| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
Departments of Obstetrics and Gynecology,3 Scott Urology,4
Molecular and Cellular Biology,5
and Molecular & Human Genetics,6
Baylor College of Medicine, Houston, Texas 77030
ABSTRACT
In the mouse, Utp14b is a retrogene transposed to an intron of Acsl3 (long-chain-fatty-acid coenzyme A ligase 3) on mouse chromosome 1. It represents a copy of Utp14a, a ubiquitously expressed, X-linked gene involved in 18S rRNA synthesis. The Utp14b is specifically expressed in male germ cells and, when mutated in the jsd (juvenile spermatogonial depletion) mouse, results in early spermatogenic arrest and male infertility. To understand the function and relevance of the orthologous human gene in testis pathology, we mapped transcripts and searched for mutations within the gene in infertile males. In humans, the strict ortholog of UTP14b has degenerated and is no longer functional. However, a second active retroposon, UTP14c, is found within a widely expressed, putative glycosyl transferase-containing gene, GT8, on human chromosome 13. Unlike mouse Utp14b, which is only expressed in the male germ line, human UTP14c is expressed in testis and ovary, which is consistent with having a gonad-specific function. To determine if UTP14c is functionally equivalent to mouse Utp14b and essential to spermatogenesis in humans, we screened DNA from 234 nonobstructive, azoospermic/severely oligospermic males and 208 proven-fertile controls for mutations within UTP14c. We identified a mutation in three unrelated patients that introduces an in-frame stop codon truncating the UTP14c protein near the carboxyl terminus. These data indicate that UTP14c may be functionally equivalent to mouse Utp14b and required for normal male fertility in humans. The novel evolution of retroposed UTP14 genes supports the hypothesis that retrogenes play an important role in evolution via regulation of male reproductive fitness.
sperm, spermatogenesis, testis
INTRODUCTION
Despite the relative high frequency of infertility in the human population (10–15% of couples [1]), relatively little is known about the cause. Nonobstructive azoospermia resulting from spermatogenic failure represents a type of human male infertility that is difficult to treat except with intracytoplasmic sperm injection. Between 25% and 50% of all male infertility is idiopathic, which may have a genetic basis. One cause of defective spermatogenesis is attributed to deletions of the Y chromosome, particularly in the Yq regions AZFa, AZFb, or AZFc. These deletions occur at a frequency of approximately 1 in 4000 [2], and they account for 8–12% of nonobstructive azoospermia cases. Klinefelter syndrome (XXY) is the most common known cause, accounting for approximately 14% of nonobstructive azoospermia (for review, see Matzuk and Lamb [3]). Nevertheless, most cases of testicular failure are of unknown etiology. Although many mouse models exist for male infertility [3, 4], translation of these findings to the human condition has been slow. Our basic understanding of the key processes controlling human spermatogenesis, sperm function, and fertilization is deficient. Hence, patients with nonobstructive azoospermia are diagnosed according to the description of their infertility defect, such as testicular failure, spermatogenic failure, maturation arrest, hypospermatogenesis, and Sertoli cell-only syndrome, rather than according to the genetic or physiological cause.
The understanding of gene function and its relevance to the human condition is a major objective in the postgenome era. Mouse mutational models, either spontaneous or deliberately created, are increasingly used to identify genes that are causative of disease and to ascertain information regarding gene function. One such mouse model is the recessive jsd (juvenile spermatogonial depletion) mutant [5]. The observed phenotype arose spontaneously in a colony of C57BL/6 mice. It is characterized by a single postnatal wave of spermatogenesis in developing jsd/jsd testes at approximately 3 wk of age, followed by a failure of type A spermatogonia to continue differentiating so that by 8 wk of age, spermatogenesis has ceased [6, 7]. In jsd/jsd males, the loss of type A spermatogonia results in sterile adult males with small testes, approximately normal numbers of Sertoli cells, and a few type A spermatogonia. The genetic basis for this phenotype is a small, jsd-specific rearrangement in the Utp14b retrogene located within the first intron of Acsl3 (long-chain-fatty-acid coenzyme A ligase 3) on mouse chromosome 1 [8, 9]. The Utp14b represents a retroposed copy of the X-linked Utp14a gene, which is the mouse homolog of the yeast UTP14 gene. Whereas Utp14a is ubiquitously expressed, the retrogene Utp14b has acquired a testis-specific expression pattern, being expressed in germ cells from zygotene to round spermatids.
