Proc. Natl Acad. Sci. USA Vol. 80, pp. 477-481, January 1983 Genetics Isolation and characterization of a full-length expressible cDNA for human hypoxanthine phosphoribosyltransferase (Lesch-Nyhan disease /gene cloning /traasfection /DNA sequence analysis) D. J. Joty*, H. Oxayamat, P. Berct, A. C. Esty*, D. Firpuca*, P. BouLen?, G. G. Jounson§, J. E. Suivety!, T. HUNKAPILLARI, AND T. FRIEDMANN* *Department of Pediatrics, University of California at San Diego, La Jolla, California 92093; *Department of Biochemistry, Stanford University School of Medicine, Stanford, California $4305; #The Salk Institute, Post Office Box 85800, San Diego, California 92138, Department of Biology, San Diego State University, San Diego, California 92182; "Department of immunology, City of Hope Medical Research Center; Duarte, California 91010; and Division of Biology, California Institute of Technology, Pasadena, Californi 25 Communicated by J. Edwin Seegmiller, October 7, 1982 ABSTRACT We have cloned a full-length 1.6-kilobase cDNA of a human mRNA coding for hypoxanthine phosphoribosyltrans- ferase (HPRT; IMP:pyrophosphate phosphoribosyltransferase, EC 2.4.2.8) into a simian virus 40-based expression vector and have determined its full nucleotide sequence. The inferred amino acid sequence agrees with a partial amino acid sequence deter- mined for authentic human HPRT protein.. Transfection of HPRT- deficient mouse LA9 cells with the purified plasmid leads to the expression of human HPRT enzyme activity in cells stably trans- fected and selected for enzyme activity in hypoxanthine /ami- nopterin /thymidine medium. Current methods of molecular biology, including the -tech- niques of recombinant DNA construction and cloning, rapid nucleotide sequence analysis, the design and construction of transducing vectors, and techniques of transfection of eukaryot- ic cells with foreign genes, have made it possible to clone and characterize a large number of eukaryotic genes. One gene of particular interest, not only for basic studies of eukaryotic gene regulation but also for understanding of several important hu- man genetic diseases, is the gene encoding the enzyme hypo- xanthine phosphoribosyltransferase (HPRT) (1). This enzyme catalyzes vital steps in the reutilization pathway for purine bio- synthesis, and its deficiency leads to forms of gouty arthritis and to the devastating Lesch-Nyhan disease (2, 3). The HPRT locus is known ‘to be X linked in the human and other mammalian genomes, and the availability. of a cloned HPRT gene would facilitate studies of the organization of a particularly interesting region of the human X chromosome and of the mechanisms of inactivation of specific and well-mapped regions of the X chro- mosome. Recently, we succeeded in isolating a human genomic clone containing a portion of the HPRT gene together with some intervening sequence (4). This fragment of the HPRT gene has been used to isolate a full-length cDNA clone encoding the human HPRT-enzyme from a human cDNA library. Nucleotide sequence analysis of the cloned cDNA segment established that it encodes the entire HPRT protein. METHODS Isolation of the HPRT Gene Fragment. A genomic segment of human DNA containing sequences from the extreme 5’ end of the HPRT gene together with a portion of an intervening sequence has been isolated and cloned as plasmid p6B2aE2 (4) by a combination of gene transfection into enzyme-deficient mouse cells and localization of human sequences by hybridiza- The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertise- ment” in accordance with 18 U. S. C. §1734-solely to indicate this fact. 417 tion with probes of middle repetitive