Expression of endogenous and exogenous growth hormone (GH) messenger (m) RNA in a GH-transgenic...

Expression of endogenous and exogenous growth hormone (GH) messenger

(m) RNA in a GH-transgenic tilapia (Oreochromis niloticus)

Antje Caelers1, Norman Maclean2,*, Gyulin Hwang2, Elisabeth Eppler1 &

Manfred Reinecke11Division of Neuroendocrinology, Institute of Anatomy, University of Zurich, Zurich, Switzerland2Division of Cell Science, School of Biological Sciences, University of Southampton, Hampshire, SO16 7PX, UK

Received 7 May 2004; revised 31 August 2004; accepted 6 September 2004

Key words: absolute quantification, growth hormone, mRNA, ocean pout antifreeze promoter, transgenetilapia

Abstract

We have previously produced transgenic fish from crosses between a wild-type female tilapia (Oreochr-omis niloticus) and a G1 transgenic male. This line of growth-enhanced tilapia carries a single copy of achinook salmon (s) growth hormone (GH) gene spliced to an ocean pout antifreeze promoter (OPA-FPcsGH) co-ligated to a carp b-actin/lacZ reporter gene construct, integrated into the tilapia genome.Because little is known about the expression sites of transgenes, we have characterised the gene expres-sion patterns of sGH and tilapia (t)GH in transgenic tilapia using a newly established real-time PCR tomeasure the absolute mRNA amounts of both hormones. The sGH gene, which was expected to beexpressed mainly in liver, was also found to be expressed in other organs, such as gills, heart, brain,skeletal muscle, kidney, spleen, intestine and testes. However, in pituitary no sGH mRNA but only tGHmRNA was found. Tilapia GH mRNA in wild-type pituitary amounted to 226 ± 30 pg/lg totalRNA but in transgenics only to 187 ± 43 pg/lg total RNA. Liver exhibited the highest level of sGHmRNA (8.3 ± 2.5 pg/lg total RNA) but the extrahepatic sites expressed considerable amounts of sGHmRNA ranging from 4.1 ± 2.0 pg/lg total RNA in gills to 0.2 ± 0.08 pg/lg total RNA in kidney. Thewidespread expression of the sGH gene is assumed to be due to the tissue specificity of the type III AFPgene promoter. It is assumed that our transgenic experiments, which in contrast to some otherapproaches caused no obvious organ abnormalities, mimick the GH expression during ontogeny.Because sGH mRNA is expressed both in liver and in extrahepatic sites it may not only promote secre-tion and release of liver-derived (endocrine) IGF-I leading to an overall growth enhancement but alsostimulate IGF-I expression within the different organs in a paracrine/autocrine manner and, thus, fur-ther promote organ growth.

Introduction

Transgenic fish are of value both as model spe-cies in fundamental research and as potentiallygenetically superior brood-stock for commercialfood production (see Fletcher & Davies, 1991;Hew et al., 1995; Iyengar et al., 1996). The firsttransgenic fish were produced about 20 years ago

(Maclean and Talwar, 1984; Zhu et al., 1985),but considerable progress has been made sincethat time. Now, fish of many species have beensuccessfully transformed. One attractive commer-cially relevant scenario regarding transgenic fishlines is the production of growth-enhanced fishfollowing the introduction of novel growth hor-mone (GH) coding genes. Growth enhancementof transgenic fish using novel piscine GHsequences has been achieved by a few researchers*Author for correspondence

E-mail: nm4@soton.ac.uk

Transgenic Research 14: 95–104, 2005. � Springer 2005

in different fish species. Thus, Du et al. (1992)have demonstrated that, following introductionof a DNA construct (OPAFPcsGH), in which anocean pout (Macrozoarces americanus) antifreezepromoter (AFP) drives a chinook salmon (On-corhynchus tschawytscha) GH cDNA sequence,transgenic Atlantic salmon (Salmo salar) exhib-ited impressive growth enhancement. Later, Dev-lin et al. (1994) and Mori and Devlin (1999)observed a dramatic improvement in growth oftransgenic coho salmon (Oncorhynchus kisutch)by the expression of a sockeye salmon(Oncorhynchus nerka) GH gene driven by ametallothionein promoter (MT-B) of the samespecies. Nam et al. (2001) have also demon-strated dramatic acceleration of growth in trans-genic mud loach (Misgurnus mizolepis) aftermicroinjecting the mud loach GH gene fused tothe mud loach b-actin promoter. By using thepreviously mentioned OPAFPcsGH construct(kindly provided by Prof. C. Hew) which was co-injected with a carp ß-actin/lacZ reporter geneconstruct, dramatic growth enhancement hasbeen demonstrated in the tilapia, Oreochromisniloticus, in which the mean weight of the7 month old G2 transgenic fish was more than3-fold that of their non-transgenic siblings (Rah-man et al., 1997). Other commercially significantspecies in which transgenic growth enhancementhas been attempted include channel catfish(Dunham et al., 1987), rainbow trout (Chour-rout et al., 1986; Penman et al., 1990), commoncarp (Zhang et al., 1990; Rosochacki et al.,1993) and northern pike (Gross et al., 1992;Guise et al., 1992). Although probable integra-tion of the transgenes into the genome wasdemonstrated in all of these examples, in somecases no transgene expression was detectable.The aberrant or nil expression was sometimesattributed to the use of promoter sequencesfrom very distantly related species (Chourroutet al., 1986; Brem et al., 1988; Rokkones et al.,1989; Penman et al., 1990).

