Introduction of a porcine growth hormone fusion gene into transgenic

pigs promotes growth

PETER D. V I Z E ' » t , ANNA E. MICHALSKA2, ROD ASHMAN2, B. LLOYD3, B. A. STONE2,

P. QUINN2, J. R. E. WELLS' and R. F. SEAMARK2

Departments of 'Bwchemisliy and ^Obstetrics and Gynaecology, University of Adelaide, GPO Box 498, Adelaide, SA. 5001, Australia•'Metmfanns, Wasley, SA. 5400, Australia

•Present address: Laboratory of Embryogenesis, National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AAf Author for correspondence

Summary

Six transgenic pigs have been produced bymicroinjecting a human metallothionein pro-moter/porcine growth hormone gene constructinto the pronuclei of fertilized eggs which weretransferred to synchronized recipient sows. Theresulting transgenic animals contained between0-5 and 15 copies of the gene construct per cell,and at least one of the animals expressed theintroduced gene and grew at an increased rate

compared to both transgenic and non-transgeniclittermates. Some of the transgenic animals thatdid not appear to grow at increased rates werefound to contain rearranged gene sequences. Twoof the transgenic pigs have been shown to pass onthe introduced genes to their offspring.

Key words: transgenic pigs, gene rearrangement, growthhormone, growth improvement.

Introduction

Transgenic mice containing integrated copies of a widerange of different genes have been produced (Palmiter& Brinster, 1986). These animals, and their transgenicoffspring, usually express the introduced transgenes ina manner appropriate to the promoter/enhancer util-ized in the introduced gene. When the expressedtransgene encodes a growth promoting hormone, suchas growth hormone or growth hormone releasingfactor, transgenic mice can grow up to twice normalsize (Palmiter et al. 1982a,b, 1983; Hammer et al.1985a). The application of this approach to improvingthe growth of livestock has been hindered by technicaldifficulties encountered in producing transgenic farmanimals, and techniques that enable the generation oftransgenic sheep and pigs have only recently becomeavailable (Hammer et al. 19856). The introduction of afusion gene containing the human growth hormonegene linked to a murine metallothionein promoter intotransgenic rabbits, sheep and pigs has previously beendemonstrated (Hammer et al. 19856). Although someof these animals were shown to express the introducedgene by both RNA and protein analysis, none of them

Journal of Cell Science 90, 295-300 (1988)Printed in Great Britain © The Company of Biologists Limited 1988

grew at enhanced rates. In this report we describe theproduction of transgenic pigs using a gene construct,which contains a porcine growth hormone (pGH)fusion gene under the transcriptional control of thehuman metallothionein-IIA (hMT-IIA) promoter.This construct has previously been shown to markedlyenhance growth rates when introduced into the germline of transgenic mice (unpublished results).

Materials and methods

DNA manipulationsAll DNA manipulations involved well-established techniques(Maniatis et al. 1982). Plasmid pHMPG.4 was constructedby inserting a ///ndlll-jEcoRI fragment containing hMT-IIA promoter sequences from —765 to +60 (Karin &Richards, 1982) into pUC19. This plasmid was then digestedwith £coRI and ligated to the 800-bp £VoRI insert of the full-length pGH cDNA clone pPG.3 (Vize & Wells, 1987«). Ofthe resulting plasmids, one which contained the EcoR\fragment in the correct orientation was selected. This clonewas then digested with Smal, and ligated to the Klenow-treated 1-kb Sma\-BcunW\ fragment from cosmid cPGH.l,which contains the 3' end of the pGH gene (Vize & Wells,

295

19876). The resulting clones were screened by restrictionmapping and a clone that contained the fragment inserted inthe desired orientation identified and named pHMPG.4(HM, human metallothionein; PG, porcine growth hor-mone).

The DNA fragment used for microinjection was a 2-7-kb/•//i/dlll—fV«I fragment isolated from pHMPG.4. Followingdigestion with these two restriction enzymes the desiredfragment was separated from other reaction products byclectrophoresis through a low melting temperature agarosegel, and purified by one phenol followed by one phenol:chloroform (1:1) extraction and ethanol precipitation. Thepurified fragment was resuspended in 10mM-Tris- HCI(pH7-5), 1 mM-EDTA and diluted to a concentration of2ngj(l~' with phosphate-buffered saline for microinjection.

Southern blots (Southern, 1975) were performed usingZeta-probe membranes and the alkaline transfer procedure ofChomczynski & Qasba (1984) as modified by Reed & Mann(1985). Dot and slot-blots were performed using Gene-Screen membranes under the conditions described by themanufacturer. In all cases the hybridization probe was a nick-translated (Mamatis el al. 1982) Hindlll-Aval fragmentfrom the 5' end of the human metallothionein-UA promoter(Karin & Richards, 1982).

Generation of transgenic animals

Multiparous Large White sows were stimulated with7S0IUPMSG (Folligen Intervet) followed by 500IU hCG44 h later. Sows in oestrus were hand mated to a fertile boarand slaughtered for egg collection 10 h after the expected timeof ovulation. The fertilized pig eggs were then recoveredfrom the reproductive tracts of the slaughtered animals. Thecytoplasm of eggs was stratified by centrifugation (7000£ for3 min) to allow the visualization of pronuclei (Hammer e? al.19856), and one of the pronuclei injected with 2 pi of thepHMPG.4 insert (containing approximately 600 copies of the2-7-kb linear DNA fragment). Following overnight culture(Stone et al. 1984) the surviving eggs were surgicallytransferred to the oviducts of synchronized recipient sows.

Results

The organization of the gene construct used in theseexperiments, pHMPG.4, is illustrated in Fig. 1. Thisplasmid contains the hMT-IIA promoter (Karin &Richards, 1982) fused to a hybrid pGH gene consistingof pGH cDNA sequences (Vize & Wells, 1987/7)encoding amino acid residues 1-158 of pre-pGH, andpGH genomic sequences (Vize & Wells, 19876) en-coding pre-pGH residues 158-216. The genomic se-quence also contains approximately 700 bases of pGHgene 3' non-coding sequences. The 3' non-codingsequences were included to ensure the efficient process-ing and polyadenylation of the hybrid messenger RNA.The 2-7-kb HindlU-Pvul insert of pHMPG.4 waspurified and used for microinjection into fertilized pigeggs-

In a preliminary series of experiments, 189 single-cell embryos, which had each been injected withapproximately 600 copies of the plasmid insert, weretransferred to 13 recipient sows, at an average of 13embryos per sow. Four of the surrogates becamepregnant as assessed by estrone sulphate determination(Stone et al. 1986) but none farrowed. In a secondseries of transfers, a total of 423 microinjected embryoswere transferred to 14 synchronized recipient sows,this time at an average of 30 per sow. Six of therecipients returned to oestrus during the fourth weekfollowing the transfer, and four had to be culledbecause of vaginal bacterial infection (not connectedwith the embryo transfer procedure). The remainingfour sows completed the pregnancy and produced atotal of 17 piglets. Thus, the frequency of producingpiglets from injected embryos was one live birth per 16embryos excluding the embryos from infected sows.This is similar to the efficiency obtained with mouse

pUC19 2639bp - - Pvul

Human MT promoter 825 bp

EcoRI

PGHcDNA522bp

Smal

P,GHcosmid~1000bp

EcoRISmal/BamHl

PGHcDNA299bp

Fig. 1. Structure of plasmidpHMPG.4. The construction ofthis plasmid is described in detailin Materials and methods. Theplasmid contains 765 bp ofpromoter and 60 bp of 5'untranslated sequence from thehuman metallothionein-I IApromoter (Karin & Richards, 1982)linked to a hybrid pGH geneconstructed by fusing pGH cDNA(Vize & Wells, 1987«) and pGHcosmid (Vize & Wells, 19876)sequences. Thef/;;/dIIl and Pz'iilrestriction sites used to excise thehMT-IIA/pGH fusion gene areindicated.

