Human Fetal Cardiac Mesenchymal Stromal Cells Differentiate In Vivo into Endothelial Cells and Contribute to Vasculogenesis in Immunocompetent Mice

Abstract

Mesenchymal stromal cells (MSCs) have shown great potential as a treatment for systemic inflammatory diseases, but their local regenerative properties are highly tissue- and site specific. Previous studies have demonstrated that adult human MSCs respond to inflammatory cytokines through the release of paracrine factors that stimulate angiogenesis, but they do not themselves differentiate into vascular structures in vivo. In this study, we used human fetal cardiac MSCs (hfcMSCs) harvested during the first trimester of heart development and injected them into the subcutaneous tissue of normal immunocompetent mice treated with short-term costimulation blockade for tolerance induction. When hfcMSCs were transplanted subcutaneously together with Matrigel matrix, they contributed to vasculogenesis through differentiation into endothelial cells and generation of the basal membrane protein Laminin α4. These characteristics of hfcMSCs are similar to the mesodermal progenitors giving rise to the developing heart and they may be useful for treatment of ischemic injuries.

Introduction

Mesenchymal stromal cells (MSCs) have received vast attention during the last decade due to their regenerative and immunomodulatory properties [1,2], mainly mediated through paracrine factors [3,4]. Clinical studies have demonstrated their efficacy in the treatment of systemic inflammatory diseases, such as graft versus host disease [5] and severe acute respiratory distress syndrome [6], where bone marrow-derived MSCs (BM-MSCs) were administered intravenously. Due to the immunomodulatory, as well as antifibrotic and pro-angiogenic properties of these cells, they would theoretically constitute an ideal treatment option whenever there is a condition of tissue damage [7]. This includes wound healing, limb ischemia, and myocardial infarctions, where a combination of restorative inflammatory response and angiogenesis/vasculogenesis is a crucial step in the healing process.

In contrast to the positive effects of human MSCs on systemic inflammatory diseases, local delivery has thus far been conducted with limited success [8]. This can potentially be ascribed to limited engraftment due to loss of anchorage [9], immunologically mediated rejection [10], and also limited capacity of the implanted MSCs to respond to signals in the environment and differentiate into vascular structures.

MSC engraftment may be augmented by limiting the immunologically mediated rejection [11]. We and others have demonstrated long-term acceptance of allo- and xenogeneic transplants through a short-term costimulation blockade treatment [12 –17]. Costimulation blockade administered transiently, with the aim to induce anergy or tolerance to the transplanted cells, is therefore well suited for the study of the in vivo differentiation capacity of implanted MSCs.

An additional approach for increasing MSC retention after implantation is to reduce cell death due to loss of anchorage (anoikis) [9]. Matrigel, which is an extracellular matrix (ECM) product produced by Engelbreth-Holm-Swarm mouse sarcoma cells, resembles the complex ECM composition [18 –21]. It has been extensively used as both cell culture substrata and to improve cell engraftment after injection [22]. Since Matrigel besides matrix proteins also contains growth factors [18], three-dimensional (3D) cultures with Matrigel and the MSCs to be used were produced to study the in vitro effects of Matrigel before implantation.

The source of MSCs is also of importance. Previous studies have demonstrated that BM-MSCs have the ability to differentiate into endothelial cells in in vitro cultures [23,24] and that implantation of these predifferentiated endothelial cells participated in the neovascularization of severe combined immunodeficiency (SCID) mice [24]. Nevertheless, to our knowledge differentiation of BM-MSCs into endothelial cells in vivo has not been shown. In our study we focus on human fetal cardiac MSCs (hfcMSCs), which based on our previous findings contain subpopulations of endothelial progenitors and can differentiate in vitro into endothelial cells [25]. Their capacity to respond to inflammatory stimuli after implantation and contribute to angiogenesis/vasculogenesis has not been investigated before.

Our study is the first to demonstrate that specific human MSCs have the capacity to engraft after implantation into normal immunocompetent mice and in response to inflammatory stimuli differentiate into vascular structures containing endothelial cells in combination with appropriate basement membrane proteins. This opens up for new regenerative strategies where MSCs besides their immunomodulatory capacities can be used for improving vascularization after tissue injury.

Materials and Methods

Derivation and expansion of hfcMSCs

Fetal hearts (gestational weeks 4.5–10.5) were obtained from legal terminations of pregnancy after the donor's informed consent and ethics approval from the Regional Ethics Board in Stockholm, ethics approval number 2015/1369-31/2. hfcMSCs were isolated as described previously [25] and expanded in Dulbecco's modified Eagle's medium/nutrient mixture F-12 (DMEM/F12; Gibco, Thermo Fisher Scientific, Gothenburg, Sweden) supplemented with 2% fetal bovine serum (PAA Laboratories, Thermo Fisher Scientific, Gothenburg, Sweden), B27 (Gibco, Thermo Fisher Scientific), MycoZap (Lonza, Stockholm, Sweden), epidermal growth factor, EGF (R&D Systems, Minneapolis, MN), 10 ng/mL and recombinant human Wnt3a 100 ng/mL (R&D Systems). The cells used in our study were harvested at passages 6–10.

