| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Mechanisms of Hormone Action |
a Laboratoire de Génétique Moléculaire,
b INSERM U427, Faculté des Sciences Pharmaceutiques et Biologiques, Université René Descartes—Paris V, F-75006 Paris, France
c Laboratoire d'Oncogénétique, Centre René Huguenin, F-92211 St-Cloud, France
d Laboratoire BGBP-UMR CNRS 5558, Université Claude Bernard, Lyon 1, F-69622 Villeurbanne Cedex, France
e Institut d'Embryologie EA3428, F-67085 Strasbourg, France
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
45 million years ago). Transient transfection experiments showed that the placenta-specific expression of INSL4 is mediated by the 3' LTR of the HERV element, and that the latter may have a major role in INSL4 up-regulation during human cytotrophoblast differentiation into syncytiotrophoblast. Finally, we identified an INSL4 alternatively spliced mRNA species that encodes putative novel INSL4-like peptides. These data support the view that ancient retroviral infection may have been a major event in primate evolution, especially in the functional evolution of the human placenta.
gene regulation, insulin, placenta, pregnancy, relaxin
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We have previously shown that RLN1, RLN2, and INSL4 map to chromosome region 9p23.3–p24.1 [5]. INLS6 has also been mapped using a human-rodent radiation hybrid panel to chromosome arm 9p24 [3], whereas the other insulin-like genes (INS, IGF-I, IGF-II, INSL3, INSL5, and RNL3) are located on other chromosomes. These findings suggest that RLN1, RLN2, INSL4, and INSL6 may be clustered in the same chromosome region and may have arisen through a mechanism of cis local duplication [6]. These four genes are expressed in tissue-specific fashion: RLN1 and RLN2 are primarily expressed in the prostate [7], INSL6 in the testis [3], and INSL4 in the placenta [8].
INSL4 was identified by screening a human cytotrophoblast-subtracted cDNA library, and its expression was found to be highly placenta-specific [8]. INSL4 is expressed more strongly during the first trimester of pregnancy [8] and in the differentiated syncytiotrophoblast than in cytotrophoblast cells [9], suggesting that INSL4 may be important in human placental morphogenesis.
To further characterize the placenta-specific expression of INSL4 relative to the other three insulin-like genes located in chromosome region 9p23.3–p24.1 (RLN1, RLN2, and INSL6), we characterized the genomic organization of INSL4, INSL6, RLN1, and RLN2 at this chromosome region. We identified a human endogenous retrovirus (HERV) element inserted in the INSL4 promoter, and examined the significance of its 3' long terminal repeat (LTR) in the placenta-specific expression of INSL4. We also discuss the evolutionary history of the insulin-like gene emergences and the HERV element integration in chromosome region 9p24.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Term placentas were obtained from healthy mothers with uncomplicated pregnancies following elective caesarean deliveries. Villous tissue was dissected free of membranes and vessels, rinsed, and minced in Ca2+-free and Mg2+-free Hanks balanced salt solution for cytotrophoblast cell isolation and culture.
Total RNA from normal human tissues (brain, thymus, placenta, liver, testis, breast, prostate, stomach, colon, skin, leukocytes, adrenal gland, pancreas, and uterus) and genomic DNA from Macaca mulatta (rhesus monkey) were purchased from Clontech (Palo Alto, CA).
Cercopithecus aethiops (Africa green monkey) cell line (kidney COS-1 cell line) was obtained from the American Tissue Type Culture Collection (Manassas, VA). Eulemur macaco (black lemur) and Calimico goeldii (Goeldi's marmoset) tissues were kind gifts from Dr. Jean-Luc Fausser (Institut d'Embryologie, EA3428, Strasbourg, France) and Dr. Eric Denamur (Laboratoire d'études de génétique bactérienne dans les infections de l'enfant, EA3105, Hôpital Robert Debré, Paris, France), respectively.
Isolation of INSL4 Genomic Clones
The human P1-derived artificial chromosome (PAC) DNA library (UK Human Genome Mapping Project [HGMP] Resource Center, Hinxton, Cambridge, U.K.) was screened at high stringency using a nick-translated (
-32P)dCTP-labeled INSL4 cDNA probe.
Genomic DNA and RNA Extraction
Total genomic DNA from a human P1-derived artificial chromosome and from C. aethiops, Macaca mulatta (rhesus monkey), E. macaco, and C. goeldii tissues was prepared according to the Protein K-phenol method [10].
