`Vol. 79, pp. 7862-7865, December 1982
`Immunology
`
`Regulated expression of an immunoglobulin K gene introduced into
`a mouse lymphoid cell line
`(Abelson murine leukemia virus/DNA transfection/gpt selection/lipopolysaccharide induction)
`DoUGLAs RICE AND DAVID BALTIMORE
`The Whitehead Institute for Biomedical Research, the Center for Cancer Research, and the Department of Biology, Massachusetts Institute of Technology,
`Cambridge, Massachusetts 02139
`Contributed by David Baltimore, September 16, 1982
`
`We have introduced a functionally rearranged
`ABSTRACT
`murine c light chain immunoglobulin (Ig) gene into an Abelson
`murine leukemia virus-transformed lymphoid cell line. Plasmid
`pSV2gpt-ic41, containing the c light chain gene from the myeloma
`MOPC41 and the selectable marker gene gpt, was introduced into
`81A-2 cells by the calcium phosphate coprecipitation technique.
`Cells expressing the gpt gene were selected by growth in medium
`containing mycophenolic acid. One transfected cell line, K-2, was
`shown to make K mRNA and polypeptide chains and to assemble
`the K chain product with y2b heavy chains to form an apparently
`complete IgG2b. When bacterial lipopolysaccharide was added to
`the growth medium, levels of K mRNA and polypeptide increased,
`showing regulated expression of the introduced K gene.
`
`B cell differentiation proceeds from the "pre-B" lymphocyte,
`which synthesizes A immunoglobulin (Ig) heavy chains but no
`light chains, to the mature B lymphocyte, which synthesizes
`both heavy and light chains and expresses surface Ig, and finally
`to the Ig-secreting plasma cell (1-5). The availability of trans-
`formed cell analogs has allowed biochemical characterization of
`these stages of cellular differentiation (6-11). Recently, such
`studies have contributed greatly to our understanding of the
`structure of Ig gene segments and the joining ofthese segments
`to produce a functionally rearranged Ig gene (12-17).
`Although much is now known about Ig gene structure, rel-
`atively little is known about the molecular mechanisms that
`control Ig gene expression. One approach to study such controls
`is to introduce Ig genes into various cell lines, including both
`lymphoid cells at various stages of differentiation and nonlym-
`phoid cells. One might then be able to identify control mech-
`anisms unique to lymphoid cells that allow the cells to express,
`assemble, and secrete Igs. To begin such studies, we have at-
`tempted to introduce a functionally rearranged murine K light
`chain gene into an Abelson murine leukemia virus (A-MuLV)-
`transformed lymphoid cell line.
`Previous studies have shown that A-MuLV infection ofbone
`marrow or fetal liver cells transforms cells of the B-lymphoid
`lineage, usually "pre-B" cells (18, 19). Derivatives of one A-
`MuLV transformant, 18-8, have been shown to switch from a
`to y2b heavy chain synthesis while in culture (20-22). One such
`derivative, 81A-2, synthesizes y2b protein, but has lost its K
`constant region light chain gene segments (unpublished data).
`Here we report that, after the introduction of a functionally
`rearranged K gene into 81A-2 cells, the K gene is expressed in
`a regulated manner.
`MATERIALS AND METHODS
`Cells. The A-MuLV-transformed cell line 81A-2, a derivative
`of the line 18-8, synthesizes y2b heavy chain protein, but no
`
`light chain, and has lost its K constant region genes (refs. 18 and
`22; unpublished data). Cells were grown and analyzed for Ig
`protein synthesis by metabolic labeling and immunoprecipita-
`tion as described (18). Nonreduced samples were prepared for
`electrophoresis as described by Margulies et al. (23).
`DNA Procedures. The phage ACh4A-41KC21, containing
`the rearranged genomic K light chain gene from the myeloma
`MOPC41, was obtained from P. Leder (12). The 7-kilobase-pair
`(kbp) EcoRI/BamHI fragment containing the K gene was in-
`serted into EcoRI- and BamHI-cleaved plasmid pSV2gpt, ob-
`tained from R. Mulligan (24). The resulting plasmid, shown in
`Fig. 1, is called pSV2gpt-K41. Ten micrograms of DNA from
`this plasmid was transfected into 5 x 107 81A-2 cells by a mod-
`ification ofthe calcium phosphate technique ofGraham and Van
`der Eb (25). Cells were washed in phosphate-buffered saline
`(0.14 M NaCl/2.5 mM KCl/16 mM Na2HPO4/1.4 mM
`KH2PO4), resuspended in 1 ml of transfection cocktail [made
`by adding DNA to 1 ml of Hepes-buffered saline, then adding
`62.5 j1. of 2 M CaCl2 (26)] and incubated 15 min at room tem-
`perature. Then 10 ml of medium was added and the cells were
`incubated at 370C for 4 hr. Cells were then washed in phos-
`phate-buffered saline, incubated at 37°C for 2 min in 2 ml of
`Hepes-buffered saline with glycerol (26), and washed again in
`phosphate-buffered saline. Cells were then resuspended in 10
`ml of nonselective medium, grown for 3 days, and then trans-
`ferred to selective medium [RPMI 1640 medium supplemented
`with mycophenolic acid at 2 ug/ml, xanthine at 250 ug/ml,
`hypoxanthine at 15 ,g/ml, and glutamine at 150 ,ug/ml (27)].
`Transfected and mock-transfected 81A-2 cells were passaged in
`selective medium for approximately 3 weeks, until the mock-
`transfected cells had died. The transfected cells were then
`cloned by limiting dilution in nonselective medium.
`RNA. Total cellular poly(A)-containing RNA was isolated by
`the guanidine-HCl procedure (28), fractionated according to
`size by electrophoresis in formaldehyde gels (29), transferred
`to nitrocellulose, and hybridized with 32P-labeled DNA probes
`as described (30).
`
`RESULTS
`To examine expression of a K gene from transfected plasmid
`DNA, the plasmid pSV2gpt-K41 was constructed to contain the
`rearranged chromosomal K light chain gene from the myeloma
`MOPC41 (12) and the selectable marker gene gpt, the gene
`from Escherichia coli that codes for the enzyme xanthine-gua-
`nine phosphoribosyltransferase [GPT; EC 2.4.2.22 (27)] (Fig.
`1). In mammalian cells grown in media containing inhibitors of
`purine synthesis (here, mycophenolic acid), expression of the
`gpt gene allows selective cell growth using xanthine as the pre-
`
`The publication costs ofthis 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.
`
`Abbreviations: A-MuLV, Abelson murine leukemia virus; GPT, xan-
`thine-guanine phosphoribosyltransferase; kb, kilobase(s); kbp, kilobase
`pair(s); LPS, bacterial lipopolysaccharide; SV40, simian virus 40.
`
`7862
`
`Merck Ex. 1017, pg 655
`
`
`
`Immunology: Rice and Baltimore
`
`Proc. Natd Acad. Sci. USA 79 (1982)
`
`7863
`
`w
`
`NV
`
`14 0
`
`SV40 ear
`promoter
`
`'-4
`
`C,
`
`Structure of the plasmid pSV2gpt-K41 (11.4 kbp). Mouse
`FIG. 1.
`DNA sequences containing the rearranged K light chain gene from the
`myeloma MOPC41 are represented by the heavy dark line. The leader
`(L), variable plus joining region (VJ), and constant region (CK) of the
`K gene are indicated. Simian virus 40 (SV40) sequences, represented
`by hatched regions, include DNA segments containing the early pro-
`moter, the small tumor antigen intervening sequence, and sequences
`for termination and polyadenylylation of SV40 early transcripts. The
`gpt gene fromEscherichia coli is shown as a stippled region. ori, Origin
`of replication; ampr, ampicillin resistance. Transcription units are in-
`dicated by wavy lines.
`
`cursor for synthesis ofguanine nucleotides (27). In pSV2gpt, the
`parental plasmid used for this construction, the gpt gene is tran-
`scribed from the SV40 early promoter and is followed by a re-
`gion of SV40 DNA containing the small tumor antigen inter-
`vening sequence and signal sequences for transcript termination
`and polyadenylylation (24). To reduce the possibility of tran-
`scription of the K light chain gene from promoters other than
`its own, plasmid pSV2gpt-K41 was designed so that transcrip-
`tion from the SV40 promoter is in opposite orientation from that
`required for K gene expression.
`The 81A-2 cell line used as recipient ofthe transfected DNA
`is an A-MuLV-transformed murine lymphoid cell that synthe-
`sizes y2b heavy chain but no light chain [no CK alleles can be
`detected by hybridization (ref. 22 and unpublished data)]. Plas-
`mid pSV2gpt-K41 DNA was introduced into 81A-2 cells by the
`calcium phosphate coprecipitation technique (25). Cells ex-
`pressing the gpt gene were selected by growth in medium con-
`taining mycophenolic acid and then cloned by limiting dilution.
`When DNA from three selected cell lines was prepared and
`analyzed by hybridization with a K probe, all three lines were
`found to have acquired one or a small number ofthe introduced
`K genes. From the pattern ofthe hybridizing bands, at least two
`of three lines were judged to be independent transfectants.
`Eight cell lines were assayed for GPT enzyme activity by the
`in situ gel assay ofMulligan and Berg (24); all eight were positive
`(data not shown).
`When the eight gpt+ cell lines were assayed for production
`of K protein by metabolic labeling with [3S]methionine and im-
`munoprecipitation with anti-K antiserum, all eight were found
`to synthesize a polypeptide which (i) was precipitable with anti-
`K antiserum (Fig. 2, lane d for clone K-2 and data not shown for
`the others); (ii) comigrated with the K chain produced by the
`myeloma MPC1l (apparent Mr 23,000) (Fig. 2, lane a); and (iii)
`was not evident in the nontransfected 81A-2 parent cell line
`(Fig. 2, lane b). In the original autoradiogram, the background
`bands in the Mr 23,000 region are much fainter than reproduced
`here. Because the 81A-2 cells lack CK alleles, none of the back-
`ground bands are K light chain. Precipitation of the Mr 23,000
`
`Mr
`x lo-,
`
`- 120
`
`- 85
`-65
`- 48.5
`
`r.
`cz
`
`+c
`
`q
`
`00
`
`w
`
`I N C1
`±
`
`cq C
`+
`+
`k
`k
`c
`I
`c C9
`cq
`cq
`cq
`I.¢ :< :
`<
`<:
`w
`wOD) D
`oo c
`:
`
`k
`
`.-:.~~lw
`
`1
`
`I
`
`Em~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1
`
`A_
`
`_-
`
`-24
`
`-18.5
`
`a
`
`b
`
`c
`
`d
`
`e
`
`f
`
`g
`
`h
`
`Polyacrylamide gel electrophoresis of Ig polypeptides syn-
`FIG. 2.
`thesized by 81A-2 cells and the transfectant K-2. Cytoplasmic extracts
`were prepared from cells labeled with [35S]methionine, immunopre-
`cipitated, and analyzed by NaDodSO4/polyacrylamide gel electropho-
`resis. Lane a, myeloma MPC11 extract immunoprecipitated with anti-
`K antiserum; lanes b and c, extract from parent A-MuLV-transformant
`81A-2 grown in the absence (lane b) and presence (lane c) ofSalmonella
`typhimurium lipopolysaccharide (LPS) and immunoprecipitated with
`anti-K antiserum; lanes d and e, extract from the transfectant K-2
`grown in the absence (lane d) and presence (lane e) of LPS and im-
`munoprecipitated with anti-K antiserum; lanes f and g, extract from
`the transfectant K-2 mixed with unlabeled MPC11 extract (lane f) or
`unlabeled MOPC104E extract (lane g) and immunoprecipitated with
`anti-K antiserum; lane h, extract from the transfectant K-2 immuno-
`precipitated with anti-AT antiserum. Sizes of molecular weight marker
`proteins are indicated.
`
`polypeptide by anti-K serum was blocked by competition with
`an unlabeled MPC11 protein extract (containing authentic K
`light chains) but not by an unlabeled MOPC104E protein ex-
`tract (containing AI light chains) (Fig. 2, lanes fand g). Also, the
`apparent K chain was not precipitable by anti-AI antiserum (Fig.
`2, lane h). Therefore, the Mr 23,000 polypeptide appears to be
`the protein product of the K light chain gene transfected into
`the 81A-2 cells.
`To examine the RNA produced from the transfected K gene,
`poly(A)-containing mRNA was prepared from the parent 81A-
`2 cell line and the transfectant K-2. The RNA was size-fraction-
`ated by agarose gel electrophoresis, transferred to nitrocellu-
`lose, and hybridized with a 32P-labeled plasmid DNA probe
`containing the K constant region (CK), the K joining (J,) seg-
`ments, and the sequence that intervenes between them. No
`hybridization was detected to the RNA prepared from the par-
`
`Merck Ex. 1017, pg 656
`
`
`
`7864
`
`Immunology: Rice and Baltimore
`
`Proc. Natl. Acad. Sci. USA 79 (1982)
`
`cc~~~r
`
`Cl1
`<
`¢c
`
`<1
`¢c
`IG
`
`kh
`
`I - 1.2
`q-.M
`
`0.8
`
`'I
`
`kb
`26-
`1,9-
`
`1.2
`
`I
`
`bD
`
`c
`
`Analysis of K RNA transcripts in the transfectant K-2.
`FIG. 3.
`Poly(A)-containing RNA was isolated from cells, size-fractionated by
`agarose gel electrophoresis, transferred to nitrocellulose paper, and
`hybridized with 32P-labeled DNA from a plasmid containing the se-
`quence from J. through CK. Lanes a and b, RNA from the transfectant
`K-2 grown in the absence (lane a) or presence (lane b) of LPS. Lane c,
`RNA from the K-producing A-MuLV-transformed cell line 18-48.
`
`ent line 81A-2 (not shown), but RNA from the K-2 line contained
`hybridizing species ofapproximately 1.2, 1.9, and 2.6 kilobases
`(kb) (Fig. 3, lane a). The smaller RNA comigrated with authentic
`K mRNA from the A-MuLV-transformed cell line 18-48 (Fig.
`3, lane c). [The smaller 0.8-kb RNA in 18-48 is an aberrantly
`small K transcript (5).] In other experiments, both the 1.9-kb
`and the 2.6-kb RNAs were found to hybridize strongly to a probe
`specific for the intervening sequence between JK and CK and
`hybridize weakly to a pBR322 DNA probe. Hence, these higher
`molecular weight species are some type of aberrant RNA. The
`1.2-kb species, however, appears to be an authentic K mRNA
`transcript in that it hybridizes only to the probe containing K
`coding sequences and not to the intervening sequence probe
`or pBR322.
`Because many A-MuLV-transformed lymphoid cell lines in-
`crease Ig production when LPS is added to the medium (18, 31),
`we investigated the effects of LPS on K chain synthesis in the
`transfected line. The parent line 81A-2 increases synthesis of
`y2b heavy chain protein and mRNA upon induction by LPS
`(22) (Fig. 2, lane c). When LPS was added to the K-2 cells, K
`light chain synthesis increased approximately 5-fold, to a level
`approximately 1/15th that ofthe myeloma MPC1l (Fig. 2, lane
`e). To determine if the LPS-induced increase in K chain syn-
`thesis was due to an increased content of specific mRNA, the
`mRNA fraction was prepared from LPS-treated 81A-2 parent
`cells and transfected K-2 cells. Again, no K mRNA species was
`
`¢
`<~~~CC
`
`Rael
`
`l
`
`cCa
`
`ct
`
`cc
`
`X
`
`I_'
`
`Ml1
`x l20
`
`205
`
`- 116
`
`- 97
`
`- 66
`
`- 45
`
`.-~~~~90
`-
`d
`b
`e
`
`a
`
`Polyacrylamide gel electrophoresis of nonreduced Ig syn-
`FIG. 4.
`thesized by the K-2 transfectant. Cytoplasmic extracts were prepared
`as for Fig. 2. Lanes a and b, extracts from the parent 81A-2 cells grown
`without (lane a) and with (lane b) LPS and immunoprecipitated with
`anti-K antiserum; lanes c and d, extract from K-2 transfectants grown
`without (lane c) and with (lane d) LPS and immunoprecipitated with
`anti-K antiserum; lane e, myeloma MPC11 extract immunoprecipi-
`tated with anti-K antiserum.
`
`detected in the parental cells, but the K-2 cells contained in-
`creased levels of the 1.2- and 1.9-kb species (Fig. 3, lane b).
`Interestingly, the level of the 1.9-kb RNA species increased
`even more than that of the presumably authentic 1.2-kb RNA
`species.
`Because the transfected cells were producing both y2b heavy
`chains and K light chains, it was possible that the cells could
`assemble the heavy and light chains into IgG. To examine this
`question, samples of [35S]methionine-labeled protein extracts
`were immunoprecipitated with anti-K antiserum and the non-
`reduced samples were subjected to NaDodSO4/polyacryl-
`amide gel electrophoresis. The parental 81A-2 cells produced
`a protein of approximately the correct size for y2b heavy chain
`dimers (Fig. 4, lanes a and b; the darker appearance of lane a
`is due to more labeled extract present). The K-2 cells produced
`a protein that migrated slightly faster than the IgG2b produced
`by the myeloma MPC11 (Fig. 4, lanes d and e) but slower than
`the bulk of the rabbit IgG antiserum visualized by staining (not
`shown). In other experiments (not shown) no free K chain was
`found in the K-2 cells, although a significant amount was present
`in MPC1L cells. Essentially all of the K chain produced in the
`K-2 cells appears to be assembled into IgG2b.
`
`DISCUSSION
`The major result of these studies is the demonstration that a
`functional K gene can be introduced into a lymphoid cell line
`in which it will be continuously expressed. This opens the pos-
`sibility of examining control and rearrangement mechanisms in
`
`Merck Ex. 1017, pg 657
`
`
`
`Immunology: Rice and Baltimore
`lymphoid cells by using inserted genetic elements.
`The K gene introduced into 81A-2 cells apparently functions
`normally in spite of being in a very unusual context. The gene
`was in an SV40/pBR322 vector that then integrated into a pre-
`sumably random site in the cell genome, a site unlikely to be
`related to the normal location of the K gene in chromosome 6.
`In spite of its unusual context, the introduced gene was ex-
`pressed at about the same level as the resident y2b heavy chain
`gene. The K gene was apparently using its own promoter be-
`cause in the construction no promoter was provided that faced
`-in the correct direction. It is possible that the SV40 DNA se-
`quences present might have provided some enhancing function
`for K expression (32).
`The introduced K gene not only was expressed at a basal level
`but also- was inducible by LPS. The mechanism and function
`of this induction system are far from clear, but the ability ofthe
`introduced K gene to respond indicates that sufficient K-related
`DNA sequences to provide for LPS inducibility were included
`in the construct. The construct contained, in addition to the VK,
`JK1 and CK coding segments, the intervening sequence between
`the coding regions and about 1-L5 kb of DNA both 5' of VKJK
`and 3' of CK. Any of this extra DNA could be involved in pro-
`moter and control functions, but the results make it unlikely that
`any sequences important for K expression exist more than 1.5
`kb to either side of the coding region.
`LPS control ofheavy chain expression in 81A-2 cells is allele
`specific and correlates with a deletion in the intervening se-
`quence between VHDJH and CM (22, 33). The productively rear-
`ranged heavy chain allele is inducible by LPS and contains this
`deletion, whereas the other allele, containing a nonproductive
`rearrangement, lacks the deletion and is not inducible by LPS.
`Therefore, LPS inducibility of heavy chain seems to be deter-
`mined at the DNA level. Whether the introduced K gene is
`responding directly to LPS or to the product ofthe heavy chain
`allele is an open question. The possibility that transcription of
`the light chain gene is controlled by a product ofthe heavy chain
`locus is an interesting possibility and needs further investigation.
`We thank Drs. F. Alt, M. Boss, S. Lewis, and R. Mulligan for helpful
`discussions. We thank Dr. R. Mulligan for plasmid pSV2gpt and Dr.
`P. Leder for the cloned MOPC41 K gene. This work was supported by
`Grant MV-34N from the American Cancer Society, Grant CA14051
`(core grant to S. E. Luria) from the National Cancer Institute, and a
`contribution from the Whitehead Charitable Foundation. D.R. was
`supported by a Helen Hay Whitney Postdoctoral Fellowship. D. B. is
`an American Cancer Society Research Professor.
`Melchers, F., Von Boehmer, H. & Phillips, R. A. (1975) Trans-
`1.
`plant Rev. 25, 26-58.
`Rosenberg, Y. & Parish, C. R. (1977)J. Immunol. 118, 612-617.
`Burrows, P. D., Jeune, M. & Kearney, J. F. (1979) Nature (Lon-
`don) 280, 838-841.
`
`2.
`3.
`
`Proc. Natl. Acad. Sci. USA 79 (1982)
`
`7865
`
`Levitt, D. & Cooper, M. D. (1980) Cell 19, 617-625.
`Siden, E., Alt, F. W., Shinefeld, L., Sato, V. & Baltimore, D.
`(1981) Proc. Nat. Acad. Sci. USA 78, 1823-1827.
`Potter, M. (1972) Physiol. Rev. 52, 631-719.
`Rosenberg, N. & Baltimore, D. (1976) J. Exp. Med. 143, -1453-
`1463.
`Cantor, H. & Boyse, E. A. (1977) Immunol. Rev. 33, 60-124.
`Paige, C. J., Kincade, P. W. & Ralph, P. (1978)J. Immunol 121,
`641-647.
`Raschke, W. C., Mather, E. L. & Koshland, M. E. (1979) Proc.
`Natl Acad. Sci. USA 76, 3469-3473.
`Strober, S., Gronowicz, E. S., Knapp, M., Slavin,.S., Vitetta, E.
`S., Warnke, R. A., Kalzin, B. & Schroeder, J. (1980) Immunol.
`Rev. 48, 169-195.
`Seidman, J. G. & Leder, P. (1978) Nature (London) 276, 790-
`795.
`Brack, C., Hirama, M., Lenhard-Schuller, R. & Tonegawa, S.
`(1978) Cell 15, 1-14.
`Early, P., Huang, H., Davis, .M., Calame, K. & Hood, L. (1980)
`Cell 19, 981-992.
`Sakano, H., Maki, R., Kurosawa, Y., Roeder, W. & Tonegawa,
`S. (1980) Nature (London) 286, 676-683.
`-Shimizu, A., Takahashi, N., Yaoita, Y. & Honjo, T. (1982) Cell
`28, 499-506.
`Marcu, K. (1982) Cell 29, 719-721.
`Siden, E., Baltimore, D., Clark, D. & Rosenberg, N. (1979) Cell
`16, 389-396.
`Alt, F., Rosenberg, N., Lewis, S., Thomas, E. & Baltimore, D.
`(1981) Cell 27, 381-400.
`Burrows, P., Beck, G. & Wabl, M. (1981) Proc. Natl. Acad. Sci.
`USA 78, 564-568.
`Alt, F. W., Rosenberg, N. E., Lewis, S., Casanova, R. J. & Bal-
`timore, D. (1979) in B Lymphocytes in the Immune Response,
`eds. Klinman, N., Mosier, D., Scher, I. & Vitetta, E. S. (Elsev-
`ier/North-Holland, New York), pp. 33-41.
`Alt, F., Rosenberg, N., Casanova, R., Thomas, E. & Baltimore,
`D. (1982) Nature (London) 296, 325-331.
`Margulies, D. H., Kuehl, W. M. & Scharff, M. D. (1976) Cell 8,
`405-415.
`Mulligan, R. C. & Berg, P. (1980) Science 209, 1422-1427.
`Graham, F. L. & Van der Eb, A. J. (1973) Virology 52, 456-467.
`Chu, G. & Sharp, P. (1981) Gene 13, 197-202.
`Mulligan, R. C. & Berg, P." (1981) Proc. Natl. Acad. Sci. USA 78,
`2072-2076.
`Strohman, R. C., Moss, P. S., Micou-Eastwood, J., Spector, P.,
`Przybyla, A. & Paterson, B. (1977) Cell 10, 265-273.
`Maniatis, T., Fritsch, E. & Sambrook, J. (1982) Molecular Clon-
`ing, A Laboratory Manual (Cold Spring Harbor Laboratory, Cold
`Spring Harbor, NY), pp. 202-203.
`Thomas, P. S. (1980) Proc. Natl. Acad. Sci. USA 77, 5201-5205.
`Rosenberg, N., Siden, E. & Baltimore, D. (1979) in B Lympho-
`cytes in the Immune Response, eds. Cooper, M., Mosier, D.,
`Scher, I. & Vitetta, E. (Elsevier/North-Holland, Amsterdam),
`pp. 379-386.
`Moreau, P., Hen, R., Wasylyk, B., Everett, R., Gaub, M. P. &
`Chambon, P. (1981) Nucleic Acids Res. 9, 6047-6068.
`Alt, F. W., Rosenberg, N., Enea, V., Siden, E. & Baltimore, D.
`(1982) Mol. Cell. Biol. 2, 386-400.
`
`4.
`5.
`
`6.
`7.
`
`8.
`9.
`
`10.
`
`11.
`
`12.
`
`13.
`
`14.
`
`15.
`
`16.
`
`17.
`18.
`
`19.
`
`20.
`
`21.
`
`22.
`
`23.
`
`24.
`25.
`26.
`27.
`
`28.
`
`29.
`
`30.
`31.
`
`32.
`
`33.
`
`Merck Ex. 1017, pg 658