MSC phenotyping and differentiation

In this study we used the same batches of hfcMSCs that were previously thoroughly characterized [25] and that fulfill the International Society for Cellular Therapy (ISCT) guidelines for MSCs [26]. In brief, the hfcMSCs were determined to express CD73, CD90, CD105, and HLA-1, while negative for HLA-DR, the hematopoietic lineage markers CD14, CD11, CD80, CD83, CD34, CD45, and the endothelial marker CD31 using the Human MSC Phenotyping Kit (Miltenyi Biotec, Lund, Sweden), where the labeled cells were examined with FACSCanto I (BD Biosciences, San José, CA).

Osteogenesis was induced by adding 50 mg/mL ascorbate and 10 mM β-glycerophosphate. Adipogenesis was induced by adding 1 mM dexamethasone, 0.5 mM 3-isobutyl-1-methylxanthine, 10 mg/mL insulin, and 1 mM rosiglitazone (Cayman Chemicals, Ann Arbor, MI) and withdrawing 3-isobutyl-1-methylxanthine and dexamethasone after 2 days. Chondrogenesis was induced with pellet culture in Chondrocyte Differentiation Medium (Lonza) supplemented with 10 ng/mL TGF-b3 for 21 days. When clonally expanded, these cells were able to form colonies stained by crystal violet.

Ratio of CD31⁺ cells among the initial hfcMSCs

To assess the ratio of CD31⁺ cells present in the hfcMSC cultures, 10,000 cells/cm² were seeded in a 48-well plate (n = 5) and given the same culture medium as described above. After 4 days, the confluent cultures were fixed with 4% formaldehyde in phosphate buffered saline (PBS) and subsequently stained with anti-CD31 antibodies (clone JC70A; Sigma-Aldrich, Stockholm, Sweden).

3D Matrigel cultures

One hundred thousand cells were resuspended in 50 μL of a 1:1 mixture of Matrigel (Corning, Thermo Fisher Scientific, Gothenburg, Sweden) and DMEM/F12, which was allowed to polymerize as hanging drops. The spherical 3D cultures were given the same basal medium as during two-dimensional (2D) cell culture, without the addition of Wnt3a and EGF. The growth factors were removed with the attempt to mimic the situation in vivo, where the presence and concentration of these growth factors are difficult to predict.

Animal experiments

All procedures were approved by the Linköping Ethics Committee on Animal Experiments, Sweden (Ethical approval No. S22-15). Female Naval Medical Research Institute (NMRI) mice were purchased from Scanbur, Karlslunde, Denmark.

The in vivo experiments are outlined in Table 1. All surgical interventions were performed under isoflurane anesthesia. A total of 500,000 cells resuspended in 30 μL of Matrigel or DMEM/F12 were injected subcutaneously in the upper neck. The subcutaneous graft containing hfcMSCs delivered in either Matrigel or DMEM/F12 was excised and embedded in Optimal Cutting Temperature compound, OCT (Histolab, Gothenburg, Sweden), and subsequently snap frozen in ethanol on dry ice.

Table 1.

Outline of In Vivo Experiments

Time (weeks)	Matrigel	DMEM	Blockade	Control	Graft	n
4	x		x		8	8
4	x			x	0	6
8	x		x		2	6
8	x			x	0	6
12	x		x		0	6
4		x	x		2	3

Overview of the in vivo experimental setup and results. Time from cell transplantation until the experiment was finished is given in weeks. The delivery vehicle (Matrigel, DMEM) and whether costimulation blockade (Blockade) or isotype control antibodies (Control) were administered are indicated by a cross. In the right column the total number of animals per group (n) and number of animals with engrafting human cells (grafts) are presented.

Blockade, T cell costimulation blockade; Control, isotype control antibodies; DMEM, Dulbecco's modified Eagle's medium; n, number of animals.

On the day of surgery (d0) and on d2, d4, and d6, all NMRI mice were given either a combination of 0.5 mg cytotoxic T lymphocyte-associated 4 Ig (CTLA4Ig), 0.5 mg anti-CD40 ligand (anti-CD40L), 0.2 mg anti-Leukocyte function-associated antigen-1 (anti-LFA-1), or isotype control antibodies, consisting of 0.5 mg Human Fc-IgG1, 0.5 mg Hamster IgG, and 0.2 mg Rat IgG2a as an intraperitoneal injection. All antibodies were purchased from Bio X Cell, West Lebanon, NH.

Immunohistochemistry

Cryo-sections (5 μm) were stained with antibodies against FoxP3 (clone FJK-16S; eBiosciences, AH Diagnostics, Stockholm, Sweden), CD4 (clone YTS 191.1; Dako, Agilent, Santa Clara, CA), CD8 (clone KT15; Dako, Agilent), CD31 (polyclonal rabbit; Abcam, Cambridge, United Kingdom; clone JC70A; Sigma-Aldrich; clone WM59 and clone MEC13.3; BioLegend, San Diego, CA), α smooth muscle actin (αSMA; polyclonal rabbit, Abcam and clone 1A4, Sigma-Aldrich), Troponin T (clone 1C11; Abcam), and Laminin α4 chain (polyclonal rabbit and clone CL3183; Atlas Antibodies, Stockholm, Sweden).

The sections were fixed in formaldehyde 4% in PBS except for clone WM59 and MEC13.3 for CD31, where ice-cold acetone was utilized. The primary antibody was added in PBS with 5% serum (blocking buffer) and incubated at room temperature overnight. When staining for Laminin alpha 4 chain and FoxP3 heat induced antigen retrieval was performed before addition of antibody, by boiling the slides in citric buffer. Depending on the origin of the primary antibody, secondary antibodies produced in goat, rabbit, or donkey were chosen (A11008, A21121, A11081, A11059, A21208; Thermo Fisher Scientific, Gothenburg, Sweden) and added to the blocking buffer. After incubation for 1–2 h the slides were washed and mounted in VECTASHIELD containing 4′,6-diamidino-2-phenylindole.

