HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 2, 1148-1157, January 11, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/2/1148    most recent M107882200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Bathgate, R. A. D. Articles by Tregear, G. W. Search for Related Content PubMed PubMed Citation Articles by Bathgate, R. A. D. Articles by Tregear, G. W. Social Bookmarking                      What's this?

Human Relaxin Gene 3 (H3) and the Equivalent Mouse Relaxin (M3) Gene

NOVEL MEMBERS OF THE RELAXIN PEPTIDE FAMILY^*

Ross A. D. Bathgate, Chrishan S. Samuel, Tanya C. D. Burazin^§, Sharon Layfield, Antonia A. Claasz, Irna Grace T. Reytomas, Nicola F. Dawson, Chongxin Zhao, Courtney Bond^¶, Roger J. Summers^¶, Laura J. Parry, John D. Wade, and Geoffrey W. Tregear

From the Howard Florey Institute of Experimental Physiology and Medicine, University of Melbourne, Victoria 3010 and the ^¶ Department of Pharmacology, Monash University, Clayton, Victoria 3800, Australia

Received for publication, August 16, 2001, and in revised form, October 4, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We have identified a novel human relaxin gene, designated H3 relaxin, and an equivalent relaxin gene in the mouse from the Celera Genomics data base. Both genes encode a putative prohormone sequence incorporating the classic two-chain, three cysteine-bonded structure of the relaxin/insulin family and, importantly, contain the RXXXRXX(I/V) motif in the B-chain that is essential for relaxin receptor binding. A peptide derived from the likely proteolytic processing of the H3 relaxin prohormone sequence was synthesized and found to possess relaxin activity in bioassays utilizing the human monocytic cell line, THP-1, that expresses the relaxin receptor. The expression of this novel relaxin gene was studied in mouse tissues using RT-PCR, where transcripts were identified with a pattern of expression distinct from that of the previously characterized mouse relaxin. The highest levels of expression were found in the brain, whereas significant expression was also observed in the spleen, thymus, lung, and ovary. Northern blotting demonstrated an ~1.2-kb transcript present in mouse brain poly(A) RNA but not in other tissues. These data, together with the localization of transcripts in the pars ventromedialis of the dorsal tegmental nucleus of C57BLK6J mouse brain by in situ hybridization histochemistry, suggest a new role for relaxin in neuropeptide signaling processes. Together, these studies describe a third member of the human relaxin family and its equivalent in the mouse.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Relaxin is a 6-kDa polypeptide hormone that is secreted by the ovary into the peripheral circulation in highest amounts during pregnancy and has a number of functions in mammals that are generally associated with female reproductive tract physiology (1). To date, only one relaxin gene has been characterized in most mammalian species, with the exception of the human where two separate genes have been described, designated H1 (2) and H2 (3) relaxin. The peptide encoded by the H2 gene is the major stored and circulating form in the human (4). H1 relaxin expression is restricted to the decidua, placenta, and prostate (5); however, the H1 peptide has similar biological activity to that of H2 relaxin in a rat atrial bioassay (6). The actions of relaxin include an ability to inhibit myometrial contractions, to stimulate remodeling of the connective tissue, and to induce softening of the tissues of the birth canal. Additionally, relaxin increases growth and differentiation of the mammary gland and nipple and induces the breakdown of collagen, one of the main components of connective tissue. Relaxin decreases collagen synthesis and increases the release of collagenases (7). These findings were recently confirmed by the establishment of the relaxin gene-knockout mouse (8), which exhibited a number of phenotypic properties associated with pregnancy. Female mice lacking a functionally active relaxin gene failed to relax and elongate the interpubic ligament of the pubic symphysis and could not suckle their pups, who in turn died within 24 h unless cross-fostered to relaxin wild type or relaxin heterozygous foster mothers.

Evidence has accumulated to suggest that relaxin is more than a hormone of pregnancy and acts on cells and tissues other than those of the female reproductive system. Relaxin causes a widening of blood vessels (vasodilatation) in the kidney, mesocecum, lung, and peripheral vasculature, which leads to increased blood flow or perfusion rates in these tissues (9). It also stimulates an increase in heart rate and coronary blood flow and increases both glomerular filtration rate and renal plasma flow (9). The brain is another target tissue for relaxin, where the peptide has been shown to bind to receptors (10, 11) in the circumventricular organs to affect blood pressure and drinking (12-14). Finally, binding sites have been identified in the prostate gland, implicating increased relaxin production with increased prostatic hyperplasia in older mammals (15).

Although the progressively increasing actions of relaxin appear to now affect many tissues and cells outside the female reproductive tract, to date only limited sources of relaxin production have been reported. Within the female reproductive tract, relaxin is primarily produced by the corpus luteum (CL)¹ in both pregnant and nonpregnant mammalian species, although it attains the highest plasma levels during pregnancy (1). Depending on the species, relaxin is also produced by the decidua, placenta, and endometrium and in thecal and granulosa cells (16) of the ovarian follicle (1). Another minor source of relaxin production in the female guinea pig is the mammary gland (17), whereas the prostate appears to be the major source of relaxin in the male (18, 19). Immunoreactive relaxin levels and binding sites for relaxin have also been detected in many organs and cells, such as in the heart and brain (1). Even though relaxin is clearly a paracrine factor, the known sources of relaxin production do not correlate with all the sites of expression. Hence, it is quite possible that other relaxin-like molecules exist, which may contribute to the functions of relaxin described to date and possible additional new functions.

Relaxin sequences from many species have been characterized, and it is now well established that relaxin is structurally related to the insulin/IGF family of peptide hormones (1). All relaxins have a two-chain structure comprising an A- and B-peptide chain linked by disulfide bonds with an intra-chain disulfide bond in the A-chain, analogous to that of insulin. Recently, two new insulin-like members of this peptide family have been identified in the expressed sequence tag (EST) data base (20, 21). The availability of the human genome sequence provides a unique opportunity to discover other novel relaxin and relaxin-related genes. This study describes the use of the Celera Discovery System and Celera Genomics associated data bases to identify a new human and mouse relaxin gene.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Sequence Identification-- The Celera Discovery System and Celera Genomics associated data bases were searched for novel relaxin homologues. Six sequences classified by computational annotation were identified in the biomolecular library, designated as protein subfamily of signaling molecules, protein/peptide hormone, relaxin-related. Two of these sequences corresponded to the previously characterized H1 and H2 relaxins, and three others corresponded to human insulin 4 (INSL4), insulin 5 (INSL5), and insulin 6 (INSL6). The final sequence encoded a peptide with relatively low homology to H1 and H2 relaxin but containing the key structural elements of relaxin peptides. This sequence (Celera accession number CSN002; www.celera.com/publicationlibrary) did not correspond to any sequences in the public data base but had significant homology to a rat EST clone. A subsequent BlastN search of the incomplete mouse genome data base in Celera, using the human sequence, identified genomic fragments with high homology to the human gene. A complete coding sequence was assembled from multiple genome fragments that displayed a similar genomic structure to the human gene. Subsequently, the availability of the complete mouse genome enabled confirmation of this coding sequence and assembly of the entire gene locus. Like the human gene, this sequence (Celera accession number CSN003) did not correspond to any sequence in the public data base but aligned with the same rat EST clone. Phylogenetic analysis was conducted on the human and mouse sequences compared with mouse relaxin and human members of the relaxin/insulin/IGF family using the Megalign program in Lasergene (DNASTAR Inc., Madison, WI). The peptides and their appropriate GenBank^TM accession numbers are as follows: mouse relaxin (Z27088), human insulin (XM 028180), INSL3 (XM 009297), INSL4 (L 34838), INSL5 (XM 001861), INSL6 (NM 007179), IGF1 (NM 000618), and IGF2 (NM 000612).

