Hormonal Regulation of Chimeric Genes

4 downloads 0 Views 17MB Size Report
Gilboa, E., Eglitis, M. A., Kantoff, P. W., and Anderson, W. F.. 2. Mann, R., Mulligan, R. C. .... Cepko, C. L., Roberts, B. E., and Mulligan, R. C. (1984) Cell 37,. 30.
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistryand Molecular Biology, Inc.

Vol. 263, No. 33, Issue of November 25, pp. 1779617808,1988 Printed in W.S.A.

Hormonal Regulation of Chimeric Genes Containing the Phosphoenolpyruvate Carboxykinase Promoter Regulatory Region in Hepatoma Cells Infected by Murine Retroviruses* (Received for publication, June 13, 1988)

Maria Hatzoglou, Edwards Park$,Anthony Wynshaw-Boris$$, Hue-lee Cheng Kaungs, and RichardW. Hanson From the Departments of Biochemistrv and $Pediatrics and the Edison Animal Biotechnology Center, Case Western Reserve University School of Medicine, Clevelckd,Ohio 44106

Hepatoma cells were infected with replication-inThe liver plays a centralrole in metabolic processes includcompetent murine retroviruses containing the selecta- ing the synthesis and storage of body fuels, the metabolism ble genefor amino-3‘-glycosyl phosphotransferase and modification of drugs, toxins, and waste products, and (neo) and/or thenonselectable gene for bovine growth the synthesis of serum proteins. There are a variety of liverhormone (bGH). Expression of these genes was con- specific genetic disorders, resulting from single gene defects trolled by the promoter regulatoryregion of the gene affecting one or more of these pathways, making the liver an for the cytosolic form of phosphoenolpyruvate car- important target for gene transfer and genetic modification. boxykinase(GTP) (EC 4.1.1.32) (PEPCK)from the Retroviral infection has been used to efficiently transfer rat, which contains hormone and tissue-specific regu- single copies of intact genes into a variety of mammalian cell latory elements. Expression of the transducedPEPCKneo gene was stimulated by BtZcAMP and glucocorti- lines (for review see Ref. 1). A series of highly efficient retroviral vectors have been developed, which use the long coids and inhibited by insulin. The amount ofRNA which initiated within the retroviral5’ long terminal terminal repeats (LTRs)’ of the Moloney murine leukemia repeat (5’ LTR) was inhibited wheninternal promoters virus. The viral gag, pol, and enu genes can be expressed in were present in the retroviral vector. When no internal trans in packaging cell lines, such as +Z cells (Z), to produce promoter was present, expression from 5’ theLTR was defective retroviruses capable of a single round of infection of murine cells. Recombinant retroviruses have been used to higher and stimulated by glucocorticoids, due to the presence of a glucocorticoid regulatory element in the transfer genes into primary cultures of pluripotent bone mar5’ LTR. Infection of cells with retroviruses altered the row stem cells (3-9), fibroblasts (10-13), and keratinocytes basal expression and hormonal regulation of the en- (14), and often these genes were expressed. However, when dogenous PEPCK gene, but had no effect on the expres- these cells were transplanted intoanimals, the level of expression of the tyrosine aminotransferase gene, which is sion of the gene of interest was found to be low in some cases regulated in a similar manner by cAMP and glucocor- (3-7), or the transplanted cells lost the ability to express the ticoids. A segment of the PEPCK promoter acted as a gene(13-15).More information on the behavior of genes hormonallyregulatedenhancer,bringing the SV40 introduced into different cell types by retroviral infection is early promoter under the control of Bt2cAMP. A sec- required in order to maximize the expression of transferred ond, nonselectable gene (PEPCK-bGH), contained in genes. the retroviral vector together with PEPCK-neo, was In order to study the properties of genes transferred into expressed andregulatedappropriatelywhenintrohepatoma cells by retroviral infection, we have designed a duced into hepatoma cells. The proviruses were ini- series of recombinant retroviral vectors, containing a strong, tially integrated randomly into the host cell genome, liver-specific promoter which is regulated by hormones. We but after prolonged selection for expression of the transduced PEPCK-neogene, cells were selected which have used the promoter regulatory region of the gene for the contain a predominant site(s) of integration. Among cytosolic form of P-enolpyruvate carboxykinase (GTP) (EC populations of cells, however, the predominant site(s) 4.1.1.32) (PEPCK) todrive the expression of two linked genes: of proviral integration was different. The selection of the gene for the selectable marker amino-3’-glycosyl phoscells with a specific site of integration from a popula- photransferase (neo) andthe gene for bovine growth hormone tion was accelerated by the presence of PEPCK pro- (bGH). A 548-bp segment of the PEPCKpromoter regulatory moter sequences in the provirus. Despite the need to region was used in these studies since it contains elements better characterize their effects on the host cell, retro- capable of directing expression of genes specifically in the viruses appear to be versatile tools for the specific liver (16), aswell as hormonal regulatory elements for cAMP (17-ZO), glucocorticoids (18-21),and insulin (20). From these introduction of regulated genes intocells. experiments we conclude that: 1) genes introduced into hep* This work was supported by Grants DK21859 and DK24451 (to atoma cells were expressed at high levels from the PEPCK R. W. H.) from the National Institutesof Health and by a grant from promoter and regulated by hormones; 2) transcription from the American Diabetes Association (to H. C. K.). Funds were also the 5’ LTR was inhibitedin the presence of aninternal provided by the Thomas A. Edison Program of the Stateof Ohio and the Pew Charitable Trust. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solelyto indicate this fact. $ Supported by the Metabolism Training Program Grant AM07319 from the National Institutes of Health.

The abbreviations used are: LTRs, long terminal repeats; PEPCK, phosphoenolpyruvate carboxykinase; neo, amino-3’-glycosyl phosphotransferase; bGH, bovine growth hormone; bp, base pairb); kb, kilobase pair(s); DMEM, Dulbecco’s modified Eagle’s medium; Bt,cAMP, dibutyryl cyclic AMP.