The regulatory sequence used by Maclean &Rahman (1997) is an AFP. Antifreeze proteins(AFPs) are synthesised in fish of many species,both from Arctic and Antarctic waters(Fletcher et al., 2001). All are small moleculessecreted into blood and tissue, which adsorb tosmall ice crystals and prevent their furthergrowth, thus lowering the serum freezing point

to that of the seawater. They are quite diverseand of at least five distinct types, the Antarcticfishes secreting glycoproteins with glycotripep-tide repeats (DeVries, 1983; Davies & Sykes,1997) known as type II antifreezes, while Arcticfishes of different families produce a range ofdifferent small AFP proteins. Although exten-sive studies have been carried out on the struc-ture and organisation of the antifreeze genes,much less is known about their expression. Thebest characterised of the latter is type I AFPfrom winter flounder (Pleuronectes americanus)and daddy sculpin (Cottus scorpius). This AFPis synthesised in liver and secreted into theblood (Hew et al., 1999) both in the winterflounder and transgenic Atlantic salmon (Salmosalar) expressing this gene, although a variantform is secreted in peripheral tissues (Gonget al., 1996). Type I AFP is made during andjust prior to the onset of Arctic winter. In con-trast, Northern Blot analysis in sea raven(Hemitripterus americanus) indicated that typeII APF mRNA is restricted to liver (Gonget al., 1992). The ocean pout (Macrozoarcesamericanus) synthesises a type III AFP which ismade year round, but the winter protein levelsare several fold higher than those in summer.Gene dosage of the AFP gene is also of inter-est, because Fletcher (2001) report 150 copiesin Newfoundland populations, but only 30–40copies in New Brunswick populations.

The promoter driving the GH gene in theconstruct used here is a type III AFP genefrom ocean pout. In 1992, Gong et al. showedby northern blot analysis that significantamounts of ocean pout type III AFP mRNAare present in many other tissues besides theliver. Until now nothing is known about themode of action of the AFP type III promoterinjected into a fish species, which normally doesnot produce AFPs. We therefore decided todetermine in which tissues (and also in whichamounts) the exogenous salmon GH (sGH),which is spliced to the ocean pout type IIIAFP, is expressed in tilapia fish transgenic forthe construct. We have chosen to work on thetilapia (O. niloticus) partly because of its majorimportance for developing world aquaculture(Nash, 1988), and also because of its naturallyrapid growth, omnivorous diet, and shortgeneration time.

96

Materials and methods

Production of transgene fish

The transgenic fish used in these experimentswere previously produced from crosses between awild type female O. niloticus and a transgenicmale bred at the University of Southampton,UK. This line of growth-enhanced tilapia (C86)carries a single copy of a chinook salmon (On-corhynchus tschawytscha) sGH gene spliced to anocean pout (Macrozoarces americanus) antifreezepromoter (OPAFPcsGH) co-ligated with a carpß-actin/lacZ reporter gene construct (19 copies),integrated into the tilapia genome (Rahmanet al., 1997).

Detection of the transgene in the fish

In order to verify the transgenic or non-trans-genic state of the individuals, PCR was carriedout on DNA from fin clips. An approximately3 · 2 mm clip was taken from the caudal fin ofeach individual investigated and immediately fro-zen in liquid nitrogen. Prior to fin clipping, fishwere tagged with transponders (Fish CultureResearch Institute, FCRI growth trial) to allowidentification of each fish after PCR analysis.Genomic DNA was isolated. The standard proce-dures for isolation and purification of DNA(Rahman & Maclean 1992) and subsequent PCRand Southern Blotting were performed as previ-ously described (Rahman et al., 1998) by usingprimers to detect novel junction fragmentsbetween the exogenous sGH gene and the repor-ter gene.

Animals

Seventeen months old males of the C86 strain(n ¼ 10) and of the non-transgenic siblings(n ¼ 10) were investigated. Fish of both groupswere reared in 260 l tanks at 24–25�C under a13 h/11 h light/dark cycle and fed with trout pel-lets three times a day to satiation as describedpreviously (Rahman et al., 1998). At that age,non-transgenic male tilapia have a mean lengthof 19 ± 5.5 cm, and a mean weight of179 ± 57 g. Also at that age, the transgenic fishshowed an average length of 26.5 ± 8.2 cm, anda mean weight of 398 ± 127 g.

Preparation of RNA

Fish were anaesthetised with 2-phenoxy-ethanol(Sigma, St. Louis, MO) added to water. Tissuespecimens were excised and immediatelytransferred to 1.5 ml of RNAlaterTM (Ambion,Austin, Texas, USA). Samples were kept at4�C overnight and stored at )20�C until RNAisolation. Total RNA was extracted using theRNAqueousTM-4 PCR Kit for isolation ofDNA-free RNA (Ambion) or TRIzolTM reagent(Invitrogen, Merelbeke, Belgium). With pitui-tary, five specimens were pooled to obtain asufficient amount of RNA. Total RNA wastreated with 1 U of RQ1 RNAse-free DNAse(Catalys AG, Wallisellen, Switzerland; whenTRIzolTM reagent was used) to digest DNAresidues, re-suspended in DEPC-treated waterand photospectrometrically quantified at A260.Purity of total RNA was assessed by the260/280 nm ratio (between 1.8 and 2.1).Additionally, integrity was assured with ethidi-um bromide-stain analysis of 28S and 18Sbands by formaldehyde-containing agarose gelelectrophoresis (data not shown).

