`
`Control mice
`VAA-fed mice
`Student‘s t-test
`
`% Lymph node cells expressing
`Lyt 1.1
`Lyt 2.1
`81621.1
`24.4214
`90.1i0.6
`24.2tO.7
`P < 0.001
`Non significant
`difference
`
`Transfer of a cloned irnmunoglohulin
`light-chain gene to
`mutant lryhridonta cells
`restores specific antibody production
`
`Immunofluorescently stained CBA lymph node cells were zénalysed
`using FACS—II (Becton-Dickinson) as described previously’ '3 , but
`with the argon ion laser set at 300 rnW, 488 nm and the photomultiplier
`tube (EMI 9524A) at 750 V with a fluorescence gain of 8 and a light
`scatter gain of 4. Data for each individual mouse were derived from
`the analysis of 5 x 104 lymph node cells. The fluorescein isothiocyanate—
`coupled rabbit anti-mouse immunoglobulin (obtained from Nordic) was
`selected for its binding to viable cells expressing immunoglobulin at
`the cell surface while producing negligible staining of other cells”. The
`values in the table were calculated after subtracting the count for
`irnmunoglobulin-positive cells and represent the mean :s.e. calculated
`iromseven determinations. These results have been repeated in two
`subsequent experiments.
`
`Our data are consistent with recent studies of Loveland and
`McKenzie’°'2‘ who have clearly shown that Lyt 1”’ T cells alone
`(depleted of Lyt 2" T cells) were sufiicient to restore immune
`responses generated against H-2 and non-H-2 alloantigens in
`ATXBM (adult thymectomized, irradiated and bone-marrow
`reconstituted) mice. Our data also corroborate their view" that
`Lyt 1"‘ cells have a central role in allograft responses. In either
`case, vitamin A acetate seems to be unique in that it enhances
`immune responses against an allograft when given by mouth
`and the mechanism of this action will require much further
`attention.
`
`Atsuo Ochi, Robert G. l-lawley,
`Marc .l. Shulrnan* & Nobnmiclii l-lozurni
`
`Ontario Cancer Institute and Department of Medical Biophysics,
`University of Toronto, 500 Sherbourne Street, Toronto,
`Canada M4X 1K9
`* Rheumatic Disease Unit, Wellesley Hospital, and
`Department of Medical Biophysics, University of Toronto,
`Toronto, Canada M4Y 1.13
`
`The expression of immunoglohulin (lg) genes is regulated at
`several
`levels. For example, although is-chain production
`requires a DNA rearrangement that juxtaposes variable and
`joining segments‘, this rearrangeent is not suficient for rc-
`chain gene expression; that is, some cell types do not permit
`immunoglobnlin production“. The mechanisms responsible
`for the regulation of the expressiomol rearranged immuno-
`globulin genes are poorly understood. The technique of modify-
`ing cloned genes in vitro and transferring the modified genes
`to cells in culture provides a tool for identifying the structural
`ieatures required for gene expression. To analyse immuno-
`glohnlin genes in this manner, however, it is lirst necessary to
`use, as recipients, cells that normally permit immunoglobulin
`production. We report here that a cloned is-chain gene is
`expressed in immunoglohulin-producing hyhridoma cells. Fur-
`thermore, the product oi the transferred K-chain gene is capable
`oi restoring specific antibody production to the transformed
`cells.
`The hybridoma Sp603 produces IgM(:<) specific for the hap-
`ten 2,4,6-trinitrophenyl ('I'NP)5. The rearranged gene encoding
`the TNP—specific K chain (Kywp) has been cloned . As recipient
`cells, we used the mutant cell line igk-14 which was derived
`from Sp603 and does not produce the K7-N1: chain. As shown
`in Fig. 3, the x-mp gene is apparently deleted from the igk-14
`cell line. Because the igk-14 cells still produce the TNP—specific
`)1. heavy chain, it would be expected that the expression of the
`Kjwp gene in these cells would restore the production of TNP-
`specific IgM. To select for cells that have taken up the rm
`gene, the cloned Kntrp gene, designated TK1, was inserted into
`the plasmid pSV2-neo, which contains a bacterial phos-
`photransferase gene (neo) and confers resistance to the amino-
`glycoside antibiotic G418 (ref. 7). In the present study, we have
`used two recombinant plasmids: in the plasmid pT-TK1, TK1
`was inserted so that the direction of transcription would be in
`the same direction as transcription initiated from the SV40
`early promoter;
`in pR-TK1,
`the orientation of the gene is
`reversed (Fig. 1).
