SEARCH

SEARCH BY CITATION

Keywords:

  • Cell transplantation;
  • In vivo tracking;
  • Mouse;
  • Embryonic stem cells;
  • Heart

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Despite rapid advances in the stem cell field, the ability to identify and track transplanted or migrating stem cells in vivo is limited. To overcome this limitation, we used magnetic resonance imaging (MRI) to detect and follow transplanted stem cells over a period of 28 days in mice using an established myocardial infarction model. Pluripotent mouse embryonic stem (mES) cells were expanded and induced to differentiate into beating cardiomyocytes in vitro. The cardiac-differentiated mES cells were then loaded with superparamagnetic fluorescent microspheres (1.63 μm in diameter) and transplanted into ischemic myocardium immediately following ligation and subsequent reperfusion of the left anterior descending coronary artery. To identify the transplanted stem cells in vivo, MRI was performed using a Varian Inova 4.7 Tesla scanner. Our results show that (a) the cardiac-differentiated mES were effectively loaded with superparamagnetic microspheres in vitro, (b) the microsphere-loaded mES cells continued to beat in culture prior to transplantation, (c) the transplanted mES cells were readily detected in the heart in vivo using noninvasive MRI techniques, (d) the transplanted stem cells were detected in ischemic myocardium for the entire 28-day duration of the study as confirmed by MRI and post-mortem histological analyses, and (e) concurrent functional MRI indicated typical loss of cardiac function, although significant amelioration of remodeling was noted after 28 days in hearts that received transplanted stem cells. These results demonstrate that it is feasible to simultaneously track transplanted stem cells and monitor cardiac function in vivo over an extended period using noninvasive MRI techniques.

Disclosure of potential conflicts of interest is found at the end of this article.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Stem cell therapies have the potential to radically improve current treatment options for patients suffering from heart disease and cardiac dysfunction [1, 2]. Recent advances in stem cell research have created much excitement and interest in this field, spawning many exciting new developments [3, [4], [5]6] and numerous clinical trials [7, [8], [9]10]. Different types of stem cells are being tested, including adult stem cells isolated from bone marrow [11, [12], [13], [14], [15], [16], [17]18], adipose [19, 20], and cardiac [21, [22], [23]24] tissues. Other studies have tested pluripotent [25, [26]27] and differentiated [28, [29]30] embryonic stem cells.

The results from these studies have been mixed. In some cases, dramatic improvements in cardiac function have been reported [12, [13]14, 25, 31], but in other cases there appeared to be little or no improvement [16, 32, [33]34]. Moreover, despite some reports showing that the stem cells can differentiate into cardiomyocytes in the injured heart, the relative number of these cells has remained rather small, and there is growing suspicion that some of the observed functional improvements in cardiac performance may be due to effects other than regeneration of myocardial tissue [35, [36], [37]38].

Many of the stem cell studies evaluate cardiac function in vivo to determine whether the stem cell therapeutic intervention results in significant improvement in performance. Subsequent immunohistochemical analysis of post-mortem tissue sections is then generally conducted to identify the stem cells and determine the degree, if any, of myocardial regeneration/repair. Although this approach is effective in animal studies, it has limited utility for clinical studies. Thus, for both animal studies and clinical trials, it would be advantageous to be able to detect and evaluate stem cells over time in vivo.

At present, relatively few studies have attempted to identify and track stem cells in the heart in vivo using noninvasive imaging methodologies. Internal bioluminescence has been used successfully in small animals [8, 39, 40], although low spatial resolution is a limiting factor with this technique. Positron emission tomography improves spatial resolution [39, 40], but short-lived radioisotopes preclude long-term studies [41]. Currently, the most promising long-term, high-resolution cell-tracking technique is T2-weighted magnetic resonance imaging (MRI) of iron-labeled cells. Transplanted stem cells of various origins have been labeled by incubation with iron-oxide nanoparticles and tracked in murine [42, 43], rat [44, 45], and swine [46, [47]48] hearts in vivo. Potential concerns regarding iron nanoparticles are that they may be toxic to the loaded cell [49] and that their collective signal strength can weaken with cellular metabolism [50] or mitosis [49]. Another potential concern is that the particles may remain after the cells die and may either remain in situ or be taken up by macrophages and/or other types of cells, which could lead to nonspecific tracking. To resolve limitations inherent with nanoparticle strategies, the feasibility of survival, differentiation, and location of mesenchymal stem cells loaded with larger iron-containing microspheres has been successfully demonstrated in murine [51] and swine [52, 53] hearts.

In the present study, we have used relatively large (1.63 μm) superparamagnetic microspheres as contrast agents to detect transplanted mouse embryonic stem (mES) cells using noninvasive MRI techniques. An advantage of this approach is that we can identify transplanted stem cells and simultaneously measure cardiac performance indices repeatedly in the same animals over an extended period of time. Our results indicate that this strategy is effective and may constitute a generally feasible approach for similar studies and other therapeutic interventions involving stem cell transplantation.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Superparamagnetic Microspheres

Superparamagnetic microspheres (Bangs Laboratories, Fishers, IN, http://www.bangslabs.com) were used as the magnetic resonance (MR) contrast agent. These spheres of magnetite (Fe3O4) are coated with an iron-free polymer shell and tagged with fluorescent dyes. Two types of superparamagnetic microspheres were used for this study. The first were the COMPEL Flash Red (excitation:emission peaks, 660:690) microspheres (catalog code UMC3F) that had a mean diameter of 2.79 μm and a magnetite composition of ∼12.2%. The second were the 1.63-μm Dragon Green (excitation:emission peaks, 480:520) microspheres (catalog code MEO3F), which harbored magnetite compositions of ∼42.5%. The microspheres were kindly provided by Drs. Eric Shapiro and Alan Koretsky of the Laboratory of Functional and Molecular Imaging (National Institute of Neurological Disorders and Stroke, NIH, Bethesda, MD).

Cardiac Differentiation of mES Cells

Pluripotent mES cells (R1-derived) [54] were induced to undergo cardiomyocyte differentiation by the hanging-drop method [55] in growth medium (Dulbecco's modified Eagle's medium supplemented with 25 mmol/l d-glucose, 1 mmol/l sodium pyruvate, 15% fetal bovine serum, 55,000 U of penicillin G, 55 mg of streptomycin, 55 μM β-mercaptoethanol, 5.75 mM glutamine, and 1% of 100× α-minimum essential medium nonessential amino acids [Gibco, Grand Island, NY, http://www.invitrogen.com]). Briefly, 400 mES cells were suspended in 20-μl droplets from the lid of a 150-mm culture dish for 2 days to produce EBs, followed by 5 days in suspension (7 + 0 days [d]) [55, 56]. EBs were plated in T-75 flasks pretreated with fibronectin and cultured in growth medium plus 1 nM endothelin-1 at 37°C until intracellular loading of microspheres (7 + 1d). Approximately 1%–2% of the mES cells started to beat by this stage of the differentiation process, consistent with earlier studies that described the development of this method for inducing cardiac differentiation in vitro [57].

Intracellular Microsphere Loading of mES Cells

At 7 + 1d of cardiomyocyte differentiation, the microspheres were applied directly to the cultures at a 1:1,000 dilution from the commercial stock for 2–4 hours at 37°C. At the end of this incubation period, the microsphere-containing medium was removed and the cells were exposed to a 15% dimethyl sulfoxide boost for 2 minutes in serum-free medium to facilitate permeabilization of the plasma membrane. Fresh growth medium was then applied and the cells were allowed to recover for at least 24–48 hours. In some experiments, loaded cells were separated from unloaded cells using a MagneSphere magnetic separation stand (Z5343; Promega, Madison, WI, http://www.promega.com) for 30 seconds.

Myocardial Infarction and Intramyocardial Injections

This study conformed to the Guide for the Care and Use of Laboratory Animals (NIH publication no. 85-23, 1985) and was approved by Institutional Animal Care and Use Committee of the University of Virginia. The surgical induction of myocardial infarction was performed as described previously [58, 59] in 16 adult male C57BL/6 mice, ∼25 g (Jackson Laboratory, Bar Harbor, ME, http://www.jax.org). Briefly, mice were anesthetized with sodium pentobarbital (100 μg/g i.p.) and placed in a supine position. The third and fourth ribs and intercostal muscles were cut with a cautery pen, and the pericardial membrane was removed to provide access to the left anterior descending coronary artery (LAD). The LAD was occluded with 7-0 silk suture for 60 minutes, after which time the ligature was removed to permit reperfusion. Eight of the 16 mice were left uninjected, whereas the remaining 8 were injected with microsphere-loaded cardiac-differentiated mES cells at 30 minutes after reperfusion. The stem cells were injected into opposite edges of the infarcted left ventricular wall via two separate injections of ∼10 μl each (∼75,000 cells per injection). Half of the mES cell-injected mice (n = 4) were given ATL313 (Adenosine Therapeutics, LLC, Charlottesville, VA, http://www.adenrx.com; 8 μg/kg body weight, i.p.) 15 minutes prior to mES cell injection. ATL313 is a highly selective agonist of the adenosine 2A receptor with potent anti-inflammatory properties [60] that was applied here in an attempt to protect the transplanted mES cells against reperfusion injury. After mES injection, the chest wall was closed in layers around a chest tube left in the thoracic cavity to remove air and fluids under syringe vacuum. The chest tube was removed, and the mouse was given 100% O2 via nasal cone upon the return of spontaneous breathing.

MRI

Baseline cardiac parameters were obtained 1 day prior to infarction surgery, and experimental readings were obtained 1, 7, and 28 days following surgery, as described previously [58, 59]. Briefly, mice were anesthetized with isoflurane (1% by volume in oxygen), and the forelimbs were shaved for the attachment of ECG surface electrodes. During MR image acquisition, anesthesia was maintained with isoflurane, and body temperature was maintained at 37°C ± 1°C using circulating warm water.

MR image acquisition was triggered by the R-wave in the ECG signal. Imaging was performed on a horizontal bore, Varian Inova 4.7T MR scanner with Magnex gradients (800 mT/m maximum strength with 667 mT/m × millisecond rise rates at 100% gradient switching) (Varian, Palo Alto, CA, http://www.varianinc.com/cgi-bin/nav?). For orthogonal long-axis images, a two-dimensional (2D), cine-fast low-angle shot (FLASH) sequence with first-order gradient moment rephasing (GMR) in the readout and slice-select directions was performed. To obtain contiguous 1-mm-thick, short-axis images, a 2D, cine-FLASH sequence (without GMR) was performed from apex to base. Flip angle was 30°. Echo time was 3.2 milliseconds, and repetition times were 8–10 milliseconds, yielding 14 phase acquisitions per cardiac cycle. Field of view was 30 mm. For contrast-enhanced imaging performed on day 1 post-myocardial infarction (post-MI), a 0.3–0.6 μmol/g bolus of gadolinium-diethylenetriaminepentaacetic acid (Gd-DTPA) was injected through the i.v. line while the mouse was inside the magnet, and images were acquired after a 15-minute delay by increasing the flip angle to 60 degrees to enhance T1 weighting [36].

MR Image Analysis

MR images were converted from free induction decay to Digital Imaging and Communications in Medicine format and processed using a set of MATLAB image analysis tools developed in-house. Briefly, left ventricular ejection fraction (LVEF) was calculated from left ventricular end-systolic (LVES) and end-diastolic (LVED) volumes as determined from endocardial contours on contiguous, 1-mm-thick image slices covering the entire heart from apex to base [35]. Left ventricle (LV) mass was similarly determined from endocardial and epicardial contours. MI size was determined by threshold analysis of Gd-DTPA-enhanced infarct regions and was expressed as percentage of LV mass [36].

Statistical Analyses

Data are expressed as mean ± SEM. Statistical comparisons were made using either the Student t test for single comparisons or one-way analysis of variance for multiple comparisons, with p < .05 required to reject a null hypothesis.

Post-Mortem Analysis

Mice were euthanized after the final MRI session at 4 weeks post-MI. The hearts were removed, fixed with 4% formalin in phosphate-buffered saline, embedded in paraffin, and cut into serial 10-μm sections. Images were viewed and collected using a Fluoview FV300 laser-scanning confocal fluorescence microscope (Olympus, Tokyo, http://www.olympus-global.com). Standard H&E histological staining was subsequently performed to assess cell/tissue integrity.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Stem Cell Loading with Superparamagnetic Microspheres

To identify and track transplanted stem cells in vivo, we first attempted to load superparamagnetic microspheres into cultured mES cells that had been induced to differentiate into beating cardiomyocytes in culture. The loading conditions were initially established using Flash Red COMPEL microspheres (supplemental online Fig. 1). In parallel preliminary experiments, it was determined that microspheres containing higher magnetite compositions were preferred over those with lower magnetite compositions since they were more readily detectable by MRI. Thus, we adopted the smaller Dragon Green (1.63 μm in diameter) rather than the larger Flash Red (2.79 μm) microspheres for the in vivo MRI studies because the magnetite composition of the Dragon Green particles was ∼42.5% compared with only ∼12.2%, on average, for the Flash Red particles.

We were able to effectively load cardiac-differentiated mES cells with Dragon Green microsphere particles using the dimethyl sulfoxide boost protocol (Fig. 1). A series of confocal images was collected and images were overlaid to provide a more three-dimensional rendering of the microsphere-loaded cells than would otherwise have been possible when viewing a single image plane. Note that the nuclei appear relatively dark (i.e., devoid of microspheres), whereas the cytoplasmic domains of the cells contain many microspheres (Fig. 1A). The number of microspheres per cell varied greatly from 0 to more than 100. Examples of mES cells loaded at lower density are shown in Figure 1B. These mES cells were dissociated from beating EBs after being loaded with the microspheres. The loaded cells were then magnetically separated from the unloaded cells, and the two groups were cultured in parallel for several days. Although we could readily identify the heavily loaded cells, such as those shown in Figure 1A, we also found numerous lightly loaded cells, such as those shown in Figure 1B, where there was often only one or a few microspheres per cell. By counting the number of loaded versus unloaded cells (magnetic separation verified by careful microscopic inspection), we were able to estimate that approximately 20% of the mES cells in these cultures were effectively loaded with one or more microspheres.

thumbnail image

Figure Figure 1.. Cardiac-differentiated mouse embryonic stem (mES) cells loaded with superparamagnetic Dragon Green microspheres (1.63 μm in diameter) in culture. Pluripotent mES cells were seeded onto collagen-coated glass coverslips, induced to differentiate into beating cardiomyocytes using the hanging-drop method [55] for 7 + 1 days, and loaded with the microspheres as described in Materials and Methods. The medium containing the unincorporated microspheres was removed, and the cells were rinsed with fresh medium before being refed with fresh growth medium. Following a 2-day recovery period, the cells were fixed with 4% paraformaldehyde in phosphate-buffered saline, pH 7.4, and analyzed by laser-scanning confocal fluorescence microscopy. (A): A series of 7 × 1 μm optical sections were captured and stacked to provide a more three-dimensional view of the microspheres (green) inside the cells (indicated by arrows). Note that the microspheres are primarily cytoplasmic and that the nuclei are dark. (B): Overlay of phase-contrast and fluorescence microscopy images showing dissociated cells containing only one or a few microspheres per cell (in contrast to the group of heavily labeled cells shown in [A]).

Download figure to PowerPoint

Importantly, the cardiac-differentiated mES cells continued to spontaneously contract in culture after microsphere loading. An example of this can be seen in Figure 2 (still-shot single video frame) and the corresponding online video clip (supplemental online Fig. 2A). In Figure 2A, the microspheres were visualized in a beating EB using reverse phase-contrast microscopy such that the microspheres appeared white (Fig. 2A, arrows). In a second example, we dissociated the cells from the beating EBs and were able to identify a few small groups of microsphere-loaded cells that were clearly beating (supplemental online Fig. 2B). A still shot from this video clip is shown in Figure 2B in normal phase contrast such that the microspheres appeared dark (Fig. 2B, arrow). These data indicate that cardiac-differentiated mES cells were capable of sustaining beating activity for up to 7 days (the longest period evaluated in vitro) or more after loading of the Dragon Green superparamagnetic microspheres.

thumbnail image

Figure Figure 2.. Video snapshots of Dragon Green superparamagnetic microspheres in beating mouse embryonic stem (mES) cells in culture following cardiac differentiation. Phase-contrast microscopic still images were captured from a DVD recording of beating cardiomyocytes loaded with the supermagnetic microspheres. (A): The phase-contrast images were reversed to enhance visualization of microspheres as white spheres (indicated by arrows) within the gray-appearing cells of the beating EB. A video recording of these beating mES cells in culture is given in supplemental online Figure 2A. (B): Video-still photograph of dissociated beating EB cells loaded with microspheres (arrow), as shown in normal phase contrast. A video recording of these beating mES cells in culture is given in supplemental online Figure 2B.

Download figure to PowerPoint

Tracking Stem Cells In Vivo

To determine whether microsphere-loaded cardiac-differentiated mES cells could be identified and followed (tracked) in vivo after transplantation into an infarcted region, we performed cardiac MRI on mice that had received such treatment. Scans were performed on days 1, 7, and 28 following transplantation. Representative images showing long and short axes from the same heart are shown in Figure 3. The arrows indicate regions of the heart containing the transplanted mES cells that were loaded with superparamagnetic microspheres, which appear as dark regions along the borders of the infarct zone. It is also possible to observe the progression of ventricular remodeling that developed as a consequence of the ischemia-reperfusion injury, including thinning of the anterior wall of the left ventricle and expansion of the left ventricular chamber due to infarct expansion and left ventricular wall thinning (Fig. 3, compare day 1 vs. days 7 and 28).

thumbnail image

Figure Figure 3.. In vivo magnetic resonance images of mouse hearts injected with superparamagnetic microsphere-loaded, cardiac-differentiated mouse embryonic stem cells. Scans were performed at 1, 7, and 28 days post-myocardial infarction in both the long-axis (top row) and short-axis (bottom row) orientations. Iron-rich regions containing microspheres are shown as dark areas (arrows) in the left ventricular wall in or near the infarct zone.

Download figure to PowerPoint

Another way to assess these images is to contour and assemble a three-dimensional stack of 2D short-axis (transverse) MR image planes. An example of such an analysis is shown in Figure 4. In the mice imaged at day 1 post-MI, Gd-DTPA was administered as a contrast-enhancement agent to aid in the identification and quantification of the infarcted region(s), which appear white in these images. The blue regions represent concentrations of superparamagnetic microspheres and presumably the mES cells that harbored them. The red regions indicate intact cardiac muscle. These data demonstrate that superparamagnetic microspheres remain in and around the infarcted region throughout the 28-day evaluation period.

thumbnail image

Figure Figure 4.. Three-dimensional reconstructions from in vivo magnetic resonance imaging scans of a mouse heart containing superparamagnetic microsphere-loaded mouse embryonic stem (mES) cells. Scans were performed at 1, 7, and 28 days post-myocardial infarction. On day 1, gadolinium-diethylenetriaminepentaacetic acid was used as a contrast agent to determine infarct size (white regions in left ventricle wall). Iron-rich regions containing the transplanted mES cells loaded with the superparamagnetic microspheres are represented by the blue regions, and intact muscle is shown in red.

Download figure to PowerPoint

LV Remodeling and Cardiac Performance

To determine whether transplantation of cardiac-differentiated embryonic stem cells affected LV remodeling and cardiac performance, we measured various cardiac parameters at baseline prior to infarction and transplantation and then again at days 1, 7, and 28 post-transplantation. On day 1 post-MI, delayed enhancement MRI was performed using Gd-DTPA to enhance infarcted regions of myocardium. Infarct size as percentage of LV mass was determined by image analysis of the complete stacks of contiguous, 1-mm-thick image, short-axis image slices. The mean infarct size in the mES cell-treated group of mice (37.7% ± 0.5% LV mass; n = 8) was similar to that of the untreated control group (38.1% ± 1.5% LV mass; n = 8). Half of these mice (n = 4) had been given an anti-inflammatory drug (ATL313; Adenosine Therapeutics) 15 minutes prior to the time of stem cell transplantation. However, no significant differences were observed between the mice that had received the ATL313 compound and those that had not, so the two subgroups were combined as one mES-treated group (n = 8) and used for comparative analyses with an untreated (i.e., no mES cells) control group of mice (n = 8) that had also been subjected to the LAD ligation/reperfusion procedure.

As shown in Figure 5, LVED and LVES volumes increased markedly in both mES cell-treated and untreated groups between days 1 and 7. Interestingly, both LVED and LVES volumes decreased slightly between days 7 and 28 in mES cell-treated mice, with LVED volume showing a significant absolute difference from control (p < .05; n = 8 per group), whereas chamber volumes in untreated mice showed progressive increases in size. Comparison of the degree of change between days 7 and 28 for both volume measurements shows much more dramatic and highly significant differences between the untreated and mES-treated groups. For example, LVED volume increased by 21.3% ± 3.5% in the untreated group but decreased 7.8% ± 4.2% in the mES-treated group between days 7 and 28 (p < .001; n = 8/group). Similarly, LVES volume increased by 19.7% ± 4.4% in the untreated group but decreased −3.9% ± 3.3% in the mES-treated group over the same 2-week period (p < .001; n = 8/group). In contrast, LVEF declined within 1 day after infarction in both groups and changed little in either group thereafter. These results indicate that treatment with mES cells protected hearts against the late phase of LV remodeling (between days 7 and 28 postinfarction in mice) but did not appear to improve cardiac performance over this same time period.

thumbnail image

Figure Figure 5.. Evaluation of LV chamber volumes and ejection fraction at baseline and 1, 7, and 28 days post-myocardial infarction (post-MI). LV chamber volumes and ejection fraction were measured using cardiac magnetic resonance imaging analyses as previously described [58, 59]. (A): Left ventricular end-diastolic (LVED) volume (mm3) measurements. (B): Left ventricular end-systolic (LVES) volume (mm3) measurements. (C): Left ventricular ejection fraction (percentage). Untreated mice (solid lines) illustrate the progressive increases in LVED and LVES volumes typically observed during the first 28 days following MI in mice. Mouse hearts transplanted with superparamagnetic microsphere-loaded, cardiac-differentiated mES cells behaved similarly during the first 7 days post-MI but then stabilized, illustrating resistance to the late phase of LV remodeling in mice (between days 7 and 28 post-MI). *, significantly different from control (untreated group at same time point); p < .05; n = 8 per group. Abbreviations: LV, left ventricle; mES, mouse embryonic stem.

Download figure to PowerPoint

Post-Mortem Analysis

To verify that the superparamagnetic microspheres remained associated with cells in the infarcted region during the 4-week study period, the hearts were collected and fixed for histological evaluation when the animals were euthanized after the 28-day assessment in vivo. The tissues were processed either for standard histological evaluation (H&E staining) or for fluorescence microscopy. As shown in Figure 6A, the infarcted zone could easily be identified in H&E-stained sections because of differential staining of infarcted tissue (light purple) compared with the surrounding viable myocardium (deep red). Within the infarcted region (Fig. 6A, boxed portion of the low-magnification panel), two cells that each contained a 1.63-μm Dragon Green fluorescent superparamagnetic microsphere were identified (Fig. 6B, 6C, arrows). The image depicted in Figure 6B was intentionally overexposed to enable visualization of the surrounding cells and tissue due to autofluorescence. Although there are several bright spots apparent in this image (Fig. 6B), most are artifacts arising from autofluorescence. This was confirmed by switching the excitation-emission filter sets into the red spectrum, which typically captures autofluorescence artifacts similar to those seen in the green spectrum but does not capture the Dragon Green particles (not shown). A second confirmation was achieved by reducing the exposure time in the green spectrum to minimize autofluorescence, thereby permitting capture of Dragon Green-specific fluorescence from the microspheres, as shown in Figure 6C (arrows). Note also that the microspheres appear almost perfectly round and are of the approximate expected size (1.63 μm in diameter). Numerous such examples could be found, and most of these were similar in that only one or a few microspheres per cell were observed, as also shown in Figure 6D and 6E. These results demonstrate that the Dragon Green superparamagnetic microspheres that were delivered to the ischemic mouse heart in transplanted mES cells could still be found in the ischemic region at least 4 weeks after transplantation.

thumbnail image

Figure Figure 6.. Identification of cells containing superparamagnetic microspheres in mouse heart sections isolated at 28 days post-myocardial infarction (post-MI). (A): Low-magnification H&E section of an infarcted region of mouse left ventricle. The dark red regions (right) indicate intact (normal) myocardial tissue, whereas the lighter-stained regions (left) depict mature scar tissue resulting from MI. The boxed region is magnified in (B) and (C), which show identical fluorescent images processed for long and short exposure periods, respectively. Arrows indicate the locations of two cells containing one Dragon Green supermagnetic microsphere each. (D): Low-magnification view from another mouse heart containing mature scar tissue as a result of MI. Note that in this fluorescent image, intact muscle tissue appears bright, whereas scar tissue appears relatively dark. The boxed region is depicted in a magnified view in panel (E), where a cell containing a Dragon Green superparamagnetic microsphere was found near the epicardial surface of the ventricle. Scale bars = 10 μm. Abbreviation: BV, blood vessel.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

The data presented in the present study demonstrate that (a) beating cardiac-differentiated mES cells were effectively loaded with fluorescent superparamagnetic microsphere particles, (b) such loading did not appear to adversely disturb the ability of the myocytes to contract, (c) the microsphere-loaded mES cells were transplanted into the mouse heart and tracked in vivo using noninvasive MRI techniques over a period of at least several weeks, (d) simultaneous assessments of cardiac performance were performed using the same MRI instrumentation, and (e) transplanted microsphere-containing mES were found in the heart over the entire 28-day study period as shown by MRI and confirmed by post-mortem histological analysis.

Superparamagnetic iron oxide is a readily detected MR contrast agent because of its high T2 relaxivity [46]. Currently, the most-studied form of intracellular iron loading has used nanoparticles, 3–150 nm in diameter [61, 62]. Various coatings, such as carbohydrate dextran, are frequently used to prevent particle aggregation [46] and acute iron toxicity, but the enhanced biodegradability [49, 63] causes the MR signal strength to diminish as the iron is metabolized [50]. Signal strength may also be reduced as intracellular nanoparticle concentration declines within an imaging voxel, which can occur with cell proliferation or exocytosis [49].

The recent introduction of coated magnetite microspheres potentially addresses these concerns. Originally designed to magnetically isolate cells and biomolecules by reactive surface molecule interactions [64], internalized microspheres are resistant to metabolism and exocytosis [51] and have been detected in vivo at concentrations as low as one microsphere per cell [65]. Thus, it is possible to detect individual labeled cells in vivo [66], thereby suggesting that these coated superparamagnetic microspheres may be ideally suited for long-term cell tracking.

ES Cell Loading with Microspheres

Microsphere-loaded cells have been imaged in the larger mammalian heart for up to 3 [52] and 16 [51] weeks, but microsphere effects on cardiac physiological characteristics have not been quantified in the living organ during that period, nor was the fluorescence capacity of microspheres visualized post mortem. Moreover, the effects of microsphere loading on stem cell cardiac differentiation and function were previously unknown.

Our results show that beating activity persisted for several days in culture following loading of cardiac-differentiated mES cells with the microspheres, thus indicating that there was no apparent adverse effect from either the loading process or the microspheres themselves on the physiological performance of these cells. It is important to note, however, that beating activity is a rather crude index, and more in-depth evaluation of cardiomyocyte function will need to be examined in future studies to fully evaluate the potential influence (or lack thereof) of superparamagnetic microsphere loading on these cells.

Another issue associated with microsphere loading was the variability in the number of microspheres loaded per cell. Depending on the size and type of microspheres used, the number varied from 0 to more than 100 per cell. In addition, the differentiation and physiological status of the cells in culture may have also influenced which cells ultimately took up the microspheres. For example, even though the entire culture is induced to differentiate, only a relatively small percentage of the total number of mES cells actually began to beat, which is typical for this culture system [56]. Nonbeating cells may be undifferentiated or in an earlier or perhaps an alternate state of differentiation. Clear morphological and functional differences exist in the characteristics of the cardiac-differentiating mES cultures. It is, therefore, conceivable that some of these different cellular qualities may have influenced their ability to take up the microspheres.

Although many of the mES cells were maximally or near-maximally loaded, we also found many cells that had far fewer microspheres and some that had none. In our in vivo experiments, we made no effort to separate cells that contained microspheres from those that did not. Consequently, it is potentially noteworthy that we did not find any of the maximally loaded cells in post-mortem tissue sections. We typically found only one or a few (<10) microspheres per cell 28 days after transplantation. There are several possible explanations that could account for this observation. For example, the heavily loaded cells could have successively divided and partitioned the microspheres to daughter cells such that cells containing only one or, at most, a few microspheres per cell could be identified after 28 days in vivo. Alternatively, the cells containing large numbers of microspheres may have simply not been able to tolerate the load and may have died off or been removed by macrophages. At the time of assessment (4 weeks post-transplantation), we found little evidence for extracellular microspheres during our histological analysis of the tissue specimens, thereby suggesting that observed MRI signals were most likely from microspheres within cells. At least some of these cells are presumed to be those that were transplanted, although we cannot at present rule out the possibility that they could have been taken up by host cells. Additional study is needed to distinguish between these possibilities.

In Vivo Assessments

A major advantage of the approach used here is that MRI affords the opportunity to simultaneously acquire data about cardiac structure and function, in addition to the cell-tracking capabilities. Thus, we could assess the effectiveness of the stem cell therapy in real time in vivo by measuring various indices of LV remodeling and cardiac performance during the course of the 4-week study. The use of delayed enhancement MRI with Gd-DTPA on day 1 post-MI enabled us to establish that infarct size was similar between treated and untreated groups. In addition to this, Gd-DTPA-enhanced cardiac MRI made it possible to assess the position of the transplanted cells relative to the infarct zone. To the best of our knowledge, this is the first reported use of Gd-DTPA as a contrast-enhancing agent for cardiac MRI evaluation in mice.

In our longitudinal study of cardiac structure and function, mES cells appeared to modify the progression of LV remodeling, with mES cell-treated hearts showing a significant decline in chamber volumes between days 7 and 28 post-MI in contrast to the progressive increases seen in untreated mice (Fig. 5). Despite the resistance displayed by mES cell-treated hearts against the late phase of LV remodeling, LVEF was nearly identical between the two groups by the end of the 28-day study.

The finding that this stem cell transplantation experiment resulted in little change in cardiac performance was not entirely unexpected. This result was consistent with the post-mortem tissue histology showing clear infarct zones where muscle tissue had been replaced by scar tissue. Thus, there appeared to be little, if any, replacement or regeneration of damaged muscle tissue in this experiment. Although it is certainly possible that some limited myocyte regeneration may have occurred, we did not examine this because of the lack of significant improvement in LVEF in response to the stem cell transplantation. Nevertheless, it is important to note that there was no significant deterioration, toxicity, or development of adverse side effects observed in response to transplantation of the microsphere-loaded mES cells in this experiment.

Our primary objective for the present study was to evaluate the feasibility of tracking mES cells in the heart using noninvasive MRI, and this was accomplished. An important feature of this approach is that it is not limited to a particular type of stem cell but rather should be widely applicable to many types of cell transplantation studies. Future studies using different stem cell approaches may show greater promise in terms of cardiac muscle regeneration and improvement of cardiac function. Indeed, promising results in this area have recently been described from a number of studies using preselected subpopulations of either adult or embryonic stem cells [1, [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14]15], and these should be readily amenable for the in vivo cell tracking approach described here.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

In the current study, we demonstrated the practical use of superparamagnetic microspheres as a contrast agent to label-transplanted stem cells in a small rodent model of chronic myocardial infarction. Despite their greater size than most solid contrast media, microspheres could be internalized within cardiac stem cells without apparent adverse effects on their differentiation or mechanical function. Noninvasive serial MRI studies clearly identified microsphere-enriched tissue and showed that microsphere-loaded stem cell locations were largely stable over an extended (4-week) period while simultaneously quantifying major cardiac structural and functional characteristics. Finally, the fluorescent properties of the microsphere coating facilitated identification of their location both prior to transplantation and in post-mortem tissue sections.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

This work was supported by Grant R01-HL078716 from the NIH and a postdoctoral fellowship (to D.G.T.) from the American Heart Association (0625631B).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information
FilenameFormatSizeDescription
Supplemental_Figure_1.tif16104KSupplemental Figure 1
Supplemental_Figure_2A.wmv2060KSupplemental Figure 2A
Supplemental_Figure_2B.avi4739KSupplemental Figure 2B

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.