- Email: [email protected]
The human vasopressin-oxytocin gene family: no evidence for additional neurophysin-related genes
Molecular and Cellular Endocrinology, 98 (1993) 61-66 0 1993 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/93/$06.00
61
MCE 03121
The human vasopressin-oxytocin gene family: no evidence for additional neurophysin-related genes Sofia Lopes da Silva, Ardy van Helvoort and J. Peter H. Burbach * Rudolf Magnus Institute, Department of Pharmacolog): (Received
Key words: Vasopressin
gene; Oxytocin
Utrecht Unicersity, Vondellaan 6, 3521 GD Vtrecht, Netherlands
24 May 1993; accepted
6 September
1993)
gene; Neurophysin
Summary Over the last 20 years several observations at the peptide level have indicated the possible existence of additional members of the vasopressin (VP)-oxytocin COT) gene family in mammals. In this study, the human genome was analyzed for the existence of genes structurally related to the VP and OT genes. Human genomic blots probed under low stringency conditions with exon B of the human OT gene, that codes for the conserved constant region of neurophysin, revealed the presence of two distinct bands in addition to the known VP and OT gene fragments. Five clones were obtained from a library of genomic EcoRI fragments ranging from 4-8 kb, that comprised both low stringency signals, by low stringency hybridization with the OT exon B probe. One clone of 7 kb hybridized at high stringency conditions to bands of the same size as previously detected with OT exon B on a human genomic blot. However, no similarity was observed between the open reading frames of this clone and the neurophysin portion of the OT gene. Another clone of 4.8 kb was identical to a fragment of the gene for the human bone morphogenetic factor hBMP-6, a member of the TGF-B family. The hBMP-6 gene was not detected by low stringency hybridization of the human genomic blot with the OT exon B probe. No significant similarity was found between the amino acid sequences of human OT neurophysin and hBMP-6. Therefore, no evidence can be provided that the human genome contains additional neurophysin-related genes. This finding suggests that if additional vasopressin and oxytocin-like peptides exist, they will not be transcribed from genes with close similarity to the VP and OT genes.
Introduction The mammalian vasopressin (VP)-oxytocin (OT) gene family consists of two members, the VP gene and the OT gene (Gainer and Wray, 1992). These genes have a similar three exon structure and display a high degree of homology, especially in exon B. Exon B, which encodes the constant region of the neurophysins, shows more than 95% nucleotide sequence identity between the VP and OT genes. In all investigated species (mouse, rat and human), the VP and OT genes are physically linked in a head-to-head orientation with an intergenic distance of 3 to 12 kb (Hara et al., 1990; Mohr et al., 1988; Sausville et al., 1985). Over the last 20 years, suggestions that the VP-OT gene family extends beyond these two known members have persisted, but direct proof has failed as yet. Most
* Corresponding
author.
Tel.: 31-30-880521;
Fax: 31-30-896034.
of these suggestions were based on observations made at the peptide level. The claim that vasotocin exists in mammals, in particular in the foetal pineal gland, was the first suggestion of a third VP-OT-related peptide (Pave1 et al., 1973). Although this claim was refuted by a number of investigators (Negro-Vilar et al., 1979; Fisher and Fernstrom, 1981; Dogterom et al., 1980; Nieuwenhuis, 1984) and explained by the identification of modified forms of VP and OT in the pineal gland (Liu et al., 1988), a number of independent observations have given support to the hypothesis that additional members of the VP-OT family may exist in mammals. These include the detection of two VP-like peptides in marsupials (Chauvet et al., 1983) and the description of litocin, a reportedly VP-OT-related variant in humans (Lee, 1988). Based on the specificity of monoclonal antibodies, a peptide related to, but different from VP and OT has been proposed to exist in the thymus (Geenen et al., 1991). Similar suggestions for the existence of a second gene for vasopressin or a
62
related peptide were forwarded by Bonner and Brownstein (1984) in order to explain the paradoxical phenomenon of the presence of normal levels of VP in the adrenals, ovaries, testes and anterior pituitary of the homozygous (di/ di) Brattleboro rat. The Brattleboro rat is a mutant strain of the Long Evans rat that contains a single base deletion in exon B of both alleles for VP, resulting in an impaired processing of the VP-precursor (Schmale and Richter, 1984). This suggestion could be in agreement with the finding that pigs have only Lys*-vasopressin (LVP) in the pituitary gland, but both LVP and AVP in the testes (Nicholson et al., 1988) and ovaries (Choy and Watkins, 1988). Recently, the partial sequence of a tumor antigen of small cell lung carcinoma with an identical primary structure to neurophysin except for one amino acid has further stimulated the speculation of a third gene belonging to the VP-OT gene family (Rosenbaum et al., 1990). Another observation in this direction is the detection of an unknown band on genomic blots of human DNA using exon B of the OT gene as probe (Sausville et al., 1985). In fish multiple genes encoding vasotocin and isotocin have been demonstrated which has led to the speculation that multiplicity of genes may have been conserved evolutionary (Morley et al., 1990; Urano et al., 1992). Experimental proof for the existence of a third gene in mammals is lacking. These observations prompted us to analyze the human genome in detail for the existence of VP-OT-related genes. In this study, human chromosomal DNA was analyzed by hybridization on genomic blots under low stringency conditions using the most conserved, neurophysin-encoding part of the VP-OT genes (exon B) as probe. Low stringency signals were isolated from a size-restricted human genomic library and characterized in order to establish their degree of similarity to the VP-OT genes. Although DNA with sequence homology to the VP-OT gene family was identified, the results do not support the existence of additional genes that are structurally related to the VP-OT genes. Materials and methods Isolation and analysis of high molecular mass cellular DNA
Frozen placenta was minced using scissors. The tissue was digested for 16 h at 56°C with 0.1 mg of proteinase K per ml in TE buffer (10 mM Tris-HCl, pH 7.5/l mM EDTA) containing 0.1% sodium dodecyl sulfate (SDS). The nucleic acids were extracted twice with an equal volume of 1:l (vol/vol) phenol/ chloroform. After precipitation with ethanol, the DNA was spooled on a glass rod and dissolved in TE buffer. RNA was hydrolyzed for 1 h at 37°C with 10 pg of RNase A per ml. Phenol/chloroform extractions were repeated, and after precipitation the DNA was dis-
solved in TE buffer and stored at 4°C. DNA was digested overnight in buffers recommended by the suppliers. For Southern blot analysis, usually 20 pg of DNA was digested with a 5-fold excess of restriction enzymes, electrophoresed on 0.8% agarose gels and transferred to Zetaprobe nylon membrane (Bio-Rad, Richmond, CA) using the alkaline Southern blot transfer technique recommended by the suppliers. DNA cloning from a size-restricted human genomic library
DNA fragments of EcoRI-digested human DNA ranging from 4-8 kb were electro-eluted from agarose gel and ligated in the dephosphorylated EcoRI arms of the lambda ZAP II phage vector (Stratagene, La Jolla, CA). Phages, packaged with the Gigapack II Gold packaging extract, were plated out on E. coli strain ER1647. The 5.105 recombinant plaques were transferred in duplicate to Colony/Plaque Screen hybridization transfer membrane (DuPont, Boston, MA) following instructions from the supplier. Filters were screened with the human oxytocin exon B probe under low stringency hybridization conditions (described below). The genomic inserts of positive phages were rescued from the lambda ZAP II phage vector by means of in vivo excision of the pBluescript II plasmid as described by Stratagene. Fragments of clones, digested with HpaII or with Sau3A, that hybridized with the human oxytocin exon B probe were subcloned in pGEM7Zf( + > and sequenced by means of primers specific for the phage SP6 and T7 promoters. AI1 sequencing reactions were performed on denatured double-stranded templates using Sequenase version 2.0 according to the manufacturer’s instructions (United States Biochemical, Cleveland, OH). High and low stringency hybridizations
High stringency hybridizations were performed in 3 x SSC (1 x SSC is 0.15 M NaCl, 0.015 M sodium citrate) containing 0.1% polyvinylpyrrolidone, 0.1% bovine serum albumin, 0.1% Ficoll, 0.5% SDS, 10% dextran sulfate and 50 pg/ml of denatured salmon sperm DNA at 65°C overnight. The filters were washed 3 X 30 min at 65°C in O.lxSSC; 1.0% SDS and exposed to Kodak XS or XOMAT-AR film at -70°C with intensifying screens for l-4 days. Low stringency hybridizations were performed at 50°C in the described solution supplemented with 8% formamide. Filters were washed at 50°C with 0.5 X SSC; l.O%SDS. Probes were labeled by random priming using a32P dCTP to a specific activity of approximately lo9 dpm/pg of DNA (Feinberg et al., 1983).The exon B probe of the human OT gene was a 179 bp SmaI fragment comprising the neurophysin portion from nucleotides 855 to 1034 (Sausville et al., 1985).
63
Barn HI
Hind Ill
Eco RI H
Xba I H
L
L
24.0-
.. -It8
19.0ll.O7.46.0-
Fig. 1. High and low stringency hybridizations on a human genomic blot with a human oxytocin exon B probe. Hybridization conditions are as described in methods. H = high stringency, L = low stringency.
A
OTexB
clone#4
GGGCGGCGGtAGGCGATGCiGGACGGGGG;:..GGGCGGAiGGAGC IIIIIII IllI I I CGGGGGCAAAGGCCGCTGCTTC
clone#4
GGGGC.........:.GGGCGGAGiGAGCGGGGC:.......GGi III
I
I
IIIII
III1
III1
43 078 69
I I I
OTexB
GGGCCCAATATCTGCTGCGCGGA.AGAGCTGGGCTGCTTCGTGGG
922
clonet4
C..GGAGGGiGCG..GGGCiGGCGGAGGGiG.....CGGiGCG.G
104
I
I
III1
I
I
II
I
I
I
IllI
I
Ill
OTexB
CACCGCCGAAGCGCTGCGCTGCCAGGAGGAGMCTACCTGCCGTC
clone#4
GCGGiGGGAG..CGh II
OTexB
II
II
967 141
. . . . ..GGCiGGCGGTGGGiGCGGGGCGGi /III
III
II
IIIlIIIII
II
GCCCTGCCAGTCCGGCCAGAAGGCGTGC ...GGGAGCGGG..GGC
1007
. clonel4
CGGTGGGAGCGGGGCGGGGACGCCG II
II
I
166
I I I I
OTexB
CGCTGCGCCTTGGGC
donet
APAGDAGRGiAEGAGRAEGAGRAEGAGRAiGAGRAVGAGRAVGAGRAVGiGRGR
1022
B_
. I neuro-
II
II
I
I
I
I
54 I I
PGGKGRCFGPNICCAEELGCFVGTAEALRCQEENYLPSPC~SG~~CGSGGRCAVLGL
58
physin Fig. 2. Sequence analysis of clone #4. (A) Nucleotide sequence of the HpuII fragment of clone #4 that hybridized to exon B of the human OT gene and homology to human OT exon B. The sequence comparison was performed with the program Bestfit (Devereux et al., 1984). A region of 172 nucleotides with 71% homology to exon B is shown. (B) Amino acid sequence alignment between the sequenced HpaII fragment of clone #4 and human OT exon B. This analysis was performed with the program Bestfit (Devereux et al., 1984). Of all six reading frames, the amino acid sequence with highest similarity to the neurophysin sequence is shown.
64
Results Analysis of human genomic blots
Blots containing restriction enzyme digests of total human chromosomal DNA were hybridized under high and low stringency conditions with a human OT exon B probe (Fig. 1). Under high stringency conditions, the known restriction fragments of the VP and OT genes were detected (Sausville et al., 1985). The VP fragment, which has 7% base pair mismatch to the used OT probe gave a weaker hybridization signal than the OT fragment. Under low stringency conditions two additional bands were found in each restriction enzyme digest. The sizes were 7.4 and 6.0 kb in the EcoRI digest, 14.0 and 18.0 kb in the Hind111 digest and 17.0 and 12.0 kb in the XbaI digest. In the BamHI digest, only one band was detected at 2.2 kb. In order to examine these low stringency signals, a genomic library was constructed using EcoRI fragments ranging from 4-8 kb.
The other four clones had inserts of approximately 7 kb and could therefore be responsible for one of the low stringency hybridization signals detected on genomic blots (Fig. 1). The EcoRI fragment of each clone was labeled and used as probe on human genomic blots under high stringency hybridization conditions. The pattern of hybridization of clone #4 corresponded to all the low stringency bands previously detected with the human OT exon B probe. In the BamHI digest fragments of 5.5 kb, 2.2 kb and 1.4 kb were detected; in the EcoRI digest 7.4 kb and 6.0 kb fragments were detected, in the Hind111 digest the detected fragments were 14.0 and 18.0 kb in size and in the XbaI the fragments were 17.0 and 12.0 kb. Clone #3 contained a repetitive DNA sequence. Clones #l, #2 and #5 showed dissimilar patterns, showing that they were different from the genomic sequences to which the human OT exon B probe had bound on genomic blots. Characterization of clones #4 and #5
Selection and cloning of genomic DNA with homology to VP and OT genes
Screening of a size-selected genomic library (5.105 independent plaques) under low stringency conditions with the human OT exon B probe yielded five positive clones. The clone showing the strongest hybridization signal (clone #5) contained a genomic insert of 4.8 kb.
Barn HI #4
OT
Eco RI #4
OT
DNA fragments that hybridized to exon B of the human OT gene were subcloned and sequenced. Figure 2 shows the sequence and analysis of a 157 bp HjraII fragment of clone #4. Comparison of the nucleotide sequences of the HpaII fragment and exon B showed an identity of 57.4% over a GC-rich region and thus explained the ability of the OT probe to hybridize
Hind Ill #4
OT
Xba I #4
OT
Fig. 3. Low stringency hybridization with 32P-labeled HpnII fragment of clone #4. Results shown are from the same human genomic blot as presented in Fig. 1. Hybridization conditions are as described in Materials and methods.
Discussion
to clone #4. The fragment had six open reading frames, but no amino acid similarity to human OT exon B was detected in any of them. A search for similarity of the sequenced fragment with nucleotide sequences deposited in the GenEMBL bank (updated 4/91) showed the closest resemblance, 71.1% identity over 1.52 nucleotides, with bovine herpes virus type 1 ~135 cDNA (Accession M84464). The 157 bp &@a11fragment of clone #4 was used as probe on human genomic blots under high and low stringency conditions. The pattern of hybridization revealed that this clone represented the low stringency signal of 7.4 kb detected previously by exon B of the OT gene in the EcoRI digest. Both under low and high stringency conditions, the 6.0 kb EcoRI fragment was also detected indicating a close similarity of this band with the &xzII fragment of clone #4. However, the VP and OT gene fragments were not detected by the 157 bp H&II fragment of clone #4 under the conditions used (Fig. 3). The nucleotide sequence of the 167 bp Sau3A fragment of clone #5, that hybridizes to exon B of the OT gene, is presented in Fig. 4. This Sau3A fragment of clone #5 was identical to nucleotides 405-571 of human bone morphogenetic factor hBMP-6, a member of the TGF-8 family (Celeste et al., 1990, Accession M38694). This fragment displayed a 67% identity over 85 nucleotides with exon B of the OT gene, but no significant amino acid similarity was observed between the identified region of hBMP-6 and the constant region of the human OT-associated neurophysin.
A
clone#S
This study was initiated to answer the question whether additional members of the VP-OT gene family exist in mammals. Several suggestions in this direction, based on experimental findings at the peptide level have persistently been made (Sausville et al., 1985; Pave1 et al., 1973; Chauvet et al., 1983; Lee, 1988; Geenen et al., 1991; Bonner and Brownstein, 1984; Nicholson et al., 1988; Choy and Watkins, 1988; Rosenbaum et al., 1990; Morley et al., 1990; Urano et al., 1992). In a previous study, Liu et al. (1988) ruled out the presence of vasotocin in the rat pineal gland by isolating modified peptides, rather than new forms of vasopressin and oxytocin. In this way the claim of an additional VP-OT-related peptide was refuted by experimental results. Here, we approached the question from the nudeotide level. Human genomic bIots were screened with exon B of the human OT gene as probe. This sequence codes for the evolutionary most conserved region of the VP and OT genes. It displays even to the distant vasotocin and isotocin genes in fish homologies between 64% (VT-l) and 69% (VT-21 (Morley et al., 1990). The mesotocin gene has 71% nucleotide homology to the used human OT exon B probe. The hybridization conditions, which were set to detect sequences with 60% or more homology to the human OT exon B probe, allowed detection of these sequences or sequences in the human genome as distantIy related to the human OT exon B probe as fish sequences. Sequences related to neurophysin but with
OTexB
AGATCTTGTiGGTGCTGGG&TCCCGCAC;GGCCCCGGC:...CC IIIII I I III III I GGGCTGCTTCGTGGGCACCGCCGAAG;G
clone#S
CTGC~CGGCCTCC~CAGCCG~G~CCCCGGCGC~CCGGCAG..: II0
I
III
I
II
ii
lflllllIlll
41 935 83
OTexB
CTGCGCTGCCAGGAGGAGAACTACCTGCCGTCGCCCTGCCAGTCC
980
clonet5
...CAGGAGiAGCAGCAGCiGCAGCAGCAiCTGCCTCGCdGAGAG
125
III
III
I
II
I
III
I
II
lllll
OTex
GGCCAGAAGGCGT.GCGGGAGCGGGGGCCGCTGC
cloneX5
CCCC~TCCCGGGCG~CTG~GTCC~CGCCCCTCT~CATGCTG
clone?5
EXLSVLGLPHRPRPLHGLQ:.QPQPPALR~EEQQQQQQiPRGEPPPGR~KSAPLFML I III Ill II PGGKGRCFGPNICCAESLGCFVGTAERLRCPEENYLPSPC
1013 167
B neuro-
56 58
physin
Fig. 4. Sequence analysis of clone #5. (A) Nucleotide sequence of the Suu3A fragment in clone #5 that hybridized to exon B of the human OT gene and homology to human GT exon B (Bestfit program, Devereux et al., 1984). Within this fragment, a region of 107 nucleotides was found with 63% sequence identity to the OT exon B probe. (B) Amino acid sequence alignment between the sequenced Suu3a fragment of clone #5 and human OT exon B (Bestfit program, Devereux et al., 1984). Gaps were introduced to optimize the alignment.