`To transfer the DNA, bacteria harbouring the pT-Tn-1 and
`pR- T/<1 plasmids were converted to protoplasts and fused with
`igk-14, after the method of Schafiner‘. G418-resistant transfor-
`mants of igk—14 were then selected as described in the legend
`to Table 1. The frequency of G418-resistant cells ranged from
`5 ><10“‘ to 10‘5 per input igk—14 cell, and was comparable for
`the pT-T;c1, pR-TK1 and pSV2-neo plasmids. The resistant
`cells were then tested for the ability to make TNP—specific
`-plaques. By this criterion about 10°/o of the G418-resistant
`transformants from pT-TK1 and about 30% of the transfor-
`mants from pR-TK1 made TNP—specific lgM. By contrast, of
`30 G418-resistant transformants from the pSV2-neo vector
`alone, none made TNP—specific plaques.
`_
`Representative transformants were selected, cloned by limit-
`ing dilution and studied further. The production of the Km?
`chain has been assayed by TNP—specific haemagglutination and
`
`Qmag:-~:na_e_»-unvvn-+
`
`In summary, these findings support the hypothesis" that the
`anti—can_cer action of vitamin A acetate 17'" is mediated through
`an immunological process and we have shown that it acts by
`increasing the representation of the Lyt 1*2' phenotype in the
`T-cell population.
`We thank Dr G. L. Asherson for the gift of the F(ab')2 rabbit
`anti-mouse immunoglobulin and Prof. I. F. C. McKenzie for
`the monoclonal antibodies. We thank Drs E. Simpson
`and B. E. Loveland for discussions and Mrs Pat McFarlane
`for preparation of the manuscript. This work was supported by
`the _CRC and MRC.
`Received 1 December 1982; accepted 21 January 1933.
`1. Moore, T. in The Vitamin: (eds Sebrell, W. H. & Harris, R. S.) 245-266 (Academic, New
`York, 1967).
`2. Goodman, D. S. Fedn Prac. 38, 2501-2503 (1979).
`3. Bollag, W. in Retinaids, Advances in Basic Research and 'l7IempY (eds Orfanos, C. E. et
`:21.) 5-11 (Springer, Berlin, 1981).
`.
`J)
`. Spom. M. B. in Rerinaids, Advances in Basic Research and Therapy (eds Ortanos, C. E.
`eral.) 73-76 (Springer, Berlin, 1981).
`Pawson, B. A. Ann. N. Y. Acad. Sci. 359, 1-8 (1981).
`Dresser, D. W. Nature 217, 527-529 (1968).
`Floersheim, G. L. & Bollag, W. Transplantation 15, 564-567 (1972).
`Jurin, M. & Tannock, I. F. Immunology 23, 283-287 (1972).
`Cohen, B. E. & Cohen. 1. K. J. Immun. 111, 1376-1380 (1973).
`Blalcck, J. E. & Gifiord, G. E. Proc. natn. Acad. Sci. U.S.A. 74, 5382-5386 (1977).
`Dennert, G. & Lotan, R. Eur. J. Jmrnun. 8, 23-29 (1978).
`Dennert. G., Crowley, C., Kouba, J. & Lotan, R. J. nam. Cancer Inst. 62, 89-94 (1979).
`Goldfarb. R. H. & Herberman, R. B. J. Immun. 126, 2129-2135 (1981).
`Abb. 1., Abb, H. & Deinhardt, F. Immunophdnnacnlogy 4, 303-310 (1982).
`Spam. M. B. & Newton, D. L. Fedn Prue. 38, 2528-2534 (1979).
`Bollag, W. & Matter, A. Ann. N. Y. Acad. Sci. 359, 9-23 (1981).
`Medawar, F. B. & Hunt, R. Immunology 42, 349-353 (1981).
`Medawar, P. B., Hunt, R. .5; Merlin, J. Prac. R. Soc. Land. B206, 265-280 (1979).
`Malkovsky, M, er al. J. Immun, 130, 785-790 (1983).
`Loveland. B. E., Hogarth, P. M., Ceredig, Rh. & McKenzie, l.F.C. J. exp. Med. 153.
`1044-1057 (1981).
`. Loveland, B. E. & McKenzie. I. F. C. Transplantation 33, 217-221 (1982).
