Interactions between chemokines and chemokine receptors have been proposed recently to be of importance in the development and progression of cancer. Human breast cancer cells express the chemokine CXCL10 (IP-10) and also its receptor CXCR3. In this study, we have investigated the role of Ras activation in the regulation of CXCL10 and its receptor splice variant CXCR3-B in two human breast cancer cell lines MDA-MB-435 and MCF-7. In cotransfection assays, using a full-length CXCL10 promoter-luciferase construct, we found that the activated form of Ras, Ha-Ras(12V), promoted CXCL10 transcriptional activation. Ras significantly increased CXCL10 mRNA and protein expression as observed by real-time PCR, fluorescence-activated cell sorting analysis, and ELISA. Selective inhibition of Ha-Ras by small interfering RNA (siRNA) decreased CXCL10 mRNA expression in a dose-dependent manner. Further, using effector domain mutants of Ras, we found that Ras-induced overexpression of CXCL10 is mediated primarily through the Raf and phosphatidylinositol 3-kinase signaling pathways. We also observed that the expression of the splice variant CXCR3-B, known to inhibit cell proliferation, was significantly down-regulated by Ras. Selective inhibition of CXCR3-B using siRNA resulted in an increase in CXCL10-mediated breast cancer cell proliferation through Gi proteins and likely involving CXCR3-A. Finally, we observed intense expression of CXCL10 and CXCR3 in association with human breast cancer in situ, indicating that these observations may be of pathophysiologic significance. Together, these results suggest that activation of Ras plays a critical role in modulating the expression of both CXCL10 and CXCR3-B, which may have important consequences in the development of breast tumors through cancer cell proliferation. (Cancer Res 2006; 66(19): 9509-18)
Ras proteins act as molecular switches that cycle between active GTP-bound and inactive GDP-bound forms (1, 2). The three isoforms of Ras, H-Ras, N-Ras, and K-Ras, are ubiquitously expressed in mammalian cells (3). The hyperactive Ras can promote breast cancer growth and development without being mutated (4, 5), where it may be activated by persistent upstream signaling events. In particular, the HER-2 receptor tyrosine kinase is overexpressed and persistently activated in approximately 20% to 25% of human breast cancers, which plays an important role in Ras activation (4, 6). Activated Ras transmits its signal to a cascade of protein kinases (1, 6, 7) and can promote the growth of different tumors through multiple factors, including matrix metalloproteinases, angiogenic factors, and chemokines (8–11).
Chemokines are small, cytokine-like, secreted proteins (8-11 kDa; refs. 12, 13). The chemokines can be subdivided into four classes, the C-C, C-X-C, C, and C-X3-C chemokines, depending on the location of the first two cysteines in their protein sequence. The biological effects of these chemokines are mediated through specific receptors, which belong to the superfamily of seven-transmembrane domain G-protein–coupled receptors (14). The chemokines are well established to be involved in leukocyte recruitment into inflamed tissues (12, 13). Moreover, recent studies have proposed that they are of critical importance in the trafficking of tumor and vascular cells in the development and progression of cancer (15, 16).
It has been suggested that the chemokines play important roles in breast cancer development (17, 18). The breast tumor-promoting chemokines, secreted by tumor cells, stroma, or inflammatory cells, include CCL2 (MCP-1), CCL5 (RANTES), CXCL8 [interleukin-8 (IL-8)], CXCL12 (SDF-1), and CXCL14 (BRAK) (19–24). The release of CCL2 and CCL5 by breast cancer cells can mediate the migration of monocytes within the tumor. Tumor-infiltrating monocytes express numerous tumor growth factors, angiogenic mediators, extracellular matrix–degrading enzymes, and inflammatory cytokines (21, 22, 25). It has also been shown that CXCL12 and CXCL14 overexpressed in tumor myoepithelial cells and myofibroblasts, respectively, bind to receptors on breast tumor epithelial cells and enhance their proliferation, migration, and invasiveness (19).
Breast cancer cells have been found recently to express the chemokine CXCL10 (IP-10) and its receptor CXCR3 (26), although the factors regulating their expression have not been defined. Moreover, the function of CXCL10-CXCR3 interactions in breast tumor development is not established. CXCL10 has been shown to possess tumor-inhibitory properties (27, 28) but has also been suggested to be advantageous for tumor growth (26, 29, 30). Recently, it has been shown that CXCL10 may promote tumor cell proliferation and invasion (31). This controversy in the function of CXCL10 could be related to the observation that the CXCL10 receptor (CXCR3) is alternatively spliced in different human tissues to produce two known variants, CXCR3-A and CXCR3-B (32–36). Some tissues, such as heart, kidney, liver, skeletal muscle, and human airway epithelial cells, express both splice variants, whereas other tissues, such as placenta and human mesangial cells, express only CXCR3-A, and human microvascular endothelial cells selectively express CXCR3-B. CXCL10 binding to CXCR3-A leads to cell proliferation and chemotaxis, whereas binding of CXCL10 to CXCR3-B has been shown to mediate growth inhibition (33, 35, 36). Thus, the expression patterns of the two splice variants of CXCR3 may be of importance in regulating the effect of CXCL10 on tumor growth.
It has been shown that Ras can activate chemokines. Liang et al. (37) showed that mob-1 is a target gene of oncogenic Ras. Sparmann et al. (11) showed that activation of Ras in cervical cancer cells induces the expression of the proinflammatory chemokine CXCL8, which plays an important role in tumor growth and angiogenesis. Recently, Melillo et al. (31) observed that, in thyroid cancer, the RET-RAS-BRAF signaling cascade promotes the expression of CXCL1 (GRO-α) and CXCL10, which in turn stimulates cell proliferation and invasion. However, no studies, to our knowledge, have evaluated the direct effect of Ras on the expression of CXCL10 and CXCR3 splice variant in human breast cancer.
In this study, we find that activation of Ras in MDA-MB-435 and MCF-7 breast cancer cells promotes CXCL10 expression. We also observe that Ras down-regulates the expression of CXCR3 and especially the CXCR3-B splice variant to promote tumor cell proliferation. Our observations suggest that Ras-induced modulation of CXCL10 and its receptor splice variant might play a crucial role in the development of human breast cancer.
Materials and Methods
Reagents. Raf-1 kinase inhibitor I (5-iodo-3-[(3,5-dibromo-4-hydroxyphenyl)methylene]-2-indolinone), phosphatidylinositol 3-kinase (PI3K) inhibitor (LY294002), and Gi inhibitor [pertussis toxin (PTX)] were purchased from Calbiochem (La Jolla, CA). The mammalian target of rapamycin (mTOR) inhibitor (rapamycin) was obtained as a gift from Wyeth-Ayerst Research (Princeton, NJ). The small interfering RNA (siRNA) for Ha-Ras, CXCR3-B, and their respective controls were purchased from Invitrogen (Carlsbad, CA). Recombinant CXCL10 and CXCL4 (PF-4) were purchased from R&D Systems (Minneapolis, MN).
Cell culture. Human breast carcinoma cell lines, MDA-MB-435 (a generous gift from Debabrata Mukhopadhyay, Beth Israel Deaconess Medical Center, Boston, MA; ref. 38) and MCF-7 (obtained from American Type Culture Collection, Manassas, VA), were maintained in DMEM with 10% fetal bovine serum (Hyclone Laboratories, Logan, UT).
Plasmids. The CXCL10 promoter-luciferase construct was obtained as a generous gift from Richard M. Ransohoff (Cleveland Clinic, Cleveland, OH; ref. 39). All Ras expression constructs encode mutant versions of the transforming human Ha-Ras(12V) and were obtained as generous gifts from Roya Khosravi-Far (Beth Israel Deaconess Medical Center; ref. 1). The pDCR-ras(12V), pDCR-ras(12V,35S), pDCR-ras(12V,37G), and pDCR-ras(12V,40C) mammalian constructs encode effector domain mutants of Ha-Ras(12V), in which expression is under the control of the cytomegalovirus promoter.
Transfection assays. For all transfection assays, the cells were plated at 2 × 105 per well in six-well plates. The cells were transfected with the expression plasmids using the Effectene transfection reagent (Qiagen, Valencia, CA). The total amount of transfected plasmid DNA was normalized using a control empty expression vector. For luciferase assay, cells were harvested 24 hours after transfection and luciferase activity was measured using a standard assay kit (Promega, Madison, WI).
RNA isolation and reverse transcription-PCR analysis. Total RNA was prepared using the RNeasy isolation kit (Qiagen). The cDNA synthesis and PCR were carried out by SuperScript one-step reverse transcription-PCR (RT-PCR) kit (Invitrogen) using gene-specific primers and following the manufacturer's protocol. The oligonucleotide primers used are as follows: for human CXCL10, 5′-GGAACCTCCAGTCTCAGCACC-3′ (sense) and 5′-GCGTACGGTTCTAGAGAGAGGTAC-3′ (antisense) and for β-actin, 5′-GTGGGGCCGCCCCAGGCACCA-3′ (sense) and 5′-GTCCTTAATGTCACGCACGATTTC-3′ (antisense).
Real-time PCR. Total RNA was prepared as described before, and cDNA was synthesized using cloned avian myeloblastosis virus first-strand synthesis kit (Invitrogen). Real-time PCR was done by using real-time gene expression assay kit (SuperArray Bioscience Corp., Frederick, MD) containing gene-specific primers (for CXCL10 or total CXCR3 or CXCR3-B; ref. 33) and SYBR Green dye (Molecular Probes, Eugene, OR) following the manufacturer's protocol. As an internal control, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was amplified and analyzed under identical conditions using specific primers. Melting point analysis was done in all cases to confirm the right amplification of the expected gene. Ct (the cycle number at which emitted fluorescence exceeded an automatically determined threshold) for gene of interest was corrected by the Ct for GAPDH and expressed as ΔCt. Data were measured as fold changes of mRNA amount, which was calculated as follows: (fold changes) = 2X [where X = ΔCt (for control group) − ΔCt (for experimental group)].
Fluorescence-activated cell sorting analysis. To measure intracellular chemokine expression, GolgiStop (BD Biosciences, San Diego, CA) was added to the cell culture medium 12 hours before analysis. The cells were fixed/permeabilized in cytofix/cytoperm solution (BD Biosciences) for 20 minutes at 4°C and washed with perm/wash solution (BD Biosciences). The fixed cells were then incubated with phycoerythrin-conjugated antibody to CXCL10/CCL2/CXCR3 or matched IgG isotype (BD Biosciences) for 30 minutes at 4°C. To measure CXCR3 surface expression, the cells were stained with phycoerythrin-conjugated CXCR3 antibody (BD Biosciences). The stained cells were analyzed in a FACSCalibur (Becton Dickinson).
ELISA. The concentrations of CXCL10 in tissue culture supernatants were determined by using Quantikine human CXCL10 immunoassay kit (R&D Systems).
Western blot analysis. Protein samples were run on SDS-polyacrylamide gel and transferred to a polyvinylidene difluoride membrane (NEN Life Sciences Product, Inc., Boston, MA; ref. 10). The membranes were coated with anti-Ras (BD Biosciences) and subsequently incubated with peroxidase-linked secondary antibody. The reactive bands were detected by chemiluminescence (Pierce, Rockford, IL).
Cell proliferation assay. Tumor cells (5 × 103) were seeded and grown in 96-well plates and transfected with the siRNA of either CXCR3-B or its control using Lipofectamine 2000 (Invitrogen). Some cells were also transfected with Ha-Ras(12V) or the empty expression vector. Following transfection, the cells were treated with CXCL10 or CXCL4 for 24 hours and [3H]thymidine (0.5 μCi/well) was added for the final 15 hours before cell harvesting. [3H]thymidine incorporation was measured using a microplate scintillation and luminescence counter (Perkin-Elmer/Wallac, Boston, MA).
Phosphorylated Akt assay. The phosphorylation of Akt was measured by using human phosphorylated Akt (S473) immunoassay kit (R&D Systems).
Immunohistochemistry. Immunohistochemistry was done on 4-μm-thick serial frozen sections of human breast tumors [invasive mammary carcinoma (IMC; n = 3) and ductal carcinoma in situ (DCIS; n = 3)] and normal breast tissue (n = 3). Tissues were obtained from patients at the University Hospital (Wuerzburg, Germany). Briefly, acetone-fixed sections were incubated first with either goat anti-human CXCL10 (R&D Systems) or rabbit anti-human CXCR3 (recognizing both A and B isoforms; GenWay Biotech, Inc., San Diego, CA) and second with a species-specific horseradish peroxidase–conjugated secondary antibody. Specimens were washed thoroughly in between incubations, developed in 3,3′-diaminobenzidine (BioGenex, San Ramon, CA), and counterstained in Gill's hematoxylin using standard techniques.
Statistical analysis. Statistical evaluation for data analysis was determined by Student's t test. Differences with P < 0.05 were considered statistically significant.
Activation of Ras promotes the overexpression of CXCL10 in human breast cancer cells. It has been shown that human breast adenocarcinoma cells express CXCL10 (26). Here, we first examined whether the activation of Ras can regulate the expression of CXCL10 in two well-established human breast cancer cell lines MDA-MB-435 and MCF-7 (38, 40). Each tumor cell line was transfected with increasing concentrations (0.5-1.5 μg) of a plasmid containing the activated form of Ras, Ha-Ras(12V), and the intracellular expression of CXCL10 and CCL2 (as a control) was measured by fluorescence-activated cell sorting (FACS) analysis. As shown in Fig. 1A (left), at all concentrations, Ras significantly increased the expression of CXCL10 in MDA-MB-435 cells (and MCF-7 cells; data not shown) compared with the empty vector–transfected controls. However, there was minimal increase in the expression of CCL2 in these cells following Ras activation (Fig. 1A , right).
We also wished to determine whether Ras activation promotes the secretion of CXCL10 protein as measured by ELISA. MDA-MB-435 cells were transfected with two different concentrations (1.0 and 1.5 μg) of Ha-Ras(12V) or the empty expression vector, and following 48 hours of transfection, the release of CXCL10 protein in culture supernatant was measured. As shown in Fig. 1B, activation of Ras significantly increased the release of CXCL10 in the culture supernatants. The expression of Ha-Ras(12V) in the transfected cells was confirmed by Western blot analysis (Fig. 1C). Together, these results suggest that Ras plays an important role in the regulation of CXCL10 in human breast cancer.
Ras promotes the transcriptional activation of CXCL10. First, by PCR, we found that the transfection of increasing concentrations (0.5-1.5 μg) of Ha-Ras(12V) in MDA-MB-435 and MCF-7 cells resulted in a significant increase in CXCL10 mRNA expression compared with the empty vector–transfected cells (Fig. 2A and B). To determine whether Ras-mediated overexpression of CXCL10 involves transcriptional regulatory mechanism(s), we used a human full-length CXCL10 promoter-luciferase construct. MDA-MB-435 and MCF-7 cells were cotransfected with the promoter-reporter construct and either the plasmid expressing the activated form of Ras, Ha-Ras(12V), or the empty expression vector. The effect of Ras on CXCL10 promoter activation was assessed by measurement of luciferase activity in cell lysates. As shown in Fig. 2C, activation of Ras significantly increased CXCL10 promoter activity in MCF-7 cells in a dose-dependent manner compared with the empty vector–transfected cells. Ras also promoted a similar pattern of CXCL10 transcriptional activation in MDA-MB-435 cells (data not shown).
Next, to examine whether inhibition of Ras can block the expression of endogenous CXCL10, we made use of a Ha-Ras-specific siRNA, which we found to significantly inhibit the expression of the gene (data not shown). We observed that the transfection of MDA-MB-435 cells with this siRNA significantly inhibited CXCL10 mRNA expression compared with the controls (Fig. 2D). Overall, our data indicate that Ras regulates CXCL10 overexpression in breast cancer cells through transcriptional mechanism(s).
