High MEIS3 expression intimates poor prognosis for clinical stage II/III colorectal cancer patients

DOI: https://doi.org/10.21203/rs.3.rs-126615/v1

Abstract

Background

The middle stages colorectal cancer (CRC) patients are severely differentiated, so suitable biomarkers are required to distinguish the cohort with high recurrence risk. We hypothesized that among the specimens retained after surgery, those from cohort who relapsed quickly after surgery may be more capable to find the high-risk markers.

Methods

A lable-free analysis was employed to identify candidate proteins markers in CRC tissues relapsed within three years since surgery. Combined with MEIS3 immunohistochemistry characteristic, we analyzed MEIS3 (myeloid ecotropic viral insertion site 3) expression level in clinical stages. Kaplan-Meier Analysis was adopted to analyze the correlation between gene expression and 5-year disease-free survival (DFS) of CRC patients. Finally, cell biology methods were employed to analyze molecular mechanism for tumor cell migration and metastasis.

Results

Through proteomics immunohistochemistry analysis, we identified protein MEIS3 are mainly located in growth front and other highly invasive areas in cancer. And its expression level is closely associated to the clinical stage of CRC. Furthermore, high MEIS3 expression in cancer tissues is closely related to the 5-year DFS expectations of CRC patients. The 5-year DFS of stage II patients with high MEIS3 expression was significantly lower than that of low MEIS3 expression (60.2% % vs 81.1% p < 0.01). The 5-year DFS III patients with high MEIS3 expression was also significantly lower than those of low MEIS3 expression (38.1% vs 56.7% p < 0.01). The analysis on SW480 cells showed that MEIS3 may promote migration and invasion of CRC cells by regulating the expression of LAMB1.

Conclusion

This study intimates that MEIS3 overexpression is associated to poor prognosis for stage II and III patients, and MEIS3 can promote the CRC cells invasion through regulating LAMB1 expression.

1. Introduction

The CRC prevalence has been increasing in recent decades, and becoming one of the most common malignancies and death factors in the world. In China and the United States, the incidence rate of CRC ranks third in all tumors.[1–3] Up to now, the therapy and prognosis of CRC has become an important project in clinical medicine and biological research. With the benefits from successive insight for CRC, the survival rate and life quality of stage IV patients have been significantly improved following the targeted treatment of radiotherapy, chemotherapy and related therapy measures.[4]

However, due to the complex pathology, a considerable number of stage II and III CRC patients are prone to relapse. However, it is difficult to identify these patients based on current diagnostic techniques. As a result, they rarely receive the necessary consistent therapy such as and/or chemotherapy radiotherapy after surgery. Ultimately, patients at these stages haven not get enough benefits from new therapy strategy, so these patients present a high risk of cancer recurrence after surgery. Therefore, finding an effective and simple identification method has become a practical urgency for clinical and biological research.

To address this problem, researchers have also done a lot of work to identify specific biomarkers that can effectively indicate the prognosis for CRC patients. [5–8] Through analyzing the function and mechanism of related genes, the expression fingerprints of gene transcriptome and proteome in peripheral blood and/or tumor tissues, the existed large database system, and the perfect bioinformatics technology, researchers and doctors have found some specific mutated or differentially expressed genes to identify the CRC patients with high-risk .[5, 9–11] Mutations in KRAS and TP53 genes, as well as abnormal DNA methylation in promoter regions, have been found to indicate poor prognosis and shortened survival in CRC patients.[12–14] The abnormally expressed genes such as carcino-embryonic antigen, CA 19 − 9 and TP53 also showed poor prognosis for patients with colorectal cancer.[12, 15] In addition, FOXRED1, FOXM1, L1CAM, CKS2 and other genes have been found to be used to evaluate the prognosis of patients with CRC.[16–18] However, the actual effect of these genes in suggestion for clinical adjuvant therapy still needs to be evaluated. Eventually, a large number of candidate biomarkers may be discarded in clinical practice on a large scale. In short, for CRC patients with stage II and III, there is still a need for more clinical indicators and molecular biomarkers that can accurately predict recurrence after surgical operation.

The probable cause of the recurrence of CRC patients is that very early and distal metastasis has occurred before the operation, which is difficult to detect based on current clinical, genetic and protein methods. But the corresponding genes and proteins have been fully expressed in cancer tissues, but these genes or proteins are still in frog waiting to be identified. Therefore, we believe that by proteomic detection of the tissues of relapsed patients, and corresponding analysis, we can find corresponding biomarkers to distinguish patients who are prone to relapse.

For researching the CRC proteomic fingerprints, we selected the tissues in stage II patients who had relapsed after surgery, according to the follow-up investigation in our department over the years. Through the analysis of the relative protein abundance difference, we screened out a batch of proteins which were significantly up-regulated or down-regulated, that is, DEPs in CRC tissues. With the aid of bioinformatics technology and literature review, we have screened a number of DEPs which may participate in the distant metastasis of cancer cells. Then, we verified the relationship of MEIS3 expression level and the clinical stages of CRC patients. We also found that MEIS3 expression level will affect the 5-year DFS of clinical stage II patients because this gene can promote cancer cell invasion through LAMB1 activation. We expect that parts of stage II patients can receive reasonable radiotherapy and/or chemotherapy in clinic through checking this biomarker.

2. Material And Method

2.1 Clinical Samples

This study was approved by the ethics committee of the Third Affiliated Hospital of Soochow University with No.2017-SY-12, and in accordance with the International Ethical Guidelines for Human Biomedical Research standards, signed informed consent with patients. The clinical staging of patients with colorectal cancer is carried out in accordance with the American Joint Committee on Cancer (ACJJ) standard, that is, the clinical staging is based on the cancer's tissue invasion, lymph node metastasis and distal metastasis[19]. All patients were treated in accordance with the colorectal cancer guidelines of the Chinese Society of Clinical Oncology (CSCO). During follow-up, once the patient was found to relapse, it was recorded as the endpoint of DFS, and if the patient died due to other than colorectal cancer, the data will be deleted. From 2015 to 2018, a total of 403 patients participated in the study, and finally 346 patients were included in the analysis. The characteristics of some cases related to this study are summarized in Table 1. And Kaplan Meier analysis was used to evaluate the 5-year DFS of patients with different factors.

After surgical excision of tumor specimens, one part was stored in paraformaldehyde for histological analysis; one part was stored in TRIZOL to extract total RNA for analyzing gene expression levels, and the remaining part was treated with liquid nitrogen. Store in -80oC refrigerator for Western Blot analysis and other purposes.

Table 1

Clinical and pathological features of CRC patients

Age(years)

Number

Percent

< 50

128

38.32%

50–70

176

52.69%

> 70

30

8.98%

Median age

55 years

Gender

    

male

184

53.18%

Female

162

46.82%

JACC Clinical Stage

    

I

45

13.01%

II

109

31.50%

III

137

39.60%

IV

55

15.90%

Tumor size (cm)

    

< 2

17

6.91%

2–5

95

38.62%

> 5

134

54.47%

2.2 Lable-free proteomic analysis

In order to study clinical II patients who are prone to relapse after surgery, we selected cancer tissues and para-cancerous tissues of 9 patients from patients who relapsed within three years after surgery, and selected 0.1 g of non-connective solid tissue in each sample. The cancer specimens of every 3 patients are mixed into a new sample, and the same 3 para-tissues are mixed into a new analysis sample. Then put the specimen in 8M urea, 50 mM TEAB lysis buffer for full lysis. The protein lysate was reduced by 10 mM DTT,then 30 mM Iodoacetamide (IAM) to alkylate the sample. Then, the lysis were transfered into the ultrafiltration tube (10K, Millipore), and was centrifuged for 12,000 rpm, 20 minutes to remove the remaining IAM and protein impurities.

The100ug of total protein was digested by trypsin for 8hrs at 37oC. After digestion, the product was lyophilized in vacuum freeze dryer at -80oC, and then the lyophilized powder was dissolved with 30 ul of 10% formic acid. The lyophilized peptide was successively transferred to the Dionex Acclaim PepMap100 C18 nm inverse column (5 m, 100 A, 100um I.D. X 2 cm,Thermo Fisher Scientific, sunnvvale, CA, USA) and RSLC C18 separation column (2um, 100 A, Isolation of peptide on 75 m I.D.X 25 cm (Thermo Fisher Scintific) by nanoflow CHIP-LC (CHIP-nLC, AB Sciex). The eluted peptides are directly transferred to the LTQ orbitrap Elite LC/MS system for analysis. The mass spectrometry data acquisition parameters were as follows: 2.2 kV electrospray, mass spectrum acquisition range (350–2000 m/z), primary detector: Oribit -Trap.

Proteome Discover Version 1.4 (Thermo Fisher Scientific) software was used for mass spectrometry data analysis, Mascot Server (Version 2.3 Matrix Science London, UK) was used for search engine, and Uniprot protein database was used for database retrieval, finally, the protein and relative abundance were defined.[20, 21] The protein abundance difference between cancer tissues and para- tissues was analyzed by t-test. When the ratio was greater than 1.75 times, P < 0.05, it was defined as differential expression protein. The differentially expressed proteins were analyzed using online tools such as gene ontology (GO), Visualization and Integrated Discovery (DAVID) and STRING for bioinformatics analysis such as protein clustering and protein interaction network.[22–24]

2.3 Antibodies and Media

In this study, we used the following antibodies:Rabbit anti- MEIS1 (Atlas) ,Rabbit anti-MEIS2 (Proteintech), Rabbit anti-MEIS3 (Atlas), Rabbit anti-LAMB1 (Proteintech), Rabbit anti-POSTN (BosterBio), Rabbit anti- GPNMB (CST), Rabbit anti-ACTB (Proteintech), HRP-conjugated Affinipure Donkey Anti-Mouse IgG(H + L) (Proteintech), HRP-conjugated Donkey Anti-Rabbit IgG(H + L)(Proteintech), Rabbit anti-E-Cadherin (Thermo Fisher Scientific) and Mouse anti-Vimentin (CST). DTT, IAM and TEAB were purchased from SIGMA Co Ltd. Cell culture medium and serum were purchased form Thermo Fisher Scientific.

2.4 WB and Q -PCR

The total RNA in tissues or cells was extracted by Trizol method, and then RNA reverse transcription was performed according to PROMEGA's AMV reverse transcriptase system product manual with RNase-free operation guidelines. The Q-PCR analysis was performed according to the instructions of TOYOBO's SYBR Green Master-mix kit. According to the Ct value generated by the amplification result, the -Δ/ΔCt method was used to quantify the expression abundance of the target sample and the reference sample. The primers involved are shown in table S1.

The Western Blot (WB) was carried out according to the description operation of molecular cloning: A Laboratory Manual.[25] After the development is completed, use Quanitive one-4.40 to analyze the strips according to the relative concentration.

2.5 Immunohistochemistry

The tissues were fixed for 24 hours after 4% PFA, and dehydrated in 70%, 80%, 95%, 95%, 100% and 100% alcohol in turn, and then transparent with xylene, immersed in paraffin wax and embedded. The tissue is cut into slices with a thickness of 8um by the LEICA microtome. The tissue sections were then reverse-treated with 100–50% alcohol from xylene, immersed in sodium citrate antigen repair solution, and repaired with microwave method. The tissue sections were soaked in 5% H2O2 for 10 minutes, washed in tap water, and added with 10% rabbit serum (PBS dilution), then, 1 hour later, it was replaced by antibody-10% serum and incubated overnight at 4oC. The next day, the primary antibody was discarded, washed with PBS three times, and the secondary antibody diluted in 1 × PBS was added. After incubating for 1 hour at room temperature, the color was developed by SABC method. After dehydration with gradient ethanol and xylene, the slices were sealed with neutral gum, and the results were photographed and recorded by microscope.

2.6 Construction of virus particles

The MEIS3 silencing lentiviral particles and the LAMB1 over-expression adenovirus particles were packaged with the assistance of ZQXZ Bio Tech. (Shanghai) Corp. Ltd., and preserved in liquid nitrogen. The knockdown expression vector is composed of three U6 promoters connected in series, each of which directed a shRNA sequence. The three shRNA sense sequences were as follows: 5'-CUUGGAAGGAGAAUGGCAUUAUCTA-3', 5'-CUGCAAGUCAACAACUGGUUCAUTA-3' and 5'-CUGGUGGAGAAGAUGAGGACUUGGA-3'.

2.7 Cell invasion capabilities analyze

The cells were first cultured in DMEM without FBS for 24 hours. During this period, the Transwell chamber was soaked and moistened with 1 × PBS, then matrigel was added, and the gel was incubated in a CO2 incubator at 37 °C to solidify the gel. Then add 600 µl of complete medium containing FBS to the lower chamber, add 200 ul of DMEM hanging cells (density 5 × 10E5 cells/ml) to the upper chamber, and continue to culture for 48hrs, stain with 0.5% crystal violet, and count the results.

3. Results

3.1 DEPs in tissues of relapsed CRC patients

For identifying candidate biomarkers that may represent the high recurrence risk of CRC patients in clinical stage II, we selected tissue specimens from surgical resection of stage II patients who relapsed within 3 years after surgery. For proteomic study, Each 3 cancer growth front tissues were mixed in equal quality, and 3 para-cancer tissues were also mixed in equal amounts to form a test sample. Compared with para-cancer tissues, when the abundance of protein in cancer tissue is higher or lower than 1.75 folds, it is selected as DEPs. A total of 136 DEPs were screened, of which 77 proteins were up-regulated and 59 proteins were down-regulated (Table S2).

3.2 Search new candidate biomarkers of prognosis by bioinformatics technology

For the purpose to analyze the distribution characteristics of DEPs in the tissues of patients with recurrence, we used Database for Annotation, DAVID (version 6.8) and the database of GO (http://geneontology.org/) to perform cluster analysis on all the DEPs found.[22, 24, 26] According to the distribution and clustering of DEPs in tissues and cell component, we found that structure-related proteins such as membranes (GO:0016020,5.90E-10), mitochondrion (GO:0005739, 3.30E-02) extracellular spaces (GO:0005615 9.10E-03), extracellular exosomes (GO:0070062༌9.00E-23) and cytoplasm (GO:0005737༌3.50E-06) accounted for the highest proportion (Fig. 1A), indicating that these subcellular structures played an important role in cancer metastasis and recurrence. From the biological processes involved in related proteins, it can be seen that the signal pathways such as cell-cell adhesion (GO:0098609,5.60E-03), positive regulator of Cell migration (GO:0030335, 3.30E-03), transcription from RNA polymerase II promoter (GO:0045944,2.20E-02), negative regulation of cell proliferation (GO:0008285, 4.10E-03), Wnt signaling pathway, planar cell polarity pathway (GO:0060071, 8.60E-05) and MAPK cascade (GO:0000165, 1.70E-02) show a strong degree of enrichment (Fig. 1B). Finally, these proteins perform their functions by RNA binding (GO:0003723, 4.80E-02), cadherin binding involved in cell-cell adhesion (GO:0098641, 4.90E-04), actin filament protein (GO:0051015, 3.90E-03), protein binding (GO:0005515, 1.00E-10) and transcription factor binding (GO:0008134, 1.90E-03) (Fig. 1C). Since Wnt signaling pathways participate in the regulation of cell proliferation, migration and invasion, metabolism and other processes in an all-round way, it implies that the DEPs related to the Wnt signaling pathway may play a crucial role in CRC.[27–29]

In order to screen new biomarkers more comprehensively and effectively, we used network analysis tools such as STRING to establish a protein-protein interaction network (PPI) for all DEPs.[23] The results showed that FARSA, MEIS3, RER1, VEZT and PLP2 and other 11 DEPs were not adequately involved in the net (Fig. 1E, Fig. S1). To further confirm this result, we established independent PPIs for individual protein of MEIS3, RER1 and FARSA by employing the STRING again (Fig. 1D, Fig S2, S3). After reviewing literature, we found that there were no report about whether these genes related to the recurrence risk for CRC.

3.3 MEIS3 Protein localization in colorectal cancer

Among these proteins, MEIS3 is a direct target gene of the Wnt/beta-Catenin signaling pathway, which participates cell migration during development.[30, 31] Therefore, we selected MEIS3 as candidate biomarker for prognosis of CRC patients. Firstly, we analyzed the MEIS3 localization in tissues by immunohistochemistry firstly. As shown in Fig. 2, MEIS3 protein was expressed in the nucleus, and not in differentiated tissues (Fig. 2A, B). Histologically, MEI3 were located in growth and invasion front (Fig. 2C, D), which is the main area for tumor growth and invasion into normal tissues. Another major area was the intervening region of the middle-highly differentiated area (Fig. 2E, F), which is full of poorly differentiated cancer cells. Such cells also have strong proliferation and invasion capabilities, and cause cancer progression. In contrast, in tumor cells are densely growing areas, the MEIS3 intensity decreases (Fig. 2G, H). During cancer metastasis, it is not differentiated, but poorly differentiated, especially cancer cells of growth front dominate the process, [32] Thus, the immunohistochemistry results suggested that MEIS3 was related to cancer deterioration.

3.4 MEIS3 expression increases with the CRC stages Progresses

Considering the relationship between ACJJ staging and cancer metastasis, [19] We screened MEIS3 expression in protein level in 346 pairs of tissues (Table 2). As a result, the MEIS3 expression level was significantly increased (2.8 folds) in CRC tissues than para-tissues (Fig. 3A, B). And MEIS3 expression increased with the disease stages progresses (Fig. 3C), which implied that MEIS3 could be a candidate biomarker for surgical prognosis of CRC patients. Although several genes expression were reported to be gender-related, we did not find this trend on MEIS3 (Fig. 3D).

3.5 Association of MEIS3 expression and prognosis in stage II/III patients

The DFS of CRC patients after surgery are related to many factors including the clinical stage, gender, gene mutations and expression levels. At present, the clinical stage is believed to be the main factor. After Kaplan-Meier Analysis, we found that the 5-year DFS of CRC patients in different clinical stages during surgery are significantly different (p < 0.01). The 5-year DFS of clinical stage I CRC patients was 87.5% (Fig. 4A). Obviously, the stage IV patients have a poor prognosis, and the 5-year DFS was 23.4% only (Fig. 4A). The 5-year DFS of patients if stage II and stage III are 69.8% and 54.9%, respectively.

The MEIS3 expression was closely correlated with the clinical stages of CRC, so we hypothesized that MEIS3 was also associated with DFS of CRC patients after operation. As shown in the Fig. 4B, in the overall study population, the 5-year DFS of patients with high MEIS3 expression was statistically lower than that of patients with low MEIS3 expression (44.8% vs 59.0% p < 0.05).These suggested that although MEIS3 expression level in CRC tissues was significantly correlated with the DFS, but this prognosis results did not show advantage compared with the DFS based on clinical stages (Fig. 4A).

Then, we performed a new 5-years DFS analysis after we simultaneously introduced the clinical stage and MEIS3 expression level into CRC prognostic assessment. Since stage I patients had a good survival, and Stage IV patients have benefited from modern biomedicine, we focused on the stage II and III patients. As shown in the Fig. 4C, the 5-year DFS of stage II patients with low MEIS3 expression was 81.1% (Fig. 4C), which was comparable to that of the overall stage I patients (87.5%) (Fig. 4A). The 5-year DFS of stage II patients with high MIES3 expression was 60.2%, which was significantly lower than those of high MEIS3 expression (p < 0.01), and roughly same to that of stage III patients with low meis3 expression (56.7%) (Fig. 4C). However, the expected 5-year disease-free survival of patients with stage III and high expression of MIES3 was significantly lower than those of low MEIS3 (38.1% vs 56.7 p < 0.01) (Fig. 4C), indicating a poor prognosis for this part patients.

3.6 Association of MEIS3 and gender on prognosis of stage II and III patients

Since gender is the most stable and convenient factor, and the male and female groups have different responses to CRC, [33] we introduced gender into the survival analysis. As shown in Fig. 4D, the survival rate of female was higher than that of male (45.3% vs 62.7%), but it was not statistically significant (p > 0.05). Then, after introducing MEIS3 and gender at the same time, we performed new analysis again. The stage II female patients with low MEIS3 expression had a significantly high 5-year DFS than those with high MEIS3 expression (90.2% vs 62.5%, p < 0.05) (Fig. 5A). While the stage III patients with high MEIS3 expression have a similar survival level to patients with low MEIS3 expression (59.1% vs 40.6%, p > 0.05) (Fig. 5B).

In the male cohort, the 5-year DFS of stage II patients with low MEIS3 expression was significantly higher than that of high MEIS3 expression (72.8% vs 50.8%, p < 0.05) (Fig. 5C). In the stage III male cohort, the 5-year DFS of patients with MEIS3 low expression was expected to be 52.2%, while that of MEIS3 high expression of was 30.6%, both of which were significantly different (Fig. 5D). In short, regardless of female or male patients, high expression of MEIS3 predicted the deterioration of patients' condition after surgery.

3.7 MEIS3 increases tumor progression tendency through LAMB1

Since MEI3 expression was associated with prognosis of stage II and III CRC patients, located in highly invasive cancer cells, and play a crucial role for inducing cell migration in development, [34, 35] we suspected that MEIS3 may contribute to further cancer progression by enhancing the migration of cancer cells.

Therefore, we constructed lentiviral particles carrying tandem U6 promoter-MEIS3 shRNA and transfected the virus into SW480 cells to silence MEIS3. (Fig. 6A-B) Then, for exploring the downstream regulated genes that promotes cell migration, we returned to the proteomics DEPs again. Among Wnt signaling pathway related proteins, LAMB1, GPNMB and POSTN expression levels were increased (Fig. 1F), which was consistent with the trend of increased cell invasion ability. Accordingly, we also found that LAMB1 expression was most significantly decreased, followed by POSTN and GPNMB (Fig. 6C), indicating that MEIS3 could regulate the expression of POSTN, GPNMB and LAMB1. We also constructed lentiviral particles overexpressing MEIS3 mRNA and transfected the virus into SW480 cells (Fig. 7A). Accordingly, LAMB1 also overexpressed significantly (Fig. 7A), meaning MEIS3 can regulate LAMB1.

The wound healing and transwell assay was used to detect the changes in cell biological function. As MEIS3 knockdown treatment, the invasion and migration ability of cells decreased significantly (Fig. 7B-C). When MEIS3 was overexpressed through lentiviral treated, the migration and invasion ability of SW480 cells has also been significantly enhanced (Fig. 7B-C). These results indicate that MEIS3 can promote the invasion ability of cancer cells through regulating the expression of LAMB1, which may lead to deterioration of patients' condition.

4. Discussion

Prognostic diagnosis is the key to estimate the recurrence risk of CRC patients after therapeutic tumor excision.[5] Today, the prognosis of patients is mainly based on cancer stage, tumor metastasis based on magnetic resonance imaging and/or positron emission tomography, and other markers,such as histopathological grade, gene stability and expression level.[13, 15, 19, 36] Once the lymphatic metastasis or nerve infiltration signs were detected, the patient have been already in a high-risk state.[19] From the perspective of tissues, cells in stable regions and highly invasive regions in cancer tissues have different activated gene groups.[37] Yet, unless specific sample is selected, genomic, transcriptome, and proteome data are difficult to reflect this characteristic. In theory, the cancer tissues of patients who are prone to relapse may contain more abundant such malignant cells. Therefore, we believed that genes, proteins and other biomarkers components related to CRC duplication can be found more effectively in this type of patient tissue.

In this study, we used proteomics and bioinformatics methods to find DEPs from surgical tissues of relapsed stage II CRC patients. As a result, we found 136 DEPs which may closely relate to the recurrence of CRC patients after surgery excision. Especially, through patient survival retrospective, biological information analysis, and reported studies, it can be shown that certain genes/proteins are closely related to cancer cell energy metabolism, proliferation, metastasis, invasion and immune escape. Therefore, we think there may contain protein biomarkers for poor prognosis of CRC patients in the DEPs.

During cancers development, the highly invasive fronts play important roles for recurrence. These areas contain a large number of pluripotent tumor stem cells, which is responsible for rapid growth, invasion and distant metastasis of the cancer.[38] Several biomarkers have been focused on this area and have shown good clinical effects.

In addition, the regions of differentiated cancer tissues are often separated by connective tissue. Based on connective tissue, accompanied with the entry of blood vessels and nerve tissues, the differentiated areas enter a relatively stable state. Thus, these areas are not easily transformed into highly invasive state, and the patient will not deteriorate due to this type of tumor tissue in a short time. Another case between the differentiated areas is poorly differentiated cells, which are scattered or distributed in a dense state with high invasiveness and proliferation. These cells are also an important factor in the distant metastasis, and this corresponding region often become a key factor for cancer recurrence. Therefore, we think it meaning to pay attention to the biological characteristics of the pathological tissues, especially genes which have positive guiding value for prognosis.

MEIS3, is mainly expressed in invasive fronts and the spaces between differentiation areas of CRC tissues, forming scattered or dense nuclear staining characteristic. The distribution range of these kinds of cells have positively correlated with the MEIS3 expression level. At the molecular level, we confirmed that MEIS3 may promote the migration and invasion of tumor cells by promoting the LAMB1 expression. Therefore, the high expression of MEIS3 is not only significantly related to the clinical stage of CRC patients, but also hinting a high-risk for CRC metastasis and recurrence. The 5-year DFS of clinical stage II patients with high MEIS3 expression was comparable to the overall level of stage III patients. And 5-year DFS expectation of the corresponding stage III patients with high MEIS3 expression was approximately closer to the overall level of the stage IV patients. Especially when combined with gender, the differentiation was further enlarged, and the risk of male was greater than that of female.

In the therapy of patients with advanced CRC, adjuvant chemotherapy and radiotherapy have played an important role in clinical practice, and the expected 5-year DFS rate has been significantly improved.[4, 19] However, due to the severely differentiated outcomes in stage II and III patients, it is controversial whether adjuvant radiotherapy and/or chemotherapy is needed for these patients. Ultimately, the benefits of these patients from modern medicine are far less than advanced stage patients.[5, 39] Here, stage II and III patients with high expression of MEIS3, especially males, had poor prognosis. It is need to treat these patients according the therapeutic measure of high-risk patients to improve their survival rate and life quality.

Of course, we also found that MEIS3 can promote the transformation through activation of LAMB1 which play crucial roles in Epithelial-Mesenchymal Transition,[38, 40] yet, it is still a superficial result. We don't know the specific genes involved in this process, nor the specific regulation process mode between these genes. Therefore, further research is needed to identify the specific molecular mechanism of MEIS3 in determining the fate of tumor/cancer cells.

In addition, we need to study the feasibility of MEIS3 as a biomarker of poor prognosis with larger retrospective CRC patients groups, especially in patients with concurrent therapy and follow-up, before applied in clinical therapy for improving their survival rate and life quality.

5. Conclusion

In summary, MEIS3 is highly expressed in the poorly differentiated areas and growth fronts of tumors/cancers. And the 5-year DFS of stage II and III CRC patients with high MEIS3 expression are significantly lower than those with low MEIS3 expression. In a short word, MEIS3 has great promise as a biomarker for surgical prognosis in patients with colorectal cancer.

Abbreviations

CRC: colorectal cancer, DFS: disease-free survival, DEPs: differentially expressed proteins, MEIS3: myeloid ecotropic viral insertion site 3, ACJJ: American Joint Committee on Cancer, DAVID: Visualization and Integrated Discovery, PPI: protein-protein interaction network, WB: Western Blot, Q-PCR: quantity polymerase chain reaction, ACJJ:  American Joint Committee on Cancer, CSCO:Chinese Society of Clinical Oncology

Declarations

Conflict of interest: The authors have no conflict of interest to declare.

Competing interests

The authors declare that they have no competing interests.

Acknowledgements

Not applicable.

Authors’ contributions

LingY and LiHT designed the study; MaJ, WangMH,ChenJ and ZhuM performed the experiments, MaJ, LingY, ChenJ,PKJ, WangMH,LiA and ZhuM analyzed and

interpreted the data; MaJ, LingY, LiTH and LiA wrote the manuscript; all authors

commented and approved the final manuscript.

Funding

Not applicable.

Availability of data and materials

All data generated or analyzed during this study are included in this published article and its supplementary information files.

Consent for publication

Informed consent form for publication was obtained from all authors.

Competing interests

The authors state no competing of interest.

Ethics approval and consent to participate

This study is approved by Medical Ethics Committee of Changzhou Cancer Hospital.

References

  1. Siegel, R.L., et al., Colorectal Cancer Statistics, 2017. CA CANCER J CLIN 2017. 67: p. 177–193.
  2. Chen, W., et al., Cancer incidence and mortality in China, 2014. Chin J Cancer Res, 2018. 30(1): p. 1-12.
  3. Ronshou, Z., et al., Report of cancer epidemiology in China,2015. China J Oncol 2019. 41(1): p. 19-28.
  4. Lieu, C.H. and W.A. Messersmith, Cetuximab or Bevacizumab With First-Line Chemotherapy in Advanced KRAS Wild-Type Colorectal Cancer: No Difference, but Not the Same. JAMA, 2017. 317(23): p. 2376-2378.
  5. Aasebo, K., et al., CDX2: A Prognostic Marker in Metastatic Colorectal Cancer Defining a Better BRAF Mutated and a Worse KRAS Mutated Subgroup. Front Oncol, 2020. 10: p. 8.
  6. Arscott, W.T. and K.A. Camphausen, Exosome characterization from ascitic fluid holds promise for identifying markers of colorectal cancer. Biomark Med, 2011. 5(6): p. 821-2.
  7. Hutchinson, L., Gastrointestinal cancer: CDX2: prognostic marker for high-risk colon cancer. Nat Rev Clin Oncol, 2016. 13(3): p. 134-5.
  8. Benhaim, L., et al., Gender-specific profiling in SCN1A polymorphisms and time-to-recurrence in patients with stage II/III colorectal cancer treated with adjuvant 5-fluoruracil chemotherapy. Pharmacogenomics J, 2014. 14(2): p. 135-41.
  9. Averaimo, S., et al., CLIC1 functional expression is required for cAMP-induced neurite elongation in post-natal mouse retinal ganglion cells. J Neurochem, 2014. 131(4): p. 444-56.
  10. Liu, Y., et al., Competitive endogenous RNA is an intrinsic component of EMT regulatory circuits and modulates EMT. Nat Commun, 2019. 10(1): p. 1637.
  11. Christou, N., et al., E-cadherin: A potential biomarker of colorectal cancer prognosis. Oncol Lett, 2017. 13(6): p. 4571-4576.
  12. Nasif, W.A., et al., Implications of CEA and p53 overexpression in the poor prognosis of colorectal cancer. Med Oncol, 2006. 23(2): p. 237-44.
  13. Bazan, V., et al., Molecular detection of TP53, Ki-Ras and p16INK4A promoter methylation in plasma of patients with colorectal cancer and its association with prognosis. Results of a 3-year GOIM (Gruppo Oncologico dell'Italia Meridionale) prospective study. Ann Oncol, 2006. 17 Suppl 7: p. vii84-90.
  14. Gonzalez-Aguilera, J.J., et al., Simultaneous mutations in K-ras and TP53 are indicative of poor prognosis in sporadic colorectal cancer. Am J Clin Oncol, 2004. 27(1): p. 39-45.
  15. Sisik, A., et al., CEA and CA 19-9 are still valuable markers for the prognosis of colorectal and gastric cancer patients. Asian Pac J Cancer Prev, 2013. 14(7): p. 4289-94.
  16. Yu, M.H., et al., Up-regulated CKS2 promotes tumor progression and predicts a poor prognosis in human colorectal cancer. Am J Cancer Res, 2015. 5(9): p. 2708-18.
  17. Chu, X.Y., et al., FOXM1 expression correlates with tumor invasion and a poor prognosis of colorectal cancer. Acta Histochem, 2012. 114(8): p. 755-62.
  18. Fei, W., S. Liu, and X. Hu, High FOXRED1 expression predicted good prognosis of colorectal cancer. Am J Cancer Res, 2016. 6(11): p. 2722-2728.
  19. Edge, S.B., et al., AJCC 7th Ed Cancer Staging Manual 7th. 2015: Springer New York Dordrecht Heidelberg London.
  20. Consortium, T.U., UniProt: the universal protein knowledgebase. Nucleic Acids Res, 2017. 45(D1): p. D158-D169.
  21. Apweiler, R., et al., UniProt: the Universal Protein knowledgebase. Nucleic Acids Research, 2004. 32( suppl_1): p. D115–D119.
  22. Huang, D.W., et al., DAVID Bioinformatics Resources: expanded annotation database and novel algorithms to better extract biology from large gene lists. Nucleic Acids Res, 2007. 35(Web Server issue): p. W169-75.
  23. Szklarczyk, D., et al., STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res, 2019. 47(D1): p. D607-D613.
  24. Huang da, W., B.T. Sherman, and R.A. Lempicki, Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc, 2009. 4(1): p. 44-57.
  25. Green, M.R. and J. Sambrook, Molecular Cloning：A Laboratory Manual. 2012.
  26. Ashburner, M., et al., Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet, 2000. 25(1): p. 25-9.
  27. T, Z., R. N, and B. M., Wnt signaling in cancer. Oncogene, 2017 36(11): p. 1461-1473.
  28. Krishnamurthy, N. and R. Kurzrock, Targeting the Wnt/beta-catenin pathway in cancer: Update on effectors and inhibitors. Cancer Treat Rev., 2018. 62: p. 50-60.
  29. Nusse, R. and H. Clevers, Wnt/β-Catenin Signaling, Disease, and Emerging Therapeutic Modalities. Cell, 2017. 169(6): p. 985-999.
  30. Elkouby, Y.M., et al., Mesodermal Wnt signaling organizes the neural plate via Meis3. Development, 2010. 137(9): p. 1531-41.
  31. Uribe, R.A. and M.E. Bronner, Meis3 is required for neural crest invasion of the gut during zebrafish enteric nervous system development. Mol Biol Cell, 2015. 26(21): p. 3728-40.
  32. Chen, Q., et al., MiR-124-5p inhibits the growth of high-grade gliomas through posttranscriptional regulation of LAMB1. Neuro Oncol, 2014. 16(5): p. 637-51.
  33. Butcher, D., et al., Female gender is a major determinant of changing subsite distribution of colorectal cancer with age. Cancer, 1985. 56(3): p. 714-716.
  34. Dibner, C., et al., The Meis3 protein and retinoid signaling interact to pattern the Xenopus hindbrain. Dev Biol, 2004. 271(1): p. 75-86.
  35. Choe, S.K., et al., Meis cofactors control HDAC and CBP accessibility at Hox-regulated promoters during zebrafish embryogenesis. Dev Cell, 2009. 17(4): p. 561-7.
  36. Tampakis, A., et al., L1CAM expression in colorectal cancer identifies a high-risk group of patients with dismal prognosis already in early-stage disease. Acta Oncol, 2020. 59(1): p. 55-59.
  37. Barretina, J., et al., The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature, 2012. 483: p. 603–607.
  38. Govaere, O., et al., The PDGFRalpha-laminin B1-keratin 19 cascade drives tumor progression at the invasive front of human hepatocellular carcinoma. Oncogene, 2017. 36(47): p. 6605-6616.
  39. Cunningham, D., et al., Cetuximab Monotherapy and Cetuximab plus Irinotecan in Irinotecan-Refractory Metastatic Colorectal Cancer New Egland J Medicine, 2004. 351: p. 337-345.
  40. Horejs, C.M., et al., Biologically-active laminin-111 fragment that modulates the epithelial-to-mesenchymal transition in embryonic stem cells. Proc Natl Acad Sci U S A, 2014. 111(16): p. 5908-5913.

Supplementary Tables

Table S1 Primers used in this study

Gene

Primer

Primer Sequences

Tm (℃)

Product Length

ACTB

ACTB-F

5'- CGCAAAGACCTGTACGCCAA -3'

58.4

216bp

ACTB-R

5'- CTCCTGCTTGCTGATCCAC -3'

MEIS1

MEIS1-F

5'- AAGCAGTTGGCACAAGACAC -3'

55.1

186bp

MEIS1-R

5'- AATTCCCATATGTTGCTGACCG -3'

MEIS2

MEIS2-F

5'- CGATAACTTCTGCCACCGAT -3'

57.3

266bp

MEIS2-R

5'- TGTCTAAACCATCCCCTTGC -3'

MEIS3

MEIS3-F

5'- GACATCGCTGCCTTTGCCAA -3'

59.8

328bp

MEIS3-R

5'- TCCCCAAATGTACAGACCCACT -3'

LAMB1

LAMB1-F

5'- CAGAAGCCGAAAGCCTAGACA -3'

55.8

232bp

LAMB1-R

5'- AACTGGGATTCTCGCTCCATC -3'

POSTN

POSTN-F

5'-ATTTGGGGCTTTTCATTAACCAT -3'

56.6

270bp

POSTN-R

5'- TCTCAAAAGCCTCATTGGTG -3'

GPNMB

GPNMB-F

5'- ACCAGGAAAAGGATCCGCTA -3'

51.7

259bp

GPNMB-R

5'- TATCAGTTTCACAACATGGCT -3'

Tabel S2. Differentially expressed proteins in colorectal cancer tissues

Symbols

Uniprot Accession

Folds

T-Test

CCT5

P48643

1.96

0.001

CDC25B

P30305

1.98

0.001

EIF3B

P55884

2.35

0.001

NME2

P22392

3.09

0.001

RAB2A

P61019

4.31

0.001

EIF3K

Q9UBQ5

0.5

0.001

ETFA

P13804

0.54

0.001

LMNA

Q3BDU5

0.49

0.001

CTNNB1

P35222

2.01

0.000

MEIS3

Q99687

3.07

0.001

PARP1

P09874

1.96

0.001

SNRPF

P62306

6.59

0.001

UBA1

P22314

3.65

0.001

PRELP

P51888

0.32

0.001

LGALS3

P17931

0.26

0.001

PRKACB

P22694

0.34

0.001

MAPK3

P27361

0.27

0.001

GGT5

P36269

0.20

0.001

CDC37

Q16543

1.98

0.002

EMILIN1

A0A0C4DFX3

2.35

0.002

NUCKS1

Q9H1E3

3.25

0.002

PITPNA

I3L471

2.24

0.002

SSBP1

Q04837

3.79

0.002

MT2A

P02795

0.29

0.002

RER1

O15258

2.49

0.002

ETFB

P38117

0.51

0.002

MAPRE1

Q15691

2.85

0.003

POSTN

Q15063

1.96

0.003

PSME2

Q9UL46

3.91

0.003

SUB1

P53999

9.54

0.003

TXNDC17

I3L0K2

1.95

0.003

ILK

Q13418

0.23

0.003

CLDN7

O95471

0.54

0.003

CRYAB

E9PNH7

0.52

0.003

IPO5

O00410

2.73

0.004

PSMC2

P35998

4.07

0.004

THY1

P04216

4.77

0.004

CDKN1A

P38936

0.54

0.004

ECI1

P42126

0.49

0.004

PRDX3

P30048

0.47

0.004

FZD6

O60353

2.43

0.005

HNRNPR

O43390

2.52

0.005

PSMB4

P28070

2.28

0.005

RPL11

P62913

2.23

0.005

SF3B3

Q15393

5.59

0.005

SMPDL3B

F8VWW8

0.37

0.005

MATR3

D6R991

3.05

0.006

SFRP2

Q96HF1

0.36

0.006

PDLIM5

H0Y8Y3

0.47

0.006

SUN2

Q9UH99

0.42

0.006

GPNMB

Q14956

2.21

0.006

MMP1

P03956

2.24

0.007

CALD1

E9PGZ1

1.86

0.008

SULF1

Q8IWU6

9.94

0.008

CAMK2D

Q13557

0.53

0.008

MYLK

Q15746

0.30

0.008

AP2B1

P63010

1.81

0.009

DHX9

Q08211

2.28

0.009

LAMB1

P07942

2.22

0.009

CA2

P00918

0.50

0.009

CAND1

Q86VP6

0.54

0.009

LASP1

C9J9W2

0.18

0.009

BMP4

P12644

1.86

0.010

CDK4

P11802

1.98

0.010

PTBP1

P26599

4.28

0.010

SND1

Q7KZF4

6.23

0.010

LTF

P02788

0.26

0.010

MMP11

P24347

2.99

0.011

PML

H3BVD2

3.35

0.011

B2M

P61769

0.43

0.011

TCF7L2

Q9NQB0

0.43

0.011

FZD5

Q13467

0.18

0.011

CLTA

P09496

1.99

0.012

CSNK2B

P67870

1.99

0.012

HYOU1

E9PJ21

2.62

0.013

RAB7A

P51149

4.42

0.013

RPN2

P04844

5.22

0.013

SELENBP1

Q13228

0.35

0.013

CA1

P00915

0.55

0.013

CCND1

P24385

1.90

0.014

HMGCL

P35914

0.23

0.014

FOXO4

P98177

0.13

0.015

EPCAM

P16422

1.93

0.016

KPNA2

P52292

2.93

0.016

POLR2H

C9JLU1

3.43

0.016

ST13

H7C3I1

7.80

0.016

PSMD11

O00231

0.34

0.016

FBLN5

Q9UBX5

0.11

0.016

RRP12

Q5JTH9

5.11

0.017

FHL1

Q5JXI8

0.53

0.018

CCT2

P78371

1.92

0.019

CLIC1

O00299

1.99

0.019

FARSA

K7EPH2

2.49

0.019

IFI16

Q16666

2.20

0.019

BTD

C9JSN9

0.43

0.019

TPM1

H0YKX5

0.43

0.019

GPX3

A0A087X1J7

0.21

0.019

PLP2

Q04941

3.14

0.020

KHDRBS1

Q07666

0.24

0.020

CCNB1

Q9NPC3

1.88

0.021

C16orf89

Q6UX73

0.51

0.021

APEX1

P27695

2.12

0.022

NME1

P15531

3.02

0.022

VAPA

Q9P0L0

0.49

0.023

ARPC4

P59998

2.12

0.024

CYFIP1

Q7L576

1.90

0.026

H1F0

P07305

2.50

0.026

AQP1

P29972

0.50

0.026

SPINK4

O60575

0.39

0.027

GSN

P06396

0.32

0.027

RPL8

P62917

4.73

0.029

SOD3

P08294

0.38

0.029

ASPN

Q9BXN1

1.85

0.030

SFXN1

Q9H9B4

0.48

0.032

SNX2

O60749

0.48

0.032

CCNB1IP1

P27695

2.13

0.033

SLC1A5

Q15758

5.79

0.033

STT3A

P46977

8.62

0.033

ANKRD22

Q5VYY1

1.81

0.034

CKB

P12277

0.55

0.034

CRABP2

Q5SYZ4

1.99

0.035

AKR7A3

O95154

0.55

0.035

GANAB

Q14697

2.67

0.036

ARHGDIB

F5H3P3

1.82

0.037

WDR1

O75083

0.48

0.037

LRPPRC

P42704

2.91

0.039

ADH1B

D6RHZ6

0.53

0.040

ALDH1A1

P00352

0.53

0.040

MGST3

O14880

0.38

0.040

MVP

Q14764

0.44

0.043

SAE1

Q9UBE0

2.19

0.045

CDC20

Q12834

1.96

0.046

IPO9

Q96P70

2.21

0.048

CALM1

P62158

0.50

0.048

VDAC1

P21796

1.89

0.049

VEZT

F8W8C2

3.31

0.049