Total RNA was extracted from cultured placental cells according to the acid-phenol guanidinium method [11]. The quality of the RNA samples was determined by electrophoresis through denaturing agarose gels and staining with ethidium bromide, with visualization of the 18S and 28S RNA bands under ultraviolet light.
Sequencing of PCR Products and Genomic Clones
PCR products and genomic clones were sequenced using the Applied Biosystems (ABI) PRISM Dye Terminator Cycle Sequencing Ready Reaction kit (Perkin Elmer Applied Biosystems, Foster City, CA) and the Applied Biosystems model 377 DNA sequencer (Perkin Elmer).
Real-Time Reverse Transcriptase-Polymerase Chain Reaction
The theoretical and practical aspects of real-time quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) on the ABI Prism 7700 Sequence Detection System (Perkin-Elmer) have been described in detail elsewhere [12]. Briefly, total RNA is reverse-transcribed before real-time PCR amplification. Quantitative values are obtained from the threshold cycle (Ct) number at which the increase in the signal associated with exponential growth of PCR products begins to be detected using PE Biosystems analysis software, according to the manufacturer's manuals. The precise amount of total RNA added to each RT reaction mix (based on optical density) and its quality (i.e., lack of extensive degradation) are both difficult to assess. We therefore also quantified transcripts of the RPLP0 gene (also known as 36B4, encoding human acidic ribosomal phosphoprotein P0) as the endogenous RNA control, and each sample was normalized on the basis of its RPLP0 content. The relative target gene expression level was also normalized to a calibrator, or 1;ts sample. Calibrators were the normal human adult tissues that contained the smallest quantifiable amount of target gene mRNA (for the expression pattern of target genes in normal human adult tissues) and the matched cytotrophoblasts from eight placentas (for the expression of target genes during in vitro human cytotrophoblast differentiation).
Final results, expressed as N-fold differences in target gene expression relative to the RPLP0 gene and the calibrator, termed "Ntarget", were determined as follows:
 |
Ct values of the sample and calibrator are determined by subtracting the average Ct value of the target gene from the average Ct value of the RPLP0 gene. The nucleotide sequences of the oligonucleotide primers are shown in Table 1.
|
PCR was performed using the SYBR Green PCR Core Reagents kit (Perkin-Elmer). The thermal cycling conditions comprised an initial denaturation step at 95°C for 10 min, and 50 cycles at 95°C for 15 sec and 65°C for 1 min. Experiments were performed with duplicates for each data point.
Construction of INSL4 Promoter/Luciferase Reporter Gene Plasmids
Different lengths of the INSL4 gene promoter were obtained by PCR amplification of human genomic DNA (Roche Diagnostics, Mannheim, Germany) using specific primers containing an artificial XmaI site. PCR products were purified and inserted into the multiple cloning site of the promoterless luciferase vector pGL3-basic (Promega, Madison, WI). Cloning details are available from the authors. Deletion mutant del(-719) (-719 to +95) was generated by AlfII/NheI digestion of del(-1707) (-1707 to +95). The identity of isolated clones was confirmed by nucleic acid sequencing.
Cell Culture and Luciferase Assay
The human choriocarcinoma cell line JEG-3 was purchased from American Type Culture Collection and cultured in Minimum Essential Medium (Invitrogen, Cergy Pontoise, France) with 10% fetal calf serum, 2 mM L-glutamine, 50 IU/ml penicillin, and 50 µg/ml streptomycin.
Cells were cotransfected with 10 µg of the appropriate INSL4 promoter plasmid and 10 µg of pSV-ß galactosidase as an internal control plasmid. The cells were harvested 48 h after transfection, and cell extracts were prepared using a cell culture lysis reagent (Promega).
After centrifugation, 20 µl of the supernatant was assayed for luciferase activity by using the Promega assay system on a Lumat LB 9501 luminometer (EG Instruments, Evry, France). The luciferase activity of each construct was normalized for differences in transfection efficiency, based on the results of the ß-galactosidase assay. Each INSL4 promoter construct was tested in duplicate in a minimum of three independent transfections.