Ras-induced CXCL10 overexpression is mediated through the Raf and PI3K signaling pathways. Raf, Rho, and PI3K are critical effector molecules of Ras-induced signaling pathways. We next wished to determine which effector is functional for Ras-mediated CXCL10 overexpression. To this end, we used three effector domain mutants of Ras in transfection assays. Ras(12V,35S) retains full-length Raf-1 binding activity, Ras(12V,37G) retains Rho binding activity, and Ras(12V,40C) retains PI3K binding activity (1). MDA-MB-435 cells were transfected with one of these Ras effector domain mutants (or empty vector), and following 48 hours of transfection, the secretion of CXCL10 protein was measured in the culture supernatants by ELISA (Fig. 3A). We found that Ras(12V,35S) and Ras(12V,40C) significantly increased the secretion of CXCL10 protein compared with the empty vector controls. In contrast, there was no detectable increase in CXCL10 protein following transfection of the cells with Ras(12V,37G), although all the Ras mutants were significantly expressed in the transfected cells (Fig. 3B).
Next, we treated the cells with pharmacologic inhibitors of Raf and PI3K (Raf-1 kinase inhibitor I and LY294002, respectively) to examine their effects on endogenous CXCL10 protein expression by FACS analysis (Fig. 3C). We found that these agents, alone or in combination, significantly inhibited the expression of CXCL10 in MDA-MB-435 cells compared with the vehicle-treated controls. However, rapamycin, which blocks mTOR (one of the effector molecules of PI3K), did not inhibit CXCL10 expression in these cells (data not shown). Together, these findings suggest that Raf and PI3K are critical effector molecules for Ras-induced CXCL10 expression in breast cancer.
Activation of Ras down-regulates the expression of CXCR3-B on breast cancer cells. Overexpressed CXCL10 may bind its receptor CXCR3 on breast cancer cells in an autocrine manner (26). As mentioned earlier, the CXCR3 gene is alternatively spliced to generate two splice variants, CXCR3-A (promoting cell proliferation) and CXCR3-B (mediating growth inhibition) (33, 36). The CXCR3 gene contains a single 978-bp intron. The 1627-bp CXCR3-A transcript is formed by removal of the entire 978-bp intron. In contrast, the 1860-bp CXCR3-B transcript uses an alternative splice acceptor site that is 233-bp upstream (5′) of the splice acceptor site used to form CXCR3-A. The sequence downstream of CXCR3-A acceptor site is identical in the two splice variants.
Here, we sought to determine whether the activation of Ras could also modulate CXCR3 expression in breast cancer cells. We first evaluated the CXCR3 expression in MDA-MB-435 cells by FACS analysis using an antibody that recognizes both CXCR3-A and CXCR3-B isoforms (36). Under nonpermeabilized conditions, we detected two populations of cells, one of which expressed high levels of CXCR3 on their surface and another without any surface expression (Fig. 4A). However, when the cells were permeabilized to allow detection of both surface and intracellular receptors, we observed that most of the cells expressed CXCR3 (Fig. 4B), indicating that cell surface expression of the receptor is regulated. A similar pattern of CXCR3 expression was also observed in MCF-7 cells (data not shown). Next, we evaluated whether activation of Ras promotes any changes in the expression of total CXCR3 (both A and B isoforms) or the CXCR3-B isoform alone, using gene-specific primers. MDA-MB-435 cells were transfected with two different concentrations (1.0 and 1.5 μg) of Ha-Ras(12V), and the expression of either total CXCR3 or CXCR3-B mRNA was measured by real-time PCR. Empty expression vector-transfected cells were used as controls. We found that activation of Ras resulted in some inhibition of total CXCR3 mRNA expression (Fig. 4C), but the expression of CXCR3-B isoform was significantly down-regulated in a dose-dependent manner compared with controls (Fig. 4D). Therefore, we suggest that, although Ras induces the expression of CXCL10, it also limits its ability to bind CXCR3-B on breast tumor cells in an autocrine manner.
