DC‐CIK cells derived from ovarian cancer patient menstrual blood activate the TNFR1‐ASK1‐AIP1 pathway to kill autologous ovarian cancer stem cells

Abstract Ovarian cancer stem cells (OCSCs) are highly carcinogenic and have very strong resistance to traditional chemotherapeutic drugs; therefore, they are an important factor in ovarian cancer metastasis and recurrence. It has been reported that dendritic cell (DC)‐cytokine‐induced killer (CIK) cells have significant killing effects on all cancer cells across many systems including the blood, digestive, respiratory, urinary and reproductive systems. However, whether DC‐CIK cells can selectively kill OCSCs is currently unclear. In this study, we collected ovarian cancer patient menstrual blood (OCPMB) samples to acquire mononuclear cells and isolated DC‐CIK cells in vitro. In addition, autologous CD44+/CD133+ OCSCs were isolated and used as target cells. The experimental results showed that when DC‐CIK cells and OCSCs were mixed and cultured in vitro at ratios of 5:1, 10:1 and 50:1, the DC‐CIK cells killed significant amounts of OCSCs, inhibited their invasion in vitro and promoted their apoptosis. The qPCR and Western blot results showed that DC‐CIK cells stimulated high expression levels and phosphorylation of TNFR1, ASK1, AIP1 and JNK in OCSCs through the release of TNF‐α. After the endogenous TNFR1 gene was knocked out in OCSCs using the CRISPR/Cas9 technology, the killing function of DC‐CIK cells on target OCSCs was significantly attenuated. The results of the analyses of clinical samples suggested that the TNFR1 expression level was negatively correlated with ovarian cancer stage and prognosis. Therefore, we innovatively confirmed that DC‐CIK cells derived from OCPMB could secret TNF‐α to activate the expression of the TNFR1‐ASK1‐AIP1‐JNK pathway in OCSCs and kill autologous OCSCs.


| INTRODUCTION
Ovarian cancer is a gynaecological cancer with a very high malignancy, and this disease severely endangers the health of women. In recent years, the onset of ovarian cancer has trended towards younger patients, and the malignancy of this cancer has also increased. 1,2 Many studies in recent years have demonstrated the presence of a subpopulation of cells in ovarian cancer tissue samples that are similar to embryonic stem cells and expresses high levels of CD44, CD133 and c-Kit (CD117) markers. These cells have the "stemness" of stem cells but also have the high proliferation, high invasiveness and high tumorigenicity features of cancer cells; therefore, these cells are known as ovarian cancer stem cells (OCSCs). [1][2][3][4] OCSCs have high heterogeneity and very strong resistance to traditional chemotherapeutic drugs; therefore, research and development of drugs that can be used for highly efficient and targeted killing of OCSC are particularly important.
Cytokine-induced killer (CIK) cells refer to the CD3+/CD56+ T lymphocytes that account for only 1%-5% of normal human peripheral blood. Currently, some studies have enriched peripheral blood mononuclear cells (PBMCs) in vitro and induced CIK cell production using cytokines intended for cancer immunotherapy. [5][6][7][8][9][10][11] CIK cells have 2 major heterogeneous cell subpopulations, CD3+/CD56+ and CD3+/CD8+. [5][6][7][8][9] Some studies have indicated that CIK cells can readily recognise cancer cells and have the advantages of both the strong anti-tumour activity of T lymphocytes and the non-MHC-restricted tumour-killing ability of NK cells. [5][6][7][8][9] Dendritic cells (DCs) are professional antigen presenting cells (APCs) that have received attention in recent years. DCs can uptake, process and present antigens to initiate T cell-mediated immune responses. [5][6][7][8][9] DCs are the strongest known and only professional APCs that can activate resting T cells in the body and are the central link for the initiation, regulation and maintenance of immune responses. [5][6][7][8][9] DCs are important components in tumour immunotherapy. [5][6][7][8][9] Some studies have already confirmed that DCs combined with CIK cells (DC-CIK) have significant killing effects on tumour cells across many systems including the blood, digestive, respiratory, urinary and reproductive systems. However, the mechanisms of action behind this phenomenon are currently still unclear. [5][6][7][8][9]11 It has been confirmed that the anti-tumour tumour-killing activity of DC-CIK cells is not affected by immunosuppressive agents (such as cyclosporine and FK506), and DC-CIK cells have very minimal toxicity against normal bone marrow haematopoietic precursor cells (less than 30%). [5][6][7][8][9] The DOC2/DAB2 (DAB2IP) gene is localised to the 9q33.1-q33.3 region of human chromosome 9. This gene is a tumour suppressor, and its inactivation is closely associated with tumours in the reproductive system. [12][13][14][15] The C2 domain of the DAB2IP protein reportedly interacts with the apoptosis signal-regulating kinase 1 (ASK1) protein; therefore, the DAB2IP protein is also known as ASK1-interacting protein (AIP). [12][13][14][16][17][18] Di Minin et al showed that when the TNF-a-mediated activation of the ASK1/JNK pathway was inhibited by NFjB, a mutant p53 was activated. This effect depended on the activation of the cytoplasmic DAB2IP inhibited by mutant p53. 16 In addition, some studies have reported that the ASK1-AIP1-JNK/p38MAPK pathway is downstream of TNF-a. External TNF-a can stimulate the activation of the intracellular ASK1-AIP1-JNK/p38MAPK pathway to eventually induce cell apoptosis or necrosis. 19,20 Some studies have already confirmed that many types of cells are present in menstrual blood, including mononuclear cells and endometrial stem cells. [21][22][23][24][25] Our previous studies confirmed that transplantation of endometrial mesenchymal stem cells derived from menstrual blood could treat premature ovarian failure in mice. 22 Furthermore, endometrial stem cells derived from menstrual blood could also be directionally induced to differentiate into various somatic cells, including neurons, osteoblasts and adipose cells, in vivo and in vitro. [21][22][23][24][25] Therefore, cells in menstrual blood have many potential biological functions.
Based on the above findings, we proposed a scientific hypothesis. Mononuclear cells were collected from ovarian cancer patient menstrual blood (OCPMB) in order to prepare DC-CIK cells in vitro.
Whether these cells were capable of targeted killing effects of autologous OCSCs was investigated. In addition, the relevant mechanism was studied in depth to examine whether DC-CIK cells derived from menstrual blood exerted killing effects on OCSCs by targeting the TNFR1-ASK1-AIP1-JNK pathway.

