CA2174860A1 - Myocardial grafts and cellular compositions useful for same - Google Patents

Abstract

Described are preferred myocardial grafts of skeletal myoblasts or cardiomyocytes, and cellular compositions amd methods useful in obtaining the grafts. The myocardial grafts are stable and can be used, for example, to deliver recombinant proteins directly to the heart.

Description

-Wo 95/14079 PCr/US94113141 ~1 748~n MYOC7~RDIAL GRAFT~ AND CELLUL~R
COMPOSlTIONS USEFUL FOR SAME
BACKGROUND OF TIIE INVENTION
T~le present inverltion resides generally in t~le field 5 of cardiology, and more particularly relates to stable myocardial grafts and methods and cellular compositio1ls useful for achieving such grafts.
As further background, organ transplantation has been widely used to replace diseased, nol1fu1lctional tissue.
10 More recently, cellular transplantation to augment deficiencies in host tissue function has emerged as a pote11tial tllerapeuLic paradigm. One eYample of tllis approach is the well publicized use of fetal tissue in il1dividuals with Parkinsonism ~reviewed in (1), see 15 reference list, i,~l~;a,), wllere dopamine secretion ~roln transplanted cells alleviates the deficiency in patients.
I1l other studies, transplanted myoblasts from uneffected siblings fused witll endogenous myotubes in Duchem1e's patients; importantly the grafted myotubes e2~pressed zo wild type dyotrophin ~

WO 95/1407g PCr/US94/13141

2~7 4~ 2-Despite their releva1lce ill other areas, ~hese earlier stu~ies do rlot describe any cellular ~ransplanta~ioll tec~1nology which can be succesfully aplied to the heart, w11ere the ability to replace damaged myocardium would have 5 obvious clinical relevance. Additionally, t~le use of intra-cardiac grafts to target the long-term e2~pression of angiogenic factors and ionotropic peptides would be of t11erapeutic value for individuals with myocardial ischemia or congestive heart failure, respectively.
In light of t~lis background there is a need for the development of cellular transplantation technology in the 11ea~t. Desirably, such technology would not only provide stable grafts in the heart but also enable the delivery of useful reco1nbina11t proteins or other molecules directly to L5 the ileart. The present invention addresses these needs.
-wo 95/14079 PcTn~594~1~141 _3_ 2~ ~4~G~
SUMMARY OF THE INVENTION
The applicant has established cellular grafts in themyocardium w~1ich are viable long-term Cardio!11yocytes and skeletal myoblasts llave been grafted directly into the 5 myocardium of syngerleic animals. Viable grafts were detected at least one-half year post-implantation (L1le latest time point assayed). T1le presence of t~le grafts was not accompanied by overt cardiac arrhythmia, and the majority of the grafts were juxtaposed directly to the 10 host myocardium and not encapsulated. It has thus been discovered that tlle myocardium can serve as a stable i~latform for cellular transplants. Tilese transplants can be used for the local delivery of recombinant molecules to the 11eart and~or for replacing diseased tissue to 15 supplement myocardial function.
Accordingly, one preferred embodiment of the inven~ion provides a myocardial graft in an animal WiliC
includes a stable graf t of skeletal myoblasts or cardiomyocytes incorporated in myocardial tissue of the 20 animal.
Another preferred embodiment of the invention provides a method for forming a stable myocardial graft in a~1 animal. The inventive method includes the step of introducing skeletal myoblasts or cardiomyocytes in 25 myocardial tissue of the a11imal so as to form a stable myocardial graft. The cells can be conveniently introduced, for example, by injection.
Anotller preferred el11bodiment of the inven~iorl - provides a method for delivering a recombinant molecule to 30 myocardial tissue of an animal. This metllod includes the step of establislling a stable graft of skeletal myoblasts or cardiomyocytes incorporated in myocardial tissue of tile 2~74~6~ _4_ --animal, whereirl tlle myoblasts or cardiomyocytes deliver the recombinant molecule to the myocardial tisfiue. In this embodiment t11e myoblasts or cardiomyocytes will carry transgenes encodir1g tlle recombi~lant molecule.
A110t11er preferred e111bodimer1t of the invention provides a cellular composition comyrisiny a substantially 11omogeneous population of non-immortalized cardiomyocytes. This and otller cell populations can be obtained utilizing a preferred inventive method that includes ti) tra1lsfecting embryonic stem cells to introduce a marker gene enabling selection of one cell lineage from other cell lineages resulting frorn differentiation of the stem cells, (iii) causing tlle stem cells to differentiate, and (iv) selecting said one cell lineage based on ~he marker gene. The cells used in and resulLing fronl such methods also for1n a part of tlle present il~ver1tion.
still anotl~er preferred embodiment of tlle invention provides a non-human animal having a stable graft of skeletal myoblasts or cardiomyocytes incorporated in nyocardial tissue of the animal.
T1le invention tl~us provides myocardial yrafts, metllods and cellular compositions useful for formir1~
~nyocardial grafts, and animals which 11ave the myocardial grafts. The grafts will find use both as a vellicle for delivering tllerapeu~ic substances sucll as reconlbinant proteins and other molecules, and as a means for replaci11g diseased ~issue o supplement myocardial function.
Cellular compositions of t11e invention can be used directly to prepare grafts, and will also be useful in screening drug substance effects on cardio1nyocytes and ~or e~pressiny and obtai11ing recombillant ~roteiris. Grafted ar1i1~1als can be used, for e~ample, to screen the effects of .

Wo 95114079 Pcr/l~ss4n3l4l 2l 74~60 recombinallt molecules on tlle lleart.
,, These arld other objects and a~vantages of the invention will be be apparent from the following description .

WO 95/14079 PCr/US94/13141 7 48~ --13RIEF DESCRIPTION OF TIIE rIGURE
Figure l is a sc~lematic diagram illustrating D~A used to generate MHC-nLAC trans~elliC mice ill Example 3, ill~

-WO95/14079 2 ~ 74 8 6~
DESC~IPTION OF TIIE ~REFERRED EMBODIMENT
For ~11e purpose of pro1noLil1g al1 understanding of ~1le principles of t~1e invention, referel1ce will now be made to certain embodiments t11ereof and specific language will be 5 used to describe the same. It will nevertheless be understood t~lat no limitation of the scope of the invention is thereby intended, such alterations, further modif ications and applications of the principles of tlle invention as illustrated herein being contemplated as 10 would normally occur to one skilled ir1 the art to wi1ich tlle invention relates.
As ill~icated above, tlle presellt invention provides stable myocardial grafts of skeletal myoblasts and/or cardiomyocytes. In this regard, as used l1erein the ter1n 15 "stable myocardial graft" is intended to mean a myocardial graf t whose cells are viable for a period of at least about 2 weeks. Surprisingly, such stable grafts i1ave been readily achieved in accordance with t~le invention, wiLh preferred grafts 11aving cells viable for six montlls or 20 more. Myocardial grafts of tlle invention can thus provide for long-term delivery of recombinant proteins or other molecules ~o the ~leart and/or long-term sui~plementation u~
nyocardial tissue.
Tlle skeletal myoblasts and cardiomyocytes used in ~11e 25 invention can be obtained or isolated from any suitable source. Skeletal myoblasts, including for e~ample C2C]2 skeletal myoblasts, are available fronl public depositories SI~CIl as t11e American Type Culture Collection (ATCC) (Rockville, MarylaIld). Skeletal myoblasts can also be 30 isolated from skeletal muscle using techniques well known to tlle art and literature. Cardio1nyocytes useful ~or the invention can be obtained using techrliques described in t~le literature (3) or using methods described more Wo 95/14079 PcrluS94113141 ~14~6~ ~
--~3--particularly in tlle Examples below. ~riei~ly, olle sucl met1~od involves digestion of 11eart tissl1es to obtair cardiomyoctes .
Al10ther method involves using an appropriate marker 5 to select specific cell lineages, such as cardiomyocytes, rom other cell lineages resulting froln the differel1tiation of embryonic stem cells (totipotent cell lines derived from the inner cell mass of blastocysts as described in (22) ) . The preferred method involves a 10 positive selection sc~leme. Thus, a marker gene, such as a gene conferring antibiotic resistance (e.g. I1eomycin or lygrolllyCill), i5 irltroduce~ into the stem cells ul1der appropriate col1trol such that expression of the gene occurs only in the desired cell lineage. For example, L1le 15 marker gene can be under the control of a promoter whic is active only in t~e desired cell lineage. Upon diferentiation of tlle stem cells, the desired lineage is ~hen selected based upon t}le marker, e.g. by contacting the Ini2~ed differentiated cells with the appropriate 20 antibiotic to which t~le desired lineage has been conferred resistance. Cell lines other than the desired line will t~lus be killed, and a substantially pure, homogeneous population of t11e desired line can be recovered. In more preferred met~lods, two marl~ers are introduced into the 25 parellt stem cells, one allowing selection of transfected stem cells froln non-transfected cells, and one allowing selection of the desired cell lineage from other lineages. A double positive selection scheme can tllus be used w~lere each selectable marker confers antibiotic 30 resistance. Using this selection metllodology, populations comprised about 90% and even about 95-100% of ~he desired cell lineage can be obtained, as demonstrated i11 tlle Exal11ples ~elow.
To obtain grafts of the invention, t~1e skeletal 35 111yoblasts or cardiomyocytes will be i~1troduced illtO the .. . . .