Down-regulation of CXCR3-B promotes CXCL10-mediated proliferation of breast cancer cells. We next evaluated whether inhibition of CXCR3-B expression can promote CXCL10-induced proliferation of breast cancer cells. We used a siRNA (25-100 nmol/L) that specifically inhibited CXCR3-B mRNA expression without altering the expression of CXCR3-A in MDA-MB-435 cells (Supplementary Fig. S1A and B); we also noted that our siRNA partially reduced the total CXCR3 surface expression, indicating that part of the surface receptors belongs to CXCR3-B (Supplementary Fig. S1C). Here, we first transfected the MDA-MB-435 cells with the CXCR3-B-specific siRNA and did a cell proliferation assay. We observed that CXCR3-B siRNA markedly increased cell proliferation compared with the control siRNA-transfected cells (Fig. 5A). Second, to analyze the effect of CXCL10 on breast cancer cell proliferation, we cultured either control siRNA-transfected or CXCR3-B siRNA-transfected MDA-MB-435 cells with two different concentrations (10 and 20 ng/mL) of CXCL10 (Fig. 5B). In control siRNA-transfected cells, CXCL10 treatment inhibited cell proliferation, whereas, in CXCR3-B siRNA-transfected cells, CXCL10 treatment markedly increased cell proliferation compared with untreated cells (Fig. 5B). Thus, in absence of CXCR3-B, CXCL10 promotes the proliferation of breast cancer cells likely through CXCR3-A.
Different signaling mechanisms mediate the reciprocal effects of CXCR3 splice variants on breast cancer cell proliferation. We next examined the reciprocal effects of CXCR3-A versus CXCR3-B on breast cancer cell proliferation and evaluated their signaling mechanism. Although most chemokine receptors, including CXCR3, are coupled to PTX-sensitive Gi proteins (12, 33), it has been reported that the two CXCR3 splice variants may mediate their reciprocal function through different G protein coupling (33, 41). To evaluate specific signals through CXCR3-B, we took advantage of the chemokine CXCL4 that binds selectively to this receptor splice variant (33, 41). As shown in Fig. 5C, CXCL4 treatment significantly inhibited the proliferation of control siRNA-transfected MDA-MB-435 cells compared with untreated controls; pretreatment of these cells with PTX did not alter the inhibitory effect of CXCL4. Thus, CXCR3-B expressed on breast cancer cells mediates a growth-inhibitory effect independent of Gi proteins.
We next wished to evaluate whether CXCL10-induced cell proliferation involves the coupling of Gi proteins. We have shown that following knockdown of CXCR3-B, CXCL10 can promote cell proliferation likely through CXCR3-A (Fig. 5B). To this end, CXCR3-B siRNA-transfected MDA-MB-435 cells were treated with CXCL10 in the presence or absence of PTX. As shown in Fig. 5C, CXCL10-induced cell proliferation was significantly inhibited by PTX. As expected, CXCL4 did not mediate any significant growth-inhibitory effect in these cells in absence of CXCR3-B (Fig. 5C). Therefore, CXCL10-induced breast cancer cell proliferation is dependent on Gi proteins.
Finally, we observed by immunoassay that, after knockdown of CXCR3-B in MDA-MB-435 cells, CXCL10 markedly increased the phosphorylation of the PI3K substrate Akt, known to be activated by CXCR3 (Supplementary Fig. S2; ref. 42). In contrast, CXCL4 failed to produce any significant change in Akt phosphorylation through CXCR3-B in these cells (Supplementary Fig. S2). Together, these experiments suggest that CXCR3 splice variants induce different signals in breast cancer cells and that each of these signaling pathways mediate reciprocal effects on cell proliferation.
Activation of Ras promotes CXCL10-mediated proliferation of breast cancer cells. Our earlier experiments show that activation of Ras in human breast cancer cells inhibits CXCR3-B expression and that knockdown of CXCR3-B promotes CXCL10-mediated cell proliferation. We next sought to determine whether Ras activation could promote CXCL10-mediated proliferation of breast cancer cells. To this end, MDA-MB-435 cells were transfected with Ha-Ras(12V) and then cultured in presence or absence of CXCL10 to assess cell proliferation. Control cells were transfected with the empty expression vector. As shown in Fig. 5D, activation of Ras itself increased cell proliferation compared with controls. Ras-induced cell proliferation was further increased in the presence of CXCL10 compared with untreated controls (Fig. 5D). These results again indicate that activation of Ras may lead to down-regulation of CXCR3-B and promotes CXCL10-mediated breast cancer cell proliferation in an autocrine manner.
Expression of CXCL10 and CXCR3 in situ in human breast tumors. We next evaluated the expression of CXCL10 and total CXCR3 (both A and B isoforms) in human breast tissues obtained from six patients with breast carcinoma (three IMC and three DCIS). Expression was compared with that found in normal breast tissue. In normal breast tissue, we observed that the expression of CXCL10 was sparse on isolated cells throughout the tissue; CXCR3 was also expressed on occasional isolated cells in these normal tissues (Fig. 6, bottom). In contrast, the distribution and intensity of expression of both CXCL10 and CXCR3 were markedly increased in breast tissues with evidence of carcinoma. As shown in Fig. 6 (top), ductal cells in specimens with evidence of IMC expressed significant amounts of both CXCL10 and CXCR3; it seems that some of the tumor-infiltrating immune cells also expressed CXCR3. In specimens with evidence of DCIS, ductal cells expressed high levels of both CXCL10 and CXCR3, but the intensity and distribution of staining was less compared with that of IMC (Fig. 6, middle). These observations show that both CXCL10 and CXCR3 are expressed in situ in human breast tumors; however, the relative expression of CXCR3-A and CXCR3-B are yet to be determined. These findings indicate that our in vitro observations are likely of pathophysiologic importance for the development of breast cancer in vivo.
In the present study, we show that activation of Ras promotes the overexpression of the chemokine CXCL10 in human breast cancer cells primarily through the Raf and PI3K signaling pathways. We also find that Ras down-regulates the expression of the growth-inhibitory CXCR3-B splice variant on these cells to promote CXCL10-mediated breast cancer cell proliferation. To our knowledge, this is the first study that has evaluated the role of Ras in the regulation of CXCL10 and its receptor splice variant in breast cancer development.
Although CXCL9 (MIG), CXCL10, and CXCL11 (I-TAC) are known as the members of IFN-γ-inducible CXC chemokines having tumor-inhibitory properties (27, 33), they have been reported to be expressed in significant amounts in aggressive tumors, indicating a possible role in tumor progression. For instance, Suyama et al. (29) showed recently that CXCR3 and its ligands are overexpressed in human renal cell carcinoma and they suggested that CXCR3-mediated signaling in tumor cells is critical for tumor progression and metastasis. Kawada et al. (43) reported that melanoma cells constitutively express CXCR3 and that its ligand CXCL10 induces cell survival, invasion, migration, and lymph node metastasis. Goldberg-Bittman et al. (26) suggested that simultaneous expression of both CXCR3 and CXCL10 in breast tumor cells indicates that CXCL10 may act not only on cells with the breast tissue microenvironment but also on the tumor cells themselves in an autocrine manner to promote tumor growth. Teichmann et al. (30) raised the possibility that high endogenous CXCL10 expression in Hodgkin's lymphoma and nasopharyngeal carcinoma may not exert an antitumor effect as ascribed to it in other studies; in another report, Mellillo et al. (31) showed that, in thyroid carcinomas, overexpressed CXCL10 can promote cell proliferation and invasion; however, the mechanism for CXCL10-induced tumor development was not defined.
Classically, CXCL10-CXCR3 interactions play important roles in the pathogenesis of autoimmune and inflammatory diseases through the recruitment of CXCR3+ T-cells (14, 32, 44). The recruitment of cytotoxic T cells (perhaps via CXCL10-CXCR3 interactions) could also facilitate antitumor immunity, resulting in tumor regression (27, 45). Thus, the overexpressed CXCL10 in human breast cancer cells may attract CXCR3+ T cells within the tumor to promote antitumor immunity. However, our observations suggest an alternative possibility, where the overexpressed CXCL10 may act in an autocrine manner to promote tumor growth through a novel mechanism.
Recently, it has been shown that Ras can induce tumor growth through chemokines (11, 31). Here, we find that Ras activation significantly induces CXCL10 in human breast cancer cells. Ras can also promote some induction of the other CXCR3-binding chemokine CXCL11, but not CXCL9, in these cells (data not shown). Moreover, the effect of Ras on CXCL10 induction does not seem to be indirect because we found no significant increase in the expression of CXCL10-inducing IFNs in Ras-transfected breast cancer cells (data not shown). However, we cannot also rule out a Ras-independent pathway for CXCL10 overexpression in these cells. Here, we suggest that the Ras-induced CXCL10 may act on CXCR3 expressed on the same tumor cells.
Lasagni et al. (33) showed that the CXCR3 gene is alternatively spliced to generate two splice variants CXCR3-A and CXCR3-B. CXCR3-B possesses growth-inhibitory properties, whereas CXCR3-A can promote cell proliferation (32, 33, 35, 36). In this study, we propose that the relative expression of these two splice variants on breast cancer cells is important in regulating their proliferation through CXCL10. The surface expression of CXCR3 on these cells may be regulated by different factors (42, 46). Similar findings have been obtained for CXCR1 and CXCR2 in human mast cells and T lymphocytes, where the cell surface expression may be regulated by extracellular stimuli (47). Recently, Aksoy et al. (46) showed that, although most human bronchial epithelial cells express cytoplasmic CXCR3, only cells in the late S to G2-M phase of the cell cycle express it on their surface. Here, we find that the activation of Ras down-regulates the mRNA expression of CXCR3-B isoform in MDA-MB-435 cells and show that the inhibition of CXCR3-B expression can promote CXCL10-induced proliferation of these cells. Thus, a lower expression level of CXCR3-B on breast tumor cells may reduce or withdraw the growth-inhibitory signals in these cells, and in absence of CXCR3-B, the CXCL10-induced tumor cell proliferation is likely mediated through CXCR3-A.
The interaction of CXCL10 with either CXCR3-A or CXCR3-B results in activation of distinctly different signaling pathways. This may explain the reciprocal effects of these two receptors on breast cancer cell proliferation. Postreceptor signals via CXCR3-B result in an increase in intracellular cyclic AMP levels, whereas CXCR3-A mediates calcium influx within the cell (33, 41, 42). Although coupling of the G proteins to chemokine receptors is important for their biological activities, CXCR3-A and CXCR3-B may mediate their functions through different G protein coupling (33, 41, 42). In this study, we show that selective activation of CXCR3-B through CXCL4 (33, 41, 48) inhibits the growth of breast cancer cells independently of Gi proteins. This suggests that CXCR3-B may couple other types of G proteins (like Gs proteins), reported to be of importance for its function (33, 41). In contrast, following knockdown of CXCR3-B, CXCL10 promotes the proliferation of breast cancer cells through Gi proteins and possibly involving the PI3K-Akt signaling pathway. These findings suggest that Ras-induced and CXCL10-mediated breast cancer cell proliferation likely involves CXCR3-A and the associated coupling of Gi proteins. However, we cannot rule out the possible role of another novel CXCL10 receptor (49), other CXCR3 splice variant (50), or differential receptor coupling mechanism (42) that may also promote CXCL10-mediated cell proliferation.
In summary, we show that the activation of Ras is of importance in regulating the expression of CXCL10 and its receptor in human breast cancer. Interactions between overexpressed CXCL10 and CXCR3 splice variant may have important consequences in terms of cellular proliferation. Targeting Ras and CXCL10/CXCR3 may represent future therapeutics for the treatment of human breast cancer.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
Grant support: NIH grants DK64182 (S. Pal), HL74436 (D.M. Briscoe), and DBU16011 (A.M. Waaga-Gasser), and Emerald Foundation (D.M. Briscoe).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Roya Khosravi-Far for the Ras constructs and also for helpful suggestions.