The UTP14 protein is an essential component of a recently discovered cellular entity termed the small subunit (SSU) processome [10]. Comprising at least 28 functional proteins, this large ribonucleoprotein complex is bound to the U3 small nucleolar RNA and plays a key role in pre-rRNA processing into 18S rRNA and the biogenesis of ribosomes. We suggest that Utp14b, having retroposed and acquired a germ cell-specific expression pattern, may provide a mechanism for increasing the efficiency and/or numbers of germ cells produced by meeting the need for more 18S rRNA and protein during the early stages of spermatogenesis [9]. Such a mechanism would be of obvious reproductive advantage and under strong positive evolutionary selection.
We present the identification and analysis of the UTP14 genes in the human. Examination of the human syntenic region within ACSL3 shows evidence for an orthologous UTP14b gene that has degenerated and is no longer functional. However, a functional retroposon, UTP14c, produced by a second retroposition event that occurred after separation of the primate and rodent lineages is located within a widely expressed, putative glycosyl transferase-containing gene, GT8, on human chromosome 13. The UTP14c exhibits a gonad-specific expression pattern, being detectable only in the ovaries and testes. We present evidence that mutations within UTP14c are associated with human male infertility, indicating that as in the mouse, UTP14 retrogenes are essential for normal spermatogenesis to occur in men.
MATERIALS AND METHODS
Structural and Expression Analysis
Human and mouse genomic sequence data were based on sequences available through Ensembl (http://www.ensembl.org) and NCBI (http://ncbi.nlm.nih.gov) databases. Primers homologous to UTP14a and UTP14c were designed on the basis of gene predictions, known expressed sequence tags, or complete cDNAs present in the databases. A series of clones covering the entire UTP14c gene were generated using high-fidelity RT-PCR from human testis RNA (Supertaq; Ambion) as five overlapping segments using the primer pairs HumJSDLend/HumJSD1AR, HumJSD1AF/HumJSD4R, HumJSDmcDNAF/HumJSD8R, HumJSD8F/HunJSD3'NTR2 R, and HumJSD3'NTR2F/HumJSDRend. In addition, a single fragment covering the intact open reading frame was generated (primer pair HumJSD1AF/HumJSD8AR) to establish the authenticity of the gene transcripts. All primer sequences are given in Table 1. The RT-PCR products were cloned into pCR4-TOPO vector (Invitrogen) and sequenced. The tissue-specific expression pattern of UTP14a and UTP14c was determined by RT-PCR using DNase-treated total human RNA (Human Master Panel; Clontech) and primers specific for UTP14a, UTP14c, and GT8. The RNA was reverse transcribed using a RETROscript Kit (Ambion) according to the manufacturer's instructions. Specific primers (Table 1) for UTP14a were hUTP14 x 3F/hUTP14 x 3R, spanning exons 9–11 and giving a 395-bp product. For GT8, the primer pair HumJSD5'test/HumJSDseq1R was used, spanning exons 3 and 4 and giving a 532-bp product. For UTP14c, the primer pair HumJSDLend/HumJSD3R was used, spanning exons 1 and 2 and giving a 1680-bp product.
|
Patient Population
The present study was performed with the full approval and oversight of Baylor College of Medicine's Institutional Review Board for Human Subjects. Samples were obtained from the Tissue Procurement Core Facility as part of "The Genetic Basis of Male Infertility" Program Project supported by the National Institute of Child Health and Human Development (D.J.L. and C.E.B.). The DNA was extracted from peripheral blood samples using a Gentra Systems DNA extraction kit (Gentra) according to the manufacturer's instructions. Patients sought treatment for male infertility and were diagnosed with nonobstructive azoospermia (no sperm in the ejaculate) or severe oligospermia (<5,000,000 sperm/ml semen). Individuals with Sertoli cell-only syndrome, hypospermatogenesis, early and late maturation arrest, and mixed phenotypes were included in the study group. The majority of the patients underwent extensive clinical and laboratory evaluation in the Division of Male Reproductive Medicine and Surgery, Scott Department of Urology, at Baylor College of Medicine (Dr. Larry I. Lipshultz and D.J.L.). Known causes of male infertility, including Y chromosome microdeletions, cystic fibrosis transmembrane conductance regulator mutations, and chromosomal abnormalities were excluded, and their diagnosis was idiopathic. Approximately 68% of the 234 patients tested were Caucasian, 16% Hispanic, 11% Arab, and 5% Asian, generally representing the ethnic and racial distribution of the Baylor College of Medicine referral population. Control samples consisted of DNA from 208 proven-fertile males with a similar ethnic distribution. Because most male infertility patients chose to keep their identity and diagnosis confidential, our consent did not permit us to approach other family members for specimen collection and pedigree analysis. This limited the potential scope of desired sample collecting but was necessary for patient participation.
