CA2221710A1 - Gene therapy for myocardial ischemia - Google Patents

Abstract

The transgene-inserted replication-deficit adenoviral vector is effectively used in in vivo gene therapy for myocardial ischemia in a protective way, by a single intracoronary injection directly conducted deeply in the lumen of the coronary arteries in an amount sufficient for transfecting all cell types in the affected region, including cardiac myocytes.

Description

CA 0222l7l0 l997-ll-l9 W O 96/40195 PCT~US96~03~8 GENE T~ERAP~ FOR MYOCARDIAL ISCl~

This invention was made with Govc~ ent support under Grant Nos.
R0l HL,49343 and K14 HL 03150, awarded by the National Institute of 5 Health. The Cove~ cnt has certain rights in this invention.

~IELD OF TEIE INVEN~ION
The ylGsellt il~vellLioll relates to a recombinant adenovirus vector which is used in gene therapy for myocardial ischemi~, a method for producing same, 10 and a method of providing myocardial protection during revascularization or non-rev~c~ n7~ti~n procedures with the use of the vector. The vector effic,iently ~,A~lcsses a transgeme in the myocardiulll.

BACKGROUND OF TIIE ART
Myocardial ischc~uia occurs when the heart muscle does not receive an adequate blood supply and is thus deprived of rcccss~ y levels of oxygen and mltrients. The most cornmon cause of myocardial ischemia is atherosclerosis, c~ ing blockages in the blood vessels (co,olla,y arteries) that provide blood flow to the heart muscle. Present Lle~l~..cnt modalities inchllle pharmacologic 20 therapies, colo,laly artery bypass ~u~ y and percutaneous revascularization using techniques such as balloon angioplasty. In the setting of acute coronary occlusion (usually secondary to in-situ thrombosis of a coronary artery segment pre~riously nall~,wed by atherosclerosis) treatment of acute myocardial ischemiais often achieved by using thrombolytic agents ("clot busters") to open the 25 occluded arteries. Standard pharmacologic therapy is predicated on strategiesthat involve either increasing blood supply to the heart muscle or decreasing the demand of the heart muscle for oxygen and nutrients. Increased blood supply to the myocardium is achieved by agents such as calcium channel blockers or nitroglycerin compounds that increase the diameter of diseased 30 arteries by causing rel~Y~tion of the smooth muscle in the arterial walls.
Decreased demand of the heart muscle for oYygen and nutrients is accomplished either by agents that decrease the hemodynamic load on the W O 96/40195 PCTAUS9G/0~8 heart or those that decrease the contractile resyul~se of the heart to a given hemodynamic load. Surgical treatment of ischemic heart disease is based on the bypass of ~ e~e~ arterial segments with ~ tcgically placed bypass graft (usually saphenous vein or int~ l mZ~ y arter~ grafts). Percutaneous S revascularization is based on the use of catheters to reduce the nalluw..lg intli~e~e~l corUllal~ arteries. All of these strategies are based on the er~lic~tio~
of i~rht~mic episodes as the ~ ly treatment evidence, and all have limit~tion~
of their effe.;LivGlles~ in this regard.
Australian Patent Publication No. 27902/92, corresponding to 10 W093/06223, ~ closes adenovirus vectors for c~lession of desired genes in muscle cells to treat muscular dystrophy and thromboses. Although the '902 application discloses adelluvilu~ type S, it discloses specific vector constructs which are used for the treatment of muscular dy~LlO~hy. Texas Heart Ins~itute Joumal article 21:104-11 (1994) discloses the advantages of use of adenoviral 15 vectors in metli~ting efflcient direct gene transfer for ~ el~lillg restenosis. In particular, this article te~rh~s that the AdS virus transfected into 293 cells is an c,~llel,lely useful vector for gene ~ rel in coronary arteries. Herz, 18(4):222-229 (1993) discloses the advantages of use of replication-deficient adenoviral vectors in direct gene therapy. Like the prece~ling article, this article tç~ches 20 general advantages in use of the AdS virus. Amencan Joumal of Medical Science, 306(2):129-36 (1993) discloses the advantages of use of recombinant adenoviral vectors in gene Llall~rel. This article teSlches general advantages in use of adenoviral vectors, direct intravascular injection, and bFGF gene for treating colollaly occlusion. ~Iowever, all of the te~hin~ of the above 25 documents are too general to address the in vivo expression efficiency of a certain vector in myocardial protection.
In particular, none of the treatment modalities of the prior art addresses the issue of protection of the myocardium against irreversible damage when ischemia does occur. Protection of heart muscle against ischemia has been 30 demonstrated in the setting of ischemic pre-conditioning. This phenomenon occurs when the heart is exposed to brief periods of ischemic stress prior to a prolonged ischemic episode. During the brief periods of ischemia, production W~ 96140195 ~ ~ AUS96/ag8~8 of specific stress related factors is induced. These stress factors protect the myocar~liulll against subsequent and potentially more harmful, prolonged ischemic episodes. To date, attempts to induce these same factors by pharmacologic means have been unsuccessful.

BRIEF DESCR~PTION OF T~IE FIGURES
lFlGURE 1 is a schematic figure which shows rescue recombination con~lluction of a transgene encoding adenovirus.
lFlGURE 2 schem~tir~lly presents the strategy for introducing a foreign lû gene into the E1 region of a replication-deficient adenoviral vector.
~ IGURE 3 graphically presents the lactate dehydrogenase released by adenoviral infected H9c2 cells following simulated ischernia.
EIGURE 4 graphically ~ies~llL~ the creatine kinase released by adenoviral infected neonatal rat myocytes following simulated ischemia.
SUMMARY OF l~E INVENTION
The present invention has exploited a gene therapy approach to treat heart ~ e~e. An objective of the present invention is to provide a method of providing myocardial protection in which a stress related factor is produced in 20 the myocardium and is present at the time of ischemia so as to protect the myocardium against subsequent, potentially more harmful, prolonged ischemic episodes. This objective concerns protective effects, rather than therapeutic effects on myocardial ischemia.
Namely, one important aspect of the present invention is a method of 25 providing myocardial protection, comprising: delivering a replication-deficient adenoviral vector to a myocardium by intracoronary injection into the coronary arteries, ~lefelably a single injection of the vector, directly into the coronary arteries, so as to transfect cardiac myocytes, which do not undergo rapid turrlover, in the affected myocardium, said vector CO~U~ illg a transgene 30 coding for a stress related factor such as heat shock proteins HSP70i, HSP27, H[SP40 and HSP60, and the adenosine A3 receptor; and ~l es~ing the transgene in the myocardiull-, thereby raising the level of stress related factor in the affected region of the myocardium. By injecting the vector stock c.)..l;1;..i..g no wild-type virus deeply into the lumen of the colullaly arteries, erG.~bly into both the right and left colonaly arteries, of the myocardiu ~ rerelably in an ischemic milieu, ~lerel~bly in an amount of 10l~-10l3 viral 5 particles as determined by optical ~1e . ,~ilo.~ ,etry (more ~l crel~bly 10ll-10l2 viral particles), it is possible to locally transfect most of the cells, especially cardiac myocytes, which do not undergo rapid turnover, in the affected myocardiulll with the genes for a stress related factor, thereby Illhxillli,illg myocardial protection efficacy of gene ll~ rer, and Ini. i---i,i,-g the possibility of an 10 infl~""~toly response to viral proteins. If a ventricular myocyte-specific promoter is used, the promoter more securely enables ~A~l ession limited to the cardiac myocytes so as to effectively avoid the potentially harmful effects of angiogenesis in non-cardiac tissues such as the retina.
In the above method, myocardial protection is expected to be more effective in cases that (a) said patient has non-revasculari_ed ischemic heart disease and said protection is desired during planned non-cardiac ~ulge~y, wherein said vector is ~1. " i~ lel ed a plurality of days prior to the planned non-cardiac ~Ulg~lyj (b) said protection is desired in anticip~tion of complex percutaneous revasculari7~tion~ and wherein said vector is delivered at the timeof a ~ gnostic catheterization a plurality of days prior to the revasculari7~tinn;
(c) said protection is desired in anticipation of complex cardiac ~ul~ely, and wherein said vector is delivered at the time of a diagnostic cardiac catheteri7~tion; (d) said protection is desired in a donor heart to be transplanted into a host patient with a coronary ~ e~e, and wherein said vector is delivered at the time of a diagnostic coronary angiography prior to explanation to rule out coronary (li~e~e; and (e) said protection is desired in a patient with diffuse, nonrevascularizable coronary artery ~ e~e7 at the time of a diagnostic coronary angiography prior to explanation to rule out coronary tii~e~e, wherein said vector is delivered a plurality of times.
Another aspect of the present invention is an injectable adenoviral vector preparation, colll~ g a recombinant adenoviral vector, preferably in a final viral titer of 10l~-10l2 viral particles, said vector conL~ g no wild-type W O 96140195 PCT~US96 ~irus and colllpli~ing a partial adenoviral sequence from which the ElA/ElB
genes have been deleted, and a transgene coding for a stress related factor such as heat shock proteins HSP70i, HSP27, HSP40 and HSP60, and the adenosine A3 receptor, driven by a promoter flanked by the partial adenoviral 5 sequence; and a pharmaceutically acceptable carrier. By using this injectable adenoviral vector ~lepalatiorl, it is possible to perform effective adenovirus-netli~te~ stress related factor-coding gene transfer for the tre~tment of clinical myocardial ischemia without any undesirable effects.
A further aspect of the present invention is a method of production of 10 a viral stock con~ail~ g a recombinant vector capable of ~Lessillg a stress l~lated factor in vivo in the myocardium, CO~ liSillg the steps of cloning a transgene coding for a stress related factor such as heat shock proteins HSP70i,HSP27, HSP40 and HSP60, and the ~ no~ine A3 receptor into a plasmid co" 1~ ;" i 11~ a promoter and a polylinker fl~nkç~l by partial adenoviral sequences 15 of the left end of the human adenovirus 5 genome from which the ElA/ElB
genes have been deleted; co-transfecting said plasmid into m~mmz~ n cells transformed with the ElA/ElB genes, with a plasmid which contains the entire hulman adenoviral S genome and an ~d~itinn~l insert m~king the plasmid too large to be encapsulated, whereby rescue recombination takes place between 20 the transgene-inserted plasmid and the pl~mi~l having the entire adenoviral genome so as to create a recombinant genome containing the transgene without the ElAlElB genes, said recombinant genome being sufficiently small to be encapsidated; ide..Liryillg successful recombinants in cell cultures;
prop~g~ting the resulting recombinant in m~mmz~ n cells transformed with the 25 ElA/ElB genes; and ~ulirying the propagated recombinants so as to contain the recombinant vector, without wild-type virus therein.
Based on the present invention, effective protection to the heart muscle against myocardial ischemia such as that encountered during threatened myocardial infarction (heart attack) can be sul~ gly achieved. That is, the 30 present invention allows for protection against tissue necrosis secondary to prolonged ischemic episodes. Technical details are ~ieline~ted below.

W O 96/40195 PCT~US96/09858 DETAILED DESCRIPIION OF ~13~; PREFERRED EMBODIMENTS

Transgenes Coding for Stress R~ t~l Factors In the present invention, various stress related factors which are capable of protecting myocardial ischemia can be used; heat shock ~loteills HSP70i, HSP27, HSP40 and HSP60, and the ~Aeno~ine A3 rcce~ can be exemplified.
Adenosine plays an important role in m~ ting the phenomenon of ischemic preconditioning. The function of adenosine appears to be me~ te-l via A3 type ~ no~ine lece~lol~. In cell culture experiments in which the number of A3 receptors per cell was increased, the efficacy of an adenosine analogue (Gensia Pharmaceutical) to mitigate protection against ischemia was increased.
The coding regions for these factors are known in the art, and it is possible todownload these cDNA sequences from Genebank and other ~i~tzlb~nk~ over the internet, for example. Full or partial length cDNAs coding for the above factors can be used in the present invention. Other than above, sarcoplasmic reticular calcium ATPase can be used for the purpose of studying myocardial calcium h~n(1ling/hypertrophy.

EIelper T~ l~nt Replication Deficient E~uman Adenovirus 5 System The cDNA of interest is transferred to the myocardium, including cardiac myocytes, in vivo and directs constitutive production of the re-encoded o~ei~l. Viral vectors provide a means for highly efficient gene transfer.
Several dirrelelll gene llallsre~ approaches are feasible. The present inventorsinitially used the helper-independent replication deficient human adenovirus 5 system which has previously demonstrated transfection greater than 60% of myocardial cells in vivo by a single intracoronary injection. Non-replicative recombinant adenoviral vectors are particularly useful in transfecting coronary endothelium and cardiac myocytes resulting in highly efficient transfection after illll~venous injection. The recombinant adenoviral vectors based on the human adenovirus 5 {rrolo~, 163:614-617 (1988)} are mi~sing essential early genes from the adenoviral genome (usually E1A/ElB), and are therefore unable to replicate unless grown in "pellllissiv~" cell lines that provide the mi~.~ing gene W O ~6J40195 PCTnUS9C/~3~5 products in trans. In place of the mi~in~ adenoviral genomic sequences, a transgene of interest can be cloned and will be ~ essed in tissue/cells infectedwith the replication deficient adenovirus. Although adenovirus-based gene trans~er does not result in integration of the transgene into the host genome S (less than 0.1~o adenovirus-mediated transfections result in ll~llsge"e incorporation into host DNA), and therefore is not stable, adenoviral vectors can be prop~g~te~l to high titer and allow gene ll~n~rer to non-replic~ting cells.
Although the transgene is not passed to daughter cells, in the case of the adultcardiac myocytes, which do not divide, this is not an important limitation.
10 Rell OVil al vectors provide stable gene transfer, and high titers are now obtainable via LCl1UVAAUS pseudotyping ~Burns, et al., Proc. Natl. Acad. Sci.
(UI~A)~ 90 8033-8037 (1993)}~ but ~Ull~llt r~llovil~l vectors are unable to tr~n~ ce nonreplir~ting cells (adult cardiac myoc,vtes) efficiently. In addition, the potential haA~i.ls of llallsgelle incorporation into host DNA are not 15 warranted if short-term gene transfer is sufficient. Thus, a limited durationexpression of a stress related factor may be sufficient for temporary myocardialprotectinn, and transient gene transfer for some cardiovascular disease processes may be adequate and possibly ~lefe~ble.
Human 293 cells, which are human embryonic kidney cells lldl,srolllled 20 with adenovirus E1A/ElB genes, typify the permissive cell lines. However, other cell lines which allow replication-deficient adenoviral vectors to propagate therein can be used. Thus, other cell lines useful for this purpose include HeLa cells.

25 ~on~tPl~ctiQn Qf R~mhinant .4denov~ra! v. ecto:s All adenoviral vectors used in the present invention can be constructed by lhe rescue recombination technique developed by Frank Graham {~rology, 163:614-617 (1988)}. Briefly, the transgene of interest is cloned into a shuttlevector that coll~ains a promoter, polylinker and partial fl~nking adenoviral 30 sequences from which E1A/ElB genes have been deleted. As the shuttle vector, plasmid pACI (l~rology, 163:61~617 (1988)} (or an analog) which encodes portions of the left end of the human adenovirus S genome {T~rrology, CA 02221710 1997-ll-l9 W O 96/40195 PCT~US96/09858 163:614-617 (1988)} minus the early ~loteill encoding ElA and ElB sequences that are essential for viral replication, and plasmid ACCMVPLPA {J. Biol.
Chem., 267:25129-25134 (1992)} which contains polylinker, the CMV promoter and SV40 polyadenylation signal flanked by partial adenoviral sequences from 5 which the ElA/ElB genes have been deleted can be exemplified. The use of plasmid pAC1 or ACCMVPLA facilit~tes the cloning process. The shuttle vector is then co-transfected with a plasmid which collLaills the entire human adenoviral 5 genome with a length too large to be encapsi-l~te-l, into 293 cells.
Co-transfection can be conducted by calcium phosphate precipitation or 10 lipofection under conditions such as those disclosed byBiotechniques 15:868-872 (1993). As the plasmid having the entire adenoviral 5 genome, plasmid JM17 which encodes the entire human adenovirus 5 genome plus portions of the vector pBR322 including the gene for ampicillin rç~ict~n~e (4.3 kb) is exemplified. Although JM17 encodes all of the adenoviral proteins necessary 15 to make mature viral particles, it is too large to be encapsidated (40 kb versus 36 kb for wild type). In a small subset of co-transfected cells, rescue recombination between the transgene containing the shuttle vector such as plasmid pAC1 and the plasmid having the entire adenoviral 5 genome such as plasmid pJM17 takes place so as to create a recombinant genome that is 20 deficient in the E1A/ElB sequences, and that contains the transgene of interest but secondarily loses the additional sequence such as the pBR322 sequences during recombination, thereby being small enough to be encapsidated (see Figure 1). With respect to the above method, we have reported successful results {Giordano, et al., Circulation, 88:1-139 (1993) and Giordano and 25 ~m mo~d, Clin. Res., 42:123A (1994)}. The CMV driven ,l~-g~l~çtosidase encoding adenovirus HCMVSPlLacZ {C~in. Res. (Abs), 42:123A (1994)} can be used to evaluate efficiency of gene transfer using X-gal treatment.
The initial mode of gene transfer uses adenoviral vectors as delineated above. The advantages of these vectors include the ability to effect high 30 efficiency gene transfer (more than 60% of target organ cells transfected in vivo), the ease of obtaining high titer viral stocks and the ability of these vectors to effect gene transfer into cells such as cardiac myocytes which do not W O 96/40195 PCTAUS~G~3 undergo rapid turnover. One potential disadvantage is that the current generation of this vector does not result in stable gene transfer. Genes rel,~d to tlle myocardium by adenovirus vectors do not integrate into the host cell DNA and, therefore, do not get passed on to the progeny of dividing ' 5 cells (fibroblasts, endothelial cells, smooth muscle cells, etc. ). Genes transferred to the myocardium by ~;UllCllt generation adenoviral vectors remain active only for a period of weeks to months. This may actually be advantageous for certain clinical applications such as myocardial protection to induce a controlled amount of a stress related factor.
Alternatively, newer generation adenoviral vectors that have further de!letions in the adenovirus genome (in ~cl~lition to E1A/ElB) are under development. These vectors have the potential to effect longer term gene transfer and to be less immunogenic. If it is delell..ined that longer term genetransfer would be more efficacious and/or infl~mm~tory response to first 15 generation vectors becnm~s problem~tiç, these newer generation vectors could be used. In addition, if gene t"l.,sre~ limited to the arterAal wall proves as effic~rjnus as myocardial gene l~ rel to effect myocardial protection, ~ltern~tive method of gene ~ rel could be used including electroporation, use of hydrogel coated balloon catheters, use of liposomes or use of alternate 20 viral vectors including reL~vi~us or adeno associated viral vectors.

Cardiac-Speci~lc Promoters It is also proposed in the present invention to use cell l~rgelillg not only by delivery of the transgene into the coronal~ artery, but also, in additional 25 experiments, by using a ventricular myocyte-specific promoter. By fusing the tissue-specific transcriptional control sequences of left ventrAcular myosin light chain-2 (MLC2v) to a transgene such as the FGF-5 gene within the adenoviral construct, transgene expression is limited to ventAcular cardiac myocytes. The efficacy of gene c;~lession and degree of specificity provided by the MLC~V
30 promoter with lacZ have been determined, using the recombinant adenoviral ~ystem of the present in~rention. Cardiac-specific ~ es~ion has been documented previously by Lee, et al. {J. Biol. Chem., 267:15875-15885 (1992)}.

W O 96/40195 PCTrUS96tO9858 The MLC2V promoter is co.-l~,ised of 250 bp, and easily fits within the adenoviral-5 p~ck~ging con~L~ ts. The myosin heavy chain promcter, known to be a vigorous promoter of transcription, cannot be used because its large r si_e (5.5 kb) cannot fit within the adenoviral vector. Other promoters, such as 5 the troponin-C promoter, while highly efficacious and sufficiently small, lacks adequate tissue specificity. By using the MLC2V promoter and delivering the transgene in vivo, it is believed that the cardiac myocyte alone (that is witbout conco. . .i~ eApression in endothelial cells, smooth muscle cells, and fibroblasts within the heart) will provide adequate e AL~l~ssion of a stress related factor 10 such as heat shock ~lolei,ls HSP70i, HSP27, HSP40 and HSP60, and the adenosine A3 recep~or to promote myocardial protection. Limiting eA~le~sion to the cardiac myocyte also has advantages regarding the utility of gene transfer for the treatment of clinical myocardial ischemia. By limiting eA~lG~ion to the heart, one avoids any potentially harmful effect in non-cardiac tissues. In 15 addition, of the cells in the heart, the myocyte would likely provide the longest transgene e A~,ession since the cells do not undergo rapid turnover; t;Ayie~sionwould not therefore be decreased by cell division and death as would occur with endothelial cells. Subsequent studies will determine whether targeting gene eA~Ies~ion to the endothelial cells, and limiting ~A~,ession somewhat to 20 the colonaly endothelium by intracoronary injection, will be a sufficient means to deliver the transgene. Endothelial-specific promoters are already available for this purpose {Lee, et al., J. Biol. Chem., 265:10446-10450 (1990)}. As yet there are no fibroblast or smooth muscle cell promoters available that would efficiently limit eA~Iession of the transgene to smooth muscle or fibroblasts 25 within the heart.
In the present invention, targeting the heart by intracolul.aly injection with a high titer of the vector, and transfecting all cell types can m~ mi7e the probability for success. Namely, it is believed that a more dramatic result can be achieved if not only myocytes but also cell types other than myocytes are 30 targeted as well, although they are dividing cells.

CA 0222l7l0 l997-ll-l9 W O 96/40195 PC~AUS96JO9~5 ProI ~ti~-n and Purification of Adenoviral Vectors Successful recombinant vectors can be plaque purified accoldil,g to standard methods. The resulting viral vectors are prop~g~te-l on 293 cells which provide ElA and ElB functions in trans to titers in the 10l~-10l2 viral ~ S particles/ml range. Cells can be infected at 80% confluence and harvested at 3~48 hours post-infection. After 3 freeze-thaw cycles the cellular debris is pelleted by standard cel,llir.~g~tinn and the virus further purified by CsCl gradient ultracellllirugation (double CsCl gradient ultracGllLlirugation is ~ler~ d). Prior to in vivo injection, the viral stocks are rles~lte-l by gel filtration through sepharose columns such as G25 sephadex. The resulting viral stock has a final viral titer in the range of 10l~-10l2 viral particles/ml. The adenoviral con~Lluct must be highly purified, with no wild-type (potentially replicative) virus. Impure Coll~lluCtS can cause an intense immune response in the host ~nim~l From this point of view, propagation and purification must be conducted to exclude col~t~ nt~ and wild-type virus by, for example, idellLiryillg successful reconnbinants with PCR using a~lu~iate primers, conducting two rounds of plaque purification, and double CsCl gradient ultrace,ll~ifugation.

Delive~y of Recombinant Adlelloviral Vectors The viral stock can be in the form of an injectable preparation containing pharmaceutically acceptable carrier such as saline, as necessary.
The final titer of the vector in the injectable preparation is ~rere,ably in therange of 10l~-10l2 viral particles which allows for effective gene L,~ rer. The adenovirus transgene collsllucts are delivered to the myocaldiu", by direct in1racoronary injection using standard percutaneous catheter based methods under fluoroscopic guidance, at an amount sufficient for the transgene to be r~ essed to a degree which allows for highly effective therapy. The amount of the vector to be injected is preferably in the range of 10l~-10l3 viral particles ~more ~rere,~bly 10ll-10l2 viral particles). The injection should be made deeply (such as 1 cm within the arterial lumen) into the lumen of the coronary arteries, and ~lerel~bly be made in both coronary arteries, as the growth of W O 96/40195 PCT~US96/09858 collateral blood vessels is highly variable within individual patients. By injecting the material directly into the lumen of the coronary artery by coronary catheters, it is possible to target the gene rather effectively, and to .--i..i,..;,e loss of the recombinant vectors to the ~luAil,lal aorta during injection. Gene 5 expression when delivered in this manner is minim~l in the liver, and viral RNA
cannot be found in the urine at any time after intracoronary injection. Any variety of colo..aly catheter, or a Stack perfusion catheter, and so forth can be used in the present invention.

Protective Applications The replication deficient recombinant adenoviral vectors of the present invention allow for highly efficient gene transfer in vivo without evidence for cytopathic effect or i"ll~.""l~tion in the areas of gene delivery. Based on these results, it is believed that a high enough degree of in vivo gene transfer to effect in vivo functional changes is achieved. In particular, protective use of the vectors can be advantageous. In order to provide optimal protection to the myocardium, stress ~luteins must be present at the time of ischemia. This requires gene llhll~rer prior to anticir~te~lisch~mi~. Although the timing of many prolonged ischemic episodes is unpredictable, there are specific settings during which ischemia is anticipated. These ~ h .;ullls~allces specifically allow for gene transfer prior to the ischemic event. The following include some of the clinical settings in which a role for a therapeutic gene transfer approach is anticipated:
1. Gene transfer to provide myocardial protection duAng non-cardiac ~uigely in patients with non-revascularized ischemic heart e~e. This is common clinical problem which often requires a coronary revascularization procedure (e.g, angioplasty or bypass ~ulgely) before procee~1ing with the non-cardiac ~ulgely (hip replacement, gall bladder ~ulgely, etc.). If revascularization is not possible because the coronary vasculature is diffusely diseased or the Ask of cardiac ~urgely is thought to be unacceptably high, the non-cardiac ~ulgely is often precluded thus exposing the patient to further -CA 02221710 1997-ll-l9 W O 96/40195 PCT~US96~09858 morbidity. In these settings, gene ~rallsrer could be effected by intracoronary injection of the viral COll~lluCt several days prior to the planned non-cardiac ~ulgely such that levels of protective stress factors in the myocardium would be high during the anticipated ~ulgely.
Cardiac catheterization, necessary for gene delivery, does not require anesthesia and is very well tolerated by otherwise clinically colll~lol,lised patients.

2. Gene llal,~rel to provide myocardial protection during complex ~el~;ul~neous rev~c~ ri7~tion procedures (angioplasty, atherectomy, etc. ) during which prolonged ischemia is anticipated.
Percutaneous revascularization of the coronary v~clll~tllre is complicated 4% of the time by abrupt total closure of the target vessel.
Although ischemia can often be aborted by use of intracoronary thrombolytic agents, placement of intracolo.laly stents or emergent bypass surgeries, frequently associated with irreversible myocardial damage. Even when abrupt vessel closure does not occur, a significant number of procedures are complicated by "slow-flow" secondary to non-occlusive in situ thrombosis or micro-emboli~tirn to the distal coronary vasculature (common when treating diseased bypass grafts). These patients are also at high risk of peri-procedural myocardial damage.
Usually, these patients undergo a diagnostic cardiac catheterization several days prior to pe--;ulaneous revascularization. Intracol..l,a.y gene delivery to the myocardium at risk at the time of the tli~gnostic catheterization in anticipation of revascularization in high risk patients is believed to be conspicuously effective.

3. Gene trans~er to provide myocardial protection during complex cardiac ~u-~,e-y (complex revascularization procedures, valve - ~ulgely, complex congenital heart corrective surgeries, etc.). Coronary artery bypass ~u~ely is associated with a 3-6.5% incidence of peri-operative myocardial infarction. When peri-operative infarction does occur, peri-operative mortality is higher and in patients with residual left ventricular function and incomplete revascularization the long-term W O 96/4019S PCTrUS96/09858 prognosis is poorer. Gene L~ fcr by intracoronary injection at the time of rli~gnostiC cardiac catheterization just prior to ~ulgely is believed to be especially effective. We anticipated this approach would be helpful both in high-risk valve ~ul~ely and congenital heart disease surgeries.
~_ Gen~tra~sfer to ~rQ~d~ myQGardial protection to dc3ncr hearts prior to cardiac transplZlnt~tion. Damage to donor hearts as a result of unavoidable delays between the time of ~oypl~n~tion and the time of grafting into the host patient is responsible for a significant proportion of transplant related morbidity and failed transplantation procedures. Donor hearts often undergo ~ gnostic colonaly angiography prior to explanation in order to rule out coronary ~ e~e Gene llall~rel to the myocaldiulll by intracoronary injection at this time is believed to be particularly effective.
5. Gene transfer to provide myocardial protection to patients with diffuse, nonrevascularizable coronary artery ~ e~e. A subset of pzlti~nt~ with coronary artery disease cannot be safely revascularized.
This subset includes patients with diffuse coronary disease in whom bypass ~Ulgt;ly is technically not feasible, and patients with preclusive co-morbidity such as severe lung ~ e~e. In these patients, long-term gene transfer to protect the myocardium against chronic recurrent ischemia is believed to be particularly effective.

Animal Model of Myocardial i~ h~min Important prerequisites for successful studies on gene therapy are (a) constitution of an adequate animal model which is applicable to myocardial ischemia of an enormous patient population, and which can provide useful data regarding mech~ni~m~ for myocardial protection in the setting of myocardial ischemia, and (b) accurate evaluation of the effects of gene transfer. From this30 point of view, none of the prior art is s~ti~f~ctory. It is proposed in the present invention to use a porcine model of myocardial ischemia that mimics clinical coronary artery disease. Placement of an ameroid constrictor around the left W O 96/40195 PCTnUSg6~09858 c:i,L;.. n~Y (LC~c) coronary artery results in gradual complete closure (within 7 days of placement) with minimzll infarction (1% of the left ventricle, 4+ 1% of the LCx bed) {Roth, et a~, Circulation, 82:1778 (1990), Roth, et al., Am. J.
Physiol., 235:H1279 (1987), White, et al., cirG Res., 71:1490 (1992), Hammond, S et aL, Cardio~, 23:475 (1994), and Hammond, et al., J. Clin. Invest., 92:2644 (1993)}. Myocardial function and blood flow are normal at rest in the region previously perfused by the occluded artery (refell~ d to as the ischemic region), dlue to collateral vessel development, but blood flow reserve is incllfficient to IJlevellt ischemia when myocardial oxygen demands increase. Thus, the LCx 10 becl is subject to episodic ischemia, analogous to clinical angina pectons.
Collateral vessel development and flow-function rel~tinn~hips are stable within 21 ldays of ameroid placement, and remain unchanged for four months {Roth, e~ al., Circula~ion, 82:1778 (1990), Roth, et aL, Am. J. Physiol., 23S:H1279 (1987), White, et aL, cirG Res., 71L:1490 (1992)}. It has been documented by 15 telemetry that ~nims~ls have period ischemic dysfunction in the bed at risk throughout the day, related to abrupt increases in heart rate during feeding, el~ u~lions by personnel, etc. (unpublished data). Thus, the model has a bed with stable but inadequate collateral vessels, and is subject to periodic ischemia.
Another distinct advantage of the model is that there is a normally perfused 20 andl filnctioning region (the LAD bed) adjacent to an abnorrnally perfused and fimctioning region (the LCx bed), thereby offering a "control" bed within each Z11nilll~l Myocardial contrast echocardiography can be used to estim~te regional myocardial perfusion in the present invention. The contrast material is 25 composed of microaggregates of ~ ctnse and increases the echogenicity ("whiteness") of the irnage. The microaggregates distribute into the coronaly alteries and myocardial walls in a manner that is proportional to blood flow {Sl~yba, et al., Circulation, 90:1513-1521 (1994)}. Although it is difficult to obtain precise quantitative inforrnation with this technique, it has been shown 30 that peak intensity of contrast is closely correlated with myocardial blood flow as measured by microspheres {Skyba, et al., Circulation, 90:1513-1521 (1994)}.
Since the echocardiographic images can accurately identify the LCx bed, and W O 96/40195 PCTrUS~6J'O~~a8 myocardial contrast echocardiography can be used to evaluate myocardial blood flow, a hydraulic cuff occluder can be placed around the proximal LC~c adjacent to the ameroid.
PCR can be used to detect stress related factor DNA and mRNA in S myocar-liu-,- from ~nim~l~ that has received gene transfer. In addition, two weeks after gene l.~ re., myocardial samples from all five lacZ-infected ~nim~1~ show sllbst~nti~ cto~ e activity on histological inspecticn- In ~ litinn, using a polyclonal antibody to a stress related factor such as heat shock protein eA~ressed in cells and in myocardium from ~nim~l.c that have 10 received gene transfer can be demonstrated.

EXPE~IlVlENT 1: Ade..~ ~.1 C~l~slru~ls A helper independent replication deficient human adenovirus 5 system was used. The genes of interest were lacZ and FGF-5. The full length cDNA
15 for human FGF-S was released from plasmid pLTR122E {Zhen, et al., Mol.
Cell. Biol., 8:3487 (1988)} as a 1.1 kb ECOR1 fragment which includes 981 bp of the open reading frame of the gene, and cloned into the polylinker of plasmid ACCMVPLPA which col~laills the CMV promoter and SV40 polyadenylation signal fl~nk~cl by partial adenoviral sequences from which the 20 ElA and ElB genes (essential for viral replication) had been deleted. This plasmid was co-transfected (lipofection) into 293 cells with plasmid JM17 which contained the entire human adenoviral 5 genome with an additional 4.3 kb insert m~king pJM17 too large to be encapsidated. Homologous rescue recombination resulted in adenoviral vectors containing the transgene in the 25 absence of ElA/ElB sequences. Although these recombinants were nonreplicative in m~mm~ n cells, they could propagate in 293 cells which had been transformed with E1A/ElB and provided these essential gene products in trans. Transfected cells were monitored for evidence of cytopathic effect which usually occurred 10-14 days after transfection. To identify succçs~ful 30 recombinants, cell supernatant from plates showing a ~;y~o~atllic effect was treated with proteinase K (50 mg/ml with 0.5% sodium dodecyl sulfate and 20 mM EDTA) at 56~C for 60 minutes, phenol/chlorofollll extracted and ethanol _ W O 96~40195 Pcrnuss6/ossss precipitated. Successful recombinants were then identified with PCR using primers {Bio~echniques, 15:868-872 (1993)} complementary to the CMV
promoter and SV40 polyadenylation sequences to amplify the insert (the e.~,ccled 1.1 kb fragment), and primers {Biotechn~ues, 1~:868-872 (1993)}
~ S ~le~igned to conco. . ~ lly amplify adenoviral sequences. Successfulrec:ombinants then underwent two rounds of plaque pllnfic~tinn. Viral stocks were proF~g~tç-l in 293 cells to titers ranging between 101~ and 10l2 viral particles, and were purified by double CsCl gradient ce,lLlifugation prior to use.
Recombinant adenuvi- uses encoding ~3-galactosi~ e, or HSP70i were constructed using full length cDNAs. The system used to generate recombinant adenoviruses imposed a p~cking limit of 5 kb for transgene inserts. The genes proposed, driven by the CMV promoter and with the SV40 polyadenylation sequences were less than 4 kb, well within the p~çk~ging constraints.
Rçc--mbinant vectors were plaque purified by st~ncl~rd procedures. The resulting viral vectors were propagated on 293 cells to titers in the 101~ - 10l2 viral particles range. Cells were infected at 80% confluence and harvested at 36-48 hours. After freeze-thaw cycles the cellular debris was pelleted by standard ce,ll,irugation and the virus further purified by double CsCl gradient ultrac. llllifugation (~i~co~.Li"uouS 1.33/1.45 CsCl gradient; cesium prepared in 5 mM Tris, 1 mM EDTA (pH 7.8); 9Q000 x g (2 hr), 105,000 x g (18 hr)).
Prior to in vivo injection, the viral stocks were desalted by gel filtration through sepharose columns such as G25 sephadex. The resulting viral stock had a final viral titer in the 101~-1012 viral particles range. The adenoviral construct washighly purified, with no wild-type (potentially replicative) virus.
EXP~RIMENI 2: Adult Rat Cardiomvocvtes in Cell Culture Adult rat cardiomyocytes were prepared by Langendorf perfusion with - a collagenase containing perfusate according to standard methods. Rod shaped cells were cultured on l~minin coated plates and at 24 hours were infected with the ,B-galacto~id~e-encoding adenovirus obtained in the above Experiment 1 a~ a multiplicity of infection of 1:1. After a further 36 hour period the cells were fixed with glutaraldehyde and incubated with X-gal.

W O 96/40195 PCTrUS96/0~58 Consis~el.~ly 70-90% of adult myocytes e~ressed the ~-galacto~ e tr~n~gene after infection with the recombinant adenovirus. At a multiplicity o infection of 1-2:1 there was no cytotoxicity observed. ~s 5 EXPERIMENT 3: P;~ Myocardium In Y'vo The ,l3-g~l~ctocidase-encoding adenoviral vector obtained in the above Experiment 1 was prop~g~te~l in pel~ siv~ 293 cells and purified by CsCl gradient ultracell~lirugation with a final viral titer of 1.5 x 10l~ viral particles, based on the procedures of Experiment 1. An anesthetized, ventilated 40 kg 10 pig underwent thoracoto..ly and isolation of the left ci. ~;u~llrlex and left anterior descending colonaly arteries. A 26 gauge buLLellly needle was inserted in the mid left anterior descending (LAD) coronary artery and the vector (1.5 x 10l~
viral particles) was injected in a 2 ml volume. The chest was closed and the animal allowed to recover. On the fourth post-injection day the animal was 15 sacrificed. The heart fixed with perfused glutaraldehyde, sectioned and incubated with X-gal for 16.5 hours. After imbedding and sectioning the tissue was coul.lel~Lained with eosin.
Microscopic analysis of tissue sections (trans~lul~l sections of L AD bed 72 hours after intracoronary injection of adenovirus col,~ail,i.lg lacZ) revealed 20 a significant magnitude of gene transfer observed in the LAD coronary bed with many tissue sections demonstrating greater than 50-60% of the cells staining positively for ~3-g~l~ctosirl~ce. Areas of the myocaldiull- remote fromthe L.AD circulatory bed did not demonstrate X-gal staining and served as a negative control, while diffuse ~L"ession of a gene was observed in myocytes 25 and in endothelial cells. The majority of myocytes showed ,l3-galactosidase activity (blue stain), and, in subsequent studies using closed-chest intracoronary injection, similar activity was present 14 days after gene transfer (n=6). Therewas no evidence of infl~mm~tion or necrosis in the areas of transfection.

30 ~XPERIMENT 4: Pi~ Constriction Model ~nim~l.c and In~ u~ ntation Details are based on previous studies {Hammond, et al., J. Clin. Invest., W 096140~95 PCTrUS~G~'~9~58 9~2:2644-2652 (1993) and Roth, et al., J. Clin. Invest., 91:939-949 (1993)}.
~nim~l~ includes 14 domestic pigs, (30-40 kg). A left thoracotomy is performed under sterile conditions for instrumentation. Catheters are placed in the left atrium and aorta, providing a means to mP~llre regional blood flow, and to monitor ~res~ules. Wires are sutured on the left atrium to permit ECG
recording and atrial pacing. Finally, an ameroid is placed around the ~lu~.lual LC~x. After a stable degree of ischemia has developed, this treatment group (n=8) receives an adenoviral COlIS~l uct that includes genes for HSP70i (a heat slhock protein), driven by a CMV prornoter. Control ~nim~ (n=5) receives gene transfer with an adenoviral coll~Ll uct that includes a reporter gene, lacZ, driven by a CMV promoter.

Adenovirsl Constructs The helper independent repli~tinn deficient human adenovirus 5 system coll~ll ucted in the above Experiment 1 is used. The genes of interest are lacZ
and hsp70i. The m~tçri~l injected in vivo is highly purified and conl~ s no wild-type (replication competent) adenovirus. Thus the possible in vivo adenoviral infection and infl~mm~tory infiltration in the heart are ",illi",i,erl By injecting the material directly into the lumen of the coronary artery by coronary catheters, it is possible to "target" the gene rather effectively. Geneexpression when delivered in this manner in minimzll in the liver, and viral RNA cannot be found in the urine at any time after intracoronary injection.

Delive~ of the Trsnsgene Techniques for large animal ~ulgely are described in ~mmond, et al.
J. Clin. Invest., 92:2644-2652 (1993), Hammond, et al., J. Amer. ColL CardioL, 23:475-482 (1994), Roth, et al., J. Clin. Invest., 91:939-949 (1993), and Ping, et al., ~m. J. P~ysioL, 267:H2079 (1994). Injection of the collslluct (4.0 ml cont~ining 10ll viral particles of adenovirus) is made by injecting 2.0 ml into both the left and right coronary arteries (collateral flow to the LC~ bed W O 96/4019S PCTrUS96/09858 appeared to come from both vessels). Animals are anesthetized, and arterial access acquires via the right carotid by cut-down; a 5F Cordis sheath is placed.A SF Multipurpose (A2) coronary catheter is used to engage the colol~a,y arteries. Closure of the LCx ameroid is conrilllled by contrast injection into the S left main coronary artery. The catheter tip is then placed 1 cm within the arterial lumen so that minim~l material will be lost to the proximal aorta during injection. This procedure is carried out for each of the pigs.

.~cses~ nt of Myocardial Protection The strategy for myocardial protective studies include the timing of transgene delivery, the route of ~l.,.i,.i~l.~tion of the transgene, and choice of the stress related gene, using the aforesaid construct including a reporter gene(lacZ) and that including a stress related factor gene as well as the aforesaid pig models. The ameroid model of myocardial ischemia is chosen, and gene lS ll~llsrer is performed after stable. Gene transfer are effected by intracoronary injection of the viral collsLluct several days prior to non-cardiac ~ulgely or agnostic cardiac catheterization such that levels of protective stress factors inthe myocardiull, will be high during the anticipated ~ulgely or percutaneous rev~cc~ ri7~tinn. In addition, gene Ll~n~rei by intraculullaly injection is con-lucted at the time of diagnostic cardiac catheterization just prior to ~ulgGly.
Myocardial protection can be ~cceccetl by the aforesaid echocardiography and microscopic analysis.

EXPEI~IMENT 5: Adenovirus Mediated Gene Transfer of a Heat Shock Protein 70 (~SP70i) Protects Against Simulated Ischemia In the following experiment, applicants inserted the heat shock protein 70 gene into an adenoviral vector and showed that they could infect neonatal rat cardiomyocytes and the myogenic rat cell line H9c2, and could further achieve very high levels of ~ ~lession of the introduced gene (hsp70i).
Moreover, the cells infected with the adenoviral-hsp70i construct were also rendered tolerant to simulated ischemia as compared to cells infected with a control recombinant adenoviral construct.

CA 0222l7l0 l997-ll-l9 wo 96/40195 PCTnJS96/09858 The experiment showed that the adenovirus mediated ll~n~rer of hsp70i is not only efficient, but also highly effective in providing protection againstsimulated ischemic injury. The following describes the experiment in detail.

S l~TElRIALS AND Mh~CPDS
~ en Culture: Neonatal rat cardiomyocytes were cultured as previously described {Iwaki, et al., Circulation, 87:2023-2032 (1993)}. The embryonic rat heart-derived cell line H9c2(2-1) and the human embryonic kidney cell line 293 were both obtained from the AnneAcan Type Culture Collection, Rockville, 10 MD, and were maintained in DMEM supplemented with antibiotics (penicillin/slle~Lc~llycill/fungizone) and 10% fetal calf serum (FCS). Cells were infected in 60 cm tissue culture plates at about 80% confluency by adding enough of the adenoviral infectious stock to 1 ml of DMEM col.l&inil.g 2%
~eat inactivated FCS. To obtain a multiplicity of infection (MOI) of 10:1 or 15 1: 1, cells were incubated with viral C~ cts for 60 minutes with mild constant .ch~kin~; 2 ml of DMEM/2~ heat-inactivated FCS was then added and the plates incuh~ted for 2 days in a 37~C, CO2 incubator. Simulated ischemia of tlle infected neonatal rat cardiomyocytes and H9c2 plates were done as previously described {Mestril, et al., J. Clin. ~nvest., 98:759-767 (1994), hereby 20 incorporated by lererellce in its ent~ly)}.
Co,.~... lion of Ropl;~nt;on-DeficientAdenoviral Vectors: The inducible rat hsp70 described previously {Mestril, et al., Biochem. J., 298:561-569 (1994)} was inserted into the El region of an adenoviral vector construct using the general strategy previously described in Graham and Prevec, "Manipulation of 25 Adenovirus Vectors", in Methods in Molecular Biolo~v~ Vol. 7, pp 109-128, Murray, E. J. (eds), The Humana Press, Clifton, NJ (1991). Briefly, the rat hsp70 gene was cloned into the multiple cloning site of the adenoviral shuttle plasmid pACCMVpLpASR- (kindly provided by Dr. Robert D. Gerard, UllivelsiLy of Texas, Southwestern Medical Center) {Gomez-Fox, et al., J. Biol.
30 Gltem., 267:25129-25134 (1992)}. This plasmid contains the 5' end of the adenovirus serotype 5 genome (map units 0 to 17) where the El region has been substituted with the human cytomegalovirus enhancer-promoter followed W O 96/40195 PCTrUS9G~ 53 by the multiple cloning site from pAC19 and the polyadenylation region from SV40. The resulting plasmid was co-transfected with pJM17, a plasmid that contains the complete adenovirus 5 genome, into the human embryonic kidney cell line 293 using the calcium phosphate transfection method. Infectious viral 5 particles cull~ainillg the inserted hsp70 were generated by in vivo recombination in the 293 cells and were isolated as single plaques seven days later.
In addition, applicants also generated a control recombinant adenoviral Co~ uct that consisted of the pACCMVpLASR- plasmid without any insert.
The isolated plaques were propagated in 293 cells for several passages to 10 obtain high titer stocks. Viral particles were purified by CsCl ultracGl~Llifugation. The titer of viral stocks was determined either by plaque assay or deprotein~tinn of an aliquot of the viral stock and amount of DNA
determined by optical density {Barr, et al., Gene Therapy, 1:51-58 (1994)}.

Protein Analvsis: Cellular protein extracts were pre~ared from neonatal cardiomyocytes and H9c2 cells infected with adenoviral-hsp70i, the control adenoviral-SR- cull~Llucts or non-infected as previously described {Mestril, et al., J. Clin. Invest., 93:759-767 (1994)}. Protein concentration was determined by the Bradford Assay (BioRad Laboratories, Richmond, CA). Protein samples 20 (40 ~g each) were fractionated on an 8% SDS-polyacrylamide gel and electrotransferred onto nitrocellulose usinga semi-dry ele-;LluLl~llsrer apparatus (BioRad Laboratories). The nitrocellulose blots were reacted either with a monoclonal antibody C92F3A-5 (StressGen, Biotechnologies Corp., Victoria, BC) which binds specifically to the m~mm~ n inducible HSP70 or with a 25 polyclonal antiserum which binds to the COOH terminal of the m~mm~ n HSP70s and HSP90s {Mehta, et al., Circ. Res., 63:512-517 (1988)}. Blots were subsequently reacted with biotinylated secondary antibodies and ~l~e~Lavidin-horseradish peroxidase-conjugated systems (Vectastain, ABC kit; Vector Laboratories, Burlin~me, CA) and developed with diaminobenzidine, 30 tetrahydrochloride (DAB kit, Vector Laboratories).
Indirect I .".l~ .ofltlo~ f ~ ,cY: Plates of infected and non-infected neonatal cardiomyocytes and H9c2 cells were washed twice with ice cold PBS

W O 96140195 PCT~US96/09858 and iïxed with 100% ice cold methanol for 2 l.linules. The fixed cells were l:hen rehydrated with TBS containing 0.1% bovine serum albumin and reacted either with the monoclonal antibody against the inducible HSP70 (C92F3A-5) and subsequently developed with an ABC kit and VectorRed kit (Vector S l aboratories) or a FlTC-col-jug~ted polyclonal antibody raised against the hexon coat ~roleil, of adenovirus (AB1056F, Chemi-.on International, l[emecula, CA).
Analvtical Tech~iq~s Creatine kinase (CK) activity was measured spectrophotometrically using a commercial CK kit (Sigma Immunochemicals, 10 St.]_ouis, MO). CK activity release was ~A~le~scd as the percent of the totalCK: activity present in each plate norm~li7~d by the amount of ~lc lei,l in eachplate. Lactate dehydrogenase (LDH) activity was dete~lllined spectrophotometrically using a LDH test kit (Sigma). LDH activity released was eA~ressed as the percent of the total LDH present in each plate 1~ nonm~li7er1 by the amount of ~loteill present.
~ tntr~ri( n~Analysis: Results are expressed as the mean i standard error.Statistical signific~nce was ~ec~e-l by the Student's two-tailed test, unpaired t test and a probability value of < 0.05 was considered ~i~nifiç~nt.

Several studies have shown that the sole e A~res~ion of exogenous copies oiF hsp70 in cardiac tissue is sufficient to render the heart tolerant to ischemic injury {Marber, et al., J. Clin. I~zvest., 95:1446-1456 (1995); Plumier, et al., J.
Clin. Invest., 9S:1854-1860 (1995)}. This increased t;A~ression of the exogenous25 h~p70 does not occur only in cardiomyocytes, but also in non-myocytic cells, such as fibroblasts, endothelial and smooth muscle cells, present in the heart.
There~ore, applicants were interested in introducing and ~Apl~ ssing exogenous copies of hsp70i specifically in neonatal rat cardiomyocytes. For this purpose, appIicants collsllucted a replication-deficient recombinant adenoviral vector 30 ccntaining the inducible rat hsp70 gene {Mestril, et al., Biochem. J., 298:561-569 (19~4)}. The general strategy used to introduce a foreign gene into the El region of the replication-deficient adenoviral vector is represented schematically in Figure 2 (see also Materials and Methods). In addition, the control adenoviral cons~ ct was generated using the same scheme with the exception that it lacks an insert.
In order to characterize the levels of infection and ~ ession achieved S with this adenoviral-hsp70i vector, protein extracts were prepared from neonatal rat cardiomyocytes 48 hours after infection. The protein extracts were eY~minP~l by Western blot analysis. During the course of this study, three Western blots pro~ ce~l iclçntic~l results. A represe,.taLive Western blot was developed with a polyclonal antibody that binds to both HSP70 and HSP90.
10 The Western blot has three lanes. The first lane contained proteins from non-infected myocytes. The second lane contained ~loteins from myocytes infected with the control adenoviral vector (adenoviral-SR) at a MOI of 10:1. The third lane conL~ ed ~,otei"s from myocytes infected with the adenoviral-hsp70i (MOI of 10:1). The Western blot showed that the adenoviral-hsp70i co,.~,uct 15 infected myocytes con~LiLuLively expressed a large amount of the exogenous hsp70i. To better eY~mine the level of expression of the virally introduced hsp70i gene, applicants developed a second Western blot with a monoclonal antibody which binds specifically to the intlnçible HSP70. The second Western blot showed that while at a MOI of 1:1, the level of c~y'c~ion of HSP70 20 obtained with the adenoviral-hsp70i was lower than at a MOI of 10:1, it was still comparable to the normal expression of hsp70i in non-infected heat shocked cardiomyocytes (42~C, 60 minutes).
Since the control adenoviral vector (adenoviral-SR) lacked an insert, indirect immunofluorescence was used to detect infection by this adenoviral 25 co,l~L~ uct as well as that of the adenoviral-hsp70i co~,sL, uct in neonatal myocytes and H9c2 cells by using a polyclonal antibody that binds to the hexon assembly protein of adenovirus. The result was obtained of such an analysis on H9c2 cells that were infected with the adenoviral con~Ll ucts 48 hours prior to fixation of the cells. Panels A and B of the indirect immunofluorescence were infected 30 with the adenoviral-hsp70i construct (MOI of 1:1), panels C and D were infected with the control adenoviral-SR co,l~L,uct (MOI of 1:1) and panels E
and F were non-infected cells. In panels A, C and E, cells were reacted with W O 96/40195 PCTrUS9 the monoclonal antibody against the inducible HSP70. In panels B, D and F, cells were reacted with the polyclonal antibody against the adenoviral hexon semhly protein. High levels of e,~lession of hsp70i could only be observed in cells infected with the adenoviral-hsp70i and reacted with the monoclonal S antibody specific to the HSP70i (panel A). While the polyclonal antibody ~ ,in~t the adenovirus hexon assembly protein reacted with cells previously :infected with either adenoviral-hsp70i or adenoviral-SR cOll~Llucts (panels B
and D), this indirect immunofluorescent analysis was done in three dirrer~llt occasions during the course of this study to monitor the reproducibility of the 10 nnfection protocol. The results were identical in all three occasions. Identical results were obtained with neonatal rat myocytes.
In order to test if the adenoviral trall~lled HSP70i preserves its protective function against stress, H9c2 cells were infected either with the adenoviral-hsp70i (de~ign~te~l "Adhsp70" in Figure 3) (MO~I of 1:1) or the 15 ad~enoviral-SR (~le~ign~tecl "AdSR-" in Figure 3) (MOI of 1:1), and 48 hours later these cells were submitted to simulated ischemia. Applicants then measured the amount of lactate dehydrogenase activity released and rem~ining after simulated ischemia as a parameter of cellular damage. Figure 3 shows the results obtained from six independent experime~t~. In Figure 3, lactate 20 dehydrogenase (LDH) released is expressed as a pelcenl~ge of LDH released in control plates (infected but mot submitted to simulated ischemia) which is taken as 100%. The amount of LDH released was calcnl~te-l as the amount of LDH activity released, norrn~li7~1 by the amount of ~loteill released (Units/mg) over the amount of total IDH activity normalized by the total 25 amount of protein in each plate (total Units/mg). The p value is less that 0.05, i~dicating a statistically signifi~nt dirrerence, and denoted by the "*" in Figure 3. A similar series of experiments was performed with neonatal rat cardiomyocytes which were either infected with the adenoviral-hsp70i or the adenoviral-SR constructs (both at MOI of 1:1) and 48 hours later submitted to 30 simulated ischemia. Creatine kinase activity released and rem~inin~, after simulated ischemia, was measured to assess cellular damage to cardiomyocytes.
Figure 4 shows the results obtained in six independent experiments. In Figure CA 02221710 1997-ll-l9 W O 96/40195 PCT~US9C/0~358 4, the creatine kinase (CK) released is expressed as a pelce"lage of CK
rele~e-l in control plates (infected but not submitted to simulated ischemia) which is taken as 100%. The amount of CK released was calc~ te-l as the amount of CK activity released, norm~li7~1 by the amount of protein released 5 (Units/mg) over the amount of total CK activity, normalized by the total ~mount of protein in each plate (total Units/mg). The p value is less than 0.05,in-lir~ting a st~ti~ti~lly signific~nt difreience, and denoted by the "*" in Figure 4. In both of the above sets of eA~Jel i,l,ents, it was observed that the ession of the exogenous hsp70i seemed to render the cardiomyocyte and 10 H9c2 cells more tolerant to cellular damage due to the simulated ischemia.
The above results show that not only are neonatal rat cardiomyocytes easily infectable by adenoviral vector particles, but also that this infection does not seem to have any deleterious effects on the myocytes. Both cardiomyocytes and the myogenic H9c2 cells were readily and reproducibly infectable by the 15 above adenoviral constructs. One important point is that the infection of these cells with adenoviral vectors does not, in itself, elicited a stress response which can readily be noted by the lack of induction of the endogenous hsp70i gene upon infection with the control adenoviral-SR construct. In addition, both cardiomyocytes and H9c2 cells presented no delete,ious effects two days after 20 infection with adenoviral particles. Su,~ i"gly, no a~are"l morphological changes or noxious effects to the cell were evident even in cells infected with the adenoviral-hsp70i con~lluct that generated a large amount of HSP70i.
The sole presence of the exogenous hsp70i, in both neonatal cardiomyocytes and H9c2 cells, was capable of confe"illg protection against 25 simulated ischemia in vitro to these cells (Figures 3 and 4). It should be noted that the level of protection obtained by the adenoviral-hsp70i construct in H9c2cells was less than in the rat neonatal cardiomyocytes (Figures 3 and 4). One probable explanation for this di~erellce in the level of protection may be due to the nature of these two cells. While the rat neonatal cardiomyocytes are 30 non-dividing cells, the H9c2 cells are an established proliferating cell line.
Therefore, at two days post-infection (the time needed to obtain sufficient t;A~,es~ion of the exogenous protein, HSP70i), the number of adenoviral-hsp70i infected H9c2 cells may have been dilated out to a certain extent, resulting in a lower number of cells protected against simulated ischemia. Nonetheless, this would seem to prove that increased levels of HSP70i in the cardiomyocyte itself is able to enhance myocardial protection. Thus, the experiment supports the - 5 i ntrorlllction of adenoviral constructs of the present invention into the hearts of ;~nim~l~ to confer protection against myocardial ischemia.

Claims (18)

WE CLAIM:

1. A method of providing myocardial protection, comprising:
delivering a replication-deficient adenoviral vector to a myocardium by intracoronary injection into the coronary arteries, so as to transfect non-dividing cardiac myocytes in the affected myocardium, said vector comprising a transgene coding for a stress related factor selected from the group consisting of HSP70i, HSP27, HSP40, HSP60 and the adenosine A3 receptor; and expressing the transgene in the myocardium, thereby raising the level of stress related factor in the affected region of the myocardium.

2. The method of Claim 1, wherein said patient has non-revascularized ischemic heart disease and said protection is desired during planned non-cardiac surgery, wherein said vector is administered a plurality of days prior to the planned non-cardiac surgery.

3. The method of Claim 1, wherein said protection is desired in anticipation of complex percutaneous revascularization, and wherein said vector is delivered at the time of a diagnostic catheterization a plurality of days prior to the revascularization.

4. The method of Claim 1, wherein said protection is desired in anticipation of complex cardiac surgery, and wherein said vector is delivered at the time of a diagnostic cardiac catheterization.

5. The method of Claim 1, wherein said protection is desired in a donor heart to be transplanted into a host patient with a coronary disease, and wherein said vector is delivered at the time of a diagnostic coronary angiography prior to explanation to rule out coronary disease.

6. The method of Claim 1, wherein said protection is desired in a patient with diffuse, nonrevascularizable coronary artery disease, at the time of a diagnostic coronary angiography prior to explanation to rule out coronary disease, wherein said vector is delivered a plurality of times.

7. The method of providing myocardial protection according to Claim 1, wherein said transgene is driven by a CMV promoter which is contained in the vector, and said vector is delivered in an amount sufficient for transfecting all cell types in the affected region.

8. The method of providing myocardial protection according to Claim 1, wherein said transgene is driven by a ventricular myocyte-specific promoter which is contained in the vector, and said vector is delivered in an amount sufficient for transfecting exclusively cardiac myocytes in the affected region.

9. The method of providing myocardial protection according to Claim 1, wherein said vector is delivered in the form of a viral stock having a final viral titer of 10 10 - 10 12 viral particles.

10. An injectable adenoviral vector preparation, comprising:
a recombinant adenoviral vector, said vector containing no wild-type virus and comprising:
a partial adenoviral sequence from which the E1A/E1B gene have been deleted, and a transgene coding for a stress related factor selected from the group consisting of HSP70i, HSP27, HSP40, HSP60 and the adenosine A3 receptor, driven by a promoter flanked by the partial adenoviral sequence; and a pharmaceutically acceptable carrier.

11. The preparation of Claim 10, wherein said vector is in a composition having a final viral titer of 10 10 - 10 12 viral particles.

12. The injectable adenoviral vector preparation according to Claim 10, wherein said promoter is a CMV promoter or a ventricular myocyte-specific promoter.

13. A method of production of a viral stock containing a recombinant vector capable of expressing a stress related factor in vivo in the myocardium, comprising the steps of:
cloning a transgene coding for a stress related factor into a plasmid containing a promoter and a polylinker flanked by partial adenoviral sequences of the left end of the human adenovirus 5 genome from which the E1A/E1B genes have been deleted;
co-transfecting said plasmid into mammalian cells transformed with the E1A/E1B genes, with a plasmid which contains the entire human adenoviral 5 genome and an additional insert making the plasmid too large to be encapsidated, whereby rescue recombination takes place between the transgene-inserted plasmid and the plasmid having the entire adenoviral genome so as to create a recombinant genome containing the transgene without the E1A/E1B genes, said recombinant genome being sufficiently small to be encapsidated;
identifying successful recombinants in cell cultures;
propagating the resulting recombinants in mammalian cells transformed with the E1A/E1B genes; and purifying the propagated recombinants so as to contain the recombinant vector, without wild-type virus therein.

14. The method of Claim 13, wherein the viral stock is at a final viral titer of between 10 10-10 12 viral particles range.

15. The method of production of a viral stock according to Claim 13, wherein said plasmid into which the transgene is cloned is plasmid pAC1 or plasmid ACCMVPLPA.

16. The method of production of a viral stock according to Claim 13, wherein said identification comprises the steps of:
monitoring transfected cells for evidence of cytopathic effect;
treating the cell supernatant from cell cultures showing a cytopathic effect with proteinase K, followed by phenol/chloroform extraction and ethanol precipitation;
identifying successful recombinants with PCR using primers complementary to the CMV promoter and primers complementary to adenoviral sequences; and undergoing two rounds of plaque purification.

17. The method of production of a viral stock according to Claim 13, wherein said purification comprises the steps of:
propagating the resulting recombinants in cells transformed with the E1A/E1B genes to titers in the 10 10-10 12 viral particles range;
purifying the propagated recombinants by double CsCi gradient ultracentrifugation; and filtering the purified recombinants through sepharose columns.

18. The method of Claim 13, wherein said stress related factor is selected from the group consisting of HSP70i, HSP27, HSP40, HSP60 and the adenosine A3 receptor.