High expression of Solute Carrier Family 1, member 5 (SLC1A5) is associated with poor prognosis in clear-cell renal cell carcinoma

Solute Carrier Family 1, member 5 (SLC1A5), also named as ASCT2, a major glutamine transporter, is highly expressed in various malignancies and plays a critical role in the transformation, growth and survival of cancer cells. The aim of this study was to assess the clinical significance of SLC1A5 in patients with clear-cell renal cell carcinoma (ccRCC). SLC1A5 expression was evaluated by immunohistochemistry on tissue microarrays. Kaplan-Meier method was conducted to compare survival curves. Univariate and multivariate Cox regression models were applied to assess the impact of prognostic factors on overall survival (OS). A nomogram was then constructed on the basis of the independent prognosticators identified on multivariate analysis. The predictive ability of the models was compared using Receiver operating characteristic (ROC) analysis. Our data indicated that high expression of SLC1A5 was significantly associated with advanced TNM stage, higher Fuhrman grade and shorter OS in ccRCC patients. Multivariate analysis confirmed that SLC1A5 was an independent prognosticator for OS. A nomogram integrating SLC1A5 and other independent prognosticators was constructed, which showed a better prognostic value for OS than TNM staging system. In conclusion, high SLC1A5 expression is an independent predictor of adverse clinical outcome in ccRCC patients after surgery.

Scientific RepoRts | 5:16954 | DOI: 10.1038/srep16954 pyrimidine synthesis, glutamine is converted to glutamate, which is the main nitrogen donor for the synthesis of nonessential amino acids 5 . Also, glutamate is the precursor of the major cellular antioxidant, glutathione 6 . Glutamate can be exploited to form α -ketoglutarate, an important tricarboxylic acid (TCA) cycle metabolite, when losing its amine group. Then, α -ketoglutarate can be converted to citrate via several enzymatic steps, which would both produce acetyl-coA for the lipid synthesis and be converted to malate for the subsequent production of NADPH 4 . Furthermore, glutamine has been regarded as an essential activator for the mammalian target of rapamycin complex (mTORC)1, which regulates protein translation, cell growth and macroautophagy 4,7 . Besides being utilized in anabolic metabolism, the exported glutamine serves as stimuli for the uptake of essential amino acids, which directly activates mTORC1 and protein synthesis 8 . Consequently, glutamine plays a pivotal role in many cancers including ccRCC. Cross-platform molecular analyses have revealed that increased glutamine transport expression correlated with poor prognosis in patients with ccRCC 9 .
Glutamine can be transported by four families of amino acid transporter systems including sodium-independent system L and system b 0,+ and sodium-dependent system A, B 0,+ , y + L, ASC and N. Among them, system ASC is the most ubiquitously expressed one in human cancer cells 10 . Solute Carrier Family 1, member 5 (SLC1A5), also named as ASCT2, belongs to system ASC and performs as a high-affinity glutamine transporter in cancer cells 11 . Consistent with the functions exerted by glutamine, SLC1A5 is profoundly involved in uptake of essential amino acids, activation of mTORC1 and glutamine-dependent tumor cell survival and growth 8 . To date, SLC1A5 has been regarded as an indispensable "switch" of glutamine metabolism and thus to be a critical regulator for cancer development 12 . Several studies have revealed that SLC1A5 is highly expressed in several cancer types and its expression is closely correlated with tumor development and prognosis [13][14][15][16][17][18] . It also has been determined that the mRNA level of SLC1A5 is significantly higher in tumor tissues than normal tissues of patients with ccRCC 9 . However, the protein level and clinical significance of SLC1A5 expression in ccRCC remains unclear.
Here, we sought to investigate the association of SLC1A5 expression with clinicopathologic features and patient outcomes. Moreover, a nomogram integrating SLC1A5 expression and pathologic variables was established to help predict prognosis and guide management for ccRCC patients after surgery.

Results
Immunohistochemical finding and association between SLC1A5 expression and clinicopathologic characteristics in patients with ccRCC. To investigate whether the expression pattern of SLC1A5 is altered during tumorigenesis of ccRCC, we evaluated SLC1A5 expression levels in normal kidney tissues (n = 3), ccRCC tumor tissues (n = 10) and corresponding peri-tumor tissues (n = 10) by qRT-PCR analysis. As presented in Figure S1, SLC1A5 expression was significantly upregulated in tumor tissues when compared with corresponding peri-tumor tissues (P = 0.003) and normal kidney tissues (P = 0.007). To further evaluate the protein level of SLC1A5 in ccRCC tumor tissues, we detected the expression of SLC1A5 by immunohistochemical staining analysis in 187 patients with ccRCC. As presented in Fig. 1a, the expression of SLC1A5 was mainly appeared in the membrane of tumor cells and the staining intensity and distribution were variable in different specimens. The H-score of SLC1A5 expression ranged from 9 to 255, According to the "minimum P value" approach conducted by X-tile, 149 was determined as the cutoff to dichotomize the patients into SLC1A5 low group (score, 9-149; n = 101) and SLC1A5 high group (score, 150-255; n = 86) (Fig. 1b).
As summarized in Table 1, patients with higher expression of SLC1A5 significantly tended to diagnosed with higher TNM stage (P = 0.032) and higher Fuhrman grade (P = 0.015). No other clinicopathologic variables were presented to have a significant correlation with SLC1A5 expression. SLC1A5 expression correlated with OS, but not RFS, of patients with ccRCC. Kaplan-Meier method and log-rank test were performed to assess the relationship between SLC1A5 expression and clinical outcome of patients with ccRCC. The median follow-up was 106 months (range, 6-120 months). As shown in Fig. 2a, SLC1A5 expression was significantly associated with OS (P < 0.001) of ccRCC patients, and so that means higher SLC1A5 expression indicated earlier death. Nevertheless, SLC1A5 expression presented no significant correlation with RFS of patients (P = 0.083, Fig. 2b).
Then the above significant factors were brought into the multivariate cox analysis. Results in Table 2 indicated that pT-stage (P = 0.005), pN-stage (P = 0.043), metastasis (P < 0.001), Fuhrman grade (P = 0.048), sarcomatoid (P = 0.038) and SLC1A5 (P = 0.001) in multivariate analysis still had statistical significance and were determined as independent prognostic factors in ccRCC.

Prognostic nomogram for survival of patients with ccRCC. Further multivariate analysis was
conducted among the independent prognostic factors ( Table 2). Thus we constructed a prognostic nomogram via integrating all these independent prognostic indicators for OS (Fig. 3a). The calibration plot for the probability of overall-survival at 3-, 5-or 10-year after surgery presented an optimal agreement between actual observation and the prediction by nomogram (Fig. 3b-d).
Superior performance of the nomogram-based model. According to the nomogram, each subgroup of every independent prognostic indicator had been assigned a risk score (Table 3). Then each patient in this study was scored by the nomogram based risk score.
Patients were preliminarily stratified according to the nomogram-based risk scores by Kaplan-Meier curves (Fig. 4) and eventually grouped into four subgroups via combining the curves with similar prognosis. The newly established nomogram-based model included four risk subgroups which represented significant distinct prognosis with each other (Fig. 5a, P < 0.001).
To further determine the effectiveness of the nomogram-based model, receiver operating characteristic analysis was conducted. As presented in Fig. 5b, the nomogram-based model (Area under curve

Discussion
To our knowledge, this study was the first to report an association between high SLC1A5 expression and poor prognosis in ccRCC patients following surgery. Moreover, SLC1A5 expression has been identified as an independent prognostic factor and could be used to construct a nomogram with established pathologic factors. This work indicated that SLC1A5 might play an important role in the development of ccRCC. Recent studies demonstrated that SLC1A5, the primary glutamine transporter, could promote tumor cell growth, cell cycle progression and survival in neuroblastoma, colorectal cancer, breast cancer and prostate cancer [19][20][21][22] . As glutamine is an important immunomodulatory nutrient, SLC1A5 also has been determined to involve in inflammatory T cell responses, which might exert key functions in tumor immunity 23 . Thus, the molecular mechanism and functional significance of SLC1A5 in ccRCC merits further investigation. SLC1A5 expression has been detected in brain, lung, skeletal muscle, testis, adipose tissue, large intestine and kidney 24 . As an important neutral amino acid transporter, SLC1A5 has been documented to transport serine, alanine, cysteine, threonine, glutamine, asparagine and so on 24 . Its ubiquitous tissue expression, along with its ability to transport certain key amino acids, indicates that SLC1A5 plays a crucial role in physiological processes including glutamine homeostasis, embryogenesis and retroviral infection 12,25,26 . In the current study, our initial work revealed that SLC1A5 expression was markedly upregulated in tumor tissues when compared with corresponding peri-tumor tissues and normal kidney tissues, which drawn our attention on the clinical significance of SLC1A5 expression in tumor tissues of ccRCC patients.
Consistent with previous findings in other cancers types, high SLC1A5 expression was identified as an independent adverse prognostic factor for OS in ccRCC. Considering the heterogeneity of ccRCC and its unpredictable natural history, SLC1A5 might be useful in risk stratification and personalizing postsurgical surveillance. Furthermore, the nomogram constructed in our study performed well in outcome prediction and provided a novel prognostic system for clinical practice.
Besides its significance in prognosis, the value of SLC1A5 in cancer treatment has also drawn more and more attention these years. Indeed, studies have demonstrated that blocking glutamine uptake might be an attractive strategy for cancer therapy and SLC1A5 represents as a potential therapeutic target in this pathway 27,28 . Since it is profoundly involved in mTOR activation, SLC1A5 exerts functions partly via mTOR signaling and thus inhibiting SLC1A5 expression could diminish the oncogenic effect of mTOR pathway in several cancer types [29][30][31] . mTOR signaling has been revealed to be maladjusted in ccRCC and molecularly targeted therapies against mTOR have gradually become one of the mainstream strategies in treatment of advanced ccRCC 32 . Hence, SLC1A5 might be an appealing target in the treatment of ccRCC, which also needs our further exploration.
Although the clinical significance of SLC1A5 in ccRCC has been revealed, several limitations of this study warrant further discussion. Firstly, the number of patients enrolled in this study was small, especially for the patients with advanced disease. Secondly, an independent external cohort is necessary to confirm our findings. Thirdly, the association between SLC1A5 expression and mTOR activation in ccRCC needs to be demonstrated. Finally, functional studies should be conducted to uncover the biological mechanisms of SLC1A5 in ccRCC.
In summary, the current study demonstrated that high SLC1A5 expression was an independent adverse prognostic factor for OS in ccRCC patients. A prognostic nomogram integrating SLC1A5 expression and pathologic factors may improve the postsurgical management of ccRCC patients. Functional studies are needed to evaluate the molecular mechanisms of SLC1A5 in the tumorigenesis of ccRCC and its role as a therapeutic target.

Methods
Patients. All methods were approved by the research medical ethics committee of Fudan University and were carried out in accordance with the approved guidelines. Informed consent on the use of clinical specimens had been received from each patient. A total of 187 consecutive patients with ccRCC who underwent radical or partial nephrectomy at Zhongshan Hospital, Fudan University (Shanghai, China) were enrolled to construct the tissue microarray in the current study. These specimens of patients were collected between January 2003 and December 2004. Another three normal kidney tissues and ten pairs of tumor tissues with corresponding peri-tumor tissues were obtained to isolate RNAs from the same institution. Patients who received preoperative neoadjuvant and/or postoperative adjuvant therapy were excluded in this study. Tumor stages were histologically classified according to 2010 AJCC TNM classification 33 . Overall survival (OS) and recurrence-free survival (RFS) were calculated from the date of surgery to the date of death (or the last follow-up) or to the date of recurrence (or the last follow-up), respectively. Patients with tumor metastasis at the time of surgical operation were excluded from the RFS analysis as indicated by the end point of RFS.
Immunohistochemistry. Tissue microarrays and immunohistochemistry analysis were performed as previously described 35 . Primary anti-SLC1A5 antibody (1:300; Abcam, Cambridge, MA) was used for immunohistochemistry staining. The intensity of immunostaining was evaluated by two independent pathologists without the knowledge of clinicopathological data. A semi-quantitative H-score which ranged from 0 to 300 was calculated by multiplying the staining intensities (0: negative, 1: weak, 2: moderate, 3: strong) by the distribution areas (percentage of positive staining cancer cells, 0-100%) at each intensity level for each sample. Statistical Analysis. X-tile plot analysis was conducted to select the optimum cutoff of the H-score to dichotomize the patients into low and high groups 36 . Comparisons between SLC1A5 expression and clinicopathologic variables were evaluated using Student's t test, χ 2-test and Wilcoxon rank-sum test, as appropriate. Survival curves were conducted by Kaplan-Meier method and compared by log-rank test. Univariate and multivariate Cox proportional-hazard models were exploited to evaluate the hazard ratios and 95% confidence intervals of clinicopathologic variables. Nomogram was set to construct the prognostic model. Calibration plot was used to evaluate the prognostic accuracy of the models. Receiver operating characteristic analysis was conducted to compare the sensitivity and specificity for the prediction of OS by the prognostic models. All statistical tests were two-tailed and differences were  Table 3. Point assignment and nomogram-based risk score. SLC1A5 = Solute Carrier Family 1, member 5.