Cytotrophoblast Cell Culture
Cytotrophoblast cells were isolated and purified from term chorionic villi after trypsin-DNase digestion and discontinuous Percoll gradient fractionation, as previously described [13]. The cells were plated in triplicate in 60-mm culture dishes (3 x 106 cells/dish) in 3 ml of Dulbecco modified Eagle medium, 2 mM glutamine, 10% heat-inactivated fetal calf serum, and antibiotics (100 IU/ml penicillin and 100 µg/ml streptomycin). The dishes were incubated at 37°C in a humid atmosphere containing 5% CO2/95% air and were allowed to aggregate, fuse, and form syncytia. The purity of the cytotrophoblast cell population was checked by cytokeratin 7 immunostaining. Syncytium formation was followed by the distribution of desmoplakin and nuclei in cells after fixation and immunostaining [13]. The staining of desmoplakin present at the intercellular boundaries in aggregated cells progressively disappears with syncytium formation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A continuous genomic sequence determined by the Human Genome Project made it possible to examine the genomic structure of the INSL6, INSL4, RLN1, and RLN2 genes located on chromosome band 9p23.3–p24.1. We used the INSL6 (GenBank accession number NM_007179), INSL4 (NM_002195), RLN1 (NM_006911), and RLN2 (NM_134441) cDNA sequences to run a search in the nr (all non-redundant GenBank + EMBL + DDBJ + PDB sequences) and htgs (unfinished high throughput genomic sequences) database with the Gapped basic local alignment search tool (BLAST) program [14]. The four genes, identified on three contiguous overlapping human bacterial artificial chromosomes (BACs) (AL161450, AL133547, and AL135786), were tightly clustered within a 176-kilobase (kb) genomic segment between the JAK2 gene (NM_004972, coding for a protein tyrosine kinase belonging to the Janus kinase family) and the MDS030 gene (NM_018465, coding for an uncharacterized protein expressed in hematopoietic stem/progenitor cells).
Figure 1 depicts the order of the genes (JAK2, INSL6, INSL4, RLN2, RLN1, and MDS030) from telomere to centromere. All the insulin-like growth factor genes, apart from INSL4, are transcribed from their centromeric end to their telomeric end.
|
It is interesting that INSL6 was physically linked to JAK2 in chromosome region 9p23.3–p24.1, as are INSL5 (NM_005478) and JAK1 (NM_002227) in chromosome region 1p31.3, INSL3 (NM_005543) and JAK3 (NM_000215) in chromosome region 19p13.1, and RLN3 (NM_080864) and TYR2 (NM_003331) in chromosome region 19p13.2, which are additional members of the insulin gene superfamily (INSL6, INSL5, INSL3, and RLN3) and the JAK gene family (JAK2, JAK1, JAK3, and TYR2). These findings point to trans regional duplication during evolution of the vertebrate genome.
Identification of Two Novel Alternatively Spliced INSL4 mRNA Species
Gunnersen et al. [7] detected an alternatively spliced mRNA species incorporating an extra exon in the human RLN1 and RLN2 genes, which encodes novel relaxin-like peptides (named RLN1-like and RLN2-like, respectively). We therefore postulated that the human INSL4 gene, like the RLN genes, was also composed of three exons and two introns.
We used the 1933-base pair (bp) genomic sequence of the INSL4 intron to run a search in the dbEST database with the Gapped BLAST program. We detected two human expressed sequence tags (ESTs) (GenBank accession numbers H60640 and H02449), which are composite sequences containing either exon 1 (H60640) or exon 2 (H02449) plus intronic sequences of INSL4. Analysis of these ESTs and PCR products obtained with EST-specific primers identified two alternative forms of INSL4 mRNA, INSL4-like1 and INSL4-like2, generated by alternative splicing (Fig. 2A).
|
The INSL4-like2 mRNA isoform is generated by the inclusion of an extra 149-bp exon, with exon/intron junctions corresponding to the consensus sequences of donor/acceptor splice sites. Splicing-in of the 149-bp exon introduces a reading-frame shift, leading to a novel predicted peptide (INSL4-like2) that would be identical to the INSL4 peptide in the amino-terminal domain but differ in the carboxy-terminal domain, in the same way as relaxin-like peptides [7].
The INSL4-like1 mRNA isoform is generated by splicing with an alternative splice acceptor site located 27 bp downstream of the new exon. The use of this alternative splice acceptor site results in the insertion of nine in-frame amino acids relative to the INSL4-like2 peptide.
