Long Noncoding RNA AL161729.4 Acts as an miR-760 Sponge to Enhance Colon Adenocarcinoma Proliferation via Activating PI3K/Akt Signaling

Background: Colon adenocarcinoma is one of the most common gastrointestinal malignancies with poor prognosis and high mortality. The mRNA-miRNA-lncRNA regulatory network mediated by m6A methylation plays an important role in a variety of cancers including colon adenocarcinoma. Methods: We integrated and analyzed the gene expression data and clinical information of 473 patients with colon adenocarcinoma and 41 normal samples in The Cancer Genome Atlas database. The luciferase reporter gene experiment is used to detect the targeting effect between gene, miRNA and lncRNA. Real-time PCR and Proliferation assays were performed to detect the biological function of gene, miRNA and lncRNA. Results: A risk model and Nomogram that could accurately predict the survival time of patients were constructed through informatics analysis. HYOU1, AL161729.4, miR-760 are differentially expressed in COAD patients and normal samples and are signi�cantly related to survival, and there is a targeted binding effect between the three. Conclusions: HYOU1-AL161729.4-miR-760 ceRNA regulatory network could regulate the proliferation of SW620 cells through the PI3K/Akt signaling pathway.


Background
Colorectal cancer (CRC) is one of the most common gastrointestinal malignancies with poor prognosis and high mortality; it is also the second leading cause of cancer-related deaths worldwide [1].More than 2.2 million new cases and 1.1 million deaths are predicted by 2030, and the global burden caused by this is also increasing year by year [2].Approximately 20%-25% of patients with CRC metastases to distant organs at the time of initial diagnosis [3].As a key tool for early detection, biomarkers have made great progress in the last few decades and have had a positive impact on the treatment of patients with CRC [4].Colon adenocarcinoma (COAD) is the most common histological subtype of CRC [5].
N6-methyladenosine (m6A) methylation is one of the most common RNA modi cations and plays an important role in various life activities and diseases [6][7][8][9].The m6A modi cations mainly include methyltransferase (m6A "writer"), demethylase (m6A "eraser"), and m6A "reader" protein [10,11].Recent studies have shown that m6A methylation plays a key role in cancer through various mechanisms and can be used as a marker for early cancer diagnosis [12,13].m6A-related lncRNA and miRNA play an important role in a variety of diseases, including cancer [14,15].MicroRNA (miRNA) is a small RNA that regulates the expression of complementary messenger RNA.It plays an important role in developmental timing, cell death, cell proliferation, hematopoiesis, and patterning of the nervous system [16].As RNAs that are longer than 200 nucleotides and do not encode proteins, lncRNAs also play an important role in a variety of life activities and diseases [17].Recent studies have shown that competing endogenous RNA (ceRNA) networks based on mRNA-miRNA-lncRNA play an essential role in various diseases, including cancer, and are expected to become new early cancer diagnostic markers and therapeutic targets [18][19][20].Taking m6A-related genes and lncRNA as the entry point, constructing and verifying a new mRNA-miRNA-lncRNA regulatory network is urgent for the early diagnosis and treatment of COAD.
In this study, we downloaded and compiled the gene, miRNA, and lncRNA expression data of patients with COAD and their corresponding clinical information from The Cancer Genome Atlas (TCGA) database.A new COAD prognostic risk nomogram and mRNA-miRNA-lncRNA regulatory network were constructed through bioinformatics analysis.The experiment veri ed that HYOU1-AL161729.4-miR-760passed through the PI3K/Akt signal to mediate the occurrence of COAD.

Patients and dataset
The gene expression pro les and clinical data of 473 patients with COAD and 41 normal samples were obtained at TCGA.Then, the gene transfer format le with re-annotated gene expression data was used and integrated with clinical information.Gene expression pro les mainly included gene, lncRNA, and miRNA.Clinical information mainly included survival time, survival status, age, sex, and tumor stage.