Creation of TaqMan primers and probes

The mRNA sequences of O. niloticus growth hor-mone (tGH) (GenBank Accession No. M2916;Rentier-Delrue et al., 1989), Oncorhynchustshawytscha growth hormone (sGH) (GenBankAccession No. S50867; Song et al., 1992) andO. niloticus b-actin (GenBank Accession No.AY116536; Hwang et al., 2003) were used to gen-erate the corresponding cDNA sequences. Theintron–exon junctions were found by comparisonof the mRNA sequences with the complete genesequences of tGH (Genbank Accession No.M84774; Ber & Daniel, 1992) and Oncorhynchusnerka GH (GenBank Accession No. U14551;Devlin, 1993).

The sGH, tGH, and b-actin primers as well asan internal oligonucleotide probe were designedusing the Primer Express software version 1.5(PE Biosystems, Foster City, CA, USA). Theprocedure resulted in the selection of the forwardand reverse primers as summarised in Tables 1and 2. The internal probes were labelled at the5¢end with the reporter dye FAM (6-carboxyfluo-rescein), and at the 3¢end with the quencher dye

97

TAMRA (6-carboxytetramethyl-rhodamine, bothfrom Eurogentec, Herstal, Belgium).

Generation of cDNA

For cDNA synthesis 5 lg RNA were annealed

with 1 lM of a poly(dT) primer(5¢-CCTGAATTCTAGAGCTCAT(dT17)-3¢) at70�C for 3 min. The RNA/primer mix wasincubated at 37�C for 1 h with 10 mM dNTPsand 100 U Moloney Murine LeukemiaVirus (MLLV)-reverse transcriptase (Promega,

Table 1. Primers and probes

Primer/probe Sequence (5¢! 3¢) PCR fragment length (nt)

tGHs TCGACAAACACGAGACGCA

tGHas CCCAGGACTCAACCAGTCCA 75

tGHasT7 CTAATACGACTCACTATAGGGCCAGGACTCAACCAGTCCA 95

tGHprobe (i)CGCAGCTCGGTCCTGAAGCTGC(ii)

sGHs TTGGCTCAGAAAATGTTCAATGA

SGHas GGAATATCTTGTTCAGCTGTCTGC 76

sGHasT7 CTAATACGACTCACTATAGGGATATCTTGTTCAGCTGTCT 94

sGHprobe (i)TTTGACGGTACCCTGTTGCCTGATGAA(ii)

ß-actins GCCCCACCTGAGCGTAAATA

ß-actinas AAAGGTGGACAGGAGGCCA 65

ß-actinasT7 CTAATACGACTCACTATAGGGAAAGGTGGACAGGAGGCCA 86

ß-actinprobe (i)TCCGTCTGGATCGGAGGCTTCATC(ii)

(i) Reporter dye (FAM) labelled nucleotide.(ii) Quencher dye (TAMRA) labelled nucleotide asT7: antisense primer with T7 polymerase promoter (italics).

Table 2. Position of primers and probes

Actin

1018 ATGCAGAAGGAGATCACAGCCCTGGCCCCATCCACCATGAAGATCAAG

tActins ! Probe tActin !1066 ATCATCGCCCCACCTGAGCGTAAATACTCCGTCTGGATCGGAGGCTTC

tActinas

1114 ATCCTGGCCTCCTGTCCACCTTTCAGCAGATGTGGATCAGCAAGCAGG

tGH

231 CAGGACTTCTGCAACTCTGATTACATCATCCAGCCCGATCGACAAACAC

tGHs !�

Probe tGH !279 GAGACGCAGCGCAGCTCGGTCCTGAAGCTGCTGTCGATCTCCTATGGAC

tGHas

327 TGGTTGAGTCCTGGGAGTTTCCCAGTCGCTCTCTGTCTGGAGGTTCCTC

sGH

152 GGCTCTTCAACATCGCGGTCAGTCGGGTGCAACATCTCCACCTATTGGC

sGHs ! � Probe sGH !201 TCAGAAAATGTTCAATGACTTTGACGGTACCCTGTTGCCTGATGAACGC

sGHas

250 AGACAGCTGAACAAGATATTCCTGCTGGACTTCTGTAACTCTGACTCCA

�Intron between two nucleotides.

98

Wallisellen, Switzerland) in 1 · reactionbuffer (50 mM Tris-HCl (pH 8.3 at 25�C),75 mM KCl, 3 mM MgCl2, 10mM DTT).

Verification of primer accuracy

Aliquots of cDNA (1 ll) were incubated with1 lM sense (s) and antisense (as) primers,200 lM dNTPs and 1.25 U Taq-Polymerase(Quantum, Appligene, Illkirch, France) in 1 ·incubation buffer (10 mM Tris-HCl (pH 9.0 at25�C), 50 mM KCl, 1.5 mM MgCl2, 0.1% Tri-tonX 100, 0.2 mg/ml BSA or gelatine). Theamplification program was optimised for a Strat-agene RoboCycler Gradient 40 as follows: onecycle of 10 min at 94�C, 1 min at 60�C, 2 min at72�C; 40 cycles of 1 min at 94�C, 1 min at 60�Cand 2 min at 72�C followed by a final extensionof 5 min at 72�C. The PCR products weresequenced, and their integrity was assured by gelelectrophoresis on a 2.5% agarose gel stainedwith ethidium bromide.