WO95/14079 21 74~6~
myocardial tissue of a living animal such as a mammal.
Tlle cells call be introduced in any suitable manner, but it is preferred that t~le mode of irltroduction be as non-invasive as possible. Thus, delivery of the cells by 5 injection, catheterization or similar means will be more desired.
The resulting gra~t-bearing allimals have exllibited normal sinus rhythms, indicating that the graft per se, as well as the graft-llost myocardium border zone, does not 10 induce arrhythmias. This is in stark contrast to the remodeling ~llat frequentLy occurs following infarcts in humans; the border zone of the infarct may give rise to circus loops which result in clinically signif icant arrllyLlllTlias (4, 5).
Grafts of Lhe invention can be proliferative or non-proliferative~ For example, the AT-l grafts established in the specific Examples below are proliferative. On the other hand, the skeletal myoblast-derived grafts formed in the Examples are 20 non-proliferative, with the absence of tritiated thynlidine uptake demonstrating that the formation of stable intra-cardiac grafts was not dependent upon sustained cell proliferation .
Preferred grafts will be characterized by Lhe 25 presellce of direct illtracellular coupling and tlle formation of gap junctions betweel~ Ilost and grafted cells. Moreover, such grafts will not cause immune response in tlle l~ost, and will exhibit terminal differentiation of grafted cells arld a non-tumorogenic 30 nature.
Grafts of tlle inventiol~ are useful i;~L ~liil to deliver T herapeutic proteins and the like via secretion WO95/14079 7 48~ PCTNS94/13141 Erom grafted cells, arld to replace diseased or damaged tissue to supplement myocardial f~1nction. As examples oE
therapeutic protein deliverie.s, grafts may express an(Jiogenic factors (as exemplified by basic and acidic 5 Fibroblast Growth Factor; Transforming Growtil Factor-~eta, Vascular Endothelial (;rowth Factor and Hepatocyte Growtl Factor) to induce neovascularization. Similarly, graFts expressing neurotrop~lic agents near an infarcted region may be used to ameliorate the arrilytllmogenesis associated 10 wiLh t11e border zone. Tl1ese and many other candidate substances for targeted delivery to the 11eart will be apparent to ti~ose skilled in the area.
To promote a furt}ler understandirlg of ~11e inverltion and its principles and advantag~s, the following specific 15 Examples are provided. It will be understood that these Exa~nples are illustrative, and not limiting, in nature.
RYAMPL~E 1 Generation of Stable AT-l Cardiomyocyte Grafts A. MET11ODS
20 AT-l Cell Culture an~l rlvocardial Graftin~l rrotocol. AT-l cardiomyocytes were isolated from subcutarleous tumors by seql1ential collagenase digestion and cultured in PC-l medium (Ventrex, Coon Rapids MN) containing 10% ~etal calf serum as previously described in (6). Cells were labeled 25 wi th l0 ,uM 8-chlorometllyl-9, 4-dif luoro-l, 3, 5, 7, -teLramethyl-4-bora-3a,4a-diazai11decene (BODIPY, Molecular Probes, Eugene OR) for 30 min at 37C to facilitate localization of the injection site. Immediately beEore illjecl:ioll, cells were harvested witi1 trypsin arld 30 collagena8e, washed t}lree times wit~1 serum-free~PC-l mediuln and direc~ly injec-ed il1to tlle ve11tricular 1nyocardium of sy11ge11eic B6D2/Fl mice (Jackson ... . . _ . . ... . .. . . . _ _ . . . _ _ . .

Wo 9~/14079 2 1 7 4 ~ ~ O PCT~JS94~13141 Laboratories, Bar llarbor MA~ under opell lleart surgery as described in (7). Cells (4-10 x 104) were injected in a volume of 2-3 ~11 using a plastic syringe fitted Witll a 30 gauye needle.
5 llistoloav. Hearts were removed following cervical dislocation and cryoprotected in 30% sucrose, embedded and sectioned at 10 llm with a cryomicrotome as described (8). For llematoxylin and eosin (H and E) staining, sections were post fixed in acetone:methanol (1:1) and 10 stained according to manllfacturer's specifications (Sigma Diagnostics, St. Louis MO). For immuno-llistology, unfixed sections were reacted witll polyclonal rabbit anti-T-Ag alltibodies (eitller 161-T, see (3) or 162-T) followed by llorseradish peroxidase-conjugated goat anti-rabbit 15 antisera (Boe~lrinyer Mannheim, Indianapolis IN), and visualized by dialllinobenzidine reaction with nickel enhancement as described in (9). Monoclonal antibodies against t~le comlnon leukocyte antigen (CD45; antibody Ml/9.3HL, Boellringer M~nr~ im) and against the macrophage 20 Mac-l antigen (CDllb; antibody Ml/70HL, Boehringer Mann~leim) were used to monitor intra-cardiac graf t rejection. Tlle Mac-l antibody has 75-90% cross reactivity with lymphocytes . Af ter ~reatment with primary antibody, sections were incubated with llorseradisll 25 peroxidase-conjugated rabbit anti-rat antisera (Boe~lrillger Man~llleim), and visualized by diamillobenzidine reaction with nickel ~n~ L. For [3H]-thymidine incorporation, mice were given a single bolus injection of isotope (400 ~LCi at 28 Ci/mM, Amersham, Arlington l~eiyhts 30 IL) and eighteen hours later sacrificed by cervical dislocatioll. Tlle ~leart was removed, cryoprotected in 30%
sucrose, embedded and sectioned with a cryomicrotome.
Sections were post-fixed in methanol:acetolle (1:1), stained wi~h ll and E, and a thin layer of pllotographic 35 emulsion (Ilford L.4, rolysciences~

Wo 95/14079 PCI/US94/13141 ~ 7 4& ~ --Warrillgton PA) diluted 1:1 Wit~l distilled water was applied. Sections were exposed for 5-7 days at 4C, and ~eveloped in Kodak D-l9 at 20OC for 4 minuLes, washed witl distilled water for 1 minute, fixed in 30% sodium 5 thiosulfate for 10 minutes, and washed in distilled water.
Electron l~'licroscor~y (~M), Tisslle Llocks were fixed in 2%
glutaraldehyde in O.lM cacodylate buffer (p~l ~.4) and post-fixed in 2% osmium tetroxide (Stevens Metallurgical Corp., New York NY). All other EM cllemicals were obtained 10 from Ladd Research Industries, Inc. (Burlington VT).
Tissue was stained en bloc witll 2% uranyl acetate in pH
5.2 maleate buEfer (0.05 M), dehydrated, and elllbedded in Ladd LX-112 . Graf ts were located using l~lm sections stailled wiLll toluidine blue. After ~rimming, tlle block 15 was tllin sectiorled, and staine~ wit:h uranyl acetate and lead citrate. Specimens were viewed on a Pllillips 400 transmission electron microscope.
Electrocardio~ra (ECG~ Alvses. For surface ECG
records, mice were anestlletized (2.5% Avertin,0.015 mlJg 20 body weigllt, IP, Fluka Chemicals, I.ake Ronk~ NY), surf ace electrodes were placed in the standard lead 1 pOsitiotl, and ECGs were recorded witll a Narco Biosystems (~loustoll TX) lligh gain amplifier coupled to an A/D
corlverter tCoulbourll Ins~ruments, Le~ligh Valley PA).
2~ Pla~m~ En7Yme A~sav (Pl'~. For lacta~e dehydrogenase (LDH) isoform assay, plaslrla was isolated by reLro-orbital sinus bleeds ~Inder anesthesia (2.5% Avertin,0.015 ml/g body weig~lt, intraperitoneally (IP)). Plasma was fractionated Oll 1% agarose gels (CK Isoenzyme 30 eLectrophoresis system, ClBA-Corning l)iagnostics, Cornirlg N.Y. ) and tlle LDH isoforrns visualized by a TNBT-Forlllazan l~istoc~lemical assay (LDI~ Assay Xit, Sigma Diagnostics, St.
Louis MO).

