Bone Marrow Derived Pluripotent Stem Cells in Ischemic Heart Disease: Bridging the Gap Between Basic Research and Clinical Applications

Ahmed Abdel-Latif1, Ewa Zuba-Surma2 and Mariusz Z. Ratajczak3 1Gill Heart Institute and Division of Cardiovascular Medicine, University of Kentucky, and Lexington VA Medical Center, Lexington, KY 2Department of Medical Biotechnology, Faculty of Biochemistry, Biophysics and Biotechnology, Jagiellonian University, Cracow, 3Stem Cell Biology Institute, James Graham Brown Cancer Center, University of Louisville, Louisville, KY, 1,3USA 2Poland


Introduction
The prevalence of ischemic heart disease and acute myocardial infarction (AMI) has increased to alarming rates in the United States and the western world (Roger et al., 2011).Patients who survive the initial AMI suffer ischemic cardiomyopathy (ICM) which is often complicated by high mortality and poor overall prognosis (Braunwald et al., 2000;McMurray et al., 2005).Despite significant advances in medical therapy and revascularization strategies, the prognosis of patients with AMI and ischemic cardiomyopathy remains dismal (Levy D et al., 2002;Roger VL et al., 2004).The last decade has demonstrated significant progress and rapid translation of myocardial regenerative therapies particularly those utilizing stem cells isolated from adult tissues (Abdel-Latif et al., 2007).
Studies examining the potential therapeutic use of bone marrow (BM)-derived cells in myocardial regeneration have overshadowed the growing evidence of innate cardiac reparatory mechanisms.Several studies have demonstrated the capability of cardiomyocytes to replenish through poorly understood innate mechanisms.Follow up of cardiac transplantation patients have demonstrated continuous replenishment of cardiomyocytes by recipient's derived cells through poorly understood mechanisms (Quaini et al., 2002).There is growing evidence that BM-derived cells are responsible, at least in part, for organ chimerism including cardiomyocyte chimerism (de Weger et al., 2008;Deb et al., 2003).Animal studies have confirmed this to be a dynamic process responding to significant injury such as myocardial infarction and peaks in the peri-infarct zone (Hsieh et al., 2007).Although this process appears to be robust enough to achieve the renewal of approximately 50% of all cardiomyocytes in the normal life span, very little is known about its underpinnings (Bergmann et al., 2009).
Complex innate reparatory mechanisms are initiated by myocardial ischemia interacting with different elements of the immune system, the infarcted myocardium and bone marrow stem cells, culminating in BM-stem and progenitor cells' (SPCs) mobilization as we and others have demonstrated (Abdel-Latif et al., 2010;Walter et al., 2007;Wojakowski et al., 2009).However, very little is known about the mechanisms and clinical significance of this mobilization.Animal studies show that mobilized BM-derived cells (BMCs) repopulate the infarct border, however the significance of this mobilization is unclear given the low rate of their differentiation into cardiomyocytes (Fukuhara et al., 2005).

Isolation and functional characteristics of BM-derived pluripotent stem cells
The bone marrow acts as a reservoir for a heterogeneous pool of tissue-committed and non-committed stem cells.These populations contain progenitors that aid in the chimerism and cellular turnover of different organs as well as very rare populations of pluripotent and non-committed stem cells.The old dogma that adult tissues lack pluripotent stem cells (PSCs) has been continuously challenged during the last decade through multiple studies that isolated PSCs from adult humans' and animals' tissues.These populations were distinguished based on their morphology with small cell size, large nucleus demonstrating euchromatin and large nucleus to cytoplasm ratio.Furthermore, cell surface markers as well as nuclear transcription factors, such as SSEA1/4, Oct4 and Nanog, have been deployed.