Human Relaxin (H3) Studies

Solid Phase Synthesis-- A putative peptide sequence encoded by the H3 gene was assembled by solid phase synthesis procedures based on the predicted signal peptide and proteolytic enzyme cleavage sites between the signal peptide and the B-chain, and the B/C- and C/A-chain junctions of the H3 relaxin prohormone (see "Results" for details). For ease of synthesis we chose to prepare the A- and B-peptides as their C-terminal-amide derivatives. Selectively S-protected A- and B-chains were synthesized on a 0.1 mmol scale by the continuous flow Fmoc (N-(9-fluorenyl)methoxycarbonyl) solid phase method as described previously (22). Selective S-protection was afforded for the following cysteine residues: trityl (Trt) for A^10,15 and B²², tert-butyl for A²⁴, and acetamidomethyl for A¹¹ and B¹⁰ (see Fig. 2A for numbering of amino acid residues).

On completion of the syntheses, the S-protected A- and B-chains were cleaved from the solid supports and simultaneously side chain-deprotected by treatment with trifluoroacetic acid in the presence of scavengers. Selective disulfide bond formation was achieved essentially as described for the synthesis of bombyxin (23).

Peptide Characterization-- Peptides were quantitated by duplicate amino acid analysis of 24-h acid hydrolysates on a GBC automatic analyzer (Melbourne, Australia). MALDI-TOF mass spectrometry (MS) was performed in the linear mode at 19.5 kV on a Bruker Biflex instrument (Bremen, Germany) equipped with delayed ion extraction.

Other Relaxin and Insulin Peptides-- Human INSL3 was synthesized using the same methodology used for ovine INSL3 (22) and was characterized by MS and amino acid analysis as outlined above. H1 relaxin was synthesized previously (24); recombinant H2 relaxin was a gift from the Connetics Corp. (Palo Alto, CA), and bovine insulin was purchased from Roche Molecular Biochemicals.

THP-1 Cell Bioassay-- The ability of H3 relaxin to induce cAMP production in the human monocytic cell line (THP-1) was compared with H1 and H2 relaxin following the procedure of Parsell and colleagues (25), with the following modifications: THP-1 cells that had been viability tested using trypan blue were resuspended in media and transferred to a 96-well plate at a density of 60,000 cells/well. Peptides (H1, H2, H3 relaxin, human INSL3, and bovine insulin) were added to the wells together with 1 µM forskolin and 50 µM isobutylmethylxanthine in RPMI media and incubated at 37 °C for 30 min. The plate was then briefly centrifuged; the media were removed and the cells resuspended in lysis buffer. cAMP levels were measured in the lysates using the cAMP Biotrak EIA system (Amersham Biosciences). The results are expressed as the maximum relaxin response (percent) in comparison to the maximum stimulation of cAMP achieved with H2 relaxin. Data represent the mean ± S.E. of three experiments performed in quadruplicate and are plotted using PRISM (Graphpad Inc., San Diego, CA).

THP-1 Cell Binding Assay-- THP-1 cells were spun down and resuspended in binding buffer (20 mM HEPES, 50 mM NaCl, 1.5 mM CaCl₂, 1% bovine serum albumin, 0.1 mg/ml lysine, 0.01% NaN₄, pH 7.5) (25) to give 2 × 10⁶ cells/well in a 96-well plate. The cells were incubated in binding buffer with ³³P-labeled H2 (B33) relaxin (100 pM), labeled as previously described (11) at 25 °C for 90 min in the absence or presence of increasing concentrations of unlabeled H1, H2 and H3 relaxin (100 pM to 30 nM)). Nonspecific binding was defined with H2 relaxin (1 µM). Cells were harvested using a Packard 96-well plate cell harvester, and Whatman GF/C glass fiber filters were treated with 0.5% polyethyleneimine. The filters were washed three times with modified binding buffer (20 mM HEPES, 50 mM NaCl, 1.5 mM CaCl₂), dried in a 37 °C oven, and the radioactivity counted by liquid scintillation spectrometry (TopCount^TM, Packard Instrument Co.).

Antibody Cross-reactivity-- The ability of well characterized human relaxin antibodies to recognize H3 relaxin was tested in comparison to H1 and H2 relaxins by radioimmunoassay. Briefly, goat anti-H2 relaxin (26) was coated onto 96-well ELISA plates (Disposable Products, Adelaide, Australia) at a dilution of 1:1000 with 0.05 M sodium carbonate buffer at 4 °C overnight. After washing twice with PBS-T (phosphate buffered saline; 0.05% Tween 20, pH 7.4), dilutions of human relaxin peptides dissolved in 50 µl of assay buffer (1% bovine serum albumin in PBS-T) were added together with 50,000 cpm ¹²⁵I-labeled relaxin in 50 µl of assay buffer. H2 relaxin was ¹²⁵I-labeled and purified by high pressure liquid chromatography (27). After an overnight incubation at 4 °C, the plates were washed twice with PBS-T. The antibody-bound ¹²⁵I-labeled H2 relaxin was collected by the addition of 1 M NaOH and decanted into tubes for counting on a Packard 5010 gamma counter (Packard Instrument Co.). Experiments were performed at least twice and similar results obtained. Data were plotted as the mean ± S.E. from one representative experiment performed in triplicate and plotted using PRISM.

Mouse Relaxin (M3) Studies

Animals-- All male and female mice used in these studies were age-matched and had the same background (C57BLK6J). Animals were housed in a controlled environment and maintained on a 14-h light, 10-h dark schedule with access to rodent lab chow (Barastock Stockfeeds, Melbourne, Australia) and water. Female mice (3.5 months old) were mated, and pregnancy was timed from the identification of the vaginal plug. At day 7.5, 10.5, and 18.5 of pregnancy, mice were sacrificed for tissue collection. Tissues were also collected from nonpregnant female and male mice (4 months old). These experiments were approved by the Howard Florey Institute's Animal Experimental Ethics Committee, which adheres to the Australian Code of Practice for the care and use of laboratory animals for scientific purposes.

