ICN-UCLA Symposia on Molecular and Cellular Biology
`Volume XIV, 1979
`
`)roduction
`
`EUCARYOTIC GENE
`REGULATION
`
`edited by
`
`RICHARD AXEL
`College of Physicians and Surgeons
`Columbia University
`New York, New York
`
`TOM MANIATIS
`Division of Biology
`California Institute of Technology
`Pasadena, California
`
`C. FRED FOX
`Department of Micobiology and Molecular Biology Institute
`University of California, Los Angeles
`Los Angeles, California
`
`.posia on
`~ystone,
`
`@ :~:::~:~ :.~:.:s.'"' '~··"'''"· ... ::"::
`
`New York London Toronto Sydney · San Francisco
`
`Merck Ex. 1033, pg 1088
`
`

`
`COPYRIGHT @ 1979, BY ACADEMIC PRESS, INc.
`ALL RIGHTS RESERVED.
`NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR
`TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC
`OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY
`INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT
`PERMISSION IN WRITING FROM THE PUBLISHER.
`
`ACADEMIC PRESS, INC.
`111 Fifth Avenue, New York, New York 10003
`
`United Kingdom Edition published by
`ACADEMIC PRESS, INC. (LONDON) LTD.
`24/ 28 Ova·! Rood, London NW! 7DX
`
`ISBN 0-12 - 068350-4
`
`PRINTED IN THE UNITED STATES OF AMERICA
`
`79 80 81 82
`
`987654321
`
`•.
`
`Merck Ex. 1033, pg 1089
`
`

`
`tGE SCANGOS eta/.
`
`EUCARYOTIC GENE REGULATION
`
`1. Acad. Sci.
`
`tl. Acad. Sci.
`
`netic Inter-
`) , Brookhaven
`
`8). Proc.
`
`Adelberg, E.
`
`ange, R.
`
`atl. Acad.
`
`ange, R.
`
`Natl. Acad.
`
`(1979).
`
`cer, A.,
`3.
`and Ruddle, F.
`s .
`nd Ruddle, F.
`
`A.,
`e, K., and
`
`•phys. Res.
`
`Mann, J.
`
`·s, W., and
`
`verstein, S.
`
`TRANSFORMATION OF MAMMALIAN CELLS
`
`WITH PROKARYOTIC AND EUKARYOTIC GENESl
`
`Michael Wigler2, Raymond Sweet, Gek Kee Sim, Barbara
`*
`Wold, Angel Pellicer, Elizabeth Lacy , Tom
`Maniatis*, Saul Silverstein, and Richard Axel
`
`College of Physicians and Surgeons,
`Columbia University, New York, N.Y. 10032
`*
`Division of Biology, California Institute
`of Technology, Pasadena, Calif. 91125
`
`ABSTRACT Cellular genes coding for selectable biochemical
`functions can be stably introduced into cultured mammal(cid:173)
`ian cells by DNA-mediated gene transfer (transformation).
`Biochemical transformants are readily identified by the
`stable expression of a gene coding for a selectable mar(cid:173)
`ker. - These transformants represent a subpopulation of
`the competent cells which integrate other physically un(cid:173)
`linked genes for which no selective criteria exist.
`In
`this manner, we have used a viral thymidine kinase gene
`as a selectable marker to isolate mouse cell lines which
`we have stably transformed with the tk gene along with
`bacteriophage ~X 174, plasmid pBR 322, or the cloned
`chromosomal rabbit S-globin gene sequences. ~X co(cid:173)
`transformants were studied in greatest detail. The fre(cid:173)
`quency of co~transformation is high, 15 of 16 tk+ tr~ns­
`formants contain the ~X sequences. Further, from one to
`more than fifty <PX sequences are stably integrated into
`high molecular weight nuclear DNA isolated from indepen(cid:173)
`dent clones. The introduction of cloned eukaryoticgenes
`now provides an in vivo system to study the functional
`significance of various features of DNA sequence organ(cid:173)
`ization. We have analyzed the ability of the mouse
`fibroblast transformant to transcribe and process the
`heterologous raubit G-globin gene. Hybridization exper(cid:173)
`iments indicate that in at least one transformant, rabbit
`S-globin sequences are expressed in the cytoplasm as a
`discrete 9S species, suggesting that mouse fibroblast
`may contain the enzymes necessary t o transcribe and cor(cid:173)
`rectly process a rabbit gene whose expression is usually
`restricted to erythroid cells. These studiesdemonstrate
`the potential value of co-transformation systems in the
`analysis of eukaryotic gene expression.
`
`457
`
`Copyrighl c 1979 by Academic P ress, Inc.
`All ri&ht of reproduction ln any fonn reserved.
`ISBN ()-12~6835().4
`
`Merck Ex. 1033, pg 1090
`
`

