Background
The use of surface markers, such as CD34 for the prospective identification and isolation of hematopoietic stem cells (HSC) has fundamentally improved the clinical development and standardization of stem cell transplantation [
1‐
3]. In contrast, most, if not all knowledge regarding the immunosuppressive capacity of mesenchymal stem cells (MSC) derives from studies using starting cultures of mixed cell populations [
4,
5]. Therefore, the International Society for Cellular Therapy has published criteria for the definition of in vitro expanded heterogeneous MSC including positivity for CD73, CD90 and CD105, negativity for hematopoietic surfaces markers and other parameters [
6]. Although these criteria are helpful to facilitate the comparison of heterogeneous MSC cultures, the prospective identification and direct isolation of homogenous MSC populations on the basis of surface markers will improve the standardization, comparability and reproducibility of MSC research.
Morikawa et al. has recently identified surface markers for the prospective identification of purified murine multipotent MSC from adult BM [
7,
8]. Their results suggested that the co-expression of Platelet-derived growth factor receptor α (PDFGRα, CD140a) and stem-cell antigen 1 (Sca-1) on CD45 and TER119 negative BM cells effectively identifies primary MSC and these were termed double positive PDFGRa
+ SCA1
+ CD45
− TER119
− (PαS) MSC [
7]. PαS MSC exhibited the highest numbers of colonies in a colony-forming unit-fibroblasts (CFU-F) assay and represented the only BM subset yielding MSC typical fibroblast/spindle-shaped cells. Only PαS cultures demonstrated a potent multilineage differentiation capacity into adipogenic, chondrogenic and osteogenic lineages indicating a high enrichment for MSC [
7].
Heterogeneous results have been reported with regard to the immunosuppressive capacity of MSC [
9,
10]. With respect to T cells, different pathways have been suggested including the direct inhibition of T effector cell proliferation and promotion of regulatory T cells [
5]. Similarly, MSC have been suggested to inhibit the cytotoxicity of NK cells and CD8
+ T lymphocytes [
11,
12]. With respect to dendritic cells (DC), MSC inhibit the monocyte-derived DC differentiation and suppress the costimulatory molecule expression, the pro-inflammatory cytokines IL12p70 and TNF-α as well as the priming of responder T cells [
5,
13,
14].
However, an increasing number of reports indicate that MSC are not immunosuppressive by themselves but can also exert immunostimulatory activity. For instance, Waterman et al. suggested that MSC can be polarized in immunosuppressive MSC1 or pro-inflammatory MSC2 by TLR3 and TLR4 triggering [
15]. Moreover, some groups have reported accelerated heart allograft rejection and a failure to inhibit graft-versus-host disease after MSC treatment [
16,
17]. Potential mechanisms of immunostimulatory MSC may be related to the release of activatory exosomes or CCL2 production.
With respect to these heterogeneous immunological results and the increasing number of MSC studies suggesting therapeutic activity in acute pneumonia [
18‐
20] the objective of this study was to investigate the in vivo immunomodulatory capacity of prospectively defined MSC in an animal model of gram negative pneumonia. Furthermore, to address more precisely the impact of this treatment on respiratory cellular inflammation we have used multiparameter flow cytometry allowing the dissection of key respiratory leukocyte subsets after MSC therapy [
21‐
23].
Methods
Mice, Klebsiella pneumoniae infection and MSC treatment
Specific-pathogen-free C57BL/6 (C57BL/6NCrl) and B6.Cg-Tg(TcraTcrb)425Cbn/J mice (20–25 g each) were purchased from Charles River, Germany and maintained under specific-pathogen-free conditions. The mice were infected intratracheally with 3.5 × 10
5 CFU of
Klebsiella pneumoniae (
K. pneumoniae) serotype 2 (
American Type Culture Collection (ATCC) 43816) in 50 μl sterile 0.9 % NaCl as previously described [
23]. Four hours p.i. the anesthetized mice received 1 × 10
6 washed PαS MSC in 50 μl 0.9 % NaCl intratracheally and were analyzed at the indicated time points (KpN/MSC). The control mice were treated identically but received 50 μl 0.9 % NaCl (KpN/NaCl). Some mice received 1 × 10
6 washed mouse lung fibroblasts (MLg;ATCC CCL-206) in 50 μl 0.9 % NaCl and were indicated as such (KpN/MLg). Experiments were approved by the regional animal authority board (#75/2011).
BM preparation, PαS MSC sorting and in vitro expansion
Femura and tibiae were prepared as previously described with minor modifications [
7]. The bone fragments were collected and digested for 1 h at 37 °C in alpha-MEM with L-Glutamine (PAN Biotech, Germany), 10 % FBS (PAA, Germany), 1 % penicillin/streptomycin (PAN Biotech) containing 3.92 U/ml collagenase (Wako Chemicals, Japan), 10 mM Hepes (Gibco, Germany) and 3 mM CaCl2. The cell suspension was filtered through a 70 μm cell strainer (BD Falcon, Germany) and collected by centrifugation at 400 g for 5 min at 4 °C. Red blood cells were lysed using 155 mM NH
4Cl/10 mM KHCO
3 buffer (pH 7.4) and washed with HBSS (PAN Biotech, Germany). After digestion, leucocytes were depleted with CD45 magnetic beads (Miltenyi Biotech, Germany) and stained with fluorochrome-labelled monoclonal antibodies and sorted by a BD ARIAIII cell sorter (Becton Dickinson, San Jose, CA, USA). PαS MSC were defined as positive for CD140a and Sca-1 and negative for CD45 and TER119 and were expanded in PureCoat Amine plates/flasks (BD, Germany) in alpha-MEM medium supplemented with L-Glutamine, 5 % FBS (mesenchymal stem cell-qualified, Life technologies, Germany), and 5 % human platelet lysate. Medium was changed every 3–7 days depending on cell growth. Human platelet lysate was prepared as previously described [
24].
Lung preparation
Lung single cell suspensions (lung homogenates) were prepared after enzymatic digestion as previously described in detail [
23]. In brief, the mice were euthanized and the lungs were perfused via the right ventricle with HBSS (PAA, Germany) to remove the intravascular pool of cells. The tissues were minced and digestion was performed in 0.09 U/ml type A collagenase (Roche, Germany) and 9.09 U/ml DNase (Roche, Germany) in IMDM (PAA, Germany) with 10 % FCS (PAA, Germany) at 37 °C for 1 h. The single cell suspensions were prepared by tissue resuspension with 20 G 1 ½ cannulas (0.9 × 40 mm; BD, Germany) and by mashing through a 70 μM cell strainer (BD, Germany). Red blood cells were lysed by ammonium chloride lysis. The cells were washed with HBSS for flow cytometry staining, or the leukocytes were magnetic-bead sorted after washing with PBS/2 % BSA/2 mM EDTA (PAA, Germany). Bronchoalveolar lavages (BAL) were performed as previously described [
25]. The bacterial load in lung homogenates and BAL were defined by preparing a two-fold dilution series in sterile HBSS after centrifugation (2500 g, 15 min, 4 ° C) according to the method developed by Schott [
26].
Flow cytometry
Cellular phenotyping and sorting were performed on a BD ARIAIII cell sorter (Becton Dickinson, San Jose, CA, USA). The following fluorochrome-labelled mAbs conjugated to FITC, PE, PeCy7, PerCPCy5.5, APC, APC-Cy7, Brilliant Violet 510, Brilliant Violet 605, Pacific Blue and Alexa700 or appropriate isotype controls were used for cell surface staining: CD11b, CD11c, CD45 (clone 30-F1), CD86, CD103, CD140a (clone APA5), CD274 (PD-L1), MHC-class II (I-Ab), GR-1, F4/80, NK1.1, SCA-1 (clone D7), Siglec-F (BD Biosciences, Germany), TER119 and 120G8 (Dendritics, France). The surface staining time was 30 min on ice and cells were washed with staining buffer (1× HBSS, PAA, Germany) at 400 g for 5 min at room temperature (RT) before analysis. The number of acquired events was ≥ 500,000.
The viability of sorted cells was >90 % as indicated by Sytox blue (Life Technologies, Germany) staining. All mAbs were ordered from Biolegend, Germany unless indicated otherwise. Absolute cell counts were determined with AccuCount Fluorescent Particles 7.7 μm (Spherotech, Lake Forest, USA).
CD4 T cell proliferation assay
CD4+ T cells were isolated from the spleens of OT-II mice (B6.Cg-Tg(TcraTcrb)425Cbn/J) using the CD4+ T Cell Isolation Kit (Miltenyi Biotec, Germany). CD4+ T cells were labeled with 0,5 μM CFSE (eBioscience, Germany) for 15 min at 37 °C with frequent agitation, then washed before being used in proliferation assays. DC were generated from C57BL/6 BM hematopoietic stem cells by culturing for 7 days in 10 ng/ml murine GM-CSF and 10 ng/ml murine IL-4 as previously described [
27]. CD4+ T-cells (1 × 10
5) were cultured with DC (1 × 10
4) and MSC (1 × 10
4) in RPMI 1640 containing 10 % FCS (PAA, Germany) for 5 days. Ovalbumin (OVA, 100 μg/ml) protein (Hyglos, Germany) was added to the culture medium. OVA-specific proliferation was evaluated as a CFSE dilution by flow cytometry.
Intracellular cytokine and Foxp3 staining
For cytokine staining cells were stimulated for 6 h at 37 °C with 50 ng/ml PMA (Sigma-Aldrich, Germany), 1 μg/ml Ionnomycin (Sigma-Aldrich, Germany) and 3 μg/ml brefeldin A (eBioscience, Germany) in RPMI with 10 % FBS (PAA, Germany). For intracellular cytokine or Foxp3 staining samples were first stained for surface antigens, washed with PBS (PAN Biotech, Germany), and centrifuged for 5 min at RT and 400 g. The cell pellets were vortexed for dissociation and incubated with fixation/permeabilization buffer (BD, Germany) 20 min at RT. After washing twice with 2 ml of permeabilization/washing buffer (BD, Germany) the cells were resuspended in 100 μl permeabilization buffer. Intracellular mAbs (IFN-γ, IL-4, IL-10, IL-17 from Biolegend, Germany; Foxp3, eBioscience, Germany) or isotype controls were added at the recommended concentrations and incubated 30 min at RT. Cells were washed two times with permeabilization/washing buffer and immediately analyzed by flow cytometry. The number of acquired events was ≥ 500,000.
Respiratory leukocyte subset discrimination
The gating strategy has been described recently with minor modifications [
23]. Briefly, out of the CD45
+ cells, neutrophils were identified by GR1
brightCD11b
bright expression. Subsequently, out of the neutrophil negative fraction, macrophages were identified as SiglecF
++/CD11c positive cells. DC were identified according to CD11c
+Siglec-F
neg NK1.1
neg expression to exclude autofluorescent macrophages and NK cells. Then they were further dissected into plasmacytoid DC (120 g8
+ CD11b
neg) and after MHC-class II
+ gating into CD103 DC (CD103
+ CD11b
neg) and CD11b DC (CD11b
+ CD103
neg). CD3
+ T cells and CD19
+ B cells were identified within the CD45
+ SSC
low fraction. Out of the CD3
+ T cell fraction, CD4
+ and CD8
+ T cells were identified on the basis of CD4 and CD8 expression, respectively. NK cells were identified as CD3
neg NK1.1
+ cells and γδ T cells were identified as CD3
+ γδ TCR
+ cells (Additional file
1: Figure S1). T regulatory cells were identified as CD3
+CD4
+CD25
+Foxp3
+ cells.
BAL protein and cytokine quantification
Protein quantification was performed with Pierce BCA Protein Assay Kit (Thermo Scientific, Germany). Mouse TNF-α, IL-10 and IL-12p70 were quantified by cytometric bead arrays according to the manufacturer’s instructions (Flowcytomix, eBioscience, Germany.
Statistical analyses
Statistical analyses were performed using the GraphPad Prism software version 5.02 (Graphpad Software, Inc., USA). The significance of any differences between groups were analyzed by the one-way ANOVA and Tukey post-test for multiple comparisons. Survival curve comparison and analysis was performed using the logrank test. A p-value of < 0.05 was considered statistically significant.
Discussion
In this study we report on the immunomodulatory capacity of prospectively defined MSC in a preclinical mouse model of acute bacterial pneumonia. Many studies have highlighted the immunomodulatory capacity of MSC but most, if not all studies have used heterogeneous mixtures of starting cells for the in vitro expansion of MSC. Recently Matsuzaki et al. has identified a BM-derived MSC subset, expressing PDGFRα and Sca-1, so-called PαS cells, that are highly enriched for CFU-Fs with differentiation potential [
7]. Here we have demonstrated that these prospectively defined PαS cells exhibited marked immunomodulatory capacity in vivo. The prospective identification of MSC based on surface markers facilitates the development of standardized protocols and additionally improves the comparability of scientific studies [
4,
8]. In accordance with Morikawa et al., we have found that cultured PαS cells uniformly expressed conventional MSC markers CD29, CD49e and SCA-1 and were positive for CD90 and CD105 [
7].
Our results using prospectively defined MSC showed that expanded PαS cells effectively suppressed acute lung injury and acute alveolitis caused by respiratory K. pneumoniae. We selected a
Klebsiella species as the pathogen because they represent an important group of bacteria that causes life threatening nosocomial infections [
28]. Furthermore, multidrug resistant
Klebsiella strains are of increasing clinical relevance worldwide due to limited treatment options [
29]. We showed that PαS MSC treatment significantly inhibited alveolar protein leakage, granulocytosis and TNF-α production indicating a marked in vivo anti-inflammatory capacity. Furthermore, PαS MSC treatment inhibited not only acute lung injury but additionally suppressed expansion of pro-inflammatory T helper subsets expressing IL-17 and IFN-γ in the post-acute pneumonia phase. Our flow cytometry analysis of respiratory DC subsets provides a rationale for understanding the marked inhibition of IL-17 and IFN-γ expressing T helper cells. Our results suggested that PαS MSC treatment effectively inhibited accumulation and CD86 upregulation on respiratory CD103
+ DC. Different reports have highlighted the critical role of CD103
+ DC for the expansion of IL-17 and IFN-γ driven T effector cell responses [
30,
31].
Various studies have investigated the effect of MSC in models of LPS- or bleomycin-induced lung injury and have reported significant amelioration after lung injury [
18,
32,
33] as well as a significant survival benefit in MSC-treated animals [
18,
33].
Only a few studies have investigated the therapeutic activity of MSC in pneumonia models with live bacteria [
19,
20,
34]. Gupta et al. and Kim et al. reported that intratracheal MSC treatment improved survival and bacterial clearance in murine
Escherichia coli pneumonia [
19,
20]. Here we have extended these findings by using prospectively defined PαS MSC in a
K. pneumonia-infected animal model. Independent animal models employing live bacteria are of clinical relevance because there may be concerns that “immunosuppressive” MSC might promote bacterial growth due to an impairment of host defences. Our results using prospectively defined MSC are in agreement with other data indicating that intratracheal MSC therapy during infectious pneumonia promotes rather than inhibits bacterial clearance [
19]. Matthay et al. has suggested that the antibacterial effect of MSC are due to the upregulation of the antibacterial proteins lipocalin 2 and LL37 [
19,
35]. Recent evidence by Devaney et al. further supports this finding in a rat E. coli pneumonia model. In their study, MSC-treatment significantly reduced acute lung injury, improved overall survival and decreased lung bacterial load. With respect to the potential anti-bacterial activity of MSC they reported enhanced macrophage phagocytic capacity and increased lung and systemic concentrations of the antimicrobial peptide LL37 in MSC-treated animals [
34]. In agreement with these findings, our histopathological analyses indicated a strikingly reduced number of bacteria in the lungs of MSC-treated animals. However, mean values of bacterial loads were decreased in MSC-treated animals at d2 and d5 p.i. but these differences were statistically not significant due to the inter-individual variation. One may speculate that the antibacterial activity of MSC may play a key role in models with live bacteria. With respect to the immunomodulatory activity of MSC, several studies using different models have reported that MSC induced increased levels of IL-10 or Foxp3+ regulatory T cells [
18,
36,
37]. However, in our study both IL-10 production as well as Foxp3+ CD4+ regulatory T cell frequencies were unaffected after MSC-therapy. Recently, Bustos et al. reported that pre-activation of human MSC with serum from patients with ARDS markedly increased their therapeutic capacity in a murine pneumonia model concomitant with increased IL-10 and IL-1 receptor antagonist levels [
38]. Therefore, different levels of MSC pre-activation in different animal models may impact on their immunomodulatory capacity. With respect to the immunosuppressive capacity of MSC we observed significantly reduced CD86 surface expression on lung CD103+ DC in MSC-treated animals early after infection. CD86 represents a major T cell costimulatory molecule that is critically involved in T cell activation [
39]. However, CD86 suppression was not significantly suppressed on CD11b + DC which may be related to different costimulatory molecule expression kinetics or differential migratory kinetics of CD11b + DC into inflamed tissue areas.
In summary, we have investigated the in vivo immunomodulatory capacity of prospectively defined, purified PαS MSC in a clinically relevant K. pneumoniae model. PαS MSC efficiently suppressed acute lung injury and promoted overall pneumonia survival. Flow cytometry analysis revealed impaired lung DC infiltration and CD103+ DC maturation representing known key drivers of T cell-mediated inflammation, which provided a rationale for the sustained anti-inflammatory effects of MSC therapy. These findings support the potential of using prospectively defined MSC-based therapies for patients with acute bacterial pneumonia and provide new insight into the immunomodulatory capacity of PαS cells.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
HH designed the research, analyzed and interpreted data and wrote the manuscript; AL, PK, IS, SK, KD, AG and NB performed research, analyzed and interpreted data; MH, GB and CB interpreted data and critically read the manuscript. All authors read and approved the final manuscript.