Cystine/glutamate antiporter xCT (SLC7A11) facilitates oncogenic RAS transformation by preserving intracellular redox balance

"VSports app下载" Oncogenic Transformation with KRAS Protects Fibroblasts from Oxidative Stress.

To define the impact of oncogenic RAS signaling on the cellular response to oxidative stress and the mechanisms involved, we initially used 3T3 fibroblasts transformed by activated KRAS^G12V (3T3 KRAS^V12) or the ETV6-NTRK3 (EN) chimeric tyrosine kinase (3T3 EN) (29). As both oncoproteins constitutively activate Ras-Erk and PI3K-Akt signaling, we used both KRAS^V12 and EN-transformed cell lines to avoid cell line-specific effects (30). Expression of oncogenic KRAS and EN were confirmed in the transformed fibroblasts (Fig. 1A).

To assess the impact of oncogenic KRAS and EN transformation on the susceptibility of fibroblasts to exogenous oxidative stress, we subjected the cells to increasing concentrations of hydrogen peroxide (H₂O₂) and quantified cell death. While H₂O₂ treatment induced massive cell death in nontransformed control cells (3T3 MSCV), especially at higher concentrations, 3T3 KRAS^V12 and 3T3 EN cells were relatively protected (Fig. 1B), indicating that oncogenic transformation by KRAS and EN provide protection against oxidative stress. Similar effects were observed for mutant KRAS using a second oxidative stress inducer, diethyl maleate (DEM) (SI Appendix, Fig. S1A). These differences in cell death were linked to amounts of intracellular ROS; indeed, while 3T3 MSCV cells displayed higher levels of intracellular ROS under ambient conditions and showed rapid accumulation of ROS following H₂O₂ treatment (Fig. 1C) or DEM treatment (SI Appendix, Fig. S1B), 3T3 KRAS^V12 cells maintained lower levels of intracellular ROS throughout these conditions. This is in line with previous work demonstrating that KRas^G12D expression actively suppresses ROS due to basal activation of the Nrf2 detoxification program (11). Similar effects were also seen with 3T3 EN cells (SI Appendix, Fig. S1C). This was further validated by analyzing levels of protein oxidation, which indicated that following exposure to H₂O₂, 3T3 MSCV cells showed higher levels of protein oxidation relative to 3T3 KRAS^V12 cells, as measured by dityrosine levels (Fig. 1D). Moreover, together with reduced ROS, 3T3 KRAS^V12 and EN cells consistently displayed higher reduced glutathione (GSH) levels (Fig. 1E) and higher total glutathione (GSH+GSSG) levels (SI Appendix, Fig. S1D) relative to 3T3 MSCV cells, both under ambient conditions and in response to oxidative stress by H₂O₂ or DEM treatment. Interestingly, while 3T3 KRAS^V12 and EN cells exhibited a lower GSH/GSSG ratio relative to 3T3 MSCV cells under ambient conditions, these cells had similar GSH/GSSG ratios as 3T3 MSCV cells following H₂O₂ treatment (Fig. 1F). This suggests that in our cellular model, protection of cells against oxidative stress following oncogenic transformation with KRAS or EN may be attributed to enhanced GSH biosynthesis rather than increased GSH-GSSG recycling. Supportive of this, 3T3 KRAS^V12 and EN cells showed higher total GSH and GSH/GSSG ratios compared with 3T3 MSCV cells when treated with DEM, which conjugates and, therefore, depletes GSH availability (Fig. 1F and SI Appendix, Fig. S1D). Together, these results suggest that oncogenic transformation with KRAS or EN protects fibroblasts against exogenous oxidative stress by preserving the redox balance and enhancing intracellular GSH capacity.

Oncogenic RAS Leads to Enhanced Induction of xCT.

