Mesenchymal stromal cell therapy reduces lung inflammation and vascular remodeling and improves hemodynamics in experimental pulmonary arterial hypertension

Background Experimental research has reported beneficial effects of mesenchymal stromal cell (MSC) therapy in pulmonary arterial hypertension (PAH). However, these studies either were based on prophylactic protocols or assessed basic remodeling features without evaluating possible mechanisms. We analyzed the effects of MSC therapy on lung vascular remodeling and hemodynamics and its possible mechanisms of action in monocrotaline (MCT)-induced PAH. Methods Twenty-eight Wistar rats were randomly divided into two groups. In the PAH group, animals received MCT 60 mg/kg intraperitoneally, while a control group received saline (SAL) instead. On day 14, both groups were further randomized to receive 105 adipose-derived MSCs or SAL intravenously (n = 7/group). On day 28, right ventricular systolic pressure (RVSP) and the gene expression of mediators associated with apoptosis, inflammation, fibrosis, Smad-1 levels, cell proliferation, and endothelial–mesenchymal transition were determined. In addition, lung histology (smooth muscle cell proliferation and plexiform-like injuries), CD68+ and CD163+ macrophages, and plasma levels of vascular endothelial growth factor (VEGF) and platelet-derived growth factor (PDGF) were evaluated. Results In the PAH group, adipose-derived MSCs, compared to SAL, reduced mean RVSP (29 ± 1 vs 39 ± 2 mmHg, p < 0.001), lung tissue collagen fiber content, smooth muscle cell proliferation, CD68+ macrophages, interleukin-6 expression, and the antiapoptotic mediators Bcl-2 and survivin. Conversely, expression of the proapoptotic mediator procaspase-3 and plasma VEGF increased, with no changes in PDGF. In the pulmonary artery, MSCs dampened the endothelial–mesenchymal transition. Conclusion In MCT-induced PAH, MSC therapy reduced lung vascular remodeling, thus improving hemodynamics. These beneficial effects were associated with increased levels of proapoptotic markers, mesenchymal-to-endothelial transition, reduced cell proliferation markers, and inflammation due to a shift away from the M1 phenotype. Electronic supplementary material The online version of this article (doi:10.1186/s13287-017-0669-0) contains supplementary material, which is available to authorized users.

subpopulations) and molecular factors [22] responsible for the beneficial effects of MSCs on right-heart hemodynamics in PAH.
The present study uses a rat model of monocrotalineinduced PAH to investigate the effects of MSC therapy on vascular histology and hemodynamics and to attempt to elucidate the possible mechanisms of MSC action in PAH, through analyses of macrophage subpopulations, expression of growth factors, levels of proapoptopic and antiapoptotic mediators, endothelial-mesenchymal transition, and cell proliferation.

Methods
This study was approved by the Institutional Animal Care and Use Committee of the Health Sciences Centre, Federal University of Rio de Janeiro (CEUA 114/14), and registered with the Brazilian National Council for Animal Experimentation Control. All animals received humane care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Society for Medical Research and the US National Academy of Sciences Guide for the Care and Use of Laboratory Animals. The present study followed the ARRIVE guidelines for reporting of animal research [23].

Animal preparation and experimental protocol
On day 0, echocardiography analysis of the right ventricle was performed in 28 male Wistar rats (weight 235 ± 21 g), which were then randomized into two groups: monocrotaline-induced PAH (MCT, n = 14), in which animals received 60 mg/kg of MCT intraperitoneally (i.p.) (C2401; Sigma Chemical Co., St Louis, MO, USA); and control (CTRL, n = 14), in which animals received a similar volume of saline solution i.p. On day 14, cardiovascular function impairment was assessed by repeat echocardiography in the CTRL and MCT groups. Both groups were randomized again to receive either MSCs (n = 7, 10 5 cells intravenously (i.v.) through the jugular vein diluted in 50 μl saline solution) or an equivalent volume of saline solution (SAL, n = 7). On day 28, echocardiography was repeated, hemodynamic parameters were measured, and animals were killed for postmortem analyses.

Adipose cell culture
For isolation of adipose-derived MSCs, male Wistar rats were killed with sevoflurane (Sevorane® 2.0 vol.%; Cristália, Itapira, São Paulo, SP, Brazil) and the inferior vena cava was transected. Fat tissue around the epididymis was identified, removed, and placed in 1× phosphate-buffered saline (PBS). The tissue was then mechanically dissociated and placed in 2 ml of type 1 collagenase solution (1 mg/ml) for 35 minutes at 37°C, with homogenization every 10 minutes. DMEM (Invitrogen Life Technologies, Grand Isle, NY, USA) supplemented with 10% fetal bovine serum (FBS; Hyclone, Rockford, IL, USA), 1% penicillin/streptomycin (Pen/Strep; Invitrogen Life Technologies), and 2 mM L-glutamine (Invitrogen) was added to the solution for collagenase inactivation, and the whole solution was centrifuged at 300 × g for 5 minutes. The pellet was transferred into tissue culture dishes containing the same medium. The culture medium was changed on the day after extraction and three times a week thereafter. MSCs were expanded in culture using Iscove's Modification of Dulbecco's Medium (IMDM; Hyclone), supplemented with 10% FBS, 10% horse serum (Hyclone), 1% Pen/Strep, and 2 mM Lglutamine; MSCs were used at the third to sixth passage. They were kept in culture at confluence no greater than 70%. Immediately before injections, cells were lifted using 2.5% Trypsin/EDTA (Invitrogen Life Technologies). Cell density and viability was determined using Trypan blue staining and counted using a hemocytometer. Cell pellets were then resuspended in 1× PBS to a final concentration of 1 × 10 5 cells per 50 μl immediately prior to injection. Approximately 1 × 10 5 cells at the third passage were characterized as MSCs according to the International Society of Cellular Therapy Consensus; that is, adhesion to plastic under standard conditions, expression of specific surface markers (CD73, CD90, and CD105) and lack of others (CD11b, CD19, CD34, CD45), and ability to differentiate into mesenchymal lineages (such as adipocytes, chondrocytes, and osteoblasts) under in-vitro conditions [24]. For flow cytometry analysis, we used antibodies against CD90-PE (thymocytes), CD34-PE (hematopoietic precursors), and CD29-FITC (nonhematopoietic precursors) (BD Biosciences, USA). The absence of CD90 and CD34 and the presence of CD29 were used to identify MSCs [25]. The different MSC populations were further characterized by their capacity to differentiate into osteoblasts and chondroblasts. Osteogenic differentiation was induced by culturing MSCs for up to 3 weeks in DMEM 10% FBS and 15 mM HEPES (Sigma Chemical Co., St Louis, MO, USA), supplemented with 10 -8 M/L dexamethasone (Sigma Chemical Co.), 5 μg/ml ascorbic acid 2-phosphate (Sigma Chemical Co.), and 10 mM/L β-glycerolphosphate (Sigma Chemical Co.). To observe calcium deposition, cultures were stained with Alizarin Red S (Nuclear, SP, Brazil). To induce chondrogenic differentiation, MSCs were cultured in DMEM supplemented with 10 ng/ml TGF-β1 (Sigma, St. Louis, MO, USA), 50 nM ascorbic acid 2-phosphate (Sigma Chemical Co.), and 6.25 mg/ml insulin for 3 weeks. To confirm differentiation, cells were fixed with 4% paraformaldehyde in PBS for 1 hour at room temperature, and then stained with Alcian Blue (pH 2.5) [26].

Echocardiography
For echocardiographic assessment of cardiovascular function, rats were anesthetized with sevoflurane (Sevorane® 2.0 vol.%; Cristália), shaved over the precordial region, and examined with a Vevo 770 system (VisualSonics, Toronto, ON, Canada) coupled to a 30-MHz transducer. Images were obtained from parasternal views. Pulsed-wave Doppler was used to measure the pulmonary artery acceleration time (PAT) and pulmonary artery ejection time (PET), and the PAT/PET ratio was used as an indirect index of PAH. All parameters followed the American Society of Echocardiography and European Association of Cardiovascular Imaging guidelines, as validated in a previous article on noninvasive assessment of murine pulmonary arterial pressure [27,28].

Hemodynamic measurements
Rats were anesthetized with fentanyl 100 μg/kg i.p. (União Química, São Paulo, SP, Brazil) and midazolam 5 mg/kg i.p. (Dormicum, União Química, São Paulo, SP, Brazil). If required, additional anesthesia was provided by inhaled sevoflurane through a nasal device. Animals were placed in dorsal recumbency and the skin and muscle tissue of the anterior chest wall were removed until the ribs were visible. The thorax was then opened and a heparinized 19-G winged infusion set (Embramac, Itapira, SP, Brazil) was inserted into the RV. The RV systolic pressure (RVSP) and RV contraction rate (RV dp/dtmax) were assessed with a physiological data acquisition system (LabChart 7.0; ADInstruments, Sydney, Australia). Blood samples were collected through inferior vena cava puncture into a heparinized syringe. These samples were then centrifuged at 1500 × g for 5 minutes at 4°C for plasma retrieval for enzyme-linked immunosorbent assay (see later). At the end of the experiment, animals were euthanized by i.v. injection of sodium thiopental 25 mg (Cristália) followed by complete exsanguination through transection of the abdominal aorta.

Determination of pulmonary artery function
After hemodynamic measurements, the heart and lung were removed. The main pulmonary artery was carefully dissected and cleaned of connective tissue. Arteries were attached to a force transducer and placed in vertical chambers filled with modified Tyrode's solution (123 mM NaCl, 4.7 mM KCl, 1.2 mM MgCl 2 , 1.2 mM KH 2 PO 4 , 15.5 mM NaHCO 3 , 1.2 mM CaCl 2 , and 11.5 mM dextrose), which were oxygenated and kept at 37°C. After a stabilization period of 2 hours in 1.5 g tension, the solutions were exposed to increasing doses of phenylephrine (Phe, 1 nM-10 μM). After the contraction plateau, the arteries were exposed to increasing doses of acetylcholine (ACh, 1 nM-10 μM) to evaluate endothelial function by the maximum relaxation capacity.

Quantification of lung collagen
The right upper lobe of the lung was homogenized in Tris-HCl 0.05 M and NaCl 1 M containing protease inhibitor (Hoffmann-La Roche) (pH 7.4). Total soluble collagen was extracted overnight at room temperature and quantified using the Sircol™ kit (Biocolor, Newton Abbey, UK). The results were expressed as micrograms of collagen/milligram of total protein by Bradford's technique.

Optical microscopy
For optical microscopy (OM) analysis, an additional 24 Wistar rats (n = 6/group) were subjected to the same protocol as already described. To keep lungs inflated before fixation, a similar end-expiratory pressure (3 cmH 2 O) was achieved among all groups by a brief period of mechanical ventilation (Flexivent; SCIREQ, Canada). Heparin (1000 IU) was injected into the tail vein to allow better visualization of lung structures under OM. To extract the lungs, a laparotomy was performed and the inferior vena cava transected to kill the animals by exsanguination. The left lung was then quickly isolated, frozen in liquid nitrogen, and placed in Carnoy's solution. The lungs were dehydrated in a graded ethanol series and embedded in paraffin. Slices (4-μm thick) were cut, deparaffinized, and stained with hematoxylin-eosin (H&E). The histopathological features of the pre-acinar and intra-acinar pulmonary arteries were classified according to Pietra et al. [21] on the basis of: medial hypertrophy with increase in crosssectional area; intimal thickening, defined as concentric laminar, eccentric, or concentric nonlaminar (both lesion types composed of fibroblasts and connective tissue matrix); and dilatation lesions. These histopathological features were graded on a five-point, semiquantitative, severity-based scoring system as follows: 0 = normal, 1 = changes in 1-25% of examined tissue, 2 = changes in 26-50% of examined tissue, 3 = changes in 51-75% of examined tissue, and 4 = changes in 76-100% of examined tissue. This semiquantitative five-point score has been validated in different animal models [26,29]. Scoring was assessed independently by one coauthor (VLC), who is an expert in lung pathology and was blinded to group allocation.
Lung slices were also stained with the modified Russel-Movat method to highlight the muscular, connective, and elastic fibers of lung vessels in histological slides [30]. In brief, a mixture of five stains-Alcian Blue, Verhoeff hematoxylin, and Crocein Scarlet combined with acidic fuchsine and saffron-was used. At pH 2.5, Alcian Blue was fixed by electrostatic binding with the acidic mucopolysaccharides. Verhoeff hematoxylin has a high affinity for nuclei and elastin fibers, which are negatively charged. The combination of Crocein Scarlet with acidic fuchsine stains acidophilic tissue components in red. Then collagen and reticulin fibers were unstained by a reaction with phosphotungstic acid, and stained yellow by saffron. Smooth muscle hypertrophy and vascular endothelial growth factor (VEGF) expression were quantified using a weighted scoring system, as described elsewhere [31,32]. Briefly, values from 0 to 4 were used to represent the severity of smooth muscle hypertrophy and VEGF expression, with 0 standing for no effect and 4 for maximum severity. Additionally, the extent of smooth muscle hypertrophy and VEGF expression per field of view was determined using values of 0-4, with 0 standing for no appearance and 4 for complete involvement. The final score for each feature was calculated as the product of extension and severity, ranging from 0 to 16. Scoring was assessed independently by one coauthor (VLC), who is an expert in lung pathology and was blinded to group allocation.

Transmission electron microscopy of pulmonary arteries
Three slices measuring 2 mm × 2 mm × 2 mm were cut from three different segments of the right inferior lung. They were then fixed in 2.5% glutaraldehyde and phosphate buffer, 0.1 M (pH 7.4), for transmission electron microscopy (TEM) analysis (JEOL 1010 Transmission Electron Microscope; Japan Electron Optics Laboratory Co, Tokyo, Japan). For each TEM image (20 per animal), an injury score was determined. The following parameters were analyzed, considering the pulmonary arteries: medial hypertrophy and concentric laminar intimal thickening in the pre-acinar artery; and basement membrane thickening, myofibroblast proliferation, endothelial hyperplasia, and smooth muscle fibers in the intraacinar artery. These parameters were graded on a fivepoint scoring system similar to that already described for histopathological features.

Enzyme-linked immunosorbent assay
Plasma levels of VEGF and platelet-derived growth factor (PDGF) (PeproTech, Rocky Hill, NJ, USA) were quantified 14 days after MSC therapy or saline administration in the CTRL-SAL, CTRL-MSC, MCT-SAL, and MCT-MSC groups. The results were normalized by total protein content by Bradford's technique, and expressed as picograms per microgram. All measurements were performed in accordance with manufacturer guidelines.

Expression of TGF-β and type I and III procollagen in lung fibroblasts
Lung fibroblasts were extracted and RT-PCR was performed to measure biological markers associated with fibrosis (transforming growth factor (TGF)-β, type I procollagen (PC-1), and type III procollagen (PC-3)). Lung fibroblasts were isolated as described previously [37]. Briefly, the right upper lung was removed, cut into small fragments, digested enzymatically in Liberase solution at 37°C for 30 minutes, rinsed three times with warm DMEM/F12 media with 10% FBS, 1× penicillin/streptomycin, centrifuged at 524 × g, resuspended in complete DMEM/F12 media, transferred to a tissue culture dish, and, finally, placed in a humidified tissue culture incubator at 37°C and 5% CO 2 . After the cells reached 80-90% confluence, they were lysed and total RNA was extracted from frozen tissues using the RNeasy Plus Mini Kit (Qiagen), following the manufacturer's recommendations. Again, the RNA concentration was measured by spectrophotometry in Nanodrop ND-1000 (Thermo Scientific, Wilmington, DE, USA). First-strand cDNA was synthesized from total RNA using a Quantitec reverse transcription kit (Qiagen). Relative mRNA levels were measured with a SYBR green detection system using ABI 7500 real-time PCR (Applied Biosystems, Foster City, CA, USA). Samples were measured in triplicate. For each sample, the expression of each gene was normalized to 36B4 housekeeping gene expression [34] and expressed as fold changes relative to CTRL-SAL, using the 2 -ΔΔ Ct method described earlier.

Statistical analysis
Each variable was tested for normality and variance using the Kolmogorov-Smirnov and Levene tests respectively. Data are presented as mean ± SD unless otherwise specified. On day 14, parametric data were compared by a one-sample Student's t test (α < 0.05). For comparisons between MCT at 14 days vs MCT-SAL and MCT-MSC at 28 days, paired t tests were used (α < 0.05). On day 28, parametric and nonparametric data were analyzed by multiple Student t tests and Wilcoxon signed-rank tests respectively, followed by Bonferroni's correction (adjusted α* < 0.0167 for three comparisons: CTRL-SAL vs CTRL-MSC, CTRL-SAL vs MCT-SAL, and MCT-SAL vs MCT-MSC). The Bonferroni-adjusted significance level was used to counteract multiple comparisons [38]. Pearson's correlations were calculated and p < 0.05 was adopted as significant. All tests were performed in GraphPad Prism version 6.01 (Graph-Pad Software, La Jolla, CA, USA).

Echocardiography and hemodynamic measurements
Fourteen days after MCT induction of PAH, the PAT/ PET ratio was reduced in PAH animals compared to CTRL, suggesting pulmonary hypertension (0.38 ± 0.09 vs 0.47 ± 0.08 respectively; p = 0.016). The ratio remained lower at day 28 (MCT-SAL, 0.27 ± 0.07 vs CTRL-SAL, 0.40 ± 0.05; p < 0.001). In the MCT-SAL group, the PAT/ PET ratio was lower at day 28 compared to day 14 (p = 0.03). On the other hand, in the MCT-MSC group the PAT/PET ratio did not differ between days 14 and 28 (p = 0.54), which may suggest mitigation of the course of PAH (Fig. 1a).

Endothelial dysfunction of pulmonary hypertension
Arteries from MCT-SAL animals exhibited less of a vasodilator response to increasing doses of ACh compared to those of CTRL-SAL animals (48 ± 5% vs 64 ± 2% respectively). This result was reinforced by calculation of the effective dose 50 (ED50) for ACh-induced dilation in both groups, in which CTRL-SAL animals showed a lower ED50 (-6.99) than MCT-SAL animals (-6.11). After MSC therapy, no additional vasodilator response of the pulmonary arterial rings was observed in MCT animals (Additional file 1: Figure S1).
On TEM analysis, MCT-SAL animals showed concentric laminar intimal thickening and prominent basement membrane in the pre-acinar and intra-acinar pulmonary arteries. In addition, MCT-SAL animals exhibited external fibronectin fibers and intracellular actin microfilaments. After MSC therapy, concentric laminar intimal and basement-membrane thickening decreased (Additional file 1: Figure S2).
MCT-SAL animals showed higher IL-6 (p < 0.0167) and survivin mRNA expression in lung tissue compared to CTRL-SAL. In addition, Smad-1 and GSK3β mRNA expressions were higher in MCT-SAL than CTRL-SAL animals (p < 0.0167 for both). In pulmonary artery tissue, VE-cadherin mRNA expression was lower, while vimentin and α-actin were higher in MCT-SAL animals compared to CTRL-SAL.
In lung fibroblasts, MCT-SAL compared to CTRL-SAL animals showed higher procollagen I, procollagen III, and TGF-β mRNA expressions, which were reduced in the MCT-MSC group in comparison to MCT-SAL (p < 0.0001) (Fig. 5b-d).

Discussion
In the monocrotaline-induced experimental model of PAH used herein, MSC therapy reduced right ventricular systolic pressure, histological damage and smooth muscle cell thickening, CD68 + macrophage density in lung tissue, lung arteriolar and plasma VEGF levels, total lung collagen, mRNA levels of procollagen type I, type III, and TGF-β in lung fibroblast culture, and inflammatory and anti-apoptotic markers in lung tissue. MSC therapy increased proapoptotic mediators in lung tissue   (10 4 μm 2 ). a, c Very few CD68 + cells were observed. e CD68 + cells were observed (black arrows). g Few CD68 + cells were observed. b, d, f, Very few CD163 + cells were observed. h CD163 + cells were observed (black arrows). Lines represent medians, and whiskers are the interquartile range of seven animals in each group. Comparisons by Student's t test followed by Bonferroni's corrections for three comparisons (*p < 0.0167). CTRL control, MCT monocrotaline, SAL saline  Monocrotaline has been widely used to induce PAH in small rodents [6,39,40]. In this line, we observed several pathophysiological features associated with PAH in animals exposed to MCT, such as an increase in RV hypertrophy index (RV/LV + S; data not shown), RVSP, and medial hypertrophy and concentric laminar intimal thickening in both pre-acinar and intra-acinar arteries. Since we were concerned about the functional and histopathological features of PAH, the usage of MCT would fit with our research question, as well as with the respective time course alterations [41] after MSC treatment.

Cell therapy and inflammation
In the present study, we used 10 5 adipose-derived MSCs (a lower dosage than in previous studies [16]), obtained from healthy animals, and observed an attenuation of inflammatory markers, including IL-6, as in previous preclinical studies using MSCs in PAH animals. IL-6 is a key factor that facilitates a shift in pericytes from the quiescent state into their contractile, hyperproliferative state [42], which would increase pulmonary vascular resistance in PAH. Nevertheless, the association of IL-6 with macrophage polarization in PAH has not been clear. By modulating macrophage polarization and therefore IL-6 production, induced pluripotent stem cell (iPSC)-based therapy has been shown to reduce inflammation and M1 phenotype subpopulations in human polarized macrophages in vitro [11]. Herein, we observed not only IL-6 and CD68 + macrophage attenuation but also an increase in the CD163 + macrophage subpopulation in vivo. Although we did not evaluate the precise mechanism of mitigation of inflammatory process, it may rely on inhibition of the NF-κB pathway [11]. By modulating the M1 phenotype and IL-6 mRNA expression, MSCs may minimize the intense proliferation observed in PAH. In this line, we observed a reduction in mRNA expression of Bcl-2 and survivin, which are known to have antiapoptotic properties [43], leading to cell survival, and intense remodeling [44]. Likewise, we observed a relative increase in procaspase-3 mRNA expression after MSC therapy in MCT animals, which, by increasing apoptosis, may counteract the intense cell proliferation observed in PAH.

Cell therapy and remodeling
Previous studies have reported a reduction in vascular remodeling after MSC therapy [16][17][18]20]. Despite using a lower dose of MSCs than prior studies [16][17][18][19][20], we observed a reduction in the remodeling process. Furthermore, previous studies have focused only on pulmonary vessel wall thickness, which is a more general measurement [16][17][18]20]. Conversely, we were interested in the main histopathological features of PAH seen in the pre-acinar and intra-acinar pulmonary arteries [21]. In the pre-acinar arteries, after MSC therapy, we observed a reduction in medial hypertrophy and plexiform-like lesions, which represent specific features in PAH [21,45], while in the intra-acinar arteries there was a reduction in medial hypertrophy and concentric laminar intimal thickening. Similar results were observed after MSC therapy in atherosclerosis lesions [46]. In pulmonary artery smooth muscle cells, GSK3β represents a mediator of the Akt pathway toward cell proliferation [47]. It has been shown that GSK3β increases in monocrotaline-induced PAH, and its levels are negatively modulated by HIV protease inhibitors in line with decreased proliferation in pulmonary arteries [48]. We observed a reduction in GSK3β but not Smad-1 levels in pulmonary artery tissue after MSC therapy. This suggests that MSCs may interfere with the Akt-GSK3β axis, one of the major non-SMAD signaling pathways [49] toward cell proliferation [47]. Since Smad-1 belongs to the bone marrow protein receptor (BMPR)/Smad signaling pathway, the lack of effect observed after MSC therapy may be explained by the minor role of BMPR-2 in pulmonary artery smooth muscle cells in monocrotalineinduced PAH [50]. We also isolated fibroblast cells after MSC therapy in order to provide further support to cell therapy and remodeling. Fibroblast cells can change their phenotype into myofibroblasts, acting on mRNA expression of procollagen and TGF-β, which have been associated with intimal arterial injury during PAH [2]   and intense endothelial-to-mesenchymal transition after MCT administration [51]. We observed a reduction in TGF-β, procollagen type I, and procollagen type III mRNA expression in lung fibroblast cell cultures after MSC therapy. Previous studies have demonstrated that MSC-conditioned medium significantly downregulates procollagen type I and procollagen type III mRNA expression, which reflects the paracrine antifibrotic effects of MSCs [52]. In addition, along with increased VEcadherin, we also detected reductions in vimentin and α-actin, which suggests that the change from an endothelial into a mesenchymal phenotype observed in PAH [51] was dampened after MSC administration.

