Glucocorticoid Regulation of Mouse Mammary Tumor Virus Sequences in Transgenic Mice (Gene Regulation/DNA Transformation) SUSAN R

Proc. Natl. Acad. Sci. USA Vol. 82, pp. 5880-5884, September 1985 Genetics

Glucocorticoid regulation of mouse mammary tumor virus sequences in transgenic mice (gene regulation/DNA transformation) SUSAN R. Ross* AND DAVOR SOLTER The Wistar Institute of Anatomy and Biology, 36th Street at Spruce, Philadelphia, PA 19104 Communicated by Salome G. Waelsch, May 17, 1985

ABSTRACT We have introduced a chimeric plasmid, fected with this construct are grown in the presence of pLTR2TK, containing the mouse mammary tumor virus dexamethasone (this report; refs. 17 and 18). We find that the (MTV) long terminal repeat (LTR) linked to the herpes simplex TK gene is expressed in the lactating mammary gland as well virus type 1 thymidine kinase gene into the mouse germ line by as in the ovaries of females of one line of mice. In addition, micro ijection. In one mouse line, the thymidine kinase gene is glucocorticoid-inducible transcription of the TK gene is appropriately expressed in the lactating mammary glands of detected in the testes of two independently derived males heterozygous females; expression also occurs in the ovaries of containing pLTR2TK in their germ line. these mice. In heterozygous males of this line, and in a male derived from another microiijection, transcription of these new germ line genes occurs in the testes and is specifically MATERIALS AND METHODS stimulated by glucocorticoid hormones. Thus, the MTV LTR Construction of pLTR2TK Plasmid. The pLTR2 plasmid is may contain elements that allow its expression both in lactating subcloned from a permuted plasmid, pMTV2, which is cloned mammary glands and in testicular tissue. from viral DNA from an MTV-infected rat hepatoma cell line (4). For construction ofthe pLTR2TK plasmid see Fig. 1. The Transcription of mouse mammary tumor virus (MTV), an pLTR2 plasmid was cut with EcoRI and combined with an endogenous murine retrovirus, can be induced by glucocor- EcoRI fragment [excised from a pFG5 (22) plasmid] contain- ticoid hormones in infected tissue culture cells or in explants ing the entire HSV-1 structural gene for TK and the TATA of mammary carcinomas (1, 2). This stimulation of expres- box (23). Transfection of the TK gene EcoRl fragment alone sion is probably the result of direct binding of the glucocor- into TK- cells results in a greatly decreased transfection ticoid receptor to sequences within the MTV genome (3, 4). frequency (23); ligation ofthe MTV sequences upstream from Although several cell types can be infected in vitro with MTV this fragment restores this frequency to that of the entire TK and can express these sequences, and although there are gene (unpublished data). active and functional glucocorticoid receptors in most mouse Pronuclear Injections and DNA Transfections. Supercoiled tissues, transcription of proviral sequences in vivo occurs or HindIII-linearized pLTR2TK was microinjected into one predominantly in lactating mammary glands and in mammary pronucleus of fertilized one-cell eggs recovered from the tumors ofthe mouse (5-7). Thus, it appears that although the oviducts of either CD-1 (Charles River Breeding Laborato- sequences necessary for MTV transcription in tissue culture ries) or C57BL/6 (The Jackson Laboratory) females (see cells are encoded within the provirus, the regulation of this Results) that had mated with males of the same strain the expression in the animal may be determined by sequences previous night. Approximately 5 pl of DNA or 1000-5000 that control tissue specificity. copies of the plasmid were microinjected. After overnight Recently, enhancer sequences, which seem to determine cultivation to the two-cell stage in Whitten's medium (24), the preferential transcription in specific cell types, were detected eggs were implanted into pseudopregnant CD-1 females. in introns (8-10) or in 5' flanking regions (11). We were Plasmid pLTR2TK was transfected into LMTK- mouse interested in whether the tissue-specific expression of MTV, cells (25) by calcium phosphate precipitation. Twenty-four as well as the functional response of the glucocorticoid hours after the addition ofthe DNA precipitate, the cells were receptor binding sites, was due to MTV-encoded sequences selected in HAT medium (0.5 mM hypoxanthine/0.4 ,M or to the chromosomal position of the endogenous provirus. aminopterin/16 ,M thymidine), TK-positive colonies were Microinjection of several different cloned genes into mouse selected (-2-3 weeks after selection) and grown. Cultures zygotes has resulted in mice that show the same tissue derived from single colonies were grown both in the presence specificity of expression as the endogenous copies of these and absence of 1 ,uM dexamethasone for 20-24 hr, and RNA genes (12-16). was isolated from these. As a first step to dissecting any MTV sequences involved Isolation of RNA and DNA. Mice were anesthetized by in the mammary gland specific expression, we microinjected intraperitoneal injection of Nembutal, and a lobe of the liver into murine embryos a chimeric plasmid that contains the was removed through a small ventral incision. After homog- MTV long terminal repeat (LTR) ligated to the herpes virus 1 kinase enization in phosphate-buffered saline in a tissue grinder simplex type (HSV-1) thymidine (TK) gene fitted with a Teflon pestle, the tissue was pelleted at 1000 rpm with the "TATA" box and transcription initiation site and in an International centrifuge and lysed in 10 mM Tris, pH derived mice that contain germ-line copies of this plasmid. mM mM EDTA The ligation ofthese MTV sequences to the TK gene confers 7.5/100 NaCl/0.5% NaDodSO4/1 containing increased transcription over a basal level when cells trans- Abbreviations: MTV, mouse mammary tumor virus; LTR, long terminal repeat; TK, thymidine kinase; HSV-1, herpes simplex virus The publication costs of this article were defrayed in part by page charge type 1; kb, kilobase(s); GRE, glucocorticoid regulatory element(s). payment. This article must therefore be hereby marked "advertisement" *Present address: Department ofBiological Chemistry, University of in accordance with 18 U.S.C. §1734 solely to indicate this fact. Illinois, College of Medicine, Chicago, IL 60612. 5880 Downloaded by guest on October 3, 2021 Genetics: Ross and Solter Proc. Natl. Acad. Sci. USA 82 (1985) 5881 r R M ,1 1 1 RESULTS -, MWI Derivation of Transgenic Mice Containing pLTR2TK Se- quences. Supercoiled or HindIII-linearized pLTR2TK plas- / mid (1,000-5,000 molecules per embryo) was microinjected R/H R,' R R into the male pronucleus of one-cell fertilized embryos pBR322 HSV-TK isolated from either CD-1 random-bred Swiss or C57BL/6 pLTR2 females. Approximately 50-60o of the one-cell embryos injected with this molecule developed to the two-cell stage (183 of 317 one-cell embryos injected with supercoiled pLTR2TK and 76 of 151 injected with the linearized plasmid). R Ofthe two-cell embryos transferred, 5-7% developed to term (13 and 4 from the supercoiled and linearized injections, 7.8kb respectively). Approximately 4 weeks after birth, DNA was isolated from liver lobes, obtained by partial hepatectomy and analyzed for retention of the plasmid by dot blots (27). Two ofthe mice injected with supercoiled pLTR2TK (mice R/H LTK 12 and LTK 3) and one injected with the HindIII- FIG. 1. Map of pLTR2TK. Plasmid pLTR2 is shown relative to linearized molecule (mouse LTK 15) retained the injected the entire MTV genome: the double asterisks represent the sites of sequences (Fig. 2A). LTK 12, afemale C57BL/6, died shortly binding to glucocorticoid receptor in the LTR (19-21); the arrow- thereafter; LTK 15, a male C57BL/6, died at age 8 weeks head, the start of MTV transcription; and the directional arrows without siring any offspring; however, tissues from this above pLTR2TK, the direction of transcription. EcoRI digestion of animal were analyzed for pLTR2TK expression (see below). the plasmid generates a 2.3-kb fragment, which hybridizes to TK LTK 3 was mated with a normal CD-1 male; of the seven gene, and a 5.5-kb fragment, which hybridizes to MTV and pBR322. offspring born, four retained the microinjected sequences. Restriction sites ofEcoRI, Pst I and HindIII are designated R, P, and These offspring were also mated to normal mice and -50% H, respectively. Both supercoiled and HindIII-linearized plasmids were used for microinjection. of the offspring in each generation inherited the pLTR2TK DNA. From this transmission ratio, we assume that integration into one chromosome probably occurred. The hetero- proteinase K at 0.4 mg/ml. After incubation for 3-10 hr at zygotes generated from these crosses have also been mated 50°C, the samples were extracted with phenol and chloroform with each other and appeared to produce homozygous and dialyzed extensively against 10 mM Tris, pH 7.5/1 mM offspring, judging by the intensity of hybridization to dot EDTA. The DNA was used at this point for restriction blots of DNA isolated from these offspring (not shown). digestion and Southern blot analysis (26). For dot blots, the DNA Analysis. The DNA from these three mice and the DNA was treated with 0.3 M NaOH for 5 min at 95°C, offspring of LTK 3 were characterized by Southern blot neutralized, brought to 1.5 M NaCl on ice, and spotted on analysis. EcoRI digestion of the parental plasmid generated nitrocellulose as described by Thomas (27). RNA was iso- two fragments of 2.3 and 5.5 kb (Fig. 1). Hybridization ofthe lated from fresh tissue by the method of Chirgwin et al. (28). blots containing EcoRI-digested liver DNA from mouse LTK For dot blots, 5 ,tg were spotted onto nitrocellulose. For 3 and her offspring with three different probes revealed, as blot-hybridization analysis, 20 ,tg of RNA was separated on 1% formaldehyde gels and transferred to nitrocellulose (29). Mouse ribosomal RNAs from ethidium bromide-stained gels MTV TK served as markers. A B C D E F G A B C D E F G Hybridization and Probes. After transfer of DNA or RNA id I U to nitrocellulose, the filters were hybridized for 24-36 hr at A 420C with 32P-labeled probes (30) in 50o formamide containing 3 x NaCl/Cit (lx NaCl/Cit = 0.15 M NaCl/0.015 M Na * -8.0 2 citrate, pH 7), 5 x Denhardt's solution (lx = 0.02% poly- - 5.5 vinylpyrrolidone/0.02% Ficoll/0.02% bovine serum albu- 3 min), and salmon sperm DNA at 50 ,ug/ml. The filters were washed briefly at room temperature in 0.1% NaCl/Cit con- 5 taining 0.1% NaDodSO4 and then two times for 30 min at 50°C 6 . 0 and were exposed to Kodak XAR film with DuPont Cronex 7 * 0 add*-2.3 Lightning-Plus intensifying screens. All probes were isolated 2jig 4 .,g after restriction enzyme digestion as a specific band from low-melting agarose and were labeled by nick-translation (30). The following probes were used in this work: (i) pBR322; (ii) a 2.1-kilobase (kb) Pvu II fragment from the pTK plasmid that contains HSV-1 TK gene (22); (iii) pPCK2, a 600-base-pair cDNA clone of the rat FIG. 2. DNA analysis of transgenic mice. (A) Detection of phosphoenolpyruvate microinjected sequences in mice. The dot blots were hybridized with carboxykinase gene, provided by R. Hanson (31); (iv) Pst I probe made to the TK gene EcoRI fragment (see Fig. 1). Lanes: 1, fragment from pLTR1 plasmid, which contains MTV LTR normal CD-1 liver DNA; 2, 3, and 5, liver DNA from normal (19) and was provided by J. Majors; (v) a 1.3-kb fragment littermates; 4, 6, and 7, liver DNA from mouse LTK 3, 15, and 12, containing the coding sequences, isolated from m1pEE3.8 respectively. Each dot blot contains 2 and 4 g.g ofDNA. (B) Southern plasmid by digestion with Bgl II and HindIII [mlpEE3.8 is a blot of DNA from LTK 3 and six offspring. Lanes: A-C, DNA from genomic clone of mouse metallothionein I and was three normal littermates; D-F, DNA from LTK 3 offspring contain- gene (32) ing the microinjected sequence; G, LTK 3 DNA. MTV and TK provided by R. Palmiter]; and (vi) a 1.8-kb Pst I fragment indicate 32P-labeled nick-translated pLTRI and TK gene hybridiza- from pMTV4 plasmid, which contains the MTV env gene (19) tion probes, respectively. Each lane contains 10 ,ug of DNA. Sizes and was provided by J. Majors. are shown in kb. Downloaded by guest on October 3, 2021 5882 Genetics: Ross and Solter Proc. Natl. Acad. Sci. USA 82 (1985) expected, a band of 2.3 kb hybridizing only to the TK gene corresponding to a transcript initiating within the MTV LTR probe (Fig. 2B) and a 5.5-kb band corresponding to the MTV and reading through the TK gene sequences also could be LTR and pBR322 sequences after hybridization of the DNA detected (Fig. 3A), as has been shown for other plasmids of with either pLTR1 (Fig. 2B) or pBR322 probes (results not similar construction (17, 18, 33, 34). To determine if tissue- shown). This blot also shows the endogenous MTV bands specific expression ofthe microinjected pLTR2TK sequences that hybridize to the DNA from all the mice. Since these mice occurred in the transgenic mice, a lactating female offspring are random-bred, these bands vary between individual mice. of mouse LTK 3, age -2 months, was sacrificed 1 week after In addition, two bands, one of 8 kb and the other of >12 kb, giving birth to her first litter, and RNA was extracted from hybridize to all three probes. These bands are probably the twelve tissues. To detect transcription from the injected result of a recombination event between the injected super- sequences, total RNA (5 ,g) from LTK 3 and an ICR mouse coiled plasmids such that one of the two EcoRI sites is (not shown) were spotted onto nitrocellulose, and the filters deleted; the bands are not seen in Southern blots of DNA were hybridized with nick-translated probe made against the from mouse LTK 12 or LTK 15 (not shown). The intensity of TK gene insert. Of 12 different tissues examined (liver, hybridization ofthe 8-kb band is similar to that ofthe 2.3- and spleen, kidney, brain, thymus, lung, heart, thyroid, skeletal 5.5-kb bands, which we estimate contain 10-20 copies of the muscle, uterus, ovary/oviduct, and mammary gland), only plasmid by comparison with the intensity of hybridization of the lactating mammary gland and ovary/oviduct from this the endogenous MTV bands, while the larger band contains mouse appeared to synthesize significant amounts of RNA fewer hybridizing copies. Therefore, it is unlikely, at least, that hybridized to the TK gene probe (Fig. 3B) (analysis of 5 that the 8-kb band represents the terminal copies of the tissues is shown). When this dot blot was melted and tandemly integrated plasmids linked to host flanking se- hybridized with probe made to an internal fragment of MTV quences; however, the band of>12 kb may contain a terminal DNA that lacks LTR sequences (pMTV4), only the lactating fragment. mammary gland synthesized appreciable amounts of MTV- From the intensity of hybridization of the LTR and TK specific sequences. Blot-hybridization analysis of mammary gene probes to Southern blots ofDNA from LTK 12 and LTK gland RNA, in comparison with RNA from a pLTR2TK- 15 (not shown), we estimate that LTK 12 has -5-10 copies transfected cell line, revealed an RNA of 1.3 kb that hybrid- and LTK 15 has 30-50 copies of the microinjected pLTR2TK ized to the TK gene probe (Fig. 3A). Preliminary analysis plasmid. Southern blot analysis of HindIll-digested DNA indicates that less transcription ofthe newly introduced DNA from all three transgenic mice (HindIII cuts once within the by at least a factor of 10 occurred in either mammary glands plasmid; Fig. 1) indicates that the majority of the copies are or ovaries of nonlactating females and that adrenalectomy identical and are arranged tandemly within the genome (one further decreased this transcription (data not shown). We major band of 7.8 kb hybridizes to the TK gene, LTR, and have not yet adrenalectomized lactating females and exam- pBR322 probes; data not shown). Thus, the MTV glucocor- ined the RNA in their mammary glands. ticoid regulatory sequences are located 5' to the TK gene Mouse LTK 15 and several male offspring of LTK 3 were transcription site in most of the integrated copies of analyzed also for expression of TK gene sequences. RNA pLTR2TK. was isolated from the spleen, liver, heart, skeletal muscle, Expression of the MicroinJected Sequences. After introduc- lymph nodes, epididymis, kidney, lung, and testes of these tion of pLTR2TK into mouse LMTK- cells by calcium animals and dot-blotted along with RNA from the tissues of was and a control CD-1 mouse (not shown). Transcription levels were phosphate precipitation, the TK gene transcribed, high in the testes ofboth ofthese independently derived mice this transcription was induced by dexamethasone (Fig. 3A). (Fig. 4A) (about 10% of the level of tubulin expression in this The major TK gene transcript produced in these cells, both tissue; data not shown); a low level of transcription (less than in the presence and absence of dexamethasone, corresponds that seen in testes) also was detected occasionally in to the TK transcript of 1.3 kb, although a minor band 1/10th 1 2 3 A - + L U M B A B MTV TK 15 5.0 1g * .A L 2.5vg 4

0 l ,.- 3 5.0 m S 0 2.5 pg S L H MU T 2.2-

1.3

FIG. 4. (A) Analysis of RNA from the tissues of mouse LTK 15 (A Upper) and the offspring of mouse LTK 3 (A Lower). Total RNA (2.5 and 5 ,g) from spleen (S), liver (L), heart (H), skeletal muscle FIG. 3. Analysis of RNA from the tissues of a female offspring of (MU), and testes (T) were analyzed by dot-blot hybridization. The mouse LTK 3. (A) Blot-hybridization analysis of the RNA, after spot on the 5 Ig spleen sample from mouse LTK 15 is a procedural hybridization with the TK gene probe. Lanes: + and -, RNA from artifact and does not line up with the RNA spot on the nitrocellulose an LMTK- tissue cell line transfected with pLTR2TK and grown in paper. (B) RNA from testes of an adrenalectomized (lane 1), an the presence (+) and absence (-) of 1 ,uM dexamethasone, respec- untreated (lane 2), and an adrenalectomized and dexamethasone- tively; L, RNA from the liver; U, RNA from uterus; and M, RNA injected (lane 3) mouse in G2 progeny of mouse LTK 3. Two male from mammary gland of a nursing female offspring of LTK 3. (B) heterozygous littermates (second generation) were adrenalecto- RNA from thymus (T), liver (L), uterus (U), ovary and oviduct (0) mized, and 5 days later one of the mice was injected subcutaneously and lactating mammary gland (M) was analyzed by dot blots. TK and with 100 ,g of dexamethasone. Five hours after this injection, these MTV denote the hybridization probes. The MTV probe was a cloned two mice and a nonadrenalectomized littermate were sacrificed, and fragment of the MTV envelope gene, pMTV4. RNA was isolated from their tissues. Downloaded by guest on October 3, 2021 Genetics: Ross and Solter Proc. Natl. Acad. Sci. USA 82 (1985) 5883

the spleens and/or livers of the LTK 3 offspring. The ectomy causes the basal level of TK gene transcription in the transcript produced in these animals had a molecular size of testes to decrease. 1.3 kb, corresponding to the TK-sized transcript (Fig. 4B). Transcription levels in the testes of an adrenalectomized DISCUSSION LTK 3 offspring (second generation), after injection with 100 ,ug ofdexamethasone, were 5- to 10-fold greater than those in In these studies, we have introduced a chimeric molecule an untreated mouse (Fig. 4B), although the basal level of TK containing the MTV LTR linked to the HSV-1 TK gene into gene expression in untreated mice varied from animal to the mouse germ line by microinjection to determine whether animal (not shown), probably due to different levels of any sequences encoded by MTV are responsible for its endogenous glucocorticoids between animals. A male tissue-specific expression. Similar studies have been carried heterozygous littermate, adrenalectomized to control for the out in tissue culture cells to define the sequences responsible effect of endogenous glucocorticoids on transcription, for the glucocorticoid induction ofMTV transcription (17, 18, showed very low levels of transcription (about 1/20th of the 33-36) and have shown that the sequences required for level of the dexamethasone-treated mouse), possibly due to hormone induction, termed the glucocorticoid regulatory low levels of circulating hormone still present in the mouse. element(s) (GRE) (33) can be dissected from the MTV No transcription of endogenous MTV in the testes of any promoter. This element(s) can act on both the MTV promoter mice (untreated or hormone-injected) was detected (data not (34-36) and on heterologous promoters (17, 18, 33). We have shown). shown that the GRE is also capable of functioning in In order to eliminate the possibility that the differences in transgenic mice; this report demonstrates hormone regula- TK gene transcription between hormone-treated and adre- tion of an injected gene. It is possible that this hormone nalectomized mice were due to random variations in the level regulation is due to the site of integration; however, if this is of expression in individual animals, we examined the effect so, then the integration site must also determine the tissue(s) ofendogenous hormones on the levels ofTK RNA in a single in which this regulation occurs, since we see little if any mouse. One testis was removed from a male G2 offspring of expression of pLTR2TK in tissues other than lactating LTK 3; the animal was adrenalectomized and sacrificed 7 mammary gland, ovary, and testis even in hormone-treated days later. RNA was isolated from the testis that had been mice. We also see transcription of this plasmid in the testes removed before adrenalectomy and from the remaining testis of two independently derived mice, and this leads us to and the liver of the animal after sacrifice. An untreated male conclude that the regulation of transcription is controlled by littermate, positive for pLTR2TK sequences, also was sac- the MTV sequences. rificed, and RNA was prepared from the testes and liver of The MTV LTR appears to contain sequences in addition to this animal. As seen in Fig. SA, the level of TK gene the GRE that direct the tissue-specific transcription of the transcription in the testis after adrenalectomy was decreased virus in vivo, since we now show that a transgenic mouse 5- to that seen either in the same animal containing the MTV LTR linked to the HSV-1 TK gene to 10-fold compared lacking most of its promoter elements but retaining its prior to adrenalectomy or in the untreated littermate. A transcription initiation site transcribes these sequences in the concomitant decrease in the RNA levels of two liver proteins appropriate tissue-the lactating mammary gland. A recent whose transcription also is induced by glucocorticoids, report by Stewart et al. (37) has shown that transgenic mice phosphoenolpyruvate carboxykinase (31) and metallothio- microinjected with molecules containing the MTV LTR with nein (32), was seen in the adrenalectomized animal relative to both the GRE and promoter elements linked to the murine his untreated sibling (Fig. 5B). Thus, a decrease in the levels c-myc oncogene, develop mammary tumors after multiple of glucocorticoids in a single animal as the result of adrenal- pregnancies, although transcription of the microinjected DNA is seen in several tissues. Our assay for expression, blot A hybridization to total RNA from tissues, is probably less 5.0 sensitive than the S1 nuclease protection assay used by pg Stewart et al. (37). However, transcription in other tissues 2.5 pg would have to be at levels <30-fold that seen in the mammary gland or testes to escape our detection. Such low-level 66 48 48a transcription might be the result of the integration site of the injected DNA, such as that seen in transgenic mice contain- B ing the rabbit f3-globin gene (38). 5.0Kg * Although our experiments and those of others have shown that the predominant site of MTV transcription is in lactating 2.5P mammary gland, we find that MTV also can be expressed in the testes. Previous reports have identified MTV antigens 5.0/Lg * (39) and RNA sequences (5) in a murine Leydig cell tumor. Endogenous MTV sequences may not be expressed in the 2.5pg * testes because of position effects; that is, they may be in a 66 48a chromosomal location not able to become transcriptionally active during testicular development. By moving the se- FIG. 5. Analysis of RNA from the testes and liver of LTK 3 male quences to a new chromosomal position using microinjec- offspring. (A) Total RNA (5 and 2.5 jig) from the testes of an tion, we may have identified another tissue in addition to the untreated (column 66) G2 progeny of LTK 3 and from an adrenal- lactating mammary gland in which MTV sequences act as ectomized littermate both before (column 48) and after (column 48a) transcriptional enhancers. It is possible that sequences en- adrenalectomy was analyzed by dot-blot hybridization to TK gene coded in the TK genome determine this testis-specific tran- probe. (B) Total RNA (5 and 2.5 jig) from the livers of the untreated of TK RNA; however, we feel that this is unlikely (column 66) and adrenalectomized (column 48a) animals were scription analyzed by dot-blot hybridization with pPCK2 (B Lower)- and for several reasons: (i) the portion of the TK gene used m1PEE3*8 (B Upper)-derived probes. One testis was removed from a contains only the TATA box and has been shown to have male G2 progeny of LTK 3 prior to adrenalectomy. Seven days after greatly reduced activity in both in vivo and in vitro transcrip- the adrenalectomy, this animal was sacrificed, along with an un- tion assays (23); (ii) transcription of constructs containing treated littermate, and RNA was isolated from their tissues. similar amounts of the TK gene 5' flanking sequences can be Downloaded by guest on October 3, 2021 5884 Genetics: Ross and Solter Proc. Nati. Acad. Sci. USA 82 (1985)

regulated in tissue culture assays when MTV sequences (17, 12. Palmiter, R. D., Chen, H. Y. & Brinster, R. L. (1982) Cell 29, 18) or the HSV-1 a control region (40) are placed upstream; 701-710. (iii) transgenic mice containing TK gene sequences alone 13. Brinster, R. L., Ritchie, K. A., Hammer, R. E., O'Brien, have been constructed in which TK RNA was detected in R. L., Arp, B. & Storb, U. (1983) Nature (London) 306, 332-336. whole embryos; and it is unlikely that this transcription 14. Swift, G. H., Hammer, R. E., MacDonald, R. J. & Brinster, would have been seen if only testicular expression occurred R. L. (1984) Cell 38, 639-646. in these animals (41); (iv) transcription of pLTR2TK in our 15. Grosschedl, R., Weaver, D., Baltimore, D. & Costantini, F. transgenic mice is controlled by glucocorticoids and, thus, (1984) Cell 38, 647-658. remains under the hormonal control of the LTR. The testes- 16. Harbers, K., Jahner, D. & Jaenisch, R. (1981) Nature (Lon- specific transcription was seen in two males derived from don) 293, 540-542. independent injections; therefore, it is unlikely that this 17. Groner, B., Kennedy, N., Rahmsdorf, V., Herrlich, P., Van tissue-specific expression is the result of integration into a Ooyen, A. B. & Hynes, N. E. (1982) eds. Dumont, J. E., particular of the host chromosome concerned with Nunez, J. & Schutz, G. (Elsevier/North-Holland, Amster- region dam), Vol. 6, pp. 217-228. testes-specific gene expression. 18. Hynes, N., Van Ooyen, A. J. J., Kennedy, N., Herrlich, P., Construction of additional transgenic mice containing Ponta, H. & Groner, B. (1983) Proc. Natl. Acad. Sci. USA 80, pLTR2TK DNA and plasmids with deletions in the MTV 3637-3641. LTR should reveal which sequences are involved in the 19. Payvar, F., Firestone, G. L., Ross, S. R., Chandler, V. L., mammary gland- and testes-specific transcription and wheth- Wrange, O., Carlstedt-Duke, J., Gustafsson, J. A. & er these MTV enhancers map to the same sequences. Further Yamamoto, K. R. (1982) J. Cell. Biochem. 19, 241-247. testing should also reveal whether the transcription in the 20. Geisse, S., Scheidereit, C., Westphal, H. M., Hynes, N. E., ovary/oviduct is specific to the LTK 3 mouse or generalizes Groner, B. & Beato, M. (1982) EMBO J. 1, 1613-1619. to other derived mice these se- 21. Pfahl, M. (1982) Cell 31, 475-482. independently containing 22. Colbere-Garapin, F., Chousterman, S., Horodniceau, F., quences and if hormones other than glucocorticoids are Kourilsky, P. & Garapin, A.-C. (1979) Proc. Natl. Acad. Sci. involved in this expression. By identifying sequences within USA 76, 3755-3759. MTV itself that are responsible for its tissue-specific expres- 23. McKnight, S. L., Gavis, E. R., Kingsbury, R. & Axel, R. sion, we can begin to understand why it is that an endogenous (1981) Cell 25, 385-398. retrovirus, which has the potential to be expressed in any 24. Whitten, W. K. (1971) Adv. Biosci. 6, 129-139. tissue in the mouse, does so only in one tissue, the lactating 25. Wigler, M., Sweet, R., Sim, G. K., Wold, B., Pellicer, A., mammary gland. Lacy, E., Maniatis, T., Silverstein, S. & Axel, R. (1979) Cell 16, 777-785. The technical assistance and thoughtful comments of Dr. Barbara 26. Southern, E. M. (1975) J. Mol. Biol. 98, 503-517. B. Knowles are gratefully acknowledged. This work was supported 27. Thomas, P. S. (1980) Proc. Natl. Acad. Sci. USA 77, by Grant 83-55 from the American Cancer Society, Illinois Division, 5201-5205. and by grants from the National Institutes of Health (Biomedical 28. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J. & Rutter, Research Support Grant 48314 and HD 19444 to S.R.R. and HD W. J. (1979) Biochemistry 18, 5294-5299. 12487 and HD 17720 to D.S.). S.R.R. was the recipient of a Special 29. Lehrach, H., Diamond, D., Wozney, J. M. & Boedtker, H. Fellowship from the Leukemia Society of America. (1977) Biochemistry 16, 4743-4751. 30. Rigby, P. W. J., Dieckmann, M., Rhodes, C. & Berg, P. (1977) 1. Parks, W. P., Scolnick, E. M. & Kizikowski, E. H. (1974) J. Mol. Biol. 113, 237-251. Science 184, 158-161. 31. Lamers, W. H., Hanson, R. W. & Meisner, H. M. (1982) 2. Lasfargues, E. Y., Kramarksy, B., Lasfargues, J. C. & Proc. Natl. Acad. Sci. USA 79, 5137-5141. Moore, D. H. (1976) Cancer Res. 36, 67-72. 32. Durnam, D. M., Perrin, F., Gannon, F. & Palmiter, R. D. 3. Schedereit, C., Geisse, S., Westphal, H. M. & Beato, M. (1980) Proc. Natl. Acad. Sci. USA 77, 6511-6515. (1983) Nature (London) 304, 249-252. 33. Chandler, V. L., Maler, B. A. & Yamamoto, K. R. (1983) Cell 4. Payvar, F., DeFranco, D., Fireston, G. L., Edgar, B., 33, 489-499. Wrange, O., Okret, S., Gustafsson, J.-A. & Yamamoto, K. R. 34. Majors, J. & Varmus, H. E. (1983) Proc. Natl. Acad. Sci. USA (1983) Cell 35, 381-392. 80, 5866-5870. 5. Varmus, H. E., Quintrell, N., Medeiros, E., Bishop, R. C., 35. Lee, R., Mulligan, R., Berg, P. & Ringold, G. (1981) Nature C. & N. H. J. Mol. (London) 294, 228-232. Nowinski, R. Sarkar, (1973) Biol. 79, 36. Huang, A. L., Ostrowski, M. C., Berard, D. & Hager, G. L. 663-679. (1981) Cell 27, 245-255. 6. Marcus, S. L., Smith, S. W. & Sarkar, N. H. (1981) J. Virol. 37. Stewart, T. A., Pattengalt, P. K. & Leder, P. (1984) Cell 38, 40, 87-95. 627-637. 7. Hu, W.-S., Fanning, T. G. & Cardiff, R. D. (1984) J. Virol. 49, 38. Lacy, E., Roberts, S., Evans, E. P., Burtenshow, M. D. & 66-71. Costantini, F. D. (1983) Cell 34, 343-358. 8. Gillies, S. D., Morrison, S. L., Oi, V. T. & Tonegawa, S. 39. Nowinski, R. C., Sarkar, N. H., Old, L. J., Moore, D. H., (1983) Cell 33, 717-728. Scheer, D. I. & Hilgers, J. (1971) Virology 46, 21-38. 9. Baneji, J., Olson, L. & Shaffner, W. (1983) Cell 33, 729-740. 40. Mackem, S. & Roizman, B. (1982) Proc. Natl. Acad. Sci. USA 10. Queen, C. & Baltimore, 0. (1983) Cell 33, 741-748. 79, 4917-4921. 11. Walker, M. D., Edlund, T., Boulet, A. M. & Rutter, W. J. 41. Wagner, E. F., Stewart, T. & Mintz, B. (1981) Proc. Natl. (1983) Nature (London) 306, 557-561. Acad. Sci. USA 78, 5016-5020. Downloaded by guest on October 3, 2021