LncRNA KCNQ1OT1 enhanced the methotrexate resistance of colorectal cancer cells by regulating miR‐760/PPP1R1B via the cAMP signalling pathway

Abstract We aimed to explore the mechanism of the KCNQ1OT1/miR‐760/PPP1R1B axis acting to regulate methotrexate (MTX) resistance of colorectal cancer (CRC). Differentially expressed mRNAs and lncRNAs in MTX‐sensitive CRC cell lines and MTX‐resistant cell lines were determined through microarray analysis. Application of bioinformatics analysis was aimed to uncover the relationships among the lncRNAs/miRNAs/mRNAs, and to demonstrate the effects of cAMP signalling pathway in MTX‐resistant CRC. The expression level of RNA and proteins was, respectively, detected using qRT‐PCR and Western blot assays, whereas the dual‐luciferase reporter gene assay was implemented to verify the targeted relationship. The influence of the lncRNA/miRNA/mRNA axis on biological functions of MTX‐resistant cells and on the growth of tumours determined through both vitro and vivo experiments. LncRNA KCNQ1OT1 and PPP1R1B mRNA were overexpressed in MTX‐resistant CRC tumour cells. KCNQ1OT1 functioned as a sponge of miR‐760, which targeted PPP1R1B. Knockdown of KCNQ1OT1 enhanced chemosensitivity towards MTX through the sponging of miR‐760. MiR‐760 expressed at low levels targeted PPP1R1B in the activated cAMP signalling pathway under MTX treatment. Knockdown of KCNQ1OT1 dampened the proliferation of MTX‐resistant (HT29/MTX) cells by regulating the miR‐760/PPP1R1B axis, which also induced cell cycle arrest together with apoptosis. KCNQ1OT1 regulated the expression of PPP1R1B and the downstream genes CREB and CBP in the cAMP signalling pathway. MTX showed a suppressive function on CRC progression. KCNQ1OT1 enhanced the MTX resistance of CRC cells by regulating miR‐760‐mediated PPP1R1B expression via the cAMP signalling pathway.

structure of oncogene and gene pathways. 5 Unfortunately, there is insufficient information on the specific mechanism of the regulation process. To improve therapies for CRC and unclose the underlying molecular mechanism of drug resistance in CRC is necessary. 6 As one of the earliest developed cytotoxic drugs, methotrexate (MTX) is a chemotherapy agent and an immune system suppressant which functions as an antimetabolite and antifolate therapy for treatment of various human cancers. However, resistance to MTX brings challenges to the clinical application in cancers. Therefore, determining the mechanism of MTX resistance in cancers is extremely urgent in the further investigations and beneficial to studies on how to overcome drug resistance to assure the efficacy of MTX therapies. 7 Long noncoding RNAs (lncRNAs) are considered noncoding RNAs whose length extends beyond 200 nucleotides, and they lack protein-coding functions. 4 Increasing evidence indicates that lncRNAs are deeply involved in human cancers through the regulation of diverse cellular functions and biological processes including epigenetic modification and mRNA stabilization. 8 LncRNAs also function as competing endogenous RNAs (ceRNAs) that sponge microRNAs (miRNAs), by which they influence the expression of target messenger RNA (mRNA). 2 Chemoresistance to anticarcinogens that is associated with this process has been found in abnormal cancer tissues. 9 KCNQ1OT1, an lncRNA connected with tumorigenesis and the development of cancers such as Wilms' tumour and hepatocellular carcinoma, 10,11 was observed to be highly expressed in human CRC cells in a recent report. 12 However, the detailed mechanism by which KCNQ1OT1 regulates drug resistance in CRC is still unclear.
MicroRNAs are classified as highly conserved single-stranded noncoding RNA molecules that down-regulate gene expression by establishing sequence-specific linkages with the 3′ untranslated regions (3′ UTRs) of homologous mRNA targets. 13 MiRNAs are always aberrantly modulated in tumorigenesis and malignant transformation, acting as a part of the pathogenesis of cancer together with lncRNAs, as mentioned above. 14 In terms of CRC cells, some altered miRNAs have been verified to control proliferation, metastasis, apoptosis and chromosomal instability. 13,15 In particular, miR-760 affects the proliferation and invasion of CRC cells by involving in corresponding signalling pathways. 16 However, its precise mechanism is still uncovered.
PPP1R1B is a gene which encodes protein phosphatase 1 regulatory subunit 1B. It performs a crucial role in brain functions. Additionally, the downstream proteins of PPP1R1B have been reported to be overexpressed in numerous cancers, such as oesophageal, gastric, colon, prostate, and breast cancers. 17 This gene has an alias called DARPP32, which is a dopamine and cyclic AMP-regulated phosphoprotein. 18 The phosphorylation or dephosphorylation of DARPP32 induced by cAMP signalling is still disputed. Furthermore, former studies on PPP1R1B have focused more on neurological and psychiatric disorders than on cancer, especially CRC, offering us a novel research direction.
Based on previous studies and our assumptions, the present experiments were devised to explore the regulatory mechanism of the KCNQ1OT1/miR-760/PPP1R1B axis in MTX-resistant CRC via the cAMP signalling pathway. This exploration may contribute to the discovery of new therapeutic targets and prognostic factors for CRC in the future.