RT-PCR and in vitro transcription for creation ofabsolute standard curves

For in vitro transcription of sGH and tGHcomplementary (c)RNAs we directly added byprimer extension to the 5¢ end of the as primers(Table 1) using RT-PCR a T7 promoter genesequence to our target gene sequences asrecently established for the insulin-like growthfactors (Caelers et al., 2004). In brief: to createtemplates (95 nt for tGH, 94 nt for sGH, 86 ntfor b-actin) suitable for in vitro transcription,conventional RT-PCR of total RNA from O.niloticus liver (for promoter-driven sGH) andpituitary (for endogenous tGH) was performed.After RT reaction as decribed above 1 ll aliqu-ots of cDNA were added to a 50 ll PCR reac-tion using the Thermo-Start� PCR Master Mix(Abgene, NY, USA). The following amplifica-tion conditions were used: one cycle of 10 minat 94�C, 1 min at 60�C, 2 min at 72�C; 29cycles of 1 min at 94�C, 1 min at 59�C and2 min at 72�C, followed by a final extension of10 min at 72�C. Correct length of the templateswas assured on a 2.5% agarose gel stained byethidium bromide.

In vitro transcription was performed using theT7-MEGAshortscriptTM Kit (Ambion) with some

modifications. The cRNA transcripts were dis-solved in DEPC-treated water, and quantified byspectrophotometrical analysis and dot blot. Integ-rity of the probes was verified by UV-shadowingof a 10% TBE/Urea gel (wavelength 245 nm).Quantification was performed by spectrophotom-etry and dot blot comparing the quantity ofcRNA with that of standardised RNA (kanamy-cin positive control RNA, Catalys) by scanningthe optic densities using ImageJ software (ImageJ,NIH Software Program, USA). From this definedamount of cRNA encoding for our genes of inter-est real-time PCR (10-fold dilutions) was per-formed in triplicates to obtain standardisedthreshold cycle (CT) values. Standard curves weregenerated on the basis of the linear relationshipexisting between the CT value and the logarithmof the starting amount.

Quantification of sGH and tGH transcripts byone-step RT-PCR TaqMan system

About 10 ng of total RNA from O. niloticus tis-sues were directly subjected to a RT-PCR stepwithin the same tube using a One-Step-RT-PCRMastermix (Applied Biosystems, Rotkreuz, Swit-zerland) whereby 25 ll RT-PCR mixture (onereaction) contained 12.5 ll 2 · Master Mix (Amp-liTaqGold� DNA Polymerase, dNTPs withdUTP, Passive Reference 1, optimised buffercomponents), 0.625 ll 40 · RT enzyme mix(Multi-ScribeTM Reverse Transcriptase andRNase inhibitor), 300 nM of each primer, and 150nM of the fluorogenic probe. A reverse transcrip-tion step at 48�C for 30 min and a denaturationstep at 95�C for 10 min were performed followedby 40 cycles, each 15 s at 95�C and 1 min at 60�Cin a single tube using ABI PRISMTM 7700Sequence Detection System Perkin Elmer (AppliedBiosystems) without modifying or moving thesamples between RT and PCR.

Determination of absolute amounts of tGH andsGH mRNA in tilapia organs

The standard curves obtained for tGH and sGHcRNAs were used to determine the absoluteamounts of tGH and sGH mRNA in total RNAwhereby different lengths of cRNA and mRNAwere considered by correction factors which we

99

determined by division of lengths of mRNA andcRNA (tGH: 615 nt/75 nt ¼ 8.2; sGH: 631 nt/76nt ¼ 8.3). Copy numbers were calculatedby the formula: copies/lg of RNA ¼ (amountRNA) · 6.022 · 1023/length of cRNA transcript(nt) · (340 for RNA).

Results

Generation of standard curves for tilapia (t)and salmon (s) GH

For measurement of absolute amounts of GHgene message expression, the standard curvemethod using primer extension was adopted asalready described (Caelers et al., 2004). Amplifi-cation of total liver RNA from transgenic andwild-type tilapia resulted in CT values used forthe generation of standard curves (Figure 1).Sample normalisation was performed using b-actin as established with the standard curvemethod.

Expression patterns of the sGH and tGH genes inthe pituitary

Endogenous O. niloticus tGH mRNA was foundexclusively in O. niloticus pituitary and not in anyother tilapia tissue. In wild-type pituitary, tGHmRNA amounted to 226 ± 30 pg/lg total pitui-tary RNA and in transgenics to 187 ± 43 pg/lgtotal pituitary RNA. The difference was statisti-cally not significant (P ¼ 0.055). On the otherhand, chinook salmon (O. tshawytschwa) sGHmRNA was not detected in O. niloticus pituitaryor in any other organ investigated in the wild-typefish.

Expression of the sGH gene in liver andextrahepatic sites

The transcriptional activity of the sGH gene wasnot only measured in the liver but also in various

Figure 1. Standard curves plotting CT values of sGH andtGH cRNAs against decreasing dilutions of total liver RNA.