Cell therapy and growth factors
Because of the controversial roles of different growth factors in PAH [53], we decided to evaluate VEGF and PDGF levels over time in plasma and small pulmonary vessels. VEGF is the most important factor regulating both physiological and pathogenic angiogenesis [54]. Serum and endothelial levels of VEGF, as well as VEGF receptor 2 (VEGFR-2) expression, are known to be increased in PAH patients [4,55]. In PAH pathogenesis, VEGF may act as a potent stimulator of the disordered angiogenic process through a phenotype shift of pulmonary vascular smooth muscle cells from a quiescent state into a proliferative and antiapoptotic profile [7]. In this line, balancing this pathway might provide a desirable therapeutic avenue.
In ischemia-reperfusion injury [56], MSC therapy has been shown to increase lung tissue VEGF levels, which was followed by vascularization and angiogenesis in vitro [56,57]. In contrast, after MSC therapy, we observed lower VEGF levels in both plasma and small pulmonary vessels. The relationship between inflammation and angiogenesis may explain these contradictory findings. MSCs reduced inflammation by changing the macrophage phenotype from M1 (proinflammatory) toward M2 (antiinflammatory) and, consequently, reducing IL-6 mRNA expression. In this scenario, lower proliferation and angiogenesis and, therefore, lower VEGF levels are to be expected. In addition, at the time of MSC administration (i.e., on day 14), circulating levels of VEGF were higher in MCT compared to CTRL animals. Administration of MSCs on day 14 would balance out the intense proliferative state and might act at the endothelial niche by the interaction of pericytes and the innate immune system [58], by increasing plasticity [59], and by balancing pulmonary vessel density [60].
In contrast, since no major alterations in plasma PDGF levels were observed, we did not evaluate PDGF in lung tissue after MSC therapy. Therefore, the observed reduction in proliferative status may be largely attributable to VEGF downregulation. It should be pointed out that the effects of MSCs may be attributable to their paracrine effects rather than to their presence in the injured lung area, since beneficial effects have been observed after administration of MSC-conditioned medium [61] and extracellular vesicles [62,63] in PAH and other respiratory diseases.

Cell therapy and functional parameters
In this experiment, MSC therapy may have improved hemodynamics, as represented by a reduction in RVSP, mainly due to attenuation of vascular remodeling [64]. Accordingly, we did not observe a significant vasodilatory response of the pulmonary arterial rings after MSC therapy, which provides further evidence of this effect. Previous studies have reported reductions in pulmonary arterial pressures after MSC administration, followed by a vasodilatory response to ACh [19]. The methods used in these prior investigations were based on the in-vivo setting, and involved i.v. injection of acetylcholine after administration of a prostaglandin agonist [19]. This has several limitations, including the fact that the systemic effects of ACh and prostaglandin might prevent proper measurement of endothelial dysfunction and that ACh dosages used in vivo are fourfold higher than those used in the present study during in-vitro measurements. In addition, the in-vitro method employed in our study has been widely used in prior research [65,66].

Limitations
The limitations of the present study must be taken into account. First, instead of performing repeated punctures for direct (invasive) measurement of RV pressures, we used echocardiography to choose the best timing for therapeutic interventions, according to indirect but reliable markers of pulmonary hypertension, such as the PAT/PET ratio [27,67,68]. Second, monocrotaline-induced PAH is not able to mimic the entire complex features observed in human PAH. Nevertheless, it is a well-established model of moderately severe PAH [69] and has been widely used both to elucidate its pathophysiological mechanisms [70,71] and to propose new therapeutic interventions [65]. Third, we chose to analyze the roles of VEGF and PDGF specifically because these mediators are associated with angiogenesis [8,72], as shown in prior studies with the MCT experimental model [71] and in human patients with PAH [2]. Nevertheless, we cannot rule out a role for other growth factors.

Conclusion
In experimental monocrotaline-induced PAH, MSC therapy improved hemodynamics by mitigating lung vascular remodeling. These beneficial effects seem to be associated with an increase in proapoptotic markers, modulation of endothelial-mesenchymal transition, and decrease in cell proliferation markers and inflammation due to reduction in the M1 macrophage subpopulation.