HB-EGF is a Potential Growth Factor for Cardiomyocyte Metabolism. The Possibility for a Novel Therapeutic Target for Heart Failure.

Seiji Takashima
Osaka University Graduate School of Medicine, Osaka, Japan

Heparin binding EGF-like growth factor (HB-EGF), one of the EGF family growth factors, was first discovered as a fibroblast growth factor from macrophage conditioned media. It is secreted as a membrane-anchored form, and truncated by processing enzyme to exert its activity through EGF receptor binding. This HB-EGF processing mechanism is important for cardiac hypertrophic signaling by GPCR agonist. These investigators developed an HB-EGF transgenic mouse to clarify the role of HB-EGF processing.

EGF-receptor transactivation is mediated in two ways: 1) intracellular signaling molecules, such as SERC, and 2) via an alternative pathway, in which a membrane-bound EGF-family ligand metalloproteinase inhibitor is activated, binds soluble EGF to the EGF receptor from the outside and transactivates the EGF receptor.

The alternative transactivation also occurred in cultured rat cardiomyocytes and is involved in cardiac hypertrophic signaling mediated by catecholamine, angiotensin, and endothelin. These neurohormonal factors bind to the receptor and transactivate metalloproteinase, and thus transactivate the EGF receptor. The metalloproteinase inhibitor KB-R7885 inhibits this transactivation and the cardiac hypertrophy induced by the neurohormonal factors. Sharp19, the HB-EGF neutralizing antibody, inhibited the transactivation by the neurohormonal factors and also protein synthesis. The metalloproteinase inhibitor KB-R7885 is also active in vivo, and has been shown to inhibit cardiac hypertrophy induced by continuous infusion of angiotensin II and endothelin.

Thus, these investigators concluded that neuorhormonal factors activate metalloproteinase and shed the membrane-anchored HB-EGF, and then soluble HB-EGF binds to the EGF receptor and induces cardiac hypertrophy.

Their subsequent work in wild type and HB-EGF pro/pro mice showed that 1) blocking HB-EGF shedding from cell membranes causes severe heart failure and early death by cardiomyocyte cell death in the HB-EGF pro/pro mice, 2) in HB-EGF del/del mice the HB-EGF null mouse shows a similar phenotype with the HB-EGF pro/pro, and 3) both phenotypes were similar to human dilated cardiomyopathy (DCM). Interestingly, none of the other ligands in the EGF family show this cardiac phenotype. Although HB-EGF function was abrogated in the entire body in HB-EGF pro/pro and HB-EGF del/del mice, only the heart was affected. The investigators concluded that HB-EGF is an essential growth factor in cardiac myocyte metabolism.

To examine the in vivo role of HB-EGF the investigators made a targeting mouse in which the HB-EGF is not truncated from the cell membrane. In the targeting construct, the exon 1 through 3 region was replaced by HB-EGF full-length CDNA, which has a 2-point mutation that does not allow the HB-EGF to become truncated from the cell membrane and remains as HB-EGF in transmembrane form (Figure 1). The homozygote mouse expressing uncleavable mutant HB-EGF is called HB-EGF pro/pro mouse. The HB-EGF pro/pro mouse was healthy at birth. The promoter region of the HB-EGF pro/pro mouse was intact and it expressed the same amount of mutant HB-EGF at the same organ as did the wild-type mouse. Wild-type HB-EGF was null in the HB-EGF pro/pro mouse. The HB-EGF pro/pro mouse died early compared to the wild-type mouse, with one-half dead at 120 days.

At 12-weeks, the body size and weight of the wild-type and HB-EGF pro/pro mouse were nearly the same. Although the HB-EGF mutant was expressed in the entire body, the abnormal phenotype was observed only in the heart. The heart size was enlarged (Figure 2) and the lung weight increased and pulmonary infusion was detected in the HB-EGF pro/pro mice, indicating death was from heart failure.

Echocardiographic analysis indicated that the heart of the HB-EGF pro/pro mouse was enlarged and fractional shortening was reduced remarkably. Blood pressure was slightly lower than in wild-type mice, but the heart rate was comparable. The extracted heart showed remarkable enlargement in size, and the coronary vessel was intact in HB-EGF pro/pro mice. Cross-section of the LV chamber showed enlargement of chamber size. Histologically, HB-EGF pro/pro showed massive degradation of cardiomyocytes and repression of fibrosis. The coronary vessel was nearly intact. High-magnification HE staining showed massive myocyte degradation, which was repressed by fibrosis. No leukocyte infiltration was seen, indicating an autoimmune or infectious mechanism was not involved in this phenotype.

Histological changes at 6 weeks in the HB-EGF pro/pro mouse were hypertrophic cardiomyocytes and detection of vacuole formation. Cross-sectional examination showed that the vacuole was pressed on the intercalated disk, and further examination showed it to be a dissection of the intercalated disk. Similar vacuole formation and intercalated disk dissection was observed in human DCM biopsy samples.

In a HB-EGF null mouse (HB-EGF del/del) mouse, in which HB-EGF CDNA was knocked out the whole body, HB-EGF expression was abrogated in the entire body. The null mouse showed a similar phenotype to the HB-EGF pro/pro mouse that had a dilated heart chamber and reduced contractility.

Thus, the investigators hypothesize that HB-EGF is partly mediating the cardiac hypertrophy signals through EGF-receptor phosphorylation. However, it is completely lost towards the cardiac cell death, indicating that HB-EGF is essential for cardiomyocyte metabolism. So, for heart failure treatment, drugs that block this pathway are being used, with their beneficial effect probably mediated by reducing contractility, heart rate, and blood pressure. However, adding HB-EGF or another signaling molecule may be beneficial for the survival of the cardiomyocyte. The combination of a blocking agent, such as beta blocker or angiotensin II blocker, and EGF or another growth factor, may be a better therapeutic way to treat heart failure.

Kruppel-Like Zinc-Finger Transcription Factor KLF5/BTEB2 is a Target for Angiotensin II Signaling and an Essential Regulator of Cardiovascular Remodeling

Takayuki Shindo
University of Tokyo, Tokyo, Japan

The transcription regulation and molecular links between stress and cardiovascular (CV) remodeling remain to be clarified. The novel transcription factor KLF5/BTEB2 was discussed in this lecture.

Phenotypic modulation of smooth muscle cells from the contractile to synthetic type is important for the pathogenesis of atherosclerosis. In this process, SMemb gene expression is induced. This group originally identified KLF5 as a transcription factor that binds to the promoter region of the SMemb gene and upregulates its transcription. KLF5 is a member of a Kruppel-like transcription factor, and possesses a transcription activating domain and DNA binding Zn-finger motif.

KLF5 is highly expressed in the cardiac smooth muscle cells or during development in the vascuole, but is downregulated in adults. In contrast, KLF5 induces activated smooth muscle cells in the neointima. In the rabbit balloon injury model, high expression of KLF5 is detected in the neointima. KLF5 transactivates various genes in addition to SMemb in vitro, for example, Egr-1, PAI-1, iNOS, and VCAM. Thus, KLF5 might be involved in cardiovascular (CV) lesions through the activation of these factors.

To better understand the in vivo function of KLF5, Shindo and colleagues generated a KLF5 knockout mouse. KLF5 is essential for fetal development; homozygotes died in utero prior to 8.5 dpc. Heterozygotes are apparently normal and fertile. KLF5 knockout mice seem to have thinner vascular walls, with thinning of the medial and advential layers. In spite of these changes, no major abnormalities were found in the tension development of the aortic ring by angiotensin II (Ang II) or endothelin 1 (ET-1) or in systolic blood pressure.

In the cuff injury model of the femoral artery, after 5 weeks formation of granulation tissue and microvessels was detected around the polyethylene cuff in wild-type mice, but was limited in the knockout mice. The arteries of the KLF5 knockout mice were thin-walled and dilated, in striking contrast to the thickened medial and intimal layers in the wild-type mice. The area of neo-intima formation was less than 2000 in the knockout mice, compared to 12000 in the wild-type mice. KLF5 is important for angiogenesis as well as neointima formation.

