Relationship between LAPTM4B Gene Polymorphism and Prognosis of Patients following Tumor Resection for Colorectal and Esophageal Cancers

Background Lysosome-associated transmembrane-4 beta (LAPTM4B) is an oncogene that participates tumorgenesis in a variety of human solid tumors, and it has two alleles named as LAPTM4B*1 and *2. The present study aimed to identify the association of LAPTM4B genotype with clinicopathological features and prognosis in colorectal and esophageal cancer patients. Method Genotypes of LAPTM4B were determined by PCR in 167 colon cancer cases (72 patients in a discovery cohort and 95 patients in a testing cohort), 160 rectal cancer cases and 164 esophageal cancer cases. Association between the LAPTM4B gene polymorphism and clinicopathological variables was calculated by Chi-square test or Fisher’s exact test. Patient survival differences were calculated by the Kaplan-Meier method. Prognostic factors were determined with Log-rank test and Cox regression model. Results LAPTM4B *1/1 was more frequently detected in colon cancer patients with lymph node metastasis and TNM III+IV stages in total colon cancer (discovery + testing cohorts). LAPTM4B *2/2 decreased in recurrent patients in total colon cancer patients (P = 0.045). Kaplan-Meier survival curves and Log-rank test showed that LAPTM4B*1 was correlated with shorter overall survival (OS) in discovery and testing cohorts of colon cancer (P = 0.0254 and 0.0292, respectively), but not in rectal and esophageal cancer cases (P = 0.7669 and 0.9356, respectively). Multivariate analysis showed that LAPTM4B genotype was an independent prognostic factor for OS in total colon cancer [P = 0.004, hazard ratio (HR) = 0.432; 95% confidence interval (CI) = 0.243–0.768], but not in rectal and esophageal cancers (P = 0.791, HR = 1.073, 95% CI = 0.638–1.804 and 0.998, HR = 1.000, 95% CI = 0.663–1.530, respectively). Conclusion These findings suggested that LAPTM4B allele *1 was a risk factor associated with poor prognosis in patients with colon cancer, but not in patients with rectal or esophageal cancers. LAPTM4B genotype status might be a useful prognostic indicator for patients that need surgical operation in colon cancer.


Introduction
Colorectal and esophageal cancers are common malignant digestive diseases with high incidence and mortality worldwide [1][2][3]. Nearly one-half patients are diagnosed at an advanced stage, and there are still lack effective targeted therapies in colorectal and esophageal cancers, so survival rates in these cancers have not been improved markedly compared with hematopoietic and lymphoid malignancies. However, it is beneficial to monitor cancer progression for cancer patients [4].
According to our previous report, the LAPTM4B Ã 2 allele frequency was 33.2% in colon cancer group, 25.5% in rectal cancer group, 22.7% in esophageal cancer group and 24.1% in health control group, indicating that LAPTM4B Ã 2 was correlated with increased risk of colon cancer (P = 0.0016), but not with that of rectal and esophageal cancers [20]. However, there was no report about the association between the existence of two variant alleles of LAPTM4B with the prognosis in patients with colorectal and esophageal cancers. The present study aimed to investigate whether there is a correlation of LAPTM4B gene polymorphism with prognosis in colorectal and esophageal cancer patients after surgical resection.

DNA extraction and PCR analysis
Total genomic DNA was isolated from peripheral white cells using Blood Genomic DNA extraction kit following the manufacturer's instructions (Tiangen Beijing, China). DNA was dissolved in elution buffer, and its concentration was measured with a Nanodrop 2000 spectrophotomer (Thermo Fisher Scientific, Wilmington, Delaware, USA) and stored at -80°C until use.

Statistical analysis
Statistical analysis was carried out by SPSS20.0 software (SPSS Inc., Chicago, IL). Chi-square test or Fisher's exact test was used to assess the correlation between the genotype and clinical parametric distributions in colorectal and esophageal cancer patients. The association of LAPTM4B gene polymorphism with overall survival (OS) was analyzed using Kaplan-Meier curves and log-rank test. Multivariate analysis determined the potential independent prognostic factors with Cox regression model. All tests of statistical significance were two-sided. P < 0.05 was used as statistically significant level.

Association between LAPTM4B genotypes and clinicopathological parameters in colorectal and esophageal cancers
The distribution of different genotypes of LAPTM4B was analyzed in clinicpathological parameters, including age, gender, lymph node metastasis, depth of invasion, distant metastasis, differentiation degree, gross type, TNM stage, location for colon cancer, CEA (carcinoembryonic antigen) for colorectal cancer and recurrence ( Table 1).The association of different genotypes of LAPTM4B with these clinical variables of colorectal and esophageal cancer patients did not reach statistical significance in our study. LAPTM4B Ã 1/1 was more frequently detected in colon cancer patients with lymph node metastasis and TNM III+IV stages compared with non-lymph node metastasis and TNM I+II stages in total colon cancer (discovery + testing cohorts) (P = 0.106, 29.6% vs. 46.6% and P = 0.157 31.4% vs. 46.6%, respectively). LAPTM4B Ã 2/2 decreased in recurrent patients compared with non-recurrent ones in total colon cancer patients (P = 0.045, 7.4% vs. 20.5%) ( Table 1) and in discovery and testing cohort (P = 0.203,  7.1% vs. 22.7% and P = 0.368, 7.6% vs. 17.2%, respectively) (S1 Table). However, LAPTM4B Ã 1 was more frequently detected in colon cancer patients with moderate and well differentiation in colon discovery cohort (P = 0.011) (S1 Table). Such association was not found in colon testing cohort and total colon cancer cases.
The same tendency of LAPTM4B Ã 1/1 and LAPTM4B Ã 1 was observed in the colon discovery and testing cohorts of colon cancer (P = 0.0417 and 0.0444, P = 0.0254 and 0.0292, respectively) ( Fig 2C-2F). However, we did not find a marked relation between LAPTM4B genotypes or alleles and OS for rectal and esophageal cancer cases (P = 0.7418 and 0.8520 for various genotypes vs. P = 0.7669 and 0.9356 for different alleles, respectively) (Fig 3A-3D).
LAPTM4B genotype was an independent prognostic marker in patients with colon cancer, but not for rectal and esophageal cancer patients The univariate Cox's model for OS of colorectal and esophageal cancer patients displayed that LAPTM4B genotype was one of the prognostic factors in total colon cancer patients (P = 0.006; HR: 0.565, 95% CI: 0.377-0.846), not in rectal and esophageal cancer patients (P = 0.692 and 0.958, respectively) ( Table 2). Furthermore, LAPTM4B genotype was a novel independent prognostic factor of OS for colon cancer (P = 0.004, HR = 0.432, 95% CI: 0.243-0.768), even in the colon discovery cohorts (P = 0.007, S2 Table), but not for rectal and esophageal cancer patients (P = 0.791, HR = 1.073, 95% CI: 0.638-1.804 vs. P = 0.998, HR = 1.000, 95% CI: 0.663-1.530, respectively). Depth of invasion, distant metastasis and recurrence were also independent prognostic factors for colon cancer (P = 0.014, P<0.001 and P<0.001, respectively). For rectal cancer patients, distant metastasis and recurrence were also independent prognostic factors (P = 0.021 and P<0.001, respectively). In addition, distant metastasis (P = 0.025) was also an independent prognosis factor in esophageal cancer patients (Table 3).

Discussion
Previous studies have demonstrated that LAPTM4B can play critical roles in various solid tumors, including proliferation, migration, invasion, apoptosis and angiogenesis [12,[27][28][29]. It also motivated multidrug resistance through promoting drug efflux by interacting with P-gp and activating PI3K/AKT signaling pathway [30]. In addition, new evidence has also revealed that LAPTM4B can participate in the autophagy initiation through binding with inactive epidermal growth factor receptor (EGFR) [31,32].
In the present study, we revealed an independent prognostic role of LAPTM4B gene polymorphism in colon cancer patients who received surgical resection, but not in rectal and esophageal cancers. In our study, the LAPTM4B Ã 2 allele frequency rate are 33.3%, 27.8% and 23.8%, nearly the same as previous report in colon, rectal and esophageal cancers [20], respectively. For the clinicopathological parameters, LAPTM4B genotype was correlated with recurrence in total colon cancer, especially for LAPTM4B Ã 2/2 which decreased in recurrent colon cancer patients. There was not a relationship between LAPTM4B genotype and other clinical factors in colorectal and esophageal cancer patients. However, LAPTM4B Ã 1 was more frequently detected in colon cancer patients with moderate and well differentiation in colon discovery cohort. This phenomenon might be caused by incomplete clinicopathological parameters in the present work. Furthermore, LAPTM4B Ã 1/1 tended to be frequently detected in patients with lymph node metastasis and TNM III+IV stages in total colon cancer cases. Patients with LAPTM4B Ã 1 (genotypes Ã 1/1) had a significantly poorer overall survival when compared with LAPTM4B Ã 2 (genotypes Ã 1/2 or Ã 2/2) patients in colon cancer (discovery and testing cohorts), but not in rectal and esophageal cancers. This is the first time demonstrating the LAPTM4B Ã 1 allele as a poor prognostic indicator. The association of LAPTM4B Ã 2 allele with colon cancer prognosis is not consistent with recent reports including in hepatocellular [23], ovary [27], lung [23], breast cancer [25] and etc. In hepatocellular carcinoma, Yang et al indicated that LAPTM4B Ã 2 was correlated with tumor recurrence, poor histopathological differentiation and also an independent prognostic factor. Previous studies indicated that the 19-bp difference in 5' untranslated region of the first exon of the LAPTM4B gene can alter the ORF, resulting in two different protein isoforms: LAPTM4B-35 and -40 [14]. It might suggest that the 19-bp sequence plays an important role in transcriptional regulation or new isoform produced by LAPTM4B Ã 2 may influence physiological activity and function of cancer cells. Whereas in our study, LAPTM4B Ã 1 allele shows a significant correlation with overall survival of colon cancer patients, but not in rectal and esophageal cancer patients. One explanation might be: the diverse expression patterns or isoforms of LAPTM4B in epithelial cells might demonstrate the difference of LAPTM4B genotype in prognosis in colon cancer vs rectal and esophageal cancers; the other explanation might be: the 19-bp sequence may play an important role in transcriptional regulation such as binding with the transcription factors or non-coding linker RNA in different cancers. Furthermore, in patients with gastric cancer, we have found that even though LAPTM4B genotype was correlated with susceptibility of gastric cancer, this polymorphism did not correlate with prognosis (data unpublished). The phenomenon illustrated the tumor heterogenicity between LAPTM4B genotype and its function, which discriminates with that in hepatocellular carcinoma, breast cancer and etc.
LAPTM4B was obviously up-regulated in various types of cancers [6]. Its overexpression might be caused by gene amplification and transcriptional up-regulation. However, the specific  [29,33,34]. Their balance may affect malignant transformation. However, a recent report has shown that LAPTM4B-24 isoform can stimulate mTORC1 via V-ATPase by influx of leu through binding with LAT1-4F2hc to lysosomes [35]. LAPTM4B-24 can also promote cell growth and proliferation [35]. As a result, the different LAPTM4B isoforms may play diverse functions in LAPTM4B Ã 1 patients when compared with LAPTM4B Ã 2 patients with colon cancer. Our findings on LAPTM4B alleles in colon cancer provide additional evidence that different LAPTM4B isoforms might play various roles, that is, LAPTM4B-35 can activate PI3K/Akt pathway and LAPTM4B -24 can activate mTORC1 pathway. Different isoform pattern might induce the function of 19 bp sequence in various cancers. Further studies should be carried out to elucidate this phenomenon. However, LAPTM4B genotype will be a useful biomarker for colon cancer patients when considering curative surgical resection.
Supporting Information S1 Table. Correlation of distribution of various genotypes of LAPTM4B with clinicopathological parameters in discovery and testing cohorts of colon cancer patients.