Immunohistochemistry stainings in cell culture dishes, including 3D Matrigel cultures, were done for CD31, αSMA, and Troponin T. Antigen retrieval was not necessary. Fixation and addition of primary and secondary antibodies were performed as described for tissue sections.

Fluorescence in situ hybridization

Cryosections (5 μm) were fixed in 4% formaldehyde in PBS (Histolab), boiled in citric buffer (Invitrogen, Thermo Fisher Scientific, Gothenburg, Sweden), and subsequently digested with Pepsin (Sigma-Aldrich) 100 μg/mL in 0.01 M HCl at 37°C. Hybridization probe (07J04-005; Abbot, Stockholm, Sweden) in Vysis LSI/WCP Hybridization buffer (06J67-011; Abbot) was added, and slides were incubated overnight at 37°C. The slides were washed in 1× SSC (Gibco, Thermo Fisher Scientific) in dH₂O with 0.3% IGEPAL (Sigma-Aldrich) at 72°C and mounted in VECTASHIELD containing DAPI (Vector Laboratories, Burlingame, CA).

TaqMan polymerase chain reaction

RNA was extracted from cells using the PicoPure RNA Isolation Kit (Life Technologies, Thermo Fisher Scientific, Gothenburg, Sweden), and DNase (Qiagen, Stockholm, Sweden) was used to digest contaminating genomic DNA. Complementary DNA (cDNA) was synthesized using the High Capacity cDNA Reverse Transcription Kit (Life Technologies, Thermo Fisher Scientific) according to the manufacturer's instructions.

Gene expression assays (Life Technologies, Thermo Fisher Scientific) for human Troponin T (TNNT2), CD31 also known as Platelet Endothelial Cell Adhesion Molecule (PECAM1), α-smooth muscle actin (ACTA2), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; Life Technologies, Thermo Fisher Scientific) were used in polymerase chain reaction, which were performed in triplicates on an ABI Prism 7000 Sequence Detection System (Applied Biosystems, Thermo Fisher Scientific, Gothenburg. Sweden). The comparative ΔCt method was applied to calculate the ratio between the gene of interest and endogenous control, GAPDH.

Statistics

All statistical calculations were performed using GraphPad Prism (San Diego, CA). For identification of differences we used the Mann–Whitney U-test. P < 0.05 was considered to indicate statistical significance. Gene expression was given as the percentage of GAPDH ± standard deviation (SD). Engraftment was given as the mean number of cells per tissue section ± standard error of the mean. Ratio of CD31⁺ cells to total number of hfcMSCs in culture and in in vivo sections is presented as mean ± SD.

Results

hfcMSCs express high levels of αSMA at the messenger RNA and protein level

We have previously demonstrated that hfcMSCs fulfill the requirements of being MSCs, including expression of mesenchymal surface markers, in addition to their clonogenic expansion potential and multilineage differentiation capacity [25]. In the present study hfcMSCs in 2D cultures were found to have an eightfold increase in expression of αSMA at the messenger RNA (mRNA) level (18.47% ± 15.06% of GAPDH), compared to Human Carotid Artery Smooth Muscle Cells (2.30% ± 0.57% of GAPDH, P < 0.001) (Fig. 1A). The mRNA expression of CD31 was limited (0.04% ± 0.03% of GAPDH), with a 500-fold decrease compared to Human Aortic Endothelial Cells (22.65% ± 7.77% of GAPDH, P < 0.0001) (Fig. 1B).

FIG. 1.

hfcMSCs express αSMA and form vessel-like structures in 3D cultures. RT-PCR was used to quantify mRNA from hfcMSCs in 2D cultures. αSMA (A) and CD31 (B) expression levels were assessed and compared to those of HCtASMC and HAECs, respectively. **P < 0.001, ***P < 0.0001. Immunohistochemical stainings for αSMA (C) and CD31 (D), both AF-488 (green), performed in the cell culture dish corroborated the RT-PCR results, identifying single CD31⁺ cells as indicated by the arrow in (D) while αSMA was ubiquitously expressed. hfcMSCs in Matrigel 3D cultures were stained for αSMA (AF-488) (E) and tube-like structures were visualized by phase-contrast microscopy of live cells in the culture dish (F). Scale bars = 100 μm (C, E), 60 μm (F), and 50 μm (D). 2D, two dimensional; 3D, three dimensional; αSMA, α smooth muscle actin; HAECs, human aortic endothelial cells; HCtASMC, human carotid artery smooth muscle cell; hfcMSCs, human fetal cardiac mesenchymal stromal cells; mRNA, messenger RNA; RT-PCR, reverse transcription polymerase chain reaction.

These findings were confirmed by the immunohistochemical stainings, where the αSMA expression was ubiquitous (Fig. 1C) and 0.008% ± 0.008% CD31⁺ cells were detected (Fig. 1D and Supplementary Table S1). The mRNA levels of Troponin T (TNNT2) were barely detectable, and no antibody staining was visible (data not shown).