Very small embryonic like stem cells (VSELs) represent a rare yet pluripotent population of adult stem cells.They have been initially described by Dr. Ratajczak's group in the murine BM based on their expression of Sca1 (murine stem cell marker) and lack of expression of CD45 (pan-leukocytic marker) and differentiated lineage (Lin) markers (Kucia et al., 2006;Zuba-Surma et al., 2008).Following their isolation from murine tissues, VSELs were subsequently isolated from human BM, umbilical cord blood (CB) and peripheral blood based on the lack of expression of Lin/CD45 and the expression of the stem cell markers CD133, CXCR4 and CD34. Figure 1 illustrates the flow cytometry protocol for identifying and isolating VSELs from murine and human samples.VSELs were further characterized using a multi-dimensional approach comprising molecular, protein and cell imaging techniques to confirm their pluripotent features (Zuba-Surma et al. 2008).VSELs are morphologically similar to embryonic stem cells demonstrating small diameter compared to more committed progenitors/stem cells with large nucleus containing open-type chromatin surrounded with thin rim of cytoplasm and multiple mitochondria (Zuba-Surma et al. 2008).
VSELs exhibit multiple embryonic and pluripotent surface and nuclear embryonic markers such as Oct4, SSEA1/4, Nanog, and Rex1.In vivo and in vitro studies have demonstrated the capability of VSELs to differentiate into multiple cell lines across germ lines including cardiomyocytes (Kucia et al. 2006).
The bone marrow harbors other multi-and pluri-potent stem cell populations such as the mesenchymal stem cells (MSC) (Hattan et al., 2005;Kawada et al., 2004), multipotent adult progenitor cells (MAPC) (Jiang et al., 2002), and marrow-isolated multilineage inducible cells (MIAMI) (D 'Ippolito et al., 2004).Similar populations with cardiac differentiation potential have been also isolated from skeletal muscle and other tissues (Abdel-Latif et al., 2008a).However, it is conceivable that different investigators have isolated, using different methods, the same or very similar populations and named them differently.It is also possible that these populations at least in part contain VSELs which might explain their pluripotent potential.
Panel A: Gating strategy for isolating human cord blood (CB)-derived VSELs.Morphology of total CBderived nucleated cells is shown on dot-plot representing FSC and SSC parameters related to cell size and granularity/ complexity, respectively.All objects larger than 2m are enclosed in region R1 and further visualized on histogram showing the expression of markers of mature hematopoietic cells (lineage markers; Lin).Cells not expressing differentiated hematopoietic markers (Lin-in region R2) are then analyzed for CD34 and CD45 expression.VSELs are identified as CD45-/Lin-/CD34+ cells (region R3), while hematopoietic stem cells (HSCs) as CD45+/Lin-/CD34+ cells (region R4).Panel B: Sorting of murine bone marrow (BM)-derived VSELs.Morphology of total murine BM-derived nucleated cells is shown on dot-plot presenting FSC and SSC parameters and all objects in range of 2-10m in diameter are included in region R1.Lymphocytic cells including stem cell fraction is further analyzed for Sca-1 and differentiated hematopoietic lineages markers (Lin) expression and only Sca-1+/Lin-cells are included in region R2.Cells from this region are further seperated based on CD45 expression.Murine VSELs are distinguished as CD45-/Lin-/Sca-1+ cells (region R3), while HSCs as CD45+/Lin-/Sca-1+ cells (region R4).Fig. 1.Strategy for flow cytometric analysis of human and murine Very Small Embryonic-Like and hematopoietic stem cells.Briefly, BM is flushed from the femurs and tibias.Nucleated cells are isolated by lysis of red blood cells and cells are then gated on based on the cell size (>2 m).Of note, lysis is preferred for isolating VSELs rather than Ficoll gradient that we have shown to lose some of the VSELs due to their small size.

BM-derived pluripotent stem cells are mobilized in the peripheral circulation following myocardial ischemia in animal models and humans
Myocardial ischemia, particularly large myocardial infarction, produce multiple stimuli include various chemokines, cytokines, kinins, bioactive lipids and members of the complement cascade, that lead to the mobilization and subsequent homing of BMSPCs.Indeed, several reports have confirmed that mobilization of stem cells originating from the BM occurs in response to myocardial ischemic injury (Grundmann et al., 2007;Kucia et al., 2004b;Leone et al., 2005;Massa et al., 2005;Shintani et al., 2001;Wojakowski et al., 2006) and heart failure (Valgimigli et al., 2004).Similar observations were noted in patients with acute neurological ischemia (Paczkowska et al., 2005) and patients with extensive skin burn (Drukala et al., 2011).
