Gene, 113 (1992) 245-250 © 1992 Elsevier Science Publishers B.V. All rights reserved. 0378-1119/92/$05.00

GENE 06399

245

Structure and sequence of the growth hormone-encoding gene from Tilapia nilotica

(Recombinant DNA; fish; polymerase chain reaction; cloning; exon/intron arrangement; promoter regions)

Raphael Ber and Violet Daniel

Department of Biochemistry, The Weizmann hlstitute of Science, Rehovot, 76100 Israel

Received by A. Kohn: 17 November 1991 Accepted: 3 January 1992 Received at publishers: 31 January 1992

SUMMARY

We report here the nucleotide (nt) sequence of the growth hormone (GH)-encoding gene (GH) of the tilapia fish (Tilapia nilotica). The T. nilotica GH gene, similar to that of the salmonidae fish, Atlantic salmon and rainbow trout, contains six exons and five introns. However, despite the presence of an additional intron (intron V), the size of the primary transcript of T. nilotica GH (1666 nt) is significantly shorter than that of all other currently characterized fish GH genes. Compari- son of sequences upstream from the transcription start point of the tilapia, carp, rainbow trout and Atlantic salmon GH genes shows a region of high homology preceding the typical TATA box. This homology does not seem to extend to the regions further upstream of the compared fish GH genes and is not observed to be present in the corresponding region of the mammalian GH genes. A sequence search for putative DNA-binding domains for transcription factors shows the presence of short nt stretches similar to those considered to be involved in the tissue-specific expression of mammalian GH

genes.

INTRODUCTION

Growth hormone (GH), a single-chain polypeptide which is synthesized and secreted by the somatotrophs in the anterior pituitary gland, plays an important role in the growth and development of vertebrates. The growth hor- mone has been extensively studied at the levels of protein, mRNA and genomic genes in a variety of species. The nt sequence of GH genes has been determined for human (DeNoto et al., 1981; Seeburg, 1982), rat (Barta et al., 1981;

Correspondence to: Dr. V. Daniel, Department of Biochemistry, Weiz- mann Institute of Science, Rehovot, 76100 (Israel)Tel. (972-8)343639; Fax (972-8)344118.

Abbreviations: bp, base pair(s); GH, growth hormone; GH, gene (DNA) encoding GH; kb, kilobase(s) or 1000 bp; nt, nucleotide(s); oligo, oligode- oxyribonucleotide; PCR, polymerase chain reaction; tsp, transcription start point(s).

Page et al., 1981), porcine (Vize and Wells, 1987), bovine (Woychik et al., 1982), ovine (Orian et al., 1988), rainbow trout (Agel!on et al., 1988a), Atlantic salmon (Johansen et al., 1989), and carp (Chiou et al., 1990) genomes. These studies have shown that while the coding regions of the GH genes are highly conserved in mammals, the fish GH se- quences present less similarity to those of the other species. The size and structure of the GH gene were observed to vary among vertebrate species. Thus, the GH gene of salmonids, rainbow trout (Salmo gairdneri) and Atlantic salmon (Salmo salar) contain an additional intron (intron V) and the size of their primary transcript is about twice as compared to that of mammalians (Agellon et al., 1988a; Johansen et al., 1989). On the other hand the size and structure of another fish GH gene, that of the carp (Cyprinus carpio), was found to be more similar to the mammalian GH genes (Chiou et al., 1990). Due to the fact that the study of the sequence and structure of GH genes provides useful data for evolutionary analysig of vertebrate species

246

1 2 3 4

Fig, 1. PCR amplification of ~hc red tilapia GH mRNA. First-strand eDNA prepared from brain arrd pituitary total RNA, was used for PCR amplification using GH specific oligos 5'-GAACTGATGCCAG- CCATGA, 5'-AGCTACAG,~GTGCAGTTTG and Taq polymerase en- zyme (Saiki et al., 1988). The products of 30 cycles of PCR amplification (2 rain denaturation at 95°C, 2 5 rain annealing at 50°C, and 2 min elon- gation of primers at 72°C), were analyzed by 1% agarose gel electro- phoresis. Lanes I and 2 are controls to the PCR reactions: in the absence of RNA (lane I) and 5 ltg RNA without eDNA synthesis (lane 2). Lanes 3 and 4 show the PCR products of reactions prepared from 5 ltg and 1/ag total RNA, respectively.

we were interested to ct'.aracterize the GH gene from an- other teleost fish, the tilapia.

