Clinical value and potential association of Rab1A and FoxM1 aberrant expression in colorectal cancer

Colorectal carcinoma (CRC) is one of the most common malignancies with a dismal 5-year survival rate. Our recent study indicated that Rab1A expression was closely related to GLI1 expression. A previous study shows that aberrant overexpression of GLI1 promotes colorectal cancer metastasis via FoxM1 overexpression. However, the potential correlation between Rab1A and FoxM1 in CRC remains elusive. Immunohistochemistry was performed to investigate the association of the expression of Rab1A and FoxM1 and to determine the prognosis in 135 CRC tissue and adjacent normal tissues. Using Oncomine datasets, we found that Rab1A and FoxM1 mRNA were obviously upregulated in CRC tissues compared to normal tissues. Additionally, the expression of Rab1A and FoxM1 was significantly higher in CRC tissues than that in normal tissues. Rab1A expression was positively correlated with FoxM1 expression in CRC, especially in TNM stage III. In addition, Rab1A and FoxM1 overexpression was found to be significantly correlated with poor prognosis in CRC patients. Besides, both high expression of Rab1A and FoxM1 led to a worse prognosis than anyone low group, and both low expression of Rab1A and FoxM1 had a better prognosis than the anyone low group. Therefore, Rab1A and FoxM1 play crucial roles and could be used as clinical biomarkers in CRC.

Colorectal cancer (CRC) is a common malignant tumor of the digestive tract, which is associated with high morbidity and mortality worldwide 1,2 . Compared to that in other developed countries, CRC has a high incidence in China 3 . Recently, comprehensive treatment modalities, such as resection and adjuvant chemotherapy, have been widely used. However, the long-term prognosis in patients remains relatively unsatisfactory mainly due to metastasis and resistance to chemotherapy 4,5 . Thus, it is necessary to identify novel targets to improve treatment outcomes in patients with CRC.
Rab proteins have been identified as housekeeping proteins that regulate intracellular membrane dynamics 6 . As a member of the RAB family, Rab1A has been verified to mediate vesicular trafficking 7 . Recent studies reveal that Rab1A is involved in regulating signal transduction 8 , cell autophagy 9 . Previous studies indicated Rab1A acted as an oncogene and Rab1A overexpression promoted a worse prognosis in multiple cancer types such as gastric cancer 10 and hepatocellular carcinoma 11 . In addition, Rab1A overexpression has been positively associated with proliferation, migration, and drug-resistance in various cancers 12,13 .
Forkhead box M1 (FoxM1) belongs to a large family of Forkhead transcription factors 14 . Recent studies indicate that FoxM1 plays a significant role in cell-cycle progression, where endogenous FoxM1 expression peaks at the S and G2/M phases 15 . In addition, FoxM1 overexpression has been linked with cancer progression and poor prognosis in several human cancers, suggesting that FoxM1 might play a crucial role in tumor initiation 16,17 .
A study shows that aberrant overexpression of GLI1 promotes the proliferation and migration of colon cancer cells through the transcriptional activation of FoxM1 18 . Nowadays, our research revealed Rab1A expression was closely related with GLI1 expression in gastric cancer 19 . However, the potential correlation of Rab1A and FoxM1 in both expression and prognosis remains largely unknown in CRC. The aim of this study was to explore the

Results
Expression levels of Rab1A and FoxM1 in gastrointestinal cancers. We determined the expression of Rab1A and FoxM1 in gastrointestinal cancers via the GEPIA Platform. These findings suggested that FoxM1 was significantly overexpressed in esophageal carcinoma (ESCA), stomach adenocarcinoma (STAD), colon adenocarcinoma (COAD), and rectal adenocarcinoma (READ) tumor tissues compared to paired normal tissues (P < 0.05) (Fig. 1A). Rab1A expression was moderately higher in ESCA, STAD, COAD, and READ tumor tissues compared to normal tissues; however, this difference was not significant (P > 0.05) (Fig. 1B). Pooled analysis of www.nature.com/scientificreports/ CRC and normal tissues across 9 Oncomine datasets indicated significant upregulation of FoxM1 mRNA in CRC tissues (P < 0.001) (Fig. 1C). Additionally, using 7 Oncomine datasets, we determined that Rab1A mRNA was obviously upregulated in CRC tissues compared to that in normal tissues (P = 0.006) (Fig. 1D).
Expression levels of Rab1A and FoxM1 in different TNM stages of CRC . First, immunohistochemistry (IHC) staining was performed in 135 pairs of colorectal cancer and adjacent normal tissues to detect Rab1A and FoxM1 expression ( Fig. 2A,B). We found that the expression of Rab1A and FoxM1 was significantly higher in colorectal cancer tissues than in normal tissues (P < 0.001, P < 0.001) (Fig. 2C,D). In addition, a subgroup analysis was performed to detect Rab1A and FoxM1 expression in different TNM stages. Our results revealed that Rab1A expression was remarkably higher in TNM stages II, III, and IV compared to that in TNM stage I (P < 0.01, P < 0.001, P < 0.001) (Fig. 2C). The levels of FoxM1 were significantly higher in TNM stages III and IV compared to that in TNM stage I (P < 0.01, P < 0.001) (Fig. 2D). There was no obvious difference in FoxM1 levels between TNM stages I and II (P > 0.05) (Fig. 2D).
The association between Rab1A and FoxM1 expression and clinicopathological parameters in patients with CRC . We investigated the association between Rab1A and FoxM1 expression and clinicopathological parameters, and the outcomes are shown in Table 1. The results showed that the expression of Rab1A and FoxM1 was obviously linked with several clinicopathological variables, such as venous invasion (P = 0.029) and TNM staging (P < 0.001). However, no obvious differences existed between Rab1A and FoxM1 expression and the left clinicopathological variables (P > 0.05).

Rab1A expression is positively related to FoxM1 expression in CRC .
We first generated a heat map displaying Rab1A and FoxM1 expression detected in 135 CRC patients based on IHC analysis. Then, we compared the levels of Rab1A and FoxM1 detected in 135 CRC tissues and paired normal tissues in a scatter plot of IHC scores (Fig. 3A,B). The outcome suggested that no obvious association existed between Rab1A and FoxM1 expression in normal tissues (P = 0.834) (Fig. 3C). Interestingly, Rab1A expression was closely related to FoxM1 expression in tumor tissues (P < 0.001) (Fig. 3D), which was similar to the results in Table 2 (P < 0.001). Subgroup analysis of associations between Rab1A and FoxM1 expression was performed based on TNM staging. The outcome showed that Rab1A expression was positively associated with FoxM1 expression in TNM stage III (P < 0.001) (Fig. 3G), while no significant associations between Rab1A and FoxM1 expression were obtained  3H). According to the Rab1A and FOXM1 IHC score of human CRC and adjacent normal tissues, it was illustrated as a cluster analysis diagram to demonstrate the differences in expression of the two proteins in tumor and normal tissues (Fig. 4A). The tumor and normal samples were distributed in the PC1 and PC2 axes; the location of the dot was significantly different between tumor and normal tissues in the PC1 axis (P < 0.001) (Fig. 4A).
The influence of Rab1A and FoxM1 overexpression in the prognosis of patients with CRC . We first investigated the influence of Rab1A and FoxM1 overexpression on overall survival (OS) using Kaplan-Meier analysis in patients with CRC. Our findings indicated that patients with higher Rab1A and FoxM1 expression had shorter survival time compared to those with lower expression of Rab1A and FoxM1 in CRC (P < 0.001, P < 0.001) (Fig. 4B,C). In addition, both high (BH) expression group of Rab1A and FoxM1 led to a worse prognosis than any one low (AL) expression group (P < 0.001) (Fig. 4D). Conversely, both low (BL) expression group of Rab1A and FoxM1 had a better prognosis than any one low (AL) group (P < 0.001) (Fig. 4D). The annual survival rates among the three groups are shown in Fig. 4E. In addition, clinicopathological parameters and Rab1A and FoxM1 expression were used to estimate the 3-and 5-year OS (Fig. 5). The nomogram shows each prognostic factor as a score via a point scale and to get the total score (Fig. 5). Results revealed that the expression of Rab1A and FOXM1, which played a significant role in predicting the prognosis in patients with CRC, ranked second. The BL expression group had the best prognosis while the BH expression group had the worst 3-year and 5-year survival rates compared to the other two groups.

Discussion
Rab1A is a small GTPase that regulates ER-to-Golgi vesicular transport 20 . Aberrant Rab1A expression has been linked to a range of human diseases such as Parkinson's disease and cardiomyopathy 21,22 . Previous studies report that Rab1A acts as an oncogene and is overexpressed in various types of cancers, including tongue squamous carcinoma 23 and colorectal cancer 24 . In our study, we first explored Rab1A expression in CRC and paired normal tissues using 7 Oncomine datasets. Our results indicated that Rab1A mRNA was obviously upregulated in CRC tissues compared to the normal tissues. Besides, using IHC, we determined that Rab1A expression was significantly overexpressed in CRC tissues compared to normal tissues, which was consistent with the findings  25 . Subgroup analysis showed that Rab1A expression was remarkably higher in TNM stages II, III, and IV compared to that in TNM stage I, which indicated that Rab1A expression might play a critical role in tumor initiation and development.
A study reports that Rab1A overexpression is positively related to clinicopathological parameters 19,26 . Our results indicated that Rab1A expression was obviously linked to the depth of tumor invasion, lymph node invasion, venous invasion, and TNM staging. Rab1A overexpression is positively correlated with poor prognosis in several gastrointestinal cancers, such as CRC 6 and GC 13 . Consistent with these reports, our findings indicated that lower Rab1A expression indicated better prognosis in patients with CRC.
Our previous study indicates that Rab1A expression is closely related to GLI1 expression in gastric cancer 19 . A recent study reported that GLI1 expression is closely associated with FoxM1 expression and that GLI1 expression elevation promotes colorectal cancer metastasis via FoxM1 overexpression 27 . Based on this perspective, the potential correlation of Rab1A and FoxM1 in expression and prognosis was explored in the current study. Pooled analysis of CRC and normal tissues across 9 Oncomine datasets revealed significant upregulation of FoxM1mRNA in the CRC tissues compared to that in normal tissues. In addition, using IHC analysis, the outcomes revealed that FoxM1 expression was significantly higher in CRC tissues than normal tissues, which   28 . In addition, we observed that FoxM1 expression was significantly higher in TNM stages III and IV compared to that in stage I, which was similar to the results of Rab1A expression level in vary TNM stage. Consistent with previous reports 29 , we found that increased expression levels of FoxM1 led to significantly worse overall survival in patients with CRC.
To investigate the potential correlation of Rab1A and FoxM1 in both expression and prognosis, first, we compared the levels of Rab1A and FoxM1 in a scatter plot of IHC scores. We found that Rab1A expression was closely related to FoxM1 expression in tumor tissues, whereas no obvious expression association in normal tissues was observed. Then, subgroup analysis of expression associations was performed based on different TNM stages. Results revealed that Rab1A expression was positively associated with FoxM1 expression in TNM stage III, indicating that Rab1A and FoxM1 might have promoted tumor progression. In addition, BH expression of Rab1A and FoxM1 led to a worse prognosis than AL expression, and BL expression had a better prognosis than AL expression. The above outcomes revealed that the overexpression of Rab1A and FoxM1 had a greater impact on OS than negative expression. However, our study had some limitations. No in vitro or in vivo experiments were performed. To summarize, Rab1A and FoxM1 may be used together in predicting patient prognosis and may be potential novel targets for individualized therapy of CRC.

Conclusion
We used Oncomine datasets and showed that Rab1A and FoxM1mRNA were obviously upregulated in CRC tissues compared to that in normal tissues. Additionally, the expression of Rab1A and FoxM1 was obviously higher in CRC than in normal tissues. Rab1A expression was closely associated with FoxM1 expression in CRC, especially in TNM stage III. In addition, Rab1A and FoxM1 overexpression led to a worse prognosis in CRC patients. Besides, overexpression of both Rab1A and FoxM1 led to a worse prognosis than AL group, and BL expression of Rab1A and FoxM1 had a better prognosis than AL group. Thus, Rab1A and FoxM1 play crucial roles and could be used as clinical biomarkers of CRC.

Methods
Patients and tissue specimens. The Rab1A and FOXM1 expression in gastrointestinal cancers was obtained via the GEPIA Platform (https ://gepia .cance r-pku.cn), an online analysis website for The Cancer Genome Atlas (TCGA) and Gene Tissue Expression (GTEX) databases. The expression of Rab1A and FOXM1 was obtained from the Oncomine platform (https ://www.oncom ine.org/resou rce/login .html). Data extraction and analysis were performed according to the instructions for authors stated on the website. CRC tissues were None of these patients had received radiotherapy or chemotherapy before radical surgery. All tissue specimens were verified using histopathological studies. Clinical data were obtained from hospital records and all specimens were obtained after permission. The study was supported by the Independent Ethics Committee (IEC) of Suzhou Municipal Hospital and all patients provided written informed consent. After surgery, each patient was followed-up regularly, as described in our previous research 19 .

Immunohistochemistry (IHC). IHC was performed to investigate the expression of Rab1A and FoxM1
in 135 CRC tissue samples and adjacent normal tissues. All analyses were performed in accordance with the relevant guidelines and regulations. Surgical specimens were fixed in 10% formalin, embedded in paraffin, cut into 5-μm sections, dewaxed, and then rehydrated. Next, the processed sections were blocked with 10% goat serum for 30 min and incubated with the polyclonal antibodies of Rab1A and FOXM1 at room temperature overnight. IHC was performed using a tissue staining kit (Zhongshan Biotechnology, Beijing, China) following the manufacturer's protocol. Immunostaining was independently examined by two clinical pathologists who were blinded to the patient outcome. The staining score was calculated using Yang, K's method 28 . The staining scores were as follows: 0 was deemed as (−),1-4 as (+), 5-8 as (++), and 9 -12 as (+++). In our current study, we classified all samples into a high expression group (++ or +++) or a low expression group (− or +) based on protein expression.
Statistical analysis. Data are expressed as the means ± S.E.M. All statistical analyses were performed using SPSS22.0 software (SPSS Inc, Chicago, IL, USA), GraphPad Prism 8 (San Diego, CA) and R programs. Student's t-test (unpaired, two-tailed) or one-way ANOVA were used to compare means between groups. Survival duration was calculated using the Kaplan-Meier method and compared using the log-rank test. IHC results were analyzed using Chi-squared or Fisher's exact tests. Self-developed R program (version 3.6.2 for Windows, https ://cran.r-proje ct.org/) was used for Cluster analysis and Nomogram analysis. A value of P < 0.05 was considered to indicate a statistically significant difference.