Inhibition of CYR61-S100A4 Axis Limits Breast Cancer Invasion

Background and Objective: Matricellular proteins modulate the micro environment of tumors and are recognized to contribute to tumor cell invasion and dissemination. The cysteine-rich angiogenic inducer 61 (CYR61) is upregulated in mesenchymal transformed and invasive breast cancer cells. CYR61 correlates with poor prognosis of breast cancer patients. The signaling mechanism that causes invasive properties of cancer cells regarding to epithelial-mesenchymal transition (EMT) needs further research. In this study, we investigated the signaling mechanism, which is responsible for reduced cell invasion after suppression of CYR61 in mesenchymal transformed breast cancer cells and in triple negative breast cancer cells. Methods: We addressed this issue by generating a mesenchymal transformed breast cancer cell line using prolonged mammosphere cultivation. Western blotting and quantitative PCR were used to analyze gene expression alterations. Transient gene silencing was conducted using RNA interference. Proliferation was assessed using AlamarBlue assay. Invasiveness was analyzed using 2D and 3D invasion assays. Immune-histochemical analysis of patient tissue samples was performed to examine the prognostic value of CYR61 expression. Results: In this study, we investigated whether CYR61 could be used as therapeutic target and prognostic marker for invasive breast cancer. We discovered an interaction of CYR61 with metastasis-associated protein S100A4. Suppression of CYR61 by RNA interference reduced the expression of S100A4 dependent on ERK1/2 activity regulation. Non-invasive breast cancer cells became invasive due to extracellular CYR61 supplement. Immune-histochemical analysis of 239 patient tissue samples revealed a correlation of higher CYR61 and S100A4 expression with invasive breast cancer and metastasis. Conclusion: Our data suggest that suppression of CYR61 impedes the formation of an invasive cancer cell phenotype by reducing ERK1/2 phosphorylation thereby suppressing S100A4. These findings identify mechanisms by which CYR61 suppresses cell invasion and suggest it to be a potential therapeutic target and prognostic marker for invasive breast cancer and metastasis.


INTRODUCTION
In 2019, approximately 271270 woman and men in the United States will be diagnosed with breast cancer. Due to improved early detection techniques and treatment options 5year-survival rates for local and regional breast cancer are 84-99 %. However, only 27of patients diagnosed with distant metastasis survive a period of 5 years (1). Consequently it is necessary to identify prognostic markers for the early detection of breast cancer metastasis and new treatment options for this indications which accounts for more than 90% of cancer related death (2).
The first key event in the multi-step process of metastasis is the separation of tumor cells from the primary tumor and the dissemination into the surrounding tissue. Cells gain the ability to migrate and invade by altering their cytoskeletal organization, cell-cell-contacts, contacts with the extracellular matrix (ECM) and surrounding stroma (3). Epithelial-mesenchymal transition (EMT) is a transient dynamic program induced by different transcription factors (TFs). EMT-TFs orchestrate tumor-promoting micro environmental changes, cancer cell stemness, and chemo resistance (4,5). The contribution of EMT programs to the metastatic cascade regarding breast cancer is supported by several publications (6)(7)(8). However, it is still under debate if an involvement of EMT programs is indispensable for creating an invasive phenotype (4).
Therefore it is necessary to study cancer cell invasion with regards to EMT complexity (9,10).
The cysteine rich angiogenic inducer (CYR61) belongs to the CCN family (CYR61, CTGF /CCN12, NOV/CCN3, WISP-1/CCN4, WISP-2/CCN5, WISP-3/CCN6) of matricellular proteins and is localized on cell surface, cytoplasm and as a secreted protein in the extracellular matrix. The functions of CYR61 are cell type and context-dependent (11). They are transmitted through binding to integrin and heparin sulfate proteoglycans (HSPGs). CYR61 was shown to be involved in facilitating EMT programs in different cancer entities (12)(13)(14). It is known that elevated CYR61 expression promotes tumor progression, proliferation, migration and invasion of breast cancer (15,16), whereas the role of CYR61 in breast cancer EMT programs remains elusive. Otherwise, CYR61 can act as a tumor suppressor in non-small cell lung cancer (17) and in fibroblasts by inducing apoptosis and senescence during wound healing (18,19). The role of CYR61 signaling in cancer invasion and EMT programs regarding to a potential use as therapeutic target and prognostic marker needs further evaluation.
We hypothesize that CYR61 is a key regulator of breast cancer invasion. We want to identify the mechanisms by which CYR61 facilitates an invasive phenotype. Furthermore, we want to investigate the value of CYR61 as a therapeutic target and prognostic marker for invasive and metastatic breast cancer.

Cell Lines and Cell Culture
Human breast cancer cell lines MCF-7, T47D, MDA-MB-231, and HCC1806 were obtained from American Type Cell Collection (ATCC; Manassas, VA, USA) and cultured in minimum essential medium (MEM; biowest, Nuaillé, France) supplemented with 10% fetal bovine serum (FBS; biochrom, Berlin, Germany), 1% Penicillin/Streptomycin (P/S; Gibco, Carlsbad, CA, USA), 0,1% Transferrin (Sigma, St. Louis, USA) and 26 IU Insulin (Sanofi, Frankfurt, Germany). Human osteosarcoma cell line MG-63 was purchased from ATCC and cultured Dulbecco's modified eagle medium (DMEM; Gibco) supplemented with 10% FBS (biochrom) and 1% Penicillin/Streptomycin (Gibco). To retain identity of cell lines, purchased cells were expanded and aliquots were frozen in liquid nitrogen. A new frozen stock was used every half year and Mycoplasma testing of cultured cell lines was performed routinely using PCR Mycoplasma Test Kit I/C (PromoCell GmbH, Heidelberg, Germany). All cells were cultured in a humidified atmosphere with 5% CO 2 at 37 • C.

Transwell Invasion Assay
Using co-culture transwell assay as described earlier (21), 1 × 10 4 breast cancer cells were seeded in DMEM w/o phenol red (Gibco), supplemented with 10% cs-FCS into a cell cultural insert (upper well) with a polycarbonate membrane (8 µm pore diameter, Merck Millipore, Cork, Ireland) coated with 30 µL of a Matrigel R (BD Bioscience, Bedford, MA, USA) solution (1:2 in serum-free DMEM). The osteosarcoma cells were seeded (2.5 × 10 4 ) in DMEM supplemented with 10% cs-FCS into the lower well (24-well-plate). After 24 h cells were co-cultured for 48 h or 96 h. Stably transfected cells (overexpressing CYR61 or S100A4) were seeded at a density of 1 × 10 4 per well in DMEM w/o phenol red cell cultural insert (upper well, Matrigel-coated with a polycarbonate membrane), with the lower well containing DMEM w/o phenol red supplemented with 10% cs-FBS and cultured for 96 h. Invaded cells on the lower side of the insert were stained with hematoxylin and the number of cells in four randomly selected fields of each insert was counted.

3D Spheroid Assay
Assessment of 3D cell invasion was pursued as describes earlier with minor changes (22). Briefly 1 × 10 3 breast cancer cells were seeded in 100 µL in a well of an ultra-low-adherence 96-well plate (ULA; Nexcelom, Cenibra GmbH, Bramsche, Germany). After 48 h spheroid formation was visually confirmed and 50 µL of media was removed. Thereafter 50 µL Matrigel were added to the spheroid wells. Central position of the spheroids was checked visually and Matrigel was allowed to solidify for 1 h at 37 • C and 5% CO 2 . Afterwards 50 µL media were added to each well and a picture was taken marking time point 0 (t0h). When indicated 1µg/ml rhCYR61 or 10µM U0126 were added. Spheroid growth area was analyzed using ImageJ polygonal selection and measurement. Mean values were calculated and compared to respective control.

ECM Degradation
Wells of a 96-well plate were coated at room temperature for 20 min with 0.05 mg/ml Poly-L-lysine in DPBS (Sigma) and 15 min with glutaraldehyde 0.5% in DPBS. Gelatin (2 mg/ml; G9391; Sigma) was FITC conjugated as recommend by manufacture (#343210; EMD Millipore Corp., Billeria, MA, USA). Wells were coated with 60 µL FITC-conjugated gelatin (2 mg/ml; Invitrogen, Milpitas, CA, USA) diluted 1:5 with unlabeled gelatin (Sigma) and incubated for 10 min at RT. Solution was discarded and wells were incubated for 30 min in 70% ethanol and afterwards free aldehydes were quenched with culture media for 30 min at room temperature before cells were seeded. Cells were seeded (4.4 × 10 3 cells per Well) and treated with rhCTGF (1 µg/ml; R&D systems). After 24 h proteolytic activity was detected by measuring fluorescence (extinction 490 nm/emission 520 nm) using Synergy (BioTek Instruments, Bad Friedrichshall, Germany). Each experiment was performed in duplicates for at least three times. Mean values were compared to the respective control.

AlamarBlue Assay
3D spheroids were grown as described above and 48 h after adding Matrigel AlamarBlue (BioRad, Hercules, USA) was added and incubated for 4 h at 37 • C 5% CO 2 . The colorimetric change of resazurin to resorufin upon cellular metabolic reduction was measured by absorbance reading at 540 nm and 630 nm, using Synergy (BioTek Instruments). Relative AlamarBlue Reduction was calculated as indicated by manufacturer.

KM Plotter Analysis
The database of the Kaplan-Meier plotter (www.kmplot. com) downloads information of gene expression and overall survival from Gene Expression Omnibus (GEO; only Affymetrix microarrays), the European Genome-Phenome Archive (EGA) and The Cancer Genome Atlas (TCGA). To be able to analyze the prognostic value (overall survival) of CYR61 in 1,402 patient samples, the samples were split into two cohorts according to the expression of quantiles of CYR61 where all possible cutoff values between the lower and the upper quantiles are computed and the best performing threshold is used as a cutoff. These groups are compared by a Kaplan-Meier survival plot and the hazard ratio with 95% confidence intervals. Redundant samples were removed, biased arrays excluded and the proportional hazard assumption was set to zero (24).

Statistical Analysis
All experiments were performed at least in three biological and technical replicates. Data were analyzed by GraphPad Prism (GraphPad software Inc., v. 7.03, La Jolla, Ca, USA) using unpaired, two-tailed, parametric t-test comparing two groups (treatment to respective control) by assuming both populations have the same standard derivation. P < 0.05 was considered statistically significant.

CYR61 and S100A4 as Prognostic Markers for Invasive and Metastatic Breast Cancer
To assess the value of CYR61 and S100A4 as prognostic marker meta-analysis were conducted. Reduced CYR61 expression increases the probability of distant-metastasis free survival (DMFS) of breast cancer patients with a lymph node positive status ( Figure 5A; dataset 213226_at; n = 382; FDR 1%; P = 2.8 e −07 ). Reduced S100A4 expression increases the probability of DMFS of breast cancer patients with a lymph node positive status but shows a higher FDR (Figure 5B; dataset 203186_s_at; n = 382; FDR > 50%; P = 0.024, cut-off values see Figure 5, Figure S6). Analyzing the effects of decreases CYR61 or S100A4 expression with regards to the relapse free survival (RFS) shows comparable results (Figures 5C,D; CYR61: dataset 213226_at; n = 1133; FDR 1%; P = 6.8 e-09 ; S100A4: dataset 203186_s_at; FDR > 50%; P = 0.0012, cut-off values see Figure 5, Figure S6). These data demonstrate that CYR61 could act as a prognostic marker in breast cancer.
CYR61 and S100A4 as Therapeutic Target for Invasive and Metastatic Breast Cancer CYR61 and S100A4 are drivers for breast cancer cell invasion in vitro. Consequently, we examined the value of CYR61 and/or S100A4 as a potential therapeutic target for advanced breast cancer. Analyzing the expression in 239 paraffin-fixed patient breast tissue sections (104 invasive breast cancer sections with corresponding metastatic lymph node section and progesterone receptor-, estrogen receptor-and Her2neu expression, BR20837; 17 invasive ductal,1 medullary carcinoma and 6 normal breast tissue sections, BR248a; 2 invasive ductal carcinomas, 1 invasive lobular carcinoma and 2 normal breast tissue section, T087a). Analyzing if expression was detected (immunofluorescence signal for CYR61 and/or S100A4 1-5 spots +; 5-10 spots ++; >10 spots + + +) or not (−), we find the following pattern: 90.2% of invasive ductal carcinomas were positive for CYR61 expression, 82% were positive for S100A4 expression and 78% showed both CYR61 and S100A4 expression ( Figure 6A and Figure S7). Corresponding metastatic lymph node sections were in 96% positive for CYR61, in 75% positive for S100A4 and in 74% for both CYR61 and S100A4. TNBC tissue sections were in 97% positive for CYR61, in 75.8% positive for S100A4, and in 75.8% expressing both CYR61 and S100A4. Interestingly, CYR61 expression was only detected in 12.5% of normal breast tissue samples and S100A4 expression in none ( Figure 6D, detailed list Figure S7). Visual expression of CYR61 and S100A4 in blood vessels ( Figure 6D) could be found throughout all tissue sections. We find that the CYR61 and S100A4 expression appeared in very close localization to each other (Figure 6, white arrows) or even co-localized (Figure 6, white stars). These studies demonstrate that CYR61 and S100A4 could be valuable therapeutic targets and prognostic marker for invasive breast cancer and metastasis.
cancer progression (12,15,16,23). The question remains how CYR61 facilitates invasion in breast cancer and which role it possesses regarding EMT complexity (4). Since CYR61 has known oncogenic functions in serval tumor entities (12,13), including breast cancer (15,16), the question appeared if CYR61 might be a valuable therapeutic target in aggressive breast cancer   Probability of DMFS in 382 breast cancer patients with lymph node positive status according to S100A4 expression level. Kaplan-Meier plots were generated using Kaplan-Meier plotter (www.kmplot.com) data set 203186_at with a false-discovery rate (FDR) over 50%. Black line illustrates high S100A4 expression group and red line illustrates low S100A4 expression group. (C) Probability of remission free survival (RFS) in 1133 breast cancer patients with lymph node positive status according to CYR61 expression level. Kaplan-Meier plots were generated using Kaplan-Meier plotter (www.kmplot.com) data set 213226_at with a false-discovery rate (FDR) of 1%. Black line illustrates high CYR61 expression group and red line illustrates low CYR61 expression group. (D) Probability of RFS in 1133 breast cancer patients with lymph node positive status according to CYR61 expression level. Kaplan-Meier plots were generated using Kaplan-Meier plotter (www.kmplot.com) data set 203186_at with a false-discovery rate (FDR) over 50 %. Black line illustrates high S100A4 expression group and red line illustrates low S100A4 expression group. HR, hazard ratio. and if it could be a prognostic marker for these indications. We report that a higher CYR61 expression correlates with a poor prognosis of breast cancer patients. Moreover, we found that reducing the CYR61 expression leads to a decreased invasion in 2D and 3D invasion analysis setups, showing comparable results. Suggesting that reduced invasion upon CYR61 suppression is due to reduces ERK1/2 phosphorylation and S100A4 expression. CYR61 might be a valuable therapeutic target and prognostic marker for invasive and metastatic breast cancer. Triple negative breast cancers (TNBC) account for 15-20% of all breast cancer incidents and there is no specific targeted therapy available (33). There is a need for identifying new targets for future therapy options. Consistent with previous published results we could demonstrate that CYR61 expression is increased in TNBC cell line MDA-MB-231 (34) and further more in the TNBC cell line HCC1806. The contribution of EMT-induced expression changes to the invasion and metastatic cascade regarding cancer progression is highly debated and needs to be interpreted cell and tissue specific (4,8,35). Mesenchymal transformed breast cancer cells show an increased expression of CYR61 and S100A4 (23), which we could reproduce in our setting. It was shown that S100A4 facilitates breast cancer invasion (36). Gründker et al. demonstrated that suppressing extracellular