Wo 95114079 2 1 7 4 8 6 0 PCT/US94/13141 B. RESULTS
In these studies, AT-l cardiomyocytes (derived from transgerlic animals t:llat expressed ~he T-Ag oncoprotein i t11e heart) were injected directly into the myocardium of 5 3yngeneic mice and t11e viability of tl1e graf~ed material was assessed. To facilitate localization of ~he injection site in preliminary experilllents, AT-l cardiocytes were incubated briefly with BODIPY prior to grafting. BODIPY
is a nontoxic glutat11ione reactive dye which E~ermits 10 fluoresce~1t tracking of liviny cells. The graft site was easily visualized by fluorescence microscopy usi11g a FITC
cube. Subsequent experime11ts did not utilize BODIPY.
Fifty percent (l9/2~) of t11e animals receiving AT-l cardiomyocyte injectiorls developed intra-cardiac grafts.
15 In most instances, the grafts were neit~ler encapsulated nor surrounded by scarred myocardium. At t~le level of light microscopy, grafted AT-l cardiomyocytes were observed directly ju~taposed with host cardiomyocytes.
The identity of the AT-l cardiomyocytes was confirmed by 20 imml~llo-PerOXidase assay using an anti-T-Ag antibo~iy primary antibody (162-T) followed by a horseradis~1 peroxidase conjugated secondary antibody. Specificity of t~1e anti-T-Ag antibody has been established previously (l0, ll). Black precipitate to was observed over 25 cardiomyocyte nuclei in the graft but not in the host myocardium, confirming that the graft was comprised of AT-l cardiomyocytes. Similar results were obtained witll other anti-T-Ag antibodies, a11d no signal was observed in l:~le absence of primary antibody.
Viable AT-l cardiomyocytes were observed at least as long as four months post-implantatio11. During this period, some degree of graft proliferation occurred;

Wo 95/14079 PCr/US94/13141 ~ 486

3 [~1] -~11ymidine irlcorporatio11 analyses detected DNA
synt~1esis in t11e grafted cells. Ten percel1t of tlle AT-l cardiomyocyte nuclei were synthesizing DNA as evidenced by isotope incorporation into ~11e nucleus. However, ~1le rate 5 w~s appreciably less than that obserYed for cultured AT-l cardiomyocyte.s, where 50% of the cells synthesized D~A
following a similar 3[~1]-t11ymidine pulse. In several instances, the grafted AT-l cardiomyocytes were localized within the subpericardial space.
Immunohistologic experiments were employed to determine if the intra- cardiac grafts were subject to cllronic rejection. Grafts olcler than one monLII failed ~o react wit11 antibodies specific for mouse leukocytes;
sigrlals observed in blood vessels located on the same 15 sectio11 provided a positive control for the experiment.
Similarly, arl antibody w11ich detects mouse macroplla~1es and lymphocytes did not react with the intra-cardiac graft;
once again positive signal was observed in a blood vessel located on the same section. Collectively, these results 20 indicate the absence of chronic graft rejection by the syngeneic hosts. This result is supported by the observation that cyclosporine treatment (50 mg/kg body weigllt, administered irltraperitoneally daily) did not influence significal~tly the frequency of intra-cardiac 2s grafting (50% success rate, n~6). Sex of t~1e 11ost animal also did not appear to influence slg11ificar1tly ~1le rate o~
graft formation (46% success rate in males, n-13; 53%
success rate in emales, n.l5). The frequency of grafting was similar in animals examined at early time points (1-40 30 days post-grafting, 47%, n.l5) as compared to t11ose r~Yr~minerl at later time points (40-120 days post-grafting, 54%, n-13). Finally, similar frequencies of intra-cardiac grafting were observed when cells were delivered to either the left ventricular free wall or ~11e apex of tlle ~leart.

WO 95/14079 2. 1 7~ ~ 6 ~ PCT/lrS94113141 .

Electron microscopic analysis of tlle AT-l cardiomyocyte grafts con~irmed the absence of encapsulation. I~igll power views revealed well-develop~d junctional complexeE between adjacent cells wit~in tlle 5 graft. Graft cardiomyocytes contained numerous polyribosomes and the dedifferentiated myofibrillar ultrastructure typical o AT-l tumors ill vivo (6).
Electron-dense secretory granules were also observed in tlle AT-l cardiomyocyte grafts, as would be expected for 10 myocytes of atrial origin. Host cardiomyocytes borderirlg tl~e grafts ~lad normal ultrastructure with well-formed sarcomeres. Althollgh only a thin basement membrane separated AT-l and host cardiomyocytes, no junctional connectiolls between these two cell types were observed.
Surface electrocardiograms were performed to determine if tlle presence of AT-l cardiomyocyte graf ts inf luenced t}le autonomic rllytllm. No appreciable differences were observed between records from sllam animals and those which harbored grafts. In each case, 20 tlle experimental animals exhibited normal sinus rhytilm, with an anesthetized heart rate of approximately 400 beats per minute. Normal P-QRS coupling was maintained, indicating tllat the grafted AT-l cardiomyocytes did not act as an ectopic pacenlaker. Tilis latter result is 25 illlportant in light of the observation that AT-l cardiomyocytes exhibit spontaneous electrical activity bot:~l in vivo (12) and i~l culture (3). Tlle absence of overt arrllytllmia a~so indicated tllat graft-illduced myocardial remodeling was not associated with the 30 generation of sigllificallt circus rhythms.
In addition to surface ECG, plaslna LDII levels were assessed in mice carrying AT-l cardiomyocyte grafts. Tlle presence of tlle cardiac LDH isoform il tl~e circulation is a well established l~allmark of myocardial infarction. No Wo 95/14079 PCT/US94/13141 ~1 7 ~60 cardiac LD1~ (isoform-l) was apparent in mouse plasma prior to grafting. After t11e introduction of AT-l cardiomyocytes, tllere was a transient appearance of t~1e cardiac isoform in the plasma, which rLlOSt likely reflected 5 damage to the host myocardium as w~ll as damaged AT-l cardiomyocytes. A transient increase in plasma skeletal LDH isoform was also observed following grafting surgery, presu1nably reflecting damage caused by the trans-Lhoracic incision. The plasma LD~I profiles returned to normal by 7 10 days post-implantation. Thereafter, the plasma LDH
profiles remained normal despite the presence of grafts.
EXAMPl~E 2 GeneraLion of Stal~le C2Cl2 Myul~last GraLts A. NeT~IODS
15 ~7(:17 CR11 Cultllre aT~ lyocarrl;al ~rafting ~rotncol.
C2C12 myoblasts were obtained from ATCC. Cells were mai11tained in the undifferentiated state by culturing at low density in ~ligh glucose Dulbecco's Modified Eagle Media (DMEM) supplemerlted with 20% fetsl boYine serum, l96 20 chicken embryo extract, l00 units/ml penicillin ~nd l00 ~1g/1111 streptomycin. For some studies, myogenic differentiation was induced by cultuling ill DMEM
supplemented with 2% horse serum and antibiotics.
Immediately before injection, myoblasts were harvested 25 wit~1 trypsin, washed three times wit11 serum free DMEM and directly injected illtO the ventricular myocardium of adult syngeneic C3Heb/FeJ mice (Jackson Laboratories) under ope ~leart surgery as described in (7). Cells (4-l0 x 104) were in~ected in a volume of 2-3 Ill using a plastic 30 syringe fitted with a 30 gauge needle.
i stoloclv. Hearts were removed, cryoprotected, enlbedded al1d sectiol1ed as i n Exal11ple l . ~I al1d E stai11il1rJ and WOgS/14079 2 1 7 4 8 6 0 PCI/US94~13141 .