Tissue Collection-- Animals were killed with an overdose of Isofluorane (Abbott Australasia Pty. Ltd., Sydney, Australia). The brain, heart, thymus, spleen, lung, liver, kidneys, skin, and gut were collected along with the reproductive organs from female (ovary, endometrium, myometrium, cervix, vagina; n = 2 for each pregnancy stage) and male (testes, epididymis, prostate; n = 3) mice. From additional animals, male brains (n = 3) were dissected into specific regions including the hypothalamus, cortex, hippocampus, thalamus, medulla, and cerebellum and immediately placed in liquid nitrogen and stored at 80 °C until used for RNA preparation. Female brains (n = 3) were collected and immediately frozen over dry ice for in situ hybridization histochemistry (28). Human CL from women in early pregnancy undergoing surgery for ectopic pregnancies were utilized with the approval of the Howard Florey Institute Human Ethics Committee and the written consent of the patients.

Tissue RNA/DNA Extraction and RT-PCR-- Human genomic DNA was extracted from human CL using standard protocols (29). Human CL and mouse tissues were finely diced in the presence of liquid nitrogen and immediately homogenized with RNAWiz reagent (Ambion Inc., Austin, TX), and the RNA was extracted according to the manufacturer's instructions. Total RNA (5 µg) from each sample was used for the reverse transcription (RT) reaction, which was performed using the Superscript II RT-PCR kit (Invitrogen) in a 20-µl volume according to the manufacturer's instructions. A 50-µl reaction containing 100 ng of primers and 150 ng of the cDNA template was used for all PCRs. Mouse tissues were screened for M3 relaxin expression using specific forward (5'-TGCGGAGGCTCACGATGGCGC-3') and reverse (5'-GACAGCAGCTTGCAGGCACGG-3') primers, which generated a 319-bp product. Mouse relaxin (M1) expression was determined using a specific forward (5'-GTGAATATGCCCGTGAATTGATC-3') and reverse (5'-AGCGTCGTATCGAAAGGCTCT-3') primer based on the published sequence (30), generating a 150-bp product. Human CL cDNA was used in RT-PCRs with specific primers for H3 relaxin, forward 1 (5'-ACGTTCAAAGCGTCTCCGTCC-3'), forward 2 (5'-CGGTGGAGACGATCAGACATC-3'), and reverse (5' ATGGCAGGACTGGGGCATTGG-3'), generating products of 504- and 310-bp for forward 1/reverse and forward 2/reverse, respectively. All primer combinations crossed the single introns in the mouse and human relaxin sequences, respectively, so as to control for genomic DNA contamination. In all experiments GAPDH forward (5'-TGATGACATCAAGAAGGTGG-3') and reverse (5'-TTTCTTACTCCTTGGAGGCC-3') primers generating a product of 246 bp were used in separate PCRs to control for quality and equivalent loading of the cDNA. M3 relaxin expression by RT-PCR was performed on cDNA samples extracted from at least two animals, although the results from only one representative experiment are shown. The mouse PCRs were completed in a PerkinElmer Life Sciences Gene Amplifier using the following (touch down) annealing temperatures: 64 (2 cycles), 63 (2 cycles), 62 (2 cycles), 61 (2 cycles), and 60 °C (32 cycles). H3 relaxin expression in human CL cDNA was performed by RT-PCR at the following annealing temperatures: 60 (2 cycles), 59 (2 cycles), 58 (2 cycles), 57 (2 cycles), and 56 °C (32 cycles). Aliquots of the PCR products were electrophoresed on 2% (w/v) agarose gels stained with ethidium bromide and photographed. Mouse tissue samples were transferred to Hybond NX membranes (Amersham Biosciences) for Southern blot analysis.

An additional PCR was performed using mouse brain and ovarian cDNA using the reverse M3 primer (above) and a forward primer from in front of the ATG start codon (5'-GGGTCGCAGGCATCTCAACTG-3'). The resulting product contained the full coding sequence and therefore confirmed the Celera-based sequence. PCR was performed as above but with the following annealing temperatures: 60 (2 cycles), 59 (2 cycles), 58 (2 cycles), 57 (2 cycles), and 56 °C (32 cycles). To generate a specific H3 relaxin cDNA probe for ³²P labeling and to utilize it for subsequent probing of a human multitissue array, RT-PCR was performed on human genomic DNA (50 ng). Specific forward (5'-CGGATGCAGATGCTGATGAAG-3') and reverse (5'-GTGCCTGAGCCCACAGTGCCT-3') primers from the exon II sequence of the H3 relaxin gene were used at the following annealing temperatures: 60 (2 cycles), 59 (2 cycles), 58 °C (2 cycles), 56 (2 cycles), and 54 °C (32 cycles). These products as well as the mouse PCR products described above were separated on 2% agarose gels. Bands were detected of the expected size under UV light (mouse 319 and 478 bp; human 374 bp), excised, and eluted from the gel using the Ultraclean TM 15 DNA purification kit (Geneworks Pty. Ltd., Adelaide, Australia). The bands were subsequently subcloned into the pGEM-T vector (Promega, Madison, WI), and multiple subclones were then sequenced on both strands using the ABI PRISM 377 automatic DNA sequencer, according to the manufacturer's instructions (Applied Biosystems, Melbourne, Australia).

Southern Blot Analysis-- PCR products on membranes were hybridized against specific internal oligonucleotide primers for the M1 relaxin (5'-CAAGCAGAGCTGGCTCCTCCTGGCTCAAAGCCAATCTTC-3') and M3 relaxin (5'-AATTTGGCTCTTGCTACAGCCCCACTCGCAGCAACTGCT-3') cDNA sequences, which had been labeled using T4 polynucleotide kinase and [-³²P]ATP. Hybridization was performed at 55 °C overnight in 5× SSC (1× SSC: 0.15 M NaCl, 15 mM sodium citrate, pH 7), 5× Denhardt's, 1% SDS, and 100 µg/ml sonicated herring sperm. Membranes were washed three times for 5 min in 2× SSC, 0.1% SDS at room temperature followed by a 30-min wash at 55 °C in 0.1× SSC, 0.1% SDS, before being exposed to BioMAX MR film (Eastman Kodak Co.) for 24 h at room temperature.

