WO2002064152A1 - Cardiomyocytes and methods of culture of same - Google Patents

Abstract

We have established a heart slice primary culture system that allows the mechanical separation of distinct cardiac cell populations and the assay of their relative mitogenic and trophic effects on cardiomyocyte proliferation and survival. Using this system, we have found that a signals from the epicardium, but not the trabeculae and endocardium, is required in heart slices for continued myocyte proliferation and survival. Blockade of retinoic acid (RA) signaling from the epicardium inhibits cardiomyocyte proliferation and survival. Interestingly, this inhibition is reversed by the addition of another epicardial secreted factor, erythropoietin (Epo). These factors do not act directly on myocardial cells, but rather induce yet another soluble epicardial factor that directly rescues proliferation of cardiomyocytes. Thus, the epicardium controls normal heart growth in ventricular segments of the heart by secreting a cardiomyocyte mitogen whose expression or activity is regulated by both RA and erythropoietin signaling within the epicardium.

Description

CARDIAC MYOCYTES AND METHOD OF CULTURE OF

SAME

FIELD OF THE INVENTION [01] This invention relates generally to heart muscle, and specifically to cardiomyocytes.

BACKGROUND OF THE INVENTION

[02] Cardiovascular disease is the killer number one in the western world, in both women and men. Evanoski, Crit. Care Nurs. Clin. North Am. 9(4): 489-96 (Dec. 1997); Smith, Am. J. Cardiol. 82(10B): 10T-13T (1998).

Cardiovascular disease and heart infarction are characterized by cardiac cell death. In contrast to adult mammalian skeletal muscle, which has capacity for tissue repair due to the presence of satellite cells, the adult mammalian myocardium lacks the ability to regenerate following an injury. Soonpaa & Field, Circ. Res. 83(1): 15-26 (1998). Fully differentiated adult mammalian myocardial cells generally do not divide. As a consequence, myocardial loss is an irreversible event that leads to scar formation, ineffectual heart function, and cardiac failure.

[03] Currently, allogenic heart transplantation is the only effective therapy for end-stage heart failure. In the United States alone, 2,340 heart transplants were performed in 1998. According to the American Heart Association, however, between 20,000 and 40,000 Americans each year could benefit from heart transplantation or alternative therapies.

[04] An increase in cardiomyocyte cell number in diseased hearts could improve function. Cell transplantation is a promising new approach to help patients with end-stage organ failure, but an important limitation of cell transplantation into damaged myocardium is the small amount of viable cardiac cells available for transplantation. For this reason, cell expansion is usually performed prior to cell transplantation. For cardiac cells, cell expansion depends upon the proliferation capacity of cardiomyocytes. [05] In previous cardiomyocyte cultures, primary cultures of cardiomyocytes are usually generated by dissociating and culturing cardiac cells. To assay for proliferation, primary cultures of dissociated rat neonatal cardiac cells are often used. To be precise, "cardiac cells" rather than "cardiomyocytes" or "cardiac myocytes" are used. Although poorly distinguished in the literature, the described cultures of neonatal rat cardiac cells only initially consist mostly of cardiomyocytes. During culture, the cardiomyocytes are rapidly overgrown by non-cardiomyocytes, due to the increased rate of proliferation of the latter cell types. According to our own experiments, after only 4 days more than 50% of the cells in culture are not stained for cardiomyocyte markers. Cultured dissociated mouse (mid-gestation) cardiac cells are also overgrown by non- cardiac cells. In our experiments, after 5 days in culture, more than 80% of the cells are negative for cardiomyocyte markers.

[06] What is needed in the art is a method for increasing the number and survival of cardiomyocyte or cardiomyocyte-like cells, either in culture or in diseased hearts, which could greatly improve heart muscle function and prevent or delay heart failure.

SUMMARY OF THE INVENTION

[07] The invention provides a heart slice primary culture system that allows cardiomyocytes to continue to proliferate in culture. The heart slice culture system can be prepared as described below. The heart is from a vertebrate, preferably a bird or mammal. In addition, the invention provides a population of cardiac cells that enriched for epicardial cells as evidenced by the expression of epicardial cell markers as well as by the absence of beating cells (cardiomyocytes).

[08] The invention also provides a heart slice culture bioassay for cardiomyocyte proliferation and survival.

[09] The invention further provides a conditioned medium that leads to cardiomyocyte proliferation and survival. [10] The mvention provides a method for repairing damaged heart tissue, h one embodiment, a composition comprising cardiomyocytes that have

2

BST99 1231787-1.061039.0010 been proliferated in vitro is administered to the damaged heart. In another embodiment, conditioned medium, or a factor that cardiomyocyte proliferation and survival is administered to the damaged heart

[11] The invention provides a method for investigating signaling pathways in cardiomyocytes. The heart slice culture system of the mvention is useful for understanding the molecular and cellular biology of these cells, which ultimately might lead to therapeutically important insights for cell transplantation into the human heart.

BRIEF DESCRIPTION OF THE DRAWINGS [12] FIG. 1 shows that cardiomyocytes proliferate in the heart slice primary culture system of the invention. FIG. 1A is a schematic drawing describing the heart slices used in the following experiments. Coronal sections of the ElO chick heart were dissected and cultured as intact slice or after removing either the epicardium alone or both the endocardium and trabeculae. FIG. IB is a photograph showing immunofluorescence for MF-20 (upper panel,

MF-20) and nuclear stain with DAPI (lower panel, DAPI) of ElO chick heart slices after 2 and 4 days of culture (2d, 4d). Note that 99% of the cells in the heart slice are MF-20 positive at all culture periods. FIG. 1C is a photograph showing immunofluorescence for BrdU (upper panel, BrdU) and nuclear stain with DAPI (lower panel, DAPI) of a two (2d), three (3d) and four day (4d) culture of ElO chick heart slices. The sequence of reagent additions is described schematically (upper part of the Figure). FIG. ID is a quantitation of FIG. lC. All experiments depicted in this Figure, as well as the following Figures, are representative experiments. [13] FIG 2 shows that the epicardium, not trabeculae or endocardium, is required for cardiomyocyte proliferation and for survival. FIG. 1 A is a four- day "slice" culture. Note that removing the epicardium dramatically reduces the number of cells that migrate out of the slice. FIG. 2B is a photograph showing immunofluorescence for BrdU (upper panel) and nuclear stain with DAPI (lower panel) of an ElO chick heart "slice"-culture. The intact slice (slice), removing the epicardium (-epicardium) or removing the endocardium and

3

BST99 1231787-1.061039.0010 trabeculae (-endocardium -trabeculae) were cultured for three (3d) and four days (4d). The sequence of reagent additions is described schematically (upper part of the Figure). FIG. 2C is a quantitation of the experiment depicted in FIG. 2B. Note that nuclei stain for BrdU in an intact slice (slice) and after removal of the endocardium and trabeculae (-trabeculae -endocardium), but do not stain for

BrdU after removal of the epicardium (-epicardium). Note also that the number and appearance of nuclei is dramatically reduced after removal of the epicardium (-epicardium), but not after culture of the intact slice (slice) or after removal of the endocardium and trabeculae (-trabeculae -endocardium) after four days of culture.

[14] FIG. 3 shows that retinoic acid (RA) is required for cardiomyocyte proliferation and for survival. FIG. 3 A is a photograph showing immunofluorescence for BrdU (upper panel, BrdU) and nuclear stain with DAPI (lower panel, DAPI) of ElO chick heart slices with (+Ro) or without (control) addition of the RA antagonist Ro415253 after three (3d) or five days (5d) of slice culture. The sequence of reagent additions is described schematically (upper part of the Figure). Note that the nuclear staining for BrdU (BrdU) is dramatically reduced upon addition of the RA antagonist (+Ro) after three (3d) and five days (5d) of slice culture. The number of nuclei in the slice treated with the RA antagonist (+Ro) is comparable to the number of nuclei in the untreated slice (control, DAPI) after three days (3d), but decreases dramatically after five days (5d) of culture in the presence of the RA antagonist (+Ro, DAPI). FIG. 3B is a quantitation of FIG. 3 A.