Angiogenesis is reduced in the KLF5 knockout mice, as shown by the delayed vessel recovery in the unilateral hindlimb ischemia model (Figure 1). In the tumor transplantation model using Sarcoma 180 (S180), tumor growth at 2 weeks, blood flow, and capillary formation were reduced in the knockout mice. In the Ang II infusion model, no differences were detected after 14 days in the systolic blood pressure or angiotensin II receptor expression levels. But, cardiac hypertrophy, perivascular fibrosis, and interstitial fibrosis were reduced in the knockout mice. KLF5 is a critical transcription factor for cardiovascular remodeling, including cardiac hypertrophy, fibrosis, angiogenesis, and atherosclerosis, based on these data.

Downstream factors in the pathogenesis of CV remodeling

PDGF-A expression is selectively reduced in KLF5 knockout mice. KLF5 is also abundantly expressed in the intestinal tract. Karlson and colleagues showed that KLF5 shows abnormal structure of the digestive tract, such as misshapen villi, reduction in the number of mesenchymal cells, and reduction in the amount of extracellular matrix in knockout mice. Notably, this phenotype is very similar to those of PDGF and knockout mice.

Shindo and colleagues analyzed the relation between KLF5 and PDGF-A. Upregulation of KLF5 was detected 2 hours after Ang II-stimulation of cardiac fibroblast and continued over a 4-hr period, and then was decreased because of PDGF-A regulation.

KLF5 overexpression significantly upregulated the transcription of PDGF-A (Figure 2). In Ang II-stimulated cardiac fibroblasts, KLF5 was shown to directly bind to the PDGF-A promoter in an Ang II-dependent manner (Figure 3).

The synthetic retinoic acid receptor-alpha (RAR) ligands LE135 and AM80 can modulate the transcriptional activity of KLF5. LE135 is a synthetic RAR- and RAR-ß specific antagonist, and AM80 is a synthetic RAR- specific agonist. PDGF-A reporter assay was transfected with an KLF expression vector. LE135 dose-dependently upregulated transcription of PDGF-A, and AM80 dose-dependently decreased the activity of KLF5.

The protein-protein interaction of KLF5 and RAR- was enhanced by LE135 and decreased by AM80, suggesting that LE135 promotes the association of RAR with KLF5 and upregulates PDGF-A transcription, and AM80 promotes the dissociation of RAR from KLF5 and downregulates PDGF-A transcription (Figure 4).

Modifying CV remodeling

The administration of LE135 to KLF5 knockout mice in the cuff injury model increased granulation tissue and neointimal formation, while these were decreased with the administration of AM80 to wild-type mice.

The characteristics of AM80, a possible modulator of KLF5, are shown in Figure 5. In a clinical trial of 66 patients with acute promyelocytic leukemia, AM80 was associated with an 81% remission rate. AM80 suppressed angiogenesis in in vitro and in vivo assays, as shown in Figure 6. Capillary formation enhanced by PDGF-A was suppressed dose-dependently by AM80. In the hindlimb ischemia model, angiogenesis was reduced in AM80-treated mice. In a tumor transplantation model with S180, tumor weight, tumor blood flow, and tumor capillary formation were reduced in AM80-treated mice. AM80 also suppressed cardiac hypertrophy. In the Ang II infusion model, AM80-treated mice had thinner ventricular wall on M-mode echocardiography.

Based on these data, these investigators conclude that KLF5 is a key element linking external stress and cardiovascular remodeling (Figure 7). Drugs that target KLF5 are effective to control CV remodeling. Modulation of KLF5 activity can control CV remodeling including atherosclerosis, angiogenesis, cardiac hypertrophy, and fibrosis.

Critical Roles of Fas/Fas Ligand Interaction and Beneficial Effect of Soluble Fas Gene Therapy for Post-Infarct Ventricular Remodeling and Dysfunction

Genzou Takemura
Gifu University School of Medicine, Gifu, Japan

The preservation of granulation tissue cells of subacute myocardial infarction (MI) through blockade of apoptosis may result in the improvement of post-infarction left ventricular remodeling and heart failure, hypothesized these investigators based on the findings of their previous work.

In a rat model of MI, a pan-caspase inhibitor (BAF) effectively blocked apoptosis of myocardial granulation tissue cells during the subacute stage of infarction. In the granulation tissue of 2-week post-infarction rats, apoptosis of endothelial cells, myofibroblasts, and macrophages were identified on electronmicroscopy. TUNEL positive cells were significantly reduced in the BAF-treated rat hearts compared to control infarct hearts. The apoptotic index of the control infarcted heart was 1.87% while that of BAF-treated heart was 0.69%. The total number of interstitial cells was significantly greater in the infarct area of BAF-treated hearts, probably reflecting the reduced apoptosis of these cells.

At 12 week, the rats treated with BAF for the first 4 weeks had a higher survival rate compared to the control rats, had greater left ventricular (LV) wall thickness, smaller LV end diastolic diameter, and greater percent fraction shortening. Treatment with the caspase inhibitor improved LV remodeling and cardiac contractility of the post-infarct heart in the chronic stage. BAF treatment also relieved congestion and improved cardiac performance of the LV of the post-infarct heart during the chronic stage, as shown by the lower central venous pressure and LV end diastolic pressure and greater LV dPdt.

Severe post-infarct LV remodeling was attenuated by BAF treatment, and the body weight to heart weight ratio was significantly decreased. BAF-treated rats had a thicker infarct wall containing collagen fibers and abundant cellular components and many small vessels compared to control rats, which had thinner walls with fibrous scar tissue containing scanty cell component. The non-myocyte cell population was more than 2-fold greater in the BAF-treated rats compared to control rats. The percent area of -smooth muscle actin positive cells in the infarct area was significantly greater in the BAF-treated rat hearts than in the control.

The number of vessels obtained was about 3-fold greater in the treated heart compared to the control. In contrast, the macrophage, a major cell component of granulation tissue, was not preserved at 12 weeks post-infarction. Electromicroscopic examination revealed myofibroblasts and many groups of mature smooth muscle cells with contractile phenotype in the BAF-treated rats. These smooth muscle cells may originate from the preserved myofibroblasts.

Although the pancaspase inhibitor was found to beneficially effect post-infarct remodeling and dysfunction, there are several limitations of this reagent. Apoptosis of many cell types, both physiological and pathological apoptosis, is largely caspase-dependent. Thus, possible adverse effects by apoptosis inhibition may occur, even if used for a short time. This study did not refer apoptotic machinery specific for granulation tissue cells of MI. Hence, a more specific way to inhibit granulation tissue cell apoptosis is desirable.

The Fas ligand system is important during the initial stage of apoptosis. Both Fas and Fas ligand are overexpressed in the post-infarct granulation tissue cell of mice (Figure 1). The lymphoproliferative (Lpr) mouse strain congenitally lacks Fas and Fas interaction, because of Fas gene defects. The generalized lymphoproliferative disease (Gld) strain contains a point mutation in the Fas ligand gene, thus Fas ligand is non-functioning. In an MI model of these strains, at 4 weeks post-MI, ventricular remodeling was attenuated and cardiac function was improved.
This group then showed that soluble Fas interferes with Fas and Fas ligand interaction and thus is an inhibitor of Fas-mediated apoptosis. They generated a recombinant adenoviral vector of mouse soluble Fas genes. On day 3 post-MI, the adenoviral vector was injected into the hindlimb. The control mice were injected with LacZ gene. The level of soluble Fas in the plasma reached 31.8 mg/ml at 4 days after viral injection. The normal range in humans of soluble Fas is about 2 nanogram/ml. The results were strikingly similar to the case of the treatment with pancaspase inhibitor.

TUNEL positive cells in the granulation tissue post-infarction were significantly reduced in the soluble Fas treated mouse hearts compared with control (Figure 2) At the chronic stage of MI, LV end diastolic and end systolic diameters were smaller and the percent fractional shortening was greater in the treated hearts.

Soluble Fas attenuated post-infarct ventricular remodeling and dysfunction in the treated mice (Figure 3). The body weight to heart rate was significantly decreased and the infarct wall was thicker in the treated compared to the control mice.

The Flk-1 positive vessels were preserved more abundantly in the infarct of the soluble Fas treated hearts. In the control infarct tissue, -smooth muscle actin positive cells were rarely observed, but they were preserved in the infarct area of the soluble Fas treated hearts. However, macrophages were equally rare in both groups. Interestingly, there were bundles of smooth muscle cells with the contractile phenotype in the old infarct area. The beneficial effects of the soluble Fas treatment were observed at 10 weeks post-MI (Figure 4).

In another experiment, anti-apoptotic therapy was begun 3 weeks post-MI. At 10 weeks, no beneficial effect of soluble Fas therapy was found on LV remodeling and dysfunction. This negative finding confirms the concept that granulation tissue cell apoptosis critically influences the post-infarct ventricular remodeling and dysfunction.

Figure 5 illustrates the factors considered responsible for the beneficial effects of anti-apoptotic treatment of granulation tissue cells. Wall stress is an important factor that accelerates ventricular remodeling and is directly proportional to wall thickness. Thus, a thick infarct wall as a result of anti-apoptosis therapy can reduce ventricular wall stress. Therapy may interfere with the vicious cycle between wall stress and remodeling. Bundles of smooth muscle cells with contractile phenotype running in parallel with the surviving myocytes might help cardiac contractility. Preservation of vessels might relieve ischemia of the surviving tissue.

In summary, the Fas and Fas ligand interaction is critical in the apoptosis of granulation tissue cells after MI. Soluble Fas gene therapy begun on day 3 post-MI alleviated post-infarct LV remodeling and heart failure at the chronic stage. The present finding may imply a novel therapeutic strategy against chronic progressive heart failure after a large MI, and can be given to patients who have lost the opportunity for coronary reperfusion therapy at the acute stage.

Role of Proinflammatory Cytokines in Cardiac Remodeling

Toru Kubota
Kyushu University Graduate School of Medical Sciences , Fukuoka, Japan

Myocardial production of proinflammatory cytokines, especially tumor necrosis factor alpha (TNF-) is increased in patients with heart failure. To investigate the role of proinflammatory cytokines in the myocardium, Kubota and colleagues developed transgenic mice with cardiac-specific overexpression of TNF-. The transgene was driven by murine -myosin heavy chain (-MHC) promoter to ensure cardiac overexpression.

Previous work by this group has shown that overexpression TNF- causes: myocardial inflammation with extracellular matrix remodeling, ventricular hypertrophy, with four-chamber dilation, impaired contractility with diminished ß-adrenergic inotropic responsiveness, reactivation of the fetal gene program with down-regulation of calcium handling genes, and premature death with heart failure. These results indicate, they conclude, that myocardial production of TNF- may play an important role in the pathogenesis of myocardial dysfunction and ventricular remodeling.

However, the mechanism by which TNF- induces cardiac remodeling and dilated cardiomyopathy is unclear. Data from this group shows that nitric oxide (NO) exerts versatile effects on cardiovascular (CV) function. Recent studies have shown that iNOS is increased in the failing human heart. A small amount of NO produced by nNOS and eNOS seems cardioprotective by improving myocardial perfusion and inhibiting apoptosis. In contrast, a large amount of NO produced by iNOS may be cardiotoxic by suppressing myocardial contractility and promoting apoptosis. These investigators hypothesize that since TNF- is a potent inducer of iNOS, the negative inotropic effect of TNF- may be mediated by the enhanced production of NO in the myocardium.

They first investigated the effects of iNOS inhibition on cardiac function, using the selective iNOS inhibitor ONO-1714. Blood pressure or heart rate was not changed with ONO-1714 in wild-type or transgenic mice. Although ONO-1714 did not affect baseline contractility estimated by dP/dt max, it significantly improved beta adrenergic inotropic hyporesponsiveness in TNF- transgenic mice, but not in wild-type mice in which iNOS was not unregulated.

Because iNOS inhibition significantly improved beta adrenergic inotropic responsiveness in transgenic mice, these investigators hypothesized that knocking out iNOS may improve cardiac function and prolong survival of TNF- transgenic mice. In TNF-alpha transgenic mice crossed with iNOS knockout mice, eNOS expression was not affected but iNOS activity was completely abolished in the myocardium. Blood pressure or baseline contractility was not affected in this double crossover model. However, disruption of the iNOS gene improved ß-adrenergic inotropic responsiveness only in TNF- transgenic mice.

No change in interstitial infiltration was seen with NO. Cytokine expression was not affected by iNOS knockout. In the iNOS knockout mice, heart failure development and survival was not changed, indicating that although myocardial expression of iNOS plays a key role in the attenuation of ß-adrenergic inotropic responsiveness, NO-independent mechanisms might be more important in the development of heart failure (Figure 1).

Matrix metalloproteinases (MMPs) are a family of proteolytic enzymes that degrade the extracellular components. MMPs are increased in the failing human heart and may play an important role in the process of cardiac remodeling. Because TNF- stimulates the expression of MMPs in vitro, these investigators hypothesized that cardiac remodeling in TNF- transgenic mice may be mediated by the activation of MMPs in the myocardium.

Anti-TNF treatment with adenoviral vector expressing soluble TNF receptor type I (AdTNFRI) attenuated MMP-2 and MMP-9 activity. In transgenic myocardium, collagen I and collagen III were increased, and this increase was prevented by anti-TNF treatment with soluble TNF receptor, indicating that overexpression of TNF- increases MMP activity and promotes myocardial extracellular remodeling.

To examine whether inhibition of MMP may prevent cardiac remodeling and heart failure in the transgenic mice, they treated 4-week-old male transgenic mice with the synthetic MMP inhibitor BB-94 for 8 weeks. Ventricular fibrosis and hypertrophy were significantly reduced and survival was prolonged in the transgenic mice (Figure 2).

In summary, crossing transgenic mice with iNOs knockout mice abolished the iNOS activity and significantly improved ß-adrenergic inotropic responsiveness. However it did not prevent cardiac remodeling or improve survival. In contrast, treatment with the MMP inhibitor BB-94 significantly reduced ventricular fibrosis and hypertrophy and prolonged survival. These investigators conclude that these results indicate that activation of MMP but not iNOS may play a pivotal role in the pathogenesis of cardiac remodeling. TNF- transgenic mice may provide a unique model to study myocardial inflammation and remodeling and to explore novel therapeutic strategies for heart failure.

Mammalian Target of Rapamycin (mTOR): A New Molecular Target for Cardiac Hypertrophy

Tetsuo Shioi
Kitasato University School of Medicine, Sagamihara, Japan

Cardiac hypertrophy, typically defined as increased myocardial size, is a major risk factor for heart disease. The role of insulin in the phosphoinositide 3-kinase (PI3K) pathway in determining myocardial size and the mammalian target of rapamycin (mTOR) in cardiac hypertrophy was reviewed.

The mechanism of organ size regulation has been studied in several models. Work on the wing size of Drosophila suggests that organ size is not regulated at the cellular level, but at the organ level. All of the genes identified to control organ size in Drosophila are in the PI3K pathway. Most of the genes control organ size by regulating both cell size and cell number in the organ, but some regulate organ size by regulating only cell size.

Insulin and insulin growth factor receptors phosphorylate insulin receptor substrates (IRS). Phosphorylated IRS activate PI3K. Expression of activated PI3K in Drosophila wing promotes wing growth and expression of dominant-negative PI3K mutant results in a smaller wing. Other developmental processes are not disturbed. The role of insulin and insulin-like growth factor 1 (IGF 1) in physiological and pathological conditions have been studied extensively (Figure 1).

To study the role of the insulin signaling pathway, his group moderated the activity of the members of the pathway by genetic methods. IGF 1 receptor gene was overexpressed in a heart-specific manner. Heart size was increased in IGF 1 receptor mice, with increased wall thickness while the proportion of the size of each chamber was preserved. Cardiomyopathic changes, such as necrosis or fibrosis, were not observed. The mice survived normally and cardiac function was preserved on echocardiography.