| Isolation and culture of tissue-derived OCSCs
The experiment was performed according to previously described methods. 2 Briefly, surgically isolated tissues from 6 ovarian cancer patients (Table S1)   2.2 | Isolation and stimulation of DC and CIK cells derived from OCPMB A sterile one time menstrual cup (S-Evans Biosciences) was used to collect menstrual blood. The patient was first in the aseptic ward. The aseptic menstrual cup was then removed from the sealed package. In order to ensure that menstrual cup was used under sterile condition, before collecting menstrual blood, 75% ethyl alcohol was used to wipe menstrual cup for 3 times. Then, the whole menstrual cup was put into the vagina of the patient after it was completely dried. After the menstrual blood was collected, the menstrual cup was removed in the sterile environment and collected menstrual blood in the aseptic biological safety cabinet. The experiment was performed according to previously described methods. 2 Briefly, 10 mL of menstrual blood was collected from each of the 6 ovarian cancer patients. PBMCs were isolated from the menstrual blood sample using the Cobe Spectra Apheresis System (CaridianBCT, USA). For the induction of CIK cells, PBMCs were cultured in the serum-free X-VIVO 20 culture medium (Cambrex, USA) supplemented with 50 ng/mL anti-human CD3e antibody (e-Bioscience), 100 U/mL recombinant human interleukin (rhIL)-1a (Gibco) and 1000 U/mL recombinant human interferon (rhIFN)-c (Gibco) at 37°C and 5% CO 2 for 24 hours; subsequently, 300 U/mL recombinant human IL-2 (rhIL-2, Gibco) was added.
For the induction of DCs, PBMCs derived from menstrual blood were cultured in serum-free X-VIVO 20 culture medium supplemented with 1000 U/mL rhIL-4 (Gibco) and 500 U/mL granulocyte/macrophage colony-stimulating factor (GM-CSF) (Gibco) at 37°C and 5% CO 2 for 7 days continuously; subsequently, autologous tumour lysates (100 lg/mL) were added and co-cultured for 24 hours to activate the DCs.  Table S2.