Comparison of the carboxy-terminal domains of the putative INSL4-like1, INSL4-like2, RLN1-like, and RLN2-like molecules revealed considerable homology, with the same carboxy-terminal end (Fig. 2B). Database searches using PSI-BLAST [14] revealed no significant homology of these four domains with any other human amino acid sequence.
The expression patterns of the INSL4-like1 and INSL4-like2 mRNA isoforms were determined by real-time quantitative RT-PCR at the mRNA level in a series of normal adult tissues including brain, thymus, placenta, liver, testis, breast, prostate, stomach, colon, skin, leukocytes, adrenal gland, pancreas, and uterus. Strong expression of the INSL4-like1 and INSL4-like2 mRNA isoforms was detected only in placental total RNA. The levels of INSL4-like1 and INSL4-like2 variants in placenta were similar to each other, but were markedly lower (10 times) than the level of the standard INSL4 transcript (data not shown).
Expression of the INSL6, INSL4, and RLN1/2 Genes in Normal Human Adult Tissues
Previous studies of INSL6, INSL4, and RLN1/2 gene expression in human tissues have given somewhat inconsistent results [3, 7–9]. To better understand the regulation of INSL4 gene expression in the insulin-like growth factor gene cluster at 9p23.3–p24.1, we developed real-time quantitative RT-PCR assays to analyze INSL6 and INSL4 alone and RLN1 and RLN2 together in a variety of normal human adult tissues (Table 2).
|
INSL4 was expressed strongly only in placenta. Much lower expression (10000 times lower than in placenta) was detected in thymus, testis, breast, stomach, and uterus. The other tissues contained very little or no detectable INSL4 mRNA.
INSL6 was expressed strongly only in testis. Much lower expression was found in placenta (100 times lower) and in thymus and prostate (1000 times lower). The other tissues contained very little or no detectable INSL6 mRNA.
RLN1 and RLN2 were expressed strongly only in prostate. Much lower expression (1000 times lower) was detected in thymus, placenta, and testis. The other tissues tested contained very little or no detectable RLN1/2 mRNA.
Identification of an HERV-K Element in the 5' Flanking Region of the Human INSL4 Gene
To better understand the placenta-specific expression of the INSL4 gene, we sequenced 4 kb of the 5' flanking region of the INSL4 gene from a P1-derived artificial chromosome (PAC number 75A3) obtained from a human PAC DNA library (UK Human Genome Mapping Project, HGMP). Using RepeatMasker version 2 (available at http://ftp.genome.washington.edu/cgi-bin/RM2), we found an unexpected HERV element inserted very close to the INSL4 gene at position -381, with a sequence similar to the env gene, flanked by two LTRs. Including the two LTRs, the full-length element is 2434 bp in length.
Compared with the full-length 3' LTR (489 bp), the 5' LTR (275 bp) shows a major deletion at its 3' end, including the tRNA binding site. The shared portion of the two LTRs differs by a few deletions, insertions, and nucleotide substitutions, resulting in a homology of 89%.
The presence of multiple stop codons in the env reading frame, and the absence of matches with ESTs in the dbEST database, suggests that the env gene of the HERV is no longer transcriptionally active.
A search of the nr and htgs databases using the HERV genomic DNA as the query identified multiple sequences, including HERV-K(C4) (GenBank accession number X80240) located in the complement C4 gene cluster at 6p21.3 [15]. The most significant sequence hit (GenBank accession number AC010203, map position chromosome 12q22) contained a HERV element (6091 bp long) with sequences similar to gag and env genes, flanked by two LTRs. It is interesting that this HERV element showed identities to multiple ESTs, suggesting that it might be the "master" gene of the INSL4-HERV family (i.e., the HERV family member that remained transcriptionally active).
Finally, we tested three contiguous overlapping human 9p23.3–p24.1 BAC clones (AL161450, AL133547, and AL135786), including the INSL6, INSL4, RLN1, and RLN2 genes, for additional interspersed repeat sequences. No additional HERV elements were detected.