17798

Retroviral Hepatoma into Transfer

Cells

17799

kb transcript of the PEPCK-bGH gene, initiating in the PEPCK promoter and terminating in the polyadenylation site of the bGH gene, after processing and splicing. Fig. 1 contains a schematic representation of these vectors. Cell Culture, DNA Transfection, and Rescue of Recombinant Virus"NIH3T3 and $2 cells were grownin Dulbecco's modifiedEagle's medium (DMEM),supplemented with 10% calf serum, whereas FTO2B and Hepal-6C cells weregrown in DMEM supplemented with 10% fetal calf serum. Transfection of DNA into the cells was carried out by calcium phosphate precipitation (28). To rescue recombinant EXPERIMENTALPROCEDURES retrovirus, apprcximately 2 X lo5 $2 cells (seeded 24 h earlier) were Materials-All DNA-modifying enzymes, nucleotides, and dextran transfected with 10 pgof retroviral vector DNA, and colonies consulfate were purchased from Boehringer Mannheim and Pharmacia taining stably integrated recombinant DNA were obtained by selecLKB Biotechnology Inc. [ c x - ~ ' P ] ~ C (3000 T P Ci/mmol), [ Y - ~ * P ] ~ C T tion P in media containing 1 mg/ml of G418. After the cells reached (7000 Ci/mmol), and Genescreen Plus were purchased from Du Pont- confluence, they were grown in DMEM media supplemented with New England Nuclear. Restriction enzymes were used according to 10% calf serum, for 48 h. The media containing the retrovirus was the specifications of the manufacturer. Nitrocellulose membrane (BA- removed and filtered through 0.45-pm Nalgene filter. The titer of 85) was from Schleicher and Schuell. All media, sera, and G418 were retrovirus in the medium produced by the $2 cells was measured by from Gibco Laboratories. The other reagents used in this study were the ability to transmitG418 resistance to NIH3T3 cells 3 weeks after infection. To infect cells, 1ml of the media containing the virus was of the highest purity available. Thymidine kinase-deficient rat hepatoma (FTO-2B) cells, provided added to 5 X lo5 NIH3T3 cells in a 100-mm2dish and incubated for by Dr. Keith Fournier, Fred Hutchison Cancer Center, Seattle, WA, 5 h at 37 "C, in the presence of 8 pg/ml polybrene. DMEM, suppleare well differentiated, with hepatocyte morphology, and express mented with 10% calf serum was added, the cells incubated for 2 liver-specific proteins, such as PEPCK andalbumin (17, 22). Mouse days, andthen split into media containing 1 mg/mlG418 (29). hepatoma (Hepa 1-6C) cells, from Dr. Gretchen Darlington, Baylor Transfected $2 cells and infected NIH3T3 cells were grown in media College of Medicine, Houston, TX, areless differentiated and do not with G418, containing 10% calf serum, whereas growth of infected express PEPCK (23). The $2 cells (2) and the retroviral vector pLJ FTO-2B and Hepa 1-6C cells in G418 required the addition of 5% (also called DOL-) (24) were a generous gift from Dr. Richard C. fetal calf and 5% calf serum to themedium. DNA Probes-The following probes used for DNA/DNA or DNA/ Mulligan, Whitehead Institute, M. I. T., Cambridge, MA. The cDNA for tyrosine aminotransferase (25) was provided by Dr. Gunther RNA hybridization experiments: PEPCK cDNA, 1.1-kb PstI-PstI Schutz, German Cancer Research Center, Heidelberg, Federal Re- fragment from the 3' end of the PEPCK cDNA, pPCKlO (30); 5'public of Germany. The gene for bGH (26) was from Dr. Fritz PEPCK, 620-bp BamHI-BglII fragment from the 5'end of the PEPCK gene (30); neo, 1-kb, BglII-EcoRI fragment from the 5' end Rottman, Case Western Reserve University, Cleveland, OH. of the neo gene (Fig. lA); TAT cDNA, 600-bp PstI-PstI fragment Construction of Recombinant Retrovirus Vector Plasmids-Four recombinant vectors were derived from the parent vector, pLJ (Fig. which includes the 3' end of the cDNA (25). All DNA probes were labeled using [ c x - ~ ' P ] ~ C Tby P ,the method l ) , which containsa chimeric genecomposed of the SV40 early promoter (including the 72-bp enhancer repeats), ligated to the neo of random oligo-priming as described by the manufacturer (Boehringene, and inserted between the Moloney murine leukemia virus LTRs. ger Mannheim, Inc). The specific activity of the DNA probe labeled The viral splicing sites have been inactivated by site-specific muta- in this manner was lo9 cpm/pg of DNA. For S1 nuclease mapping of the PEPCK-bGH RNA, the 690-bp tions. Transcripts initiating in the 5' LTRor the SV40 promoter are polyadenylated within the 3' LTR, producing RNAs of 3.9 and 3.2, BamHIIPstI fragment containing 620 bp of the promoter regulatory respectively. The vector contains a polyoma early region gene for region of the PEPCK gene and 70 bp of the 5' end of the bGH gene more efficient replication during transient expression and a pBR322 was used as a probe. This segment of DNA was labeled to a specific origin of replication. $ sequences needed for viral encapsidation are activity of10' cpm/fmol at the PstI site (5' end). Details of the included in the vector, allowing rescue of infectious virus after trans- labeling procedure have been published previously (17). Isolation and Analysis of CellularDNA and RNA-Isolation of fection of $2 cells, a retroviral packaging cell line, which contain the gag, pol, and enu genes stably integrated into their genome (2). A genomic DNA and total cellular RNA were carried out by standard further description of this retroviral vector is presented by Korman procedures which have been described in detail previously (17). Southern and Northern blotting analysis, agarose gel electrophoresis, and et al. (24). pW(-SV40)-The SV40 promoter sequences were removed from S1 nuclease analysis were performed by standard methods (31, 32). Hormonal Treatment of Cells-Cells were changed to medium pLJ by digestion with BamHI and HindIII. These DNA ends were made blunt with the Klenow fragment of DNA polymerase I and without serum and incubated for 4 or 24 h in the presence of 0.5 mM ligated. Transcription initiating within the 5' LTR and terminating BtzcAMP plus 1 mM theophylline, 1 p~ dexamethasone, or 50 nM porcine insulin or without hormones, as indicated in the various in the 3' LTR would generate a 3.4-kb RNA. pWPCK-The SV40 promoter DNA sequences were removedfrom figures. pLJ by digestion with BamHI and HindIII and replaced by the Measurement ofbGH-The concentration of bGH released into BamHI-BglII restriction fragment from the PEPCK promoter regu- the incubation medium by cells infected by retrovirus containing the latory region (positions -548 to +73). The transcriptional orientation PEPCK-bGH gene was determined by an enzyme-linked immunoabof the PEPCK promoter regulatory region was the same as the 5' sorbant assay (16). LTR. Two transcripts should be produced by the provirus: one of 4.2 kb which initiates in the 5' LTR and one of 3.2 kb from the PEPCK RESULTS promoter. Both transcripts terminate in the 3' LTR. ViralProduction by $2 Cells-Infectious retrovirus was pWPF"A PuuII-PuuII fragment (positions -416 to -61) from the PEPCKpromoter regulatory region was inserted intothe BamHI produced from the recombinant retroviral vectors (Fig. 1) site of pLJ by blunt-end ligation. This fragment was ligated in the after transfection into $2 cells. The cells were selected in same transcriptional orientation as the SV40 promoter. Two tran- media containing G418, and surviving clones containing stascripts should be produced one of 4.3 kb initiating in the 5' LTR bly integrated plasmid DNA were combined and expanded and the second of3.2 kb initiating in the PEPCK-SV40 chimeric as a mass culture. The virus produced from cells transfected promoter. Both transcripts terminate in the 3' LTR. pLJPP+,pLJPCKbGH pWPCKbGH-A chimeric gene containing an EcoRI-BglII frag- with pLJ,pLJ(-SV40),pLJPCK, ment of the PEPCK promoter regulatory region (positions -461 to were designated vLJ, vLJ(-SV40), vLJPCK, vLJPP+, +73), linked to the entire structural gene for bGH (26), was ligated VLJPCKbGH, respectively. The viral titer from these cells into the unique ClaI site of pLJPCK by the use of a polylinker flanked was determined by the ability of the virus to transmit G418 by ClaI sites (27). The transcriptional orientation of the chimeric resistance to NIH3T3 cells by infection (Table I). The level PEPCK-bGH gene was opposite to the 5' LTR. Three transcripts should be produced from this provirus: 1)a 7.2-kb RNA, initiating in of virus production by various $2 cell lines ranged from lo3 the 5' LTR andterminating inthe 3' LTR; 2) a 6.2-kb RNA initiating to IO7 colony-forming units/ml and contained no wild-type in the PEPCK promoter and terminating in the 3' L T R and 3) a 1- virus as determined by the inability of media from infected

promoter contained in the provirus; 3) the expression and regulation of the endogenous PEPCK gene was altered by retroviral infection; 4) two genes introduced into cells by the same virus were expressed at high levels and regulated by hormones; and 5 ) a predominant site of proviral integration in the genome was noted in cells grown for 3 months under selective conditions, suggesting clonal overgrowth.

17800

Retroviral Transfer into HepatomaCells A

Born HI SmoI Sal I

FIG, 1. Retroviral vectors. A , pLJ (DOL-), is a Moloney murine leukemia virus-based retroviral vector containing the SV40 early promoter, linked to the structural gene for neo. The dotted box represents the polyoma early region gene. Numbers refer to sizes of the DNA fragments produced by digestion with the restriction enzymes shown in the figure. This figure was redrawn from one provided by Dr. Richard Mulligan and details on the construction of the vector is provided by Korman et al. (24). B , pLJ-derived retroviral vectors containing the PEPCK promoter regulatory region. The arrows beneath each vector refer to the direction of transcription from internal promoters. A detailed description of the construction of these vectors is contained under “Experimental Procedures.”

TABLE I

M-MuLV LTR

1 .

Eco RI

determined after infection by recombinant retrovirus. Cells were infected with vLJPCK and selected in G418-containing media until resistant colonies appeared (3 weeks). Genomic Infectious retroviral stocks were produced from mass cultures of DNA wasisolated from the combined colonies (approximately G418-resistant $2 cells which had been transfected with the vectors listed above. One ml of medium collected from these cells was used 5 X lo3 cells) and analyzed by Southern blotting after digesto infect 5 X 10’ NIH3T3 cells. Forty-eight h afterinjection, the cells tion with appropriaterestriction endonucleases. Southern were removed from the plates with trypsin and replated in medium blots were hybridized with the neo or 5’ PEPCK probes. containing 1 mg/ml of G418. The retroviral titer is expressed as Infected cells contained the intact provirus, since after Sac1 colony-forming units (cfu) and represents the number of G418-resist- digestion, only the predicted 4.1-kb fragment hybridized with ant NIH3T3 cells found 3 weeks after infection. either probe (Fig. W ) .In uninfected cells, no DNAhybridized Retroviral vector Titer of $2 cells with the neo probe, whereas the 5’ PEPCK probe hybridized cfulml only to the endogenous PEPCK gene. PLJ 4 X 104 In order to determine thesite(s) of integration of the pLJ(-SV40) 2 X 10‘ provirus in the host cell genome, we digested the same DNA pLJPCK 8.5 X lo6 with two restriction enzymes: BglII, which cuts once within pLJPP+ 6.3 X 103 the provirus, and BarnHI, which is not presentin the provirus. pLJPCKbGH 5 X lo7 Digestion with either enzyme should yield multiple bands of various sizes which hybridize with the neo DNA probe, after cells to transmit G418 resistance to uninfected NIH3T3 cells. analysis by Southern blotting. Nospecific bands wereobIt is of interest that a retroviral vector, such as pLJPCKbGH, served when DNAfrom infected cells was digested with either which contains two internal promoters, generates infectious of the two restriction endonucleases and hybridized to the retrovirus with a titer similar to that of vectors containing a neo probe. The only specific bands which hybridized with the single internal promoter. The +2 cells containing the PEPCK- 5’ PEPCK probe corresponded to the endogenous PEPCK bGH gene also produced bGH in the medium (datanot gene (Fig. W). Thus, the pLJPCK provirus was integrated shown). randomly into the genome of FTO-2B cells after 3 weeks of Integration of Recombinant Retrovirus into Cells by Viral selection in media containing G418. Infection-The integrity of the provirus and its site of inteThe mixed population of infected cells, 3 weeks after infecgration into the genome of FTO-2B cells at various times was tion, was grown in medium containing G418 for an additional Titer of retrovirus produced by $2 cells transfected with various retroviral vectors

17801

Retroviral Transfer into HepatomaCells B

A FIG.2. Analysis of the pattern of integration of the provirus into the genome of infected cells. A , FTO-2B cells were infected with v U P C K and selected for 3 weeks in a medium containing G418. Surviving cells (approximately 5 X lo3 cells) were pooled, and the genomic DNA was isolated. Control cells were not infected. Approximately 20 pg of genomic DNA wasdigested with SacI, which cuts once in each of the LTRs, yielding an intact provirus. The DNA wasalso digested with EglII, which cuts once within the provirus, and with BamHI, a restriction site not included within the provirus. The DNA fragments were separated by electrophoresis on 0.8% agarose gel and hybridized with either a neo or 5’ PEPCK probe. E, the FTO-2B cells from the experiment shown in the left panel weregrown in culture medium containing G418 for an additional 2 or 3 months (as indicated at the top of the figure). The DNAwas extracted from these cells, digested with SacI and BglII, and analyzed by Southern blotting using either neo or the 5’ PEPCK DNA probes. The arrows indicate the size of the DNA fragments which hybridize with the DNA probes. The 2.4- and 7-kb bands represent fragments from the endogenous PEPCK gene which hybridize with the 5’ PEPCK probe. A drawing of the provirus, indicating the sites of restriction endonuclease digestion, is included at the bottom of the figure, together with the predicted size fragments generated from the digestion. The DNA probes are shown as boxes in the drawing: open box, 5’ PEPCK probe; dark box, neo probe.

nn

-n

2 and 3 months. Genomic DNA was isolated, digested with the SacI and BglII, and analyzed by Southern blotting, using the neo and 5’ PEPCK probes. Predominant bands were detected inDNA from the same populationof cells which had a random pattern of proviral integration when analyzed at 3 weeks. After BglII digestion, predominant DNA bands of 6.2 kb (neo probe) and 9.2 kb (5’ PEPCK probe) were observed in one population (Fig. 2B). Additional DNA bands, which hybridized with these probes, were also observed. The intensity of these additional bands decreased between 2 and 3 months in culture, whereas the predominant bands (6.2 and 9.2 kb) became more intense. The provirus was contained in a 4.1-kb segment of the genome as demonstrated by SacI digestion (Fig. 2B), and the appropriate restriction sites were maintained withinthe provirus (data notshown), demonstrating that therewere no gross rearrangements. The intensityof this 4.1-kb band did not change in cells cultured over this period. Proviral integration in three separate populations of cells derived from individual infections with vLJPCK (approximately 5 x lo3 individual colonies in each population) was analyzed as described above. DNA from infected cells was digested with BglII, analyzed by Southern blotting, and hybridized with the neo probe. In DNA from cells in culture for

t70kb 6.2kbb-,

C 9 . 2 kb 4- 70 kb

4.1kb+ t4.Ikb

t 2.4 kb t2.4 kb

2 months, we noted specific bands which were different for each population of infected cells. In one population of infected cells (infection I inFig. 3), we analyzed the patternof proviral integration at 2 and 6 months after the cells were placed in culture. After 2 months, we noted a predominant DNA band a t 7.8 kb, together with other, less intense bands a t 5.6 and 4.4 kb. After 6 months inculture, these low abundance bands were less intense than at2 months, whereas the predominant band at 7.8 kb was more intense. A different set of predominant bands were noted in two other populations of infected cells: 6.2 and 4.2 kb in infection I1 and 17, 5.6, and 3.7 kb in infection I11 (Fig. 3). Thus, in multiple, separate populations of infected cells, the growth of individual cells results in predominant DNA bands, reflecting the position of the provirus in the host cell genome after clonal expansion. These bands were different for each population of cells tested. The patternof proviral integration was determined in cells infected with vLJ and vLJ(-SV40), which lack the PEPCK promoter regulatory region. A mass culture of infected cells, maintained in G418 for 3 months, had the intact provirus integrated in a random manner (Fig. 4). However, after prolonged selection (more than 4 months), these cells contained a low percentage of specifically integrated provirus (data not shown). Thus, FTO-2B cells infected with all retroviruses

Retroviral Transfer into HepatomaCells

17802

IIm . I

2

6

months

2

2

months

. I I

t 17kb

78 k b 4

c 5.6 kb+

t 6k. 2b c

4.4 kb “1

t 5.6 kb

4.2 kb t 3.7 kb

i-

e

3.0kb+ FIG. 3. Pattern of integration of vLJPCK into the genome of separate populations of FTO-2B cells. DNA was isolated from FTO-2B cells infected with vLJPCK, digested with BglII, and analyzedby Southern blotting. I, I I , and III represent three separate populations of infected cells maintained in culture for 2 or 6 months. The size of the DNA fragments hybridizing with the neo probe are indicated in the figure. Experimental detailsare provided in the legend to Fig. 2.

I)

c)

t3.4kb

tested contained predominant sites of proviral integration after prolonged growth in G418, probably due toselection and neo 1 clonal expansion. FIG. 4. Pattern of integration of provirus which does not Regulation of Gene Expression of the pL.JPCK Provirus by contain PEPCK promoter sequences. FTO-PB cells were infected Bt2cAMP Dexamethasone and Insulin-The expression and with either VUor vLJ(-SV40) and selected in G418 containing hormonalregulation of the chimeric PEPCK-neo gene in medium for 3 weeks (as in Fig. 2) and then maintained in the same FTO-2B cells infected with vLJPCKwas analyzed by North- medium for an additional 2 months. DNA was extracted from the ern blotting. An RNA transcriptof 3.2 kb hybridized with the cells and digested with SacI and BglII as indicated in the figure. The neo probe. This would be expected if transcription initiated resultant fragments were analyzed by Southern blotting using a neo The arrows indicate the provirus generated after SacI digesat the PEPCK promoter and the newly synthesized RNAwas probe. tion. polyadenylated within the 3’ LTR (Fig. 5). Treatment of the cells with Bt2cAMP plus theophyllineor dexamethasone in(Fig. 6). There was a marked induction of the endogenous creased the expression of the PEPCK-neo gene by 6- to 7PEPCK gene by both Bt2cAMP and dexamethasone incells fold after 4 h of treatment. Insulin, when added together with which were not infected with the virus (Fig.6). However, in Bt2cAMP, blocked the induction of neo gene expression by cells infected with vLJPCK, the endogenous PEPCK gene both the cyclic nucleotide and by dexamethasone. If the cells was only slightly increased (less than 2-fold) by the addition were pretreated with insulin for 2 h before the addition of of Bt2cAMP or dexamethasone. These cells were responsive Bt2cAMP and then harvested2 h later, thelevel of neo RNA to Bt2cAMP or dexamethasone since the PEPCK-neo gene, was lower than that noted in controlcells. We did not detect introduced into FTO-2B cells by infection, was induced by a 4.2-kb species by Northern blotting, which would be exboth compounds. Insulin addition caused a marked decrease pected if transcription initiated at the 5‘ LTR. However, a in endogenous PEPCK mRNA in infected cells treated with low level of transcription from the 5‘ LTR was detected when these hormones (Fig. 5). The amount of RNA in each lane the RNAwasanalyzed by S 1 nuclease mapping (data not was identical, based on hybridization of the blots with pshown). tubulin (data not shown). We have also noted that insulin Expression of the Endogenous PEPCK and Tyrosine Ami- decreased the basal level of PEPCK gene expression in cells notransferase in Cells Infected with vL.JPCK-The effect of infected with retrovirus, in the absence of BtzcAMP or dexahormones on the expressionof the endogenous PEPCK gene methasone (data not shown). in FTO-2B cells infected with vLJPCK was also measured Since infection of FTO-2B cells with vLJPCK was associusing the identical Northern blot. As predicted, a 2.8-kb RNA ated with a marked increase in the basal concentration of hybridized with the PEPCK cDNA probe (Figs 5 and 6). PEPCK RNA, we measured the level of mRNA for tyrosine However, the basal concentration of PEPCK mRNA in inaminotransferase, a gene regulated by CAMP andglucocortifected cells was very high, compared with uninfected cells coids in a manner similar to PEPCK (25). The infection of

Retroviral Hepatoma into Transfer

Cells

17ao3

3.2KbJ 2.8Kb+

probe”.)

-PEPCK-~DNA~-~~~-

FIG. 5. Effect of hormones on RNA from FTO-2B cells infected with vLJPCK. FTO-2B cells were infected with vWPCK and maintained in culture with G418 for 2 months. The level of RNA for neo and for endogenous PEPCK was determined by Northern analysis of RNA isolated from cells which were treated for 4 h with the various hormones; except for cAMP/insulin and insulin/cAMP, in which case thecells were treated for 2 h with Bt2cAMP or insulin, respectively, followed by the addition of the other hormone for 2 h. The multiple bands which hybridize with the PEPCK cRNA probe may be the precursor RNAs noted previously (54). Details on the procedure for the extraction and analysisof RNA and the concentrations of the hormones used in this experiment are provided under “Experimental Procedures.” CAMP, Bt2cAMP;Dex, dexamethasone; INS, insulin; CON, control (no added hormones).

“PEPCK-CDNA~

-TAT

cDNAA

FIG. 6. Effect of retroviral infection on the expression of the endogenous PEPCK and tyrosine aminotransferase genes in FTO-2B cells. Cells were maintained in culturemedium containing G418 for 2 months and then treated with the various hormones as indicated.Detailsonthehormonaltreatment of the cells are presented in the legend t o Fig. 4. RNA was extracted from the cells and analyzedby Northern blotting, usingas probes cDNAsto PEPCK andtyrosineaminotransferase. CONTROL, notinfectedwiththe retrovirus; p W P C K , infected with vWPCK; CON, not treated with hormones; CAMP,Bt2cAMP; Dex, dexamethasone. The 2.8- and 2.6kb bands are PEPCK and tyrosine aminotransferase mRNA, respectively.

FTO-2B cells with this virus had noeffect on the basallevel of tyrosineaminotransferasemRNAandtheaddition of the level of this 3.2-kb RNA. Insulin, on the other hand, had Bt2cAMP or glucocorticoids to thecells stimulated the expres- no effect on SV40-neo gene expression.Based on these obsersion of the endogenous tyrosine aminotransferasegene in the vations, the marked stimulatory effect of Bt2cAMP and the expected manner (Fig. 6). negative effect of insulin on expression of the neo gene in cells infected with vLJPCK appears to be dependent on the Expression and Regulation of Proviruses Which Did Not PEPCK promoter regulatory element. The effect of dexaContain PEPCK Sequences-Cells were infected with vLJ(-SV40) and vLJ to determine the contribution of the methasone may be due to sequences in the5’ LTR or in the PEPCK sequences to the hormonal regulationof transduced PEPCK promoter. Thenegative effect of Bt2cAMP on SV40genes and their effect on the expression of the endogenous neo expression ispossibly the resultof a negative element for PEPCK gene. These proviruses have either no internal pro- CAMPpresentintheSV40promoter.Thebasal level of moter (vLJ(-SV40)) or the SV40 promoter ( v u ) linked to expression of the endogenous PEPCK gene was increased the neo gene. Two identical Northern blots were hybridized after infectionof FTO-2B cells with vLJ, and treatment with Bt2cAMP and dexamethasone did not cause a further increase with either a neo or a PEPCK cDNA probe. In vLJ(-SV40) of this gene (Fig. 7B).The endogenous gene infected cells (Fig. 7 A ) ,the neo probe hybridized with a 3.4- in the expression kb RNA whichcorresponded to the full-length viral transcript was expressed and regulated in cells infected with containing neo. Treatment of cells with Bt2cAMP caused a vLJ(-SV40) (Fig. 7 A ) , in a manner identical to uninfected small increase in neo RNA, whereas insulin did not signifi- cells (Fig. 7B). Insulin decreased the expression of the endogcantly alter neo gene transcription. In contrast, the addition enous PEPCK gene in cells infected with either virus. These of dexamethasone to the infected cells caused a more than results suggest that both SV40 and PEPCK sequences can 10-fold induction of gene transcription from the 5’ LTR. alter the expressionof the endogenous PEPCKgene. Thus, dexamethasone has the potential to control the expresThe Expression and Regulation of Proviruses Containing a sion of a structural gene included in the provirusby altering Chimeric PEPCK-SV40Promoter Regulatory Element-A transcription from the 5’ LTR, probably due to the presence segment of the PEPCK promoter regulatory region (positions of a glucocorticoid regulatory element in the LTR (33). In -416 to -61) was linked to the 5‘ end of the SV40-neo gene cells infected with vLJ (Fig. 7 B ) ,transcription from the 5’ (pLJPP’, Fig. 1B) and introduced by infection into FTO-2B LTR (3.9-kb RNA) was undetectable and the internal SV40 and Hepa 1-6C cells (Fig. 8). This segment of the PEPCK promoter (3.2-kb RNA) was expressed a t high levels. The gene can actas a glucocorticoid- and CAMP-regulated enhanaddition of BtZcAMP decreased and dexamethasone increasedcer element when linked to a heterologous gene (18).The

-

Hepatoma into Transfer Retroviral

17804 A

pLJ(-SV40)

v)

v)

I *

Cells B

i v)

z

E

.L

-

CONTROL

PLJ I

+

3.4 kb +

-nco-

LPEPCKCDNAJ

-nco-

“PEPCKcDNA-

FIG. 7. Effect of hormones on RNA from FTO-PB cells infected with vLJ and vLJ(-SV40). FTO-ZB cells were infected with VUor vW(-SV40) and grown in medium containing G418 for 2 months. The level of RNA for neo and the endogenous PEPCK gene was determined by Northern analysis of RNA isolated from these cells after treatment with the various hormones for 4 h. Control refers to RNA from FTO-2B cells which were not infected with retrovirus andCON to cells infected with retrovirus but not treated with hormones; CAMP, BhcAMP; Dex, dexamethasone; INS, insulin. A , cells infected with VU.B, cells infected with vLJ(-SV40). Arrows indicate probes. The 3.4- and 3.2-kb RNA the sizes of the RNA whichhybridizes with either neo or PEPCK cDNA transcripts initiate within the5’ LTR or theSV40 promoter, respectively, whereas the 2.8-kb band represents the mRNA from the endogenous PEPCK gene. Furtherdetailsare provided in the legend to Fig. 5 and under “Experimental Procedures.”

SV40 early promoter containing its own enhancer sequences is a strong promoter in many cell types and is capable of driving the expression of linked genes in retroviral vectors (24). FTO-2B or Hepa 1-6C cells infected with vWPP+ expressed the neo gene from the SV40 promoter (Fig. 8). The basal level of neo gene expression in vLJPP+ infected cells waslower than notedin cells infected with v W (Fig. 8, compare first two lanes). The addition of BbcAMP toinfected FTO-2B cells caused a marked induction in the levels of neo RNA transcribed from the SV40 promoter, but the cyclic nucleotide inhibited transcription from the same chimeric gene, when infected into Hepa 1-6C cells. The reason for these differences in the regulation of transcription of this chimeric gene in two different cell lines is not clear. However, Hepa 16C cells are less differentiated than the FTO-2B cells and do not express the endogenous PEPCK gene. Finally, the basal level of expression of the endogenous PEPCK gene is not altered by infection of FTO-2B cells with v U P P + (Fig. 8). Infection of Hepatoma Cells with a Retrovirus Containing TwoGenes between the LTRs-The usefulness of the p W system for introducing a second, nonselected gene into cells was tested by infecting FTO-2B and Hepa 1-6C cells with VWPCKbGH, which contains two internal chimeric genes (Fig. 1).The provirus was integrated into thegenome of these cells and the PEPCK-neo gene was expressed and regulated as described above for cells infected with vWPCK (data not shown). In order to determine the level of expression of the PEPCK-bGH gene included within theLTRs of the pWPCKbGH provirus, we measured the concentration of PEPCK-bGH mRNA in infected cells. Low levels of this

mRNA were detected by Northern blotanalysis of total RNA from cells treated for 24 h with various hormones (data not shown). Quantitative S1 nuclease analysis of the RNA from these cells demonstrated that PEPCK-bGHmRNA initiated at the correct transcription start site within the PEPCK promoter (Fig. 9). The patternof regulation of gene expression by Bt2cAMP, insulin, and dexamethasone in both FTO-2B and Hepa 1-6C cells was the same as that noted previously for PEPCK-neo (Fig. 5). Both cell linessynthesized and released considerable quantities of immunoreactive bGH into the culture medium (Fig. 10). The amount of bGH produced by Hepa 1-6C cells is six times that from FTO-2B cells (Fig. IO), even though the level of PEPCK-bGH mRNA was only 2-fold higher in Hepa 1-6C cells (Fig. 9). The addition of BbcAMP to bothFTO-2B and Hepa 1-6C cells increased the concentration of bGH in the medium after 24 h, whereas dexamethasone was effective only in FTO-2B cells. Insulin, added together with BbcAMP, had little effect on the secretion of bGH by either cell line. These differences in the response to hormones between the level of bGH mRNA and bGH in the medium, noted in Figs. 9 and 10, may be due to effects on bGH synthesis, processing, and release by the cells. A lack of correlation between mRNA levels and protein secretion of human Factor IX in cells infected with virus containing Factor IX cDNA has been observed (13). DISCUSSION

Retroviral vectors represent a valuable means of introducing genes into thegenome of cells and animals(1).However, the expression and regulation of only a limited number of

- -

Retroviral Transfer into HepatomaCells FTO-2B

Hepa I-6C I

I

Hepo 1-6 C

FTO-2B

v)

z

2 %

n n

r n n

17805

+

n

o r

u-

’ U

pLJ pLJ-PP* pLJ pLJ-PP* probes

+ I-

neo-l

tPEPCKcDNAi

PLJ PLJ-PP’ I--neo-i

FIG. 8. Effect of regulatory elements in the PEPCK promoter regulatory region on the expression of the SV4O/neo gene introduced into hepatoma cells by retroviral vectors. FTO-2B and Hepa 1-6C cells were infected with v U P P + or vLJ (as indicated in the figure) and grown in medium containing G418 for 2 months. The cells were treated for 4 h with the hormones indicated in the figure, the RNA isolated, and analyzed by Northern blotting. CON refers to cells not treated with BtZcAMP; CAMP is BtZcAMP. The DNA probe used for Northern analysis was the neo probe or the PEPCK cDNA, as indicated. The 3.2-kb RNA is the transcript from the PEPCK promoter and the 2.8-kb band is mRNA for the endogenous PEPCK gene. The drawing a t the bottom of the figure represents the provirus and shows the position of the PEPCKpromoter regulatory region in the retroviral vector. The experimental conditions for hormonal treatment of the cells are described in the legends to Figs. 5 and 6.

chimeric genes introduced into cells via retroviral vectors have been tested, so that problems related to the interaction of promoters witheach other andwith the viral promoter(3436), the responsiveness of promoters containing various regulatory elements (37-39), and thespecificity of integration of the provirus in the host cell genome (40-46) remains to be more fully explored. In this paper, we report that retroviral vectors can be used to introduce chimeric genes containing the PEPCK promoter regulatory region (19, 20) into hepatoma cells. Transcription from the PEPCK promoter was high and was regulated by hormones, whereas transcription from the viral promoter was extremely low. Expression of the endogenous PEPCK gene was altered by infection. In addition, cells containing the provirus integrated into its genome at specific site(s) overgrew the population of cells after prolonged selection in G418. Expression of Chimeric Genes Introduced into Cells with Retrouiral Vectors-The usefulness of retroviral vectors for the introduction of genes into cells and animals depends on

-Pst, I 0*

6 9 0 b p probe

FIG. 9. S1 nuclease mapping of PEPCK-bGH RNA from hormonally treated hepatoma cells which were infected with VLJPCKbGH. FTO-2Band Hepa 1-6C cells were infected with vUPCKbGH and grown in medium containing G418 for 2 months. The cells were treated for 24 h with the hormones indicated in the figure and the RNA isolated and analyzed by S1 nuclease mapping. A 681-bp DNA probe, containing 61 bp of bGH coding sequences, together with 620 bp of the PEPCK promoter, which was labeled at the 5’ end, was used for mapping the start siteof gene transcription. The DNA fragment, protected after S1 nuclease digestion, is shown schematically at thebottom of the figure. CON, control (no hormones added); CAMP, BtZcAMP; Der, dexamethasone; INS, insulin.

the ability of the provirus to express multiple genes included between the LTRs. For some vectors, the inclusion of an internal promoter linked to a selectable gene greatly reduces the promoter activity from the 5’ LTR, whereas with other vectors, the internal promoter is less active transcriptionally than theLTR. In general, the activity of an internalpromoter is dependent upon whether it is driving the expression of the selectable gene included in the retroviral vector. The negative effect of one promoterupon another (“promoter suppression”) has been analyzed by Emerman and Temin (34, 35) using LTRs from avian and murine retroviruses, together with the promoter for the Herpes virus thymidine kinase gene. Transcription from the LTRsof both viruses was suppressed when the thymidine kinase promoter was included in theprovirus. This suppression was not related tothestrength of the thymidine kinase promoter or to the distance between the promoters but was inversely related to theDNase I sensitivity of the chromatin surrounding the integrated provirus. When vectors which contain the PEPCKpromoter regulatory region linked to the neo gene were transfected into $2 cells, both the 5’ LTR and the PEPCK promoters were active and high-titer virus was produced. However, when FTO-2B cells were infected with the virus, the PEPCK promoter was expressed at much higher levels than the 5‘ LTR. This was

17806

Retroviral Transfer into Hepatoma Cells 70

60

FIG. 10. Effects of hormones on

the productionof bGH by hepatoma cells infectedwith VLJPCKbGH. FTO-2B and Hepa 1-6C cells, which had been infectedwithvLJPCKbGHand grown in medium containing G418 for 2 months, were treated for 24 h with var9. ioushormonesasindicatedinFig. Immediately before RNA was extracted from the cells for S1 nuclease analysis (Fig. 9),media was removed and the level of bGH determined by the immunological assay(19). The concentration of bGH in the medium is expressed relative to the concentration of total RNA extracted from the cells. Note that here is a 10-fold difference in the ordinates for bGH concentration in the left and right panels. CONTROL, no hormones added; CAMP, BhcAMP; Dex, dexamethasone.

T

-

FTO 28

1

Hepa I - 6 C

700 600

2 50

500 ; i

z

z

L

(L

m

0

& 40

400

-

0

&

-r

0

X

c3

(3

a 30

300 n

20

200

IO

100

CONTROL cAMP

De x

cAMP 4-

Insulin

CONTROL

cAMP

Dex

cAMP

+

Insulin

observedfor all viruses containing sequences fromthe latory element in FTO-2B cells. Clearly, the presence of the PEPCK promoter regulatory region. Thus, the PEPCK pro- cAMP regulatory element from the PEPCK promoter can override the negative effect of Bt’cAMP on the transcription moter, when driving the expression of a selectable neo gene is more active than the5‘ LTR of the provirus. The 5’ LTR of the SV40 promoter, but only in FTO-2B cells, since the was active in $2 cells, since cells transfected with vectors same gene was negatively regulated by Bt2cAMP in Hepa 1containing the PEPCK promoter produced virus at a level 6C cells. Despite the inducibility of the SV40 promoter by equal to or greater than those without this promoter (Table Bt2cAMP when PEPCK regulatory sequences were present I). The 5‘ LTR was also transcriptionally active in NIH3T3 ( v U P P + infected cells), the basal level of transcription from and normal rat kidney cells infected with these viruses, as the chimeric PEPCK-SV40 promoter was markedly reduced, demonstrated by Northern blot analysis.’ Since the 5‘ LTR relative to the SV40 promoter. Thus, the chimeric PEPCKis extremely responsive to glucocorticoids in the absence of SV40 promoter has potential usefulness in studies requiring an internal promoter, it may be possible to exploit this effect a strong promoter,which has a low level of basal transcription t o produce high-titer retrovirus inpackaging cell lines which but which can be markedly and rapidly induced by CAMP. This type of “on-off’ mechanism might have application for are sensitive to glucocorticoids, even when strong, internal chimeric genes introducedintoanimaltissues, where the promoters are present in the provirus. expression of the linked structural gene must be rigorously Hormonal Regulation of Chimeric Genes Introduced into Cells with Retroviral Vectors-An attractive feature of this controlled. system of gene insertion intocells via retroviral vectors is the Previously, we reported that a chimeric PEPCK-neo or possibility of including chimeric genes containing regulatable PEPCK-TK gene, containing segments of the PEPCK propromoters in the vectors. The PEPCK promoter regulatory moter regulatory region of varyinglength (we have used fragments from positions -2000 to +73) was not responsive element is a n excellentcandidatefor use withretroviral systems, since it is a strong promoter, especially in liver and to insulin when transfected into FTO-2Bcells (17). In these kidney, which can be acutely regulated by a variety of hor- studies, the effect of insulin was measured by its ability to mones (16-21). We have shown in this study that retroviral block the induction of gene transcription byBtzcAMP in vectors containing 547 bp of 5”flanking sequence of the stably transfected cells. The endogenous PEPCK gene, preas PEPCK promoter, a region known to contain CAMP,gluco- sent in FTO-2Bcells is responsive to insulin and was used corticoid (17-19), and tissue-specific regulatory elements (16), a control for the effectiveness of the hormone. In contrast, gene was introduced intocells conferred hormonalresponsiveness to linkedgenes. The effect when the chimeric PEPCK-neo of hormones was noted even in a retroviral vector such as by retroviral infection, insulin hada marked inhibitory effect pPLJPCKbGH, which contains two PEPCK promoters, both on the ability of BtzcAMP and dexamethasone to increase transcription. When the same retroviral vector, containing of which were regulated by hormones. Transcription from the proviralSV40 promoter in the two the chimeric gene, was transfected into FTO-2B cells, no hepatoma cell lines (FTO-2B and Hepa 1-6C) infected with effect of insulin on gene transcription was noted.3 This sugvLJ (a virus containing no PEPCK sequences), was decreased gests that the insulin regulatory element is contained within by theaddition of Bt’cAMP. However, a segment of the -547 to +73 bp of the promoter regulatory region of the PEPCK gene, a finding which is in agreement with a study PEPCK regulatory region (between positions -416 and -61), when linked to the SV40 promoter, acted as a positive reguG. R. Vandenbark, E. Park, M. Hatzoglou and R. W. Hanson, unpublished observations. * M. Hatzoglou and R. W. Hanson, unpublished data.

