Sage Journals: Discover world-class research

Abstract

BACKGROUND:

The prognosis of lung cancer patients is poor without useful prognostic and diagnostic biomarker. To search for novel prognostic and diagnostic markers, we previously found homeobox-A13 (HOXA13) as a promising candidate in lung cancer.

OBJECTIVE:

To determine the precisely clinical feature, prognostic and diagnostic value, possible role and mechanism of HOXA13.

METHODS:

Gene-expression was explored by real-time quantitative-PCR, western-blot and tissue-microarray. The associations were analyzed by Chi-square test, Kaplan-Meier and Cox-regression. The roles and mechanisms were evaluated by MTS, EdU, transwell, xenograft tumor and luciferase-reporter assays.

RESULTS:

HOXA13 expression is increased in tumors, and correlated with age of patients. HOXA13 expression is associated with unfavorable overall survival and relapse-free survival of patients in four cohorts. Interestingly, HOXA13 has different prognostic significance in adenocarcinoma (ADC) and squamous-cell carcinoma (SCC), and is a sex- and smoke-related prognostic factor only in ADC. Importantly, HOXA13 can serve as a diagnostic biomarker for lung cancer, especially for SCC. HOXA13 can promote cancer-cell proliferation, migration and invasion in vitro, and facilitate tumorigenicity and tumor metastasis in vivo. HOXA13 acts the oncogenic roles on tumor growth and metastasis by regulating P53 and Wnt/ $\beta$ -catenin signaling activities in lung cancer.

CONCLUSIONS:

HOXA13 is a new prognostic and diagnostic biomarker associated with P53 and Wnt/ $\beta$ -catenin signaling pathways.

Keywords

Overall survival diagnostic marker homeobox-A13 oncogene P53 signaling Wnt/β-catenin signaling non-small cell lung cancer

Abbreviations

NSCLC	Non-small cell lung cancer
ADC	Adenocarcinoma
SCC	Squamous cell carcinoma
OS	Overall survival
RFS	Relapse-free survival
HR	Hazard ratio
ROC	Receiver operating characteristic
TMA	Tissue microarray
TCGA	The Cancer Genome Atlas
EdU	5-ethynyl-2’-deoxyuridine
CTNNB1	$\beta$ -catenin
CCND1	CyclinD1

1. Introduction

Lung cancer is the most frequent malignant cancer in the worldwide [1]. Non-small cell lung cancer (NSCLC) is the common type of lung cancer, accounting for more than 80% of total lung cancer. The main types of NSCLC are lung squamous cell carcinoma (SCC) and lung adenocarcinoma (ADC) [2, 3, 4]. The prognosis of patients with NSCLC is very poor and the rate of 5-year survival is less than 15% [5, 6]. This poor survival of NSCLC patients is mainly due to lack of useful prognostic and early diagnostic marker. Therefore, identification of new, useful prognostic and diagnostic biomarker is crucial for early treatment and prolonging the survival of NSCLC patients.

Homeobox (HOX) family genes, as the nuclear transcription factors, play important roles in embryogenesis and cellular differentiation [7]. HOX family was divided into four HOX gene subgroups, which are HOXA, HOXB, HOXC and HOXD clusters. Previous study has revealed that the position of HOX genes on chromosome is closely associated with site and time of their expression, the genes at 3’ end of the subgroups (for example HOXD1 and HOXA1) are usually expressed early in proximal and anterior regions, however the genes at 5’ end (for example HOXD13 and HOXA13) are generally expressed later in more distal and posterior regions [8]. Increasing evidences have indicated that the expression of HOX genes is usually abnormal in many tumors [9, 10, 11, 12, 13, 14].

HOXA13 (homeobox-A13), as the most distal member of HOXA cluster in HOX family, acts key roles in distal limb and genitourinary development [15]. HOXA13 has been found expressed highly in the hindgut region, but extremely low or absent in anterior areas in human adult [16]. HOXA13 may involve in the carcinogenesis of esophageal carcinoma and hepatocellular carcinoma [17, 18, 19]. In the recent studies, HOXA13 is associated with cancer progression and may be a predictive factor for outcome in hepatocellular cancer, gastric cancer, prostate cancer, cervical cancer, ovarian cancer and glioma [20, 21, 22, 23, 24, 25]. However, the possible role, potential mechanism, prognostic and diagnostic significance of HOXA13 are largely unexplored in human cancers, including lung cancer.

In our present study, HOXA13 was identified as a preferential high expression molecule in tumor tissues by RT-PCR, qRT-PCR and immunohistochemistry in lung cancer. Then the clinical relevance, prognostic significance, diagnostic value, functional role and possible mechanism of HOXA13 were further explored. Our study reveals that HOXA13 is an important prognostic and diagnostic biomarker, and acts as an important oncogene by transcriptional regulating P53 and Wnt/ $\beta$ -catenin signaling pathways in lung cancer.

2. Materials and methods

2.1 Patient samples and cell lines

A total of 130 lung cancer patients (who undergone surgical resection from 2008 to 2010) were acquired from Southwest Hospital in Chongqing, China. The clinico-pathologic data including gender, age, tumor size, histological grade, lymph node status, tumor location, overall survival (OS) and clinical stage was obtained. The present study was approved by the ethics committee from Southwest Hospital which affiliated to Army Medical University. The experiments were performed in accordance with approved guidelines of Southwest Hospital and Army Medical University. The informed consents were all signed by the patients.

The cell lines LTEP-a-2 (LTEP) and SPC-A-1 obtained from the Chinese Academy of Sciences Cell Bank of Type Culture Collection (Shanghai, China) were cultured in RPMI-1640 medium (HyClone, USA) which was supplemented with 10% FBS (fetal bovine serum, HyClone).

2.2 RNA isolation

The total RNA was extracted by Trizol reagent (Invitrogen, Carlsbad, CA) from the frozen tissues according to the manufacturer’s instruction. The RNA (5.0 $\mu$ g) was treated using DNase I to eliminate genomic DNA contamination. Then, the treated RNA was reverse transcribed into cDNA and stored at $-$ 80 ${}^{\circ}$ C.

2.3 RT-PCR and qRT-PCR

The expression of HOXA13 was evaluated by RT-PCR and quantitative real-time PCR (qRT-PCR). A series of different PCR cycles were performed to determine the appropriate cycles. The products of PCR were subjected to electrophoresis in 3.0% agarose gel. An endogenous control of $\beta$ -actin was used to amplify. The qRT-PCR was performed with GoTaq ${}^{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}$ qPCR Master Mix (Promega, USA) by iQ5 real-time detection system (Bio-Rad Laboratories, CA, USA). All PCR was carried out in triplicate. The primers used are listed in Supplemental Table S1.