Stimuli responsible for the mobilization and homing of BMSCs in the setting of myocardial ischemia show similarities and differences with those involved in hematopoietic stem cells (HSCs) homing to the BM.The role of stromal cell derived factor (SDF-1) and its receptor (CXCR4) axis in the retention of hematopoietic stem/progenitor cells (HSPCs) in bone marrow is undisputed (Kucia et al., 2005;Lapidot et al., 2002), however, its role in the mobilization and homing of BM-SPCs to a highly proteolytic microenvironment, such as the ischemic/infarcted myocardium, is somewhat less certain.Studies have demonstrated that multiple members of the metalloproteinases (MMP) family, such as MMP2, MMP9 and MMP13, are upreagulated in the myocardium following infarction (Peterson et al., 2000).The elevated levels of the MMPs contribute to the degradation of chemokines such as SDF-1 and the byproduct of this degradation acts as an inhibitor the sole SDF-1 receptor, CXCR4 (McQuibban et al., 2001;McQuibban et al., 2002).In support of this hypothesis, Agrawal et al demonstrated that the conditional deletion of CXCR4 in cardiomyocytes did not influence the recovery of left ventricular (LV) function, reduce the scar size or alter the homing of MSCs to the myocardium following myocardial infarction (Agarwal et al., 2010).Thus, there is growing evidence that other mechanisms beside the SDF-1/CXCR4 axis are contributing to the mobilization and homing of BM-SPCs in AMI and other conditions (Jalili et al., 2010;Ratajczak et al., 2010).These data suggest an important interplay between the complement cascade, the immune system, cathelicidins, low levels of SDF-1, and sphingosine-1 phosphate (S1P) and other bioactive lipids in the mobilization and homing of HSPCs.Our preliminary data suggest that these complex interactions might be involved in the mobilization of BM-SPCs in acute myocardial ischemia as well (unpublished data).Clinically, pharmacological modulators of S1P receptors are already approved by the FDA and can be utilized to enhance BM-SPC mobilization in the setting of ischemic heart disease.Similarly, modulation of the complement cascade can be also utilized in this process similar to their role in the mobilization of HSPCs.
The first evidence for the mobilization of CD34+ mononuclear cells in AMI was demonstrated by Shintani et al (Shintani et al. 2001).The authors demonstrated successful in vitro differentiation of circulating BM-SPCs into endothelial cells that expressed CD31, VEcadherin and the kinase insert domain receptor (KDR) (Shintani, et al. 2001).Leone et al demonstrated that the levels of circulating CD34+ cells in the setting of AMI were higher when compared to patients with mild chronic stable angina and healthy controls.The magnitude of CD34+ cell mobilization correlated with the recovery of regional and global LV function recovery as well as other functional LV parameters (Leone et al. 2005).Similarly, Wojakowski et al demonstrated the mobilization of multiple BM-SPCs populations in patients with AMI and found significant correlation between the number of circulating CD34+ cells and plasma SDF-1 levels (Wojakowski et al., 2004).In their following publication, the authors demonstrated the correlation between circulating BM-SPCs and ejection fraction at baseline and lower brain natriuretic peptide (BNP) levels (Wojakowski et al. 2006).Interestingly, the mobilization of BM-SPCs is reduced by the successful revascularization of the culprit vessel in acute STEMI (Müller-Ehmsen et al., 2005).However, the majority of the above mentioned studies have focused on the mobilization of partially committed stem cells such as HSPCs and endothelial progenitor cells (EPCs).
We and others have demonstrated the mobilization of pluripotent stem cells (PSCs) including VSELs in the setting of myocardial ischemia (Abdel-Latif et al. 2010;Wojakowski et al. 2009).The number of circulating VSELs was highest in patients with ST-elevation myocardial infarction (STEMI), particularly in the early phases following the injury, when compared to patients with lesser degrees of ischemia such as non STEMI (NSTEMI) and those with chronic ischemic heart disease (Abdel-Latif et al. 2010).The mobilization of PSCs appears to be related to the extent of myocardial ischemia and the degree of myocardial damage.Moreover, the ability of patients to mobilize PSCs in the peripheral circulation in response to AMI decreases with age, reduced global LV ejection fraction (LVEF) and diabetes supporting the notion of an age/comorbidity related decline in the regenerative capacity (Abdel-Latif, et al. 2010;Wojakowski, et al. 2009).Indeed, animal models confirm the reduction of number as well as pluripotent features of BM-derived VSELs with age (Zuba-Surma et al. 2008).Similarly, studies have demonstrated the reduction of number as well as functional capacity of EPCs in diabetic patients (Fadini et al., 2006).
The pluripotent features of mobilized VSELs, including the presence of octamer-binding transcription factor-4 (Oct4) and stage specific embryonic antigen-4 (SSEA4), were confirmed both on the RNA and protein levels.Utilizing the capabilities of the ImageStream system, we demonstrated that circulating VSELs during AMI have very similar embryonic features similar to their BM and CB counterparts including the small size (7-8 m), large nucleus and high nuclear-to-cytoplasm ratio (Figure 2).Furthermore, circulating VSELs during AMI express markers of early cardiac and endothelial progenitors that suggest that the mobilization is rather specific and that circulating VSELs are destined to aid in the myocardial regeneration following injury (Abdel-Latif et al. 2010;Kucia et al. 2004b;Wojakowski et al. 2009).
The above evidence suggest an innate, yet poorly understood, reparatory mechanism that culminates in the mobilization of BMSCs including pluripotent and embryonic like stem cells in acute myocardial injury.This mobilization correlates with the recovery of LV function and other LV functional parameters.Therefore, mobilization of PSCs in myocardial ischemia is a relevant and clinically significant process.Future studies aiming at selective mobilization of PSCs rather than the non-selective actions of agents such as granulocyte colony stimulating factor (G-CSF) may prove beneficial in the field of myocardial regeneration.
Indeed, there is evidence that the mobilization of CXCR4+ cells in the setting of AMI is correlated with LVEF recovery as well as myocardial reperfusion when assessed with cardiac MRI in humans (Wojakowski, et al. 2006).Fig. 2. Representative ImageStream images of VSEL and hematopoietic stem/ progenitor cell (HSPC) circulating in peripheral blood following acute ST-elevation myocardial infarction.Cells were stained against: 1) hematopoietic lineages markers (Lin) and CD45 to be detected in one channel (FITC, green), 2) marker of pluripotency Oct4 (PE, yellow) and 3) stem cell antigen CD34 (PE-Cy5, cyan).Nuclei are stained with 7-aminoactinomycin D (7-AAD, red).Scales represents 10 m.VSELs are identified based on the lack of expression of both Lin and CD45 markers and positive staining for CD34 antigen and nuclear appearence of Oct-4 transcription factor (Upper Panel).HSCs are identified as cell expressing Lin and/ or CD45 markers as well as CD34 antigen; however, negative for Oct-4 (Lower Panel).BF: Bright field.

Therapeutic mobilization of BM-derived stem cells in myocardial regeneration
Hematologists have used the concept of BM-derived stem cell mobilization using pharmacological agents such as G-CSF for a long time.Based on the available clinical experience and safety profile of these therapies, pharmacological stem cell mobilization in the setting of AMI has gained increasing enthusiasm.Mobilized BM-SPCs are either harvested for further transplantation or allowed for spontaneous homing to the infarcted myocardium and has demonstrated various degrees of success (Engelmann et al., 2006;Ince et al., 2005;Zohlnhöfer et al., 2006).Similar to BMCs transplantation studies, the heterogeneous methodologies of the included studies diluted the effect.The overall lack of efficacy with G-CSF BMCs mobilization in the setting of acute myocardial infarction is somewhat incongruent with the salutary effects of BMCs transplantation in humans and G-CSF therapy in animal models for myocardial regeneration.
The largest study utilizing G-CSF in the setting of acute myocardial infarction was the REVIVAL-2 trial that included 114 patients (Zohlnhöfer, et al. 2006).The study randomized AMI patients to 10 g/kg of G-CSF vs. placebo and left ventricular functional parameters were assessed using cardiac MRI (CMR).The study demonstrated no significant difference in the tested parameters between patients treated with G-CSF or placebo.However, baseline characteristics in the study population showed normal or near normal LV function and therefore the expected benefit is minimal.Patient selection was a methodological flaw that plagued some of the studies that utilized G-CSF.Indeed, with careful examination of the available literature, patients with reduced LV function at baseline as well as those treated within the first 36 hours following AMI benefited the most (Abdel-Latif et al., 2008b;Achilli et al., 2010).On the other hand, safety concerns regarding a potentially increasing evidence of instent restenosis (Kang et al., 2004) and recurrent ischemia (Hill et al., 2005) have halted subsequent clinical trials.However, it is important to note that these safety concerns were not confirmed in large studies (Zohlnhöfer, et al. 2006) or in the cumulative meta-analyses (Abdel-Latif et al. 2008b).
Beyond the methodological flaws encountered in human trials, this lack of efficacy can be explained by multiple factors.While G-CSF and similar therapies mobilize a wide array of BMSPCs in the peripheral blood, homing factors may not be sufficient to guide them to the myocardial infarct zone.Indeed, the homing of c-Kit+ cells to the infarcted myocardium improved when G-CSF therapy was combined with local administration of SDF-1 (Askari et al., 2003).The myocardial levels of chemoattractants peaks within 24-72 hours following injury (Kucia et al., 2004a;Ma et al., 2005;Wang et al., 2006) and therefore delayed therapy in some human trials may have missed the homing window to the infarct zone.Similarly, the addition of Flt-3 to G-CSF therapy improved outcomes in animal models (Dawn et al., 2008).Moreover, different cytokines are known to preferentially mobilize somewhat different subsets of BMCs (Hess et al., 2002;Neipp et al., 1998).Future studies investigating the characteristics of G-CSF-mobilized cells will be necessary to glean additional mechanistic insights in this regard.
Recently, a combined approach with stem cell mobilization and enhanced homing using therapies known to increase local SDF-1 or CXCR4 antagonists have been proposed and is currently being tested (Jujo et al., 2010;Zaruba et al., 2009).Going forward, the beneficial effects of BM-derived stem cell mobilization may be augmented by selective mobilization of undifferentiated BMSCs rather than differentiated inflammatory cells.It is also important to remember that some of the G-CSF arbitrated effects can be mediated by its direct effect on cardiomyocytes which are known to express G-CSF receptor (Shimoji et al., 2010).G-CSF therapy may be inducing the proliferation of cardiomyocytes or the differentiation of resident cardiac stem cells.On a similar note, G-CSF therapy upregulates Akt (Ohtsuka et al., 2004) and may result in reducing apoptosis of ischemic cardiomyocytes if utilized early following the acute event.

BM-derived stem cell transplantation for myocardial repair
The use of BM-derived cells in myocardial regeneration has moved rapidly from the basic research lab to the clinical arena.The results from these studies varied widely probably secondary to the heterogeneous methodologies used with an overall marginal benefit with BM-derived cell transplantation compared to placebo.The majority of studies, however, utilized unselected populations of BMCs and these studies provide the longest follow-up of up to 5 years (Assmus et al., 2010;Schachinger et al., 2009).The underling mechanisms leading to the beneficial effect of transplanted BMCs are unclear.The observed benefits of BMCs transplantation is out of proportion of the observed rates of newly formed cardiomyocytes from BMCs' origin (Zuba-Surma et al., 2011).Indeed, recent evidence suggest a primarily paracrine effect of BM-derived stem cells following their transplantation by recruiting and stimulating resident cardiac stem cells (CSCs) (Loffredo et al., 2011).Furthermore, human purified CD34+ cells are a source of several growth factors including VEGF, cytokines and chemokines that may prevent apoptosis of dying cardiomyocytes and promote angiogenesis in damaged myocardium (Majka et al., 2001).Cell membrane derived microvesicles or exosomes that are enriched in S1P may contribute to regeneration of myocardium and its re-vascularization (Baj-Krzyworzeka et al., 2002).Hence, transplanted CD34+ cells may contribute to regeneration of damaged heart by paracrine signals and released microvesicles (Ratajczak et al., 2008) and was recently confirmed by others (Sahoo et al., 2011).
