Frequent loss expression of dab2 and promotor hypermethylation in human cancers: a meta-analysis and systematic review.

Objective : Disabled-2 (Dab2) is an important endocytic adaptor which plays an inhibition role in cancer cell growth. The objective of this study was to systematically review expressions of Dab2 in human cancers. Methods : Eligible studies about Dab2 in human cancers were retrieved from databases of PubMed, Embase, Web of Science. Odds Ratios (ORs) with 95% confidence intervals (CIs) were calculated using Review Manager 5.0 software and statistical analyses were performed by the SPSS 13.0 software. Results : Fourteen case-control studies with a total of 689 human tumor tissues, 332 control tissues and 32 cancer cell lines were included in the meta-analysis study. The results indicated loss expressions of Dab2 were observed in 74.9% and 46.9% in human malignant cancer tissues and cancer cell lines, respectively. The ratio of Dab2 promotor hypermethylation is 34.54% in cancer tissues which Dab2 expression are lost, but none in the control tissues or cells by Methylation-specific PCR (MSP). Conclusions : The expressions of Dab2 are frequently lost in human malignant cancer tissues, and promotor hypermethylation of Dab2 are common in human malignant cancer tissues, which is an important factor for the loss expression of Dab2 in human cancers tissues.


INTRODUCTION
The gene Disabled-2 (Dab2) which also names DOC-2 (Differentially expressed in the Ovarian Carcinoma 2, DOC-2) owns two different splicing formats, and encodes two isoforms (p96-Dab2 and p67-Dab2). 1,2 The main functional domain is the phosphotyrosine binding domain (PTB) of the N-terminal, which is a highly conserved sequence and plays a variety of functional roles in endocytosis, cell mitosis, and growth factor signaling. 3,4 Especially, the p96-Dab2 is essential for the development of visceral endoderm during mouse embryogenesis and homologous with 93% full-length of mouse Dab2. [5][6][7] Mechanistically, Dab2 is shown to bind with the growth factor receptor binding protein 2 (Grb2), consequently uncouple the activation of c-Fos expression and Ras/mitogen activated protein kinase (MAPK). [8][9][10][11] Accumulated studies have shown that Dab2 is reduced or lost in human cancers, containing lung cancers, nasopharyngeal carcinomas, the breast cancers, and colorectal cancers, et al. [12][13][14][15][16][17][18][19] Thus, it is gradually considered as a tumor suppressor gene. However, there are no complete credible studies to explain the concrete mechanisms, except a few of epigenetic studies about promoters or exons of Dab2. 16,17,19 In the current study, we firstly conducted a meta-analysis of cohort studies to evaluate Dab2 expression level and its epigenetic variations in various human malignant cancers or cell lines. Furthermore, systematically investigated the concrete reasons for Dab2 expression loss, and its correlation with human oncogenesis.

Literature search strategy:
We adapted the Cochrane Central Register of Controlled Trials, and searched relevant papers published before September 31 st 2013 in Medline, Embase, Web of Science, and Scopus with relevant text words and medical subject headings that included all spellings of "Disabled-2'' or "DOC-2'' and ''cancer'' and ''human''. In addition, we firstly performed an initial scanning of the titles or abstracts, reference lists of retrieved papers and reviews to identify other potential relevant studies. Disputes were resolved through discussions among three authors.

Criterions of Inclusion and Exclusion:
We selected the eligible studies in the present meta-analysis using the following criterions: (1) use of an cohort study or case-control study design, and focusing on the correlations between Dab2 or DOC-2 and human cancers; (2) basic researches with big size of tissue samples. Studies were excluded using the following criterions: (1) not a case-control study or cohort study; (2) not a primary document, such as a letter, meta-analysis, review, duplicate or editorial article; (3) literatures with insufficient samples or incomplete data, and the conclusions are out of date. Data extraction: Three authors recorded the following details of each included research cooperatively, containing: authors, year of publication, the country of study, journal, materials and methods, study design, age of study population, pathological type of tumors, detecting sample size, source of participants, confounders adjusted for, effect sizes and 95% Confidence Index (CI) or standard errors of effect sizes. Differences were resolved by discussion among three authors in cases of conflicting evaluations. Statistical Analysis: The present case-control study was performed by Review Manager 5.0 software, the direct count method was used to estimate the expression level of Dab2 in human cancers. The odds ratios (OR) and 95% confidence intervals (CI) were computed by means of the Fisher's exact probability test (two-tailed p-values). The data was analyzed by means of SPSS version 13.0 (SPSS Inc; Chi-cago, USA). This merged ORs and the 95% CI were obtained by means of the fixed or random effects model for each kind of human cancers. The heterogeneity was computed by Cochrans Q test, which P-value > 0.05 or I 2 > 50% indicated the existence of heterogeneity among studies. The subgroup analysis was used to explore sources of heterogeneity.

Search of Eligible Studies:
We initially retrieved 112 relevant papers in September, 2013. Finally, 15 studies were included in the present meta-analysis. Fig.1.

Frequent loss expression of Dab2 was observed in various human cancers tissues:
The immunostained percentage of Dab2 expression were detected in five different kinds of human cancer tissues. Mok SC et al found that Dab2 expression were significantly down-regulated in ovarian cancers than normal ovarian tissues (the strong positive rate was 4.5% Versus 84.1%, and weak positive rate was 92.3% versus 0.0%). 12 Analogously, Xu et al reported that Dab2 expression was significantly reduced in lung cancers than the non-cancerous tissues (the strong, moderate, and weak positive rate were 56.2% Versus 24.76%, 37.1% Versus 47.62% and 6.7% Versus 27.62%, respectively) 15 Table-II. Furthermore, Tong et al found that Dab2 was un-detectable in 72% nasopharyngeal carcinomas of Chinese people. 16 After pooling eligible data, absent expression of Dab2 were detected in 221 of 295 (74.9%) human malignant cancer tissues (OR = 0.28, 95% CI: 0.22 -0.35, P < 0.001; I 2 = 20.3%, P heterogeneity = 0.28) Fig.2A. Especially, XIE et al reported that the p96-Dab2 was expressed only in the nuclei of 31 cases (31/50, 62.0%) of normal lung tissues, and was lost in all the lung cancer tissues; which suggested that the two isoforms of Dab2 were differentially expressed in a tissue-specific manner.   Du145, et al. 14,16,17,22,[24][25][26][27] There was no significant difference between weak positive and absent expression of Dab2 in human cancer cells (OR = 1.04, 95% CI: 0.66 -1.65, P = 0.85; I 2 = 39.9%, P heterogeneity = 0.11) (Fig. 2B). Results of the Pearson χ 2 test revealed that abnormal expression of Dab2 was not significant correlated with the types of cancers from which cancer cell lines originated (χ 2 = 3.23, P = 0.36).

Reduced expression of Dab2 was correlated with the aberrant promotor hypermethylation in human cancers:
Bisulfite sequencing and methylation specific PCR (MSP) were employed to explore the correlations between promoter aberrant hypermethylation of Dab2 and expressing reduction in 4 studies. 16,17,19 Subgroup analysis was applied to discriminate the discrepancies of aberrant promoter hypermethylation of Dab2 in cancer tissues and cell lines. Results showed that ratio of Dab2 promoter hypermethylation is 34.54% in cancer tissues which Dab2 expression are lost, and Dab2 promoter hypermethylation might play a key role in the down-regulated expression of Dab2 in human cancer tissues (OR = 24.45, 95% CI: 11.00 -54.32, P < 0.001; I 2 = 45%, P heterogeneity = 0.14) (Fig.3).
Publication bias: Egger test 28 was performed to observe potential publication bias in each meta-analysis, and results showed no evidence of publication bias for each outcome: expression loss in human cancers tissues (P Egger = 0.092), expression reduced in various cancer cell lines (P Egger = 0.086) and aberrant promoter hypermethylation (P Egger = 0.061).

DISCUSSION
In this first systematic review, Dab2 expression was analyzed in approximately 789 human tumor and 432 normal tissues of 15 included papers. Some studies demonstrated that Dab2 protein was un-detectable in 70% ~ 90% human malignant cancers, including nasopharyngeal carcinomas, breast cancers, and gestational choriocarcinomas. 16,17,20 However, other studies have suggested the weak to moderate positive immunostained of Dab2 expression in lung cancers, and ESCCs, et al. 15,19,21 These diversities may be correlated with the tissue-specific differentially expression patterns of Dab2. 10,11 Similarly, our previous studies on lung cancers suggest that there are different functions between p96-Dab2 and p67-Dab2 in the process of oncogenesis. 14 Unfortunately, some weaknesses of current researches on Dab2 in cancer cell lines are identified in this meta study. Bagadi et al. reported that Dab2 was lost in all the breast cancer cell lines containing MDA-MB-231 17 ; conversely, Cheong et al. hold opinions that it was weak positive not absent expression of Dab2 in MDA-MB-231. 27 More interestingly, we and other researchers found that both lung cancer and 60% of TCC cell lines showed weak to moderate positive expression of Dab2 protein. 14,25 Thus, there is not enough evidence to determine that whether the loss or weak positive of Dab2 expressions are analogous in all the human malignant cancer cell lines (P = 0.85). Furthermore, new standards must be ruled out for the evaluation of Dab2 expression in cancer cell lines in the near future.
Our research reveals that the promoter hypermethylation of Dab2 is an important factors for the loss expression of Dab2 in human cancers tissues (OR = 24.45, P < 0.001). Although Dab2 promoter hyper-methylation have been observed in some cancer cells, there are still few reasons to attribute down-regulated expression of Dab2 to the promoter hypermethylation unless further credible evidences emerge from other cancer cells (P = 0.19).
In conclusion, frequent loss expressions of Dab2 are common in human malignant cancer tissues, and significantly correlated with the promoter hypermethylation. More studies would be conducted to enhance the expression of Dab2, and eliminate the aberrant hypermethylation of Dab2, which would offer some potential therapeutic treatment methods for human malignant cancers.