Identi cation of prognosis-related lncRNAs
We rst analyzed the co-expression of m6A-related genes and lncRNAs through R packages of "limma" to nd the lncRNAs related to m6A and used R packages of "igraph" to draw the co-expression network.
Then, we used R packages of "survival" to analyze the survival of m6A-lncRNA and obtained m6A-lncRNA signi cantly related to the survival of patients with COAD.Finally, the difference in the expression of survival-related m6A-lncRNA was analyzed using the R packages of "limma" in patients with COAD and normal samples.

Construction and veri cation of the prognostic risk model
We used the least absolute shrinkage and selection operator (LASSO) method to analyze survival-related m6A-lncRNA and selected key lncRNAs to construct a prognostic risk model for patients with COAD.The formula for predicting the risk score of patients using the prognostic risk model was as follows: where Coef denotes the coe cient of this gene and x denotes the expression level of this gene.
All patients with COAD were randomly divided into training and test groups.In each group, patients with COAD were divided into high-risk and low-risk groups according to the median value of the risk score of the training group.The receiver-operating characteristic (ROC) curve of 1-year, 3-year, and 5-year survival curves, single-factor independent prognostic analysis, multi-factor independent prognosis, and correlation analysis of the survival time and risk score were used to evaluate the constructed prognostic risk model.
The area under the curve (AUC) value in the ROC curve was greater than or equal to 0.6, indicating that the model had good prediction accuracy.

Nomogram
We performed Cox proportional-hazards analysis on the gene expression data and information such as age, sex, cancer stage, risk score, and so forth, of patients to construct a nomogram so as to more conveniently and accurately predict the prognostic survival time of patients.At the same time, the ROC curve and calibration curve were drawn to evaluate the accuracy of the analysis.

Functional analysis of the prognostic risk model lncRNA
We used the R packages of "limma" to analyze the correlation between the expression of lncRNA and the risk score of the risk model so as to explore the relationship between the two, and performed survival analysis on these lncRNAs one by one to explore the relationship between these lncRNAs and overall survival.We found the lncRNA that was positively correlated with the risk score and negatively correlated with the overall survival as the follow-up research object (interest lncRNA).Finally, we used the lncATLAS (https://lncatlas.crg.eu/) and lncLocator (http://www.csbio.sjtu.edu.cn/bioinf/lncLocator/)databases to perform subcellular location of the interest lncRNA so as to evaluate whether lncRNA gene knockdown experiments could be performed in the future.
Prediction and functional analysis of lncRNA target miRNA and gene First, we used The Encyclopedia of RNA Interactomes (ENCORI) (http://starbase.sysu.edu.cn/)online database to predict the target miRNA of the interest lncRNA, and crossed it with the low-expressed miRNA in patients with COAD to obtain the target miRNA of the interest lncRNA.Second, we predicted the target gene of target miRNA through miRDB (http://mirdb.org/),miRTarBase (http://starbase.sysu.edu.cn/starbase2/index.php), and TargetScan (http://www.targetscan.org/vert_71/)databases.Then, Gene Expression Pro ling Interactive Analysis (GEPIA, http://gepia.cancer-pku.cn/)database was used to analyze the expression difference of the target genes so as to nd the target genes highly expressed in patients with COAD.Finally, an lncRNA-miRNA-gene ceRNA regulatory network signi cantly related to the survival of patients was constructed based on the principle of base complementary pairing.