296 P. D. Vize et al.

B

1

2

3

4

5

Fig. 2. Gene copy number in transgenic pigs. DNAsamples from transgenic pigs were examined by slot-blotanalysis. Samples were filtered onto a membrane using aslot-blot apparatus, which also contained pig negative andhuman positive controls, and plasmid DNA samplescorresponding to known gene copy numbers. The samples(5 fig) are: row 1, transgenic pigs 177, 180 and 295; row 2,transgenic pigs 375, 736 and 739; row 3 (no sample in firstcolumn) pig negative control and human positive control;row 5, plasmid standards corresponding to 40, 10 and 4gene copies per cell; row 6, as for row 5 exceptcorresponding to 30, 8 and 2 gene copies per cell; row 7, asper row 6 except amounts correspond to 20, 6 and 1 genecopies per cell.

one-cell embryos injected with the same plasmid con-struct, where we achieved one mouse pup from anaverage of 11 injected embryos.

Dot-blot analysis was performed on DNA isolatedfrom tail tissue (Palmiter et al. 1982a) from each of the17 piglets produced, and revealed that six of theanimals that developed from injected eggs containedsequences homologous to the hMT-IIA promoter hy-bridization probe, with copy numbers (determined byslot-blot analysis) ranging from 0-5 to 15 copies of thepHMPG.4 insert per cell (Fig. 2). The two animalswith less than one copy per cell (see Table 1) will not bediscussed further.

Southern analysis was performed on the DNA ofeach transgenic pig to determine the chromosomalorganization of the introduced sequences. PlasmidpHMPG.4does not contain BamHl restriction sites, sothe digestion of transgenic animal DNA with thisenzyme will produce one band per integration site onSouthern analysis. The data show that each of the

transgenic pigs contains a single band, which hybrid-izes to the hMT-IIA promoter probe (Fig. 3A). Inthree cases, (pigs 180, 295 and 736) the detected bandsare of greater than 25 kb in length, and so appear to beof identical size. Longer gels using lower percentages ofagarose also produce a similar result to that seen inFig. 3A, and we therefore conclude that each of thesefour animals probably contains the transgenes inte-grated into a single chromosomal site in each case.

EcoRI cuts twice within the pHMPG.4 insert, sodigestion with this enzyme followed by hybridization tohMT-IIA sequences can identify the presence of head-to-head multimers, head-to-tail multimers, and flank-ing EcoRI restriction sites (Fig. 1). Digestion of DNAfrom each of the transgenic pigs with EcoRI followedby Southern analysis (Fig. 3B) revealed that only oneanimal, pig number 295, produced a pattern of bandsthat indicated that all integrated copies of thepHMPG.4 insert were in a tandem head-to-tail array.Of the remaining three pigs, one, number 177, showeda random arrangement (or possibly tail-to-tail), whilethe remaining two animals produced multiple bandsindicative of rearranged sequences.

Serum pGH concentrations (Table 1) were deter-mined for each of the transgenic and control pigs usingan ELISA assay system (Signorella & Hymen, 1984) todetermine if the introduced genes had produced anyeffect on serum growth hormone levels. One of thetransgenic pigs, number 295, contained an elevatedlevel of serum pGH, which was over twice that of non-transgenic and transgenic littermates both at birth andat 50 days post-partum (Table 1).

All piglets were monitored under conventional com-mercial rearing conditions with ad libitum feeding.Growth was monitored by weighing at weekly inter-vals. No differences were evident in the growth rate ofpiglets up to 20 kg live weight. Subsequently, one ofthe two transgenic sows, number 177, continued togrow at a similar rate to non-transgenic control litter-mates, while the other, number 295, began to growsubstantially faster (Table 1), and had achieved thetarget market weight of 90 kg at 17 weeks of age asopposed to 22-25 weeks required by her littermates. Itis difficult to judge if either of the two transgenic malespossess enhanced growth rates as only two non-trans-genic male controls were produced in the four litters.Neither of the transgenic males showed evidence ofmarkedly increased serum pGH levels (Table 1).Following breeding (see below) the largest of thetransgenic boars, number 180, was sacrificed at 35weeks of age, and a number of tissues examined for thepresence of pGH mRNA by both Northern andnuclease protection analysis. No expression of theintroduced gene was detected in liver, kidney, spleen,brain, testis or pituitary (data not shown).

Growth hormone fusion gene in transgenic pigs 297

Table 1. Weight gain and serum pGH levels in transgenic pigs

Animal

177180295375736739Non-transgenic

littermates

The mean daily weight gain

Sex

FMFFMFF + M

was determined

No. oftransgencs

per cell

36

150-560-5-

between 20 and 90 kg

Plasma pGH

At birth

2-54-2

27-86-31-10-5

6-4 ± 5 - 2

live weight. F, female

(mlUmr1)

At 50 days

10-415-327-8NDi i - i6-9

11-3 ± 2-7

; M, male; ND,

Weightgain

(gday-1)

758845

1273680646700

781 ±44

not determined.

Considering the high incidence of gene rearrange-ment in the founder transgenic pigs, it was importantto demonstrate that, once incorporated, the introducedgenes could be stably transmitted to offspring. Two ofthe transgenic pigs, one with a random arrangement(female 177) and one with rearranged (male 180)sequences, were therefore mated with control animals,and the offspring examined for the presence of theforeign gene by dot-blot analysis. As Fig. 4 illustrates,both of the animals were able to pass on the transgeneto a proportion of their offspring (pig 177, one out offive; pig 180, six out of eight), indicating that thegeneration of lines of transgenic pigs is feasible.Southern analysis of DNA from each of the transgenicoffspring revealed that each of the pigs produced

restriction enzyme patterns identical to those of theirtransgenic parent (data not shown).

Discussion

The finding that only one of the four transgenic pigsstudied contained the introduced genes organized in ahead-to-tail array is surprising, as all seven transgenicmice containing the same gene construct, which havebeen studied by Southern blotting, contain the trans-genes in head-to-tail arrays (for example, see Mclnneset al. 1987). Also, the results of a large number of otherinvestigators indicate that nearly all transgenic micecontain integrated transgenes in this conformation (forreviews see Palmiter & Brinster, 1985, 1986; Gordon &

A U

cCO

al xkb

_23-1-6-6

.2-32"2-03

BoCO

Oic\j

CDCO

Q_

0E3X kb

9-46-64-36

2-32•203

0-92 •

-0-56

Fig. 3. Southern analysis of transgenic pigs. DNA samples (3jUg) from each of the transgenic pigs along with pig negativeand human positive controls was digested with either BainHl (A) or EcoRl (B), electrophoresed through an agarose gelfollowed by transfer to a membrane and hybridization to a nick-translated human metallothionein-IIA promoter probe. The0-92-kb band (pig 295) expected from EcoRl digestion of a head-to-tail array is indicated in B. Molecular weight markers(in kb) are shown on the right of each gel.

298 P. D. Vize et al.

177

177-1

177-2

177-3

177-4

177-5

180

180-1

180-2

180-3

180-4

180-5

180-6

180-7

180-8

Pig

Human

Fig. 4. Inheritance of the transgene. DNA samplesfrom the two transgenic parents along with that of theiroffspring, plus appropriate negative (pig) and positive(human) controls, were applied to a membrane using a dot-blot apparatus and hybridized to a nick-translated humanmetallothionein-HA promoter probe. Pig 180, andoffspring 177-1, 177-2 and 180-1 to 180-3 were male, andpig 177 and offspring 177-3 to 177-5 and 180-4 to 180-8female. Despite appearances here, Southern analysis ofpigs 177 and 177-4 indicates that both of these animalscontain an equal number of gene copies per cell (data notshown).

Ruddle, 1985). The only available data on integrationpatterns in transgenic pigs other than those presentedhere are those of Hammer et al. (19856), who haveshown that in the two transgenic pigs studied the DNAwas integrated in an ordered fashion and inserts werefound mostly in head-to-tail arrays. As the mouse dataindicate that our construct is not inherently unstable,and the results of Hammer et al. (19856) indicate thatother constructs integrate in a similar fashion in theporcine and murine genomes, we concluded that therearrangement of the pHMPG.4 sequences in trans-genic pigs is probably a phenomenon specific for thisconstruct in the porcine genome.

The data presented here demonstrate for the firsttime that the expression of a transgene in a transgenicfarm animal can result in markedly improved dailyweight gain and body weight. As transgenic pigs

expressing high levels of human growth hormone donot appear to grow at increased rates (Hammer et al.19856), we propose that the enhanced growth of ourtransgenic female number 295 is due to the use of agene construct that directs the expression of authenticporcine growth hormone, rather than a heterologoushormone. It is interesting to note that the only pigshowing evidence of improved growth performancewas also the only animal that contained the introducedgene sequences integrated in a random head-to-tailarray. The increased serum pGH and improved growthperformance of this animal did not result in anydetrimental effects on health, and there were noindications of liver damage or arthritis as are observedafter prolonged exposure to high levels of injected pGH(Machlin, 1972). This is probably due to this animalexpressing growth hormone at less than toxic levels(Chung et al. 1985). Unfortunately, this animal con-tracted pneumonia following an uncommonly severeperiod of cold weather and was killed at 18 weeks ofage, precluding studies on the fertility of this animal orits ability to transmit its transgenes to offspring. Suchstudies are of great importance, as transgenic femalemice expressing high levels of human growth hormoneare infertile (Palmiter et al. 1983). This may not be assevere a problem in pHMPG.4 transgenic pigs, as theseanimals will be expressing porcine growth hormone,not a hetefologous growth hormone, and transgenicmice with elevated murine growth hormone levels donot display any indications of reduced fertility (Ham-mer et al. 1985a). However, the study of the effect ofincreased pGH levels in transgenic pigs will have toawait the production of a large number of transgenicpigs containing functional transgenes.

Our results also demonstrate for the first time thattransgenes integrated into the procine genome arcstable in the germline, and can be passed onto aproportion of the offspring, indicating that the pro-duction of stable lines of transgenic farm animals withenhanced growth performances is feasible.

Our future research is aimed at the generation of alarger number of transgenic pigs using the pHMPG.4construct and generating homozygous lines of trans-genic pigs from these founder animals to determine thepotential of this approach for improving pig pro-duction. Studies on the activity of the hMT-IIApromoter with transgenic animals fed on diets contain-ing increased levels of heavy metals (such as zinc) arealso planned as transcription rates from this promoterare enhanced by heavy metals (Karin et al. 1984).Experiments aimed at determining the region of thepHMPG.4 construct responsible for the observed in-stability in the porcine genome are also in progress.

This work was supported by grants from the Common-wealth Research Centre for Gene Technology and the

Growth hormone fusion gene in transgenic pigs 299

Australian Pig Industry Committee. P.D. V. was the recipientof a Commonwealth Postgraduate Research Award and anAustralian Pig Industry Research Committee Special Re-search Fellowship, A.E.M. was the recipient of a Universityof Adelaide Research Award. We would like to thank Dr L. S.Coles for critical discussions on the manuscript.

References

CHOMCZYNSKI, P. & QASBA, P. K. (1984). Alkaline transferof DNA to plastic membrane. Biochem. biophys. Res.Commun. 122, 340-344.

CHUNG, C. S., ETHERTON, T. D. & WIGGINS, J. P. (1985).

Stimulation of swine growth by porcine growthhormone. J. Anim. Sci. 60, 118-130.

GORDON, J. W. & RUDDLE, F. H. (1985). DNA-mediatedgenetic transformation of mouse embryos and bonemarrow - a review. Gene 33, 121-136.

HAMMER, R. E., BRINSTER, R. L., ROSENFELD, M. G.,

EVANS, R. M. & MAYO, K. E. (1985a). Expression ofhuman growth hormone-releasing factor in transgenicmice results in increased somatic growth. Nature, Land.315, 413-416.

HAMMER, R. E., PURSEL, V. G., REXROAD, C. E., WALL,

R. J., BOLT, D. J., EBERT, K. M., PALMITER, R. D. &

BRINSTER, R. L. (19856). Production of transgenicrabbits, sheep and pigs by microinjection. Nature, Land.315, 680-683.

KARIN, M., HASLINGER, A., HOLTGREVE, H., CATHALA,

G., SLATER, E. & BAXTER, J. D. (1984). Activation of aheterologous promoter in response to dexaniethasone andcadmium by metallothionein gene 5'-flanking sequences.Cell 36, 371-379.

KARIN, M. & RICHARDS, R. I. (1982). Humanmetallothionein genes - primary structure of themetallothionein-II gene and a related processed gene.Nature, Land. 299, 797-802.

MACHLIN, L. J. (1972). Effect of porcine growth hormoneon growth and carcass composition of the pig. J. Anim.Sci. 35, 794-800.

MANIATIS, T., FRITSCH, E. F. & SAMBROOK, J. (1982).

Molecular Cloning: A Laboratory Manual. Cold SpringHarbor, New York: Cold Spring Harbor Laboratory.

MCINNES, J. L., VIZE, P. D., HABILI, N. & SYMONS, R.

H. (1987). Chemical biotinylation of nucleic acids with

photobiotin and their use as hybridization probes. Focus(BRL/Life Technologies Inc.) 9, 1-4.

PALMITER, R. D. & BRINSTER, R. L. (1985). Transgenic

mice. Cell 41, 343-345.PALMITER, R. D. & BRINSTER, R. L. (1986). Germline

transformation of mice. A. Rev. Genet. 20, 465-499.PALMITER, R. D., BRINSTER, R. L., HAMMER, R. E.,

TRUMBAUER, M. E., ROSENFELD, M. G., BIRNBERG, N.

C. & EVANS, R. M. (1982a). Dramatic growth of micethat develop from eggs microinjected withmetallothionein-growth hormone fusion genes. Nature,Land. 300, 611-615.

PALMITER, R. D., CHEN, H. Y. & BRJNSTER, R. L.

(19826). Differential expression of metallothionein-thymidine kinase fusion genes in transgenic mice andtheir offspring. Cell 29, 701-710.

PALMITER, R. D., NORSTEDT, G., GELINAS, R. E.,

HAMMER, R. E. & BRINSTER, R. L. (1983).

Metallothionein-human growth hormone fusion genesstimulate growth of mice. Science 222, 809-814.

REED, K. C. & MANN, D. A. (1985). Rapid transfer ofDNA from agarose gels to nylon membranes. Nucl.Acids Res. 13, 7207-7221.

SIGNORELLA, A. P. & HYMEN, W. C. (1984). An enzyme

linked immunoabsorbant assay for rat prolactin. Analvt.Biochem. 136, 372-381.

SOUTHERN, E. M. (1975). Detection of specific sequencesamong DNA fragments separated by gel electrophoresis.J. molec. Biol. 98, 503-517.

STONE, B. A., QUINN, P. & SEAMARK, R. F. (1984).

Energy and protein sources for development of pigembryos cultured beyond hatching /// vitro. Anim.Reprod. Sci. 6, 405-412.

STONE, B. A., SEAMARK, R. F., GODFREY, B. M., QUINN,

P. & LLOYD, B. (1986). Oestrone sulphate levels inplasma of sows as a basis for prediction of litter size atterm. Anim. Reprod. Sci. 11, 51-62.

VIZE, P. D. & WELLS, J. R. E. (1987a). Spacer alterationswhich increase the expression of porcine growthhormone in E. coli. FEBS Lett. 213, 155-158.

VIZE, P. D. & WELLS, J. R. E. (19876). Isolation andcharacterization of the porcine growth hormone gene.Gene 55, 339-344.

(Received 20 October 1987 -Accepted, in revised fonn,4 January 1988)

300 P. D. Vize et al.