The correlations between CXCL12, CXCR4, EAAT1 and GS in malignant pleural mesothelioma, and CXCL12 regulates cell invasion and migration via CXCR4/EAAT1/GS pathway Author names and affiliations

Purpose: To elucidate the mechanism of CXCR4/EAAT1/GS pathway in CXCL12 regulating invasion and migration in malignant pleural mesothelioma (MPM). Methods: Immunohistochemistry for CXCL12, CXCR4, EAAT1 and GS stainings and correlation analysis between them were conducted in MPM and normal tissues. Western blot and real-time PCR were performed to examine the CXCR4, EAAT1 and GS expression in H2052 cells. Wound healing and transwell assay were applied to determine the cell migration and invasion. MTT was utilized to assess cell viability. Results: CXCL12, CXCR4, EAAT1 and GS were highly expressed in MPM tissues and correlated with each other. CXCL12 upregulated both in protein and mRNA levels of CXCR4, EAAT1 and GS in H2052 cells. The EAAT1 and GS expression upregulated or not by CXCL12 were decreased by CXCR4 and EAAT1 knockdown. CXCR4 antagonist AMD3100 and EAAT1 antagonist TFB-TBOA also resulted in the same effects as CXCR4 and EAAT1 knockdown, respectively. CXCL12 promoted cell invasion and migration and increased the Matrix metalloproteinase 9 (MMP9) mRNA level. CXCR4 and EAAT1 knockdown suppressed all these functions. Furthermore, CXCL12 promoted H2052 cells growth in nude mice, both AMD3100 and TFB-TBOA inhibited this promotion. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 17 September 2020 doi:10.20944/preprints202009.0380.v1


Introduction
Malignant pleural mesothelioma (MPM) is a relatively rare and aggressive cancer, originating from mesothelial surfaces of the pleural lining. The prognosis of MPM is poor for decades due to its nonspecific clinical manifestations and the relatively late diagnosis [1]. Treatments for MPM in the past rarely extend patients survival beyond 12 months [2,3]. However, recent studies have elucidated genetic alterations in mesothelioma. Some genes have been found to be significantly espressed in MPM, such as BAP1, NF2, TP53, SETD2, SMARCC1, PBRM1, SF3B1 and TRAF7 [4][5][6][7]. Pastorino et al. [8] reported that a subset of MPM patients who were not aware of asbestos exposure and carried pathogenic germline mutations of BAP1 had significantly improved survival. Nevertheless, recent clinical trials targeting molecular pathways have failed to significantly improve the prognosis of MPM [9]. Therefore, a better understanding of the cytogenetic and molecular pathogenesis of MPM is still needed to develop effective therapeutic approaches for the MPM patients [10][11][12].
Chemokines are a class of chemotactic cytokines that are categorized based on the positioning of well-conserved N-terminal cysteine residues and classified as C, CC, CXC, CX3C chemokines. Chemokines play a critical role in many biological processes, including hematopoiesis, immune cell trafficking and stem and progenitor cell mobility by activating their corresponding chemokine receptors, which belong to the rhodopsin-like seven transmembrane G protein-coupled receptor (GPCR) family [13,14]. Among these receptors, CXCR4, a kind of very important receptor, is widely expressed in various types of tissues, where it plays a central role in development-related processes [15,16] and tumor metastasis.
CXCR4 is the specific chemokine receptor which can interact with the chemokine Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 17 September 2020 doi:10.20944/preprints202009.0380.v1 4 of 42 CXCL12, also known as Stromal-Derived Factor-1 (SDF-1) ligand [17]. CXCL12 was expressed in breast tumor, brain tumor and colorectal cancer tissues [18][19][20], etc. and also usually expressed in fibroblasts and endothelial cells [21,22] or produced by stromal cells of lymph nodes, lung, liver, and bone marrow, which were the most commom sites for metastasis [23]. CXCL12/CXCR4 axis is also a key factor in the tumor development, metastasis and cross -talking between tumor cells and their microenvironment of various malignancies [24,25]. This axis not only promotes the survival and proliferation of cancer cells but also attracts the CXCR4-expressed tumor tissue/organ-specific metastasis [26,27].
In addition, previous work suggests that CXCL12 and CXCR4 are involved in MPM tumorigenesis [28,29].
Glutamate (Glu) is an important excitatory neurotransmitter of mammalian central nervous system (CNS) [30]. The presynaptic concentration of Glu in human CNS is regulated by Glutamate transporters (excitatory amino acid transporters, EAATs) [31], which utilize the transmembrane gradients of Na + , K + , and H + as a driving force to uptake glutamate from extracellular space into cells [32], and subsequently converted into glutamine by glutamine synthetase (GS), which is a glia -specific enzyme [33,34], and transported back to the neurons for recycling into glutamate. To data, five EAATs (EAAT1 ~ 5) have been characterized [35,36]. Evidence has suggested that EAATs play a critical role in maintaining extracellular glutamate homeostasis and regulating glutamate neurotransmission [37][38][39]. Among these EAATs, EAAT1 is a ubiquitous subtype. GS is the only key enzyme that is capable of de novo synthesis of glutamine and also functions to detoxify glutamate and ammonia [40], which is involved in nitrogen metabolism [41]    As shown in Table 4, CXCL12, CXCR4, EAAT1 and GS have no strong correlations with each other in normal pleura samples, In addition, the positive staining areas of CXCL12, CXCR4, EAAT1 and GS were similar in some of other mesothelioma tissue samples. CXCR4 staining ------

The effects of CXCL12 on the CXCR4, EAAT1 and GS expression
In order to determine an optimal time point for the CXCL12 stimulation.  To explore the potential regulatory effects of CXCL12/CXCR4 axis on EAAT1/GS.
Next, H2052 cells were transfected with CXCR4 siRNA or EAAT siRNA alone, in the presence or absence of CXCL12 treatments or vehicle. As shown in Figure 2B, CXCR4 and EAAT1 silencing apparently suppressed CXCL12-upregulated the levels of EAAT1  Figure 2C). Similarly, as shown in Figure 3B, QPCR results showed that TFB obviously inhibited CXCL12-increased the expression of CXCR4, EAAT1 and GS mRNA.

The effects of CXCL12 on cell invasion and migration
The transwell assay was carried out to determine the ffects of CXCL12 on the cell invasion and migration. Exposure to various concentrations of CXCL12 (0 -100 ng/mL) promoted the invasion ( Figure 4A) and migration ( Figure 5A) in H2052 cells in a does dependent manner and both of them reached their highest peak in 100 ng/mL. In addition, MTT was also performed to evaluate the cell proliferation. As shown in Figure S2A and S2B, CXCL12 (0 -100 ng/mL) gradually promoted the proliferation in H2052 cells.
AMD3100 (0 -10 µ M) gradually inhibited the proliferation in H2052 cells ( Figure S2C). Representative images from different groups were shown in left panal and the corresponding quantitative analysis was also presented in right panal. * p < 0.05, ** p < 0.01 versus control. # p < 0.05 versus CXCL12 treatment, as indicated. The data were expressed as mean ± SD from repeated three independent experiments.
To further define the effects of CXCR4 and EAAT1 on the cell invasion and migration, H2052 cells were transfected with CXCR4 siRNA or EAAT siRNA alone, in the presence or absence of CXCL12 treatments or vehicle. The results showed that CXCR4 siRNA or EAAT1 siRNA significantly suppressed CXCL12-promoted the H2052 cells invasion. CXCR4 siRNA or EAAT1 siRNA alone also slightly inhibited H2052 cells invasion ( Figure 4B). Similar effects were also observed in H2052 cells migration ( Figure   5B). Furthermore, H2052 cells were treated with the CXCL12, AMD3100 alone, AMD3100 followed by CXCL12. The results showed that CXCL12-promoted the cells invasion remarkably suppressed by AMD3100 (Figure 4C). Similar effects were also observed in H2052 cells migration (Figure 5C). These results indicated that CXCL12 regulated the invasion and migration also through CXCR4/EAAT1 pathway in H2052 cells. Representative images from different groups were shown in left panal and the corresponding quantitative analysis was also presented in right panal. * p < 0.05, ** p < were expressed as mean ± SD from repeated three independent experiments.
In addition, wound healing assay in vitro was also performed to assess the cell migration ability. H2052 cells were treated with the indicated agents for 24 h and 48 h. As shown in Figure S3A and 3B, the results were consistent with H2052 cells invasion and migration by transwell assay.

QPCR for MMP9
As is known to all, MMPs are frequently expressed in nearly all tumors, where they accelerates tumor progression. Among these MMPs, MMP9 is particularly known to play an important role in tumor progression, including tumor angiogenesis and growth as well as metastasis to distant organs [45,46]. Therefore, we focused on MMP9 and detected its expression in H2052 cells invasion. The result showed that CXCL12 slightly promoted the MMP9 mRNA level in 5 min ( Figure S4). Moreover, H2052 cells were transfected with CXCR4 siRNA or EAAT siRNA alone, in the presence or absence of CXCL12 treatments or vehicle. As shown in Figure 6A, the CXCL12 significantly increased the MMP9 mRNA level in 24 h, while this effect was obviously suppressed by CXCR4 siRNA or EAAT1 siRNA. CXCR4 siRNA or EAAT1 siRNA alone also decreased the MMP9 mRNA level, respectively. Similarly, AMD3100 also inhibited CXCL12 -increased the MMP9 mRNA level ( Figure 6B). All these data indicates that MMP9 is involved in the cell invasion.

In vivo experiments
Two weeks after inoculation, the tumors were obviously visible on more than half of the mice's back ( Figure S5A).

Discussion
Chemokines are thought to play a critical role in tumor growth, invasion and progression [47]. Previous studies have showed that there is obvious correlation between multiple chemokine receptors in many kinds of human malignant tumor metastases, including gastric carcinoma [48], T-cell leukemia [49], osteosarcoma [50], colorectal cancer [51,52], hepatocellular carcinoma [53], ovarian cancer [54], malignant melanoma [55,56], and renal cell carcinoma [57]. Our previous studies [43,44]  GS proteins compared to control siRNA, but conversely, EAAT1 silencing alone failed to decrease the CXCR4 protein level compared to control siRNA, which indicates that EAAT1 is a downstream factor of CXCR4 and GS is a downstream factor of EAAT1. It is noticeable that according to our results (Figure 2B), EAAT1 silencing conversely slightly inhibited CXCL12-upregulated the expression of CXCR4. This suggests that one possible mutual regulatory relationship exists between them. However, the impact of CXCR4 on EAAT1 is greater than that of EAAT1 on CXCR4.
Moreover, to assess the role of EAAT1 in invasion and migration of MPM cells, we used specific EAAT1 siRNA to kncockdown the EAAT1 expression and found that CXCL12promoted the H2052 cells invasion and migration were significantly suppressed by EAAT1 silencing, and to a certain extent, EAAT1 silencing alone also inhibited H2052 cells invasion and migration. Besides, CXCL12-increased the expression of MMP9 mRNA was remarkably inhibited by EAAT1 silencing. These results denoted that CXCL12 regulated the invasion and migration also through CXCR4/EAAT1/GS pathway in H2052 cells.
Our recent study showed that AMD3100 could inhibit the growth of MPM in vivo [68]. In this study, the results showed that CXCL12 promoted the growth of MPM in vivo.
AMD3100 also inhibited the growth of MPM in vivo, which was consistent with our Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 17 September 2020 doi:10.20944/preprints202009.0380.v1 23 of 42 previous report. This indicates that the progression of MPM is associated with CXCL12/CXCR4. Besides, in vivo results showed that TFB treatment also inhibited the growth of MPM in vivo. All these results suggested that CXCL12/EAAT1 may be another axis that affects the cell proliferation, invasion and migration in MPM cells. The CXCL12 binding the CXCR4 receptor activated divergent signals on multiple pathways, such as ERK1/2, p38, SAPK/JNK, AKT, mTOR [17], which reminds us that CXCL12/EAAT1 axis proposed also may activate certain signaling pathway molecules in cancer cell, and thus represents a potential target for cancer treatment.
In sum up, Although we obtained the first evidance that EAAT1 is a downstream factor of CXCR4 and GS is a downstream factor of EAAT1. CXCL12 regulated the invasion and migration through CXCR4/EAAT1/GS pathway in H2052 cells. There are still some limitation in our research, for example, CXCL12/CXCR4-or CXCL12/EAAT1mediated downstream signaling pathways involved in cell invasion and migration needs further study. Therefore, our next research will attempt to explore the molecular signal pathways and mechanisms involved. It will bring us a deeper insights of clinical significance and provide us with many potential candidate targets and theoretical basis for drug development. according to the manufacturer's instructions. The IHC, HE staining procedures and the scoring system were carried out referring to our previous study [43].

Western blot
The H2052 cells in 6-well culture plates were washed three times with cold PBS and   Table S1.

Cell invasion and migration
Transwell

Statistical analysis
Statistical analysis was carried out with SPSS 21.0 (SPSS Inc, Chicago, USA). Data were expressed as mean ± S.D. Student's t-test or one way ANOVA was used to detect the differences between different groups. For correlation analysis, 0, 1, 2 and 3 represented -, +, ++ and +++, respectively during the calculation. P < 0.05 was considered to be of significance.

Conclusions
CXCL12 regulated the invasion and migration through CXCR4/EAAT1/GS pathway in H2052 cells.

Patents
Fresh primary human MPM samples from 41 patients were obtained (the patients diagnosed with MPM underwent surgical removal of the primary tumor). The normal pleura tissue samples were obtained simultaneously to serve as controls from seven of these patients. We  The data were expressed as mean ± SD from quadruplicates and repeated three times independently.