ARMCX Family Gene Expression Analysis and Potential Prognostic Biomarkers for Prediction of Clinical Outcome in Patients with Gastric Carcinoma

Hindawi BioMed Research International Volume 2020, Article ID 3575038, 16 pages https://doi.org/10.1155/2020/3575038

Research Article ARMCX Family Gene Expression Analysis and Potential Prognostic Biomarkers for Prediction of Clinical Outcome in Patients with Gastric Carcinoma

TingAn Wang,1 HuaGe Zhong,1 YuZhou Qin,1 WeiYuan Wei,1 Zhao Li,1 MingWei Huang ,1 and XiaoLing Luo 2

1Department of Gastrointestinal Surgery, Guangxi Medical University Cancer Hospital, Guangxi Clinical Research Center for Colorectal Cancer, Nanning 530021, Guangxi Zhuang Autonomous Region, China 2Guangxi Medical University Cancer Hospital, Nanning 530021, Guangxi Zhuang Autonomous Region, China

Correspondence should be addressed to MingWei Huang; [email protected] and XiaoLing Luo; [email protected]

TingAn Wang and HuaGe Zhong contributed equally to this work.

Received 12 February 2020; Accepted 20 May 2020; Published 30 June 2020

Academic Editor: Ji-Fu Wei

Copyright © 2020 TingAn Wang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Armadillo gene subfamily members (ARMCX1-6) are well-known to regulate protein-protein interaction involved in nuclear transport, cellular connection, and transcription activation. Moreover, ARMCX signals on cell pathways also implicated in carcinogenesis and tumor progression. However, little is known about the associations of the ARMCX subfamily members with gastric carcinoma. This study investigated the prognostic value of ARMCX subfamily mRNA expression levels with the prognosis of gastric carcinoma (GC). We retrieved the data of a total of 351 GC patients from TCGA database. Survival and gene set enrichment analyses were employed to explore the predictive value and underlying mechanism of ARMCX genes in GC. The multivariate survival analysis revealed that individually low expressions of ARMCX1 (adjusted P =0:006, HR = 0:620, CI = 0:440 − 0:874) and ARMCX2 (adjusted P =0:005, HR = 0:610,95%CI = 0:432 – 0:861) were related to preferable overall survival (OS). The joint-eﬀects analysis shown that combinations of low level expression of ARMCX1 and ARMCX2 were correlated with favorable OS (adjusted P=0:003, HR = 0:563, 95%CI = 0:384 – 0:825). ARMCX1 and ARMCX2 were implicated in WNT and NF-kappaB pathways, and biological processes including cell cycle, apoptosis, RNA modiﬁcation, DNA replication, and damage response. Our results suggest that mRNA expression levels of ARMCX subfamily are potential prognostic markers of GC.

1. Introduction The armadillo genes are clustered on the X chromosome, also known as X-linked (ARMCX or ALEX). In 1989, it was Gastric carcinoma, one common type of malignant tumors, is first discovered in the segment polarity gene armadillo in the fifth highest incidence and the second highest mortality Drosophila [3, 4]. Since then, more and more related proteins after lung cancer worldwide [1]. Each year, more than have been identified and classified as armadillo repeat family. 300,000 newly diagnosed cases and about 260,000 people die The common feature of these proteins is an amino acid in China. The poor prognosis is due to a high incidence of sequence (arm repeats) approximately 42 residues, identified advanced disease, high recurrence rate, high metastasis, and as 6-13 repeat units in all members of the family [5, 6], and abnormal gene expression. In addition, despite great advances each repeat domain consists of three helices, designated as in the surgery and chemotherapy technology, the death rate H1, H2, and H3 [7–9]. remains high [2]. Therefore, new strategies to improve diag- The armadillo domain protein has the functions of cell nosis and prognosis of gastric cancer are shortly needed. contact and cytoskeletal-related protein and signal 2 BioMed Research International

3 3

2 2

1 1 Log10 (TPM) Log10 (TPM) 0 0

–1 –1 Liver Lung Lung Liver Testis Testis Ovary Ovary Spleen Spleen Uterus Uterus Vagina Vagina Bladder yroid Bladder yroid Prostate Prostate Stomach Stomach Pituitary Pancreas Pituitary Pancreas Whole blood Whole Whole blood Whole Nerve - tibial Nerve Nerve - tibial - Nerve Artery tibial - Brain - cortex Brain Artery - aorta Artery - tibial Brain - cortex Brain Artery - aorta Adrenal gland Adrenal Fallopian tube Adrenal gland Adrenal Fallopian tube Fallopian Kidney - cortex Kidney - cortex Colon - sigmoid Colon Colon - sigmoid Colon Muscle - skeletal Muscle Muscle - skeletal Muscle Brain - amygdala Brain Brain - amygdala Brain Artery - coronary Artery - coronary Colon - transverse Colon Colon - transverse Colon Brain - cerebellum Brain Brain - cerebellum Brain Cervix - ectocervix Cervix - ectocervix Cervix - endocervix Cervix - endocervix Heart - le ventricle Heart Heart - le ventricle Heart Esophagus - mucosa Esophagus Esophagus - mucosa Esophagus Minor salivaryMinor gland Minor salivaryMinor gland Brain - hippocampus Brain Brain - hippocampus Brain Brain - hypothalamus Brain Brain - hypothalamus Brain Brain - substantia nigra - substantia Brain Brain - substantia nigra - substantia Brain Esophagus - muscularis Esophagus Esophagus - muscularis Esophagus Adipose - subcutaneous Adipose Adipose - subcutaneous Adipose Heart - atrial appendage Heart - atrial appendage Breast - mammaryBreast tissue Breast - mammaryBreast tissue Brain - frontal cortex (BA9) cortex - frontal Brain Brain - frontal cortex (BA9) cortex - frontal Brain Brain - cerebellar hemisphere - cerebellar Brain Brain - cerebellar hemisphere - cerebellar Brain Adipose - visceral (omentum) Adipose Adipose - visceral (omentum) Adipose Skin - sun exposed (lower leg) - sun exposedSkin (lower Cells - transformed ﬁbroblasts Cells - transformed Skin - sun exposed (lower leg) - sun exposedSkin (lower Brain - caudate (basal ganglia) - caudate Brain Cells - transformed ﬁbroblasts Cells - transformed Brain - caudate (basal ganglia) - caudate Brain Brain - putamen (basal ganglia) - putamen Brain Brain - putamen (basal ganglia) - putamen Brain Small intestine - terminal ileum - terminal intestine Small Small intestine - terminal ileum - terminal intestine Small Brain - spinal cord (cervical cord c-1) - spinal Brain Brain - spinal cord (cervical cord c-1) - spinal Brain Skin - not sun exposed - not Skin (suprapubic) Skin - not sun exposed - not Skin (suprapubic) Cells - EBV-transformed lymphocytes Cells - EBV-transformed Cells - EBV-transformed lymphocytes Cells - EBV-transformed Esophagus - gastroesophageal junction - gastroesophageal Esophagus Esophagus - gastroesophageal junction - gastroesophageal Esophagus Brain - anterior cingulate cortex (BA24) cortex cingulate - anterior Brain Brain - anterior cingulate cortex (BA24) cortex cingulate - anterior Brain Brain - nucleus accumbens (basal accumbens ganglia) - nucleus Brain Brain - nucleus accumbens (basal accumbens ganglia) - nucleus Brain (a) (b)

2.5 2

2.0 1 1.5

1.0 0 Log10 (TPM) Log10 (TPM) 0.5 –1 0.0

–2 Lung Liver Testis Ovary Lung Liver Spleen Uterus Testis Vagina Ovary Spleen Bladder yroid Uterus Prostate Vagina Stomach Pituitary Pancreas Bladder yroid Prostate Stomach Pituitary Pancreas Whole blood Whole Nerve - tibial Nerve Artery - tibial Brain - cortex Brain Artery - aorta Adrenal gland Adrenal Whole blood Whole Fallopian tube Fallopian Nerve - tibial Nerve Artery - tibial Brain - cortex Brain Artery - aorta Adrenal gland Adrenal Fallopian tube Fallopian Kidney - cortex Colon - sigmoid Colon Muscle - skeletal Muscle Kidney - cortex Brain - amygdala Brain Colon - sigmoid Colon Artery - coronary Muscle - skeletal Muscle Brain - amygdala Brain Colon - transverse Colon Brain - cerebellum Brain Artery - coronary Cervix - ectocervix Colon - transverse Colon Brain - cerebellum Brain Cervix - endocervix Heart - le ventricle Heart Cervix - ectocervix Esophagus - mucosa Esophagus Minor salivaryMinor gland Brain - hippocampus Brain Cervix - endocervix Heart - le ventricle Heart Esophagus - mucosa Esophagus Brain - hypothalamus Brain Minor salivaryMinor gland Brain - hippocampus Brain Brain - hypothalamus Brain Brain - substantia nigra - substantia Brain Esophagus - muscularis Esophagus Adipose - subcutaneous Adipose Heart - atrial appendage Brain - substantia nigra - substantia Brain Breast - mammary tissue Esophagus - muscularis Esophagus Adipose - subcutaneous Adipose Heart - atrial appendage Breast - mammary tissue Brain - frontal cortex (BA9) cortex - frontal Brain Brain - frontal cortex (BA9) cortex - frontal Brain Brain - cerebellar hemisphere - cerebellar Brain Adipose - visceral (omentum) Adipose Skin - sun exposed (lower leg) - sun exposedSkin (lower Cells - transformed ﬁbroblasts Cells - transformed Brain - caudate (basal ganglia) - caudate Brain Brain - cerebellar hemisphere - cerebellar Brain Adipose - visceral (omentum) Adipose Skin - sun exposed (lower leg) - sun exposedSkin (lower Cells - transformed ﬁbroblasts Cells - transformed Brain - caudate (basal ganglia) - caudate Brain Brain - putamen (basal ganglia) - putamen Brain Small intestine - terminal ileum - terminal intestine Small Brain - spinal cord (cervical cord c-1) - spinal Brain Brain - putamen (basal ganglia) - putamen Brain Small intestine - terminal ileum - terminal intestine Small Brain - spinal cord (cervical cord c-1) - spinal Brain Skin - not sun exposed - not Skin (suprapubic) Skin - not sun exposed - not Skin (suprapubic) Cells - EBV-transformed lymphocytes Cells - EBV-transformed Esophagus - gastroesophageal junction - gastroesophageal Esophagus Cells - EBV-transformed lymphocytes Cells - EBV-transformed Brain - anterior cingulate cortex (BA24) cortex cingulate - anterior Brain Esophagus - gastroesophageal junction - gastroesophageal Esophagus Brain - anterior cingulate cortex (BA24) cortex cingulate - anterior Brain Brain - nucleus accumbens (basal accumbens ganglia) - nucleus Brain Brain - nucleus accumbens (basal accumbens ganglia) - nucleus Brain (c) (d)

2 2.0

1.5 1

1.0

0 0.5 Log10(TPM) Log10 (TPM) 0.0 –1

–0.5

–2 Liver Lung Testis Lung Liver Testis Ovary Spleen Uterus Vagina Ovary Spleen Uterus Vagina Bladder yroid Prostate Bladder yroid Stomach Prostate Pituitary Pancreas Stomach Pituitary Pancreas Whole blood Whole Nerve - tibial - Nerve Whole blood Whole Nerve - tibial Nerve Artery tibial - Brain - cortex Brain Artery - aorta Adrenal gland Adrenal Fallopian tube Fallopian Artery - tibial Brain - cortex Brain Artery - aorta Adrenal gland Adrenal Fallopian tube Fallopian Kidney - cortex Kidney - cortex Colon - sigmoid Colon Muscle - skeletal - Muscle Colon - sigmoid Colon Brain - amygdala Brain Muscle - skeletal Muscle Brain - amygdala Brain Artery - coronary Artery - coronary Colon - transverse Colon Brain - cerebellum Brain Colon - transverse Colon Brain - cerebellum Brain Cervix - ectocervix Cervix - ectocervix Cervix - endocervix Esophagus - mucosa Esophagus Heart - le Ventricle Heart Heart - le ventricle Heart Minor salivaryMinor gland Cervix - Endocervix Esophagus - mucosa Esophagus Brain - hippocampus Brain Minor salivaryMinor gland Brain - hippocampus Brain Brain - hypothalamus Brain Brain - hypothalamus Brain Brain - substantia nigra - substantia Brain Esophagus - muscularis Esophagus Adipose - subcutaneous Adipose Brain - substantia nigra - substantia Brain Heart - atrial appendage Esophagus - muscularis Esophagus Adipose - subcutaneous Adipose Heart - atrial appendage Breast - mammary tissue Breast - mammaryBreast tissue Brain - frontal cortex (BA9) cortex - frontal Brain Brain - frontal cortex (BA9) cortex - frontal Brain Brain - cerebellar hemisphere - cerebellar Brain Adipose - visceral (omentum) Adipose Brain - cerebellar hemisphere - cerebellar Brain Skin - sun exposed (lower leg) - sun exposedSkin (lower Cells - transformed ﬁbroblasts Cells - transformed Brain - caudate (basal ganglia) - caudate Brain Adipose - visceral (omentum) Adipose Skin - sun exposed (lower leg) - sun exposedSkin (lower Cells - transformed ﬁbroblasts Cells - transformed Brain - caudate (basal ganglia) - caudate Brain Small intestine - terminal ileum - terminal intestine Small Brain - putamen (basal ganglia) - putamen Brain Small intestine - terminal ileum - terminal intestine Small Brain - spinal cord (cervical cord c-1) - spinal Brain Brain - spinal cord (cervical cord c-1) - spinal Brain Brain - Pputamen (basal ganglia) - Pputamen Brain Skin - not sun exposed - not Skin (suprapubic) Skin - not sun exposed - not Skin (suprapubic) Cells - EBV-transformed lymphocytes Cells - EBV-transformed Cells - EBV-transformed lymphocytes Cells - EBV-transformed Esophagus - gastroesophageal junction - gastroesophageal Esophagus Esophagus - gastroesophageal junction - gastroesophageal Esophagus Brain - anterior cingulate cortex (BA24) cortex cingulate - anterior Brain Brain - anterior cingulate cortex (BA24) cortex cingulate - anterior Brain Brain - nucleus accumbens (basal accumbens ganglia) - nucleus Brain Brain - nucleus accumbens (basal accumbens ganglia) - nucleus Brain (e) (f)

Figure 1: Matrix graphs of the relative expression levels of ARMCX genes in multiple normal tissues were determined with the GTEx Portal. ARMCX5 was expressed at a medium level (e), whereas the other ARMCX genes (ARMCX1, ARMCX2, ARMCX3, ARMCX4, and ARMCX6) were expressed at low levels (a-d, f). transmission by producing and transmitting signals that Recent studies have shown a strong implication of diﬀer- aﬀect gene expression [5, 9]. Studies have revealed that ent members of the Armcx1-6/Armc10 family in human armadillo repeat proteins regulate protein interactions tumorigenesis [14–16]. For instance, some members of the through multiple binding domains such as nuclear trans- Armcx cluster can be regulated through the WNT signaling port, transcriptional activation, and cell connectivity [10]. pathway by interacting with transcription factors of the For example, bioinformatics analysis shows that ARMCX1, E-cadherin and T cytokine/lymphoid enhancement factor ARMCX2, and ARMCX3 are encoded by an single exon, con- (TCF/LEF) families [17, 18], which is also implicated in taining some ARM repeat domains, a DUF634 (domain 634 carcinogenesis and tumor progression [19–21]. function unknown) and an N-terminal transmembrane Although the ARMCX family plays an important role domain [11–13]. in many biological processes including cell adhesion, BioMed Research International 3

6 ⁎ ⁎ 6

5 5

4 4

3 3 Expression level 2 Expression level 2

1 1

0 0

STAD STAD Tumor tissue (n=408) Tumor tissue (n=408) Normal tissue (n=211) Normal tissue (n=211) (a) (b) ⁎ 6 5

5 4

4 3 3

Expression level 2 = Expression level 2

1 1

0 0

STAD STAD Tumor tissue (n=408) Tumor tissue (n=408) Normal tissue (n=211) Normal tissue (n=211) (c) (d)

Figure 2: Continued. 4 BioMed Research International

⁎ 5 6

4 5

4 3

3 2 Expression level

Expression level 2

1 1

0 0

STAD STAD Tumor tissue (n=408) Tumor tissue (n=408) Normal tissue (n=211) Normal tissue (n=211) (e) (f)

Figure 2: Boxplots of ARMCX family gene levels in primary gastric tumors and adjacent tissues from databases. ARMCX1, ARMCX2, and ARMCX4 were lowly expressed in primary gastric tumors (a, b, d). ARMCX5 was less expressed in normal gastric tissues (e). The expression levels of ARMCX3 and ARMCX6 have no significant difference between gastric tumors and normal gastric tissues (c, f). ∗P <0:05. tumorigenesis, and embryogenesis [22]. However, the rela- accessed Sept 25, 2019) [24].The analysis of ARMCX mRNA tionship between ARMCX genes and gastric cancer is expression between primary gastric cancer tissue and adja- poorly understood. Therefore, in this study, we determined cent normal tissue was done by Gene Expression Profiling the associations between expression levels of ARMCX Interactive Analysis (GEPIA, http://gepia.cancer-pku.cn/, genes and clinical outcomes of GC prognosis, with the accessed Sept 25, 2019) [25]. aim of providing insightful information regarding ARMCX genes as a novel prognostic biomarker for GC patients. 2.3. Bioinformatics Characteristic of ARMCX Genes. Gene function enrichment analysis of ARMCX genes was 2. Material and Methods performed to disclose the biological processes and signal pathways using the Database for Annotation and Enrich- 2.1. Data Source and Patient Information. First, we identified ment KOBAS 3.0 (http://kobas.cbi.pku.edu.cn/index.php). the genes differentially expressed between normal gastric The analysis included biological processes and molecular tissue and primary tumors of the ARMCX family using an function, but no results for the ARMCX family were online database (http://merav.wi.mit.edu/; accessed Sept 25, obtained. GeneMANIA was employed to reveal the gene- 2019). Then, we obtained mRNA expression levels of gene and protein-protein interactions of ARMCX family ARMCX1, ARMCX2, ARMCX3, ARMCX4, ARMCX5, and (http://www.genemania.org/, accessed Sept 26, 2019) [26, ARMCX6 by using The Cancer Genome Atlas (TCGA, 27]. Additionally, the relationship among ARMCX1, http://tcga-data.nci.nih.gov/tcga) and OncoLnc website ARMCX2, ARMCX3, ARMCX4, ARMCX5, and ARMCX6 (http://www.oncolnc.org/; accessed Sept 25, 2019) [23]. was evaluated using Pearson’s correlation coefficient. Results We downloaded the clinical information of 415 gastric with a P value < 0.001 were considered to be statistically cancer patients from UCSC Xena (http://xena.ucsc.edu/, significant. accessed Sept 25, 2019), including age, gender, tumor stage, survival time, and survival status. Next, a total of 351 cases 2.4. Survival Analysis. According to the database, the 351 GC were included for follow-up analysis after excluding the cases patients were, respectively, divided into low- and high- with missing medical data and 0-day survival time. expression groups for survival analysis. Overall survival (OS) and median survival time (MST) were used to assess 2.2. Characteristics of Gene Expression Levels. The high- the prognosis of patients with gastric cancer, to evaluate the expression and low-expression groups of ARMCX genes correlation of ARMCX member mRNAs with patient sur- were distinguished according to the median of each gene. vival by Kaplan-Meier estimator with a log-rank test. The The relative expression levels of ARMCX genes in multiple relative risk of survival in gastric cancer patients was assessed normal tissues were determined with the Genotype- by calculating the hazard ratio (HR) and 95% confidence Tissue Expression Portal (http://www.gtexportal.org/home/, interval (CI). BioMed Research International 5

Catalytic complex

Protein modiﬁcation by small protein conjugation or removal

Protein modiﬁcation by small protein conjugation

Protein ubiquitination

Transferase complex

Ubiquitin ligase complex

Cullin-RING ubiquitin ligase complex

Cul4-RING E3 ubiquitin ligase complex

0 0.5 1 1.5 2 2.5 3 3.5 –log10 (a)

CTNND1 CTNND2 CTC-554D6.1 ARMCX1 ARMCX2 ARMCX4 ARMCX6 ARMCX3 ARMCX5 1 INSC GPRASP1 ARVCF ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ARMCX1 1 0.77** 0.33 0.6 –0.17 0.23 0.8

0.6 GPRASP2 ARMCX4 BHLHB9 ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ PKP3 ARMC8 ARMCX2 0.77 1 0.43 0.54 –0.22 0.3 0.4 ARMCX5 ARMCX3 ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ 0.2 ARMCX3 0.33 0.43 1 0.34 0.25 0.63 RSPH14 ARMC7 0 ARMCX6 ARMCX2 ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ARMCX4 0.6 0.54 0.34 1 –0.09 0.18 – RTTN ARMCX1 ARMC12 0.2

USO1 ARMC5 –0.4 ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ARMCX5 –0.17 –0.22 0.25 –0.09 1 0.14 ARMC10 – ZER1 ARMC3 0.6 ARMC2 ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ –0.8 Co-expression ARMCX6 0.23 0.3 0.63 0.18 0.14 1 Co-locolization –1 (b) (c)

Figure 3: Functional assessment and bioinformatics analysis of ARMCX family genes (a). GeneMANIA constructed the gene-gene interaction network of the ARMCX family (b). Matrix graphs of Pearson’s correlations of ARMCX family gene expression levels (c). ∗∗P <0:001.

2.5. Joint-Effects Analysis. By analyzing the TCGA data, the evaluate the individual prognosis. Furthermore, the probable results have shown that only ARMCX1 and ARMCX2 had utility of the ARMCX family in predicting clinical grade was statistical significance. The combination of ARMCX1 and evaluated. In terms of clinical data and survival analysis, age, ARMCX2 was investigated by joint-effects analysis. The tumor stage, and ARMCX expression level were included combination included group 1 (low ARMCX1 and low in the risk model after Cox proportional risk regression ARMCX2 expression), group 2 (low ARMCX1 and high model adjustment. Scores for each factor could be ARMCX2 expression, High ARMCX1 and low ARMCX2 counted, and 1-year, 5-year, and 10-year survival rates also expression), and group 3 (high ARMCX1 and high ARMCX2 can be calculated [28]. expression). In addition, according to the results of TCGA database, age and tumor stage were adjusted in the Cox 2.7. Gene Set Enrichment Analysis. In order to explore the proportional hazards regression model. difference in biological functions and pathways in the survival of GC between low- and high-ARMCX gene expression 2.6. Nomogram Model. Due to the clinical characteristics and groups, the potential mechanism in the molecular signature risk score, a nomogram prediction model was constructed to database (MSigDB) of c2 (c2.all.v6.1. Symbols) and c5 6 BioMed Research International

Table 1: Clinical data characteristic of 351 GC patients.

Items Cases (total n = 351) No. of events (%) MST (days) Crude P Crude HR (95% CI) Age ≥60 239 108 (45.2) 766 Ref. 0.017 <60 106 35 (33.0%) 1811 0.629 (0.429-0.923) Missing 6 Gender Male 236 100 (42.4%) 869 Ref. Female 125 44 (35.2%) 1043 0.184 0.787 (0.552-1.122) Missing 0 Tumor stage i 47 11 (23.4%) 2197 0.260 (0.126-0.537) ii 109 34 (31.2%) 1686 0.424 (0.247-0.728) iii 147 69 (46.9%) 779 <0.001 0.643 (0.397-1.042) iv 35 Missing 13 22 (62.9%) 476 Ref. Abbreviations: MST: median survival time; Ref.: reference; HR: hazard ratio; 95% CI: 95% conﬁdence interval.

Table 2: Univariate and multivariate survival analyses of the ARMCX family.

Items Cases (total n = 351) No. of events (%) MST (days) Crude P Crude HR (95% CI) Adjusted P Adjusted HR (95% CI) ARMCX1 Low 175 61 (34.6%) 1294 0.667 (0.479-0.929) 0.620 (0.440-0.874) 0.016 0.006 High 176 83 (47.6%) 766 Ref. Ref. ARMCX2 Low 176 58 (33.0%) 1686 0.655 (0.469-0.915) 0.610 (0.432-0.861) 0.012 0.005 High 175 86 (49.1%) 762 Ref. Ref. ARMCX3 Low 176 66 (37.5%) 1095 Ref. Ref. 0.366 0.690 High 175 78 (44.6%) 794 1.163 (0.838-1.616) 1.072 (0.763-1.506) ARMCX4 Lo 176 70 (39.8%) 1095 Ref. Ref. 0.570 0.220 High 175 74 (42.3%) 801 1.099 (0.793-1.525) 1.238 (0.880-1.741) ARMCX5 Low 176 76 (43.2%) 728 Ref. Ref. 0.271 0.507 High 175 68 (38.9%) 1153 0832 (0.599-1.155) 0.891 (0.634-1.253) ARMCX6 Low 176 73 (41.5%) 1153 Ref. Ref. 0.786 0.929 High 175 71 (40.6%) 874 0.956 (0.689-1.326) 1.016 (0.722-1.428) Abbreviations: MST: median survival time; Ref.: reference; HR: hazard ratio; 95% CI: 95% conﬁdence interval; Notes: Adjusted P, adjustment for age and tumor stage.

(c5.all.v6.1. Symbols) was studied by GSEA (http://software Hochberg procedure was employed for multiple tests of FDR .broadinstitute.org/gsea/index.jsp, accessed Sept 27, 2019) in GSEA to control [31–33], and P <0:05 was considered [29–31].The nominal P value < 0.05 and the false discovery statistically signiﬁcant. GraphPad Prism v.6.0 (La Jolla, CA) rate (FDR) <0.25 for the enriched gene sets in GSEA were was used to draw vertical scatter plots and survival curves. statistically signiﬁcant. SPSS software v.22.0 (IBM, Chicago, IL, USA) was employed for statistical analysis. 2.8. Statistical Analysis. Survival analysis was carried out by Kaplan-Meier and the log-rank test to calculate MSTs and 3. Results P values. The crude or adjusted HR and 95% CI were calculated using the Cox proportional risk regression model for 3.1. ARMCX mRNA Expression Analysis. In human normal univariate and multivariate survival analyses. The Benjamini stomach tissue, ARMCX5 was expressed at a medium level BioMed Research International 7

100 100

80 80

60 60

40 40 Percent survival 20 Percent survival 20 Log-rank P=0.016 Log-rank P=0.013 0 0 HR =0.6 67 (0.4 79- 0.92 9) HR=0.655 (0.469-0.915) 0 1000 2000 3000 4000 0 1000 2000 3000 4000 Overall survival (days) Overall survival (days) Low ARMCX1 Low ARMCX2 High ARMCX1 High ARMCX2 (a) (b) 100 100

80 80

60 60

40 40 Percent survival Percent 20 Percent survival 20 Log-rank P=0.367 Log-rank P=0.570 0 0 HR=1.163 (0.838-1.616) HR=1.009 (0.793-1.525) 0 1000 2000 3000 4000 0 1000 2000 3000 4000 Overall survival (days) Overall survival (days) Low ARMCX3 Low ARMCX4 High ARMCX3 High ARMCX4 (c) (d) 100 100

80 80

60 60

40 40 Percent survival Percent survival 20 20 Log-rank P=0.271 Log-rank P=0.786 0 0 HR=0.832 (0.599-1.155) HR=0.956 (0.689-1.326) 0 1000 2000 3000 4000 0 1000 2000 3000 4000 Overall survival (days) Overall survival (days) Low ARMCX5 Low ARMCX6 High ARMCX5 High ARMCX6 (e) (f)

Figure 4: Univariate curves of OS of ARMCX family genes in GC. The lower expression levels of ARMCX1 and ARMCX2 were signiﬁcantly associated with satisfactory OS results (a, b). The expression of ARMCX3, ARMCX4, ARMCX5, and ARMCX6 mRNA did not have signiﬁcant prognostic value for OS (c, d).

(Figure 1(e), whereas the other ARMCX genes (ARMCX1, ARMCX1, ARMCX2, and ARMCX4 were lowly expressed ARMCX2, ARMCX3, ARMCX4, and ARMCX6) were in primary gastric tumors and has a high expression in nor- expressed at low levels (Figures 1(a)-1(d) and 1(f)), com- mal gastric tissues. However, conversely, ARMCX5 was less pared with other normal tissues. Box plots of ARMCX1-6 expressed in normal gastric tissues than in primary gastric genes were downloaded from GEPIA as shown in Figure 2. tumors. The expression levels of ARMCX3 and ARMCX6 8 BioMed Research International have no significant difference between gastric tumors and Table 3: Grouping information for joint-effects analysis. normal gastric tissues. Group Combinations 3.2. Bioinformatics and Functional Annotation Analyses of 1 Low ARMCX1+low ARMCX2 the ARMCX Genes. Enrichment and functional analyses by Low ARMCX1+high ARMCX2 2 KOBAS revealed that ARMCX genes were significantly High ARMCX1+low ARMCX2 enriched in ubiquitin ligase complex and the process of 3 High ARMCX1+high ARMCX2 protein modification (Figure 3(a)). However, we have not Abbreviations: ARMCX: arm protein lost in epithelial cancers, X found any associations of the ARMCX family using Kyoto chromosome. Encyclopedia of Genes and Genomes (KEGG) and Database for Annotation, Visualization, and Integrated Discovery (DAVID) analyses. By analyzing gene-gene and protein- shown in Table 4. Group 1 had the longest MST of 1686 protein interaction networks, we confirmed that the ARMCX days (adjusted P =0:003), while group 3 had the shortest family had strong protein homology and coexpression at MST of 762 days (adjusted P =0:012). Kaplan–Meier sur- both gene and protein levels, as shown in Figure 3(b). vival analyses of ARMCX1 and ARMCX2 are shown in Figure 5. Low expression levels of ARMCX1 and 3.3. Correlation Analysis Value Assessment of the ARMCX ARMCX2 in group 1 were significantly correlated with Family. Coexpression analyses of individual ARMCX genes ’ ffi better clinical outcome. In group 3, high expression of were analyzed using Pearson s correlation coe cient. The ARMCX1 and ARMCX2 was correlated with poor OS expression level of ARMCX1, ARMCX2, ARMCX3, and (log-rank P =0:007). ARMCX6 was correlated with each other. Furthermore, there was no significant correlation between the expressions of ARMCX4 and ARMCX5, but both the expressions of 3.7. Nomogram Model. Nomogram risk scoring includes ARMCX4 and ARMCX5 all related to the other members age, tumor stage, and the expression level of ARMCX1 of the ARMCX family (∗∗P <0:01; Figure 3(c)). and ARMCX2 to calculate 1-year, 5-year, and 10-year related survival rates. The higher total points, the lower 3.4. Clinical Characteristics of GC Patients. There were 351 survival rate, and the results substantiated that high GC patients who had prognosis information included in expression levels of ARMCX1 and ARMCX2, age of the the current study; UCSC Xena dataset is shown in patient (>60 years old), and advanced tumor stage estab- Table 1. The univariable survival analysis revealed that lished a prognostic feature that conduced to the highest age and tumor stage were correlated with MST in combi- risk for poor OS (Figure 6). nation with clinical data (P =0:017 and P <0:001, respectively), and preliminary stage was significantly correlated with favorable MST (2197 days, P <0:001, HR = 0:260, 3.8. Gene Set Enrichment Analysis. In order to further 95%CI = 0:126 – 0:537). On the other hand, gender was explore the underlying mechanisms of ARMCX genes not associated with MST. in GC prognosis, we used the PAAD genome-wide RNA sequencing dataset for GSEA. GSEA results of 3.5. Survival Analysis of the ARMCX Gene Family. Survival the c2 reference gene set revealed that a low ARMCX1 analysis is shown in Table 2 and Figure 4. Due to the age expression was involved in the WNT signaling pathway, and tumor stage that were related with MST, both age and regulation of cell metastasis (Figure 7(a)) and cell cycle tumor stage were analyzed using the multivariate Cox biological processes (Figures 7(b)-7(h)), and poor sur- proportional risk regression model. In univariate survival vival of lung cancer (Figure 7(i)). Also, the enrichment analysis, lower expression levels of ARMCX1 and ARMCX2 of c5 indicates that low ARMCX1 is also involved in were significantly associated with satisfactory OS results cell division (Figure 8(c)), cell cycle (Figures 8(a) and (log-rank P =0:016, HR = 0:667, 95%CI = 0:479 – 0:929; 8(b)), gene silencing (Figure 8(d)), RNA modification log-rank P =0:013, HR = 0:655, 95%CI =0:469 – 0:915, (Figure 8(i)), and NF-kappaB signaling pathway respectively; Figures 4(a) and 4(b)). The expression of (Figures 8(g) and 8(h)). GSEA results of c2 enrichments ARMCX3, ARMCX4, ARMCX5, and ARMCX6 mRNA reveal that the low expression of ARMCX2 was correlated to did not have a significant prognostic value for OS (log- the cell cycle biological process (Figures 9(a), 9(b), and 9(f)), rank P =0:367, 0.570, 0.271, and 0.786, respectively; regulation of apoptosis (Figure 9(h)), DNA replication Figures 4(c)-4(f)). (Figure 9(c)) and damage response (Figure 9(g)), and E2F, WNT, and NF-kappaB signaling pathways (Figures 9(d), 3.6. Joint-Effects Analysis of ARMC1 and ARMCX2. Based on 9(e), and 9(e)), whereas the c5 enrichments suggest that the findings in the multivariate survival analysis, ARMCX1 low ARMCX2 expression is involved in the biological process and ARMCX2 were associated with a significantly different of cell division (Figure 10(d)), cell cycle (Figures 10(b), 10(c), survival. A joint-effects analysis was employed to further and 10(i)), apoptosis (Figure 10(e)), gene silencing determine the combined effects in prognostic prediction of (Figure 10(g)), DNA damage checkpoint (Figure 10(f)), and ARMCX1 and ARMCX2 (grouped as summarized in the NF-kappaB signaling pathway (Figure 10(a)). Moreover, Table 3). The combination of ARMCX1 and ARMCX2 the remaining results of this study can be seen in Supplemen- included group 1, group 2 and group 3, and results are tary Tables 1 and 1. BioMed Research International 9

Table 4: Joint-eﬀects analysis of the combination of ARMCX1 and ARMCX2.

Items Cases (total n = 351) No. of events (%) MST (days) Crude P Crude HR (95% CI) Adjusted P Adjusted HR (95% CI) Group 1 143 46 (32.2%) 1686 0.007 0.600 (0.414-0.870) 0.003 0.563 (0.384-0.825) Group 2 65 27 (41.5%) 766 0.549 0.873 (0.559-1.362) 0.506 0.854 (0.537-1.358) Group 3 143 71 (49.7%) 762 0.025 Ref. 0.012 Ref. Abbreviations: MST: median survival time; Ref.: reference; HR: hazard ratio; 95% CI: 95% conﬁdence interval; Notes: Adjusted P, adjustment for age and tumor stage.

100

80 Log-rank P=0.007 60

Percent survival Percent 20 Log-rank P=0.024 0

0 1000 2000 3000 4000 Overall survival (days) Group1 Group2 Group3

Figure 5: Survival curves for joint-eﬀects analysis of the combination of ARMCX1 and ARMCX2 genes in the TCGA database. Low- expression levels of ARMCX1 and ARMCX2 in group 1 were signiﬁcantly correlated with better clinical outcome. In group 3, high expression of ARMCX1 and ARMCX2 was correlated with poor OS.

0 102030405060708090100 Points High expression ARMCX1 Low expression High expression ARMCX2 Low expression Low expression ARMCX3 High expression Low expression ARMCX4 High expression High expression ARMCX5 Low expression Low expression ARMCX6 High expression ≥60 Age <60 ii iv Tumor stage i iii Total points 0 20 40 60 80 100 120 140 160 180 200 1−year survival 0.95 0.9 0.85 0.8 0.75 0.7 0.6 0.5 3−year survival 0.85 0.8 0.75 0.7 0.6 0.5 5−year survival 0.8 0.75 0.7 0.6 0.5 10−year survival 0.75 0.7 0.6 0.5

Figure 6: Nomogram for predicting the 1-, 3-, 5-, and 10-year events (death) with risk scores and clinical parameters. The high expression levels of ARMCX1 and ARMCX2, age of the patient (>60 years old), and advanced tumor stage established a prognostic feature that conduced to the highest risk for poor OS. 10 BioMed Research International

Enrichment plot: BIDUS_METASTASIS_UP Enrichment plot: FISCHER_G2_M-CELL_CYCLE Enrichment plot: KEGG-CELL_CYCLE 0.6 0.6 NES=2.04NES=2E 040 0.6 NES=2.07NES=2E 070 NES=1.99NES=1.E 999 0.5 0.5 NOMO P=0.=0.00200020022 0.5 NOMO P=0.006=0 0066 NOMO P=0.004= 4 0.4 0.4 FDRD q=0.=0.01200120 012 0.4 FDRDRD q=0.=0.01400140 014 FDRD q=0.015=0.0 015 0.3 0.3 0.3

0.2 0.2 0.2 Enrichment score (ES) score Enrichment

0.1 (ES) score Enrichment

Enrichment score (ES) score Enrichment 0.1 0.1 0.0 0.0 0.0

‘Low’ow (positively correl correlated)ated)ed) 0.50 0.50 ‘Low’ow (positively(posit ivelyv correlated) 0.50 ‘Low’ow (positively correlated) 0.25 0.25 0.25 0.00 0.00 0.00 – Zero crossr s at 108388 Zero crosso aat 108383 Zero crosscrossr s aatt 1083108388 8 0.25 –0.25 –0.25 – 0.50 –0.50 –0.50 – 0.75 –0.75 –0.75 – ‘High’ (negatively correlated) ‘High’ (negatively(n correcorrelated)lated) ‘High’i (negativelyi correlated) 1.00 –1.00 –1.00 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked metric list (Signal2Noise) Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (a) (b) (c)

Enrichment plot: REACTOME_CELL_UP Enrichment plot: REACTOME_CELL_CYCLE_MITOTIC Enrichment plot: 0.6 REACTOME_MITOTIC_G1_G1_S_PHASES NES=2.09NESS=2. 209 9 0.6 NES=2.10NES=2E 1 10 0.7 0.5 NES=2.03N 2 NOMM P<0.<0.00100010 001 0.5 NOMO P=0.002=00022 0.6 0.4 NOMN P=0.0040 FDRFFDDDRR q=0.=0.01500150150 0.4 FDRFD q=0.01500150 0.5 0.3 q 0.3 0.4 FDRF =0.0130.0133 0.3 0.2 0.2 0.2

Enrichment score (ES) score Enrichment 0.1 Enrichment score (ES) score Enrichment 0.1

Enrichment score (ES) score Enrichment 0.1 0.0 0.0 0.0

0.50 ‘Low’ow (positively o correlated) r 0.50 ‘Low’ow (positivelys correlated)r 0.50 ‘Low’ow (positively s correlated) r 0.25 0.25 0.25 0.00 0.00 0.00 Zero crossro at 108383 Zero crossr s at 1083810838 –0.25 –0.25 Zero crosscrosso aatt 10831083838 –0.25 – –0.50 –0.50 0.50 –0.75 – –0.75 ‘High’i (negatively(negativelyi correcorrelated)lated) 0.75 ‘High’ (negatively( correcorrelated)lated) ‘High’ (negatively(negativelyi correcorrelated)lated)l –1.00 –1.00 –1.00 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (d) (e) (f)

Enrichment plot: Enrichment plot: REACTOME_SIGNALING_BY_WNT Enrichment plot: REACTOME_MITOTIC_G1_G1_S_PHASES SHEDDEN_LUNG_CANCER_POOR_SURVIVAL_A6 0.6 0.7 NES=1.90NESE 19 90 NES=2.06NESS=2.066 0.5 0.6 NES=2.13NESE 2.131 0.6 NOMO P=0.008=00088 NOMM P<0.0010 0.4 0.5 NOMO P=0.002==0 022 0.5 FDRFD q=0.=0.0240 024 0.4 FDRFFDDRR q=0.=0=0.0130130 0.3 0.4 FDRFD q=0.021=00 021 0.3 0.3 0.2 0.2 0.2

Enrichment score (ES) score Enrichment 0.1 Enrichment score (ES) score Enrichment Enrichment score (ES) score Enrichment 0.1 0.1 0.0 0.0 0.0

0.50 ‘Low’ow (positively(posit o ively correlated)correl r ated) 0.50 ‘Low’ow (positively (poso t ve y cocorrelated)r e ated) 0.50 ‘Low’ow (positively s correlated)r 0.25 0.25 0.25 0.00 0.00 0.00 –0.25 Zero crosscr s ata 1083108388 8 –0.25 Zero crossro at 108383 –0.25 Zero crossro at 108383 –0.50 –0.50 –0.50 –0.75 – –0.75 ‘High’ (negatively( correlated)l 0.75 ‘High’h (negatively correlated)t – ‘High’ (negatively( correlated) – 1.00 –1.00 1.00 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (g) (h) (i)

Figure 7: GSEA results of low ARMCX1 expressed in GC patients, using gene set c2. The low ARMCX1 expression was involved in the WNT signaling pathway, regulation of cell metastasis (a) and cell cycle biological processes (b-h), and poor survival of lung cancer (i). Abbreviations: FDR: false discovery rate; GSEA: gene set enrichment analysis; KEGG: Kyoto Encyclopedia of Genes and Genomes; NES: normalized enrichment score; NOM: nominal.

4. Discussion 2 both alone and in combination may serve as potential biomarkers of GC. In our present study, we elucidated the associations In 1989, the armadillo family proteins were first discov- between the expression levels of ARMCX 1-6 genes with ered in the polar gene fragment of Drosophila [3]. Subse- the prognosis of GC patients. Our research disclosed that quently, more and more proteins containing arm repeats ARMCX 1 and ARMCX 2 contribute significantly to OS, have been analyzed and sequenced. Armadillo repeats con- but ARMCX 3-6 show no significant association with taining x-chain (ARMCX 1-6) are involved in many biologi- OS. Thus, the expression levels of ARMCX 1 and ARMCX cal processes, such as mediating protein-protein interactions BioMed Research International 11

Enrichment plot: GO_CELL_CYCLE_CHECKPOINT Enrichment plot: Enrichment plot: GO_CELL_DIVISION GO_CELL_CYCLE_G1_S_PHASE_TRANSITION 0.5 0.5 NES=2.04 0.6 NES=1.67 P NES=1.99 0.4 0.4 NOM <0.001 NOM P=0.=0.022022 q 0.5 NOM P=0.002 FDRF =0.013 0.3 FDR q=0.=0.064064 0.3 0.4 FDR q=0.013 0.3 0.2 0.2 0.2 0.1

0.1 (ES) score Enrichment Enrichment score (ES) score Enrichment 0.1 0.0 (ES) score Enrichment 0.0 0.0

‘Low’Low (positively(positively correlatecorrelcorrelated)ated) 0.50 0.50 ‘Low’’( (positivelyo l correlated)r l d) 0.50 ‘Low’Low (positively(posits ively correlcorrelated)r ated)ate 0.25 0.25 0.25 0.00 0.00 0.00 Zero crosscrossr s aatt 1083108388 8 Zero crosscrossr s aatt 1083108388 8 –0.25 –0.25 Zero crossc s at 10838 10838 –0.25 –0.50 –0.50 –0.50 –0.75 –0.75 –0.75 ‘High’ (negatively correlated) ‘High‘High’’ (negatively correcorrelated)lated) ‘High’ (negatively correlated) –1.00 –1.00 –1.00 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked metric list (Signal2Noise) Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset Ranked metric list (Signal2Noise) Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (a) (b) (c)

Enrichment plot: GO_GENE_SILENCING Enrichment plot: GO_GENE_SILENCING_BY_RNA Enrichment plot: GO_NEGATIVE_REGULATION_OF_CELL_DIVISION 0.5 NES=1.99NES=1.99 NES=2.04 0.5 P 0.6 NES=2.11NES=2.11 0.4 NOM P<0.001 NOM <0.001 0.4 0.5 P FDR q=0.=0.013013 FDR q=0.013 NOM <0.001<0.001 0.3 0.3 0.4 FDR q=0.011 0.3 0.2 0.2 0.2 0.1 0.1 Enrichment score (ES) score Enrichment Enrichment score (ES) score Enrichment 0.1 Enrichment score (ES) score Enrichment 0.0 0.0 0.0

0.50 ‘Low’ow (positively o correlated) r 0.50 ‘Low’Low (positively correlcorrelated)ate 0.50 ‘Low’‘L ’ (positivelys l correlated)r l 0.25 0.25 0.25 0.00 0.00 0.00 Zero crossr s at 108388 –0.25 Zero crossc s at 1083108388 8 –0.25 –0.25 Zero cross s at 10838 10838 –0.50 –0.50 –0.50 – –0.75 –0.75 0.75 ‘High’h (negatively correlated) ‘High’h (negatively(negatively correcorrelated)lated) – ‘High’ (negatively correcorrelated)lated) –1.00 –1.00 1.00 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (d) (e) (f)

Enrichment plot: GO_NF_KAPPAB_BINDING Enrichment plot: GO_NIK_NF_KAPPAB_SIGNALING Enrichment plot: GO_RNA_MODIFICATION 0.6 NES=1.88NES=1.88 0.6 NES=1.94 0.6 NES=1.99 0.5 NOM P=0.004 0.5 NOM P=0.002 0.5 NOM P<0.001 0.4 FDR q=0.020 0.4 FDR q=0.015 0.4 FDR q=0.=0.013013 0.3 0.3 0.3

0.2 0.2 0.2 0.1 0.1 Enrichment score (ES) score Enrichment Enrichment score (ES) score Enrichment 0.1 (ES) score Enrichment

0.0 0.0 0.0

0.50 ‘Low’Low (positively correlatecorrelated) 0.50 ‘Low’Low (positively(posits ely correlcorrelated)r ated)ate 0.50 ‘Low’Low (positively(posits ivelyivev correlated)correlr ated 0.25 0.25 0.25 0.00 0.00 0.00 Zero crossr s at 108383 –0.25 Zero crosscrossr s aatt 1083108388 8 –0.25 Zero crossr s at 108388 –0.25 –0.50 –0.50 –0.50 –0.75 –0.75 –0.75 ‘High’h’ (negatively correcorrelated)lated) ‘High’h (negatively correlated) ‘High’i (negatively(negatively correcorrelated)lated) –1.00 –1.00 –1.00 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (g) (h) (i)

Figure 8: GSEA results of low ARMCX1 expressed in GC patients, using gene set c5. The low ARMCX1 is also involved in cell division (c), cell cycle (a, b), gene silencing (d), RNA modiﬁcation (i), and the NF-kappaB signaling pathway (g, h). Abbreviations: FDR: false discovery rate; GSEA: gene set enrichment analysis; KEGG: Kyoto Encyclopedia of Genes and Genomes; NES: normalized enrichment score; NOM: nominal. and intervening in cell assembly, nuclear transport, and tran- ARMCX1, ARMCX2, and ARMCX3 are located in the scriptional activation [34]. Many studies have demonstrated chromosome region xq21.33-q22.2, respectively. Their that ARMCX is associated with the risk and prognosis of sev- amino N-terminal region has a transmembrane domain, eral diseases. For instance, the ARMCX family plays an indicating that these proteins may be located in the mem- important role in embryogenesis and tumorigenesis [22]. brane structure of cells. ARMCX3 has been found to be a Scholars have found that some members of the ARMCX complete membrane protein of the mitochondrial outer protein family (Armcx1-3) were underexpressed in several membrane, which functions by interacting with transcription cancers of epithelial origin, including the lung, prostate, regulator Sox10 [12]. In addition, ARMCX4, ARMCX5, and colon, and pancreatic [11]. ARMCX6 were located in chromosome regions xq22.1, 12 BioMed Research International

Enrichment plot: FISCHER_G2_M_CELL_CYCLE Enrichment plot: KEGG_CELL_CYCLE Enrichment plot: KEGG_DNA_REPLICATION 0.8 0.6 0.6 NES=2.09 NES=2.09 0.7 NES=2.01NES=2.01 0.5 0.5 NOM P<0.002 NOM P<0.001 0.6 NOM P<0.001 0.4 0.4 FDRF q=0.005 FDR q=0.=0.005005 0.5 FDR q=0.007 0.3 0.3 0.4 0.3 0.2 0.2 0.2 Enrichment score (ES) score Enrichment Enrichment score (ES) score Enrichment 0.1 (ES) score Enrichment 0.1 0.1 0.0 0.0 0.0

‘Low’ (positively correlated) 0.5 0.5 ‘Low’ (positively correlated) 0.5 ‘Low’ (positively correlated)

0.0 0.0 0.0 Zero cross at 10838 Zero cross at 10838 Zero cross at 10838 –0.5 –0.5 –0.5

–1.0 –1.0 – ‘High’ (negatively(negatively correcorrelated)lated) ‘High’ (negatively correcorrelated)lated) 1.0 ‘High’ (negatively correlated) 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset metricRanked (Signal2Noise) list Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (a) (b) (c)

Enrichment plot: PID_E2F_PATHWAY Enrichment plot: Enrichment plot: REACTOME_G0_AND_EARLY_G1 REACTOME_ACTIVATION_OF_NF_KAPPAB_IN_B_CELLS 0.8 0.6 NES=2.11 0.7 0.5 NESNES=2.03=2.03 P 0.7 0.6 NOM <0.001 NOM P<0.001<0.001 0.4 0.6 NES=2.01NES=2.01 0.5 FDR q=0.005 q 0.3 0.5 NOM P<0.001 0.4 FDR =0.=0.006006 0.4 0.3 0.2 FDR q=0.007 0.3 0.2

Enrichment score (ES) score Enrichment 0.1 0.2 (ES) score Enrichment 0.1 0.0 0.1 0.0 Enrichment score (ES) score Enrichment 0.0

0.5 ‘Low’ (positively correlated) 0.5 ‘Low’ (positively correlated) 0.5 ‘Low’ (positively correlcorrelated)ated) 0.0 0.0 0.0 Zero crosscross aatt 1083108388 Zero cross at 10838 – Zero cross at 10838 0.5 –0.5 –0.5 – – 1.0 ‘High’ (negatively correlated) 1.0 ‘High’ (negatively correlated) –1.0 ‘High’ (negatively correlated) 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset metricRanked (Signal2Noise) list Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (d) (e) (f)

Enrichment plot: Enrichment plot: Enrichment plot: REACTOME_SIGNALING_BY_WNT REACTOME_P53_DEPENDENT_G1_DBA_DAMAGE_RESPONSE REACTOME_REGULATION_OF_APOPTOSIS 0.7 0.7 0.6 NES=2.09 0.6 NES=2.02 0.7 0.5 NOM P<0.001 0.5 NOM P<0.001 0.6 0.4 FDR q=0.005 0.5 0.4 FDR q=0.=0.007007 0.3 0.4 0.3 0.2 0.3 0.2

0.2 (ES) score Enrichment 0.1

Enrichment score (ES) score Enrichment 0.1 0.1 Enrichment score (ES) score Enrichment 0.0 0.0 0.0

‘Low’ (positively correlated) ‘Low’ (positively correlated) 0.5 ‘Low’ (positively correlated) 0.5 0.5

0.0 0.0 0.0 Zero cross at 10838 Zero cross at 10838 Zero cross at 10838 –0.5 –0.5 –0.5

–1.0 –1.0 – ‘High’ (negatively correlated) ‘High’ (negatively correlated) 1.0 ‘High’ (negatively correlated) 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (g) (h) (i)

Figure 9: GSEA results of low ARMCX2 expressed in GC patients, using gene set c2. The low expression of ARMCX2 was correlated to the cell cycle biological process (a, b, and f), regulation of apoptosis (h), DNA replication (c) and damage response (g), and the E2F, WNT, and NF-kappaB signaling pathways (d, i, and e). Abbreviations: FDR: false discovery rate; GSEA: gene set enrichment analysis; KEGG: Kyoto Encyclopedia of Genes and Genomes; NES: normalized enrichment score; NOM: nominal. xq22.1-q22.3, and xq21.33-q22.3, respectively. Studies have Here, we downloaded and analyzed data from GEO online shown that ARMCX5 can be activated by binding to the database to determine the potential relationship between oncogene ZnF217 [35] and ARMCX6 upexpressed at least ARMCX mRNA expression and clinical outcomes of patients 2-fold in peripheral blood monocytes of rheumatoid arthritis with gastric cancer. We observed significant differences in the patients compared to those identified using oligonucleotide expression of ARMCX1, ARMCX2, ARMCX4, and ARMCX5 array [36]. Moreover, regardless of their function in other between primary tumors and adjacent normal tissues, without diseases, they are associated with tumorigenesis and were ini- ARMCX3 and ARMCX6. More importantly, ARMCX1 and tially described as presumed tumor suppressors [11]. ARMCX2 are more highly expressed in adjacent normal BioMed Research International 13

Enrichment plot: Enrichment plot: Enrichment plot: GO_ACTIVATION_OF_NF_KAPPAB_INDUCING_ GO_CELL_CYCLE_G1_S_PHASE_TRANSITION GO_CELL_CYCLE_PHASE_TRANSITION KINASE_ACTIVITY 0.6 NES=2.04 NES=2.09 NES=1.91 0.5 0.6 0.5 NOM P<0.001 NOM P<0.<0.001001 P 0.4 0.5 NOM =0.006 0.4 FDR q=0.=0.009009 FDR q=0.008 0.4 0.3 FDR q=0.015 0.3 0.3 0.2 0.2 0.2 0.1 Enrichment score (ES) score Enrichment 0.1 (ES) score Enrichment 0.1

Enrichment score (ES) score Enrichment 0.0 0.0 0.0

‘Low’ (positively correlated) 0.5 ‘Low’ (positively correlated) 0.5 ‘Low’ (positively correlated) 0.5

0.0 0.0 0.0 Zero cross at 10838 Zero cross at 10838 Zero crosscross aatt 1083108388 –0.5 –0.5 –0.5

–1.0 – –1.0 ‘High’ (negatively(negatively correcorrelated)lated) 1.0 ‘High’ (negatively correlated) ‘High’ (negatively(negatively correcorrelated)lated) 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset Ranked metric list (Signal2Noise) Rank in ordered dataset Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (a) (b) (c)

Enrichment plot: GO_CELL_DIVISION Enrichment plot: Enrichment plot: GO_G1_DNA_DAMAGE_CHECKPOINT GO_EXECUTION_PHASE_OF_APOPTOSIS 0.5 NES=1.96 NES=2.04NES=2.04 0.5 0.4 0.5 NES=2.00NES=2.00 NOM P=0.002 NOM P<0.001 0.4 NOM P<0.001 q 0.3 FDR q=0.=0.009009 0.4 FDR =0.=0.011011 q 0.3 0.2 0.3 FDR =0.010 0.2 0.2 0.1 0.1 Enrichment score (ES) score Enrichment 0.1 (ES) score Enrichment 0.0 (ES) score Enrichment 0.0 0.0

‘Low’ (positively correlated) ‘Low’ (positively correlated) 0.5 0.5 ‘Low’ (posit(positivelyively correlated)correlated) 0.5

0.0 0.0 0.0 Zero cross at 1083108388 Zero cross at 10838 Zero cross at 10838 –0.5 –0.5 –0.5

– – – 1.0 ‘High’ (negatively correlated) 1.0 ‘High’ (negatively correlated) 1.0 ‘High’ (negatively correlated) 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (d) (e) (f)

Enrichment plot: GO_GENE_SILENCING_BY_RNA Enrichment plot: GO_NIK_NF_KAPPAB_SIGNALING Enrichment plot: 0.7 GO__REGULATIOIN_OF_CELL_CYCLE_PHASE_TRANSITION 0.5 0.5 NES=1.94 0.6 NES=2.18 NES=1.93 0.4 NOM P<0.001 0.5 NOM P<0.001 0.4 P q 0.4 q NOM <0.001 0.3 FDR =0.013 FDR =0.045 0.3 0.3 FDR q=0.013=0.013 0.2 0.2 0.2 0.1

Enrichment score (ES) score Enrichment 0.1

0.0 0.0 Enrichment score (ES) 0.0

‘Low’ (positively(positively correlated) ‘Low’ (positively correlcorrelated)ated) 0.5 0.5 0.5 ‘Low’ (positively correlated)

0.0 0.0 0.0 Zero crosscross aatt 1083108388 Zero cross at 1083810838 Zero cross at 10838 – –0.5 0.5 –0.5

– – – 1.0 ‘High’ (negatively correcorrelated)lated) 1.0 ‘High’ (negatively correlated) 1.0 ‘High’ (negatively correlated) 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 0 2,500 5000 7,500 10,000 12,500 15,000 17,500 Ranked list metricRanked (Signal2Noise) list Ranked list metricRanked (Signal2Noise) list Rank in ordered dataset Rank in ordered dataset Rank in ordered dataset Ranked list metric (Signal2Noise) Enrichment profile Enrichment profile Enrichment profile Hits Hits Hits Ranking metric scores Ranking metric scores Ranking metric scores (g) (h) (i)

Figure 10: GSEA results of low ARMCX2 expressed in GC patients, using gene set c5. The low ARMCX2 expression is involved in biological process of cell division (d), cell cycle (c, i), apoptosis (e), gene silencing (g), DNA damage checkpoint (f), and the NF-kappaB signaling pathway (a). Abbreviations: FDR: false discovery rate; GSEA: gene set enrichment analysis; KEGG: Kyoto Encyclopedia of Genes and Genomes; NES: normalized enrichment score; NOM: nominal.

tissues than in tumor tissues, leading to better OS in The results indicated that high ARMCX expression was an patients with gastric cancer, although the mechanism of independent risk factor as a prognostic characteristic for action needs further clarification. patients with gastric cancer, and the relevant risk score could In addition, a comprehensive survival analysis of the be used as a prognostic indicator. Nomogram, composed of current prognostic characteristics of ARMCX was performed risk score and other clinical information such as age and by establish a nomogram, and stratified joint-effects survival tumor stage, is an important prognostic risk assessment analysis was conducted to explore its potential application. system for gastric cancer. 14 BioMed Research International

To explore the underlying mechanism of ARMCX genes Ethical Approval in gastric cancer prognosis, we used a genome-wide RNA sequencing dataset in GSEA. The NF-kappaB, E2F and This article does not contain any studies with human partic- WNT signaling pathways, the cell cycle, and gene silencing ipants or animals performed by any of the authors. Since all were significantly enriched in the ARMCX1 and ARMCX2 datasets included in the present study were downloaded from low-expression groups. TCGA, additional approval by an Ethics Committee was not It is well established that, as members of the armadillo needed. The procedures were in accordance with the Helsinki (Arm) family, β-catenin and adenomatous polyposis coli declaration of 1964 and its later amendments. (APC) are important components of the WNT signaling pathway. Moreover, WNT signaling plays an important Conflicts of Interest role in a variety of biological processes, such as tumori- fl genesis, embryonic development, and stem cell mainte- The authors declare that they have no con icts of interest. nance [37–39]. β-Catenin, which is a multifunctional protein, plays an essential role in a variety of biological Authors’ Contributions responses. For instance, in the WNT signaling pathway, β-catenin works by interact with E-cadherin and TCF/LEF TW, YQ, and XL conceived and designed the study. ZH, transcription factors, respectively [40, 41]. APC can regu- WW, and ZL interpreted the data and performed the analy- fi late the WNT signaling pathway by synergistically acting sis. TW wrote the article, and MH approved the nal version fi with casein kinases 1, glycogen synthase kinase-3b, and of the manuscript. All authors approved the nal version of AXIN to induce degradation of β-catenin [37, 42, 43]. the manuscript. TingAn Wang and HuaGe Zhong contrib- On the basis of GSEA results, we deduced that both uted equally to this study. ARMCX1 and ARMCX2 were involved in the pathway and biological processes that are associated with the prog- Acknowledgments ress and treatment of gastric cancer and may serve as a GC prognostic marker. Once these results are verified, We sincerely appreciate TCGA (https://cancergenome.nih ARMCX1 and ARMCX2 may be used as biomarkers in .gov/) and UCSC Xena (http://xena.ucsc.edu/) for sharing combination with other clinical factors to facilitate the the GC data and for Dr Liucheng Wu for his advice with selection of diagnosis and treatment decisions for GC the language consulting of the manuscript. The study was and to benefit patients with better clinical outcomes. supported by the National Natural Science Foundation of Although significant results have been achieved in the GuangXi (No. 2019GXNSFBA185019) and National Natural current study, there are still some deficiencies to be con- Science Foundation of China (No. 81760521). sidered. First, the results of this study were obtained from a single cohort in the TCGA database, and its demo- Supplementary Materials graphic characteristics may not be representative of all patient groups. Therefore, genetic changes may have devi- Supplementary 1. Supplementary Table 1: GSEA KEGG ation and require further validation in other GC groups. (c2.all.v6.2.symbols.gmt) results of ARMCX1 and ARMCX2. Second, the clinical information from TCGA was incom- Supplementary 2. Supplementary Table 2: GSEA GO plete. Therefore, we cannot conduct a comprehensive (c5.all.v6.2.symbols.gmt) functional analysis results of stratification analysis in the Cox proportional risk regres- ARMCX1 and ARMCX2. sion model including all layers. Third, the mechanism between the above ARMCX and WNT signaling pathways References affecting the clinical prognosis of GC still needs to be determined. [1] E. Van Cutsem, X. Sagaert, B. Topal, K. Haustermans, and H. Prenen, “Gastric cancer,” The Lancet, vol. 388, no. 10060, pp. 2654–2664, 2016. 5. Conclusion [2] R. L. Siegel, K. D. Miller, and A. Jemal, “Cancer statistics, 2019,” CA: a Cancer Journal for Clinicians, vol. 69, no. 1, pp. 7–34, 2018. Our present study has determined that ARMCX has a “ potential prognostic value for gastric cancer and may have [3] B. Riggleman, E. Wieschaus, and P. Schedl, Molecular analysis of the armadillo locus: uniformly distributed transcripts clinical application value. In addition, further basic fi and a protein with novel internal repeats are associated with research is needed to clarify the speci c mechanisms of a Drosophila segment polarity gene,” Genes & Development, ARMCX in GC. vol. 3, no. 1, pp. 96–113, 1989. [4] G. López-Doménech, R. Serrat, S. Mirra et al., “The Eutherian Armcx genes regulate mitochondrial trafficking in neurons Data Availability and interact with Miro and Trak2,” Nature Communications, vol. 3, no. 1, p. 814, 2012. The datasets used and/or analyzed during the current study [5] M. Peifer, S. Berg, and A. B. Reynolds, “A repeating amino acid are available from the corresponding author on reasonable motif shared by proteins with diverse cellular roles,” Cell, request and TCGA data base. vol. 76, no. 5, pp. 789–791, 1994. BioMed Research International 15

[6] M. A. Andrade, C. Petosa, S. I. O’Donoghue, C. W. Müller, and [22] M. Hatzfeld, “The armadillo family of structural proteins,” P. Bork, “Comparison of ARM and HEAT protein repeats1,” International Review of Cytology, vol. 186, no. 1, pp. 179– Journal of Molecular Biology, vol. 309, no. 1, pp. 1–18, 2001. 224, 1999. [7] A. H. Huber, W. J. Nelson, and W. Weis, “Three-Dimensional [23] J. Anaya, “OncoLnc: linking TCGA survival data to mRNAs, Structure of the Armadillo Repeat Region of β-Catenin,” Cell, miRNAs, and lncRNAs,” PeerJ Computer Science, vol. 2, vol. 90, no. 5, pp. 871–882, 1997. no. 1, article e67, 2016. [8] H. J. Choi and W. I. Weis, “Structure of the armadillo repeat [24] L. J. Carithers, K. Ardlie, M. Barcus et al., “A novel approach to domain of plakophilin 1,” Journal of Molecular Biology, high-quality postmortem tissue procurement: The GTEx pro- vol. 346, no. 1, pp. 367–376, 2005. ject,” Biopreservation and Biobanking, vol. 13, no. 5, pp. 311– [9] H. Striegl, M. A. Andrade-Navarro, and U. Heinemann, 319, 2015. “Armadillo motifs involved in vesicular transport,” PLoS [25] Z. Tang, C. Li, B. Kang, G. Gao, C. Li, and Z. Zhang, “GEPIA: a One, vol. 5, no. 2, article e8991, 2010. web server for cancer and normal gene expression profiling ” [10] F. Parmeggiani, R. Pellarin, A. P. Larsen et al., “Designed and interactive analyses, Nucleic Acids Research, vol. 45, – Armadillo Repeat Proteins as General Peptide-Binding Scaf- no. W1, pp. W98 W102, 2017. folds: Consensus Design and Computational Optimization of [26] D. Warde-Farley, S. L. Donaldson, O. Comes et al., “The Gen- the Hydrophobic Core,” Journal of Molecular Biology, eMANIA prediction server: biological network integration for vol. 376, no. 5, pp. 1282–1304, 2008. gene prioritization and predicting gene function,” Nucleic – [11] I. V. Kurochkin, N. Yonemitsu, S. I. Funahashi, and Acids Research, vol. 38, suppl_2, pp. W214 W220, 2010. H. Nomura, “ALEX1, a Novel Human Armadillo Repeat Pro- [27] S. Mostafavi, D. Ray, D. Warde-Farley, C. Grouios, and tein That Is Expressed Differentially in Normal Tissues and Q. Morris, “GeneMANIA: a real-time multiple association Carcinomas,” Biochemical and Biophysical Research Commu- network integration algorithm for predicting gene function,” nications, vol. 280, no. 1, pp. 340–347, 2001. Genome Biology, vol. 9, Suppl 1, p. S4, 2008. [12] Z. Mou, A. R. Tapper, and P. D. Gardner, “The armadillo [28] V. P. Balachandran, M. Gonen, J. J. Smith, and R. P. DeMatteo, repeat-containing protein, ARMCX3, physically and function- “Nomograms in oncology: more than meets the eye,” The Lan- ally interacts with the developmental regulatory factor Sox10,” cet Oncology, vol. 16, no. 4, pp. e173–e180, 2015. The Journal of Biological Chemistry, vol. 284, no. 20, [29] A. Subramanian, P. Tamayo, V. K. Mootha et al., “Gene set – pp. 13629 13640, 2009. enrichment analysis: a knowledge-based approach for inter- [13] C. A. Smith, P. J. McClive, and A. H. Sinclair, “Temporal and preting genome-wide expression profiles,” Proceedings of the spatial expression profile of the novelarmadillo-related gen- National Academy of Sciences of the United States of America, e,Alex2, during testicular differentiation in the mouse vol. 102, no. 43, pp. 15545–15550, 2005. ” – embryo, Developmental Dynamics, vol. 233, no. 1, pp. 188 [30] V. K. Mootha, C. M. Lindgren, K. F. Eriksson et al., “PGC-1α- 193, 2005. responsive genes involved in oxidative phosphorylation are [14] F. Zeng, K. Liao, J. Wu et al., “ALEX1 may be a novel bio- coordinately downregulated in human diabetes,” Nature marker for human cervical squamous cell carcinoma,” Inter- Genetics, vol. 34, no. 3, pp. 267–273, 2003. national Journal of Clinical and Experimental Pathology, [31] A. Liberzon, C. Birger, H. Thorvaldsdóttir, M. Ghandi, J. P. – vol. 8, no. 8, pp. 9434 9439, 2015. Mesirov, and P. Tamayo, “The Molecular Signatures Database [15] H. Iseki, A. Takeda, T. Andoh et al., “ALEX1 suppresses colony Hallmark Gene Set Collection,” Cell Systems, vol. 1, no. 6, formation ability of human colorectal carcinoma cell lines,” pp. 417–425, 2015. – Cancer Science, vol. 103, no. 7, pp. 1267 1271, 2012. [32] A. Reiner, D. Yekutieli, and Y. Benjamini, “Identifying differ- [16] C. Montavon, B. S. Gloss, K. Warton et al., “Prognostic and entially expressed genes using false discovery rate controlling diagnostic significance of DNA methylation patterns in high procedures,” Bioinformatics, vol. 19, no. 3, pp. 368–375, 2003. ” grade serous ovarian cancer, Gynecologic Oncology, vol. 124, [33] Y. Benjamini and Y. Hochberg, “Controlling the false discov- – no. 3, pp. 582 588, 2012. ery rate: a practical and powerful approach to multiple test- [17] H. Iseki, A. Takeda, T. Andoh et al., “Human Arm protein lost ing,” Journal of the Royal Statistical Society: Series B in epithelial cancers, on chromosome X 1 (ALEX1) gene is (Methodological), vol. 57, no. 1, pp. 289–300, 1995. β transcriptionally regulated by CREB and Wnt/ -catenin sig- [34] B. Rubinfeld, I. Albert, E. Porfiri, C. Fiol, S. Munemitsu, and ” – naling, Cancer Science, vol. 101, no. 6, pp. 1361 1366, 2010. P. Polakis, “Binding of GSK3beta to the APC-beta-Catenin [18] E. Zeller, K. Mock, M. Horn, S. Colnot, M. Schwarz, and complex and regulation of complex assembly,” Science, A. Braeuning, “Dual-specificity phosphatases are targets of vol. 272, no. 5264, pp. 1023–1026, 1996. β β the Wnt/ -catenin pathway and candidate mediators of - [35] S. R. Krig, V. X. Jin, M. C. Bieda et al., “Identification of ” catenin/Ras signaling interactions, Biological Chemistry, genes directly regulated by the oncogene ZnF217 using – vol. 393, no. 10, pp. 1183 1191, 2012. chromatin immunoprecipitation (chip)-chip assays,” The [19] H. Yao, E. Ashihara, and T. Maekawa, “Targeting the Wnt/β- Journal of Biological Chemistry, vol. 282, no. 13, pp. 9703– catenin signaling pathway in human cancers,” Expert Opinion 9712, 2007. – on Therapeutic Targets, vol. 15, no. 7, pp. 873 887, 2011. [36] C. J. Edwards, J. L. Feldman, J. Beech et al., “Molecular Profile [20] J. R. Jessen, “Noncanonical Wnt signaling in tumor progres- of peripheral blood mononuclear cells from patients with sion and metastasis,” Zebrafish, vol. 6, no. 1, pp. 21–28, 2006. rheumatoid arthritis,” Molecular Medicine, vol. 13, no. 1-2, – [21] B. T. MacDonald, K. Tamai, and X. He, “Wnt/β-Catenin Sig- pp. 40 58, 2007. naling: Components, Mechanisms, and Diseases,” Develop- [37] F. H. Brembeck, M. Rosário, and W. Birchmeier, “Balancing mental Cell, vol. 17, no. 1, pp. 9–26, 2009. cell adhesion and Wnt signaling, the key role of β-catenin,” 16 BioMed Research International

Current Opinion in Genetics & Development, vol. 16, no. 1, pp. 51–59, 2006. [38] T. Reya and H. Clevers, “Wnt signalling in stem cells and cancer,” Nature, vol. 434, no. 7035, pp. 843–850, 2005. [39] M. Katoh, “WNT signaling pathway and stem cell signaling network,” Clinical Cancer Research, vol. 13, no. 14, pp. 4042– 4045, 2007. [40] A. E. Bass-Zubek, L. M. Godsel, M. DelSept, and K. J. Green, “Plakophilins: multifunctional scaﬀolds for adhesion and signaling,” Current Opinion in Cell Biology, vol. 21, no. 5, pp. 708–716, 2009. [41] W. Xu and D. Kimelman, “Mechanistic insights from structural studies of beta-catenin and its binding partners,” Journal of Cell Science, vol. 120, no. 19, pp. 3337–3344, 2007. [42] S. Munemitsu, I. Albert, B. Souza, B. Rubinfeld, and P. Polakis, “Regulation of intracellular beta-catenin levels by the adenomatous polyposis coli (APC) tumor-suppressor protein,” Pro- ceedings of the National Academy of Sciences of the United States of America, vol. 92, no. 7, pp. 3046–3050, 1995. [43] K. H. Goss and J. Groden, “Biology of the adenomatous polyposis coli tumor suppressor,” Journal of Clinical Oncology, vol. 18, no. 9, pp. 1967–1979, 2000.