 2004.
Accepted: 9 March 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science 2001 292:468-472 targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science 2001 292:464-468. Biochem Biophys Res Commun 1996 223:54-59[CrossRef][Medline]. Genomic organization, expression and characterization of an alternative first exon and 5' flanking sequence. Eur J Biochem 1997 246:155-165[Medline] is encoded by two different mRNA isoforms: expression from a tissue-specific and a housekeeping-type promoter. Blood 1998 91:3471-3480 during the late stages of mouse spermiogenesis. Mol Endocrinol 2002 16:234-243 in activated T lymphocytes. J Biol Chem 2001 276:48754-48763 is dependent on heat shock protein-90 activity. J Biol Chem 2002 277:9262-9267. FASEB J 1999 13:81-88 stability, VEGF expression, and angiogenesis. FASEB J 2003 17:761-763 (HIF-1) protein in the adult rat testis following ischemic injury occurs without an increase in HIF-1 messenger RNA expression. Biol Reprod 2002 67:995-1002 locus. J Biol Chem 2002 277:32405-32408. Genes Dev 1998 12:149-162 gene: HIF1A structure and evolutionary conservation. Genomics 1998 52:159-165[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  N. Tajima, K. Schonherr, S. Niedling, M. Kaatz, H. Kanno, R. Schonherr, and S. H. Heinemann Ca2+-activated K+ channels in human melanoma cells are up-regulated by hypoxia involving hypoxia-inducible factor-1{alpha} and the von Hippel-Lindau protein J. Physiol., March 1, 2006; 571(2): 349 - 359. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Giaccia, M. C. Simon, and R. Johnson The biology of hypoxia: the role of oxygen sensing in development, normal function, and disease Genes & Dev., September 15, 2004; 18(18): 2183 - 2194. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