Mutation Screening
The coding region of UTP14c was divided into four overlapping regions, and gene-specific primers were used to generate four amplicons suitable for mutation screening by denaturing high-performance liquid chromatography (DHPLC) [11] using a Transgenomic WAVE instrument (Transgenomic Nebraska). Primer pairs used to generate fragments for analysis were HumJSD1AF/HumJSD2R for amplicon A, HumUTP14-3F/HumJSD4R for amplicon B, HumJSDmcDNAF/HumJSD6R for amplicon C, and HumJSD7F/HumJSDcDNAR for amplicon D. The PCR products that were identified by DHPLC as carrying potential mutations were directly sequenced and cloned into pCR4-TOPO vector (Invitrogen). A minimum of eight individual clones were sequenced to confirm the presence of mutations and to determine whether the alleles were heterozygous or homozygous.
RESULTS
UTP14c Evolution: Human UTP14c Is a Retrotransposed Copy of UTP14a
The mouse jsd gene, Utp14b, represents a retrotransposition of Utp14a from the mouse X chromosome to the first intron of Acsl3 on chromosome 1 (Fig. 1A). The mouse chromosome 1 region containing Utp14b and the syntenic region within human chromosome 2 were aligned using the ClustalW algorithm (http://www.ebi.ac.uk/clustalw/). Unexpectedly, no evidence for an intact, active retroposon within ACSL3 was found, although the ClustalW alignment was able to identify remnants of the complete gene. As shown in Figure 1B, Blast alignment of Utp14b with approximately 7000 bp of ACSL3 intron 1 identified homologies of greater than 80% with Utp14b. This suggests that before the divergence of rodents and primates, a copy of UTP14b was present within ACSL3 but has subsequently degraded and become nonfunctional. In addition, the host gene, ACSL3, has acquired two additional noncoding exons at the 5' end of the gene compared to mouse Acsl3 (Fig. 1A).
|
Further database searching revealed a complete retroposed copy of the ancestral X-linked human UTP14a gene, termed UTP14c, within the 3' nontranslated region of a novel, putative glycosyl transferase-containing gene, provisionally termed GT8, on human chromosome 13 (Fig. 1C). The UTP14c syntenic region in the mouse on chromosome 8 completely lacks any sequence related to Utp14 genes, indicating that human UTP14c is the result of a second retrotransposition event that occurred after the divergence of the rodent and human lineages [9]. Database analysis indicated that the chimpanzee shares the human genomic structure of Utp14c and ACSL3, which is consistent with their common phylogenic origin and evolutionary history.
Structure and Expression Pattern of UTP14c
To obtain the complete sequence of the UTP14c transcript, a series of primers covering the Ensembl-predicted transcript and several longer expressed sequence tags containing additional UTP14c sequences were used for RT-PCR from human testis RNA. The products were then sequenced, and the presence of a complete open reading frame, together with 5' and 3' untranslated region (UTR) sequences, was confirmed. Comparison of the UTP14c transcript with genomic sequences showed that UTP14c transcription is initiated within the third exon of GT8, 248 bp upstream of the splice junction (Fig. 2A), constituting the first noncoding exon of the gene. The second and terminal exon of UTP14c (5244 bp) completely overlaps the 2348-bp fourth and terminal exon of GT8. The first 458 bp of UTP14c exon 2 is 5' UTR, followed by a 2301-bp coding region and ending with a 2458-bp 3' UTR. The coding region of UTP14c starts 214 bp downstream of the termination codon of GT8.
