Potent antitumor effects of the conditioned medium of bone marrow‐derived mesenchymal stem cells via IGFBP‐4

Abstract Cell transfer therapy using mesenchymal stem cells (MSCs) has pronounced therapeutic potential, but concerns remain about immune rejection, emboli formation, and promotion of tumor progression. Because the mode of action of MSCs highly relies on their paracrine effects through secretion of bioactive molecules, cell‐free therapy using the conditioned medium (CM) of MSCs is an attractive option. However, the effects of MSC‐CM on tumor progression have not been fully elucidated. Herein, we addressed this issue and investigated the possible underlying molecular mechanisms. The CM of MSCs derived from human bone marrow greatly inhibited the in vitro growth of several human tumor cell lines and the in vivo growth of the SCCVII murine squamous cell carcinoma cell line with reduced neovascularization. Exosomes in the MSC‐CM were only partially involved in the inhibitory effects. The CM contained a variety of cytokines including insulin‐like growth factor binding proteins (IGFBPs). Among them, IGFBP‐4 greatly inhibited the in vitro growth of these tumors and angiogenesis, and immunodepletion of IGFBP‐4 from the CM significantly reversed these effects. Of note, the CM greatly reduced the phosphorylation of AKT, ERK, IGF‐1 receptor beta, and p38 MAPK in a partly IGFBP4‐dependent manner, possibly through its binding to IGF‐1/2 and blocking the signaling. The CM depleted of IGFBP‐4 also reversed the inhibitory effects on in vivo tumor growth and neovascularization. Thus, MSC‐CM has potent inhibitory effects on tumor growth and neovascularization in an IGFBP4‐dependent manner, suggesting that cell‐free therapy using MSC‐CM could be a safer promising alternative for even cancer patients.


| INTRODUC TI ON
Mesenchymal stem cells (MSCs) are adult SCs with self-renewal and multipotency, which can be isolated from a variety of tissues and are culture expandable. [1][2][3][4][5] MSCs have low immunogenicity and potent immunomodulatory properties, but no tumorigenicity. MSCs show pronounced therapeutic effects against a variety of diseases including inflammatory diseases and autoimmune diseases and therefore have attracted great interest in the field of regeneration medicine. 3,6 The therapeutic effects of MSCs are mediated by their ability to migrate toward damaged tissues, as well as subsequent secretion by engrafted cells of bioactive molecules such as growth factors, cytokines, and extracellular vesicles including exosomes. [7][8][9] Although cell transfer therapy using MSCs has promising therapeutic potential, there are still concerns about its risk of tumor promotion, 10,11 immune rejection of cells, and possible induction of embolism or thrombosis, as well as ethical issues. 12 Distinct from induced pluripotent SCs that have intrinsic tumorigenic properties to induce teratomas, 13 MSCs do not have such intrinsic tumorigenicity. However, tumor stromal cells such as MSCs in the bone marrow (BM) supporting hematopoietic SC niches facilitate the formation of cancer SC niches, thereby contributing to progressive tumor development. 14 Indeed, MSCs have the intrinsic property of homing toward tumor sites 15 and thus can be used as tumor-specific vectors for tumor therapy. 11 Within the tumor microenvironment (TME), MSCs can exert both stimulatory and inhibitory effects on tumor cell growth, invasion, and metastasis by affecting tumor cells through direct cell-cell interactions and secretion of various factors including cytokines and extracellular vesicles, and also by modulating the innate and adaptive immune responses. 10,11 In response to soluble factors such as transforming growth factor beta secreted from cancer cells, MSCs can differentiate into cancer-associated fibroblasts, a cell type capable of promoting tumorigenesis within the TME. 16 Thus, reciprocal modulation of MSCs and tumor cells in vivo is highly likely to be biased toward the promotion of tumor progression and metastasis, 17 and the net effects of MSCs are considered to be predominantly protumorigenic. 10,11 By contrast, cell-free therapy using the conditioned medium (CM) of MSCs (MSC-CM) has fewer concerns but similar therapeutic effects to cell transfer therapy using MSCs, and thus is as an attractive alternative. 9,18 MSCs have the capacity to migrate to injured sites in response to environmental signals and differentiate into damaged tissue and cells to promote tissue regeneration. Because, in most cases, only a small number of surviving cells are able to migrate in a short period of time, [19][20][21][22] the mode of action of MSCs highly relies on their strong secretion of bioactive molecules upon interaction with neighboring cells and tissues. 21,22 To date, however, limited studies have investigated the antitumor properties of MSC-CM and the underlying molecular mechanisms. It was recently reported that the CM, but not exosomes, of MSCs derived from the BM of young healthy donors significantly inhibited the proliferation of hematological malignant cells such as myeloma and lymphoma cells in vitro, but the molecular mechanism remains unknown. 23 In this study, we investigated the effects of the CM of human BM-derived MSCs on in vitro and in vivo tumor growth and angiogenesis using human myeloma, breast cancer, lung cancer, and mouse squamous carcinoma cell lines and its possible underlying molecular mechanisms. The CM of MSCs greatly inhibited in vitro tumor growth and angiogenesis. To elucidate the molecular mechanisms by which CM exhibits antitumor effects, antibody array analysis was performed and consequently revealed that CM contains a variety of cytokines and growth factors. Among them, insulin-like growth factor binding protein-3 (IGFBP-3), IGFBP-4, and IGFBP-6 ranked at higher positions. The IGFBP family consists of six members (IGFBP-1-6) sharing 50% homology and binds to IGF-1 and IGF-2 but not insulin. 24,25 The IGF axis plays important roles in the promotion of cell proliferation and the inhibition of cell death. 26 Therefore, we examined the effects of recombinant (r) IGFBP-3, IGFBP-4, and IGFBP-6 and their immunodepletion from CM on tumor growth and angiogenesis and found that IGFBP-4 in CM is critically involved in the inhibitory effects on tumor growth and antiangiogenic activity. This is the first report on the antitumor effects of MSC-CM, which showed that IGFBP-4 in the CM plays important roles. Thus, cell-free therapy using the MSC-CM could be a safer and promising alternative for even cancer patients.

| Flow cytometry
Single-cell suspensions of MSCs were stained with Pacific Blue- Dead cells were discriminated using 7-aminoactinomycin D (Sigma-Aldrich). The flow cytometry gating strategy for the tetramer staining assay is shown in Figure S1.

| Preparation of CM
Mesenchymal stem cells cultured in αMEM supplemented with 5% PLTMax were washed three times with phosphate-buffered saline (PBS) and then cultured in unsupplemented αMEM for 72 h.
As control CM, the culture supernatant that was immediately col-

| Cell growth analysis
Tumor cells (1 × 10 3 cells/100 μL) were incubated with 0%-40% con-  The lower fraction was resuspended in the same volume of medium before ultracentrifugation. The content of exosomes was analyzed by Western blotting and the NanoSight NS300 Nanoparticle Tracking Analysis System (Malvern Panalytical).

| Cytokine measurement
The cytokine content in the CM was broadly analyzed using the Quantibody Human Cytokine Array Q1000 (RayBiotech), a multiplex ELISA array that can quantitatively determine the concentration of 80 different cytokines (Table S1). The intensity of each signal was determined by laser scanning using the GenePix 4400A scanner (Molecular Devices).

| Western blot analysis
The content of exosomes was analyzed by Western blotting using antibodies against CD63 and CD81. Briefly, the sample was separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis and electrotransferred to polyvinylidene difluoride membranes (Millipore). Then, the membrane was blocked, probed with antibodies against CD63 (clone H5C6; BioLegend) or CD81 (clone 5A6; BioLegend), incubated with anti-mouse IgG conjugated to horseradish peroxidase (HRP), and visualized with the enhanced chemilumi- CST), total IGF-1Rβ (clone D23H3; CST), p-p38 (Thr180/Tyr182, clone D3F9; CST), or total p38 (clone D13E1; CST). Then the membranes were incubated with an appropriate secondary antibody conjugated to HRP and visualized with the enhanced chemiluminescence detection system. Immunoreactive bands were detected with the iBright FL1500 Imaging System (Thermo Fisher Scientific).

| Tube formation assay
The angiogenic activity of MSC-CM was determined with a tube formation assay using HUVECs. 27 Briefly, HUVECs were placed on a Matrigel (BD Biosciences) or Cultrex Basement Membrane Extract (BME, R&D Systems) in a 24-well plate and incubated for 40-60 min at 37°C. SSCVII tumor cell-derived concentrated CM (10% of ~10fold concentrated), rIGF-1 (10 ng/mL), or rVEGF (10 ng/mL) with or without MSC-CM that had been untreated, treated with control antibody, or depleted of IGFBP-4 with its antibody, or rIGFBP-4 was then added and incubated for further 7 h. The tube formation ability of HUVECs was monitored under the DM IL inverted microscope (Leica), and each image was captured at different time points.
The tube length was determined by drawing lines along the tubelike structure and measuring the length of the line in pixels 28 using ImageJ software (National Institutes of Health). 29 The number of nodes of the tubular structures was quantified as the mean pixel density obtained from image analysis of multiple random microscopic fields using ImageJ software. 29

| In vivo tumor growth
SSCVII tumor cells (1 × 10 6 cells/mouse) were injected subcutaneously into the right flank of mice. The MSC-CM, control CM (100 μL), or antibody-treated CM of MSCs was intradermally injected into the mice starting on day 5, when the tumor became palpable, followed by similar injections daily up to five or seven times. Tumor size was measured daily using electronic calipers and was expressed as volume (mm 3 ) using the volume equation 0.5(ab 2 ), where a is the long diameter and b is the short diameter.

| Immunohistochemical analysis
Mice were euthanized in a CO 2 chamber, and tumor masses were harvested and fixed in 10% neutral buffered formalin and embedded in paraffin. Deparaffinized sections (4 μm) were treated by boiling for 10 min at 121°C in 1 mM EDTA/10 mM Tris-HCl buffer (pH 9.0) for antigen retrieval of CD31. After blocking in 2% skim milk, the sections were incubated with primary antibodies, rat anti-mouse CD31 (clone SZ31; Dianova), followed by incubation with HRP-conjugated goat anti-rat IgG and colorimetric detection using diaminobenzidine substrate and counterstaining with hematoxylin and eosin (H&E). Immunostained cells were quantified with the EVOS M7000 Imaging System (Thermo Fisher Scientific) or Keyence BZ-X810. The positive area was calculated using ImageJ software.

| Statistical analyses
Data are expressed as the mean ± standard deviation (SD) for each group. Statistical analyses were performed with GraphPad Prism software (version 9; GraphPad Software) using the unpaired twotailed Student's t-test for comparisons of two groups and one-or two-way analysis of variance (ANOVA) with Tukey's or Dunnett's multiple comparison test for comparisons involving three or more groups. P < 0.05 was considered statistically significant.

| IGFBP-4 in MSC-CM contributes to the inhibitory effects on tumor growth
To explore the molecular mechanism by which MSC-CM exerts inhibitory effects on tumor growth, the MSC-CM was analyzed for the production levels of various cytokines using a cytokine antibody array, revealing that the CM contained a variety of cytokines and growth factors of which tissue inhibitors of metalloproteinase and

| The CM of MSCs blocks IGF signaling in a partly IGFBP-4-dependent manner
Insulin-like growth factor signaling pathways are important for cell proliferation, 26 and IGF-1R is a receptor for IGF-1 and IGF-2 and consists of two α-and β-subunits derived from the same precursor. 31 The β-subunit contains a kinase domain and is phospho-

F I G U R E 1 The conditioned medium of mesenchymal stem cells (MSC-CM) shows growth inhibitory effects against various tumors in
vitro and in vivo together with reduced neovascularization. Human multiple myeloma RPMI8226, human breast cancer MCF-7, human lung cancer A549, and mouse squamous carcinoma SSCVII cell lines were incubated with various percentages of concentrated MSC-CM#2 (A) or #3 (B) obtained from different donors or control medium in the medium supplemented with 1% FBS for 72 h, and cell proliferative activity was determined in triplicate. Data are shown as the mean ± SD and are representative of more than two independent experiments. SSCVII tumor cells were injected subcutaneously into the right flank of C3H/He mice. The CM of MSCs or control medium was intradermally injected into the mice starting on day 5, when the tumor became palpable, followed by similar injections every day up to five times. Tumor size was measured and tumor volume was calculated (C). (D) Representative images of CD31 + blood vessels in each tumor section are shown. (E) The number per field and percentage of CD31 + blood vessels were calculated. Data are shown as the mean ± SD (n = 5) and are representative of two independent experiments. p-values were determined using unpaired two-tailed Student's t-test. *P < 0.05, **P < 0.01, ***P < 0.001.
Of note, preincubation of MSC-CM and FBS also similarly, but less effectively, enhanced the growth inhibitory effects on these tumors with the exception of MCF-7 ( Figure 4B). These results suggest that the inhibitory effects of MSC-CM on tumor growth are highly likely to be mediated by binding to IGFs present in the FBS and consequently inactivating them.
IGF-1 and IGF-2 bind to IGF-1R and induce phosphorylation of IGF-1Rβ followed by phosphorylation of AKT and ERK, as one of the critical signaling pathways leading to cell proliferation. 36

F I G U R E 2
Exosomes in the conditioned medium of mesenchymal stem cells (MSC-CM) were only partially involved in the inhibitory effects on tumor growth. MSC-CM was separated into two fractions by ultracentrifugation: The supernatant contained MSC-CM without exosomes, and the precipitate contained exosomes derived from MSC-CM. The precipitate was resuspended in the same volume before ultracentrifugation. The content of exosomes was analyzed by scanning electron microscopy analysis (A) and nanoparticle-tracking analysis (B) with NanoSight and Western blotting using antibodies against CD63 and CD81 following sodium dodecyl sulfate polyacrylamide gel electrophoresis (C). Respective fractions and MSC-CM were analyzed for their tumor growth inhibitory activities in vitro using human lung cancer A549 (D) and mouse squamous carcinoma SSCVII (E). Cell proliferative activity was determined in triplicate. Data are shown as the mean ± SD and are representative of two independent experiments. p-values were determined using one-way ANOVA with the Tukey test for multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001.

| The CM of MSCs exerts antiangiogenic activity in an IGFBP-4-dependent manner
Antitumor effects are highly related to not only direct growth inhibitory effects but also antiangiogenic effects. IGFBP-4 has antiangiogenic activity as well. 33

| The CM of MSCs inhibits angiogenesis induced by IGF-1 and VEGF in an IGFBP-4-dependent manner and IGFBP-4-independent manner, respectively
Because the MSC-CM contains both anti-angiogenic molecules such as IGFBPs and pro-angiogenic molecules such as growth factors such as VEGF, the overall outcome may depend on the assay system in vitro or the microenvironmental conditions in vivo due to the balance of them. Therefore, we further investigated the ef-

| DISCUSS ION
In this study, we addressed the question of whether the CM of BM-derived MSCs exerts protumorigenic or antitumorigenic effects and the possible underlying molecular mechanisms.
MSC-CM greatly inhibited the in vitro growth of several human cancer cell lines and in vivo growth of SSCVII mouse squamous F I G U R E 3 Insulin-like growth factor binding protein-4 (IGFBP-4) in the conditioned medium of mesenchymal stem cells (MSC-CM) contributes to the inhibitory effects on tumor growth. (A) Content of cytokines in the CM was broadly analyzed using the cytokine antibody array, and the concentrations of 80 different cytokines were quantitatively determined. Cytokines whose concentration is more than 1 pg/mL were shown. (B) Human lung cancer A549 and mouse squamous carcinoma SSCVII were incubated with recombinant (r)IGFBP-3, rIGFBP-4, or rIGFBP-6 in the medium supplemented with 1% FBS for 72 h. Cell proliferative activity was determined in triplicate. (C) Human breast cancer MCF-7, A549, and SSCVII were incubated with 40% control medium, untreated MSC-CM (−), MSC-CM treated with control antibody, or MSC-CM depleted of IGFBP-3, -4, or -6 in the medium supplemented with 1% FBS for 72 h. Cell proliferative activity was determined in triplicate. (D) These tumor cells together with human multiple myeloma RPMI8226 were incubated with 40% control medium, untreated MSC-CM (−), MSC-CM treated with control antibody, or MSC-CM depleted of IGFBP-4 in the medium supplemented with 1% FBS for 72 h. Cell proliferative activity was determined in triplicate. Data are shown as the mean ± SD and are representative of more than two independent experiments. p-values were determined using one-way (C, D) or two-way (B) ANOVA with the Tukey test for multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001. Insulin-like growth factors play a variety of roles as endocrine, paracrine, and autocrine factors that promote cell growth, proliferation, differentiation, survival, and wound healing. 36 The IGFBP family comprises six related members, IGFBP1-6, which bind IGF-1 and IGF-2 but not insulin with high affinity. 24 10,11 and that the mode of action of MSCs highly relies on strong paracrine effects F I G U R E 4 The conditioned medium of mesenchymal stem cells (MSC-CM) blocks insulin-like growth factor (IGF) signaling in a partly IGF binding protein-4 (IGFBP-4)-dependent manner. (A) Human multiple myeloma RPMI8226, human breast cancer MCF-7, human lung cancer A549, and mouse squamous carcinoma SSCVII were incubated in the presence of various concentrations of recombinant (r)IGFBP-4 with and without preincubation of 1% FBS and rIGFBP-4 at 4°C for 1 h. After 72 h, the cell proliferative activity was determined in triplicate. (B) These tumor cells were incubated in the presence of various percentages of MSC-CM with and without preincubation of 1% FBS and MSC-CM at 4°C for 1 h. After 72 h, the cell proliferative activity was determined in triplicate. Data are shown as the mean ± SD and are representative of more than two independent experiments. SSCVII cells were incubated with 40% control medium, undertreated MSC-CM (−), MSC-CM treated with control antibody, or MSC-CM depleted of IGFBP-4 in the medium supplemented with 1% FBS for 24 h. Resultant cell lysates were subjected to Western blotting using antibodies against pAKT (C), pERK (D), and pIGF-1Rβ (E) together with antibodies against respective total molecules. Data are representative of more than two independent experiments. (F) Each band intensity in (E) was quantified using ImageJ, and relative expression of pIGF-1Rβ per total pIGF-1Rβ was calculated and statistically compared among three independent experiments. Data are shown as the mean ± SD. p-values were determined using unpaired two-tailed Student's t-test (A), one-way (F), or two-way (B) ANOVA with the Tukey test for multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001.

F I G U R E 5
The conditioned medium of mesenchymal stem cells (MSC-CM) exerts antiangiogenic activity in an insulin-like growth factor binding protein-4 (IGFBP-4)-dependent manner. Angiogenic activities of 40% control medium, untreated MSC-CM (−), MSC-CM treated with control antibody, or MSC-CM depleted of IGFBP-4 were determined by a tube formation assay induced by the culture supernatant of SSCVII tumor cells using HUVECs. Representative photographs of the tube formation are shown (A). Total length of tube formation (B) and number of nodes per field (C) were quantified using ImageJ. Data are shown as the mean ± SD in triplicate and are representative of two independent experiments. p-values were determined by one-way ANOVA with the Tukey test for multiple comparisons test. **P < 0.01, ***P < 0.001.

F I G U R E 6
Legend on next page F I G U R E 7 Inhibitory effects of the conditioned medium of mesenchymal stem cells (MSC-CM) on the tumor progression and neovascularization in vivo are dependent on insulin-like growth factor binding protein-4 (IGFBP-4). Mouse squamous carcinoma SCVII tumor cells were injected subcutaneously into the right flank of C3H/He mice. The MSC-CM treated with control antibody or MSC-CM depleted of IGFBP-4 (A) was intradermally injected into the mice starting on day 5, when the tumor became palpable, followed by similar injections every day up to seven times (B). The extent of depletion of IGFBP-4 was determined to be approximately 60% by measuring the residual concentration by ELISA (A). Tumor size was measured, and tumor volume was calculated (B). (C) Representative images of CD31 + blood vessels in each tumor section are shown. (D) The number of CD31 + blood vessels per field was calculated. Data are shown as the mean ± SD (n = 5) and are representative of two independent experiments. p-values were determined using one-way ANOVA with the Tukey test for multiple comparisons (A) or unpaired two-tailed Student's t-test (B, D). *P < 0.05, **P < 0.01, ***P < 0.001.

F I G U R E 6
The conditioned medium of mesenchymal stem cells (MSC-CM) inhibits angiogenesis induced by insulin-like growth factor-1 (IGF-1) and VEGF in an IGF binding protein-4 (IGFBP-4)-dependent manner and IGFBP-4-independent manner, respectively. Angiogenic activities of 40% control medium, untreated MSC-CM (−), MSC-CM treated with control antibody, or MSC-CM depleted of IGFBP-4 were determined by a tube formation assay induced by recombinant (r)IGF-1 or rVEGF using HUVECs. Representative photographs of the tube formation are shown (A, C). Total length of tube formation (B, D) was quantified using ImageJ. Data are shown as the mean ± SD in triplicate and are representative of two independent experiments. Serum-starved HUVECs were incubated with the culture supernatant of SSCVII tumor cells, rIGF-1, or rVEGF in the presence or absence of 40% control medium, undertreated MSC-CM (−), MSC-CM treated with control antibody, or MSC-CM depleted of IGFBP-4 for 15 min. Resultant cell lysates were subjected to Western blotting using antibodies against p-p38 and total p38 (E, G, I). Each band intensity was quantified using ImageJ, and relative expression of p-p38 per total p38 was calculated and statistically compared among three independent experiments (F, H, J). Data are shown as the mean ± SD. p-values were determined using one-way ANOVA with the Tukey test for multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001. NS, not significant. through their high secretion of bioactive molecules, 21,22 cell-free therapy using just the CM of MSCs may be a safer and promising alternative for even cancer patients.

ACK N OWLED G M ENTS
The

CO N FLI C T O F I NTER E S T S TATEM ENT
The authors declare no conflict of interest.