Age-Related Impaired Efficacy of Bone Marrow Cell Therapy for Myocardial Infarction Reflects a Decrease in B Lymphocytes

Abstract

Treatment of myocardial infarction (MI) with bone marrow cells (BMCs) improves post-MI cardiac function in rodents. However, clinical trials of BMC therapy have been less effective. While most rodent experiments use young healthy donors, patients undergoing autologous cell therapy are older and post-MI. We previously demonstrated that BMCs from aged and post-MI donor mice are therapeutically impaired, and that donor MI induces inflammatory changes in BMC composition including reduced levels of B lymphocytes. Here, we hypothesized that B cell alterations in bone marrow account for the reduced therapeutic potential of post-MI and aged donor BMCs. Injection of BMCs from increasingly aged donor mice resulted in progressively poorer cardiac function and larger infarct size. Flow cytometry revealed fewer B cells in aged donor bone marrow. Therapeutic efficacy of young healthy donor BMCs was reduced by depletion of B cells. Implantation of intact or lysed B cells improved cardiac function, whereas intact or lysed T cells provided only minor benefit. We conclude that B cells play an important paracrine role in effective BMC therapy for MI. Reduction of bone marrow B cells because of age or MI may partially explain why clinical autologous cell therapy has not matched the success of rodent experiments.

Implantation of bone marrow cells into mouse hearts after myocardial infarction is therapeutic, but if the cells are from donors that are older or post-MI (mimicking autologous cell therapy), they are less effective. This report presents evidence that a decrease in B lymphocytes is responsible for the reduced therapeutic response.

Introduction

Cardiovascular disease including myocardial infarction (MI) has become the leading cause of mortality and morbidity in the world.1, 2 Despite advances in revascularization for MI patients, post-MI remodeling still leads to a substantial heart failure burden, and cell therapy is viewed as a potential approach to regenerate, or at least to preserve, viable myocardium. Experimental cell therapy approaches have been based on a variety of cell types,³ including bone marrow mononuclear cells, mesenchymal stem cells (MSCs), resident cardiac stem or progenitor cells, cardiosphere-derived cells, stem cells from other tissues such as adipose tissue, embryonic stem cells, and myoblasts,4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 with and without genetic modification to increase functionality.17, 18, 19 We and others reported that treatment with secreted products or cell lysates from various populations of bone marrow cells (BMCs) substantially improves post-MI cardiac function in animal models.20, 21, 22, 23 While considerable controversy has surrounded claims of bone marrow-derived regenerated myocardium,²⁴ therapeutic benefits in rodent models have been frequently observed regardless of whether endogenous myocardial regeneration is involved.²⁵

However, attempts to translate this approach to the clinic have led to modest functional improvement in some trials and poor efficacy of cell therapy in others.5, 8, 26, 27, 28, 29, 30, 31, 32, 33, 34 Notably, while laboratory experiments typically use young healthy mice as BMC donors, MI patients undergoing autologous cell therapy (i.e., they receive their own BMCs) are older and are post-MI. Recently, in two high-profile clinical trials from the NIH Cardiovascular Cell Therapy Research Network (CCTRN), acute MI patients were treated with their own BMCs 3 and 7 days post-MI (TIME trial)⁸ and 2–3 weeks post-MI (LateTIME trial).³² These two clinical trials were based, in part, on observations from the REPAIR-AMI trial²⁹ that patients receiving their own BMCs 5–7 days post-MI fared better than those at 3–5 days post-MI, which was attributed chiefly to the state of the recipient heart rather than the BMCs. However, neither TIME nor LateTIME were successful.

We previously reported that BMCs from mice that are themselves post-MI or old are less therapeutic than BMCs from healthy mice or young controls.35, 36 We found that MI causes acute inflammation mediated by interleukin-1 leading to changes in BMC composition, including a decrease in B lymphocytes, and contributes to the poor efficacy of post-MI donor BMCs. Interestingly, BMCs became progressively less therapeutic over the first week post-MI, as the inflammatory response developed, and then became more therapeutic by 21 days post-MI as acute inflammation was subsiding. These results support the hypothesis that the 3–7 days post-MI BMCs in the TIME trial were still in the inflammatory state and problematic for therapy, while the 2–3 weeks post-MI BMCs in LateTIME may have actually regained some of their therapeutic capacity, but not enough to affect the outcome.

