
Citation: | Pei Chen, Shiwen Wang, Bo Chen, Bing Lu, Xiaojing Zhang, Xiaohui Jiang, Yao Wang. High LIMP2 expression serves as a predictor for improved clinical outcomes in gastric cancer patients[J]. Blood&Genomics, 2020, 4(2): 142-150. DOI: 10.46701/BG.2020022020119 |
Gastric cancer (GC) is the fifth most frequently occurring malignancy and the third leading cause of cancer-related mortality worldwide[1]. Although rates of gastric cancer incidence have decreased in Western countries, they are increasing in China[1]. During the early stages of GC, most patients are asymptomatic, and consequently, diagnosis is frequently made when the disease is at an advanced stage[2]. The 5-year net survival rate of patients with GC in China is 30%-39% due to postsurgical recurrence and metastasis[3]. Thus, it is of paramount importance to develop methods for evaluating and improving patients' survival.
The Cancer Genome Atlas (TCGA) program collected over 11 000 primary cancer samples, which provided abundant information for researchers. When we analyzed the TCGA data, we found that LIMP2 expression positively associated with oncogenes and tumor suppressor genes in the stomach adenocarcinoma database and is upregulated in human GC compared with normal human gastric samples.
Scavenger receptor class B member (SCARB2) is the gene of LIMP2 protein. LIMP2 is an abundantly expressed type Ⅲ double-transmembrane glycoprotein (MIM: 602257, Locus: NM_005506) with N-as well as C-terminal cytoplasmic tails consisting of 20 amino acids, and LIMP2 is primarily expressed in lysosomes and late endosomes in diverse cell types[4-5]. Sequence analysis classified LIMP2 as a member of the class B scavenger receptor family that includes CD36 and SR-B1[6-7]. LIMP2 and glucocerebrosidase(GCase) interact in a highly specific and pH-dependent manner in the endoplasmic reticulum, to mediate transport to lysosomes and late endosomes[8].
Deficiency in LIMP2 decreases the levels and activity of lysosomal GCase, accompanied by the accumulation of glucosylceramide, which eventually causes Gaucher disease and contributes to Parkinson disease[9-11]. Furthermore, LIMP2 functions as an efficient plasma membrane-localized receptor that mediates the uptake of enterovirus 71 and coxsackievirus A16, which are major causative agents of hand, foot and mouth disease[12-13]. Moreover, LIMP2 is required for the maintenance of endosomes and lysosomes and lipid transport[14-15]. LIMP2 mutations in humans may represent the major cause of myoclonus-renal failure syndrome, and evidence indicates that LIMP2 deficiency in mice causes deafness and peripheral neuropathy[16-17]. The level of LIMP2 mRNA increases approximability 2-4 fold in prostate cancer cells vs. nonmalignant cells[18]. Upregulation of LIMP2 expression can predict lymph node metastases in oral squamous cell carcinoma[19]. However, the expression levels and effects of LIMP2 on GC have never been studied before and are worthy of exploration.
In this study, we showed that LIMP2 mRNA expression levels correlated with upregulated tumor-associated genes in corresponding tumors in the TCGA database. The levels of LIMP2 were elevated in human gastric cancer tissues compared with normal tissues. Further, univariate and multivariate analysis revealed the potential of LIMP2 expression for predicting metastasis and survival of patients with GC.
Paraffin-embedded tumor tissues (n = 329) were collected from patients with gastric cancer who underwent curative primary tumor resection with or without lymph node dissection at the Affiliated Hospital of Nantong University (Jiangsu, China), between April 2004 and December 2010. Clinical data including name, sex, age, pathological diagnosis, histological type, differentiation grade, TNM stage, lymph node metastasis, and preoperative serum CEA level were retrospectively collected from medical records. Patients who met the criteria as follows were selected: ①histological and clinical diagnosis of GC; ②no preoperative chemotherapy, radiotherapy, immuno-therapy, or molecularly targeted therapy; ③no serious complications; ③clinical and follow-up information available.
Data for survival, death, and the causes of death after surgery were updated through a review of the patients' medical records and telephone calls. We calculated the OS from the date of surgery to the date of death or last follow-up. The Human Research Ethics Committee of the Affiliated Hospital of Nantong University approved the study protocol. The research conducted in this study complied with the terms of the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study.
As previously described[20], tissue microarrays were baked for 2 hours at 60℃, deparaffinized in xylenes, and rehydrated through an ethanol series (70%, 80%, 95% and 100%). Endogenous peroxidase activity was quenched using 3% H2O2 and methanol for 15 min. Next, the sections were boiled in citric acid (MVS-0101, MXB Biotechnologies, FuZhou, China) for 30 min in a pressure cooker. Tissues were then incubated with primary mouse anti-LIMP2 antibody (ab176317, Abcam, Cambridge, UK), diluted 1:400, overnight at 4℃ in a moist chamber. The primary antibody was removed using PBS (3 × 5 min) and incubated with a horseradish peroxidase (HRP)-conjugated secondary antibody (kit-0015, MXB) for 30 min at room tem-perature. Subsequently, the sections were stained using 3, 3´-diaminobenzidine (kit-0015, MXB) and counterstained with hematoxylin. The sections were then dehydrated and mounted on a glass slide.
The Vectra Multispectral Imaging System (Perkin Elmer, Hopkinton, MA, USA) was used to analyze the IHC slides. Briefly, entire stained images were scanned, and multispectral image cubes were acquired using the Phenochart Whole Slide Contextual Viewer (Perkin Elmer). InForm v2.2 software was trained to identify segment tissue and extraneous substances. The nucleus of each cell was visualized using hematoxylin staining, and a tissue size range 30–50 µm was used to identify the nucleus to further improve the accuracy of identification. The cytoplasmic signal was obtained by sampling the perinuclear area. With the software trained to identify tissue and cell segmentation, the InForm system can effectively identify the nucleus and cytoplasm of each cell, and the data can be automatically applied to the 329 samples. To evaluate the accuracy of our image analysis system for identifying cell segmentation, two pathologists evaluated the image segment results. The accuracy rate for identifying cells reached 95%.
Staining intensity scores included staining that was undetectable (score: 0), weak (score: 1), moderate staining (score: 2), and strong (score: 3)[21]. The frequency of each intensity category (0, 1+, 2+, and 3+) cell was automatically counted in each core. The H score of LIMP2 expression was equal to the sum of the percentages of cells (0%–100%) in each intensity category. The final score was assigned using a con-tinuous scale between 0 and 300. With the help of Xtile software, a score of 66 was used as a cutoff for classifying patients into either the LIMP2high or LIMP2low groups.
Normalized gene expression data for normal and cancer tissues, profiled using the TCGA database were downloaded on April 19, 2019. Gene expression values are presented as X = log2(X+1), where X is the normalized value of the fragments per kilobase transcript per million mapped reads. RNAseq data for 407 unmatched GC samples were used to analyze the levels of LIMP2 expressed by GC and normal gastric tissue samples.
GSEA is a threshold-free method for analyzing genes according to their expression ranks without gene filtering[22]. The levels of LIMP2 mRNA expressed by TCGA samples of stomach adenocarcinomas were previously grouped as LIMP2high or LIMP2low according to the median value. GSEA was performed according to the correlations of a list of genes with LIMP2 expression. The pathway comprising genes enriched at the top or bottom of the ordered gene list was searched using GSEA. This process was quantized using an enrichment score that reflects pathway enrichment in the top or bottom of the ranked gene list. The Q value (Q < 0.05) of the false-discovery rate was used to select pathways, and a normalized enrichment score reflecting pathway enrichment in this list was used to rank the identified pathway.
TIMER is computational tool used to investigate the molecular characteristics of tumor-immune inter-actions[23]. The expression level of six types of tumor-infiltrating immune cells (macrophages, neutrophils, B cells, CD4+ T cells, CD8+ T cells and dendritic cells) were precalculated for 10 897 tumors in the TCGA database, which allows users to infer the abundance of tumor-infiltrating immune cells form specific gene expression profiles. Here, we calculated the correlation between LIMP2 expression and the abundance with the six types of tumor-infiltrating immune cells listed above. Further, we applied GEPIA to reconfirm this correlation[23]. The Pearson correlation coefficient was determined to assess the correlated LIMP2 levels with a set of gene markers for tumor-infiltrating immune cells.