Influence of the HERV 3' LTR on Human INSL4 Promoter Activity
To examine the possible promoter activity of the HERV 3' LTR, we analyzed the capacity of promoter fragments; namely, a 1502 bp fragment including the entire 3' LTR element (fragment -1407/+95) and shorter fragments with 5' to 3' deletions, to drive the expression of the luciferase gene after cloning into the promoterless pGL3-basic vector. The resulting plasmids were transiently transfected into JEG-3 choriocarcinoma cells, and the promoter activity of the constructs was assessed by measuring luciferase activity relative to the transcriptional activity of the construct containing the full-sized 3' LTR (fragment -1407/+95) (Fig. 3). The results clearly indicated that the 3' LTR was important for INSL4 expression in JEG-3 cells. Deletion of the retrovirus-derived promoter portion abolished its activity.
|
Expression of the INSL4, INSL6, RLN1/2 Genes During Human Cytotrophoblast Differentiation In Vitro
Recent studies suggest that HERV elements expressed in the syncytiotrophoblast may be involved in placental morphogenesis [16, 17]. Using an in vitro model of human villous cytotrophoblast differentiation into syncytiotrophoblast [13], we examined the possible role of the INSL4 gene, via its HERV, in human placental morphogenesis. We quantified INSL4, INSL6, and RLN1/2 expression at the mRNA level in matched cytotrophoblasts and syncytiotroblast from eight placentas. Isolated cytotrophoblasts aggregated and fused together in vitro to form a syncytiotrophoblast within 72 h.
We also analyzed the Leptin gene, because its placenta-specific expression may be due the presence of an LTR element in its promoter [18].
As shown in Figure 4, INSL4 expression was up-regulated more than 10-fold in syncytiotrophoblast relative to the corresponding cytotrophoblasts. Similar results were obtained with the Leptin gene (more than 40 times higher). In contrast, INSL6 and RLN1/2 mRNA levels did not change during cytotrophoblast differentiation.
|
Evolutionary Conservation of the INSL4 Gene
The human INSL4 peptide revealed 44% identity to the human RLN1 peptide, 43% identity to the human RLN2 peptide, and 15% identity to the human INSL6 peptide, suggesting that the INSL4 gene likely arose from the common precursor RLN gene (not from INSL6) by a process of gene duplication.
By genomic sequence analysis, we confirmed that Old World monkeys, including the rhesus monkey (M. mulatta) and the Africa green monkey (C. aethiops), as well as New World monkeys, including Goeldi's marmoset (C. goeldii) and prosimians including the black lemur (E. macaco) have only one RLN gene.
The INSL4 gene is probably the only insulin-like growth factor gene to be primate-specific. Indeed, 1) a search of the dbEST database using INSL4 cDNA as the query identified only multiple human expressed sequence tags, whereas the other insulin-like growth factor cDNAs showed multiple murine ESTs; 2) hybridizing fragments using INSL4 cDNA were observed in human and chimpanzee genomic DNA but not in horse or mouse genomic DNA (data not shown); 3) PCR using INSL4 degenerated primers and sequence analysis identified the INSL4 gene in the Old World monkeys (rhesus monkey and Africa green monkey) but not in mouse, horse, prosimians (black lemur), or New World monkeys (marmoset); 4) finally, a search of the nr and htgs databases using the murine RLN cDNA (GenBank accession number Z27088) as the query identified a 215-kb mouse genomic clone (AC093339; 10 unordered contigs) and a 145-kb rat genomic clone (AC096324; 30 unordered contigs) in the "working draft" of the complete murine and rat genomes, which respectively include the mouse and rat genomic DNAs of the unique common precursor RLN gene, but no other significant homology (at the protein or nucleic acid levels) with other known insulin-family nucleic acid or protein sequences. Confirmation of the absence of murine INSL4 must await full sequencing of the genomic clone AC093339.
Finally, by genomic sequence analysis, we identified the HERV element in the rhesus monkey (Old World monkeys), suggesting that HERV insertion into the INSL4 promoter occurs relatively soon after the emergence of this gene (Fig. 5).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Initially, the common precursor insulin-like growth factor gene at 9p23.3–p24.1 arose by trans duplication of an ancestor chromosome region, including one member of the JAK family and one member of the insulin gene family. Indeed, like RLN1, RLN2, INSL4, INSL6, and JAK2 in chromosome region 9p23.3–p24.1, INSL5 and JAK1 are physically linked at 1p31.3, INSL3 and JAK3 at 19p13.1, and RLN3 and TYR2 (a additional member of the JAK gene family) at 19p13.2.
Subsequently, cis duplication at 9p23.3–p24.1, probably mediated by a series of local rearrangements, generated four insulin-like growth factor genes in chromosome region 9p23.3–p24.1. From telomere to centromere, INSL6, INSL4, RLN2, and RLN1 are tightly clustered within a 176-kb genomic segment between the JAK2 gene and the MDS030 gene. We failed to detect other insulin-like growth factor genes within this cluster, or within the 100- to 200-kb sequences at the telomeric and centromeric ends of this cluster, despite systematic searches.
Contrary to the other insulin-like growth factor genes that have homologous murine genes, INSL4 appears to be primate-specific. Indeed, we failed to identify any INSL4 nucleic acid sequences in mouse, horse, and lemur samples despite careful searches, including cross-hybridizing bands, PCR analysis using degenerate INSL4 primers, murine ESTs in the dbEST database, and INSL4 nucleic acid or protein sequences in two genomic clones (AC093339, AC096324) recently revealed by the working draft of the complete murine and rat genomes, and which contain the mouse and rat common precursor RLN genes, respectively.
However, we found INSL4 homologue genes in the rhesus monkey and African green monkey genomes, suggesting that INSL4 emerged after the divergence of New and Old World monkeys (45 million years BP). INSL6 and the common precursor RLN gene are present in the mouse genome [19, 20]. We confirmed the recent duplication event leading to the generation of two RLN genes, after the divergence of Old World monkeys and great apes (25 million years BP). Evans et al. [21] showed that all great apes (chimpanzees, gorillas, and orangutans) have two RLN genes, whereas Old World monkeys (rhesus monkey) have only one RLN gene, and suggested that RLN1 is probably nonfunctional in the orangutan and gorilla.
INSL4 probably arose through duplication of the common precursor RLN gene rather than INSL6. Indeed, the INSL4 peptide sequence is closer to that of RLN1/2 than INSL6. Moreover, we identified extra exons in the INSL4 gene, as occur in RLN genes [7]. We failed to identify such exons in INSL6 despite careful searches, including human intronic ESTs in the dbEST database and intronic homologies with the extra exons of INSL4 and RLN1/2 both at the protein and nucleic acid levels. The conservation of the open reading frames (ORFs) and the homologies between these INSL4-like peptides and relaxin-like peptides suggests that these conserved peptides may have a significant biological role. It is noteworthy that these INSL4-like and relaxin-like molecules lack the A-chain and, thus, conserved cysteines involved in disulfide bonds, which are characteristic of other members of the insulin-related protein family [22].
Surprisingly, we identified a HERV element inserted in the INSL4 promoter. The 3' LTR of this HERV appears to mediate the placenta-specific expression of the INSL4 gene. The HERV element is inserted very close to the INSL4 gene, at position -381, generating a promoter with new functional features. This HERV element is also present in the promoter of the rhesus monkey (M. mulatta) INSL4 homologue gene, suggesting that genomic insertion occurred throughout Old World monkey evolution (between 25 and 45 million years BP), after duplication of the common precursor RLN gene gave rise to the INSL4 gene (Fig. 5). Indeed, we failed to identify a HERV element in the 5' flanking regions of the human RLN1 and RLN2 genes. This is in keeping with results reported by Medstrand and Mager [23], which showed that the oldest HERV, in evolutionary terms, is found in Old World monkey lineages.
This 2434-bp HERV element comprises a classical retroviral structure, except for the truncation of the 5' LTR (including the tRNA binding site) and the absence of the gag and pol genes. This proviral genome could not produce a functional protein, as the env ORF is interrupted by frame shifts and stop codons. Sequence homology with other HERV sequences suggests that this HERV should be included among the HERV-K elements [24]. We also identified a very closely related HERV sequence on chromosome region 12q22 (GenBank accession number AC010203); it possesses sequences similar to those of the gag and env genes, is flanked by two LTRs, and shows similarities to multiple ESTs. This HERV element is transcriptionally active and could thus be the "master" gene of the INSL4-HERV element.
We postulate that the placenta-specific expression of the INSL4 gene is mediated by the 3' LTR of the HERV element. Several published findings support this hypothesis. First, several HERV families, especially HERV-K, HERV-W, HERV-F, HERV-R (ERV-3), and HERV-H (RTLV-H), are mainly expressed in the placenta [16, 25–29]. In addition, Mi et al. [16] and Blond et al. [17] identified a HERV-W coding for an envelope protein (syncytin) that has fusogenic activity and therefore could be involved in syncytiotrophoblast formation. Second, alternative chimeric transcripts with placenta-specific expression can be created by HERV insertion into various gene promoters (the pleiotrophin and endothelin B receptor genes) [30, 31]. Third, isolated HERV LTR possess bidirectional promoter activity as well as both young LTR (emergence in the human genome 5 million years BP) and old LTR (emergence in the human genome 45 million years BP) show similar promoter activity [32]. Fourth, HERV LTRs contain elements that can regulate the transcription of neighboring genes. [18, 33]. For example, Bi et al. [18] identified a MER11 repetitive element in the Leptin promoter that mediates its placental expression. Because the MER11 elements are not present in the murine genome, this might explain why the human but not the murine placenta expresses leptin. We found that INSL4 is the only member of the insulin-like growth factor gene cluster at 9p23.3–p24.1 to be specifically expressed in the placenta, and the only gene to have a HERV element inserted in its promoter.
Our results strongly support the conclusion that the 3' LTR promotes human INSL4 expression in JEG-3 cells. Indeed, transfection experiments demonstrated that the LTR was capable of directing luciferase expression in JEG-3 cells, and deletion of the retrovirus-derived promoter portion abolished its activity. HERVs (or isolated LTRs) may be present in the 5' flanking region of other human genes with placenta-specific expression.
Several reports suggest a role for HERV in the creation of the syncytiotrophoblast layer [16, 26, 27]. HERV could mediate neighboring gene function. We studied INSL4 expression in an in vitro model of human villous cytotrophoblast differentiation into the syncytiotrophoblast and found that expression of INSL4 (but not that of INSL6, RLN1, or RLN2) was up regulated more than 10-fold in the syncytiotrophoblast relative to the corresponding cytotrophoblasts; in the same way, the Leptin gene, the placental expression of which is under control of a MER11 repeat element [18], was up-regulated more than 40-fold.
In conclusion, our results confirm that ancient retroviral infection may have been a major event in primate evolution, especially the functional evolution of the human placenta. The human hemochorial placenta is considered more invasive than the mouse counterpart in which trophoblast and maternal cells are rather demarcated from each other [34]. Some genes (such as INSL4, Leptin, Pleiotrophin, Endothelin B receptor, and Syncytin) with new placenta-specific expression in primate species via HERV element integration could contribute to the aggressive growth phenotype of the normal primate placenta.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Michel Vidaud, Laboratoire de Génétique Moléculaire—UPRES EA 3618, Faculté des Sciences Pharmaceutiques et Biologiques, Université René Descartes—Paris V, 4 Avenue de l'Observatoire, F-75006 Paris; France. FAX: 33 1 44 07 17 54; mvidaud{at}teaser.fr
3 These authors contributed equally to this work
Received: 14 August 2002.
First decision: 5 September 2002.
Accepted: 30 October 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. A. Bathgate, R. Ivell, B. M. Sanborn, O. D. Sherwood, and R. J. Summers International Union of Pharmacology LVII: Recommendations for the Nomenclature of Receptors for Relaxin Family Peptides. Pharmacol. Rev., March 1, 2006; 58(1): 7 - 31. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Millar, N. Streiner, L. Webster, S. Yamamoto, R. Okabe, T. Kawamata, J. Shimoda, E. Bullesbach, C. Schwabe, and G. Bryant-Greenwood Early Placental Insulin-Like Protein (INSL4 or EPIL) in Placental and Fetal Membrane Growth Biol Reprod, October 1, 2005; 73(4): 695 - 702. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Dunlap, M. Palmarini, D. L. Adelson, and T. E. Spencer Sheep Endogenous Betaretroviruses (enJSRVs) and the Hyaluronidase 2 (HYAL2) Receptor in the Ovine Uterus and Conceptus Biol Reprod, August 1, 2005; 73(2): 271 - 279. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Prudhomme, G. Oriol, and F. Mallet A Retroviral Promoter and a Cellular Enhancer Define a Bipartite Element Which Controls env ERVWE1 Placental Expression J. Virol., November 15, 2004; 78(22): 12157 - 12168. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Bannert and R. Kurth Retroelements and the human genome: New perspectives on an old relation PNAS, October 5, 2004; 101(suppl_2): 14572 - 14579. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Dunn, P. Medstrand, and D. L. Mager An endogenous retroviral long terminal repeat is the dominant promoter for human {beta}1,3-galactosyltransferase 5 in the colon PNAS, October 28, 2003; 100(22): 12841 - 12846. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