EXPERIMENTAL AND DISCUSSION

(a) Cloning and characterization of Philippine red tilapia GH cDNA

Total RNA isolated from Philippine red tilapia brain and pituitary tissues was incubated with reverse transcriptase and deoxynueleoside triphosphates, to synthesize th¢ first eDNA strand. Then, using two tilapia GH-specifie oligo primers, one corresponding to nt 21-39, 5 ' -GAACTG- ATGCCAGCCATGA, and the other complementary to nt 652-634, 5 '-AGCTACAGAGTGCAGTTTG, derived from tilapia (Oreochromis niloticus) GH eDNA (Rentier- Delrue et al., 1989), and Taq DNA polymerase, we have amplified by the PCR method the eDNA corresponding to

G A T C

,2/,.

Fig. 3. Determination of the tsp of tilapia GH gene. Using the .oligo 5 ' - CATGGCTGGCATCAGTTC, primer-extension products were gener- ated from tilapia brain and pituitary total RNA (lane i) and tilapia gills total RNA (lane 2) as negative control. Lane~ G, A, T, and C are the dideoxy sequencing product markers using the same oligo primer and the tilapia GH gene 5'-flanking DNA fragment of tGHLIB7 clone as tem- plate. Arrows indicate the adenine residues corresponding to the 5' end of the GH mRNA.

tilapia OH mRNA. Analysis of the PCR synthesis prod- ucts reveals the formation of a single DNA chain of 631 nt (Fig. 1). This DNA molecule was subsequently cloned into the Smal site of pGEM-I plasmid and sequenced. The sequence of the cloned red tilapia GH eDNA was found to contain the entire translated region of the GH which, as expected, spans the two oligo boundaries used as PCR primers. This sequence is found to be identical with that of

I t

the tila.pia Oreochromis niloticus GH eDNA described by Rentier-Delrue et al. (1989). The PCR products were also directly subjected to sequencing before cloning, to verify whether one or more populations of eDNA fragments have been amplified. Since the analysis of independent PCR

iv"

II

m

ql-

-~ A f./) u )

I-,,,,---4 O0 bp

9 4 . . . .

Fig. 2. Restriction map of the tilapia GH gene and DNA sequencing strategy of subcloned fragments in pGEM-I. Arrows indicate direction and extent of dideoxy (Sanger et al., 1977) DNA sequencing. P represents tt,e TATA box, and A represents the putative polyadenylation signal. An open box in- dicates transcribed and untranslated regions, a blackened box indicates the coding region. Thin lines between the boxes indicate introns.

247

reactions reveals the same sequence, it appears that only one eDNA species was amplified. This finding, together with Northern-blot analysis (data not shown) suggests that the tilapia genome contains only one expressed GH gene, differing in this respect from the common carp (Koren et al., 1989; Chiou et al., 1990) and rainbow trout genomes (Agellon et al., 1988b) which contain two GH genes. How- ever, since this comparison relates only to the translated region arJd not to the full length of GH eDNA, one can- not exclude the possible existence of more than one GH gene in tilapia.

(b) Cloning and endonuclease mapping of the tilapia GH gene

DNA was purified from gonads of a single T. nilotica fish and a genomic library was prepared by partial digestion

with Sau3AI, and cloning into 2EMBL3 phage vector. The genomic library was screened by plaque h.Jbridization (Benton and Davis 1977) using two probes corresponding to nt 63-187 and 340-651 of the tilapia GH eDNA. From the screening of 4.4 x 105 2EMBL3-tilapia recombinant DNA clones, three phages were found to eor~tain tilapia GH sequences as detected by hybridization to both eDNA probes. All three clones (designated tGHLiB 1, tGHLIB3 and tGHLIB7) were analyzed by restriction mapping, sub- cloning of DNA fragments into pGEM- 1, and sequencing. The bidirectional sequencing strategy used for the study of one of these clones (tGHLIB7) is illust:'ated in Fig. 2.

(c) Transcription start point (tsp) The tsp was determined by primel-extension analysis

using the oligo 5'-CATGGCTGGCATCAGTTC, which

gtcgaectttatttcagaattcagtttaatgactgaatgttcagtcaacaaaaggtgattgtgagtc -407 atcaaagcgagcactactgtctgtcaccatgaaaacctgaacacatgtcaaactgtcggtgtgaaca -340 tttctgtcccgactcattcgtgagcatgtgcgttcaaagtgacttctctgtggatgaaagctcgaat -273 aaaatcttttcacctgaaaaatttgatttgatgtaaatattttggagtttttgaaaactttacatta -206 gatctccttttaaatgttaacaaacatgttggatgtcagagcactttagtcaccagcggtgtgtttt -139 tcatgtagcatcaatatgatgattgaggtgttaatgttaattagacttagacacacgtgtttgctca -72 tgtcacgtttctcctgatgaatttaaacatctagttttcaactataaaagcaaaaactctgagctga -5

+1 aaacATCAGAACCACCGACTCACATCATAATCATCTGAGCCGCAAACAGAGCCTGAACTGATGCCAG 63

M N S CC ATG AAC TCA Ggtaagacatctgggctcccccacgagaagggaccactqctttatgatgttt 126

V V L L L S aacaaagtctgaaactgtctgtctgtctgtctgtctgt~tgtcagTC GTC CTC CTG CTG TCG 188 V V C L G V S S Q Q I T D S Q R L

GTT GTG TGT TTG GGC GTC TCC TCT CAG CAG ATC ACA GAC AGC CAG CGT TTG 239 F S I A V N R V T H L H L L A Q R

TTC TCC ATT GCA GTC AAC AGA GTC ACG CAC CTG CAC CTG CTC GCC CAG AGA 290 L F S D F

CTC TTC TCG GAC TTT gtaagcctgcagcagctcaacaatctttcttctttctgaaaaagacc 352 aaatgttacctaaatcaaagctaatgcacaggacagaaactaggttcaaaatacgttcaacaaaatg 419 ttctggatattcagtgtgtgcagtgagtttagatgcacacagacatatggacacatttcacatttga 486 tgtcaagggaaccgagacactttgtagdctgtcactgctaaaaacacagcagacgtttacactttac 553 attttaatgacagtcagctataatctcaggatattcagttaagaattatgaaacgatattaaaattt 620 gctgtcagtcaataaaacacaggtattcctgtgtattttgtgtatttctgtatgcacctgtgtattt 687

E S S L Q T E ttgtgtgttaccattctttaattctacacatgtcatag GAG AGC TCT CTG CAG ACG GAG 746 E Q R Q L N K I F L Q D F C N S D

GAG CAA CGT CAG CTC AAC AAA ATC TTC CTG CAG GAC TTC TGC AAC TCT GAT 797 Y I I S P I D K H E T Q R S S

TAC ATC ATC AGC CCG ATC GAC AAA CAC GAG ACG CAG CGC AGC TCG gtcatt 848 aaactacacagtactacgcaacactgcacagcagtacacagtactacacagtaccagaggtactctg 915

V L K L cccatgcgataatcttctgaagctgttgtctgcttattgatgggtcacag GTC CTG AAG CTG 977 L S I S Y G L V E S W E F P S R S

CTG TCG ATC TCC TAT GGA CTG GTT GAG TCC TGG GAG TTT CCC AGT CGC TCT 1028 L S G G S S L R N Q I S P R L S E

CTG TCT GGA GGT TCC TCT CTG AGG AAC CAG ATT TCA CCA AGG CTG TCT GAG 1079 L K T G I L L L I R

CTT AAAACG GGA ATC TTG CTG CTG ATC AGG gtgagaagataaattgcacaaacatat 1136 A N Q D E

tgctccatgcatggtaccaatgcgactaaccacctgactctacag GCC AAT CAG GAT GAA 1196 A E N Y P D T D T L Q H A P Y G N