Data are expressed as the mean±SD from at least three independent experiments and were analyzed using a two-tailed Student t test for binary measures or ANOVA for more than two measures by using GraphPad 7.0 (GraphPad Software Inc., San Diego, CA, USA). Pearson's chi-square test was applied to investigate the relationships between LIMP2 levels and clinicopathologic variables using Stata 15.0 (Stata-Corp, College Station, TX, USA). The 5-year OS curves were generated and analyzed using the Kaplan-Meier method and the log-rank test. The Cox multivariate proportional hazards regression model was used to identify independent factors that influenced survival and recurrence by using SPSS 23.0 (SPSS, Chicago, USA). P < 0.05 was considered statistically significant.
By analyzing the TCGA stomach adenocarcinoma database, we found that LIMP2 mRNA was highly expressed in GC compared with gastric mucosa tis-sues (P < 0.0001, Fig. 1A) and significantly correlated with tumor associated genes such as BRAF (P < 0.0001, r=0.63), PTEN (P < 0.0001, r=0.57), APC(P < 0.0001, r=0.61), and VHL(P < 0.0001, r=0.5) (Fig. 1B). To verify the expression of LIMP2 in GC, we performed IHC analysis of a tissue array representing 260 GC tissue samples, 33 normal tissue samples and 54 benign samples. The automatic analysis by InForm software helped score tissues. The LIMP2 positive staining rate was significantly higher in GC tissue samples (59.61%, 155/260) than normal and benign samples (36.78%, 32/87) (P < 0.0001, Fig. 1C, Fig. 1D). As shown in Fig. 2, LIMP2 was more frequently expressed in GC tissues, but was negatively expressed in normal tissues. Meanwhile, the localization of LIMP2 expression was mainly in the cytoplasm of tumor cells.
The association between LIMP2 protein expression and clinicopathologic parameters of patients with GC are presented in Table 1. LIMP2 levels significantly correlated with Laurén classification (P=0.042), suggesting that downregulation of LIMP2 was more frequently associated with a worse classification. Moreover, decreased LIMP2 levels were significantly associated with advanced TNM stage (P=0.027), particularly with advanced T stage (P=0.016) and lymph node metastasis (P=0.039). There was no significant difference between LIMP2 expression and sex (P=0.562), age(P=0.185), differentiation (P=0.073), preoperative CEA levels (> 5 ng/mL, P=0.054), and distant metastasis (P=0.310). This analysis reveals that LIMP2 levels might play a certain role in GC.
Characteristics | n | LIMP2 low or no expression[n(%)] | LIMP2 high expression[n(%)] | Pearson χ2 | P |
Sex | 0.3367 | 0.562 | |||
Male | 183 | 76(41.53) | 107(58.47) | ||
Female | 077 | 29(37.66) | 048(62.34) | ||
Age | 1.7569 | 0.185 | |||
< 60 | 094 | 43(45.74) | 051(54.26) | ||
≥60 | 166 | 62(37.35) | 104(62.65) | ||
Laurén classification | 6.3183 | 0.042 | |||
Intestinal type | 080 | 26(32.50) | 054(67.50) | ||
Diffuse type | 160 | 74(46.25) | 086(53.75) | ||
Mixed type | 020 | 05(25.00) | 015(75.00) | ||
Differentiation | 3.2253 | 0.073 | |||
Poor | 151 | 68(45.03) | 083(54.97) | ||
Well-middle | 082 | 27(32.93) | 055(67.07) | ||
othersa | 027 | 10(37.04) | 017(62.96) | ||
pT status | 8.2975 | 0.016 | |||
T1 and T2 | 077 | 25(32.47) | 052(67.53) | ||
T3 | 084 | 29(34.52) | 055(65.48) | ||
T4a and T4b | 099 | 51(51.52) | 048(48.48) | ||
pN status | 4.2601 | 0.039 | |||
N0 | 104 | 34(32.69) | 070(67.31) | ||
N1, N2 and N3 | 156 | 71(45.51) | 085(54.49) | ||
pM status | 1.0300 | 0.310 | |||
M0 | 247 | 98(39.68) | 149(60.32) | ||
M1a and M1b | 013 | 07(53.85) | 006(46.15) | ||
TNM stage | 9.2203 | 0.027 | |||
0 and Ⅰa and Ⅰb | 055 | 18(32.73) | 037(67.27) | ||
Ⅱa and Ⅱb | 076 | 23(30.26) | 053(69.74) | ||
Ⅲa and Ⅲb and Ⅲc | 110 | 54(49.09) | 056(50.91) | ||
IV | 019 | 10(52.63) | 009(47.37) | ||
Preoperative CEA | 3.7056 | 0.054 | |||
≤5 ng/mL | 159 | 59(37.11) | 100(62.89) | ||
> 5 ng/mL | 041 | 22(53.66) | 019(46.34) | ||
Unknown | 060 | 24(40.00) | 036(60.00) | ||
Othersa : mucinous adenocarcinoma, signet ring cell carcinoma and undifferentiated carcinoma. |
Among the 260 patients, 124 (47.69%) were alive and 136 (52.31%) had passed away at the last follow-up (median observation period, 58 months; range, 2 months to 113 months). Univariate analyses revealed that differentiation (HR: 0.594, P < 0.001), T stage (HR: 1.623, P < 0.001), lymph node metastasis (HR: 1.904, P < 0.001), distant metastasis (HR: 2.470, P=0.006), preoperative CEA (HR: 1.707, P=0.018), and LIMP2 level (HR: 0.508, P < 0.001) were significantly related to 5-year OS (Table 2). Multivariate analysis of survival revealed that differentiation (HR: 0.603, P=0.001), TNM stage (HR: 2.326, P < 0.001), and LIMP2 level (HR: 0.559, P=0.001) were independent predictors of OS (Table 2). Moreover, patients with high-levels of LIMP2 had better clinical outcomes compared with those with low-levels (Fig. 3).
Factors | Univariate analysis | Multivariate analysis (adjusted for age) | |||||||
HR | P > |z| | 95% CI | HR | P > |z| | 95% CI | ||||
LIMP2 expression | |||||||||
High vs. low and none | 0.508 | < 0.001 | 0.361 | 0.714 | 0.559 | < 0.001 | 0.395 | 0.790 | |
Age (years) | |||||||||
≤60 vs. > 60 | 0.945 | 0.755 | 0.664 | 1.346 | |||||
Sex | |||||||||
Male vs. female | 0.781 | 0.179 | 0.544 | 1.120 | |||||
Laurén classification | |||||||||
Intestinal type vs. Diffuse type vs. mixed type | 1.052 | 0.726 | 0.791 | 1.401 | |||||
Differentiation | |||||||||
Well and middle vs. poor | 0.594 | < 0.001 | 0.448 | 0.788 | 0.603 | < 0.001 | 0.451 | 0.806 | |
TNM stage | |||||||||
0 and Ⅰa and Ⅰb vs.Ⅱa and Ⅱb vs.Ⅲa and Ⅲb and Ⅲc vs. IV | 2.350 | < 0.001 | 1.912 | 2.887 | 2.326 | < 0.001 | 1.883 | 2.874 | |
T | |||||||||
1 vs. 2 vs. 3 vs. 4a and 4b | 1.623 | < 0.001 | 1.342 | 1.963 | |||||
N | |||||||||
0 vs. 1 vs. 2 vs. 3a and 3b | 1.904 | < 0.001 | 1.640 | 2.210 | |||||
M | |||||||||
M0 vs. M1a and M1b | 2.470 | 0.006 | 1.289 | 4.736 | |||||
Preoperative CEA | |||||||||
≤5 ng/mL vs. > 5 ng/mL | 1.707 | 0.018 | 1.097 | 2.655 |
We used GSEA to analyze expression data acquired from the TCGA. Eight pathways were significantly associated with the LIMP2high group, and five of eight pathways were associated with the phenotypes of cancer cells. The TGF-β signaling pathway, the ERBB signaling pathway, and Toll-like receptor signaling were significantly enriched in the LIMP2high group (Fig. 4).
LIMP2 participates in innate immunity by inducing the production of proinflammatory cytokines and tumor-infiltrating lymphocyte cells, which are important factors that affect survival[25-26]. TIMER analysis revealed that SCARB2 levels were significantly associated with macrophages (P < 0.0001, r=0.345) (Fig. 5A). Spearman analysis confirmed the significance of the correlation between SCARB2 levels and immune cells markers determined by GEPIA. CD80 and CD86 are hallmarks of M1 TAMs, and CD163 and CD200R1 are markers of M2 TAMs[27-28]. We found that LIMP2 levels were significantly associated with CD80 (P < 0.0001, r=0.37) and CD86 (P < 0.0001, r=0.45), but not with CD163 (P=0.067, r=0.074) and CD200R1 (P=0.00025, r=0.16) (Fig. 5B). Next, we performed immunohistochemistry to evaluate CD86+ macrophages and LIMP2+ cells in five patients with gastric cancer. It turned out that elevated CD86+ macrophages correlated with abundant LIMP2 expression (Fig. 6). These results indicate that LIMP2 is significantly related to M1 TAMs.
In the present study, we showed that relatively high levels of LIMP2 were significantly associated with a longer OS in patients with GC, which demonstrates that the expression level of LIMP2 is an independent prognostic factor. This result is the first data, to our knowledge, indicating that the level of LIMP2 expression is significantly associated with the prognosis of patients with GC who undergo curative surgery. Furthermore, our IHC analyses revealed that LIMP2 levels were higher in GC tissues compared with those detected in noncancerous tissues. Moreover, LIMP2 levels were significantly associated with Laurén classification, depth of invasion, lymph node metastasis, and TNM stage, but not with distant metastasis, the preoperative CEA level, age, sex, or differentiation. Therefore, our data identified that LIMP2 levels were negatively associated with the progression of gastric adenocarcinoma.
Our univariate analysis of survival indicated that LIMP2 expression, tumor differentiation, depth of invasion, lymph node metastasis, distant metastasis, TNM stage, and preoperative CEA level were significantly associated with the 5-year OS of patients with GC. Moreover, multivariate COX regression analysis showed that LIMP2 expression, tumor differentiation, and TNM stage can serve as prognostic markers for 5-year OS and that LIMP2 was an independent prog-nostic factor of OS.
We gained new information on the biological functions of LIMP2 using TIMER analysis. Here, we found that SCARB2 levels were significantly associated with macrophages' infiltration of tumors, indicating that SCARB2 induced macrophage-mediated innate immunity. GSEA revealed that the ERBB signaling pathway had a high enrichment score. In a previous study, LIMP2 was recognized to induce immune cells to produce cytokines in response to bacterial infections and regulate macrophage activation in response to infection with Helicobacter pylori through EGFR signaling[25, 29]. Moreover, EGFR influences the polarization of macrophages in the tumor microenvironment. Specifically, upregulation of EGFR expression inhibits the expression of the M2-related marker[30]. SCARB2 expression was also discovered significantly associated with M1 but not M2 related hallmark. Therefore, we assumed that upregulated LIMP2 expression facilitated the infiltration of macrophages with M1-related markers while inhibiting EGFR activity in GC. Thus, M1 TAMs in the tumor environment produced proinflammatory cytokines that enhanced the antitumor response to prolong survival.
However, our study had certain limitations. Firstly, the effects of LIMP2 on cancer pathogenesis, angiogenesis, and metastasis were not investigated, and animal models of the effect of LIMP2 on tumor progression has not been established. Secondly, the association between LIMP2 and macrophages requires experimental confirmation. Thirdly, our data indicating the prognostic value of LIMP2 must be validated before translation to the clinic, particularly in view of the heterogeneity of GC and associated mechanisms.
In conclusion, the present study showed for the first time that LIMP2 was highly expressed in GC tissues, and that patients with high LIMP2 levels experienced longer overall survival. Moreover, LIMP2 can serve as an independent prognostic factor for OS of GC.
This work was supported by the grants of National Natural Science Foundation of China (81572390), the 13th Five Years Talent's Subsidy Project in Science and Education of Jiangsu Province (ZDRCA2016051).
Conflict of interests: All authors have no conflict of interests.
[1] |
Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA Cancer J Clin, 2018, 68(6): 394-424. doi: 10.3322/caac.21492
|
[2] |
Van Cutsem E, Sagaert X, Topal B. Haustermans kand prenen H gastric cancer[J]. Lancet, 2016, 388: 2654-2664. doi: 10.1016/S0140-6736(16)30354-3
|
[3] |
Allemani C, Weir HK, Carreira H, et al. Global surveil-lance of cancer survival 1995-2009: analysis of individual data for 25 676 887 patients from 279 population-based registries in 67 countries (CONCORD-2)[J]. Lancet, 2015, 385(9972): 977-1010. doi: 10.1016/S0140-6736(14)62038-9
|
[4] |
Saftig P, Klumperman J. Lysosome biogenesis and lysosomal membrane proteins: trafficking meets function[J]. Nat Rev Mol Cell Biol, 2009, 10(9): 623-635. http://www.nature.com/nrm/journal/v10/n9/fig_tab/nrm2745_F1.html
|
[5] |
Fujita H, Ezaki J, Noguchi Y, et al. Isolation and sequencing of a cDNA clone encoding 85kDa sialogly-coprotein in rat liver lysosomal membranes[J]. Biochem Biophys Res Commun, 1991, 178(2): 444-452. doi: 10.1016/0006-291X(91)90127-S
|
[6] |
Endemann G, Stanton LW, Madden KS, et al. CD36 is a receptor for oxidized low density lipoprotein[J]. J Biol Chem, 1993, 268(16): 11811-11816. doi: 10.1016/S0021-9258(19)50272-1
|
[7] |
Acton SL, Scherer PE, Lodish HF, et al. Expression cloning of SR-BI, a CD36-related class B scavenger receptor[J]. J Biol Chem, 1994, 269(33): 21003-21009. doi: 10.1016/S0021-9258(17)31921-X
|
[8] |
Reczek D, Schwake M, Schröder J, et al. LIMP-2 is a receptor for lysosomal mannose-6-phosphate-independent targeting of beta-glucocerebrosidase[J]. Cell, 2007, 131(4): 770-783. doi: 10.1016/j.cell.2007.10.018
|
[9] |
Kampine JP, Brady RO, Kanfer JN, et al. Diagnosis of gaucher's disease and niemann-pick disease with small samples of venous blood[J]. Science, 1967, 155(3758): 86-88. doi: 10.1126/science.155.3758.86
|
[10] |
Tayebi N, Callahan M, Madike V, et al. Gaucher disease and parkinsonism: A phenotypic and genotypic characterization[J]. Mol Genet Metab, 2001, 73(4): 313-321. doi: 10.1006/mgme.2001.3201
|
[11] |
Aharon-Peretz J, Rosenbaum H Gershoni-Baruch R. Mutations in the glucocerebrosidase gene and Parkinson's disease in Ashkenazi Jews[J]. N Engl J Med, 2004, 351(19): 1972-7. doi: 10.1056/NEJMoa033277
|
[12] |
Yamayoshi S, Yamashita Y, Li J, et al. Scavenger receptor B2 is a cellular receptor for enterovirus 71[J]. Nat Med, 2009, 15(7): 798-801. doi: 10.1038/nm.1992
|
[13] |
Yamayoshi S, Iizuka S, Yamashita T, et al. Human SCARB2-dependent infection by coxsackievirus A7, A14, and A16 and enterovirus 71[J]. J Virol, 2012, 86(10): 5686-5696. doi: 10.1128/JVI.00020-12
|
[14] |
Kuronita T, Eskelinen EL, Fujita H, et al. A role for the lysosomal membrane protein LGP85 in the biogenesis and maintenance of endosomal and lysosomal morphology[J]. J Cell Sci, 2002, 115(Pt 21): 4117-4131. http://www.ncbi.nlm.nih.gov/pubmed/12356916
|
[15] |
Conrad KS, Cheng TW, Ysselstein D, et al. Lysosomal integral membrane protein-2 as a phospholipid receptor revealed by biophysical and cellular studies[J]. Nat Commun, 2017, 8(1): 1908. doi: 10.1038/s41467-017-02044-8