2.4 Tissue microarray (TMA)

In order to construct lung cancer TMA slides, a core was taken from tumor and adjacent non-tumor tissue within a distance of approximately 30 mm, respectively. The adjacent noncancerous tissues were stained with H&E (hematoxylin-eosin) and evaluated by comparing with the normal tissues. Then, the TMA slides containing lung cancer were constructed as described previously [26].

2.5 Immunohistochemical and quantified analyses of HOXA13

Immunohistochemical (IHC) analysis was performed to evaluate HOXA13 expression in patient samples using HOXA13 antibody with 1:200 (Abcam, ab106503), P53 antibody with 1:100 (Santa Cruz, sc-71820), P21 antibody with 1:100 (Santa Cruz, sc-6246), BAX antibody with 1:100 (Santa Cruz, sc-20067), $\beta$ -catenin (CTNNB1) antibody with 1:100 (Santa Cruz, sc-7963) and MYC antibody with 1:150 (Cell Signaling Technology, MA, USA, #13987) as described previously [27]. The staining of HOXA13 was quantified based on positive staining and intensity of staining. The positive cells of staining were classified into 5 grades ( $<$ 5% positive cells score for 0; 5% to 25% positive cells score for 1; 26% to 50% positive cells score for 2; 51% to 75% positive cells score for 3 and $\geqslant$ 76% positive cells score for 4). The intensity of staining was graded into 4 groups (negative score as 0; weak score as 1; moderate score as 2 and strong score as 3). The product of the score for positive staining and the score for intensity of staining was used to define the expression levels of HOXA13 in patient samples.

2.6 Construction of HOXA13 expression vector

Full-length of HOXA13 gene was amplified using genomic DNA extracted from human normal blood by PCR, verified by sequencing and then cloned to pIRES2-EGFP vector (stored by Invitrogen, USA). The vector was transfected into cancer cells using the transfection reagent of lipofectamine 2000 (Invitrogen, USA).

2.7 Construction of stable transfected cells

The transfection was performed and the stable transfected cells were selected as described in previous study [28].

2.8 HOXA13 siRNA assay

The HOXA13 siRNA of this study was purchased from Santa Cruz Biotechnology (sc-45666), and was transfected into cancer cells according to the manufacturer’s instruction.

2.9 Western blot analyses

Proteins of transfected cancer cells were run on 12% SDS-PAGE which were transferred into the PVDF membrane (Millipore, USA), and then incubated in the primary antibody (HOXA13, 1:1000, Abcam, ab106503; $\beta$ -actin, 1:1500, Sigma). The target protein was detected using chemiluminescence (Pierce, USA).

2.10 MTS assay

The cancer cells transfected with HOXA13 expressing vector were plated at 3, 500 cells per well in 96-well plate. Cell viability was determined via MTS (Promega, USA) at 1, 2, 3, 4 and 5 days posttransfection as described previously [28]. These experiments were done in triplicate.

2.11 5-ethynyl-2’-deoxyuridine (EdU) assay

EdU assay was carried out as described previously [28]. Briefly, cancer cells were cultured in 96-well plates at 4,000 cells per well and transfected with pIRES2-EGFP-HOXA13 or empty vectors. After 46 h of transfection, cancer cells were incubated in EdU for about 2 h, and fixed with 4% formaldehyde for 0.5 h. The fixed cells were incubated with glycine for 5 min, and then treated for 10 min with the Triton X-100. The cancer cells were thereafter incubated for 30 min in the Apollo reaction cocktail after phosphate-buffered saline washing, and then treated with the Triton X-100 twice. DNA was stained in Hoechst 33342 stain for 30 min and was observed with fluorescence microscopy.

2.12 Migration and invasion assay

Transwell assays were carried out in 24-well plates without or with the matrigel (8 $\mu$ m, Corning) [29, 30]. Briefly, cancer cells transfected with pIRES2-EGFP-HOXA13 or empty vectors were suspended with serum-free medium at 3 $\times$ 10 ${}^{4}$ cells in the upper well of chamber. The lower well of the chamber contained the media which is supplemented with full (10%) fatal bovine serum. The cancer cells which migrated to the lower well of the chamber were stained by 0.1% crystal violet after 18 h and were then counted. All the experiments were done in triplicate.

2.13 Clinical microarray database analyses

The Kaplan-Meier Plotter (http://kmplot.com/analysis/index.php), which contains gene expression data and the information of survival for lung cancer patients, was used to analyze in this study [31]. The probe used was 231786_at. The lung cancer patients in the database were grouped based on the best cut-off using auto selection. When $p$ value is less than 0.05, it was considered as the statistically significant.

2.14 ROC (receiver operating characteristic) curve analyses

The diagnostic values of HOXA13 expression were evaluated in NSCLC, ADC and SCC patients using the expression data of tumor and normal tissues from TCGA database. The x and y axes in the ROC curve are the specificity and the sensitivity, respectively. The area under curve (AUC) in the ROC curve was calculated to evaluate the ability of HOXA13 expression in predicting the diagnostic outcome of the patients.

2.15 Tumorigenicity assay in nude mice

The LTEP stably cells with control and HOXA13 vector were suspended in PBS and were subcutaneously injected into the flank of male Balb/c nude mice (6-week-old) using 6 $\times$ 10 ${}^{6}$ cells/mouse for the tumorigenicity assay. The tumor diameter of these nude mice was measured weekly and the nude mice were killed at 6 weeks after the cell injection. The volume of the tumors was calculated as previous study [28]. The mouse experiment was approved by the Institutional Animal Care and Use Committee of Army Medical University, China.

2.16 Experimental metastasis assay in nude mice

The LTEP stably cells with control and HOXA13 vector were suspended in PBS and were injected into tail vein of male Balb/c nude mice (6-week-old) using 2 $\times$ 10 ${}^{6}$ cells/mouse for the metastasis assay. These mice were killed at 8 weeks after the cell injection. The metastasis of tumor cells into lung and liver was pathologically observed and quantitatively measured. The mouse experiment was approved by the Institutional Animal Care and Use Committee of Army Medical University, China.

2.17 P53 and Wnt/ $\beta$ -catenin signaling activity and luciferase reporter assays

LTEP and SPC-A-1 cells were co-transfected with TOP/FOP-Flash plasmids or pp53-TA-Luc plasmid with pIRES2-EGFP-HOXA13 and empty vectors, respectively. The plasmids of TOP/FOP-Flash were previously purchased from the Addgene (No. 12456 and 12457, Cambridge, USA), and the plasmids of pp53-TA-Luc was purchased from Beyotime Biotechnology (D2223, Shanghai, China). The Varioskan-LUX (a fluorescence microplate reader measurement system, Thermo Fisher, MA, USA) and the Dual-luciferase reporter kit (Promega) were used to analyze the luciferase reporter assays as described previously [28]. The activities of luciferase were measured at 32 h post transfection, and the experiment was performed thrice.

