*3.6. No Di*ff*erence in Therapeutic E*ff*ectiveness of ECFCS Derived from Diabetic or Nondiabetic Donors after Transplantation into Ischemic Myocardium*

To evaluate the hypothesis that diabetes mellitus may impair the protective potential of ECFCs, we stratified cell donor patients into "diabetic" versus "nondiabetic". Interestingly, there was no statistical difference in the amount of CD45+/CD34<sup>+</sup> BMCs within the ischemic myocardium between animals treated with either ECFCs from diabetic or nondiabetic patients (1.1 ± 0.4 (non DM-ECFCs) vs. 0.8 ± 0.2 (DM-ECFCs) vs. 0.7 ± 0.1% (control); for non DM-ECFCs vs. DM-ECFCs *p* = 0.18). Furthermore, LV function assessed by catheterization revealed no significant difference between animals transplanted with ECFC from diabetic or nondiabetic patients after MI. However, both ECFCs from diabetic or nondiabetic patients were able to significantly improve cardiac contractility compared to saline-treated control animals (dPdt max: 5154.65 ± 469.37 (non DM-ECFCs) vs. 5240.02 ± 447.51 (DM-ECFCs) vs. 3896.24 ± 534.95 (control) mmHg/s, for nonDM-ECFCs vs. DM-ECFCs *p* = 0.28; CO: 3821.92 ± 238.14 (non DM-ECFCs) vs. 2006.71 ± 243.18 vs. 3236.54 ± 384.78 (DM-ECFCs); for nonDM-ECFCs vs. DM-ECFCs *p* = 0.28; Figure 6).

**Figure 6.** No difference in therapeutic effectiveness of ECFCs derived from diabetic or nondiabetic donors after transplantation into ischemic myocardium. (**A**) Bar graphs representing the percentage of CD45+/CD34<sup>+</sup> stem cells in the ischemic hearts of control animals (white bar) after ECFC transplantation from nondiabetic donors (black bar) and after ECFC transplantation from diabetic donors (grey bar) 2 days after MI. Data represent mean ± SEM; n.s., not significant. (**B**) Bar graphs representing cardiac output and contractility (**C**) of control animals (white bar) after ECFC transplantation from nondiabetic donors (black bars) and after ECFC transplantation from diabetic donors (grey bars) 30 days after MI. Data represent mean ± SEM; n.s., not significant. (**D**) Bar graphs representing the percentage of CD45+/CD34<sup>+</sup> subpopulations in the ischemic hearts of control animals (white bar) after ECFC transplantation from nondiabetic donors (black bars) and after ECFC transplantation from diabetic donors (grey bars) 2 days after MI. Data represent mean ± SEM; n.s., not significant.

#### **4. Discussion**

This study aimed to evaluate a proregenerative/reparative potential of human ECFCs in a murine model of MI. Our main findings were the following: (1) transplantation of human ECFCs into ischemic myocardium resulted in an improved cardiac function and reduced size of infarction 30 days after MI; (2) the attenuated remodeling after ECFC transplantation was associated with enhanced neoangiogenesis; (3) treatment with ECFCs increased the number of resident Sca-1<sup>+</sup> stem cells without significant homing of migrated BMCs; and (4) there was no difference in therapeutic effectiveness between transplanted ECFCs derived from diabetic or nondiabetic donors.

In the present study, we used immunocompromised SCID animals to prevent cell death due to cell rejection. However, 30 days after MI, we were not able to detect any transplanted cells by histological analysis. Sheikh et al. [6] investigated recently the survival kinetics and gene expression changes of transplanted BMMCs after transplantation into ischemic myocardium. Utilizing molecular-genetic bioluminescence imaging, they demonstrated short-lived survival of cells following transplant, with less than 1% of cells surviving by six weeks post-transplantation [6]. In line with these data, Higuchi et al. [22] showed in a rat study of chronic myocardial ischemia that human EPCs transduced with the sodium iodide symporter (NIS) gene for reporter gene imaging by (124)I-PET and labeled with iron oxides for visualization by MRI showed poor cell engraftment after injection. The (124)I uptake decreased on day three and was undetectable on day seven, which was confirmed by histological analysis with CD31 and CD68 antibodies [22].

However, in accordance with previous preclinical [23,24] and clinical studies, we were able to see functional improvement after transplantation of ECFCs into ischemic myocardium. These functional effects were associated with an attenuation of adverse post-MI cardiac remodeling reflected by smaller size of infarction and less pronounced thinning of the LV wall after treatment with ECFCs. Moreover, our data showed an improved neovascularization reflected by a high number of CD31<sup>+</sup> vessels at the infarct border zone. It is an ongoing matter of debate how transplanted stem and progenitor cells are able to increase neovascularization. On the one hand, it has been shown that EPCs are carriers of proangiogenic cytokines and growth factors like VEGF, HGF, angiopoietins, and IGF-1 [25–27]. In this regard, we detected increased expression of proangiogenic transcription factors HIF-1alpha and MMP-2 in ECFCs in vitro. Dubois et al. [28] investigated the paracrine activation after ECFC transplantation in a pig model of acute MI. After infusion of autologous ECFCs, they observed decreased infarct size and a greater vascular density, which was associated with higher levels of the proangiogenic placental growth factor (PLGF) [28]. Likewise, in a rat model of transient middle cerebral artery occlusion (MCAO), injection of ECFCs was associated with a reduction in the number of apoptotic cells and an increase in capillary density at the site of injury. These effects were associated with an increased expression of growth factors VEGF and IGF-1 in the infarct area 14 days after MCAO [12]. Furthermore, neovascularization also increased cell survival, and decreased apoptosis is an important contributor of cardio-protective paracrine effects after cell transplantation. In this regard, Kim et al. [29] showed a significant decrease of terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) positive nuclei after transplantation of ECFCs compared to control animals in tissue sections collected from the peri-infarct zones at day seven [29].

On the other hand, there is some evidence that migration and homing of BMCs as well as expansion of resident cardiac stem cells beneficially influences cardiac remodeling after MI [3]. We used cardiac FACS analysis to quantify the number of migrated CD45+/CD34<sup>+</sup> BMCs after ECFC transplantation. We were not able to see a significant increase in the absolute number of BMCs and their subpopulations. However, we found a higher number of CD45−/CD34−/Sca-1<sup>+</sup> cardiac-resident cells within the myocardium of human ECFC treated animals injected with ECFCs. In the group of mice treated with huECFC, we observed a more than 2-fold increase in the number of CD34−/CD45−/Sca1<sup>+</sup> cells 2 days after MI when compared to control animals. Sca1<sup>+</sup> cells within the mouse heart have been described as a multipotent stem cell population that are able to differentiate into cardiomyocytes, endothelial cells, and smooth muscle cells in vitro and after cardiac grafting. In fact, more recently and in contrast to

previous reports, it has been shown that adult Sca-1<sup>+</sup> cardiac-resident cells appear to lack significant cardiomyogenic potential [30]. Lineage tracing of Sca-1 expressing cells in the mouse heart revealed only a very limited ability to significantly contribute new cardiomyocytes to the heart after MI in the mouse. In fact, they demonstrated a mostly endothelial expression pattern in Sca-1mCmR26dTomato hearts that colocalized with CD31 expression [31]. Tang et al. [32] by targeting endogenous Sca1<sup>+</sup> cells in Sca1-2A-CreER; R26-tdTomato mice through pulse labeling experiments showed that 94.25 ± 0.54% tdTomato<sup>+</sup> cells were CD31<sup>+</sup> endothelial cells, indicating that most Sca1<sup>+</sup> cells adopt an endothelial cell fate during cardiac homeostasis. By immunostaining for tdTomato and VE-cad on post-MI murine heart sections, they detected a substantial number of tdTomato<sup>+</sup> coronary endothelial cells in the injured region. Roughly, two thirds of VE-cad<sup>+</sup> cells expressed tdTomato, whereas 91.77 <sup>±</sup> 1.12% tdTomato<sup>+</sup> cells expressed VE-cadherin in the injured myocardium. They concluded that Sca1<sup>+</sup> cardiac progenitor cells mainly differentiate into cardiac endothelial cells and fibroblasts but not cardiomyocytes during cardiac homeostasis and after injuries. In a different experimental approach, Vagnozzi et al. [33] generated Ly6a gene-targeted mice containing either a constitutive or an inducible Cre recombinase to perform genetic lineage tracing of Sca-1<sup>+</sup> cells in vivo. Similarly, they observed that the contribution of endogenous Sca-1<sup>+</sup> cells to the cardiomyocyte population in the heart was < 0.005% throughout all of cardiac development, with aging, or after MI. Moreover, pulse labeling of Sca-1<sup>+</sup> cells with an inducible Ly6a-MerCreMer allele revealed that Sca-1<sup>+</sup> cells rather represent a subset of vascular endothelial cells that expand postnatally with enhanced responsiveness to pathological stress in vivo [33]. In our study, huECFCs treatment were a potent stimulus for Sca-1<sup>+</sup> endogenous angiogenic cells. Anti-CD31 immunohistochemistry confirmed improved neovascularization within the infarct border zone, thereby contributing to protection of nonischemic areas. Sca1<sup>+</sup> cells also have been described as differentiating into fibroblasts that may positively act during the early phase of infarct healing, thereby mediating positive effects on post-MI adverse remodeling.

In our study, we did not use specific cardiac stromal cell markers to define the phenotype of CD45−/CD34−/Sca-1<sup>+</sup> cells found in the myocardium (along with absence of CD31 expression). It must be the goal of future studies to further characterize our population of Sca-1<sup>+</sup> cells and to completely rule out that they are not vessel-related cells or part of the mobilized pool from a noncardiac origin.

After we had established the cardioprotective potential of ECFCs, we further evaluated our ECFC patient collective for cardiovascular risk factors and discovered that a fraction of these patients had diabetes mellitus Typ II (DM Typ II) according to current guidelines. Interestingly, we were not able to detect functional or cellular differences within the ischemic myocardium between animals treated with either ECFCs from diabetic or nondiabetic patients. In this context, Tan et al. [34] injected ECFCs of diabetic and healthy rabbits intramyocardially in diabetic rabbits. They found that injection of transplantation of diabetic ECFCs was not able to rescue the ischemic myocardium of diabetic rabbits. In comparison, transplantation of healthy ECFCs resulted in increased neovascularization and paracrine activation. However, methodological differences might explain the different findings. Tan et al. [34] used recipient animals with severe, untreated diabetes, suggesting that the recipient environment plays an important role for the magnitude of therapeutical effects of cell-based therapies. Likewise, others have reported that the dysfunction of diabetic ECFCs can be restored by improved glycemic control [35,36].

A sole mechanism of action for ECFC beneficial effects remains to be elucidated. In some of these studies, ECFCs facilitate vascular repair by directly integrating within the host vasculature, while in others, vascular engraftment of these cells is low or completely absent [11]. Herein, we could show that ECFC transplantation is associated with a secondary increase in Sca-1<sup>+</sup> cardiac resident progenitor cells and propose a novel proreparative mechanism of ECFC action.

In the present study, we were not able to perform noninvasive and repetitive analyses of the animals because mice were sacrificed after invasive hemodynamic measurements and for histological analyses. Therefore, it will be the goal of future studies to use noninvasive imaging techniques such as cardiac MRI to facilitate serial assessments of cardiac function [37].

The rapid donor cell death 30 days after transplantation seen in our study is consistent with previous studies showing poor donor cell survival after transplantation of different cell types [6]. However, reporter gene PET imaging could be employed in future studies to monitor stem cell survival in vivo. Likewise, molecular imaging will possibly help to unravel the mechanistic puzzle about the contributory roles of inflammation, ischemia, apoptosis, anoikis, and autophagy after cell transplantation [38].