|
[16] |
Berkovic SF, Dibbens LM, Oshlack A, et al. Array-based gene discovery with three unrelated subjects shows SCARB2/LIMP-2 deficiency causes myoclonus epilepsy and glomerulosclerosis[J]. Am J Hum Genet, 2008, 82(3): 673-684. doi: 10.1016/j.ajhg.2007.12.019
|
[17] |
Gamp AC, Tanaka Y, Lüllmann-Rauch R, et al. LIMP-2/LGP85 deficiency causes ureteric pelvic junction obstruction, deafness and peripheral neuropathy in mice[J]. Hum Mol Genet, 2003, 12(6): 631-646. doi: 10.1093/hmg/ddg062
|
[18] |
Johnson IR, Parkinson-Lawrence EJ, Shandala T, et al. Altered endosome biogenesis in prostate cancer has biomarker potential[J]. Mol Cancer Res, 2014, 12(12): 1851-1862. doi: 10.1158/1541-7786.MCR-14-0074
|
[19] |
Pasini FS, Maistro S, Snitcovsky I, et al. Four-gene expression model predictive of lymph node metastases in oral squamous cell carcinoma[J]. Acta Oncol, 2012, 51(1): 77-85. doi: 10.3109/0284186X.2011.620619
|
[20] |
Chen P, Liu C, Li P, et al. High RhCG expression predicts poor survival and promotes migration and proliferation of gastric cancer via keeping intracellular alkaline[J]. Exp Cell Res, 2020, 386(2): 111740. doi: 10.1016/j.yexcr.2019.111740
|
[21] |
Sun MJ, Qi Y, Jin Q, et al. Increased SLC4A11 expression is associated with poor prognosis of gastric cancer[J]. Blood and Genomics, 2020, 4(1): 61-70. doi: 10.46701/BG.2020012019119
|
[22] |
Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles[J]. Proc Natl Acad Sci U S A, 2005, 102(43): 15545-15550. doi: 10.1073/pnas.0506580102
|
[23] |
Li T, Fan J, Wang BB, et al. TIMER: A web server for comprehensive analysis of tumor-infiltrating immune cells[J]. Cancer Res, 2017, 77(21): e108-e110. doi: 10.1158/0008-5472.CAN-17-0307
|
[24] |
Tang Z, Li C, Kang B, et al. GEPIA: a web server for cancer and normal gene expression profiling and interactive analyses[J]. Nucleic Acids Res, 2017, 45(W1): W98-W102. doi: 10.1093/nar/gkx247
|
[25] |
Carrasco-Marín E, Fernández-Prieto L, Rodriguez-Del RE, et al. LIMP-2 links late phagosomal trafficking with the onset of the innate immune response to Listeria monocytogenes: a role in macrophage activation[J]. J Biol Chem, 2011, 286(5): 3332-3341. doi: 10.1074/jbc.M110.146761
|
[26] |
Savas P, Salgado R, Denkert C, et al. Clinical relevance of host immunity in breast cancer: from TILs to the clinic[J]. Nat Rev Clin Oncol, 2016, 13(4): 228-241. doi: 10.1038/nrclinonc.2015.215
|
[27] |
Kalkman HO, Feuerbach D. Antidepressant therapies inhibit inflammation and microglial M1-polarization[J]. Pharmacol Ther, 2016, 163: 82-93. doi: 10.1016/j.pharmthera.2016.04.001
|
[28] |
Biswas SK, Mantovani A. Macrophage plasticity and interaction with lymphocyte subsets: cancer as a paradigm[J]. Nat Immunol, 2010, 11(10): 889-896. doi: 10.1038/ni.1937
|
[29] |
Hardbower DM, Singh K, Asim M, et al. EGFR regu-lates macrophage activation and function in bacterial infection[J]. J Clin Invest, 2016, 126(9): 3296-3312. doi: 10.1172/JCI83585
|
[30] |
Zhao G, Liu L, Peek RJ, et al. Activation of epidermal growth factor receptor in macrophages mediates feedback inhibition of M2 polarization and gastrointestinal tumor cell growth[J]. J Biol Chem, 2016, 291(39): 20462-20472. doi: 10.1074/jbc.M116.750182
|
Characteristics | n | LIMP2 low or no expression[n(%)] | LIMP2 high expression[n(%)] | Pearson χ2 | P |
Sex | 0.3367 | 0.562 | |||
Male | 183 | 76(41.53) | 107(58.47) | ||
Female | 077 | 29(37.66) | 048(62.34) | ||
Age | 1.7569 | 0.185 | |||
< 60 | 094 | 43(45.74) | 051(54.26) | ||
≥60 | 166 | 62(37.35) | 104(62.65) | ||
Laurén classification | 6.3183 | 0.042 | |||
Intestinal type | 080 | 26(32.50) | 054(67.50) | ||
Diffuse type | 160 | 74(46.25) | 086(53.75) | ||
Mixed type | 020 | 05(25.00) | 015(75.00) | ||
Differentiation | 3.2253 | 0.073 | |||
Poor | 151 | 68(45.03) | 083(54.97) | ||
Well-middle | 082 | 27(32.93) | 055(67.07) | ||
othersa | 027 | 10(37.04) | 017(62.96) | ||
pT status | 8.2975 | 0.016 | |||
T1 and T2 | 077 | 25(32.47) | 052(67.53) | ||
T3 | 084 | 29(34.52) | 055(65.48) | ||
T4a and T4b | 099 | 51(51.52) | 048(48.48) | ||
pN status | 4.2601 | 0.039 | |||
N0 | 104 | 34(32.69) | 070(67.31) | ||
N1, N2 and N3 | 156 | 71(45.51) | 085(54.49) | ||
pM status | 1.0300 | 0.310 | |||
M0 | 247 | 98(39.68) | 149(60.32) | ||
M1a and M1b | 013 | 07(53.85) | 006(46.15) | ||
TNM stage | 9.2203 | 0.027 | |||
0 and Ⅰa and Ⅰb | 055 | 18(32.73) | 037(67.27) | ||
Ⅱa and Ⅱb | 076 | 23(30.26) | 053(69.74) | ||
Ⅲa and Ⅲb and Ⅲc | 110 | 54(49.09) | 056(50.91) | ||
IV | 019 | 10(52.63) | 009(47.37) | ||
Preoperative CEA | 3.7056 | 0.054 | |||
≤5 ng/mL | 159 | 59(37.11) | 100(62.89) | ||
> 5 ng/mL | 041 | 22(53.66) | 019(46.34) | ||
Unknown | 060 | 24(40.00) | 036(60.00) | ||
Othersa : mucinous adenocarcinoma, signet ring cell carcinoma and undifferentiated carcinoma. |
Factors | Univariate analysis | Multivariate analysis (adjusted for age) | |||||||
HR | P > |z| | 95% CI | HR | P > |z| | 95% CI | ||||
LIMP2 expression | |||||||||
High vs. low and none | 0.508 | < 0.001 | 0.361 | 0.714 | 0.559 | < 0.001 | 0.395 | 0.790 | |
Age (years) | |||||||||
≤60 vs. > 60 | 0.945 | 0.755 | 0.664 | 1.346 | |||||
Sex | |||||||||
Male vs. female | 0.781 | 0.179 | 0.544 | 1.120 | |||||
Laurén classification | |||||||||
Intestinal type vs. Diffuse type vs. mixed type | 1.052 | 0.726 | 0.791 | 1.401 | |||||
Differentiation | |||||||||
Well and middle vs. poor | 0.594 | < 0.001 | 0.448 | 0.788 | 0.603 | < 0.001 | 0.451 | 0.806 | |
TNM stage | |||||||||
0 and Ⅰa and Ⅰb vs.Ⅱa and Ⅱb vs.Ⅲa and Ⅲb and Ⅲc vs. IV | 2.350 | < 0.001 | 1.912 | 2.887 | 2.326 | < 0.001 | 1.883 | 2.874 | |
T | |||||||||
1 vs. 2 vs. 3 vs. 4a and 4b | 1.623 | < 0.001 | 1.342 | 1.963 | |||||
N | |||||||||
0 vs. 1 vs. 2 vs. 3a and 3b | 1.904 | < 0.001 | 1.640 | 2.210 | |||||
M | |||||||||
M0 vs. M1a and M1b | 2.470 | 0.006 | 1.289 | 4.736 | |||||
Preoperative CEA | |||||||||
≤5 ng/mL vs. > 5 ng/mL | 1.707 | 0.018 | 1.097 | 2.655 |