[15] FIG. 4 shows that retinoic acid (RA) does not overcome the inhibition of cardiomyocyte proliferation by removal of the epicardium. FIG. 4A is a photograph showing immunofluorescence for BrdU (upper panel, BrdU) and nuclear stain with DAPI (lower panel, DAPI) of ElO chick heart after three days of culture of intact slices (control) or after removal of the epicardium (- epi), without (-RA) or with addition of RA (+RA). The sequence of reagent additions is described schematically (upper part of the Figure). Note that the nuclear staining for BrdU is dramatically reduced after the removal of the

RST^QQ 19117R7-1 (161039 0010 epicardium (-epi) with (+RA) or without addition of RA (-RA). FIG. 4B is a quantitation of FIG. 4 A.

[16] FIG. 5 shows that Epo overcomes the inhibition of cardiomyocyte proliferation by retinoic acid receptor antagonists and is regulated by retinoic acid. FIG. 5 A is the results of a RT-PCR of E 13 mouse heart slices (slice culture) or non-cardiac cells, epicardial cells (Non-CM) cultured for one day with or without Retinoic acid, all-t-RA 10-7 M and 9-cis RA 10-9 M. FIG. 5B is a photograph showing immunofluorescence for BrdU (left panel, BrdU) and nuclear stain with DAPI (right panel, DAPI) of ElO chick heart slices after three days of culture with (+Ro) or without (control) addition of RA antagonist Ro 415253, or, with addition of RA antagonist and erythropoietin (+Ro+EPO). The sequence of reagent additions is described schematically (upper part of the Figure). FIG. 5C is a quantitation of FIG. 5C. Note that the dramatic reduction of the nuclear stain for BrdU upon treatment with the RA antagonist (+Ro) is overcome by the addition of erythropoietin

(+R0+EPO). Note also, that the addition of erythropoietin (+R0+EPO) reverses the reduction in the number of nuclei (DAPI) observed after RA antagonist (+Ro) treatment alone.

[17] FIG. 6 shows that Epo does not overcome the inhibition of cardiomyocyte proliferation by removal of the epicardium. FIG. 6A is a photograph showing immunofluorescence for BrdU (upper panel, BrdU) and nuclear stain with DAPI (lower panel, DAPI) of ElO chick heart after three days (3d) of culture of intact slices (control) or after removal of the epicardium (-epi), with (+EPO) or without addition of erythropoietin (-EPO). The sequence of reagent additions is described schematically (upper part of the Figure). FIG. 6B is a quantitation of FIG. 6 A. Note that the nuclear staining for BrdU is dramatically reduced after the removal of the epicardium (-epi) with (+EPO) or without addition of erythropoietin (-EPO).

[18] FIG. 7 shows that Epo induces the secretion of a soluble mitogen/survival factor for cardiomyocytes. FIG. 7A is a photograph showing immunofluorescence for BrdU (upper panel, BrdU) and nuclear stain with DAPI

RCTQ ) ι-m S7-ι nfiin3Q nmo (lower panel, DAPI) of ElO chick heart after four days of culture of intact slices (control, panel a, b) or after removal of the epicardium (-epicardium, panel c-j). Slices after removal of the epicardium (-epicardium, panel c-j) were cultured without (panel c, d) or with the addition of Epo (+epo, panel e-j) to the culture medium, and, medium conditioned by epicardial cells in the absence (panel g, h) or presence of Epo (w/epo, panel i, j) was added to slices after removal of the epicardium. The sequence of reagent additions is described schematically (upper part of the Figure). FIG. 7B is a quantitation of FIG. 7 A. Note that the nuclear staining for BrdU is dramatically reduced after the removal of the epicardium (panel c, d), which can be rescued by the addition of medium conditioned in the presence of Epo (panel i, j). Note also, that the addition of conditioned medium (panel i, j) reverses the reduction in the number of nuclei (DAPI) observed after removal of the epicardium (panel c, d).

[19] FIG. 8 is a schematic drawing outlining the regulation of cardiomyocyte proliferation, showing that erythropoietin mediates retinoic acid- dependant proliferation of cardiomyocytes via secretion of a soluble factor. The epicardium secretes RA, which stimulates Epo production in the epicardium. Epo, via its receptor, stimulates the secretion of a soluble mitogen survival factor ("X") from the epicardium, which stimulates cardiomyocyte proliferation directly.

DETAILED DESCRIPTION OF THE INVENTION

[20] Heart slice culture system of the invention. As shown in FIG. 1 and FIG. 2, we have established a heart slice primary culture system that allows cardiomyocytes to continue to proliferate in culture. Previously, cardiomyocytes have generally neither proliferated nor survived in culture. However, the cardiac heart slice culture system of the invention allows the mechanical separation of distinct cardiac cell populations and the assay of relative mitogenic and trophic effects of the cardiac cell populations on cardiomyocyte proliferation and survival in culture. [21] Methods for making the heart slice culture system of the invention are provided in EXAMPLE I.

6

RST99 1231787-1.061039.0010 [22] The heart slice culture system of the invention is a dramatic improvement over previous cardiac primary cultures, h the heart slice culture system of the mvention, cardiomyocytes remain in a tissue context more closely resembling physiological conditions as compared to single cells in culture. Non- cardiomyocytes (which are mostly epicardial cells) migrate out of the slice and form a halo around the slice. Within the slice 99% of the cells are positive for cardiac markers even after several days of culture (See, EXAMPLE I, for example, titin and myosin heavy chain (MHC)). Very few cells in the heart slice culture system of the mvention (typically at the outermost layer of the cultured slice) are non-cardiomyocytes. Moreover, heart slices sustain cardiomyocyte proliferation for several days, which allows the assay of modulators of cell proliferation and survival.

[23] Furthermore, heart slices cultured after the removal of the epicardium do not proliferate and do not survive. This allows one to assay for substances that rescue proliferation and/or survival.

[24] Therapeutic utility of the heart slice culture system of the invention. Cell transplantation and tissue engineering are relatively new, pre- clinical, very promising approaches to help patients in end-stage organ failure. An important limitation of cell transplantation into damaged myocardium is the small amount of viable cells available for transplantation. For this reason, cell expansion is usually performed prior to transplantation. Cell expansion relies on proliferation capacity, which had previous to this invention been lacking in adult cardiomyocytes.

[25] Culturing of adult cardiomyocytes by the heart slice culture system of the invention or by methods of this invention permits an opportunity to restore function and blood flow to regions of the heart which were destroyed by acquired heart disease or absent because of congenital heart disease. For example, an increase in cardiomyocyte cell number in diseased hearts can greatly improve function and potentially prevent or at least delay heart failure. After heart infarction, the resulting scar in the heart can be removed, cardiomyocytes can be grown in culture and expanded in number, and the

R5T^QQ 1911787.1 (IfilmQ 11111(1 resulting cells can be transplanted back into the damaged heart to increase the number of myocardial cells.

[26] Cardiomyocytes can be transplanted into non-scar heart tissue by using the methods of U.S. Pat. No. 5,602,301; 3. Koh et al, Journal of Clinical Investigation 92:1548-54 (1993); Soonpaa et al, Science 264:98-101 (1994). In accordance with the method taught in U.S. Pat. No. 5,602,301, cellular grafts are introduced into the myocardium by injection.

[27] Cell transplantation can also be accomplished using a system to form channels in the heart wall and a means to deliver a therapeutic or diagnostic agent into the channels, as taught by U.S. Pat. No. 5,840,059. The system can be configured so that it can be introduced into the channels percutaneously or intraoperatively. The system generally comprises an ^' elongated, flexible lasing transmission catheter that emits laser radiation and has delivery lumen opening at the distal end. The system is used for forming channels in the heart wall and delivering a therapeutic agent (here, the transplanted cells) into the channel.