| Microarray analysis
LncRNA and mRNA expression profiles from 3 pairs of MTX-sensitive and MTX resistant CRC cells were selected from the Gene Expression Omnibus database (https://www.ncbi.nlm.nih.gov/geo/) (Series Accession Number GSE16066; Platform GPL570). Differentially expressed mRNAs and lncRNAs were sifted according to a threshold of | log 2 (fold change) | > 1 and adjusted P-value <0.05. Analyses was performed with R version 3.4.3 (https://www.r-project.org/) and the program package "Limma," and the results are presented in a heatmap (green for RNAs with low expression and red for RNAs with high expression).

| Coexpression network for lncRNAs and mRNAs
A coexpression network for differentially expressed mRNAs and lncRNAs was built using Cytoscape version 3.5.1 (http://www.cy toscape.org/). Node and edge files were produced by R according to the thresholds of Pearson coefficient factor >0.7 and P < 0.05. The solid line indicates a positive correlation between lncRNA and mRNA, whereas a dotted line indicates a negative correlation.

| Gene set enrichment analysis
Gene set enrichment analysis (GSEA) was set about using GSEA software version 3.0. Significantly enriched signalling pathways were analysed based on the differentially expressed mRNAs and on data from the Kyoto Encyclopedia of Genes and Genomes database (threshold P < 0.05). Dot plots and ridge plots of the enriched pathways were generated using R with the "ggplot2," "easygplot2," and "clusterProfiler" program packages.

| Western blot
Radio immunoprecipitation assay lysis buffer (Sigma-Aldrich) was employed for protein extraction. Then, the proteins were were added to the membranes, and the membranes were then incubated at 4°C overnight. After being using TBST to wash three times the membranes were subsequently incubated for 2 hours at 37°C, with goat anti-rabbit IgG H&L secondary antibody (1:2000).
The immunological reaction was detected with ECL Plus reagent (Amersham, Piscataway, NJ, USA). ImageJ 1.8.0 software was conducted to analyse data. β-actin was employed as an internal reference protein.

| Statistical analysis
The experiments mentioned above were repeated at least three times. GraphPad Prism 6.0 software (GraphPad Software, La Jolla, CA, USA) was used for data analysis. The data presented are the mean values (± SD) of three independent experiments. Differences between two groups were evaluated though Student's t test, whereas differences among multiple groups were calculated using ANOVA. P < 0.05 was deemed statistically significant.  Figure 1A,B). Furthermore, a coexpression network of differentially expressed lncRNAs and miRNAs indicated that KCNQ1OT1 was closely associated with PPP1R1B ( Figure 1C). Subsequently, specific miRNA targets were identified with TargetScan and miRanda, suggesting that the existence of binding sites between miR-760 and KCNQ1OT1/PPP1R1B ( Figure 1D). Additionally, some crucial path-

| LncRNA KCNQ1OT1 was markedly overexpressed in MTX-resistant CRC cells and tissues, and silencing of KCNQ1OT1 suppressed MTX-resistant CRC cell survival and proliferation
qRT-PCR analysis revealed that lncRNA KCNQ1OT1 was markedly

