The relationship between LAPTM4B polymorphisms and cancer risk in Chinese Han population: a meta-analysis

LAPTM4B is a newly cloned gene that shows an active role in many solid tumors progression in substantial researches, mainly through the autophage function. Accumulated studies have been conducted to determine the association of LAPTM4B polymorphism with cancer risk. While the results are inconsistent, we conducted the meta-analysis to determine the strength of the relationship. Results showed that allele*2 carriers exhibited a significantly increased risk of cancer development with comparison to allele*1 homozygote (for *1/2, OR = 1.55, 95% CI 1.367-1.758; for *2/2, OR = 2.093, 95%CI 1.666-2.629; for *1/2 + *2/2, OR = 1.806, 95%CI 1.527-2.137). We also observed a significant association between *2/2 homozygote and cancer risk with comparison to allele*1 containing genotypes (OR = 1.714, 95%CI 1.408-2.088). Allele*2 is a risk factor for cancer risk (OR = 1.487, 95%CI 1.339-1.651). Stratified analysis by tumor type exhibits the significant association of this genetic variants with various cancers. In conclusion, LAPTM4B polymorphism is associated with cancer risk and allele*2 is a risk factor. Electronic supplementary material The online version of this article (doi:10.1186/s40064-015-0941-7) contains supplementary material, which is available to authorized users.


Background
Lysosomes are organelles that contain hydrolytic enzymes such as proteases, nucleases and lipases. The stability of lysosomal membrane and the normality of lysosomal function are essential in cellular biology. Lysosomal-membrane permeabilization (LMP) can pose a threat to cellular homeostasis through release of lysosomal contents, such as lysosomal cathepsins B or D, and is a recognized trigger of cell death (Johansson et al. 2010). Autophagy is a conserved lysosome-mediated intracellular trafficking pathway that degrades and recycles intracellular components, thereby promoting cell survival as well as the adaption to a variety of cell stresses as cancers develop and disseminate (Burman and Ktistakis 2010;Jin and White 2007;Degenhardt et al. 2006;Kenific et al. 2010). Noticeably, too much autophagy may conversely result in autophagic cell death (Scarlatti et al. 2009;Eisenberg-Lerner et al. 2009). Lysosomal retention of drugs is responsible for drug resistance and lysosomal concentration of anthracyclines is considered to increase drug efflux and decrease drug nuclear localization, thereby preventing effective chemotherapyinduced DNA damage (Larsen et al. 2000).
Lysosome-associated protein transmembrane 4 beta (LAPTM4B) is a recently discovered gene mapped to chromosome 8q22.1 with 7 exons and 6 introns spanning at least 50 kb . LAPTM4B protein has a lysosome localization motif and co-localizes with markers of late endosomes and lysosomes Li et al. 2011;Vergarajauregui et al. 2010). LAPTM4B is required for the later stages of autophagy maturation in which autophagosomes are fused with lysosomes to form autolysosomes, the place where cytosol and damaged organelles are discomposed into micro molecules and new macro molecules and ATPs are synthetized, thereby maintaining the cellular homeostasis (Li et al. 2011). The transcriptional products of the gene have been observed over-expressed in many solid tumors (Kasper et al. 2005) and different expression levels exist in tumors of different status , exerting various functions. Up-expression of LAPTM4B inhibits lysosome-mediated death pathways, promotes autophagy and leads to stress tolerance, thereby enhancing tumor cell growth, survival and resistance to apoptosis. Down regulation of LAPTM4B triggers lysosome membrane permeabilization (LMP), weekening the capability of lysosomes to retain the drug and provoking lysosomal-mediated programmed cell death (Li et al. 2011). Evidences from Maeda, et al. study suggest that the transcriptional product of this gene results in an active role in the progression of cell proliferation other than results from tumor genesis (Maeda et al. 2005) and LAPTM4B proteins play an influential role in occurrence, development, progression and prognosis of cancer cells Yang et al. 2010b;Li et al. 2010b;He et al. 2003;Yang et al. 2010a).
Allelic loss of essential autophage gene can lead to tumor formation in beclin1+/− mutant mice (Qu et al. 2003;Yue et al. 2003;Liang et al. 1999), indicating the role for polymorphisms of autophage-associated gene in tumor suppression. LAPTM4B has two alleles, LAPTM4B allele*1 and LAPTM4B allele*2, encoding proteins with 35kD and 40kD respectively . Allele*1 contains one copy of a 19-bp sequences at the 5'UTR of the first exon, while the segment is duplicated and tandemly arranged in allele*2. This polymorphism may alter its opening reading frame (ORF) and perform as a cis-acting elememt and participate in genetic translation or tissue specific expression with some transcriptional factors or nucleoproteins Yang et al. 2010b). The relationship between the genetic polymorphisms and cancer susceptibility has been studied while the conclusions are not accordant (Deng et al. 2005;Li et al. 2006;Liu et al. 2007;Sun et al. 2007;Cheng et al. 2008;Sun et al. 2008;Wang and Zhang 2010;Qi 2010;Meng et al. 2011;Fan et al. 2012;Li et al. 2012;Yang et al. 2012;Wang et al. 2012;Zhai et al. 2012;Xu et al. 2012;Meng et al. 2013;Wang et al. 2013;Tang et al. 2014). To evaluate the association of LAPTM4B polymorphisms and the risk of cancer development, we conducted a meta-analysis of 18 enrolled case-control studies including gynecological tumors (Meng et al. 2011;Meng et al. 2013;Xu et al. 2012), digestive system cancers (Wang and Zhang 2010;Zhai et al. 2012;Yang et al. 2012;Qi 2010;Cheng et al. 2008;Liu et al. 2007;Wang et al. 2012;Sun et al. 2008), lung cancer (Li et al. 2006;Deng et al. 2005;Tang et al. 2014), breast cancer (Fan et al. 2012;Li et al. 2012), nasopharyngeal carcinoma (Wang et al. 2013), and lymphoma (Sun et al. 2007).

