Shock

INTRODUCTION

Accumulating evidence has demonstrated that mesenchymal stem cells (MSCs) may protect the myocardium from ischemia through paracrine actions, such as the induction of growth factor production (^{1, 2}). These growth factors may mediate several beneficial effects including improvement of myocardial function, reduction of myocyte apoptosis, and conduction of positive ventricular remodeling (^{3, 4}). Among these, vascular endothelial growth factor (VEGF) plays an important role in regulating MSC paracrine action. Increased VEGF has been shown to mediate MSC-induced protection in the ischemic heart (⁵). Overexpressing VEGF in MSCs provides greater cardiac protection (⁶). Therefore, genetic modification of MSCs to increase stem cell VEGF production may augment paracrine effects of stem cell cardioprotection.

After myocardial I/R injury, a substantial amount of TNF is released in the heart (^{7, 8}). The locally produced TNF may exert important effects on implanted stem cell function. TNF acts by binding to two distinct receptors, a 55-kd receptor (TNFR1) and/or a 75-kd receptor (TNFR2) (⁷). Previous studies have shown that TNFR1 signaling mediates deleterious effects on TNF in cardiac cells (^{4, 9}), whereas the TNFR2 pathway may mediate beneficial effects in the heart (^9-11). This led to the appreciation that different TNF receptors (TNFRs) may mediate MSC function differently. Indeed, recent evidence has indicated that ablation of TNFR1 in stem cells increases VEGF secretion, whereas the absence of TNFR2 decreases MSC VEGF production (^{12, 13}). This suggests that TNFR1 signaling may play a detrimental role in MSC VEGF release, whereas TNFR2 may favor stem cell function. However, it is unknown whether genetically modified MSCs with TNFR1 deficiency (TNFR1 knockout [TNFR1KO]) and/or TNFR2KO will affect MSC-mediated protection of myocardial function after acute I/R.

Therefore, in this study, we hypothesized that 1) preischemic infusion of MSCs derived from TNFR1KO mice would further increase myocardial function compared with wild-type (WT) MSCs after I/R and that 2) preischemic treatment with TNFR2KO or TNFR1/2KO MSCs would neutralize MSC-improved myocardial function after I/R.

MATERIALS AND METHODS

Animals

Normal, adult, male Sprague-Dawley rats were obtained from Harlan (250-275 g; Indianapolis, Ind). Both TNFR1KO and TNFR2KO mice with a background of C57BL/6J (WT1) and TNFR1/2KO mice with a background of B6129SF2 (WT2) (male, 6-8 weeks old) were obtained from Jackson Laboratory (Bar Harbor, Me). Receptor-deficient mice are viable, normal in size, and do not display any gross physical abnormalities. Animals were fed a standard diet and acclimated in a quiet quarantine room for 1 week before the experiments. The animal protocol was reviewed and approved by the Indiana Animal Care and Use Committee of Indiana University. All animals received humane care in compliance with the Guide for the Care and Use of Laboratory Animals (NIH publication No. 85-23, revised 1996).

Preparation of mouse bone marrow stromal cells

A single-step stem cell purification method using adhesion to cell culture plastic was used, as previously described (¹⁴). Briefly, mouse bone marrow stromal cells were collected from bilateral femurs and tibias after sacrifice by removing the epiphyses and flushing the shaft with complete media (Iscove modified Dulbecco medium; GIBCO Invitrogen, Carlsbad, Calif) and 10% fetal bovine serum (GIBCO Invitrogen) using a syringe with a 26-gauge needle. Cells were disaggregated by vigorous pipetting several times and were passed through a 30-μm nylon mesh to remove remaining clumps of tissue. Cells were washed by adding complete media, centrifuging for 5 min at 300g at 24°C, and removing the supernatant. The cell pellet was then resuspended and cultured in 75-cm² culture flasks with complete media at 37°C, 5% CO₂. The bone marrow stromal cells preferentially attached to the polystyrene surface; after 48 h, nonadherent cells in suspension were discarded. Fresh complete media was added and replaced every 3 days thereafter. When the cultures reached 90% of confluence, MSCs were passaged with the addition of a solution 0.25% trypsin-EDTA (GIBCO Invitrogen) and replated in flasks. Cells were used for experimentation between passages 3 and 8.

Isolated heart preparation (Langendorff)

Hearts were isolated as previously described (^15-17). Briefly, rats were anesthetized (sodium pentobarbital, 60 mg/kg i.p.) and heparinized (500 U i.p.), and hearts were rapidly excised via median sternotomy and placed in 4°C Krebs-Henseleit solution. The aorta was cannulated, and the heart was perfused under constant pressure (mean, 75 mmHg) with oxygenated (95% O₂/5% CO₂) Krebs-Henseleit solution (37°C). A left atrial resection was performed before insertion of a water-filled latex balloon, which was initially adjusted to a desired mean end-diastolic pressure (EDP) of 5 mmHg. Hearts were allowed to equilibrate for 15 min and paced at 350 beats/min to ensure a standard heart rate between groups. A three-way stopcock above the aortic root was used to create warm global ischemia, during which time the heart was placed in a 37°C degassed organ bath. After 25 min of ischemia, hearts were reperfused for 40 min. The left ventricular developed pressure (LVDP), the maximal positive and negative values of the first derivative of pressure (+dp/dt and −dp/dt), and EDP were continuously recorded using a PowerLab 8 preamplifier/digitizer (AD Instruments Inc, Milford, Mass) and an Apple G4 PowerPC computer (Apple Computer Inc, Cupertino, Calif).

Experimental groups

Rat hearts were divided into the following groups: 1) vehicle control (n = 6), 2) WT1 MSC infusion (n = 6), 3) WT2 MSC infusion (n = 5), 4) TNFR1KO (R1KO) MSC infusion (n = 4), 5) TNFR2KO (R2KO) MSC infusion (n = 6), and 6) TNFR1/2KO (R1/2KO) MSC infusion (n = 5). Cells were collected from flasks by trypsinization and centrifuged. The cell pellet was resuspended in Krebs-Henseleit solution (37°C), and a count was performed. The appropriate amount of Krebs-Henseleit solution is then added to the cell suspension to create a solution with a final concentration of 1 × 10⁶ cells/mL. During the course of 1 min before ischemia, 1 mL of this MSC solution was infused into the coronary circulation.

TNF-α and VEGF enzyme-linked immunosorbent assay

Ventricular myocardial tissue was homogenized, and TNF-α and VEGF levels were determined by enzyme-linked immunosorbent assay (ELISA) using commercially available ELISA sets (R&D Systems, Inc, Minneapolis, Minn). The ELISA was performed according to the manufacturer's instructions. All samples and standards were measured in duplicate. Protein concentrations of samples were measured via biophotometer using the Bradford protocol. The ELISA values were normalized using the protein concentrations of each sample, and results were reported as picograms per milligram protein.

Presentation of data and statistical analysis

All reported values are mean ± SEM, and P < 0.05 was considered statistically significant. The LVDP, +dp/dt, and −dp/dt are presented as a percentage of equilibration. Myocardial functional parameters were analyzed at time 80 (end reperfusion). Data were compared using one-way ANOVA with post hoc Tukey test.

RESULTS

MSCs improved cardiac function after I/R injury

Infusion of MSCs from WT1 significantly increased postischemic myocardial function compared with control, as exhibited by improved LVDP (Figs. 1A and 4A) and +dp/dt (Figs. 1B and 4B). A trend toward increased myocardial recovery compared with control after infusion of WT2 MSC was also noted with respect to improved LVDP (95% confidence interval [CI], −28.47 to 0.1503), +dp/dt (95% CI, −18.80 to 5.68), and −dp/dt (95% CI, −23.77 to 2.33). In general, infusion of WT MSCs seemed to protect cardiac contractility in hearts subjected to acute I/R, which is in line with our previous observations (¹⁸).

Absence of TNFR1 improved MSC-mediated cardioprotection

Treatment with MSCs from TNFR1KO mice before ischemia further improved postischemic cardiac function when compared with WT1 MSC group. Ablation of TNFR1 in MSCs showed an increased LVDP by 33% (Fig. 2A), +dp/dt by 39% (Fig. 2B), and protected −dp/dt by 43% (Fig. 2C) at the end of reperfusion compared with control.

Deficiency of TNFR2 neutralized MSC-provided cardioprotection

Notably, ablation of TNFR2 in MSCs did not increase MSC-mediated myocardial function as shown by TNFR1KO MSCs. In fact, it decreased cardiac function in terms of +dp/dt compared with WT1 MSC group after I/R (Fig. 3B). The absence of TNFR2 reduced MSC-mediated protection of LVDP (Fig. 4A), -dp/dt (Fig. 4C), and EDP (Fig. 4D) to the levels seen in the control group in response to I/R.

Double KOs of TNFR1 and TNFR2 (TNFR1/2KO) abolished MSC-mediated improvement of cardiac function

Hearts treated with WT2 MSCs exhibited a slight trend toward improved recovery compared with control (Figs. 5 and 7). Hearts treated with TNFR1/2KO MSCs exhibited a slight trend toward worse recovery as compared with hearts treated with WT2 MSCs (Figs. 6 and 7). No difference in recovery at end reperfusion was detected between TNFR1/2KO MSCs and control (Fig. 7).

Improved functional recovery after I/R is associated with decreased levels of myocardial TNF-α

Ventricular tissue from all hearts was assayed for TNF-α and VEGF levels via ELISA. No differences in myocardial VEGF levels were detected (Fig. 8, B and D). Differences in ventricular TNF-α levels were detected with WT1, WT2, and TNFR1KO MSC-treated hearts, exhibiting lower levels of TNF-α as compared with control (Fig. 8, A and C).

DISCUSSION

In this study, by using genetically modified MSCs with ablation of TNFR1, TNFR2, or TNFR1/2, we determined the effects of TNFRs on MSC-mediated protection of myocardial function after acute I/R. Herein, our results further confirmed that preischemic infusion of MSCs improved cardiac function in response to I/R. In addition, the absence of TNFR1 increased MSC-provided cardioprotection as exhibited by further improved LVDP and ±dp/dt in TNFR1KO MSC-treated group. However, deficiency of TNFR2 or TNFR1/2 seems to have no effect or, in the case of the latter, may even negate the association of MSCs with improved myocardial recovery after I/R. In addition, improved functional recovery correlated with reduced levels of ventricular myocardial TNF-α levels.

Stem cells have been reported to repair irreversible tissue damage via differentiation to the injured cells (¹⁹). However, accumulating evidence has questioned this mechanism because of the low numbers of stem cells engrafted into target tissue (^{20, 21}). Recently, there have been a growing number of studies indicating that cardiac protection by MSCs may be mediated through their paracrine actions, including production of protective molecules (²²). Indeed, our group has previously demonstrated that treatment with MSCs before ischemia significantly improved myocardial function (^{16, 23, 24}). The infused MSCs were not able to differentiate into cardiomyocytes during such a brief time (25 min of ischemia + 40 min of reperfusion), suggesting that MSC paracrine actions exert primary effects on protection of myocardial function after acute I/R. In this study, we further confirmed that paracrine effects of MSCs provide cardiac protection directly with respect to preischemic infusion of MSCs, resulting in increased functional recovery after I/R.

Recently, much attention has been directed at how to optimize the therapeutic effectiveness of stem cells for clinical use. Among these, genetic modification of MSCs is an appealing method. The MSCs engineered to overexpress myocardial protective factors such as VEGF and Akt have been shown to significantly augment their ability to protect the ischemic myocardium (^{6, 25}). TNF, a proinflammatory cytokine, is substantially produced in the heart subjected to I/R injury. TNF exerts different effects on cardiac myocytes through binding to different receptors (TNFR1 and/or TNFR2) and activating different downstream signaling pathways (detrimental versus beneficial) (^{4, 9, 11}). Therefore, one would assume that TNF may play different roles in MSC paracrine effects depending on activation of different TNFRs. On one hand, TNF has been reported to induce MSC apoptosis and reduce MSC survival (^{12, 26}). Conversely, the addition of TNF is able to stimulate MSC production of growth factors, including VEGF, hepatocyte growth factor, and insulin-like growth factor 1 in vitro (²⁷), suggesting that TNF may benefit MSC paracrine action (²⁸). In the present study, we found that deficiency of different TNFRs led to different effects of MSCs on the protection of myocardial function after I/R. Ablation of TNFR1 significantly increased MSC-mediated cardioprotection, suggesting that TNFR1KO may exert a beneficial effect on MSC paracrine actions and, thus, may augment MSC-provided protection. On the other hand, the absence of TNFR2 did not improve MSC-mediated protection of myocardial function, and deletion of both TNFRs was associated with a trend toward negation of MSC-mediated protection.

In fact, our previous study has indicated that genetic deletion of TNFR1 increased MSC VEGF production (¹³). The VEGF is an important mediator of stem cell paracrine effects. Genetically modified MSCs to overexpress VEGF have been shown to enhance MSC-mediated protection in the ischemic heart (⁶). In addition, adult MSCs provide greater protection of myocardial function after I/R because of their higher VEGF production when compared with neonatal MSCs (²⁴). Furthermore, decreasing VEGF levels in MSCs by using VEGF siRNA neutralizes MSC-provided cardiac protection in the heart subjected to acute I/R injury (²⁴). These findings suggest that modification of MSCs to increase VEGF production may augment protective paracrine effects of MSCs. Therefore, it is possible that increased VEGF production in TNFR1KO MSCs may provide greater cardioprotection after I/R. On the other hand, deficiency of the TNFR2 gene has been noted to decrease MSC VEGF production under normal cell culture conditions and stimulation with TNF or hypoxia (¹³). Therefore, it is not surprising that in this study, preischemic infusion of TNFR2KO MSCs abolishes MSC-mediated protection of cardiac function after acute I/R.

Importantly, however, differences in myocardial VEGF levels among groups in these experiments were not detected. It is possible that the use of mouse MSCs in the rat myocardium may confound the ability to detect changes in VEGF levels particularly when the proposed differences in VEGF are hypothesized to occur because of changes in MSC secretion of this growth factor. In this case, an antibody specific for rat VEGF was used for the ELISA assay. However, there is significant cross-reactivity between both the mouse and rat antibodies and the mouse and rat VEGF protein in commercially available kits and, therefore, changes in VEGF levels related to differences in MSC production of this growth factor may be difficult to detect via this method. Nevertheless, the previously obtained in vitro data previously described do support a potential role for VEGF in the changes observed among the different MSCs studied with these experiments.

Ablation of TNFR1 has also been shown to decrease MSC production of TNF-α and IL-6 in response to hypoxia in vitro (¹²). This provides another possibility that decreased levels of TNF and IL-6 in TNFR1KO MSCs may alleviate MSC damage when they are infused into the ischemic heart. Aside from the effect of TNF-α on MSCs, it is also of interest to consider the effect that MSCs can have on TNF-α production by other cells types. Accumulating data suggest that MSCs have a unique ability to modulate the immune system and the inflammatory response (²⁹). Recent evidence obtained from coculture experiments suggests that MSCs can reduce TNF-α production by cells of the immune system (³⁰). This is a possible explanation for the findings of reduced TNF-α levels in the myocardium of MSC-treated hearts observed with these experiments. It is possible that TNFRs differentially regulate the ability of MSCs to attenuate TNF-α production by other cell types, which could provide an explanation for these findings. Additional experiments will have to be performed to confirm this.

In summary, genetic modification of MSCs is able to enhance MSC paracrine effects for widespread clinical use. This study indicates that genetically engineered MSCs with TNFR1KO increases MSC-mediated protection of myocardial function, whereas the absence of TNFR2 did not improve MSC-mediated protection of myocardial function, and deletion of both TNFRs was associated with a trend toward negation of MSC-mediated protection. Further investigations are required to determine the detailed mechanisms by which TNFR1 or TNFR2 MSCs act via paracrine actions and to develop new methods for enhancing the therapeutic effectiveness of stem cells.

REFERENCES