To uncover the mechanisms underlying KRAS^V12- or EN-mediated oxidative stress resistance, we performed whole-transcriptome microarray analysis of 3T3 KRAS^V12, EN, and MSCV cells under basal conditions (UT) and H₂O₂ treatment to identify differentially regulated transcripts potentially involved in the response to oxidative stress. When 3T3 KRAS^V12 and 3T3 EN cells were individually analyzed following H₂O₂ treatment, we found overlap of a core of 90 probesets (corresponding to 67 unique genes) that were significantly up-regulated >twofold in both cell lines relative to 3T3 MSCV cells. Gene ontology (GO) overrepresentation analysis of these 67 genes showed significant enrichment for functional categories relating to cellular stress response and cell death, including the GO categories “cellular response to hypoxia,” “endoplasmic reticulum unfolded protein response,” and “apoptotic process” (Fig. 2A, Left). These categories comprised many up-regulated genes of the integrated stress response pathway such as Ppp1r15a, Atf3, Trib3, Chac1, and Ddit3 as well as oxidative stress response genes such as Hmox1, Plk3, and Trp53inp1 (Dataset S1).

To better characterize individual genes uniquely up-regulated by KRAS^V12 and EN in response to oxidative stress, we subdivided genes to identify those that were differentially up-regulated by KRAS^V12 and EN transformation relative to control cells only under H₂O₂ treatment, and not under basal conditions. From this analysis, we identified only 15 genes with overlap across both cell lines (Fig. 2A, Right). Among these, solute carrier family 7 member 11 (Slc7a11; also commonly referred to as xCT) was of interest due to its known role in redox regulation through GSH biosynthesis. Differential expression of xCT mRNA was validated by qRT-PCR, confirming that while xCT expression was induced in all cell lines in response to DEM or H₂O₂, it was fourfold and twofold higher in 3T3 KRAS^V12 compared with 3T3 MSCV cells at 6 h treatment with DEM and H₂O₂, respectively (Fig. 2B). To corroborate these findings in another cell type, we analyzed levels of xCT expression in a publicly available gene expression dataset of human mammary epithelial cells (HMEC) ectopically expressing either activated HRAS or a panel of oncogenes [β-catenin (BCAT), E2F3, MYC, or SRC] (31). In line with our findings, we observed that xCT is significantly up-regulated in HRAS^V12 and SRC-transformed cells, but not in other oncogene expressing cells, compared with control GFP HMECs (Fig. 2C). Furthermore, xCT induction is not RAS isoform specific in 3T3 cells, as HRAS^V12 transformation also resulted in enhanced induction of xCT in response to H₂O₂ (SI Appendix, Fig. S2A). To further corroborate the link between RAS signaling and xCT expression, we used Nras/Hras double knockout MEFs with 4-OHT–inducible knockout of endogenous Kras (32). As shown in SI Appendix, Fig. S2B, in the presence of 4-OHT, the conditional Kras alleles become fully excised rendering the cells “Rasless.” This complete knockout of Kras led to a dose-dependent decrease in xCT expression, further linking xCT induction to RAS signaling (SI Appendix, Fig. S2B). In contrast, levels of GPX4, another oxidative stress response protein, were unchanged. Together, these data support the notion that oncogenic RAS pathway activation leads to induction of xCT expression.

We next investigated the impact of oncogenic KRAS on xCT activity to determine whether increased xCT mRNA levels led to enhanced protein activity. We opted for this approach rather than Western blotting since it has been documented that at present, all commercially available antibodies display nonspecific immunoreactivity when used to detect xCT protein in mouse cell lines and tissues (33). We therefore performed in vitro uptake assays of the xCT-specific PET tracer ¹⁸F-5-fluoroaminosuberic acid (FASu), which was developed as a diagnostic tracer of oxidative stress via system x_c⁻ activity (34, 35). In sharp contrast to nontransformed 3T3 cells, which showed no significant increase of FASu uptake following H₂O₂ and DEM exposure, 3T3 KRAS^V12 cells consistently exhibited a five- to sixfold induction of system x_c⁻ activity in response to these compounds (Fig. 2D), in keeping with higher levels of xCT mRNA detected in transformed cells. FASu uptake was effectively blocked with a known xCT inhibitor, erastin, confirming the specificity of FASu for xCT. Moreover, increased FASu uptake observed in KRAS^V12 transformed cells was not linked to altered expression of the heavy-chain subunit of system x_c⁻, i.e., CD98, as expression of the latter was similar in transformed versus nontransformed cells under basal and H₂O₂ treatment conditions (SI Appendix, Fig. S2C). This points to a major role for xCT in KRAS^V12-mediated induction of system x_c⁻ activity. Together, these results indicate that oncogenic KRAS^V12 leads to enhanced levels of xCT expression and activity in response to exogenous oxidative stress.