To simulate an in vivo environment, hfcMSCs were allowed to form 3D structures by mixing with Matrigel. hfcMSCs cultured in 3D demonstrated a similar αSMA staining intensity as in 2D cultures (Fig. 1E), with the difference that they now formed tube-like structures (Fig. 1F), with few CD31⁺ cells. This indicates that the Matrigel itself does not stimulate endothelial differentiation but support cell survival and adherence with tube formation.

hfcMSCs delivered in Matrigel survive up to 8 weeks in the subcutaneous tissue of mice treated with triple costimulation blockade

To assess the survival of hfcMSCs in a stringent model of T cell-mediated rejection, we utilized a subcutaneous in vivo model. Based on the results from the previous in vitro experiments, we chose Matrigel as the delivery matrix. NMRI mice were used, which received either costimulation blockade (CTLA4Ig, anti-CD40L and anti-LFA-1) or isotype control antibodies for 1 week after implantation and were euthanized at 4, 8, or 12 weeks after the treatment. Mice subcutaneously injected with hfcMSCs in culture medium alone (DMEM/F12) and treated with costimulation blockade were used as negative controls and sacrificed at 4 weeks postinjection.

No surviving human cells could be detected after 4 weeks when NMRI mice were given isotype control antibodies (Supplementary Fig. S1). In contrast, when costimulation blockade was administered, eight out of eight mice with hfcMSCs delivered in Matrigel and two out of three mice with hfcMSCs delivered in DMEM/F12 displayed engrafting human cells at 4 weeks (Fig. 2A, D, E, H). The use of Matrigel as a delivery matrix significantly increased the number of engrafting human cells (188.3 ± 31.9 cells per tissue section), compared to DMEM/F12 (20 ± 5.6 cells per tissue section), P < 0.0001 (Fig. 2N). Neither group showed any sign of lymphocytic infiltration that could indicate immunologically mediated rejection (Fig. 2I, L).

FIG. 2.

hfcMSCs delivered in Matrigel survive up to 8 weeks in the subcutis of immunocompetent mice treated with costimulation blockade. (A–L) Show histological sections from NMRI mice treated with costimulation blockade and harvested at different time points. H&E staining was used to identify grafts and/or inflammatory infiltrates representing immunological rejection (A–D). FISH (E–H) was applied to consecutive sections to analyze the presence of human cells, where SpectrumRed-positive cells contain hDNA. Immunohistochemical staining with anti-CD4 antibodies was performed on consecutive sections to assess lymphocytic infiltration, AF-488 (green), (I–L). Scale bar = 100 μm. The percentage of NMRI mice with surviving hfcMSCs are shown over time (M), Matrigel was used as delivery matrix. The dot plot (N) depicts the increased engraftment of hfcMSCs delivered in Matrigel, compared to DMEM/F12, in mice treated with costimulation blockade. ***P < 0.0001. DMEM/F12, Dulbecco's modified Eagle's medium/nutrient mixture F-12; FISH, fluorescence in situ hybridization; H&E, hematoxylin and eosin; hDNA, human DNA; NMRI, Naval Medical Research Institute.

After 8 weeks hfcMSCs delivered in Matrigel were detected in two out of six animals (Fig. 2B, F); however, at this time point the injected cells were surrounded by a vigorous lymphocytic infiltrate (Fig. 2J), which was also apparent at the injection site in the four mice with no retaining cells. The immunological response was dominated by CD4⁺ cells, in combination with a limited number of CD8⁺ and FoxP3⁺ cells (regulatory T cells). After 12 weeks, no human cells could be detected (Fig. 2C, G). Instead there was a substantial immune reaction at the injection site, similar to what was seen in mice analyzed after 8 weeks (Fig. 2K).

hfcMSCs express vascular markers in vivo

hfcMSCs delivered subcutaneously in Matrigel demonstrated vessel-like structures at 4 weeks, as well as 8 weeks, after implantation. To find out whether this indicated de novo vessel formation, antibody stainings were performed against the endothelial marker CD31, vascular smooth muscle marker αSMA, and Laminin α4 chain, which is an extracellular basal membrane component commonly expressed in the vessel wall [27,28]. All three markers were present in the graft; however, αSMA was not as abundant as expected from the in vitro stainings. Several CD31⁺ cells were observed in all analysed tissue sections and in the same region, which together with α4-containing laminins formed vessel-like structures.

To investigate whether CD31, αSMA, and Laminin α4 chain were derived from the implanted hfcMSCs, or from the resident mouse cells, tissue sections subsequent to the fluorescence in situ hybridization (FISH) staining (identifies human cells) were used for these immunohistochemical analyses. Several human cells coexpressed CD31 in combination with Laminin α4, forming vessel-like structures. Furthermore, CD31⁺ cells and Laminin α4 expression were mainly observed in direct proximity to the human cells (Fig. 3A–D), whereas αSMA seemed to be expressed by both human and mouse cells (Supplementary Fig. S2).

FIG. 3.

hfcMSCs express the vascular markers CD31 and Laminin α4 in vivo. NMRI mice implanted subcutaneously with hfcMSCs in Matrigel and treated with costimulation blockade were found to express CD31 and Laminin α4 in the area of the graft at 4 weeks after injection. To analyze whether CD31 and Laminin α4 originated from the hfcMSCs, double stainings were performed using FISH (SpectrumRed) and antibodies against CD31 (A, C) (AF-488, green) or Laminin α4 (B, D) (AF-488, green). (A–D) Tissue sections of near proximity. In addition, antibodies specific for either mouse or human CD31 were used for double stainings (E), where mouse CD31 stains red (AF-546) and human CD31 stains green (AF-488). Scale bar = 50 μm. hCD31, human CD31; LAMA4, Laminin α4; mCD31, mouse CD31.

To corroborate the notion that the CD31⁺ cells, largely forming the vessel-like structures, were derived from hfcMSCs immune stainings with antibodies specific for mouse and human CD31, respectively, were performed as double stainings. The area containing engrafted human cells exclusively stained positive for human CD31, whereas mouse vessels located in other areas of the subcutaneous tissue stained positive with the mouse specific antibody (Fig. 3E). hfcMSCs delivered in DMEM/F12 did not form any vessel-like structures and did not stain positive for CD31, Laminin α4, or αSMA.

To exclude the possibility that CD31⁺ cells identified in the tubular structures were generated from the CD31⁺ cells present among the implanted hfcMSCs, we calculated both the total number of human cells per visual field (20× magnification) and the ratio of CD31⁺ human cells to the total number of human cells (Supplementary Table S2). On average there were 68.5 ± 44.7 human cells per visual field, where 43% ± 12% were CD31⁺ cells. This substantial increase in ratio indicates that hfcMSCs have the ability to differentiate into CD31⁺ cells with endothelial characteristics, after implantation.

Laminin α4 originates from hfcMSCs

The absence of Laminin α4 staining after implantation of cells in DMEM/F12 alone raised the question whether Laminin α4 originates from the Matrigel matrix or from the hfcMSCs. Matrigel is known to contain Laminin 111 [29], but whether it contains other isoforms as well has not been thoroughly studied. Mice injected subcutaneously with hfcMSCs in Matrigel and treated with isotype control antibodies showed evidence of remaining Matrigel matrix after 4 weeks, identified through hematoxylin and eosin (H&E)-stained sections (Fig. 4A). However, all mice treated with isotype control antibodies rejected hfcMSCs, and the remaining Matrigel was inhabited by mouse cells (FISH-negative) (Fig. 4B).

FIG. 4.

Laminin α4 is absent in the remaining Matrigel devoid of retaining hfcMSCs. The images show sections of subcutaneous tissue from representative mice that received hfcMSCs subcutaneously in Matrigel. (A–D) Are from a mouse treated with isotype control antibodies, and (E–H) are from a mouse given costimulation blockade. (A, E) Show H&E from the area of injection with remnants of Matrigel matrix. FISH was performed on sections of close proximity, (B, F) to identify human cells (SpectrumRed). An additional section was then used for immunohistochemistry to identify Laminin α4 expression (AF-488, green) (C, G). (D, H) Images of the tissue sections shown as (C, G) but applying the red filter to identify autofluorescence. Scale bar = 50 μm in (A, B, E, F) and 100 μm in (C, D, G, H).

Immunohistochemical stainings revealed that inside the remaining Matrigel matrix there was a number of mouse cells expressing αSMA, whereas Laminin α4 (Fig. 4C, D) and CD31 were absent. In contrast, in mice treated with costimulation blockade, surviving hfcMSCs (Fig. 4E, F), Laminin α4 (Fig. 4G, H), and CD31 were present in the area of the FISH-positive human cells. This strongly indicates that Laminin α4 is produced by the injected hfcMSCs and not derived from either Matrigel or the mouse cells.

Discussion

Human MSCs are an attractive clinical cell source mainly due to their immunomodulatory effect on systemic inflammatory diseases [9,10]. In contrast, the inborn regenerative potential of MSCs seems to be tissue dependent [30]. This might be explained by the “pericyte concept,” where tissue specific MSCs are suggested to be derived through activation of pericytes surrounding the micro- and macrovessels [31]. According to this concept, MSCs act as sentinels that react to tissue damage and respond through the secretion of bioactive trophic factors that inhibit scar formation and apoptosis and induce angiogenesis [31 –33]. Through the alignment of pericytes along the exterior of vessels, they form a tripartite arrangement with both the endothelial cells of the vessels and the tissue-specific stem cells. This might explain the tissue- and site specificity of MSCs generated from activated pericytes [31]. Accordingly, activated MSCs control the regenerative response through release of paracrine factors, but they do not seem to differentiate themselves into endothelial cells and contribute to vasculogenesis in adult tissues. This is further supported by previous studies, where adult MSCs rather stimulate angiogenesis through the release of exosomes that transfer pro-angiogenic RNA to endothelial cells [34].

Vasculogenesis is the embryonic process that entails direct formation of vessels by in situ differentiation of angioblastic precursor cells [35,36]. This is a property that has not been described in adult MSCs. We have used MSCs isolated and expanded from first trimester fetal hearts, which is the first functioning organ to develop during human embryogenesis [37]. These cells express the early cardiac progenitor markers Isl1, Nkx2.5, PDGFα, KDR, and SSEA1 [25], but their in vivo regenerative capacity has not previously been explored.

Within 4 weeks after injection into the subcutaneous tissue of immunocompetent mice, the hfcMSCs differentiated into endothelial cells and deposited α4-containing laminins, which are important basement membrane proteins of the vessel walls [28]. The human origin of the CD31⁺ cells found within the graft was demonstrated using FISH and immunohistochemistry for CD31, which was subsequently corroborated with human specific CD31 antibodies. The large increase in the ratio of human CD31⁺ cells observed 4 weeks after transplantation, compared to the initially delivered hfcMSCs, clearly speaks in favor of in vivo differentiation of hfcMSCs into CD31⁺ cells, rather than a mere contribution of the initial fraction of CD31⁺ cells among the hfcMSCs.

The mechanism by which hfcMSCs contribute to formation of vascular structures cannot be regarded as an angiogenic process, where vessel formation is initiated by sprouting from the already existing microcirculation [36]. Instead, the new vessel formation represents vasculogenesis [38], where both endothelial cells and matrix proteins are produced from the implanted hfcMSCs. At the same time the expression of αSMA is reduced as a sign of differentiation of activated MSCs. However, although ECM molecules typically found in vessel wall basement membranes were detected by immunohistochemistry, we cannot claim with certainty that the CD31⁺ structures formed by hfcMSCs represent functional vessels. To assess vessel functionality, in situ perfusion experiments would be required, but were not performed in the present study.

An interesting finding is that hfcMSCs only contribute to vasculogenesis when the cells are co-implanted with Matrigel matrix. In contrast, when hfcMSCs were injected with medium alone, the cells survived for at least 4 weeks, but we could not detect any differentiation into endothelial cells. Probably Matrigel protects the hfcMSCs from anoikis, which is supported by our in vitro cultures, where the hfcMSCs adhere to the Matrigel matrix and form tube-like structures. The Matrigel itself did not stimulate endothelial differentiation in the in vitro tests.

To avoid using mice with several inborn defects, like Nod SCID Gamma [39], we chose normal immunocompetent mice. To prevent immune rejection of the xenogeneic hfcMSCs, we used a combined blockade of the surface molecules CD28, CD40L, and LFA-1. These molecules are important receptors involved in the costimulation of T cells [40,41]. If these receptors are blocked, the naive T cell that encounters a specific antigen will not receive proper costimulation and may develop into a regulatory T cell (T_reg), which induces tolerance to its antigens through direct interaction with other immune cells and through the release of cytokines [42,43].

Short-term blockade of CD28 and CD40 signals through CTLA4Ig and anti-CD40L antibodies induces long-term acceptance of allogeneic grafts in mice [16]. With additional blockade of LFA-1 through anti-LFA-1 antibodies, xenografts have been accepted long term in the brain, muscle, heart, and kidney of mice [12 –15,17].

In the present study, at 4 weeks all mice treated with costimulation blockade displayed surviving cells, with no evident immune rejection. However, long-term acceptance was not achieved and at 8 weeks immune rejection was apparent, with all human cells lost at 12 weeks. These results differ from abovementioned findings, where triple costimulation blockade provided long-term survival of xenografts. However, liver, neurological tissue, and heart differ from the skin and subcutaneous tissue, where the latter provide a barrier between the body and the outer world.

Barrier organs, such as skin and lungs, are generally considered to mount a more aggressive immune response. In this model, the immune response to the transplant was initially suppressed. Since the costimulation blockade is only given transiently, anti-LFA-1, CTLA4Ig, and anti-CD40L will be lost from the circulation after a few weeks. All NMRI mice in the control group that received isotype control antibodies rejected human cells within 4 weeks. This strongly indicates that hfcMSCs are fully immunogenic, which corroborates previous studies using other MSCs [10,11].

In conclusion, our findings are interesting from a developmental perspective, where we demonstrate that first trimester human MSCs behave differently in response to local regenerative signals than adult sources, which not by themselves differentiate into endothelial cells and contribute to vasculogenesis in vivo. Furthermore, the stem cell like characteristics of the fetal MSCs corroborate with properties of mesodermal progenitors giving rise to pericytes during heart development [44 –47], and cells with these properties might be important for treatment of a variety of ischemic injuries where vasculogenesis and angiogenesis are crucial.

Footnotes

Acknowledgment

This study was supported by grants from the Swedish Research Council 2013-03590.

Author Disclosure Statement

K.-H.G., S.R., and M.C. are cofounders of the company IsletOne AB. All other authors declare no conflicts of interest.

Supplementary Material

Supplementary Figure S1

Supplementary Figure S2

Supplementary Table S1

Supplementary Table S2

References

Pittenger

, Mackay

, Beck

, Jaiswal

, Douglas

, Mosca

, Moorman

, Simonetti

, Craig

and Marshak

. (1999). Multilineage potential of adult human mesenchymal stem cells. Science, 284:143–147.

Bernardo

and Fibbe

. (2013). Mesenchymal stromal cells: sensors and switchers of inflammation. Cell Stem Cell, 13:392–402.

Phinney

and Pittenger

. (2017). Concise review: MSC-derived exosomes for cell-free therapy. Stem Cells, 35:851–858.

Samper

, Diez-Juan

, Montero

and Sepúlveda

. (2013). Cardiac cell therapy: boosting mesenchymal stem cells effects. Stem Cell Rev, 9:266–280.

Le Blanc

, Frassoni

, Ball

, Locatelli

, Roelofs

, Lewis

, Lanino

, Sundberg

, Bernardo

, et al. (2008). Mesenchymal stem cells for treatment of steroid-resistant, severe, acute graft-versus-host disease: a phase II study. Lancet, 371:1579–1586.

Simonson

, Mougiakakos

, Heldring

, Bassi

, Johansson

, Dalén

, Jitschin

, Rodin

, Corbascio

, et al. (2015). In vivo effects of mesenchymal stromal cells in two patients with severe acute respiratory distress syndrome. Stem Cells Transl Med, 4:1199–1213.

Le Blanc

. (2006). Mesenchymal stromal cells: tissue repair and immune modulation. Cytotherapy, 8:559–561.

Kean

, Lin

, Caplan

and Dennis

. (2013). MSCs: delivery routes and engraftment, cell-targeting strategies, and immune modulation. Stem Cells Int, 2013:732742.

Grossmann

. (2002). Molecular mechanisms of “detachment-induced apoptosis—Anoikis”. Apoptosis, 7:247–260.

10.

Nauta

, Westerhuis

, Kruisselbrink

, Lurvink

, Willemze

and Fibbe

. (2006). Donor-derived mesenchymal stem cells are immunogenic in an allogeneic host and stimulate donor graft rejection in a nonmyeloablative setting. Blood, 108:2114–2120.

11.

Grinnemo

, Månsson

, Dellgren

, Klingberg

, Wardell

, Drvota

, Tammik

, Holgersson

, Ringdén

, Sylvén

and Le Blanc

. (2004). Xenoreactivity and engraftment of human mesenchymal stem cells transplanted into infarcted rat myocardium. J Thorac Cardiovasc Surg, 127:1293–1300.

12.

Larsson

, Corbascio

, Pearson

, Larsen

, Ekberg

and Widner

. (2003). Induction of operational tolerance to discordant dopaminergic porcine xenografts1. Transplantation, 75:1448–1454.

13.

Kumagai-Braesch

, Ekberg

, Wang

, Osterholm

, Ehrnfelt

, Sharma

, Lindeborg

, Holgersson

and Corbascio

(2007). Anti-LFA-1 improves pig islet xenograft function in diabetic mice when long-term acceptance is induced by CTLA4Ig/anti-CD40L. Transplantation, 83:1259–1267.

14.

Pearl

, Lee

, Leveson-Gower

, Sun

, Ghosh

, Lan

, Ransohoff

, Negrin

, Davis

and Wu

. (2011). Short-term immunosuppression promotes engraftment of embryonic and induced pluripotent stem cells. Cell Stem Cell, 8:309–317.

15.

Ljung

, Simonson

, Felldin

, Wärdell

, Ibarra

, Antonsson

, Kumagai-Braesch

, Hovatta

, Lampela

, Grinnemo

and Corbascio

. (2013). Costimulation blockade induces Foxp3(+) regulatory T cells to human embryonic stem cells. Biores Open Access, 2:455–458.

16.

Larsen

, Elwood

, Alexander

, Ritchie

, Hendrix

, Tucker-Burden

, Cho

, Aruffo

, Hollenbaugh

, et al. (1996). Long-term acceptance of skin and cardiac allografts after blocking CD40 and CD28 pathways. Nature, 381:434–438.

17.

Grinnemo

, Genead

, Kumagai-Braesch

, Andersson

, Danielsson

, Månsson-Broberg

, Dellgren

, Strömberg

, Ekberg

, et al. (2008). Costimulation blockade induces tolerance to HESC transplanted to the testis and induces regulatory T-cells to HESC transplanted into the heart. Stem Cells, 26:1850–1857.

18.

Hughes

, Postovit

and Lajoie

. (2010). Matrigel: a complex protein mixture required for optimal growth of cell culture. Proteomics, 10:1886–1890.

19.

Kleinman

and Martin

. (2005). Matrigel: basement membrane matrix with biological activity. Semin Cancer Biol, 15:378–386.

20.

Kleinman

, McGarvey

, Liotta

, Robey

, Tryggvason

and Martin

. (1982). Isolation and characterization of type IV procollagen, laminin, and heparan sulfate proteoglycan from the EHS sarcoma. Biochemistry, 21:6188–6193.

21.

Orkin

, Gehron

, McGoodwin

, Martin

, Valentine

and Swarm

. (1977). A murine tumor producing a matrix of basement membrane. J Exp Med, 145:204–220.

22.

Laflamme

, Chen

, Naumova

, Muskheli

, Fugate

, Dupras

, Reinecke

, Xu

, Hassanipour

, et al. (2007). Cardiomyocytes derived from human embryonic stem cells in pro-survival factors enhance function of infarcted rat hearts. Nat Biotechnol, 25:1015–1024.

23.

Oswald

, Boxberger

, Jørgensen

, Feldmann

, Ehninger

, Bornhäuser

and Werner

. (2004). Mesenchymal stem cells can be differentiated into endothelial cells in vitro. Stem Cells, 22:377–384.

24.

Allameh

, Jazayeri

and Adelipour

. (2016). In vivo vascularization of endothelial cells derived from bone marrow mesenchymal stem cells in SCID mouse model. Cell J, 18:179–188.

25.

Månsson-Broberg

, Rodin

, Bulatovic

, Ibarra

, Löfling

, Genead

, Wärdell

, Felldin

, Granath

, et al. (2016). Wnt/β-catenin stimulation and laminins support cardiovascular cell progenitor expansion from human fetal cardiac mesenchymal stromal cells. Stem Cell Rep, 6:607–617.

26.

Horwitz

, Le Blanc

, Dominici

, Mueller

, Slaper-Cortenbach

, Marini

, Deans

, Krause

and Keating

; International Society for Cellular Therapy. (2005). Clarification of the nomenclature for MSC: the International Society for Cellular Therapy position statement. Cytotherapy, 7:393–395.

27.

Iorio

, Troughton

and Hamill

. (2015). Laminins: roles and utility in wound repair. Adv Wound Care, 4:250–263.

28.

Yurchenco

. (2011). Basement membranes: cell scaffoldings and signaling platforms. Cold Spring Harb Perspect Biol, 3:a004911.