L311~-thymidine i11corporation assays were also cor1ducted as in Example 1. For immunol1istology, meLIIanol fixed sections (-20C, 10 min. ) were reacted wit}1 tlle monoclonal anti-skeletal myosill heavy chain antibody (MY-32, Sigma 5 Cllemical Corp. ) followed by rhodamine-conjugated sheep anti-mouse IgG F(ab')2 fragment (Boellringer Mann~leim), and visualized by epif luorescence .
Electron Mi croscor~Y. EM was performed as ill Example 1.
Electrocardior,lrAm AnAlYses. ECG anlyses were performerl as 10 in Exalllple 1.
Plas-nq pn7,yme ~ssaY. ~'EA was performed as in Example 1.
B. RESULTS
SeYeral myoblast cell lines are known which, as exemplified by C2C12 cells, have the capacity to 15 differentiate into myotubes in culture (13). C2C12 myoblasts were derived from cultured expla11ts of injured t}1igh m-rscle of C3H mice. When maintained in serum-ricl media, the r~lyoblasts proliferate rapidly and retain an undifferentiated phenotype. However, when cultured in 20 serllm-poor media myogenic differentiation is ir1duced. T}le C2C12 cells witl1dIaw from the cell cycle and fuse, tilereby fornling multinucleated myotubes. Myogenic differerltiation is also induced, as evidenced by the appearance ~f numerous muscle-specific gene products. Thus, in t~lis 25 model proliferation and myogenic differentiation are mutually exclusive (14). Myoblast differer~tiation in vitro is thought to mimic satellite cell IQediated rnyofiber regeneratior1 irl vivo.
Myoblasts were irljected directly into tl1e myocardiu 30 of syngeneic C3Heb/FeJ IQice and the viabili~y of tlle Wo 95/14079 PCr~ss~/13141 .
7 486~ -18-gra~ted 0aterial was assessed. Olle llundred percerlt (13/13) of t~le mice receiving intra-cardiac implallts of C2C12 myoblasts developed grafts in the heart. Viable grarLs were observed as long as six moll~hs 5 post-implantation (Lllis was the last time point assayed).
1ll all instances, tlle grafted material was not ellcapsulated. The diEferentiated status of the grafted C2C12 cells was determined by immunohistological assay with an anti-myosill lleavy chain antibody (MY-32). This 10 antibody does not react with myoblasts llor wi~ll cardiac myosin lleavy chain. Althougll differentiated C2C12 cells were observed in every heart receiving myoblast injections, tlle grafting efficiency of illdividual cells was IIOt deternlined. As an addi~iollal control, ~learts 15 bearing AT-l intra-cardiac grafts (see Example 1) were examirled with tlle MY-32 arltibody. No stairling was observed, thereby ruling out the possibility that the signal seen in the C2C12 grafts was due to skeletal myosi lleavy chain induction in llost cardiomyocytes.
Example 1 above demonstrates that AT-l cardiomyocytes form stable grafts in syngeneic myocardium. Ilowever, the observation that these cells retained the capacity for proliferation in vivo raised the possibility that fiustained cell division might be required for successful irltra-cardiac graf~ills. The proliferative status of tlle C2C12 grafts was tllerefore ~lr;lminPd. Virtually IlO DNA
syntllesis (as assessed by tritiated tllymidine incorporation) was observed, indicating ~llat ~lle majority of tlle grafted C2C12 cells had indeed witlldrawn from tlle cell cycle. Examination of serial sections illdicated tllat less tllan 0.1% of t~le cells in or near the yrafts were synthesizing DNA. This result Inost likely reflects fibroblast proliferation during the remodeling process. As witll tlle AT-l grafts, immunohistological analyses of C2C12 grarts failed to detect macrop~lage, WO 95/14079 2 ~ 7 4 ~3 6 D PCT/US94n31JI

illflamlllatory leukocyte or lympilocyte i1lfilLraLio:l at two months post-implantation, indicating ~lle absence of cllronic graft rejection by tlle syllgeneic hosts.
At the level of ligilt microscopy, tlle C2C12 5 intra-cardiac grafts exllibited cellular heterogeneity wil:h both H and E and MY-32 ill~munof luorescence staining .
E:lectron microscopic analyses were employed ill an effort to further characterize the cellular Illake-up of the C2C12 grafts. Toluidine-stained 1 ~lm sections were surveyed at 10 100 ,um intervals to locate graft sites for ~M analysis.
Once localized, ~llin sections were prepared rom the block. Cells with morphology typical of skeletal myocytes were observed throughout the graft. Abundant mitochondria localized between well developed sarcomeres were readily 15 detected. Prolninent Z bands and tllick and thill filaments were observed. Occasionally, expanded t-tllbules alld ruffled cell melllhranes were detected in the grafted n~yocytes. In addition to well developed myocytes, a second less differentiated cell type was observed in C2C12 20 grafts. Most notably, these cells exhibited a large nucleus to cytoplasnl ratio, with a prominent band of heterochromatin at the lluclear periphery. Moderate amounts of centrally located ileterochromatin were also detected.
Limited rough endoplasmic reticululn and few mitochondria 25 were observed in Lllese cells. Similar ultrastructural characteristics ~lave been ascribed to satellite cells i vivo and in culture (15, 16).
Two studies were initiated to assess any deleterious effects of C2C12 intra-cardiac grafts on ilOst lleart 30 function. In tlle first study, surface electrocardiograms failed to detect any appreciable differences between records from control alld experimenta1 nlice. All allimals examined llad normal P-QRS coupling, and exllibitecl normal sinus rilytllm witll an anesthetized heart rate of -Wo 95tl4079 2 1 7 4 ~ 6 o PcrluS94tl3141 appro~imately 400 beats per minute. Ttlese data indicate tllat the iIltra-cardiac myoblast graf ts did ~lot induce overt cardiac arrlly~hmias. Irl tlle second study, plasma LDH levels were monitored ill graf t-bearing animals . The 5 presellce of the cardiac LDI~ isoform in the circulation is a well establislled llallmark of myocardial infarction. Tl1e cardiac-specific LDH isoforms (isoforms l, Z, and 3) were not observed in plasma prior to grafting. Immediately after grafting, an increase in t~e cardiac isoforms was 10 observed in plasma, which most likely reflected damage to tlle host myocardium. A transient increase in the plas~rla skeletal LDH isoform (isoforln 5) was also observed, presumably ref lecting damage caused by the trans-thoracic il1cision. Plasma LD~I profiles returned to normal by 7 15 days post-illlplantation EXAMPI,13 3 Generation of Stable Fetal Cardiollryocyte Grafts A. METIIODS
Cardi, cYte Cell Clllture ~n~ yocardial Graftinn 20 Protocol. Transgenic mice were generated which carry a fusion gene comprised of the o~-cardiac myosin lleavy chain (M~IC) promoter and a modified B galactosidase (nLAC) reporter. To generate the MHC-nLAC transgenic mice, MIIC-IILAC insert DNA (see Figure l) was purified by 25 absorption onto glass beads, dissolved at a concentration of 5 ~Ig~ml, and rllicroinjected illtO the nuclei of one cell irlbred C31~3B/FeJ embryos according to established protocols (17). Polymerase Cllain Reaction (E'CR) allalysis was employed to ideIltify founder animals and to monitor 30 transgene segregation. The sense sLrand primer 5'-GGTGGGGG~:lcll~:ACCCCCAGACCTCTCC-3 ' was localized to t ~le MHC
promoter and the antisense strand primer 5'-GCCAGG~ ll~C~:AGTCACGACGTTGT-3' was localized to tlle , ., ., ... ,,,,,,, . , .. ,, .,, . , ,,, _, _ _ _ _ _ _ _ WO 95/14079 2 ~ 7 4 8 6 0 PCT/l~S94~13141 .

LAC reporter. PCR analyses were as described in (18) ~he MHC promoter consisted of 4.5 kb of 5' flankiIIg sequence and 1 kb of the yene eIlcompassing exons 1 Lhrough 3 up to but not including tlle initiation codon. The IILAC
5 reporter was modif ied so as t:o carry bo~ll a eukaryotic translation initiatiorI site and the SV90 nuclear localization signal (19). TIle mPl sequences carried an iIItron, as well as Lranscriptional termination aIId po3yadenylation signals froln the mouse protamine 1 gene.
For preparations for examinatiolI of B galactosidase (13GAL) activity and DAPl epifluorescence, trans~Jenic animals were heparinized (10,000 U/kg IP) prior to sacrifice by cervical dislocation. Hearts were l~laced in a beaker of gassed (95% 2' 5% CO2) KIIB buffer (105 mM
15 NaCl , 20 mM NaIIC03 , 3 . 8 mM KCl , 1 nlM RI~2PO4 , 1 . 2 nlM
MgS04, 0.01 mM CaC12, 1 mM mannitol, 10 IrlM taurine, 10 M dextrose, 5 n~ N~-pyruvate). Hearts were t~Iell IIung by the aorta and perfused with gassed KHB (0.5 ml/miII at 37C) containing 2.5 mM EGTA for five minutes, followed by 20 0.17% collagenase (Type I, Worthin~ton Biocllemical, Freehold NJ) in KHB. Hearts were perfused until flaccid and tIIe ventricles were minced witI~ scissors and isolated cells obtained by triturating with a Pasteur pipe~te.
After at least one hour of formalin fixation, suspensions 25 were filtered and smeared onto positively c~larged slides (Superfrost Plus, Fisher, Pittsburg~l PA), and allowed to rlry .
For isolation of single cells for injection, Lemales wit~I 15 day embryos (onset of pregnancy determined by 30 vayinal plugs) were sacrificed by cervical dislocation.
EIlIbryos were removed, decapitated, aIld hearts were arvested Ilnder E'BS, and velltricles alld atria were seE~arated. TrarIsgenic ventricles (identified by cardiac BGA~ activity) were digesLed in 0.1% collagenase Wo 95/14079 PCTNS94/13141 ~17ft~60 -22-~Wor~llingtoll) ill DPBS (Dulbecco's P1losphate Buffered Salille, Sigma) for 45 mi11utes, and were tril:urated wit.ll a Paste~1r pi pette in PC-l medium (Ventrex, Coons Rapids MN) wit~l 10% FBS, resulting in a suspension of single cells.
lmmediately after isolation, embryonic cardiomyocytes were washed t}1ree times with DPBS and directly injected irlto ~}1e Ye11tricular myocardium of syngeneic mice (Jackson Laboratories) under open heart surgery as in Example l.
1-10 x 104 cells were illjecterl in a volume of 2-3 Ill using a plastic syringe fitted wi~h a 30 gauge needle.
1~ist~loay. For H and E, X-GAL, immuno}1istology and t11yn1idine analyses, }1earts were removed following cervical dislocation and cryoprotected in 30~r sucrose, ~ lrlr~d a11d sectioned at lO ~lln with a cryomicrotome as in Example l.
H and E stai11ing, monitoring for intra-cardiac graft rejection, and assay for [311]-thyrnidine incorporation were also conducted as in l~xample l. ~o assay BGA~
activity, sections were }Iydrated in PBS, post-fixed in acetone:111et~1anol (l:l) a11d t}len overlaid Wit}1 mixtl1re containing l mg/ml X-GAL
(5-bromo-4-chloro-3-indolyl-B-D-galactoside), 5 mM
potassium ferricyanide, 5 mM potassium ferrocya11ide and 2 nM m~gnesiurn chloride in PBS. Positive staining is indicated by tlle appearance of a blue c}lromop~lore. Af ter treat1nent with primary antibody, signal was visualized by an avidin-biotin (ABC) kit (Vector Labs, Burlingame CA).
The 11eart was processed as described above, and sections were post-fixed ir1 methanol:acetorle (l:l), stained with 11 and E, alld coated Witl1 a t}lin layer of pllotograp~lic emulsior1 (Ilford L.9, Polysciences) dil11tea l:l wi~1 distilled water. Sections were exposed, developed, was~1ed, fixed alld washed as in Example l. X-GAL staining of single cell preparal:ions was as described above. For visualization of nuclei in single cell preparations, ., . ,, _ . . , . ,, . , . ,,, .. , , . , _ _ _ _ , . . . .

WO 95/14079 2 ~ 7 ~ ~ ~ O ~T/US94113141 slides were stained with DAPI ill PBS (0.2~3 IIM, tllree min.
at room temperature, Boehringer Mr~nnl-~im), waslled three times irl PBS, and wet-moullted in 2% propyl gallate dissolYed in glycerol. To obtain coronal lleart sections, 5 mice were sacrificed by cervical dislocation, hearts were llarYested and perfused on a Langendorff apparatus with 2%
glutaraldellyde in 0.1 M cacodylate buffer (pH 7.4). Af~er immersion fi~ation overnight in the same buffer, 200 llm coronal sections were made witll a vibratolne (Campden, 10 London, United Kingdom) . To localize the graf t, sections were pooled and stained for BGAL activity with X-GAL as described above.
Electron Microsco~Y. MHC-nLAC embryonic grafts were localized i~l coronal lleart sections as described above.
15 Ater trimming, t}le ~issue was post- fixed in 2% osmium tetroxide (Stevens Metallllrgical Corp., New York NY).
Tissue was ~llen dellydra~ed and ~ led in Ladd LX-112 (Ladd Research Industries). Grafted areas were further trimmed, thin sectioned, and stained with uranyl acetate 20 and lead citrate. Specimens were viewed on a P~lillips 400 translllission electron microscope as in Example 1.
Electrocardior,~r~ nalvses ECG analyses were collducted as in E2~ample 1.
B. RESULTS
Transgenic mice generated as above carried a fusio gene comprised of tlle M~IC promoter and a nLAC reporter.
nLAC carries the SV~0 I~uclear transport signal, w~licll results in the accumulation of B galactosidase activity in ~le nucleus of targeted cells. Four transgenic lineages were produced, and two (desigllated MllC-rlLAC-2 alld MIIC-IILAC-~) were selected for furtller analyses. To ensure Wo 95/14~79 PCr/US94113141 ~.1 7 4~60 ~hat t~le ME~C-nLAC transgene would provide a s~1itable cell lineage marker, B galactosidase (BGAL) activity was assessed in transgenic cardiomyocytes. Single cell preparations generated by retrograde collagenase perfusion 5 were P -~ed simultaneously for 13GAL activi-y al1d DAPI
epiEluorescellce. 99.0 ~ 0.~5% (n-400) of the transgenic:
cardiomyocyte nuclei expressed BGAL, wllereas no 13GAL
activity was detected in noncardiomyocytes. In additiol1, no nuclear BGAL activity was detected in nontransgenic 10 col1trol cardiomyocytes.
Single cell suspensions were prepared by collagenase digestion of hearts }larvested f rom embryonic day 15 ~ransgenic mice. Grea~er than 95% of the cardiomyocytes isolated by this tec1migue were viable as evidenced by dye 15 excl~lsion assay. Cardiomyocytes were delivered to left ve11~ricular free wall of syngeneic 11ontransgenic animals.
Gra~ted cardiomyocytes were readily and unambiguously identified by virtue of the nuclear 13GAL activity e1lcoded by ~11e MHC-IILAC transgel1e . Graf ted cardiomyocytes were 20 frequel1tly observed at sites distal to t11e point of delivery; it presently is not clear if this distribution of grafted cells reflects cardiomyocyte migratio11 or E~assiVe diffusion along dissection rlaneS produced by tlle injection process. Approxi1nately 50% (7/13) of tlle 25 animals receiving intra-cardiac inJeCtionS of embryol1ic cardiomyocytes developed grafts. T1lis freque11cy of successful graft formation is likely to increase as cell preparation and implantation pro~ocols are optimized.
Ligllt microscopic analyses of 1~ alld E stained 30 sections processed for 13GAL activity indicated ~hat grafted cardiomyocytes (1~lue lluclei) were ju~taposed directly wit~ ost cardiol11yocytes (purple nuclei).
Additional ~1 and E analyses failed to detect significant graf t encapsulation. T11e observed proximity of graf t a11d WO 95114079 2 1 7 4 8 6 o P~/rJsg4n3l4l I~ost cardiomyocytes alld absence of encapsulation are prerequisites for successful coupling between t~le two cell types .
Consecutive sections of a 19 day old intra-cardiac 5 graft were processed for H and E, BGAL activity, and macrophage and leukocyte immunoreactivity. No evidence for graft rejection was observed, despite the fact that the animals were not immune suppressed. As a positive control for the immunollistology, grafts of incompatible 10 MHC llaplotype were produced; graft rejection was clearly evident in these llearts. Tritiated thymidille uptake analyses indicated that only 0 . 6% (n-156) of the BGAL-positive nuclei were syntllesizing DNA, althougll noncardiomyocyte DNA synt~lesis was apparent. Since the 15 embryonic day 15 dollor cells were still mitotically active w~len grafted ( labeling inde~c of ca. 29%), the e~ceedinyly low level of DNA syntllesis observed in BGAL positive cells at 19 days post-graf ting suggested that t~le MHC-nLAC
embryonic cardiomyocytes had undergone terminal 20 ~lifferentiation.
Tlle ju~itaposition of graft and llost cardiomyocytes observed by ligllt luicroscopic analyses protnpted a determination whether direct i~ltercellular coupling coulll be detected between tlle two cell types. The X-(;AL
Z5 reaction product is an electron-dense precipitate wllich can be detected by transmission electron microscopy (TEM, see 19). Vibra~ome sectiolls from glutaraldehyde perfusion-fixed llearts were stained for BGAL activity, alld grafted regions t~lUs identified were trimmed and embedded 30 for TEM. 13GAL positive nuclei were readily observed by lig~lt microscopic analysis of 1 ,um sections. Tlle X-GAL
reaction product llad a perinuclear appearance due to a slig~lt degree of lluclear leac~ling wllich occurred during the elllbeddirlg process. Tlle same groups of cardiomyocytes Wo 95/14079 PCT/US94/13141 .
2 l 7 4 8 6~ -26-were idel1tified by TE:M arlalysis of a collseclltive t~lin section. Host cardiomyocytes, wi1ich were not readily ide11tified in tl1e lig11t micrographs due to the absence o perilluclear BGAL actiYity, were observed by electron icroscopy to be juxtaposed Witl1 t~1e grafted cells.
Numerous junctional complexes were present between the ost and graft cardiomyocytes, indicating a 11iyh degree of intercellular coupling. Many examples of intercellular coupling between host and graft cardiomyocytes were 10 observed throug11out the grafted regions. Importantly, intercellular connections could be traced from BGAL
positive cardiomyocytes through numerous host cells, thus demonstrating t~1at grafted cardiomyocytes could be participating in a functional syncytium.
In addition to docume11ti11g Ll1e presence of abundar1t i11tercellular coupling between grafted and host cardiomyocytes, the TEM analyses revealed that the grafted cardiomyocytes were highly differentiated. Normal characteristics of adult cardiomyocytes were observed 20 including myofibrillae forming complete sarcomeres, n~lmerous junctional complexes between cells and abundarlt mitocllondria. Indeed, aside from t11e presence o~ t~1e X-GAL ~eaction product, graf ted cardiomyocytes were indistingllishable from host cells. Further, binucleated, 2~ BGAL positive cells could be detected in the illtra-cardiac grafts. Because bil1ucleation is a cllaracteristic of adult rodent cardiomyocytes, this observation further supports that tlle grafted cardiomyocytes have undergone terminal differentiation.
Surface ECG ~ecordings were elllployed to determille if the presence of coupled embryonic cardiomyocyte grafts negatively influer1ced 11ost ~leart automaticity. ECG traces rom graft-bearing animals were indistinguishable from s11am operated controls, and exhibited P and QRS complexes WO95114079 2 ~ 74 ~o PCTIUsg4,l3l4, typical for 111ice. Tt1ere was 110 evidence or c~rdiac arri1yt~ 1ia in grat-i~earing animals, despite Lhe prese11ce of a 11iyi1 degree oE ir1tercellular coupling between graf ted a11d ilOst cardiomyocytes.
S EX~IPLE 4 Preparation of Substantially Pure Cardiu. lu~yLe Culture Embryonic steln cells were genetically modified in a manner enabling tlle production of a substantially i1omogeneous popula~ion of non-immortalized cardiomyocytes. T11e paren~al ES cell line (D3) was cotransfected with a pGI~-HYG (llygromycin) plasmid and a plasmid containir1g a M11C-neor gene. The pGR-HYG plas1nid provides selection for transfected ES cells, while the mMHC-11eor gene facilitates a second round of selection on differenLiated cells: incubal.ion in tlle presellce of G418 eliminates no~1-cardiomyocytes (ti1at is, cells in whic11 the MilC promoter is not active).
Stably tranfected ES cells were selected by growt~
t~le presence of hygromycin. The plasmids were linearized and intrûduced into the stem cells via electroporation at ll~0 ~1Farad, Z20 volts. The transfected cells were maintailled in DMEM supplemented wit:h 10% preselected FBS, -0.1 mM B-mercAptoethanol, nonessential amino acids, PenStrep a11d LIF, and transformants selected by t11e additio11 of 11ygromycin illtO the medium. Co-transfecta11ts were then iden~ified by PCR analysis specific for both transgenes. The transfections produced a cell line, designated 9A, wllich carries both transgenes.
Cardiogenesis was induced in 9A ES cells by plati11g 2 x l06 cells onto uncoated l00 mm bacterial petri dislles in t11e absence of LIF. Ater ~ days i~l culture, r1u1nerous patclles of cells ex11ibi~ing spontalleous co11~ractile Wo 95/14079 PCT/IJS94/13l4l activity (i.e. cardiomyocytes) were observed. At tllis pOillt, G418 was added to tlle media, alld the cells incubated for an additional 9 days. Durillg l:l~is Lreatment it was apparent that m~ny of the lloll-cardiomyocytes were being killed by tlle G418. Importalltly, the G418 had no discernible effects on the cardiomyocytes, which retained their spontaneous beating activity throuyhout the course of Llle experiment.
After a total of 9 days of G418 selection, t~le surYivins cells were dissociated wi~h collagenase an~
trypsin, and tllen replated onto fibronectin coated microscope slides. The cells were cultured an additional 24 llours to allow them to recover from passage, and then fixed for illlmunocytologic analysis. Cells were reacted witll MF20, a monoclonal antibody whicll recogllizes sarcomeric myosin. Cells were then individually counted for the presence or absence of sarcomeric myosin, a marker for cardiomyocytes. The results were as follows:
Total nulnber of cells counLed: 794 Number of MF20+ cells: 791 Number of MF20- cells: 3*
Percent caldiomyocytes: 99 . 6%
*The 3 cells whicll did not stain witll MF20 may still have been cardiomyocytes, since this antibody will not react with monomeric myosin.
It was thus demonstrated that tlle expression of drug ~neomycin) resistance under a cardiac-specific promoter ellables tlle selection of an essentially pure population of FS derived cardiomyocytes in culture. Sucll populations 30 can be used to fornl myocardial graf ts usillg procedures as discussed in Lhe Examples above.

WO 95114079 2 1 7 4 8 ~0 ~cT~usg4n3l4l E~AMPI,E S
Delivery of E'rotein via Graft A. METHODS
C2~ Cell Culture ~r~l tL~n~fectioll. C2C12 myoblasts ~ATCC) were maintained in tlle urldifferell~ia~ed state by culturing at low density in higll glucose Dulbecco's Modified Eagle Media (DMEM) supplemented with 20% fetal bovine serum, 1% chicken embryo e~tract, 100 ullits/ml eenicillin and 100119/ml streptomycin. For some studies, myogenic differentiation was induced ty culturing in DMEM
supplemented Witll 2% horse seru~n and allLibiotics.
A fusion gene comprised of tlle metallotllioneill (MT) ~romoter drivillg a modified Transforming Growth Factor-~3eta 1 (TGF-Bl) cDNA was obtained from Samllel arld colleagues (20). Transcriptional activity of Lhe metallotllionein promoter call be regulated by modulaLing tlle lleavy metal content of cell culture media. The TGF-Bl cONA carried site-directed mutations wl~ich reslllted in the conversion of Cys 23 and Cys 2 to serines. T~lis modification (described further in (21) ) results in t}le elAboration of a TGF-131 molecule which is unable to form dimers, and consequently is not subject ~o normal post trarlslational regulation. Cells e~rpressirlg t}le modified cDNA constitutively secrete processed, active TGF-Bl (20). The MT-TGF fusion gene was introduced illtO C2C12 myoblasts by calcium pllosehate transfection; stable tra~lsfectants were selected by virtue of co-transfectio wi~h an SV40-neor transgene. Four illdependerlt clolles were isolated, and presence of the transgene was collfirmed by Southern blot analysis. T~le relative levels of T~F-131 expression in the different clonal cell lines was illitially assessed by Nortllerll blot ana~ysis, and one line, designated C2(280), was utilized for subsequent expel iments .
_ _ _ _ _ , . , . , . . . . _ .. . . .. . . _ _ Wo 95/14079 PCrNS94/13141 MYscardial Graftina Frotocol. Tlle grafting protocol was as described in Example 1. Fourteen days post-surgery, graft bearillg animals were givell lleavy metal (25 mM ZnSO4 in drinking water). Zinc treatment was colltinued ulltil the terminatiorl of ~I~e experilllellt (1-4 weeks ) .
IlistoloQY. For paraffin sections, i~earts were fixed ill 10% lleutral buffered formalin, dehydrated througll graded alcohols, and infiltrated with paraffin. Tissue blocks were then sectioned at 6~1m. H and E staining was performed directly after sectioning according to manufacturer's specifications (Sigma Diagnostics).
For [3H~-L~lymidirle illcorporatioll, Illice were given a bolus and sacrificed as in Example 1. The lleart was removed and processed for paraffin embedding as described above. AuLoradiograplly was likewise conducted as in E~ample 1.
B. RESULTS
Expression of recombinant TGF-Bl in response to lleavy metal induct:ion was r~minr~rl in C2(280) myoblasts and myotubes. Transgene transcripts (1.8 kb) were readily disLillguished fron~ ~llose origirlating frorn the endogenous TGF-131 gelle (2.5 kb) by Northern blot analysis. Additior of lleavy metal to tlle culture media resulted in a marked illcrease of recombillant TGF-~31 transcripts in C2(280) myoblasts and myotubes. As indicated above, Illodified T(;F-~31 expressed by C2(280) cells sllo~lld have constitutive biological activity. To directly test Lhis, conditioned media from C2(280) Illyoblasts and myot:ubes was ~ milled by growth inhibition assay.

Wo 95/14079 PCr/USs4/13141 21 7~60 C2(280) myoblasl s were used ~o produce illtra-cardiac graf ts in syngeneic C311ebfFeJ rnice. The ~resence of grafts was readily detected i~ H and E stained sections.
100% (n > 50) of t11e animals receiving intra-cardiac 5 illjections of C2(280) cells went on i:o develop grafts.
Interestingly, 11 and ~: analysis suggested that the C2(280) grafts were somewhat less di~fere1ltiated as compared to t~ose produced with unmodified C2Cl2 cells. This result was confirmed by immunohistologic analysis with a 10 mo110clonal antibody whic11 recognizes skeletal myosin heavy cllain, C2(280) graEt transgene expression was assessed by imrnunollistology with an anti-TGF-Bl antibody; TGF-Bl expression was readily detected in C2(280) grafts. As a 15 llega~ive control, TGF-i3l expression was assessed in grafts produced by C2Cl2 D1yoblasts. As expected, tlle relative levels of TGF-Bl expression were markedly reduced in C2Cl2 grafts as compared to C2(280) grafts.
TGF-Bl is a well known angiogenic Eactor.
20 ~-t11ymidine incorporation analyses in vascular endot11elial cells was therefore assessed to determine iE
an en11anced angiogenic response occurred in graf ts expressing the MT-TGF transgene. ONA synt~lesis in vascular endothelial cells was readily apparent in C2(280) 25 grafts under administration of a single bolus injection of 3H-thymidine (H-THY). I11 contrast, vascular endothelial l~NA syntllesis was markedly reduced in non-transfected C2Cl2 grafts (Table l). To rule out tlle possibility that tlle angiogenic responses was due solely to graf t mass, 30 3H-thymidine incorporation was compared between similar size and ayed C2Cl2 a11d C2(280) graEts (Table l). A
narked increase ill tlle number of vascular endothelial cells synthesiziny DNA was apparel1t ill all oE ~1le -2~ 1~8~ -32-analyses . Finally, the thymidine incorporation assay also revealed tllat a percentage of the cJraf ted myo~lasts continued to proliferate. This observation is c~onsistent wiLII tl~e known ir~ itory effect of TCF-Bl 5 Illyodiferentiation, and most likely accounts for t~le ulldifferentiated appearance of the C2(280) grafts.

WO 9~114079 PCT~US94n3141 2 1 l4860 TABI.E 1 TGF-n:l I)elivery alld Vascular ,Fn~ hel; al DNA SYnl:~t~cis Time E'081: Zn C2C12 C2(280) 5 ~ ctioll T(.F-Bl(-) T~F-Bl(+) 1 Week Tot~l # 3H-'I'HY+ 0 8 Endotilt31ial Cells Total # of Vessel 35 17 Sections Counted # Synthetic Cells/ O.o + 0.00 0.46 + 0.036 Vessel SectioII
2 Wee~s Total # 3H-TIIY+ 1 7 Elldo~llelial Cells Total ~ of Vesæel 18 21 Sections Coulltetl # Syl~t.be~:ic Cells/ 0 . 06 + O . 056 0 . 34 + O . 052 Vessel SectioIl While t~le inYelltion Ilas been illustrated arld described in detail in the foregoing description, the same is to be 20 considered as illustrative and not restrictive in cllaracter, it being understood t~lat only t~le preferred elnbodiments have been described and that all modifications trIat come witlIin tlle spirit of L~le invention are desired to be p ro tec ted .

SUBSTITUTE SH~ (RU~

Wo 95/14079 PCTIUS94113141 ~114~ 34-REFERENCES
The following reerences, to tlle e~rtent tilat tlley ~rovi~ie exemplary procedural or other details s~lpplementary to those set forth llerein, are specifically 5 incorporated llereill by reference.
1. Tompson, 1.. Fetal transplants show promise. Science 257: 868-870, 1992.
2. Gussoni, E., Pavlath, G.K., Lanctot, A.M., Sllarma, K.R., Miller, R.G., Steinman, L. and ~lau, i-l.M. Normal 10 dystrophin transcripts detected in Duchenne muscular dystrophy patients after myoblast transplantation. Nature 356: 435-438, 1992.
3. Steinilelper, M.E., ~anson, N., Dresdller, K., Delcarpio, J.B., Wit, A., Claycomb, W.C. and Field, L.J.
15 Proliferation in vivo and in culture of diferelltiated adult atrial cardiomyocytes from transgenic mice. American Jourllal oi Physiology 259 (lleart and Circulatory ?llysiology 28): 111826-H1834, 1990.

4. Janse, M.J., Cinca, J., Morena, il., Fiolet, J.W., 20 Kleber, A.G., deVries, G.P., 8ecker, A.E. and Durrer, D.
Tl~e horder zone in myocardial ischemia. An electrophysiological, metabolic and histocilemical correlation in t}le pig heart. Circulation Researcl 44: 576-588, 1979 .
25 5. Spear, J F-, Mic~lelson, E.L., and More, E.N.
Cellular electrophysiologic characteristics of cilronically illfarcted myocardium in dogs susceptii~le to sustained velltricular ~achyarrhy~llmias. Journal of tlle Ainerican College of Cardiology 4:1099-1110, 1983.

2l 74860 Wo 95/14079 PCTIUS94/13141 6. Delcarpio, J.B., Lallson, N.A. Jr., ~ield, L.J. al~d Claycomb, W.C. Morpllological characterization of cardiomyocytes isolated fronl a transplantable cardiac tumor deriYed from transgenic mouse atria (AT-l cells).

5 Circulation Researcll 69:1591- 1600, 1991.
7. Roclcman, H.A., Ross, R.A., llarris, A.N., Kllowltorl, K.U., Stein~lelper, M.E., Field, L.J., Ross, J. Jr. and Cllien, K.R. Segregation of atrial-specific an~ inducible e~rpression of an ANF transgene in an in YiYo murine model 10 of cardiac llypertroplly. Proc. NaLl. Acad. Sci. USA
88:8277-8281, 1991.
8. Bullock, G.R. and Petrusz, P., in Techniques in Immunocytochemistry, Vol. II, Academic Press, New York, 1983 .
15 9. Fiel~:l, L-J. Atrial natriuretic factor-SV40 T antigen transgenes produce tumors and cardiac arrllyt}lmias in mice.
Science 239 :1029-1033, 1988 .
10. Katz, E., Steir~lelper, M.E., Daud, A. Delcarpio, J.B., Claycomb, W.C. and Field, L.J. Ventricular 20 cardion~yocyte proliferation in transgenic mice expressing -Cardiac Myosin Heavy Chain-SV40 T antigen fusion genes. American Journal of E'hysiology 262 (l~eart alld Circulatory Pllysiology 31) :H1867-1876, 1992.
11. Stei~ elper, M.E. and Field, L.J. "SV40 large 25 T-Antigen induces myocardiocyte proliferation in transgerlic mice", in The deYelopment and regenerative poterltial of cardiac muscle, John Oberpriller and Jean OLerpriller, eds. ~larwood Academic press, 1990.

Wo 95114079 PCrlUS94113141 ~1 7486~ -36-12. Stei~ elper, M.E. and Field, L.J. Cardiac Tumors and dysr~lythloias in trallsgenic mice. Toxicologic Pathology 18:
464-469, 1990.
13. Yaffe, D., and O. Saxel. Serial passaging and 5 differentiation of myogenic cells isolated from dystrop~lic mouse muscle. Nature 270:725-727, 1977.
lq. Nadal-Ginard, B. Commitment, fusion, and biocllemical differentiation of a myogenic cell line ill t:~e absence of DNA synthesis. Cell 15:885-864, 1978.
10 15. Brulli, C. Mitotic activity of Illuscle satellite cells during the early stages of rllabdomyosarcomas induction with nickel subsulfide. In Muscle Regeneration, A. Mauro, editor. Raven Press, New York, NY, 1979.
16. Rubin, L.L.,.C.E. Keller and S.M. Schuetze. Satellite 15 cells in isolated adult muscle fibers in tissue culture.
In Muscle Regeneration, A. Mauro, editor. Raven Press, New York, NY, 1979.
17. B. Hogan, F. Costarltini and E. Lacy, in Manipulating the Mouse Embryo - A Laboratory Manual, Cold Spring Harbor 20 Laboratory Press, Cold Spring Harbor, New York, 1986.
18. M.E. Steillhelper, K.L. Cochralle and L.J. Field.
Ilypotension in transgenic mice expressing atrial natriuretic factor fusion genes. I~ypertension 16:301-307, lg90 .
25 19. A.D. Loewy, P.C. Bri~gman and T.C. MettenleiteL.
Brain Researcll 555:346, 1991.

wo 95~l407g 2 1 7 4 8 6 ~ PCT/US94/13141 20. S.K. Samuel, et al. Autocrine induction of tutnor protease production and invasion by a Illetallotllionein-regulated TGF-beta 1 (Ser223,225). Elnbo Journal [JC:el~lb] 11(4) :1599-1605, 1992.
5 21. A.M. Brunner, et al. Site-directed mutagenesis of cysteille residues in t~le pro region of t~le transformillg growth factor beta 1 precursor. E~pression and cll~racterization of mutant ~roteins. Journal of Biological Chemistry 264(23) :13660-13664, 1989.
lû 22. E.J. Robertson. Embryo-derived stem cell lines, in Teratocarcinomas and embryonic stem cells: a practical approacil, E.J. Robertsorl, editor. IRL i~ress, Wasllingl:on DC, 1987.

~ 7 4~

(1) General Information:
(i) Applicallt: I.oren J. Fieldii) Title of Invelltioll: Myocardial Grafts Alld Celltllar Compositions Useful for Sal~le (iii) Number of Seqtlences: 2 ( iv) Corresponding Ad~ress:
(A) Addressee: Tllomas O. tlellry (B) Street: Bar~ Orle Tower, Suite 3700, 111 Mon~ment ~ircle (C) City: Indi.allapolis (D) State: ll)~li.al-la (E) Cotlntry: USA
(F) Zip: 46209v) Computer Readable Form:
(A) Medium Type: Diskette, 3.50 illcll, 1. 4 Mb storage ( B ) Comp u t e r: COMPAQ
( C ) Ope r a t i llg Sys t ell~: MSDO S
(D) Software: ASCI~
(vi) C~lrrent Application Data:
(A) Application Number: OB/153,664 (B) Filirlg Date: NovembeL 16, 1993 (C) Classification:
(vii ) Prior Application Data: None (viii) Attorney Information:
(A) Name: T~lo~nas Q. Hellry (B) Registratioll Number: 28,309 (C) Reference/Docket Numt~er: IU-30 (ix) Telecommunication Informatioll:
(A) Telephone: (317) 639-3456 (B) Telefax: (317) 637-75Gl (2) Information for Seq ID No:l:i) Sequence Characteristics (A) Lengtl~: 3U
(B) Type: N~lcleotide Seq~ler~ce (C) Strandedness: Single (D) Topology: Linear (ii) Molec~ule Type: Primer (xi) Sequence Descriptioll: SEQ ID NO:l:

40 (3) IrlEormatioll for Seq Jr~ l~o:2:
(i) Seq~lence Cllarac ~-eristis~
(A) Lellgt~l: 2~3 (B) Type: NucleDtide Seq~lellce (C) Strandedrless: Sillgle (D) Topology: Lillear (ii) Molecule Type: Prilner (xi) Sequence Descriptioll: SEQ ID NO:2:
GCC AGG GTT TTC CCA GTC ACG ~CG TTG T 2 8

Claims (28)

1. A myocardial graft in an animal, comprising:
a stable graft of viable skeletal myoblasts or cardiomyocytes incorporated in myocardial tissue of said animal.

2. The myocardial graft of claim 1 wherein said stable graft comprises cardiomyocytes.

3. The myocardial graft of claim 1 wherein said stable graft comprises skeletal myoblasts.

4. The myocardial graft of claim 1 which is non-tumorogenic.

5. The myocardial graft of claim 1 wherein said stable graft delivers recombinant molecules to the myocardial tissue.

6. The myocardial graft of claim 5 which is non-tumorogenic.

7. A method for forming a stable myocardial graft in an animal, comprising:
introducing viable skeletal myoblasts or cardiomyocytes in myocardial tissue of the animal so as to form a stable myocardial graft.

8. The method of claim 7 wherein said introducing comprises injecting the skeletal myoblasts or cardiomyocytes into myocardial tissue of the animal.

9. The method of claim 7 in which skeletal myoblasts are introduced into the myocardial tissue.

10. The method of claim 7 in which cardiomyocytes are introduced into the myocardial tissue.

11. The method of claim 7 wherein the myocardial graft is non-tumorogenic.

12. The method of claim 11 wherein the myocardial graft delivers recombinant molecules to the myocardial tissue.

13. A method for delivering recombinant molecules to myocardial tissue o an animal, comprising:
establishing a stable graft of viable skeletal myoblasts or cardiomyocytes incorporated in myocardial tissue of the animal, wherein the myoblasts or cardiomyocytes carry a DNA sequence which is expressed so as to deliver recombinant molecules to the myocardial tissue.

14. The method of claim 13 wherein the stable graft comprises skeletal myoblasts.

15. The method of claim 13 wherein the stable graft comprises cardiomyocytes.

16. The method of claim 13 wherein the graft is non-tumorogenic.

17. A cellular composition comprising a substantially homogeneous population of non-immortalized cardiomyocytes.

18. A method of obtaining a substantially homogeneous population of cells, comprising:
transfecting embryonic stem cells to introduce a marker gene enabling selection of one cell lineage from other cell lineages resulting from the differentiation of the stem cells;

causing the stem cells to differentiate in culture; and selecting said one cell lineage based on said marker gene.

19. The method of claim 18, comprising:
transfecting the stem cells to introduce (i) a first marker gene enabling selection of transfected stem cells from non-transfected stem cells and (ii) a second marker gene enabling selection of said one cell lineage from said other cell lineages;
selecting transfected stem cells based on the first marker gene;
causing the selected stem cells to differentiate; and selecting said one cell lineage based on said marker gene.

20. The method of claim 19 wherein said one cell lineage is cardiomyocytes.

21. A non-human animal having a stable graft of viable skeletal myoblasts or cardiomyocytes incorporated in myocardial tissue of the animal.

22. The animal of claim 21 which is a mammal.

23. The animal of claim 22 wherein the graft is non-tumorogenic.

24. The animal of claim 23 wherein the graft includes cardiomyocytes.

25. The animal of claim 23 wherein the graft includes skeletal myoblasts.

-41a-

26. A cellular composition comprising a population of cells at least 90% of which are non-immortalized cardiomyocytes.

27. The cellular composition of claim 26 wherein the cardiomyocytes are mammalian.

28. The cellular composition of claim 27 wherein the population of cells is about 95 to 100% comprised of cardiomyocytes.