[28] Alternatively, forming a stable myocardial graft in a mammal, by transplanting cells into scar tissue in a heart, can be performed according to the methods of U.S. Pat. No. 6,110,459. Different from normal myocardial tissue, scar tissue in the heart has no cardiac muscle cells and is composed on connective tissue cells, such as fibroblasts, and non-cellular components, such as collagen and fibronectin. The scar tissue is formed after necrosing the ventricular wall of the heart. The mature scar tissue also a limited blood supply. [29] In accordance with the method taught in U.S. Pat. No. 6,110,459, cardiomyocytes can be successfully transplanted into the scar tissue formed after ventricular necrosis and into tissue membranes and porous synthetic membranes, h part of the method taught by U.S. Pat. No. 6,110,459, in a cell suspension (0.25 ml) is injected into the scar tissue of the heart using a tuberculin syringe. Cultured cells are transplanted into the center of a mature scar, even when there is no contact between the transplanted cells and the host

RKT99 1231787-1.061039.0010 cells. The cell grafts can form tissue that survives, improves myocardial function, and stimulates angiogenesis.

[30] Cultured cells can be employed to restore regional cardiac function and blood flow to regions in the heart damaged by acquired disease processes or absent because of congenital defects. During reconstructive surgery, the defective portion of the heart is removed and replaced with inert materials to secure a watertight seal. Cardiac surgeons frequently remove segments of the heart that have been damaged or are defective due to congenital abnormalities. A cellular graft permits an opportunity to restore function to these regions. The cells can be used to create a functioning myocardial graft that will replace the region removed at the time of surgery.

[31] The optimal time for transplantation is immediately after the acute inflammatory response to the myocardial injury has disappeared. Co- transplantation of growth factors such as IGF-I, IGF-II, TGF- 1 and PDGF-B may increase the survival of transplanted cells, induces transplanted muscle hypertrophy, and stimulates angiogenesis.

[32] For transplant, cultured cardiomyocytes can be seeded on the biological mesh, such as a collagen membrane, and non-biological membranes, such as non-degradable membranes (such as Dacron) or degradable membranes (such as polyglycolic acid polymers). See, U.S. Pat. No. 6,110,459. The mesh and the cells are cultured in the cell culture medium.

[33] Also for transplant, cultured cardiomyocytes can introduced using a cell gluing technique for cell transplantation. The product of such a process would be a patch that can have various clinical and therapeutic uses. Such membranes may be made from Dacron or biodegradable sheets such as polyglycolic acid polymers with or without polylactic acid polymers. Such a patch can be associated with a pacemaker and be implanted close to a cardiac defect thereby providing a means of paced cardiomyoplasty.

[34] Thrombin and cryoprecipitate (fibrin glue) derived from human blood clots quickly. This fibrin glue for cell transplantation, according to the methods taught is U.S. Pat. No. 6,110,459. The biological glue can be applied

9

R TQ 1231787-1 061039.0010 onto the injection site. The injection needle is withdrawn after the glue clots. Accordingly, leakage of transplanted cells is prevented.

[35] Following cardiomyocyte transplantation, heart function of the transplanted animals can be measured using a Langendorff preparation. Stephen et al, Annals of Thoracic Surgery 59:1127-1133 (1995). Measurement of ventricular remodeling can be quantified using computerized planimetry as described by Wu et al, Cardiovascular Research 27:736-39 (1993). The histology of the transplanted cells can be followed, for example, using stains for haematoxylin and eosin and antibodies to cardiac myosin heavy chain (Rougier Bio-Tech, Montreal, Canada) and other antibodies.

[36] Another use of the heart slice culture system of the invention includes transplanting blocks of tissue. Tissue engineering laboratories are pursuing the idea of transplanting a block of tissue into the infarcted heart after removal of the scar that had formed as a result of the infarction. The heart slice system of the invention can be used to obtain tissue blocks, rather than growing single cells in culture.

[37] Another use of the heart slice culture system of the invention includes the use of heart slices in the construction of an artificial heart. After injecting a reversibly polymerizable substance into the heart arteries, one can digest the tissue off and may obtain a "skeleton" of the blood vessels of the heart. This "skeleton" can then be seeded with endothelial cells to form new blood vessels. Once this vasculature of the heart is constructed, one would seed cardiomyocytes, for example the cardiac slices of the invention, to create a new organ in culture, which would then be transplanted into infarcted areas of patient hearts.

[38] Enriched population of epicardial cells (non-cardiomyocyte cells). We have observed that culture of ElO chick or E12-14 mouse slices gives the following phenotype after one to several days of culture. When an intact heart slice is cultured, cells migrate out of the heart slice, forming a halo around the slice. These cells do not stain for cardiomyocyte markers (see, EXAMPLE I) and are most likely epicardial cells (because at this stage of heart development,

10

RST99 1231787-1.061039.0010 fewer than 1% of the cells in the ventricular wall are fibroblasts). As shown in FIG. 1, the cells that are in the slice after one to several days in culture are 99% cardiomyocytes, as evidenced by the expression of cardiac markers, as well as beating. [39] When slices are cultured after removal of the epicardium, fewer cells migrate out of the slice forming the mentioned halo like ring around the slice. When slices are cultured after removal of the endocardium and trabeculae, cells still migrate out of the superficial layers of the slice, as was the case in the intact slice culture. This is evidence that the cells that migrated out of the slice are indeed coming from the epicardium, not from the endocardium.

[40] Conditioned medium of the invention. In addition to the heart slice culture system, the invention provides cultured medium, which contains a factor that directly regulates cardiomyocyte proliferation and survival and which is useful for the proliferation and survival of cardiomyocytes. In the heart slice- culture system of the invention, conditioned medium can promote both the proliferation and survival of cardiomyocytes that would otherwise neither proliferate nor survive in the absence of signals from the adjacent epicardium. See, FIG. 7. To increase the effectiveness of the conditioned medium, one can induce the secretion of factors from these epicardial cells, by the addition of Epo to the culture medium. Moreover, one can expand the population of epicardial cells that produces the mitogen to obtain large volumes of conditioned medium starting.

[41] The conditioned medium of the invention can be used to expand populations of proliferating cardiomyocytes derived from several sources. These sources include: (a) fetal cardiomyocytes (e.g. from pigs or other animals for xenotransplantation; see, EXAMPLE 1 for chick and mouse); (b) embryonic stem cells; (c) other stem cells (e.g. bone marrow); (d) young (proliferating) human cardiomyocytes from people that suffered fatal accidents; (e) adult cardiomyocytes that have been made to reenter the cell cycle. The relevance of adult cardiomyocytes is that upon heart surgery, a small part of the heart muscle

11

BST99 1231787-1.061039.0010 is usually removed. These cells could be expanded in culture before reimplantation, thus avoiding transplant rejection.

[42] Human recombinant Epo has been used in all assays with ElO chicken heart slices in EXAMPLE I. Epo conditioned medium from epicardial cells has been obtained from both chicken and from mouse epicardial cells.

Both were active in chicken cardiac slice cultures (with or without prior removal of the epicardium). Therefore, medium conditioned by chicken epicardial cells is active for human cardiomyocytes. This is useful, because chicken epicardial cells grow much faster than mouse epicardial cells, allowing great expansion of these cells, which enables us to obtain large volumes of conditioned medium starting from a very limited amount of starting material.

[43] Also, the conditioned medium of the invention could be used for cell transplantation and tissue engineering. As described above, cell transplantation and tissue engineering are relatively new, pre-clinical, very promising approaches to help patients in end-stage organ failure or after heart infarction. Infusion or local application of the conditioned medium can enhance the survival of cardiomyocytes in situ after infarction.

[44] Furthermore, the conditioned medium can be used to enhance the survival of mitogen signal-responsive cell types (e.g. neurons) during or after degenerative diseases (e.g. neurodegenerative diseases).

The effects of conditioned medium can be tested in an animal model of heart failure. Cardiotoxic agents, including doxorubicin, have been used in several species including rat and dog to induce heart failure. For an analysis of these models, see Czarnecki, Comparative Biochemistry and Physiology 79C: 9- 14 (1984); Smith & Nutall, Cardiovascular Research 19: 181-186 (1985).

Several experimental procedures have been utilized to effect coronary artery occlusion, myocardial ischemia and resultant heart failure primarily in rats and dogs. These procedures include direct coronary ligation, embolism with liquid mercury, injection of preformed thrombus, wedged catheters, and sequential coronary microembolization with microspheres. Khomaziuk et al, Kardiologiya

5: 19-23 (1965); Rees & Redding, Cardiovascular Research 2: 43-53, (1968);

12

R<5T^QQ 1911787-1 061039 0010 Lumicao et al, American Journal of Medical Science 261: 27-40 (1971); Millner et al, Annals of Thoracic Surgery 52: 78-83, 1991; Sabbah et al, American Journal of Physiology 260: H1379-H1384 (1991).

An assessment of whether the conditioned medium is therapeutically useful for the treatment of heart failure can also be performed in an animal with simulated heart failure, including congestive heart failure. See, U.S. Pat. No. 6,158,438, which teaches a chronically instrumented conscious pig with heart failure wherein the heart failure is induced by myocardial ischemia and intermittent rapid cardiac pacing and provides a method for assessing the effects of the test compound on cardiac function and systemic vascular function, h addition, the effects of the conditioned medium can be tested in a pig model of myocardial ischemia that mimics clinical coronary artery disease. See, U.S. Pat. No. 6,174,871; Roth et al, Circulation 82:1778 (1990); Roth et al, Am. J. Physiol. 235:H1279 (1987); White et al, Circ. Res. 71:1490 (1992); Hammond et al, Cardiol 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 (referred to as the ischemic region), due to collateral vessel development, but blood flow reserve is insufficient to prevent ischemia when myocardial oxygen demands increase. The pig 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 and functioning region (the LAD bed) adjacent to an abnormally perfused and ftmctioning region (the LCx bed), thereby offering a control bed within each animal. [45] Bioassay of the invention. The heart slice culture system (intact slice or after removal of the epicardium or after application of Ro415253) also provides a screening assay: The bioassay of the invention usefully and advantageously provides a dramatic improvement over existing bioassays for cardiomyocytes. The bioassay of the invention can be used to screen substances for the following activities: (1) increase, induce or decrease cardiomyocyte proliferation, (2) increase, induce or decrease cardiomyocyte survival, (3)

13

BST99 1231787-1.061039.0010 increase, induce or decrease cardiomyocyte apoptosis, (4) increase, induce or decrease cardiomyocyte necrosis or other cell or nuclear degradation.

[46] Any substance or medication (e.g. combination of drugs) can be rapidly screened with the bioassay of the invention to test for the ability to promote cardiomyocyte proliferation or survival. Of particular interest are the existing drugs and medications for the treating heart diseases. Given the acknowledged importance of heart diseases to society, we expect that many of these substances and medications would easily and rapidly be screened for activity. Substances and medications found to be positive in the bioassay of the invention may generally be useful to modulate cell proliferation, survival, apoptosis, necrosis or other cell or nuclear degradation of cardiomyocytes. These substances and medications might therefore be useful for other therapeutic applications, specifically: anti- cancer treatments, treatment of cytostatica, anti-neurodegenerative treatment, regeneration of tissues, and treatment of rare heart cancers.

[47] The biological basis for the bioassay of the invention comes from the developmental events that occur during cardiomyocyte cell proliferation. Cardiomyocytes proliferate until around birth in mouse and rat or hatching in chick. The major proliferation-driven increase in thickness of the ventricular wall of the heart takes place from El 1.5 until E14.5 in the mouse, and from E8-

E14, in the chick. The maintained proliferation of cardiomyocytes at these embryonic stages results in the expansion of the ventricular wall. The region of the heart that shows the highest rate of proliferation lies adjacent to the epicardium and is teπned the compact zone. Postnatally, the further increase in the size of the heart is generally attributed to hypertrophy of cardiomyocytes.

[48] The precise mechanisms that regulate cardiomyocyte cell proliferation are largely unknown. By analyzing the clonal progeny of chick heart cells infected with viruses encoding dominant-negative FGF-receptors, Mima et al, Dev. Biol 167(2): 617-20 (Feb. 1995) demonstrated an FGF- dependent, early phase and an FGF-independent, late phase for cardiomyocyte proliferation. Several knockout mice have been described that show a

14

BST99 1231787-1.061039.0010 hypoplastic phenotype associated with a thin ventricular wall, or reduced trabeculation, suggesting a possible role in cardiomyocyte proliferation for such diverse molecules as: RAR , RXR , VCAM-1, 4-integrin, erythropoietin, erythropoietin receptor, neuregulin, erbB2, erbB4, N-rnyc, TEF-1, WT-1, gp 130, Jak2, β-ARK, p300, Pax-3 and FOG-2. In the case of VCAM-1, α4- integrin or FOG-2 deficient animals, the interaction between the epicardium and the myocardium is disrupted. Thus, communication between these tissues may affect expansion of the ventricular wall. Interestingly the hypoplastic phenotype for RXR , and gp 130-defϊcient animals is non-cell-autonomous with respect to the cardiomyocyte lineage, suggesting that these signaling molecules may be required in cells other than cardiomyocytes to promote inductive interactions between cardiomyocytes and non-myocyte tissue layers. Indeed several of the above mentioned gene products are primarily expressed within the epicardium ( 4-integrin, WT-1, FOG-2) or expressed in the endocardium (neuregulin), or in both these tissues (erythropoietin receptor), but are not expressed in the myocardium of the heart, although its deficiency results in absence of compact zone expansion or reduced trabeculation within the ventricular myocardium. However, none of the affected gene products in these mutant mice have been implicated directly in cardiomyocyte proliferation per se. [49] Thus, this invention provides solutions to some of the problems previous encountered in cardiomyocyte culture. Among the solutions are the following:

[50] Signals from the epicardium are required for cellular proliferation in the myocardium. The involvement of the epicardium in modulating cardiomyocyte proliferation has been suggested by previous studies using several knockout mice lacking either V-cam, -integrin, erythropoietin, or the erythropoietin receptor, h these genetically altered mice, there was a correlation between either a loss (V-cam, -integrin) or at least a partial detachment (erythropoietin, erythropoietin receptor) of the epicardium from the myocardium and the subsequent development of a hypoplastic ventricular wall around mid-gestation. However, it was not clear from these studies whether the

15

RSTQQ 1931787-1 061039.0010 epicardium directly regulated cardiomyocyte proliferation, or if the ensuing cardiomyopathy resulted from a disturbance of cardiac vascularization that originates from epicardium or from an upstream obstruction of blood flow. [51] Using the heart slice culture system of the invention, we show that signals from the epicardium are required for cellular proliferation in the myocardium. However, removal of the endocardium and trabeculae from the heart slice cultures of the invention does not affect myocardial cell proliferation.

[52] Epicardial retinoic acid (RA) is apleiotropic regulator of cardiomyocyte differentiation, survival and proliferation. Recently, the epicardium had been recognized to be the principal source of RA in the developing heart, because RALDH2, (retinaldehyde dehydrogenase 2, a retinaldehyde dehydrogenase that generates retinoic acid, and the rate-limiting enzyme in RA synthesis), is highly expressed in the epicardium, but not in cardiomyocytes. Moss et al, Dev. Biol. 199(1): 55-71 (My 1, 1998). Although RA is produced in the epicardium, expression of RA-responsive reporter constructs in transgenic mice has indicated that both epicardial and myocardial cells in the heart respond to RA signaling during development. Moss et al, Dev. Biol. 199(1): 55-71 (July 1, 1998).