Gene expression vector construction
The base sequence of the CD region of HYOU1 [full length 3000 base pairs (bp)] was chemically synthesized using the whole gene synthesis technology, and then constructed into the pIRES2-EGFP vector (Tsingke, Beijing, China) to obtain the overexpression vector of HYOU1.The binding site of miR-760 in the 3′-untranslated region (UTR) of HYOU1 was found, 500-bp upstream and downstream of the binding site were chemically synthesized, and the site was constructed into the pSi-Check2 vector to construct the HYOU1 wild-type 3′-UTR vector.In the same way, a wild-type vector of AL161729.4(length 476 bp) was constructed.Using the point mutation method, HYOU1 mut-type 3′-UTR vector and AL161729.4mut-type vector were constructed.HYOU1 siRNA1#, HYOU1 siRNA2#, AL161729.4siRNA1#, AL161729.4siRNA2#, siRNA negative control, miR-760 mimics, mimics negative control, miR-760 inhibitors, and inhibitor negative control were chemically synthesized (Gene Pharma, Shanghai, China).The sequence is shown in Table 1.
Cell culture and luciferase reporter gene experiment HT29 and SW620 cells were purchased from the National Collection of Authenticated Cell Cultures (Shanghai, China) and cultured in a 37°C cell incubator containing 5% CO 2 .The medium of HT29 cells was McCoy's 5A medium (HyClone,Waltham, MA, USA), and the medium of SW620 was Leibovitz's L-15 (HyClone).Media were supplemented with 10% fetal bovine serum (Invitrogen, Carlsbad, CA, USA), 100 U/ml penicillin, and 100 ug/ml streptomycin (Invitrogen).Lipofectamine 3000 transfection reagent (Invitrogen) was used to transfect the corresponding gene expression vector and oligonucleotides into the cells, the cell culture was continued for 48 h, and the cells were then lysed.The Dual-Luciferase Reporter Assay System (Promega, WI, USA) was used to detect the uorescence level so as to detect the targeted binding between lncRNA-miRNA-gene.

RNA extraction and quantitative polymerase chain reaction
The gene expression vector and oligonucleotides were transfected into the HT29 and SW620 cells and cultured for 48 h.TRIzol reagent (Invitrogen) was used to lyse the cells and extract total RNA.Then, a RevertAid First-Strand cDNA Synthesis Kit (Thermo Fisher Scienti c, MA, USA) was used for reverse transcription to obtain cDNA.Finally, SYBR Green Polymerase Chain Reaction (PCR) Master Mix (Thermo Fisher Scienti c) was used to perform uorescence quantitative PCR, and the relative expression of genes was determined by the 2 -ΔΔCt method.All primer sequences are listed in Table 2.

Cell proliferation assay
The HT29 and SW620 cells were seeded into 96-well plates at a density of 2 × 10 4 cells per well, and the culture was continued for 24 h.Then, we changed the fresh medium and transfected the gene expression vector and oligonucleotides into cells.These cells were mixed with 10 µL of MTT staining solution and incubated for 4 h on days 1, 3, 5, and 7 after transfection.Then, the MTT cell proliferation assay (Solarbio, Beijing, China) was used to detect cell proliferation.The absorbance was measured at 490 nm.
All experiments were performed in triplicate.

Statistical analysis
In the co-expression analysis of m6A-related genes and lncRNA, Pearson correlation coe cient ≥0.4 and P <0.001 are the criteria for judging whether they were related.In the analysis of survival difference and expression difference, P <0.001 indicated a signi cant difference.All experiments were performed independently at least three times with similar results, and representative experiments were shown.P <0.05 was considered statistically signi cant (NS > 0.05; * P < 0.05; ** P < 0.01; *** P < 0.001).
A signi cant difference in survival was found between the high-risk and the low-risk groups in the training group; the survival rate in the low-risk group was signi cantly higher than that in the high-risk group (Fig. 2D).The AUC value of the 1-year, 3-year, and 5-year survival curves was 0.766, 0.785, and 0.726, respectively, which showed that the survival curve was credible (Fig. 2E).The calibration curve of the survival curve further veri ed its credibility (Fig. 2F).The trend in the test group was the same as that in the train group (Fig. 2G-2I).The survival status results showed that the higher the risk score of patients, the greater the number of deaths (Fig. 2M and 2N).The results of correlation analysis showed that the overall survival rate of patients was negatively correlated with the risk score, although the statistical difference was not signi cant (Fig. 2O and 2P).Single-factor independent prognostic analysis and multivariate independent prognostic analysis showed that whether in the train group or in the test group, the risk score could be used as a key factor to predict the prognostic survival rate of COAD (Fig. 2Q-2T).

Nomogram
By integrating factors such as age, sex, cancer stage, and risk score of patients, we constructed a nomogram that could predict the 1-year, 3-year, and 5-year survival rates of patients (Fig. 3A).For each patient, the overall score could be calculated according to the scores of the four factors of age, sex, cancer stage, and risk score.Finally, the survival rate could be obtained according to the score and the horizontal axis of the survival rate.The AUC values of the ROC curve of 1 year, 3 years, and 5 years were all greater than 0.8, indicating that the constructed nomogram had good prediction accuracy.The calibration curve of the ROC curve also illustrated this point.

Functional analysis of the prognostic risk model lncRNA
The analysis between the risk score and the expression of lncRNA showed a positive correlation between AC003101.2,LINC02657, AL161729.4,AP006621.2,AC156455.1,ZKSCAN2-DT, AC245041.1,and the risk score.The higher the level of expression, the higher the risk score.This implied that these lncRNAs might be involved in the occurrence and development of COAD (Fig. 4A-4G).The results of survival analysis showed that L161729.4,AP006621.2,and AC156455.1 were signi cantly related to the survival of patients, and the higher the expression of lncRNA, the lower the survival rate of COAD (Fig. 4H-4N).The full length of AL161729.4was only 476 bp with only one transcript; therefore, we determined it as the target of follow-up research.The results of subcellular localization showed that AL161729.4mainly existed in the cytoplasm in GM12878, MCF7, and other cells (Fig. 4O).The subcellular localization results of the prediction model showed that about 40% of lncRNA AL161729.4was located in the cytoplasm (Fig. 4P).The results of subcellular localization showed that the knockdown vector designed and constructed for lncRNA could decrease the content of AL161729.4 in vivo.Then, the function of lncRNA AL161729.4was examined.

AL161729.4-miR-760-HYOU1 was involved in regulating the PI3K/Akt signaling pathway
In the detection of the knockdown e ciency of the knockdown vector, it was found that the knockdown effect of AL161729.4siRNA 1# was signi cantly better than that of siRNA 2# (Fig. 7A), and the knockdown effect of HYOU1 siRNA 2# was better than that of siRNA 1# (Fig. 7B).In subsequent studies, the knockdown vectors with better effects were used for experiments.MiR-760 mimics could signi cantly reduce the mRNA levels of AL161729.4,HYOU1, PI3K, and Akt (Fig. 7C), and miR-760 inhibitors could increase their mRNA levels (Fig. 7D).