GCA GAG AAT TAT CCT GAC ACC GAC ACC CTC CAG CAC GCT CCT TAC GGA AAC 1247 Y Y Q S L G G N E S L R Q T Y E L

TAT TAT CAA AGT CTG GGA GGC AAC GAA TCG CTG AGA CAA ACT TAT GAA TTG 1298 L A C F K K D M H K

CTG GCT TGC TTC AAG AAG GAC ATG CAC AAG gtgaggtagtggataatggtgatgtca 1355 V E T Y

ctgtgatgatgacaatgatgtaatgatggtgaagatgacatttttgttgcagGTG GAG ACC TAC 1419 L T V A K C R L S P E A N C T L ;LM

CTG ACG GTA GCT AAA TGT CGA CTC TCT CCA GAA GCA AAC TGC ACT CTG TAG 1470 CTCCACCTAATATTGATACTGATACGTGCTCTGTAGCCCCACCCTCATGTTGGCAAACTCTGCTTAC 1537 ATGTGTTAGCATTAGCAATAGGATAATAATAGCAGTGGTAATCGTGACATCAGAAGTTTTTCTGACA 1604 TAACTGTGATGCAAGGTGTGAACGGGAATAATGTTATTCTGTGAAATAAATGTGTTGCATTGatgtg 1671 tggagtctgttctttctgatttatttggagaatttattctctat~acttcttacacagaaac 1738 acatgctgacgtgagctgaggtccatatcgatggcagaaaaactgactcgaacatgtcagcttagaa 1805 atgatcctcttatggcctctgacagtggactcgtctctgtgcttgtcctgctagacctcagtgcagc 1872 gttcaatactgtcgac 1888

Fig. 4. The nt sequence of tilapia GH gene. The numbers on the right margin refer to the last nt on the corresponding line, with respect to the distance from the putative cap site ( + 1), which was determined by primer extension. Exons are shown in upper-case letters; introns and flanking regions are in lower-case letters. The TATA box and the putative poly(A) signal are underlined. Encoded aa residues, represented by one-letter symbols, are placed above the second nt of each codon. The sequence data in this paper have been deposited in the EMBL/GenBank libraries under accession No. M84774.

248

is complementary to nt 38-21 of the tilapia GH eDNA sequence. Primer extension reactions were can%d out with AMV reverse transcriptase after annealing the oligo probe with total RNA prepared from tilapia tissues. RNA was taken from brain containing also the pituitary, and from gill tissue as a negative control (Fig. 3). Using the same oligo as primer in the sequencing reaction of a DNA fragment containing the 5' end of the isolated GH gene as a marker, we located the tsp at an adenine residue, 23 bp downstream from the putative TATA box. The published tilapia GH eDNA sequence (Rentier-Delrue et al., 1989), which is about 30 nt shorter in the 5' region, represents an incom- plete eDNA probably due to reverse transcriptase disso- ciation during eDNA preparation.

(d) Nucleotide sequence of the tilapia GH gene Alignment of the genomic tilapia GH nt sequence with

that of the tilapia GH eDNA (Rentier-Delrue et al., 1989), together with the results of primer-extension analysis, al- lowed us to determine the structure of the tilapia GH gene (Fig. 4). Comparing the coding sequences determined by us with those from the GH eDNA, we observed only six nt differences, all in nontranslated (three nt in the 5' and three nt in the 3') regions. The GH gene of T. nilotica is com- posed of six exons (exon I, 75 bp; exon II, 134 bp; exon Ill, 117 bp; exon IV, 144 bp; exon V, 147 bp; exon VI, 259 bp) and five introns (intron I, 96 bp; intron It, 420 bp; intron Ill, 123 bp; intron IV, 72 bp; intron V, 79 bp). All five introns start with a GT dinucleotide and end with an AG, in agreement with the consensus splice site sequences ($eif et al., 1979). By having six exons and five introns, the tila- pia GH gene appears to be structurally similar to the GH gene of salmonids, rainbow trout and Atlantic salmon (Age,on et al., 1988a; Johansen et al., 1989). It differs therefore from the mammalian GH gene by having an ad- ditional intron interrupting exon V, thus forming one more exon (exon VI). The transcription unit of the tilapia GH gene spans a region of about 1.7 kb, which is by 2 kb smaller than that of salmonid GH gen¢ (Fig. 5). The tilapia gene is smaller than the carp GH gene which comprises only five exons and four introns spanning about 2.4 kb, in a struc- ture resembling that of mammalian GH genes. Tilapia car- ries therefore the currently identified smallest fish GH gene, with structural motives similar to those of the largest GH

Atlantic salmon--J3 ~.~ = ~ . _ _ ~ _ _

rainbow trout --13 ~.~ ~= ~ ~.~ _ ~ - -

t i lapia ---13 - - I---

carp --13 ~.~ = : i - -

human --t3

rat ---t3 ~ _ _.

bovine --13 I,-,--t

ovi ne ---t3 ~ ~o bp

Fig. 5. Comparison of the GH gene structure of rainbow trout, Atlantic salmon, tilapia, common carp, human, rat, bovine and ovine. An open box represents transcribed and untranslated regions, a blackened box indi- cates the coding region. Thin lines between the boxes indicate introns.

genes thus far identified. This finding supports the hypo- thesis that the introduction of intron V into the fish GH gene took place after the divergence of fish and tetrapodes, and tilapias may represent a middle stage in the evolution of fish GH genes.

(e) Analysis of 5' and 3' flanking sequences of the GH gene The 5' flanking region contains the sequence TATAAA

at a distance of 23 nt upstream from the tsp. An identical TATA box sequence (Breathnach et at., 1981) has been identified in all currently characterized fish GH genes, i.e., rainbow trout (Agellon et ai., 1988a), Atlantic salmon (Jo- hansen et ai., 1989), and common carp (Chiou et at., 1990). A putative polyadenylation signal AATAAA (Proudfoot et al., 1976) is located at position + 1649 in the 3'-flanking region, It is identical to the polyadenylation signal in the Sabno species but differs from the carp poly(A) signal AT- TAAA.

Analysis of the 473-nt sequence 5' upstream from the tsp of the tilapia GH gene did not reveal any GH consensus promoter elements other than the TATA box, nor any ob- vious DNA-binding sites for either glucocorticoid or thy- roid hormone-receptor complexes which have been local- ized and shown to play a role in the hormonal regulation of the mammalian GH genes (Moore et al., 1985; Barlow et al., 1986; Martial et al., 1977; Samuels et at., 1979; Dob- ner et al., 1981). A sequence search for putative DNA- binding domains for transcription factors, revealed the presence of sequences CTTTACAT, ACAAACAT and TTAAACAT at the locations of -2i5, -186 and -49, respectively, which are similar to the consensus core DNA

ATLANTIC SALMON

RAINBOW *ROUT

TILAPIA

COITION CARP

GGAAATCTCATGTTTCCTCCTG ....... TTGATACATTAAAACATGGGTTATCCATC~ IIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIII III lllllllllllll GGAA**C*CATGTT******** ....... ************************************ I II III IIII IIIIII III III IIIIII I I I II IIIIIIII GCTC**G*CACGTT*-***C** ....... *****-****************************** IIIIIII I I III II III III IIIIII I I II ~IIIIII

TCTC**G*TTCAAC*-***A**CATTAAGA**C---***A******GTTCATGC**CA|******

A A C A - 2 3 I I I I

A A * * - 2 3 I I I

&G** - 2 3

TATC -22

Fig. 6. Alignment of sequences flanking the TATA box in the promoter region of fish GH genes. Asterisks represent nt that are identical in all four se- quences, Dashes were introduced to obtain best alignment. The number at the right of each sequence indicates the distance from the tsp. The putative TATA box is framed.

binding site of the mammaliar, GHF- 1/PIT- 1 transcription factor (Bodner and Karin, 1987; Nelson et al., 1988). How- ever, the significance of these sequences for the transcrip- tion of the GH gene in the fish genomes is yet unknown, and is under investigation in our laboratory. A sequence search for perfect and imperfect direct and inverted repeats revealed many such structures throughout the gene. We note a direct repeat of the octanucleotide 5'-aatgttaa be- tween nt -193 and -186 and nt -106 and -99 in the promoter region. The significance of these structures is, however, unknown.

(f) Comparison of the 5'-flanking region of tilapia, Atlantic salmon, rainbow trout and carp GH genes

Tilapia promoter region shows a very low similarity to mammalian GH gene promoters. However, when com- pared to the currently known fish genes a conserved se- quence in the region flanking the TATA box is noted (Fig. 6). Highly conserved regions usually reflect evolution- ary pressure to maintain structures of functional signifi- cance. The relatively small size of the conserved region may indicate that the genetic information for the tissue specific expression of the fish GH gene is exclusively associated with the proximal region of the 5'-flanking region of the gene. Such a situation would be similar to that of the human GH gene where observations indicate that the regulatory sequences are located within the proximal 200 bp of the 5'-flanking region. Since the highly conserved region in fish has no obvious homologous counterpart in the mammalian GH gene, nor any obvious thyroid/glucocorticoid respon- sive elements, we cannot exclude the possibility that the tissue specificity and expression regulation mechanisms in fish pituitary may be different from that in mammalian. This hypothesis is under current investigation in our labo- ratory.

(g) Conclusions (1) Philippine red tilapia GH eDNA isolated using the

PCR technique, is found to be identical in the translated region to the eDNA of tilapia Oreochromis niloticus.

(2) Primer extension and nt sequencing of the tilapia GH gene, indicate a primary transcript of 1666 nt.

(3) The tilapia GH gene contains one additional intron and exon as compared to mammalian or carp GH genes, and resembles the salmonid GH genes in this respect. The additional intron separates exons V and VI, which are equivalent to the region encoded by exon V in the mam- malian gene.

(4) Analysis of upstream sequences of the tilapia, carp, rainbow trout and Atlantic salmon GH genes shows that they all contain a small highly conserved region which in- cludes the consensus TATA box.

(5) The upstream sequences oftilapia, like those of other

249

fish GH genes, present a low similarity to the correspond- ing region of the mammalian GH genes.

ACKNOWLEDGEMENTS

We thank Prof. Avtalion Rami for supplying the tilapia pituitary and gonad tissues. This work was supported in part by a grant from the National Council for Research and Development Israel and the GKS S, Geesthacht, Germany.

REFERENCES

Agellon, L.B., Davies, S.L., Chen, T.T. and Powers, D.A.: Structure of a fish (rainbow trout) growth hormone gene and its evolutionary im- plications. Prec, Natl. Acad. Sci. USA 85 (1988a) 5136-5140.

Agellon, L.B., Davies, S.L., Lin, C.-M., Chen, T.T. and Powers, D.A.: Rainbow trout has two genes for growth hormone. Mol. Rep. De- velop. I (1988b) !1-17.

Barlow, J.W., Voz, M.LJ., Eliard, P.H., Mathy-Hartert, M., De Nayer, P., Economidis, I.V., Belayew, A., Martial, J.A. and Rousseau, G.G.: Thyroid hormone receptors bind to defined regions of the growth hormone and placental lactogen genes. Prec. Natl. Acad. Sci. USA 83 (1986) 9021-9025.

Barta, A., Richards, R.I., Baxter, J.D. and Shine, J.: Primary structure and evolution of rat growth hormone gene. Prec. Natl. Acad. Sci. USA 78 (1981) 4867-4871.

Benton, W.D. and Davis, R.W.: Screening 2gt recombinant clones by hybridization to single plaques in situ. Science 196 (1977) 180-182.

Bodner, M. and Karin, M.: A pituitary-specific trans-acting factor can stimulate transcription from the growth hormone promoter in extracts of nonexpressing cells. Cell 50 (1987) 267-275.

Breathnach, R. and Chambon, P.: Organization and expression of eu- caryotic split genes coding for proteins. Annu. Rcv. Biochem. 50 (1981) 349-383.

Chiou, C.-S., Chen, H.-T. and Chang, W.-C.: The complete nucleotide sequence of the growth-hormone gene from the common carp (Cyprinus carpio). Biochim. Biophys. Acta 1087 (1990) 91-94.

DeNote, F,M., Moore, D.D. and Goodman, H.M.: Human growth hor- mone DNA sequence and mRNA structure: possible alternative splic- ing. Nucleic Acids Res. 9 (1981) 3719-3730.

