CA2098827A1 - Tumor susceptible non-human animals - Google Patents

Abstract

A desired non-human animal or an animal cell or human cell which contains a predefined, specific and desired alteration in at least one of its two p53 chromosomal alleles, such that at least one of these alleles contains a mutation which alters the expression of the allele, and the other of the alleles expresses either a normal p53 gene product, or comprises an identical or different p53 mutation.

Description

WO 92/1 1~74 PCI/USg2/00295 TITL~ ~ E-lE~ENTION:

TW~OR s~scEæTIBL~ NO~-~n~aN a~o}oL~s FI~L~ OF ~E~ l~V~lIQe:

The in~ention is directed toward tu~or-susceptible non-human animals. The invention further pertains to the use of such animals in the de~elopment of anti-cancer agents and therapies.

:
This application is a continuatio`n-in-part of U.S.
Patent Application Serial No. 07/637,563 (filed January 4, 1991) . ' ' ' _ OF T~E IhVP~ION:

I. ~himeric and Tran~g~nic ~ni~al~ .

: Recent advances :in recombinant DNA~ and genetic technologies hav- made it pos~ibl~ to introduce and express a ~esired gene seguen~e in a r~cipient animal. Through the u e of such methods, animal~have been engineered to contain ~ene aequences that are not normally or naturally present in an unaltered animal. ~h~ techniques have al80 been used to produce animals which ~xhibit altared ~expression of aturally pres~nt gene sequencee.

:: :

~ ~ SuBs~lTuTE:;~sHEET

W092/11874 PC~/US92/0~295

2 O 9 ~ 8 ~

, : -2-The animals pr ~ a ~hrough the use of these methods are known as eith~r "chimeric" or "transgenic" animals. In a "chimeric" animal, only some of the animal's cells contain and express the introduced gene s2quence, whereas other cells have been unaltered. The capacity of a chimeric animal to transmit the introduced gene sequence to its progeny depends upon whether the introduced gene sequences are present in the germ cells of the animal. Thus, only certain chimeric animals can pass along the desired gene sequence to their progeny.
In contrast, a}l o~ the cells o~ a ~'transg~nic" animal :~
: contain the introduced gene sequence. Con~equently, every transgenic ani~al i8 capable o~ transmitting ~e introduced - .
gene sequence to it~ progeny.
: :
II. Production of TransgQnic An;~ls:
~iGroinjection ~ethods The most widely used~ method through which transgenic animals have been produced involves injecting a DNA molecule into the male pronucleus of a~fertilized egg (Brin~ter, R.L.
~iaL , Cell ~7:223 (1981); Costanti~i, F~ al., ~3~YE~

3~:92~ (1981); Harbers, K. et 3Ll., Na~re 293 :540 (1981);
: ~ Wagner, E.F. t al., Proc. ~L ~çad. Sci. (U.$.~.~ 78:5016 25 (1981); Gordon, J.W. et al,, ~prQç~ Ç~,d.. S~ ~S,A.) 73:1260 (1976); Stewart, T.A. et al., Sçienc~ 17:1046-1048 (~982); Palmiter, R.D. e~ al., Scienc~e ~:809 (1983);
Evans, R.M et :~1. (U.S. Patent 4~,870,009)).
The gene~ ~ quence b~ing in~roduced need not be incor-30 porated into any~Xind~ of self-replicatirlg plasmid or virus SUBS~ETUTE SHE~ ~
~ . -: . .

. ~ . . , . . . . , .. , . . . . . . . ~ . . .

4 PCT/US92/00295 ~9~:8~7 (Jaenisch, R., Science, 240:1468-1474 (1988)). Indeed, the presence of vector DNA has been found, in many cases, to be undesirable (Hammer, R.E. et ~1., Science 235:53 (1987);
Chada, K. et al., Na~u~ 31g:685 (1986); Kollias, G. et al., Cell 46:89 (1986); Shani, M., :9l~C.~ i2l. 6:2624 (1986); Chada, ~. et al., Nat~ C377 (1985); Townes, T.
et al., EM~.O J. 4:1715 (1985)).
~fter being injected into the recipient fertilized egq, the DNA molecules ara believed to recombine with one another to form extended head-to-tail concatemers. It has been proposed that such concate~ers occur at sites ~of double-stranded DNA breaks at random sites in the egg's chromoso~es, and that the concate~ers are inserted and integrated into such sites (~rinster, R.L. et ~1,, Pro~.
: 15 Na~l Aca~. Sci. tU.S.. A,~ 82:4438 ~1985)). Although it is, - thus, possible for the injected DNA molecules to be incorporated at ~everal sites within the chromcsomes of the fertilized egg, in most instances, only a single site of insertion is observed (Jaenisch, R., S~ien$e,~240:1468-1474 (19~8)). ~ :
once the: D~A molecule~: has :been injected into the ~ fertilized egg cell,~thc c`ell is implanted into the uterus~
; o~ a recipient femàle,:~and allowed to::develop into an : animal. Since all of the animal's cells are derived from the implanted ferti}ized egg, all of the cells of the resulting animal tinclud~ng the g~rm line cells) shall .
contain the: introduced~gane 8equence. If, as occurs in :: :
about 30% of~ events, the ~irst cellular division occurs before the introduced~gen- sequence has integrated into the llB~lTUTE~S~ ET~

~ WO9~11874 PCTt~5g2/00~95 2~)9~27 _4~

cell's genome, the resulting animal will be a chimeric animal.
By breeding a~d inbreeding ~uch animals, it has been . possible to produce heterozygous and homoxygous transgenie animals. Despite any unpredictability in the ~ormation of such transgenic animals, the animals have generally been found to be stable, and to be capable of producing o~spring which retain and express the introduced gene sequence.
Since microinjection causas the injected DNA to be incorporated into the genome of the fertilized egg through a process involving the disruption an,d alteration of the nucleotide sequence in the chromo~ome o~ the egg at the insertion site, it has been observed to result in the alteration; disruption, or los~ of function of the endogenous egg gene in which ~he injected DNA is inserted.
Moreover, substantial alterations (deletions, duplications, rearrangements, and translocations) o~ the endogenous egg seguences flanking the inserted DNA have been observed (Mahon, K.A. ~al~, PrQ ~ U.S.A.~ ~5:1165 (1988); Covarrubias, Y. ~ aL~, PrO~. NatL~ Acad. sci.
83:6020 (1986j; ~ark, W~ et ~l., Cold Sp~. Harb.
: ~ ~ 50:453 (1985)). Indeed, lethal mutations or gross :morphologica} abno~malities have been observad (~aenisch, R., ~cience 240:1468-1474 (1988); First, N.L. et ~l " ~er. ~At Sçi~ssn~ th Re~i~rs~l Me~Con~ 39:41 : (1986)))-Significantly, it has been ob~erved that even if the~
desired gene sequence of the microinjected DN~ ~olecule is one that is naturally ~ound in the recipient egg's genome, integ~ation of the desired~gene sequence rarely occurs at :

SUBSTITUTE SHEET
:: :

W092/11~74 PCT/US92/00295 ~ 20~g8~27 :
:A' .~) 'I' the site of the natural gene (Brinster, R.L. et al , Pr~ç.
Natl. Acad. Sci. ~U.S~A~) 86:7087-7091 ~1989)). Moreover, introduction of the desired gene sequence does not generally alter the sequence of the originally present egg gene. .
Although the site in the ~ertilized egg's geno~e into which the inject~d DNA ultimately integrates cannot be predetermined, it is possible to control the expression of the desired gene sequence such that, in the animal, expression of the sequence will occur in an orqan or tissue specific manner (reviewed by We6tphal, H., ~g~ 3:117 (1989); Jaenisoh, R., ~ 240:1468-1474 (1988); Msade, H. t al. (U.S. Patent 4,873,316)).
The SUCCeS5 rate for producing transgenic animals is greatest in mice. Approximately 25% of fertilized mouse eggs into which DNA has been injected, and which have been i~planted in a ~emale, will becom2 ~ransgenic mice. A lower rate has been thus far achieved with rabbits, sheep, cattle, ~;
and pigs (Jaenisch, R.,: Scie~çe ~ 146~-1474 (1988);
Hammer, R.E. et ~L, ~ L~~ 63:269 (1986); Hammer,~
R.E. et al., ~e 31S:680 (1985~; Wagner, ~.E. et al., Thç~io~enQlo~y ~1:29. (1984)). ~ The lower~rate~ may reflect greater familiarity~wi~h~the ~ouse as~a genetic system, or ~:
may reflect the~ difficulty o~ visualizing the male pronucleus :of ~he fertilized eggs of many farm animals : ~: 25 (Wagner, T.E. e~ al~, Theri~enoloov ~1:29 (1984)).
Thus, the production o~: transgenic animals by ::
microinjection o~ DNA su~fers ~rom at least two major drawbacks. First, it can be accomplished only during the . .
single-cell stage of an animal's life. Second~ i~ requires 30 ~ the~disruption~of~the:natural se~uence Or the DNA, and thus: .~;
- :-.
:.

SU~3$TITUTE SHET

WO~2~11874 PCT~US92/00295 2~98~7 is often mutagenic sr teratogenic (Gridley, T. et al., Trenqs Genet. 3:162 (1987)).

III. Production of Chi~OEiC and ~ransgenic Ani~alæ:
Reco~binant Y~ral ~nd Re~roviral ~e~h~ds Chimeric and transgenic animals may also be produced using recombinant viral or retroviral techniques in which the gene sequence is introduced into an animal at a multi-ce}l stage. In such methods, the desired gene sequence is introduced into a virus or retrovirus. Cells which are in~ected with the virus acquire ~he introduced gene sequence. If the virus or retroviru~ infects every cell of the animal, then the ~ethod r2sults in the production of a transgenic ani~al. If, however, the virus infects only some of the animal's cells, then a chimeric animal is produced.
The general advantage of viral or retroviral methods o~
producing transgenic animals over thos2 methods which involve the ~icroinjection of non-replicating DNA, is that it is not necessary to perform the genetic ~anipulations at a single cell st~ge. ~oreover, infQction is a highly effi~iant means for introducing the DNA into a desired cell.
Raco~binant retroviral ~ethods for producing chimeric or transgenic animals have the advantage that retroviruses integrate into a hos~' genome in a precise manner, resulting generally in the presence of only a single integrated r~troviru~ (although multiple insertions may occur). Rearrangement~ o~ the host chromosome at the site of integratio~ are, in gen~ral, li~ited to minor ; 30 duplication~ (4-6 base pairs) o~ host DN~ at the ho~t virus : SlJB5TITUTE SHE~E T ~

W09Vll874 P~r/US92/00295 ~3 ~ ~ 7 junctions (Jaenisch, R., sci~nce 240:1468-1474 (1988); see also, Varmus, H., In: ~NA Tu~r Viru~es (Weiss, R. çt al., Eds.), Cold Spring Harbor Press, Cold Spring Harbor, NY, pp.
369-512 (}982)). The method is, ho~ever, as mutagenic as microinjection methods.
Chimeric animal~ have, for example, bsen produced by incorporating a desired gene sequence into a virus (such as bovine papilloma virus or polyoma~ which is capable of infecting the cells o~ a host animal. Upon infection, the virus can be maintained in an in~ected cell as an extrachromo~omal episome (Elbrecht, ~. et al-, ~QlÇs_5C~
~iQl~ 7:1276 t~987); Lacey, M. ~_~al~, ~3~E~ 609 (19861; Leopold, P. ~ ÇÇ11 51:8~5 (1987~). Although this method decrsa e8 ~he mutagenic nature of 15 chimeric/tran~genic animal for~ation, it does so by ~:
decreasing germ line stability, and increa~ing oncogenicity.
Pluripotsnt embryonic stem cells (referred to as "ES"
cells) are cells which may be obtained from embryos until the early post-implantation stage of embryogenesi~. The :.:
20 cells may be propagated in culture, and are able to .
dif~erentiate either in vitro or in ~ivo upon implan~ation into a mouse as a tu~or. ES cells have-a nor~al karyotype - ~Evans, ~.J. ~ $~ ~ :154-156 (1981~; Martin, G.R.
~ ~. Na~l A~::ad. S~i . ._~A,~ 78: 7634-7638 25 (1981) ) . ~

Upon injection into a bla~tocyst of a developing embryo, :~ :
ES cells will proliferate and di~ferentiate, thus r2~ulting in the production of a chimeric ani~al. ES cells are capable of colonizing both the so~atic and gçrm-line -.
lineages of such a chimeric animal ~Robært80n, Æ~
: . - -,:

::. :
. .
. ,: .
.
~: - ..

: SUBSTITlJTE S~IEET
.

.

W0~2~11874 P~T/~S92/~02~5 2098~27 Cold sDrin~ Harb. Conf. Cell Prolif. lO:647-663 (~83);
Bradley A~ et al., ~ure 309:255-256 ~l984); Bradley, A. et al., Curr. To~. Devel. Bio~ ~0:357-371 (1986); Wagner, E.F.
et al., Co~d S Ei~ Ek $y~p. Ouant._ Biol 50:691-700 (19B5); (all o~ which references are incorporatad herein by reference).
In this method, ES cells are cultured in vitFo, and infected with a viral or re~roviral vec~or containing the gene sequence of interest. Chi~ric animals generated with retroviral vectors have been found to have germ cells which either lack the introduced gsne sequence, or contain the introduced sequence but lack the ~apacity to produce progeny cells capable of expre~sing the introduced sequence (Evans, M.J. et al.l Col~ sli~c~ $Ymp. Q~nt. ~io~. s0:685-689 (1985); Stewart, C.L. e~ al., EMBO ~. 4:3701-3709 (1985);
Robertson, L. et~ , N~ (1986); which references are incorporated herein by reference). . -Because ES cell6 may be propagated in vitro, it is possible to manipulate such cells using the techniques of somatic cell genetics. Thus,~it is possible to'select ES
cells which carry mu~ations (such as in the ~rt gene :
(encoding hypoxanthine pho~phoribosyl transferase) (Hooper!
. çt al., ~ X~ 326:292-295 (1987); Kuehn, M.~. ç~iaL~, N~t~re ~ 295-298 (1987)). Such selected cells can then be u~ed to produce chimeric or tran~genic mice which fail to express an active HPRT enzyme, and thu8 provide animal models for diseases (such as the Lesch-Nyhan syndrome which is characterized by an HPRT ~eficiency) (Doetschman, T. et ~1~, ~ 85:8583-~587 (1988)).

`~

n~l~ ~FFT --Wos~11874 PCT/US92/00~95 ~ 2`09~27 ~ s indicated above, it i~ possible to generate a transgenic animal from a chimeric animal (whose germ line cells contain the introduced gene sequence) by inbreeding.
The above-described methods permit one to screen ~or the desired genetic alteration prior to introducing the trans-fected ES cells into ~he blas~ocyst. One drawbac~ o~ these methods, however, is the inability to control the site or nature of the in~egration of the vector.

IV. Production of Chi~eric and Transgenic AnL~als~
Plas~id ~ethoas The inherent drawbacks of the above-described methods for producing chimeric and transgenic ani~als have caused researchers to attemp~ to identify additional methods through which such ani~al could be produced.
Gossler, A. et 31~, for example, have described the use of a plasmid vector which had been modified to contain the gene for neomycin phosphotransferase: (nE~ gene) to transfect ES cells in culture. The presence of the rE~
gene conferred resistance to the an~ibiotic G418 to ES cells that had been infected by the plasmid ~Gossler, A. e.~
83:9065-9069 ~1986), which reference is incorporated herein by reference). The chimeric animals which recei~ved the pla~mid and which beaa~e resistant to G~18, were found to have integrated the vector into their chromosomes.
Takahashi, Y. et al. have described the use of a plasmid to produce chimeric mice cells which ~xpressed an avian crystallin gene ~ ev-lcp~ent 10~:258-26~ ~1i88), incor-~ `

C` l: l n ~TlTt lTI~

W092t11874 PCT/U592/00295 2 0 9 ~ g ~ 7 porated herein by reference). The avian gene w~s incorporated into a plasmid which contained the n~I gene.
Resulting chimeric animals were found to express the avian gene.
V. Introductio~ of Gene ~eQUenCes into So~atic Cells '' ':
DNA has been introduced into so~atic cells to produce variant (chimeric) cell line~. hprt-deficient Chinese hamst~r ovary (CH0) cells have been transformed with the CH0 haE~ gene in order to produce a prototrophic cell line (Gra~, L.H. et ~ Som~ Cell G~net. 5:1031-1044 (1979)).
Folger et a~Y examined the fate of a thymidine kinase gene (~ gene) which had been microinjected into the nuclei of cultured mammalian cells. Recipient ce ls were found to contain from 1 to 100 copie~ of the introduced gene sequence integrated as concatemers at one or a few sites in the cellular genome (Folger, K.R. 9~.~aL~, Molec,_~ell. BlQl.
; ~ 2:1372-1387 (1982)). DNA-mediated transformation of an RNA
polymerase II gene into;Syrian~ ha~ter cells~has also been reported (Ingles, C. et ~l~, Molec~ Cel~ ~ ~ 2:666-673 (1982)). ~ ~
Plasmida~ conferring~ host neomycin reslstance and guanosine phosphotrans~erase ac~ivity have been tran fected into Chinese hamster ovary cell~ to qenerate novel cell lines (Robson, C.N. et al., ~at. Rçs. ~ 201-208 (1986j).

.:

:
:. : : ~ ~ : ~ .

slll3sTlTuTllsHEET ~ ~ ~

WO92/~1~74 PCT/USg2/00295 2~%~27 VI. Oncogene~ and tu~or supprefi~or gene~

O~e mechanism through which cancer may axise is through a cell's exposure to a carcinoqenic agent, either chemical or radiation. Such exposure may da~age the DNA sequence of critical genes present in the genome of a cell of an animal.
If this damage leads to either an i~pairment in the expression of the gene, or in the production of a mutant gene product, the cell may then proceed to proliferate, and ultimately result in the ~ormation o~ a tumor.
O~e class of such critical genes has been referred to as l'oncogenes." Onco~ene~ are gene6 which are naturally in an "inactivated'l state, but which, throu~h the ef~ect of the DNA damage are converted ~o an "activated" state capable of inducing tumorigenesis (i.e. tumor formation). Oncogenes have been identified in 15-20~ of human tumors. The products o~ oncogene~ ("oncoproteins") can be divided into two broad classes according to their location in the cell.
Oncogene products which a~t in the cytoplasm of celIs have readily identifiable biochemical or: biological : .activities ~Green, ~.R., C ll :1-3 (1989)). Those that act iD the nucleus of a: cell have been:~ore di~ficult to c~aracterize. Some nuclear oncoproteins (such as ElA and : . ~y~) hava transcriptional regulatory activity, and are balieved to ~ediate ~ eir activi~ies by the transcriptional activa~ion of celIular genes ~King~ton, ~.E., Ç~ll 41:3-5 ~1985)). Other ~nuclear: o~coproteins appear to have a complex array o~ activities (such a~ DNA binding activity, ability to initiate viral DNA synthesis, ATPase activity, ~ ~ :

St.113STlTUTE SHEET

''` ~`~

20'gg'8~27 helicase activity, and transcriptional regulatory activity) (Green, M.R., Cell 56:1-3 (1989)).
The creation of a mutant oncogene is only one of the requirements needed for tumor ~ormation; tumorigenesis appears to also require the additional inactivakion of a second class of critical genes: the "amti-oncogenes" or "tumor-suppressing genes." In their natural state these genes act to suppress cell proliferation. Damage to such genes leadis to a loss of this suppression, and thereby results in tumorigenesis. Thus, th~ deregulation of cell growth may be mediated by either the activation of oncogenes or the inactivation o~ tumor-suppre~sing genes (Weinberg, R~A-~ ~gl~ L~ C~, Sept. 1988, pp ~4-51).~
Oncogenes and tumar-suppres~ing genes have a basic distinguishing feature. The oncogenes identified thus far have arisen only in somatic c~11i5 ~ and thus have been incapable o~ transmitting their effects to the germ line of ~ ~ the host animal. In contrast, mutations in tumor-:: sup~ressing genes can be identified in ger~ line cells, and are thus ~ransmissible to an:ani~al's progeny.
The classic example o~ a . hereditary cancer is retinoblastomas~: in ~childr~n. The incidence of retinoblastomas is determined by~a t~mor suppres~or ~ene, the retinoblastoma (RB) gene~(Weinberg,~ R.A.,. 9Sah~L~ Q
~ç~, Sept.~ 19~8, pp 44-51); Han~en M.F. et al., Trends Genet. 4:125-128 (19~8)). Individuals born with a lesion in on~ o~ the RB all~les are predispoaed to early childhood development of retinob}astomas (W~inberg, R.A.~ Scienti$ic A~er., Sept. 1988, pp 44-51); Hansen M.F. et~aL , Trends : 30 :~ Genet. ~:125-128~(19~8)). Inactivation or mutation of lthe :
SUBSTITUTE: SHEET

WO9~/llX74 PCTtU~2/0~295 ~ 2 ~ , g ~
.; , .

second RB allele in one of the somatic cells of these susceptible individuals appears to be the molecular event that leads to tumor formation (Cavenee, WoK~ et al., Nature 305:779-784 (1983); Friend, S.H. et al., Proc. Nat'l. Acad.
Sci. (U.S.~,) 84:9059-9063 (1987)).
The RB tumor-suppressing gene has been localized onto human chromosome 13. The mutation ma~ be readily transmitted through the germ line of a~flicted individuals (Cavenee, W.K. ~ l., New ~nal. JO Med.. 314:1201-1207 (1986)~. Individuals who have mu~ations in only one of the two naturally present ~lleles of this tumor-suppressing gene are predisposed to retinoblastoma. Inactivation of the s~cond of the two allele~ is, howaver, required for tumorigenesis (~nudson, A7G., PXQC. Nat'l. Acad~_Sci.
!u.s~A. L 68:820-823 (1971)).
A second tu~or-suppressing gene is the p53 gene (Green, M.R., Cell S6:1-3 (1989); Mow~t et al., ~ E~ 31~:633 636 (1985)). The protein encoded by the p53 gene is a nuclear protein that forms a stable complax with both the~SV40 large T antigen and the adenovirus ElB 55 kd protein. The p53 gene product ~ay be inactivated by binding ~o these proteins.
Initially, the p53 gene was thought to be an oncogene : rather than a tumor~suppressing gene since it is capable of i~mortalizing primary rodent cells and can cooperata with the E~ oncogene to cau~e transformation. Subsequent research revealed that the p53 genes used in those early experiments was a mutant allele of the normal p53 gene (Green, M.R., Ç~ll 56:1-3 (1989)). Thus, the p53 gene is a ~umor-suppressing gene rather than an oncogene.
.

SUBSTITUTE SHEET .;

' wos2/l1874 pcT~us92/oo29s 2098~2''i ,', Mutations at any of a large number of positions in the p53 gene can result in the activation of the transforming potential of the p53 gene product (Eliyahu et al., N3~ure 312:646-649 (1984); Finlay ~ , Molec. ~ell. Bi~l, 8:531-539 (1988)). This has sugge~ted that the activation of the p53 transforming activity is due to the inactivation o~ the normal p53 activity (Green, M.R., Cell 5~ 3 (1989)).
The p53 gene has been implicated as having a role in colorectal carcinoma (~aker, S.J. ~ iÇnÇ~ 244:217-221 (1989)~. studies had shown that allelic deletions of the short arm of chromosome 17 occurred in over 75% of colorecta~ carcinomas. The re~ion dele~ed was subseq~ently found to enco~pass the p53 gene loc~s (Baker, S.J. et ~l., Science ~:217 221 (1989)). The deletion of the region was found to mark a transition from a (benign) adenocarcinoma stage to a (malignant) carcinomatous stage (Vogelstein, B.
et al., New ~nal. J. M~d. 319:525 (1988)).
Similar deletions in chromosome 17 have been identified in a wide variety ~o~ canc-rs includi~g breast and lung : ~ 20 cancers (Mackay, J. et al~ ançet~ 3~4~(1988); James, : C.D.~et alO t Canc. Res.;~ 5546:(~1988); Yakota, J. et.al., Proc. Nat'l.:~cad. Sci. rU.~ ) 84:9252 (1987); Toguchida e__al~, Canc,~ R~ 8 3~939~1983)).~ In addition to p53 ~ allele loss, Nigro et al. ~ 705-708 ~I9~9)) have : 25 demonstrated that ~the single remaining p~3 allele in a variety of human t ~ rs (brain, colon, brea~t, lung) undergo a point mutation which renders it tumorigenic. Fearon et al- (ÇÇlL 61:759-767 (1990)) have hypothesized that both point~ mutations and d~letion in the p53 alleles may be :~ 30 required for a~ully:tumorigenic phenotype. These:findings .
- ~

- -SUE~STIT~UTE~SHEET ~

WO 92/11874 PCI'/US92/00295 ~4~`~
2 0.
, --~ 5--suggest that the p53 gene may have a role in many types of cancers.

VII. Conclusions The application of the above-described technologies has the potential to produce animals which cannot be produced throuqh classical genetics~ For example, animals can be produced which suffer ~rom hu~an diseases (such as AIDS, lo diabetes, cancer, etc.), and may be valuable in elucidating therapies for such diseasQs. Chimeric and transgenic animals have substantial use as probes of natural gene expression.
Leder, P. et al. (U.S. Pa~ent 4,736,866) di~close the produ~tion of transgenic non-human mammals which contain cells having an eXOgQnOU~ly added activated oncogene se~uence. Although the animals are disclosed as being useful for assaying for carcinogenic materials, the precise lo~ation and structure of the added oncogene sequence in the animals is unknown, and cannot be experimentally controlled.
: Thus, the value of the animals as a model for oncogenesis is significantly impaired.
Despite the successes o~ the~above-described techniques, the methods hav~ not yet led to the development o~ a model tran~genic animal which can be u~ed to study ~he conditions responsible for the initiatio~ of naoplasia, and which can be used as a ~eans ~or developing suitab}e antineoplastic agents and therapies. Indeed, prior to the present inven~ion, research on transgenic or chimeric animals 30 containing mutations in oncogene and ~critical~ genes had SUBST`ITUTESHE~:ET : ~

~098~7 suggested that it would not be possible to produce viable animals containing mutations in the chromosomal alleles of their tumor-suppressing genes. It was believed that such animals would be non-viable, or would not survive to maturity. See, Soriano, P. ~_hl~l Ç~ll 64:693-702 (1991) [c-src]; McMahon, A. and Bradle~, B., Ç~ 1073-10~5 (1990) [Wnt-l]; Koller, B. et ~1~, Science 248:1227-}230 (19~0) [MHÇ-I]; Tybulewicz, V.L.3., Cell 65:1153-1164 (1991) [c-abl]; Mucenski, M.L., .Cell 65:677-690 (1991) ~c-myb].
If however, animals predisposed to cancer could be obtai~ed, they would ~acilitate a better understanding of cancer; they could be used ~o assay ~or the presence o~
cancer-causing agents in food, wa~te products, etc.; they could also be us~d to identify agents capable of suppressing or preventing cancer. Such animals would, therefore, be extremely de~irable. The present invention provides such animals, and the methods ~o produce and use them.

Figure 1 shows the modifications made to a 3.7- kb fragment spanning exons 2 ~o 10 o~ the 11 eYon p53 gene in order to ~aci}itate~the construction of the transgenic anLma~s o f ~he present invention.
~: 25 Figure ~ shows the structure o~ the targeting construct used ~to succeasfully inactivate the p53 gene in mouse embryonic ~tem cells. The arrow~ labelled 5' and 3~
repre~ent the oligonucleotide primers used to screen the re~istant tem c~ll colonies by PCR. ~ represents the ~: ~ 30 hybridization probe for the~PCR fragment.

~ ~tT1IT~ C~FE~ :
:

WO92/lt874 PCT/U~92/00295 2 0 9 8;~

Figure 3 shows a comparison Southern ~lot o~ the p53 alleles present in two of the constructed embryonic stem cells (designated 1 and 2) and in wild type (wt) cells.
Figure 4 shows a Southern blot analysis of tail DNA of the progeny of the F1rl~sxC57BL/6~ chimeric mice. The analysis reveals that mou~e pup pl and p3 contain DNA which hybridizes to the neo gene, and thus contains the mutated p53 construrt~ "+" indicates positive control (i.e.
embryonic stem cell with disrupted p53 gene); "-" indicates negative control ~i.e. normal mouse DNA).
Figure 5 shows the procedures which are used to produce any desired mutation into a p53 allele of a cel}.
Figure 6A shows the location and orientation o~ the primers used in the PCR amplification of the p53 mRNA.
Figure 6B shows ~he results of the PCR amplification. The gel shows 19 lanes, of which the last ~M) contains molecular ~ize markers. The initial 18 lanes d~pict the amplification obtained in normal mice (++), p53 heterozygote mice (+~
and p53 homozygote mice (--), respectively, with each of six dif~erent pairs o~ PCR primers (1 and 4a; 7 and 10; B-actin ; control; 4b and 10; 1~and 10; and 4b an~ 6). The appearance o a band ,indicat~s that: the primer pairs ampli~ied a :~ segm-nt of a~n mRNA species present in the preparation.
.
G~- IY~L ~L '~ 7Ir~
...... ................................................................... ........ .:.
The present invention provides a desired non-human animal or an animal (including human) cell which contains a ~
predefined,: specific~ and~desired:alteration xendaring~the .~.
3 a ~ non-human ~animal or animal celI predispo~ed to cancer.
: , . . .
. :
, .

~1 ID~TIT1 tTI~ * LI~T :::
,: :: -WO 921tl874 Pcr/vs92~oo295 g~

Specifically, the invention pertains to a genetically altered non-human animal ~most pre~erably, a mouse), or a cell (either non-human animal or human) in culture, that is de~ective in at least on~ of two alleles of a tumor-suppressor gene such as the p53 gene, the rb gene/ etc., andmos~ pre~erably, ~he p53 gene. The inactivation of at least one of these tumor suppressor alleles results in an animal with a high~r susceptlbility to tu~or induction. A
genetically altered mouse of ~his type is able to serve as lo a useful model for hereditary cancers and as a test animal for carcinogen studies. The invention additionally psrtains to the use of such non-human animals or animal cells, and their progeny in research and medicine.
In detail, the invention provides a transgenic or chimeric animal cell who~e genome comprises two chromosomal alleles of a tu~or-~uppres ing gene ~especially, the p53 gene), wherein at least one of the two alleles contains a mutation, or the proqeny of this cell.
The invention includes the embodiment of the above --~
animal cell, wherein one o~ the alleles expresses a normal tu~or-suppressing gene product.
The invention includes the e~bodiments whérein the above : ani~al cells are hu~an cells, or the cells of a non-human animal. The invention includes the embodiment wherein the 25 cell is an embryonic ~te~ cell, and in particular, wherein . .
the tumor-suppres~ing gene is a p53 gene~ and the embryonic stem cel} is ATCC CRL 10631.
The invention also includes a non-human ~ransgenic or chimeric animal having an animal cell whose genome comprises -:
~: 3~ ~wo ~hro~osomal alleles Or a tumor-suppressing gene, wherein . .:

- : : ..

SUBSTITUTE SHEET

:

W092t1187~ P~T/US92/0~295 . .
2 ~ 7 at least one of the two alleles contains a mutation, or the progeny of the animal, or an ancestor of the animal, at an embryonic stage (preferably the one-cell, or fertilized oocyte stage, and generally, not later than about the 8-cell stage).
The invention also includes the embodi~ent wherein the tumor suppressing gene of the non-human animal is a p53 gene.
T~e invention is also direated to the embodiments wherein the animal cell of the non-human animal is a human cell, or-a cell o~ a non-human animal (of either the same species as the non-human animal or a different species).
The invention is also directed to the embodiments wherein the animal cell of the non-human animal is a germ- :
line cell, or a somatic cell.
The invention is also directed to the embodiment wherein the animal cell of the non-hu~an animal i5 an embryonic stem cell (especially the embryonic stem cell, ATCC CRL 10631). ~ .
: The invention additionally provides a method for identi~ying the presence of an agent suspected of being capable o~ affecting a characteristic of an animal cell that is attributabIe to the presence or expression of a tumor-suppressing gene, the method comprising:
A) administering an amount of the agent to an animal 25 cell in cell culture, the cell having a genome that ~ ~:
comprises two chromoso~al alleles of the tumor-suppressing gene, wherein at least one of the two allele~ contains a mutation; :
. B) :maintaining the cell culture for a desired period of time after the administration;
.
.

STi~ E SI~EET

WO~2/~lX74 PCT/US92/00295 ~ J

c) ~etermining whether the administration of the agent has affected a characteristic of the animal cell that is attributable ~o the presence or expression of ~he alleles of the tumor-suppressing gene.
In particular, the invention include~ the embodiment of the above method wherein the tumor-suppressing gene is a p53 gene. .
The invention also includes ~he embodiments of this method wherein the ~gen~ is su~pected of ~being able to increase a tumorigenic potential of the animal cell, wherein the agent is su3pected of being able to decrease a ~umorigenic potential of ~he animal cell, wherein the animal cell is a human cell, and wherein the animal cell is a non-human animal cell (such as an em~ryonic s~em cell, and in particular, the embryonic stem cell, ~TCC CRL 10631).
The invention also provides a method for identifying the presence of an agent suspected of being capable of affecting : -a characteristic of an animal cell that is attributable to the presence or expression of a tumor-~uppressing gene, the method comprising~
A~ admi~nistering an~ amount of the ag~nt to an animal, the animal having a cell who-se genome comprises two chromosomal~allel:es~of the tumor-suppressing gene, wherein : at least one o~ the two alleles contains a mutation;
B) maintaining the animal for a desired period of time after the administration; ..
C) determining w~ether the administration of the ; agent has affécted~ a~characteristic of the cell that is attributable~to the presence oF expression of the alleles of : 30 the tumor-suppréssing gene.

~: ~
:~ ~ : SllB~;TlTliTE ~S~EET

WO9~/11874 PCT/U~92/00295 ~ `'~
.~"J~

~ .: . . .

In particular, the invention includes the embodiments of the above method wherein the tumor-suppressing gene is a p53 gene.
The invention also includes the embodiments of this method wherein the agent is ~u~pected of being able to increase a tumorigenic potential oP the animal cell, wherein the agent is suspected of ~eing able to decrease a tumorigenic potential of the animal cell, wherein the animal cell is a human cell, and wh~rein the animal cell is a non~
human animal cell (such Z8 an embry~onic stem cell, and in particular, thQ e~bryonic ~tem cell, ATCC CRL 10631).
The invention also includes the embodiments o~ this method wherein the ani~al and the animal cell are of the same species, of different species.
15The invention further provides a method of gene therapy comprising altering the genome of a cell of either a human or a non-human animal, wherein the cel} has a genome that comprises two chromoso~al all~le~ of a tumor-suppressing gene, and wherein at l~ast one o~ the two alleles contains a mutation, to thereby form a cell wherein at last one of the alleles e~presses a normal tumor-suppressing gene product.
~ ~ In particular, the invention includes the embodiments of : the above method of gene ~herapy wherein the tumor-suppressing g~ne is a p53 gene.

D~S~IP~ION OF qH~ PF~rl~lU~L B~BOD ~ S:

~: ~As is well known, the cells of humanc and animals 30. (especially, rodents (i.e. mouse, rat, hamster, etc.), ` SU~BSTITUTI~SHEET

~ . , w092/tl874 PCTtVS92/0~29s 2 ~i~ 8 8 2 7 rabbits, shaep, goat6, f ish, pigs, cattle and non-human primates3 are "diploid" cells, and thus naturally contain two copies ("alleles") of each and every gene of their genome. A cell~s "geno~e" consists of all of its heritable DNA (either chromosomal or non-chromosomal (i.e. episomal, viral, etc.). one of the two alleles of a gene is provided by the animal~s or cell~s maternal parent; ~he other se~ is provided by its paternal parent. ~he diploid nature of human and animal cells is described by DeRobertis, E.D.~., et al. (~ell ~ioloov, 6th Ed., W.B. Saunders company, Philadelphia, tls7s))~ and in other ~imilar treatises of cell biology.
cancer in hu~ans develops th~ough a multi-step process, indicating that multiple changes must occur to convert a normal cell into one with a ~alignant phenotype. One class of involved genes include6 cellular oncogenes. When activated by mutation or when expressed inappropriately, dominant-acting onc~genes override normal cellular control mechanisms and pro~ote unbridled cell proliferation. A
newly recognized class of cellular genes that appears to be equally important;in cancer developDent inclutes ~he tumor suppressor genes, 50metime5 called "an~i-oncogenes." These g~nes act to dampen cell gro~th; inactivation of their normal function appears to be a common denomina~or in the evolution of tumor~ cells. ~oth alleles o~ a tumor suppressor gene must be inactivated to result in loss of function in the cell. Inactivation of one allele (i.e., the gene copy on one of~ the two chromosomes) increases the probability that an event~will damage the survivLng allele, : ~ :

~ ~U8SrlTUTeSH~EET ~

WO92/11~74 PCT/US92/00295 - 2 ~;9 ~ 7 .. .

in effect making the ho~t more susceptible to tumor induction.
The present invention relates to the production of non-human transgenic and chimeric animals and cells which

5 contain at least on~ mutated chromo~omal allele o~ a tumor suppressor gene ~and, in a preferred embodiment, one normal allele of tbat gene). Where the cells and non-human animals of thQ present invention con~ain mutations in both of their - chromosomal alleles, such m~ltations may be the same, or they ~ .
lO may be different ~ro~ one anoth2r. ~:
As is well Xnown, an allele may be ~apable of being expressed by the natural proces~es operating in a cell. The e~pression of an allele results in the production of a gene . product. The term "allele" a6 u~ed herein is intended to denote any nucleotide sequence that affects the expression of a particular gene. It thus is in~ended to refer to any enhancer, promoter, processing, intervening, coding or termination sequence or region of the gene, or any sequence that stabilizes the gene product,: or its mRNA, etc.
~ 20 : An allele of a gene is~said to be mutated if (l) it is .~ : not expressed:in a~cell or anima:l,~ (2) the expression:of the allele is altered ~with~respect~to the expression of the normal allele of the gene, or- (3)~ the~ allele expresses a gene product,~ ~ut that gene product has altered structure, activity, or characteristics relative to the gene product of a nor~al allele of that gene.
Thus, the te~rms "mutation" or "mutated" as used herein are intended to denote an:alteration in the "normal" or : -~: "wild-type"~nuoleotide~sequence of~ any nucleotide sequence : 30~ ~or rcg1on o~ the~al1e1e. ~o~ue~d~herein, the terms "normal"

SUB~CiTlTI ITE~ EET~

W092/11874 PCT/US92tO0295 ,.. . . ...
2~98~
-24- .

and "wild-type" are intended to be ~ynonymous, and to denote any nucleotide sequence typically found in nature. The terms ~mutated~ and ~normal~ are thus defined relative to one another; where a cell has two chromofiomal alleles of a gene that differ in nucleotide sequence, at least one of these alleles is a "mutant" allele as that term is used herein. A "normal tumor-8upprQssing gene product" is the gene product ~ha~ i~ expressed by a "normal" tumor-suppre~sing gene.
A mutation may be "cryptic. It A cryptic mutation does not affect either the expression of the ~utated gene, or the activity or function of the expre6sed gene product. Cryptic mutations may be detected through nucleotide saquence anal-ysis. Examples of cryptic mutations include mutations that do not result in a change in the amino acid sequence of the expressed gene product, as well as mutatlons that result in the substitution of an equivalent amino acid at a particular position in the expres~ed gene product. Most preferably, the mutation will be "non-cryptic" and will therefore intro-: 20 duce a change in the nucleotide sequence of the allele that detectab}y ~alters either the expression or the activity or ~unction of the allele. A "mutation that detectably alters the expression of an allele " as used herein denotes any change in nucleotide sequence af~ecting the extent to which the allele is transcribed, processed or translated. Suchalterations may bej for example, in an enhancer, promoter, coding or termination region of the allele, mutations which stabilize the gene product, or its mRN~, etc. A "mutation that detectabIy alters the activity:o~ an allele, 1l as used herein denotes any change in nucleotide sequance that alters ~: , SUE~STITUTE SHE~T

wos2/11x74 PCT/U~92/~0295 2098~2 ~

the capacity of the expressed gene product to ~ediate a function of the gene product. Such mutations include changes that diminish or inactivate one or more functions of the expressed pro uct. Significantly, s~ch mutations also include changes that re~ult in an increase the capacity of the yene product to mediate any function ~for example, a catalytic or binding activity) of that gene product. A
"mutation that detectably alters the function of an allele,"
as used herein denotes any change in nucleotide ~equence - lo that alters the capacity of a binding molecule (such as a binding protein) to specifi~ally bind to the all~le.
Any o~ a wide variety of methods (treatment with mut~genic compounds, spontaneou~ isolation, insertional inactivation, site-~pecific insertions, deletions or substi-tutions, homologous recombination, etc.) may be used toproduce mutations in accordance with the present invention.
As indicated above, a large number of such mutations are known, and mutations can be readily identified by seguenc-ing, tumorigenicity, resilience to tumorigenicity, binding 20~ activity, etc. (see, for exa~ple, Eliyahu et al~, a~YFe 312:6~6-649 (1984); Finlay et ~l., Molec. Cell. ~iol. 8:531-539 (1988); Nigro, J.M. ç~_~l~, Na~Ee 42:705-708 (1989), . all herein incorporated by re~erence).
~n allele i5 said to be "chromosomal" if it elther is, or replaces, one of the two alleles of a gene which a cell inherits from its ance~tors, or which an animal inherits ~rom its parents. An allele is not "chromosomal," as that term is used herein, i~ ~he allele increases the copy number of the total number o~ alleles of a particular gene which are present in a cell.

$~QSTOTUT~ SHEET

~ . . .
.

-.. -. - . ~ .i , . ... .. - .

wos2/ll874 PCT/US92/00295 .
3~827 The cells that can be produced in accordance with the present invention i~clude both "germ-line'~ and llso~atic"
cells. A germ-line cell is a sperm cell or egg cell, or a precursor or progenitor of either; such cells have the potential of transmitting their ~enome (including the altered tumor-suppre~sor allele) in the for~ation of progeny animals. A somatic cell is a cell that is not a germ-line cell. Such c~lls may be "sub~tantially ~ree of naturally occurring contaminants," or may b~ present in an animal of : 10 the same or of different species. A cell is 9~substantially free of naturally occurring contaminants" when it, or a pracursor or ancestor cell, has been purified from tissue (normal, tumor, etc.~ in which the ~ell is, or would be, naturally associated. Two ~pecie~ are said to be the same i~ they are capable of breeding with one another to produce ~ertile offspring. Two species are ~aid to be different i~
they are either incapable of breeding to produce viable of~spring, or are substantially incapable of produci~g fertile offspring.
The pre ent invention encompasses the ~ormation o~ such cell~ and non-hu~an ani~als for any tumor suppressor gene.
Exa~ples of such gene~ include the b gene (Weinberg, R.A., Sçi~n~Lc ~e~, Sept. 1988, pp 44-51); Hansen ~.F. et~al., Trends_Çe~et. 4:125-128 (1988)), the gene responsible for Wilms:tumor (Maitland, N.J. et al., ~ei5 ~ . 9:1417-1426 (1989); Shaw, A.P.W. e~ ~1., gnçggçn~ 3:143-150 (1988);
Xovacs, G. ~ Q~ Natl. Acad. $~ LU-S,A,~ 85:1571-1575 (1988); Xumar, S. et al., I~ J9unç~ 40:499-504 (1987); Allen, D~Mo ~et~1., lin. es. ~5:416A (1987)j and the p53 ge~ Lane, D.P. ~h~ , Gen~$ ~ DeveloD. 4:1-8 :

Cl IRCT(TI IT~

WO92/11~74 PCT/~S92/00295 9~,7 (1990)). Other tumor suppre~sor gene~ may, however, be used in accordance with the present invention (Hiti, A.L., ~
Cell. Biol. 9:4~22~47~0 (1989); Hastie, N., Antlcanc. ~es.
8:1074 (1988); Boada. M~.. QncolQqi~ lQ:9-30 (~987);
Gallie, B.L., J. Cell~_~iochem. ~:215-222 (1986); Alt, F.W.
et ~al., Cold ~r. H~rb. Svm~._ Quaat. ~iol. 51:931-942 (1986); Knudson, A.G., $v~ MoLeo. Cell. ~iol.~ BiQchem.
Molec. ~idemiol Canc, ~Q:6~13 (1985~; ~alcolm, 5. Molec.
Med. 1:79-94 (1984); Tricoli, J. e~ al- ~YI ~.
36:121S (1984); all herein incorporated by reference).
The ~ and p53 gene~ are ghe preferred tumor suppressor genes of the present invention. The cDNA and genomic forms of the rb gene have been cloned (Friend, S. et al., Nature 323:643 (1986); Lee, W. et ~l~, Sgiencg 235:1394 (1987);
Fung, y. et alJ ~ science ~ 1657 (1987); Hong, F. et al., Proc. Na~'L~ A.ca~ Sci. LU.S.A.l 86:5502 (1989), all herein incorporated by reference). Potential methods and animal (including ~ransgenic) mode~s for the study of retinoblast-oma are discu~sed in Lee, W. ~ al~, WO 90/05180, herein ~:20 incorporated~ by reference. The invention is illu~trated : below wit~ re~erence to the p53 tu~or suppressor qene. The ability to manipulate ~his gene and to produce non:-human transgenic animals which carry such ~utated alleles is ill-ustrated with respect to a particular mutated allele. It is to be understood, however, ~hat the invention and the me-thods disclosed herein, can be uged to produc~ any possible mutation in the p53 gene. In particular, the invention in-cludes the production o~ animal cells and non-human trans-ge~ic or chimeric animall which carry the particular muta--:
..

Cl lRcTlTllTE S~EET ; `

W0~2/11874 PCT/US92/00295 ' .' 2 0`9~`'g;;~7` `

tions of the p53 gene that are responsible for the Li-Fraumeni Syndrome discu~ed below.
I. The p~3 Gbne The present invention concern~ a non-human animal or an animal (including human) cell in which one of the ~wo natur-ally pres~nt copies of the p53 ~ene o~ such non-human animal or animal cell has been rendered non-~unctional through a mutation (such a~ a deletion, insertion, or substitution in the naturally occurring p53 gene ~e~uencs).
The normal p53 gene product is a tumor-suppressing protein (sager, R., scie~ce 24~:1406-1412 (1989); Fi~lay, C., Cell 57:1083 (1989), both of which references are herein incorporated by refersnce). The nature and characteristics of the p53 gsne i8 reviewed by Lane, D. et al. (~enes pçvel.
4:1-8 (lssO), herein incorporated by reference). The p53 gene plays a protective role~against the trans~orming effects of Friend erythrole~kemia virus (Munroe, D.
oncoaene 2:621 (1~8~ t is also believed~to play a role in chromosome ~tability, dirferentiation and senescence, and cell~ proli~eration (Sager, R., : Science 246:1406-141 (I989)).
As indicated above, allelio deletions of the p53 locus bave been identified in colorectal carcinomas. Mutations in the p53 gene have also been identified in tumor~ of the brain, breast, and lung (including :bronchioalveolar carcino~a, extrapulmonary small cell carcinoma, adenocarcinoma,:~ s~all ~cell :carcinoma, adenosquamous 30: carcinoma~, pul~onary carcinoid tumors, and in a -.
; : SlJBSTlTl~TE SHEEY ~: ~

W092/l1~74 PCT/U~92tO0295 2 ~ 9 3 ~ 2 ~ !

neurofibrosarcoma (Nigro, J.M. et al., Nat~re 342:705-708 ~1989); Takahashi, T. et al~, Sçience 246:491-494 (1989)).
~08t tumors with such allelic deletions contain pS3 poin~
mutations resulting in amino acid subfititution~. Such mutations are not con~ined to tumors with such allelic daletions, but also occur in so~e tumors that have retained both parental allele~. ~utations in the p53 gene which give rise to tumors are as~ociated in four "hot-spots~' which roincide with the ~08t highly conserved se~uences of the p53 gene (Nigro, J.M. et ~L, ~a~u~e ~ 705-708 (1989), herein incorporated by referenoe).
In summary, there is now convinci~g evidence that the human p53 gene iæ a tumor suppres~or gena (Weinberg, R.A., SCi~ati~ic Am~, Sept. 1988, pp 44-51). Like the RB
protein, discus~ed above, p53 is a nuclear protein that forms a complex with SV40 large T antigen (De~aprio, J.A. et al., Cell S4:27s-283 (1988); Crawford, L.V., Int, _~ev.
Exper..Pathol~ 25:1-50 (1983)). The binding of these two prot~ins by viral tumor antigens presumably inactivates them and contributes to transformation. p53~ gene déletions have besn~ noted m s-veral mou6e erythrole~emiG cell lines, reminiscent of:the RB gene deletions in retinoblastomas (Mowat et a~ , Nat~e ~1:633-636 (1985); Chow et al., J.
V-~Ql. 61:27770278l~ (1987); Hicks, G.G. et al., J. ViroJL, 62:4752-4755 (1988)):. Cell lines and tumors derived from human osteogenic sarco~as ortan contain gro88 rearrangements of the p53 gene, including deletion~ (Masuda, H., Proc.
Nat'l. Acad. SCi~-Lu.s.~L 84:7716-7719 (1987)). Allelic deletions in chromo~ome 17p twhich contains the p53 qene) : 30 occUr in oYer~ 75%:or colorectal carcinomas (Baker, S.~. et .

SUE~STlrUTE~S~EET ` - -2~98S27 ~1., Sciençe ~ 217-221 (1989)). In two tumors, the remaining non-deleted p53 allels was shown to contain mutant~ in highly con~¢rved regions previously found to be mutated in murine p53 gene (~aker, S.J. ~ L~, Science ~44:217-Z21 (1989)). Loss o~ hQterozygosity in chromosome 17p has been noted in a high per~entage of individuals with small cell lung carcinoma (Yo~oto, J. et al~, Proc._Nat~l.
~cad~ Sci. rU.S.~.) 84:925~-9526 (1987); ~arbour, J.~. et ~1~, Sc~ence ~ 353-356 (1388)). In the HL-60 human leukemic cell line, major dele~ions in the p53 gene and absence of the pS3 protein have ~ee~ noted (Wol~, D- ~
R2:790-7g4 (1985)). These results indicate that the ab~ence of a functional p53 allele is highly coxrelated with some forms of cancer in humans and mice, strongly suggesting a tumor-suppressor role for p53.
Finally, Finlay et alt (Ç~ll S7:1083-1093 (1989)) have rec~ntly demonstrated that the wild-type mouse p53 gene suppresses transformation in vi~ro after cotransfection of rat embryo cells with Ela and activated ras, indicating that : 20 the presence of the normal p53 gene acts negatively to block transfor~ation. :
: The cDNA and genomic ~orms o~ the p53 gene have been . cloned (Pennica, D. et ~1., Vi~ol. 134:477-482 (1984);
Jenkins, J. ~ ., N~tu~ ~12:651-654 (1984); Oren, M. et al., EM~0 J. ~:1633-163g (1983); Zahut-Houri, R. e~ al., Na~ure 306:594-597 t1983j, all of which references are herein incorporated by reference)~
Recent evidence has cugges~ed that a mutation in the p53 gene may be responsible for the Li-Fraumeni Syndrome, a rare human genPtic disorder (Malkin, D. ~ L-, Soience 250:1233--SU13STITUTE SI~EET : :

.

WO92/1~874 PCT/US92/00295 " '~09~8~7 ~, .

-3~

1238 ~1990); Narx, J., ~ÇiÇL~ ~Q:1209 ~1990), both references herein incorporated by refe~ence). Indivi~uals afflicted with this di~ease are highly su~ceptible to several malignant tumors - breast carcinomas, soft tissue .
sarcomas, brain tumors, osteo~arcomas, leukemia, and adrenocortical carcinoma. The disease is also associated with a higher incidence of melanoma, gonadal ger~ cell tumors, and carcinoma~ of the lung, pancreas and prostate (~i, F.P. Q~ L~ ~ 71:747 (1969); Birch, 10J-M- ~ l., J. ~lLn~_Qncol. ~:583 (1990); Birch, J.M. et al., ~it. ~._Canc. 49:32~ ~1984); Li, F.P. ~ c.
~s ~8:5358 (1988); Williams, W.~. Ç~_Ll-, Famili~l Canc., ~st~Int. ~çs Conf. p. 151 (~arger, Basel, 1985); Strong, L-C- ~_31-, J~ Na~L~_~n~ Inst. 79:1213 (1987)).
15~he Li-Fraumeni Syndrome i8 associated with a particular se~ of mutations in exon 7 of the p53 gene (~alkin, D. et 31-, Science 2S~:1233-1238 (199O), herein incorporated by reference). Knowledge of the s~uence and location of the mutation which causes the Li-Fraumeni syndrome penmits one to use the methods of the pre~ent inventio~ to produce a ~non-human ~ransgenic animal which contains any of these :
mutations. As indicated above, such an animal,~and its uses - in diagnostics and gene~herapy, is an smbodimen~ of the~
present invention.
II. The Inter~ction:o~ Nutant and Nor~al p53 Gene Pro~iuct6:

A cascade of mutational events is believed to be required for~ tumorigenesis. It is believed that this ; ~ :
Sllf35TlTUTE SHEET: -W092/1l874 PCT/US92~00295 2088~7 cascade involves both the activation o~ one or more oncogene(s) and the inactivation of one or more tumor-suppressing gene(s) (Sager, R., Science ~246:1406-1412 (1989); Fearson, E.R. et~ , ÇÇll ~1:759-767 (1990), both herein incorporated by re~erence)~ Indeed, mutations in at least 4-5 genes are believed to be required for the ~o~mation of a malignant tumor.
Studies of the clonal nature of tumox formation have suggested that tumors have a monoclonal composition, and lo hence arise by the clonal propagation of a single progenitor cell (Fearson, E.R. et aL , gSi~n~ 247:193-197 (1987)).
The loss (either by mutation or deletion) of functional p53 gene function has been found to be a characteristic of colorectal tumors (Fearson, E.R. et_~l., Cçll 61:759-767 (1990)).
The simplest model to explain the mechanism of action of a tumor-suppressing gene is that malignancy requires two separate genetic events (i.e. loss by deletion or mutation o~ both functional p53 alleles in a cell). Inactivation of only one o~ the two natural p53 alleles causes a phenotypically silent, re~essive mutation which does not resul~ in ~alignancy in the absence o~ the second mutational event~
It has ~een observed that, by i~troducing DNA containing a mutated p53 gene into rodent cells, such cells can be trans~ormed, This transfor~ation is due to a cooperative action o~ the mutant p53 gene product with the gene product o~ the. ras oncogene. Since ~uch cells also express the normal p53 gene,~i~ has been proposed that the inactivation of only one p53 :allele ls not phenotypically silent ' SUE~TiTUTE S~
-W~9~ 74 PCT/US92/00295 ~ . .
~098~2~ ' . .

(Fearson, E.R. et ~l , ÇÇLl Çl:759-767 (1g90)). The ability of the cells to undergo trans~ormation in the pre~ence of both the normal and the mutant p53 gene products suggests that the two gene produc~s may in~eract to exert ~heir ef~ects. One explanation o~ the mixed reces~iv¢-dominant characteri~tic of the p53 mutant/nor~al allele interao~ion is that oligomerization i8 needed ~or actiYity~ and ~hat the oligomer ~ormed in cells haYing both a ~utant and a no~mal p53 allele contains a ~ixture of the gene products of the~e all~les (Eliyahu ek al" Oncoqe~e 3:313-321 (1988); Kraiss, ~_31~ QL~ ~2:4737-4744 (1988)). Due to the presence of the mutant p53 gene product in such mixed oligomers, ~he oligomers would be expected to ~ave a diminished ability to supprass tumor formation (Fearson, E.R. et al.,~ 753-767 ~1990~, herein incorporated by reference). Con~istent with this explanation has been the ~inding tha~ mutant p53 gene products form heterodimers with the wild-type p53 gene ; product (Finlay, C.A., ~ 57:10~3 (1989)).:
Transgenic ani~als ~ay be u ed to investigate the biological i~plications ~of tumor-suppressing qenes (Capecchi~ M.R., Sçie~ 1288-1292 tI989)). Lavigueur, A. et al. constructed~a transgenic mouse which had a singIe added mutant p53 gene in addition to the endogenous two wild-type p53 alleles. The ~ou~e, and its progeny overexpres~ed the added p53 gene. The mice were found to have a high incidence of lung, bone, and Iymphoid tumors (Lav.igueur, A. et ~L, Molec. ~ell. aio}~ 9:398Z-3991 (1989 The~hlgh l-vels of mutant p53:in these mice led to tumor ~ormation in~22 of 112 ~ice examineq. Due to the overexpression of the mutant p53 gene product, an~ to the - .

~.CI IR;~, ~T~ FFT~ ~ ~

WO92/~187~ PCT/USg2/00295 .~'"'!
2 09~g~r~ f ability of this mutant protein to impair the biological activity of the normal p53 gene product, the transgenic mice of Lavigueur et al~ exhibit the phenotype of an animal in whlch all nor~al, endogenou~ p53 gene product activity has been functionally inactivated (LaYigueur, A. et al., Molec.
Cell. Biol~ 9:39~2-3991 (1989); Sager, Ro~ ~ienc@ 246:1406-1412 (1989)).
Thus, the mice described by Lavigueur, A. et al. ( olec.
Cell. Bi~l. 9:3g82-39gl (1989)) con~ained an added oncogenic p53 allele in addition ~o its ~wo normal p53 genes.
Importantly, the e~act of gene dosage in such animals is unknown. Moreover, it is not known whether the added p53 gene in such ani~al6 contain~ all of the normal upstream and downstream regulatory ele~en~s. Even if all of such elements were present, the context of the integration of the construct may render the regulation of the gene abnormal.
For these reasons, the utility of the mice in assessing the role of the p53 gene on oncogenesis is unclear.
It would be desirable to produce a transgenic animal whose genome possesses one normal and functional p53 all le ~nd one non-functional (mutant) pS3 allele. Such animals could be used to study the consequences resultlng from the Ioss of one p53 allele, and thus would more clearly aid in elucidatin~ the processe~ of oncogenesis and tumorigenesis.
Such animals would also be userul in screening potential carcinogens, in developing novel antineoplastic therapeutics, and in gene therapy. The prQsent invention provides such an ani~a}.

..

lTUTE SHEET

.

WO92/1l874 PCT/US92/0029s 2~9~:8.~,7;

III. ~o~ologou~ Reco~bination The present invention U8~8 the process of homologous recombination to introduce a specific mutation into the nat~rally present p53 ~equence o~ an animal cell, most preferably an embryonic stem (ES) cell. ~he mutated E5 cells of non-human animals can th~n be either cultured in suitable cell culture medium, or introduced into the uterus of ~ suitable recipient and permitted to develop into a non-human animal.
Alternatively, the mathods of the present invention maybe used to alter the 60matic cells of a non-human animal to produce a chimeric non-human animal.
An understanding of the process of homologous reco~bination (Watson, J.D., In: ~lecula~ ~ioloqy o~_the Ç~, 3rd Eq., W.A. Benjamin~ Inc., Menlo Park, CA (lg77), which re~erence is incorpoxated herein by reference) iæ thus desirable in order to fully appreciate the present invention. ~ ~
~ In brie~, homologous recombinatio~ is a well-studied natural cellular proces~ which~results in the scission of : ~ tw~ nucleic acid molec~les having identicaI or substantially similar sequences (i.e. "ho~ologous"), and the;ligation of the two ~olecules such tha~ one region of each initially ~ .
pre~ent molecule is now liga~ed to a region o~ the other initially present molecule (Sedivy, J~ io-~h~ol.

6:1192-1196 (19~8), which rererence is incorporated herein by reference).
Homologous recombination~is, thus, a sequence specific ~; 30 process~ by which cells can transfer a "region" of DNA rrOm : ; : :

TlT~TE S.L~EE~1 ~
~ .

wos2/11874 P~T/US9~/00295 ~ ~ b ~ ~

-3~-one DNA molecule to another. As used herein, a "region" of DNA is intended to generally refer to any nucleic acid molecule. The region ~ay be of any length from a single base to a substantial fragment of a chromosome.
5For homologous recombination to occur between two DNA
molecules, the molecules must pos~es~ a ~region o~ homology"
with respect to one another. Such a region o~ homology must be at least two base pairs long. Two DNA molecules possess such a "region o~ homology" when one contains a region whose sequence is so similar to a region in the second molecule that homoloqous recombination can occur.
Recombination is catalyzed by enzymes which are naturally present in bo~h prokaryotic and eukaryotic cells.
The transfer o~ a region of DNA may be envisioned as occurring through a multi-step process.
If either of the two participant molecules is a circular molecule, then the recombination event results in the : integration of ~he circular molecule into the other participant.
20Importantly, if a particular region is flanked by regions of homology (which may be the same, but are pre~erably differentj, then tvo ~eco~binational events may occur, a~d result in the exchange o~ a region o~ DNA between two DNA molecule~. Reoo~bina~ion may~be i'reciprocal," and thus resuits in an exchange of DNA regions between tWQ
recofmbining DNA molecul~s. Alternatively, it may be "non-reciprocal," (also referred to as l!gene conversion") and result in both recombining nucleic acid molecules having the same nucleotide se~uence. There are no constraints : .

i .
IRi:~TlTl ITF ~

WO92/l1874 PCr/US92/00295 2 ~ g~ `7 regarding the size or sequence of the region which is exchanged in a two-event recombinational exchange.
The frequency of recombination betwsen two DNA molecules may be enhanced by treating the introduced DNA with agents which sti~ulate reco~binatio~. Examples of such agents include ~ri~ethylp~oralen, W light, etc.

IV. Prcduction of ChiPeric and Transgenic Ani~als:
&ene Targeting ~e~hod~

One approach to producing animals having defined and specific genetic alterations has used homologou~
recombination to control the site o~ integration of an introduced marksr gene ~equence in tumor cells and in fusions be~ween diploid human fibroblast and tetraploid mouse erythroleukemia cells (Smithies, O. et a~,, Na~e ~.
317:230-234 (1985)).
This approach was further expIoited by Thomas, K. R., and co-workers, who ~escribed a general method, known as ;'gene targeting," for targeting mutations to a preselected, de~ired gene s~quence o~ an ES cell in order to produce a transgenic animal (Mansour, S.L. ~ , Nature 336:348-3S2 (1988~; Capecchi, ~.R " ~ç~s ~enet~. 5:70-76 (1989);
:- Capecchi, ~.R., Scie~ce ~ 1288-1292 (1989); Capecchi, M.R.
~ , In: ~UE~.gnt Communica~l~ns._in Molecular ~ioloqy, Capecchi, M.~. (ed.), Cold Spring Harbor Press, Cold Spring Harbor, NY (1989), pp. ~5-52; Frohman, M. A. et al., ~
S6:145-147 (1989~; all of which ra~erences are incorporated herein by r~eference). :~

~,.

.:
':

.

SlJlillSTlTUTE SHEET

WO92tll~74 PCT/US92/00295 2;a ~
.

It may now be fea~ible to deliberately alter any gene in a mouse (Capecchi, M.R., Trends Gen~. 5:70-76 (1~89);
Frohman, M. A. ~ , Ç~ll 56:145-147 (1989)). Gene targeting involves the use of standard recombinant DNA
techniques to introduce a de~ired ~ut~tion into a cloned DNA
sequence o~ a chosen locus. ~hat mutation is then trans~erred through homologous recombination to the genome of a pluripotent, e~bryo-derived ~tem (ES) cell. The altered stem cells are microinjected into mouse blastocysts and are incorporated into the developing mouse e~hryo to ultimately develop into chimeric animals. In some cases, germ line cells of the chimeric animals will be derived from the ge~etically altered ES cells, and the mutant genotypes can be transmitted throuqh breeding.
Gene targeting has been used to produce chimeric and transgenic ~ice in which an np~II gene has been inserted into the ~2-microglobulin }ocu8 (Xoller, B.~. et al., Proc.
Natl._Acad. Sci. ~U.S~a~ 86:8932-8935 (1989); Zijlstra, M.
et al., Nature ~ 435-438 (1989); Zijlstra, M. et al~, Na ure 34~:742-746 (1989); DeChiaba al~ 78-80 (1990)). Similar experim~nts have enabled the production of chimeric and transgenir animals ~aving a c-abl gene which ~: has b~en disrupted by the ins~rtion of an ~ gene ~Schwartzberg, P.L. et al., Sciçnce ~:799-803 ~1989)).
The technique has been used to produce chimeric mice in which the en-2 gene has been disrupted by the insertion of an p~II gene ~Joyner, A.L. e~ ature 338:153-155 t1989)). :
Gene targeting has also been~u~ed to correct an ~E~
deficiency in an ~E~~ ES cell lin-. Ce}ls corrected o~ the -SOBSTiTUTE SHEET

~:

WO92/1l874 PCT/US9~/00295 2 ~`~ 8~

deficiency were used to produce chimeric animals. Signifi-cantly, all of the corrected cells exhibited gross disrup-tion of the regions flanking the ~ locus; all of the cells tested were found tc contain at least one copy of the vector used to correc~ the deficiency, integrated at the hpE~ locus (Thompson, S. g~al~, Ç~ll 56:313-321 (1989);
Koller, B.H. e~ al,, 8~:~927-8931 (1989~
In order to utilize khe "gene targetlng" method, the gene of interest must have been previou~ly cloned, and the intron-exon bo~ndaries determined. The method results in the insertion of a marker gene (i.e. the n~II gene) into a translated region of a particular gene of interest. Thus, use of the gene targeting method results in the gross destruction of the gene of interest.
Recently, chi~eric mics carrying the homeobox hox l.1 allele have been produced using a modification of the gene targeting method (Zimmer, A. et al., ~ 338:150-154 (1989). In this modification, the integration of vector sequences was avoided by microinjecting ES cells with linear DNA containing only~a:portion of the hox ~1 allele, without any accompanying vector sequences. The DNA was found to cause the gene conYersion of the cellular h~ allele.
Selection was not used to facilitate the recovery of the "converted" ~S cells, which were identified using the polymerase chain reac~ion (~PCR~). Approximately 50% of cells which had been clonally purified from "converted"
cells were found to contain the introduced hg~ allele, suggesting to Zimmer, A. e~ al. either chromosomal instability or aonta~ination of sample. None of the ' ,` `

.

~ ~ : SUBST~TUTE S~E~T ~ -W~92/11874 PCT/USg2/0029s ~d~8~`

chimeric mice were found to be able to transmit the "converted" gene to their progeny (Zim~er, A. et ~l., In:
~ent Com~u~L3~ions in Molç~u~ iolQ~y, Capecchi, ~.R.
(ed.), Cold Spring Harbor PresE~, Cold Spring Harbor, NY
~1989), pp. 53-58).
Significantly, the use of gene ~argeting to alter a gene of a cel} results in the formation o~ a gross alteration in the sequence of that gene. The ef~iciency o~ gene targeting depends upon a number of variables, and is different from construct to construct. ~or the p53 gene constructs used herein, such efficiency is approximately l/300.

V. Production o~ ~hi~eric and Tran~,genic Animal8:
~e of Ins,ertion Vector~
~
In contrast to the a~ove-described methods, the present invention uses methods capable of producing subtle, precise, and predetermined mutations in the sequence of one of the two alleles o~ the p53 gene of a human or animal cell.
Although the ~ethods-di cussed balow ~re capable of mutating both alleles ;of the cell's p53 gene, it is possible to readily identify ~for exa~ple :throu~h the use o~ PCR
: (discussed below), or other methods) such~ dual mutational events. Since the frequency of such dual mutational events - 25 is the square of the ~requency of a single mutàtional event, cells having mutations in both of their p53 alleles will be only a very small proportion o~ the total population of mutated cells.
The present;invention has several e~bodiments. In the simplest embodiment, an:~nsertion vector is used to mutate -:

I JF~.~TIT~ ITF C~

W0 92/11874 P~/US~2/002g5 209~27 . .

the nucleotide sequQnce o~ one of the two alleles o~ the p53 gene of a cell. ~he u~e of this vector type in combination with a second selectable rever~ion event prevents the disruption of the chromosome by the selectable marker gene (an n~II gene, for exa~ple) of the vector or by other vector sequences. Thus, gro8 distortio~s of the recipient chro~osome are avoided. ~he efficiency of the gene targeting is sub~tantially greater than in the gene targeting me~hod~ discussed above.
~ost pre~erably, the DNA molecule(s) which are to be introduced into the racipien~ cell contains a region of homology with a re~ion of the p53 gene. In a pre~erred embodiment, the DNA molecule will contain two regions of.
homologn~ with the cell t 5 ps3 gene. T~ese regions of homology will preferably flank ~the precise sequence whose incorporation into the pS3 gene is desired. As stated above, the regions of homology may be of any size greater than two ba~es long. Most preferably, the regions of homology will be greater ~han 10 bases Iong.
The DNA molecule(s) may be single stranded,~ but are pref~rably double stranded. T~e DNA molecule(s) ~ay ~e introduced ~o the ~Gell as one or `more RNA: molecules which may be converted to DNA by reverse transcrip~as~ or by other m~ans. Pre~erably, the:DNA molecule will be double ~tranded linear molecule. ~ ~n the best mode for conducting the invention, ~uch a molecule is obtained by cleaving a closed covalent circular molecule to form a linear molecule.
Preferably, a restriction endonuclease capable of cleaving :~ the molecule~at~a single~ e to produce either a blunt end or~ staggere~ ~end: linear molecule is emp~loyed. Most : -SUE~STITUTE SHE~T

wos~/11874 PCT/US92/00295 2098~7 preferably, the nucleotides on each side of this restriction site will co~prise at least a portion of the preferred two xegions of ho~ology between the .~NA molecule being introduced ~nd the p53 gene.
The invention thus provide~ a ~ethod for altering the natural p53 gene sequence through the introduction of a "desired gene seque~ce" into ~hat gene. ~he "desired gene sequence" may be o~ any length, and have any nucleotide sequence. It may co~prise o~e or ~ore gene sequences which encode complete proteins, frag~ent~ of such gene sequences, requlatory equences, etc. Signi~icantly, the d~sired gene sequence may difPer only slightly fro~ a native gene of the recipient cell (for example, it may contain single, or multiple ~ase alterations, insertion~ or deletions relative to the native gene). The use of such desired gene ~equences permits one to create subtle and precise changes in the p53 gene of the recipient cell. Th~s, the present invention provides a means for manipulating and modulating the expression and regulation of the p53 gene.
In particular, the i~vention provides a mean ~or manip~lating and modulating p53 gene expres~ion and protein structure through the replace~ent of a naturally present p53 gene sequence with .a ~non-selectable" -"desired gene seguance." A gene sequence is non-selectable if its presence or expression in a recipient cell provides no survival advantage to the cell under the culturing conditions employed. ~hus, by definition, one cannot select for cells which have received a "non-selectable" gene seque~ce in:their p53 gene. In contrast, a "dominant" gene sequence is one which can un~er certain circumstances Sl)~3STITIJTE SHEET ~

WO~2~11874 PCT/US~2/00295 2~827`
' ! ,, ., , :

provide a survival advantage to a recipient cell. The neomycin resistance conferred by the nPtII gene is a survival advantage to a cell cultured in the presence of neomycin or ~418. The na~II gene is thus a dominant, rather than a non-selectable gene sequence.
In particular, the invention permits the replacement of the naturally present p53 g~ne sequence o~ a recipient cell with an "analog" sequence. ~ seguence is said to be an analog of another sequence if the two sequenceis are substantially similar in sequence, but have minor changes in saquence corresponding to ~ingle base substitutions, deletions, or insertions with respect to one another, or if they pos~ess '3minor" multiple base alterations. Such alterations are intended to exclude insertions of dominant selectable marker genes.
When the desired gene sequence, flanked by regions of homology with the p53 gene seqUQnce of the recipient cell, is introduced into the recipient cell as a linear double stran~ed molecule, whose termini correspond to the regions of homology, a single recombination event with the p53 gene o~ the cell will occur in approximately 5% of the transfected cells. Such a single recombinational event will lead to the integration of the entire linear molecule into the genome of the recipient cell.
The structure generated by the integration of the linear molecule will undergo a sub equent, second recombinational event (approximately 10-5 - 107 per cell generation). This second recombinational event will result in the elimination of all DNA except ror~the flanking regions of homology, and the desired DN~sequence from the integrated structure.
"- . ' SUBSTITUTE SHEET : ~
: . . .

7~ PCT/US92/00295 2~988?~

Thus, the consequence of the second recombinational event is to replace the DNA s~quence which is normally present between the flanking regions of homology in the c811~ S p53 gene, with the desired DNA sequence, and to eliminate the instability of gene replacementO
The DNA molecule containing the desired gene sequence may be introduced into the pluripotent cell by any method which will permit the introduced molecule to undergo recombination at its regions of homology. Some methods, such as direct microinjection, or ca}cium phosphate transformation, may c~use ~he introduced molecule ~o form concatemers upon integration. These concatemers may resolve themselves to form non-concatemeric integration structures.
Since the presence of concatemers is not desired, methods which produce them are not preferred. In a preferred embodiment, the DNA is introduced by electropor~tion (TonPguzzo, F. et al., Nucleic A~i~s ~s. 16:5515-5532 (1988); Quillet, A. çe_~lL_, J. I~ncl~ 17-20 (1988);
Machy, P. ~ , Pro~. Na~l~. Açad. Sçi~.~(U.S.a.) 85:8027-8031 (1988); all of~w:ich~references ar~ incorporated here m by reference). ~ ~ ~
After permitting the introduction of the~ DNA
molecule(s)~, the cells ~are cultured under conventional conditions, as are kn~wn in the art.
In order to facilitate the recovery of those cells which have received the DNA molacule containing the desired gene sequence, it is~preferable to introduce the DNA containing the desired~gene~Dequence in combination with a second gene sequence which~would~contain `a detectable marker gene sequence. ~or the~purposes of- the present invention, any , SUBSTITI.ITE:SHEET:

WO92/11874 PCT~US92/00295 ~882~

gene seguence who~e presence in a cell permits one to recognize and clonally isolate the cell may be employed as a detectable ~arker gene sequence.
In one embodiment, the pre ence of the detectable mar~er sequence in a recipienk cell i8 recegni2ed by hybridization, by detection of radiolabelled nuclectides, or by other assays of datec~ion which do not require the expre~sion o~
the detectable marker ~equence. Pre~erably, such sequences are detected using PCR t~ullis, K. e~ al., Cold ~Prina 10Har~or Svm~ Quan~. Biol. 51:263-273 ~1986); Erlich ~. et ~1~, EP 50,424; EP 84,796, EP~258,017, EP 237,3~2; Mullis, K~, EP ~01,184; Mullis K. ~ j US 4,683,202; Erlich, H., US 4,582,788; and Saiki, R. Ç~ L~, US 4~6e3~194)~ which references àre incorporated herein by reference).
15PCR achieves the amplification of a specific nucleic acid sequence using two oligonucleotide primers complementary to regions of the sequence to be amplified.
Exten~ion products incorporating the primers then become templates for subsequent replication steps. PCR provides a : 20 method for selectively increasing the concentration of a nucleic acid molecule having a partiaular sequence even when that molecule has not been previously purified and is present only in a single copy in a particular sample. The method can be used to amplify either single or double stranded D~A.
Most preferably, howev~r, the detectable marker gene sequence will be expressed in the recipient cell, and will result in a selectable phenotype. Examples of such preferred detectable gen~ se~uences include the hprt gene (LittlefiQld, J.W., S~i~nce ~ 709-710 ~1964), herein -~: .
SUe5TlTUTE~ 5HEF.T ~ ::
..
.

.~ . .
..

2~82~
-46- . :

incorporated by reference), the tk gene ~i.e. thymidine kinase gene) and e~pecially the tk gene of herpes simplex virus (Giphar~-Gas~ler, M. et al., ~a~. Res. 214:223-232 (1989) herein incorpora~ed by re~ere~ce), ~he ~E~II gene (Thomas, K.R. e~ al~, ÇÇll ~1:50~-512 ~1987); Man~our, S.L.
et ~l., Na~u~ ~36:348-352 ~19~8), both re~erences herein incorporated by reference), or other genes whi¢h confer resistance to amino acid or nucleoside analogues, or anti~iotics, etc.
Call~ which express an active HPRT enzyme are unable to grow in the pre~ence of certain nucleo~ide analogues (such as 6-thioguanine, 8-azapurine, e~c.), but are able to grow in media suppl~mented with HAT ~hypoxanthine, aminopterin, and thy~idine). Conversely, cell6 which fail to expres~ an active HPRT enzyme are unable to grow in media containing HATG, but are resistant ~o analogues such as 6-thioguanine, etc. (Littlefield, J.~., Science l45:709-710 (1964)). Cells expressing active thymidine kinase~are able to grow in media :.
containing HATG, but are ùnable to grow in media containing .
~ nucleoside analogues eUCh as;s-azaC~ idi~e (Giphart-Gassler, ~ çt al., Mutat. R~9.~ 2l4:223-~32~ (1989)~)~. Cells : :~ containing an~active HSV~ gene are:incapablé of growing in the~presence of gangcy}oyir or similar agents. ~:.
: : The detectable:marker gene may be any gene which aan complement for a ~recognizable cellular deficiency. Thus, for example, the gene~for HPRT could be used as the detectable marker gene se~uenae when employing cells lacking HPRT activity. Thus, this agen~ is an xample of agents may be used: to select mutant:cells, or to ~nègatively select"
30 for cells whi~h~have regalned:normal function.~ - .

SllBS'rl l UTE Sl~ED

WO92/11874 PCT/US92/002~5 2 0 9 8 ~

' The aE~I~ gene (Southern, P.J., et al., J. Molec. Appl.
Ge~et. l:327-34l (1982); Smithies, O. ~ , N~ture 3l7:230-234 (l985), which references are incorporated herein by reference) is the most preferred detectable marker g~ne sequence. Constructs which contain both an n~tII gene and either a tk gene or an hp~t gene are especially preferred.

A. U~e of a Single DNA ~ol~cule Containing Both the Detectable NaIker Sequence and the Desired Gene Se~uence In a first preferred embodiment, the detectabIe marker gene sequence, flanked by the regions of homology to the p53 gene, is provided to the recipient cells on the same DNA
molecule which contains the desired gene sequence. As discussed previously, it is preferred that this DNA molecule be a linear molecule.
After selection for cells which have incorporated the desired DNA molecule (for example by selection for Gil8 20 resistant cells when the detectable marker gene s~quence is ;~
an expre~si~le ~ gene sequence~, the cells are cultured, and the presence of the in~roduced DNA molecule is confirmed as described above. Approxim~tely I~7 celis are cultured and screened for cells which have undergone the second recombinational event (discussed above) resulting in the replacement of a native sequence (i.e. a gene sequence which is normally and naturally precent in the recipient cell~ with the desired gene sequence.
Any of a varie~y of methods may be used to identify cells which have undergone the second recombinational event.
Direct screening of clones, use of PCR, use of hybridization ~ ;

:
- :

-: . .

WO92/1~X74 PCT/US92/00295 .. .; ., ~

probes, etc., may all be employed for ~his purpese. In a preferred embodi~ent, the DNA ~olecule will, in addition to the desired gene sequence, the flanking regions of homology and the detectable marker gene sequence, contain an addition~l gene sequence which will permit the selection or recognition of cells which have undergon~ the second recombinational event. This additional gene sequenc0 will be excised from the cell's p53 gene as a direct consequence of the second recombinational event. Thus, gene sequence~
which are suitable for this purpose include any gene se~uence whose loss from a cell can be detected or selected for. Examples of such "negative selection" gene sequences include tha h~rt g~ne, and the tk gene (especially the tk gene of herpes simplex virus).
In the first preferred e~bodi~ent, the frequency of the second recombinational event is approximately 10~5. However, the use of a "negative selection" gene sequence permits one to identify such recombinant cells at a frequenoy of approximately ~00%.
The DNA molecule may have a region of heterology located at the proposed i~sertion site. Insertion of such a vector permits one to select for reco~binants which have recombined at the insertion si~e (and not at~other potential sites).
If recombination occurs at the desired inaertion site, it will lead to the loss o~ the sequence of heterology located at the proposed insertion site of the DNA~olecule (HSVtk, for example). Insertions whiCh result from other recombinational events will retain the sequence of heterology. Thus, by;e~ploying a region of he~erology~which ~;~ 30 encodes an as~aya~le gene product, ar~which can be used as Sl~eSTlTUTE SREET

.

WO92/11874 PCT/US~2/00295 ~ 209~27 ':,, ~;

a "negative selectable" marker, one can readily determine that the locus of inser~ion of the recipient cell contains the precise sequence desired. The ef ficiency of æuch a vector is approximately 0.5%.

The region of heterology which may be introduced at the insertion site of the DN~ molecule may be either short or of substantial size (for example, 2 kb). The site o~

linearization may be 5', 3', or within the region of heterology. When the site of linearization is within the 0 region of heterology, the efficiency of gene targeting is approximately 2%.
The region of heterology may be located at a site internal to the region o~ homology where the desired recombination shall occur. Such a construct can be used when one desires to in~ro~uce a subtle mutation into a locus of the cellular gene at a site other than that of the site of desired recombination.

B. Use of a Different DNA ~olecule~ to Provide the Detectable ~arker Sequenc~ and ~he Desired Gene Seguence :

: ~ In a seoond preferred ~mbodi~ent, the detectab}e marker gene sequence, flanked by the regions o~ homology, will be provided to the recipient cell on a different DNA molecule from that which contains the desired gene sequence. It is preferred tha~ these molecules be linear molecules.
When provided on separate DNA molecules, the detectable marker gene sequence and the desired gene sequence will most ~ 30 preferably be provided ~o the recipient cell by co-electroporation, or by other equiYalent techniques.
. .
. .
, .
, :SUBSTITUTE $tllEET
- :

WO~2/11874 PCT/US92/00295 .` ,,l~j .

After selection o~ such recipients (preferably through th~ use of a detectable marker sequence which expresses the nptII gene and thus confers cellular resistance to the antibiotic G418), the cells are yrown up and ~creened to confirm the insertion event (preferably using PCR).
In the absence o~ any selection, only one cell in 107 would be expected to have the predicted recsmbinant structures. If, however, one selects for recipient cells which contain and express a detectable marker sequence (such as the n~II gene), it is possible to obtain a 103 to lOs fold enrichment for cells which have taken up both DNA
molecules. ~ypically, such enrichment enables one to identify the desired recipient cell (in which the introduced DNA has integrated into the cell's genome) by screening only 800 -l,500 cells. Such screening is preferably done using PCR, or other: equivalent methods. Using such negative selection techniques, one may manipulate the vector copy number.
The two introduced DNA molecules will generally not have : 20 integra~ed into the~same site in the genome of the recipient cell. Thus, in some cases, the desired gene sequence will have integrated in a manner~so as to rep}ace the nati~e cellular gene: sequence between the ~lanking regions of : homology. The locus of integ~ation of the deteatable marker gene is unimportant for the purposes o~ the present invention, provided it is not genetically linked to the locus of the p53 gene. If desired, however, it is possible to incorporate a gene sequence capable of. negative selection ~ along with the DNA c ontaining the detectable marker 30 sequence, Thus, one can ultimately~selec~ for cells which ;

SU~3STITUTE SHEET ~ :

: ~: : ; ~ : ` . .

WO92/11874 ~CT/US~2t0~295 ~! .
`9$~8~

have lost the introduced selectable marker gene sequenGe DNA.

C. U~e of Direct Selection to Identi~y ~o~ologous Reco~bination ~vent~ .

Although all of the above-described preferred . ~.
embodiments enable the isolation o~ cells in which one of the cell's p53 alleles has been mutated to contain a desired gene sequence, each embodiment requires the screening of a significant number of candidate cells in order to identify the desired recombinant cell. I~ is, however, possible to directly select for the desired recombinant cell by .-employin~ a variation o~ the above embodiments. ~- -The method for direct selection of the desired cells :
relies upon the phenotypic difference in targeted and non-targeted cells and the use of a single gene which can be used for both positi~e and negative selection. .
Typically, in any homologous.reco~bination experiment performed with an insertion vector, three populations of celIs wil} be created. The first class o~ cells will be those which have failed to receive the desired DNA molecule. ~ .
This class~will comprise virtually all of the candidate cells isolated on completion of the experi~ent. The second class of cells will be those c~lls in which the desired gene sequence has been incorporated at a random insertion site (i.e. a site other than in the p53 gene). Approximately one cell in 103-104 total cell~ will be in this class. The thir~
class of cells will be those cells in which the desired gene : 30 sequence~has~been incorporated~by homologous reco~bination .

SUIBSTITUTE SHEET

WO92/11~74 PCT/VS92/0029S

into a site in the p53 gene. Approximately one cell in 105-o6 total cells will be in this class.
In ~he above-described embodiments, the cells of the fir~t class tnon-trans~ected cells) can be eliminated by positive selection, thus necessitating the screening o~ only about l,OOO cells in order to identify the desired recombinant cell. ~n the present e~bodiment, cells of ~he third class (homologous recombinants) may be ~elected from the cells of the second clas~ (random insertions) i~ a phenotypic difference exists between the cells of the two classss.
Since random integration sites are likely to be concatemeric with few single copy clones (depending upon the DNA concentratian with which the cells were transfected), such integration events are inheren~ly unstable. Thus, such concate~eric constructs will typically undergo in~rachromo-so~al recombination. Such recombination will always leave one intact co W of the vector in the genome. Thus, all rando2 in~ertion events may be neyatively selected from the population i~ a negatively ~electable marker is included on the vector.
In contrast,:cells in which the dasired gene sequence :~ ~has been ;~incorporated into the: p53 gene by homologous recombination will revert with a:relatively high frequency (approximately l in lo4-lOs per cell division (depending upon the size of the duplicated structure) to produce a mutated p53 gene that does not contain vector sequences. Therefore, even if the~ vector contained a negatively selecta~le gene sequence, ;such~:ce~lls w~ urvive negative~ selection, and : ~ 30 can be recovered.: The ~small percentage of homologous . .

SUE~STITUTE SHEET
~: : : :

~VO~2~11874 PCT/US92/00295 2 ~ 8.~

recombinant ce}ls which have not undergone reversion will also be eliminated by the negative selection.
A preferred negative selectable marker is the hnr~ gene (cells e~pressi~g an active ~PRT enzyme are unable to grow in the presence of certain nucleoside analogues such as 6-thioguanin~ etc.). When using 6-thioguanidine as a negative ~election agent, a den~ity o~ 107 sells is : :
pre~erably u~ed since the efficiency of 6-thioguanidine selection is cell density depende~t. A typical experiment with 107 transfected cell~ would yield approximately lO
revertant cells after successive ~election. The relative yield of rev rtant clones can be substantially increased by using "~oly A Selection" ~or the first round of selection.
In such a "Poly A Sqlection!' one exploits the fact that, if an introduced DNA molecule were to integrate at random into the host chromosome, it would generally not integrate at a site adjacent to a nece6sary 3' polyadenylation site.
Thus, the mRNA produ~ed by the transcription of such randomly inserted constructs would generally lack polyadenylation. This fact can be exploited by using ; ~vectors which permit one to select for a recombinational avent that results in inte~ration adjacent to the natural polyadenylation 6ite of the introduced gene sequence (i.e.
by homologous recombination rather than by random insertion). As stated above, the frequency of obtaining a deæired recombinant cell~is approximately 10-3. By using Poly A Selection, desired cells can be recovered a~ a ~reguency of approximately lOZ. ~hus, the poly A selection results in an approxi~ate increase;of overall efficiency of nearly lO
fold. Poly A selection may therefore be advantageously used .

SUBSTIT~ITE S~EE~

.~ ; : . . .
, : ~ . ...

WO 9~/11874 PCl`tUS9~00295 in situations where one desires to minimize or avoid the screening of colonies to identify random versus homologous recombinants.

5D. Production of altered p53 alleles containing heterologou~ sequence~

As stated above, the desired gene sequence may be of any length, and have any nucleotide sequence. In particular, it is possible to design the sequence of the desired gene sequence in order to crea~e single, or multiple ba~e alterations, insertions or deletions in any preselected ~e~e of a cell.
If such changes are within a translated region of the p53 gene sequence, then a new protein variant of the p53 gene product can be~obtained.
: The present inventlon may be used to produce cells i~
~ which the n~tural p53 gene has been replaced with an altered ;: gene seque~ce, or a heterologoùs p53 gene~. A ~p53 gene is : 20 said to:be heterolog3us to:a transgenic:cell if:it is deriv-able from a~spe~ies:other~than;that of~the~transgenic cell.
In one embodiment~,~this~replace~en~ may be accomplis~ed :in::a single s~ep.~ To~accomplish such :replacement, ~ DNA
: molecule containing a~desired gene sequence and a region of homology with the p53 gene is introduced into a recipiPnt : ~cell. A selectable marker gene is also intrOdUCed into the cell, and use~ to select for cells whic~ have underwent recombination. ~he~method results in the replacement of the normal seguences~adjacent:to the~region of~homologY with the 30 : heterologous~ se~ ences~of~the~deslred~DNA~sequence.

: 3UEiSTiTUTEsHEET

WO9~118~4 PCT/~S92/0029S

0988~7 ~, ~ / .

In a second e~bodiment, this replace~ent may be accomplished in two steps. As in the embodiment described above, a cell is provided with a DNA molecule containing a desired gene sequence and a region o~ homology with the p53 gene. The DNA molecule also contains a selectable marker gene used to select for cells which have undergone a re~ombinational event that has re~ulted in the in~ertion of the introduced DNA molecule into their chromosomes at the site of homology.
Significantly, in this embodiment, the introduced DNA
molecule will also contain a "negative selectable" marker gene which can be u~ed to select for cells which undergo a second recombinational event that results in the loss of the inserted DNA.
A secon~ DNA molecule is employed to complete the gene replacement. This ~econd DNA molecule ne~d not contain any selectable marker gene. Upon receipt of the second DNA
molecule, a second recombinational event occurs which exchanges the "second" DN~ molecule for the integrated "first'i DNA molecule (including the desired DNA sequence, the selectable: marker sequence, and the "negative selec~able" marker sequence contained on that molecule).
:In another embodi~ent of the invention, subtle mutations may be introduced into a desired locus using~a l'cassette"
construct containing both a positive selection marker (such as the ÆE~II gene or the q~t gene) and a negative s~lection marker ~such as the ~k gene). In this embodiment, ona first uses the positive selection capaci~y of the cons~ruct to : ~introduce the:two:selection markers into a desixed locus.
~ :30 One then introduces the desired subtle mutations ( eUbsti-.
, : :

SUBSTITUTESHEET ~ ~
~ : .

W~ ~tllX74 PCT~US92/OOZ95 ., . ~

tutions, insertions, deletion~, etc.) by providing a cell with a DNA molecule that contains the desired mutation. By selecting for the loss o~ the "cassette" (using the negative selection marker~, one can select for recombinational events which result in the replacement of the "cassette" sequence with the DNA sequence containing the desired mutation.
The present invention may also be used to replace co~tiguous regions o~ a chromo~om~ with any desired gene sequence. ~hus, the present invention is not limited in the size of the DNA re~ions which may be altered or replaced.
This aspect of the present invention may be considered as a series of ~ steps. The first step in replacing a large region of a chromosome with a de ired sequence involves setting up an initial target. In this step, a recipient cell is provided with a DNA molecule which contains a "fir~t fragment" o~ the total desired replace~ent sequence. This "first fragment~ o~ the desired replacement sequance contains a selectable marXer sequence (most preferably the ntII gene3 at its end.
: 20 The DNA molecule also contains a "dual selection" gene sequence which encodes a non-functional fragment of a gene sequence for ~hich both a positive and a negative selection exists. An 2xa~ple o~ such a~gene is the g~ gene when used in :the context of an hE~' cell. CelIs which expresS a ~unctional qDt gene can be ~elected for by their ability to grow in HAT medium; Cells which lac~ a ~unctional q~ gene can be selected ~or by their abillty to grow in the presence o~ 6-thioguanine.
Homologous::recombination results in the insertion of the ~ DNA ~olecule ~into the cellls genome at ~he region of - -$1JeSTlT~TE S~ T
. ..

@~ , 209:~827 , homology. Importantly, since this step results in the creation of a cell whose genome contains the selectable marker gene, it is possible to select for the desired recombinational event.
In the second step o~ the method, a second DNA molecule is provided to the cell. This second DNA molecule contains a ~second ~rag~ent" of the dasired repla~ement sequence as well as a sequence of tha dual selection gene that, due to an internal d~letion, is incapable o~ encoding a functional ge~e product. Homologous recombination results in the insertion of the second DNA molecule into the cell's genome in a manner so as to crea~e a functional dual selection gene. Recombination also results in the inteqration of a non-functional fragment of the dual selection gene.
Importantly, sinc~ this step re ults in the creation of a cell whos~ genome contains a functional dual selection gene, it is possible to select for the desired recombinational event.
In the third:step of the~method, a third DNA mol~cule is 2~0 provided to the cell. This third DN~ molecule contains both : the:; ~'first" and ~second~ fragmen~s of the des~ired : replacement sequence.~ Homologou~ recombination results in the in~ertion of ~he third DNA mole~iule into ~he cell's genome in a manner so~as ~o delete the functional dual selection gene. The non-functional fragment of the dual : : selection gene (formed in step 2) is not affected by the recombination, and is retained. Importantly, since this step results in ~he creation of a:~cell whose genome lacks th-~dual selection;:~:g~ne,~ it: is possible to select for the desir d recombina~lona~l::event.:~
~ ~

SUI~STITUTE~SHEET ~

....

W092/11874 PC~/~Sg2/00295 '~

;~o~ g~2~7 In the fourth step of the method, a fourth DNA molecule is provided to the cell. This fourth DNA molecule contains a "third fragment" of the desixed replarement sequence as well as a sequence of the dual selection gene that, as in step 2, is incapable of encoding a functional gene product due to an internal deletion. Homologous recombination results in the insertion o~ the ~ourth DNA molecule into the cell's genome in a manner so as to create a functional dual selection gene. Racombination also results in the integration of a non-functional fragment of the dual selection gene. Importantly, since this step results in the creation of a cell whose genome contains a functional dual selection gene, it is possible to select for the desired recombinational event.
In the fifth step of the method, a fifth DNA molecule is provided to the ceIl. This fifth DN~ molecule contains both the "second" and nthird" fragments of the desired replacement sequence. Homologous recombination results in : the insertion of the fifth D~A molecule into the cell's genome in a manner so as to delete the functisnal dual selec~ion g-ne. ~ ~he non-~unctional fragment of the dual ~- selection gene (formed~-in~step 4) is~ not affected by the : reco~bination,~ and is retained. I~portantly, since this step results in the creation of a C211 whose genome lacks the dual selection gene, it i8 posfiible to select for the : desired recombinational event.
As will be appreciated, the net effect of the above-: described steps is to~produce a c~ll who:se genome~has been engin~ered to~ contain a~ nfirs~ "sec~nd,": and l'third"
~: 30 "fragment" of a particular desired gene ln a contiguous SUBSTlTilJTE SH~ET

:

. ., ; 2 ~

manner. The steps may be repeated as desired in order to introduce additional "fragments" into the cell's gennme. ~n this manner, cells can be constructed which contain heterologous genes, chromosome fragment~, or chromosomes, that could not be introduced u~ing a single vector. As indicated above, it is possible to select ~or each step of the method.

YX. The Production o~ Chi~eric and ~ransgenic Animals The chimeric or transgenic animal cells of the present invention are prepared ~y introducing one or more DNA
molecules into a pracursor pluripotent cell, most preferably an ES cell, or equivalent (Robertson, E.~., In: Cu~ent Co~m~liLcations ~ , Capecchi, ~.R. (ed.), Cold Spring Harbor Press, Cold Spring Harbor, NY (1989~, pp.
39-44, which reference is incorporated herein by reference).
The term "precursor" is intended to denote only that the pluripotent cell is a precursor to the desired ("transfected") pluripotent cell which~ is prepared in accordance with the teachings o~ the present inventionO The pluripotent ~precursor or tran3fected) cell may be cultured iIL~LLYe, in a manner known in the art (Evans, M.J. e~ al., ~ 154-156 ~1981)) to form a chimeric or transgenic 2~ animal.
Any ES cel} may be used in accordance with the present invention. It i5, however, praferred to use primary isolates of ES cells, Such isolates may be obtained dire tly:~ro~ e~bryos~such as the CCE cell line disclosed by 30 Robertson, E.J., In: Cur~e~ ions in ~e~ular .

.

: : :
~ ~T~I iTF ~ ~EE~r W~92/llX74 PCT/US92100295 2'0`'~'~'J~'æ''~`''''`"

Bioloqy, capecchi, M.R. (ed.), Cold Spring Harbor Press, Cold Spring Harbor, NY (1989), pp. 39-44), or from the clonal isoIation of ES cells from the CCE cell line (Schwartzb~rg, P.A. ~ , Ç~ 799-~03 (1989), which reference is incorporated herein by r~f~rence). Such clonal isola~ion may be accomplished according to the method o~ E.J. Robertson (In: ~
, (E.J. Robertson, ~d.), IR$
Press, Oxford, 1987) which reference and method are incorporated herein by reference. The purpose of such clonal propagation is to obtain ES cells which have a greater efficiency for dif~erentiating into an animal.
Clonally selected ES cells are approximately lO-fold more effec~iva in producing transgenic animals than the progenitor cell line C~E. For the purposes o~ the recombination ~ethods of the present invention, clonal se}ectlon provides no advantage. An example of ES cell lines which have been:clonally derived from embryos are the ES cell lines, AB1 ( ~t~) or AB2.1 (~
20: : The ES cells are preferably ~ultured:~on stromal cells (such as STO cells ~ (especially SNC4 STO cells) and/or primary embryonic fi~broblast :~ cellsj as described by E.J.
Robertson ~(In: Tera~Qca~ir~2mas and E~L~rYonic S~er~ Cells: A
.I=i~: h, ~E.J. Robertson, Ed. ), IRI. Press, Ox~ord, 1987, pp 71-112), which~reference is incorporated herein by re~erence. ~ethods for the production and analysis of chimeric mice are di~closed by Bradley, A. (In:
Tera~Qç ~ as . and ~mbryQr1ic ~ CeIls: A _Practical Anproach,~(E.J.~Robertson, Ed,), }~L Press, Oxford, ~987~, pp 113-151), ~which re~orence is inco ~ orated herein by ' -SU~STltUTE SH~T~ :

WO92/11~74 PCT/US~2~00~95 ' ~9~827 reference. The stromal (and/or fibrob~ast) cells serve to eliminate the clonal overgrowth of abnormal ES cells. Most . preferably, the cells are cultured in the presence of }eukocyte inhibitory factor ("lif") (Gough, N~. a~
Rep~Qd ~e~iL~_~a~ 28l-288 (1989); Yamamori, Y. ~
Science 2~46:1412-1416 tl989), both of which re~erences are . incorporated herein by re~erence). Since the gene encoding ; lif has been oloned (Gough, N.M. e ~ ~ bL ~O~L~
Dev. 1:281-288 (1989)), it is especially pre~erred to transform stromal cells with this gene, by means known in the art, and to then culture the ES cells on trans f ormed s~romal cells tha~ secrete li~ in~o the culture medium.
ES cell lines may be derived or isolated from any species (for example, chicken, e~c.), although cells derived or isolated from mam~als such as rodents (i.e. mouse, rat, hamster, etc.), rabbits, sheep, goats, fish, pigs, cattle, primates and humans are preferred.
::
VII. Uses o~ the Present Invention ~ .
The present invention provides human or animal ce}ls :
which contain a desired gene sequence in one of the two p53 alleles of the celI's genome.
In a first e~bodiment, the invention also provides a means for producing non-human chimeric or transgenic animals whose cells contain ~uch a s~quence. The an~mals which may be produced ~hrou~h applicat~on o~ the described method include chicken, non-human mammals (especially, rodents (i.e. mouse, rat, hamster, etc~), rabbits, sheep, goats, :30 ~ish, pigs, cattle and non-human pr~mat~5).
. .
:: , ;: . .

SU~3STiTUTE SHEET ~ ~ -W~92~11874 P~T/US92/00295 7 j~

The cells and non-human animals o~ the present invention have both diagnostic and therapeutic utiliky.

A. Diagno~tic ~tility Since the invention provides a cell, or a transgenic or chimeric non-human anima~ that contain~ a single functional allele of the p53 gene, and since such cells will b~come tumor cells upon the mutation of the func~ional allele to ~
non-functional form, the present invention can be used to identify an agent that is capable of affecting a character-istic of an animal cell that is attributable to the presence or expression of a tumor-suppressing gene. A characteristic of an animal cell is said to be "attributable to the presence or expression of a tumor-suppressing gen~," if the characteristic is altered by the absence or lack of expression of the tumor-suppressing gene. ~xamples of such characteristics include tumorigenesis, resilience to tu~origenesis, the : extent, distribution, incidence, 20 location, grade, etc. o~ tu~ors, etc. . ~ :
In :one e~bodiment, such agents can decrease the tumorigenic ~or neoplas~ic) potential of ~he ce}ls or animals. Such agents are discussed below with regard to the therapeutic potential of the invantion.
In a second embodiment, such~ agent can increase the tumorigenic ~or neoplastic) pote~tial of the cells or animals. Thus, the cQlls and non-human animals of the present invention have uti~lity~in testing po~ential or suspected carcinoge~s for:~tumorigenic acti~ity. They~may be used :to identify and assess~ the ~u~origenic efPects of : :

E;l.IE35TITUTE SH~

wos2/l1874 PCT/US92/00295 ~, agents that may be present, for example, in the environment (such as environmental pollutants in air, water or soil), or resulting ~rom environmental exposures to chemicals, radioisotopes, etc. They may also be used to facilitate studies of the effects of diet on oncogenesis. They may be usad to determine whether potential or present food additives, chemical waste products, chemical proces~ by-prod~cts, water sources, proposed or presently used pharmac~uticals, cosmetics, etc., have tumorigenic activity.
~They may also be used to determine the tumorigenic potential of various ener~y forms (such as W rays, X-rays, ionizi~g radiation, ga~ma rays of elemental isotopes~ eto.~.
The frequency at which a mutational event occurs is dependent upon the concentration of a mutagenic chemical agent, or the intensity of a mutagenic radiation. Thus, since the frequency of a single cell receiving two mutational events is the sguare of the fraquency at which a single mutational event will occur, the cells and non-human : animals of the~pre~ent invention shall be able to identify neoplastic (mutagenic) agents at concentrations far below those needed to~induce neoplastic changes in natural cells or animals.
One especially preferred cell is a non-human cell in which one of~the natural p53 alleles has been replaced with Z5 a functional human p53 ~lle:le and the other of the natural p53 alleles~ha~ baen mutated to a non-functional form.
Alternatively, one may amp}oy a non-human cell in which the two natural p53 alleles have~be-n replaced with a functional : and a ~on-functional~211elo:`0f the human~p53 gene~.
, ~ ~ ~WE~STITUTE;SHeET ~;

, ~ ' 2~)~g~3~7 ''' Such cells may be used, in accordance with the methodsdescribed above to assess the neoplastic potential of agents in cells containing the human p53 allele. ~ore preferably, such cells are u~ed to produce non-human animals which do not contain any natural functional p53 all~les, but which contain only one functional human p53 allele. 8uch non-human animals can be used ta as~ess the tumorigenicity of an agent in a non~human animal expressing the human p53 gene product.

1. In yi~ro As~ay . , .
In one em~odime~t, one ~ay employ the cells of the present invention, in in vi~r_ cell culture, and incubate such cells in tXe presence; of an amount of the agent whose tu~origenic potential is to be measured. This embodiment therefore comprises an in vitro assay sf tumorigenic acti~ity.
Although many carcinogenic agents ~ay directly mediate their ~tumoriy-nic e~fects, ;~ome agents will not exhibit tumorigenic po~ential until metabolized, or until presented to a :susceptible cell~ along with~ one or :more "co-carcinogenic" ~actors~. ~he pre~ent invèntion permits the identification of such nlatent" carcinogeni~ and "co-carcinogenic" agents. In accordance with this embodiment o~the invent~on, the pr~sence:of a:"la~ent" carcinogen can be identified by~merely maintaining cell or animal expo~ure to a candidate agent.. Alternatively, the~cells of the present invention:can:`be i~cubated~in "sp-nt" culture~medium ;(i.e.
:~ 30~ medium contain~ing ~he ~candidate agen~ that ~was used to :~
....
, SUBSTI~UTE S~tEE~

~, , ~.
2a9~2~

-65- .:

cultur~ other cells before being used to culture the cells of the prese~t invention).
The present invention permits the identification of co-carcinogenic factors capable of inducing ne~plastic effects in the presencR of a second agent. Such factors can be identified by culturing the cells of the present invention in the pre~ence of two or more candidate agents simultaneously, and then assaying for neoplasia.
The transfor~ation of the cells to a neoplastic state would be indicative o~ tumorigenic (or neoplastic) activity of the assayed agent. Such a neoplastic state may be - evidenced by a change in cellular morphology, by a loss o~
contact inhibition, by the acquisition of the rapacity to grow in soft agar, or 1~108t preferably, by the initiation of 15 expression of tumor antigens.
The use of tumor antigens as a means of detecting neoplastic activity is preferred since such antigens may be readily detected.
As is well known in the art, antibodies, or fragments of antibodies, may be used to quantitati~ely or qualitatively : ~ detect: the presence tumor of antigens on cell surfaces.
Since any cell type ;~i.e. lung, kidney, colon, etc. ) may be . .
employed to form ~ the pS3- mutated cells of ~he present : .
invention, it is possib}e to determine whether an agent has a tissue specific tumorigenia potential. To accomplish this goal, one would incubate a candidate agent in the presence of;p53-mutated cells derived from~any of a variety of tissue types. Since ~umors have tumor~specific antigens, and since antibodies capable of~ binding to such antigens have been : 30 isolated, it~ i8 possible: to use such~ an~ibodies to ~:
~ : , :; :: ~:

WO92/tl874 P~T/US92/00295 ~ ` ' ` ' 1 ` ! ` .
2 ~
-6~-characterize any tu~or antigens which may be expressed by the p53-mutated cells.
Such detection may be accomplished u~ing any of a variety of immunoassays. For example, by radioactively labeling the antibodies or antibody fragments, it i5 possible to detect the antigen through the uqe of radioimmune assays. A good dascription of a radioimmune assay (RIA) may be found in abor~ory Tech~ es ~nd ~iochemistry in Mole~ula~ ~iolQ~y, by Work, TuS~
North ~olland Publishing Company, NY (1978), with particular reference to ~he chapter entitled "An Introduction to Radioimmune A~say and ~elated Techniques" by Chard, T., incorporated by reference herein. Examples of suitable radioisotopic labels include 3H, 1~1In, 125I, 131I, 32p, 35S, 14C, Cr, 57To, 58Co, 59~e, ~Se, 152EU, ~y, 67CU 217Ci 211At Z12p~
47Sc 1~Pd ~tc A' Alternatively, enzyme labels, non-radioa~tive isotopic labels, fluorescent labels, c~emiluminescent labels or other suitable labels can be employed.
Examples of suitable enzy~e labels include malate dehydrogenase, staphylococcal nuclease, delta-5-steroid isomerase, yea~-alcohol dehy~rogena~e, alpha-glycerol phosphate dehydrogena e, triose phosphate isomerase, peroxida8e, alkaline phosphatage, asparaginase, glucose oxidase, beta-galacto~idase, ribonuclease, urease, catalase, glucose-6-pho~3phate dehydrogena~;e, glucoamylase, acetylcholine esterase, etc.
Examples of suitable non-radioactive isotopic labels include 157Gd, 55Nn~ 1eDy S~rr ~Fe ~t " ,.

- SUI~STITUTE SHg~ET

: . :
W092/lt874 PCT/US92/002g5 e,~ y 2 ~ $.~
.' . ' ;.

Examples of suitable fluorescent labels include an 152Eu label, a fluorescein label, an isothiocyanate label, a rhodamine label, a phycoerythrin label, a phycocyanin label, an allophycocyanin label, an o-phthaldehyde label, a fluorescamine label, etc.
Examples of chemiluminescent labels include a luminal label, an isoluminal l~bel, an aromatic acridinium ester label, an imidazole label, an acridinium salt label, an oxalate ester label, a luciferi~ label, a luciferase label, an aequorin la~el, etc~
Those o~ ordinary skill in the art will know of other suitable labels which may be employed in accord~nce with the `present invention. The binding of these labels to antibodies or frag~ents thereof can be accomplished using standard techniques commonly known to tho~e of ordinary skill in the art. Typical technique5 are described by Kennedy, J.H., et ~l. (Clin.~Chi~. Acta 70:1-31 (1976)), and Schurs, A.H.W.~., et al. ~ClinO ~him. Acta 8~ 40 (1977)).
Coupling techniques mentioned in the latter are the glutaraldehyde method, the periodate ~ethod, the di~aleimide method, the ~-maleimidobenzyl-N-hydroxy-succinimide ester method, all of which methods ar~ incorporated by reference herein.
The above-described ln ~itro assay has the advantageous feature~ of potentially lower cost than presently used assays, and the capacity to readily screen large numbers of agents. Use of this~embodi~ent facilitates comparisons of test results obtained at different times and conditions.
~:Moreover, becaus~ it~is possib~e~to u~e very large numbers : 30 Or cells in~ such asgay8~ i~ is possible to detect the .~. .

SUE~STITUTE SHEET

WO92/1187~ pcT/us92/oo29s ~ ' .
2098..827.. .
,, ~ " !

~68-tumorigenic activity of tumorigenic agents even at very low concentrations. Lastly, since this embodiment can be per~ormed using human cells, it provides a means for determining the tumorigenic (or neoplastic) potential of a test compound on human cells.

2. I~_Y~ Assay~

In a seoond embodiment, one may employ the non-human animals of the present inven~ion, and provide to such animals (by, ~or exa~ple, inhalation, ingestion, injectionj implantation, e~c.) an amount o~ the agent whose tumorigenic potential is to b~ ~easured. The formation of tumors in such animals (as evidenced by direct visualization by eye, or by biopsy, 1maging, detection of tumor antigens, etc.) would be indicative of tumorigenic activity of the assayed agent.
The use of the non-human animals of~ the present invention is pre~erred over na~urally occurring non-human animals since such natural animals contain two ~unctional p53 a1leles, and thus~require two~mutational~events in order to lead to loss o~ ~unctional p53 activity~. In contrast, since the non-hu~an animals of ~he present invention have only one functional p53 allele, only one ~utational event is needed to cause total lo85 ~of p53 funotion.
The d~tection of tumors in such ani~als can be accomplished by biopsy, imaging, or by assaying the animals for the presence of cells which express tumor antigens.
For~ exa~ple, such d~tection may be accomplished by re~ovin ;a samplo of tis~ue from a subject and then treating~
:
, .

WO92~11874 PCT/US92~00295 .,; .; ~, ~
2~827 the isolated sample with any suitably labeled antibodies (or antibody fragments) as discussed above. Preferably, such in situ detection is accomplished by removing a histological specimen from the subject, and providing the labeled antibody to such specimen. The antibody (or fragment) is preferably providad by applyi~g or by overlaying the lab~led anti~ody (or fragm2nt) to a sample of tissue. Through the us~ of such a procedure, it is possible to determine not only the presence o~ antigen, but also the distribution of the antigen on the examined ti~sue. Using ~he present invention, those of ordinary skill will readily percei~e that any of a wide variety o~ histolo~ical methods (such as staining procedures) can be ~odified in order to achieve such in sit~ detection.
Alternatively, the detection of tumor cells may be accompli~hed by in v~o imaging techniquQs, in which the la~eled antibodies (or fragments thereof) are provided to the æubject, and the presence o~ the tu~or is detected without the prior removal of any tissue sample. Such in vivo : 20 detection procedures have the advantage of being less invasive than other detection methods, and are, moreover, capable of detecting the presence of antigen-expressing cells in ~issue which cannot be 2asily removed from the patient.
Additionally, it i~ possible to assay ~or the presence of tu~or antigens in body fluid~ ~uch as blood, lymph, etc.), stools, or cellular extracts. In such immunoassays, the antibodies (or antibody frag~ents) may be utilized in liquid phase or bound to a solid-phase carrier, as de~cribed :~ 30 b~low.
~ :
, , . . .

: ~ SUB~;TITIJTE SHEEI ~

WO92/1~874 PCT/US92/00295 ., ,,, ~ ~.
2 ~

--, o--The use of an in Vi~Q assay has several advantageous features. The in ViYo assay permits one not only to identify tumorigenic agents, but also to assess the kind(s) of tumors induced ~y the agent, the number and location (i.e. whether organ or tissue specific) of any elicited tumors, and the grade (clinical slqnificance~ of such elicited tumors. It further pe~mits an assessment of :
tumorigenicity whioh inherently oonsiders the possible natural metabolism of the introduced agent, the possibility that the introduced agent (or it~ metabolic by-products) might selectively accumulate in speci~ic tissues or organs :`~
of the recipient ani~al, the possibility that the recipien~
animal might recognize and repair or prevent tumor ;-formatio~. In short, such an assay provides a true biological model for studying and evaluating the tumorigenio potential of an agent in a living non-human animal.

3. ~oa88ay8 of Tu~or Antigens .
The in vi~o, in ~i~u, or i~l vivo detection of tumor antigens using::antibodies (or fragments:of antibodies) can .:.
: : ~ be improved ~through the use o~ carriers. Well-known carriers includ~ ~gla~s, polystyrene, polypropylene, polyethylene,~dextran, ny~on, amylases, natural and ~odified celluloses, polyacrylzmides~, agaro8es, and magnetite. The nature of the carrier can be ei~her 801uble to some extent or insoluble for the pUrpo8e8 of the present invention. The suppor~ material ~ay have virtually any possible structural :con~iguration ~o~long as the~coupled molecule is cap:~ble of .-~ binding:to sn antigen.~m us, t~e 6upport~configuration may .

~: ~

c~ TlTl JTE~ ~il 4EE~

~ .
2 ~ 2~i ~ . . ., s s .

be spherical, as in a bead, or c~}indrical, as in the inside surface of a test tube, or the external surface of a rod.
Alternatively, the surface may be flat such as a sheet, test strip, etc. Those skilled in the art will note many other suitable carriers for binding monoclonal antibody, or will be able to ascertain the same by use of routine experimentation.
The binding molecules of the present invention may also be adapted for utilization in an immunometric assay, also known as a "two-site" or "sandwich" assay. In a typical immunometric assay, a quantity of unlabeled antibody (or fragment of antibody) is bound to a solid ~upport that is insoluble in the fluid being tested (i.e., blood, lymph, liquified stools, tl~sue homogenate, etc.) and a quantity of detectably labeled soluble antibody is added to permit detection and/or quantitation of t~he ternary complex formed between solid-phase antibody, antigen, and labeled antibody.
Typical immunometric assays include "forwardi' assays in which the an~ibody bound to ~he solid phasa is first contacted with the sample being tested to extract the :~ antigen fro~ the sample by~formation of a binary solid phase an~ibody-antigen complex. A ff er a suitable incubation period, the ~olid support is washed to remove the residue of the fluid sample, including unreac~ed antigen, if any, and then contacted with the solution containing an unknown quantity of labeled an~ibody (~hich functions as a "reporter molecule")~ ~fter a second incubation period to permit the labeled antibody to complex with the antigen bound to the :~ solid support through the unlabeled antibody, ~the solid : : 30 support i8 wa Xed: a second time to remove the unreacted .
.

, T~ JTE Sl4EET
.

WO 92/11874 PCI`/VSg2/0029~;

labeled antibody. This type of forward sandwich assay may be a simple "yes/no" assay to determine whether antigen is present or may be made quantitative by comparing the measure of labeled antibody with that obtained ~or a standard sample S containing known quantities of antigen. Such "two-site" or "sandwich" assays are described by Wide at pages 199-206 o~
R~dioi~e ~say MethQ~, adited by Kir~lam and Hunter, E.
S. Livingstone, Edinburgh, 1970.
In another type of "sandwich" assay, which may also be useful to identify tum~r antigens, the so-called "simul-taneous" and "reverse" assays are ~ed. A simultaneous assay involves a single incubation step as the antibody bound to the solid support and labeled antibody are both added to the sample being tested at the same time. After }5 the incubation is completed, the solid support is washed to remove the residu~ of fluid sample and uncomplexed labeled antibody. The pre~ence of labeled antibody associated with the solid support is then determined as it would be in~a conven~ional ~'forwardH sandwich assay.
In the "r verse" assay~, stepwise addition first of a s~lution of labeled antihody to the ~luid sample followed by : the addition of unlabeled antibody bound to a solid support after a suitable incubation pe~iod is utilized. After a second incubation, the~solid phase is washed in conventional rashion to free it of the residue of the sa~ple being tested and the solution o~ unr~act~d labeled antibody. The determination of labe}ed antibody associated with a solid support is then determined as in the "simultaneous" and "forward" assays.
' ~

.

:SUE~STITUTE SHEET

W~92/11874 PCT/US92/00295 20988`27 As explained above, the i~munometric assays for antigen require that the particular binding molecule be labeled with a "reporter ~olecule." These reporter molecules or labels, as identified above, are conventional and well-known to the art. In the practioe of the present invention, enzyme labels are a preferred e~odiment. No single en2yme is ideal for use as a label in every conceivable immunome~ric assay. Instead, one must determine which enzyme is suitable for a particular assay sy~tem. Criteria impar~ant for the choice of enzymes are turnover nu~ber of the pure enzyme (the number o~ substrate molecules converted to product per enzyme site per unit of time), purity of the enzyme preparation, sensitivity of detection of i~s product, ease and speed of detection of the enzy~e reaction, absence of ~.
interfering factors or of enzy~e-like activity in the test fluid, sta~ility of the enzy~e and its conjugate, availability and cost of the enzyme and its conjugate, and the like. Included among the enzy~es used as preferred labels in~the immunometric~:~assays of the~present~:invention ~20 are peroxidaY , alkaline p~osphatase, be~à-galactosidase, urease, glucose oxidase, ~lycoamylase, malate dehydrogenase, and gluco~e-6-phosphata dehydrogenase. ured e is among the -~
more pre~erred~enzyme labels, particularly because of ~:~: chromogenic pH indicàtors which make its activity readily visi~le to the nake~ eye.

B~ TherapQutic Utll~ty ` .

: Signlficantly,~ the:~cells and animals of:t~e precent ~inventlon cdn be ~used; to ldentify dgents~thdt decrease the : ~ `

SUBS ~ iTUTE SHEET

WO92/11874 PCT/USg2100295 ~,~
2 0' tumorigenic (or neoplastic) potential of the cells or animals. Such agents can be "anti-tumor agentsl~ and/or "chemoprev2ntative agents." 'IAnti-tumor agents" act to de~rease the proliferation of the cells (or the growth, ::
dissemination, or metastasis of tumors in the chimeric or transgenic animals~. "Chemopreventative agents" act to inhibit the formation of new tu~ors. Such agents may have general activity (inhibiting all new tumor formation), or may have a specific activity inhibiting the distribution, frequency, grade, etc.) of sp~cific types of tumors in speGi~ic organs and tissue. Thus, the present invention permits the identification of novel antineoplastic therapeutics. Any of the above assays o~ tumor-suppressing activity may be used for ~his purpose.
~he transgenic cells and non-human animals of the present invention can be used to study human gene regulation . .;
oP the p53 gene. For example, such cells and animaIs can be used to investigate the interaction3 of the p53 gene with oncogenes or other tumor ~uppre~Bor gene t- ~hus, they may be used to identify~ thterapeutic ag2nts which have the ability to impair or prevent neoplastic or tumorigenic developmen~. Such agents have utility in the treatment and ::
~:: cure of cancer in humans and:animal5. :.
: Significantly, potential therapeutic agents are : 25 fre~uently found to i~duce toxic effects in one animal model but not in another animal ~todel. To resolve the potential of such agents, it i8 o~ten n~cessary to determine. the me~abolic pat~erns in various 8pecies, and to then de~ermine the toxicities:o~ thte~metabolites. The pre~ent invention ',,":

: : :

SI~E~;TITUTE SHEET

;: :

WO92/t1874 PCT/US92/00295 2~g~827 -75- .

permits one to produce transgenic cells or animals which could facilitate such determinations.
When providing the therapautic agents of the present invention to the cells of an animal, pharmaceutically acceptable carriers (i.e. liposomes, etc.) are preferably employed. Such agents can be ~ormulatPd according to known mathods to prepare pharmaceutically useful compositions, whereby these materials, or their functional derivativ~s, are combined in admixture with a pharmaceutically acceptible carrier vehicle. Suitable vehlcles and their for~ulation, are described, for example, in Nicolau, C. et al. (ÇEi~
Rev. ~her. Druq Carrier Svst. 6:239-271 (1989)), which re~erence is incorporated herein by re~erence.
In order to form a pharmaceutical.ly acceptable composition suitable for ef~ective administration, such compositions will contain an effective amount o~ the desired gene sequence together with a suitable amount of carrier vehicle O
Additional phar~aceutical methods may be employed to control the duration of ac~ion~ Control release preparations may be achie~ed through the use o~ polymers to :complex or absorb the desired yene sequ2nce (either with or without any a sociated carrier). The controlled dslivexy - may bQ exercised by selecting appropriate macromolecules (for example polyesteræ, polyamino acids, polyvinyl, pyrrolidone, ethylenevinylacetate, methylcellulose, carboxymathylcellulose, or protamine, sul~a~e) and the concentration of ~acromolecule~ as well as the methods of ~ incorporation in order to control re:lease. Anoth~r possible 30 method to control th~ duration of action by controlled ;:

-.
:

- ~ SUE3STITUTE S~EET

W~ 9~tl l874 . PCT/VS92/00295 2 0 9 8 8 ~ 7 release preparations is to incorporate the agent into particles of a polymeric material such as polyesters, polyamino acids, hydrogels, poly(lactic acid) or ethylene ~inylacetate copolymers. Alternatively, instead of incorporating these agents into polymeric particles, it is possible to entrap the~e materials in microcapsules prepared,.for example, by coacervation technique~t or by interfacial polymerization, for example, hydroxymethylcellu-lose or gelatine-microcapsules and poly(methylme~hacylate) microcapsules, respectively, or in colloidal drug delivery systems, for example, liposomes, albumin micro~pheres, microemulsions, nanoparticles, and nanocapsules or in macroemulsions.

C. ~ in Re~earch and in Gene Therapy The cells and non-human animals of the present invention, quite apart from their uses in veterinary and human medicine, may be used ko investigate gene regulation, expression and organization in animals. The methods of the present inve~tion may be used to produce~alterations in a regulatory region for the native p53 gene sequence. Thus, the invention provide~ a means ror alteri.ng the nature or : contrvl of transcription or tr2nslation of the pS3 gene, and o~ altering the pS3 gene its~l~. For example, the invention enables one to introduce ~u~ations which result in increas~d or decrea8ed gene expreasion. Similarly, it enables one to impair or enhance the transcriptional capacity of the natural p53 gene i~ order to decrease or increa6~ its :

:

: .- el l~C!tTlT~ IT~ C~t~

W092/11~74 PCT/US92tO0295 ~8~

expression. Thus, the present invention p~rmits the manipulation and dissection o~ the p53 gene.
Such abilities are especially valuable in gene therapy protocols, and in the d~velopment of improved animal mod~ls of cancer.
In one embodiment o~ the present invention, DNA encoding either a functional pS3 gene, variants of that gene, or other genes which in~luence the activity of the p53 gene, . may be introduced into the somatic cells of an animal (particularly mammals including humans) in order to provide a treatment for canc~r (i.e. "gene therapy"). Most preferably, viral or retroviral vectors are emplo~ed for this purpose.
The principles of gene therapy are disclosed by Oldham, R.K. tIn: Pri~çiDles Q~_~iotherapy, Raven Press, NY, 1987), and si~ilar texts. Disclosures of the me~hods and uses for gene therapy are pro~ided by Boggs, S.S. (Int.J . Cell CLQn.

8:80-96 (l990)); Karson, E.M. (Biol~ ReprQd. 42:39-49 (1990)); Ledley, ~.D., In: Bio~ç$hnologyl_A Comprehensive Trea~isffæ,~oiLl~g_zB ! ~ene ~echnolooy, VCH Publishers, Inc.
NY, pp 399-~58 (19R9) ); all of which references are incorporated herein by reference. ~ .:
Although, a~ indicated above, such gene therapy can be provided to a recipient in order to treat (i.e. suppress, attenuate, or cau~e regre~ion) an axisting n~oplastic state, the principle~ o~ the present invention can be used to provide a prophylac~ic gene therapy to individuals who, due to inherited genetiC mutatlons, or somatic cell ~uta~ion, contain cells~having i~paired p53 gene expression (~or exa~ple, only a ~ing1e ~ ~ ctional allele of ~he p51 ':

J~ J ~ ~T

WO92/~1874 P~T/US92/00295 ':

-7~-.. . . ..
gene). Such therapy would be administered in advance of the detection of cancer in order to lessen the individual's predisposition to the disease.
Having now generally described the invention, khe same wil} be more readily understood through reference to the following examples which are provided by way of illustra~ion, and are not intended to be limiting o~ the prQsent invention, unless specified.
, INACTIVATION OF 'TE~ p53 '~UN~ S~PPRESSOR G~ME
IN TE~ MO~S~: OVER~I~W OF ~E~ EYP~RINE~T

15Structural alterations of the p53 ~umor suppressor gene have b~en associated with a wide array of human cancers. 'ro examine the role of p53 in tumorigenesis and mammalian development, embryonic stems (ES) cell lines were generated in which one of the two endogenous p53 alleles has been - -inactivated by the insertion of a neo~i gene (ea. the n~tII
gene of tn5) following ho~ologous recombination.
The gene targeting strategy utilized a murine p53 3.7 kb genomic construct flanked by an HSV TK gene and interrupted in exon 5 by a polyA - neoR gene driven by a ~Ql_11 pro~oter.
2S The 3.7 kb fragment is derived from a genomic plasmid clone of murine p53 (Oren, M. s~ , EMB~ J. ~:1633-1639 (1983);
Pinhasi e~ Molec. Cell. Biol. 4.2180-2186 (1984)). The DN~ was cloned ~rom a liver cell of a normal ~alb/c mouse.
Following gen~ transfer into ES cells and G418/FIAU
selection ~as described~;abov~), two clones were identified ~by PCR and Southern analysis;~which have the expected altered ~: . .
~:

:: . . ..:

~ CIIR~TITIITt S~EE~r WO92/11874 PCT/US92/0~295 `` 20~2~
.... I ..~ .

,9 p53 gene structure (Figure 3). Three probes were used for this purpose. The first probe (neo probe) indicates the presence of the neomycin resistance determinant in the stem cells. The second and third probes (exon 1-2 and exon 5-10 S probes) reveal tha~ the wild ~ype and mutant alleles are present in the two embryonic stem cells, but that the mutant allele is not present in the wi1d type ~Wt) cell. The ~act that the three probes identi~y the same bands i~ the mutant embryonic stem cells indicate~ that, as expected, the neo determinant in these cells is linked to the p53 gene.
one of the heterozygous ES clones was injected into C57BL/6 recipient blastocysts and implanted into pseudopregnant CS7BL/6 female mice. This ES clone was designated ES~p53.
15Embryonic stem cell line ESQp53, (containing the p53 targeting construct of Figure 2, as verified in Figure 3) was deposited with the American Type Culture Collection, Rockville, MD. on December 27, 1990 and was accorded the ATCC acce sion number: CRL 10631. ~ -20The offspring resulting from the injection of the above-described ES cell into a blastocyst, an~ the : implantation of the blastocyst into a mouse uterus, were :highly chimeric. As discussed below, the chimeric males have been bred, and germ line het-rozygotes have been generated. These heterozygotes are then examined ~or increased tumor susceptibility and bred to deter~ine the effects of a p53 nuIlizygous gtate on embryonic development.
; ' . ~ ', ; i~ ~ . , . ':
' :

' ~' ' ' .

.

-q? ~ TlT!~TF~ SI~EET

WO 92tl 1 1374 PC~/US9;~/00:~95 209~8~7 SP~ 2 TEIE GE:NER.ATION OF q~E RECONINOG~NIC p53 ~ONS~lRVCTS

To maximize an ability to obtain and selet ES cells containing homologous recombination even~s, the strategy of Mansour et a~. was utilized tMansour, S.L. e~_a~ re 336:348-352 (1988), herein incorporated by reference). The procedure used a positive selection method for isolating cells that had stably integrated the introduced mutating DNA
and a negative selection agains~ cells that did not contain a homologous recombination event, allowing an enrich~ent for cells containing the desired homologous recombination event.
. The adaptation of these proc~dures for obtaining homologous reco~bination into the p53 gene is outlined in Figure 1.
15As indicated above, a segment of the genomic sequences from the mouse p53 gene ~a 3.7-kb frag~ent spanning exons 2 to 10 of the 11 exon ~ene) was obtained from plasmid pSVpcG3 ; (Pinhasi et al~ olec~._Ce~L Biol. ~:2180-21~6 (1984)). As shown in Figure 1, this ~equence of DNA wa~ ~odi~ied in two ways. First, a neo marker gene driven by the:MCl promoter e ~ ancer wa~ designed to obtai~ high le~els of expre~sion in ES cells (Thomas, K.R.~ et ~ 503--512 (1988)), This sequence was in~erted into the unique Bal 1 si~e in exon 5. ~his insertion provided a po~itive ~electable marker (neo) for ~table gene tr~n~er into ~S cells and disrupted the coding sequQnce o~ ps3, producing an inactive allQle ~ollowing success~ul ho~ologous recombination. The second alteration entailed the at~achmen~ of a herpes simplex ~irus thymidin~ kina~e gene (HSV TX) to the 3' end o~ the gene-targeting construct (~igure 2). Thi~ attach~ent ..

:: : SUE~ITUTF S~EE~T

WO92/11874 PCT~S92/00295 ~$~27 provided the negative selection (using the HSV-TK-speci~ic thymidine analogue FIAU (1-(2 deoxy, 2 fluoro, ~-D
arabinofurano~yl)-5-iodouracil) against cells that have random integrationæ of the targeting construct (Figure 2).
EXa~PL~ 3 T~ANSFER OF CON5~R~C~S INTO BS C~LLS AND ID~NTIFICATION OF
CFT,LS CON~AINING ~OLOGO~S RECO~BINATION ~V2NTS

After generating the above-described p53 tarqeting construct (figure 2), the construct was introduced into cultured ES cells by electroporation. ES cells were cultured as de~cribed by ~.J. Robertson tIn: Ter~ocar-~inom~s and Emkrvon~c s~e~-5~lL~LL~ ac~içal B~Lgas~, 15(E.J. Robertson, Ed.), IRL Press; Oxford, 1987, pp 71-112).
Cells for electroporation were collected by trypsinization at 60-80~ confluency, ~edimented and resuspended in buffered saline with DNA at 25 U/ml. 107 cells/ml are treated with a single pulse from a Bio-Rad Gene Pulser (~240 volts, 500 uF, 0.4 cm cuvette) in order to achieve elec~roporation.
Under these conditions, efficiencies of stable gene transfer of approximately 103 were obtainad. After~pulsing, the ES
cells were pla~ed onto feeder cells (as described aboYe~ at 5 X 106 cell~/6-cm plate for ~418/gancyclovir selection.
~Selection was allowed to; proceed for 10-12 days (until control plate showed no coloniQs).
G418/FIAU-re~istant cQmpanles were isolated, amplified, and genomic DNA purified from about 50-100 individual isolates. These DNAs~were restricted with Hind III, Eco RI, and Pvu II. ~Colonies~were separated into two halves; one .

. , .
T~ TE SH~EI

2 S~C~J

half of each colony was lysed with detergent and subjected to PCR with oligonucleotide probes derived from the neo gene and p53 gene sequences outside the borders of the targeting construct. If the insertion of the construct had occurred through homologous recombination, then a 1.2 kb fragment would be generated. ~wo of 100 colonies tested gave the expected 1.2 kb PCR band. ~he DNA from these positive colonies were then subjected to agarose gel electrophoresis, Southern blotting, and hybridization to three probes. The 10probes used were a p53 exon 4 probe (130 bp), an exon 11 probe (200 bp3, and the neo probe (1 kb). Each o~ these prob~s hybridized to Southern ~ragments diagnostic for a homologous reco~binatio~ event. Taken together, the Southern hybridization conclusively identified colonies with p53 gene disruptions generated by homologous recombination.
Initial constructs utilized an ~Cl-neo gene; these constructs did not work well. ~he successful constructs were a - Pol II - neo - polyA~ - gene construct.
.

BXAXP$E 4 ~: CONSTR~CTION OF NO~S~ CEI~R~8 FRO~
: ~: SPECIFIC ~S CELL CLO~ES
.
25Chimeras were constructed frcm suitable clones as described by Bradley, A. (In: T~ratQ~inc~Ls and E=kryonic Stem ~çll$: A Pr~ical App~çh, (~ obertson, Ed.), IRL
Press, Oxfor~, lg87, pp 113-151). Briefly, 3.5-day-old blastocyst-stage embryos were collected from the dissected 30uterine tracts of C57BL ~emales 3 days after plug~ing. 12-;
: : ~ : ~.. .
SU~STIT~JTE S~E~ :

2~988~7 `

15 individual cells were microinjected into the cavity of the blastocyst-stage embryos and, after a brief culture period, transferred back into the uterine horns o~
pseudopregnant F1(CBAxC57BL) foster mothers 2 days after matin~ with a vasectomized ~ale.
The methods for introducing the ES cell into the blastocyst, and for producing offspring have been described above, and comprise techniques which are well-known to those o~ ordinary ~kill (Mansour, S.L. e~_al., Na~e~336:348-352 (198~); Capecchi, ~.R.; ~rends Genet. 5:70-76 (1989);
Capecchi, M.R., $cien~e ~ 1288-1292 (1989); Capecchi, M.R.
et ~1., In: Current Com~unic~t~ L_~Lj~ =y~a~ QqY, Capecchi, M.R. (ed.), Cold Spring Harbor Press, Cold ~pring Harbor, ~Y (1983), ppo 45-52; Frohman, M. A. et aL., Ce~
Sh:145-147 (1989); all of which re~erences have been incorporated herein by reference).
In order to generate ~n ad~quate number of chimeras of high quality (i.e., high contribution of donor cells) the blastocyst injection experiments were repeated over a period ; 20 of a few weeks. Approximately 53 blastocyst injections were : per~ormed.
Following embryo transfer, the you~g are born 17 days later. The levels of~ chimerism could not be assessed at this time because the~ cell line used carries the Black Agouti coat color ~arkers and the embryos are ~lack non-Agouti. The Agouti marker becomes visible in the coat at day 8. At this stage the chimera~ were scored and non-chimeric animals. discard~d. The presence of the Y
chro~ossme in the donor cell line ensured that the majority Or any~ germ 11ne; contrlbutlon was through ~he germ line of :: .

SUI~STITUTE SHE~ ~:
.

-WO9~/11874 PCr/US92/00295 ; ~ "'''' ~$~

male chimeras. In a~dition, dominant effects of the Y
chromosome distort the sex ratio among the chimeric offspring in favor of males by ~he conVersion of some female embryos to phenotypic males. The sex ratio consequently is an effective early mea~ure of the levels of chimerism in ~he experimental population, and ~his tend~ to reflect germ line chimarism.
The mice ranged from total black (i.e. no chimerism~ to greater than 90% agouti (i.e. very high chimerism). When the above-described stromal feeder cell procedur2s were empIoyed, greater than 50% of the injected pups showed high degrees of chi~erism (as determined by the ratio of agouti oat color to black Coat color in the offspring).

15 . ~ ~ ~XAHPL2 5 .~ - l~ST BR~DING OF CIIIII~S TO ASæAY FOR G ~ LIN~:
CONTRlB~TIONS AND I~ENTIFICaTION OF ~ETEROZYGOTES

.
Of the 53 injected blastocysts, approximately 12 male mice were cbtained which exhibited ~high to very high chi~eris~ ti.e. levels of contributions of~ agouti hair in their coat;that exceeded~50%).~Th-se mice were~subjected to ; inbreeding and~:further;~analysis~as described below.
The~ 1~ male: chimeric mice were test bred when they reached sexual maturi*y (about 8 weeks of age). Test breeding proceeded by caging a ~ingle ger~ line~chimera with two virgin ~57BL females, and permitting tha ani~als to naturally breed. Successive litters were scored after 8 days of age.: The pre ence of the~ dominan~t Agouti coat color among the litter;de~onstra~ed~h~ 3ucces~ful colonization of ITI~IT~ c~U~r W~9~/11874 PCT/US92/0~295 209882~
-8s-the germ line. This approach assumed, that, as in the case of the rb gene, a single normal germ line allele would be able to supply normal p53 function and that gene dosage effects would not be important.
Unexp~ctedly, this assumption was not con~irmed. Of the 12 chimeric mice, only four mice (#96, ~lO1, #102 and #103) were found to produce progeny mice. The great majority of these mouse pups (approximately 95-97%~ died at about 24 hours post partum. Nevertheless, after several litt~rs, lo five viable pups, designated pl-p5 were obtained ~rom parents #101 and ~102. These parental mice (#101 and #lOZ) were found to be equivalent for the purposes of the invention. The five mic~ pup~ were grown to adolescence, and are fully viable.
The 5 mice pups were tested by Southern blot hybridization of tai} DNA to determine whether they contained the mutated p53 allele. Thus, when the pups reached six to eight weeks of age, approxi~ately one inch of tail was snipped orf and used:to prepare high molecular weight DNA according to :the foIlowing protocol for the : preparation of h~gh molecular weight DNA ~rom mouse embryo and/or yolk sacs~
: In the procedure, it is d~sirable~o reserve materials and solutions ~such as ~Eppendor~ tubes, proteinase K, distLlled water ~dH20), phenol, ph~nol/chloroform, : chloroform, isopropanol,~ TE, etc.) for genomic use, and to then use such mat~rials only when preparing sa~ples o~
genomic D~A.
:

~ : ~: ....
: , ' SUBSTlTUTlE S~ E~

Wo92~ 74 PC~ S92tO0295 .' 2~

--8 6 r .
Steps of the Protocol: ~.

1. The embryos of the required gestation are care~ully isolated rom maternal tissue. The embryos andlor yolk sacs are then placed into Eppendorf tubes containing 250 ~l of Cutting Buf~er t50 mM Tris pH 7.5, 50 mN EDTA pH
8.0, 100 mM NaCl, 5 mM DTT, O.5 mM Spermidine). Th embryos can either be proc~s~ed singly or fixed for staining, using their respecti~e yolk saes for the isolation o~ DNA.

The embryonic tissue is ~inely minced with a scissors.
It may, alternatively, be broken up by passing it twice through a 26G 5~8~ Sub-~ needle attached to a lcc syringe. Yolk sac is very soft tissue that is easily broken up during lysi~; therefore, no further manipulation is~necessary. : .

3. 250 ~l of Lysis Buffer (50 mM ~ris:pH 7.S, 50 mM EDTA pH
. . .
:20 8.0, 100 mM ~aCl, 5 mM:DTT, 0.5 m~ Spermidine, 2% SDS) -; ~and 10 ~l of ~ L~ ~C~r~ 10 mg/ml Pr~teinase X is .
added to each tube. ~ .

4. The tubes are gently rocked overnight at 5S~C.
:
: ~ 5.~ ~00 ~l of phenol is added to each tubé, and the tubes are rotated at 5-10 rpm for 15-20 ~inutes. The tubes are thèn:centrifuged in a ~icrocen~ri~uge at maximum :~ spoed~ for 10 ~i:nu~es.~The DNA :is then tr~nsferred-~o ; : 30 : clean ~ppendor~tubes uslng pip-tte tips ~ro~ which the : .

`

;SUBSTITUTE~ SHEET

WO92/11874 PCT/US~2~00295 20~

tips have ~een snipped off. ~When pipetting geno~ic DNA
samples, it is preferable to use pipette tips from which the tips have been snipped o~f. This will prevent shearing the DNA and k~ep the average size in excess of - 5 80-lO0 kb.~

6. 500 ~1 of phenol/chloro~or~ is added to each tube, and step 5 is repeated.

}O 7. 500 ~1 of chloroform is added to each tube, and step 5 is repeated. When transferring thè DNA to clean tubes, the volume transferred is to be recorded.
, 8. A volume of isopropanol equal to ~he volume of DNA is lS added to each tube. The tubes are inverted a few times to precipitate the DNA.
, ~,.

9. The tubes are centxifuged in a ~icrocentrifuge at maximum speed for 10 ~inutes. ~-.
~ -lO. The supernatant~ is casefully~ aspirated off. The ~:
pelleted:DN~ is rin~ed in 1 ml :of 70~ of ethanol (EtOH).
The tubes are centrifuged again i~ the microcentrifuge at ~aximum speed for lO minutes.
~: 25 11. The EtOH is poured of~ a~d allow the pellets are allowed ~:
to air dry until th~re is no visible sign o~ remaining EtOH. The DNA i8 resu pended in 15 ~1 of TE buf~er ~lO
~mM Tris pH~.O, l ~M EDTA pM 8.0). : - :
::
: ~ 30 :: ~ : ~

~ : , -' .
SUBSrlTUTE ~SHEET ~ - ~

W092/tl874 PC~/US92~0~295 ~91~7. ~ :

12. Enough DNA can be prepared for at least one lane of a 20 cm x 15 cm agarose gel with 30 sample wells; usually, a couple of hundred micrograms of DNA is obtained. The ' DNA is digested by adding 2 ~l of the appropriate lOX
s buffer, 2 ~l of dH20, and 1 ~l of the appropriate restriction enzyme (i.e., such that the reaction is in 20 ~1 total volume).

13. Since genomic DNA is difficult to load in thl~ size of well (in which it has to displace th~ electrophoresis buffer), the gel is loaded at the bench before it is submerged in the running buffer. -Once the DNAs were obtained, their concentrations were deter~ined by A~ determination on a W spectrophotometer.
5-10 ~g of the:DNAs:were then cleaved with either BamHI or EcoRl restriction endonucleases for several hours at 37C.
The cut DNAs were loaded~ on a 0.8% agarose gel and electrophoresed at 100V for ~approxlmately 3 hours on a 2~ horizontal ~submarine gel appara~us.~ Th-:~DNAs were then blotted to Zetaprobe~(BioRad?~nylon membranes and hybridized in 20% dextran~sulfate-3XSSP~ bo~h~staps~according to~:the protocol o~ Reed~and~ann~(Reed; K.C. e~ al., ~u~leic A~ids 3:7207-7221 ~1985)~). The probe ~or hybridization was 3ZP-la~ell~d according: to~ the protocol of F-inbery and Vogelstein;(Feinberg,:~A. ç~_al,, ~nal. Biochem. 1~2:6-13 : t1988))- Hybridization was~permitted to ~ontinue overnight - at 68C and the. filter3 were washed according to the : protocol of Reed and~Mann tReed, ~.C. et al.,:~gçleic ~ids Bçe~:13:7207-7:221~(198s)))~ he probe for F~lgure 4 was:a IR~TlT~ lT~ D~T

WO 92~11874 PCI'/US92/00295 ` 20~8:82~

pol II-neo probe. The Pol II part of the probe hybridizes to a 4.0 kb fragment of the Pol II gene which is present in all mice. The Neo part of the probe hybridizes to a larger 6.7 kb fragment present only in the p53 targeked germ lin~
heterozygote mice.
Two of the five mice te~ted, pups pl and p3, were found to contain the mutated p~3 all~le in its germ line (Figure 4). Figure 4 shows a Southern blot analysis of tail DNA of the progeny o~ the F1[12~xC57BL/6] chimeric mice. The DNA
was restricted with B~m~I and hybridized to the a~ove-described 32P~labelled Pol II-Neo probe. The band between the offspring and the controls is a marker which happened to hybridize to the probe. The analysis thus rev~aled that m~use pups pl and p3 contain DNA which hybridizes to the neo gene, and thus contain the mutated p53 construct.
Following the demonstration of germ line chimericism, the initial F~(C57BLxl29) male~ were cros~ed with s~rain 129 females to generate a larger population of heterozygote mice. Germ line transmission o~ the mutated p53 allele to offspring may be confirmed by restriction mapping analysis.
: Thus, in sum~ary, 4 of the 12 mice exhibiting very high ~: chimerism (~96, ~lOlr ~102 and #103) were found to be good bre~ders, when pcrmitted to breed naturally. 0~ these ~our ~ice, mice #lOl and #102 were found to produce viable offspring having the agouti phenotype. Thus, these chimeric mice have been demonstrated to be germ line chimeric animals (i.e. they contain germ cells which pos~es~ the desired p53 construct, and are.capable to transmitting this construct to progeny ani~als). :Indee~, progeny animals of mice #lOl and ~:102 were obtained, through~routine breeding.: Two of these ,. - .
~ ~ ~ SU~STIT~)TE:S~FI~
.
.

2~`~g~` s progeny animals (pl and p3) contained the ~utated p53 alle}e in its germ line, and were thus capable of transferring it indefinit~ly .to future generations by further routin breeding.
.
EXANPL~ 6 .
PROD~CTION OF rRANSGENIC ANr~aLS ~AVING
LI-FRA~ENI D = p53 N~TANT ALLEI~S

10The above-described non-human transgenic animal contains a mutation in the p53 gene. The same methods can be used to produce animals having other mutations in this gena (Fi~ure S). Thus, a targeting construct is prepared which contains ~-.
a desired mutation (deletion, insertion or point mutation).
The construct is introduced into embryonic stem cells, and permitted to und~rgo homologous recombination with a p53 allPle of that cell. This process leads to the integration of the construct into;the p53 locus of the stem cell, and results in a duplica~ion Qr the p53 gene at that locus. The: .
20 duplication is resolved:, via intrachromosomal reco~bina~ion .
to yield an embryonic Btem ell having the desired mutated p53 allsle (Figure 5).

EXaNPLB 7 25ASSAY OF TQ~OR FR~Q~Q XN TH~ L~TEnozyGo~s ANINALS

A sizable group (n=50-lO0) of F1 mice with a single : :
disrupted p53 allele are observed over a period of 12-24 months:~for signs of;spontaneous~tumors,;abnormaI sympto~s, ..
30~: ~or :earIy~ deaths:. Careful~records of deaths and t~mor :: ~

WO92/11874 PCT/~S92/~0295 2O~ 2~

--91-- .-incidence are kept and compared to a control population of F1mice that are ho~ozygous for the normal p53 allele. ~ice are sacrificed at 3, 6, 12, and 18 months of age and autopsied for any unusual histopathology. Mice exhibiting 5 distress, disease symptoms, or evidence of tumors are sacrificed and examined to establish the exact nature of ~.
disease.
Groups of mice are ~reated with a known carcinogen, such as DMBA, fol}owing standard procedures (Bu~el, J~So et al., lo J. Virol. 38:571-580 (1~81); Xnepper, J.E. et~al~, Intl. J.
canc. ~0:414-422 (1987), both o~ which references are herein incorporated by reference) and observed ~or tumor develop-ment.
The transgenic mice are infected with Fxiend ~urine leukemia virus (F-MuLv) and assayed for time of onset and severity of erythroleukemia. F-MuLv integration into one or both alleles of the p53 gene is associated with :~.
erythroleukemia in virus-in~ected mice (~owa* et al., Nature - ~.
~14:633-636 (1985); Chow et al~, J. Virol. 61:2777-2781 (1987); Hicks, G.G. et ~, J. v~r~. 62:4752-4755 (1988)).
When differences in tumor incidence between nor~al mice and heterozygotes are observed, the data:is analyzed by : ~ :standard statistical procedures to determine the signi~icance Or the results. Significant increases in tumor : 25 incidence in the p53 heterozygote ~ice is pre~erably established using larger scale ~onfirmatory studies and carcinogen sensiti~ity studies documenting the utility of the mice as research and carcinogen screenin~ tools. ~ -: :The transgenic mice::are inter-bred,~ by conventionaI, non-recombinant methods, in order to determine whether p53 ' -SIil0STtTlJTE~ SHE~T

2~9~
;92-is a recessive lethal gene and~ if so, to determine what stage of embryogenesis is affected. If development proceeds significantly, the organs and issues ~ost affected are identified. Tumors derived from p53 heterQzygot~s are screened to dete~mine the range of mutations in the remaining "normal" p53 allele. E~bryo fibroblast cultures are derived from the altered animals to determine which parameters of transformation are affected under normal culture conditions and following trans~ection with oncogenes ~ :

10 or treatment wi~h carcinogens. -The p53 heterozygotes are interbred to obtain transgenic animals in which both of the animal's p53 alleles contàin ~:
the mutated construct.
The p53 heterozygotes are bred to other transgenic mice containing acti~ated dominant oncoqenes (e.g., the transgenic mouse of Leder, P. e~ al! (U.S. Ratent 4,736,~66)) to determine whether offspring containing both ~ .
=utaAt genes have an accelerated r~te of tumor ~ormation.

:.,~..

.

~:' ; SUE~STITUTE SHEET

:
, WO~2/11874 PCT/US92/002~5 . . , . ,;, ~ ~
CHARACT2RiZATION OF TE~ p53-DEPICIENT ~I OE

The above-described p53 heterozygous mice were o~tained and interbred to produce progeny. Among the progeny identified were p53 homozygous animals. Thus, unexpectedly, the present in~ention permitted the isolation and propagation of both p53 heterozysous mice and p53 homozygous mice.

A. p53-Homozygotes ~ave No Intact p53 RNA or Protein To conclusively show that the gene targeting of the p53 allele in ~he mouse germ line generated a null mutation, p53 RNA and protein were analyzed from homozygote ~issues and cells. The rationale was ~hat no in~act p53 RNA or protein should be present if the disrupted p53 gene was completely inactivated. To assess p53 RNA leYels in the homozygotes, RNA-PCR procedure~ were used because tissue p53 RNA levels are often so~low~that Northern analyses are~inadequate for ~: detection. Five sets of primers (~igure 6~) wer~ used to amplify cDNA synthesized ~rom total RN~ purified from : : spleen, kidney,~and tostes of~wild type~, heterozygote, and ho~ozygote mice.~ The re~ults o~:t~1ese RNA-PCR assays using total spleen RNA (Figure 6B) clearly indicated that an intact p53 mRNA was ~ynthesized only in the wild type and heterozygote mice. ~ote that ~or the primers which ampli~y a ~rag~ent spanning the exon 5 neo insertion si~e there is no apparent PCR ~ragment in the homozygot~s, indicating the ~allure of ~the homozygotes to produce an intact ~-ssage.

: ~ :
. .

TITI ITF: ~FFl~ ~ :
~ :

j " ~, ~ ", ~,, i "~ ,"" ., .. , .~. ~ , ................................ .

However, exon 7 and lO primers generate a strong fragment from the homozygotes cDNA. The data supports the conclusion that the RNA generated 3' of the neo insertion site is part of a hybrid neo p53 message initiated by the pol II
promoter. Exon l a~d 4 primers amplified a very faint PCR
band from the homozygote cDNAs relative to the wild type cDNAs. The low amounts of this ban~ indica~es that transcription from the endogenous pS3 promoter in the homozygotes appaars to result in an uns~able truncated }0 transcript as a result of the neo ca~sette insertion.
To demonstrate that a nor~al p53 protein is not e~pressed in the homo~ygotes, an immunoprecipitation assay was performed on proteins derived ~rom tertiary embryo fibroblasts obtained from normal, heterozygote, and homozygote embryos. Using a polyclona} pS3 antiserum, the wild type and heterozygote cell~ w~re found to contain a precipitable p53, while the ho~ozygote cells did not.
In order to further investi~a~e whether any intact or truncated p53 molecule was expre ~ed in the homozygotes, an immunob}ot analysis was performed using a control antibody (Ela-specific antibody ~73 t~arlow, E. et_ al., J~ Virol.
~5:533-546 (l985)~ and three different p53~specific ~onoclonal antibodie-~ ~s~I~cted from a set of six such antibodies: PAb 2~2 (Yewdell, J.W. et_al~ Y1~Q1~ 59:444~
:~ 25452 (1986)); PAb 246 (Yewdell, J.W. et ~l., J ~ V~r~l.
444-452 (19~6)); PAb 248 ~Yewdell, J.W. et al., ~. Virol.
59:444-452 (19~6)); PAb 421 (Harlow, E. et al., ~. Virol.
39:861-869 (1981)~; R~3.2C2 (Coffman, R.L. et al-, Z__E~
~ 269-279 (1981)); and 200.47 (Dippold e~ Q~.
;~ ~ 30 Natl~ Aca~ Sci. ~U.... ~.L ~ 95-1699 (1981)). The ,:

` ~ SUBSTiTU~E SI~E~

~"
2098~
. ,' ``-:~ ., --ss- : .

epitopes identified by these antibodies have been mapped (Gannon, J.V. et ~1-, EMB0 J. 9:1595-1602 (1990); Wade-Evans, A. et al! ~ EM~O J. 4:699-706 (1985); Mole, S.E. et al., Nucl. Acids Res. 17:3319 (1989); Yewdell, J.W. et al., S ~ L~ 59:444-452 (1986)). Cell lysates were initially immunoprecipated with a collection of either (1~ PAb 242, PAb 246, and PAb 248, mapping to the p53 epitope located at amino acids 1~-25, 88-lOg, and 14-S9, respectively, or (2) :
PAb 421, RA3.2C2, and 200.47, mapping to the p53 epitope located at amino acids 370-378, 41-69, and 41-111, ;
respectively. The i~munoprecipitates were then subjected to SDS polyacrylamide gel ~lectropbore~is, followed by tran~fer ..
o~ the proteins ~o nylon. These blots wer~ then probed with all six of the above-mentioned anti-p53 monoclonal anti~odies. The results clearly indicated that homozygotes did not contain any intact or truncated p53 detectable by this assay.
In concIusion, the RNA and protein data indicate that :
neither intact or truncated p~3 is produced in the homo2ygotes and that the targeted p53 mutation indeed represents a null mutation.

B. p53-~omozygote Ni~ ~re Susceptible to Tu~ors Although: the p53-negative homozygous mice appeared developmentally normal, they were clearly susceptible to spontaneous tumor formation (naturally occurring tumors are rare in mice less than 6 months o~ age (Madison,~R.M et al., J. N~ st. 40:683-685 (lg68); Maita, K. et al., Toxicol. PathQl~ 6:340-349 ~19~8); Squire, R.~. et al., In:
.
,:

R.~TuTE s~E~ : ~

WO92~11874 PCT/U~92/0029S

2 0 9 8 ~ 2 t Patholoqy of ~aboratorY Ani~als, Volume l I, (Eds., Benirshke, R. et al.), 1054-1055 (Springer-Verlag, NY, 1978)).
The chimeric mice and the germ line p53 heterozygotes and homozygotes were ~onitored daily ~or tumors or other abn~rmalities. The incidence of tu~ors was noted when the chimeric mice were 14 months old, the oldest hetarozygotes were o~er nine months old and the oldest homozygotes were about six months old (Table 1, age is in weeks (wks) or ~onths (mons)). ~t that time, no tumors were observed in the normal control mice,.a variety o~ tu~ors were detected in the homozygote mice (Table 1; ca~es # 1-20), two tumors were detected in the hetero~ygote mice (Table 1, cases # 21 a~d # 22), and three ~umors were d~tected in the chimeric mice (Table 1, cases # 23 and # 24; in case # 23, the phenotypically male mouse had a choriocarcinoma surrounded by recognizable ovarian tissue).

.

.

C~ TlTlJ~E S~

W ~ g2/11874 PCT/US92/~0295 2098~J7 , . .; .. .

-97~ :
, ..
TABLR 1: T~HORS OBSERN~D I~ p53--D~FICI~T~I OE
~ Sex A~e Hi~tologic type Anatomic nite I 1 M 18 wk8 undifferentl~ted ~arcoma ~ubcutls ~ . . . . , . . , _ ._ ¦ 2 ~ 8 wks gonadobla~toma teste~ ..
l mallonant lYmPho~a thv~us ~ __ ~ . . __ .
¦ 3 M 18 wk~ heman~lo~rcoma ~2) 0ubcuti~/llver _ _ _ . ,..... ., ~ ... __ _. __ 1 4 F 9 wk~ mallanant ly~phoma aeneralized I _ _ . _ . , . _.____ _ _ _ F 15 wks al~qnant lym~h_ a _ generaltzed _ ¦ 6 M 24 wk~ undlfferent~ted ~arco~A vertebra I heman~io~arcoma ~ubcuti~
t--__ , ~
¦ 7 M 24 wk~ hemangioQarcoma ~ubcuti~/mu~le malionant lYmphoma ~eneralized _ ~ , ," : . .
¦ ~ M 15 wk~ hemanqioaarcoma ~ubcutia l - . . . ~ , _ , . ~
l 9 M 24 wk~ ha~anglo~rcoffa aubcutl~
I ~ali~nant lymphoma ~enerallzed .
.. ...... ~ , ..... . ..
M _ 16 wk0 o~teoaascom~ _ _ pelvi~ _

11 F 15 wk~ malignant lympho~a generalized hemanqioma heart _ ... ..... __ l 12 F 19 wk~ mal~onant lYmPhoma thvmu~
. . . _ , __ _ ~_ __ ¦ 13 M 14 wk~ maliqnant lymphoma qeneralized . - . . .. - - - - , . -14 F14 wks alignant lymphom~ generalised 1 15 F15 wks mAm~ rY adcnocarcincma ~mammar~ qland -.
__ _ ~ _. .
16 M16 wk~ malignan~ lympho~a generalized .. -.
: : hem~ngIo~rcoma ~boutia : _ ~hem~ngio~a _ _ muscle .: .

17 _M17 ~X- mallqnan~ lympho~a eneralis~d ¦ 18 M17 wks m2li~nant lympha~n generalized l , ~ hemangio~arcom3 ~2) heart/perlr~nal :
. .. . .
19 M21 wk~ mallgnant ly~phoma thymu~ _ : :
M _ 22 wks ma *qnant lymphoma general~zed ~ .
.
21 M24 Wk8 embryonaI car~tnom~ te~ti~ -~
_ ... . ~ - . :::
22 M37 w~-- ~ ~Ii9D~Dt lg~pho~ ; ~ _ -':
. .
:..
~;UBSTITI JTE S~IE~ ~:

W~92/1~X74 PCT/US92/00295 2098~27 TAB~ T~SORS OB5~ I~ p53--DlSFICIl~T ~C~
. , . _ . . __ - 1 . ~ Sex A~ Histologic_typa _ 23 M 14 monu oflteoo~coma ~xum _ chort ocarclno~a ova~Y~
_, .
24 M 14 mon~ Leyd~g cell tumc~r te~
_ _~ ~

~ore than 70 ~ild type mice, lOO heterozygotes, and 60 homozygotes have been analyzed. The data collected on the mouse tumors indicate that homozygote mice are susceptible to tumors at an early age. By six months, greater than 50%
o~ homozygotes develop tumors. Thus, the tumor incidence in such mice is considerably higher than the 20% incidence observed by 18 months of age in mice carrying a mutant p53 transgene (Lavigueur, A. et al.1 M~lçç ~CCLl~lRiQl~ 9:3982-3991 (1989)). Four of the homozygote mice were ~ound to have had multiple primary tumors of different cell type . .
origin. The tumors in the older chimeric and heterozygote mice suggest that mice with a single mutant p53 allele are also susceptible to tumors, but at a decreased rate compared to homozygotes. No wiId type control mice have developed tumors curre~tly under ob~ervation. These results thus 20. support the conclusion that 10~8 of one or both p53 alleles predispose the mice to cancer. The variety o~ tumors identified in the homozygote mice demonstrate that p53 is involved in tumorigenesis in many tissues and cell t~pes. :
Despite haYing a developmen~ally normal appearance, a signi~icant nu~ber o~ the homozygous ani~als succumbed to pr~mature death in the absence of obvious tumors (tissue ~:
au*olysis preven~ed a complete pathological analy~is in some cases). Several o~ the deaths appeared associated with ' ' .
~. ....
SlJBSTITUTE S~EEl~

~ ; .
2~9~7 _99_ :

unresolved infections, suggesting the presence of a subtle defect of the immune system in the homozygotes. Such immune defects, thus. appear able to contxibute to the accelerated rate of tumor development in these animals.
Whi}e the invention has been described in connection with specific em~odiments thereof, it will be understood that it is capable of ~urther modifications and this application is intended to oover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or : -:
customary practice within the art to which the invention pertains and as may be applied to the essential features .. .
hereinbefore set ~orth and as follows in the scope of the :
appended claims.

'~'. '., : ' , ' ':.

-- , - .: .

~; ~ , ..

fiUE~STlTUTE Sl~EET~

Claims (39)

WHAT IS CLAIMED IS:

1. A transgenic or chimeric animal cell whose genome comprises two chromosomal alleles of a tumor-suppressing gene, wherein at least one of said two alleles contains a mutation.

2. The animal cell of claim 1, wherein one of said alleles expresses a normal tumor-suppressing gene product.

3. The animal cell of any of claims 1 or 2, wherein said tumor suppressing gene is a p53 gene.

4. The animal cell of claim 1 which is a human cell.

5. The animal cell of claim 1 which is a non-human animal cell.

6. The animal cell of claim 5 which is an embryonic stem cell.

7. The embryonic stem cell of claim 6 which is ATCC
CRL 10631, and wherein said tumor-suppressing gene is a p53 gene.

8. A non-human transgenic or chimeric animal having an animal cell whose genome comprises two chromosomal alleles of a tumor-suppressing gene, wherein at least one of said two alleles contains a mutation, or a progeny of said animal, or an ancestor of said animal, at an embryonic stage.

9. The non-human animal of claim 8, wherein said tumor suppressing gene is a p53 gene.

10. The non-human animal of claim 8, wherein said animal cell is a human cell.

11. The non-human animal of claim 8, wherein said animal call is a non-human cell.

12. The non-human animal of claim 11, wherein said animal cell is a germ-line cell.

13. The non-human animal of claim 11, wherein said animal cell is a somatic cell.

14. The non-human animal of claim 11, wherein said animal and said animal cell are of the same species.

15. The non-human animal of claim 11, wherein said animal and said animal cell are of different species.

16. A non-human animal containing the embryonic stem cell of claim 6, or a progeny of said animal, or an ancestor of said animal, at an embryonic stage.

17. A non-human animal containing the embryonic stem cell of claim 7, or a progeny of said animal, or an ancestor of said animal, at an embryonic stage.

18. A method for identifying the presence of an agent suspected of being capable of affecting a characteristic of an animal cell that is attributable to the presence or expression of a tumor-suppressing gene, said method comprising:
A) administering an amount of said agent to an animal cell in cell culture, said cell having a genome that comprises two chromosomal alleles of said tumor-suppressing gene, wherein at least one of said two alleles contains a mutation;
B) maintaining said cell culture for a desired period of time after said administration;
C) determining whether the administration of said agent has affected a characteristic of said animal cell that is attributable to the presence or expression of said alleles of said tumor-suppressing gene.

19. The method of claim 18 wherein said tumor-suppressing gene is a p53 gene.

20. The method of claim 18 wherein said agent is suspected of being able to increase a tumorigenic potential of said animal cell.

21. The method of claim 18 wherein said agent is suspected of being able to decrease a tumorigenic potential of said animal cell.

22. The method of claim 18 wherein said animal cell is a human cell.

23. The method of claim 18 wherein said animal cell is a non-human animal cell.

24. The method of claim 23 wherein said non-human animal cell is an embryonic stem cell.

25. The method of claim 24 wherein said embryonic stem cell is ATCC CRL 10631.

26. A method for identifying the presence of an agent suspected of being capable of affecting a characteristic of an animal cell that is attributable to the presence or expression of a tumor-suppressing gene, said method comprising:
A) administering an amount of said agent to an animal, said animal having a cell whose genome comprises two chromosomal alleles of said tumor-suppressing gene, wherein at least one of said two alleles contains a mutation;
B) maintaining said animal for a desired period of time after said administration;
C) determining whether the administration of said agent has affected a characteristic of said cell that is attributable to the presence or expression of said alleles of said tumor suppressing gene.

27. The method of claim 26 wherein said tumor-suppressing gene is a p53 gene.

28. The method of claim 26 wherein said agent is suspected of being able to increase a tumorigenic potential of said animal cell.

29. The method of claim 26 wherein said agent is suspected of being able to decrease a tumorigenic potential of said animal cell.

30. The method of claim 26 wherein said animal cell is a human cell.

31. The method of claim 26 wherein said animal cell is a non-human animal cell.

32. The method of claim 31 wherein said non-human animal cell is an embryonic stem cell.

33. The method of claim 32 wherein said embryonic stem cell is ATCC CRL 10631.

34. The method of claim 26, wherein said animal and said animal cell are of the same species

35. The method of claim 26, wherein said animal and said animal cell are of different species.

36. A method of gene therapy comprising altering the genome of a cell of an animal, wherein said cell has a genome that comprises two chromosomal alleles of a tumor-suppressing gene, wherein at least one of said two alleles contains a mutation, to thereby form a cell wherein at least one of said alleles expresses a normal tumor-suppressing gene product.

37. The method of claim 36, wherein said animal is a non-human animal.

38. The method of claim 36, wherein said animal is a human.

39. The method of claim 36, wherein said tumor-suppressing gene is a p53 gene.