They generated transgenic (TG) mice that expressed contributive active PI3K or dominant-negative PI3K in a heart-specific manner. Expression of contributive active PI3K resulted in a larger heart, and dominant-negative PI3K resulted in smaller heart.

Increase or decrease in heart size was associated with a comparable increase or decrease in cardiac myocyte cell size. All of the mice survived normally and cardiac function was preserved on echocardiography, even in the smaller hearts.

They also characterized cardiac specific PTEN knockout mice. Antagonism of PI3K function resulted in Akt activation in the heart tissue of the knockout mice. Heart weight was increased by 35% in PTEN knockout mice. This group has shown that IGF-1 receptor, PI3K, PTEN, and Akt are important in heart size examination. Hence, the insulin-PI3K pathway is an important signaling module for organ size regulation in mammals. To identify target genes of PI3K in intact heart tissue, they prepared RNA from heart tissue from PI3K TG mice and performed CDNA chip analysis. If a gene is upregulated or downregulated in dominant-negative PI3K mice and the gene is regulated in the opposite way in constitute active PI3K mice it is very likely that the gene is a direct target of PI3K. For example, cardiotropin gene is likely to be regulated by PI3K. Cardiotropin is an important growth factor for cardiomyocyte. So, PI3K may regulate heart size by regulating such growth factors. Work by Izumo and colleagues shows that a single nucleotide polymorphism (SNP) of the IGF 1 gene appears to be associated with heart size, suggesting that insulin signaling may be involved in cardiac hypertrophy in humans.

This group showed in transgenic mice studies that the insulin signaling pathway is important in the development of heart growth. To examine the role of the pathway in cardiac hypertrophy induced by pathological stress, rapamycin was used as a pharmacologic probe.

Rapamycin inhibits the activity of mTOR, which regulates downstream effectors of insulin signaling pathways. 4E-BP1 regulates the activity of the translation-initiation complex. Liposomal S6 kinase (S6K) is another important target of mTOR. S6K phosphorylates 40S ribosomal S6 protein and increases the translation of selective mRNAs (ribosomal proteins, translation elongation factors, etc). Importantly, it is activated by most of the hypertrophic agonists, such as angiotensin, calcium, protein kinase C. Body size of the S6K 1 knockout mice is smaller than control.

Because rapamycin potently inhibits cellular growth and cytokine production and regulates effectors of the insulin pathway, rapamycin may effectively suppress cardiac hypetrophy. To test this possibility, rapamycin was given to adult mice with ascending aortic constriction. Heart weight increased by 35%, mostly occurring within 48 hours after the operation. S6K activity was highest 4 hours after the operation and returned to nearly normal at 48 hours. This result may be comparable with the natural growth curve of the heart, because growth of the heart stops at 48 hours. S6 phosphorylation was highest 4 hours after the operation. Rapamycin completely suppressed S6 kinase activation in the binded heart 4 hours after the operation. S6 phosphorylation was also completely suppressed.

Rapamycin did not induce lethality or body weight loss in adult mice. Heart weight increased about 35% in the banded mice and rapamycin significantly attenuated the heart weight increase by 60%. On echocardiographic examination, rapamycin significantly reduced the LV end diastolic diameter. Rapamycin did not affect cardiac contractility.

To examine the effects of rapamycin on primary disease, rapamycin was given to a rat autoimmune myocarditis model. Cardiac myosin immunization into Lewis rats induced autoimmune myocarditis. Three weeks after immunization, extensive myocardial inflammation was observed and heart weight was increased (Figure 2). Rapamycin almost completely suppressed the increase in heart size.

In summary, mTOR appears to be a good candidate for heart failure treatment. But several questions must be resolved, including whether rapamycin can reverse established cardiac hypetrophy and prevent or reverse cardiac dilation. Despite the significant amount of work to study the signaling mechanism of cardiac hypertrophy and failure, a gap between basic information and clinical practice remains. Methods to deliver drugs or genes in a heart-specific manner to avoid systemic toxicity are needed. Rapamycin target molecules will be useful for improved treatment of heart failure.

Therapeutic Myo-Angiogenesis for Ischemia-Induced Myocardial Remodeling by Transplantation of Autologous Bone Marrow Mononuclear Cells

Hiroaki Matsubara
Kansai Medical University, Moriguchi, Japan

The current concept of angiogenesis is that endothelial progenitor cells in the peripheral circulation migrate to the inflammatory areas and adhere and differentiate into the endocardium. Bone marrow mononuclear cells (BMMC) contain cardiac progenitor cells and release angiogenic factors, such as bHGF and VEGF. Matsubara and colleagues hypothesized that implanting BMMC into ischemic limb or heart may enhance angiogenesis or cardiomyogenesis. Culturing human marrow CD34+ cells with human endothelial cells resulted in a network formed with endothelial cells.

In the pig acute or chronic ischemia model, BMMC were injected into the ischemic areas by open thoracotomy and catheter-based injection. On coronary angiography, collateral vessels were observed prior but not after the injection. Neocapillary vessel formation was increased about 2-fold after injection. The bone marrow cells were differentiated in the endothelium. In a further experiment, in which GFP-positive bone marrow stem cells were injected into the scar area in the pig model, there was little differentiation into cardiomyocytes from the GFP lin-, cKit-, Sca-1 bone marrow hematopoietic stem cells. Cell therapy using bone marrow cells results in angiogenesis, not myogenesis.

This group has performed pre-clinical studies of marrow cell implantation in the ischemic limb model in the rabbit, acute myocardial infarction in the pig, and in the chronic myocardial ischemic model in the pig. These studies showed that there is no cardiac injury by inflammatory cytokines released from implanted bone marrow cells, no differentiation of other lineage cells (osteoblasts, fibroblasts), and no malignant arrhythmia on Holter monitoring. This shows the safety and feasibility of bone marrow implantation into ischemic myocardium. The cells should be injected into the hibernating myocardium.

The Therapeutic Angiogenesis by Cell Transplantation (TACT) clinical trial studied the effect of bone-marrow derived cells implantation in patients with ischemic heart disease (IHD) and peripheral artery disease (PAD). This multicenter study in Asia and the US studied 96 patients.

The results of the Sole Cell Therapy Phase I arm of this trial in patients with IHD were reviewed. NOGA™ 3-D electromechanical mapping is used to map the hibernating area for the catheter-based cell delivery. Voltage and mechanical mapping was analyzed simultaneously. Wall motion was improved at 6 weeks in the pig injected with bone marrow cells compared to saline control injection.

One patient with CCS class IV angina has been studied: a 64-year-old male patient with an old MI with no medical or therapeutic revascularization options and who was using nitroglycerin spray 10-15 times per day. On left coronary angiogram, before the implantation, no blood vessels were seen around the left circumflex and the injection was made in this area. This is the first patient in the world to receive this type of implantation, which was done using the open heart method.

Spect-sestamibi scan showed improvement of myocardial ischemia at 2 months. At 6 weeks, on NOGA mapping, improvement was seen. The left ventricular ejection fraction improved from 42% before the procedure to 53% at 6 weeks post-procedure. Chest pain disappeared and cardiac function improved. The use of nitroglycerin spray decreased from 10 times per day before the procedure to 2 times in the 6 months and 6 times in the 12 months after the procedure. The number of PVCs decreased from 500-1000/day before the procedure to <100/day at 6 months post-procedure. There were no side effects.

The second clinical case was a 59-year-old woman with medically refractory angina (CCS 3-4), with diabetes, ASO, hypertension, heart failure, and an ejection fraction of 45%. She had a history of PCI and bypass surgery. Catheter injection of bone marrow cells were made into 20 different sites in the heart. At week 8 post-procedure, wall motion was improved. Spect-sestamibi scan showed improvement of the distribution areas.

In summary, for future regeneration therapy for IHD, angiogenesis is best for the ischemic portion. The tools are bone marrow cells or angiogenic factors (VEGF, bFGF, HGF). For the infarcted scar portion, myocyte regeneration is needed, and the tools used in the US and Europe are skeletal myoblasts. Presently, autologous transplantation is prevalent. Another potential cell is the cardiac stem cells, which may come from the bone marrow or from cells originating from the heart or skeletal myoblasts.