Northern Blot Analysis-- To examine further the expression of M3 relaxin compared with M1 relaxin mRNA, total RNA (5-25 µg) from the heart, brain, lung, thymus, and spleen of male mice and ovary, endometrium, myometrium, cervix, and vagina of female mice pooled from day 7.5, 10.5, and 18.5 of pregnancy were run on standard MOPS/formaldehyde gels. RNA was then transferred to optimized Hybond-NX membranes and probed for M3 and M1 relaxin and GAPDH mRNA with ³²P-labeled probes corresponding to the PCR products generated by specific primers (see above). These products were labeled with [-³²P]dCTP using the specific reverse primers (above) and T7 polymerase as described previously (31). The membrane was hybridized at 65 °C overnight in 0.25 M NaH₂PO₄, pH 7.2, 1 mM EDTA, 20% SDS, followed by three washes for 5 min in 2× SSC, 0.1% SDS at room temperature, and finally a 30-min wash at 65 °C in 0.1× SSC, 0.1% SDS. Membranes were first exposed to a PhosphorImager plate for 48 h at room temperature before being analyzed in a FujiX 2000 PhosphorImager (Fuji Photo Co., Japan) and then exposed to BioMAX MS film (Integrated Sciences, Melbourne, Australia) together with a Hyperscreen (Amersham Biosciences) at 80 °C. In a separate experiment, total RNA (200 µg) from the male brain was purified to poly(A) RNA using an mRNA purification kit (Amersham Biosciences), and Northern blotting together with total RNA from day 18.5 ovary (5 µg) was performed as described above. A human multiple tissue expression array (CLONTECH laboratories, Palo Alto, CA) was hybridized with a ³²P-labeled H3 relaxin-specific probe according to the manufacturer's recommendations. The 374-bp fragment of the H3 relaxin sequence isolated from genomic DNA was labeled with [-³²P]dCTP using the H3 relaxin-specific reverse primer (described above) and T7 polymerase (31). The membrane was exposed to a PhosphorImager plate and BioMAX film as described above.

In Situ Hybridization Histochemistry-- Coronal sections (14 µm) were cut on a cryostat at 16 °C and mounted on silane-coated slides. Sections were delipidated in chloroform for 10 min, rinsed, and stored in 100% ethanol at 4 °C. Three oligonucleotides (39-mers), 5'-GGTGGTCTGTATTGGCTTCTCCATCAGCGAAGAAGTCCC-3', 5'-AATTTGGCTCTTGCTACAGCCCCACTCGCACGAACTGCT-3', and 5'-TAAGGAGACAGTGGACCCCTTGGTGCCTCGCCTGTAGGA-3', of the M3 relaxin mRNA sequence, and three oligonucleotides, 5'- GCACATCCGAATGAATCCGTCCATCCACTCCTCCGAGAC-3', 5'-CAAGCAGAGCTGGCTCCTCCTGGCTCAAAGCCAATCTTC-3', and 5'-GTTGTAGCTCTGGGAGCGAGGCCTGAGCCTCAGACAGTA-3', of the previously known M1 relaxin sequence (30) were prepared commercially (Geneworks Pty. Ltd.). Probes were labeled with [-³⁵S]dATP (1200 Ci/mmol; Amrad, Melbourne, Australia) to a specific activity of 1 × 10⁹ dpm/µg using terminal deoxynucleotidyltransferase (Roche Molecular Biochemicals (32)). Screening of the sequences used against gene sequence data bases (Celera, EMBL. and GenBank^TM; NCBI/National Institutes of Health Blast Service) revealed homology only with the appropriate M1 and M3 relaxin mRNAs.

Sections were incubated overnight at 42 °C with multiple ³⁵S-labeled probes (30 fmol each probe/slide) in hybridization buffer containing 50% formamide, 4× SSC, 10% dextran sulfate, and 0.2 M dithiothreitol. Slides were washed in 1× SSC at 55 °C for 1 h, rinsed in 0.1× SSC, and then dehydrated before being apposed to Kodak BioMAX MR for 10 days.

The authenticity of the hybridization was confirmed by the demonstration that the signal could be successfully blocked in all areas by the addition of a 100-fold excess of unlabeled probes to the hybridization buffer, except those that corresponded to nonspecific or background hybridization (data not shown). In addition, three oligonucleotide probes were used that were complementary to different, nonoverlapping regions of the M3 relaxin gene sequence.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Characterization of H3 and M3 Relaxin Gene Sequences-- Searches of the Celera Discovery System and Celera Genomics associated data bases resulted in the identification of a novel human relaxin gene. A novel mouse relaxin gene was also discovered by using the human sequence to search by using BlastN the incomplete mouse data base. The mouse sequence was initially constructed from the incomplete mouse data base by assembly of aligned 500-700-bp fragments of the mouse genome. Subsequently, the fully annotated mouse sequence has become available in the Celera data base and has allowed the confirmation of the gene sequence including determination of the exact intron size (1446 bp). Furthermore, the mouse coding region sequence has been confirmed by sequencing of multiple subclones of RT-PCR products from the mouse ovary and brain (see below). The human sequence has been designated H3 relaxin, for human gene 3, as it is the third relaxin gene to be discovered in the human. We have designated the mouse sequence as M3 relaxin for consistency, although there is currently only one other relaxin gene identified in the mouse (30).

Both novel relaxin sequences contain features representative of functional genes (Fig. 1, A, human, and B, mouse). Each contains a putative TATA box for initiation of transcription 65 and 59 bp upstream of putative ATG start codons for human and mouse, respectively. A polyadenylation signal is present in the 3'-untranslated region of both genes, in a position 582 and 448 bp downstream from an in-frame TAG stop codon for the human and mouse genes, respectively. A single intron interrupts the coding region in an identical position in the sequence of both genes, corresponding to a similar position to that of other relaxin and insulin family members (2, 30, 33). The H3 relaxin gene is localized on chromosome 19 at 19p13.3, whereas the mouse gene is located on chromosome 8 at 8C2. The derived coding regions of the H3 and M3 relaxin genes were 142 and 141 amino acids, respectively.

View larger version (50K): [in this window] [in a new window]   Fig. 1.   Assembled DNA sequence of the H3 (A) and M3 (B) genes. Start and stop codons as well as predicted TATA boxes and polyadenylation sequences are underlined. The positions of the putative signal peptide and B-, C-, and A-chain peptide sequences are indicated by arrows. A- and B-chain sequences are boxed, and the residues implicated in relaxin receptor binding are shaded. The intron sequence, which is at an identical position in the C-chain in both sequences, is indicated by lowercase letters, and the exact size of the intron is marked. These data were generated through use of the Celera Discovery System and Celera Genomics associated data bases.