| Western blot
Briefly, total proteins from the cells in each group were subjected to 12% denaturing sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were then transferred onto a polyvinylidene fluoride (PVDF) membrane (Millipore). The membrane was blocked, washed and incubated with primary antibodies at 37°C for 45 minutes (Table S3). After the membrane was fully washed, it was incubated with secondary antibodies at 37°C for 45 minutes (Table S3).
The membrane was washed with Tris-buffered saline/Tween 20 (TBST) 4 times at room temperature for 14 minutes each time. Next, the samples were exposed and imaged (Sigma-Aldrich Chemical) using the enhanced chemiluminescence (ECL) method (Pierce Biotechnology).
The steps were performed according to the manufacturer's instructions for the plasmid. Briefly, the above plasmid was transfected into cells using the Ultracruz â transfection reagent (Santa Cruz).

| Statistical analysis
Each experiment was performed as least 3 times, and data are shown as the mean AE SE where applicable, and differences were evaluated using Student's t tests. The probability of P < .05 was considered to be statistically significant.

| The TNFR1 expression level negatively correlated with ovarian cancer stage and prognosis
To explore the relationship between the TNFR1 expression level and ovarian cancer stage and prognosis, tissue samples from 10 cases of high-grade and 10 cases of low-grade ovarian cancer patients were collected. The qPCR detection results showed that TNFR1 expression in low-grade ovarian cancer tissues was significantly higher than that in high-grade ovarian cancer tissues, and this difference was statistically significant (Figure 8). In addition, analyses of prognosis showed that the survival time of ovarian cancer patients with high TNFR1 expression was significantly longer than that of those with low TNFR1 expression (Figure 8). The experimental results indicated that the TNFR1 expression level negatively correlated with ovarian cancer stage and prognosis.

| DISCUSSION
Ovarian cancer stem cells (mainly the CD44+/CD133+ or the CD44+/CD117+ subpopulations) were discovered in ovarian cancer F I G U R E 2 OCPMB-derived DC-CIK cells significantly killed CD44+/CD133+ OCSCs. A, The cytotoxicity experiment results showed that the OCSC killing rates of CIK and the OCSC killing rate of DC-CIK cells were both significantly higher than those in the simple OCSC group and the DC-OCSC group at the 10:1 and 50:1 ratios; B, DC-CIK cells and OCSCs were mixed at a ratio of 10:1 and cultured in vitro for 24 h. FACS results of Annexin V/PI staining showed that the total apoptosis rate in the DC-CIK-OCSC group was significantly higher than in the simple OCSC group and the DC-OSCS group; C, The transwell invasion experiment results suggested that the number of migrated cells in the DC-CIK-OCSC group was significantly lower than those in the simple OCSC group and the DC-OCSC group. Scale bar = 30 lm  The ELISA results suggested that the levels of IFN-c, TGFb and TNF-a in the cell culture medium in the DC-CIK-OCSC group were significantly higher than those in the OCSC group, the simple CIK-OCSC group and the DC-OCSC group; C, The qPCR results showed that the mRNA expression levels of Tnfr1, Ask1, Aip1, Jnk and caspase-3 in OCSC from the DC-CIK-OCSC group and the DC-OCSC group were significantly higher than those in the simple OCSC group; however, the mRNA expression levels of the above genes in the DC-CIK-OCSC group were significantly higher than that in the DC-OCSC group. The expression level of the nuclear proliferation factor mKi67 in the DC-CIK-OCSC group was significantly lower than that in the DC-OCSC group and the simple OCSC group. **P < 0.01 vs OCSCs; *P < 0.05 vs OCSCs; n = 6 indicated that an effective combination of DC-CIK cells can exhibit the common high-efficient killing functions that are present in many types of tumours. [5][6][7][8][9] In the DC-CIK treatment, DCs act like "radar" that can recognise antigens to activate the immune response, and CIK cells are similar to a "missile" that can exert cytotoxicity and secret cytokines to accurately kill tumour cells.
Therefore, adoptive immunotherapy of malignant tumours using Bye the way, there are many advantages of DC-CIK amplified from menstrual blood when compared with PBMC from peripheral F I G U R E 6 TNFR1 knockout OCSCs were resistant to OCPMB-derived DC-CIK cells. A, The qPCR results showed that the expression level of endogenous TNFR1 in the OCSC group transfected with the human TNFR1 CRISPR/Cas9 KO plasmid was significantly lower than that in the group transfected with an empty plasmid. **P < .01 vs OCSCs Tnfr1 WT; n = 4; B, Immunofluorescence staining results showed that OCSCs in the TNFR1KO group barely expressed any TNFR1 protein. Scale bar = 30 lm; C, The Western blot results showed that the expression of endogenous TNFR1 protein in the OCSCs from the TNFR1KO group was significantly lower than that in the TNFR1WT group; D, The cytotoxicity experiment results showed that the cytotoxicity of DC-CIK cells on OCSCs in the TNFR1KO group was significantly lower than that in the control group. **P < .01 vs OCSCs Tnfr1 WT; n = 4; E, The qPCR results showed that the mRNA expression levels of Tnfra, Ask1, Aip1, Jnk and caspase-3 in OCSCs in the DC-CIK-TNFR1KO group were significantly lower than that in the DC-CIK-TNFR1WT group; F, The transwell experiment results showed that the number of migrated OCSCs in the DC-CIK-TNFR1KO group was significantly higher than that in the control cell group. **P < .01 vs OCSCs Tnfr1 WT; n = 4. Scale bar = 30 lm; G, Western blot results showed that the expression levels of AIP1, p-ASK1, JNK, p-JNK, p38MAPK, p-p38MAPK and activated caspase-3 in OCSCs in the DC-CIK-TNFR1KO group were significantly lower than those in the DC-CIK-TNFR1WT group. **P < .01 vs OCSCs Tnfr1WT; n = 4 blood. First, for women, the source of menstrual blood is stable.
An adult female will exclude more than 20 mL of menstrual blood every month. In view of our study, we found that the mononuclear cells contained in above menstrual blood are sufficient to induce differentiation into DC-CIK cells in vitro for the treatment of tumours. Therefore, for adult women, menstrual blood collection and DC-CIK cells induction could be carried out weekly, so that a large number of DC-CIK cells could be obtained. Secondly, the method of menstrual blood collecting is very simple and safe.
Usually, the venous puncture must be carried out to collect the F I G U R E 7 TNFR1 knockout OCSCs readily formed tumours in nude mice. A, The in vivo tumorigenicity experiments in nude mice showed that the visual sizes of tumours in nude mice that developed from OCSC in the DC-CIK-TNFR1KO group were significantly larger than those that developed from the DC-CIK-TNFR1WT group; B, The volumes of tumours that developed from OCSCs in the DC-CIK-TNFR1KO group were significantly larger than those that developed from the DC-CIK-TNFR1WT group. **P < .01 vs OCSCs Tnfr1 WT; n = 4; C, The weights of tumours that developed from OCSCs in the DC-CIK-TNFR1KO group were significantly higher than those that developed from the DC-CIK-TNFR1WT group. **P < .01 vs OCSCs Tnfr1 WT; n = 4; D, Pathological analysis showed that tumours that developed from these 2 groups of cells were all mixed-type epithelial ovarian cancer. Scale bar = 30 lm; E, The immunofluorescence staining results showed that the expression of ki67 protein in tumours that developed from OCSCs in the DC-CIK-TNFR1KO group was higher than that in tumours that developed from the DC-CIK-TNFR1WT group. Scale bar = 30 lm F I G U R E 8 The TNFR1 expression level negatively correlated with ovarian cancer stage and prognosis. A, The qPCR results showed that TNFR1 expression in lowgrade ovarian cancer tissues was higher than that in high-grade ovarian cancer tissues, and this difference was statistically significant; B, The survival time of ovarian cancer patients with high TNFR1 expression was significantly longer than those with low TNFR1 expression; C, OCPMB-derived DC-CIK cells activated the TNFR1-ASK1-AIP1-JNK pathway to kill autologous OCSCs through the secretion of TNF-a peripheral blood. However, the collection of menstrual blood does not require invasive puncture behaviour. The collection of menstrual blood only needs to put the sterile disposable menstrual cup into the vagina. This operation is simple and safe. Any women who have the ability to take care of themselves can do their own menstrual blood collection. As the collection of menstrual blood does not need to be venepuncture, the risk of infection is greatly reduced.
In summary, this study proposes a new route for acquiring DC-CIK cells using PBMCs collected from OCPMB to prepare highefficient DC-CIK cells. In addition, this study also provides clues for evaluating the efficacy of DC-CIK cells; when DC-CIK immunotherapy is required, whether the TNFR1 molecule is expressed on the surface of cancer cells should be assessed beforehand.