Intriguingly, B lymphocytes have been reported to be a therapeutic BMC subpopulation,³⁷ and B cell depletion and restoration have been shown to increase and reduce infarct size in a cerebral artery occlusion model.38, 39, 40 In addition, the progressive influence of advancing age on the therapeutic efficacy of BMCs is relatively unexplored. Given the similar effects of donor age and MI on BMC therapeutic efficacy, we asked whether age also alters bone marrow B cell levels, potentially accounting for the impaired therapeutic potential of BMCs from aged donors. We show here that increasing donor age drastically and progressively impairs donor BMC therapeutic efficacy, irrespective of the state of the recipient heart, and that advancing donor age impairs therapeutic potential of BMCs by B cell reduction. We also show that partial depletion of B cells from young healthy donor BMCs reduces the BMCs’ therapeutic effect, mimicking the effects of donor age or MI. Furthermore, we demonstrate that injection of isolated bone marrow B cells or their lysate can reproduce the therapeutic effect of unfractionated BMC therapy for MI. Bone marrow B cell reduction by age or MI may explain why autologous BMC therapy in patients has been less successful in clinical trials than in rodent experiments.

Results

Progressive Impairment of BMCs Derived from Progressively Older Mice

To confirm whether donor advanced age impairs the therapeutic potential of BMCs for the treatment of MI, we implanted donor BMCs into groups of infarcted recipient mice, keeping the recipient conditions constant, but varying the donor conditions with increasing ages (10 weeks, 6 months, 1, 1.5, and 2 years) and Hank’s balanced salt solution (HBSS)/BSA vehicle control. Injection of BMCs from donors of increasing ages resulted in progressively lower day 28 recipient left ventricular ejection fraction (LVEF) and progressively larger day 28 recipient end-systolic volume (ESV) and end-diastolic volume (EDV) (Figure 1A). Moreover, injection of BMCs from donors of increasing ages also led to progressively lower wall thickness (Figure 1B) and larger infarct size on day 28 (Figure 1C), despite the recipients being of constant age. Linear regression analysis revealed a negative correlation between donor age and recipient ejection fraction (EF) change from day 2 to day 28 post-MI, as well as a positive correlation between donor age and recipient infarct size on day 28 (Figure 1D). Notably, BMCs from 2-year-old donor mice were completely devoid of therapeutic potential, comparable with the negative control vehicle injections, despite consistently good viability (96% assessed by trypan blue staining) in all BMCs assessed before implantation.

Figure 1.

Progressive Impairment of BMCs Derived from Progressively Older Mice

Donor BMCs were harvested from mice with increasing ages (from 10 weeks, 6 months, 1 year, 1.5 years, to 2 years). Increasing donor age led to progressively (A) lower recipient EF and larger ESV and EDV on day 28 post-MI, (B) reduced anterior wall thickness, and (C) increased infarct size on day 28. Linear regression analysis (D) for correlations between donor age and both recipient LVEF change from day 2 to day 28 and infarct size indicated that increase in donor age predicted a decrease in recipient LVEF change from day 2 to day 28 post-MI (p < 0.0005) and an increase in day 28 recipient infarct size (p < 0.0005). All data are means ± SD. (B) ^@p < 0.05 versus 10 weeks, *p < 0.05 versus 6 months, ^Δp < 0.05 versus 1 year, ^#p < 0.01 versus 10 weeks, and ^§p < 0.01 versus 6 months; (C) ^Δp < 0.05 versus 6 months, *p < 0.05 versus 1.5 years, ^ˆp < 0.01 versus 6 months, ^#p < 0.01 versus 1 year, ^@p < 0.01 versus 1.5 years, ^§p < 0.001 versus 10 weeks, ^†p < 0.001 versus 6 months, and ^‡p < 0.001 versus 1 year. Data are summarized numerically in Table S2.

Reduction of Bone Marrow CD19⁺ B Cells from Old Donors

To better understand the differences between BMC populations in young and old mice, we quantitatively analyzed by flow cytometry BMCs harvested from donor mice at ages of 10 weeks, 1.5 years, and 2 years. Increasing age was associated with a significant decrease in the number and percentage of CD19⁺ cells (B lymphocytes) and with significant increases in both the number and percentage of CD3⁺ and CD11c⁺ cells (T lymphocytes and potentially dendritic cells, respectively) (Figure 2). There was a trend toward increase with age of CD11b⁺ cells that did not reach significance (p values are from 0.202 to 0.520). However, no significant differences were evident in other cell populations such as CD62L⁺, Gr-1⁺, and Ly6c⁺ cells.

Figure 2.

Old Donor BMC Composition Analysis by Flow Cytometry

BMCs were harvested from donor mice at ages of 10 weeks, 1.5 years, and 2 years for quantitative analysis by flow cytometry. The key differences observed were a significant decrease in CD19⁺ B cells of total live cells and significant increases in CD3⁺ T cells and CD11c⁺ cells in the 2-year-old donor bone marrow. Cell amount is represented as absolute number of live cells per bone marrow, as well as percentage of cells per total cellular population (%).

BMC Therapeutic Impairment by CD19⁺ B Cell Depletion

Because we previously demonstrated that both donor age and donor MI decrease therapeutic efficacy of BMCs when implanted into post-MI hearts,35, 36 we searched for any common differences in BMC composition resulting from age or MI in young donors. The only difference that we detected between young and old donors that we had also observed between healthy and post-MI mice was the reduction in CD19⁺ B cells (data not shown). Therefore, we depleted B cells from young healthy BMCs (10 weeks old) on the basis of CD19⁺ expression to see whether BMC therapeutic efficacy is impaired by B cell depletion. CD19⁺ B cells were undetectable by FACS in the post-depletion BMCs (data not shown). Interestingly, young healthy donor BMCs became less therapeutic when manually depleted of B cells (Figure 3), as their implantation into young post-MI hearts led to significantly lower day 28 recipient EF (p < 0.001) and higher day 28 recipient ESV (p < 0.05) as compared with whole BMCs from donors at the same age and condition.

Figure 3.

BMC Therapeutic Impairment by B Cell Depletion

Young healthy donor BMCs (10 weeks old) were harvested and depleted of B cells on the basis of CD19⁺ expression. Implantation of the B cell-depleted donor BMCs led to less effective therapeutic outcomes in recipient day 28 EF and ESV as compared with whole BMCs from donors at the same age and condition. All data are means ± SD. Data are summarized numerically in Table S3.

Implantation of Intact B Cells Reproduces the Therapeutic Effects of Whole BMCs

Because (1) reduction in bone marrow B cells was associated with two distinct donor conditions (age and MI) that reduced therapeutic efficacy of BMCs, (2) direct depletion of B cells left the remaining BMCs in a less therapeutic state, and (3) B cells were reported to be a therapeutic subpopulation of BMCs when implanted immediately after MI in a rat model,³⁷ we tested the hypothesis that B cells account for the therapeutic efficacy of BMCs implanted 3 days post-MI. We isolated bone marrow B cells from young healthy donors by a subtractive selection strategy (see Materials and Methods) to avoid potential confounding effects of antibody binding to the cells and implanted them into the recipient infarct border zone 3 days post-MI as in our previous experiments. Similarly prepared bone marrow T cells and vehicle (HBSS/BSA) were injected as negative controls. Because we have observed B cells to account for roughly 20% of the BMCs (Figure 2 and unpublished data), B cells were suspended at 2 × 10⁷ cells/mL, and 2 × 10⁵ cells were injected split into two 5-μL injections as described in Materials and Methods, with the goal of implanting a comparable number of B cells to those in the original whole BMC implantations. T cells as a negative control were implanted at the same concentration and cell number as the B cells. As before, recipient mice were always young (10–12 weeks). Injection of B cells significantly improved recipient day 28 EF and reduced day 28 ESV and EDV; in contrast, injection of T cells improved LV dimensions but provided only a modest benefit for EF relative to injection of HBSS (Figure 4A). The effect on EF, ESV, and EDV by isolated B cells was comparable with that of the whole BMCs in our earlier experiments (see Figure 1).

Figure 4.

Therapeutic Effects of B Cells and B Cell Lysate on MI

Injection of purified bone marrow B cells (A) or B cell lysate (B) at day 3 post-MI significantly improved recipient day 28 EF, decreased day 28 ESV and EDV, and reduced infarct size. In contrast, bone marrow T cells (A) or T cell lysate (B) provided varying degrees of benefit to ESV and EDV relative to HBSS, but little or no benefit in EF. All data are means ± SD. Data are summarized numerically in Table S4.

Because the B cells were isolated by depletion of non-B cells from the harvested BMCs, but the depletion antibodies used for this approach did not include those specific for markers of MSCs, the possibility remained that a small number of MSCs contaminated the preparation, theoretically accounting for the therapeutic activity of the isolated cells. To confirm that this was not the case, we isolated the B cells from one additional donor using the same depletion approach and used flow cytometry to analyze the resulting population for MSCs defined as CD105⁺ CD29⁺ Sca-1⁺ CD45⁻ cells, based on the criteria used by R&D Systems in their mouse MSC identification panel FMC003 (see Materials and Methods). By these criteria, contaminating MSCs were undetectable in our B cell preparation (Figure S1).

Therapeutic Benefit of Bone Marrow B Cells Does Not Require the Cells to Be Alive

We showed previously that the therapeutic effects of BMC implantation into post-MI hearts are mimicked by injection of cell-free extract from lysed BMCs,²² and are thus paracrine in nature. To determine whether the therapeutic benefit of purified B cell implantation similarly was mediated by a paracrine mechanism, we repeated the implantation of B versus T cells but used cell lysate rather than intact cells. Notably, B cell lysate, but not T cell lysate, also significantly increased recipient day 28 EF. B cell lysate decreased day 28 ESV and EDV comparably with the intact B cells and intact BMCs, whereas T cell lysate bestowed partial therapeutic benefit for day 28 ESV and led to improvement in EDV (Figure 4B). Infarct size was smaller in the B cells and B cell lysate treatment groups than in the HBSS control group (Figures 4A and 4B).

Discussion

Our results indicate that increasing donor age makes donor’s BMCs progressively less able to prevent a decline in cardiac function when the aged BMCs are implanted into recipient hearts after MI. Advancing age lowers the number of bone marrow B cells, resulting in a less therapeutic BMC population, leading to lower recipient left ventricular EF and larger ESV and EDV on day 28 post-MI, along with larger infarct size and thinner anterior wall, independent of the recipient status. Implantation of intact and lysed B cells improved recipient post-MI cardiac function in EF, ESV, and EDV, as well as reduced infarct size, whereas implantation of BMCs with fewer B cells provided less therapeutic effect. Bone marrow B cells play a crucial role in BMC-based cell therapy via a paracrine mechanism involving intracellular mediators that can bestow benefit after a single intramyocardial administration.

It has long been understood that cells progressively lose function with advancing age, playing a large role in the loss of cardiac functionality.⁴¹ Increasing age can also be correlated with reductions in various molecular or functional properties of cells in the bone marrow42, 43, 44, 45, 46, 47, 48, 49 and of circulating angiogenic cells from the bone marrow and peripheral blood.50, 51, 52 Age-dependent reduction in plasticity of bone marrow-derived cells is a potential explanation for therapeutic functional impairment of those cells,⁴⁷ although our previous work²² and the results reported here indicate that cellular plasticity is not required for the therapeutic effects that we have observed (i.e., the cells are still therapeutic when dead and lysed). We and others have shown that the age of the BMC donor can also reduce therapeutic properties of the BMCs35, 53, 54 and intrinsic properties that may be mechanistically relevant.47, 50 Thus, BMCs in general from old animals may suffer from a lifetime of progressive impairment, causing gradual functional declines. The results of our current study demonstrate this progressive impairment of donor BMCs over 2 years of aging, implying that age-related decrease in cellular therapeutic potential may limit efficacy of autologous BMC therapy. Additionally, we previously reported that inflammation resulting from donor MI impairs the therapeutic efficacy of the donor BMCs, leading to cellular alterations within the bone marrow compartment.³⁶ Together, these findings imply that one reason that clinical trials have been less effective than mouse experiments may be that the bone marrow used in human acute MI trials is impaired by both the age and post-infarct condition of the patients. Thus, regardless of the mechanisms that lead to functional improvement, the age and pathophysiological background from which the donor BMCs are derived may interfere with autologous cell therapy in patients.⁵⁵

Notably, we showed previously that post-MI BMCs contain fewer CD19⁺ B cells but more inflammatory cells such as CD11b⁺ and Gr1⁺ myeloid cells,³⁶ indicative of an inflammatory state. We now show that aged-donor BMCs also contain fewer CD19⁺ B cells but more CD3⁺ and CD11c⁺ cells, and that removal of CD19⁺ B cells from young healthy donor BMCs renders the BMCs less therapeutic. Interestingly, B cells have been reported as a therapeutic BMC subpopulation for MI treatment,³⁷ and the depletion and restoration of B cells have been shown to increase and reduce infarct size in a cerebral artery occlusion model.38, 39 We show here that partial depletion of B cells from young healthy donor BMCs reduces the BMCs’ therapeutic effect, mimicking the effects of donor age or MI. Therefore, despite presumed differences in the etiology of the reduction in B cells by age or MI, the therapeutic impairment that results from both conditions may result from the alteration of the bone marrow B cell compartment that they have in common.

Our results provide an insight that aged BMCs have impaired cellular efficacy and offer less B cells, which might be central to the therapeutic impairment. Thus, further investigation has also been addressed on bone marrow B cells’ role in the therapeutic effect. As expected, injections of intact B cells or B cell lysate isolated from young healthy donor bone marrow significantly improved the post-MI cardiac function and reduced infarct size in the recipient MI hearts in mice. However, injections of bone marrow T cells or T cell lysate from young healthy donor bone marrow led to only a slight preservation of LVEF and to intermediate preservation of chamber volumes. These data raise the possibility that B cells may also play an important role in human BMC therapeutic impairment as they did in mice, and that their lower numbers in BMCs from older animals may partially explain the poor outcomes of the clinical trials. Supporting this notion, CCTRN researchers observed that the level of CD19⁺ B cells in bone marrow of the individual TIME and LateTIME participants correlated positively with clinical outcome.⁵⁶

Importantly, our studies provide insights that cannot result from the study of only autologous human trials, because by delivering human or mouse BMCs to mice, we were able to vary the condition of the donor BMCs while keeping the recipient heart condition constant. For example, another CCTRN trial, FOCUS,³³ observed in post hoc subgroup analysis that patients under age 62 years exhibited better LVEF than those over 62, but it is difficult to know whether this is related to the recipient heart or the donor BMCs in an autologous setting. This has been a powerful advantage of our approach using recipient mice that are constant while implanted BMCs are from variable conditions.

Many of the therapeutic effects of post-MI implantation of a range of cell types have ultimately proven to be paracrine in nature even if the particular cell type can differentiate into cardiomyocytes under appropriate culture conditions, and have been attributed to exosomes secreted by these cells.21, 23, 57, 58, 59 The mechanistic basis for this observation remains unclear. We observed partial benefit from injection of T lymphocytes that did not reach the therapeutic effect of the B lymphocytes, so it is possible that the repertoire of exosomal microRNAs or other factors within the cytoplasm from T cells includes a subset of the crucial factors in B cells. Still, it is notable that the reduction with age of BMC therapeutic efficacy correlated with a reduction in B cells and an increase in T cells, arguing against the T cells carrying the main therapeutic activity that is lost with age (Figure S2). The observation that B cells play a major role in the therapeutic effect is intriguing, given that the post-MI heart experiences a robust multi-stage inflammatory response involving many bone marrow-derived cells that do not bestow this benefit naturally. Nonetheless, our results indicate that a key component of bone marrow-derived B cells, possibly the exosomes, is not present in this intrinsic response, at least not at sufficient levels to cause the therapeutic effect of direct B cell delivery to the tissue. The potential therapeutic use of exosomes from bone marrow B cells for post-MI therapy is the subject of ongoing studies.

We conclude that B lymphocytes account for at least a substantial portion of the therapeutic activity of unfractionated BMCs when administered post-MI. The therapeutic effect is mediated by intracellular or transmembrane factors that improve cardiac function and preserve viable myocardium. Our observations that therapeutic effect results from a single injection of whole BMC lysate²² or B cell lysate indicate that the therapeutic benefits of these factors do not require active secretion from persisting living cells. The reduction of the number of B cells in the bone marrow is a consequence of both age and/or MI that limits therapeutic efficacy of autologous BMC therapy.

Materials and Methods

Donor and Recipient Mice

All animal experiments were approved by the Institutional Animal Care and Use Committee of the University of California, San Francisco. Male C57BL/6J mice were obtained from the Jackson Laboratory (Bar Harbor, ME, USA). Donor mice were at various ages from young (10–12 weeks) to old (2 years). The donor mice at ages of 6 months, 1, 1.5, and 2 years were purchased from the National Institute of Aging. Recipient mice were always at 10–12 weeks old. Recipient group size was 15 or greater; BMCs from one donor mouse were used per five recipient mice.

Myocardial Infarction

MI was surgically induced as described previously.⁶⁰ Mice underwent the MI surgery under anesthetization with 2% isoflurane and received analgesics (buprenorphine 0.1 mg/kg, subcutaneous injection) at time of surgery. The heart was exposed via a parasternotomy, and the left anterior descending coronary artery was permanently ligated ∼3 mm below the tip of the left atrium. Recipient mice undergoing the parasternotomy MI were all consistent from experiment to experiment. The mortality rates during and after MI surgery were 4% and 2%, respectively.

BMC Harvest and Injection

The protocol for BMC harvest and injection has been previously described.22, 60 In brief, mouse femurs were harvested, and bone marrow was flushed with cold HBSS with 0.5% BSA. The cell suspension was strained through a 70-μm filter and washed twice with HBSS/BSA. After centrifugation, the unfractionated whole BMCs were harvested for injection. The cell concentration was adjusted to 10⁸ viable cells/mL. 10⁶ cells were injected into myocardium at the infarct border zone split into two 5-μL injections under ultrasound visualization using a Vevo660 micro-ultrasound system (VisualSonics, Toronto, ON, Canada).⁶⁰ Each donor mouse provided BMCs for five recipient mice in order to have multiple donors per group. BMCs were always implanted into recipient hearts on day 3 post-MI. Injection of HBSS with 0.5% BSA served as a negative control. Optimal intramyocardial injections were judged as 98% of visible changes in localized ultrasound signal resulting from the presence of cell suspension.22, 36

Flow Cytometry

The fluorescent antibodies for cell-surface markers used for flow cytometry analysis were listed as follows: anti-panNKFITC, anti-Gr-1-PE, anti-MHC class II (major histocompatibility complex class II)-PE, anti-CD3-PerCPCy5.5, anti-CD25-PerCPCy5.5, anti-CD45R-PerCPCy5.5, anti-CD11b-PerCPCy5.5, anti-CD11c-PECy7, anti-CD69-PECy7, anti-CD11b-Pacific Blue, anti-CD44-Pacific Blue, anti-CD80-allophycocyanin (APC), anti-CD19-Alexa 647, anti-CD4-APC750, and anti-Gr-1-APC750 (eBioscience); anti-Ly6C-FITC (fluorescein isothiocyanate) and anti-CD19-APCCy7 (BD Biosciences); anti-CD86-Pacific Blue (BioLegend); anti-CD62LAlexa 647 (UCSF Hybridoma Core); and anti-CD8-Pacific Orange (Invitrogen). Anti-CD11b-Pacific Blue, anti-CD11b-PerCPCy5.5, anti-Gr-1-PE, anti-Gr-1-APC750, anti-Ly6CFITC, and anti-CD11c-PE were used for identifying monocytes, macrophages, and neutrophils (Table S1). Bone marrow was harvested from two femurs, and BMCs were isolated as indicated above. Following red blood cell lysis, cells were resuspended in PBS with 5% BSA, and Fc receptors were blocked, stained, and processed on an LSR II flow cytometer and analyzed using FlowJo v8.8.6. Absolute numbers of cells were calculated based on the percentages of total live cell numbers (cell viability accessed by trypan blue staining).

Bone Marrow CD19⁺ B Cell Depletion

Mouse CD19⁺ MicroBeads (Miltenyi Biotec, Auburn, CA, USA) were used for the positive depletion of mouse B cells from bone marrow. In brief, whole BMCs were harvested as described above from donor mice at age 10 weeks, the total cell number was determined, and the BMC suspension was incubated with CD19⁺ MicroBeads for 15 min at 4°C. The mixture was loaded onto a magnetic-activated cell sorting (MACS) LD column, which was placed in a magnetic separator (QuadroMACS) to remove the magnetically labeled CD19⁺ cells. 10⁶ cells from the B cell-depleted fraction were injected into myocardium at the infarct border zone as described above.

Isolation and Injection into Myocardium of B Cells and T Cells

Whole BMCs were harvested from the 10-week-old donor mouse and were treated with red blood cell lysis buffer (eBioscience), and the total cell number was determined. B cells were isolated indirectly by MACS using a mouse Pan B cell isolation kit II (Miltenyi Biotec) to deplete non-B cells from the BMC suspension. The antibodies used to remove non-B cells were as follows: FITC anti-mouse CD3ε⁺, FITC anti-mouse CD4, FITC anti-mouse CD8a, FITC anti-human/mouse CD11b, FITC anti-mouse Ly-6G, and FITC anti-mouse TER-119 (Tonbo Biosciences); and FITC anti-mouse CD49b (BioLegend). The remaining unlabeled fraction enriched in B cells was incubated with mixed antibodies listed above at a dilution of 1:400 for subsequent fluorescence-activated cell sorting using an Avalon cell sorter, and the un-labeled cells that were highly enriched for B cells were collected for cell injection. T cells were prepared by a similar indirect isolation approach using a mouse Pan T cell isolation kit II (Miltenyi Biotec). Antibodies used for T cell isolation were as follows: FITC anti-mouse CD19⁺, FITC anti-mouse CD4, FITC anti-mouse CD8a, FITC anti-human/mouse CD11b, FITC anti-mouse Ly-6G, and FITC anti-mouse TER-119 (Tonbo Biosciences); and FITC anti-mouse CD49b (BioLegend). For implantation of B or T cells into the myocardium at the infarct border zone, cells were suspended at 2 × 10⁷ cells/mL, and 2 × 10⁵ cells were injected split into two 5-μL injections as described above. Cell viability by trypan blue staining was assessed in all cells before implantation. B cells isolated from two donor mice were pooled and used per five recipient mice, while T cells isolated from one donor mouse were used per five recipient mice.

Assessment of Potential MSC Contamination of B Cell Preparations

B cells were isolated from one more donor using the depletion approach described above. The resulting B cell preparation (1.2 × 10⁶ cells) was divided, and one portion was stained with fluorescent antibodies in a panel based on R&D Systems mouse identification kit (catalog #FMC003) but assembled independently. The panel antibodies used for flow cytometry analysis were the following: Alexa Fluor 647 anti-CD105⁺, PE anti-CD29⁺, BV421 anti-Sca-1⁺, and FITC anti-CD45⁻ (all from BD Biosciences). The remaining portion of the preparation was stained separately with anti-CD19⁺ in Alexa Fluor 647, PE, BV421, and FITC (BD Biosciences) as positive controls for flow cytometry analysis.

Preparation of Cell-Free Extracts from B or T Cell Lysates

Isolated B or T cells were diluted with HBSS with 0.5% BSA at a concentration of 2 × 10⁷ cells/mL. This diluted cell suspension was subjected to four freeze-thaw cycles using an ethanol/dry ice bath followed by micro-centrifugation at 14,000 rpm (1 min) to remove insoluble material. Soluble cell-free lysate was collected and injected into myocardium at the infarct border zone as two 5-μL injections as described above.

Echocardiography

Echocardiography of recipient mice under anesthesia with 1.25% isoflurane was performed at baseline and at 2 and 28 days post-MI, using a Vevo660 micro-ultrasound system as described previously.22, 36 Echocardiograms were obtained at long-axis view to measure the left ventricular ESV and EDV. LVEF was calculated by the following formula: EF (%) = [(EDV − ESV)/EDV] × 100. Wall thickness was measured at the apical-segment (infarct) and mid-segment (infarct border zone) of the anterior wall. Echocardiographic parameters were measured by two blinded investigators.

Measurement of Infarct Size

Mouse hearts were arrested in diastole with saturated KCl injected into the left ventricular chamber and removed. Frozen heart sections were stained with Masson trichrome for infarct size measurement by a midline arc length method.⁶¹ The sections were read blind and scored for the extent of fibrosis.

Statistical Analysis

Power calculation based on SDs from within-group comparisons in several of our previous MI experiments determined that n = 10/group was sufficient to detect changes in cardiac function at a power of 0.8 and significance level of 0.05. For comparisons involving multiple groups and times, we fit a two-factor (treatment condition and time) repeated-measures ANOVA to all the data at once using a mixed model estimated with restricted maximum likelihood estimation with an unstructured covariance matrix of residuals, then tested for differences over time and across treatment condition using contrasts and pairwise comparisons, adjusted for multiple comparisons using the Sidak method. Calculations were done with Stata 13.1. All values are presented as mean ± SD.

Author Contributions

S.A. conducted the aging experiments; X.W. helped conceive of and performed the B and T cell experiments and data analysis with participation of M.A.R., H.J.R., D.S.K., M.V., E.L., and R.D.; S.C.K., M.L.H., and M.L.S. conceived of the project; M.L.H. and S.V.S. provided collaborative experimental support and ideas for the project; X.W. and M.L.S. wrote the manuscript.

Conflicts of Interest

The authors declare that they have no competing interests.