2.18 Statistical analyses

Statistical analyses were carried out via SPSS 13.0 software (SPSS, Chicago, IL). The differences of categorical variables were analyzed via Chi-square test. The OS was calculated and evaluated using Kaplan-Meier with log-rank test. Multivariate analysis of prognostic predictors was performed using Cox-regression. The high and low expression groups were categorized base on the average score. When $p$ value is less than 0.05, it was considered as the statistically significant.

Figure 1.

HOXA13 expression is increased in tumor tissue of lung cancer. (A) HOXA13 mRNA expression was determined by RT-PCR and qRT-PCR analyses in normal and tumor tissues. The “N” represents normal lung tissue. The “M” represents DNA marker. The “T” represents tumor tissue. $\beta$ -actin is the internal control. The error-bar indicates Standard Error of Mean (SEM). The “**” represents $p$ value $<$ 0.01. (B and C) HOXA13 protein expression was analyzed by immunohistochemistry (IHC) in 130 (seven patients were slided off) cancerous and adjacent noncancerous samples. HOXA13 expression was higher in the cancerous tissues than the adjacent noncancerous tissues. The scale bar represents 50 $\mu$ m. The “H&E” represents hematoxylin-eosin staining. The “***” represents $p$ value $<$ 0.0001. (D) HOXA13 expression was analyzed in normal and tumor tissues of NSCLC from the TCGA database (https://genome-cancer.ucsc.edu/proj/site/hgHeatmap). The meta-analysis of the expression of HOXA13 in TCGA was performed by RNAseq (IlluminaHiSeq, $N=$ 553).

3. Results

3.1 HOXA13 expression is sharply increased in NSCLC tissues

To investigate HOXA13 expression in tissues of NSCLC, HOXA13 expression was tested in normal lung and NSCLC tissues by RT-PCR and qRT-PCR. The expression of HOXA13 was elevated in NSCLC tissues (3.86 $\pm$ 0.32) compared to normal lung tissues (2.10 $\pm$ 0.18) (Fig. 1A). To confirm this altered expression at protein level, the immunohistochemistry (IHC) for HOXA13 was performed using tissue microarray (TMA) with 130 NSCLC and the corresponding tumor-adjacent normal tissues. HOXA13 protein was significantly increased in tumor tissues when compared to tumor-adjacent normal tissues (Fig. 1B). Thereafter, HOXA13 protein was quantified with intensity and positive staining percentage and found that HOXA13 expression was 7.28 $\pm$ 1.40 in tumor tissues much higher than 2.06 $\pm$ 1.52 in tumor-adjacent normal tissues (Fig. 1C). The result was further confirmed in patients of lung cancer using database of The Cancer Genome Atlas (TCGA) (Fig. 1D). These data reveal that HOXA13 expression is sharply increased in NSCLC tissues compared to normal lung and tumor-adjacent normal tissues both at mRNA and protein levels.

3.2 HOXA13 expression is associated with age of NSCLC patients

Two groups of the patients were classified according to the quantified expression of HOXA13, the high group (score greater than average expression) and the low group (score equal to and less than average expression) groups. Then the associations between HOXA13 expression and the clinico-pathologic features of the NSCLC patients were investigated, and the expression of HOXA13 was found to be closely correlated with age ( $n=$ 123, $p<$ 0.001) of these patients (Table 1). Yet the expression of HOXA13 was not correlated with gender ( $p=$ 0.712), histological grade ( $p=$ 0.054), size of tumor ( $p=$ 0.714), location of tumor ( $p=$ 0.853), lymph node ( $p=$ 0.194) and the clinical stage ( $p=$ 0.131) of NSCLC patients (Table 1).

Table 1
Correlations of HOXA13 expression with clinicopathologic features in human patients ( $n=$ 123)

Clinical feature	Total	High	Low	$P$ value
		HOXA13 expression
Age (years)
$<$ 60	63	29	34	$<$ 0.001
$\geqslant$ 60	58	45	13
Gender
Male	65	41	24	0.712
Female	58	34	24
Histological grade
1	22	14	8	0.054
2	64	45	19
3	35	16	19
Tumor size
$<$ 4	61	36	25	0.714
$\geqslant$ 4	62	39	23
Tumor location
Left	59	35	24	0.853
Right	64	40	24
Lymph node
Negative	67	37	30	0.194
Positive	56	38	18
Clinical stage
I $+$ II	76	41	35	0.131
III $+$ IV	31	22	9

The age of two patients and the clinical stage of seventeen patients are missing.

3.3 High expression of HOXA13 is associated with unfavorable OS and RFS of NSCLC patients

To determine the relationships between HOXA13 expression and survival prognosis of NSCLC patients, we analyzed the patients’ survival data. The results of Kaplan-Meier analyses revealed that the patients with high HOXA13 expression owned a poor OS compared to the patients with low HOXA13 expression ( $p<$ 0.001, Fig. 2A and B). In order to rule out other interfering factors, HOXA13 expression as well as clinical characteristics (including gender, histological grade, age, tumor location, tumor size, clinical stage and lymph node) and the survival data were analyzed using Cox-regression method. High expression of HOXA13 is an independent and unfavourable prognostic factor (HR [hazard ratio] $=$ 1.233, $p<$ 0.001) in addition to clinical stage (HR $=$ 2.060, $p<$ 0.001) for OS of these patients (Fig. 2B, Table 2). To further validate this correlation, the contributions of HOXA13 expression to the survival of NSCLC patients were evaluated using a database of clinical microarray (http://kmplot.com/analysis/index) [31]. The analysis results of the database indicated that high HOXA13 expression indeed predicts unfavourable OS as an independent factor for the patients (univariate: $p<$ 0.001/multivariate: $p=$ 0.008, Fig. 2C). Besides, high HOXA13 expression was also found to predict poor relapse-free survival (RFS) of patients in two independent cohorts (cohort-1, $p=$ 0.026; cohort-2, $p=$ 0.009, Fig. 2D). These findings demonstrate that high HOXA13 expression is a harmful and independent prognostic factor for NSCLC patients.

Table 2
Multivariate analysis of different prognostic factors in human lung cancer patients ( $n=$ 123)

Expression level	Variable	HR	95% CI	$P$ value
HOXA13 expression of	Age	1.027	0.998–1.057	0.071
lung cancer patients	Gender	0.735	0.392–1.380	0.338
	Histological grade	0.641	0.370–1.110	0.113
	Tumor size	0.909	0.685–1.207	0.511
	Tumor location	1.025	0.941–1.117	0.572
	Lymph node	1.331	0.714–2.479	0.368
	Clinical stage	2.060	1.385–3.064	$<$ 0.001
	HOXA13 expression	1.233	1.101–1.380	$<$ 0.001

Abbreviations: HR, hazard ratio; CI, confidence interval.

Figure 2.

High expression of HOXA13 is associated with adverse OS of the patients with lung cancer. (A) The high expression and low expression groups of HOXA13 was shown in human NSCLC tissues. Scale bars represent 50 $\mu$ m. (B) Kaplan-Meier and Cox-regression survival analysis of HOXA13 expression in 123 NSCLC patients. HOXA13 is associated with adverse OS and acts as an independent prognostic marker of the patients. The “HR” represents hazard ratio. (C) Kaplan-Meier and Cox-regression survival analyses between HOXA13 expression and OS of 1926 lung cancer patients in database of Kaplan-Meier Plotter (http://kmplot.com/analysis/index). The OS of the patients with high HOXA13 expression was shorter than that of the patients with low HOXA13 expression. Auto best cutoff was chosen; cutoff value was 20, and the expression range was 0–585. (D) Kaplan-Meier survival analyses between HOXA13 expression and RFS of NSCLC patients in two independent cohorts (GSE8894, Seoul [1995–2005], Lee, $n=$ 138; GSE31210, NCCRI, Okayama, $n=$ 204).

3.4 High expression of HOXA13 is associated with adverse OS in ADC patients but with favorable OS in SCC patients

The correlations were then investigated between HOXA13 expression and the survival in the patients with ADC and SCC, respectively. The ADC patients with high HOXA13 expression owned a shortened OS when compared to the patients with low HOXA13 expression (HR $=$ 1.62, $p=$ 0.0001, Fig. 3A). However, the SCC patients with high HOXA13 expression owned a prolonged OS when compared to the patients with low HOXA13 expression (HR $=$ 0.067, $p=$ 0.013, Fig. 3A). These data indicate that HOXA13 is an adverse prognostic factor in the patients with ADC, but a favorable prognostic factor in the SCC patients.

Figure 3.

HOXA13 is a sex- and smoke-related prognostic factor and potential diagnostic biomarker. (A) High expression of HOXA13 is associated with unfavorable OS in the patients with ADC, but with better OS in the patients with SCC. Kaplan-Meier survival analysis was performed between HOXA13 expression and the OS of 673 ADC patients and 271 SCC patients from database of Kaplan-Meier Plotter. Auto best cutoff was chosen. (B) HOXA13 expression is an unfavorable prognostic biomarker of the male ADC patients, but a favorable prognostic biomarker of the female ADC patients. Kaplan-Meier analysis was performed between HOXA13 expression and the OS of 328 male and 287 female patients with ADC in the database of Kaplan-Meier Plotter. Auto best cutoff was used. (C) HOXA13 is a smoke-related prognostic biomarker of the ADC patients. Kaplan-Meier analysis was performed between HOXA13 expression and OS of 140 non-smoked and 231 smoked ADC patients in the database of Kaplan-Meier Plotter. Auto best cutoff was used. (D) ROC curves and AUC analysis were performed to evaluate the diagnostic values of HOXA13 expression in NSCLC, ADC and SCC patients. The gray and red lines represent the random performance and the ROC curve, respectively.

Figure 4.

HOXA13 accelerates cancer cell growth and proliferation. (A) Cancer cell transfectants of HOXA13 expressing vector and empty vector (mock) control were identified in SPC-A-1 and LTEP cells by western blot. Strong expression of HOXA13 was detected after transfection of HOXA13, whereas not detected after transfection of empty vector. (B) MTS assay was carried out to examine the role of HOXA13 on cancer cell growth and proliferation in SPC-A-1 and LTEP cells. Cell viability was determined by the CellTiter 96 AQueous One Solution Cell-Proliferation-Assay (Promega) in triplicate. The “*” represents $p$ value $<$ 0.05. The “**” represents $p$ value $<$ 0.01. (C) EdU assay was performed to examine the role of HOXA13 on cancer cell proliferation in SPC-A-1 and LTEP cells. The “EdU” represents 5-ethynyl-2’-deoxyuridine. The “**” represents $p$ value $<$ 0.01. (D) The knockdown of HOXA13 was confirmed by WB and the effects of HOXA13 knockdown on cancer cell growth and proliferation were examined by MTS assays in stable HOXA13 expressing LTEP cells. Cell viability was determined by the CellTiter 96 AQueous One Solution Cell-Proliferation-Assay (Promega) in triplicate. The “*” represents $p$ value $<$ 0.05. The “**” represents $p$ value $<$ 0.01.

Figure 5.

HOXA13 stimulates cancer cell migration and invasion. (A) The cancer cell migration was determined by transwell assays without matrigel after transfection of HOXA13 expressing vector or empty vector in SPC-A-1 and LTEP cells. The cancer cells in lower well of chamber were fixed, stained and then quantified. The migration cells were measured with the average count of at least five random microscopic fields. Error bar indicates SEM. The “**” represents $p$ value $<$ 0.01. (B) The invasion of cancer cells was determined by transwell assays with matrigel after transfection of HOXA13 or empty vector in SPC-A-1 and LTEP cells. The cancer cells in lower well of chamber were also fixed, stained and then quantified. The invasion cells were measured with the average count of at least five random microscopic fields. Error bar indicates SEM. The “**” represents $p$ value $<$ 0.01. (C) The cancer cell migration was determined by transwell assays without matrigel after knockdown of HOXA13 by siRNA in stable HOXA13 expressing LTEP cells. The cancer cells in lower well of chamber were fixed, stained and then quantified. The migration cells were measured with the average count of at least five random microscopic fields. Error bar indicates SEM. The “**” represents $p$ value $<$ 0.01.

Figure 6.

HOXA13 promotes tumor growth and metastasis in vivo. (A) The tumor growth curve and tumor weight were compared between the nude mice with vector control and the nude mice with HOXA13. The result was got from three independent experiments ( $n=$ 6). The “*” represents $p$ value $<$ 0.05. The “**” represents $p$ value $<$ 0.01. The “***” represents $p$ value $<$ 0.001. (B) Tumor cells that metastasize into lung of nude mice were observed and determined by H&E (haematoxylin and eosin) staining and quantitative analysis. Arrow indicates the metastasis loci of tumor cells in lung. Scale bars indicate 50 $\mu$ m. The “***” represents $p$ value $<$ 0.001. (C) Tumor cells that metastasize into liver of nude mice were observed and determined by H&E staining and quantitative analysis. Arrow indicates the metastasis loci of tumor cells in liver. Scale bars indicate 50 $\mu$ m. The “***” represents $p$ value $<$ 0.001.

Figure 7.

The roles of HOXA13 on proliferation and metastasis are associated with P53 and Wnt/ $\beta$ -catenin signaling-pathways. (A) The mRNA expression of P53 signaling-pathway and Wnt/ $\beta$ -catenin signaling-pathway related genes (P53, P21, PMAIP1, BAX, CTNNB1, MYC, CCND1 and MMP7) were determined in HOXA13 transfected LTEP cells by qRT-PCR analyses. $\beta$ -actin was the internal control. Error bar indicates SEM. The “**” represents $p$ value $<$ 0.01. (B) The protein expression of HOXA13, P53 signaling-pathway and Wnt/ $\beta$ -catenin signaling-pathway related genes (P53, P21, BAX, CTNNB1 and MYC) were analyzed by immunohistochemistry (IHC) in lung cancer samples. Two representative images of high or low expression for each indicator were shown. Scale bars represent 50 $\mu$ m.

Figure 8.

HOXA13 acts as an oncogene by regulating the activities of P53 and Wnt/ $\beta$ -catenin signaling-pathways. (A) The activity of P53 response reporter construct pp53-TA-Luc was examined to determine the effects of HOXA13 on the P53 signaling-pathway in SPC-A-1 and LTEP cells. The pp53-TA-Luc plasmid was co-transfected with PRL-TK and HOXA13 or empty vector plasmids into the cancer cells to evaluate the transcriptional activity of P53 signaling-pathway. The “**” represents $p$ value $<$ 0.01. Error bar indicates SEM. (B) The activity of Wnt/ $\beta$ -catenin response reporter constructs TOP-flash (contains wild-type binding sites) and FOP-flash (contains the mutant binding sites) was examined to determine the effects of HOXA13 on the Wnt/ $\beta$ -catenin signaling-pathway in SPC-A-1 and LTEP cells. The plasmids were co-transfected with PRL-TK and HOXA13 or empty vector plasmids into the cancer cells to evaluate the Wnt/ $\beta$ -catenin signaling transcriptional activity. The “**” represents $p$ value $<$ 0.01. Error bar indicates SEM. (C) The schematic diagram of mechanism for HOXA13 acting as a prognostic biomarker and an oncogene in ADC. This schematic diagram is derived according to the data of qRT-PCR, IHC and luciferase reporter assays.

3.5 The expression of HOXA13 is a sex-related prognostic factor for ADC patients

The correlation of HOXA13 expression and the OS of different sex ADC patients was further determined. High HOXA13 expression was prominently associated with poor OS of the male patients with ADC (HR $=$ 1.53, $p=$ 0.014, Fig. 3B). Whereas high expression of HOXA13 was correlated with better OS of the female patients with ADC (HR $=$ 0.64, $p=$ 0.036, Fig. 3B). However, the correlations of HOXA13 expression and the OS of different sex patients have no difference in SCC. These data show that HOXA13 expression is an unfavorable prognostic marker in the male ADC patients, but is a favorable prognostic biomarker in the female ADC patients.

3.6 HOXA13 expression is a smoke-related prognostic factor for ADC patients

Next, we ascertain whether HOXA13 expression predicts unfavorable OS in the ADC patients related to smoking. The analysis results found that high expression of HOXA13 was closely associated with unfavorable OS of the patients with smoking (HR $=$ 1.83, $p=$ 0.017, Fig. 3C), whereas HOXA13 expression was not correlated with OS of the ADC patients without smoking (HR $=$ 0.46, $p=$ 0.058, Fig. 3C). These data reveal that HOXA13 is a smoke-related prognostic biomarker of ADC patients indicating smoking is an unfavorable factor for ADC patients with high HOXA13 expression.

3.7 HOXA13 is a promising diagnostic biomarker for NSCLC patients

HOXA13 expression was significantly increased in tumor tissues compared to normal and corresponding tumor-adjacent normal tissues. To explore whether this change of HOXA13 expression can serve as a potential diagnostic biomarker for lung cancer, we performed ROC curves and AUC analyses to evaluate the diagnostic values of HOXA13 expression in NSCLC, ADC and SCC patients. HOXA13 expression exhibited a high sensitivity and specificity with AUC being 79.6%, 72.1% and 92.8% in NSCLC, ADC and SCC patients, respectively (Fig. 3D). Moreover, HOXA13 could exhibit a higher diagnostic efficacy in SCC patients than that in NSCLC and ADC patients (Fig. 3D). These data reveal that HOXA13 can serve as a promising diagnostic biomarker for NSCLC patients, especially for SCC patients.

3.8 HOXA13 markedly stimulates cancer cell growth and proliferation in vitro

To determine possible roles of HOXA13 in NSCLC, we performed MTS and EdU assays to determine the cell growth and proliferation in SPC-A-1 and LTEP cell lines (two human NSCLC cell lines lowly expressed HOXA13) following over-expression of HOXA13. Ectopic expression of HOXA13 could significantly promote cancer cell growth and proliferation when compared to the empty vector control in both SPC-A-1 cells and LTEP cells (Fig. 4A–C). Moreover, knockdown of HOXA13 could inhibit cancer cell growth and proliferation when compared to the negative control in stable HOXA13 expressing LTEP cells (Fig. 4D). These results indicate that HOXA13 remarkably activates cancer cell growth and proliferation in NSCLC.

3.9 HOXA13 significantly promotes migration and invasion of cancer cells in vitro

To investigate the roles of HOXA13 on migration and invasion of cancer cells in NSCLC, we performed transwell assays without and with matrigel, respectively. The results of transwell assays without matrigel revealed that there was a significant rise of cancer cell migration in HOXA13-transfected cells when compared to the empty vector control in SPC-A-1 and LTEP cells (Fig. 5A and B). The data of transwell assays with matrigel showed that the invasion of cancer cells was also significantly increased in HOXA13-transfected cells when compared to the empty vector control in SPC-A-1 and LTEP cells. Furthermore, the migration of cancer cells was indeed decreased in HOXA13 knockdown cells when compared to the negative control in stable HOXA13 expressing LTEP cells (Fig. 5C). These results demonstrate that HOXA13 markedly accelerates cancer cell migration and invasion in NSCLC.

3.10 HOXA13 strongly accelerates tumor growth and metastasis in nude mice

After determining the roles of HOXA13 in vitro, we want to investigate whether the HOXA13 still plays its mentioned roles in vivo. The xenograft tumor model was used to assess the growth and metastasis of stable tumor cells containing empty and HOXA13 vectors. The data of xenograft tumor model demonstrated that the volume and weight of tumor was much larger in the nude mice receiving stable expressing HOXA13 cells than that of receiving empty vector cells by measuring and weighing (Fig. 6A), and more lung and liver metastatic tumors were found in the nude mice receiving stable expressing HOXA13 cells than that receiving stable empty vector cells by HE staining and quantitative analyses (Fig. 6B and C). These experimental results reveal that HOXA13 strongly accelerates tumor growth and metastasis in vivo.

3.11 HOXA13 plays the oncogenic roles by regulating P53 and Wnt/ $\beta$ -catenin signaling-pathways

To gain insights into the mechanism of oncogenic effect of HOXA13, the possible downstream targets were screened by qRT-PCR in LTEP cells transfected with HOXA13, and were confirmed in patient samples by IHC. The expressions of P53 and its targets P21, PMAIP1 and BAX (the P53 signaling pathway) were reduced, whereas the expressions of $\beta$ -catenin (CTNNB1) and its targets MYC, CyclinD1 (CCND1) and MMP7 (the Wnt/ $\beta$ -catenin signaling pathway) were increased (Fig. 7A). The changes of P53 and Wnt/ $\beta$ -catenin signaling pathways affected by HOXA13 were verified in the patient samples by IHC (Fig. 7B). These data suggest that HOXA13 activates cancer cell proliferation by down-regulating the activity of P53 signaling-pathway, and stimulates cancer cell metastasis by up-regulating the activity of Wnt/ $\beta$ -catenin signaling-pathway.

To further verify HOXA13 represses P53 signaling activity and activates Wnt/ $\beta$ -catenin signaling activity, we used P53 signaling response reporter vector pp53-TA-Luc and Wnt/ $\beta$ -catenin signaling response reporter vector TOP-Flash/FOP-Flash to monitor the effects of HOXA13 on P53 and Wnt/ $\beta$ -catenin signaling pathways. The experimental results indicated that HOXA13 could significantly inhibit the transcriptional activity of P53 signaling-pathway, whereas could strongly activate the transcriptional activity of Wnt/ $\beta$ -catenin signaling-pathway in SPC-A-1 and LTEP cells (Fig. 8A and B). These data further show that HOXA13 indeed regulates the activities of P53 and Wnt/ $\beta$ -catenin signaling-pathways in NSCLC.

4. Discussion

In our present study, we determined the precisely clinical relevance, prognostic and diagnostic significance of HOXA13 expression in patients with NSCLC for the first time. The expression of HOXA13 was much higher in the tumor tissues than that in the non-tumor tissues, and this expression was closely associated with age of the patients. The high HOXA13 expression patients had a short OS and RFS when compared to the low HOXA13 expression patients, and HOXA13 expression was an independent prognostic marker for the patients with NSCLC. Moreover, HOXA13 is a potential and promising diagnostic biomarker for NSCLC patients, especially for SCC patients. These data suggest that the expression of HOXA13 is an important predictor for prognosis and diagnosis of patients in NSCLC.

Previous evidences have indicated that the expression of HOXA13 is higher in tumor tissues than in tumor-adjacent normal tissues and normal tissues in esophageal squamous cell cancer, hepatocellular cancer, prostate cancer, gastric cancer and glioma [17, 19, 20, 22, 25]. In our study, we found that the expression of HOXA13 was strongly increased in tumor tissues of NSCLC, which is consistent with the previous researches [32]. This change of HOXA13 expression suggests that HOXA13 is probably associated with tumorigenesis in NSCLC. HOXA13, as a nuclear transcription factor, is expected to be usually expressed in the nuclei of cells [17, 33]. However, some studies have revealed that HOXA13 is expressed in the cytoplasm of tumor cells [19, 21]. In our study, we showed that HOXA13 was expressed both in the cytoplasm and the nuclei of tumor cells of NSCLC samples (Figs 1B, 2A, 7B), which may be associated with its interaction with a cytoplasmic protein in these samples.

HOXA13 has been reported as a potential biomarker for predicting outcome in esophageal squamous cell cancer, hepatocellular cancer, gastric cancer, prostate cancer and glioma [19, 20, 21, 22, 25]. In the present study, the patients with high HOXA13 expression owned a poor OS compared to the patients with low HOXA13 expression in NSCLC. Further analyses revealed that high expression of HOXA13 is associated with adverse OS in patients with ADC, but with better OS in SCC patients, indicating different prognostic significance of HOXA13 in the ADC and SCC patients, respectively. HOXA13 expression is an adverse prognostic biomarker of the male ADC patients, but is a favorable prognostic biomarker of the female ADC patients, implying a sex-specific prognostic factor in ADC patients. Moreover, high expression of HOXA13 is closely associated with poor OS of the ADC patients with smoking, whereas the expression of HOXA13 is not correlated with OS of ADC patients without smoking, which suggests that HOXA13 is a smoke-related prognostic biomarker in ADC patients. The results suggest that HOXA13 is an important potential and unique biomarker for predicting outcome of NSCLC patients. In addition, HOXA13 expression is a sex- and smoke-related prognostic factor of ADC patients, which suggests that HOXA13 expression may be associated with sex and smoking of the patients. However, the specific correlations and roles between them are still need to explore in the future

Accumulated studies have demonstrated that HOXA13 is involved in tumor progression in esophageal squamous cell cancer, hepatocellular cancer, prostate cancer and glioma [17, 19, 22, 25]. HOXA13 can promote cancer cell growth and proliferation in esophageal cancer, prostate cancer and glioma, and also affects cancer cell cycle, invasion and apoptosis [19, 22, 25]. In the present study, HOXA13 was found to significantly promote cancer cell proliferation, migration and invasion in vitro, and strongly accelerate tumor growth and metastasis in vivo. These data strongly suggest that HOXA13 is a real and key oncogene in lung cancer. However, the roles of HOXA13 on cancer cell cycle and apoptosis are still unknown. Further studies are needed to solve these problems.

For a long time, few studies on the mechanisms of HOXA13 function were performed in human cancers. A recent study shows that HOXA13 stimulates cancer cell proliferation and metastasis partly via activating Erk1/2 in gastric cancer [34]. Another study reveals that HOXA13 confers 5-FU resistance by MRP1 via a p53-dependent-pathway in gastric cancer [35]. However, the clear molecular mechanisms of HOXA13 roles in cancer are still largely undeveloped. In our present study, we found that HOXA13 decreased the expressions of P53 signaling critical and target genes P53, P21, PMAIP1 and BAX, and increased the expressions of Wnt/ $\beta$ -catenin signaling critical and target genes CTNNB1, MYC, CCND1 and MMP7. These results suggest that HOXA13 activates cancer cell proliferation by inhibiting the activity of P53 signaling-pathway, and promotes cancer cell metastasis by stimulating the activity of Wnt/ $\beta$ -catenin signaling-pathway. This conclusion was further verified by luciferase reporter assays using P53 signaling response reporter vector pp53-TA-Luc and Wnt/ $\beta$ -catenin signaling response reporter vector TOP-Flash/FOP-Flash. These analysis data suggest that HOXA13 acts important oncogenic roles on tumor growth and metastasis by regulating the activities of P53 and Wnt/ $\beta$ -catenin signaling-pathways in lung cancer. However, further researches are still required to explore the exact mechanism that how HOXA13 regulates P53 and Wnt/ $\beta$ -catenin signaling-pathways, such as directly or indirectly and the main functional domain for the regulation.

In conclusion, HOXA13 is a potential and promising prognostic and diagnostic biomarker, and acts as a new oncogene promoting tumor progression and metastasis through P53 and Wnt/ $\beta$ -catenin signaling-pathways in lung cancer (Fig. 8C).

Author contributions

Conception: FH, JD and JL.

Interpretation or analysis of data: YW, BH, YD, GH, XQ, YL, YY, FH, JD and YR.

Preparation of the manuscript: FH, ZC, XQ and YL.

Revision for important intellectual content: FH, JD, YW and JL.

Supervision: FH, JD and JL.

Footnotes

Acknowledgments

This work was supported by the National Natural Science Foundation of China (81672856; 81502551; 81803028), and the Natural Science Foundation of Chongqing (cstc2015jcyjBX0060).

Conflict of interest

The authors have no conflicts of interest to declare.

Supplemental Table

Table S1

Sequence of primers used in the present study

Primer	Sequence (5’-3’)	Purpose
HOXA13-F	ACGAACCCTTGGGTCTTCCC	Expression analysis
HOXA13-R	TCGTGGCGTATTCCCGTTCA
Catenin-F	ATGACTCGAGCTCAGAGGGT	Expression analysis
Catenin-R	ATTGCACGTGTGGCAAGTTC
MYC-F	GCGAACACACAACGTCTTGG
MYC-R	TGAGCTTTTGCTCCTCTGCT
CCND1-F	GATGCCAACCTCCTCAACGA
CCND1-R	GGAAGCGGTCCAGGTAGTTC
MMP7-F	CATGATTGGCTTTGCGCGAG
MMP7-R	AGACTGCTACCATCCGTCCA
P53F	GCTGCTCAGATAGCGATGGT	Expression analysis
P53R	CACGCACCTCAAAGCTGTTC
PMAIP1-F	GACTGTTCGTGTTCAGCTCG
PMAIP1-R	TAGCACACTCGACTTCCAGC
P21-F	CACCGAGGCACTCAGAGGAG
P21-R	CATTAGCGCATCACAGTCGC
BAX-F	GTCACTGAAGCGACTGATGT
BAX-R	GGAAAAACACAGTCCAAGGCA
$\beta$ -actin-F	CCACGAAACTACCTTCAACTCC	Internal control
$\beta$ -actin-R	GTGATCTCCTTCTGCATCCTGT

References

Siegel

Naishadham

and Jemal

, Cancer statistics, 2012, CA Cancer J Clin 62 (2012), 10–29.

Boyle

and Ferlay

, Cancer incidence and mortality in Europe, 2004, Ann Oncol 16 (2005), 481–488.

Ridge

C.A.

McErlean

A.M.

and Ginsberg

M.S.

, Epidemiology of lung cancer, Semin Interv Radiol 30 (2013), 93–98.

Pao

and Chmielecki

, Rational, biologically based treatment of EGFR-mutant non-small-cell lung cancer, Nat Rev Cancer 10 (2010), 760–774.

Molina

J.R.

Yang

Cassivi

S.D.

Schild

S.E.

and Adjei

A.A.

, Non-small cell lung cancer: epidemiology, risk factors, treatment, and survivorship, Mayo Clin Proc 83 (2008), 584–594.

Jemal

Siegel

and Ward

, Cancer statistics, 2010, CA Cancer J Clin 60 (2010), 277–300.

Abate-Shen

, Deregulated homeobox gene expression in cancer: cause or consequence, Nat Rev Cancer 2 (2002), 777–785.

Goodman

F.R.

, Limb malformations and the human HOX genes, Am J Med Genet 112 (2002), 256–265.

De Vita

Barba

Odartchenko

Givel

J.C.

Freschi

Bucciarelli

Magli

M.C.

Boncinelli

and Cillo

, Expression of homeobox-containing genes in primary and metastatic colorectal cancer, Eur J Cancer 29A (1993), 887–893.

10.

Tiberio

Barba

Magli

M.C.

Arvelo

Le Chevalier

Poupon

M.F.

and Cillo

, HOX gene expression in human small-cell lung cancers xenografted into nude mice, Int J Cancer 58 (1994), 608–615.

11.

Makiyama

Hamada

Takada

Murakawa

Takahashi

Tada

et al., Aberrant expression of HOX genes in human invasive breast carcinoma, Oncol Rep 13 (2005), 673–679.

12.

Jung

Kim

R.S.

Lee

S.J.

Wang

and Jeng

M.H.

, HOXB13 homeodomain protein suppresses the growth of prostate cancer cells by the negative regulation of t-cell factor 4, Cancer Res 64 (2004), 3046–3051.

13.

Kanai

Hamada

Asano

Murakawa

Takahashi

Murai

Tada

Miyamoto

Kondo

and Moriuchi

, Aberrant expressions of HOX genes in colorectal and hepatocellular carcinomas, Oncol Rep 23 (2010), 843–851.

14.

Cantile

Scognamiglio

La Sala

La Mantia

Scaramuzza

Valentino

et al., Aberrant expression of posterior HOX genes in well differentiated histotypes of thyroid cancers, Int J Mol Sci 14 (2013), 21727–21740.

15.

and Taylor

H.S.

, Molecular regulation of mullerian development by Hox genes, Ann NY Acad Sci 1034 (2004), 152–165.

16.

Takahashi

Hamada

Murakawa

Takada

Tada

Nogami

et al., Expression profiles of 39 HOX genes in normal human adult organs and anaplastic thyroid cancer cell lines by quantitative real-time RT-PCR system, Exp Cell Res 293 (2004), 144–153.

17.

Cillo

Schiavo

Cantile

Bihl

M.P.

Sorrentino

Carafa

, et al., The HOX gene network in hepatocellular carcinoma, Int J Cancer 129 (2011), 2577–2587.

18.

Huang

Chesnokov

Yokoyama

K.K.

Carr

B.I.

and Itakura

, Expression of the hoxa-13 gene correlates to hepatitis B and C virus associated HCC, Biochem Biophys Res Commun 281 (2001), 1041–1044.

19.

Z.D.

Shen

L.Y.

Wang

Chen

X.M.

Ning

and Chen

K.N.

, HOXA13 promotes cancer cell growth and predicts poor survival of patients with esophageal squamous cell carcinoma, Cancer Res 69 (2009), 4969–4973.

20.

Han

W.W.

Wen

Y.G.

D.P.

Qiu

G.Q.

Tang

H.M.

Peng

Z.H.

and Zhou

C.Z.