The human sequence was characterized within the Celera Discovery System data base as relaxin-related, based on the homology of the B-chain peptide motif to relaxin. Specifically, the cysteine residues necessary for disulfide bond formation are retained in the correct positions, together with conserved glycine residues necessary for flexibility around the cysteine linkages (34). Most importantly, the residues demonstrated to be essential for relaxin receptor binding in the core of the B-chain (RXXXRXXI) (35) have been retained in both the human and mouse sequences. Therefore, although the human sequence most closely resembles the hINSL5 peptide sequence on direct amino acid homology, the presence of this binding motif indicates that the peptide is more like a relaxin peptide. Interestingly, the M3 relaxin A-chain conforms to the cysteine pattern of family members, whereas the previously characterized M1 relaxin sequence contains an extra tyrosine residue before the final cysteine residue (Fig. 2A).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Sequence comparisons of H3 and M3 relaxin with other relaxin and insulin family members. Alignments of A- and B-chain sequences from H3 and M3 relaxin with other human and mouse relaxin sequences (A). The consensus sequences are boxed. Cons 1-3, consensus sequence between human relaxins 1-3. Cons 3, consensus sequence between H3 and M3 relaxin for the B-chain and H3, R3, and M3 relaxin for the A-chain. Cons Mouse, consensus sequence between M1 and M3 relaxin. The rat sequence is derived from an EST clone (see "Results" for details). + denotes a conservative substitution; . denotes no homology. B, phylogenetic tree of evolution of H3 and M3 relaxin full-length sequences with human sequences of the relaxin/insulin/IGF superfamily.

The H3 and M3 sequences share greater than 70% homology in the coding region at the nucleotide level. However, the homology is most striking in the derived amino acid sequence. Both derived pro-hormone sequences contain a typical signal sequence after the ATG start codon which is likely to be cleaved at an identical position between alanine and arginine in both the human and mouse peptides (36). The arginine-arginine pair of basic amino acids at the B/C junction found with other members of the relaxin family strongly suggests cleavage between tryptophan and arginine. Similarly, cleavage at the C/A junction is most likely to occur between the arginine and aspartic acid as indicated in Fig. 1, A and B, as this corresponds to a weak furin (proprotein convertase) cleavage site (37). Therefore, we predict that both H3 and M3 relaxins comprise a B-chain of 27 amino acids, a C-peptide of 66 amino acids, and an A-chain of 24 amino acids.

A comparison of the A- and B-chain sequences of H3 and M3 relaxin with H1, H2, and M1 relaxin is outlined in Fig. 2A. There are only two amino acid differences in both the A- and B-chains between the M3 and H3 sequences, of which three of these changes are conserved. In contrast, the homology between M1 and H2 relaxin is only 42 and 45% in the A- and B-chains, respectively. Furthermore, other than the key core elements in the B-chain and the key structural elements in the A-chain, there is very little homology between H2 and H3 relaxin and between M1 and M3 relaxin. Interestingly, H3 and M3 relaxin show high homology of the C-peptide domain (73%), compared with less than 20% homology in this region of other insulin/relaxin family members. The C-peptide lengths of H3 and M3 relaxin are 65 and 66 amino acids, respectively, and are much shorter than that of other relaxins (102 amino acids for H1 and H2 and 99 amino acids for M1 relaxin). The C-peptide chain length and sequence homology is most similar to INSL5 (24%).

The full-length amino acid sequences of the two genes were aligned to other members of the insulin/relaxin family, and a phylogenetic tree was generated (Fig. 2B). The H3 and M3 relaxin sequences are grouped under a separate branch, indicating that the evolution of these particular relaxins diverged from other relaxins early in evolution. This was also the case for INSL5 within this analysis, which interestingly shares closest primary structural similarity to H3 relaxin.

Significant homology at the nucleotide level to any sequences in the public data bases were not found, except for two rat EST clones. These two clones (GenBank^TM accession numbers AW 521175 and AW 523625) were identical partial sequences from a rat brain subtracted cDNA library (Research Genetics, Huntsville, AL). The predicted amino acid sequences from these partial clones encode a sequence including all of the A-chain of the rat (R3) relaxin homologue (Fig. 2A) as well as three amino acids (GRR) from the C-peptide chain. This sequence is identical to the M3 relaxin sequence, apart from the substitution of arginine for serine in the C-peptide chain.

Peptide Synthesis-- H3 relaxin was prepared by solid phase synthesis in low overall yield (0.7%). MALDI-TOF MS showed a single product with an MH⁺ of 5,494.7 (theoretical value, 5,497.5). Amino acid analysis also confirmed its correct composition.

Demonstration of Relaxin Activity of Synthetic H3 Relaxin-- Synthetic H3 relaxin C-terminal amide derivatives were tested for relaxin activity in a relaxin receptor expressing cell line, THP-1 (25). H2 relaxin produces a dose-dependent increase in cAMP production from these cells (Fig. 3A). Synthetic H3 relaxin also stimulated a dose-dependent increase in cAMP (pEC₅₀ = 8.68 ± 0.08 (2.11 nM); n = 3), albeit with slightly lower activity than H1 (pEC₅₀ = 9.10 ± 0.05 (0.794 nM); n = 3) and H2 (pEC₅₀ = 9.67 ± 0.11 (0.214 nM); n = 3) relaxin. The specificity of this response was demonstrated by the inability of bovine insulin (bINSL) or human insulin 3 (hINSL3) to stimulate cAMP responses at doses up to 1 µM.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Bioactivity of H3 compared with H1 and H2 relaxin in a human relaxin receptor expressing cell line. A, cAMP accumulation in THP-1 cells upon stimulation with peptides. Data are expressed as mean ± S.E. of the maximum response (%) to H2 relaxin (n = 3). The response to bovine insulin (bINSL) and human INSL3 (hINSL3) are also shown to highlight the specificity of the assay. H1, H2, and H3, human 1, 2, and 3 relaxin, respectively. B, the ability of H1 (n = 7), H2 (n = 11), and H3 (n = 3) relaxin peptides to compete for 33P-labeled H2 relaxin (B33) binding to THP-1 cells. Data are expressed as mean ± S.E. of the specific binding (%).

Synthetic H3 relaxin was also tested for its ability to compete for ³³P-labeled H2 relaxin binding to relaxin-binding sites in THP-1 cells (Fig. 3B), with an affinity (pK_i = 7.5 ± 0.16; n = 3) lower than that of H2 (pK_i = 8.74 ± 0.11; n = 11) and H1 (pK_i = 9.0 ± 0.11; n = 7) relaxin. Nevertheless, these data provide definitive evidence that the synthetic H3 relaxin peptide binds to and elicits a second messenger response by stimulating human relaxin receptors.

Ability of a Well Characterized H2 Relaxin Antibody to Recognize H3 Relaxin-- The ability of a well characterized anti-H2 relaxin antibody to recognize H1 and H3 relaxin was tested by radioimmunoassay. As shown in Fig. 4, H2 relaxin was able to displace ¹²⁵I-labeled H2 relaxin binding to the anti-H2 relaxin antibody with high specificity. In contrast, H1 and H3 relaxin showed poor cross-reactivity with the antisera as determined by their poor ability to displace ¹²⁵I-labeled H2 relaxin binding. Furthermore, the non-parallelism of the displacement curves indicates that not all the antibody epitopes are recognized by the two peptides.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Ability of a well characterized H2 relaxin antibody to recognize H3 relaxin. The H2 relaxin antibody was immobilized onto ELISA plates, and a competition experiment was performed using H1, H2, and H3 relaxin against 125I-labeled H2 relaxin. Results are mean ± S.E. of the specific binding (%) of triplicate determinations from a representative assay.

Relaxin Gene Expression in the Mouse-- The expression of M3 relaxin mRNA was compared with M1 relaxin mRNA expression using Southern blotting of RT-PCR products. Although this technique is only semi-quantitative, it enabled us to determine the potential sites of expression of M3 relaxin compared with M1 relaxin. The results of a representative experiment are shown (Fig. 5), and duplicate experiments gave identical results. M3 relaxin mRNA was expressed in a number of tissues in C57BLK6J mice where M1 relaxin was found, but the pattern of expression between the two mouse relaxins was different. In male non-reproductive tissues (Fig. 5A), the highest levels of M1 relaxin expression were seen in the brain, moderate levels in the thymus, heart, and kidney, lower levels in the lung, spleen, and skin, with no expression seen in the gut. Interestingly, M3 relaxin expression was detected at highest levels in brain; however, it was expressed at moderate levels in the thymus, lung, and spleen, only at very low levels in the heart and liver, and not at all in the kidney, skin, and gut. Female mice showed an almost identical pattern of expression for both genes in these tissues (data not shown). In male reproductive tissues M3 relaxin mRNA was significantly expressed only in the testis, whereas M1 relaxin mRNA was detected in the testis, epididymis, and prostate (Fig. 5B). Both relaxins were also detected in female reproductive organs in the mammary gland, ovaries of non-pregnant, pregnant, and lactating mice, and the endometrium and myometrium of pregnant mice (Fig. 5C). Significant expression of M3 relaxin mRNA was observed in all ovarian stages, whereas M1 relaxin expression was higher in ovaries of late gestation compared with ovaries from non-pregnant and lactating mice. High levels of M3 relaxin mRNA were detected in the brain, and further analysis of this tissue revealed that both relaxins were expressed in several distinct regions (Fig. 5B). Whereas M1 relaxin mRNA was consistently expressed in the hypothalamus, hippocampus, cortex, thalamus, pons/medulla, and cerebellum, M3 relaxin mRNA was found to be highly expressed in the thalamus and pons/medulla, thus suggesting that the two relaxins may play distinct roles in the mouse brain.

View larger version (48K): [in this window] [in a new window]   Fig. 5.   Southern blot analysis of M1 and M3 relaxin gene expression in the mouse. Southern blots were conducted on RT-PCR products (from various male and female mouse tissues) using specific internal oligonucleotide primers to M1 and M3 relaxin cDNA, respectively. UV photographs of GAPDH PCR products are shown as controls for quality, and equivalent loading, of the cDNA. Water (H2O) replaced cDNA in negative control reactions for each PCR. Relaxin mRNA expression was determined in a number of non-reproductive male tissues (A), in reproductive tissues and specific brain regions of the male (B), and in female reproductive tissues at different stages of pregnancy (C).

Northern Analysis-- Tissues in which M1 and M3 relaxin mRNA were positively identified by RT-PCR and Southern blot analysis were further examined by Northern blotting. Total RNA (5-25 µg) from the heart, brain, lung, thymus, spleen, ovary, endometrium, myometrium, cervix, and vagina were initially probed with a ³²P-labeled M3 relaxin specific probe, but no specific hybridizing bands were found in any tissue (Fig. 6A). In contrast a strong hybridizing signal was seen in the ovary (~1 kb) using a ³²P-labeled M1 relaxin-specific probe. Poly(A) RNA from the brain (15 µg) together with day 18.5 ovarian total RNA were then analyzed for M3 and M1 mRNA expression, and a specific ~1.2-kb hybridizing band was identified using the ³²P-labeled M3 relaxin probe in the brain only (Fig. 6B). The obtained transcript size was consistent with the predicted size based on the M3 relaxin transcript sequence (~1 kb) plus a poly(A) tail (~200-bp). In contrast, no M1 expression was seen in brain poly(A) RNA although a strong signal was seen in the ovarian total RNA of the same size as that seen in the total RNA blot and consistent with the size (~1 kb) published previously (30).

View larger version (55K): [in this window] [in a new window]   Fig. 6.   Northern blot analysis of M1 and M3 relaxin mRNA expression in the mouse. Total RNA (5-25 µg) from male non-reproductive and female pregnant reproductive mouse tissues (A) were hybridized separately with 32P-labeled single-stranded M3 relaxin, M1 relaxin, and GAPDH probes. The RNA from the female reproductive tissues was pooled from day 7.5, 10.5, and 18.5 of pregnancy. B, poly(A) RNA from male brain (15 µg; purified from 200 µg of total RNA) and total RNA (5 µg) from the pregnant (day 18.5) ovary (B) were also probed with the M3 relaxin, M1 relaxin, and GAPDH probes. The positions of a 0.24-9.5-kb RNA ladder are indicated. Note the specific signals for M1 relaxin (~1 kb) in the ovary and for M3 relaxin (~1.2 kb) in brain poly(A) RNA.

Expression of H3 Relaxin in Human Tissues-- A CLONTECH Multitissue Expression Array was used to examine the possible sites of expression of H3 relaxin in human tissues. The array contained normalized poly(A) RNA (50-750 ng) from 76 different human tissues including 8 different control RNAs and DNAs, spotted onto a nylon membrane. The array was probed with a ³²P-labeled 374-bp H3 relaxin-specific gene fragment from the 3' end of the H3 relaxin transcript, generated from genomic DNA. This DNA fragment was sequenced on both strands and found to be identical to the Celera sequence. Very weak hybridizing signals were observed in spleen, thymus, peripheral blood leukocytes, lymph node, and testis; however, these signals were barely discernible above background, and hence, the data are not shown. RT-PCR was also performed on human CL from early pregnancy using two different primer combinations based on the H3 relaxin sequence. No specific bands were observed in any PCR even after changing the PCR conditions, whereas transcripts for H2 relaxin and GAPDH were easily amplified (data not shown), confirming the integrity of the cDNA.

Distribution of Relaxin mRNA in the Mouse Brain-- Given the high levels of M3 relaxin mRNA expression detected by RT-PCR and Northern blotting in the brain, its distribution was further examined using in situ hybridization histochemistry (28). Multiple specific ³⁵S-labeled oligonucleotide probes were utilized to determine the cellular distribution of M3 relaxin mRNA throughout the rostro-caudal extent of the female C57BLK6J mouse brain. M3 relaxin mRNA was not widely detected throughout brain nuclei but was most strongly detected in the pons/medulla (Fig. 7). The strongest level of M3 relaxin mRNA was present in the pars ventromedialis of the dorsal tegmental nucleus. In addition, M3 relaxin mRNA was also detected, albeit at far lower levels, in the hippocampus and olfactory regions. Brain regions containing low levels of mRNA encoding M3 relaxin may not have been detected in the current study due to sensitivity limitations associated with in situ hybridization histochemistry. The distribution of M3 relaxin mRNA in the brain differs from that of M1 relaxin mRNA, as no M1 relaxin mRNA was detected in the pars ventromedialis of the dorsal tegmental nucleus (data not shown).

View larger version (126K): [in this window] [in a new window]   Fig. 7.   In situ hybridization histochemical localization of M3 relaxin mRNA in mouse brain. A near adjacent thionin-stained section (A) reveals the distribution of M3 relaxin mRNA (B) to be localized to the pars ventromedialis of the dorsal tegmental nucleus (vmDTg), situated ventral to the dorsal tegmental nucleus (DTg), within the pons/medulla. An x-ray film image of M3 relaxin mRNA enriched in the pars ventromedialis of the dorsal tegmental nucleus of the mouse pons/medulla (B), but not in the cerebellum (Cb). Scale bar: A, 1 mm; B, 0.5 mm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In recent years many new members of the relaxin/insulin/IGF superfamily have been identified, either by differential cloning or by screening of the EST data bases using A- and B-chain motifs (20, 21, 38). The availability of the human genome sequence provides an ideal tool to screen for other novel relaxin genes. Searching of the Celera Discovery System and Celera Genomics associated data bases resulted in the identification of a novel human and mouse relaxin gene. The derived amino acid sequences of these genes reflect the characteristics of members of the relaxin/insulin family of peptide hormones. Although the sequences show highest amino acid homology to INSL5, the presence of an amino acid motif in the B-chain (RXXXRXXI), which is essential for relaxin receptor binding (35), indicates that these sequences are more relaxin-like than insulin-like. Furthermore, the A-chain of M3 relaxin is more typical in structure to other members of the relaxin peptide family, wherein the terminal A-chain cysteine residue is separated from the intra-disulfide bond-linked cysteine by eight amino acids. The previously characterized M1 relaxin peptide stands out from all the known relaxin sequences by having nine residues in this sequence region (30). It remains to be seen if the relaxin bioactivity of M3 is greater than that of M1 relaxin. The homology of relaxin peptides between species is generally quite low with H2 and M1 relaxin sharing only 43% homology in the peptide domains. Conversely, H3 and M3 relaxin share 93% homology in the A- and B-domains. It is difficult to speculate what the significance of this homology is. The homology of the C-peptide chain between relaxins is also generally quite low; however, H3 and M3 relaxin show very high homology in this region and are also much shorter than other relaxins. Again, it is difficult to speculate on the reason for this difference, but it may reflect differences in the functioning of the C-chain. Interestingly, the H3 relaxin B-chain sequence shares 59% sequence homology to a newly discovered frog relaxin sequence, although it only shares 38% homology in the A-chain and 14% homology in the C-chain (39). Considering the high homology between the H3, M3, and R3 relaxins in the A-chain sequence, it is unlikely that this frog sequence is an H3 homologue.

The H3 relaxin gene is localized on chromosome 19 at 19p13.3 in close proximity (~5 Mb) to INSL3 (19p13.2). In contrast, H1 and H2 relaxin are in close proximity on chromosome 9 at 9p24 together with INSL4 and INSL6, whereas human insulin (11p15) and INSL5 (1p31) are at different locations. The situation is similar in the mouse where the M3 gene is localized on chromosome 8 at 8C2 about 13 Mb from mouse INSL3 at 8B3.2, whereas M1 relaxin and INSL6 are localized in a similar region of chromosome 19 at 19B. These data match the data obtained from phylogenetic analysis indicating that the novel relaxin genes are grouped under a separate branch and that this peptide family may have diverged from other relaxins early in evolution.

A series of relaxins have been successfully prepared in our laboratory by simple combination of the two fully unprotected A- and B-chains in solution at high pH (24, 40, 41). Like insulin chain folding and combination, relaxin chains have a high propensity for correct pairing and alignment in solution, although yields are often low, due largely to the poor solubility of the B-chain. However, efforts to use a similar approach for the preparation of H3 relaxin were unsuccessful despite the use of many variations of reaction conditions including temperature, pH, chain ratios, and oxidation buffer. It appears that the flexible multiserine N-terminal region of the A-chain results in less successful formation of the -helix that is an essential initiation element for subsequent chain combination (42). Consequently, alternative means of preparation of the H3 relaxin were investigated, and use of S-selective thiol protection and regioselective disulfide bond formation was selected. This approach has been successfully used for the synthesis of other members of the insulin superfamily including bombyxin and H2 relaxin (23, 43). The strategy employed for bombyxin afforded H3 relaxin synthesis, albeit in very low overall yield, due to the difficulty in assembly of the A-chain, the poor solubility of the B-chain under certain conditions, and the large number of intermediate manipulation and purification steps involved in the chain combination. Further optimization and refinement of the procedure is currently under way. Nevertheless, the synthetic H3 relaxin peptide was confirmed to be homogeneous by chemical characterization using MALDI-TOF MS and analytical reverse phase-high pressure liquid chromatography.

The synthetic H3 relaxin peptide was tested for its ability to bind to, and elicit a biological response from, a human cell line (THP-1) that expresses the relaxin receptor and that is used routinely as an in vitro relaxin bioassay in our laboratory and elsewhere (25). The synthetic H3 relaxin peptide was able to compete with ³³P-labeled relaxin-binding sites on THP-1 cells with somewhat lower affinity than H2 or H1 relaxin. Furthermore, stimulation of THP-1 cells with the synthetic H3 relaxin peptide resulted in a dose-dependent increase in cAMP production, again with slightly lower activity than seen with H1 and H2 relaxin. Importantly, both hINSL3 and bINSL (1 µM) do not stimulate cAMP production from THP-1 cells. Although the relaxin receptor has yet to be cloned, relaxin has been clearly demonstrated to bind to cell surface receptors distinct from insulin, IGF, or INSL3 receptors (44, 45). Furthermore, we have recently shown (46, 47) that a synthetic sheep INSL3 peptide modified to contain the full relaxin-binding motif in the same position in the B-chain as relaxin has only minimal relaxin activity. Therefore, our data provide conclusive evidence that this novel peptide retains the structural features necessary for interaction with, and activation of, relaxin receptors and can therefore be termed a "relaxin." Considering the sequence homology between the M3 and H3 relaxins, it is also likely that a peptide derived from the mouse sequence should also be considered a relaxin.

Many studies have reported circulating levels of relaxin in human subjects using an ELISA developed by Genentech (26). We have tested the ability of the anti-H2 relaxin antibody used in this ELISA to recognize H1 and H3 relaxins. These results clearly demonstrate poor recognition of H1 or H3 relaxin by the antisera. Therefore, it can be concluded that previous studies measuring circulating relaxin levels have only measured H2 relaxin. Studies currently under way involve the development of specific antibodies to H3 relaxin to establish an assay to measure circulating levels and to locate storage sites.

M1 and M3 relaxin gene expression were identified in tissues of the female reproductive tract and in extra-ovarian tissues such as the brain, spleen, thymus, and lung, confirming that relaxin is not just a hormone of pregnancy. Although PCR data suggest that the M3 relaxin gene is expressed in the ovary, it seems unlikely that it is a luteal cell product like H2 (3) and M1 relaxin (30). Northern blot analyses and preliminary in situ hybridization histochemistry data of ovaries from late pregnant mice failed to detect any specific expression, and PCR studies using human CL of early pregnancy failed to amplify a specific transcript. In contrast, a strong hybridizing signal was seen in the ovary with an M1 relaxin probe, consistent with strong M1 relaxin expression in the CL during late pregnancy as shown previously (30). The specific cellular localization of the ovarian source of the newly discovered relaxin is therefore still to be determined. This paper demonstrates for the first time M1 relaxin mRNA expression in the mammalian liver, lung, thymus, spleen, kidney, skin, testis, epididymis, and myometrium. Furthermore, we show that M3 relaxin mRNA is present in several tissues including the brain, thymus, spleen, lung, testis, ovary, and mammary gland and weakly expressed in the heart, liver, epididymis, prostate, and uterus, suggesting similar but separate function(s) of the two mouse relaxins. It should be noted that M3 and M1 relaxin mRNA were only detected in these tissues by RT-PCR, and M3 mRNA expression was only shown in the brain by Northern blot analysis of poly(A) RNA. Alignment of the relevant M3 probe nucleotide sequence used in the Northern blots with the nucleotide sequences in this region of other members of the insulin/relaxin family shows less than 50% nucleotide homology and, importantly, never in blocks of more than 7 nucleotides. The highest homology is in fact with the M1 relaxin sequence at 47%, and as the Northern blots did not show any specific hybridization of the M3 probe to ovarian RNA, it clearly indicates that it does not cross-react with M1 relaxin mRNA. It is therefore not possible that the specific signal that we are demonstrating for the new relaxin gene (M3) in the brain is the result of hybridization to mRNA of another family member. Furthermore, M3 relaxin mRNA has been clearly demonstrated to be expressed in distinct brain regions to that of M1 relaxin mRNA by in situ hybridization. It is also probable that M3 and M1 relaxin expression in other tissues detected by RT-PCR resides in specific cell types within each tissue. Similarly, results for H3 relaxin expression in the human multitissue array reflect very low levels of poly(A) RNA (50-750 ng) for each tissue. These levels correlate to less than 20 µg of total RNA for each tissue and therefore mirror the results obtained for total RNA in Northern blots detected with mouse tissues. These findings suggest that the newly discovered relaxin may play a role in both male and female reproduction, in addition to serving functions outside the mammalian reproductive tract.

The strongest evidence supporting a functional role for the newly discovered relaxin in mammals is the presence of M3 relaxin mRNA expression, as detected by Northern blot analysis and in situ hybridization histochemistry, in the mouse brain. In the pons/medulla, M3 relaxin mRNA was highly abundant in cells encompassing the pars ventromedialis of the dorsal tegmental nucleus and the dorsal part of the reticular tegmental nucleus of female C57BLK6J mice. M3 relaxin mRNA was also detected in CA1-3 fields of the hippocampus. In contrast, M1 relaxin mRNA was detected in the olfactory system and CA1-3 fields of the hippocampus but was not detected within the pars ventromedialis of the dorsal tegmental nucleus or the reticular tegmental nucleus, thus suggesting possible differential functions of these two relaxin genes in the mouse brain.

Little is known about the connections or functions of the pars ventromedialis of the dorsal tegmental nucleus. According to its location and cytoarchitecture, the pars ventromedialis is reported to correspond to nucleus O in the rabbit (48). Interestingly, large numbers of retrogradely labeled neurons have been reported in layer V of the prefrontal and cingulate cortices and in olfactory areas of the rat (49), which correspond to regions of the brain previously described to contain relaxin-binding sites (10). More recently, direct projections from the pars ventromedialis of the dorsal tegmental nucleus were demonstrated to the frontal and temporal cortices, and it was proposed that this nucleus may be involved in the modulation of cortical activity (50). Behavioral studies suggest an influence of this nucleus on the regulation of forebrain activity mainly in delayed motivational, emotional, and associative functions (51). Clearly, further studies are necessary and are currently under way to better define the possible role(s) of this novel relaxin neuropeptide in this specific region of the human and mouse brain.

The availability of the human and mouse genome sequences provides a unique opportunity to discover novel relaxin genes. This paper describes the identification, sequencing, activity, and expression of a novel relaxin gene in the human and mouse, and we are currently undertaking experiments to define the function of relaxin-3 in the human, mouse, and rat, particularly in the brain. These studies confirm that relaxin has many functions outside the female reproductive tract and point to a role of relaxin in neuropeptide signaling.

	ACKNOWLEDGEMENTS

We gratefully acknowledge Dr. Marc Mathieu for valuable input and critique of the manuscript, Dr. Ping Fu for synthesis of human INSL3, Kathryn Smith for amino acid analysis, and Tracey Wilkinson for help with data base searching and sequence alignments.

	FOOTNOTES

^* This work was supported in part by a block grant to the Howard Florey Institute (Reg Key 983001) from the National Health and Medical Research Council of Australia.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ Recipient of a Peter Doherty Postdoctoral fellowship (National Health and Medical Research Council of Australia).

Recipient of an Australian Research Council QEII fellowship.

Recipient of a National Health and Medical Research Council of Australia R. D. Wright fellowship. To whom correspondence should be addressed. Tel.: 61-3-8344-5648; Fax: 61-3-9348-1707; E-mail: r.bathgate@hfi.unimelb.edu.au.

Published, JBC Papers in Press, October 31, 2001, DOI 10.1074/jbc.M107882200

	ABBREVIATIONS

The abbreviations used are: CL, corpus luteum; IGF, insulin-like growth factor; EST, expressed sequence tag; MALDI-TOF, matrix-assisted laser desorption ionization/time of flight; MS, mass spectrometry; ELISA, enzyme-linked immunosorbent assay; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; RT, reverse transcription; MOPS, 4-morpholinepropanesulfonic acid.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea