1. Introduction
Inflammatory activity of the innate immune system following myocardial infarction substantially influences remodeling after myocardial infarction (MI) and the evolution of heart failure [
1]. Therefore, anti-inflammatory therapies have been under intense investigation since the early 1970s. Yet, none of the clinical trials testing pharmacological strategies aiming at unspecific reduction of myocardial inflammation has proven successful [
2]. The main reason for these disappointing findings is the complexity of the well-orchestrated activity of different leukocyte populations. The amplitude and duration of inflammation and the timely and spatial resolution within the heart define the quality of the scar following MI and the amount of tissue loss rather than the mere extent of inflammation [
1]. Hence, the concept of suppressing inflammation has changed to a further elaborated concept of “modulating cellular inflammation”. The balance of pro- and anti-inflammatory immune cells and their recruiting mechanisms are discussed as both therapeutic and prognostic targets [
1]. Further understanding of this intricate cellular immune response to myocardial ischemia in a translational setting requires non-invasive molecular imaging tools. The 18F-FDG PET (Positron emission tomography) using specific protocols for suppressing glucose uptake in cardiomyocytes has recently been introduced to detect cellular inflammation following MI in both patients and C57BL/6 mice [
1,
3,
4,
5]
Protocols to suppress physiological myocardial 18F-FDG uptake require low-carbohydrate diet the day before imaging followed by a 12 h fasting period and intravenous application of heparin before the actual scan [
3]. In mice, myocardial glucose metabolism can be effectively suppressed by replacing the commonly used anaesthetic isofluorane for a ketamine/xylazine-based protocol [
1].
However, to our knowledge, no studies have used these protocols for measuring therapeutic effects of therapies aiming at improving myocardial healing following MI. Therefore, we implanted syngeneic cardiac induced cells committed to the cardiac lineage in order to improve post-MI cardiac function in 129Sv mice as measured my cardiac magnetic resonance (CMR). Cellular inflammation was detected by 18F-FDG PET in vivo and by post-mortem flow cytometry in both infarcted and remote myocardium.
We hypothesize that PET-based imaging and quantification of cellular inflammation can be used as a molecular imaging tool for both quantification and spatial distribution of monocytes and as an early prognostic tool to predict the effect of cardiac stem cell therapies modulating post-MI inflammation.
2. Materials and Methods
2.1. Stem Cell Culture and Cardiovascular Differentiation
W4 murine embryonic stem cells (mESCs), originally isolated from the 129S6 mouse strain [
6], were grown according to standard protocols as described previously [
7]. In brief, cells were cultured in DMEM supplemented with 15% FBS Superior (Biochrom AG, Berlin, Germany), 1% Cell Shield
® (Minerva Biolabs GmbH, Berlin, Germany), 100 µM non-essential amino acids, 1000 U/mL leukemia inhibitory factor (Phoenix Europe GmbH, Mannheim, Germany) and 100 µM β-mercaptoethanol (Sigma-Aldrich GmbH, Steinheim, Germany) at 37 °C, 5% CO
2, and 20% O
2. For cardiovascular differentiation we used cardiogenic differentiation medium, containing IBM (Iscove’s Basal Medium, Biochrom AG) supplemented with 10% FBS Superior, 1% Cell Shield
®, 100 µM non-essential amino acids, 450 µm 1-thioglycerol (Sigma-Aldrich GmbH), and 213 µg/mL ascorbic acid (Sigma-Aldrich GmbH), as described previously [
7,
8]. Cardiovascular differentiation was initiated by hanging-drop culture for two days at 37 °C, 5% CO
2, and 20% O
2. 400 cells per drop were plated on the cover of a square petri dish and grown for 2 days to start formation of embryoid bodies (EB). Afterwards, EB were grown additional four days in suspension culture [
9] and then harvested for transplantation and PCR analysis. These cells will be referred to as cardiac induced cells (CiC). They were cultured till day 30 for beating foci analysis.
2.2. Animal Model
The present study was approved by the federal animal care committee of the Landesamt für Landwirtschaft, Lebensmittelsicherheit und Fischerei Mecklenburg-Vorpommern (LALLF, Germany) (registration no. LALLF M-V/TSD/7221.3-1.1-054/15). The 129S6/SvEvTac were bred in the animal facility of the Rostock University Medical Center and maintained in specified pathogen-free conditions. The mice had access to water and standard laboratory chow ad libitum and received humane care according to the German legislation on protection of animals and the Guide for the Care and Use of Laboratory Animals (NIH publication 86–23, revised 1985), and all efforts were made to minimize suffering. Mice were anaesthetized with pentobarbital (50 mg/kg, intraperitoneal). Following thoracotomy, the left anterior descending coronary artery (LAD) was ligated to induce acute myocardial infarction. The MI group received an intramyocardial injection of 10 µL PBS mixed with 10 µL Growth Factor Reduced Matrigel™ Matrix (Corning, Berlin, Germany). The MI induction plus cell transplantation (MIC) group received a suspension of 1 × 10
6 syngeneic in PBS (10µL) mixed with 10 µL Growth Factor Reduced Matrigel™ Matrix. Injections of 4 × 5 µL were given along the infarct border. The injection site was controlled visually at the time of transplantation [
10].
Healthy mice (n = 33) were divided into 7 groups for 18F-FDG PET imaging 5 days following MI using different protocols for anaesthesia:
- (1)
no intervention, isofluorane (n = 2)
- (2)
no intervention, ketamine/xylazine (n = 2)
- (3)
MI group, isofluorane (n = 4)
- (4)
MI group, ketamine/xylazine (n = 6)
- (5)
MIC group, ketamine/xylazine (n = 7)
Cardiac function was assessed in separate groups three weeks following MI by cardiac MRI:
- (6)
MI only (n = 6)
- (7)
MIC (n = 6)
Flow cytometric analysis was performed in separate groups (MI, n = 5 and MIC, n = 5).
2.3. qRT-PCR
RNA was isolated from the cells using the NucleoSpin® RNA isolation kit (Macherey-Nagel, Dueren, Germany). First strand cDNA was then synthesized using the cDNA synthesis kit (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s instructions. The qPCR reaction was then carried out with the Taqman® Universal PCR Master Mix (Thermo Fisher Scientific) and performed on a StepOnePlus Real-Time PCR system (Applied Biosystems, Foster city, CA, USA). Primers of the following target genes: Pou5f1 (Mm00658129_gH), cTnnt2 (Mm01290256_m1), MesP1 (Mm00801883_m1), and Nkx2.5 (Mm01309813_s1) were purchased from Thermo Fisher Scientific. Gene expression values of the target genes at day 6 were then normalized to the housekeeping gene Hprt (Mm00446968_m1; Thermo Fisher Scientific) and compared relative to the expression values at day 0 using the ∆∆Ct method for relative quantifications.
2.4. Beating Foci Analysis
The number of beating foci per EB was analyzed from day 7 to day 30 of differentiation. The EB were observed under a microscope (Carl Zeiss, Oberkochen, Germany) and the beating foci per each EB were then visually analyzed using the ZEN2011 software (Carl Zeiss).
2.5. Flow Cytometry
Single cell cardiac monocyte suspensions were prepared for flow cytometry, as previously described [
11] Briefly, the remote and infarct tissue of the heart was dissected and enzymatically digested separately in HBSS with Ca
2+ and Mg
2+(450 U/mL collagenase type I, 125 U/mL collagenase type XI, 120 U/mL DNase I, 60 U/mL hyaluronidase, all Sigma-Aldrich) for 30 min at 37 °C. The digested samples were then passed through a 100 µm filter and centrifuged to enrich for mononuclear cells. Red blood cells were then lysed using erythrocytes lysis buffer (eBioscience, San Diego, CA, USA) and the digest was then washed and suspended in MACS
® buffer (PBS, 2 mM EDTA, 0.5% BSA). Samples were then labeled using Zombie Aqua dye (BioLegend, San Diego, CA, USA.), washed, resuspended in MACS buffer containing FCR Block (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany), and stained (see
Table 1 for antibody list). Stained samples were then analyzed on a BD FACS LSR II
® running BD FACS Diva software (version 6.1.2, Franklin Lakes, NJ, USA). The various immune cell populations in the heart tissue were then assessed, as described in
Figure 1.
2.6. PET Imaging
For the PET study, mice of the groups 1,3, 6, and 7 were anaesthetized by inhalation of isoflurane (4% for induction and 1% to 2.5% maintenance during preparation and scanning), whereas mice of the groups 2,4, and 5 were anaesthetized by i.p. injection of ketamine/xylazine (ketamine 84 mg/kg and xylazine 11.2 mg/kg) 20 min before tracer application. All PET/CT scans were performed on a small animal PET/CT scanner (Inveon MM-PET/CT, Siemens Medical Solutions, Knoxville, TN, USA) [
12] according to a standard protocol: Mice were injected intravenously with a dose of approximately 10 MBq 18F-FDG via a custom-made micro catheter placed in a tail vein. After an uptake period of 60 min, mice were imaged in prone position for 20 min. During the imaging session, respiration of the mice was controled and core body temperature was constantly kept at 38 °C via a heating pad. For attenuation correction and anatomical reference, whole body CT scans were acquired. CT images were reconstructed with a Feldkamp algorithm. The PET images were reconstructed with the three-dimensional (3D) iterative ordered subset expectation maximization reconstruction algorithm (3D-OSEM/OP-MAP) with the following parameters: 4 iterations (OSEM), 32 iterations (MAP), 1.7 mm target resolution, and 128 × 128 matrix size. Reconstruction included corrections for random coincidences, dead time, attenuation, scatter, and decay.
2.7. PET Image Analysis
Image analyses were performed using an Inveon Research Workplace (Siemens, Knoxville, TN, USA), as described previously [
13]. PET (Positron emission tomography) and CT (Computerized tomography) images were fused by the use of an automated volumetric fusion algorithm and then verified by an experienced reader for perfect alignment. Consecutively, standardized representative volumes of interest (VOI) were manually placed in the remote area and in the infarcted region as well as the whole heart guided by anatomical landmarks, as described in detail in
Figure 2. Correct VOI positioning was visually verified in axial, coronal, and sagittal projection.
Carimas 2 software (Turku PET Centre, Turku, Finland) was used for generating polar maps of the left ventricle according to the manual provided by the developer. Results are presented using 17-segmental standardized myocardial segmentation.
2.8. Cardiac Magnetic Resonance Imaging
Cardiac magnetic resonance (CMR) measurements were performed on a 7 Tesla small animal MRI system (BioSpec 70/30, maximum gradient strength 440 mT/m, Bruker BioSpin Gmbh, Ettlingen, Germany) equipped with a 1H transmit volume coil (86 mm, volume resonator) and a 2-by-2 receive-only surface coil array (both Bruker BioSpin GmbH). After induction of anaesthesia using 2% to 3.5% isoflurane in oxygen, animals were placed in a supine position on a dedicated mouse bed and surface coil was placed on the chest of the mice. Respiration rate and body temperature were monitored using an MR-compatible small animal monitoring and gating system (Model 1030, SA Instruments, Inc., Stony Brook, NY, USA), and stable body temperature was maintained by a warm water heating. Anaesthesia was maintained during the experiment with isoflurane oxygen (1.5% to 2%) to achieve a respiration rate of about 35 to 55 breaths.
After planning sequences, for the short axes view final images of the left ventricular ejection fraction (LVEF) measurements were acquired using a IntraGate gradient-echo cine sequences (IntraGate Cine-FLASH) in six short-axis planes completely covering the left ventricle. Acquisition parameters included: echo time (TE) 2.38 ms, repetition time (TR) 5.89 ms, flip angle 15°, 14 frames per cardiac cycle, oversampling 140, averages 1, field of view (FOV) 29.4 × 25.2 mm, matrix size 211 × 180, resolution in-plane 0.14 × 0.14 mm, slice thickness 1 mm, and scan time per slice 2 min.
2.9. Cardiac Magnetic Resonance Analysis
LVEF was assessed from the cine sequences using the freely available software Segment v2.0 R5165 (Medviso, Lund, Sweden) (
http://segment.heiberg.se) [
14]. The left ventricular (LV) endocardium in these slices was manually segmented to exclude the papillary muscles. The volumes of these segments were then integrated along the six planes of the LV and LVEF was then calculated from their summated end systolic and end diastolic volumes.
2.10. Statistics
All data are presented as mean values ± standard deviation (SD). Flow cytometric data and qPCR values are presented as mean values ± standard error of mean. Student’s t-test was used for statistical analysis of parametric data and the Mann–Whitney test was used for nonparametric analysis of flow cytometric data. Values of p < 0.05 were considered statistically significant.
4. Discussion
Translational cardiovascular research has evolved at an incredible pace within the last decade and resulted in significantly higher survival rates after acute myocardial infarction. In contrast to highly efficient approaches based on early revascularization and pharmacological prevention of adverse ventricular remodeling, replacement of irreversibly lost cardiomyocytes has not been achieved yet, despite huge efforts in the field of regenerative medicine.
Recently, cellular inflammation following ischemic myocardial injury has been identified as a key player in the process of myocardial healing. Thereby, the local distribution patterns of specific monocyte and macrophage subpopulations have been proposed to determine the quality of myocardial healing [
15,
16].The vast majority of data has been obtained from rodent studies because their hearts can easily be excised for in vitro experiments such as flow cytometric measurements of single cells suspensions. Most researchers focus on the invasion of two distinct macrophage subpopulations, usually referred to as M1 and M2 macrophages. The early M1 macrophage subset is attracted to the site of myocardial injury via CCL2, expressing pro-inflammatory mediators and proteases for degradation of infarcted tissue. Subsequently, M2 macrophages are recruited via CX3CL1 for mediating synthesis of extracellular matrix and angiogenesis [
3]. However, the M1/M2 classification does not adequately explain the complete spectrum of macrophage phenotypes. Recently, macrophages which do not express CCR2 in the neonatal heart have been shown to regenerate the infarcted tissue [
15] while the adult heart involves CCR2
+ monocyte-derived macrophages also taking part in the remodeling process [
17].
Preclinical studies have examined the effect of therapeutic applications, such as stem cell injections, to modify the qualitative and quantitative composition of the post infarct cellular immune response [
18,
19,
20]. This modulation of the innate cellular immune response resulted in improved cardiac pump function, reduction of scar size, and adverse remodeling [
18]. Findings from a meta-analysis by our group have ascribed high therapeutic potential to cardiovascular cell preparations [
21]. Therefore, we sought to transplant ESC-derived cardiac induced cells to improve myocardial healing and also investigate whether they influence the ischemic cellular immune response.
The above-mentioned experiments are based on post-mortem analyses such as flow cytometry and immunohistochemistry of excised organs, and thus have been restricted to preclinical research. However, clinical translation requires methods that allow in-vivo visualization and quantification of inflammatory cells. Lee et al. established 18F-FDG PET for imaging the inflammatory cell activity in the heart, based on suppressing glucose metabolism in myocytes [
1]. Interestingly, anaesthesia with ketamine/xylazine is both sufficient and highly effective in reducing glucose uptake in cardiomyocytes, hence, enabling visualization of inflammatory activity in the heart [
1]. According to Lee et al., 18F-FDG uptake at the site of myocardial inflammation related to the content of local CD11b
+ monocyte/macrophage concentration in C57BL6 mice [
1], at day 5 after MI induction.
On the basis of these findings by other groups, we hypothesized the following:
- (A)
Distribution pattern of CD11b+ myeloid cells at day five after MI induction can be modified by intramyocardial transplantation of CiC;
- (B)
This change can be visualized and quantified by 18F-FDG PET using ketamine/xylazine anaesthesia;
- (C)
The specific 18F-FDG uptake pattern correlates with functional outcome as measured by cardiac magnetic resonance imaging.
We were able to replicate the glucose uptake suppression protocol based on anaesthesia with KX [
5] in 129sv mice and achieved an almost 88% reduction in the glucose uptake in the whole myocardium as compared with isoflurane anaesthesia. We did not find any significant difference in the 18F-FDG uptake levels between the remote and infarcted myocardium using KX anaesthesia, which is in line with findings from Thackeray et al. [
4,
5]. Furthermore, intense focal tracer uptake was localized in the border zone. Similarly, Lee et al. also reported accentuated FDG uptake in the border zone corresponding to high numbers of infiltrating CD11b
+ cells [
1]. Despite visual assessment of this high focal tracer accumulation, quantitative analysis in this region was only performed by Thackeray et al. However, their VOI positioning strategy remains unclear from the manuscript. This might be due to a certain overlap of the border zone to the adjacent regions which makes demarcation of the border zone difficult. In order to produce comparable data in this aspect, VOI positioning strategies should be provided in future studies.
Interestingly, transplantation of 10
6 cardiac induced cells after MI significantly increased FDG uptake in both the whole heart and the remote myocardium, but not in the infarct region. Moreover, CiC transplantation led to a 7% improvement in LVEF three weeks after MI. The magnitude of LVEF improvement is well in line with previous reports [
21]. Therefore, we conclude that the change in the myocardial 18F-FDG uptake pattern represents a valid tool for early PET-based in vivo prediction of myocardial healing post MI.
A deeper investigation of the immune response in the heart five days after MI using flow cytometry revealed an increase in the percentage of CD11b
+ cells and a shift towards increased monocyte-derived macrophages in the remote area of cell-treated mice. We also observed a reduction in the yolk sac-derived resident macrophages in the infarct area of cell-treated mice. We did not however observe a shift in the M1/M2 polarization phenotype, as observed in previous studies involving MSCs [
18]. It is interesting to note that increased monocyte-derived macrophages and reduced yolk sac-derived macrophages in the heart have been previously attributed to adverse cardiac remodeling and scar formation [
17]. However, these studies did not involve the transplantation of cells into the infarcted heart and it is well known that the gene expression profile and the characteristics of the immune cells are affected based on their location in the myocardium and the cells they interact with [
22].
However, unbiased expression profiling of these cells over various time points and under the influence of CiC has yet to be carried out. The influence of CiC transplantation on the immune response in the heart has not been studied until now and our work suggests possible new mechanisms and targets for improving the efficiency of CiC.
The 18F-FDG-based imaging strategies for myocardial inflammation are highly attractive and have been applied in a clinical setting despite some limitations, which have yet to be overcome [
3]. Both cardiomyocytes [
4] and infiltrating inflammatory cells [
23] in the setting of acute myocardial injury possess high levels of glucose metabolism. Metabolism of healthy cardiomyocytes is mainly based on fatty acid oxidation, whereas ischemia triggers increased anaerobic glycolysis, which requires much higher rates of glucose [
24]. This might hamper direct correlation of focal 18F-FDG uptake and high concentration of CD11b
+ cells in the infarcted region. The use of KX for anaesthesia has been shown to reduce serum insulin levels in rodents, thus, preventing translocation of GLUT4 to membranes of cardiomyocytes and reducing 18F-FDG uptake [
25]. In contrast, leucocyte glucose influx in a setting of acute inflammation depends more on GLUT1 and GLUT3, which are expressed and translocated independently of insulin [
26]. Whereas, this KX protocol is well suited for preclinical research, different strategies are used for suppression of myocardial glucose uptake in patients including prolonged fasting, dietary modifications, and heparin loading before imaging [
27]. This might hamper straightforward translation of imaging protocols established in rodents to clinical application.
Furthermore, the transplantation of CiC adds more complexity to its use in imaging the myocardium. It becomes difficult to attribute the observed 18F-FDG uptake pattern to the inflammatory cells alone since the ability of CiC to alter the cardiac metabolism is poorly understood. It is, therefore, difficult to delineate the relative contribution of 18F-FDG uptake between the host cardiomyocytes, immune cells, and the CiC. The heterogeneity of the inflammatory response and the relative 18F-FDG uptake between the different immune cell populations is also clearly not understood. Further research is necessary to understand metabolic changes in the different cells involved in healing myocardium.
Nevertheless, this is the first study to investigate 18F-FDG-based PET imaging of inflammation as an early indicator for assessing long term therapeutic efficiency in a rodent model of acute myocardial infarction. Furthermore, the current work illustrates the benefit of CiC transplantation to improved cardiac function after MI possibly through its beneficial modulation of the innate inflammatory response. Using such non-invasive techniques in the field of translational research will foster a better understanding of the inflammatory response and how its modulation could lead to altered cardiac function. It is also a valuable tool for monitoring various immune modulation and cell therapies and broadens the horizon for understanding the mechanisms of these therapeutic strategies.