Table 3. Absolute amounts of sGH mRNA in liver andextrahepatic sites from 10 individuals of the transgenic tilapia

Organ pg/lg total RNA

Brain 1.4 ± 0.5

Pituitary –

Gills 4.1 ± 2.0

Heart 1.9 ± 0.8

Gut 0.7 ± 0.63

Liver 8.3 ± 2.5

Muscle 2.6 ± 1.4

Kidney 0.2 ± 0.08

Spleen 0.6 ± 0.52

Testes 2.0 ± 1.7

Results are given as mean ± SD.

Figure 2. Absolute amounts of sGH mRNA (pg/lg totalRNA) in tilapia liver and extrahepatic sites from 10 individu-als. Results are given as mean ± SD (error bars).

100

other O. niloticus tissues, such as gills, heart,brain, skeletal muscle, kidney, spleen, intestineand testes (Table 3) with the only exception ofpituitary. Although the extrahepatic sitesexpressed considerable amounts of sGH mRNA,liver exhibited the highest level (Figure 2). ThesGH mRNA levels in the organs investigatedwere quite different (Table 3), e.g. about the halfof the amount measured in liver (8.3 pg/lg totalRNA) was present in gills (4.1 pg/lg total RNA)while kidney exhibited as little as 0.2 pg/lg totalRNA.

Discussion

In the present study we have measured the abso-lute amounts of endogenous (t) and exogenoussalmon (s) GH mRNA in a growth enhancedtransgenic tilapia (O. niloticus). The sGH genewas anticipated to be mainly exclusively in theliver under control of the ocean pout anti-freezepromoter, but considerable amounts of sGHmRNA were also measured in other tissues, suchas gills, heart, brain, skeletal muscle, kidney,spleen, intestine and testes. As expected, sGHmRNA was not detected in any organ of wild-type tilapia.

Endogenous GH mRNA was found exclu-sively in O. niloticus pituitary. In contrast, indeveloping rainbow trout (Oncorhynchus mykiss),GH mRNA has been detected in pituitary andseveral other organs, such as brain, gills, heart,liver, kidney, pyloric ceca and ovary (Yang et al.,1999). Furthermore, in coho salmon, endogenousGH mRNA was not only expressed in pituitarybut also in intestine although GH mRNA in theintestine could be detected only in small fish(Mori & Devlin, 1999). Thus, during fish devel-opment GH mRNA seems to be expressed alsoin extra-pituitary sites. This suggestion is sup-ported by a study on adult four-spine sculpin,Cottus kazika (Inoue et al., 2003), that as in ourresults also detected GH mRNA only in pitui-tary. Under our transgenic conditions no obviousorgan abnormalities were observed (Rahmanet al., 1998) unlike some other lines of transgenicfish (Mori & Devlin, 1991; Devlin et al., 1995;Devlin et al., 2001). The likely reason may bethat our approach mimicks the GH gene expres-sion during ontogeny. Thus, transgenic methods

like the one used here may predictably result inrelatively normal and healthy but growth-enhanced fish.

The amount of tGH mRNA in transgenicpituitary was about 17% lower than in wild-type.Although the difference was statistically not sig-nificant (P ¼ 0.055) the decrease in endogenousGH expression resembles that described in a pre-vious study on transgenic coho salmon contain-ing a transgene comprised of the sockeye salmonGH1 gene under the control of the sockeye sal-mon MT-B promoter (Mori & Devlin, 1999). Inthis study, the level of endogenous GH mRNAwas significantly reduced in transgenic pituitariesrelative to control fish. Thus, in transgenic GH-overexpressing fish the circulating levels of bothendogenous and exogenous GH seem to reduce(endogenous) pituitary GH production by meansof a negative feedback mechanism.

In the transgenic animals, pituitary was theonly organ among those investigated where nosGH mRNA expression was detected. On theone hand, pituitary may nevertheless containsGH mRNA but the amount may be below thedetection level of our method. However, the highsensitivity of the approach used contradicts thisidea. On the other hand, as discussed below, thedistribution pattern of the sGH gene expressionlikely reflects the tissue specificity of the type IIIAFP gene promoter and this may not beexpressed in pituitary. This hypothesis is sup-ported by the study on transgenic coho salmon(Mori & Devlin, 1999). Here by the use of in situhybridisation in both control and transgenic ani-mals, GH mRNA was localised in the same pitui-tary regions with the signal being stronger incontrols. Although the study gives no informa-tion on the potential presence of transgenic GHin pituitary, the result may indicate that onlyendogenous GH was expressed.

The expression of exogenous GH recorded inthe present study presumably reflects the tissuespecificity of the type III AFP gene promoter. Inprevious work with the ocean pout, type III AFPmRNA was detected in most tissues examined,with the highest level occurring in the liver, fol-lowed by gills (Gong et al., 1992). This is roughlyin parallel with the present results in the trans-genic tilapia where liver and gills also exhibitedthe highest amounts of sGH mRNA. Similaramounts of AFP mRNA were found in heart

101

and brain by Gong et al. (1992), which is also inaccordance with our results on sGH. In contrastto the AFP gene, which was barely expressed inmuscle and testes, both organs showed relativelyhigh sGH mRNA expressions in our experi-ments. Small but significant quantities of AFPmRNA were detected in the red blood cells(1.7%of liver levels) (Gong et al., 1992). There-fore, the differences in the expression rates couldbe attributed to the presence of red blood cellswithin the organs investigated. Thus in this work,much less expression is recorded in kidney thanin the earlier study. However, it is highly possiblethat the tissue specificity of the sGH recorded isnot an absolutely accurate reflection of the tissuepreference of the native gene sequence, since itschromosomal locus is different. Also only one ofthe many copies present in the Ocean pout gen-ome is under study, and some physiological fac-tors which may help to regulate the endogenousgene copy may be inoperative in the transgenicsituation described here. Hew et al. (1999) exam-ined the tissue expression of Winter flounderantifreeze, driven by its own promoter, in trans-genic Atlantic salmon. In these fish, antifreezeexpression, as determined by Northern blotting,was confined to the liver, but showed substantialseasonal variation. The Ocean pout antifreeze isknown to have a different expression pattern tothat of Winter flounder, and to be much less sea-sonally variant. Other work in transgenic fish inwhich reporter genes were driven by tissue spe-cific promoters, indicates apparently faithful tis-sue specific expression, in line with the expectedexpression specificity of the promoters used. Thiscan be extremely specific, as in the case of pro-moters from gonadotropin releasing hormone(GnRH) genes, which are expressed only inhypothalamus. Thus Farahmand et al. (2003),using a GnRH promoter of tilapia (O. niloticus)in zebrafish, detected reporter gene expressiononly in the hypothalamus. Both muscle specificexpression of a myosin heavy chain promoterand widespread expression of the ubiquitouslyoccurring b-actin promoter, which were bothderived from common carp (Cyprinus carpio),were observed in zebrafish embryos (Mulleret al., 1997).

The tissue expression pattern revealed in thepresent study suggests that the AFP-drivensGH gene is expressed at low to moderate lev-

els in all tissues, but most strongly in liver,whence the GH protein is presumably secretedinto the circulation. In the light of this newinformation about the tissue specific expressionof the sGH gene driven by the Ocean poutAFP, it is also interesting to consider the phys-iological effects of the novel GH. The fishexpressing this gene construct are approximately2.5-fold growth enhanced but display no obvi-ous abnormalities (Rahman et al., 1998). TheGH release in or from the liver and, if happen-ing, from the other organs expressing the sGHgene, is probably not pulsatile, unlike therelease of the native hormone from the anteriorpituitary. The plasma levels of endogenous GHfurther depend on environmental factors, suchas season, temperature and water salinity (e.g.Bjornsson et al., 2000; Gabillard et al., 2003)and on nutrition (Pottinger et al., 2003). Undernormal rearing and feeding conditions, as wereused in our study, in rainbow trout the meanGH plasma levels were about 4.8 ng/ml (Gabil-lard et al., 2003) or 2.17 ng/ml (Pottinger et al.,2003) and in Atlantic salmon about 1.6 ng/ml(Bjornsson et al., 2000). The levels of sGH incirculation in the transgenic tilapia were previ-ously observed to lie between 0.6 and 1.5 ng/ml(Rahman et al., 1998) and are in the samerange as the levels of endogenous GH in rain-bow trout and salmon. Thus, endogenous plusexogenous expression in the transgenic fishleads to an approximate doubling of GH con-centrations.

Of course growth is much more complex thana simple 1 to 1 response to circulating GH. Mostif not all growth processes, although stimulatedby GH, are mediated via insulin-like growth fac-tor I (IGF-I). As in mammals, the major site ofIGF-I gene expression in bony fish is the liver(e.g. Pierce et al., 2003; Vong et al., 2003; Caelerset al., 2004) and that is the main source of circu-lating IGF-I (Plisetskaya, 1998; Reinecke &Collett, 1998). However, as shown in the tilapia,O. mossambicus (Reinecke et al., 1997), numerousextrahepatic sites also express IGF-I in parenchy-mal cells, including the organs investigated in thisstudy. Not only the expression of IGF-I in theliver (Moriyama, 1995) but also in extrahepaticsites seem to be regulated by GH (Vong et al.,2003), rendering the situation even more com-plex. On the one hand, exogenous GH most

102

likely promotes IGF-I expression and secretionin the liver as endogenous GH also does andthus raises serum IGF-I to a higher level. Thiswill lead to an overall enhancement of growth.Because sGH was further transcribed in extrahe-patic sites as shown here it may also stimulateIGF-I expression in the different organs in aparacrine/autocrine manner and, thus, furtherpromote organ growth. In agreement, in trans-genic tilapia lines with different ectopic expres-sion of tGH the level of IGF mRNA in muscleand gonad was higher when these organs alsoexpressed tGH than in transgenic lines wherethese organs did not express tGH (Hernandezet al., 1997).

Acknowledgements

The study was supported by the Swiss NationalFoundation (project 32-061481).

References

Ber R and Daniel V (1992) Structure and sequence of thegrowth hormone-encoding gene from Tilapia nilotica. Gene113: 245–250.

Bjornsson BT, Hemre GI, Bjornevik M and Hansen T (2000)Photoperiod regulation of plasma growth hormone levelsduring induced smoltification of underyearling Atlantic sal-mon. Gen Comp Endocrinol 119: 17–25.

Brem G, Brenig B, Horstgen-Schwark G and Winnacker EL(1988) Gene transfer in tilapia (Oreochromis niloticus).Aquaculture 68: 209–219.

Caelers A, Berishvili G, Meli ML, Eppler E and Reinecke M(2004) Establishment of a real-time RT-PCR for the deter-mination of absolute amounts of IGF-I and IGF-II geneexpression in liver and extrahepatic sites of the tilapia. GenComp Endocrinol 137: 196–204.

Chourrout D, Guyomard R and Houdebine L (1986) Highefficiency gene transfer in rainbow trout (Salmo gairdneri)by microinjection into eggy cytoplasm. Aquaculture 51:143–150.

Davies PL and Sykes BD (1997) Antifreeze proteins. CurrOpin Struct Biol 7: 828–834.

Devlin RH (1993) Sequence of sockeye salmon type 1 and 2growth hormone genes and the relationship of rainbowtrout with Atlantic and Pacific salmon. Can J Fish AquatSci 50: 1738–1748.

Devlin RH, Biagi CA, Yesaki TY, Smailus DE and Byatt JC(2001) Growth of domesticated transgenic fish. Nature 409:781–782.

Devlin RH, Vesaki TY, Biagi CA, Donaldson EM, Swanson Pand Chan W-K (1994) Extraordinary salmon growth. Nat-ure 371: 209–210.

Devlin RH, Vesaki TY, Donaldson EM and Hew CL (1995)Transmission and phenotopic effects of an antifreeze/GHgene construct in coho salmon (Oncorhyncus kisutch). Aqua-culture 137: 161–167.

De Vries AL (1983) Antifreeze peptides and glycopeptides incold-water fishes. Annu Rev Physiol 45: 245–260.

Du SJ, Gong Z, Fletcher GL, Shears MA, King MJ, IdlerDR and Hew CL (1992) Growth enhancement in trans-genic Atlantic salmon by the use of an ‘all fish’ chimericgrowth hormone gene construct. Biotechnology 10:179–187.

Dunham RA, Eash J, Askins J and Townes TM (1987)Transfer of the metallothionein-human growth hormonefusion gene into channel catfish. Trans Am Fish Soc 116:87–91.

Farahmand H, Rahman MA, Sohm F, Hwang GL andMaclean N (2003) Isolation and expression of tilapia (Ore-ochromis niloticus) serine 8-type GnRH coding and regula-tory sequences. Gene 304: 97–106.

Fletcher GL and Davis PL (1991) Transgenic fish for aqua-culture. In: Setlow JK (ed.), Genetic Engineering (pp.331–370) Plenum Press, New York.

Fletcher GL, Hew CL and Davies PL (2001). Antifreeze pro-teins of teleost fishes. Annu Rev Physiol 63: 359–390.

Gabillard JC, Weil C, Rescan PY, Navarro I, Gutierrez J andLe Bail PY (2003) Environmental temperature increasesplasma GH levels independently of nutritional status inrainbow trout (Oncorhynchus mykiss). Gen Comp Endocri-nol 133: 17–26.

Gong Z, Ewart KV, Hu Z, Fletcher GL and Hew CL (1996)Skin antifreeze protein genes of the winter flounder, Pleuro-nectes americanus, encode distinct and active polypeptideswithout the secretory signal and prosequences. J Biol Chem271(8): 4106–4112.

Gong Z, Fletcher GL and Hew CL (1992) Tissue distributionof fish antifreeze protein mRNAs. Can J Zool 70: 810–814.

Gross ML, Schneider JF, Moav N, Moav B, Alvarez C,Myster SH, Liu Z, Hallermann EM, Hackett PB, GuiseKS, Faras AJ and Kapuscinski AR (1992) Molecular analy-sis and growth evaluation of northern pike (Esox lucius)microinjected with growth hormone genes. Aquaculture 103:253–273.

Guise KS, Hackett PB and Faras AJ (1992) Transfer of genesencoding neomycin resistance, chloramphenicol acetyltransferase and growth hormone into goldfish and northernpike. In: Hew CL and Fletcher GL (eds), Transgenic Fish(pp. 142–164) World Scientific, Singapore.

Hernandez O, Guillen I, Estrada MR, Cabrera E, PimentelR, Pina JC, Abad Z, Sanchez V, Hidalgo J, Mar tinez R,Lleonart R and de la Fuente J (1997) Characterization oftransgenic tilapia cell lines with different ectopic expressionof tilapia growth hormone. Mol Mar Biol Biotechnol 6:364–375.

Hew CL, Fletcher GL and Davies PL (1995) Transgenic sal-mon: tailoring the genome for food production. J Fish Biol47 (suppl A): 1–19.

Hew C, Poon R, Xiong F, Gauthier S, Shears M, King M,Davies P and Fletcher G (1999) Liver-specific and seasonalexpression of transgenic Atlantic salmon harboring the win-ter flounder antifreeze protein gene. Transgenic Res 8:405–414.

Hwang GL, Azizur Rahman M, Abdul Razak S, Sohm F,Farahmand H, Smith A, Brooks C and Maclean N (2003)

103

Isolation and characterisation of tilapia beta-actin promoterand comparison of its activity with carp beta-actin pro-moter. Biochim Biophys Acta 1625: 11–18.

Inoue K, Iwatani H and Takei Y (2003) Growth hormoneand insulin-like growth factor I of a Euryhaline fish Cottuskazika: cDNA cloning and expression after seawater accli-mation. Gen Comp Endocrinol 131: 77–84.

Iyengar A, Muller F and Maclean N (1996) Regulation andexpression of transgenes in fish – a review. Transgenic Res5: 1–19.

Maclean N and Talwar S (1984) Injection of cloned genesinto rainbow trout eggs. J Embryol Exp Morph 82: 187.

Mori T and Devlin RH (1999) Transgenic and host growthhormone expression in pituitary and nonpituitary tissues ofnormal and growth hormone transgenic salmon. Mol CellEndocrinol 149: 129–139.

Moriyama S (1995) Increased plasma insulin-like growth fac-tor-I (IGF-I) following oral and intraperitoneal administra-tion of growth hormone to rainbow trout, Oncorhynchusmykiss. Growth Regul 5: 164–167.

Muller F, Williams DW, Kobolak J, Gauvry L, Goldspink G,Orban L and Maclean N (1997) Activator effect of coinject-ed enhancers on the muscle-specific expression of promotersin zebrafish embryos. Mol Reprod Dev 4: 404–412.

Nam YK, Cho YS and Kim DS (2000) Isogenic transgenichomozygous fish induced by artificial parthogenesis. Trans-genic Res 9: 463–469.

Nash CE (1988) A global overview of aquaculture produc-tion. J World Aquacult Soc 19: 15.

Penman DJ, Beeching AJ, Penn S and Maclean N (1990) Fac-tors affecting survival and integration following microinjec-tion of novel DNA in rainbow trout eggs. Aquaculture 85:35–50.

Pierce AL, Dickey JT, Larsen DA, Fukada H, Swanson Pand Dickhoff WW (2003) A quantitative real-time RT-PCRassay for salmon IGF-I mRNA, and its application in thestudy of GH regulation of IGF-I gene expression in pri-mary culture of salmon hepatocytes. Gen Comp Endocrinol135: 401–411.

Plisetskaya EM (1998) Some of my not so favorite thingsabout insulin and insulin-like growth factors in fish. CompBiochem Physiol B Biochem Mol Biol 121: 3–11.

Pottinger TG, Rand-Weaver M and Sumpter JP (2003) Over-winter fasting and re-feeding in rainbow trout: plasmagrowth hormone and cortisol levels in relation to energymobilisation. Comp Biochem Physiol B Biochem Mol Biol136: 403–417.

Rahman MA, Iyengar A and Maclean N (1997) Co-injectionstrategy improves integration efficiency of a growth

hormone gene construct, resulting in lines of transgenic tila-pia (Oreochromis niloticus) expressing an exogenous growthhormone gene. Transgenic Res 6: 369–378.

Rahman MA and Maclean N (1992) Fish transgene expre-ssion by direct injection into fish muscle. Mol Mar BiolBiotechnol 1: 286–289.

Rahman MA, Mak R, Ayad H, Smith A and Maclean N(1998) Expression of a novel piscine growth hormone generesults in growth enhancement in transgenic tilapia (Ore-ochromis niloticus). Transgenic Res 7: 357–369.

Reinecke M and Collet C (1998) The phylogeny of the insu-lin-like growth factors. Int Rev Cytol 183: 1–94.

Reinecke M, Schmid A, Ermatinger R and Loffing-Cueni D(1997) Insulin-like growth factor I in the teleost Oreochr-omis mossambicus, the tilapia: Gene sequence, tissue expres-sion, and cellular localization. Endocrinology 138:3613–3619.

Rentier-Delrue F, Swennen D, Philippart JC, L’Hoir C,Lion M, Benrubi O and Martial JA (1989) Tilapia growthhormone: molecular cloning of cDNA and expression inEscherichia coli. DNA 8: 271–278.

Rokkones E, Alestrom P, Skjervold H and Gautvik KM(1989) Microinjection and expression of a mouse methallo-thionein fusion gene in fertilized salmonid eggs. J CompPhysiol B 158: 751–758.

Rosochacki SJ, Bialowas H, Czonlonkowska M, Gus-zkiewicz A, Kossakowski M, Sadowska, J, Siadkowska E,Zebrowska T, Szumiec J and Zwierzchowski L (1993)Introduction of human growth hormone gene into commoncarp (Cyprinus carpio). Anim Sci 11: 47–58.

Song S, Trinh KT and Hew CL (1992) Cloning and sequenceanalysis of salmon growth hormone cDNA. I Chuan HsuehPao 19: 308–315.

Vong QP, Chan KM and Cheng CH (2003) Quantification ofcommon carp (Cyprinus carpio) IGF-I and IGF-II mRNAby real-time PCR: differential regulation of expression byGH. J Endocrinol 178: 513–521.

Yang BY, Greene M and Chen TT (1999) Early embryonicexpression of the growth hormone family protein genes inthe developing rainbow trout, Oncorhynchus mykiss. MolReprod Dev 53: 127–134.

Zhang P, Hayat M, Joyce C, Gonzalez-Villasenor LI,Lin CM, Dunham RA, Chen TT and Powers DA (1990)Gene transfer, expression and inheritance of pRSV-rainbowtrout GH-cDNA in the common carp, Cyprinus carpio(Linnaeus). Mol Reprod Dev 25: 3–13.

Zhu Z, Li G, He L and Chem S (1985) Novel gene transferinto the fertilized eggs of goldfish. Z Ange Ichthyol 1:31–34.

104