[53] Using the heart slice culture system of the invention, we show that while RA signaling in the epicardium is necessary for cardiomyocyte proliferation, this is not because of a direct mitogenic effect of epicardial RA on cardiomyocytes. Although the RA-antagonist, Ro415253, dramatically reduced proliferation and subsequent survival of cardiomyocytes in slice culture in the presence of intact epicardium, RA by itself could not substitute for the epicardium. Thus, RA secreted by the epicardium either works in concert with another epicardial factor to promote myocardial cell proliferation, or RA induces the synthesis of a distinct cardiomyocyte mitogen within the epicardium itself. Consistent with this latter interpretation are the recent findings by Chen et al, Genes Dev. 8(19): 2293-2301 (1998) that mice with a cardiomyocyte- specific deletion of RXR fail to develop hypoplastic hearts. The same conclusion has been reached by the analysis of chimeric mice where

16

BST99 1231787-1.061039.0010 cardiomyocytes deficient in RXR develop normally and contribute to the ventricular chamber wall. Tran & Sucov, Development 125(10): 1951-6 (1998). [54] hi addition to indirectly modulating cardiomyocyte proliferation and survival, RA may directly affect cardiomyocyte differentiation. Consistent with this idea, we have observed that the expression of cardiomyocyte structural genes is down-regulated following treatment of heart slice cultures with Ro415253. Thus, RA signaling may indirectly control myocardial cell proliferation and affect cardiomyocyte differentiation.

[55] Epicardial Epo is a downstream target ofRA that mediates the regulation of cardiomyocyte proliferation. Epo is highly expressed in the epicardium and is not detectable by in situ hybridization in the myocardium of the developing heart. Wu et al, Development 126(16): 3597-605 (1999).

[56] We have found that RA stimulates Epo expression in epicardial cells, showing that some of the effects of RA signaling in the epicardium may be mediated by Epo. Consistent with this idea, we have found that exogenous

Epo rescued the inhibition of cardiomyocyte proliferation following administration of the RA-antagonist, Ro415253, to heart slice cultures. Thus, RA signals in the epicardium may promote myocardial growth in an Epo- dependent pathway. The observed regulation of Epo expression by RA signals in the epicardium is consistent with prior observations that the Epo promoter contains a steroid hormone-binding site capable of binding RAR or RXR.

[57] Epo signaling in the epicardium induces the synthesis or expression of a cardiomyocyte mitogen in epicardial cells. Genetically engineered mice that lack either Epo or the Epo-receptor (Epo-R) have recently been shown to develop a hypoplastic heart phenotype with a detached epicardium at midgestation. Wu et al, Development 126(16): 3597-605 (1999). Interestingly, the Epo-receptor is expressed in the epicardium, but not in cardiomyocytes, suggesting that Epo signaling in cells other than cardiomyocytes are necessary for normal cardiac growth. [58] In heart slice culture system of the invention, we have found that administration of Epo failed to rescue cardiomyocyte proliferation following

17

BST99 1231787-1.061039.0010 removal of the epicardium. Thus, Epo does not directly induce myocardial cell proliferation. By contrast, conditioned medium from Epo-treated epicardial cells, but not from non-Epo-treated epicardial cells, was able to rescue myocardial proliferation in heart slices that lacked the epicardium. Thus, the observed hypoplastic heart phenotype in Epo-/- and Epo-receptor-/- mice may be due to the requirement for Epo signaling in the epicardium to promote the expression or activity of a cardiomyocyte mitogen in the epicardial cells themselves.

[59] Summary. Our results show that the control of myocardial growth is regulated by the sequential activity of both RA and Epo signaling pathways in the epicardium, which in turn controls the expression or activity of a myocardial mitogen.

[60] The epicardium expresses RALDH2, a principle regulator of RA synthesis. Moss et al, Dev. Biol. 199(1): 55-71 (July 1, 1998). RA probably regulates multiple aspects of cardiomyocyte development, including differentiation, proliferation and survival, in a coordinate fashion. However, RA regulation of cardiomyocyte proliferation occurs indirectly via expression of Epo and its downstream targets in the epicardium. As we show, Epo in turn stimulates the secretion or activity of a soluble factor that is able to promote cardiomyocyte proliferation directly. That the control of myocardial cell proliferation is mediated by such a complex chain of molecular intermediaries is perhaps not surprising, in view of the pleiotropic effects of both RA and Epo signaling, and the need for both spatially and temporally precise control of proliferation in the developing heart. It is known that the Epo-receptor is expressed in the epicardium only from E10.5 to E13.5 (Wu et al, Development

126(16): 3597-605 (1999)), coincident with the relatively massive growth of the ventricular wall from El 1.5 to E14.5 in mouse. By contrast, the overall responsiveness to RA signaling in the heart, as judged by expression of RA- stimulated reporter transgenes, is maintained for a significantly longer period of time. The utilization of an erythropoietic growth and differentiation factor for regulating heart size is potentially intriguing from a physiologic and perhaps an

18

BST99 1231787-1.061039.0010 evolutionary standpoint. It might be advantageous to modulate the heart size of an animal according to the oxygenation status during fetal development. As it is known that Epo in the liver is stimulated by Hypoxia Inducible Factor, HIF-1 (Ebert & Bunn, Blood 94(6): 1864-77 (Sept. 15, 1999)), it is possible that a systemic increase in Epo levels, following oxygen deprivation during fetal development, could both increase rates of erythropoiesis and also increase myocardial cell proliferation. Both outcomes of Epo signaling would act to improve the oxygenation status of the developing tissues, by increased erythropoiesis and by increased tissue perfusion, respectively.

[61] The following EXAMPLES are presented in order to more fully illustrate the preferred embodiments of the invention. These examples should in no way be construed as limiting the scope of the invention, as defined by the appended claims.

EXAMPLE I

HEART SLICE CULTURE

[62] Introduction. To directly study the influence of the epicardium or endocardium on cardiomyocyte proliferation, we developed a heart slice culture system that allows the mechanical separation of the epicardium or endocardium/trabeculae from the ventricular wall. The epicardium (but not the endocardium) is required for cardiomyocyte proliferation. We also investigated whether retinoic acid or erythropoietin signaling contributes to the mitogenic action of the epicardium on ventricular myocytes. Administration of the RAR antagonist, Ro-415253, to heart slice cultures mimics the effect of removal of the epicardium and blocks cardiomyocyte proliferation. Furthermore, inhibition of cardiomyocyte proliferation by Ro-415253 can be overcome by the addition of erythropoietin (Epo), which is expressed in the epicardium and which is inducible by RA signaling. Finally, Epo induces cardiomyocyte proliferation by regulating the expression of yet another soluble factors in epicardial cells. Together, the findings of this EXAMPLE support a model in which a cascade of

19

BST99 1231787-1.061039.0010 RA and Epo signals in the epicardium induce the expression (or activity) of a mitogen that directly regulates cardiomyocyte proliferation.

[63] Cardiomyocytes proliferate in slice-cultures from young hearts. To monitor the effects of the epicardium or the endocardium on proliferation and survival of cells within the myocardium, we established an organ culture system (an embodiment of the heart slice culture system of the invention) for ElO chick or El 3 mouse hearts cultured in vitro. Hearts from these stage embryos were excised, the posterior half removed, embedded in agarose, 100 to 150 micron transverse slices were made with a vibratome, and slices were cultured.

