Edited by: Prasanna Krishnamurthy, Houston Methodist Research Institute, USA
Reviewed by: Mohsin Khan, Temple University, USA; Ramaswamy Kannappan, University of Alabama at Birmingham, USA
*Correspondence: Marcin Wysoczynski
Steven P. Jones
This article was submitted to Stem Cell Research, a section of the journal Frontiers in Cell and Developmental Biology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Cell therapy improves cardiac function. Few cells have been investigated more extensively or consistently shown to be more effective than c-kit sorted cells; however, c-kit expression is easily lost during passage. Here, our primary goal was to develop an improved method to isolate c-kitpos cells and maintain c-kit expression after passaging. Cardiac mesenchymal cells (CMCs) from wild-type mice were selected by polystyrene adherence properties. CMCs adhering within the first hours are referred to as rapidly adherent (RA); CMCs adhering subsequently are dubbed slowly adherent (SA). Both RA and SA CMCs were c-kit sorted. SA CMCs maintained significantly higher c-kit expression than RA cells; SA CMCs also had higher expression endothelial markers. We subsequently tested the relative efficacy of SA vs. RA CMCs in the setting of post-infarct adoptive transfer. Two days after coronary occlusion, vehicle, RA CMCs, or SA CMCs were delivered percutaneously with echocardiographic guidance. SA CMCs, but not RA CMCs, significantly improved cardiac function compared to vehicle treatment. Although the mechanism remains to be elucidated, the more pronounced endothelial phenotype of the SA CMCs coupled with the finding of increased vascular density suggest a potential pro-vasculogenic action. This new method of isolating CMCs better preserves c-kit expression during passage. SA CMCs, but not RA CMCs, were effective in reducing cardiac dysfunction. Although c-kit expression was maintained, it is unclear whether maintenance of c-kit expression
Following myocardial infarction, the loss of contractile units causes left ventricular dysfunction and is the major barrier to treating heart failure. The field of cardiac regeneration developed to address this question and was predicated on the concept of administering a stem or progenitor cell with sufficient potency to regenerate the damaged heart (Orlic et al.,
Adoptive transfer of various types of progenitor/mesenchymal/stromal/stem cells attenuates cardiac dysfunction in preclinical heart failure models (Keith and Bolli,
Here, we refined an uncomplicated protocol to isolate c-kit+ cells with a goal of preserving c-kit expression in the majority of cells through several passages. This involved modification of initial cell plating after tissue digestion and improvement of the sorting procedure, which allowed us to monitor cell purity directly after sorting. We immunophenotyped the resulting cells and used them along with parallel “control” cells to test whether they were effective in attenuating infarct-induced heart failure. An additional feature of our present study was the use of echocardiographically guided percutaneous delivery to the left ventricular lumen, which enabled intravascular delivery of the cells. Thus, we were able to provide details for the isolation and culture of cardiac mesenchymal cells (CMCs) with stable c-kit expression, and characterize their effectiveness in the setting of infarct-induced heart failure.
All animal procedures were performed in compliance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the University of Louisville Institutional Animal Care and Use Committee. Myocardial cells were isolated from 12 to 15 week old, male C57BL6 mice (Jackson Laboratory). Mice were euthanized by sodium pentobarbital injection (i.p. 100 mg/kg) and cardiac excision. The excised hearts were washed in room temperature PBS, minced into small pieces and enzymatically dissociated with Collagenase II (5 μg/mL in PBS; Worthington) with gentle agitation for 45 min at 37°C. After Collagenase II inactivation with DMEM/F12 medium containing 10% FBS cells were centrifuged at 600 × g for 10 min. The collected cell pellet was suspended in growth medium consisting of DMEM/F12 (Invitrogen), 10% FBS (Seradigm, VWR), bFGF (10 ng/ml), EGF (10 ng/ml), LIF (10 ng/ml), ITS (insulin/transferrin/selenium), glutamine and Pen-Strep. Two c-kitpos cell isolation methods were used. They differed in initial seeding of the digested hearts and cell labeling for c-kit sorting (direct vs. indirect method). The indirect staining procedure allowed evaluation of the c-kit sort purity directly after magnetic selection. In the standard procedure (Supplementary Figure
Myocardial cells were detached from culture dishes with 0.25% trypsin-EDTA. After incubation for 30 min at 4°C with mAbs, cells were washed then suspended in 0.5 mL of PBS, and analyzed via flow cytometry on an LSRII system. The following mAb were used for characterization of cell phenotype were used (Table
CD90 | PE | 30-H12 | eBioscience |
CD29 | PE | HMb1-1 | eBioscience |
CD105 | PE | MJ7/18 | eBioscience |
CD73 | PE | TY/11.8 | eBioscience |
CD31 | PE | 390 | eBioscience |
c-kit | APC Cy7 | 2B8 | eBioscience |
Sca1 | Per CP Cy5 | D7 | eBioscience |
CD45 | APC Cy7 | 30-F11 | BD Biosciences |
Clonogenic potential was evaluated 7 days after sorting and expansion. Using Terasaki plates, single cells were seeded in 10 μL of growth medium per well. Single cell seeding was confirmed on the day of seeding. The cells were grown for 14 days. Colonies of ≥30 cells were considered clonogenic; the rate of clonogenicity was calculated as a percent of wells with clonogenic cells per total number of seeded wells.
Total mRNA was isolated with the RNeasy Mini Kit (Qiagen) and reverse transcribed with TaqMan Reverse Transcription Reagents (Applied Biosystems). Quantitative assessment of mRNA markers of pluripotency (
Adult (12–20 week) male mice were subjected to non-reperfused,
Transthoracic echocardiography of the left ventricle was performed as previously described (Chang et al.,
After cell sorting, RA and SA c-kitpos CMCs were expanded for 3–5 passages. On the day of injection, cells were 60–80% confluent. The cell monolayer was washed two times with room temperature PBS to remove debris and dead cells. Subsequently, cells were harvested by enzymatic digestion with 0.25% trypsin-EDTA (Invitrogen). The resulting cell suspension was trypsin-inactivated with DMEM/F12 supplemented with 10% FBS. After centrifugation at 600 × g for 10 min at RT, cells were resuspended in PBS and counted using a hemocytometer. Next, cells were centrifuged for 10 min at 600 × g and RT. A total of 106 cells were resuspended in 200 μL of sterile PBS (pH 7.4) at room temperature and transported to the echocardiography laboratory for echo-guided percutaneous injection. The infusion was performed within 10 min after the last centrifugation of the cells.
At 48 h after MI, mice were anesthetized with isoflurane (1.5%), and kept supine on an examination board interfaced with the Vevo 770. Depilatory cream was applied to the mouse's abdomen and wiped clean to remove all hair. Next, each side of the lower abdomen was taped down to the platform with surgical tape to keep the injection area taught. Mice were treated with analgesic (5 mg/kg ketoprofen; subcutaneously). The echo probe was placed suprasternally to capture 2D, long-axis views of the heart. The field of view was adjusted to include the presence of the ribs (to avoid them during insertion of the needle). A 1 mL syringe was loaded with a 250 μL volume of 106 cells (SA or RA) or PBS alone, and a 27 ga (13 mm length) needle was placed onto the syringe. The syringe was loaded onto an injection mount (VisualSonics) with the needle's bevel facing up. Once the needle was in the frame of the B-mode image, the angle and height of needle were determined based on the needle-guide feature on the Vevo 770. Essentially, a diagonal line was drawn creating a visual, virtual path of the needle such that a clear intercostal path into the ventricle was made. The angle and height of the needle were adjusted through fine control adjustments of the injection mount until it approximated the needle-guide. Once positioned, the needle was advanced into the skin and through the intercostal space. The needle was slightly withdrawn to ensure successful passage between the ribs and a clear path to the ventricle was visualized. Then the needle was advanced into the ventricle. The volume of the syringe was injected (over ~2 s) into the ventricular cavity and the needle was withdrawn. Successful deployment of cells or PBS was identified by a visual surge of radio-opaque signal inside the ventricle via 2D imaging or the by the infiltration of blood into the syringe. Anesthesia was quickly discontinued and the mice were cleaned. The mice were placed under a heat lamp to maintain body temperature until recovery.
After final echocardiography, hearts were arrested in diastole with saturated KCl and CdCl2 (100 mmol/L). The heart was excised, the aorta was then cannulated, and the heart perfused with PBS followed by 10% buffered formalin at 75 mmHg while the LV was perfused through an apical cannula with formalin at 8 mmHg (to preserve overall spherical geometry). Hearts were then cut into 2 mm cross sectional slices and processed for paraffin embedding. Slices were cut into 4 μm sections for histology and immunofluorescent staining. LV area, risk area, LV cavity area, and infarct area were measured in Masson's trichrome stained sections as previously described (Tang et al.,
Short-axis, mid-ventricular, paraffin-embedded cardiac sections were heated at 70°C for 30 min, then washed in xylene twice for 5 min to deparaffinize the slides. The subsequent rehydration process consisted of washing slides in serial dilutions of ethanol (100, 96, 96, 90, and 80%) for 5 min each. Slides were then washed in dH2O for 3 min. The antigen retrieval process entailed microwaving (GE, 1.55 kW microwave) slides in citrate retrieval buffer [2.4 g/L sodium citrate tribasic dehydrate (Sigma, S4641), 0.35 g/L citric acid (Sigma, C0759), pH 6.0] at 100% power to induce boiling for 2–3 min and then at 20% power for 7–8 min. Slides were allowed to cool at room temperature for 30 min. Slides were then washed in 1 × DPBS (Sigma, D5652-10L) 3 times for 5 min each. Staining for wheat germ agglutinin (WGA), Isolectin B4, and DAPI: Precautions were taken to shield slides from light exposure during the staining process. Slides were incubated with 1 × WGA (ThermoFisher, W32464) for 30 min at room temperature and were subsequently washed with 1 × DPBS 3 times for 3 min each. Slides were then incubated with 1:25 Isolectin B4 (Vector Labs, FL-1201) for 1 h at room temperature. Slides were then washed 3 times for 3 min each with 1 × DPBS. In the final wash step, 50 μL of DAPI (1 mg/mL, ThermoFisher, D3571) was added to the 1 × DPBS. Slides were again washed with 1 × DPBS 3 times for 3 min each. Next, to reduce autofluorescence of the heart muscle, slides were incubated with Sudan Black [1 mg/mL in 70% ethanol (Sigma, 199664)] for 30 min. Slides were then washed 6 times for 3 min each with 1 × DPBS. Finally, slides were mounted, dried, stored at 4°C and protected from light until confocal imaging. Cardiomyocyte hypertrophy and capillary density were visualized using a Nikon TE-2000E microscope interfaced with a Nikon A1 confocal system. Slides were imaged with a 60 × objective and excited in series with a 405 nm laser for DAPI, a 488 nm laser for Isolectin B4, and a 561 nm laser for WGA. Emission was band-pass filtered through 450/50, 525/50, and 595/50, respectively. Cardiomyocyte areas were determined in cardiomyocytes with centrally located nuclei. All confocal analyses were blinded and performed using Nikon Elements software [64-bit, version 3.22.00 (Build 710)].
Results are shown as mean ± SD. The statistical analysis (GraphPad 5.0d) was conducted using student's
Although many groups have studied the role of c-kit-sorted cardiac cells for treatment of heart failure, data regarding the maintenance of c-kit expression during expansion is rare. First, we isolated CMCs from mouse hearts using standard isolation procedure (described in details in Materials and Methods section; see Supplementary Figure
Positively sorted cells were plated at a density of 2500 cells per cm2 and expanded. At ~75% confluence, cells were passaged (P) and evaluated for c-kit and Sca-1 expression (up to 11 passages). CMCs were initially ~70% c-kitpos, which gradually declined with further passaging (Figure
Next, c-kit expression, pluripotency, and cardiac markers were evaluated by qPCR in cells from P1 and P5. There was a significant decline in c-kit, pluripotency markers (
The standard isolation protocol is based on the principle that c-kitpos CMCs possess adherent properties; however, no systematic study has determined the time necessary for c-kitpos CMCs to adhere to plastic. Here, we performed sequential plating of freshly digested mouse hearts (described in Materials and Methods; Supplementary Figure
Next, the expression of pluripotency and cardiac markers was evaluated by qPCR in the RA and SA cells. Similar to c-kit expression, SA cells expressed higher levels of pluripotency markers,
Because the standard method of c-kit cell sorting does not allow for monitoring of cell purity and sorting efficiency directly after sort, we modified the c-kit cell sorting procedure. The typical one-step method with primary antibody conjugated with magnetic beads was replaced with a two-step method. The primary antibody directed against c-kit conjugated with FITC and secondary anti-FITC antibody conjugated with magnetic beads. This allowed for a small sample of sorted cells to be evaluated for purity by flow cytometry and the remaining cells were expanded for further experiments (Supplementary Figure
Next, we isolated RA and SA CMCs and expanded them. Once they reached 70% confluence, c-kitpos cells were sorted with the aforementioned, improved two-step method. The average purity of sorted cells was 88% for both RA and SA CMCs (Figure
Clonogenicity is a hallmark of stem/progenitor cells, and because c-kit has been touted as a cardiac stem cell marker, we queried potential clonogenic differences between SA c-kitpos and RA c-kitpos CMCs. The maintenance of c-kit expression in the SA c-kitpos CMCs led us to speculate that they are more primitive than RA c-kitpos CMCs. SA and RA c-kitpos CMCs were expanded one passage after sorting and Terasaki plates were used to test their clonogenicity. Individual cells were seeded per well, which was verified via microscopy. Single cells were incubated for 14 days and evaluated for clonogenicity; wells with 30 or more cells were considered to be clonogenic. The 1′ cells had lowest clonogenicity of 0.6%, which increased to 1.9% in 2′, 5.0% in 3′, 6.6% in 4′ and 6.1% in 5′. Overall, the clonogenicity of RA c-kitpos CMCs was lower than SA c-kitpos CMCs (Figure
Cell primitiveness is correlated with pluripotency marker expression. SA c-kitpos CMCs mRNA levels of
After the cell populations were carefully characterized
During our pilot studies to establish the percutaneous delivery protocol, we spent significant time optimizing the positioning of the probe (more suprasternal position than typical PLAX echocardiograms in mice), in coordination with the needle's angle of attack. The surgeons and sonographers worked together to establish the approach of the needle into the visual field of the echocardiography probe, while negotiating an intercostal access point. For those attempting this in their own laboratories, the use of echo-opaque contrast agent should help in confirming the locus of injection; however, the injection of the cells provided sufficient monographic disturbance to confirm their administration within the left ventricular lumen. Although we injected cells into the left ventricular lumen, this technique can also be employed for intramyocardial (i.e., directly into the muscle) injections; we confirmed the feasibility of the intramyocardial approach during our pilot studies to establish this technique. Although the focus of the percutaneous delivery of cells naturally centers on the specific aspect of manual administration of the cells, the success of this technique is undoubtedly predicated on the quality of cells being prepared for injection. As one last cautionary note, it was critical to position the warming lamps so that the cells were not exposed to excessive heat.
Two days after MI, mice were subjected to an echocardiogram to confirm sufficient depression of cardiac function (LVEF < 50%). All mice included in the study had significantly depressed cardiac function, and there were no differences among the groups at treatment (2 days post-MI). At 37 days post-MI (35 days following cell/vehicle injection), mice were subjected to their final echocardiograms. Mice treated with RA cells did not significantly improve cardiac function compared with vehicle treated mice; however, SA cell treated mice exhibited significantly higher ejection fractions (Figure
To address the issue of how the cells might improve cardiac function, we examined potential changes in vascular density. Using isolectin B4 staining, capillaries were counted in mid-ventricular cardiac sections at 37 days post-MI (35 days following cell/vehicle injection). Following infarction, there was a clear reduction in capillary density in the ischemic and border zones compared to the remote zone. In the ischemic zone, SA CMC treatment significantly augmented capillary density compared to vehicle treatment (Figure
We also assessed cardiomyocyte cross-sectional area in the same sections as the capillary counts. In the ischemic zone, myocyte cross-sectional areas were relatively preserved in the SA CMC (292 ± 46 μm2), but not the RA CMC (213 ± 35 μm2) or the vehicle (181 ± 21 μm2). In the border zone, cross-sectional areas were similar among the vehicle (365 ± 33 μm2), RA CMC (280 ± 40 μm2), and SA CMC (290 ± 36 μm2). In the remote zone, the myocyte cross-sectional areas were virtually indistinguishable among the vehicle (299 ± 34 μm2), RA CMC (303 ± 32 μm2), and SA CMC (281 ± 12 μm2). Thus, most of the pathologically identifiable changes associated with SA CMC treatment occurred in the ischemic and border, but not remote, zones.
The primary goal of this study was to establish a reproducible protocol to maintain c-kit expression during passage of CMCs, and to confirm whether the resulting cells had reparative potential. We found that differential plating enriched c-kit positivity of CMCs. This observation in the so-called SA CMCs led us to query whether these cells represented a more refined population of CMCs that are likelier to maintain c-kit expression. To address this possibility, we modified an immunologic sorting technique and passaged the resulting cells in conjunction with the differential plating protocol described in the Materials and Methods. The resulting (c-kit sorted) SA cells reproducibly retained high levels of c-kit expression. These findings, alone, are important for other investigators in the field.
After identifying a valid approach to stabilize c-kit expression, we characterized the cells and identified a number of immunophenotypic features of the SA CMCs, which are putatively favorable for use in adoptive transfer studies. There were several aspects of the
Because this study focused on c-kit sorted cells (though potentially a different cell population), we must acknowledge some of the disagreement in the field. There seems to be a popular notion that some studies (e.g., van Berlo et al.,
Because we had not previously observed significant transdifferentiation of our injected cells (Keith and Bolli,
Collectively, cell therapy studies have used a menagerie of cells. Yet, most of these cells do not convincingly transdifferentiate into significant numbers of cardiomyocytes, though they do improve cardiac function (Keith and Bolli,
The lack of a beneficial effect of RA CMCs is interesting for several reasons. These data indicate that there are populations of c-kit-sorted cells that give rise to non-reparative cells, which is an innovative concept (i.e., c-kit sorting
We predict that c-kit sorted cells may contain cells that participate beneficially, neutrally, or antagonistically to cardiac repair, and the differential plating step we described here significantly enriches for the reparative population(s) of c-kit sorted cells. Indeed, the present data indicate that cells isolated based on c-kit positivity are not necessarily reparative (i.e., c-kit sorted RA cells); perhaps the reparative fraction of c-kit cells is all (or largely) represented in the SA population. It is conceivable that such technical differences may explain apparent discrepancies in previous studies of adoptive transfer of c-kit-sorted cells. One of the motivations for performing this study was that the expression of c-kit was required for the reparative effects of SA CMCs. Although this was not tested specifically, if we assume a different view of the conclusions, we may argue that the RA/SA segregation was more important than c-kit sorting. In other words, sorting for c-kit may have been irrelevant for the reparative effects we report here; this is the subject of current efforts in the laboratory.
In its simplest form, the present study establishes a refined technique to enrich for reparative c-kit sorted cells (i.e., SA cells), and employs them in a refined, minimally invasive model of syngeneic adoptive transfer. Yet, in a broader context, this study poses new questions regarding the absolute requirement based on sorting for cell markers. We predict that the segregation of cells based on their adherent phenotype, which may be a proxy for reparative vs. non-reparative cells, may be a new and singularly sufficient approach to cardiac cells with the potential to repair the failing heart. Even if such speculation is not ultimately validated, the combination of c-kit sorting with differential plating yields a population of cells with clear reparative capacity.
Conceived study, designed experiments, composed manuscript (MW, SJ); Financial support (MW, SJ, RB); Edited manuscript (MW, SJ, BL, RB, SD, AZ, SG); Performed experiments and analyzed data (MW, SD, AZ, SG, BL, CN, AD, KB).
This work was supported by National Institutes of Health Grants R01 HL083320, R01 HL094419, R01 HL131647 (to SJ), P20 GM103492, P01 HL078825 (to RB, SJ, MW), and UM1 HL113530 (to RB); an American Heart Association (Great Rivers Affiliate) Predoctoral Fellowship 14PRE19710015 (to SD), an American Heart Association (Great Rivers Affiliate) Postdoctoral Fellowship 14POST18870020 (to AZ), and an American Heart Association Scientist Development Grant 13SDG14560005 (to MW).
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The authors thank Ms. Linda Harrison and Mr. Gregory Hunt for their technical assistance during the course of these studies.
The Supplementary Material for this article can be found online at: