Long‐term repopulation of aged bone marrow stem cells using young Sca‐1 cells promotes aged heart rejuvenation

Abstract Reduced quantity and quality of stem cells in aged individuals hinders cardiac repair and regeneration after injury. We used young bone marrow (BM) stem cell antigen 1 (Sca‐1) cells to reconstitute aged BM and rejuvenate the aged heart, and examined the underlying molecular mechanisms. BM Sca‐1+ or Sca‐1− cells from young (2–3 months) or aged (18–19 months) GFP transgenic mice were transplanted into lethally irradiated aged mice to generate 4 groups of chimeras: young Sca‐1+, young Sca‐1−, old Sca‐1+, and old Sca‐1−. Four months later, expression of rejuvenation‐related genes (Bmi1, Cbx8, PNUTS, Sirt1, Sirt2, Sirt6) and proteins (CDK2, CDK4) was increased along with telomerase activity and telomerase‐related protein (DNA‐PKcs, TRF‐2) expression, whereas expression of senescence‐related genes (p16INK4a, P19ARF, p27Kip1) and proteins (p16INK4a, p27Kip1) was decreased in Sca‐1+ chimeric hearts, especially in the young group. Host cardiac endothelial cells (GFP−CD31+) but not cardiomyocytes were the primary cell type rejuvenated by young Sca‐1+ cells as shown by improved proliferation, migration, and tubular formation abilities. C‐X‐C chemokine CXCL12 was the factor most highly expressed in homed donor BM (GFP+) cells isolated from young Sca‐1+ chimeric hearts. Protein expression of Cxcr4, phospho‐Akt, and phospho‐FoxO3a in endothelial cells derived from the aged chimeric heart was increased, especially in the young Sca‐1+ group. Reconstitution of aged BM with young Sca‐1+ cells resulted in effective homing of functional stem cells in the aged heart. These young, regenerative stem cells promoted aged heart rejuvenation through activation of the Cxcl12/Cxcr4 pathway of cardiac endothelial cells.


| INTRODUC TI ON
The idea that age itself is one of the major risk factors for the development of cardiovascular disease (CVD) has motivated interest in the field of cardiac aging. Growing clinical and experimental evidence shows that the aging process promotes structural and functional remodeling of the heart, even in the absence of overt CVD (Keller & Howlett, 2016). The current understanding of the mechanisms implicated in age-related CVD includes telomere shortening and cellular senescence (Chen et al., 2014), mitochondrial oxidative stress (Kornfeld et al., 2015), as well as genomic instability and chromatin modifications (Lord & Ashworth, 2012). An increase in senescent cells within the vascular wall and heart contributes to the structural and functional decline of the cardiovascular system with age (Huang, Alhenc Gelas, & Osborne-Pellegrin, 1998). Critical aspects associated with cellular senescence include age-dependent defects in adrenergic signaling and calcium handling which dampen mechanical efficiency and electrophysiological properties, increasing the risk of arrhythmias. Mitochondrial overproduction of reactive oxygen species (ROS) ultimately leads to the formation of highly reactive superoxide or H 2 O 2 , the accumulation and diffusion of which fosters cellular senescence, DNA mutations, inflammation, and the activation of multiple cell death pathways (Camici, Savarese, Akhmedov, & Luscher, 2015). It has been suggested that sporadic genomic mutations accumulated across the lifespan represent a major underpinning for the development of CVD (Shah & Mahmoudi, 2015). These findings, coupled with the growing aged population worldwide, highlight the urgency of understanding how aging promotes CVD in order to develop new treatment strategies.
Previously, we evaluated the importance of the function of the cardiac-resident BM-derived progenitor cell pool on cardiac recovery after injury in aged animals (Li et al., 2013). A BM transplantation model was developed which employed old C57BL/6 mice aged 20-22 months (old recipients) that were lethally irradiated (10.5 Gy) and then immediately received an infusion (through the tail vein) of fresh BM cells (5 × 10 6 ) from C57BL/6-Tg-GFP mice aged 2-3 months (young marrow), generating young chimeras or aged 20-22 months (old marrow), generating old chimeras. At steady state, these young BM cells were able to stably integrate into the aged myocardium within areas with a high density of fibronectin near other non-BM cells. These cardiac-resident BM-derived progenitor cells were characterized as positive for myeloid and mesenchymal markers (CD34, c-Kit, CD45, Tie2, CD44), but negative for cardiac lineage markers (von Willebrand factor, vWF for endothelial cells, smooth muscle actin, SMA for smooth muscle cells, and sarcomeric α-actinin for cardiomyocytes). Furthermore, we were able to create a unique mouse model with young cardiac-resident BM cells in old animals, with old peripheral BM cells. To establish this model, after recovery from primary BM reconstitution, young and old chimeric mice underwent a second lethal irradiation followed by an infusion of BM cells from old female donors. During the second irradiation, the heart was protected with a lead shield carefully positioned on the chest, which prevented the radiation-induced loss of cardiac-resident BM-derived progenitor cells introduced into the heart during the first reconstitution but eliminated the peripheral BM cells, which were subsequently replaced with old BM cells in the second reconstitution. This new cardiac-restricted chimeric protocol allowed us to isolate the origin and capability of these cardiac-resident BM-derived progenitor cells and demonstrate for the first time that their functional capacity before injury determined the extent of ventricular functional restoration after injury by paracrine mechanisms. This minority cardiac cell population elicited its effects through paracrine-directed improvement in survival, proliferation, and neovascularization. Subsequently, we found that stem cell antigen 1 (Sca-1) cells were the key BM cell type involved in the rejuvenation of the aged heart (Li et al., 2017). We isolated Sca-1 + or Sca-1 − cells from the BM of young donor mice and infused them into lethally irradiated old recipients to generate Sca-1 + or Sca-1 − chimeras, respectively. At homeostasis, these BM Sca1 + cells maintained the monocyte/progenitor cell pool in aged heart. After myocardial infarction (MI), these homed BM Sca-1 + cells stimulated proliferation of donor and host progenitor cells in the aged heart which was associated with better restoration of cardiac function. We also evaluated Sca-1 + cell differentiation from 14 days up to 4 months after MI and found that myocardial injury did not initiate their vascular or cardiac differentiation. We further demonstrated that these Sca-1 + subset of young cells improved healing of the aged heart by stimulating cell proliferation through activation of the PDGFRβ-Akt/p27 Kip1 signaling pathway. However, the mechanisms and the exact mediators underlying BM Sca-1 + cell-mediated cardiac rejuvenation of the aged heart remain elusive. In addition, the specific cell type in the aged heart that was rejuvenated by BM Sca-1 + cells has not been identified.
Chronological aging is characterized by biomarkers of cellular senescence, including the upregulation of gene products implicated in growth arrest, the downregulation of rejuvenation-related genes, and attenuation of telomerase activity (Costantino, Paneni, & Cosentino, 2016). To confirm these phenotypic changes at the organ and cellular level, in the present study, we first compared the expression of senescence-related and rejuvenation-related genes in young and old wild-type mouse hearts. Next, we investigated the possible cross-talk between BM Sca-1 + cells and the cardiac aging process using in vitro and in vivo experimental models, including reconstitution of aged mice with BM Sca-1 + or Sca-1 − cells from young or aged GFP + transgenic mice. The effects of Sca-1 + cells on rejuvenation of the aged heart before and after MI were investigated as well as the underlying molecular mechanisms involved. Host cardiac endothelial cells (GFP − CD31 + ) were identified as the primary cell type rejuvenated by young BM Sca-1 + cells. We demonstrate that homed young BM Sca-1 + cells improved aged recipient cardiac endothelial function and promoted heart rejuvenation through activation of the Cxcl12/ Cxcr4 pathway.

| Heart senescence increased with aging
To confirm cellular senescence at the organ level, we compared the expression of senescence-related and rejuvenation-related genes in young and old wild-type mouse hearts without infarction. The expression of senescence-related genes was significantly increased

| Cardiac endothelial cells most susceptible to senescence during aging
To verify whether organ senescence involved cellular senescence of all cell types or a specific cell type, we first conducted immunostaining and found more p16 INK4a -positive cells located on the inner lumen side of blood vessels which were also positive for cardiac en- As shown in Figure 2d, the expression of senescence-related genes was increased and the expression of rejuvenation-related genes was decreased in old compared with young mouse cardiac endothelial cells. Consistent with these findings, protein expression of p16 INK4a and p27 Kip1 was increased, whereas protein expression of CDK2 and CDK4 was decreased in old compared with young mouse cardiac endothelial cells (Figure 2e). In addition, both telomerase activity ( Figure 2f) and telomerase-related protein expression ( Figure 2g) were decreased in aged cardiac endothelial cells. Interestingly, the senescent phenotype observed in the aged cardiac endothelial cells did not appear in other aged cardiac cell types. There was no significant difference between young and old cardiomyocytes ( Figure   S1B,C), fibroblasts ( Figure S2B,C), or vascular smooth muscle cells ( Figure S3B,C) with regard to senescence-or rejuvenation-related gene expression. These findings indicate that cardiac endothelial cells are more susceptible to senescence during the aging process of the mouse heart.

| Young BM Sca-1 cells decreased senescence and improved cellular function of aged recipient cardiac endothelial cells
Next, young or old BM Sca-1 + or Sca-1 − cells were used to reconstitute old recipients. At 4 month after reconstitution, the young Sca-1 + subset had more uniformly replaced the stem and progenitor cells in the BM, blood, and heart of aged recipients than the other three groups. The number of GFP + cells in the BM, blood, and heart was significantly higher in the BM Sca-1 + compared with the Sca-1 − reconstituted groups, especially in the YS + group ( Figure S4). Four months after BM reconstitution, recipient cardiac endothelial cells were isolated from reconstituted mouse hearts using fluorescence- To determine whether BM Sca-1 cells not only decreased cell senescence, but also improved aged endothelial cell function, recipient cardiac endothelial cells (GFP − CD31 + ) were isolated from the reconstituted heart and cultured in vitro. The proliferation ( Figure S5A), migration ( Figure S5B), and tubular formation ( Figure S5C) abilities of recipient cardiac endothelial cells were examined. As shown in Figure S5A, the recipient cardiac endothelial cells from the OS − group had the lowest proliferative ability (BrdU + 5-bromo-2'-deoxyuridine] cells). BM reconstitution with old Sca-1 + (OS + ) or young Sca-1 − (YS − ) cells effectively increased the number of proliferative recipient cardiac endothelial cells, but the effect was greatest in the YS + group.
A similar pattern was found for migration ( Figure S5B) and tubular formation ( Figure S5C). The ratio of phospho-eNOS/total eNOS protein, which is the major mediator of endothelial function, was also increased in the recipient cardiac endothelial cells from the OS + and YS − groups relative to the OS − group with the highest ratio found in the YS + group ( Figure S5D). These data confirm that BM Sca-1 cells not only decreased cell senescence, but also improved the function of aged recipient endothelial cells.

| Young BM Sca-1 cells decreased global senescence of aged recipient hearts
To determine whether by decreasing recipient endothelial cell senescence, BM Sca-1 cells were capable of attenuating global senescence in aged recipient hearts, animal survival and a global senescent phenotype were examined in the four chimeric groups. Four months after BM reconstitution, the YS + chimeras had the highest MI further amplified the expression of all of these regulators, but the differential expression pattern among the four chimeric groups was similar to that observed at baseline.
Among these 11 regulators, Cxcl12 was identified as the most dramatically increased in the GFP + cells isolated from the YS + chimeric hearts, especially after the induction of MI. Cxcl12 is of interest because it plays an important role in BM stem cell retention, mobilization, homing, and survival (Geng et al., 2015). Cxcl12 exerts its effects by binding the receptors Cxcr7 and Cxcr4, for which it is a unique ligand. To determine whether the effects of BM Sca-1 F I G U R E 1 Heart senescence increased with aging. The phenotypic changes in cellular senescence were compared in the hearts of young (Y, 2-3 months) and old (O, 22-23 months) wild-type (C57BL/6) mice. The expression of (a) senescence-related genes p16 INK4a , p19 ARF , and p27 Kip1 and (b)

| D ISCUSS I ON
Previously, we found that functional recovery was improved in old mice after BM reconstitution with young donor cells (Li et al., 2013) and this minority cardiac cell population elicited its effects through paracrine-directed improvement in survival, proliferation, and neovascularization. Subsequently, BM Sca-1 cells were identified as the key BM cell type involved in the rejuvenation of the aged heart (Li et al., 2017). We demonstrated that BM chimerism established in aged mice with a Sca-1 + subset of young cells was associated with better restoration of myocardial progenitors and improved healing of the aged heart by stimulating cell proliferation through activation of the PDGFRβ-Akt/p27 Kip1 signaling pathway. Extending from this previous work, in the present study, we showed that 4 months after the transplantation of BM Sca-1 stem cells into aged recipients, (a) the expression of rejuvenation-related genes and proteins was increased, (b) the expression of senescence-related genes and proteins was decreased, and (c) telomerase activity and telomeraserelated protein expression were increased in Sca-1 + chimeric hearts, especially in the young Sca-1 + group. Recipient cardiac endothelial cells were identified as the primary cell type rejuvenated by young BM Sca-1 + cells. The proliferation, migration, and tubular formation abilities of recipient cardiac endothelial cells were preserved by BM Sca-1 + cells. Homed donor BM cells secreted more growth factors in aged recipient hearts, especially after the induction of MI.
Among the multiple upregulated factors, Cxcl12 was identified as the most dramatically increased factor in the homed donor BM cells isolated from the YS + chimeric hearts, especially after the induction of MI, compared with the other groups. In response to the increased level of Cxcl12, the protein expression of the Cxcr4 receptor and the downstream mediator, Akt, was increased in the recipient cardiac endothelial cells, especially in the young Sca-1 + group. We thus showed that reconstitution of aged BM with young Sca-1 + cells promoted rejuvenation of endothelial cells in the aged heart through activation of the Cxcl12/Cxcr4 pathway.
It has been suggested that chronological age is associated with telomere shortening in cardiac stem cells (CSCs), leading to the inheritance of short telomeres and quick progression to a senescent phenotype in newly formed cardiomyocytes. Senescence of CSCs and myocytes predisposes the development of an aging myopathy.
However, in the current study, we found that cardiac endothelial cells were the primary cell type most susceptible to senescence during mouse heart aging and chronological aging coincided mainly with endothelial senescence. We postulated that the status of endothelial cells, which may originate from c-Kit + cells during development, was the major determinant of cardiac senescence and aging.
Indeed, several recent preclinical studies have established endothelial dysfunction as one of the key vascular modifications that occurs during aging resulting in a predisposition for cardiovascular disease (Lakatta & Levy, 2003). Therefore, rejuvenation of aged endothelial cells could be a means by which to counteract cardiac senescence and aging. In fact, we found that BM Sca-1 cells, through decreasing endothelial senescence and improving endothelial function, effectively decreased global senescence in aged recipient hearts. Recently, we evaluated the importance of the function of cardiac-resident BM-derived progenitors, especially the Sca-1 + cells on cardiac recovery after injury in aged animals (Li et al., 2017). At homeostasis, these BM Sca1 + cells maintained the monocyte/progenitor cell pool in aged mouse heart. After MI, these homed BM Sca-1 + cells stimulated proliferation of donor and host progenitor cells in the aged heart which was associated with better restoration of cardiac function. However, these cardiac-resident BM-derived cells did not participate directly in angiogenesis or cardiomyogenesis through differentiation. We believed that the improvement in cardiac function resulted from stimulation of other cells within their sphere of influence through paracrine factors, thereby amplifying their beneficial effects beyond their number and location. Indeed, we found a variety of cardioprotective factors (Cxcl12, Vegfa, Fgf 6, Gdf5) were significantly higher in the Sca-1 + subset. We showed that a significant increase in rejuvenating and angiogenic factors combine through downstream AKT activity and the related pro-survival and pro-vascular factors contributed to the protection of the heart.
Although the mechanisms and the exact mediators underlying BM Sca-1 + cell-mediated cardiac rejuvenation of the aged heart remain elusive, we believe this minority cardiac cell population elicited its effects through paracrine-directed improvement in rejuvenation, survival, proliferation, and neovascularization.

| CON CLUS IONS
We demonstrated the effects of BM Sca-1 + cells on rejuvenating the aged heart using in vivo BM reconstitution and in vitro cell functional assays. We identified Cxcl12 as a key modulator of the interaction between homed BM Sca-1 + cells and the Cxcr4 receptor in recipient cardiac endothelial cells which led to activation of the senescencesuppressing signal pathway AKT-FoxO3a-P27. Activation of this pathway contributed to decreased recipient cardiac endothelial cell senescence and improved cellular survival and function, enabling the global rejuvenation of the aged heart following BM reconstitution with Sca-1 + cells.

| Animal procedures
The F I G U R E 5 Expression profile of growth factors in cardiac homed donor BM cells. Four months after BM reconstitution, myocardial infarction (MI) was induced and donor BM cells (GFP + ) were isolated from chimeric hearts at baseline (control without MI) and at 3 days post-MI. (a) Growth factor qPCR array to profile 84 different growth factors in control groups. Eleven factors were differentially expressed in the homed donor BM cells from the four chimeric groups. The 11 factors were further validated by real-time qPCR at control as well as at 3 days post-MI for the4 chimeric groups. These 11 factors were clustered into 5 functional subcategories: (b) cell differentiation regulators (Bmp8a, Bmp5, Lep), (c) apoptosis regulators (Gdf5), (d) angiogenic growth factors (Fgf6, Vegfa), (e) development controllers (Cxcl12,Gdnf,Nodal), and (f) others (IL2, and Tdgf1). n = 3/group for the growth factor qPCR array; n = 6/group for the other experiments; **p < .01 YS + vs. other group; ## p < .01 OS + vs. OS − ; # p < .05 OS + vs. OS − al., 2017). All aged chimeric mice were housed for 4 months to allow young/old stem cells to repopulate the BM and home to the heart prior to further experimentation. Thus, coronary occlusion was performed on chimeric mice 4 months after BM reconstitution and heart tissues were used for various analyses at the gene, protein, and cellular function levels. mice 4 months after BM reconstitution, as previously reported (Li et al., 2013). In brief, mice were intubated and ventilated with 2% isoflurane and given buprenorphine (0.05 mg/kg, S.C.) for analgesia.
Through a thoracotomy, the pericardium was dissected and the left anterior descending (LAD) coronary artery was ligated.
More detailed experimental procedures can be found in the Supporting Information file.

| Statistics
All values are expressed as mean ± SD. Analyses were performed using GraphPad InStat software . Student's t test was used for twogroup comparisons. Comparisons of parameters among three or more groups were analyzed using one-way analysis of variance (ANOVA) followed by Tukey or two-way ANOVA with repeated measures over time followed by Bonferroni post hoc tests for multiple comparisons.
Differences were considered statistically significant at p < .05.