The immunomodulatory effects of human mesenchymal stem cells on peripheral blood mononuclear cells in ALS patients

Authors

  • Min-Soo Kwon,

    1. Department of Pharmacology, School of Medicine, CHA University, Bundang-gu, Seongnam-si, Gyeonggi-do, Korea
    Search for more papers by this author
    • These authors contributed equally to this work.
  • Min-Young Noh,

    1. Cell Therapy Center and Department of Neurology, College of Medicine, Hanyang University, Haengdang-dong, Seoul, Korea
    Search for more papers by this author
    • These authors contributed equally to this work.
  • Ki-Wook Oh,

    1. Cell Therapy Center and Department of Neurology, College of Medicine, Hanyang University, Haengdang-dong, Seoul, Korea
    Search for more papers by this author
  • Kyung-Ah Cho,

    1. Cell Therapy Center and Department of Neurology, College of Medicine, Hanyang University, Haengdang-dong, Seoul, Korea
    Search for more papers by this author
  • Byung-Yong Kang,

    1. Cell Therapy Center and Department of Neurology, College of Medicine, Hanyang University, Haengdang-dong, Seoul, Korea
    Search for more papers by this author
  • Kyung-Suk Kim,

    1. Bioengineering Institute, CoreStem Inc., Seoul, Korea
    Search for more papers by this author
  • Young-Seo Kim,

    1. Cell Therapy Center and Department of Neurology, College of Medicine, Hanyang University, Haengdang-dong, Seoul, Korea
    Search for more papers by this author
  • Seung H. Kim

    Corresponding author
    1. Cell Therapy Center and Department of Neurology, College of Medicine, Hanyang University, Haengdang-dong, Seoul, Korea
    • Address correspondence and reprint requests to Dr Seung H Kim, Cell Therapy Center, Department of Neurology, College of Medicine, Hanyang University, Haengdang-dong, Seongdong-gu, Seoul 133-791, Korea. E-mail: kimsh1@hanyang.ac.kr

    Search for more papers by this author

Abstract

In a previous study, we reported that intrathecal injection of mesenchymal stem cells (MSCs) slowed disease progression in G93A mutant superoxide dismutase1 transgenic mice. In this study, we found that intrathecal MSC administration vastly increased the infiltration of peripheral immune cells into the spinal cord of Amyotrophic lateral sclerosis (ALS) mice (G93A mutant superoxide dismutase1 transgenic). Thus, we investigated the immunomodulatory effect of MSCs on peripheral blood mononuclear cells (PBMCs) in ALS patients, focusing on regulatory T lymphocytes (Treg; CD4+/CD25high/FoxP3+) and the mRNA expression of several cytokines (IFN-γ, TNF-α, IL-17, IL-4, IL-10, IL-13, and TGF-β). Peripheral blood samples were obtained from nine healthy controls (HC) and sixteen patients who were diagnosed with definite or probable ALS. Isolated PBMCs from the blood samples of all subjects were co-cultured with MSCs for 24 or 72 h. Based on a fluorescence-activated cell sorting analysis, we found that co-culture with MSCs increased the Treg/total T-lymphocyte ratio in the PBMCs from both groups according to the co-culture duration. Co-culture of PBMCs with MSCs for 24 h led to elevated mRNA levels of IFN-γ and IL-10 in the PBMCs from both groups. However, after co-culturing for 72 h, although the IFN-γ mRNA level had returned to the basal level in co-cultured HC PBMCs, the IFN-γ mRNA level in co-cultured ALS PBMCs remained elevated. Additionally, the levels of IL-4 and TGF-β were markedly elevated, along with Gata3 mRNA, a Th2 transcription factor mRNA, in both HC and ALS PBMCs co-cultured for 72 h. The elevated expression of these cytokines in the co-culture supernatant was confirmed via ELISA. Furthermore, we found that the increased mRNA level of indoleamine 2,3-dioxygenase (IDO) in the co-cultured MSCs was correlated with the increase in Treg induction. These findings of Treg induction and increased anti-inflammatory cytokine expression in co-cultured ALS PBMCs provide indirect evidence that MSCs may play a role in the immunomodulation of inflammatory responses when MSC therapy is targeted to ALS patients.

image

We propose the following mechanism for the effect of mesenchymal stem cells (MSCs) administered intrathecally in amyotrophic lateral sclerosis (ALS): MSCs increase infiltration of peripheral immune cells into CNS and skew the infiltrated immune cells toward regulatory T lymphocytes (Treg) and Th2 lymphocytes. Treg and Th2 secret anti-inflammatory cytokines such as IL-4, IL-10, and TGF-β. A series of immunomodulatory mechanism provides a new strategy for ALS treatment.

Abbreviations used
FACS

fluorescence-activated cell sorting

FBS

fetal bovine serum

HC

healthy controls

IDO

indoleamine 2,3-dioxygenase

MSCs

mesenchymal stem cells

PBMCs

peripheral blood mononuclear cells

PBS

phosphate-buffered saline

PE

phycoerythrin

qPCR

quantitative polymerase chain reaction

qRT

quantitative reverse transcription

Amyotrophic lateral sclerosis (ALS), also known as Lou Gehrig's disease, is the most fatal progressive motor neuron disease, and it is characterized by the selective and progressive loss of motoneurons of the primary motor cortex, the brainstem and the spinal cord, consequently resulting in muscle weakness, wasting, and ultimately death (Chio et al. 2013; Van Damme and Robberecht 2013). Recent studies have focused on the roles of innate and adaptive peripheral immune cells, which play pivotal and interdependent roles in regulating the rate of disease progression in ALS mice (Appel et al. 2010; Schwartz and Shechter 2010; Seksenyan et al. 2010; Beers et al. 2011). Additionally, it has been reported that T lymphocytes are able to infiltrate into the spinal cord in ALS patients (Engelhardt et al. 1993; Appel et al. 2010; Beers et al. 2011). Among immune cells, it has been demonstrated that CD4+ T lymphocytes slowed disease progression with a 50% increase in disease duration, modified microglial phenotypes and extended the survival of ALS mice (Beers et al. 2008, 2011; Chiu et al. 2008; Henkel et al. 2013). Moreover, CD4+/CD25high/forkhead box P3 (FoxP3)+ regulatory T lymphocytes (Treg), which are up-regulated during the stable disease phase in ALS mice, may be associated with the slowing of ALS progression (Beers et al. 2011; Henkel et al. 2013). Treg cells are involved in both the unbalanced innate and adaptive immune responses that are detrimental to the host, down-regulate pro-inflammatory cytokine production and suppress the proliferation and activation of CD4+/CD25 effector T-lymphocytes (Sakaguchi et al. 2010). In addition, it has been demonstrated that Treg cells are able to induce the differentiation of macrophages/microglia toward the alternative, protective M2 phenotype (Beers et al. 2011; Henkel et al. 2013).

Although ALS is a heterogeneous disease, Treg cells appear to be an attractive factor and are involved in slowing the rate of disease progression in ALS patients, as neuroinflammation is a prominent pathological feature in ALS (Henkel et al. 2013). Thus, up-regulating Treg activity or increasing the Treg ratio in the immune system of ALS patients may represent a potential therapeutic strategy for ALS. However, it is currently difficult to directly increase the number of these cells via, for example, the passive transfusion of Treg cells, because Treg cells can be transformed into other T lymphocytes and can be inactivated in vivo according to the microenvironment (Sakaguchi et al. 2010; Ohkura et al. 2011). Thus, the development of a novel molecule that can induce Treg formation in vivo appears to be a more effective and feasible solution.

Mesenchymal stem or stromal cells (MSCs) represent potential candidate factor that modulate immune-inflammatory function (Gebler et al. 2012). MSCs are multipotent non-hematopoietic progenitor cells that can be isolated and expanded from bone marrow and other tissues (Reyes et al. 2002), and these cells are capable of differentiating into multiple mesenchymal lineages (Woodbury et al. 2000; Reyes et al. 2002). MSCs have been therapeutically used because they secrete various neurotrophic factors that are required for tissue repair (Caplan and Dennis 2006; Koh et al. 2009). Based on these capacities of MSCs, several clinical trials using MSCs have been conducted to target diseases with unmet medical needs, including ALS, multiple sclerosis, and other degenerative diseases (Kim et al. 2009; Karussis et al. 2010; Mazzini et al. 2010; Lee et al. 2012). Recently, it was demonstrated that MSCs exert additional potential benefits that are expected to be responsible for the ability of MSCs to modulate the functions of immune cells (Gebler et al. 2012). One of these additional mechanisms is the induction of Treg formation, suggesting that MSCs may represent an effective therapeutic strategy for ALS treatment (Yagi et al. 2010; English and Mahon 2011; Singer and Caplan 2011).

However, the underlying mechanism by which MSCs affect ALS progression, especially with respect to the immune-inflammatory aspects of this disease, has yet to be clearly elucidated. In this study, we found that injection of MSCs into the cisterna magna vastly increased the infiltration of peripheral immune cells into the spinal cord of ALS mice [G93A mutant superoxide dismutase1 transgenic (SOD1)]. In addition, intrathecal MSC injection slowed disease progression in SOD1 mice (Kim et al. 2010) and ALS patients (Kim et al. 2009). Considering that immune cells are associated with disease progression in ALS patients and that MSCs may exert beneficial effects via Treg induction and anti-inflammatory cytokines in ALS patients, it is important to confirm whether MSCs actually induce Treg formation and immunomodulatory responses in peripheral blood mononuclear cells (PBMCs) of ALS patients because, to date, there are no data supporting this mechanism in ALS patients. Thus, to reconstruct the environment in which infiltrated immune cells interact with MSCs administered intrathecally, we co-cultured PBMCs obtained from ALS patients with MSCs to elucidate the effect of MSCs on the levels of Treg cells and cytokines according to the co-culture duration. Moreover, this interaction was compared with PBMCs from healthy controls (HCs). In addition, the correlation between the secretion of soluble factors by MSCs and the increase in Treg cells was analyzed to identify markers that indicate the capacity of MSCs to induce Treg formation in ALS PBMCs.

Materials and methods

Isolation, culture, and identification of human MSCs

Ethical approval for using MSCs was obtained from Hanyang University Hospital in Seoul, Korea (HYUH2012-05-009). The remaining MSCs after the procedure of allogeneic bone marrow-derived MSC transplantation into the ALS patients (ClinicalTrials.gov, NCT01758510) were used. The patient's CSF was collected via lumbar spinal puncture when the MSCs were collected. After using the CSF for the clinical application, the remaining CSF was stored in a deep freezer at −70°C and was used after thawing. To obtain the MSCs, mononuclear cells were isolated via aspiration of the bone marrow at the iliac crest, followed by enrichment using a density gradient (Histopaque, density 1.077 g/mL; Sigma-Aldrich, St. Louis, MO, USA) and two washes with Dulbecco's modified Eagle's medium containing low glucose concentration (DMEM-LG; GIBCO BRL, Grand Island, NY, USA). The cells were cultured at a density of 2 × 105 cells/cm2 in DMEM-LG supplemented with 10% fetal bovine serum (FBS; Hyclone, Waltham, MA, USA) and were grown at 37°C in 5% CO2 for 72 h. After removing the non-adherent cells, the culture medium was changed twice per week. For each passage, the cells were detached using 0.25% trypsin/ethylenediaminetetraacetic acid (EDTA) for 3 min at 37°C; the cells were then seeded at a density of 4 × 103 cells/cm2 and expanded to 80–90% confluence. The MSCs were used at their third passage.

To determine the mesenchymal phenotypes (Koh et al. 2012) of the MSCs, we used monoclonal antibodies against Human leukocyte antigen, CD49c, CD73, CD105, CD34, CD45, CD29, CD44, CD106, and CD54 (BD Biosciences Pharmingen, San Diego, CA, USA). Briefly, the MSCs at their third passage were immunostained in phosphate-buffered saline (PBS; Ca2+-and Mg2+-free) supplemented with 5% FBS (Hyclone). After the final wash, the cells were fixed in 1% paraformaldehyde prior to analysis using the fluorescence-activated cell sorting (FACS) FlowJo software (BD Biosciences, San Jose, CA, USA); phycoerythrin (PE)-labeled mouse anti-human immunoglobulin was used as the isotype control. To control for non-specific binding, the same fluorochrome–protein ratio was used for the isotype control. The MSC phenotype was confirmed via flow cytometry of the MSCs after co-culturing for 24 or 72 h with the PBMCs from the HCs (Fig. 1).

Figure 1.

Phenotypic characterization of mesenchymal stem cells (MSCs) alone and co-cultured MSCs. After isolating the MSCs, the MSCs were cultured alone or co-cultured with ALS peripheral blood mononuclear cells (PBMCS) for 24 or 72 h. The MSCs were analyzed to determine the phenotypes using monoclonal antibodies against Human leukocyte antigen, CD49c, CD73, CD105, CD34, CD45, CD29, CD44, CD106, and CD54. The MSCs alone expressed typical MSC phenotype markers. Almost all of the co-cultured MSCs expressed ICAM-1 (~ 100%), but MSCs alone did not express ICAM-1.

Intrathecal human MSC injection into SOD1 mice

The animal experiments were performed according to the Hanyang University guidelines for the care and use of laboratory animals and were approved by the Institutional Animal Care and Use Committee of Hanyang University (HY-IACUC-13-063).

All mice [B6SJL-Tg (SOD1-G93A) 1Gur/J over-expressing human SOD1 containing the Gly93 → Ala mutation] were obtained from the Jackson Immuno-Research, West Grove, PA, USA. The mice were identified according to the genotyping protocol of the Jackson Laboratory, in which the transgene copy numbers were evaluated via real-time quantitative polymerase chain reaction (qPCR). We used 24 SOD1 (male: 12; female: 12) mice in this study. The SOD1 mice were age- and sex-matched and were assigned to one of two groups: the CSF group (treated with 10 μL of CSF, n = 12) and the MSC group, which received a dose of cells 1 × 106 (n = 12). The cells were suspended in CSF at cellular density of 1 × 106, and the volume of the cell suspension that was administered to each group was 10 μL. The maximum cellular density was 1 × 105 cells in 1 μL of CSF, and in 10 μL of CSF, the volume was 1 × 106 cells that was transplanted into the cisterna magna of the mice, as determined in our previous study (Kim et al. 2010).

The CSF or the MSCs were administered twice to the SOD1 mice. At 60 and 67 days after birth, the mice were anesthetized using tiletamine–zolazepam (60–80 mg/kg, i.p.) and xylazine (5–10 mg/kg, i.p.) and were positioned in a stereotaxic apparatus (ASI Instruments, Heidelberg, Germany) such that the cisterna magna was the highest point. Under an operating microscope, the atlanto-occipital membrane was exposed, and 10 μL of CSF or MSCs was injected using a Hamilton syringe (25 μL, 31G) and an injection pump (1 μL/min). One drop of saline was placed on the top of the injection point to avoid cell leakage (Kim et al. 2010). The needle was withdrawn after 10 min, and the incision was closed. All animals that underwent this operation were injected with cyclosporine (10 mg/kg, i.p., per day) from the day prior to transplantation to the end point of the experiment. At 74 days, all mice were perfused, and the lumbar segment of the spinal cord was dissected for qPCR analysis (n = 6 per group) and immunohistochemistry (n = 6 per group). The qPCR method is described below, and primers specific for CD3 (F-TGCTCTTGGTGTATATCTCATTGC; R-CAGAGTCTGCTTGTCTGAAGCTC) and CD45 (F-TCATGGTCACACGATGTGAAGA; R-AGCCCGAGTGCCTTCCT) were obtained (Bionics, Seoul, Korea) and used.

Immunohistochemical analysis

The animals were killed via the inhalation of an overdose of isoflurane. The animals were then transcardially perfused with saline followed by 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). The spinal cord was carefully dissected, post-fixed for 24 h in the same solution and cryoprotected overnight in 30% sucrose solution in 0.1 M phosphate buffer. The tissue was frozen in cryostat medium and sliced into coronal 20-μm-thick serial sections. Of the slides described above, every fifth slide was processed for immunohistochemistry. The selected slide was rinsed three times in 0.1 m PBS for 5 min. To block non-specific immunoglobulin G binding, peroxidase activity in the sections on the slide was blocked with 3% hydrogen peroxide for 15 min at 20–23.5°C, and the slides were rinsed three times in PBS. In addition, the proteins were blocked with 10% normal goat serum for 30 min at 20–23.5°C. After the blocking procedures, the slides were incubated in the anti-CD3 antibody (ab5690, 1 : 100; Abcam, Cambridge, MA, USA) for 24 h at 4°C. The sections were then washed three times for 5 min each to remove unbound primary antibodies, followed by incubation in the appropriate secondary biotin-conjugated antibody for 2 h at 20–23.5°C. Unbound secondary antibody was removed by rinsing three times for 5 min each. After air-drying, the slides were coverslipped using Vectashield (Vector Laboratories, Burlingame, CA, USA). The CD3-positive cells in the ventral horn of the spinal cord were identified under a microscope (Olympus Imaging America Inc., Center Valley, PA, USA). As a negative control, the above procedures were repeated in the absence of primary antibody, and there were no stained cells in the negative controls.

Isolation of human PBMCs and co-culture of PBMCs with MSCs

Ethical approval for the collection of peripheral blood was obtained from Hanyang University Hospital in Seoul, Korea (HYUH 2013-06-012-002). A total of 16 ALS patients (eight males, eight females; Mage: 52 ± 10 years) diagnosed with sporadic ALS according to the revised El Escorial criteria (Brooks, 1994) and 9 HC volunteers (five males, four females; Mage: 45 ± 6 years) were recruited, and blood samples were collected after obtaining written informed consent. Consenting ALS patients who fulfilled the following three criteria were included in this study: (i) the patients satisfied the revised El Escorial criteria for definite or probable ALS, (ii) the patients were male or non-pregnant females between 20 and 65 years of age, and (iii) the patients displayed a revised ALS Functional Rating Scale score higher than 20 at the time of blood sampling. Thirty percent of the ALS patients were treated with riluzole. None of the ALS patients or HCs received immunosuppressant therapy or had a previous history of infectious disease, cancer, or immune disorder. The participants' information is presented in Table 1. After collecting peripheral blood from the HC and ALS patients, the PBMCs were isolated immediately for co-culture with MSCs using the Histopaque density gradient method with Histopaque®-1077 (Sigma, St. Louis, MO, USA).

Table 1. Demographic and clinical characteristics of ALS patients
 SexAgeAge of onsetALSFRS-RDurationdelta-FRSSite of onsetRiluzole
  1. (+)Riluzole: stably administrated over 3 months.

  2. ALSFRS-R: score at sampling time.

  3. delta-FRS: progression rate.

ALS1F59584171Bulbar onsetX
ALS2M49484360.83Limb onsetX
ALS3F211946200.1Limb onset0
ALS4F424046140.14Limb onsetX
ALS5F635940140.57Limb onsetO
ALS6M555341200.35Limb onsetX
ALS7M595643250.2Limb onsetO
ALS8M535141150.47Limb onsetO
ALS9M454341150.47Limb onsetX
ALS10M615938130.77Limb onsetX
ALS11M626043140.36Limb onsetX
ALS12M52514180.88Limb onsetX
ALS13F605844150.27Limb onsetX
ALS14F565443140.36Limb onsetX
ALS15F51504042Limb onsetO
ALS16F575546160.13Limb onsetX
HC1M51  
HC2M44  
HC3M51  
HC4M38  
HC5F48  
HC6F48  
HC7F41  
HC8F52  
HC9M36  

Human MSCs (5 × 104 cells/well) were seeded on 24-well plates (Nunc Multiwell plates; Sigma) 24 h prior to co-culture with PBMCs and were cultured in DMEM-LG supplemented with 1% penicillin–streptomycin and 10% FBS (Gibco). The isolated PBMCs (5 × 105 cells/well) were co-cultured in the 24-well culture plates with the pre-seeded MSCs. After co-culture for 24 or 72 h, the supernatant was collected for ELISA analysis. The PBMCs were separated via pipetting for FACS and quantitative reverse transcription (qRT)-polymerase chain reaction analyses. After the PBMCs were separated, the MSCs were detached from the substrate using trypsin–EDTA, washed and resuspended in PBS containing 1% (v/v) bovine serum albumin for FACS and qRT-PCR analyses.

Flow cytometry analysis

After co-culturing for 24 or 72 h, the PBMCs were immunolabeled using antibodies against surface proteins CD4, CD25, and intracellular protein FoxP3. Briefly, cells were incubated in CD4-FITC and CD25-Allophycocyanin or isotype-matched control antibodies (BD Pharmingen) for 20 min at 20-23.5°C in the dark; the cells were then washed with 1 mL staining buffer (BD Pharmingen) and fixed with the fixation buffer that was supplied in the FoxP3 kit (BD Pharmingen) for 10 min at 20-23.5°C in the dark. The cells were permeabilized using permeabilization buffer (BD Pharmingen) for 1 h at 20–23.5°C and then washed with 1 mL of staining buffer. After washing, the cells were incubated in PE-conjugated anti-human FoxP3 (BD Pharmingen) or the isotype control for at least 30 min at 20–23.5°C in the dark. At the end of the incubation period, the PBMCs were washed and resuspended in staining buffer. All data were collected on a FACSCanto II flow cytometer (BD Biosciences Pharmingen) and analyzed with FACSDiva (BD, NJ, USA) or FlowJo software (TressStar Inc., Ashland, OR, USA). The MSCs were also analyzed via flow cytometry using the method described above (identification of MSCs) to examine the alteration of the MSC phenotype after co-culturing with PBMCs for 24 or 72 h. Treg cells were defined as CD4+/CD25high/FoxP3+ cells, as previously described (Zahran et al. 2012).

qRT-PCR

The MSCs and the PBMCs were separated after co-culturing for 24 or 72 h to quantify the gene expression of each cell type. Total RNA was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA), and 3–5 μg of RNA was reverse transcribed using RevertAid M-MuLV reverse transcriptase (MBI Fermentas, Burlington, ON, Canada), 0.2 μg of random primer (Invitrogen), 1 mM dNTPs (Invitrogen), and the supplied buffer. First-strand cDNA was amplified using the Power SYBR Green PCR master mix (eBioscience, Hatfield, UK) and primers specific for FoxP3 (PPH00029C; Qiagen, Hilden, Germany), t-box transcription factor (Tbx21, PPH00396A; Qiagen), Gata3 (PPH02143A; Qiagen), IFN-γ (PPH00380C; Qiagen), TNF-α (PPH00341F; Qiagen), IL-17 (PPH00537C; Qiagen), IL-4 (PPH00565B; Qiagen), IL-10 (PPH00572C; Qiagen), IL-13 (PPH00688F; Qiagen), TGF-β (PPH00508A; Qiagen), indoleamine 2,3-dioxygenase (IDO, PPH01328B; Qiagen), and β-actin (PPH00073G; Qiagen). Real-time qRT-PCR was conducted using the StepOnePlus system (Applied Biosystems, Carlsbad, CA, USA) with the following cycling parameters: 95°C for 10 min followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. After amplification, dissociation curves were constructed from 60 to 90°C, and the Ct values were converted to absolute amounts of cDNA (E-Ct). To correct for the differences in the cDNA amounts between the samples, each target PCR was normalized to the geometric mean value of the reference gene, β-actin.

ELISA

After co-culturing for 24 or 72 h, the culture supernatants were collected. The production of human IL-10, IFN-γ, and prostaglandin E2 (PGE2) in the 24-h co-culture and human IL-4 and TGF-β in the 72-h co-culture was analyzed using a Quantikine ELISA kit (R&D Systems, Minneapolis, MN, USA) according to the manufacturer's instructions.

Statistical analysis

Data are presented as the means ± SD. The statistical significance of the differences between the groups was assessed using Student's t-test and one way-anova with Bonferroni post hoc analysis. R2 values of linear regression analyses were assessed using GraphPad Prism Version 5.0 software for Mac OS X (GraphPad Software, San Diego, CA, USA); < 0.05 was considered significant.

Results

Phenotypic characterization of MSCs alone and co-cultured MSCs

The demographic and clinical characteristics of the 16 ALS patients are shown in Table 1. To evaluate the effect of the MSCs on Treg induction in the HC and ALS PBMCs, we used PBMCs from 9 HCs and 16 ALS patients. MSCs obtained from one healthy donor were used in the co-culture. The MSCs displayed the typical characteristics and morphology and were confirmed using known MSC markers (Fig. 1), and there was no difference in the number of MSCs alone between 24 and 72 h of culture. The differentiation capacity of the MSCs into osteocytes, chondrocytes, and adipocytes was also validated via differentiation assays, as performed in our previous study (Kim et al. 2013).

Intrathecal MSC injection increased infiltration of CD3- and CD45-positive cells into the lumbar spinal cord of SOD1 mice

Previous studies have demonstrated that MSCs could represent an alternative therapeutic strategy for intractable neurodegenerative diseases such as ALS because of their capacity to differentiate and secrete neurotrophic factors. Recently, it has been demonstrated that the immunomodulatory effect of MSCs plays an important role in their positive effect. In addition, we found that MSCs slowed disease progression in SOD1 mice (Kim et al. 2010). Thus, we hypothesized that intrathecal MSC administration may increase both the infiltration of peripheral immune cells into the spinal cord and the differentiation of these infiltrated immune cells to an anti-inflammatory phenotype. As shown in Fig. 2a, MSC injection increased the CD3 and CD45 levels in the lumbar spinal cord of SOD1 mice, suggesting that intrathecal MSC administration attracts peripheral immune cells to the spinal cord.

Figure 2.

Intrathecal mesenchymal stem cell (MSC) injection increased the CD3 and CD45 mRNA levels in the lumbar spinal cord; flow cytometry analysis of Treg and the change in the Treg/total lymphocyte ratio in peripheral blood mononuclear cells (PBMCs) co-cultured with MSCs. (a) The SOD1 mice were age- and sex-matched and were separated into two groups: the CSF group (treated with 10 μL of CSF, n = 12), and the MSCs group, which received a dose of 1 × 106 cells (n = 12). The cells were suspended in CSF at a cellular density of 1 × 106, and the volume of the cell suspension administered to each group was 10 μL. The CSF or MSCs were administered twice into the cisterna magna of the SOD1 mice at 60 and 67 days of age. At 74 days of age, all mice were perfused, and the lumbar segment of the spinal cord was dissected for qPCR analysis (n = 6 per group) or immunohistochemistry (n = 6 per group). The MSCs displayed increased CD3 and CD45 mRNA expression and immunoreactivity in the lumbar spinal cord. **p < 0.01 compared with the CSF group. (b) CD4+/CD25+high (hi)/FoxP3+ (Treg) lymphocytes were detected in PBMCs alone (upper panel) or in PBMCs co-cultured with MSCs (lower panel). The forward and side scatter histogram was used to define the lymphocyte population (P1). CD4+ lymphocytes were defined as P2. Three gates were drawn to define the CD4+/CD25 population (P3), the CD4+/CD25+intermediate(med) population (P4), and the CD4+/CD25+hi lymphocytes (P5) among the CD4+ lymphocytes. (P5) The CD4+/CD25+hi/FoxP3+ (Q2, Treg) lymphocytes were detected in the CD4+/CD25+hi lymphocytes using the FoxP3 phycoerythrin (PE) antibody (c and d). The healthy controls (HC) and ALS PBMCs were isolated from whole blood and were co-cultured with MSCs or cultured alone for 24 or 72 h. After 24 or 72 h, the PBMCs were separated from the MSCs for flow cytometry analysis (c) and qRT-PCR (d) analyses. Co-culture with MSCs increased the Treg/total lymphocyte ratio in 24-h co-cultures, and this increase was further enhanced in 72-h co-cultures (c). There was no difference in the fold-change in the Treg/total lymphocyte ratio between the HC and ALS groups. Co-culture with MSCs elevated the FoxP3 mRNA level in both groups (d). The relative quantification (RQ) levels were calculated using the ▵▵Ct method. $p < 0.05, $$$p < 0.001 compared with HC PBMCs alone; #p < 0.05, ##p < 0.05, ###p < 0.001 compared with ALS PBMCs alone; *p < 0.05, ***p < 0.001 compared between the two groups.

MSCs increased the Treg/total lymphocyte ratio in the co-cultured PBMCs

To evaluate the effect of MSCs on the Treg/total lymphocyte ratio in the PBMCs, the PBMCs were co-cultured with MSCs for 24 or 72 h, and the Treg were detected based on CD4+/CD25high/FoxP3 immunolabeling (Fig. 2). As shown in Fig. 2b and c, 24-h co-culture with the MSCs increased the Treg/total T-lymphocyte ratio, which can be considered Treg induction in the HC and ALS PBMCs (Increment, HC: 9.6 ± 8.3; ALS: 5.8 ± 3.0). The increase in the Treg/total lymphocyte ratio was further enhanced in the PBMCs co-cultured for 72 h (Increment, HC: 15.7 ± 31.7; ALS: 11.7 ± 15.1). Moreover, as shown in Fig. 2d, co-culture with MSCs elevated the level of FoxP3 mRNA, which is currently the most reliable marker for identifying Treg cells, in a pattern similar to the change in Treg formation in the co-culture. Furthermore, although Treg/total T-lymphocyte ratio is further elevated in 72-h co-culture, we found that HC FoxP3 mRNA level was returned to basal level and ALS FoxP3 mRNA showed a trend to be reduced when compared with PBMCs only. The basal level of Treg cells between the HCs and the ALS patients was not significant, in agreement with a previous study (Henkel et al. 2013).

MSCs induced dynamic changes in the cytokine levels of PBMCs when co-cultured for 24 and 72 h

MSCs induce PBMC activation via various pathways, such as donor Allophycocyanin migration, donor cell-derived protein transfer, donor Major histocompatibility complex/peptide transfer, and direct cell–cell contact (Griffin et al. 2013). Activated PBMCs can alter the mRNA expression levels of cytokines, resulting in the release of cytokines that are involved in the immunomodulatory effect of MSCs. Therefore, the changes in the mRNA levels of T-lymphocyte-associated cytokines were measured in the PBMCs after co-culture with MSCs. As shown in Fig. 3, the IFN-γ mRNA levels were increased in the HC and ALS PBMCs when examined after co-culturing for 24 h. However, after co-culturing for 72 h, the IFN-γ mRNA level in the HC PBMCs had returned to the basal level, whereas that in the ALS PBMCs remained elevated. Indeed, we found that the mRNA expression of IL-10, one of the major anti-inflammatory cytokines in Treg cells, was very strongly increased in the HC and ALS PBMCs, and this increase returned to the basal level in both groups. The mRNA levels of IL-4 and TGF-β in both groups were increased only after co-culturing for 72 h. Based on a previous study that found an association between the Th1–Th2 lymphocyte balance and disease progression in ALS mice (Beers et al. 2011), we examined the alterations in the levels of Th1- and Th2-specific transcription factors to evaluate the effect of MSCs on the proliferation of Th1 and Th2 lymphocytes (Zhu and Paul 2010). The mRNA expression of Gata3, the master transcription factor preferentially expressed by Th2 lymphocytes, was increased only after co-culturing for 72 h, whereas that of Tbx21, the master transcription factor preferentially expressed by Th1 lymphocytes, was not changed by co-culturing with MSCs. The mean TNF-α, IL-13, or IL-17 mRNA value was not elevated by co-culturing with MSCs although several samples showed the increased pattern (Supporting Information).

Figure 3.

Changes in the T-lymphocyte subpopulations and the cytokine levels in co-cultured peripheral blood mononuclear cells (PBMCs). Healthy controls (HC) and ALS PBMCs were isolated from whole blood and were cultured alone or co-cultured with mesenchymal stem cells (MSCs) for 24 or 72 h. The PBMCs were separated for qRT-PCR analysis after co-culturing. The co-culture supernatant was used for ELISA. The IFN-γ (a) levels were increased in the HC and ALS PBMCs when co-cultured with MSCs for 24 h. Furthermore, after co-culturing for 72 h, this elevated level persisted in co-cultured ALS, but not HC PBMCs. IL-10 mRNA expression (b) in both the co-cultured HC and ALS PBMCs was increased after co-culturing for 24 h and returned to the basal level after co-culturing for 72 h. IL-4 (c) and TGF-β (d) mRNA expression was increased in 72-h co-cultured HC and ALS PBMCs, but not in the 24-h co-cultures. Although Tbx21 mRNA expression (e) was unchanged, Gata3 mRNA expression (f) was elevated in 72-h co-cultured PBMCs of both the HC and ALS groups. The RQ levels were calculated using the ▵▵Ct method. $p < 0.05, $$$p < 0.001 compared with HC PBMCs alone; #p < 0.05, ##p < 0.01, ###p < 0.001 compared with ALS PBMCs alone; *p < 0.05 compared between the two groups.

As shown in Fig. 4, the concentrations of IFN-γ and IL-10 were quantified in the supernatant of the 24-h co-culture, and elevations in IFN-γ (HC: 14.8 ± 12.8; ALS: 14.5 ± 22.4) and IL-10 (HC: 587 ± 304; ALS: 510 ± 414) were detected compared with PBMCs that were cultured alone (Fig. 4a and b). The concentrations of IL-4 and TGF-β were quantified in the supernatant of the 72-h co-culture, and the elevation in IL-4 level (HC: 2.5 ± 1.19, ALS: 3.1 ± 2.0) was detected compared with PBMCs that were cultured alone (Fig. 4c). TGF-β (HC: 1444 ± 442; ALS: 1707 ± 648) was also increased compared with PBMCs alone or MSCs alone (Fig. 4d). There was no difference in any cytokine level between the HC and ALS PBMCs.

Figure 4.

The change in the cytokine levels in the co-cultured supernatant. The healthy controls (HC) and ALS peripheral blood mononuclear cells (PBMCs) were isolated from whole blood and were cultured alone or co-cultured with mesenchymal stem cells (MSCs) for 24 (a and b) or 72 h (c and d). The PBMCs were separated after co-culture, and the co-culture supernatant was used for ELISA. IFN-γ (a), IL-10 (b), IL-4 (c), and TGF-β (d) were increased in the supernatant of both groups. The concentrations of IFN-γ and TGF-β in the PBMCs alone were < 5 pg/mL. IL-10 and IL-4 were not detected in the PBMCs alone. TGF-β was ~ 750 pg/mL in MSCs alone cultured for 72 h. *p < 0.05, **p < 0.01, ***p < 0.001 compared with PBMCs alone; +p < 0.05, +++p < 0.001 compared between two groups.

The fold-change in IDO mRNA expression in co-cultured MSCs correlated with Treg induction in PBMCs

MSCs can induce Treg by cell contact and soluble mediators. It has been well demonstrated that IDO, TGF-β, and PGE2 are representative soluble mediators of Treg induction and that these soluble mediators exert immunomodulatory effects directly and indirectly via Treg induction (English et al. 2009; Shi et al. 2012). Thus, the changes in the mRNA levels of IDO and TGF-β were investigated in the MSCs co-cultured with the HC or ALS PBMCs (co-MSCs). The protein level of PGE2 was examined instead of the mRNA level because PGE2 is synthesized in the cell from essential fatty acids (Harris et al. 2002). As shown in Fig. 5, IDO mRNA expression was increased in 24-h co-MSCs (HC: 5.7 ± 3.4; ALS: 2.9 ± 3.2) and was further enhanced in 72-h co-MSCs (HC: 50.5 ± 51.5; ALS: 37.8 ± 30.1). TGF-β mRNA expression was increased only in 24-h co-MSCs cultured with HC, but not ALS, PBMCs. The increase in the TGF-β mRNA level in the co-MSCs cultured with HC PBMCs was larger than that of the co-MSCs cultured with ALS PBMCs (HC: 2.5 ± 2.9; ALS: 0.8 ± 0.5; p = 0.0358). However, after co-culturing for 72 h, there was no difference in the expression level of TGF-β mRNA between the groups, which displayed elevated expression compared with MSCs alone (HC: 2.4 ± 1.2; ALS: 2.1 ± 1.2). In addition, the PGE2 protein level (HC: 2647 ± 27.4; ALS: 2626 ± 56.3) was increased in the co-cultured supernatant compared with that of MSCs alone.

Figure 5.

Change in the mRNA levels of indoleamine 2,3-dioxygenase (IDO), TGF-β, PGE2 in 24 and 72-h co-cultured mesenchymal stem cells (MSCs), and the correlation between these mediators and Treg induction in peripheral blood mononuclear cells (PBMCs) (Upper panel). The healthy controls (HC) and ALS PBMCs were isolated from whole blood and were co-cultured with MSCs for 24 or 72 h. The co-cultured MSCs were separated for qRT-PCR analysis. The co-cultured supernatant was used for ELISA. IDO mRNA (a) and PGE2 protein (c) were increased in 24-h co-cultures of both HC and ALS PBMCs. IDO mRNA expression was further enhanced in 72-h co-cultures compared with 24-h co-cultures. In the 24-h co-cultures, TGF-β mRNA expression (b) was only increased in the co-cultured HC, but not ALS, PBMCs. However, in the 72-h co-cultures, TGF-β mRNA expression in co-MSCs cultured with ALS PBMCs was elevated compared with that in 24-h co-cultures, and there was no significant difference in the TGF-β mRNA expression levels of the two groups. The mRNA levels were normalized to that of MSCs alone cultured for 24 and 72 h (dashed line). There was no difference in the cytokine levels in MSCs alone between 24 and 72 h. The RQ levels were calculated using the ▵▵Ct method. +p < 0.05, +++p < 0.001 compared with MSCs alone; *p < 0.05 compared between the two groups; the data are presented as the means (Lower panel). We analyzed the correlation between the fold-change in the IDO mRNA (d), TGF-β mRNA (e), and PGE2 protein levels (f) and the fold-change in the Treg/total lymphocyte ratio. The red circle and line represent the ALS group. The fold-change in IDO mRNA expression strongly correlated with the fold-change in the Treg/total lymphocyte ratio in ALS PBMCs (R2 = 0.6058).

To investigate the correlation between the candidate markers and Treg induction and identify the factor representing the Treg induction capacity of MSCs, we analyzed the R square values between the level of IDO mRNA, TGF-β mRNA, or PGE2 protein in co-MSCs and the increase in Treg in ALS PBMCs. The fold-change in IDO mRNA displayed a high correlation with the increase in Treg in the co-MSCs cultured with ALS PBMCs (R2 = 0.6058, red line). However, the fold-change in the TGF-β mRNA and PGE2 protein was not correlated with the increase in Treg.

Discussion

In this study, we found for the first time that MSCs: (i) increase the infiltration of peripheral immune cells into the spinal cord of SOD1 mice and (ii) induce an increase in the Treg/total lymphocyte ratio and cause dynamic alterations in the anti-inflammatory cytokine mRNA levels in ALS PBMCs. Additionally, MSCs induced similar patterns of cytokine expression in both HC and ALS PBMCs, except the mRNA level of IFN-γ when co-cultured with MSCs for 72 h. Furthermore, we suggest that the IDO mRNA level in co-MSCs may be an indicator of the capacity of Treg induction in ALS PBMCs. Our results demonstrate an effect of MSCs on Treg induction because of the interplay between MSCs and PBMCs and suggest that MSCs might play a role in slowing the progression of ALS via the immunomodulation of inflammation, in agreement with our ALS mouse data (Kim et al. 2010) and clinical trial results (Kim et al. 2009).

MSCs increased Treg induction in ALS PBMCs, and this increase was further enhanced after co-culturing for 72 h. It is well known that several soluble factors, such as IDO, TGF-β, and PGE2, are involved in the mechanism of Treg induction by MSCs. The TGF-β and PGE2 are secreted constitutively in MSCs and can induce Treg directly (English et al. 2009). Among these factors, IDO, a rate-limiting enzyme associated with the catabolism of the essential amino acid tryptophan, is a representative soluble factor involved in Treg induction (English et al. 2009; Shi et al. 2012; English 2013). It has been demonstrated that local reduction in the tryptophan concentration by IDO results in changes in immune cell populations (Meisel et al. 2004). However, MSCs do not constitutively express IDO mRNA (Shi et al. 2012). Human MSCs express IDO mRNA when they are stimulated by IFN-γ while murine MSCs express nitric oxide by IFN-γ and TNF-α (Meisel et al. 2004; Krampera et al. 2006; Sheng et al. 2008; Bernardo and Fibbe 2013). In this study, we found that the MSC phenotype was transformed into an activated-like form (~ 100% ICAM-1 expression in the MSCs, Fig. 1), and the IDO mRNA expression levSPELL CHECKel was increased along with the IFN-γ elevation in co-culture (Shi et al. 2012). It has been reported that the inflammatory cytokines produced by immune cells modulate MSC function, leading to the release of immunosuppressive factors, the altered expression of surface molecules and the production of growth factors. Thus, ICAM-1 expression may represent a marker of MSC activation (Shi et al. 2012). Taken together, it can be speculated that although we cannot exactly confirm which mechanism is involved in IFN-γ, MSCs can increase IFN-γ mRNA expression in PBMCs via various pathways, such as MSC-derived protein transfer, Major histocompatibility complex/peptide transfer of MSCs, and direct MSC-PBMC contact (Griffin et al. 2013), and that IFN-γ secreted by PBMCs leads to the elevation of IDO mRNA expression in MSCs. In addition, it has been reported recently that IFN-γ plays a pivotal role in CNS immune surveillance and repair by regulating the choroid plexus in the brain, which is a gate that permits T lymphocytes and M2 macrophages to enter the CNS (Baruch and Schwartz 2013; Kunis et al. 2013; Shechter et al. 2013b). Thus, it can be speculated that IFN-γ expression, via interplay between MSCs and PBMCs, might contribute to elevating the immunomodulatory activity of MSCs as well as Treg infiltration into the spinal cord with direct Treg induction by TGF-β and PGE2 (English et al. 2009). However, we cannot confirm whether the IFN-γ concentration in the co-culture supernatant exerts a detrimental or positive effect in ALS patients because IFN-γ plays an important role in determining the phenotype and function of MSCs, depending on the narrow concentration window in the microenvironment (Chan et al. 2006). Thus, further studies are required to strengthen this hypothesis.

The elevation of the IFN-γ mRNA levels was maintained in the 72-h co-cultured ALS but not HC, PBMCs while TNF-α and IL-17 mRNA were not changed in both PBMCs. These differential responses between the HC and ALS groups might be due to an exaggerated response of the ALS PBMCs to MSC-mediated stimuli. It has been demonstrated that systemic immunological activation may be associated with ALS progression (Alexianu et al. 2001; Beers et al. 2011; Rentzos et al. 2012; Zhao et al. 2012; Henkel et al. 2013). In addition, based on microarray analysis, Lipopolysaccaride/Toll-like receptor 4 signaling-associated genes, such as genes encoding type I interferons, interferon-inducible genes, and genes associated with secreted activation proteins, were up-regulated in ALS PBMCs (Zhang et al. 2011). In addition, the changes in the profiles of up-regulated genes and proteins may be associated with a possible intracellular pathogenic mechanism of ALS PBMCs (Nardo et al. 2011). Thus, it can be speculated that the prolonged elevation of the IFN-γ mRNA in ALS PBMCs may be associated with their exaggerated response to MSCs compared with HC PBMCs.

PBMCs appear to be involved in regulating the anti-inflammatory cytokine levels in co-cultured supernatant, although we did not measure the concentrations of all cytokines in the supernatant of MSC-alone cultures. Regarding the anti-inflammatory cytokines, the MSCs displayed increased mRNA levels of IL-10 in 24-h co-cultured PBMCs and of IL-4, TGF-β and Gata3 in 72-h co-cultured PBMCs. However, IL-13 mRNA level was not changed. Although we investigated the change in the cytokine levels in the PBMCs, it has been demonstrated that IL-4, IL-10, and TGF-β are the predominant cytokines that are expressed and released by Treg and Th2 lymphocytes (Fujimura et al. 2010; Sakaguchi et al. 2010; Singer and Caplan 2011). The production of anti-inflammatory cytokines, such as IL-10 and TGF-β, plays an important role in suppressing the proliferation of activated CD4+/CD25 T cells. Additionally, it has been reported that the mRNA expression of IL-4 and IL-10 was increased in the spinal cords of mSOD1 mice during the stable disease phase and was suppressed at the end disease stage (Beers et al. 2011). Reduced IL-4 and TGF-β expression levels in leukocytes have been found in rapidly progressing ALS patients and inversely correlated with the progression rate (Henkel et al. 2013). In addition, recent studies have demonstrated that microglial phenotype and activity were orchestrated and shaped by peripheral leukocytes, such as Treg and monocytes, in various neurodegenerative diseases, including ALS (Yong and Rivest 2009; Michaud et al. 2013; Shechter et al. 2013a). Previous studies have shown that IL-4 inhibited microglial activation shaped the microglial phenotype and augmented motor neuron survival in vitro (Butovsky et al. 2005; Zhao et al. 2006). Taken together, our findings that the increase in the levels of anti-inflammatory cytokines, such as IL-4, IL-10, and TGF-β, which occurs because of Treg and Th2 induction, can be stimulated in ALS PBMCs by MSCs, thus providing supporting evidence for the potential of stem cell therapy using MSCs. In addition, it can be speculated that early MSCs administration in this study, in which SOD1 mice age was 60 days and the mean ALS Functional Rating Scale point of patients was 42.3 ± 2.35, may prolong a stable disease phase (early phase) by increasing anti-inflammatory cytokines with Treg and Th2 in ALS progression (Beers et al. 2011; Noh et al. 2014). Therefore, the elevation in Treg and Th2 formation and anti-inflammatory cytokines may play important roles in retarding disease progression of ALS if MSCs are clinically targeted to ALS patients for stem cell therapy.

Another interesting finding in co-MSCs cultured with ALS PBMCs for 72 h was that it took somewhat longer for IDO and TGF-β to reach mRNA levels similar to those of co-MSCs cultured with HC PBMCs. However, a series of responses related to Treg induction and anti-inflammatory cytokine production which play critical roles in ALS progression, appeared to be induced normally in the co-cultured ALS PBMCs. In addition, the fold-change in the IDO mRNA level in the co-MSCs was highly correlated with Treg induction. The strong correlation between the fold-change in IDO mRNA expression in co-MSCs and the fold-change in Treg cells in ALS PBMCs indicates that the IDO mRNA expression level may represent a biomarker of the capacity of MSCs to induce Treg formation, although we cannot ignore the roles that TGF-β and PGE2 play in Treg induction.

In conclusion, we found that MSCs exert immunomodulatory effects in ALS PBMCs via the induction of Treg and Th2, and the increase in the levels of anti-inflammatory cytokines, including IL-4, IL-10, and TGF-β, which were not different from the results obtained for HC PBMCs although those may result from the use of early phase ALS PBMCs in co-culture. These results provide indirect evidence that MSCs may play a positive role in the immunomodulatory effects when MSC therapy is targeted to ALS patients. However, based on the MSCs being administered intrathecally, we cannot confirm the following: (i) whether the infiltrated immune cells are maintained until end stage of ALS, (ii) which subsets of CD3 cells are infiltrated into spinal cord of SOD1 mice and ALS patients, and (iii) whether the induced Treg by MSCs have suppressive function in ALS PBMCs. These undetermined issues will be recapitulated in further study with a larger sample.

Acknowledgments and conflict of interest disclosure

This study was supported by grants from the Korea Healthcare Technology R&D Project of the Ministry for Health & Welfare Affairs of the Republic of Korea (A101712 and A120182). The authors have no conflicts of interest to disclose.

All experiments were conducted in compliance with the ARRIVE guidelines.

Ancillary