[64] Embryonic heart tissue was obtained as follows: M159 fertilized eggs were incubated in a humidified egg-incubator at 37oC. The day that the incubation started was defined as El . After 10 or 11 days of incubation, the chicken embryos were harvested and staged according to Hamburger & Hamilton, J. Morphology, 88: 49-92 (1951) (see also, Sanes, Developmental

Dynamics 195: 229-275 (1993)). Whole hearts were dissected on ice in PBS + Penicillin/Streptomycin. Mouse tissue was similarly harvested from Swiss Webster mice at embryonic day 12.5 or 13.5 (E12.5; E13.5). The noon following the morning when the vaginal plug was detected was defined as E0.5. Embryos were staged according to Theiler, The House Mouse: Atlas of Mouse

Development (Springer- Verlag, New York, 1989) by overall size and particularly by limb and digit development.

[65] Slice culture was performed as follows: Prior to the dissection, 5- 10% low melting agarose (Fisher) was dissolved in PBS at 65°C for 30 min. Meanwhile chick or mouse whole hearts were cut sagitally in order to remove the dorsal part of the heart (chick) or to remove the lowermost third of the ventricles (mouse). Remaining blood cells were carefully rinsed from the lumen of the ventricles, and the hearts briefly stored in PBS on ice. The hearts were positioned in Peel-A-Way plastic molds (VWR) in agarose initially cooled to room temperature, then rapidly chilled and solidified on ice. Blocks of agarose containing embedded hearts were cut out and fixed onto a specimen holder. The

20

RST^QQ 1911787-1 Λ1 39 001 specimen holder was then cooled on ice and sections subsequently cut in ice- cold PBS + penicillin/streptomycin with a vibratome (Leica VT1000M) (speed: 1.5 units; frequency: 2.0 units). The use of non-undulating razor blades (Dickinson, NJ) was critical in obtaining thin slices. We regularly cut 150 μm slices because the number and quality of the slices dropped with thinner sections. However, it was possible to obtain living slices that were 50 μm thin. Floating slices were then collected and typically cultured on a Petri dish in a drop of at least 100 1 of medium, DMEM, 10% FBS, 5 mM pyruvate, at 37°C, 5% CO₂. 10 M BrdU was added to the culture medium at the indicated time points.

[66] A schematic of such a heart slice is depicted in FIG. 1 A. A subpopulation of cells migrated from the slices and formed a halo-like structure on the periphery of the slice. These migratory cells are not immunoreactive for the cardiomyocyte markers titin and myosin heavy chain (MHC). The migratory cells thus represent a cell population derived from non-muscle tissue. By contrast, over 99% of the non-migratory cells remaining in the slice were titin and MHC positive (FIG. IB). BrdU incorporation over 4 days of culture indicated that the cardiomyocytes continued to proliferate during the entire culture period (FIG. 1C and FIG. ID). [67] Light microscopic immunocytochemistry was performed as follows: Hearts were immersed in fixative containing 4% paraformaldehyde, 4% sucrose, 2 mM CaC12, 200 mM Hepes-NaOH pH 7.4 or 4% paraformaldehyde in PBS pH 7.4 for 1 hour. The fixed tissue was infiltrated with sucrose, frozen in Tissue-tek, and 5-7 m cryosections were prepared and processed as described by Aaku-Saraste et al, Mech. Dev. 69(1-2): 71-81 (Dec. 1997), except that the sections were collected on Superfrost-Plus slides rather than on gelatin- coated slides, and the blocking medium contained 3% fetal calf serum instead of 0.05% BSA. In the case of anti-BrdU immunofluorescence cryosections were then processed as described by Novitch et al, J. Cell Biol 135(2): 441-56 (Oct. 1996). The primary antibodies used and their dilution/concentration were as follows: Mouse rnAb MF-20 against sarcomeric myosine, ascites supernatant

21

BST99 1231787-1.061039.0010 1 :10; mouse mAb G3G4 against BrdU, hybridoma supernatant 1:200. The secondary antibodies used and their dilution/concentration were as follows: TRITC-conjugated goat anti-mouse 1:200, Cy-2 or Cy-3-conjugated goat anti- mouse 1 :200 (Jackson Laboratory, ME). Sections were mounted with Moviol and observed directly with a Zeiss Axioskop microscope. Pictures were taken with a Toshiba 3CCD camera using Phase 3 Imaging software (see, <http//www.p3i.com>).

[68] The epicardium, but neither the trabeculae nor endocardium, is required for cardiomyocyte proliferation and survival. To examine the role of the epicardium, trabeculae, and endocardium on proliferation and survival of the myocardium, we dissected ElO chick heart slices and cultured the myocardium either with or without these adjacent tissues. As schematically outlined in FIG. 1 A, epicardium or endocardium and trabeculae were mechanically dissected from the heart slice, which was then cultured for several days and assayed for proliferation. Removal of the epicardium dramatically reduced the apparent number of non-cardiac "halo" forming cells that migrated out of intact slices. This effect was not seen following removal of endocardium and trabeculae. Thus, the migratory cells were epicardial in origin. Removal of the epicardium resulted in a dramatic loss of BrdU incorporation in the myocardium between the second and third days of culture while the number and appearance of the nuclei was comparable to that in intact control slices. (FIG. 2A, compare panels a, c and e, g, FIG. 2B and FIG. 2C). After four days culture, removal of the epicardium resulted in a dramatic reduction in the number of D API-staining nuclei (in many cases a decrease of greater than 90%), as well as evidence of nuclear degradation (FIG. 2A, panels d and h, FIG. 2C). By contrast, the proliferation or survival of cardiomyocytes was not significantly affected by removal of the endocardium and trabeculae (FIG. 2A, panels b, d and i, j, FIG. 2B and FIG. 2C). Thus, signals from epicardial cells are required for cardiomyocyte proliferation as well as for survival. [69] RA is required for cardiomyocyte proliferation and survival. To identify molecules that might be required for cardiomyocyte proliferation and

22

BST99 1231787-1.061039.0010 survival, we examined the effect of secreted molecules expressed in the epicardium. We first tested the involvement of RA, which has recently been shown to be synthesized in the epicardium (Moss et al, Dev. Biol. 199(1): 55- 71 (July 1, 1998)) and to modulate ventricular cell proliferation in a non-cell autonomous fashion (Tran & Sucov, Development 125(10): 1951-6 (1998);

Chen et al, Genes Dev. 8(19): 2293-2301 (1998)). Intact heart slices were cultured with or without the RA antagonist, Ro415253, and assayed for proliferation. Three days following administration of the RA antagonist, incorporation of BrdU into cardiomyocytes was dramatically inhibited while the number and appearance of the nuclei was comparable to that in control slices

(FIG. 3 A, panels a, c and e, g, FIG. 3B and FIG. 3C). After 5 days of culture, addition of Ro415253 resulted in both a dramatic reduction in the number of nuclei and in nuclear degradation (Figure 3 A, panels b, d, and f, h, FIG. 3B and FIG. 3C). Thus, administration of the RA antagonist, Ro415253, resulted in a phenotype similar to that following removal of the epicardium.

[70] Exogenous RA does not overcome the inhibition of cardiomyocyte proliferation following removal of the epicardium. Cardiomyocytes respond to a RA signal, since a RA-stimulated reporter transgene is strongly expressed in the myocardium at El 2.5 and onwards in the mouse. Moreover this same group showed that cardiac expression of the RA biosynthetic enzyme, RALDH2 predominantly occurs in the epicardium from El 1.5 onwards. Thus, the epicardium may be the relevant source for RA in the embryonic heart. To investigate whether RA secreted from the epicardium might directly regulate cardiomyocyte proliferation, we tested whether exogenous RA rescues proliferation of myocardial cells in heart slices cultured in the absence of the epicardium (FIG. 4). Administration of exogenous RA failed to rescue cardiomyocyte proliferation in heart slices cultured in the absence of the epicardium (FIG. 4A, panels e, g and f, h; and FIG. 4B). Because RA cannot substitute for the epicardial signal that drives myocardial cell proliferation, yet RA signaling is apparently required for such mitogenic activity

(FIG. 3), it raised the possibility that RA might act on the epicardium itself to

23

BST99 1231787-1.061039.0010 induce the expression (or activity) of a myocardial mitogen. Indeed this scenario would be consistent with prior work demonstrating that RXR functions in a non-cell-autonomous fashion to modulate myocardial cell proliferation (Tran & Sucov, Development 125(10): 1951-6 (1998); Chen et a , Genes Dev. 8(19): 2293-2301 (1998)).

[71] Epo expression is modulated by RA signaling in the epicardium. Because erythropoietin (Epo) is highly expressed in the epicardium, but not in the cardiac wall, and because mice lacking Epo display a hypoplastic ventricle that is reminiscent of that seen in RXR knockout mice, we tested whether RA signaling regulates Epo expression in heart explants. To investigate whether

Epo is regulated by RA signaling in epicardial cells, we cultured epicardial cells of El 3 mouse hearts or heart slices either with or without RA and subsequently performed RT-PCR to assay Epo expression. Epo was indeed transcriptionally up-regulated in cultures treated with RA as compared to control cultures (FIG. 5 A).

[72] RT-PCR was performed as follows: Preparation of cDNA and analysis by PCR are essentially as described previously by Schultheiss et al, Development 121(12): 4203-14 (Dec. 1995), except that PCR was carried out in a volume of 25 1 instead of 50 1. The primers and conditions used were as follows, GAPDH (mouse) 5 ' primer TGC GAC TTC AAC AGC AAC TC,

3 'primer GAT GGA AAT TGT GAG GGA GA, Epo 5'primer CGG GCA CCT GCA TAG ATT CAC, Epo 3 'primer AGA AGG TAA TGA CCC GTT TGA.^'

[73] Ectopic Epo administration rescues myocardial cell proliferation in heart cultures treated with the RA antagonist, Ro415253. Because Epo can be induced by RA signals in epicardial cells, it raised the possibility that treatment of heart slice cultures with the RA antagonist, Ro415253, may indirectly block myocardial cell proliferation by modulating the expression of Epo. To investigate this possibility, we cultured ElO chick heart slices with Ro415253 in either the absence or presence of exogenous Epo, and assayed cellular proliferation by BrdU incorporation. Whereas cultures treated with Ro415253 failed to sustain myocardial DNA synthesis (FIG. 5B, panel c and FIG. 5C),

24

BST99 1231787-1.061039.0010 those treated with Ro415253 plus Epo maintained robust myocardial DNA synthesis (FIG. 5B, panel e, and FIG. 5C). In addition, Epo also improved the survival of cardiomyocytes following Ro415253 administration, as the number of nuclei in cultures treated with the combination of Epo plus Ro415253 was not reduced as it was following Ro415253 treatment alone (FIG. 5B, compare panels b, d and f).

[74] Epo is an indirect regulator of cardiomyocyte proliferation. Our results show that epicardial Epo expression, which is regulated by RA signaling, can mediate the proliferative effects of RA on cardiomyocytes. To investigate whether Epo could promote myocardial cell proliferation in the absence of the epicardium, we tested whether administration of exogenous Epo could rescue myocardial proliferation in cultured heart slices after removal of the epicardium. Administration of exogenous Epo, however, failed to promote myocardial cell proliferation in heart slices lacking the epicardium (FIG. 6A, panels a-f, FIG. 6B). The results of this EXAMPLE show that Epo signaling in myocardial cells

(devoid of the adjacent epicardium) is not sufficient for myocardial cell proliferation.

[75] Epo induces the synthesis of a cardiomyocyte mitogen in epicardial cells. While Epo administration is capable of rescuing myocardial cell proliferation in intact heart slices treated with a retinoic acid antagonist

(FIG. 5A, panels a-f), Epo administration cannot drive myocardial cell proliferation in the absence of the epicardium (FIG. 6A, B). This finding showed that Epo modulates cardiomyocyte proliferation by inducing the synthesis or activity of a cardiomyocyte mitogen within the epicardium itself. Accordingly, we prepared conditioned medium from epicardial cells that had been cultured either in the absence or presence of Epo and tested whether such conditioned medium could restore myocardial cell proliferation in heart slices devoid of epicardium.

[76] Conditioned medium was prepared as follows: ElO chick heart slices were cultured in 200 1 of medium, DMEM, 10 % FBS, 5 mM pyruvate, at 37°C, 5% CO₂. After one week the slice consisting of cardiomyocytes was

25

RST99 1231787-1 061039 0010 removed carefully so that the non-cardiomyocyte cells that had migrated out of the slice remained attached to the culture dish. Epo (2 U/ml) were added to the medium and the supernatant was collected every 2-5 days.

[77] Conditioned medium made from untreated epicardial cultures failed to restore myocardial cell proliferation in heart slices lacking epicardium

(FIG. 7 A, panels a-h, FIG. 7B). In contrast, conditioned medium made from Epo treated epicardial cultures induced robust myocardial cell proliferation in these same cultures (FIG. 7A, panels i and j, FIG. 7B). Administration of Epo in parallel with conditioned medium from non-Epo treated epicardium failed to induce proliferation in myocardial cultures (see FIG. 7 A, panels g and h, FIG.

7B). Thus, the mitogenic effects of Epo on the myocardium are mediated via the epicardial cells. Epo induces either the synthesis or activity of a soluble cardiomyocyte mitogen within epicardial cells. The findings of this EXAMPLE are consistent with the observation that the cardiac-specific expression of the Epo receptor is most predominant in the epicardium. Wu et al, Development

126(16): 3597-605 (1999).

[78] Summary. By employing the heart slice culture system of the invention, we have demonstrated in this EXAMPLE that signals from the epicardium are necessary for myocardial cell proliferation. One such signal is apparently retinoic acid (RA), which is produced by the epicardium (Moss et al,

Dev. Biol. 199(1): 55-71 (July 1, 1998)). Administration of the RA-antagonist, Ro415253, to intact heart slices can block myocardial cell proliferation.

[79] RA secreted from the epicardium may be a general regulator of cardiomyocyte proliferation in young hearts. Interestingly however, administration of RA to heart slices devoid of the epicardium failed to promote myocardial cell proliferation. Thus, RA cannot mimic the effects of the epicardium on cardiomyocyte proliferation. This EXAMPLE shows that RA signaling can induce the expression of Epo in the epicardium and that Epo signaling in turn induces either the expression or activity of a cardiomyocyte mitogen in epicardial cells that directly regulates cardiomyocyte proliferation.

Thus, RA acts in an autocrine or paracrine fashion in epicardial cells to regulate

26

RST99 1231787-1.061039.0010 epicardial expression of Epo. Epo then acts via its own receptor, which is expressed only in epicardial and not myocardial cells (Wu et al, Development 126(16): 3597-605 (1999)) to regulate a more specific, direct regulator of cardiomyocyte proliferation. This signaling cascade is schematically outlined in FIG. 8.

EXAMPLE H

PROTOCOL FOR PREPARING CONDITIONED MEDIUM

[80] This EXAMPLE provides protocols for producing conditioned medium that contains a soluble factor that that promotes cardiomyocyte proliferation and/or survival (Factor X). Factor X can be a protein or a biomolecule other than a protein.

[81] For small volumes, conditioned medium was prepared as follows: ElO chick heart or mouse slices were cultured in 200 1 of medium, DMEM, 10% FBS, 5 mM pyruvate, at 37°C, 5% CO₂. After one week the slice consisting of cardiomyocytes was removed carefully so that the non- cardiomyocyte cells that had migrated out of the slice remained attached to the culture dish. Epo (2 U/ml) were added to the conditioned medium and the supernatant was collected every 2-5 days.

[82] For large volumes, conditioned medium was prepared as follows: ElO chick or El 2- 13 mouse hearts were dissected and cut into small pieces. The pieces were digested in solution D for 10 min at 37°C. Solution D contains solution B (Hank's salts, taurine, creatine, Mg, insulin 10 g/ml; 1:500 from 5 mg/ml stock, pH 7.25) and collagenase type II 500 1 stock /10 ml. Collagenase stock is 10 mg/ml = 200 U/mg; 60 mg (=12,000 U) were dissolved in 6 ml. Trypsin is 0.01%; 2.5% stock diluted 1:250. Hyaluronidase is 0.1%, 1 mg = 350

U (Sigma)

[83] After 5 and 10 minutes the tissue pieces were triturated with a 1000 1 pipette first and then with a 200 1 pipette (pipette gently up and down a few times). The cell mixture was then centrifuged for 3 min at 2000 rpm (in a benchtop Eppendorf centrifuge). Cells were resuspended in 1 ml of medium

27

BST99 1231787-1.061039.0010 containing DMEM, 10% FBS, penicillin, and streptomycin and then grown in 6 cm dishes.

[84] To resuspend, take two ElO chick hearts for 1 x 6 cm dish. Let cells attach and grow overnight. After 1 day the medium is exchanged, and after 2-3 days cells reach confluence. At this point, tissue chunks (mostly containing cardiomyocytes) can optionally be removed. The cells are trypsinized and split 1:3 or 1:4.

[85] Starting with 1 Ox ElO chick hearts, within 2 weeks 100 10 cm dishes of cells can be obtained. [86] At the desired day, Epo (2 U/ml) is added to a confluent plate of cells and the medium is harvested after 2 days. The Epo-conditioned medium can then be tested in slice culture.

[87] After the expansion phase, cells can be grown in serum-free conditioned medium, to obtain conditioned medium. When confluent, cells can be grown in Hybridoma-SFM (Gibco, Cat. No. 12045-084) or PFHM-H (Gibco or Leibovitz). Cells can also be grown in Hybridoma-SFM, but grow more slowly.

EXAMPLE

IDENTIFICATION OF CARDIOMYOCYTE MITOGEN SECRETED BY EPICARDIAL CELLS

[88] The use of the heart slice culture system of the invention has shown the existence of a cardiomyocyte mitogen. The cardiomyocyte mitogen secreted by epicardial cells can be identified by employing two complementary methods: (1) biochemical purification of mitogenic factors from medium conditioned by epicardial cells; and (2) subfractive hybridization of cDNA from epicardial cells that can be stimulated to secrete the cardiomyocyte mitogen (erythropoietin-stimulated vs. control cultured epicardial cells).

[89] Firstly, the mitogen is characterized and purified biochemically in order to obtain a micro/nanosequence. Epicardial cells are cultured under serum-free conditions in the presence or absence of erythropoietin. The conditioned medium is harvested and the population of the constituent proteins

28

BST99 1231787-1.061039.0010 compared by two-dimensional gel electrophoresis. Analysis focuses on protein species unique to erythropoietin-induced cultures. Identification of candidate protein spots confers information with regard to molecular weight as well as isoelectric point. This information is then used to focus purification of the mitogenic activity from the medium by gel filtration and ion-exchange chromatography. The resulting fractions are assayed for biological activity using the heart slice-culture assay described above. Depending on the complexity of the purified protein fraction, it should be possible to obtain a sufficient amount of cardiomyocyte mitogen from a mass culture of epicardial cells that have been grown in the presence of erythropoietin, to obtain a candidate protein. The candidate protein is then purified by one or two-dimensional PAGE for micro/nanosequencing.

[90] Secondly, the erythropoietin-induced mitogen is cloned by subfractive hybridization of cDNAs from naϊve versus erythropoietin-treated epicardial cells. To identify erythropoietin-induced transcripts, tracer cDNA is generated from epicardial cultures treated with erythropoietin and biotinylated driver cDNA from naϊve epicardial cells that have not been treated with erythropoietin. After several cycles of annealing the tracer cDNA with the biotinylated driver cDNA, erythropoietin-induced transcripts are significantly enriched. After successful enrichment, which is monitored with slot-blots, a library is constructed from the subtracted sequences. Clones are sequenced and analyzed for differential expression in either naϊve or erythropoietin treated epicardial cell cultures. Candidate proteins that are induced by erythropoietin are expressed in COS cells and supernatant from such transfected cells are assayed for the ability to reuse cardiomyocyte proliferation in the absence of the adjacent epicardium.

[91] The details of one or more embodiments of the invention are set forth in the accompanying description above. Although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of the invention, the preferred methods and materials are now

29

BST99 1231787-1.061039.0010 described. Other features, objects, and advantages of the invention will be apparent from the description and from the claims, h the specification and the appended claims, the singular forms include plural referents unless the context clearly dictates otherwise. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. All patents and publications cited in this specification are incorporated by reference.

[92] The foregoing description has been presented only for the purposes of illustration and is not intended to limit the invention to the precise form disclosed, but by the claims appended hereto.

30

RST99 1931787-1 061039.0010

Claims

CLAIMS WE CLAIM:

1. A heart slice primary culture system, comprising: a thin slice of a vertebrate heart in a culture medium, wherein the heart slice is enriched for cardiomyocytes.

2. The system of claim 1 , wherein the epicardium has been removed from the coronal heart slice.

3. The system of claim 1 , wherein both the epicardium and the trabeculae have been removed from the coronal heart slice.

4. The system of claim 1 , wherein the coronal heart slice has been treated with an application of Ro415253.

5. The system of claim 1, wherein the coronal heart slice has been treated with an application of erythropoietin (Epo).

6. The system of claim 1,

(a) wherein the coronal heart slice has been treated with an application of Ro415253; and

(b) wherein the coronal heart slice has further been treated with an application of Epo.

7. A method for culturing cardiomyocytes, comprising the steps of:

(a) dissecting a coronal section of a heart;

(b) removing either the epicardium or both the endocardium and the trabeculae from the coronal section; and (c) culturing the coronal section as an intact slice in vitro.

31

BST99 1231787-1.061039.0010

8. The method of claim 1, wherein the heart is a chick heart.

9. The method of claim 7, wherein the heart is a mouse heart.

10. An assay for detecting a factor that promotes cardiomyocyte proliferation, comprising the steps of:

(a) obtaining a heart slice culture system;

(b) contacting the heart slice culture system with a factor suspected of being a factor that promotes cardiomyocyte proliferation; (c) detecting cardiomyocyte proliferation in the heart slice culture system, wherein the detection of cardiomyocyte proliferation in the heart slice culture system identifies the factor as being a factor that promotes cardiomyocyte proliferation.

11. An assay for detecting a factor that promotes cardiomyocyte survival, comprising the steps of:

(a) obtaining a heart slice culture system;

(b) contacting the heart slice culture system with a factor suspected of being a factor that promotes cardiomyocyte survival; (c) detecting cardiomyocyte survival in the heart slice culture system, wherein the detection of cardiomyocyte survival in the heart slice culture system identifies the factor as being a factor that promotes cardiomyocyte survival.

12. A conditioned medium, comprising a soluble factor that that promotes cardiomyocyte proliferation.

13. The culture medium of claim 12, wherein the medium is conditioned by culture in epicardial cells, following addition of erythropoietin (Epo) to the epicardial cells.

32

BST99 1231787-1.061039.0010

14. The culture medium of claim 12, wherein the medium is a serum-free medium.^"

15. A method for repairing damaged heart tissue, comprising the step of: administering to the damaged heart a composition comprising cardiomyocytes that have been proliferated in vitro.

17. The method of claim 16, wherein the cardiomyocytes are injected into the heart.

18. The method of claim 16, wherein the cardiomyocytes are inj ected into scar tissue in the heart.

19. A method for repairing damaged heart tissue, comprising the step of: administering to the damaged heart a composition comprising a soluble factor that that promotes cardiomyocyte proliferation.

33

BST99 1231787-1.061039.0010