Integrated Analysis of lncRNA–Mediated ceRNA Network in type 2 diabetes

doi:10.21203/rs.3.rs-4072483/v1

Download PDF

Research Article

Integrated Analysis of lncRNA–Mediated ceRNA Network in type 2 diabetes

https://doi.org/10.21203/rs.3.rs-4072483/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background:

The lncRNA–miRNA–mRNA ceRNA network has been theorized to play an indispensable role in many disease,however, the function and regulatory mechanisms of ceRNAs associated with lncRNA in diabetes remains unclear.We predict the key lncRNA involved in the peripheral blood ceRNA mechanism of type 2 diabetes by correlation analysis and constructing a lncRNA-miRNA-mRNA network, to discover new diabetes markers or therapeutic targets.

Methods

The expression profile of differential lncRNA in peripheral blood of type 2 diabetes was detected by gene chip technology. Then use R language and bioinformatics tools to process chip data, predict the target gene by correlation analysis and construct lncRNA-miRNA-mRNA network. Then perform KEGG pathway analysis and GO enrichment analysis with lncRNA, and predict key lncRNA.

Results

Correlation analysis obtained 2016 pairs of relationship, including 125 lncRNA and 163 mRNA. KEGG pathway and GO enrichment analysis show that there are multiple pathways which related to the occurrence and development of type 2 diabetes. The lncRNA-miRNA-mRNA network was successfully constructed according to the results of the chip and predicted data, including 21 miRNAs, 12 mRNAs, 82 lncRNAs and 187 interaction pairs. The prediction tools screened out 6 key lncRNAs.

Conclusion

LncRNA may mediate the occurrence and development of diabetes by the ceRNA mechanism, and its key lncRNA may become a new diabetes screening marker or therapeutic target in the future.

Type 2 diabetes mellitus

LncRNAs

Competitive endogenous RNA network

Bioinformatics analysis

Type 2 diabetes mellitus (T2DM) is a metabolic diseases characterized by hyperglycaemia, which is known as an important predisposing factor in the development of DM^{[1, 2]}.Type 2 diabetes is a complicated metabolic disease affecting millions of people worldwide^[3].Type 2 diabetes and its complications constitute a major worldwide public health problem with high rates of diabetes-related morbidity and mortality^[3–5]. In 2021, the number of DM patients has exceeded 530 million, which is expected to reach 700 million by 2045^[6] .According to previous research, the prevalence of type 2 diabetes in China was 10.1% from 2010 to 2014, while the prevalence of the 65–74 age group was 14.1%^[7].

In recent years, non-coding RNA(ncRNAs),which lack protein coding has attracted attention as a common phenomenon that affects many cellular processes^[8–10]. NcRNAs include microRNAs (miRNAs, 21–24 base pairs) and long non-coding RNAs (lncRNAs, longer than 200 base pairs)^[10–11]. Previous studies have found that

the ncRNAs could regulated up or down endothelial function in the vasculature, which is associated with the occurrence of diabetes^[12]. Due to its high stability in body fluids (urine, plasma, exosomes, etc.) and the development of new detection technologies, ncRNA has been recognized as a new biomarker for diagnosis, prognosis, and prediction of treatment response^[13].

At present, the commonly used method for studying the function of lncRNA is to infer the function of lncRNA through known target gene functions. Target Gene prediction is the most commonly used prediction model mediated by miRNA and based on ceRNA regulation mechanism. This method constructs an lncRNA miRNA mRNA network based on the predicted results of various prediction tools, in order to connect lncRNA with target genes ^[14]. In addition, there are other commonly used prediction methods, such as judging based on gene expression correlation analysis or predicting nucleic acid sequences bound to lncRNA based on base complementary pairing rules. Currently, most prediction tools and databases are based on this method ^[15]. Up to now, researches on ceRNA in type 2 diabetes mostly focus on a single lncRNA miRNA pair, and there is no report on the peripheral blood ceRNA network in diabetes.

In this study, Pearson correlation analysis and the construction of lncRNA miRNA mRNA network were used to predict the key lncRNA in the peripheral blood of type 2 diabetes.The expression of specific lncRNAs was validated by quantitative real-time PCR (qRT-PCR) in T2DM. It is providing a basis for the study of the molecular mechanism in the development and treatment of diabetes.

Subjects

80 patients with type 2 diabetes diagnosed in Xiaqiao Hospital, Jiawang District, Xuzhou City in 2018 were enrolled as the case group, and 50 healthy examinees matched by age and sex were enrolled as the control group. Collect peripheral blood as samples, select 8 cases (4 cases, 4 controls) for chip detection, and use the remaining samples for fluorescence PCR validation ^[24]. The chip is the Affinemetrix Human oeLncRNAs gene chip produced by Agilent company. The original data is processed by Genespring software, the transformed data is read by the Affy package of R language, and the standardized pre-processing is performed by RMA method.

Experimental methods

By utilizing the miranda tool to predict lncRNA miRNA information, the tool uses an algorithm similar to Smith Waterman to score the base complementarity of nucleic acids and screen out relationship pairs with scores greater than 140. Predict miRNA mRNA relationship pairs in the miRWalk tool, which can be correlated with other database results for comparison. Add seven databases, miRWalk, mirbridge, miRDB, RNA22, miRanda, miRMap, and Targetscan, to assist in predicting the list of assumed target genes for minRNA. Set the database search relationship as OR, and after the running, we obtain a list of predicted genes. Select genes that appear in either of the seven databases as predicted target genes.

Integrate the lncRNA miRNA relationship pairs predicted by miranda tool and the miRNA mRNA relationship pairs predicted by miRWalk tool, and combine them with the lncRNA mRNA co-expression relationship pairs screened by correlation analysis to construct a miRNA-lncRNA-mRNA network, also known as the ceRNA network. The lncRNAs screened through miranda tools are the key lncRNAs. Real time quantitative fluorescent PCR (qRT PCR) was used to validate the selected key lncRNAs. GAPDH was selected as the internal reference, with 3 wells per sample and conducted under the same experimental conditions.

We integrate the lncRNA miRNA relationship pairs predicted by miranda tool and the miRNA mRNA relationship pairs predicted by miRWalk tool, and combine them with the lncRNA mRNA co-expression relationship pairs screened by correlation analysis to construct a miRNA lncRNA mRNA network, also known as ceRNA network. The lncRNAs screened through miranda tools are the key lncRNAs. Real time quantitative fluorescent PCR (qRT PCR) was used to validate the selected key lncRNAs. GAPDH was selected as the internal reference, with 3 wells per sample and conducted under the same experimental conditions.

Principal component analysis and co-expression analysis

Principal component analysis was used on the data and create a three-dimensional graph of principal component analysis. Screening for differentially expressed genes, conducting t-tests on the expression levels of lncRNA and mRNA to obtain significant P values, and calculating the logarithmic value of fold change (FC). Screen for differential lncRNAs and mRNA using P value < 0.01 and | log2FC |>0.585 as the standard. Draw a volcanic map to display the data.

MRNA-incRNA co-expression analysis was used based on gene expression correlation, use Pearson correlation coefficient to measure the degree of correlation between lncRNA and target genes, calculating the Pearson correlation coefficient matrix, and conducting correlation testing. Conduct KEGG pathway and GO functional analysis on differentially expressed lncRNAs with over 20 targeted mRNA, and display the analysis results using bubble plots. The threshold for significant enrichment is P < 0.05.

Statistical analysis

R language was used for data analysis, with psych package as principal component analysis, and ggplot2 package as a three-dimensional map. The t-test was conducted using the Limma package and the Pheatmap package was used to create a volcanic map. The selected differential lncRNA and mRNA standardized fluorescence signal values were imported into the psych package for Pearson correlation coefficient, with a threshold r of 0.9 and P value < 0.01. KEGG pathway bubble diagram of key lncRNAs displayed through the clusterProfiler package. The data obtained from the qRT-PCR experiment were analyzed using the 2^−△△Ct method for relative quantification, and both validation and plotting were performed using GraphPad Prism software.