Discussion
The poor prognosis caused by the inability of the early diagnosis of COAD has always been a medical challenge.Hence, more comprehensive and accurate diagnostic markers and therapeutic targets are urgently needed [21][22][23].m6A modi cations regulate the occurrence and development of a variety of cancers through lncRNAs and miRNAs [24][25][26][27].For example, m6A writer-METTL14 suppresses the proliferation and metastasis of colorectal cancer by downregulating oncogenic long noncoding RNA XIST [28].m6A writers-METTL3 and eraser-ALKBH5 promoted the invasion and metastasis of cancer cells [29,30].However, the current research on the occurrence and regulation of COAD mediated by the mRNA-miRNA-lncRNA regulatory network based on m6A modi cations has not been in depth.In this study, we used bioinformatics methods to construct a new mRNA-miRNA-lncRNA regulatory network based on m6A modi cations, and experimentally veri ed how it regulated the occurrence and development of COAD.
Through single-factor Cox regression analysis, we determined that 11 m6A-related lncRNAs were signi cantly related to the prognosis of COAD, and further used LASSO regression analysis to nd seven most critical lncRNAs.The prognostic risk model and the nomogram constructed based on this were also veri ed by survival analysis, ROC curve, and calibration curve.We constructed the HYOU1-AL161729.4-miR-760regulatory network through expression difference analysis, survival analysis, and target combination prediction.HYOU1, AL161729.4,and miR-760 were all signi cantly related to the survival of COAD or a signi cant difference in expression existed between patients with COAD and normal samples.
The protein encoded by HYOU1 belongs to the heat shock protein 70 family.It is found to be highly expressed in a variety of tumors and is associated with tumor aggressiveness and poor prognosis [31][32][33][34][35]. AL161729.4has 476 nucleotides [36], which is suitable for constructing an overexpression vector.It is mainly located in the cytoplasm [37,38], and hence it is also convenient for knockdown vectors to knock it down.Currently, no report exists on the function of AL161729.4.It is a brand new lncRNA with unknown function.Studies have shown that microRNA-760 can inhibit the proliferation and invasion of colorectal cancer cells through the PTEN/Akt signaling pathway [39].Studies have shown that HYOU1 promotes cell growth and metastasis by activating PI3K/Akt signals and leads to poor prognosis [40].Combined with the targeting relationship between HYOU1-AL161729.4-miR-760,we predicted that the HYOU1-AL161729.4-miR-760regulatory network will regulate the occurrence of COAD through the PI3K/Akt signaling pathway.
We constructed overexpression vectors and knockdown vectors of HYOU1, AL161729.4,and miR-760 through genetic engineering and chemical synthesis.The experimental results of the luciferase reporter gene proved a targeted binding effect among the three.The results of qPCR experiments further con rmed the reliability of the HYOU1-AL161729.4-miR-760regulatory network and its relationship with the PI3K/Akt signaling pathway.Finally, the cell proliferation experiment con rmed that the HYOU1-AL161729.4-miR-760regulatory network was involved in the proliferation regulation of SW620 cells.
Our study involved bioinformatics analysis.We constructed a COAD prognostic risk model and nomogram based on m6A-related lncRNA and experimentally veri ed the latest mRNA-miRNA-lncRNA regulatory network.However, the ndings were still inadequate.First, because of the research conditions, we only explored the effect of the HYOU1-AL161729.4-miR-760regulatory network on the proliferation of SW620 cells, and did not conduct research on invasion and in ltration.In addition, the effect of the HYOU1-AL161729.4-miR-760regulatory network on the occurrence and development of COAD has not been investigated in vivo.

Conclusions
In this study, we integrated the COAD gene, lncRNAs, miRNAs, and clinical information to construct a risk model and nomogram that could accurately predict the prognosis of patients with COAD.The HYOU1-AL161729.4-miR-760 regulatory network was constructed, and experiments proved that it regulated the proliferation of SW620 cells by mediating the PI3K/Akt signaling pathway. Figures

Figure 2 Construction
Figure 2

Table 1 :
Our study provided new biomarkers for the early diagnosis of COAD and also a new target for further in-depth study of the occurrence and development of COAD.Declarations 37. Mas-Ponte D, Carlevaro-Fita J, Palumbo E, Hermoso Pulido T, Guigo R, Johnson R. LncATLAS database for subcellular localization of long noncoding RNAs.RNA.2017;23(7):1080-7. 3 .Cao Z, Pan X, Yang Y, Huang Y, Shen HB.The lncLocator: a subcellular localization predictor for long non-coding RNAs based on a stacked ensemble classi er.Bioinformatics.2018;34(13):2185-94. 39.Li X, Ding Y, Liu N, Sun Q, Zhang J. MicroRNA-760 inhibits cell proliferation and invasion of colorectal cancer by targeting the SP1-mediated PTEN/AKT signalling pathway.Mol Med Rep. Sequences of the primers used for this experiment.

Table 2 :
Sequences of the primers used for this experiment.