|
The novel transcript initiation of UTP14c within the glycosyl transferase gene was confirmed by RT-PCR of testis RNA. As shown in Figure 2A, two forward primers, A and C (HumJSD5'test and HumJSDLend, respectively), were used. The former is within the GT8 coding region immediately upstream of the predicted start of UTP14c, and the latter covers the predicted start site. In addition, two reverse primers, B and D (HumJSD2R and HumJSD3R, respectively), were employed. Primer B is within a region shared by both transcripts, whereas primer D is beyond the predicted termination of GT8. As shown in Figure 2B, the combination of primers A and D gave no PCR product, whereas the combination of primers C and D gave the predicted band size. This is consistent with UTP14c initiating within GT8 and primer D being outside the GT8 transcript. The primer combinations specific to GT8 (primers A and B) and sharing sequences of both genes (primers C and B) each gave PCR products of the expected size. These data confirm that both transcripts overlap and are expressed in testis and that the termination site of the GT8 and initiation site of UTP14c are as predicted in Figure 2A.
The expression pattern of the ancestral X-linked UTP14a, the UTP14c retroposon, and the GT8 host gene was determined by RT-PCR using specific primers and a panel of human RNAs (BD Biosciences). As shown in Figure 3 (top), UTP14a showed a ubiquitous pattern of expression, being detectable in all 18 tissues tested. Similarly, GT8 was expressed in all tissues tested (Fig. 3, middle). In contrast, expression of the UTP14c retrogene was restricted to the gonads, being detectable only in the testis and ovary. This indicates that UTP14c has acquired gonad-specific promoter/enhancer elements after retrotransposition that are potentially located within the coding sequences of GT8.
|
Detection of Mutations Within UTP14c
Four primer pairs spanning the open reading frame of UTP14c (shown as A, B, C, and D in Fig. 4A) were used to generate PCR products from 234 DNA samples from infertile patients. These PCR products were screened by DHPLC for the presence of mutations. Analysis of the two 5' fragments, A and B (generated from primers HumJSD1AF/HumJSD2R and Hutp14-3F/HumJSD4R, respectively), identified the presence of two known, single nucleotide polymorphisms (SNPs) in the first fragment (Table 2; rs3742289 and rs372290) and five of eight known SNPs in the second (Table 2; rs1954782, rs1054784, rs105475, rs17402034, and one that is unassigned but present in the Ensembl database). In the third amplicon, C (generated using primers HumJSDmcDNAF/HumJSD6R), a novel A-to-G substitution was found in a single patient (patient 3), leading to a glutamine-to-arginine amino acid change, Q560R. This Q560R substitution falls into the class of very conservative amino acid substitutions and, as such, probably has little, if any, functional significance.
|
|
Within fragment D (generated using primers HumJSD7F/HumJSDcDNAR), a potential mutation was detected in three samples from unrelated Caucasian (white European descent) patients (Fig. 4B) identified as patients 1, 2, and 3. Direct sequencing of the PCR products from these three patients revealed the presence of a heterozygous C-to-G substitution at 2214 bp of the open reading frame. This was further confirmed by sequencing eight individual clones of the PCR product from each patient (Fig. 5A). This substitution resulted in the introduction of a TAG stop codon in place of tyrosine at position 738 (Y738X, a nonsense mutation) (Fig. 5B). The predicted gene product is truncated by 28 amino acids compared to the wild type. This mutation was not detected in any of the 208 DNA samples from fertile males, racially matched to the patient samples, which were screened for mutations in UTP14c.
|
The infertile patients who carried the Y738X mutation in UTP14c presented either with nonobstructive azoospermia (patients 1 and 2) or severe oligospermia (patient 3). Testicular biopsy of patient 1 revealed that he had an early germ cell arrest, with pachytene spermatocytes being the most mature cell type present. His hormone values were relatively normal (FSH, 8.2 mIU/ml; LH, 1.6 mIU/ml; prolactin, 17.3 ng/ml). Patient 2 also was azoospermic, but no further clinical information beyond the semen analysis was available. Patient 3 had severe oligospermia (<0.4 x 106 sperm/ml). Testicular histology was not available, because patients with any sperm present in their ejaculate did not undergo testis biopsies.
DISCUSSION
The UTP14 protein is predicted to be part of the SSU processome complex that binds to U3 small nucleolar RNA involved in 18S rRNA production [10]. We and others have shown previously that a mutation within the mouse Utp14b retrogene leads to early germ cell failure and sterility in jsd males [8, 9]. This suggests that the ubiquitously expressed, X-linked UTP14a is used in somatic cells, whereas male germ cells are further dependant on the autosomal retrogene for successful spermatogenesis. We therefore searched the syntenic region within ACSL3 on human chromosome 2 for evidence of a functional human ortholog. Although evidence of a human UTP14 retrogene could be found, the sequence is fragmented and degraded, with no evidence of being functional. However, a complete functional retrogene, specifically expressed in the testis and ovary, was identified within GT8, a novel glycerol transferase-containing gene on human chromosome 13. Comparison of the X transcript and genomic sequence around GT8 indicates that the UTP14c retroposition involved insertion of the complete X transcript into exon 4 of GT8, approximately 188 bp downstream of the translation stop codon. Subsequently, UTP14c has incorporated additional 5' and 3' UTR sequences, lengthening them to 733 and 2458 bp, respectively. It also has acquired specific promoter/enhancer elements, restricting its expression pattern to the ovary and testis. In addition, GT8 has incorporated part of the UTP14c coding sequence into its own 3' UTR region. Therefore, production of a complete UTP14c transcript requires bypassing the polyadenylation signal of the host GT8 gene.
These data indicate that the transcriptional machinery involved in transcribing UTP14c in the gonads is functionally different from that in somatic cells. In fact, a variety of novel gene regulation strategies are used during germ cell development, reflecting the complexity and number of maturation stages involved. These include initiation of transcription at alternative start sites, differential exon splicing, changes in the site of polyadenylation to control mRNA stability, and delays in the translation of transcripts to ensure a source of protein late in germ cell development after transcription ceases [12].
To test whether UTP14c, like its mouse ortholog, was required for normal spermatogenesis in men, we scanned for gene mutations in DNA samples from patients presenting with severe oligospermia or nonobstructive azoospermia. Of the 234 samples screened, three carried a heterozygous mutation (Y738X) that introduced a stop codon within the open reading frame, resulting in the loss of 28 amino acids at the carboxyl terminus. The DNA from parents or close relatives was unavailable for study. Thus, we were unable to determine if the mutation arose de novo or was inherited through the maternal lineage. However, none of 208 normal, fertile males who were tested carried this truncating mutation, which strongly suggests it was the underlying cause of the spermatogenic defects seen. Two of the infertile patients who carried the Y738X mutation in UTP14c presented with nonobstructive azoospermia, one of whom was shown to have an early maturation arrest. The third patient presented with severe oligospermia.
Because Y738X is present in the heterozygous state, several mechanisms exist to account for its potential role in the infertility observed in these patients. The simplest explanation is that haploinsufficiency leads to a suboptimal amount of UTP14c protein, compromising spermatogenesis. A second, more likely explanation is a dominant negative effect. Truncation of the UTP14c protein could compromise its structural integrity and interfere with its ability to correctly form an active SSU complex. This would result in a dominant negative effect on the function of the wild-type allele during 18S rRNA synthesis. Precedence for such a model is set by a mutation within the human testis-specific SYCP3 gene [13] that encodes a DNA-binding protein that is a structural component of the synaptonemal complex mediating homologous chromosome pairing during germ cell meiosis. A heterozygous mutation caused by a single nucleotide deletion within SYCP3 results in the loss of 21 (of 236) amino acids at the carboxyl terminus. The resulting diminished capacity for protein-protein interaction leads to meiotic arrest during spermatogenesis [14].
The nature of the Y738X nonsense mutation is of particular interest, because it yields a shortened polypeptide. Several examples for the expression of truncated polypeptides, encoded in the heterozygous state, are associated with clinically important anatomical and physiological outcomes in humans. The dominant form of ß-thalassemia is the result of the presence of any one of two nonsense mutations in the third exon of the ß-globulin gene on chromosome 11. In this case, inheritance of a C-to-T mutation in either base 121 or 127 of exon 3, the terminal exon, results in the development of anemia in spite of the presence of a normal allele [15]. Of interest is the fact that nonsense mutations within exon 2 of the ß-globulin gene have no phenotypic effect when in the heterozygous state, which suggests that the anemia observed in the case of exon 3 nonsense mutations is not the result of haploinsufficiency but, rather, of the presence of a truncated gene product that has a dominant negative effect on the function of the wild-type allele.
Another example of a heterozygous dominant nonsense mutation within the human genome is that of a Q155X mutation within the SOX2 gene found on chromosome 3. The SOX2 gene encodes a peptide of 317 amino acids that is involved in embryonic stem cell development. Clinically, the Q155X mutation manifests as anophthalmia, hearing loss, and brain anomalies, the development of which is thought to result from haploinsufficiency of SOX2 [16]. However, the most interesting example of a dominant nonsense mutation has been identified within the SOX9 gene [17, 18] that is caused by a Y440X stop codon mutation that truncates the 509-amino-acid Sox9 peptide by 69 amino acids. The SOX9 gene is located on chromosome 17, and mutations within this gene result in campomelic dysplasia and autosomal XY sex reversal. Four independent reports of campomelia dysplasia associated with Y440X mutations have appeared in the literature [17–20]. These mutations are caused by C-to-A or C-to-G transversions in a TAC codon generating TAA or TAG stop codons, respectively. This is analogous to the C-to-G transversion, and resultant Y738X mutation, that we have observed in UTP14c. In SOX9, the recurrent Y440X mutation accounts for approximately 8% of all mutations in the protein-coding region [20]. Because XY sex-reversed patients fail to reproduce and the Y440X mutation is nearly always in the heterozygous state [20], the C-to-A and C-to-G transversions must occur at a high frequency within SOX9 in man.
These examples clearly demonstrate that nonsense mutations can arise repeatedly at the same genetic locus and have important clinical ramifications in the heterozygous state. In the case of the mouse jsd mutation, a 4-bp insertion within 300 bases of the ATG translational start codon results in a frame shift and translational alignment of premature stop codons. This mutation thus results in a peptide that is 107 amino acids long, compared to the normal length of 779 amino acids. Severe truncation of the mouse Utp14b peptide results in total loss of function, and coupled with possible mRNA instability, this could explain why the mutation is recessive in the mouse. In contrast, the Y738X mutation within UTP14c of man only truncates the peptide by 28 amino acids, so this mutation is more analogous to those found in the ß-globulin, SOX2, and SOX3 genes described above, in which the almost intact genes display a dominant phenotype.
Two of the infertile patients who carried the Y738X mutation in UTP14c presented with nonobstructive azoospermia, one of whom was shown to have an early germ cell arrest. This is consistent with the phenotype seen in the jsd mouse model. The third patient presented with severe oligospermia, which is less severe than the phenotype seen in the mouse model. This could be caused by a number of factors. The mouse jsd mutation is a recessive null, in which no RNA or protein product is made. In humans, the mutation leads to a truncated RNA and protein product, which most likely acts as a dominant negative effect. Phenotype variability also could be strongly influenced by genetic background. In the mouse, a strain-dependent variability in the jsd phenotype exists, with C57BL/6 being more severely affected than C3H/He/J (M. Meistrich, personal communication).
The retroposition of genes from the X chromosome to the autosomes occurs at a frequency 300% greater than what is expected in both human and mouse, and of these retrogenes, 77% have testis-associated expression [21]. Three general hypotheses have been proposed relating to the selection forces producing autosomal, testis-specific retrogenes: 1) to compensate for sex chromosome inactivation and transcriptional repression during meiotic prophase of spermatogenesis [22], 2) the acquisition of unique germ cell-specific functions, and 3) the haploid syncytium hypothesis, in which "back-up" autosomal retrogenes are needed to equalize gene products in X- or Y-bearing haploid spermatids [23]. In all cases, retrogene survival is driven by enhancement of fertility.
The mechanism by which such retrogenes acquire a testis-specific expression pattern is unclear. In the case of the testis-specific PGK2 (phosphoglycerate kinase 2) retrogene, derived from the ubiquitously expressed X-linked PGK1, the transposition event carried with it the original promoter at the 5' end. Its subsequent modification has resulted in the acquisition of testis-specific promoter activity [24, 25]. In the case of UTP14c, it appears that the gene retropositioned into the GT8 gene without any promoter. It is probable that ubiquitous expression of UTP14c was originally driven by the GT8 host gene promoter, with tissue-specific expression and additional 5' and 3' sequences being acquired later. Interestingly, UTP14c has acquired this specific promoter without disruption of the host gene's open reading frame.
To our knowledge, of the genes reported to date that have retroposed from the X chromosome and acquired testis-specific expression in the mouse and human [24], only UTP14 has inserted within a host gene so that the transcription of both genes is in the same direction and overlaps. This novel strategy for acquiring testis-specific function appears to have been used by UTP14 twice, once before divergence of rodents and primates (as exemplified by the mouse Utp14b retrogene mapping within Acsl3) and a second time in the primate lineage (in which a second retroposition event placed UTP14c within GT8). A highly degraded copy of the original retrogene, UTP14b, can still be found within ACSL3 on human chromosome 2 in man (Fig. 1B).
The divergence of UTP14a and UTP14c since retrotransposition is of interest, because it can reflect the evolution of novel function and differences in selective pressure on different regions of the encoded proteins. Analysis of SNPs between the two genes is a good measure of divergence, because the genes were fixed within the human population. Twelve SNPs are found in the human database that fall within the coding sequence of UTP14c, eight of which result in amino acid changes. An additional two that have been found in the DNA from infertile male patients are reported in the present paper. The known SNPs within UTP14c are not randomly distributed but, rather, are located within the first 1360 bases of the coding region. Eight of the SNPs are clustered between nucleotides 633 and 956, suggesting that this region of the gene is under reduced selective pressure. To our knowledge, no SNPs have been identified previously in the region between bases 1360 and 2310, indicating that the carboxyl terminus is stable and under strong selective pressure. The two new SNPs found within this region in the DNA from patients with male infertility, particularly the one coding for a premature stop codon, therefore are within a highly conserved part of the gene. This is consistent with the possibility that the mutation at Y738X leads to a loss of function, and its absence from the human SNP database reinforces its low frequency within the human genome. In the case of the progenitor gene, UTP14a, a single SNP is found upstream of base 2130 of the coding sequence. The remaining 10 SNPs are clustered at the 3' end, indicating that unlike UTP14c, the carboxyl terminus is accumulating mutations. Alignment of the terminal 42 amino acids, including SNP-associated substitutions, of UTP14a with the carboxyl terminus of UTP14c (Fig. 5C) shows that the amino acid changes are remnants of the original peptide sequence that has been preserved in UTP14c. This may indicate that UTP14c has substituted for some function that was originally performed by the carboxyl terminus of UTP14a, thus releasing it from selective pressure and allowing the accumulation of mutations.
The UTP14c appears to be one of a number of genes that have retroposed off the X chromosome and acquired a critical and specific role in either male germ cell maintenance or sperm development. In light of the fact that the human genome encodes only 20000 to 25000 protein-coding genes [26], the relatively large number of retrogenes that have acquired germ cell- and/or spermatogenesis-specific functions indicates that genes such as UTP14c play an important role in speciation and development of major phylogenic groups. Further study of the evolutionary mechanisms by which humans have acquired male-specific retrogenes and their role in spermatogenesis will lead to a better understanding of male fertility in general and of spermatogenesis in particular.
ACKNOWLEDGMENTS
We would like to acknowledge the expert technical assistance of Leticia Bustos.
FOOTNOTES
1 Supported, in part, by P01 HD 36289 to D.J.L. and C.E.B. from the National Institutes of Health. 
2 Correspondence: Jan Rohozinski, Department of Obstetrics and Gynecology, Baylor College of Medicine, 1709 Dryden Road, Suite 1100, Houston, Texas 77030. FAX: 713 798 5074; janr{at}bcm.tmc.edu
Received: 16 August 2005.
First decision: 7 September 2005.
Accepted: 13 December 2005.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  M.-S. Shiao, P. Khil, R. D. Camerini-Otero, T. Shiroishi, K. Moriwaki, H.-T. Yu, and M. Long Origins of New Male Germ-line Functions from X-Derived Autosomal Retrogenes in the Mouse Mol. Biol. Evol., October 1, 2007; 24(10): 2242 - 2253. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