VSports - xCT Mediates Oncogenic KRAS-Induced Resistance to Oxidative Stress by Preserving Redox Balance.

Given that xCT activity is essential to support GSH biosynthesis pathway and therefore to maintain redox balance, we next asked whether xCT promotes the oxidative stress resistance conferred by oncogenic KRAS (15). To address this, we acquired mouse embryonic fibroblasts (MEFs) derived from wild-type (xCT^+/+) or xCT^−/− mice into which we stably expressed K-Ras^V12 (36). xCT^−/− cells are routinely cultured in the reducing agent 2-mercaptoethanol (2ME), allowing these cells to circumvent the block in cysteine uptake by instead importing its reduced form, cysteine, via neutral amino acid transporters (36). Strikingly, xCT knockout cells transformed with K-Ras^V12 showed marked cell death following exposure to H₂O₂ compared with wild-type cells (Fig. 2E). Similarly, knocking down xCT with two nonoverlapping siRNAs in 3T3 KRAS^V12 cells increased cell death induced by H₂O₂ (SI Appendix, Fig. S2 D and E). Susceptibility to oxidative stress was reversed in MEFs and 3T3 models either by supplementation with the antioxidants GSH and NAC or with 2ME (specifically in xCT^−/− KRAS^V12) (Fig. 2E and SI Appendix, Fig. S2E). Further, ectopic overexpression of xCT in 3T3 MSCV cells phenocopied effects of oncogenic KRAS on susceptibility to H₂O₂ by protecting cells against this treatment (SI Appendix, Fig. S2F). Together, these data clearly indicate that xCT mediates oncogenic KRAS-induced cytoprotection against oxidative stress. This function is linked to antioxidant activity of xCT. 2ME depletion led to a significant increase of ROS in xCT^−/− KRAS^V12 relative to xCT^+/+ KRAS^V12 cells, but this was reversed by reexpressing xCT (+xCT KRAS^V12) or by supplementation with 2ME, GSH, or NAC, but less so under cotreatment with NAC and 1-buthionine-S,R-sulfoximine (BSO), a glutamate-cysteine ligase (GCL) inhibitor (Fig. 2F). This suggests that xCT promotes KRAS-mediated oxidative stress resistance by providing cystine intermediates for the synthesis of GSH via GCL. Similarly, knockdown of xCT in 3T3 KRAS^V12 cells was accompanied by an increase in ROS levels under both basal and H₂O₂ treatment conditions, which was also reversed by GSH or NAC, but not NAC and BSO together (SI Appendix, Fig. S2G). Consistent with this finding, knockout (Fig. 2G) or knockdown (SI Appendix, Fig. S2H) of xCT led to significant decreases in GSH levels under basal and oxidative stress conditions (2ME for MEFs and H₂O₂ treatment for 3T3), which could be reversed by either 2ME (specifically in xCT^−/− KRAS^V12) or NAC, but not NAC with BSO. Taken together, these data provide compelling evidence that xCT mediates resistance to oxidative stress conferred by oncogenic KRAS by providing cystine for GSH synthesis.

xCT Expression and Activity Are Induced by Mutant KRAS in Human Cancer Cells.

To assess regulation of xCT by oncogenic KRAS in human cancer cells, we analyzed various cell lines originating from mutant KRAS-driven human cancers, such as lung adenocarcinoma (LUAD), colorectal adenocarcinoma (COAD), and pancreatic ductal adenocarcinoma (PDAC). We surveyed xCT expression in mutant KRAS versus wild-type KRAS cancer cells from LUAD, COAD, and PDAC. Normal epithelial cells HPL1D and HPDE6 from lung and pancreas, respectively, were included as controls. We found that xCT expression is consistently higher in mutant KRAS cancer cell lines compared with wild-type KRAS cell lines originating from the same tumor types and in normal epithelial cells (Fig. 3A), pointing to a correlation between KRAS activation and xCT expression in human cancer cells. To test this, we performed mutant KRAS knockdown, which markedly decreased xCT expression at the mRNA level in the COAD cell line SW620 (Fig. 3B), as well as at the protein level in SW620 and in the LUAD cell line H460 (Fig. 3C). Consistent with this, knockdown of mutant KRAS led to reduction of xCT activity (FASu uptake; Fig. 3D) and reduced GSH (Fig. 3E) in SW620 cells under basal and H₂O₂ treatment conditions. The decrease in reduced GSH levels following mutant KRAS knockdown under basal conditions was partially rescued by NAC, thereby phenocopying the effect of xCT knockdown on reduced GSH levels as observed in 3T3 KRAS^V12. Consistent with 3T3 KRAS^V12, knockdown of xCT in SW620 cells (Fig. 3F and SI Appendix, Fig. S2 I and J) and H460 cells (Fig. 3G) led to significant decreases in GSH levels under basal and oxidative stress conditions, which could be reversed by either NAC, but not NAC in the presence of BSO, indicating that these cell lines are also dependent on xCT for GSH synthesis via GCL. Moreover, siRNA-mediated knockdown of xCT led to increased ROS levels under both basal and oxidative stress conditions (SI Appendix, Fig. S2K). Together, these results indicate that oncogenic KRAS promotes xCT expression and activity in human cancer cells to control the redox balance.

ETS-1 Mediates the Induction of xCT by Oncogenic KRAS via Synergistic Cooperation with ATF4.

To elucidate how oncogenic KRAS regulates xCT expression, we analyzed transcription factor networks by performing gene set enrichment analysis (GSEA) of mRNA expression data obtained in 3T3 KRAS^V12 versus 3T3 MSCV cells. This revealed significant enrichment of genes controlled by the transcription factor Ets-1 in oncogenic KRAS cells (SI Appendix, Fig. S3A). Ets-1 is a known substrate of ERK, downstream of the Ras-Raf-Mek pathway, but its effects on xCT transcription have not been reported. Confirming a role for the Ras-Raf-Mek pathway in xCT regulation, pharmacological inhibition of Mek (by PD184352), but not of Akt (Akt Inhibitor VIII), attenuated xCT expression in 3T3 KRAS^V12 cells under both basal and H₂O₂ treatment conditions (SI Appendix, Fig. S3 B and C). This suggests that Ets-1 mediates KRAS^V12 induction of xCT downstream of the Ras-Raf-Mek-Erk pathway. Indeed, knockdown of Ets-1 with two nonoverlapping siRNAs in 3T3 KRAS^V12 cells (SI Appendix, Fig. S3D) significantly restricted xCT induction in response to H₂O₂ treatment versus control siRNAs (Fig. 4A). Accordingly, Ets-1 knockdown increased ROS levels in KRAS^V12-transformed cells under basal conditions, which was rescued by xCT overexpression or NAC supplementation, and dramatically increased ROS levels under H₂O₂ treatment (SI Appendix, Fig. S3E). Moreover, blocking Ets-1 expression increased sensitivity of 3T3 KRAS^V12 cells to H₂O₂ (SI Appendix, Fig. S3F). Therefore, Ets-1 confers KRAS^V12 cytoprotection against oxidative stress, potentially by controlling xCT expression. In human cancer cells harboring KRAS mutations, ETS-1 knockdown similarly decreased xCT expression at both the mRNA and protein level compared with siRNA controls (Fig. 4 B and C). To corroborate this, we ectopically expressed ETS-1 in the DLD-1 colon cancer cell line devoid of ETS-1 expression, which was sufficient to up-regulate xCT expression at both the mRNA and protein level under H₂O₂ treatment (Fig. 4D). Together, these highlight a role for ETS-1 in mediating KRAS^V12 induction of xCT expression.

Next, to demonstrate whether ETS-1 controls the activity of xCT promoter, we performed luciferase transactivation assays in HEK293 cells. Notably, one potential ETS-1 binding site (E1BS), 5′-TGAGGAAGCT-3′ containing the consensus GGAA/T core motif at position –15 of the human xCT promoter, was detected in silico (SI Appendix, Fig. S3G). Indeed, exogenous ETS-1 activates a luciferase reporter containing the xCT promoter in a concentration-dependent manner (Fig. 4E). Furthermore, chromatin immunoprecipitation (ChIP) of 3T3 KRAS^V12 cells revealed that endogenous Ets-1 occupies the promoter region of the xCT gene under both basal and H₂O₂ treatment conditions (Fig. 4F). This strongly supports xCT as a target gene of Ets-1. However, we noticed that the levels of total and phosphorylated Ets-1 (p-Ets-1, the active form of Ets-1; ref. 37), are unchanged following H₂O₂ exposure in 3T3 KRAS^V12 cells (SI Appendix, Fig. S3H). This implies that while Ets-1 can activate xCT transcription downstream of KRAS signaling, there may be other mechanisms to explain how xCT expression is enhanced in response to oxidative stress. Therefore, to uncover other transcription factors potentially involved in xCT regulation specifically under oxidative stress, we employed GSEA in H₂O₂-treated versus untreated 3T3 KRAS^V12 cells. This revealed enrichment of genes regulated by Atf4 under H₂O₂ treatment (SI Appendix, Fig. S4A) and conversely no enrichment of genes regulated by Ets-1. This is consistent with the established role of Atf4 as a transcriptional regulator of xCT expression in response to oxidative stress (17). Indeed, knockdown of Atf4 markedly suppressed xCT induction (SI Appendix, Fig. S4B) and xCT protein expression (SI Appendix, Fig. S4C) in response to H₂O₂ in 3T3 KRAS^V12 and H460 cells, respectively. Notably, knocking down Ets-1 together with Atf4 achieved a stronger suppression of xCT induction under oxidative stress conditions than targeting Ets-1 or Atf4 alone (Fig. 4G), suggesting an interplay between these two transcription factors in regulating xCT expression. While total and phosphorylated Ets-1 are up-regulated in KRAS^V12-transformed cells compared with nontransformed cells but unaffected by H₂O₂ treatment, Atf4 expression is conversely strongly induced by H₂O₂ without any differences between MSCV and KRAS^V12 cells (Fig. 4H). These results suggest that both transcription factors may cooperate to regulate xCT expression in response to both oncogenic KRAS and oxidative stress, with Ets-1 being a component of RAS signaling and Atf4 being independently regulated by oxidative stress.

To determine potential cooperativity between Ets-1 and Atf4 in regulating xCT promoter activity, we constructed a series of xCT promoter constructs containing mutations in the putative E1BS at position −15, or in the known ATF4 binding sites (AARE) at position −95 and −78. Ectopically expressed ETS-1 and ATF4 were able to synergistically activate the wild-type xCT promoter compared with the activity induced by each alone (Fig. 4I; see top bars). However, when either the E1BS alone or the Atf4 AAREs alone were mutated, activation of the luciferase reporter was strongly decreased (Fig. 4I). Moreover, association between endogenous Ets-1 and Atf4 in a putative protein–protein complex within 3T3 KRAS^V12 cells was confirmed by coimmunoprecipitation experiments, and this association was enhanced under oxidative stress (SI Appendix, Fig. S4D). Taken together, these results provide strong evidence that ETS-1 and ATF4 synergistically transactivate the xCT promoter downstream of oncogenic RAS signaling and in response to oxidative stress, possibly as a cotranscriptional activating complex.

The master regulator of the oxidative stress response, Nrf2, has also previously been shown as being induced by oncogenic KRAS (11) and to control xCT promoter activity in response to electrophilic agents in cooperation with Atf4 (38). We therefore investigated the role of Nrf2 in Ets-1 regulation of xCT transcription downstream of oncogenic KRAS. We found that, at least in 3T3 fibroblasts, while Nrf2 was increased following ectopic expression of oncogenic KRAS (or EN), its levels remain unchanged following exposure to H₂O₂ (SI Appendix, Fig. S4E). Furthermore, siRNA-mediated knockdown of Ets-1 (upper blots) or Atf4 (lower blots) failed to alter expression of Nrf2 (SI Appendix, Fig. S4F). Conversely, siRNA-mediated Nrf2 knockdown did not alter Ets-1 or Atf4 expression, nor did ATF4 overexpression alter ETS-1 levels (SI Appendix, Fig. S4G). This indicates that Ets-1 and Nrf2 are independently regulated downstream of the Ras-Raf-Mek pathway and, therefore, likely comprise two distinct pathways controlling xCT promoter activity. As expected, Nrf2 knockdown resulted in decreased xCT transcript levels, which interestingly could be at least partially rescued by Atf4 overexpression (SI Appendix, Fig. S4G). These data suggest that while Nrf2 does not regulate Ets-1 or Atf4 to control xCT transcription, it may potentially coregulate xCT promoter together with these transcription factors. Indeed, luciferase transactivation assays reveal that NRF2 exerts an additive effect with ETS1 or ATF4 or with ETS-1 and ATF4 on xCT promoter activity (SI Appendix, Fig. S4H). However, NRF2 was unable to act in synergy with these transcription factors, contrasting with the effect observed between ETS1 and ATF4. Together, our studies provide evidence that the Ras-Raf-Mek pathway controls the xCT promoter through Ets-1 in synergy with Atf4, but independently of the Nrf2 pathway.

VSports - xCT Supports Oncogenic KRAS Transformation and Tumorigenicity in Vivo.

We next asked whether xCT contributes to KRAS oncogenic transformation and tumorigenicity. We first performed soft agar colony formation assays and found that siRNA-mediated xCT silencing results in ∼70% inhibition of colony formation in 3T3 KRAS^V12 cells (Fig. 5A). Similarly, xCT knockout cells transformed with K-Ras^V12 showed markedly reduced colony formation compared with wild-type cells transformed with K-Ras^V12 (SI Appendix, Fig. S5A). These results were recapitulated in the human cancer cells H460 (Fig. 5B), SW620, and SW480 (SI Appendix, Fig. S5 B and C, respectively) harboring KRAS mutations. Moreover, pharmacological inhibition of xCT using the inhibitor erastin significantly impaired the ability of 3T3 KRAS^V12 and H460 cells to form colonies (SI Appendix, Fig. S5 D and E). These results strongly indicate that xCT is essential for KRAS-mediated oncogenic transformation and maintenance of tumorigenicity in vitro. To explore the potential mechanism involved, we assessed whether xCT reduces oxidative stress, proposed to be critical for oncogenic transformation (39). Notably, addition of NAC partially restored the ability of xCT-deficient and xCT-targeted cells to form colonies in soft agar (Fig. 5 A and B and SI Appendix, Fig. S5 A–C), consistent with an important role for xCT in supporting oncogenic KRAS-mediated transformation through its antioxidant function. To confirm that ETS-1 and ATF4 also contribute to oncogenic KRAS transformation by regulating xCT transcription, we carried out siRNA-mediated silencing of ETS-1 and ATF4 alone, or in combination. We found these to result in more than 70% inhibition of colony formation of 3T3 KRAS^V12 cells, compared with siRNA controls, which could be partially rescued with xCT overexpression or supplementation with NAC (SI Appendix, Fig. S5F). This highlights that ETS-1 and ATF4, by regulating xCT transcription, can mediate oncogenic KRAS transformation.

To determine the in vivo relevance of these findings, 3T3 KRAS^V12 cells stably expressing scr or two individual shRNAs targeting xCT were s.c. implanted in nu/nu immunocompromised mice. Tumor xenografts established from 3T3 KRAS^V12 cells expressing xCT-specific shRNA were severely impaired in their growth compared with tumors established from control scr cells (Fig. 5C). This data were accompanied by dramatically improved survival of mice bearing tumors with xCT silencing compared with control tumors (Fig. 5D). These highlight that xCT is required for KRAS-mediated tumorigenesis in vivo. To determine whether the impairment of KRAS-driven tumor growth induced by xCT deficiency is associated with oxidative stress, we measured GSH levels and found that tissues from xCT-knockdown tumors exhibited significantly reduced GSH levels compared with control tumors (Fig. 5E). Further, immunoblots performed on xCT knockdown tumor tissues demonstrated an increase in levels of the oxidized protein marker, Dityrosine, relative to controls, providing evidence for increased oxidative stress levels in these tumors (Fig. 5F). These data support a model whereby xCT is critical for supporting KRAS oncogenic transformation and tumorigenicity in vitro and in vivo by increasing antioxidant capacity and mitigating oxidative stress.

High xCT Expression Predicts Poor Outcome in Mutant KRAS-Driven Human Tumors.

We next investigated the relevance of these findings in preclinical and clinical models of RAS-driven cancers. We first interrogated mRNA expression data from genetically engineered mouse models and found that lung tumor specimens from transgenic mice expressing Kras^G12D mutant display higher levels of xCT mRNA compared with those from oncogenic c-Myc–expressing mice or from normal lung tissue (Fig. 6A). Additionally, elevated levels of xCT mRNA were also found in lung tumor specimens from transgenic mice expressing either EGFR in-frame exon 19 deletion mutant or the EGFR^L858R mutant, providing evidence that constitutively active RAS signaling in addition to expression of mutant RAS proteins are associated with xCT induction.

These findings were confirmed in clinical cancer specimens by analyzing publicly available gene expression data from lung and colon cancer and glioma patients. This revealed that xCT expression is up-regulated in patient tumors that are positive for KRAS mutations (Fig. 6 B and C). Interestingly, xCT expression is also up-regulated in glioma tumors that are positive for EGFR amplification, further highlighting that constitutive RAS signaling is associated with xCT induction (Fig. 6D). Furthermore, hypergeometric analyses of gene expression datasets from TCGA LUAD, LUSC, and COAD cohorts revealed that the top 1% genes positively coexpressed with xCT are those that are significantly enriched in association with constitutively active MEK (SI Appendix, Fig. S6A). Similar analyses of these gene expression datasets also showed that the top 5% genes positively coexpressed with xCT are significantly enriched for genes controlled by the ETS-1 transcription factor (SI Appendix, Fig. S6B). Altogether, these results suggest that xCT is elevated downstream of overactive RAS-RAF-MEK-ERK-ETS-1 signaling axis in human tumors harboring KRAS mutations.

Finally, we assessed the prognostic value of xCT expression in colon cancer and glioma patients. We observed that high xCT levels correlates with poorer outcomes across all colon cancer subgroups (Fig. 6E) and in glioma patients (Fig. 6F). This is in line with previous reports in triple-negative breast cancer (20) as well as nonsmall cell lung cancer (18), and it suggests that xCT predicts poor prognosis in tumor types associated with increased RAS pathway activation. In summary, these data strongly suggest that xCT expression has prognostic value in KRAS mutant-expressing tumors, further reinforcing the link between oncogenic KRAS and xCT expression.

Cell Culture.

NIH 3T3 cells stably expressing EN or KRAS^V12 were as described (60). xCT^−/− (KO) and xCT^+/+ (WT) MEFs were a kind gift from Hideyo Sato (Niigata University, Niigata, Japan) and routinely cultured in DMEM supplemented with 10% FBS (36). xCT^−/− MEFs were also supplemented with 50 µM 2-mercaptoethanol (Sigma-Aldrich). H460 and SW620 cell lines were purchased from American Type Culture Collection (Rockville, MD) and maintained in RPMI 1640 media supplemented with 10% FBS. Nras/Hras double knockout MEFs with 4-OHT–inducible knockout of endogenous Kras were kindly provided by Mariano Barbacid, National Centre for Cancer Research, Madrid (32).

Tumorigenicity Assay.

Aliquots of 0.5 × 10⁶ cells were resuspended in 200 μL of PBS and injected s.c. into the flanks of 5- to 6-wk-old female Nu/Nu immunodeficient mice using standard procedures. Starting from day 13 after injection, mice euthanasia was required when tumors exceeded humane practice guidelines (500 mm³). Mice were evaluated for tumor growth every 2 d until the experimental endpoints. Tumors were measured with a caliper, and volumes were estimated using the following formula: tumor length × (tumor width)² × π/6 mm³. All animal experiments underwent ethical approval from the Animal Care Committee of the University of British Columbia (A16-0050; A16-0050-A001). Full material and methods are available in SI Appendix, Materials and Methods.

VSports最新版本 - Gene Expression of Clinical Cancer Specimens.

Analyses of gene expression data from human subjects did not require approval from the institutional review board (IRB), as the data were all obtained from publicly available databases in which each study had utilized strict human subjects protection guidelines, informed consent, and respective IRB review of protocols.