17807

Retroviral Transfer into HepatomaCells by McGrane et al. (16) using transgenic mice. In this study, the expression of a chimeric PEPCK-bGH gene, containing the same region of 5”flanking DNA from the PEPCK gene, was markedly suppressed by feeding the animals a diet high in carbohydrate. Recently, Magnuson et al. (20) showed that insulin can block the induction by glucocorticoids and cAMP of a chimeric PEPCK-CAT gene (positions -600 to +73 of the PEPCK gene), which was expressed transiently in H41IE hepatoma cells. Differences in the expression of chimeric genes introduced into cells by transfection as compared to infection have been noted by other investigators (38). These differences in the response of hormonal regulatory elements introduced into cells by different techniques underlines the importance of the method of gene transfer on the regulated expression of the transgene as well as the cellular and chromosomal environment. Effect of Retroviral Infection on the Expression of the Endogenous PEPCK Gene-Another interesting finding noted in this study is the stimulatory effect of retroviral infection on the basal level of expression of endogenous PEPCK gene in FTO-2B cells. This effect did not require the presence of PEPCK gene sequences in the provirus and was observed with all of the chimeric genes introduced by infection, except for vLJPP+ andvLJ(-SV40). One common element between these two vectors is the low level of expression of the neo gene in infected cells containing either provirus (compare Figs. 7 and 8). A high level of expression of the product of the neo gene, amino-3’-glycosyl phosphotransferase, could alter factors which regulate transcription of the PEPCK gene. The high basal level of the endogenous PEPCK gene is probably not due toan increase in intracellular cAMP caused by retroviral infection, since infection of FTO-2B cells with v U P C K did not alter the basal level of expression of the tyrosine aminotransferasegene. The effect was observed with viruses which do not contain PEPCK sequences, making it unlikely that factors which negatively regulate transcription of thePEPCK gene are removedby binding to proviral sequences. Whatever the mechanism, it is clear that retroviral infection can alterexpression of the endogenous PEPCK gene in hepatoma cells. Specificity of Proviral Integration-Recent studies have shown that there are a limited number of genomic sites for retroviral integration (40, 42, 44, 47, 48). Here, we demonstrate thatproviral integration was initially random in FTO2B cells, but when these cells were maintained in culture for 2 months, cells were selected which contained the provirus in a predominant location. The ability to select cells which have the provirus integrated in a specific location appeared to be a general property of retroviruses containing the neogene. However, the presence of the PEPCK promoter regulatory region in the recombinant retroviruses greatly accelerated this selection. When PEPCK sequences were present, the selection process and clonal expansion occurred within 2 months. In contrast, it took 4 months for this selection to occur in cells infected with virus where PEPCK sequences were absent, and the percentage of cells demonstrating this specificity was low. We have infected a variety of transformed cells, including the hepatoma cell lines FTO-2B (rat) and Hepa 1-6C (mouse), NIH3T3 (fibroblast)cells, as well as liver cells in theintact animal by microinjection of infectious retrovirus into fetal rats in utero.4 Only the hepatoma cells and the livers of the animals injected with the retrovirus demonstrated specificity in the patternof proviral integration. Battula and Temin (46)reported the integration of spleen necrosis virus in infected chicken fibroblasts early and late

* M. Hatzoglou and R. W. Hanson, manuscript in preparation.

after infection. They found that proviral integration into the host cell genome was random in acutely infected cells, but in chronically infected cells, the virus appeared to be integrated at a single site in the genome. Our observations are compatible with these findings. The ability to detect nonrandom integration events may depend on the cell line and retroviral vector used, as well as the time after infection at which integration is studied. It is probable that a clonal expansion of a subset of hepatoma cells, containing the provirus at a specific position, resulted in the ultimateintegrationpattern noted after 2 months of growth. Integration of the provirus at one or more of these sitesmay result inthe overproduction of the neo gene product. This may provide selective advantages for these cells, since they are grown in the presence of high concentrations of G418. A small population of infected cells could divide a t a rate exceeding that of other cells containing the provirus, resulting in clonal overgrowth. Barklis et al. (40) found that the site of integration of the provirus was important for the expression of transduced neo genes in embryonic carcinoma cells. The provirus was in two distinct chromosomal locations in 30% of the selected cell lines. This suggests that thesite of integration is important for proviral expression, possible because it is an actively transcribed region of chromatin. Alternatively, a growth factor may also be induced by viral infection, providing a selective advantage for a subpopulation of cells. Extensive cloning and sequencing of a number of these integration sites will be required to explain these results, as was recently reported for the integration sitesof Rous sarcoma virus in chicken cells (48). Gene Transfer into Hepatic Cells-The liver is an important organ for gene transfer, both for metabolic studies andfor the correction of genetic diseases. Retroviral transduction has been used to introduce genes into hepatoma cells (49) and primary hepatocytes (50-52). Progress in the transplantation of hepatocytes into animals (53) may provide an approach to gene transfer into animals. Recently, we have used retroviruses to directly infect fetal hepatocytes in vivo by injection of infectious retrovirus into the peritoneal cavity of fetal rats late in de~elopment.~ The provirus was integrated into the liver and thechimeric PEPCK-neo gene included in the virus was expressed and regulated by hormones. The efficient transfer of genes by retroviral transduction, coupled with the high level of expression and hormonal regulation of genes linked to the PEPCK promoter regulatory region, may provide an ideal means to alter gene expression and metabolic processes in the liver. Acknowledgments-We are indebted to Dr. Richard Mulligan for generously providing us with the retroviral vector pLJ (DOL-) prior to its publication. We also wish to acknowledge the excellent technical assistance of Yashomita Patel and the advice provided during the course of these studies by Drs. Thomas Wagner and Fritz Rottman. We are especially indebted to Drs. David Samols, Hsing-Jien Kung, and William Roesler for reviewing the manuscript and to Josh Bloom for assaying bGH. REFERENCES 1. Gilboa, E., Eglitis, M. A., Kantoff, P. W., and Anderson, W. F. (1986) Biotechniques 4, 504-512 2. Mann, R., Mulligan, R. C., and Baltimore, D. (1983) Cell 33, 153-159 3. Williams, D. A,, Lemischka, I. R., Nathans, D. G., and Mulligan, R. C. (1984) Nature 310,476-480 4. Miller, A. D., Eckner, R. J., Jolly, D. J., Friedman, T.,and Verma, I. M. (1984) Science 225, 630-632 5. Keller, G., Paige, C., Gilboa, E., and Wagner, E. S. (1985) Nature 318,149-154

17808

Retroviral Transfer into HepatomaCells

6. Eglitis, M. A., Kantoff, P.,Gilboa, E., and Anderson, W. F. (1985) Science 2 3 0 , 1395-1398 7. Dick, J. E., Magli, M. C., Huszar, D., Phillips, R. A., and Bernstein, A. (1985) Cell 42, 71-79 8. McIvor, R. S., Johnson, M. J., Miller, A.D., Pitts, S., Williams, S. R., Valerio, D., Martin, D. W., and Verma, I. M. (1987) Mol. Cell Bid. 7, 838-846 9. Lim, B., Williams, D. A., and Orkin, S. H. (1987) Mol. Cell. Biol. 7, 3459-3465 10. Sorge, J., Kuhl, W., West, C., and Beutler, E. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,906-909 11. Palmer, T. D., Hock, R. A., Osborne, W. R. A., and Miller, A. D. (1987) Proc. Natl. Acad. Sci. U. S. A . 8 4 , 1055-1059 12. Garver, R. I., Chytil, A., Courtney, M., and Crystal, R. G. (1987) Science 2 3 7 , 762-764 13. St. Louis, D., and Verma, I. M. (1988) Proc. Natl. Acad. Sci. U. S. A. 85,3150-3154 14. Morgan, J. R.,Barrandon, Y., Green, H., and Mulligan, R.L. (1987) Science 2 3 7 , 1476-1479 15. Selden, R. F., Skoskiewicz, M. J., Howie, K. B., Russell, P. S., and Goodman, H. M. (1987) Science 236,714-718 16. McGrane, M.M., devente, J., Yun, J., Bloom, J.,Park, E., Wynshaw-Boris, A., Wagner, T., Rottman, F. M., and Hanson, R. W. (1988) J. Biol. Chem. 263,11443-11451 17. Wynshaw-Boris, A., Lugo, T. G., Short, J. M., Fournier, R. E. K., and Hanson, R. W. (1984) J. Biol. Chem. 259, 12161-12169 18. Wynshaw-Boris, A., Short, J. M., Loose, D. S., and Hanson, R. W. (1986) J.Biol. Chem. 261,9714-9720 19. Short, J. M., Wynshaw-Boris, A., Short, H. P., and Hanson, R. W. (1986) J. Biol. Chem. 261,9721-9726 20. Magnuson, M.A., Quinn, P. G., and Granner, D. K. (1987) J. Biol. Chem. 262,14917-14920 21. Peterson, D. D., Magnuson, M.A., and Granner, D. K. (1988) Mol. Cell. Biol. 8 , 96-104 22. Killary, A. M.,Lugo, T. G., and Fournier, R. E. K. (1984) Biochem. Genet. 22, 201-213 23. Darlington, G. J., Papaconstanatinou, J., Sammons, D. W., Brown, C.C., Wong, E. Y., Esterman, A. L., and Kang, J. (1982) Somatic Cell Genet. 8 , 451-464 24. Korman, A. J., Frantz, J. D., Strominger, J. L., and Mulligan, R. C. (1987) Proc. Natl. Acad. Sci. U. S. A. 8 4 , 2150-2154 25. Scherer, G., Schmid, W., Strange, C. M., Rowekamp, W., and Schutz, G. (1982) Proc. Natl. Acad. Sci. U. S. A. 79,7205-7208 26. Woychik, R. P., Camper, S. A., Lyons, R.H., Horowitz, S., Goodwin, E. C., and Rottman, F. M. (1982) Nucleic Acids Res. 10,7197-7210 27. Hughes, S. H., Greenhouse, J. J., Petropoulos, C. J., andSutraue, P. (1987) J. Virol. 6 1 , 3004-3012 28. Corsaro, C. M., and Pearson, M. L. (1981) Somatic Cell Genet. 7, 603-616 29. Cepko, C. L., Roberts, B. E., and Mulligan, R. C. (1984) Cell 3 7 , 1053-1062 30. Yoo-Warren, H., Monahan, J. E., Short, J., Short, H., Bruzel, A.,

31. 32. 33. 34.

Wynshaw-Boris, A., Meisner, H. M., Samols, S. D., and Hanson, R. W. (1983) Proc. Natl. Acad. Sci. U. s. A. 80,3656-3660 Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning, A Laboratory Manual, Cold Spring HarborLaboratory, Cold Spring Harbor, NY Davis, L. G., Dibner, M. D., and Battey, J. F. (1986) Basic Methods i n Molecular Biology, Elsevier, New York Miksicek, R., Heber, A., Schmid, W., Danesch,.U., Posseckert, G., Beato, M., and Schutz, G. (1986) Cell 4 6 , 283-290 Emerman, M., and Temin, H. M. (1986) Mol. Cell. Biol. 6 , 792800

35. Emerman, M., and Temin, H. M. (1986) Nucleic Acids Res. 14, 9381-9396 36. Armentano, D., Yu, S-F., Kantoff, P. W., von Ruden, T., Anderson, W. F., and Gilboa, E. (1987) J. Virol. 6 1 , 1647-1650 37. Cone, R. D., Weber-Benarous, A., Baorto, D., and Mulligan, R. C. (1987) Mol. Cell. Biol. 7 , 887-897 38. Miller, A.D., Ong, E. S., Rosenfeld, M.G., Verma, I. M., and Evans, R. M. (1984) Science 225,993-998 39. Yu, S-F., von Ruden, T., Kantoff, P. W., Garber, C., Seiberg, M., Ruther, U., Anderson, W. F., Wagner, E. F., and Gilboa, E. (1986) Proc. Natl. Acad. Sei. U. S. A . 83, 3194-3198 40. Barklis, E., Mulligan, R.C., and Jaenisch, R. (1986) Cell 4 7 , 391-399 41. Vijaya, S., Steffen, D. L., and Robinson, H. L. (1986) J. Virol. 60,683-692 42. Rohdewohld, H., Weiher, H., Reik, W., Jaensich, R., and Breindl, M. (1987) J. Virol. 61,336-343 43. Sorge, J., Cutting, A. E., Erdman, V.D., and Gautsch, J. W. (1984) Proc. Natl. Acad. Sci. U. S. A. 81,6627-6631 44. Rasmussen, J. A., and Gilboa, E. (1987) J. Virol. 6 1 , 1368-1374 45. Shimotohno, K., and Temin, H. M. (1980) Proc. Natl. Acad. Sci. U. S. A. 77,7357-7361 46. Battula, N., and Temin, H. (1977) Proc. Natl. Acad. Sci. U. S. A. 74,281-285 47. Wilson, A. F., and Cohen, J. C. (1988) Genomics, in press 48. Shih, C-C., Stoye, J. P., and Coffin, J. M. (1988) Cell 6 3 , 531537 49. Ledley, F. D., Grenett, H. E., McGinnis-Shelnutt, M., and Woo, S. L. C. (1986) Proc. Natl. Acad. Sci. U. S. A. 8 3 , 409-413 50. Wolfe, J. A., Yee, J-K., Skelly, H. F., Moores, J. C., Respess, J. G., Friedmann, T., and Leffert, H. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,3344-3348 51. Ledley, F. D., Darlington, G. J., Hahn, T., and Woo, S. L.C. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,5335-5339 52. Wilson, J. M., Jefferson, D. M., Chowdhury, J. R., Novikoff, P. M., Johnson, D. E., and Mulligan, R.C. (1988) Proc. Natl. Acad. Sci. U. S. A. 85,3014-3018 53. Demetriou, A. A., Whiting, J. F., Feldman, D., Levenson, S. M., Chowdhury, N. R., Moscioni, A. D., Kram, M., and Chowdhury, J. R. (1986) Science 2 3 3 , 1190-1192 54. Hatzoglou, M., Sekeris, C. E., and Hanson, R. W. (1985) Proc. Natl. Acad. Sci. U. S. A. 82,4346-4350