Citation
Grauss, R. W. (2008, January 17). Regenerative medicine in cardiovascular disease: from tissue enginering to tissue regeneration. Retrieved from https://hdl.handle.net/1887/12556
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/12556
Note: To cite this publication please use the final published version (if
applicable).
SUMMARY, CONCLUSIONS AND FUTURE PERSPECTIVES
197198
SUMMARY
The general introduction in Chapter 1 of this thesis provides an overview of the emerging fi eld of regenerative medicine, and its diff erent areas of technology. Contemporary approaches to heart valve repair and replacement by mechanical and biological prosthesis, as well as their limita- tions are discussed. Furthermore, the novel approach of (stem)cell-based therapy in ischemic heart disease and a variety of potential cell types are described.
In Part I of the thesis diff erent methods to produce completely acellular porcine and rat aortic valve scaff olds for the purpose of creating a biological substitute that can restore, maintain, or improve normal valve function were assessed. In addition, the in vivo behavior of these valves was studied in a rat model. Subsequently, Part II of the thesis describes the in vivo behavior and functional improvement after transplantation of diff erent human cell types in an animal model of acute myocardial infarction. Furthermore, the feasibility of combining cell therapy with gene-therapy was studied in this same model. Finally, in Part III diff erent methods for car- diac phenotyping in mice were assessed and compared.
Par t I
In Chapter 2 we addressed the histological changes in porcine valve extracellular matrix (ECM) constitution induced by two previously described cell extraction methods; a non-ionic deter- gent Triton X-100 and an enzymatic trypsin decellularization method. (Immuno-)histochemistry demonstrated that trypsin treatment resulted in a fragmentation and distortion of elastic fi bers.
Changes in collagen distribution were observed in both groups. An almost complete washout of glycosaminoglycans and chondroitin sulfate was observed in both methods, with a smaller glycosaminoglycans reduction in the trypsin group. Furthermore, both treatments resulted in a considerable washout of the adhesion molecules laminin and fi bronectin. In addition, the possi- bility of production of ECM compounds by in vitro reseeded cells was investigated. Cultured and seeded von Willebrand factor positive endothelial cells were capable of synthesizing laminin, fi bronectin and chondroitin sulfate. All components that were lost during the decellularization treatment.
In Chapter 3 we developed a 2-step detergent–enzymatic extraction method involving sodium dodecyl sulfate (SDS) in combination with RNase and DNase to produce completely acellular rat aortic heart valve conduits. Against our expectations previously described methods to decel- lularize larger porcine aortic valves failed in removing cells from the small and compact rat aortic valves. Our novel decellularization procedure resulted in a complete loss of all cellular structures from the entire valve conduit with only minimal histological damage to the ECM. In
199 addition, we performed transplantation experiments in a heterotopic rat aortic valve implanta-
tion model. These experiments included 4 diff erent groups of animals, consisting of cellular allogeneic (n = 4), acellular allogeneic (n = 4), cellular syngeneic (n = 3), and acellular syngeneic (n = 3) valve transplantations. After 21 days, all the explanted cellular allograft leafl ets were deformed and swollen, with severe damage to the elastin and collagen fi bbers accompanied by a loss of chondroitin sulfate. The morphology and ECM constitution of the decellularized allogeneic and syngeneic conduits, however, appeared almost normal. Interestingly in both the acellular allogeneic and syngeneic valve leafl ets ingrowth of transanastomotic smooth-muscle actin–positive intimal cells occurred, that were capable of producing the ECM molecules chon- droitin sulphate and fi bronectin.
In Chapter 4 we focussed on characterizing the immune response to cellular and acellular aor- tic valve allografts, and aimed to correlate these results to structural damage. Rat aortic valve conduits were decellularized by a 2-step detergent-enzymatic extraction method described earlier in Chapter 3. Cellular allogeneic (n = 8), acellular allogeneic (n = 8), and syngeneic (n = 8) aortic valve conduits were then grafted into the descending aorta for either 7 (n = 4) or 21 (n = 4) days. Immuno-histochemical analysis revealed a severe infl ux of CD8+ T-lymfocytes and CD68+ macrophages accompanied by interstitial cell apoptosis 1 week after transplantation of cellular allograft valves, followed by extensive fragmentation of the collagen matrix at 3 weeks. However, decellularization resulted in an absent cellular infi ltration after 1 week with histological preservation of the collagen matrix after 3 weeks. In addition, chemical analysis revealed an increase in the amount of denatured collagen after decellularization (7.8 ± 0.7 % vs.