Characteristics of the studies
The flowchart of the selection of eligible studies is shown as Figure 1. The major information of the publications is listed in Table 1.Totally, 18 eligible case-control studies published from 2005 to 2014 were recruited in this metaanalysis. The countries of origin and ethnic group of these studies are not shown in the table for the study populations are all Chinese. All of the enrolled studies comply with Hardy-Weinberg Law. Table 2 shows the main results of pooled OR and 95% CI for LAPTM4B polymorphism and cancer risk. Generally speaking, allele*2 carriers showed a significant association with cancer risk. When adjusted by age, gender, et al., the association between LAPTM4B polymorphism and cancer susceptibility is also observed (Additional file 1: Table S1). The forest plots of the pooled OR and 95% CI for the five genetic models are shown as Figure 2. Another meta-analysis for the estimation of association between LAPTM4B expression and cancer prognosis was conducted. Results are shown in Additional file 2: Table S3. Overexpression of LATPM4B may be a risk predictor for poor survival. The articles eligible for this meta-analysis are listed in Additional file 3: Table S2.

Subgroup analysis results
The results of stratified analysis by tumor type are listed in Table 3. Stratified analysis results suggested that allele*2 may increase the risk of multiple cancers. Overexpression of LAPTM4B is also associated with poor prognosis in multiple cancers (Additional file 4: Table S4).

Sensitivity analysis
A sensitivity analysis was performed by omitting every study in turn to find potential outliers to check the inclusion criteria. Pooled estimates for all genetic models were insensitive to the removal of individual study, and the corresponding pooled ORs were not substantially altered, indicating that our results were stable and reliable (data not shown).

Publication bias
Begger's plot funnel and Egger's test were performed to examine the underlying publication bias. The symmetrical  plot funnel and p value >0.05 for Egger's test indicated no publication bias in this meta-analysis ( Figure 3).

Discussion
Due to insufficient angiogenesis, tumors frequently experience elevated metabolic stress from nutrient and oxygen deprivation (Jin and White 2007;Folkman 2003). Survival mechanisms such as autophagy may be more reliable for cancer cells to maintain growth or to disseminate (White and DiPaola 2009). Many studies have shown the roles of LAPTM4B in limiting cell death and promoting autophagy (Li et al. 2011;He et al. 2003;Yang et al. 2010a).
To date, the exact molecular mechanism that might be involved in progressive phenotypes caused by LAPTM4B protein remains unclear. Nonetheless, accumulated detections in other studies may demonstrate the functional mechanism to some extent.
Both the C-and N-terminal of LAPTM4B protein contain a proline-rich domain PXXP, forming a specific src-homology 2 (SH2) binding domain at C-terminal and specific src-homology 3(SH3) binding domain at N-terminal. The functional domain in C-terminal has a lysosome localization motif, sharing the same domain with the markers of endosomes and lysosomes (Liu et al. 2004;Hunziker and Geuze 1996;Honing et al. 1998;Dell'Angelica et al. 2000;Hogue et al. 2002). And the N-terminal offers sites for SH3-containing proteins such as phosphoinositide 3-kinase (PI3K), protein phosphatase 2A and protein kinase C (PKC) Liu et al. 2004). We hypothesized that Cterminal is responsible for positioning and N-terminal is responsible for the function. LAPTM4B plays a significant role in cell cycle progression through regulation of c-Myc, cyclinD1/E and p27 (He et al. 2003;Yang et al. 2010a;Liu et al. 2009). Up-expression of LAPTM4B proteins promoted the phosphorylation of AKT, the pivotal signaling pathway in cell survival and proliferation (Cantley 2002;Yang et al. 2010a;Medema et al. 2000). Moreover, LAPTM4B can decrease the expression of p16, increase the expression of c-Fos, c-Jun Liu et al. 2009;Zhou et al. 2010b), and phosphorylate Rb (Retinoblastoma protein) . Another study (Li et al. 2010a) suggested that LAPTM4B may promote drug efflux and multidrug resistance through PI3K/AKT pathway and thus results in poor prognosis.
A high level of expression of LAPTM4B protein has been observed in cells with tendency of migration and metastasis, while the knockdown of the genetic expression reverses the propensity of growth, proliferation and malignancy (Li et al. 2010b;He et al. 2003;Liu et al. 2009). In ) study, we observed that the over-expression of LAPTM4B enhanced the expression of MMP-2, MMP-9 and uPA (Milde-Langosch 2005). Among them, MMP-9 and uPA are targets of activator protein (AP-1) (Milde-Langosch 2005). Given the fact that no binding site of AP-1 was found in promoter region of MMP-2 gene (Yan and Boyd 2007) and that uPA can activate MMP-2 (Mazzieri et al. 1997), we inferred that LAPTM4B might activate MMP-2 by up-regulating expression of uPA.
The allele LAPTM4B*1differs from allele LAPTM4B*2 in that it contains a particular single 19-bp sequence, whereas LAPTM4B *2 contains two copies of these sequences in the 5'UTR of exon 1. This results in the change in the N terminal of the proteins. Li (Li et al.

2012) and
Wang (Wang et al. 2012) also observed that the allele LAPTM4B*2 is associated with higher level of LAPTM4B expression.
With the detections mentioned above, we hypothesized that the effect of the genetic polymorphism in LAPTM4B on cancer susceptibility may result from the change in N-terminal region or from the change in expression level with different genotypes.
While the results of the studies concerning the relationship between LAPTM4B polymorphism and cancer risk are inconsistent, we conducted this meta-analysis to determine the strength of the association with a relatively large sample size containing 4556 cases and 4584 controls. In our meta-analysis, the association between LATPM4B polymorphism and cancer susceptibility and the association between LAPTM4B expression levels and cancer prognosis were observed in multiple cancers. All above support our hypotheses of the association between LAPTM4B polymorphism and cancer susceptibility.
This analysis has some limitations. One of the limitations lies in the study population. All the populations are Chinese. This may because that the gene is newly discovered and the polymorphisms were initially detected in Chinese Han population. And that our analysis indicated a strong relationship between this polymorphism and cancer risk, this meta-analysis may be a reference for further researches in other countries. Second, the lack of raw data, including the information of histology, family history and smoking status, limited the assessment of interaction between gene-gene and gene-environment.
Asides the limitations mentioned above, this study presents some advantages. No publication bias and heterogeneity among studies were observed in this analysis, suggesting a stable and reliable result. Additionally, the estimation of pooled ORs and 95%CIs was based on a relatively large scale of cases and controls. In this view, the results indicated a powerful association between the genetic polymorphisms and cancer risk and may provide a strong evidence for further researches.

Conclusions
In summary, LAPTAM4B polymorphisms are associated with multiple cancer susceptibility and allele*2 is a risk factor for various tumors. More studies on the association of LAPTM4B polymorphisms and cancer risk in other countries are warranted to confirm the association.

Identification and eligibility of relevant study
All of the studies were identified through five online electronic databases including PubMed, web of science, google scholar, CNKI and WanFang database, using the search strategies "keywords or subjects or title or abstract = (LAPTM4B AND polymorphism AND (cancer OR carcinoma OR tumor OR sarcoma))". Citation indexes were performed and the reference list was searched to identify potential relevant studies. The last search update was Jun 29, 2014.
Studies involved in the meta-analysis should meet the following inclusion criteria: (i) a case-control study; (ii) frequency of genotype or allele is available; (iii) P value of Hardy-Weinberg equilibrium for the distribution of genotypes in control group should be no less than 0.05.