Dobner, P.R., Kawasaki, E.S., Yu, L.-Y. and Bancroft, F.C.: Thyroid or glucocorticoid hormone induces pre-growth-hormone mRNA and its probable nuclear precursor in rat pituitary cells. Prec. Natl. Acad. Sci. USA 78 (1981) 2230-2234.

Johansen, B., Johnsen, O.C. and Valla, S.: The complete nucleotide se- quence of the growth-hormone gene from Atlantic salmon (Sabno salar). Gene 77 (1989) 317-324.

Koren, Y., Sarid, S., Bet, R. and Daniel, V.: Carp growth hormone: molecular cloning and sequencing of eDNA. Gene 77 (1989) 309-315.

Martial, J.A., Baxter, J.D., Goodman, H.M. and Seeburg, P.H.: Regu- lation of growth hormone messenger RNA by thyroid and glucocor- ticoid hormones. Prec. Natl. Aead. Sci. USA 74 (1977) 1816- 1820.

Moore, D.D., Marks, A.R., Buckley, D.I., Kapler, G., Payvar, F. and Goodman, H.M.: The first intron of the human growth hormone gene contains a binding site for glucocorticoid receptor. Prec. Natl. Acad. Sci. USA 82 (1985) 699-702.

Nelson, C., Albert, V.R., Elshoitz, H.P., Lu, L.I.-W. and Rosenfeld, M.G.: Activation of cell-specific expression of rat growth hormone

250

and prolactin genes by a common transcription factor. Science 239 (1988) 1400-1405.

Orian, J.M., O'Mahoney, I.V.O. and Brandon, M.R.: Cloning and se- quencing of the ovine growth hormone gene. Nucleic Acids Res. 16 (1988) 9046.

Page, G.S., Smith, S. and Goodman, H.M.: DNA sequence of the rat growth hormone gene: location of the 5' terminus of the growth hor- mone mRNA and identification of an internal transposon-like ele- ment. Nucleic Acids Res. 9 (1981) 2087-2104.

Proudfoot, N.J. and Brownlee, G.G.: 3' Non-coding region sequences in eucaryotic messenger RNA. Nature 263 (1976) 211-214.

Rentier-Delrue, F., Swennen, D., Philippart, J.C., L'Hoir, C., Lion, M., Benrubi, O. and Martial, J.A.: Tilapia growth hormone: molecular cloning of eDNA and expression in Escherichia coli. DNA 8 (1989) 271-278.

Saiki, R.K,, Geifard, D.H., Stoffel, S., Scharf, SJ., Higuchi, R., Horn, G.T., Mullis, K.B. and Erlich, H.A.: Primer-directed enzymatic am- plification of DNA with a thermostable DNA polymerase. Science 239 (1988) 487-491.

Samuels, H.H., Stanley, F. and Shapiro, L.E.: Control of growth hor- mone synthesis in cultured GHI cells by 3,5,3'-triiodo-L-thyronine and glucocorticoid agonists and antagonists: studies on the indepen- dent and synergistic regulation of the growth hormone response. Bio- chemistry 18 (1979) 715-721.

Sanger, F., Nicklen, S. and Coulson, A.R.: DNA sequencing with chain- terminating inhibitors. Prec. Natl. Acad. Sci. USA 74 (1977) 5463- 5467.

Seeburg, P.H.: The human growth hormone gone family: nucleotide se- quences show recent divergence and predict a new polypeptide hor- mone. DNA 1 (1982)239-249.

Seif, I., Khoury, G. and Dhar, R.: BKV splice sequences based on anal- ysis of preferred donor and accepter sites. Nucleic Acids Res. 6 (1979) 3387-3398.

Vize, P.D. and Wells, J.R.E.: Isolation and characterization of the por- cine growth hormone gene. Gene 55 (1987) 339-M4.

Woychik, R.P., Camper, S.A., Lyons, R.H., Horowitz, S., Goodwin, E.C. and Rottman, F.M.: Cloning and nucleotide sequencing of the bovine growth hormone gene. Nucleic Acids Res. 10 (1982) 7197-7210.