`
`458
`
`39. M. WIGLER eta/.
`
`E
`
`INTRODUCTION
`
`Specific genes can be stably introduced into cultured
`mammalian cells by DNA-mediated gene transfer. The process
`of transformation results in a change in the genotype of the
`recipient cell and provides a unique opportunity to study the
`function and physical state of exogenous genes in the trans(cid:173)
`formed host.
`In our laboratories, we have developed trans(cid:173)
`formation systems which may allow the introduction of virtu(cid:173)
`ally any defined gene into cultured cells. We have therefore
`performed a series of transformation experiments with a
`1) to
`variety of different eukaryotic and prokaryotic genes:
`develop in vivo systems to study the functional significance
`of various features of DNA sequence organization;
`2) as a
`means for gene purification where now classical routes in(cid:173)
`volving recombinant DNA technology and molecular hybridiza(cid:173)
`tion are inapplicable;
`3) to examine the fluidity and prom(cid:173)
`iscuity of the eukaryotic chromosome.
`In initial studies, we developed a transformation system
`for the thymidine kinase (tk) gene of herpes simplex virus
`(HSV-1) . Through a series of electrophoretic fractionations
`in concert with transformation assays, we isolated a unique
`3.4 kb fragment of viral DNA which is capahJe of efficiently
`transferring tk activity to mutant Ltk- cells (Wigler et al.,
`1977) . Extension of these studies to unique cellular genes
`has resulted in the stable transfer of genes coding for thy(cid:173)
`midine kinase, adenine-phosphoribosyl transferase and a metho(cid:173)
`trexate resistant mutant of dihydrofolate reductase to mouse
`fibroblasts (Wigler et al., 1978, 1979a).
`The methods we have used to transfer these genes can, in
`principle, be applied to any gene for which conditional
`selection criteria are available. The isolation of cells
`transformed with genes which do not code for selectable
`markers, however, is problematic, since current transforma(cid:173)
`tion procedures are highly inefficient. We have re-
`cently demonstrated the feasibility of co-transforming cells
`with two physically unlinked genes (Wigler et .al., 1979b).
`Co-transformed cells can be identified and isolated when one
`of these genes codes for a selectable marker. We have used
`the viral tk gene as a selectable marker to isolate mouse cell
`lines which contain the tk gene along with e~ther bacterio(cid:173)
`phage ~X 174, plasmid pBR 322, or the cloned rabbit S-globin
`ge ne sequences stably integrated into cellular DNA. We have
`further demonstrated that the gene coding for the rabbit
`S-globin in transformed mouse fibroblasts is properly recog(cid:173)
`niz e d by the transcriptional and processing enzymes of the
`mouse cell to generate RNA indistinguishable from the mature
`globin mRNA of the rabbit erythroblast (Wold et al., 1979).
`
`i
`a~
`sj
`tl:
`gE
`(~
`ta
`th
`fo
`rna
`ta
`
`wi
`fo
`si
`Pu:
`cl•
`mo1
`lY
`l e •
`fr E
`ge1
`
`s e c
`wa ~
`rec
`ate
`ceJ
`nic
`19/
`
`for
`ref
`Sin
`nurr
`kar
`~X.
`4 0
`fra
`fra
`
`dom
`
`Merck Ex. 1033, pg 1091
`
`