66
less than 60% homology escaped detection in this study. Although the OT exon B probe has a high G/C content (67%) we detected under low stringency conditions only two distinct bands in each restriction enzyme digest in addition to the VP and OT gene fragments. These nucleotide sequences had 60% similarity to neurophysin. However, none of these sequences appeared to encode a protein related to the amino acid sequence of neurophysin. We therefore conclude that the human genome does not contain a third closely-related member of the VP-OT gene family. Both human genomic blots and the size-restricted (4-8 kb) human genomic library were screened under low stringency conditions revealing DNA fragments with moderate homology to the used human OT exon B probe. One of these fragments (clone #4) was identified to be responsible for both low stringency hybridization signals previously found on the Southern blot (fig. 1 and 3). A close analysis of this clone and of the clone causing the strongest hybridization signal in the genomic library (clone #5) revealed the regions with homology of 60-70% to OT exon B (fig. 2B and 4B). Clone #5 shows 67% similarity to the OT exon B probe which is concentrated in a small sequence only. This may be the reason why it was detected in the genomic library but not, at first instance, on the genomic blot. At the amino acid level, no resemblance was observed between the sequenced fragments and the translated exon B probe (fig. 2C and 40. Regarding clone #5, no significant similarity was found between either the complete cDNA or the amino acid sequences of human OT and hBMP-6. Furthermore, the exon B probe bound to a region of the hBMP-6 gene that is not conserved among members of the TGF-B family. Any resemblance between the VP-OT family and the TGF-B family seems therefore of no significance. The hybridization characteristics of clone #4 further excluded the presence of neurophysin-related sequences in the human genome. Under high stringency conditions, clone #4 (7.4 kb in size) hybridized to all the low stringency bands previously found on the Southern blot with the human OT exon B probe. Under low stringency conditions, clone #4 failed to detect the VP-OT gene fragments on the Southern blot (Fig. 3). This indicates that the other low stringency fragment (6.0 kb EcoRI fragment) has a higher degree of similarity to clone #4 than to the human OT exon B probe. Based on these results, we cannot provide evidence that the human genome contains genes encoding neu-
rophysin-related peptides in addition to the VP and OT genes. The possibility that VP-OT-like peptides are encoded by unrelated genes with a VP-OT-like nucleotide sequence that lack the neurophysin portion cannot be excluded by our results. In theory, the use of another first exon spliced to the neurophysin encoding exons might lead to the synthesis of peptides related to but different from authentic VP and OT (Foo et al., 1991). However, it is more likely that the observations of related peptides originate from posttranslational processing events as indicated by Liu et al. (1988). References Bonner, T.I. and Brownstein, M.J. (19841 Nature 310, 17. Celeste, A.J., Iannazzi, J.A., Taylor, R.C., Hewick, R.M., Rosen, V., Wang, E.A. and Wozney, J.M. (1990) Proc. Natl. Acad. Sci. USA 87, 9843-9847. Chauvet, M.T., Chauvet, D.H.J. and Acher, R. (1983) Gen. Comp. Endocrinol. 49, 63-72. Choy, V.J. and Watkins, W.B. (1988) Neuropeptides 11, 119-123. Devereux, Haeberli and Smithies (1984) Nucleic Acids Res. 12, 387-395. Dogterom, J., Snijdewint, F.G.M., Pevet, P. and Swaab, D.F. (1980) J. Endocrinol. 84, 115123. Feinberg, A.P. and Vogelstein, B. (1983) Anal. Biochem. 132, 6-13. Fisher, L.A. and Fernstrom, J.D. (1981) Life Sci. 28, 1471-1481. Foo, N.C., Carter, D., Murphy, D. and Ivell, R. (1991) Endocrinology 128, 2118-2128. Gainer, H. and Wray, S. (1992) in Oxytocin in Maternal, Sexual, and Social Behaviors, Annals of the New York Academy of Sciences, vol. 652, 14-28. Geenen, V., Robert, F., Martens, H., Benhida, A., De Giovanni, G., Defresne, M-P., Boniver, J., Legros, J-J., Martial, J. and Franchimont, P. (1991) Mol. Cell. Endocrinol. 76, C27C31. Hara, Y., Battey, J. and Gainer, H. (19901 Mol. Brain Res. 8, 319-324. Lee, J.N. (1988) Program of the 8th International Congress of Endocrinology, Kyoto, Japan, p. 203 (Abstract). Liu, B., Poulter, L., Neascu, C. and Burbach, J.P.H. (19881 J. Biol. Chem. 263, 12-15. Nieuwenhuis, J.J. (1984) Life Sci. 35, 1713-1724. Mohr, E., Schmitz, E. and Richter, D. (1988) Biochimie 70,649-654. Morley, SD., Schonrock, C., Heierhorst, J., Figueroa, J., Lederis, K. and Richter, D. (1990) Biochemistry 29, 2506-2511. Negro-Vilar, A., Sanchez-France, F., Kwiatkowski, M. and Samson, W.K. (1979) Brain Res. Bull. 4, 789-792. Nicholson, H.D., Smith, A.J., Birkett, S.D., Denning-Kendall, P.A. and Pickering, B.T. (1988) J. Endocrinol. 117, 441-446. Pavel, S., Dorcescu, M., Petrescu-Holban, R., Ghinea, E. (1973) Science 181, 1252-1253. Rosenbaum, L.C., Neuweh, E.A., Van Tol, H.H.M., Peng Loh, Y., Verbalis, J.G., Hellstrom, I., Hellstrom, K.E. and Nilaver, G. (1990) Proc. Natl. Acad. Sci. USA 87, 9928-9932. Sausville, E., Carney, D. and Battey, J. (1985) J. Biol. Chem. 260, 18, 10236-10241. Schmale, H. and Richter, D. (1984) Nature 308, 705-709. Urano, A., Hyodo, S., Suzuki, M. (1992) Progr. Brain Res. 92, 39-46.
61
MCE 03121
The human vasopressin-oxytocin gene family: no evidence for additional neurophysin-related genes Sofia Lopes da Silva, Ardy van Helvoort and J. Peter H. Burbach * Rudolf Magnus Institute, Department of Pharmacolog): (Received
Key words: Vasopressin
gene; Oxytocin
Utrecht Unicersity, Vondellaan 6, 3521 GD Vtrecht, Netherlands
24 May 1993; accepted
6 September
1993)
gene; Neurophysin
Summary Over the last 20 years several observations at the peptide level have indicated the possible existence of additional members of the vasopressin (VP)-oxytocin COT) gene family in mammals. In this study, the human genome was analyzed for the existence of genes structurally related to the VP and OT genes. Human genomic blots probed under low stringency conditions with exon B of the human OT gene, that codes for the conserved constant region of neurophysin, revealed the presence of two distinct bands in addition to the known VP and OT gene fragments. Five clones were obtained from a library of genomic EcoRI fragments ranging from 4-8 kb, that comprised both low stringency signals, by low stringency hybridization with the OT exon B probe. One clone of 7 kb hybridized at high stringency conditions to bands of the same size as previously detected with OT exon B on a human genomic blot. However, no similarity was observed between the open reading frames of this clone and the neurophysin portion of the OT gene. Another clone of 4.8 kb was identical to a fragment of the gene for the human bone morphogenetic factor hBMP-6, a member of the TGF-B family. The hBMP-6 gene was not detected by low stringency hybridization of the human genomic blot with the OT exon B probe. No significant similarity was found between the amino acid sequences of human OT neurophysin and hBMP-6. Therefore, no evidence can be provided that the human genome contains additional neurophysin-related genes. This finding suggests that if additional vasopressin and oxytocin-like peptides exist, they will not be transcribed from genes with close similarity to the VP and OT genes.
Introduction The mammalian vasopressin (VP)-oxytocin (OT) gene family consists of two members, the VP gene and the OT gene (Gainer and Wray, 1992). These genes have a similar three exon structure and display a high degree of homology, especially in exon B. Exon B, which encodes the constant region of the neurophysins, shows more than 95% nucleotide sequence identity between the VP and OT genes. In all investigated species (mouse, rat and human), the VP and OT genes are physically linked in a head-to-head orientation with an intergenic distance of 3 to 12 kb (Hara et al., 1990; Mohr et al., 1988; Sausville et al., 1985). Over the last 20 years, suggestions that the VP-OT gene family extends beyond these two known members have persisted, but direct proof has failed as yet. Most
* Corresponding
author.
Tel.: 31-30-880521;
Fax: 31-30-896034.
of these suggestions were based on observations made at the peptide level. The claim that vasotocin exists in mammals, in particular in the foetal pineal gland, was the first suggestion of a third VP-OT-related peptide (Pave1 et al., 1973). Although this claim was refuted by a number of investigators (Negro-Vilar et al., 1979; Fisher and Fernstrom, 1981; Dogterom et al., 1980; Nieuwenhuis, 1984) and explained by the identification of modified forms of VP and OT in the pineal gland (Liu et al., 1988), a number of independent observations have given support to the hypothesis that additional members of the VP-OT family may exist in mammals. These include the detection of two VP-like peptides in marsupials (Chauvet et al., 1983) and the description of litocin, a reportedly VP-OT-related variant in humans (Lee, 1988). Based on the specificity of monoclonal antibodies, a peptide related to, but different from VP and OT has been proposed to exist in the thymus (Geenen et al., 1991). Similar suggestions for the existence of a second gene for vasopressin or a
62
related peptide were forwarded by Bonner and Brownstein (1984) in order to explain the paradoxical phenomenon of the presence of normal levels of VP in the adrenals, ovaries, testes and anterior pituitary of the homozygous (di/ di) Brattleboro rat. The Brattleboro rat is a mutant strain of the Long Evans rat that contains a single base deletion in exon B of both alleles for VP, resulting in an impaired processing of the VP-precursor (Schmale and Richter, 1984). This suggestion could be in agreement with the finding that pigs have only Lys*-vasopressin (LVP) in the pituitary gland, but both LVP and AVP in the testes (Nicholson et al., 1988) and ovaries (Choy and Watkins, 1988). Recently, the partial sequence of a tumor antigen of small cell lung carcinoma with an identical primary structure to neurophysin except for one amino acid has further stimulated the speculation of a third gene belonging to the VP-OT gene family (Rosenbaum et al., 1990). Another observation in this direction is the detection of an unknown band on genomic blots of human DNA using exon B of the OT gene as probe (Sausville et al., 1985). In fish multiple genes encoding vasotocin and isotocin have been demonstrated which has led to the speculation that multiplicity of genes may have been conserved evolutionary (Morley et al., 1990; Urano et al., 1992). Experimental proof for the existence of a third gene in mammals is lacking. These observations prompted us to analyze the human genome in detail for the existence of VP-OT-related genes. In this study, human chromosomal DNA was analyzed by hybridization on genomic blots under low stringency conditions using the most conserved, neurophysin-encoding part of the VP-OT genes (exon B) as probe. Low stringency signals were isolated from a size-restricted human genomic library and characterized in order to establish their degree of similarity to the VP-OT genes. Although DNA with sequence homology to the VP-OT gene family was identified, the results do not support the existence of additional genes that are structurally related to the VP-OT genes. Materials and methods Isolation and analysis of high molecular mass cellular DNA
Frozen placenta was minced using scissors. The tissue was digested for 16 h at 56°C with 0.1 mg of proteinase K per ml in TE buffer (10 mM Tris-HCl, pH 7.5/l mM EDTA) containing 0.1% sodium dodecyl sulfate (SDS). The nucleic acids were extracted twice with an equal volume of 1:l (vol/vol) phenol/ chloroform. After precipitation with ethanol, the DNA was spooled on a glass rod and dissolved in TE buffer. RNA was hydrolyzed for 1 h at 37°C with 10 pg of RNase A per ml. Phenol/chloroform extractions were repeated, and after precipitation the DNA was dis-
solved in TE buffer and stored at 4°C. DNA was digested overnight in buffers recommended by the suppliers. For Southern blot analysis, usually 20 pg of DNA was digested with a 5-fold excess of restriction enzymes, electrophoresed on 0.8% agarose gels and transferred to Zetaprobe nylon membrane (Bio-Rad, Richmond, CA) using the alkaline Southern blot transfer technique recommended by the suppliers. DNA cloning from a size-restricted human genomic library
DNA fragments of EcoRI-digested human DNA ranging from 4-8 kb were electro-eluted from agarose gel and ligated in the dephosphorylated EcoRI arms of the lambda ZAP II phage vector (Stratagene, La Jolla, CA). Phages, packaged with the Gigapack II Gold packaging extract, were plated out on E. coli strain ER1647. The 5.105 recombinant plaques were transferred in duplicate to Colony/Plaque Screen hybridization transfer membrane (DuPont, Boston, MA) following instructions from the supplier. Filters were screened with the human oxytocin exon B probe under low stringency hybridization conditions (described below). The genomic inserts of positive phages were rescued from the lambda ZAP II phage vector by means of in vivo excision of the pBluescript II plasmid as described by Stratagene. Fragments of clones, digested with HpaII or with Sau3A, that hybridized with the human oxytocin exon B probe were subcloned in pGEM7Zf( + > and sequenced by means of primers specific for the phage SP6 and T7 promoters. AI1 sequencing reactions were performed on denatured double-stranded templates using Sequenase version 2.0 according to the manufacturer’s instructions (United States Biochemical, Cleveland, OH). High and low stringency hybridizations
High stringency hybridizations were performed in 3 x SSC (1 x SSC is 0.15 M NaCl, 0.015 M sodium citrate) containing 0.1% polyvinylpyrrolidone, 0.1% bovine serum albumin, 0.1% Ficoll, 0.5% SDS, 10% dextran sulfate and 50 pg/ml of denatured salmon sperm DNA at 65°C overnight. The filters were washed 3 X 30 min at 65°C in O.lxSSC; 1.0% SDS and exposed to Kodak XS or XOMAT-AR film at -70°C with intensifying screens for l-4 days. Low stringency hybridizations were performed at 50°C in the described solution supplemented with 8% formamide. Filters were washed at 50°C with 0.5 X SSC; l.O%SDS. Probes were labeled by random priming using a32P dCTP to a specific activity of approximately lo9 dpm/pg of DNA (Feinberg et al., 1983).The exon B probe of the human OT gene was a 179 bp SmaI fragment comprising the neurophysin portion from nucleotides 855 to 1034 (Sausville et al., 1985).
63
Barn HI
Hind Ill
Eco RI H
Xba I H
L
L
24.0-
.. -It8
19.0ll.O7.46.0-
Fig. 1. High and low stringency hybridizations on a human genomic blot with a human oxytocin exon B probe. Hybridization conditions are as described in methods. H = high stringency, L = low stringency.
A
OTexB
clone#4
GGGCGGCGGtAGGCGATGCiGGACGGGGG;:..GGGCGGAiGGAGC IIIIIII IllI I I CGGGGGCAAAGGCCGCTGCTTC
clone#4
GGGGC.........:.GGGCGGAGiGAGCGGGGC:.......GGi III
I
I
IIIII
III1
III1
43 078 69
I I I
OTexB
GGGCCCAATATCTGCTGCGCGGA.AGAGCTGGGCTGCTTCGTGGG
922
clonet4
C..GGAGGGiGCG..GGGCiGGCGGAGGGiG.....CGGiGCG.G
104
I
I
III1
I
I
II
I
I
I
IllI
I
Ill
OTexB
CACCGCCGAAGCGCTGCGCTGCCAGGAGGAGMCTACCTGCCGTC
clone#4
GCGGiGGGAG..CGh II
OTexB
II
II
967 141
. . . . ..GGCiGGCGGTGGGiGCGGGGCGGi /III
III
II
IIIlIIIII
II
GCCCTGCCAGTCCGGCCAGAAGGCGTGC ...GGGAGCGGG..GGC
1007
. clonel4
CGGTGGGAGCGGGGCGGGGACGCCG II
II
I
166
I I I I
OTexB
CGCTGCGCCTTGGGC
donet
APAGDAGRGiAEGAGRAEGAGRAEGAGRAiGAGRAVGAGRAVGAGRAVGiGRGR
1022
B_
. I neuro-
II
II
I
I
I
I
54 I I
PGGKGRCFGPNICCAEELGCFVGTAEALRCQEENYLPSPC~SG~~CGSGGRCAVLGL
58
physin Fig. 2. Sequence analysis of clone #4. (A) Nucleotide sequence of the HpuII fragment of clone #4 that hybridized to exon B of the human OT gene and homology to human OT exon B. The sequence comparison was performed with the program Bestfit (Devereux et al., 1984). A region of 172 nucleotides with 71% homology to exon B is shown. (B) Amino acid sequence alignment between the sequenced HpaII fragment of clone #4 and human OT exon B. This analysis was performed with the program Bestfit (Devereux et al., 1984). Of all six reading frames, the amino acid sequence with highest similarity to the neurophysin sequence is shown.
64
Results Analysis of human genomic blots
Blots containing restriction enzyme digests of total human chromosomal DNA were hybridized under high and low stringency conditions with a human OT exon B probe (Fig. 1). Under high stringency conditions, the known restriction fragments of the VP and OT genes were detected (Sausville et al., 1985). The VP fragment, which has 7% base pair mismatch to the used OT probe gave a weaker hybridization signal than the OT fragment. Under low stringency conditions two additional bands were found in each restriction enzyme digest. The sizes were 7.4 and 6.0 kb in the EcoRI digest, 14.0 and 18.0 kb in the Hind111 digest and 17.0 and 12.0 kb in the XbaI digest. In the BamHI digest, only one band was detected at 2.2 kb. In order to examine these low stringency signals, a genomic library was constructed using EcoRI fragments ranging from 4-8 kb.
The other four clones had inserts of approximately 7 kb and could therefore be responsible for one of the low stringency hybridization signals detected on genomic blots (Fig. 1). The EcoRI fragment of each clone was labeled and used as probe on human genomic blots under high stringency hybridization conditions. The pattern of hybridization of clone #4 corresponded to all the low stringency bands previously detected with the human OT exon B probe. In the BamHI digest fragments of 5.5 kb, 2.2 kb and 1.4 kb were detected; in the EcoRI digest 7.4 kb and 6.0 kb fragments were detected, in the Hind111 digest the detected fragments were 14.0 and 18.0 kb in size and in the XbaI the fragments were 17.0 and 12.0 kb. Clone #3 contained a repetitive DNA sequence. Clones #l, #2 and #5 showed dissimilar patterns, showing that they were different from the genomic sequences to which the human OT exon B probe had bound on genomic blots. Characterization of clones #4 and #5
Selection and cloning of genomic DNA with homology to VP and OT genes
Screening of a size-selected genomic library (5.105 independent plaques) under low stringency conditions with the human OT exon B probe yielded five positive clones. The clone showing the strongest hybridization signal (clone #5) contained a genomic insert of 4.8 kb.
Barn HI #4
OT
Eco RI #4
OT
DNA fragments that hybridized to exon B of the human OT gene were subcloned and sequenced. Figure 2 shows the sequence and analysis of a 157 bp HjraII fragment of clone #4. Comparison of the nucleotide sequences of the HpaII fragment and exon B showed an identity of 57.4% over a GC-rich region and thus explained the ability of the OT probe to hybridize
Hind Ill #4
OT
Xba I #4
OT
Fig. 3. Low stringency hybridization with 32P-labeled HpnII fragment of clone #4. Results shown are from the same human genomic blot as presented in Fig. 1. Hybridization conditions are as described in Materials and methods.
Discussion
to clone #4. The fragment had six open reading frames, but no amino acid similarity to human OT exon B was detected in any of them. A search for similarity of the sequenced fragment with nucleotide sequences deposited in the GenEMBL bank (updated 4/91) showed the closest resemblance, 71.1% identity over 1.52 nucleotides, with bovine herpes virus type 1 ~135 cDNA (Accession M84464). The 157 bp &@a11fragment of clone #4 was used as probe on human genomic blots under high and low stringency conditions. The pattern of hybridization revealed that this clone represented the low stringency signal of 7.4 kb detected previously by exon B of the OT gene in the EcoRI digest. Both under low and high stringency conditions, the 6.0 kb EcoRI fragment was also detected indicating a close similarity of this band with the &xzII fragment of clone #4. However, the VP and OT gene fragments were not detected by the 157 bp H&II fragment of clone #4 under the conditions used (Fig. 3). The nucleotide sequence of the 167 bp Sau3A fragment of clone #5, that hybridizes to exon B of the OT gene, is presented in Fig. 4. This Sau3A fragment of clone #5 was identical to nucleotides 405-571 of human bone morphogenetic factor hBMP-6, a member of the TGF-8 family (Celeste et al., 1990, Accession M38694). This fragment displayed a 67% identity over 85 nucleotides with exon B of the OT gene, but no significant amino acid similarity was observed between the identified region of hBMP-6 and the constant region of the human OT-associated neurophysin.
A
clone#S
This study was initiated to answer the question whether additional members of the VP-OT gene family exist in mammals. Several suggestions in this direction, based on experimental findings at the peptide level have persistently been made (Sausville et al., 1985; Pave1 et al., 1973; Chauvet et al., 1983; Lee, 1988; Geenen et al., 1991; Bonner and Brownstein, 1984; Nicholson et al., 1988; Choy and Watkins, 1988; Rosenbaum et al., 1990; Morley et al., 1990; Urano et al., 1992). In a previous study, Liu et al. (1988) ruled out the presence of vasotocin in the rat pineal gland by isolating modified peptides, rather than new forms of vasopressin and oxytocin. In this way the claim of an additional VP-OT-related peptide was refuted by experimental results. Here, we approached the question from the nudeotide level. Human genomic bIots were screened with exon B of the human OT gene as probe. This sequence codes for the evolutionary most conserved region of the VP and OT genes. It displays even to the distant vasotocin and isotocin genes in fish homologies between 64% (VT-l) and 69% (VT-21 (Morley et al., 1990). The mesotocin gene has 71% nucleotide homology to the used human OT exon B probe. The hybridization conditions, which were set to detect sequences with 60% or more homology to the human OT exon B probe, allowed detection of these sequences or sequences in the human genome as distantIy related to the human OT exon B probe as fish sequences. Sequences related to neurophysin but with
OTexB
AGATCTTGTiGGTGCTGGG&TCCCGCAC;GGCCCCGGC:...CC IIIII I I III III I GGGCTGCTTCGTGGGCACCGCCGAAG;G
clone#S
CTGC~CGGCCTCC~CAGCCG~G~CCCCGGCGC~CCGGCAG..: II0
I
III
I
II
ii
lflllllIlll
41 935 83
OTexB
CTGCGCTGCCAGGAGGAGAACTACCTGCCGTCGCCCTGCCAGTCC
980
clonet5
...CAGGAGiAGCAGCAGCiGCAGCAGCAiCTGCCTCGCdGAGAG
125
III
III
I
II
I
III
I
II
lllll
OTex
GGCCAGAAGGCGT.GCGGGAGCGGGGGCCGCTGC
cloneX5
CCCC~TCCCGGGCG~CTG~GTCC~CGCCCCTCT~CATGCTG
clone?5
EXLSVLGLPHRPRPLHGLQ:.QPQPPALR~EEQQQQQQiPRGEPPPGR~KSAPLFML I III Ill II PGGKGRCFGPNICCAESLGCFVGTAERLRCPEENYLPSPC
1013 167
B neuro-
56 58
physin
Fig. 4. Sequence analysis of clone #5. (A) Nucleotide sequence of the Suu3A fragment in clone #5 that hybridized to exon B of the human OT gene and homology to human GT exon B (Bestfit program, Devereux et al., 1984). Within this fragment, a region of 107 nucleotides was found with 63% sequence identity to the OT exon B probe. (B) Amino acid sequence alignment between the sequenced Suu3a fragment of clone #5 and human OT exon B (Bestfit program, Devereux et al., 1984). Gaps were introduced to optimize the alignment.
66
less than 60% homology escaped detection in this study. Although the OT exon B probe has a high G/C content (67%) we detected under low stringency conditions only two distinct bands in each restriction enzyme digest in addition to the VP and OT gene fragments. These nucleotide sequences had 60% similarity to neurophysin. However, none of these sequences appeared to encode a protein related to the amino acid sequence of neurophysin. We therefore conclude that the human genome does not contain a third closely-related member of the VP-OT gene family. Both human genomic blots and the size-restricted (4-8 kb) human genomic library were screened under low stringency conditions revealing DNA fragments with moderate homology to the used human OT exon B probe. One of these fragments (clone #4) was identified to be responsible for both low stringency hybridization signals previously found on the Southern blot (fig. 1 and 3). A close analysis of this clone and of the clone causing the strongest hybridization signal in the genomic library (clone #5) revealed the regions with homology of 60-70% to OT exon B (fig. 2B and 4B). Clone #5 shows 67% similarity to the OT exon B probe which is concentrated in a small sequence only. This may be the reason why it was detected in the genomic library but not, at first instance, on the genomic blot. At the amino acid level, no resemblance was observed between the sequenced fragments and the translated exon B probe (fig. 2C and 40. Regarding clone #5, no significant similarity was found between either the complete cDNA or the amino acid sequences of human OT and hBMP-6. Furthermore, the exon B probe bound to a region of the hBMP-6 gene that is not conserved among members of the TGF-B family. Any resemblance between the VP-OT family and the TGF-B family seems therefore of no significance. The hybridization characteristics of clone #4 further excluded the presence of neurophysin-related sequences in the human genome. Under high stringency conditions, clone #4 (7.4 kb in size) hybridized to all the low stringency bands previously found on the Southern blot with the human OT exon B probe. Under low stringency conditions, clone #4 failed to detect the VP-OT gene fragments on the Southern blot (Fig. 3). This indicates that the other low stringency fragment (6.0 kb EcoRI fragment) has a higher degree of similarity to clone #4 than to the human OT exon B probe. Based on these results, we cannot provide evidence that the human genome contains genes encoding neu-
rophysin-related peptides in addition to the VP and OT genes. The possibility that VP-OT-like peptides are encoded by unrelated genes with a VP-OT-like nucleotide sequence that lack the neurophysin portion cannot be excluded by our results. In theory, the use of another first exon spliced to the neurophysin encoding exons might lead to the synthesis of peptides related to but different from authentic VP and OT (Foo et al., 1991). However, it is more likely that the observations of related peptides originate from posttranslational processing events as indicated by Liu et al. (1988). References Bonner, T.I. and Brownstein, M.J. (19841 Nature 310, 17. Celeste, A.J., Iannazzi, J.A., Taylor, R.C., Hewick, R.M., Rosen, V., Wang, E.A. and Wozney, J.M. (1990) Proc. Natl. Acad. Sci. USA 87, 9843-9847. Chauvet, M.T., Chauvet, D.H.J. and Acher, R. (1983) Gen. Comp. Endocrinol. 49, 63-72. Choy, V.J. and Watkins, W.B. (1988) Neuropeptides 11, 119-123. Devereux, Haeberli and Smithies (1984) Nucleic Acids Res. 12, 387-395. Dogterom, J., Snijdewint, F.G.M., Pevet, P. and Swaab, D.F. (1980) J. Endocrinol. 84, 115123. Feinberg, A.P. and Vogelstein, B. (1983) Anal. Biochem. 132, 6-13. Fisher, L.A. and Fernstrom, J.D. (1981) Life Sci. 28, 1471-1481. Foo, N.C., Carter, D., Murphy, D. and Ivell, R. (1991) Endocrinology 128, 2118-2128. Gainer, H. and Wray, S. (1992) in Oxytocin in Maternal, Sexual, and Social Behaviors, Annals of the New York Academy of Sciences, vol. 652, 14-28. Geenen, V., Robert, F., Martens, H., Benhida, A., De Giovanni, G., Defresne, M-P., Boniver, J., Legros, J-J., Martial, J. and Franchimont, P. (1991) Mol. Cell. Endocrinol. 76, C27C31. Hara, Y., Battey, J. and Gainer, H. (19901 Mol. Brain Res. 8, 319-324. Lee, J.N. (1988) Program of the 8th International Congress of Endocrinology, Kyoto, Japan, p. 203 (Abstract). Liu, B., Poulter, L., Neascu, C. and Burbach, J.P.H. (19881 J. Biol. Chem. 263, 12-15. Nieuwenhuis, J.J. (1984) Life Sci. 35, 1713-1724. Mohr, E., Schmitz, E. and Richter, D. (1988) Biochimie 70,649-654. Morley, SD., Schonrock, C., Heierhorst, J., Figueroa, J., Lederis, K. and Richter, D. (1990) Biochemistry 29, 2506-2511. Negro-Vilar, A., Sanchez-France, F., Kwiatkowski, M. and Samson, W.K. (1979) Brain Res. Bull. 4, 789-792. Nicholson, H.D., Smith, A.J., Birkett, S.D., Denning-Kendall, P.A. and Pickering, B.T. (1988) J. Endocrinol. 117, 441-446. Pavel, S., Dorcescu, M., Petrescu-Holban, R., Ghinea, E. (1973) Science 181, 1252-1253. Rosenbaum, L.C., Neuweh, E.A., Van Tol, H.H.M., Peng Loh, Y., Verbalis, J.G., Hellstrom, I., Hellstrom, K.E. and Nilaver, G. (1990) Proc. Natl. Acad. Sci. USA 87, 9928-9932. Sausville, E., Carney, D. and Battey, J. (1985) J. Biol. Chem. 260, 18, 10236-10241. Schmale, H. and Richter, D. (1984) Nature 308, 705-709. Urano, A., Hyodo, S., Suzuki, M. (1992) Progr. Brain Res. 92, 39-46.