human (Alu) sequence, as described (4). A 1.5-kilobase (kb) EcoRI/BamHI fragment derived from p6B2aE2 and free of repetitive human sequences ‘has been subcloned into the BamHI and EcoRI sites of pBR322 and shown to hybridize to a discrete human cytoplasmic poly(A)"RNA approximately 1.5.kb long, presumably. repre- senting the mRNA for HPRT (4). DNA from this subcloned plasmid, called pBR1.5, was prepared by detergent lysis (5) of chloramphenicol-treated (6) transformed SF8 Escherichia coli (7) followed by cesium chloride/ethidium bromide ultracentri- fugation (8). This DNA was cleaved with EcoRI/ BamHI and the cloned insert was isolated by trough electroelution onto a di- alysis membrane (9) after agarose gel electrophoresis. It was labeled by nick-translation (10) with [*?P]dATP (3,000 Ci/ mmol; 1.Ci = 37 GBq; Amersham) in the presence of. unlabeled dCTP/dGTP/dTTP to a specific activity of approximately 2 X 10° cpm/pg and was thereafter.used in hybridization studies to screen a human cDNA library. Preparation and Screening of a Human cDNA Library. The preparation of the. complete human cDNA library and the sub- libraries ‘containing different size cDNA inserts is described in detail elsewhere (11). Plasmid DNA representative of the com- plete and sublibraries. was transformed. into actively growing cultures of E. coli SF8 (ODgo9 = 0:6) by the-method of Lerach . et al (12). Transformed colonies were selected by growth over- night at 37°C on Penassay base agar (Difco Laboratories) con- taining ampicillin. at 25-50 pg/ml: (Sigma) in 400-cm? Petri dishes (Nunc Bio-assay plates) at one dish per transformation. Approximately 300,000 transformants per library were grown and replica plated twice with velvet. E. coli SF8 containing the cloned HPRT fragment (p6B2aE2) was plated as a positive con- trol. The colonies on the replica plates were transferred to What- man 54]: paper, amplified on chloramphenicol plates overnight at'37°C by the method of Gergen et al. (13), denatured and fixed to the filters, and screened for HPRT cDNA sequences by hy- bridization with the gel-purified insert from pBR1.5 containing the HPRT fragment free of repeat sequences. The filter replicas of the random colony arrays from each library were prehybri- dized for 4 hr at 60°C: in 30% formamide/0.5 M NaCl/20 mM Hepes, pH 7.4/0.1 mM EDTA/10 mM NaH,PO,/10 mM Na,P,O, containing denatured and sonicated herring sperm DNA at 500 yzg/ml (buffer A) and then hybridized to the *°P- labeled nick-translated pBR1.5 insert fragment. overnight at 60°C in buffer A/10% dextran sulfate. The filter sheets were Abbreviations: HPRT, hypoxanthine’ phosphoribosyltransferase; kb, kilobase(s); HAT, hypoxanthine/aminopterin/thymidine; bp, base pair(s). 478 Genetics: Jolly et al then washed twice with 0.30 M NaCl/0.030 M Na citrate, pH 7.4, and then with 60 mM NaCI/6 mM Na citrate, pH 7.4, once at 60°C and twice at room temperature. Filters were dried and autoradiographed on flashed Kodak XAR-5 film with an inten- sifying screen at —70°C. Areas showing a putative positive signal on duplicate filters were identified, scraped with a toothpick onto 1 ml of Penassay broth, plated at various concentrations from the broth, and grown overnight. Single colonies were transferred with a tooth- pick into arrays, grown overnight, and rescreened on Whatman 541 paper as described above. DNA “minipreps” (14) were prepared from apparently pos- itive colonies derived from the one original positive area from the transformation with sublibrary 4 (11), which contains inserts of 1.5-2.0 kb of DNA. The cDNA inserts were screened with *P.labeled nick-translated purified insert DNA from pBR1.5 by blotting agarose gels containing whole DNA or BamHI di- gests onto nitrocellulose according to Southern (15), treatment with HCI (16), and hybridization as described above for the cDNA library, with the addition of 5-fold concentrated Den- hardt’s additives (17) to the prehybridization and hybridization buffers. All manipulations involving recombinant DNA were carried out in accordance with National Institutes of Health and insti- tutional guidelines and requirements. Transfection. of Enzyme-Deficient Cells. The isolated full- length cDNA plasmid was introduced into: enzyme-deficient mouse LAQ cells grown in Dulbecco's modified Eagle’s medium/ 10% calf serum by calcium phosphate-mediated transfection as described (4, 18), and the resulting HPRT-positive cells were selected by growth in hypoxanthine/aminopterin/thymidine (HAT) selective medium (19). Colonies were. picked through cloning cylinders and grown in the selective medium. Gel Assay for HPRT Enzyme Activity. After isoelectric fo- cusing in polyacrylamide gels, cytoplasmic extracts of trans- Fic. 1. Structure of the HPRT cDNA plasmid (p4aA8). Purified plasmid DNA was digested to completion with BamHI, electropho- resed on a 1% agarose gel, stained, and photographed, and the DNA was blotted onto nitrocellulose paper. The paper was hybridized to the 32P.labeled insert from pBR1.5 in 10% formamide/0.5 M NaCl/0.1 mM EDTA/50 mM Hepes, pH 7.4, at 60°C, washed, and exposed to x- ray film. (Left) X-ray film of the hybridized blot. (Right) Photograph of the ethidium bromide-stained gel. Lanes: 1, p4aB8 (a randomly picked control plasmid); 2, p4aA8 (the HPRT cDNA plasmid); 3, Hindlll-digested bacteriophage markers (from top to bottom): 23.6, 9.6, 6.6, 4.3, 2.3, 2.0, and 0.55 kb (29). Arrow, the heavily hybridizing insert fragment from the HPRT cDNA plasmid. Proc. Natl. Acad. Sci. USA 80 (1983) Fic. 2. Transfection of mouse LAY cells with the HPRT clone p4aA8 (Left) and a control clone, p4aB&.(Right). Approximately 10° HPRT-deficient LAQ cells per plate were transfected with 1 xg of plas- mid together with 20 ug of LA9 carrier DNA, grown in HAT medium after transfection, and stained with Giemsa. fected and control cells were assayed for HPRT activity by the method of Johnson et al. (20). Cloning of p4aA8 cDNA Fragments into M13. Plasmid p4aA8 was treated with BamHI and electrophoresed on a prep- arative 1% agarose gel in Tris acetate buffer (21). Subclones in M13 mp8 were generated by the method of P. Deininger (per- sonal communication). Briefly, the 1.6-kb BamHI fragment from p4aA8 was electroeluted (9), ligated into'a multimer, and sonicated to give random shear fragments of 200-800 base pairs (bp), which were ligated into the Sma I-cut phosphatase-treated M13 mp8 vector. The ligation mixture was used to transfect E. coli JM 101 (22). DNA Sequence Analyses. Single-stranded template was pre- pared from white M13 plaques and used for sequence deter- mination by the dideoxy chain-termination methods of Messing et al. (22) and Sanger et al. (23) using an M13 pentadecamer primer (New England BioLabs). HPRT Amino Acid Sequence Determination. Authentic hu- man HPRT enzyme was purified as described by Johnson et al. (24) and tryptic peptides were prepared and purified by HPLC (25). The amino acid sequences of several of the peptides were determined by microanalysis using a modified spinning cup method (26-28). 123 4 5& 6 OH He Fic. 3. Gel assay for HPRT activity of transfected mouse LAY cells. Polyacrylamide isoelectric focusing gels loaded with cell lysates from different cell types were assayed for enzyme activity, and results were detected by autoradiography. OH™ and H*, alkaline and acid ends of the gel; M and H, major mouse and human HPRT activities. (Both human and mouse HPRT also display a minor band of activity at a lower pI.) Lanes: 1, 3T6 mouse cells; 2, a line of cells derived from LA9 cells transfected to HAT resistance with p4aA8 and making HPRT that has human isoelectric focusing properties; 3, as in lane 2 but a sepa- rately derived cell line; 4, HeLa (human) cells; 5, 3T6 mouse cells; 6, LA9 HPRT* cells, the parent of the cell lines in lanes 2 and 3. Genetics: Jolly et al. Proc. Natl Acad. Sci. USA 80 (1983 | 479 RESULTS isolated previously (4), thought to contain a portion of the HPRT Approximately 300,000 colonies of the cDNA library were coding sequence and some HPRT intervening sequence. One transferred from nutrient-containing plates to filters and hy- reproducibly hybridizing colony was found on duplicate filters, bridized with the nick-translated insert from plasmid pBR1.5, and it was picked, colony purified, and grown to prepare large GGGGGGGGGGGGGGTCTTGCTGcCGCCcTCcGEccercetTccercTGéctTccGcccaccGgGectTTce? 10 20 30 40 50 60 met ala thr arg ser pro gly CEOTCCTGAGCAGTCAGCCCGCECGCCEECCEGCTCCGTTATGGCGACCCGCAGCCCTGGC 70 80 90 100 110 120 val val ile ser asp asp glu pro gly tyr asp leu asp leu phe cys ile pro asn his GTCGTGATTAGTGATGATGAACCAGGTTATGACCTTGATTTATTTTGCATACCTAATCAT 130 140 150 160 170 180 tyr ala glu asp leu glu arg (val phe ile pro his gly leu ile) met (asp arg thr glu TATGCTGAGGATTTGGAAAGGGTGTTTATTCCTCATGGACTAATTATGGACAGGACTGAA 190 200 210 220 230 240 arg leu ala arg asp val met) lys glu = met gly gly his his ile val ala leu cys val CGTCTTGCTCGAGATGTGATGAAGGAGATGGGAGGCCATCACATTGTAGCCCTCTGTGTG 250 260 270 280 290 300 leu lys gly gly tyr lys phe phe ala asp leu leu asp tyr ile lys ala leu asn = arg CTCAAGGGGGGCTATAAATTCTTTGCTGACCTGCTGGATTACATCAAAGCACTGAATAGA 310 320 330 340 350 360 asn ser asp arg ser ile pro met thr val asp phe ile arg leu lys (ser tyr cys asn AATAGTGATAGATCCATTCCTATGACTGTAGATTTTATCAGACTGAAGAGCTATTGTAAT 370 380 390 400 410 420 asp gln ser thr gly asp ile lys) val ile gly gly asp asp leu ser thr leu thr gly GACCAGTCAACAGGGGACATAAAAGTAATTGGTGGAGATGATCTCTCAACTTTAACTGGA 430 440 450 460 470 480 lys (asn_ val leu ile val glu = asp ile ile asp thr = gly lys) thr met gin thr leu leu AAGAATGTCTTGATTGTGGAAGATATAATTGACACTGGCAAAACAATGCAGACTTTGCTT 490 500 510 520 530 540 ser leu val arg gin tyr asn pro lys met val lys val ala ser leu leu val lys arg TCCTTGGTCAGGCAGTATAATCCAAAGATGGTCAAGGTCGCAAGCTTGCTGGTGAAAAGG 550 560 570 580 590 600 thr pro arg ser val gly tyr lys pro asp phe val gly phe glu ile pro asp lys phe ACCCCACGAAGTGTTGGATATAAGCCAGACTTTGTTGGATTTGAAATTCCAGACAAGTTT 610 620 630 640 650 660 val val gly tyr ala leu asp tyr asn glu tyr phe arg (asp leu asn his val cys val GTTGTAGGATATGCCCTTGACTATAATGAATACTTCAGGGATTTGAATCATGTTTGTGTC 670 680 690 700 710 720 ile ser glu thr gly lys ala lys) tvr lys ala aaK GCAAAATACAAAGCCTAAGATGAGAGTTCAAGTTGAGTTTGG ATTAGTGAAACTGGAAAA 730 740 750 760 770 780 AAACATCTGGAGTCCTATTGACATCGCCAGTAAAATTATCAATGTTCTAGTTCTGTGGCC 790 800 810 820 830 840 ATCTGCTTAGTAGAGCTTTTTGCATGTATCTTICTAAGAATTTTATCTGTEITTGTACTT DA 850 860 870 880 890 0 GAAATGTCAGTTGCTGCATTCCTAAACTGTTTATTTGCACTATGAGCCTATAGACTATCA 910 920 930 940 950 960 GTTCCCTTTGGGCGGATTGTTGTTTAACTTGTAAATGAAAAAATTCTCTTAAACCACAGC 970 980 990 1000 1010 1020 ACTATTGAGTGAAACATTGAACTCATATCTGTAAGAAATAAAGAGAAGATATATTAGTTT 1030 1040 1050 . 1060 1070 1080 TTTAATTGGTATTTTAATTTTTATATATGCAGGAAAGAATAGAAGTGATTGAATATTGTT 1090 1100 1110 1120 1130 1140 AATTATACCACCGTGTGTTAGAAAAGTAAGAAGCAGTCAATTTTCACATCAAAGACAGCA 1150 1160 1170 1180 1190 1200 TCTAAGAAGTTTTGTTCTGTCCTGGAATTATTTTAGTAGTGTTTCAGTAATGTTGACT GH 1210 1220 1230 1240 1250 1 ATTTTCCAACTTGTTCAAATTATTACCAGTGAATCTITGTCAGCAGTTCCCTTTTAAATS 1270 1280 1290 1300 1310 CAAATCAATAAATTCCCAAAAATTTAAAAAAAAAAA 1330 1340 1350 Fic. 4. Nucleotide sequence cf human HPRT cDNA and inferred amino acid sequence. The cDNA sequence was determined by dideoxy sequence analysis of M13 mp8 recombinant clones. Fifty-three independent cloned fragments having a random distribution within the cDNA were analyzed and their sequences were assembled according to their overlapping regions. All of the DNA sequence was confirmed on independent clones and, for about 85% of the cDNA fragment, both strands of the sequence were analyzed. Verified amino acid sequences of isolated peptides from purified human HPRT are shown in parentheses, We have assumed that translation begins at the first available AUG codon (at base 86) and terminates at the first chain-termination codon in the single open reading frame (i.e., at the UAA codon marked by asterisks at base 740), A potential poly- adenylylation site, A-A-T-A-A-A, occurs at base 1,334. 480 Genetics: Jolly-et al. amounts of plasmid DNA. Digestion of this purified plasmid (p4aA8). DNA with BamHI yielded an insert of approximately 1.6-kb (Fig. 1). The cloning vector has.two BamHI sites that flank the-cloned cDNA segment. If the cloned cDNA segment does not contain a BamHI restriction site, digestion with BamHI will excise the complete cDNA sequence flanked at the 5’ end by about 100 bp derived from simian virus 40 DNA and the G-C joint and at the 3’ end by 100-200 bp containing the A‘T joint and simian virus 40 DNA. Transfection of mouse HPRT-deficient LAQ cells with the HPRT cDNA p4aA8 resulted in the appearance of HAT-resis- tant colonies at a frequency of approximately 1 per 10° cells when 1 yg of plasmid and LAQ carrier DNA was used (Fig. 2). Transfection of these-cells with a control plasmid picked at ran- dom from the library (p4aB8) failed to produce any HAT-resis- tant colonies. Lysates from two HAT-resistant colonies were tested for HPRT activity byan isoelectric focusing gel assay (Fig. 3). Both colonies contained enzyme activities. that cofocused with the major band of authentic human HeLa marker enzyme, as opposed to the major and minor bands of enzyme activity derived from mouse 3T6 cells. Fragments of the isolated 1.6-kb insert were cloned into M13 mp8 and their sequences were determined. The nucleotide se- quence of the insert spanning the G-C to the A‘T tailing regions is presented in Fig. 4 together with the inferred amino-acid sequence of the encoded polypeptide and the tryptic and cyano- gen bromide peptides whose sequences were determined in- dependently. In the five peptides (a total of 56 amino acids), there are no disagreements between the amino acid sequences and the nucleotide sequences determined. DISCUSSION Three important prerequisites were satisfied for the successful cloning of full-length human HPRT cDNA. First, a probe con- taining HPRT coding sequence was isolated by a combination of transfection of haman DNA into enzyme-deficient mouse. cells and identification of the human sequences in such trans- fected cells by their hybridization to human repetitive (Alu) sequences or to total human DNA. Second, a cDNA library, preferably one likely to contain a full-length cDNA segment was prepared by using the procedure outlined previously (30), which has been further developed to allow expression of full- length cDNA sequences in mammalian cells (11). The simian virus 40-pBR322-based cloning vector contains a simian virus 40 early region promoter and intron 5‘ to the cDNA segment and a polyadenylation signal 3’ to the cDNA segment. These permit the expression of inserted cDNA sequences. In this in- stance, they permitted a direct test of the function of putative HPRT cDNA sequences in transfection assays and enabled us to identify and confirm full-length HPRT sequences. Third, it happened that pBR1.5 hybridizes to the 5’ end of the cDNA clone (this was shown by hybridization to M13 subclones) and hence its use as a hybridization probe selectively detected full- length or nearly full-length cDNA clones. This correlates with the observation that mouse sequences exist in this subcloned plasmid and at one end of the original genomic clone (p6B2aE2) and, hence, implies that the site of 5’ linkage of the exogenous human HPRT gene to mouse sequences is cloned in these plas- mids. When individual size classes of the cDNA library were hy- bridized with a fragment between positions 142 and 1,221 in the nearly full-length HPRT cDNA (Fig. 4), additional positive clones were found in the various sublibraries. The frequency of HPRT-positive clones in the library was about 2 x 1075 or - equivalent to approximately 2-5 copies of HPRT mRNA per Proc. Natl. Acad. Sci. USA 80 (1983) cell. We estimate the frequency of full-length cDNA clones rel- ative to the total number of positive clones in the complete li- brary to be about 5%. The sequence of the cDNA insert between the G-C and A:T linkers has been fully determined. The cDNA sequence is just under 1,350 nucleotides long, in good agreement with the size (approximately 1,600 bp) predicted from the RNA blots hy- bridized to pBR1.5 (4). It is not known whether some 5’ se- quences are missing in the cDNA clone, but the 85-bp 5’-non- translated region is in the usual size range for eukaryotic mRNA leader sequences. The coding sequence occupies 654 bp, slightly less than half of the full length of the cDNA, leaving a 3'-noncoding tail of approximately 600 nucleotides. The orga- nization is typical of eukaryotic mRNA sequences; in which the major noncoding region is usually 3’ to the coding sequences (31). The single open frame encodes a polypeptide 218 amino acids long, indicating a molecular weight of approximately 24,600. Post-translational removal of the NH,-terminal methionine would result in a protein 217 amino acids long, with a molecular weight-of 24,450. The purified human HPRT enzyme has been estimated to have a molecular weight of 24,000-26,000 (32, 33). In addition, the partial amino acid composition of the purified enzyme is in complete agreement with our inferred amino acid composition. Most impressively,. the inferred amino acid se- quence is in complete agreement with the full amino acid se- quence of the human erythrocyte HPRT enzyme recently de- termined by Wilson et al, (34). This sequence indicates that the NH,-terminal methionine is indeed cleaved in the fully pro- cessed: mature enzyme. The protein has no obvious large hydrophobic or hydrophilic domains but instead has small alternating regions of hydropho- bic and hydrophilic regions, suggesting a compact and globular structure. We predict that human HPRT is not a membrane constituent. Caskey and co-workers (35, 36) have used a different ap- proach in their isolation of a cDNA clone of the mouse HPRT gene, taking advantage of cells overproducing HPRT (35, 36). They have determined the sequence of mouse HPRT cDNA, and comparison of the inferred mouse and human amino acid sequences shows only eight differences, and the lengths of the polypeptides are identical. We expect that this cDNA clone will be useful for studying the organization of normal and mutant human HPRT genes, for examining the mechanisms of X-chromosome inactivation, for studying the fate and expression of HPRT cDNA in transfected HPRT-deficient mouse and human cells, and for examining the distribution and expression of HPRT sequences in mice after introduction of the gene into the male pronucleus of fertilized mouse eggs. T.F. thanks Dr. Wm. Nyhan for his interest and support. We thank Dorothy Miller for her administrative assistance during this work. This work was supported by grants from the Kroc Foundation, Santa Ynez, CA, and the Leon Gould Foundation, Tamarac, FL, and by National Institutes of Health Grants GM28223 and I RP] CA31928-01. D.F. was supported by Public Health Service Fellowship Training Grant CA09290. lL. Caskey, C. T. & Kruh, G. C. (1979) Cell 16, 1-9. 2. Becker, M. & Seegmiller, J. (1975) Arthritis Rheum. 18, 687- 694. 3. Seegmiller, J. E., Rosenbloom, F. J. & Kelley, W. N. (1967) Sci- ence 155, 1682-1684. 4. Jolly, D. J., Esty, A. C., Bernard, H. U. & Friedmann, T. (1982) Proc. Nail. Acad. Sci. USA 79, 5038-5041. 5. Clewell, D. B. & Helinski, D. R. (1969) Proc. Natl Acad. Sci. USA 62, 1159-1166. 10. 1]. 12. 13. 14. 15. 16. 17. 18. 19. 20. Genetics: Jolly et al Clewell, D. B. (1972) J. Bacteriol. 110, 667-676. Cameron, J. R., Panasenko, §. M., Lehman, I. R. & Davis, R. W. (1974) Proc. Natl Acad. Sci. USA 72, 3416-3421. Radloff, R., Bauer, W. & Vinograd, J. (1967) Proc. Natl. Acad. Sci. USA 57, 1514-1521. Yang, R. C. A., Lis, J. & Wu, R. (1979) Methods Enzymol. 68, 176-182. Maniatis, T., Jeffrey, A. & Kleid, D. G. (1975) Proc. Natl. Acad. Sci. USA 72, 1184-1188. Okayama, H. & Berg, P. (1982) Mol. Cell. Biol., in press. Lerach, H., Frischhauf, A. M., Hanahan, D., Wozney, J., Fuller, F., Crkvenjakov, R., Boedtker, H. & Doty, P. (1978) Proc. Natl. Acad. Sci. USA 75, 5417-5421. Gergen, J. P., Stern, R. H. & Wensink, P. C. (1979) Nucleic Acids Res. 7, 2115-2136. Holmes, D. & Quigley, M. (1981) Anal. Biochem, 114, 193-198. Southern, E. M. (1975) J. Mol, Biol. 98, 503-517. Wahl, G. M., Stern, J. & Stark, G. R. (1979) Proc. Natl Acad. Sci. USA 76, 3683-3687. Denhardt, D. T. (1966) Biochem. Biophys. Res. Commun. 23, 641-646. Wigler, M., Pellicer, A., Silverstein, S. & Axel, R. (1979) Cell 14, 725-731. Littlefield, J. E. (1966) Exp. Cell Res. 41, 190-196. Johnson, G. G., Eisenberg, L. R. & Migeon, B. R. (1979) Science 203, 174-176. Hayward, G. S. & Smith, M. G. (1972) J. Mol. Biol. 63, 383-395. Messing, J., Crea, R. & Seeburg, P. H. (1981) Nucleic Acids Res. 9, 309-321. Proc. Natl. Acad. Sci. USA 80 (1983) 481 Sanger, F., Nicklen, $. & Coulson, A. R. (1977) Proc. Natl Acad. Sci. USA 74, 5463-5467. Johnson, G. G., Kamage, A. L., Littlefield, J. W. & Kazazian, H. H.., Jr. (1982) Biochemistry 21, 960-966. Bohlen, P., Brazeau, T., Sesch, F., Ling, N. & Guillemin, R. (1981) Regul. Peptides 2, 359-369. Wittmann-Liebold, B. (1973) Hoppe Seiler’s Z. Physiol Chem. 354, 1415-1431. Hunkapiller, M. W. & Hood, L. E. (1978) Biochemistry 17, 2124-2133. Shively, J. E. (1981) Methods Enzymol. 79, 31-48. Philipsen, P., Kramer, R. A. & Davis, R. W. (1978) J. Mol. Biol. 123, 371-386. Okayama, H. & Berg, P. (1982) Mol. Cell. Biol. 2, 161-170. Breathnach, R. & Chambon, P. (1981) Annu. Rev. Biochem. 50, 349-383. Olsen, A. S. & Milman, G. (1977) Biochemistry 16, 2501-2505. Holden, J. A. & Kelley, W. N. (1978) J. Biol Chem. 253, 4459- 4463 Wilson, J. M., Tarr, G. E., Mahoney, W. C. & Kelley, W. N. (1982) J. Biol Chem. 257, 10978-10985. Melton, D. W., Konechi, D. S., Ledbetter, D. M., Hejtmancik, J. F. & Caskey, C. T. (1981) Proc. Natl. Acad. Sci. USA 78, 6977- 6980. Brennand, J., Chinault, A. C., Konechi, D. S., Melton, C. W. & Caskey, C. T. (1982) Proc. Natl Acad. Sci. USA 79, 1950-1956.