`
`19. M. WIGLER et al.
`
`EUCARYOTIC GENE REGULATION
`
`459
`
`1to cultured
`The process
`~notype of the
`:y to study the
`in the trans(cid:173)
`!loped trans(cid:173)
`:ion of virtu(cid:173)
`have therefore
`:s with a
`: genes: 1) to
`. significance
`2) as a
`m;
`. routes in-
`tr hybridiza- ·
`li ty and prom-
`
`mation system
`mplex virus
`·ractionations
`.ted a unique
`.f efficiently
`Wigler et al. ,
`llular genes
`ding for thy(cid:173)
`se and a metho(cid:173)
`tase to mouse
`
`genes can, in
`ditional
`n of cells
`lectable
`transforma(cid:173)
`tVe re(cid:173)
`sforming cells
`:!:.• 1 1979b) •
`3.ted when one
`!Je have used
`late mouse cell
`=r bacterio(cid:173)
`:Jbi t
`j3-globin
`)NA. We have
`1e rabbit
`Jperly recog(cid:173)
`rrnes of the
`the mature
`) HJ
`?-..:!:.· f 1979).
`
`These studies demonstrate the value of co-transformation sys(cid:173)
`tems in the analysis of eukaryotic gene expression.
`
`RESULTS
`
`Co-Transformation of Mouse Cells with ~X-174 DNA . The
`addition of the purified thymidine kinase gene from herpes
`simplex virus to mutant mouse cells lacking tk results in
`the appearance of stable transformants expressing the viral
`gene which can be selected by their ability to grow in HAT
`(Maitland and McDougall, 1977; Wigler et al., 1977). To ob(cid:173)
`tain co-transformants, cultures are exposed to the tk gene in
`the presence of vast excess of a well-defined DNA sequence
`for which hybridization probes are available. Tk+ transfor(cid:173)
`mants are isolated and scored for the co-transfer of unselec(cid:173)
`table DNA sequences by molecular hybridization.
`We initially used ~X DNA in co-transformation experiments
`with the tk gene as the selectable marker. ~X replicative
`form DNA was cleaved with Pst I, which recognizes a single
`site in the circular genome (Fig. 1) (Sanger et al., 1977).
`Purified tk gene (500 pg) was mixed with 1-10 ~g of Pst(cid:173)
`cleaved ~X replicative form DNA. This DNA was then added to
`mouse Ltk- cells using the transformation conditions previous(cid:173)
`lY described (Wigler et al., 1979a). After two weeks in se(cid:173)
`lective medium (HAT) ,~k~transformants were observed Qt a
`frequency of one colony per 106 cells per 20 pg of purified
`gene. Clones were picked and grown into mass culture.
`We then asked whether tk+ transformants contained ~X DNA
`sequences. High molecular weight DNA from the transformants
`\V'as cleaved with the restriction endonuclease Eco RI, whlch
`recognizes no sites in the ~X genome. The DNA was fraction(cid:173)
`ated by agarose gel electrophoresis and transferred to nitro(cid:173)
`cellulose filters, and these filters were then annealed with
`nick-translated 32p-~x DNA (blot hybridization) (Southern,
`1975; Botchan et al., 1976; Pellicer et al., 1978).
`These annealing experiments indicate-that 15 of 16 trans(cid:173)
`formants acquired bacteriophage sequences. Results with two
`representative clones, ~X 4 and ~X 5 are shown in Figure 2.
`Since the ~X genome is not cut with the enzyme Eco RI, the
`number of bands observed reflects the minimum number of eu(cid:173)
`karyotic DNA fragments containing information homologous to
`~X. The clones contain variable amounts of ~X sequences:
`4 of the 15 positive clones reveal only a single annealing
`fragment while others reveal at least fifty ~X-specific
`fragments.
`It should be noted that none of 15 clones picked at ran(cid:173)
`dom from neutral medium, following exposure . to tk and ~X DNA,
`
`Merck Ex. 1033, pg 1092
`
`

`
`¢x 174
`
`'.····
`
`1.3
`I
`
`~
`0,7
`
`I &I 'S7
`I&
`0.37
`
`l
`1.7
`
`3.7
`
`&I •
`
`•
`
`I
`
`I
`
`I l'i7
`
`I
`
`2 .0
`
`Figure 1. Cleavage Map of ~ tPX 174 Genome. ClEoavage
`sites for the restriction endonucleases Pst I, Hpa I
`(V), Hpa
`II (!), and Hae III (I) are shown for circular RFI and Pst !(cid:173)
`linearized ~X 174 DNA (Sanger et al., 1977). The numbers
`above the line refer to the si;es~f the internal Hpa I frag(cid:173)
`ments in kbp, while those below the line refer to the sizes of
`the Hpa II fragments.
`
`::r: ~ oeo PJ
`::r:
`I:.! Hl
`H~o '0 t-<:.: rn
`" ~
`~Pllll
`.....
`
`- ........
`
`::r: o 1>-3
`'0 I-' 11
`llJOPJ
`. ...
`
`- - - - - -- ----.. --·- - - -
`
`tTl ,-
`
`Merck Ex. 1033, pg 1093
`
`

`
`EUCARYOTIC GENE REGULATION
`
`461
`
`F G H
`
`J
`
`• •
`w •.
`• -
`·-
`..;!#IIi~ •
`--....
`• -
`-
`-.,.
`I
`--
`
`~···
`
`A B c
`
`D E
`
`(q~~~
`
`; ..
`
`....
`,.
`-..
`•
`
`q
`C\J
`
`I
`
`,._
`c5
`
`-
`
`Figure 2. Extent of Sequence Representation in ~X Co(cid:173)
`Transformants . High molecular weight DNA from co-transformant
`clones ~X 4 and ~X 5 was digested with either Eco RI, Bam HI,
`Hpa I or Hpa II and analyzed for the presence of ~X sequences
`as described.
`(Lanes B and I) 50 pg (4 gene equivalents) of
`~X RFI DNA digested with Hpa I and Hpa II, respectively.
`(Lanes A, D, E and H) 15 ~g of clone ~X 4 DNA digested with
`Hpa I, Eco RI, Bam HI and Hpa II, respectively, and analyzed
`for ~X sequences by blot hybridization.
`(Lanes C, F, G and
`J) 15 ~g of clone ~X 5 DNA digested with Hpa I, Eco RI, Bam
`HI or Hpa II, respectively.
`
`Merck Ex. 1033, pg 1094
`
`