| Silencing KCNQ1OT1 inhibited the proliferation of MTX-resistant CRC cells through the sponging of miR-760
The targeted relationship between KCNQ1OT1 and miR-760 was con- The cell cycle stage was detected by flow cytometry in HT29/MTX cells to explore the effect of the miR-760/PPP1R1B axis. *P < 0.05 and **P < 0.01 compared with the NC group, #P < 0.05 compared with the PPP1R1B group. (B, D) The cell cycle stage was determined by flow cytometry in Caco2/ MTX cells to explore the effect of the miR-760/PPP1R1B axis. *P < 0.05 compared with the NC group, #P < 0.05 compared with the PPP1R1B group. (E, G) Cell apoptosis was detected by flow cytometry in HT29/MTX cells to explore the effect of the miR-760/PPP1R1B axis. *P < 0.05 and **P < 0.01 compared with the NC group, ##P < 0.01 compared with the PPP1R1B group. (F, H) Cell apoptosis was detected by flow cytometry in Caco2/MTX cells to explore the effect of the miR-760/PPP1R1B axis. *P < 0.05 compared with the NC group, #P < 0.05 compared with the PPP1R1B group PPP1R1B group, whereas the opposite effect was observed in the HT29/MTX + si-PPP1R1B group ( Figure S1G). In addition, the effects of PPP1R1B on the resistance to MTX were similar in the Caco2/ MTX cells ( Figure S1H). After transfection with si-PPP1R1B or miR-760 mimics, the mRNA and protein expression of PPP1R1B were significantly suppressed. Compared with the levels in the PPP1R1B group, the PPP1R1B mRNA and protein levels in the miR-760 mimics + PPP1R1B group were clearly repressed ( Figure 7E-H). Thereafter, Western blot and qRT-PCR were utilized to reveal the influence of the miR-760/PPP1R1B axis on the PPP1R1B-mediated downstream expression of CREB and CBP in the cAMP signalling pathway. As shown in Figure 7E

MTX-resistant CRC cells together with inducing cell cycle arrest and apoptosis by regulating PPP1R1B
The

MTX-resistant CRC tumour growth in nude mice
The impact of KCNQ1OT1 on MTX-resistant CRC tumour growth and related proteins in the cAMP signalling pathway was examined with a tumour xenograft assay. As demonstrated in Figure 9A-C, the tumour mass volume and the tumour mass weight of the nude mice injected with HT29/MTX cells were much larger than those of mice injected with HT29 cells (P < 0.01  Figure 9D). Finally, Western blot analysis revealed the influence of KCNQ1OT1 on protein expression in the cAMP signalling pathway. As demonstrated in Figure 9E, Dysregulation of miR-760 has been found in tumour tissues. On the one hand, overexpression of miR-760 accelerates the proliferation of ovarian cancer cells. 25 On the other hand, miR-760 suppresses breast cancer cell multiplication and metastasis 26 and is down-regulated in the medulla osmium and primary tumour of terminal gastric cancer patients. 27 In terms of CRC, research has shown that miR-760 is closely connected with the processes of proliferation and invasion in CRC cells through the targeting of downstream mRNAs through related signalling pathways 16 ; it is also down-regulated in CRC cells and might serve as a prognostic biomarker. 28 Moreover, miR-760 mediates chemoresistance in cancer. 29 Consistent with previous findings, miR-760 acted as a tumour suppressor in CRC and that low miR-760 expression was shown in our study, and it may be involved in cancer cells' resistance to oncology drugs, including MTX.
PPP1R1B, together with its downstream proteins, is overexpressed in diverse adenocarcinomas, including colon cancer. Experiments and statistical analyses have been implemented to verify this argument and have discovered that PPP1R1B promotes chemoresistance by regulating pro-oncogenic signal transduction pathways. 17 MiRNAs down-regulate gene expression through sequence-specific interactions with mRNA targets. 13 In our study, the results indicated that miR-760 could target PPP1R1B by interacting with highly con-

CONF LICT OF I NTEREST
No conflict of interest exits in the submission of this manuscript.

AUTHOR CONTRIBU TI ON
Substantial contribution to the conception and design of the work: