Author response: Mesenchymal stem cell suppresses the efficacy of CAR-T toward killing lymphoma cells by modulating the microenvironment through stanniocalcin-1

Abstract

Article Figures and data Abstract Editor's evaluation eLife digest Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Stem cells play critical roles both in the development of cancer and therapy resistance. Although mesenchymal stem cells (MSCs) can actively migrate to tumor sites, their impact on chimeric antigen receptor modified T cell (CAR-T) immunotherapy has been little addressed. Using an in vitro cell co-culture model including lymphoma cells and macrophages, here we report that CAR-T cell-mediated cytotoxicity was significantly inhibited in the presence of MSCs. MSCs caused an increase of CD4+ T cells and Treg cells but a decrease of CD8+ T cells. In addition, MSCs stimulated the expression of indoleamine 2,3-dioxygenase and programmed cell death-ligand 1 which contributes to the immune-suppressive function of tumors. Moreover, MSCs suppressed key components of the NLRP3 inflammasome by modulating mitochondrial reactive oxygen species release. Interestingly, all these suppressive events hindering CAR-T efficacy could be abrogated if the stanniocalcin-1 (STC1) gene, which encodes the glycoprotein hormone STC-1, was knockdown in MSC. Using xenograft mice, we confirmed that CAR-T function could also be inhibited by MSC in vivo, and STC1 played a critical role. These data revealed a novel function of MSC and STC-1 in suppressing CAR-T efficacy, which should be considered in cancer therapy and may also have potential applications in controlling the toxicity arising from the excessive immune response. Editor's evaluation This study uncovers the contributions of MSC on modulating CAR T-cell behaviour. Based on the importance in basic biology and its immediate impact on translational potential, all reviewers are satisfied on the advances in this study. https://doi.org/10.7554/eLife.82934.sa0 Decision letter Reviews on Sciety eLife's review process eLife digest Immunotherapy is a type of cancer treatment that helps the immune system fight cancer. For example, chimeric antigen receptor T cell (CAR-T) therapy is used to target several types of blood cancer. It works by reprogramming patients’ immune cells to target specific tumor cells. In blood cancers, CAR-T therapy works very well, but it can cause extreme responses from the patient’s immune system, which can be life threatening. In solid tumors, CAR-T therapy is much less successful because the tumors secrete molecules into the space surrounding them, which weaken the immune processes that attack cancerous cells. Stem cells are the master cells of the body. Originating in the bone marrow, they can repair and regenerate the body’s cells. Cancer stem cells play a role in resistance to CAR-T therapy, due – in part – to their ability to renew themselves, but the role of another type of stem cell, called mesenchymal stem cells, was less clear. Mesenchymal stem cells develop into tissues that line organs and blood vessels. Although it is known that mesenchymal stem cells are present in most cancers and play a role in shaping and influencing the space around tumors, their impact on CAR-T therapy has not been studied in depth. To find out more, Zhang et al. looked at the influence of a protein, called staniocalcin-1 (STC1), on CAR-T therapy, by studying cells grown in the laboratory and human tumor cells that had been implanted in mice. Zhang et al. found that mesenchymal stem cells reduce the ability of CAR-T therapy to destroy cancer cells and that they needed STC1 to do this successfully. They also increased the expression of molecules that dampen the immune system, and suppressed molecules called inflammasomes, which are an important part of the way the immune system detects disease. Moreover, reducing the amount of STC1 that mesenchymal stem cells expressed restored the effectivity of CAR-T therapy. This study increases our understanding of the way that mesenchymal stem cells affect CAR-T therapy. It has the potential to open up a new way of improving the efficiency of this treatment and of reducing the harmful side effects that it can cause. Introduction Advances in chimeric antigen receptor modified T cell therapy (CAR-T) in recent years have shown enormous promise in cancer immunotherapy, which has produced unprecedented clinical outcomes, most notably for patients with hematologic malignancies (Singh et al., 2016; Park et al., 2018). Despite the striking achievements, CAR-T therapy is also facing many challenges such as the treatment-related severe toxicity and side effects, including cytokine release syndrome (CRS) and neurotoxicity (Hong et al., 2020; Freyer and Porter, 2020). CRS is the most common acute toxicity associated with an excessive immune response that causes fever, hypotension, and respiratory insufficiency. The neurotoxicity induced by CAR-T therapy exhibits a diverse array of neurologic symptoms such as tremors, expressive aphasia, and impaired attention. The precise mechanism that causes these life-threatening side effects remains unclear (Freyer and Porter, 2020; Jiang et al., 2019). On the other hand, the success of CAR-T therapy in treating solid tumors is still very limited (Martinez and Moon, 2019). Identifying hurdles and potential mechanisms that impede the function of CAR-T cells is of vital importance to expanding its use. The immunosuppressive tumor microenvironment (TME) is one of the obstacles that diminishes the efficacy of CAR-T therapy, especially for solid tumors. Among the many factors that can modulate TME and immune response, the impact of mesenchymal stem cell (MSC) on CAR-T therapy has been little studied. MSC is a type of adult stem cell with high proliferative activity and multidirectional differentiation capacity. However, MSCs have additional paracrine effects that are believed to underlie their therapeutic functions (Jiang and Xu, 2020). By secreting a variety of cytokines into the tissue microenvironment, it has been known that MSCs can modulate extracellular matrix, promote angiogenesis, and suppress inflammation and apoptosis (Keating, 2012; Wang et al., 2014; Regmi et al., 2019). Some MSC-secreted cytokines, such as stromal cell-derived factor 1 and stem cell factor, play important roles in hematopoietic and immune regulation (Kawaguchi et al., 2019; Markov et al., 2007). In addition, studies suggest that MSCs can modulate the function of monocytic lineages cells, especially macrophages (Németh et al., 2009; YlÖstalo et al., 2012; Choi et al., 2011). Some reports also showed that MSCs could directly affect the functionality and cellular responses of T cells, Tregs, and memory T cells (Cen et al., 2019; Tumangelova-Yuzeir et al., 2019; Luque-Campos et al., 2019). It was reported that human mesenchymal stem cells (hMSCs) could be activated by lipopolysaccharide (LPS)-stimulated macrophages to increase the expression and secretion of stanniocalcin-1 (STC1) (Oh et al., 2014). STC1 was a mitochondria-related glycoprotein originally identified as a calcium/phosphate regulating hormone in bony fishes, and later on, it was found to be a pleiotropic factor involved in various degenerative diseases such as ocular and renal disease, as well as idiopathic pulmonary fibrosis (Yeung et al., 2012; Ohkouchi et al., 2015). STC1 could improve the cell survival and regeneration of MSCs in a paracrine fashion (Ono et al., 2015). There was also evidence suggesting that STC1 played an oncogenic role in various types of tumors (Du et al., 2011; Liu et al., 2010). Based on a retrospective study of ~1500 clinical samples, it was concluded that high STC1 expression is associated with the poor clinical outcome of breast cancer (Chang et al., 2015). It was proved that STC1 is involved in several oxidative and cancer-related signaling pathways, such as NF-κB, extracellular-signal-regulated kinase (ERK), and c-Jun NH(2)-terminal kinase (JNK) pathways (Nguyen et al., 2009; Chan et al., 2017). The expression and secretion of STC1 in cancer tissue can be stimulated by external stimuli, including external cytokines and oxidative stress (Nguyen et al., 2009). Under hypoxia conditions, STC1 could be modulated by Hypoxia-inducible factor-1 (HIF-1) to facilitate the reprogramming of tumor metabolism from oxidative to glycolytic metabolism (Yeung et al., 2005). STC1 was also reported to participate in the process of epithelial-to-mesenchymal transition, which is associated with tumor invasion and the reshape of the tumor microenvironment, as well as increasing therapy resistance (Pastushenko and Blanpain, 2019). Considering the pleiotropic role of STC1, especially its intercellular linkage between MSCs, cancer cells, and macrophage stimulation, it is interesting to know what role it plays in connection to the functions of MSC in TME. Therefore, we generated a stable STC1 knockdown MSC cell line. With a cell co-culture model containing CAR-T cells, hMSCs, macrophages, and Pfeiffer lymphoma cells to partially mimic the tumor microenvironment together with a xenograft mice model, here we studied the impacts of MSC on CAR-T efficacy and the potential immune response change in the presence and absence of STC1. Results Stable knockdown of STC1 in hMSC-inhibited cell migration, slightly suppressed cell proliferation, but no increase in apoptosis To study the function of STC1, we first generated a stable knockdown cell line by lentivirus-based shRNA for the STC1 gene, and the expression of STC1 protein was evaluated by Western blot (Figure 1A). STC1 stable knockdown in hMSCs exhibited a minor effect in cell survival (Figure 1B) and slightly reduced proliferation rate based on the small increase in the proportion of cells in G0/G1 phases versus that in the S phase (Figure 1C) as determined by MTT (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide) and Fluorescence-activated Cell Sorting (FACS) analysis. To investigate whether knockdown of STC1 affects cell migration, wound healing and transwell chamber assays were performed. After creating a ‘scratch’ in a monolayer of hMSCs, the closure of the gap was determined after 24 hr. As shown in Figure 1D, compared to control hMSCs, the gap was less filled in hMSCshSTC1. The inhibitory effect on cell migration was further confirmed by a transwell assay. As shown in Figure 1E, there were significant migration and invasion observed in hMSCsshCtrl, whereas there was a >30% reduction in migration across the transwell chamber membrane in hMSCsshSTC1. To further determine whether knockdown of STC1 may have any lethal effect, apoptosis was determined by two different assays. To measure the early apoptosis, cells were stained with the Alexa Fluor 488 annexin V and the propidium iodide (PI) followed by flow cytometry to detect apoptosis-associated phosphatidylserine (PS) expression and membrane permeability (Figure 1F). Parallelly, no DNA fragmentation was detected as determined with the TUNEL assay (Figure 1G, the green dots were from the background due to overexposure). Both studies showed that knockdown of STC1 did not cause apoptosis of hMSCs. Figure 1 Download asset Open asset The impact of stanniocalcin-1 (STC1) knockdown on cell proliferation, migration, and apoptosis of hMSCs. (A) Western blot analysis of STC1 protein expression in hMSCs. (B) Cell viability determined by MTT, measurements are shown as the mean ± SD from three independent experiments. (C) FACS analysis of cell cycle progression on hMSCs w/o STC1 knockdown. (D, E) Knockdown of STC1 suppressed cell migration as determined by wound healing and transwell chamber assays. (F) Apoptosis determination by the Alexa Fluor 488 annexin V and PI detection. (G) DNA fragmentation determination by transferase-mediated dUTP nick-end labeling (TUNEL) assay. Figure 1—source data 1 Labeled original blots of Figure 1A. https://cdn.elifesciences.org/articles/82934/elife-82934-fig1-data1-v3.zip Download elife-82934-fig1-data1-v3.zip Figure 1—source data 2 Unlabeled original blots of Figure 1A. https://cdn.elifesciences.org/articles/82934/elife-82934-fig1-data2-v3.zip Download elife-82934-fig1-data2-v3.zip Figure 1—source data 3 Figure 1B in Excel file. https://cdn.elifesciences.org/articles/82934/elife-82934-fig1-data3-v3.xlsx Download elife-82934-fig1-data3-v3.xlsx The presence of hMSCs inhibited CAR-T cell killing activity, but knockdown of STC1 completely abrogated this inhibition To investigate the impact of hMSCs on CAR-T treatment, we used an in vitro cell co-culture model modified according to previous studies to mimic a simplified situation of tumor environment (Singh et al., 2017; Liu et al., 2021). The co-culture contained CD19 CAR-T cells, Pfeiffer cells that were from human diffuse large cell lymphoma, and M2 macrophages (derived from THP-1 cells by phorbol-12-myristate-13-acetate [PMA] polarization for 24 hr) at a cell number ratio of 1:3:1. The cell-killing activity of CAR-T cells toward Pfeiffer cells was determined by lactate dehydrogenase (LDH) cytotoxicity assay on total cell co-culture. As shown in Figure 2A, 67% of Pfeiffer cells were killed after being exposed to CAR-T cells for 24 hr, and 93% were killed at 48 hr as compared to mock-treated control. After adding hMSCs into the co-culture, the cell-killing activity of CAR-T was significantly inhibited (Figure 2A). The number of hMSC added was the same as the CAR-T cell. Interestingly, the inhibitory effect of hMSCs on CAR-T cytotoxicity could be completely abrogated if knockdown STC1 gene in hMSCs. These results for the first time revealed that CAR-T efficacy could be affected by the presence of MSCs, and the gene STC1 played a critical role. Figure 2 Download asset Open asset Analysis of cytotoxicity, T cell composition, and immune-suppressive markers. The cell co-culture contained chimeric antigen receptor modified T cell (CAR-T) cells, Pfeiffer cells, M2 macrophages, and control or stanniocalcin-1 (STC1) knockdown hMSCs in a ratio of 1:3:1:1. After 24 hr (or 48 hr for cytotoxicity) incubation, the following analysis was conducted: (A) The impact of hMSC (w/o STC1) on the cytotoxicity of CAR-T toward Pfeiffer cells; (B) FACS analysis of CD4+ and CD8+ composition. (C) Quantitation of the FACS data on CD4+ and CD8+; (D) FACS analysis of Treg+ cells (CD4+CD127+CD25+); (E) Quantitation of Treg+ cells. (F) Western blot analysis of indoleamine 2,3-dioxygenase (IDO) and programmed cell death-ligand 1 (PD-L1) expression in the cell co-culture. Data in bar graphs are presented as the mean ± SD from three independent experiments (p values are as indicated, n=3). Figure 2—source data 1 Figure 2A in Excel file. https://cdn.elifesciences.org/articles/82934/elife-82934-fig2-data1-v3.xlsx Download elife-82934-fig2-data1-v3.xlsx Figure 2—source data 2 Figure 2C in Excel file. https://cdn.elifesciences.org/articles/82934/elife-82934-fig2-data2-v3.xlsx Download elife-82934-fig2-data2-v3.xlsx Figure 2—source data 3 Figure 2E in Excel file. https://cdn.elifesciences.org/articles/82934/elife-82934-fig2-data3-v3.xlsx Download elife-82934-fig2-data3-v3.xlsx Figure 2—source data 4 Labeled original blots of Figure 2F. https://cdn.elifesciences.org/articles/82934/elife-82934-fig2-data4-v3.zip Download elife-82934-fig2-data4-v3.zip Figure 2—source data 5 Unlabeled original blots of Figure 2F. https://cdn.elifesciences.org/articles/82934/elife-82934-fig2-data5-v3.zip Download elife-82934-fig2-data5-v3.zip Co-culturing with hMSCs caused an increase of CD4+ T cells and Treg cells but a decrease of CD8+ T cells Previous studies have demonstrated that the composition of CD4+ and CD8+ T cell subsets was crucial for CAR-T cell efficacy (Sommermeyer et al., 2016; Turtle et al., 2016). To investigate the mechanism of how hMSC inhibited the cytotoxicity of CAR-T, the amount of CD4+ and CD8+ T cells were analyzed by flow cytometry 24 hr after co-culture. As shown in Figure 2B and C, the ratio between CD4+ and CD8+ was about 1:4 when there were no hMSCs in co-culture (Figure 2C). However, the addition of hMSC caused a significant increase of CD4+ and a decrease of CD8+ T cells (Figure 2B), resulting in a ratio change to 2:1. Similar to the change of CD4+ T cells, the percentage of regulatory T cells (Treg) was also significantly increased from ~3 to 12% when co-culture with hMSC (Figure 2D and E). When using hMSCshSTC1, all the changes were completely reversed back to the level similar to that of co-culture without hMSCs. This explains the reduced CAR-T cytotoxicity since CD8+ T cells are directly responsible for specific lytic activity against lymphoma (Sommermeyer et al., 2016). Tregs, which account for 5–10% of the total number of CD4+ T cells, are known to play a role in suppressing the function of T cells and other immune cells (Zhang et al., 2018). Therefore, the above results indicate that hMSCs’ inhibitory effect on CAR-T cytotoxicity was due to both suppression of CD8+ cells and the induction of Treg cells, and the presence of STC1 was indispensable for these impacts of hMSC. The presence of hMSC enhanced immune suppression and STC1 played a key role The immune-suppressive TME is the main cause of CAR-T cell exhaustion which attenuates its efficacy. To further investigate the function of STC1 and the molecular mechanism of hMSC on CAR-T resistance, some key regulators of TME were determined. As shown in Figure 2F, the addition of hMSC to the cell co-culture stimulated the expression of indoleamine 2,3-dioxygenase (IDO) and programmed cell death-ligand 1 (PD-L1). IDO and PD-L1 are two of the most important immunosuppressive proteins. IDO is an intracellular enzyme that converts tryptophan into inhibitory metabolites for T-cell activity (Ninomiya et al., 2015). PD-L1 is expressed in tumor cells and immune cells contributing to the immune-suppressive TME (Ribas and Hu-Lieskovan, 2016). When using hMSCshSTC1, the expression level of IDO and PD-L1 was both significantly reduced by more than 50%, though still higher than that without hMSC. These results indicated that the presence of hMSC can enhance the expression of immune suppressive proteins in Pfeiffer cells and macrophages, and the presence of STC1 is important for hMSC to exert these effects. hMSCs suppressed key components of the NLRP3 inflammasome by modulating mitochondrial ROS release In the co-culture model, M2 macrophages were included since a previous study showed that macrophages could activate hMSCs to secrete STC1 (Cen et al., 2019). In addition, the macrophage is a critical part of immune response and an important regulator of immunotherapy (DeNardo and Ruffell, 2019). To further identify the mechanisms mediating the inhibitory effects of hMSCs, the activation of the NLRP3 inflammasome was determined. The NLRP3 inflammasome is a critical component of the innate immune system mediating caspase-1 activation and proinflammatory cytokines secretion in response to harmful stimuli such as infection and endogenous stress (Menu and Vince, 2011). As shown in Figure 3A, the release of cleaved caspase-1 p20 in cell lysates, which is the indicator of caspase-1 activation, was detected after the PMA polarization of THP-1 cells to form the M1 macrophages (M-THP1). Following co-culture with CD19 CAR-T, the level of cleaved caspase-1 was significantly upregulated. The increase of active caspase-1 was abrogated when hMSCs were added into the co-culture. knock-down of STC1 led to another reverse and completely blocked the inhibitory function of hMSCs (Figure 3A). Concomitant with the reduction in active caspase-1, the cleaved IL-1β mature form and absent-in-melanoma 2 (AIM2), two key components of the inflammasome (Kelley et al., 2019), were both increasingly expressed following M-THP1 polarization and further incubation with CAR-T (Figure 3A). Compared to the partial inhibition of the active caspase-1 formation, the addition of hMSC in the cell co-culture showed a stronger inhibition of these two proteins, and their expression level was returned to the base level of Pfeiffer plus CAR-T (Figure 3A). This result suggests that the immune-suppressive effect of hMSC was through its impact on macrophages, not CAR-T or Pfeiffer cells. Knockdown of STC1 abrogated the inhibition of hMSC on IL-1β and AIM2 (Figure 3A). The levels of IL-1β in the supernatants measured by ELISA showed similar results as cell lysate (Figure 3B). Figure 3 Download asset Open asset The impact of mesenchymal stem cells (MSCs) on the expression of key components involved in the formation of NLRP3 inflammasome and mitochondrial reactive oxygen species (ROS). (A) The protein expression of IL-1β, caspase-1, and AIM2 in cell lysates was analyzed by Western blot. (B) Quantitation of IL-1β secretion in the supernatants by ELISA. (C) FACS analysis of ROS level and mitochondria mass with fluorescent dye CellROX Deep Red and MitoTracker Green. (D) Quantitation of mitochondria-specific ROS level based on the percentage of cells that were both positive for CellROX and MitoTracker. All samples were collected 24 hr post the co-culture of different cells. For the measurements of IL-β, results are shown as the mean ± SD from three independent experiments (p values are as indicated, n=3). Figure 3—source data 1 Labeled original blots of Figure 3A. https://cdn.elifesciences.org/articles/82934/elife-82934-fig3-data1-v3.zip Download elife-82934-fig3-data1-v3.zip Figure 3—source data 2 Unlabeled original blots of Figure 3A. https://cdn.elifesciences.org/articles/82934/elife-82934-fig3-data2-v3.zip Download elife-82934-fig3-data2-v3.zip Figure 3—source data 3 Figure 3B in Excel file. https://cdn.elifesciences.org/articles/82934/elife-82934-fig3-data3-v3.xlsx Download elife-82934-fig3-data3-v3.xlsx Figure 3—source data 4 Figure 3D in Excel file. https://cdn.elifesciences.org/articles/82934/elife-82934-fig3-data4-v3.xlsx Download elife-82934-fig3-data4-v3.xlsx Mitochondrial dysfunction is one of the major stimuli that activates the NLRP3 inflammasome, and it was reported that exogenous STC1 is internalized by macrophages within 10 min and localizes to mitochondria to suppress superoxide generation (Wang et al., 2009). Therefore, we determined the impact of hMSC on the intracellular level of reactive oxygen species (ROS) and mitochondria mass in macrophages by fluorescent dye CellROX and MitoTracker Green, respectively. As shown in Figure 3C and D, the presence of hMSCsshCtrl markedly suppressed both the cellular and mitochondrial ROS induced by the co-culture of CAR-T cells, tumor cells, and macrophages. Knockdown of STC1 eliminated the function of hMSC in suppressing ROS. This result correlates well with the expression of caspase-1, IL-1β, and AIM, suggesting that hMSCs inhibited NLRP3 inflammasome activation in macrophages was most likely by inhibiting the oxidative burst. hMSCs showed strong inhibition on CD19 CAR-T therapy in xenograft mice, which was abrogated by STC1 knockdown The immune-suppressive impact of hMSC on CAR-T therapy and the function of STC1 were further evaluated in a xenograft model. Upon injection of Pfeiffer cells and confirmation of engraftment, we injected hMSC into the tumor area while applying CAR-T treatment by tail vein injection. As shown in Figure 4A, CD19 CAR-T treatment combined with the injection of hMSCshSTC1 achieved a significant curative effect, and the tumors nearly disappeared at day 38. However, the hMSCshCtrl group showed a continued increase in tumor size and spreading of tumor. Figure 4 Download asset Open asset The inhibition of hMSC on chimeric antigen receptor modified T cell (CAR-T) therapy in xenograft mice relied on stanniocalcin-1. (A) The formation and progression of tumor in three groups of mice monitored with bioluminescence imaging: the control group without any treatment, CAR-T/M-THP1/hMSCsshSTC1 group, and CAR-T/M-THP1/hMSCsshCtrl group. Day 0 was set when the engraftment was confirmed after injecting the Pfeiffer cells. (B) Immunohistochemical analysis of IL-1β, CD4+, CD8+, and Treg cells (using FOXP3 as the biomarker) in tumor tissue at day 10, positive cells display brown or brownish-yellow staining color. (C) The tumor size change with time. (D) The counted average radiance, presented as the mean ± SD (p values are as indicated, n=3). Figure 4—source data 1 Figure 4C in Excel file. https://cdn.elifesciences.org/articles/82934/elife-82934-fig4-data1-v3.xlsx Download elife-82934-fig4-data1-v3.xlsx Figure 4—source data 2 Figure 4D in Excel file. https://cdn.elifesciences.org/articles/82934/elife-82934-fig4-data2-v3.xlsx Download elife-82934-fig4-data2-v3.xlsx Based on the immunohistochemical analysis of IL-1β in tumor tissue on day 10, the number of positive cells (brownish-yellow staining) ranged from 76 to 100% in the hMSCshSTC1 group, while it ranged from 5 to 20% in the hMSCshCtrl group, indicating that hMSC could suppress TME and STC1 knockdown significantly diminished this impact (Figure 4B). Consistent with the results in vitro, a large amount of CD4+ T cells were detected in the hMSCshCtrl group but much less in the hMSCshSTC1 group. On the contrary, the amount of CD8+ T cells was significantly increased in the hMSCshSTC1 group compared to that of the hMSCshCtrl group (Figure 4B). Based on the staining of FOXP3 (forkhead box P3), a master regulator involved in the development of Treg cells, the amount of Treg cells was also evidently increased in the hMSCshCtrl group compared to that of the hMSCshSTC1 group (Figure 4B). These results further confirmed that knockdown of STC1 abrogated the immune-suppressive capability of MSC. The changes in the average radiance were consistent with the changes in the tumor size (Figure 4C and D). The survival time of mice demonstrated that mice in CAR-T combined with the hMSCshSTC1 group had the longest survival with no death by day 38 (Figure 4D). Compared to the control group with no CAR-T treatment, tumor spreading in the hMSCshCtrl group was slower, and all survived for 6 days more. These results confirmed the inhibitory effects of hMSC on CAR-T therapy under in vivo situations and demonstrated that STC1 is an important factor affecting therapy efficacy. Discussion Stem cells are believed to play critical roles in resistance to cancer therapy, which is a major contributor to poor treatment responses and tumor relapse. Previous studies have been mainly focused on the role of cancer stem cells. In the current study, we presented evidences that the presence of MSCs in TME may also be an important source of cancer treatment resistance. By modulating TME, MSCs showed a strong suppressive function on CAR-T efficacy toward lymphoma cells, and interestingly, the presence of the STC1 gene played a critical role. The role of STC1 in cancer is paradoxical. Some reports showed that it exerts an oncogenic role, whereas other studies suggested the opposite (Chen et al., 2019). The aberrant expression of STC1 has been reported to impact various types of cancer, such as triggering tumor angiogenesis by upregulating the expression of VEGF in gastric cancer cells (He et al., 2011), causing tumorigenesis and poor clinical outcomes in ovarian, colorectal, and lung cancers (Yeung et al., 2012; Chen et al., 2019). To date, the potential roles of STC1 in immunotherapy are still largely unknown. Here, we demonstrated that the presence of STC1 is critical for MSC to exert its immunosuppressive role by inhibiting cytotoxic T cell subsets, activating some key immune suppressive/escape mechanisms, and crosstalk with other immune cells. First, a significant downregulation of CD8+ T Cells together with the upregulation of CD4+ T helper cell subsets and Tregs indicated that the suppressed CAR-T efficacy was at least partially associated with MSC’s function in modulating the proliferation of different T-cell subsets. Since the suppression of CD8+ T cells was completely abrogated if knockdown STC1 in MSCs, it is clear that STC1 played a key role here. Moreover, considering that STC1 is secreted into the extracellular matrix in a paracrine manner, MSCs’ modulation of the T cell subsets is most likely indirectly via altered cytokine expression or other secondary molecules activated by STC1. In line with our study, it was recently reported that STC-1 negatively correlates with immunotherapy efficacy and T cell activation by trapping calreticulin, which abrogates membrane calreticulin-directed antigen presentation function and phagocytosis (Lin et al., 2021). The presence of MSCs also stimulated the expression of IDO and PD-L1, two important immune-suppressive molecules. Upregulation of IDO is an endogenous feedback mechanism controlling excessive immune responses, which can be produced both by tumor cells and macrophages (Uyttenhove et al., 2003). IDO-mediated formation of immunosuppressive metabolites can inhibit T-cell proliferation and induce T-cell death through the dioxin receptor (Opitz et al., 2011; Frumento et al., 2002). PD-L1 is a well-characterized molecule of the major escape mechanism of immunotherapy by inhibiting PD-1-mediated effector T cell function and downregulating antigen tolerance (Ribas and Hu-Lieskovan, 2016). There have been numerous studies reporting the bidirectional interactions between MSCs and cancer cells, resulting in regulating the expression of PD-L1 on the surface of various cancer cells or TME (Aboulkheyr and Bigdeli, 2022; Krueger et al., 2019; O’Malley et al., 2018; Sun et al., 2018). Importantly, here we demonstrated that the upregulated expression of both IDO and PD-L1 by MSCs was much reduced if the STC1 gene was k