29.

Benton

, Arnaoutova

, George

, Kleinman

and Koblinski

. (2014). Matrigel: from discovery and ECM mimicry to assays and models for cancer research. Adv Drug Deliv Rev, 79–80:3–18.

30.

Chen

WCW

, Park

, Murray

, Zimmerlin

, Lazzari

, Huard

and Péault

. (2013). Cellular kinetics of perivascular MSC precursors. Stem Cells Int, 2013:983059.

31.

Caplan

. (2017). New MSC: MSCs as pericytes are sentinels and gatekeepers. J Orthop Res, 35:1151–1159.

32.

Meirelles Lda

, Fontes

, Covas

and Caplan

. (2009). Mechanisms involved in the therapeutic properties of mesenchymal stem cells. Cytokine Growth Factor Rev, 20:419–427.

33.

Caplan

and Dennis

. (2006). Mesenchymal stem cells as trophic mediators. J Cell Biochem, 98:1076–1084.

34.

Gong

, Yu

, Wang

, Liu

, Paul

, Millard

, Xiao

, Ashraf

and Xu

. (2017). Mesenchymal stem cells release exosomes that transfer miRNAs to endothelial cells and promote angiogenesis. Oncotarget, 8:45200–45212.

35.

Comşa

, Ceauşu

, Popescu

, Cîmpean

and Raica

. (2017). The human mesenchymal stem cells and the chick embryo chorioallantoic membrane: the key and the lock in revealing vasculogenesis. In Vivo, 31:1139–1144.

36.

Ribatti

, Nico

, Vacca

, Roncali

, Burri

and Djonov

. (2001). Chorioallantoic membrane capillary bed: a useful target for studying angiogenesis and anti-angiogenesis in vivo. Anat Rec, 264:317–324.

37.

Srivastava

. (2006). Making or breaking the heart: from lineage determination to morphogenesis. Cell, 126:1037–1048.

38.

Drake

and Fleming

. (2000). Vasculogenesis in the day 6.5 to 9.5 mouse embryo. Blood, 95:1671–1679.

39.

Shultz

, Lyons

, Burzenski

, Gott

, Chen

, Chaleff

, Kotb

, Gillies

, King

, et al. (2005). Human lymphoid and myeloid cell development in NOD/LtSz-scid IL2R gamma null mice engrafted with mobilized human hemopoietic stem cells. J Immunol, 174:6477–6489.

40.

Linsley

, Brady

, Urnes

, Grosmaire

, Damle

and Ledbetter

. (1991). CTLA-4 is a second receptor for the B cell activation antigen B7. J Exp Med, 174:561–569.

41.

Lederman

, Yellin

, Krichevsky

, Belko

, Lee

and Chess

. (1992). Identification of a novel surface protein on activated CD4⁺ T cells that induces contact-dependent B cell differentiation (help). J Exp Med, 175:1091–1101.

42.

Wing

, Onishi

, Prieto-Martin

, Yamaguchi

, Miyara

, Fehervari

, Nomura

and Sakaguchi

. (2008). CTLA-4 control over Foxp3⁺ regulatory T cell function. Science, 322:271–275.

43.

Ito

, Hanabuchi

, Wang

, Park

, Arima

, Bover

, Qin

FX-F

, Gilliet

and Liu

. (2008). Two functional subsets of FOXP3⁺ regulatory T cells in human thymus and periphery. Immunity, 28:870–880.

44.

, Chien

and Mummery

. (2008). Origins and fates of cardiovascular progenitor cells. Cell, 132:537–543.

45.

Bearzi

, Leri

, Lo Monaco

, Rota

, Gonzalez

, Hosoda

, Pepe

, Qanud

, Ojaimi

, et al. (2009). Identification of a coronary vascular progenitor cell in the human heart. Proc Natl Acad Sci U S A, 106:15885–15890.

46.

, Jiang

, Martin-Puig

, Caron

, Zhu

, Shao

, Roberts

, Huang

, Domian

and Chien

. (2009). Human ISL1 heart progenitors generate diverse multipotent cardiovascular cell lineages. Nature, 460:113–117.

47.

Kattman

, Huber

and Keller

. (2006). Multipotent flk-1⁺ cardiovascular progenitor cells give rise to the cardiomyocyte, endothelial, and vascular smooth muscle lineages. Dev Cell, 11:723–732.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.12 MB

0.09 MB

0.02 MB

Human Fetal Cardiac Mesenchymal Stromal Cells Differentiate In Vivo into Endothelial Cells and Contribute to Vasculogenesis in Immunocompetent Mice

Abstract

Introduction

Materials and Methods

Derivation and expansion of hfcMSCs

MSC phenotyping and differentiation

Ratio of CD31+ cells among the initial hfcMSCs

3D Matrigel cultures

Animal experiments

Immunohistochemistry

Fluorescence in situ hybridization

TaqMan polymerase chain reaction

Statistics

Results

hfcMSCs express high levels of αSMA at the messenger RNA and protein level

hfcMSCs delivered in Matrigel survive up to 8 weeks in the subcutaneous tissue of mice treated with triple costimulation blockade

hfcMSCs express vascular markers in vivo

Laminin α4 originates from hfcMSCs

Discussion

Footnotes

Acknowledgment

Author Disclosure Statement

Supplementary Material

References

Supplementary Material

Ratio of CD31⁺ cells among the initial hfcMSCs