, Identification and validation that up-expression of HOXA13 is a novel independent prognostic marker of a worse outcome in gastric cancer based on immunohistochemistry, Med Oncol 30 (2013), 564.

21.

Pan

T.T.

Jia

W.D.

Yao

Q.Y.

Sun

Q.K.

Ren

W.H.

Huang

et al., Overexpression of HOXA13 as a potential marker for diagnosis and poor prognosis of hepatocellular carcinoma, The Tohoku J Exp Med 234 (2014), 209–219.

22.

Dong

Cai

Liu

Jiao

Z.T.

Guo

D.Y.

X.W.

Wang

Y.J.

and Yang

D.K.

, HOXA13 is associated with unfavorable survival and acts as a novel oncogene in prostate carcinoma, Future Oncol, 13 (2017), 1505–1516.

23.

Liu

Tian

Zhang

and Jiang

, Long non-coding RNA DLEU1 promotes proliferation and invasion by interacting with miR-381 and enhancing HOXA13 expression in cervical cancer, Front Genet 9 (2018), 629.

24.

Zhang

and Liu

, Long noncoding RNA LUCAT1 promotes malignancy of ovarian cancer through regulation of miR-612/HOXA13 pathway, Biochem Biophys Res Commun 503 (2018), 2095–2100.

25.

Duan

Han

Wang

Wei

Chen

Zhang

Kang

and Wang

, HOXA13 is a potential GBM diagnostic marker and promotes glioma invasion by activating the Wnt and TGF-β pathways, Oncotarget 6 (2015), 27778–27793.

26.

Gao

Qiu

S.J.

Fan

Zhou

Wang

X.Y.

Xiao

Y.S.

Y.W.

and Tang

Z.Y.

, Intratumoral balance of regulatory and cytotoxic T cells is associated with prognosis of hepatocellular carcinoma after resection, J Clin Oncol 25 (2007), 2586–2593.

27.

Han

Zhang

M.Q.

Liu

W.B.

Sun

Hao

X.L.

Yin

Jiang

Cao

and Liu

J.Y.

, SOX30 specially prevents Wnt-signaling to suppress metastasis and improve prognosis of lung adenocarcinoma patients, Respir Res 19 (2018), 241.

28.

Han

Liu

Jiang

Shi

Yin

et al., SOX30, a novel epigenetic silenced tumor suppressor, promotes tumor cell apoptosis by transcriptional activating p53 in lung cancer, Oncogene 34 (2015), 4391–4402.

29.

Han

Liu

W.B.

Shi

X.Y.

Yang

J.T.

Zhang

Z.M.

et al., SOX30 inhibits tumor metastasis through attenuating wnt-signaling via transcriptional and posttranslational regulation of β-catenin in lung cancer, EBioMedicine 31 (2018), 253–266.

30.

Han

Liu

W.B.

J.J.

Zhang

M.Q.

Yang

J.T.

Zhang

et al., SOX30 is a prognostic biomarker and chemotherapeutic indicator for advanced-stage ovarian cancer, Endocr Relat Cancer 26 (2019), 303–319.

31.

Győrffy

Surowiak

Budczies

and Lánczky

, Online survival analysis software to assess the prognostic value of biomarkers using transcriptomic data in non-small-cell lung cancer, PLoS One 8 (2013), e82241.

32.

Deng

Zhang

Gan

Zhang

Chen

and Hu

, The expression of HOXA13 in lung adenocarcinoma and its clinical significance: a study based on the cancer genome atlas, oncomine and reverse transcription-quantitative polymerase chain reaction, Oncol Lett 15 (2018), 8556–8572.

33.

Knosp

W.M.

Scott

Bächinger

H.P.

and Stadler

H.S.

, HOXA13 regulates the expression of bone morphogenetic proteins 2 and 7 to control distal limb morphogenesis, Development 131 (2004), 4581–4592.

34.

Qin

Chen

Weng

Rong

and Zhou

, Elevated HOXA13 expression promotes the proliferation and metastasis of gastric cancer partly via activating Erk1/2, Onco Targets Ther 12 (2019), 1803–1813.

35.

Han

Song

Wang

Tang

Peng

and Lu

, HOXA13 contributes to gastric carcinogenesis through DHRS2 interacting with MDM2 and confers 5-FU resistance by a p53-dependent pathway, Mol Carcinog 57 (2018), 722–734.

Homeobox-A13 acts as a functional prognostic and diagnostic biomarker via regulating P53 and Wnt signaling pathways in lung cancer

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

Abbreviations

1. Introduction

2. Materials and methods

2.1 Patient samples and cell lines

2.2 RNA isolation

2.3 RT-PCR and qRT-PCR

2.4 Tissue microarray (TMA)

2.5 Immunohistochemical and quantified analyses of HOXA13

2.6 Construction of HOXA13 expression vector

2.7 Construction of stable transfected cells

2.8 HOXA13 siRNA assay

2.9 Western blot analyses

2.10 MTS assay

2.11 5-ethynyl-2’-deoxyuridine (EdU) assay

2.12 Migration and invasion assay

2.13 Clinical microarray database analyses

2.14 ROC (receiver operating characteristic) curve analyses

2.15 Tumorigenicity assay in nude mice

2.16 Experimental metastasis assay in nude mice

2.17 P53 and Wnt/ β -catenin signaling activity and luciferase reporter assays

2.18 Statistical analyses

3.1 HOXA13 expression is sharply increased in NSCLC tissues

3.2 HOXA13 expression is associated with age of NSCLC patients

Table 1 Correlations of HOXA13 expression with clinicopathologic features in human patients ( n = 123)

Table 2 Multivariate analysis of different prognostic factors in human lung cancer patients ( n = 123)

3.6 HOXA13 expression is a smoke-related prognostic factor for ADC patients

3.7 HOXA13 is a promising diagnostic biomarker for NSCLC patients

3.8 HOXA13 markedly stimulates cancer cell growth and proliferation in vitro

3.9 HOXA13 significantly promotes migration and invasion of cancer cells in vitro

3.10 HOXA13 strongly accelerates tumor growth and metastasis in nude mice

3.11 HOXA13 plays the oncogenic roles by regulating P53 and Wnt/ β -catenin signaling-pathways

4. Discussion

Author contributions

Footnotes

Acknowledgments

Conflict of interest

Supplemental Table

References

2.17 P53 and Wnt/ $\beta$ -catenin signaling activity and luciferase reporter assays

Table 1
Correlations of HOXA13 expression with clinicopathologic features in human patients ( $n=$ 123)

Table 2
Multivariate analysis of different prognostic factors in human lung cancer patients ( $n=$ 123)

3.11 HOXA13 plays the oncogenic roles by regulating P53 and Wnt/ $\beta$ -catenin signaling-pathways