Long-term follow-up studies demonstrated 'catch-up phenomenon' of the placebo treated patients, thus leading to mixed results regarding the sustainability of the BMCs treatment benefit (Assmus, et al. 2010;Meyer et al., 2009;Yousef et al., 2009).The benefit of BMCs therapy is less robust among patients with chronic ischemic heart disease (IHD) (Assmus et al., 2006;Strauer et al., 2010).Similarly, smaller studies have demonstrated the antianginal effects of BMCs in patients with non-revascularizable severe coronary artery disease (Losordo et al., 2007;Tse et al., 2007).
Selected BM-derived stem cell subpopulations represent an attractive substrate for cellular therapies since they lack the inflammatory cells, which contribute to the ongoing inflammatory response at the site of myocardial infarction, contained in the unselected BMCs populations.Furthermore, highly purified stem cell populations are more likely to induce myocardial regeneration through paracrine effects or by directly differentiating into cardiomyocytes.The largest study utilizing selected BM-derived stem cell population is the REGENT study which compared selected to non-selected populations of BMCs in patients with acute ischemic heart disease and reduced LV function at baseline (Tendera et al., 2009).While there were no significant differences between the groups, patients treated with selected CD34+/CXCR4+ cells showed trends of improvement in LV function when compared to controls.Other studies utilizing primitive populations of BM-SPCs such as CD133+ cells have reported improvement of LV function and perfusion (Bartunek et al., 2005;Stamm et al., 2004).
Nevertheless and despite the disparity in the methodologies of the conducted studies, the overall collective effect of BMCs' transplantation suggests small yet statistically significant benefit in myocardial regeneration (Abdel-Latif, et al. 2007;Martin-Rendon et al., 2008).However, these trials have been hampered by their reliance on surrogate endpoints rather than patient important endpoints such as mortality, need for repeat revascularization, recurrent MI or re-hospitalization for congestive heart failure.While surrogate endpoints are important for mechanistic studies, patient-important endpoints are quintessential for a new therapy to achieve mainstream status.

Future directions
Growing evidence suggest that a multitude of BM-derived stem and pluripotent stem cells are mobilized in the peripheral blood following AMI.However, the clinical significance and the potential therapeutic use of this mobilization are still not fully understood.Circulating PSCs can be used as markers of ischemic injury in humans or as predictors of myocardial recovery following large ischemic damage.On the other hand, the therapeutic application of VSELs in myocardial regeneration has proven beneficial although the beneficial mechanisms remain elusive and are probably mainly paracrine in nature.Given the pluripotent potential of VSELs, their transplantation at smaller numbers (10,000 cells per mouse) have proven to be more beneficial than larger numbers of the more committed HSPCs (100,000 cells per mouse) indicating their greater therapeutic potential (Dawn et al. 2008).Current efforts directed at the ex-vivo expansion and priming of VSELs have proven to be a successful strategy in animal models and their clinical applications are pending (Dawn et al. 2008;Zuba-Surma, et al. 2011).Nuclear reprogramming has opened the door for creating patientspecific autologous pluripotent stem cells with multiple therapeutic opportunities (Takahashi et al., 2006).Further studies are needed to examine the feasibility as well as the safety of inducible pluripotent stem cells (iPS) particularly their tumorigenicity and immunogenicity before they can be explored in human studies.
On the biotechnology frontier, multiple modifications of the transplanted cells (priming) and the host environment are being tested in humans to improve the efficiency of BM-SPCs' regenerative capacity.Transplanting three dimensional constructs that provide an enriched environment for the transplanted and resident stem cells are attractive modifications to the currently tested protocols [reviewed in (Mooney et al., 2008)].Similarly, the concept of multiple doses of stem cell to repair the complex process of myocardial remodeling following acute myocardial infarction is gaining traction and is very appealing.While the field of stem cell regenerative therapy for ischemic heart disease is still in its infancy, the accelerated advances in a wide array of biological and biotechnological areas have rapidly propelled the field from the bench to clinical applications.
Dr. Abdel-Latif is supported by the University of Kentucky Clinical and Translational Science Pilot Award and the University of Kentucky Clinical Scholar program.Dr. Zuba-Surma is supported by the "Polish Foundation of Science" homing program grant number 2008/15.Dr. Ratajczak is supported by NIH grant R01 CA106281, NIH R01 DK074720, and Stella and Henry Endowment.We thank Dr. Karapetyan for her technical support for this review.Conflict of interest: None.