Microarray analysis

The case group and control group samples are distributed at relatively distant locations in three-dimensional space, while the spatial distribution of samples within each group is relatively concentrated, indicating that the chip data is detected through principal component analysis, and the experimental process is normal, and the data has authenticity and reliability.(Figure 1)

After screening, 38 differentially expressed mRNAs were obtained, of which 90 were upregulated and 148 were downregulated. There are also differences in lncRNA133[24]. After de-duplication and re-annotation, there are 129 remaining.(Figure 2)

Co-expression analysis and pathway analysis

According to the Pearson correlation coefficient matrix calculation method mentioned in the previous section, 30702 pairs of differential lncRNA mRNA relationships were obtained, including 129 lncRNAs and 238 mRNA pairs. Select those with | r |>0.9 and P < 0.01 for subsequent analysis. There are a total of 2016 pairs that meet the conditions, including 125 lncRNAs and 163 mRNA. Further analysis was conducted on the co-expression relationship of differential lncRNA mRNA, and a total of 28 lncRNAs with more than 20 target genes were identified. KEGG pathway and GO functional enrichment analysis were performed on these 28 lncRNAs.(Figure 3)

The horizontal axis represents lncRNA, the number of enriched gene entries in parentheses。 A: The vertical axis is the name of the KEGG pathway ;B༚The vertical axis is the name of the GO enrichment analysis

Construction of ceRNA network

According to the method section, six potential miRNA lncRNA regulatory relationships were identified using miranda and miRWalk prediction tools, including 5 miRNAs and 6 lncRNAs; 22 potential miRNA mRNA regulatory relationships obtained includes 21 miRNAs and 12 mRNA; And 159 pairs of lncRNA mRNA relationships. Including the 2016 pairs of relationship pairs from the aforementioned lncRNA mRNA correlation analysis, take the intersection of these three results to construct a ceRNA network. It contain 115 nodes,including 21 miRNAs (gray triangle), 12 mRNA (circular, dark green down regulated, red up regulated), and 82 lncRNAs (diamond shaped, light green down regulated, pink up regulated). There are another 187 relationship pairs, with arrows representing miRNA mRNA regulation relationships, T-shaped arrows representing mRNA lncRNA co expression relationships, and solid red lines representing lncRNA miRNA prediction relationships.(Figure 4)

Key lncRNA screening and qRT PCR validation

Six key lncRNAs were screened using Miranda tool, with one upregulated and five downregulated. (Table 1) Display the KEGG pathway and GO functional bubble diagram of key lncRNAs using R language.(Fig. 5)

Four of them were selected for qRT PCR experimental verification, and the results showed that their expression in the peripheral blood of diabetes patients was significantly different from that of the control group. One of them was up-regulated, and three of them were down regulated, with P < 0.05. (Fig. 6)

Table 1

Key lncRNA
lncRNA	P value	Fold change	regulated
lnc-AL901608	0.001910364	2.116334	Down
lnc-BAI1-6:1	0.016691843	2.734740	Down
NONHSAT120819	0.007929356	3.180285	Down
NONHSAT140069	0.022604343	2.003800	Down
NONHSAT197460	0.036472443	2.060125	Up
NONHSAT214000	0.047390923	2.053059	Down

The coding protein in human body only accounts for 1.2%, and the rest 24% and 75% are Intron and multiple non-coding RNAs between genes^[16], of which the non coding RNA with more than 200 nucleotides is long chain non coding RNA (lncRNA). A large number of experiments have confirmed that it is closely related to the occurrence and development of type 2 diabetes^[17–20]. Abhishek Suwal et al.^[26]found that NONRATT021972 has multiple functions in various diseases related to diabetes. Fan Yang et al.^[20] found that the expression of Kcnq1ot1 increased in myocardial cells induced by high glucose and heart tissue of diabetes mice, while inhibiting the expression of Kcnq1ot1 can inhibit cell apoptosis.

In the ceRNA mechanism, lncRNA plays a regulatory role as a competitive endogenous RNA of miRNA. There are miRNA response elements (MREs) present in various RNAs, and there is a competitive relationship between multiple RNA molecules that bind to the same MREs ^{[21, 22]}. At present, the competitive endogenous RNA (ceRNA) mechanism has been reported in various diseases. Bo Jia et al. ^[23] found that LINC00707 can act as a ceRNA for miR-370-3p to inhibit WNT2B expression. Arianna Mangiavacchi et al. found that endogenous linc-223 can act as a competitive endogenous RNA for miR-125-5p (carcinogenic miRNA in leukemia). Reducing linc-223 expression enhances the activity of miR-125-5p, leading to a decrease in interferon regulatory factor 4 (IRF4), which can inhibit the carcinogenicity of miR-125-5p ^[24]. The ceRNA mechanism is also applicable to type 2 diabetes. For example, NONRATT003679.2 and rat pancreatic islets induced by glycolipid toxicity β Cell damage is related, as it serves as a molecular sponge for miR-34a, regulating oxidative stress and cell apoptosis mediated by the target SIRT1 of miR-34a ^[25]. LncRNA HOTAIR can also act as a molecular sponge for miR-34a, activating miR-34a to activate SIRT1 expression ^[26].

In this study, the competitive endogenous RNA network of type 2 diabetes peripheral blood lncRNA was constructed by bioinformatics for the first time. A total of 238 differentially expressed mRNA and 133 differentially expressed lncRNAs were screened based on the chip results. The Pearson correlation coefficient was calculated to obtain 2016 pairs of relationship pairs, 125 lncRNAs, and 163 mRNA. After software processing, 21 miRNAs, 12 mRNA, 82 lncRNAs, and 187 interaction pairs were obtained, and an lncRNA miRNA mRNA network was constructed. The lncRNA miRNA information in this network is predicted using miranda tools, which use sequence matching degree and minimum free energy to construct a scoring matrix. The miRNA mRNA relationship is derived from a miRWalk database that can compare the predicted results of 12 tools. This ensures the credibility of the ceRNA network in this study.

This study obtained a total of 28 lncRNAs with over 20 target genes through Pearson correlation coefficient, and conducted KEGG pathway and GO enrichment analysis on these 28 differential lncRNAs. Among them, Peroxisome and PPAR signaling pathways may be related to glucose and lipid metabolism in diabetes[27,28]. PPAR γ Its synthetic ligand is an insulin sensitizer and has been used to treat type 2 diabetes ^[28]. Complement and coagulation cascade pathways are involved in the vascular inflammatory reaction of diabetes and play a role in the pathogenesis, clinical symptoms and vascular complications of diabetes ^{[29, 30]}. In addition, Aldosterone regulates sodium reabsorption and proximal tubule Bicarbonate recovery pathway may be related to diabetes nephropathy[31]. These pathways indicate that the differential lncRNAs obtained from correlation analysis have some relationship with type 2 diabetes.

The data also shows six key lncRNAs in the ceRNA network, with one upregulated and five downregulated. KEGG pathway and GO enrichment analysis of key lncRNAs include Peroxisome, PPAR signaling pathway, cortisol synthesis and secretion, complement and coagulation cascade. This shows the effectiveness of the network constructed by this research institute, and the key lncRNA obtained is representative, which fully shows that this ceRNA network plays a potential role in the occurrence and development of type 2 diabetes. It lays a foundation for the follow-up study of the molecular mechanism of type 2 diabetes, and provides a theoretical basis for the discovery of new therapeutic targets and screening markers.

Although there are still shortcomings in this study, such as a small sample size. In future research, we will increase the sample size of the study in order to obtain more accurate results. Moreover, the balance and vitality of ceRNA networks may be influenced by various factors, including the abundance and subcellular localization of ceRNA components, the affinity between miRNAs and their molecular sponges, and so on ^[21]. In the next step, we will conduct case follow-up, collecting clinical information for further data analysis cell and animal experiments for more in-depth research on the core ceRNA network and its key lncRNAs.

Funding

This work was supported by a grant from the Xuzhou Medical Youth Innovation Project [grant number XWKYHT20200004 and XWKYHT20220126], the Xuzhou Science and Technology Project [grant number KC22226], and the Young Medical Science and Technology Innovation Project of the Xuzhou Municipal Health Commission [grant number XWKYHT20200030].The researchers had no relationships with the funder. The study funding had no influence on the study design,data collection, analysis, interpretation of data, writing of the report, or decision to submit the article for publication.

Author contributions

YW designed the study, acquired and analyzed data, and drafted and reviewed the manuscript.

XZ and YW conducted experiments, reviewed the manuscript, and contributed to the method.

ZD conceived and investigated the study, analyzed data, and reviewed the manuscript.

CQ and TL researched data, contributed to the discussion, and reviewed the manuscript.

PZ is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis, reviewed and edited the manuscript.

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical approval and consent to participate

This study was conducted with approval from the ethics committee of Xuzhou Center for Disease Control and Prevention. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study.

Consent for publication

Not applicable

Availability of data and materials

Not applicable

Wu Q, Liu Y, Ma YB, Liu K, Chen SH. Incidence and prevalence of pulmonary tuberculosis among patients with type 2 diabetes mellitus: a systematic review and meta-analysis. Ann Med. 2022;54(1):1657–66. PMID: 35703920; PMCID: PMC9225779.
Zheng Y, Ley SH, Hu FB. Global aetiology and epidemiology of type 2 diabetes mellitus and its complications. Nat Rev Endocrinol. 2018;14(2):88–98.
Zheng Y, Ley SH, Hu FB. Global aetiology and epidemiology of type 2 diabetes mellitus and its complications[J]. Nat reviews Endocrinol. 2018;14(2):88–98.
Wu Y, Ding Y, Tanaka Y, Zhang W. Risk factors contributing to type 2 diabetes and recent advances in the treatment and prevention. Int J Med Sci.
Wu Y, Ding Y, Tanaka Y, Zhang W. Risk factors contributing to type 2 diabetes and recent advances in the treatment and prevention. Int J Med Sci. 2014;11(11):1185–200. 10.7150/ijms.10001. PMID: 25249787; PMCID: PMC4166864.
Sun H, Saeedi P, Karuranga S, et al. IDF Diabetes Atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045[J]. Diabetes Res Clin Pract. 2022;183:109119.
Yang LL, Shao J, Bian YY et al. Prevalence of type 2 diabetes mellitus among inland residents in China (2000–2014): A meta-analysis. J Diabetes Investig 2016,7(6): 845–52.
Ramón y Cajal S, Segura MF, Hümmer S. Interplay between ncRNAs and cellular communication: a proposal for understanding cell-specific signaling pathways[J]. Front Genet. 2019;10:281.
Uchida S, Adams JC. Physiological roles of non-coding RNAs[J]. Am J Physiology-Cell Physiol. 2019;317(1):C1–2.
Geisler S, Coller J. RNA in unexpected places: long non-coding RNA functions in diverse cellular contexts[J]. Nat Rev Mol Cell Biol. 2013;14(11):699–712.
Bartel DP. MicroRNAs: target recognition and regulatory functions. Cell. 2009;136:215–33. 10.1016/j.cell.2009.01.002.
Zhang HN, Xu QQ, Thakur A, Alfred MO, Chakraborty M, Ghosh A, et al. Endothelial Dysfunction in Diabetes and Hypertension: Role of microRNAs and Long Non-Coding RNAs. Life Sci. 2018;213:258–68. 10.1016/j.lfs.2018.10.028.
Martinez B, Peplow PV. MicroRNAs in Blood and Cerebrospinal Fluid as Diagnostic Biomarkers of Multiple Sclerosis and to Monitor Disease Progression. Neural Regen Res. 2020;15(4):606–19. 10.4103/1673-5374.266905.
Yang F, Qin Y, Wang YQ et al. LncRNA KCNQ1OT1 mediates pyroptosis in diabetic cardiomyopathy. Cell Physiol Biochem 2018,50(4):1230–44.
Bridges MC, Daulagala AC, Kourtidis A. LNCcation: lncRNA localization and function. J Cell Biol. 2021;220(2):e202009045. 10.1083/jcb.202009045. PMID: 33464299; PMCID: PMC7816648.
Jarroux J, Morillon A, Pinskaya M. History,discovery,and classification of lncRNAs[J]. Adv Exp Med Biol. 2017;1008:1–46.
Goyal N, Kesharwani D, Datta M. Lnc-ing non-coding RNAs with metabolism and diabetes: roles of lncRNAs[J]. Cell Mol Life Sci. 2018;75(10):1827–37.
Hussein RM. Long non-coding RNAs: the hidden players in diabetes mellitus-related complications[J]. Diabetes & Metabolic Syndrome: Clinical Research & Reviews; 2023. p. 102872.
Zhang H, Xu Q, Thakur A, et al. Endothelial dysfunction in diabetes and hypertension: role of microRNAs and long non-coding RNAs[J]. Life Sci. 2018;213:258–68.
Yang F, Qin Y, Wang YQ et al. LncRNA KCNQ1OT1 mediates pyroptosis in diabetic cardiomyopathy. Cell Physiol Biochem 2018,50(4):1230–44.
Qi XL, Zhang DH, Wu N, et al. CeRNA in cancer: possible functions and clinical implications. J Med Genet. 2015;52(10):710–8.
Rehman OU, Uzair M, Farooq MS, et al. Comprehensive insights into the regulatory mechanisms of lncRNA in alkaline-salt stress tolerance in rice[J]. Mol Biol Rep. 2023;50(9):7381–92.
Jia B, Wang ZP, Sun X, et al. Long noncoding RNA LINC00707 sponges miR-370-3p to promote osteogenesis of human bone marrow-derived mesenchymal stem cells through upregulating WNT2B. Stem Cell Res The. 2019;10(1):67.
Arianna M, Melissa S, Silvia M, Oncotarget et al. 2016,7(37): 60155–68.
Kong X, Liu CX, Wang GD et al. LncRNA LEGLTBC functions as a ceRNA to Antagonize the Effects of miR-34a on the Downregulation of SIRT1 in Glucolipotoxicity-Induced INS-1 Beta Cell Oxidative Stress and Apoptosis. Oxid Med Cell Longev,2019,2019:4010764.
Gao L, Wang XF, Guo S, et al. LncRNA HOTAIR functions as a competing endogenous RNA to upregulate SIRT1 by sponging miR-34a in diabetic cardiomyopathy. J Cell Physiol. 2019;234(4):4944–58.
Guo H, Zhang Q, Yuan H et al. Nitric oxide mediates inflammation in type II diabetes mellitus through the PPARγ/eNOS signaling pathway[J]. PPAR research, 2020, 2020.
Wang SB, Edward JD, Robert LD. PPARγ signaling and emerging opportunities for improved therapeutics. Pharmacol Res. 2016;111:76–85.
Tan SM, Snelson M, Østergaard JA, et al. The complement pathway: New insights into immunometabolic signaling in diabetic kidney disease[J]. Antioxid Redox Signal. 2022;37(10–12):781–801.
Ramzi AA, Verena S. Role of complement in diabetes. Mol Immunol. 2019;114:270–7.
Lozano-Maneiro L, Puente-García A. Renin-angiotensin-aldosterone system blockade in diabetic nephropathy. Present evidences[J]. J Clin Med. 2015;4(11):1908–37.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Integrated Analysis of lncRNA–Mediated ceRNA Network in type 2 diabetes

Status:

Version 1

Abstract

Background:

Methods

Results

Conclusion

Figures

Introduction

Materials and Methods

Subjects

Experimental methods

Principal component analysis and co-expression analysis

Statistical analysis

Result

Microarray analysis

Co-expression analysis and pathway analysis

Construction of ceRNA network

Key lncRNA screening and qRT PCR validation

Discussion

Declarations

References

Additional Declarations

Status:

Version 1