signaling of CYR61 and S100A4 decreased the ability of breast cancer cell invasion in a co-cultural setting mimicking bone metastasis (23). It was not tested how the intracellular signaling is affected when CYR61 or S100A4 expression is reduced. We report here that transient gene silencing of either CYR61 or S100A4 can reduce invasiveness in mesenchymal transformed and TNBC cells. To further assess the impact of CYR61 on breast cancer cell invasion we increased extracellular CYR61 expression in non-invasive breast cancer cells and could show that this led to an increased invasive behavior. These findings indicate that CYR61 could be a regulator of breast cancer cells invasion. We showed that reversing EMTinduced upregulation of CYR61 and S100A4 leads to reduced invasive behavior in breast cancer cells in different invasion setups. This could indicate a role of EMT within this process. Further research is necessary to assess, if modulating CYR61 regulates EMT-TFs and thereby facilitates cellular plasticity. It has been suggested that targeting EMT-TFs could help to overcome chemo resistance and recent findings suggest an involvement of CYR61 in resistance to certain therapies in different tumor entities (5,37,38).
Despite, it was unclear how CYR61 regulates invasiveness of breast cancer cells. We suggest that CYR61 regulates S100A4 expression in mesenchymal transformed and TNBC cells through regulating ERK1/2 phosphorylation. Reducing S100A4 expression leads to decreased 3 D spheroid invasion and invasiveness of breast cancer cells in co-culture with osteosarcoma cells. Adding extracellular CYR61 to breast cancer spheroids with transient decreased S100A4 expression could restore the effect und led to a slightly increased invaded area. Hou et al. suggested that regulating CYR61 in osteosarcoma cells targets the MEK-ERK pathway (12). ERK1/2 signaling is gaining higher interest since the unique ERK1/2 position within cellular signaling. Targeting ERK1/2 could be valuable for therapy-resistant cancer to known clinically used BRAF and MEK inhibitors (39). We could show that inhibition of ERK1/2 phosphorylation led to decreased 3D spheroid invasion and reduced spheroid proliferation. Inhibition of ERK1/2 phosphorylation led to decreased S100A4 expression. But S100A4 decreased expression by itself had no impact on spheroid proliferation, neither had CYR61 or YAP transient suppression.
YAP is regulated negatively through the Hippo-Pathway, which regulates key events of organ size, development and angiogenesis (40)(41)(42). Regarding breast cancer YAP is reported to have dual function as oncogene and tumor suppressor (43). Higher YAP expression correlates with increased EMT marker expression (44). We suggest that reduced YAP expression leads to decreased 3D spheroid invasion by suppression of CYR61, p-ERK1/2 and S100A4. The effect of reduced YAP expression on 3D invasion could be restored by extracellular CYR61 addition.
CYR61 or S100A4 are suggested to be valuable prognostic markers regarding several tumor entities (45)(46)(47)(48). Egeland et al. suggested the use of S100A4 as a prognostic marker for earlystage breast cancer (49). We examined whether CYR61 and S100A4 could be valuable prognostic marker for invasive and metastatic breast cancer. CYR61 and S100A4 are highly expressed in invasive-ductal carcinomas, including TNBC, and both are expressed in metastatic lymph node sections. Of all analyzed tissue sections 82.2% expressing CYR61 did express S100A4, respectively, which lead to the conclusion, that CYR61 together with S100A4 would be valuable prognostic marker for breast cancer and breast cancer metastasis. Moreover, we found that expression of CYR61 and S100A4 is closely located (Figure 6, indicated by arrow) or even co-localized (Figure 6, indicated by star). Considering that CYR61 regulates cancer invasion and the findings, that it may be a valuable prognostic marker in different cancer entities (45,46,50,51), It was suggested before, that CYR61 regulates E-cadherin, N-cadherin and Twist in osteosarcoma cells (12). Further investigations should clarify if CYR61 suppression regulates EMT-TFs in breast cancer and facilitates invasion by altering ECM degradation and adhesion. Secretome analysis of co-cultured cancer cells could identify secreted proteins, like matricellular proteins, that are drivers for invasion and promote metastasis.
Our findings suggest that CYR61 plays a major role in breast cancer invasion. This impact is facilitated through the regulation of ERK phosphorylation and S100A4 expression. Moreover, targeting YAP, a CYR61 upstream regulator, regulates CYR61, ERK phosphorylation and S100A4. We could identify a close correlation between CYR61 and S100A4 expression and breast cancer invasion and metastasis in breast cancer patients. CYR61 together with S100A4 might be utilized as therapeutic target and prognostic marker for invasive breast cancer and metastasis.

DATA AVAILABILITY STATEMENT
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

AUTHOR CONTRIBUTIONS
Conception and design of the reported work was done by JH and CG. JH, GB, and CG did the development of methodology used. JH and SH performed invasion assays. JH and LG contributed to protein expression analysis. JH contributed to immune histochemical staining, gene expression analysis, proliferation analysis, in silico and network analysis. Analysis and interpretation of data was performed by JH, SH, LG, GB, GE, and CG. All authors read and approved the final manuscript.

FUNDING
Research reported in this publication was supported, by the Deutsche Krebshilfe grant 70112534.

ACKNOWLEDGMENTS
We would like to thank Sonja Blume for her excellent technical assistance. The group of Dr. Florian Wegwitz for fruitful discussions. Moreover, we appreciate the valuable suggestions of the reviewers.