12.0 ± 1.2 %, p < 0.05), which did not change 3 weeks after transplantation (12.0 ± 1.8 %, p = ns), in contrast to the increase in denatured collagen after cellular aortic valve allograft transplantation (12.3 ± 1.5 %, p < 0.05).
Par t I I
In Chapter 5 we assessed the feasibility of using human mesenchymal stem cells (hMSCs) from patients with ischemic heart disease (IHD) in cell based therapy for acute myocardial infarction, in an immune-compromised NOD/scid mouse model of acute myocardial infarction. hMSCs from IHD patients labeled with green fl uorescent protein (GFP) (hMSC group: n = 12) or vehicle only (Medium group: n = 14) were injected into infarcted myocardium of these mice. Sham-oper- ated animals were used as control (n = 10). Two and fourteen days after the operation, cardiac anatomy and function were serially assessed using a small animal 9.4-T magnetic resonance imager (MRI). At day 2, gadolinium-DPTA delayed-enhancement MRI showed no diff erence in infarct size between the hMSC and Medium groups (33 ± 2 % versus 36 ± 2 %; p = ns), with
200 a comparable increase in left ventricular (LV) volume, and decrease in ejection fraction (EF) in both MI groups. However, at day 14, the EF was signifi cantly higher in the hMSC than in the medium group (24 ± 3 % versus 16 ± 2 %; p < 0.05). After two weeks immuno-histological analysis showed an engraftment rate of 4.1 ± 0.3 % of the labelled hMSCs. These cells were positive for the human endothelial cell-specifi c protein vWF (16.9 ± 2.7 %) and the smooth muscle cell (SMC) marker ASMA (78.3 ± 4.0 %), but not for the highly specifi c human SMC marker smMHC or the cardiomyocyte-specifi c proteins α-sarcomeric actin, cardiac Troponin I and myosin heavy chain.
In addition, the total blood vessel density in the scar and borderzone areas were signifi cantly higher in the hMSC group (610 ± 78 / mm2 and 810 ± 68 / mm2 respectively) than in the Medium group (347 ± 56 / mm2 and 565 ± 50 / mm2 respectively; both p < 0.05). Finally, measurements of wall thickness showed that hMSC injection signifi cantly reduced the extent of infarct wall thin- ning (Medium group: 18.0 ± 2.1 × 10-2 mm, hMSC group: 30.5.0 ± 2.9 × 10-2 mm; p < 0.05).
In Chapter 6 we hypothesized that forced expression of the pivotal cardiomyogenic transcription factor myocardin in hMSCs may enhance their potential therapeutic eff ect. hMSCs transfected either with GFP and Myocardin (hMSCmyoc, n = 10) or GFP and an empty vector (hMSC, n = 10), were injected into the acutely infarcted immune-compromised NOD/scid mouse heart. Animals injected with medium (n = 12) and sham operated animals (n = 12) were used as controls. Detailed hemodynamic measurements were performed at day 2 and 14 using a 9.4T MRI, and at day 15 using invasive LV pressure-volume measurements. At day 14, EF in both cell-treated groups was preserved compared to the medium treated group; in addition hMSCmyoc injection also reduced LV end-systolic and end-diastolic volumes (P < 0.05), not observed after hMSC injection alone.
In addition, pressure-volume measurements revealed signifi cantly better load independent parameters of systolic and diastolic function after hMSCmyoc injection as compared to hMSC treatment. Quantitative assessment at day 15 demonstrated a signifi cant higher engraftment rate in the hMSCsmyoc group (5.8 ± 0.5 %) as compared to the hMSC group (4.2 ± 0.3 %, p < 0.05).
Furthermore, these cells expressed a number of cardiomyocyte specifi c markers not observed in the hMSC group. Finally, at day 15 the total blood vessel density in the scar area was signifi cantly higher in the hMSC and hMSCmyoc groups (11.8 ± 2.1 %, and 11.3 ± 1.3 %; p = ns) as compared to the Medium group (5.9 ± 0.9 %; p < 0.05).
In Chapter 7 we assessed the possibility that adult human epicardial derived cells (hEPDCs) could recapitulate part of their embryonic program when transplanted into diseased myocar- dium, serving as a novel progenitor cells in cardiac cell therapy. hEPDCs were isolated from human adult atrial tissue and injected into an immune-compromised NOD/scid mouse model of acute myocardial infarction. Analysis by 9.4T MRI showed that ejection fraction was signifi cantly higher after hEPDC injection after 2 (n = 22) and 6 (n = 15) weeks as compared to medium injec-
201 tion alone (n = 24 and n = 18 respectively). Furthermore, end-systolic and end-diastolic volume
were signifi cantly smaller in the EPDC-injected group than in the medium-injected group at all ages evaluated. This was accompanied by an increased survival, increased scar vascularity, and increased wall thickness at 2 weeks. We also demonstrated massive engraftment of injected hEPDCs at 2 weeks, expressing α-smooth muscle actin, von Willebrand Factor, sarcoplasmic reticulum Ca2+-ATPase and SCN5a, but no stringent cardiomyocyte markers. However, although at 6 weeks survival and wall thickness were still increased only few hEPDCs could be detected.
Par t I I I
Finally, in Chapter 8 we performed a head-to-head comparison between two diff erent methods which both enable cardiac phenotyping under physiological, closed-chest conditions in murine models of acute myocardial infarction. Namely, magnetic resonance imaging (MRI) and pres- sure-volume loops by conductance catheter (CC). After permanent occlusion of the left anterior descending coronary artery (LAD), LV enddiastolic volume (EDV), endsystolic volume (ESV) and ejection fraction (EF) were determined by MRI (day 14) and CC (day 15). We found moderate-to- strong linear correlations for EDV (CC = 0.62MRI+9.71 (R = 0.61)), ESV (CC = 0.62MRI+11.35 (R = 0.72)) and EF (CC = 0.80MRI-1.78 (R = 0.81)). Furthermore, after analysing diff erent groups (Sham oper- ated animals and small or large myocardial infarction), we found that volumes and EF were con- sistently lower by CC than by MRI, but that group diff erences were evident for both techniques.
Moreover, receiver-operating characteristics (ROC) curve analysis showed good sensitivity and specifi city for both techniques, with superior results for MRI. In addition the complementary additional features of MRI and CC and their value for heart failure research were discussed.
202
CONCLUSIONS
• Chemically induced decellularization by Triton or trypsin results in changes in the ECM con- stitution, which could lead to problems in valve functionality and cell growth and migration.
Future studies should focus more on the eff ect of chemically-induced cell extraction on the ECM of the remaining scaff old and the potential in vitro or in vivo restoration of lost compounds.
• Although procedures to detach cells from their surrounding matrix inevitably implies a cer- tain amount of damage to the ECM, the changes observed after the 2-step detergent–enzymatic extraction method involving sodium dodecyl sulfate in combination with RNase and DNase are much less than the damage caused by the recipient’s immune response.
• The 2-step detergent-enzymatic extraction method in rat aortic valve allografts results in an attenuation of the cellular immune response, and a histological but not chemical preservation of the collagen matrix.
• Mesenchymal stem cells from patients with ischemic heart disease can engraft in the inf- arcted mouse myocardium and preserve LV function 2 weeks after acute myocardial infarction, potentially through enhancement of scar vascularity and reduction of wall thinning.
• Forced expression of the cardiomyogenic transcription factor Myocardin in mesenchymal stem cells from patients with ischemic heart disease improves LV function and also limits LV remodeling, potentially through improved engraftment and diff erentiation into a cardiomyo- cyte-like phenotype.
• Adult human EPDCs preserve cardiac function and attenuate ventricular remodeling after transplantation into ischemic mouse myocardium in the short, but also in the long-term, and are promising candidates for clinical application in human infarcted hearts.
• Although Magnetic resonance imaging (MRI) and pressure-volume loops by conductance catheter (CC) are both highly valuable for evaluation of LV volume and function in murine myo- cardial infarction studies, MRI is recommended for longitudinal studies, for accurate absolute volumetric measurements, and when anatomic information is essential. In addition, the CC method has the possibilities for advanced analysis of LV function, mechanics and energetics.
FUTURE PERSPECTIVES
203Tissu e Engineering in Regenerative M e dicine
The ideal artifi cial valve is a non-obstructive, non-thrombogenic tissue valve substitute which will last the lifetime of the patient, provide ongoing remodelling and repair of cumulative injury, and potentially grow in maturing individuals. Although the concept of tissue engineering a heart valve has been shown to be promising, several problems still have to be resolved.
Mechanisms of heart valve morphogenesis during embryological development are highly regulated and complex, and are only beginning to be understood. Although much knowledge exists on the vascular wall cells of the great vessels, little is known about valvular endothelial and interstitial cells, which are essential in heart valve tissue engineering. Interestingly, the use of stem or progenitor cells for seeding of tissue engineered valves may preclude the issue of a preselected cell phenotype. Therefore, future studies could focus more on the use of stem cells in this new area of regenerative medicine.
The functional role of tissue and ECM elements, lost by the cell extraction protocols, is an addi- tional point of concern. At present it is unknown whether they really need to be replenished in the tissue engineered heart valves. In addition, more sophisticated techniques that could remove the cells from biological tissue without damaging the remaining tissue need to be explored.
Furthermore, for the production of a possible ‘off -the-shelf’ product, researchers should concen- trate on the use of allogeneic or even xenogeneic tissues or cells, in combination with cellular or molecular biological modifi cations to prevent a host specifi c immune-response.
Finally, before tissue engineered heart valves are ready to enter the clinical arena, they need to replicate some, if not all, of the biological characteristics and functions of a normal valve, and they must have been validated in appropriate in vitro and in vivo models.
Cell Based Th erapy in Cardiac Regenerative M e dicine
Although the fi eld of cell based therapy is currently under intense investigation with numerous breakthroughs achieved, several major controversies still need to be addressed. The present the- sis has demonstrated benefi cial eff ects of some novel cell types, however, the question which of all the currently described cell types is best suited for cardiac regeneration and vascularization still remains unanswered. More importantly, which cells should be used in acute MI, and which ones in heart failure or angina? In addition, the number of cells needed per unit volume of damaged cardiac muscle, as well as the optimal time window between damage and repair (e.g.
during MI vs. early vs. late) has to be elucidated. Moreover, to further complicate this fi eld, sev- eral clinical studies all have used diff erent application methods, therefore comparative studies
204 should assess the relative merits of intravenous, intracoronary, transepicardial and endocardial application methods. The best location of implantation must be compared in animal models, e.g. scar, infarct borderzone zone or remote myocardium.
Furthermore, the exact mechanisms by which these diff erent cell types display their therapeu- tic eff ects still remains controversial. Although the initial thought was that stem cells could regenerate the diseased myocardium by diff erentiating into functional cardiomyocytes, this phenomenon appears to be at best very rare.
Gene-therapy approaches in combination with cell therapy, as described in this thesis, could therefore be a promising solution. Interestingly, our laboratory recently demonstrated that forced myocardin expression in hMSCs also leads to the functional expression of cellular com- ponents involved in electrical conduction. This might result in improved electrical coupling between areas of surviving myocardium, leading to a more effi cient contraction of the scarred myocardium. Moreover, what is the effi cacy of cell transplantation relative to other currently available medical and surgical treatment options? One could argue that enhancing perfusion by cell therapy is already suffi cient to improve function. In addition, future research should focus more in how long the cell based improvement lasts, and how cell survival after injection can be enhanced. Indeed recent studies have also applied gene-therapy approaches to overcome this problem (AA Mangi, Nat Med 2003;9:1195-201).
Therefore, a better understanding of the molecular mechanisms; the complex interaction between chemokines, adhesion molecules, and extracellular matrix factors; and stem cell traf- fi cking is inevitable for the further evolution of cell based therapy in cardiovascular disease.
Cardiac Phenot yping of th e Murine Hear t
Mouse models of cardiac disease play an important role in cardiovascular research en have proven their clinical translatability. Methods to determine cardiac function and volumes are challenging due to the small size and rapid heart rate. Although the present thesis only assessed two common methods for cardiac phenotyping in mice, MRI and the conductance catheter, novel methods are emerging. Recently also 3D echocardiography (Dawson et al. Circulation, 2004) and Speckle-tracking technology (Popovic et al. AJP, 2006) have come into sight. Inter- estingly, several authors have been experimenting with micro computed tomographic (μCT) imaging (Nahrendorf et al., AJP 2007) and mouse Positron emission tomography imaging (PET) (Chang et al., J Mol Cell Cardiol, 2007). Therefore, in the near future multimodality imaging is expected to play an important role in studying mouse models of human cardiac disease.