`. Moon, R. C., Grubbs, C. J. & Sporn, M. B. Cancer Res. 36. 2626-2630 (1976).
`. Malkovskfu M., Asherson. C. I... Stockinger, B. Er Watkins, M. C. Namre 309, 652-655
`(1982).
`. Hughes, W. L. er al. Fedn Prac. 23, 640-648 (1964).
`. Julius, M. H.. Simpson. E. & Herzenbcrg, L. A. Eur. J. lmmun. 3, 645-649 (1973).
`. Mage, M. G., Mcl-lugh, L. L. & Rolhstein, '1'. L. J. Imntun. Math. 15, 47-56 (1977).
`. Mage, M. :1 al. Eur. J. Immun. 11, 223-235 (1981).
`. Zernbala, M. A., Asherson, G. L., James. B. M. 8., Stein, V. E. & Watkins, M. C. J.
`Immun. 129, 1823-1829 (1982).
`. Gorini, G.. Medgyesi, G. A. 8: Doria, G. J. Immun. 103, 1132-1142 (1969).
`30. Abehsira. 0., Edwards, A. & Simpson, E. Eur. J. Immun. 11, 275-281 (1981).
`31. Simon, M. M. er al. Eur. J. lmrnun. 11, 246-250 (1981).
`
`.°.‘°€7-‘.*'?‘!""‘.“’!".*‘.°!°?°.".°‘."'
`
`.,...._.................._..
`
`_
`
`0028-0836/83/120340-0330‘. 00
`
`© 1983 Macmillan Journals Ltd V 4
`
`Sanofi/Regeneron Ex. 1021, pg 595
`
`
`
`CH 1983
`
`‘ NATURE VOL. 302 24 MARCH 1983
`
`~fl.iro.»~.n.;~.Ur'e‘E
`
`DER 322 mi
`
`5V40m
`
`.Fig. 1 Structure of TK1 transducing plasmids. The TK1 fragment
`(9.6 kb)6 was inserted into the Baml-II site of pSV2-nea (donated
`by P. Berg)7 and transfected into Escherichia colt‘ cells (C600) to
`obtain the two types of recombinant, pT-TK1 (TK1 inserted in
`tandem with the SV40 early promoter) and pR-Tkl (Tk 1 orienta-
`tion reversed with respect to the SV40 early promoter). The
`directions of transcription of the K1-Np gene and SV40 early region
`are indicated by arrows.
`
`Identification of Kqwp-Chalfl production in hybridoma cell
`Fig. 2
`lines and pT-TK1 and pR-TK1 transformants. Lane a, iglc-14;
`lane b, T1L2; lane c, T3L2; lane d, R11L3; lane e, 11311.4; lane
`f, Sp603. Secreted irnmunoglobulin was radiolabelled by incubat-
`ing cells for 18 ii
`in leucine-free Dulbecco‘s modified Eagle’s
`medium containing “C-leucine (5 p.Cin1l") and dialysed fetal
`calf serum (5%). Immunoglobulin was reduced and culture super-
`natants analysed by SDS-PAGE as described previouslys.
`
`Table 1 Secretion of TNP-specific lgM by G418-resistant ,
`transformants
`
`ector
`V
`pT-TK1
`
`Cell lines
`
`Sp603
`igk-14
`Transformants
`TILZ
`T3I..2
`T4L1
`T12e
`T17L1
`T19I.1
`T21L2
`R2L6
`R9L3
`R10L18
`R1 11.3
`R20L1
`R22L1
`R31L4
`
`Haemagglutination
`titre
`
`TNP plaque
`formation
`
`+
`"
`
`1/1,280
`<1/'2
`<1/2
`1/160
`<1/2
`<1/2
`1/20
`<1/2
`<1/2
`1/1,280
`<1/2
`<1/2
`1/20
`<1/2
`1/640
`1/320
`
`The igk-14 cells were fused with bacterial protoplasts containing
`either the pT-TK1 or pR-TK1 plasmids. The methods for increasing
`the plasmid copy ‘number and making protoplasts have been described
`elsewhere”. About 101° protoplasts were collected by centrifugation.
`107 iglc—14 cells, grown as described previouslys, were washed‘ and
`resuspended in Gibco H21 medium, and centrifuged onto the protoplast
`pellet. The protoplasts and igk-14 cells were then fused with polyethyl-
`ene glycol 1000 (PEG), as described by Taniguchi and Miller for the
`production of T-cell hybridomas”. Briefly, after centrifugation, the
`cell—protoplast pellet was gently resuspended in 2 ml of H21 medium
`containing 40% PEG and 10% dimethyl sulphoxide. After 15 s, the
`mixture was added to an equal volume of H21 medium containing 50%
`PEG and mixing was continued for a further 15 s. The cells were then
`diluted to 50 ml with H21 medium containing 20% fetal calf serum,
`washed once in the same medium and plated in 24-well culture dishes
`at a cell density of 105 cells per well. After 24 h, the fused cells were
`selected in H21 medium containing 1 mg ml” of G418 (a gift of
`Schering Corporation). The transformarzts resistant
`to C-418 were
`screened for the production of TNP-specific lgM by testing for TNP—
`specific plaque formation. Representative transformants were then
`cloned by limiting dilution. Culture supernatants were tested for
`haemagglutination of TN}-"—coupled sheep red cells (TNP—SRC) as
`described e1sewhere5. The sensitivity of the assay was enhanced by
`including monoclonal rat anti—).r antibody C2-23, as described else-
`where (M. Potash er
`.21.,
`in preparation). Plaquing was done on
`~10“ cells, also as described previously’.
`
`1'
`
`«
`
`ii
`
`plaque formation (Table 1). Synthesis of the Kmp chain has
`g also been measured by SDS-polyacrylamide gel electrophoresis
`(PAGE) (Fig. 2). The recipient cells igk-14 still produce the K
`3, chain from the myeloma X63-Ag8 used to generate the parental
`Sp603 hybridoma, and this myeloma K chain migrates more
`slowly than the Kmp protein. Low level production of the K'rNp
`chain by R11L3 can be detected both as low titre haemaggluti-
`nation and as a weak band on SDS-PAGE. In general, the
`intensities of the bands corresponding to the K1‘Np chain are
`roughly proportional to the TNP-specific haemagglutination
`titres.
`To analyse the KTN1) genes in the transformed cell lines,
`transformant DNA was digested with the restriction endonu—
`clease BamHI, fractionated by agarose gel electrophoresis and
`transferred to nitrocellulose. The blot was then probed with a
`CDNA clone of the K constant-region gene segment (Fig. 3).
`“ Previous work has shown that this probe detects three K-chain
`' genes in DNA from the Sp603 hybridoma—the bands at 5.9
`and 5.4 kilobases (kb) correspond to K -chain genes donated by
`the myeloma parent; the band at 9.6 kb represents the K-{Np
`gene‘, and this band is not observed in the case of igk-14. For
`» all the G418-resistant transformants tested, including the T1L2
`cells which do not make TNP-specific IgM, a band at 9.6 kb
`was observed with this probe. Results from DNA blot analysis
`of high molecular weight transformant DNA before and after
`digestion with the restriction endonuclease EcoRI (which
`cleaves the recombinant plasmids once) suggest that the trans-
`ferred sequences are integrated into cellular DNA as tandem
`oligomers (data not shown). These results are consistent with
`those obtained by other investigators using the pSV2 transfer
`vectors7'9'“’.
`We can estimate the copy number of the .-<7-Np gene in the
`transformants by comparing the intensity of the band corres-
`ponding to the Km; gene in DNA from the transformants with
`the intensity of the band corresponding to the Km? gene in the
`DNA of Sp603 which apparently contains one copy of the K-,-,.,,.
`gene per cell‘. By this criterion, the transforrnants T3L2 and
`R31L4 have multiple copies. of the Km; gene per cell, while
`" ‘ T1L2 and R11L3 contain about 1 copy per cell. It is interesting
`that the transformant R3114 makes more K-nqp chain than does
`T3L2, although T3L‘2 has more copies of the Km? gene. Fur-
`thermore, R31L4 makes about 10-fold less K7-Np chain per gene
`copy than does the wild-type hybridoma, although it should be
`pointed out that we do not know if all copies of the l<-nvp gene
`in R31L4 function equally efficiently. This variability in gene
`
`Sanofi/Regeneron Ex. 1021, pg 596
`
`
`
`NATURE VOL. 302 24 MARCH 1983
`
`. Kelley, D. E., Coleclough, C. & Perry, R. P. Cell 29, 681-689 (1982).
`. Hamlyn, P. H.. Brownlee, G. G., Cheng, C. C., Gait, M. J. & Milsteln,-C. Cell 15,
`1067-1075 (1978).
`. Falkner. F. G. & Zachau, H. G. Nature 298, 286-288 (1982).
`. Sandri-Goldin, R. M., Goldin, A. 1... Levine, M. & Glorioso, I. C. Malec. czll. Biol. 1.
`743-752 (1981).
`. Taniguchi, M. & Miller, J. F. A. P. J. exp. Med. 148. 373-332 (1978).
`. Southern, E. M. J. males. Biol. 97, 503-517 (1975).
`. Rice, D. & Baltimore. D. Prcc. rmtn. Acad. Sci. U.S.A. 79, 7862-7865 (1982).
`. Oi, V. T., Morrison, S. L., Henenberg, L. A. & Berg, P. Prac. rtatn. Acad. Sci. U.5'.A.
`80, 825-829 (1983).
`
`Deduced amino acid sequence
`irotn the bovine
`oxytocin-neurophysin l
`precursor cDNA
`
`H. Land*i, M. Gt-ez*:l:, S. Rupp-ert*, l-l. Schtnalei,
`M. Rehbeini, D. Richter!‘ & G. Schiitz*
`* Institute of Cell and Tumor Biology. German Cancer Research
`Center, Im Neuenheimer Feld 280, D-6900 Heidelberg, FRG
`‘r Institut fiir Physiologische Chemie, Abteilung Zellbiochemie,
`Universitiit Hamburg, UKE, 2000 Hamburg 20, FRG
`
`The nonapeptide hormone oxytocin-like arginine-vasopressin
`(AVP)“ is synthesized as part of a larger precursor polypep-
`tide. The precursor also includes the neutrophysin molecule with
`which the hormone is associated in the neurosecretory granules
`oi the hypothalamo-pituitary tract. A protein oi molecular
`weight (M,) ~20,000 has been isolated from supraoptic nuclei
`of rat hypothalami which, after tryptic cleavage, released a
`nettrophysin-like olecule of M,~ 10,000 and an oligopeptide
`related to oxytocins. This result was complemented by in vitro
`translation of bovine inypotltalatnic mRNA”. Among the
`primary
`translation
`products
`a
`single
`polypeptide
`of
`Ill,-16,500 was shown to contain antigenic determinants rec-
`ognized by specific antisera against bovine neurophysin I and
`oxytocin. Here we report the amino acid sequence of the bovine
`oxytocin—neut-ophysin I (OT-Npl) precursor which was derived
`from sequence analysis of the cloned cDNA. As is the case for
`the bovine arginine-vasopressin—neurophysin H (AVE-Npll)
`precursor‘, the signal sequence of the OT-Nplprecursor is
`immediately followed by the nonapeptide hormone which is
`connected to neurophysin l by a Gly-Lys-Arg sequence. A
`striking feature of the nucleic acid sequence is the 197-nucleo-
`tide long perfect homology with the AVP-Npll precursor
`tnRNA sequence encoding the conserved middle part of
`nettrophysins I and ll.
`A cDNA library of bovine hypothalamic mRNA‘ was
`screened for plasmids containing the mRNA sequence for the
`OT-Npl precursor. Bovine neurophysins I and II show nearly
`80% homology in their amino acid sequences,
`including a
`complete homology between amino acids 10 and 74, so con-
`siderable cross-hybridization was expected between the mRNA
`sequences for the AVP-NpII precursor and the OT-Npl pre-
`cursor.
`
`The cloned cDNA encoding the AVP-NpII precursor‘ was
`therefore chosen as a hybridization probe for in situ colony
`screeningg. Out of 5,000 recombinants, 63 clones gave a positive
`hybridization signal. Restriction analysis of
`the plasmids
`revealed two groups of 47 and 16 members containing different
`types of CDNA inserts (see Fig. 1 for examples). The first group
`represented CDNA sequences specific for the AVP-NpII pre-
`cursor‘. Sequence analysis of members of the second group
`(Fig. 1) showed that these contained OT-Npl precursor-specific
`sequences.
`Figure 2 shows the nucleotide sequences of the cloned OT-
`Npl precursor CDNA and of the previously described AVP-
`NpII precursor CDNA‘ plus the predicted amino acid sequences.
`
`i Present addresses: MIT Center for Cancer Research, 77 Massachusetts Avenue,
`Cambridge, Massachusetts 02138, USA (H.L.); University of Southern California, School of
`Medicine, Department of Microbiology, 2025 Zonal Avenue, Los Angeles, California
`90033, USA (M.G.'J.
`
`'
`
`’
`
`' '© 1983'Macmillan Journals Ltd
`
`Fig. 3 Blot hybridization of DNAs from hybridoma cell lines
`and pT-TK1 and pR-TK1 transformants. Lane a,.igk-14; lane b,
`T1L2; lane c, T3L2; lane at, R11L3; lane e, R31L4; lane f, Sp603.
`BamHI-digested DNA samples (20 ug) were electrophoresed
`through a 1% agarose gel at 2 V cm_1 for 40 h. After transfer to
`nitrocellulose“, the blot was hybridized with a 37'?-labelled cDNA
`clone of the K constant-region gene segment (pL21-58 provided
`by R. Wall) by a method described elsewhere .
`
`'
`
`expression raises the question of whether all the regulatory
`elements of the normal Kjwp gene are present or functioning
`on the cloned fragment. The messenger RNA cap sites have
`been identified for several immunoglobulin K-chain genes“. In
`all cases, the cap site seems to be within 30 base pairs (bp) of
`the initiation codon. As the TR] fragment has 5 kb of DNA
`upstream of the initiation codon, it is likely that the DNA
`fragment carries the K1-Np gene promoter. Similarly, the poly(A)
`addition site is estimated to lie 211 bp downstream of the
`constant-region gene segment”, which is well within the 1.2 kb
`of downstream flanking DNA. Experiments are in progress to
`determine the molecular basis for the differences in the level
`of expression of the K1-Np gene in the various transformants.
`Falkner and Zachau have studied the transient expression of
`mouse immunobulin K-chain genes in African green monkey
`cells (CV1), HeLa cells and mouse L cells”. In the case of the
`monkey CV1 cells and the HeLa cells, the K-chain gene was
`expressed only when transcription of the gene was placed under
`the direct control of an SV40 promoter after removal of the
`putative K -chain gene promoter region. In the case of the mouse
`L cells, the K-chain gene was not expressed. There are many
`structural diflerences between the vectors used by Falkner and
`Zachau and those used here. On the other hand, difierences in
`the cellular environments might underlie the different require-
`ments for K-chain gene expression. It should be possible to
`resolve these differences by using the gene transfer .system
`described here on cells of different types.
`‘
`This work was supported by grants from the MRC of Canada,
`the National Cancer Institute of Canada, the Arthritis Society
`of Canada and the Allstate Foundation. A.O. was supported
`by a research fellowship of the National Cancer Institute of
`Canada. R.G.l-I. was supported by a studentship of the MRC
`of Canada.
`Note added in proof: Rice and Baltimore” and Oi et al.” have
`recently reported the expression of cloned K light chain genes
`in transformed lymphoid cells.
`Received 17 November 1982: accepted 19 January 1983.
`1. Seidman, J. G., Max. E. E. & Leder, P. Nature 280. 370-375 (1979).
`2. Pat-slow, T. G. & Granner, D. K. Nature 299, 449-451 (1982).
`3. Colfino, P., Knowles, B., Nathenson, S. G. & Scharfi, M. D. Nature new Biol. 231, 87-90
`(1971).
`4. Ar-Rushdi. A.. Tan. K. B. & Croce, C. M. Somatic Cell Genet. 8, 151-161 (1982).
`S. Kéhler, G. & Shulman. M. J. Eur. J. Immun. 10, 467-476 (1980).
`6. Hawley, R. G., Shulrnan, M. 1., Murialdo, H., Gibson, D. M. & Hozumi, N. Pmc. ttam.
`Acad. Sci. Z/.$.A. 79, 7425-7429 (1982).
`. Southern, P. J. & Berg, P. J. molec. appl. Genet. 1, 327-341 (1982).
`. Schafinet, W.AProc. natn. Acad. Sci. U.S.A. 77, 2163-2167 (1980).
`. Mulligan, R. C. & Berg. P. Pmc. natn. Acad. Sci. U.S.A. 78, 2072-2076 (1981).
`. Canaani. D. & Berg. P. Prac. mun. Acad. Sci. U.S‘.A. 79. 5166-5170 (1982).
`
`, U028 -0836/83/120342-03501.00
`
`Sanofi/Regeneron Ex. 1021, pg 597