`
`462
`
`39. M. WIGLER et a/.
`
`contain ~X information. Transformation with ~X therefore is
`restricted to a subpopulation of tk+ transformants. The addi(cid:173)
`tion of a selectable marker therefore facilitates the identi(cid:173)
`fication of co-trans formants.
`
`~X s e que nces Are Int e g rated I n t o Ce llular DNA. Cleavage
`of DNA from <l>X trans formants \oli th Eco RI (Fig . 2) generates a
`series of fragments which contain ~ X DNA sequences. These
`fragments may reflect multiple integration events. Alterna(cid:173)
`tively, these fragments could result from tandem arrays of
`complete or partial ~X sequences which are not integrated into
`cellular DNA. To distinguish between these possibilities,
`transformed cell DNA was cut with Bam HI or Eco RI, neither of
`which cleaves the <l>X genome.
`If the ~X DNA sequences were not
`integrated, neither of these enzymes would cleave the <l>X frag(cid:173)
`ments.
`Identical patterns would be generated from undigested
`DNA and from DNA cleaved with either of these enzymes.
`If
`the sequences are integrated, then Bam HI and Eco RI should
`recpgnize different sites in the flanking cellular DNA and
`generate unique restriction patterns. DNA from clones ~X 4
`and <)>X 5 was cleaved with Bam HI or Eco RI and analyzed by
`Southern hybridization (Fig. 2, clone 4, lanes D and E; clone
`5, lanes F and G).
`In each instance, the annealing pattern
`with Eco RI fragments differed from that observed with the
`Bam HI fragments. Furthermore, the profile obtained with
`undigested DNA reveals annealing only in very high molecular
`weight regions with no discrete fragments observed (data not
`shown) . Similar observations were made on clone <l>X 1 (data
`not shown). Thus, most of the <l>X sequences in these three
`clones are integrated into cellular DNA. Experiments with
`subcellular fractions demonstrate that over 95% of the <l>X se(cid:173)
`quences are localized in the high molecular weight fraction
`of nuclear DNA in transformants.
`
`Extent of Se quence Representation of the <l>X Genome. The
`annealing profiles of DNA from transformed clones digested
`with enzymes that do not cleave the ~X genome provide evidence
`that ~X sequences are integrated and allow us to estimate the
`number of <l>X sequences integrated. Annealing profiles of DNA
`from transformed clones .digested with enzymes which cleave
`within the ~X genome allow us to determine what proportion of
`the genome is present and how these sequences are arranged
`following integration. Cleavage of <l>X with the enzyme Hpa I
`generates three fragments for each integration event (see
`two "internal" fragments of 3. 7 and 1. 3 kb which
`Fig. 1) :
`together comprise 90% of the ~X genome, and one "bridge" frag(cid:173)
`ment of 0.5 kb which spans the Pst I cleavage site. The an(cid:173)
`nealing profile observed with clone <l>X 4 digested with Hpa I
`
`is
`at
`moJ
`ref
`suJ
`in
`
`mer
`clE
`bar.
`thq;
`ge~
`clc
`C).
`due
`cul
`bri
`clo
`clo
`vat
`~X
`spa
`
`the
`att
`or
`or
`Pst
`poi
`for
`fra
`amo
`ent
`wit
`Alt
`tio
`let
`int
`str
`<l>X
`dur
`the
`tra
`
`ser
`ind
`hun
`
`Merck Ex. 1033, pg 1095
`
`

`
`39. M. WIGLER el a/.
`
`!>X therefore is
`nants. The addi(cid:173)
`'l.tes the identi-
`
`~ DNA. Cleavage
`. 2) generates a
`mces. These
`mts. Alterna(cid:173)
`lem arrays of
`: integrated into
`ISSibili ties 1
`~o RI, neither of
`:quences were not
`:ave the ~X trag(cid:173)
`from undigested
`enzymes.
`If
`Eco RI should
`ular DNA and
`m clones ~X 4
`analyzed by
`D and E; clone
`aling pattern
`ved with the
`tained with
`high molecular
`rved (data not
`ne ~X 1 (data
`these three
`r imen ts with
`~ of the ~X se(cid:173)
`Lght fraction
`
`>x Genome. The
`1es digested
`>rovide evidence
`:o estimate the
`>rofiles of DNA
`rhich cleave
`tt proportion of
`tre arranged
`: enzyme Hpa I
`event (see
`.3 kb which
`"bridge" frag(cid:173)
`ite. The an(cid:173)
`ed with Hpa I
`
`EUCAR YOTIC GENE REGULATION
`
`463
`
`is shown in Fig. 2, lane A. Two intense bands are observed
`at 3.7 and 1.3 kb. A less intense series of bands of higher
`molecular weight is also observed, some of which probably
`represent ~X sequences adjacent to cellular DNA. These re(cid:173)
`sults indicate that at least 90% of the ~X genome is present
`in these cells.
`It is worth noting that the internal 1.3 kb Hpa I frag(cid:173)
`ment is bounded by an Hpa I site only 30 bp from the Pst I
`cleavage site. Comparison of the intensities of the internal
`bands with known quantities of Hpa I-cleaved ~X DNA suggests
`that this clone contains approximately 100 copies of the ~X
`genome (Fig. 2, lanes A and B). The annealing pattern of
`clone 5 DNA cleaved with Hpa I is more complex (Fig. 2, lane
`C).
`If internal fragments are present, they are markedly re(cid:173)
`duced in intensity; instead, multiple bands of varying mole(cid:173)
`cular weight are observed. The 0.5 kb Hpa I fragment which
`bridges the Pst I cleavage site is not observed in either
`clone ~X 4 or clone ~X 5. Similar analyses of DNA from these
`clones with the enzymes Hpa II and Hae III confirm our obser(cid:173)
`vation that although, in general, the "internal" fragments of
`~X are found in these transformants, "bridge" fragments which
`span the Pst I site are reduced or absent.
`The data allow us to make some preliminary statements on
`the nature of the integration intermediate. These experiments
`attempted to distinguish between the integration of a linear
`or circular intermediate.
`If either precise circularization
`or the formation of linear concatamers had occurred at the
`Pst I cleavage site, and if integration occurred at random
`points along this DNA, we would expect cleavage maps of trans(cid:173)
`formed cell DNA to mirror the circular ~X map. The bridge
`fragment, howeveT, is not observed or present in reduced
`amounts in digests of transformed cell DNA with three differ(cid:173)
`ent restriction endonucleases. The fragments observed agree
`with a model in which ~X DNA integrates as a linear molecule.
`Alternatively, it is possible that intramolecular recombina(cid:173)
`tion of ~X DNA occurs, resulting in circularization with de(cid:173)
`letions at the Pst termini (Lai and Nathans, 1974). Random
`integration of this circular molecule would generate a re(cid:173)
`striction map similar to that observed for clones ~X 4 and
`~X 5. Other more complex models of events occurring before,
`during or after integration can also be considered. Whatever
`the mode of integration, it appears that cells can be stably
`transformed without significant loss of donor DNA sequences.
`
`Stability of the Transformed Genotype. Our previous ob(cid:173)
`servations on the transfer of selectable biochemical markers
`indicate that the transformed phenotype ~emains stable for
`hundreds of generations if cells are maintained under
`
`Merck Ex. 1033, pg 1096
`
`

`
`464
`
`39. M. WIGLER et a/.
`
`EUC.
`
`selective pressure. If maintained in neutral medium, the
`transformed phenotype is lost at frequencies which range from
`<0 . 1 to as high as 30% per generation (.Wigler et al. , 1977;
`1979) . The use of transformation to study the expression of
`fore ign genes depends upon the stability of the transformed
`genotype . This is an importan.t consideration with genes for
`'"hich no selective criteria are available . We assume that the
`presence of ~X DNA in our transformants confers no selective
`advantage on the recipient cell. We therefore examined the
`stability of the 4>X genotype in the descendants of t\-10 clones
`after numerous generations in culture . Clones ~X 4 and !}iX 5,
`both containing multiple copies of 4>X DNA, were subcloned and
`six independent subclones from each clone were picked and
`grown into mass culture. DNA from one of these subclones was
`then digested with either Eco RI or Hpa I, and the annealing
`profiles of !}iX-containing fragments were compared with those
`of: the original parental clone. The annealing pattern ob(cid:173)
`served for four of the six 4>X 4 subclones is virtually identi(cid:173)
`cal to that of the parent (Fig. 3) •
`In t\..ro subclones , an
`additional Eco RI fragment appeared t..rhich is of identical
`molecular weight in both . This may have resulted from geno(cid:173)
`typic heterogeneity in the parental clone prior to subcloning .
`These data indicate that !JiX DNA is maintained \-lithin the sub(cid:173)
`clones examined for numerous generations '"i thout significant
`loss or translocation of information .
`
`Transformation of Mouse Cells with the Rabbit ~-Globin
`Gene. Transformation ~ith purified eukaryotic genes may pro(cid:173)
`vide a means for studying the expression of cloned genes in a
`heterologous host. We have therefore performed co-transfor(cid:173)
`mation experiments with the rabbit 8 major globin gene which
`was isolated from a cloned library of rabbit chromosomal DNA
`(Maniatis et al., 1978). OneS-globin clone. designated RSG-1
`(Lacy et al., 1978), consists of a 15 kb rabbit DNA fragment
`carriea-o;-the bacteriophage A cloning vector Charon 4A.
`Intact DNA from this clone (RSG-1) was mixed with the viral
`tk DNA at a molar ratio of 100:1, and tk+ transformants were
`isolated and examined for the presence of rabbit globin se(cid:173)
`quences. Cleavage of RSG-1 with the enzyme Kpn I generates
`a 4.7 kb fragment which contains the entire rabbitS-globin
`gene. This fragment was purified by gel electrophoresis and
`nick-translated to generate a probe for subsequent annealing
`experiments. The S-globin genes of mouse and rabbit are par(cid:173)
`tially homologous although we do not observe annealing of the
`rabbit S-globin probe with Kpn-cleaved mouse DNA, presumably
`because Kpn generates very large globin-specific fragments
`(Fig. 4, lanes C, D and G).
`In contrast, cleavage of rabbit
`liver DNA with Kpn I generates the expected 4.7 kb globin
`
`Tra
`<Px
`par
`eit
`
`Merck Ex. 1033, pg 1097
`
`

`
`19. M. WIGLER et a/.
`
`EUCARYOTIC GENE REGULATION
`
`465
`
`ediurn, the
`ich range from
`!.. al., 1977;
`expression of
`transformed
`ith genes for
`assume that the
`no selective
`examined the
`of two clones
`:PX 4 and <IlX 5,
`subcloned and
`picked and
`subclones was
`the annealing
`=d with those
`?attern ob(cid:173)
`r-tually identi(cid:173)
`::!lones, an
`identical
`~d from gena(cid:173)
`to subcloning.
`Lthin the sub(cid:173)
`: significant
`
`)it f3-Globin
`renes may pro(cid:173)
`led genes in a
`co-transfor(cid:173)
`_n gene which
`~omosomal DNA
`!Signated Rf3G-l
`DNA fragment
`1aron 4A.
`:h the viral
`:ormants were
`: globin se-
`I generates
`>it s- globin
`>phoresis and
`'nt annealing
`tbbi t are par(cid:173)
`tealing of the
`, , presumably
`fragments
`tge of rabbit
`kb globin
`
`A B C D E F G H
`
`J K L M N
`
`-~
`
`Figure 3. Stability of <IlX Sequences in Subclones of
`Transformants. Annealing profiles of DNA from parental clone
`<IlX 4 digested with Eco RI (lane A) and Hpa I (lane H) are com(cid:173)
`pared with DNA from six independent subclones digested with
`either Eco RI (lanes B-G) or Hpa I (lanes I-N).
`
`·.
`
`Merck Ex. 1033, pg 1098
`
`

`
`466
`
`39. M. WIGLER eta/.
`
`A 8 c
`
`-
`
`Figure 4. The Rabbit S-Globin Gene Is Present in Mouse
`DNA. Cells were expose d to RBG-1 DNA and the viral tk:gene and
`se lected in HAT. High molecular weight DNA from eight inde(cid:173)
`p e ndent c lones was dige sted with Kpn I and ele ctrophoresed on
`a
`l % agarose gel. The DNA was denatured in situ and trans(cid:173)
`ferred to nitrocellulos e filters, which w;;e then annealed
`with a 32P-labeled 4.7 kbp fragment containing the rabbit B(cid:173)
`(Lanes A and L) 50 pg of the 4.7 kbp Kpn frag(cid:173)
`globin gene.
`ment of RBG-1; (lane B) 15 ~g of rabbit liver DNA digested
`with Kpn; (lane C) 15 ~g of Ltk-aprt-DNA; (lanes D-K) 15 ~g
`of DNA f rom each of eight independently isolated tk+ trans(cid:173)
`formants.
`
`EU
`
`bat
`th<
`glc
`mat
`thE
`lat
`wh:
`loii
`t I
`
`fm
`pre
`thE
`Si:
`pr<
`In
`r e <
`gl<
`
`gl<
`RNi
`ral
`g l<
`s e <
`sh•
`44'
`on
`p s •
`s e •
`No
`pa_
`cl·
`RN
`at
`wi
`pl
`
`f o
`ge
`fl
`su
`me
`Ma
`e:x
`Rl\
`tr
`sc
`
`Merck Ex. 1033, pg 1099
`
`

`
`M . WIGLER eta/.
`
`EUCARYOTIC GENE REGULATION
`
`467
`
`J K L
`
`mt in Mouse
`- - - - -
`Ll tk gene and
`eight inde(cid:173)
`:ophoresed on
`and trans-
`l annealed
`1e rabbit (3-
`>p Kpn frag-
`L digested
`D-K) 15 )lg
`tk+ trans-
`
`band (Fig. 4, lane B). Cleavage of transformed cell DNA with
`the enzyme Kpn I generates a 4.7 kb fragment containing
`globin-specific information in six of the eight tk+ transfor(cid:173)
`mants examined. The number of rabbit globin genes present in
`these transformants is variable.
`In comparison with control
`lanes, some of the clones contain a single copy of the gene,
`while others may contain as many as 20 copies of this hetero(cid:173)
`logous gene.
`
`Rabbit (3-Globin Sequences are Transcribed in Mouse Trans(cid:173)
`formants. The co-transformation system we have developed may
`provide a functional assay for cloned eukaryotic genes if
`these genes are expressed in the heterologous recipient cell.
`Six transfur·med cell clones were therefore analyzed for t.he
`presence of rabbit (3-globin RNA sequences (Wold et al., 1979).
`In initial experiments, we performed solution hybridization
`reactions to determine the cellular concentration of rabbit
`globin transcripts in our transformants.
`A radioactive eDNA copy of purified rabbit a and (3
`qlobin mRNA was annealed with a vast excess of total cellular
`RNA from transformants under experimental condit1ons such that
`rabbit globin eDNA does not form a stable hybrid with mouse
`globin mRNA but will react completely with homologous rabbit
`sequences. A summary of these hybridization reactions is
`shown in Table 1. Total RNA from transformed clone 6 protects
`44% of the rabbit eDNA at completion, the value expected if
`only (3 gene transcripts are present. This reaction displays
`pseudo-first-order kinetics with an Rotl/2 of 2 X 103. A
`second transformant (clone 2) reacts with an Rotl/2 of 8 X 103.
`No significant hybridization was observed with total RNA pre(cid:173)
`parations from four other transformants. Further analysis of
`clone 6 demonstrates that virtually all of the rabbit {3-globin
`RNA detected in this transformant is polyadenylated and exists
`at a steady state concentration of about five copies per cell
`with greater than 90% of the sequences localized in the cyto(cid:173)
`plasm.
`
`Globin Sequences Exist as a Discrete 9S Species in Trans(cid:173)
`formed Cells.
`In rabbit erythroblast nuclei, the S-globin
`gene sequences are detected as a 145 precursor RNA which re(cid:173)
`flects transcription of two intervening sequences which are
`subsequently spliced from this molecule to generate a 95
`messenger RNA (Flavell, manuscript in preparation ; Lacy and
`Maniatis , unpubished results). Our solution hybridization
`experiments only indicate that polyadenylated rabbit globin
`RNA sequences are present in the mouse transformant .
`It \vas
`there£ore of interest to determine \~hether the globin tran(cid:173)
`scripts we detect exist as a discrete 95 species , which is
`
`Merck Ex. 1033, pg 1100
`
`

`
`468
`
`39. M. WIGLER et a/.
`
`TABLE 1
`
`Rabbit
`
`Globin Gene Transcripts in Total RNA
`
`of Ltk+ RSG-1 Transformants
`
`L-Cell
`
`Lane in
`
`~t112 of Total
`RNA and Rabbit
`
`Transcripts of
`
`Rabbit S-Globin
`
`Trans formant
`
`Figure 4
`
`Globin eDNA
`
`per Cell
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`D
`
`E
`
`F
`
`K
`
`J
`
`H
`
`> 2.0 X 10
`
`'U 8.0 X 10
`
`> 1.2 X 10
`
`> 3.0 X 10
`
`> 1.0 X 10
`
`2.0 X 10
`
`4
`
`3
`
`4
`
`4
`
`4
`
`3
`
`< 0.2
`
`'U 1.0
`
`< 0.2
`
`< 0.1
`
`< 0 . .5
`
`'U 5.0
`
`EUCA
`
`like]
`tran~
`tainj
`urea
`gel c.
`persc
`were
`rabbj
`thisj
`RNAi
`latec
`95 Rl'
`eithE
`plasn
`globi
`
`of a
`form<
`in m
`are l
`nique
`stror
`cess:
`a 95
`the (
`
`have
`EK-2
`co-tJ
`cell
`Blot
`quen•
`of p:
`or B.
`inte•
`repl
`gene
`coul•
`mati·
`
`Figu
`Ltk(cid:173)
`lect
`sere
`dono
`resi
`stre
`
`Merck Ex. 1033, pg 1101
`
`

`
`M. WIGLER et al.
`
`EUCARYOTIC GENE REGULATION
`
`469
`
`RNA
`
`.nscripts of
`
`(3-Globin
`•bi t
`Cell
`
`0.2
`
`1.0
`
`0.2
`
`0.1
`
`0.5
`
`5.0
`
`likely to reflect appropriate splicing of the rabbit gene
`transcript by the mouse fibroblast. Cytoplasmic poly A-con(cid:173)
`taining RNA from clone 6 was denatured by treatment with 6 M
`urea at 70°C, and electrophoresed on a l% acid-urea-agarose
`gel and transferred to diazotized cellulose paper (B. Seed,
`personal communication) . Following transfer, the RNA filters
`were hybridized \.,ith DNA f r om the plasmid R(3G-l containing
`rabbit (3-globin eDNA sequences (Wold et al., 1979). Using
`this 32P-labeled probe I a discrete 9Sspecies of cytoplasmic
`RNA is seen which co-migrates with rabbit globin mRNA iso(cid:173)
`lated from rabbit erythroblasts (Fig. 5). Hybridization to
`9S RNA species is not observed in parallel lanes containing
`either purified mouse 9S globin RNA or polyadenylated cyto(cid:173)
`plasmic RNA from a tk+ transformant containing no rabbit
`globin genes.
`We are unable in these experiments to detect the presence
`of a 145 precursor in nuclear RNA populations from the trans(cid:173)
`formant. This is not surprising , since the levels expected
`in nuclear RNA, given the observed cytoplasmic concentration ,
`are likely to be below the limits of detection of this tech(cid:173)
`nique. Neve.z;theless, our results 1-.'ith cytoplasmic RNA
`strongly suggest that the mouse fibroblast is capable of pro(cid:173)
`cessing a transcript of the rabbit (3-globin gene to generate
`a qs polyadenylated species \•lhich is indistinguishable from
`the (3-globin mRNA in rabbit erythroblasts.
`
`Rescue of pBR 322 DNA from Transformed Mouse Cells. We
`have extended our observations on co-transformation to the
`EK-2 approved bacterial vector, plasmid pBR 322. Using the
`co-transformation scheme outlined earlier, we have constructed
`cell lines containing multiple copies of the pBR 322 genome.
`Blot hybridization analyses indicate that the pBR 322 se(cid:173)
`quences integrate into cellular DNA "''i thout significant loss
`of plasmid DNA .
`pBR 322 DNA linearized with either Hind III
`or Barn HI, which destroy
`the tetracycline resistance gene,
`integrates into mouse DNA with retention of both the plasmid
`replication origin and the ampicillin resistance ((3-lactamase)
`gene . We therefore asked whether these plasmid sequences
`could be rescued from the mouse genome by a second transfor(cid:173)
`mation of bacterial cells .
`The experimental approach we have chosen is outlined in
`Figure 6 . Linearized pBR 322 DNA is introduced into mouse
`Ltk- cells via co-transformation using the tk gene as a se(cid:173)
`lectable marker . DNA is isolated £rom transformants and
`screened for the presence of paR 322 sequences . Since the
`donor plasmid is linearized, interrupting the tetracycline
`resistant gene , transformed cell DNA contains a linear
`stretch of plasmid DNA consisting of the replication origin
`
`Merck Ex. 1033, pg 1102
`
`

`
`470
`
`39. M. WIGLER el a/.
`
`EU
`
`A
`
`c ~· 0
`
`Figure 5. Sizing of Cytoplasmic Polyadenylated Rabbit
`Globin Transcripts from Clone ~-
`RNA was electrophoresed in
`a
`l % agarose gel in 6 M urea and the RNA transferred to dia(cid:173)
`zotized cellulose paper. The filter was hybridized with 32p(cid:173)
`labeled plasmid DNA (pSGl) containing the rabbit S-globin
`eDNA sequence. Lane A:
`2 ng of purified 98 polyadenylated
`RNA from rabbit reticulocytes, plus 25 ~g of carrier chicken
`oviduct RNA. Lane B:
`25 ~g of polyadenylated cytoplasmic
`RNA from clone 6. Lane C:
`25 ~g of polyadenylated cytoplas(cid:173)
`mic RNA from a transforrnant containing no rabbit globin genes.
`Lane D:
`2 ng of purified 98 polyadenylated RNA from mouse
`reticulocytes plus 25 ~g of carrier chicken oviduct RNA.
`
`f:
`n :
`
`Merck Ex. 1033, pg 1103
`
`

`
`~. M. WIGLER et a/.
`
`EUCARYOTIC GENE REGULATION
`
`471
`
`Rescue of pBR 322 from Transformed Mouse Cells
`
`L___...j
`
`T<•t r
`
`•
`0
`Hind ill CIPavPd pllR l22
`
`Amp'
`
`! ' """"'""" """"'' "'"'
`
`'---------'
`Ampr
`
`t
`Xb" I
`
`pllR 122 lnl<'gratPd in MOl"<' DNA
`
`C1n ulariLP \Yilh .IIP,d~l'
`Tramform E. coli X 177b
`
`• 0
`j Cl<•av<' with Xbd I
`
`mouSl' DNA in~f'rl
`
`Ampr
`
`rlated Rabbit
`~rophoresed in
`:erred to dia(cid:173)
`lized with 32p(cid:173)
`Lt f)-globin
`llyadenylated
`<rrier chicken
`cytoplasmic
`_ated cytoplas(cid:173)
`_t globin genes.
`~ from mouse
`_duct RNA.
`
`Figure 6. Scheme for the Rescue of Bacterial Plasmids
`from Transformed Cultur~Cells Using D;uble Selection Tech(cid:173)
`(For explanation,~text).
`niques .
`
`Merck Ex. 1033, pg 1104
`
`

`
`472
`
`39. M. WIGLER et a/.
`
`and the B-lactamase gene covalently linked to mouse cellular
`DNA. This DNA is cleaved with an enzyme such as Xho I, which
`does not digest the plasmid genome. The resulting fragments
`are circularized a