The Impact of lncRNA on Diabetic Kidney Disease: Systematic Review and In Silico Analyses.

Background: Long noncoding RNA (lncRNA) is involved in the occurrence and development of diabetic kidney disease (DKD). It is necessary to identify the expression of lncRNA from DKD patients through systematic reviews, and then carry out silico analyses to recognize the dysregulated lncRNA and their associated pathways.
Methods: The study searched Pubmed, Embase, Cochrane Library, WanFang, VIP, CNKI, and CBM to find lncRNA studies on DKD published before March 1, 2021. Systematic review of the literature on this topic was conducted to determine the expression of lncRNA in DKD and non-DKD controls. For the dysregulated lncRNA in DKD patients, silico analysis was performed, and lncRNA2Target v2.0 and starBase were used to search for potential target genes of lncRNA. The Encyclopedia of Genomics (KEGG) pathway enrichment analysis was performed to better identify dysregulated lncRNAs in DKD and determine the associated signal pathways.
Results: According to the inclusion and exclusion criteria, 28 publications meeting the eligibility criteria were included in the systematic evaluation. A total of 3,394 patients were enrolled in this study, including 1,238 patients in DKD group, and 1,223 diabetic patients, and 933 healthy adults in control group. Compared with the control, there were eight lncRNA disorders in DKD patients (MALAT1, GAS5, MIAT, CASC2, NEAT1, NR_033515, ARAP1-AS2, and ARAP1-AS1). In addition, five lncRNAs (MALAT1, GAS5, MIAT, CASC2, and NEAT1) participated in disease-related signal pathways, indicating their role in DKD. Discussion. This study showed that there were eight lncRNAs in DKD that were persistently dysregulated, especially five lncRNAs which were closely related to the disease. Although systematic review included 28 studies that analyzed the expression of lncRNA in DKD-related tissues, the potential of these dysregulated lncRNAs as biomarkers or therapeutic targets for DKD remains to be further explored. Trial registration. PROSPERO (CRD42021248634).

1. Background

Diabetic kidney disease (DKD) refers to chronic kidney disease (CKD) caused by diabetes, which is one of the main microvascular complications of diabetes [1]. DKD has now become the main cause of chronic kidney disease and end-stage kidney disease in the world. It is also one of the main causes of death for diabetic patients. Diabetic kidney disease accounts for 40% of end-stage renal disease in the United States and Europe. According to the 2021 Atlas of the International Diabetes Federation, the global prevalence of diabetes among people aged 20–79 is estimated at 10.5% in 2021, rising to 783.2 million by 2045 [2]. Such a large number of diabetic patients means that there will be more diabetic kidney disease patients, and if diabetic patients have renal insufficiency, it will further aggravate microvascular dysfunction and significantly increase the risk of cardiovascular disease and cognitive decline. Therefore, it is urgent to investigate the mechanism of diabetic kidney disease and explore new therapeutic targets to delay the progression of the disease and improve the quality of life of patients.

The clinical manifestations of DKD are microalbuminuria, dominant proteinuria, and decreased glomerular filtration rate, which can eventually develop into end-stage renal disease [3]. The pathological features of DKD include glomerular mesangial expansion and hypertrophy, extracellular matrix (ECM) protein deposition, and podocyte apoptosis. Further glomerular sclerosis and tubular interstitial fibrosis will appear, which will eventually lead to renal failure [4, 5]. It is well known that DKD is caused by the interaction of environmental, genetic, and epigenetic factors in many aspects. lncRNAs exist stably in body fluids and can be detected. Recent studies have confirmed that some lncRNAs are important regulatory molecules involved in the occurrence and development of DKD, but their working mechanism in DKD is not very clear yet [6].

Noncoding RNAs (ncRNAs) are a type of RNA molecules that do not encode protein. They are important biological regulatory factors for the body, which regulates gene expression at the transcription and posttranscription level, and play a key role in both physiological and pathological processes. The development process of DKD is also affected. According to their length and function, ncRNAs can be divided into different subtypes, including long ncRNAs (lncRNAs), that is, ncRNAs with a length of more than 200 nucleotides, which do not encode proteins, but exhibit structural and functional heterogeneity. It is known that lncRNAs can be detected in both nuclei and cytoplasm. lncRNAs in nuclei can act on chromatin to regulate gene expression, and lncRNAs in cytoplasm can act on mRNA to regulate the translation process. lncRNAs exert their biological functions as signal molecules, bait molecules, guide molecules, and scaffold molecules by regulating transcription, translation, mRNA shearing, and posttranscriptional modification. lncRNAs regulate gene expression at the transcription, posttranscriptional, and epigenetic levels and participate in cell proliferation, differentiation, and apoptosis. Their abnormal expression is closely related to the occurrence and development of many diseases [7, 8].

It has been discovered that noncoding RNAs are involved in the progression of many diseases. The antisense lncRNA-HOX (HOX transcript antisense RNA, HOTAIR) is upregulated in the rat model of diabetic nephropathy [9], and lncRNA myocardial infarction-related transcripts (myocardial Infarction associated transcript, MIAT) are closely related to the onset of myocardial infarction [10]. lncRNAs may play a regulatory role in almost every gene expression stage, but their roles in the human body are still unknown [11].

Therefore, in order to further study which lncRNA may be involved in the pathogenesis of DKD and used as a potential biomarker of the disease, the study conducted a systematic review of the literature on this topic. In addition, bioinformatics analysis was performed to investigate the regulatory role of dysregulated lncRNA in the pathogenesis of diabetes mellitus (DM). The flowchart illustrates the detailed flow of the study design (Figure 1).

Figure 1

Detailed flow chart of the study design.

2. Methods

2.1. Search Strategies

This systematic review was carried out according to the current guidelines [12, 13], and the research protocol was registered on PROSPERO (https://www.crd.york.ac.uk/PROSPERO) with the identification No. of CRD42021248634. PRISMA(2020) checklist was in Supplemental File 1 (Supplemental File 1). English databases (Pubmed, Embase, and Cochrane Library) and Chinese databases (CBM (China Biomedical Database), CNKI (China National Knowledge Infrastructure), VIP (China Science and Technology Database), and Wanfang Database) were searched, and the studies on lncRNA in DKD published before March 1, 2021, were identified. To seek relevant clinical research, the terms below were searched: (“RNA long noncoding” or “untranslated RNA”) AND (diabetic nephropathy or diabetic kidney disease or diabetic renal disease). The literature search strategy is in Supplemental File 2 (Supplemental Files 2). Subject words and free words were combined, and different search strategies were taken for Chinese and English language databases. There were no restrictions on the status or language of the publication.

2.2. Inclusion and Exclusion Criteria

The original articles that analyzed the expression of lncRNAs in patients with DKD and those without DM or those with DM (controls) were enrolled. Studies without control group were excluded. Exclusion criteria: (1) review, nonclinical studies, and case observations; (2) meta-analysis, case reports, and editorials; (3) repeated studies; (4) research data are incorrect, incomplete, or unavailable, or at least one of the main results has not been reported; and (5) research on fuzzy outcome indicators.

2.3. Data Extraction

Two researchers conducted a comprehensive search of relevant databases and independently reviewed the research based on the inclusion criteria. The researchers deleted duplicate records, then screened the titles and abstracts of the remaining search results for relevance, and determined exclusion or further evaluation. If there were differences in the screening process, the two researchers would discuss together or with the help of a third party. The following data were extracted from each study: first author, publication year, research design, sample size, participant characteristics, lncRNA expression, quantitative method, tissue type analyzed, lncRNA expression, the expression profile of lncRNAs in the case group and control group.

2.4. lncRNAs Target Gene Acquisition and Analysis

Potential target genes of dysregulated lncRNAs in DKD were searched using lncRNA2Target v3.0 (https://bio-annotation.cn/lncrna2target/) [14] and starBase v2.0 (https://starbase.sysu.edu.cn/index.php) [15]. Networks of lncRNA-mRNA interactions were visualized in Cytoscape.

2.5. Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway Enrichment Analysis of lncRNA Target Genes

KEGG pathway enrichment analysis was performed on lncRNA target genes using the Database for Annotation, Visualization, and Integrated Discovery (DAVID), which revealed lncRNA-mRNA-related pathways. Statistical significance was reported using the Benjamini–Hochberg correction.

2.6. Protein-Protein Interaction (PPI) Network Construction and Identification of Hub Genes

The targets genes of lncRNA were put into STRING (https://string-db.org/cgi/input.pl) to construct the PPI network. Species were limited to Homo sapiens, confidence scores were limited to >0.9 and free spots were hidden. CytoHubba and CytoNCA are network topology analysis plug-ins in Cytoscape, which are used for topology analysis of PPI networks. Degree refers to the number of connections of nodes in the entire network, which reflects the interaction information between nodes. The value of Degree is used as a reference for the importance of the core goal. The target gene in the PPI network is used as a node, the line between the two nodes represents the relevant interaction, and the strength of interaction is represented by the color of the node. Hub genes were defined as genes that played an important role in the network. Cytoscape V3.8.2 was used to construct and visualize PPI networks.

2.7. Gene Ontology (GO) Functional Annotation and KEGG Pathway Analysis of Hub Genes

In this study, GO functional annotation and KEGG pathway analysis were performed through the ClueGO plugin in Cytoscape V3.8.2. All target gene names were corrected to their official gene symbols by entering target gene names and restricting the species to human. Statistical significance in GO terms and KEGG pathways was reported with Benjamini–Hochberg corrected p values, and p values <0.01 were considered significant.

2.8. Patients and Public Involvement

It was not appropriate to involve patients or the public in the research.

3. Results

3.1. Search Results and Characteristics of the Included Patients

According to the search strategy, a total of 128 articles were retrieved from the database; 84 duplicate articles were eliminated by the exclusion criteria, and after 16 preliminary screening of titles and abstracts, 28 articles still needed further review. The remaining articles were evaluated in full text, and 28 articles were included in the final analysis. A total of 3,394 patients were enrolled in this study, including 1,238 patients in DKD group, and 1,223 diabetic patients and 933 healthy adults in control group. The flow chart illustrates the literature screening process and results of this systematic review (Figure 2).

Figure 2

Literature screening processes and results. The literature screening process and results of the systematic review.

The number of lncRNAs with differential expressions between the different study groups and the control group varied from 1 to 858, and the number of samples in the study group ranged from 12 to 120. Among the 28 studies included in this systematic review, the most analyzed samples were serum, plasma, and kidney (Table 1).

Table 1

Features of lncRNAs expression studies included in the systematic review.

Author, year	Sample size		Organization	Methods	Country	Race	Total number of lnc RNA	Significant		References
	Case	Control						Up	Down
Fan et al, 2015	21 DN patients	19 controls/9 DM patients	Serum	lncRNA microarray	China	Mongolian race	858	45	813	[16]
Fan et al., 2017	DN patients	Controls/DM patients	Serum	qRT-PCR	China	Mongolian race	1	0	1	[17]
Bai et al., 2018	30 T2DM patients	13 controls	Kidney	Microarray	China	Mongolian race	1	0	1	[18]
Gao et al., 2018	111 DN patients	111 controls	Serum	qRT-PCR	China	Mongolian race	1	1	0	[19]
Wang et al., 2018	66 DN patients	56 controls/296 T2DM patients	Serum	qRT-PCR	China	Mongolian race	1	0	1	[20]
Yang et al., 2018	46DN patients	57 controls/36 T2DM patients	Serum	qRT-PCR	China	Mongolian race	1	0	1	[21]
Feng et al., 2019	30 DN patients	58 controls	Serum	qRT-PCR	China	Mongolian race	1	1	0	[22]
Jiao et al., 2019	33 DN patients	48 controls/43 T2DM patients	Serum	qRT-PCR	China	Mongolian race	1	1	0	[23]
Liu et al., 2019	14 DN patients	60 controls	Serum	qRT-PCR	China	Mongolian race	1	1	0	[24]
Majumder et al., 2019	12 DN patients	61 controls	Kidney	RNAscope	Canada	Caucasian race	1	1	0	[9]
Yang et al, 2019	32 DN patients	14 controls/28 DM patients	Kidney	qRT-PCR	China	Mongolian race	1	0	1	[25]
Yang et al., 2019	21 DN patients	19 controls/9DM patients	Serum	lncRNA microarray	China	Mongolian race	858	45	813	[26]
Dong et al., 2020	46 DN patients	42 controls/38 DM patients	Plasma	qRT-PCR	China	Mongolian race	1	0	1	[27]
Fan et al., 2020	42 DN patients	36 controls	Kidney/Peripheral blood/Urine	lncRNA microarray/qRT-PCR	China	Mongolian race	13	7	6	[28]
Fawzy et al., 2020	90 DN patients	90 controls	Serum	qRT-PCR	Saudi Arabia	Caucasian race	1	1	0	[29]
Ji et al., 2020	30 DN patients	30 controls	Kidney	RNA scope/qRT-PCR	China	Mongolian race	4	3	1	[30]
Li et al., 2020	40 DN patients	40 controls	Serum	qRT-PCR	China	Mongolian race	1	1	0	[31]
Qin et al., 2020	50 DN patients	50 controls/50 DM patients	Plasma	qRT-PCR	China	Mongolian race	1	1	0	[32]
Shen et al., 2020	30 DN patients	32 controls	Blood	qRT-PCR	China	Mongolian race	1	0	1	[33]
Zhang et al., 2020	66 DN patients	66 controls/66 DM patients	Plasma	qRT-PCR	China	Mongolian race	1	1	0	[34]
Zhao et al., 2020	60 DN patients	60 controls/77 T2DM patients	Serum	qRT-PCR	China	Mongolian race	1	1	0	[35]
Zhou et al., 2020	27 DN patients	14 controls/20 T2DM patients	PBMCs	qRT-PCR	China	Mongolian race	1	1	0	[36]
Petrica et al., 2021	88 DN patients	25 controls/48 T2DM patients	Serum/Urine	qRT-PCR	Romania	Caucasian race	4	2	2	[37]
Lv et al., 2017	21 DN patients	19 controls/9 DM	Serum	lncRNA microarray	China	Mongolian race	858	45	813	[38]
Zhou et al., 2021	27DN patients	14 controls/20 T2DM	Serum	qRT-PCR	China	Mongolian race	1	1	0	[39]
Li et al., 2019	60 DN patients	60 controls/77T2DM	Serum	qRT-PCR	China	Mongolian race	1	1	0	[40]
Liang et al., 2019	120 DN patients	120 controls/120 T2DM	Serum	qRT-PCR	China	Mongolian race	1	1	0	[41]
Chen et al., 2017	25 DN patients	9 controls/10 DM	Serum	qRT-PCR	China	Mongolian race	1	1	0	[42]

3.2. Differentially Expressed lncRNAs in DKD

Eight types of lncRNA disorders (MALAT1, GAS5, MIAT, CASC2, NEAT1, NR_033515, ARAP1-AS2, and ARAP1-AS1) in patients from two or more studies were selected for further evaluation, of which target gene information was found in five. Among them, lncRNA MIAT and lncRNA GAS5 had different trends of regulation. Therefore, this may be explained by differences in the types of tissues analyzed (serum, urine, and kidney) (Table 2).

Table 2

Differential expression of lncRNAs in at least two studies.

lncRNA	Author, year (reference)	Samples	Tissue	Express changes
lncRNA MALAT1	Zhou et al., 2021 [39]	DN 27	Serum	Up
	Zhou et al., 2020 [36]	DN 27	PBMCs	Up
	Petrica et al., 2021 [37]	DN 88	Serum/urine	Up
	Fawzy et al., 2020 [29]	DN 90	Serum	Up
	Ji et al., 2020 [30]	DN 30	Kidney	Up
lncRNA GAS5	Chen et al., 2017 [42]	DN 25	Serum	Up
	Fan et al., 2017 [17]	DN	Serum	Down
lncRNA MIAT	Ji et al., 2020 [30]	DN 30	Kidney	Up
	Petrica et al., 2021 [37]	DN 88	Serum/urine	Down
lncRNA CASC2	Wang et al., 2018 [20]	DN 66	Serum	Down
	Yang et al., 2018 [21]	DN 46	Serum	Down
lncRNA NEAT1	Petrica et al., 2021 [37]	DN 88	Serum/urine	Up
	Li et al., 2020 [31]	DN 40	Serum	Up
lncRNA-NR_033515	Fan et al., 2020 [28]	DN 42	Kidney/peripheral blood/urine	Up
	Gao et al., 2018 [19]	DN 111	Serum	Up
lncRNA-ARAP1-AS2	Fan et al., 2015 [16]	DN 21	Serum	Up
	Yang et al., 2019 [26]	DN 21	Serum	Up
	Lv et al., 2017 [38]	DN 21	Serum	Up
lncRNA-ARAP1-AS1	Fan et al., 2015 [16]	DN 21	Serum	Down
	Yang et al., 2019 [26]	DN 21	Serum	Down
	Lv et al., 2017 [38]	DN 21	Serum	Down

3.3. Related Target Genes of Differentially Expressed lncRNAs in Human Samples

Eight kinds of lncRNA disorders (MALAT1, GAS5, MIAT, CASC2, NEAT1, NR_033515, ARAP1-AS2, and ARAP1-AS1) were found through bioinformatics, of which target gene information was identified in five kinds of lncRNA (MALAT1, GAS5, MIAT, CASC2, and NEAT1). These five lncRNAs jointly regulated the expression of 2987 related target genes. MALAT1 had the largest number of target genes (1,316), followed by NEAT1 (1,000), GAS5 (566), and MIAT, CASC2, and the least number of targets (97 and 8, respectively). Among the 2,987 target genes, 1,924 were protein-coding genes, 506 were pseudogenes, 244 were small nuclear RNA (snRNA), and 313 were other types of ncRNA, including lncRNA, microRNA, rRNA, tRNA, and mitochondrial RNA (mtRNA) (Figure 3(a)).

Figure 3

Venn diagram. (a) Shared target genes of persistently dysregulated lncRNAs in DKD. (b) Pathways of shared target genes.

3.4. KEGG Enrichment Analysis of Related Target Genes

Next, in order to further explore the functional consequences of five lncRNA dysregulations of interest, the pathway map from the KEGG repository was used to perform a functional enrichment analysis of their protein-coding target genes. As a result, a total of 115 unique pathways were enriched for lncRNA target genes. In addition, as in the five lncRNAs (MALAT1, GAS5, MIAT, CASC2, and NEAT1), no common pathway had been found. Many of the 115 pathways had been confirmed to be involved in the pathogenesis of DKD, such as PI3K/Akt, TNF, HIF-1, AGE/RAGE, apoptosis, and FoxO (Figure 3(b) and 4).

Figure 4

Important KEGG pathway in DKD that may be regulated by persistently dysregulated lncRNAs. The size of the dots represents the number of genes, and the color represents the range of the pathway q-values. The y-axis represents the KEGG pathway, and the x-axis shows the five lncRNAs involved in each selected pathway. Q-value: p value adjusted for multiple testing using the Benjamini–Hochberg method.

3.4.1. lncRNA-mRNA Interaction

After data preprocessing and analysis of the two databases, the expression of 1,921 mRNAs from five lncRNAs coregulated by the databases were obtained (Figure 5, Table 3).

Figure 5

lncRNA-target gene interaction. Interaction network of lncRNAs and target genes. Red dots represent lncRNAs, green dots represent target genes, and blue dots represent intersection target genes.

Table 3

miRNA-mRNA network.

Symbol	Up/down	Count	Target mRNA
lncRNA MALAT1	Up	1140	ABCA1, ADAMTS12, ADGRL2, AKT1, BMPER, CA2, CASP3, CASP9, CCAR2, CCL2, CCND1, CCT4, CD36, CDCP1, CDH1, CDH1, CDH1, CDH1, CDH2, CDH5, CDKN1A, CDKN1B, CLDN5, COL6A1, CPM, CSF1, CTHRC1, CTNNB1, CTNNB1, CXCL5, DRD1, EZH2, EZH2, EZH2, EZH2, EZH2, FN1, FSD1, GPC6, HMGB1, HMMR, HNF4G, LAYN, LPAR1, LTBP1, LTBP3, LY6K, MAP2K1, MAP2K2, MAPK1, MAPK10, MAPK3, MAPK8, MAPK9, MCAM, MMP14, MMP2, MMP9, MMP9, MYBL2, MYC, NNMT, OCLN, PCNA, PIK3CB, PRKCE, PTBP3, PTK2, PXN, RAP1B, RASSF6, ROBO1, SFRP1, SFRP1, SNAI2, SNCA, SP1, STC1, TJP1, TP53, ZEB1, ZEB1, ZEB2, ZNF174, ZZEF1, ZNF678, ZNF646, ZNF518A, ZNF500, ZNF354B, ZNF282, ZNF277, ZNF232, ZNF224, ZNF207, ZNF202, ZNF175, ZNF160, ZNF142, ZMIZ1, ZKSCAN1, ZIK1, ZFYVE27, ZFP91-CNTF, ZFP91, ZFP36L1, ZER1, ZBTB41, ZBED1, YWHAZ, YWHAQ, YWHAE, YWHAB, YJEFN3, YBX2, XRCC6, WLS, WIZ, WDTC1, WDR74, WDR73, WDR27, WDR26, WBP11, WASHC5, WASHC2A, WASF2, WARS, WAC, VTI1B, VPS39, VPS25, VPS13C, VOPP1, VEGFB, VEGFA, VAPA, VAMP3, UTRN, UTP6, UTP4, UTP14A, URB2, UQCRH, UPRT, UNC5B, UHRF1, UCHL1, UBL7, UBL5, UBL3, UBE4A, UBE3C, UBE2V1, UBE2R2, UBE2M, UBE2L6, UBE2J1, UBE2H, UBE2G1, UBD, UBC, UBB, UBAP2L, UBA52, UBA1, U2SURP, TYSND1, TXNDC5, TUFM, TUBB4B, TUBB, TUBA4A, TUBA1C, TUBA1B, TUBA1A, TTL, TTC38, TTC3, TTC17, TSTA3, TSSC4, TSC22D3, TRRAP, TROVE2, TRNP1, TRIP13, TRIP10, TRAPPC1, TPT1, TPR, TPPP, TPI1, TOP2B, TOMM70, TOMM7, TOMM40, TOLLIP, TOGARAM1, TOB2, TOB1, TNS4, TNPO2, TNFAIP3, TMSB4X, TMPO, TMEM50A, TMEM259, TMEM248, TMEM245, TMEM206, TMEM199, TMEM189-UBE2V1, TMEM184C, TMEM132A, TMEM123, TMEM107, TMBIM6, TM9SF1, TLN1, TKT, TJP2, TJAP1, TIMM23, TIMM10B, TIGD5, TICAM1, THOC6, TGIF1, TFAP4, TEX261, TEAD3, TCTA, TCP1, TCOF1, TCL1A, TBL1XR1, TAX1BP3, TAX1BP1, TARS, TAOK3, TALDO1, TAGLN2, TAF1D, TACC3, TAB3, SZRD1, SYT7, SYNRG, SYMPK, SUMO2, SULF2, SUCLG1, STXBP3, STT3B, STRAP, STMP1, STMN1, STK38L, STAT1, SSR4, SRSF9, SRSF6, SRSF2, SRRT, SRP72, SRP68, SRGN, SRCIN1, SPEN, SPCS2, SP100, SOX8, SOX4, SORD, SNX13, SNURF, SNU13, SNRPN, SNRNP25, SMYD5, SMYD4, SMG1, SMDT1, SMARCE1, SMARCD2, SMARCA2, SMARCA1, SMAD2, SLMAP, SLK, SLCO4A1, SLC6A9, SLC44A2, SLC43A3, SLC41A1, SLC38A1, SLC29A3, SLC29A1, SLC25A6, SLC25A44, SLC25A3, SLC25A15, SLC25A11, SLC18B1, SLC17A3, SLC16A1, SLC12A8, SLC12A7, SLC12A6, SLAMF6, SKIV2L, SIPA1L1, SIM2, SIGMAR1, SHMT1, SHC1, SH3KBP1, SGCB, SFRP1, SF3B4, SF3A3, SETD5, SETD2, SERPINB6, SERINC2, SERF2, SENP1, SEMA6C, SEMA3F, SELL, SEC61A1, SEC23A, SEC11A, SDHC, SDHB, SDHAF2, SCPEP1, SCNM1, SCD, SCAMP1, SBF2, SARS2, SAP130, S1PR2, RUFY3, RTN3, RTF1, RSPRY1, RSC1A1, RRP36, RPUSD3, RPSA, RPS9, RPS6, RPS3, RPS26, RPS23, RPS20, RPS2, RPS18, RPS16, RPS13, RPS11, RPLP2, RPLP1, RPLP0, RPL9, RPL8, RPL7L1, RPL4, RPL39, RPL37A, RPL35, RPL32, RPL30, RPL3, RPL28, RPL24, RPL23, RPL22L1, RPL22, RPL18A, RPL17-C18orf32, RPL17, RPL15, RPL13A, RPL13, RPL12, RPL11, ROMO1, RNF41, RNF26, RNF213, RNF170, RNF167, RNF150, RNF111, RNASE6, RMC1, RIPOR1, RIMBP3, RIC8A, RHOA, RHBDF2, RGS10, RGPD2, RER1, REEP3, RCOR2, RCC2, RCAN3, RBM5, RBM33, RBM25, RBM19, RASSF5, RASA4B, RAPH1, RAPGEF1, RALGPS2, RALGAPA1, RAD51AP1, RACK1, RAC2, RABGGTB, RABAC1, RAB7A, RAB6A, RAB12, RAB11B, R3HDM2, QSOX1, QRICH1, PYM1, PYCR1, PXN, PXDN, PWWP2A, PWP2, PTPRF, PTPRD, PTPRCAP, PTPN11, PTPN1, PTMA, PTK2, PTDSS2, PTBP1, PSMF1, PSMD7, PSMD5, PSMD3, PSMC6, PSMC5, PSMB5, PSMA4, PSAT1, PSAP, PRTG, PRRC2B, PRR5-ARHGAP8, PRR14, PRPF8, PRPF19, PRPF18, PRMT2, PRKRIP1, PRKDC, PRKCD, PRC1, PPT1, PPP4R1, PPP2R5D, PPP2R1A, PPP2CA, PPP1R8, PPP1R12C, PPP1R12A, PPP1CB, PPOX, PPIL3, PPIL2, PPIG, PPIF, PPIA, PPARGC1B, PPA2, POU2F1, POTEJ, POM121C, POLRMT, POLR3E, POLM, POFUT1, PNMA8A, PNISR, PMPCB, PLXND1, PLXNA4, PLXNA1, PLIN3, PLEKHM1, PLEKHG3, PLEKHF2, PLEKHA1, PLCG2, PLAGL1, PKM, PKIG, PJA2, PIP5K1A, PHTF2, PHIP, PHF2, PGS1, PGM1, PGK1, PFN1, PDZD8, PDIA6, PDIA3, PDCD6IP, PDCD4, PDCD2, PCYOX1L, PCTP, PCSK9, PCLAF, PCDH11Y, PCBP4, PCBP2, PCBD1, PBRM1, PARP10, PAPD7, PAK1, PAFAH1B1, PABPC1, P4HB, P2RY11, P2RX5-TAX1BP3, OST4, OR2A25, OGT, OCLN, OBSL1, OAS3, OAS2, NUP62, NUP153, NUMA1, NUFIP2, NUDT21, NUCKS1, NUBP1, NT5C, NRXN1, NREP, NRDE2, NRBP1, NR2F2, NPDC1, NOTCH2, NOP9, NOC3L, NOA1, NIPSNAP1, NIN, NIF3L1, NGRN, NFXL1, NFE2L2, NFE2L1, NFATC2IP, NEK9, NEK7, NEK4, NEDD8, NECAP1, NDUFV1, NDUFB8, NDUFB11, NDOR1, NDFIP1, NCAPH2, NCAM1, NARS, NAPRT, NAGPA, NACA, MYO1E, MYH9, MYDGF, MYC, MXD3, MVB12B, MUC4, MUC19, MUC16, MTRNR2L8, MTRNR2L3, MTRNR2L12, MTRNR2L11, MTRNR2L1, MTPN, MT-ND5, MT-ND4L, MT-ND4, MT-ND2, MT-ND1, MT-CO3, MT-CO2, MT-CO1, MTCH2, MT-ATP8, MT-ATP6, MRPS6, MRPS25, MRPS16, MRPL44, MRPL43, MRPL42, MRPL40, MRPL33, MRPL2, MROH1, MRI1, MPZL1, MPZ, MON1B, MOB1A, MMGT1, MLLT11, MLF2, MLEC, MKRN1, MKNK2, MIOS, MIER1, MICAL3, MGST1, MGLL, MGEA5, MGAT3, METTL9, METTL4, METTL2B, MDK, MDH1, MDFIC, MCM7, MCM3, MCL1, MBNL1, MBD2, MATR3, MATN2, MAT2A, MARS, MARK3, MARCKSL1, MARCKS, MAPK14, MAP3K4, MAN2B1, MAN1B1, MAGT1, LSM4, LSM3, LSM2, LSM12, LRRN1, LRRC61, LRRC58, LRPPRC, LRIG3, LRIF1, LRCH1, LPXN, LONP1, LNPEP, LITAF, LINGO1, LIMD1, LIMA1, LHFPL2, LGALS9B, LENG8, LDHA, LDB1, LCP1, LBHD1, LARP1, LAPTM5, LAD1, KRT7, KRT5, KRT13, KRCC1, KPNA2, KMT2A, KLHL8, KIFC1, KIF3C, KIF1C, KIAA0040, KHSRP, KHDC4, KDM5B, KDELR2, KDELR1, KCNK1, KAT7, KAT6A, JUP, JUND, JRK, JPT1, JPH4, JADE2, ITPRIPL2, ITGB1, ITGA3, ISYNA1, ISOC1, IRF6, IRF4, IPO7, INTS8, INPP5D, IMPACT, IMMT, ILF3, ILF2, IKZF2, IGSF9, IGSF8, IGF2BP2, IGF2BP1, IGF2, IGF2, IGF1R, IGDCC3, IFI6, IFI30, IARS, HSPG2, HSPA9, HSPA8, HSPA4, HSP90AB1, HSF1, HSD17B10, HS6ST3, HOXD4, HOMER1, HNRNPUL1, HNRNPL, HNRNPK, HNRNPDL, HNRNPC, HNRNPA2B1, HNRNPA0, HMMR, HMGCS1, HMGCR, HMGA1, HM13, HLA-E, HLA-DRA, HLA-DMA, HLA-C, HLA-B, HLA-A, HIST2H3A, HIST1H3A, HIST1H2AE, HIST1H1C, HIPK2, HINT1, HIF1A, HIC2, HGSNAT, HEXDC, HEMK1, HEATR1, HDGFL3, HDGF, HDDC3, HCFC1, HBP1, HADH, GUCA1A, GTPBP6, GTPBP1, GTF3C5, GTF2I, GSTA4, GSPT1, GRWD1, GPR137, GPI, GPC6, GNL2, GNB1, GNAS, GLUL, GLUD2, GLG1, GINM1, GGT7, GFRA1, GDE1, GCDH, GATA3, GAPDH, GANAB, GALNT18, GABBR1, GABARAPL1, GAB2, G6PC3, FXYD6, FUS, FTL, FREM2, FRAT2, FRAS1, FP565260.1, FMNL2, FLOT2, FLNA, FKBP5, FIG4, FEN1, FCF1, FBXW11, FBXO30, FBXO22, FBLN1, FAR1, FANCD2, FAM193B, FAM117A, F11R, EZR, EXT2, ETFA, ETF1, ESPL1, ERMP1, ERG28, ERCC1, EPS15L1, EPM2AIP1, EPHB4, EPB41L2, EPB41, ENO1, EMC8, EMC6, EMC10, EMC1, ELOVL6, ELK4, EIF5A, EIF4G2, EIF4G1, EIF4EBP2, EIF4B, EIF4A2, EIF3L, EIF3E, EIF2S3B, EHMT1, EFR3A, EEF2, EEF1G, EEF1A1, EDARADD, ECHDC3, E2F4, DYNLT1, DUSP5, DUSP23, DUSP14, DTX2, DST, DSP, DSC3, DSC2, DPY19L1, DPP3, DOLPP1, DOCK3, DNM2, DNASE1, DNAJC14, DNAJB5, DNAJA3, DMTF1, DIP2A, DIAPH1, DHX15, DGUOK, DGKH, DGCR2, DGAT1, DEF8, DEAF1, DDX56, DDX5, DDX41, DDX23, DDX17, DDIT4, DDI2, DCTN4, DCTN3, DCTN1, DCAF7, DCAF13, DBT, DBN1, DAZAP2, DAD1, CYB5D2, CXCL14, CTSD, CTNND1, CTNNA1, CSTB, CSNK2B, CSNK2A2, CSNK1D, CS, CRTC2, CRKL, CRCP, CPT2, CPSF6, CPSF1, COX6B1, COX6A1, COX5A, COTL1, CORO1C, CORO1B, COPB1, COG4, COG3, COCH, COA1, CNPPD1, CNOT9, CNOT3, CNNM4, CNNM2, CNIH4, CLNS1A, CLN3, CLIC4, CLDN4, CLASP1, CKAP2L, CIC, CHTF18, CHD8, CHAF1B, CFL1, CENPV, CDKN1A, CDK4, CDK2, CDK12, CDHR1, CDCA3, CDC7, CDC25A, CD33, CD276, CCT3, CCNT1, CCNH, CCNG1, CCND2, CCNB2, CCL4L1, CCL3L1, CCL22, CCDC6, CCAR1, CC2D1B, CBX1, CBSL, CBS, CASD1, CASC3, CARHSP1, CAPNS1, CAPN1, CALU, CALR, CACNB4, C8orf37, C6orf62, C6orf48, C5orf24, C21orf59, C20orf204, C20orf194, C1orf43, C1orf226, C19orf54, C19orf48, C17orf62, C14orf119, C11orf98, BTG2, BTBD10, BSG, BRK1, BRD4, BPTF, BORCS8-MEF2B, BORCS8, BMP8B, BLOC1S5-TXNDC5, BIRC5, BHLHE40, BDH1, BCL2L2, BCAT1, BCAP29, BBS9, BBS2, BAG6, BAG3, BACH1, BABAM2, B4GALT1, B3GAT3, B2M, ATRN, ATP6V0E2, ATP5PO, ATP5MC2, ATP5F1D, ATP5F1C, ATP1A1, ATIC, ATG9A, ATG3, ASH2L, ASH1L, ARPP19, ARMCX3, ARL6IP5, ARHGEF2, ARHGEF12, ARHGAP8, ARHGAP5, ARHGAP45, ARHGAP21, ARHGAP17, ARHGAP12, ARF6, ARF3, ARF1, ARCN1, APRT, APPBP2, APLP2, AP5M1, AP2A2, AP1B1, AP003108.2, AP002990.1, AP000781.2, ANXA11, ANP32E, ANO5, ANO10, ANLN, ANKRD50, ANKRD46, ANKRD17, ANKRD10, ANAPC16, AMMECR1L, ALG2, ALDOA, ALDOA, ALCAM, AL591806.3, AL358113.1, AL157392.5, AL136295.4, AL136295.1, AL133352.1, AL021546.1, AKR7A2, AKR1A1, AKAP8L, AK2, AGPAT5, AFDN, ADSS, ADO, ADNP, ADIPOR2, ADIPOR1, ADGRG5, ADD3, ADAT1, ACTR3, ACTG1, ACTB, ACTA2, ACOT8, ACBD6, AC138894.1, AC124312.1, AC093668.2, AC091167.2, AC087632.1, AC073610.3, AC073508.2, AC068946.2, AC068831.7, AC068580.4, AC009336.2, AC008758.4, AC007192.1, ABT1, ABR, ABCD4, ABCC4, AARS
lncRNA GAS5	Up/down	362	YBX1, VIM, VEGFA, TP53, SMAD3, PTEN, PTEN, PPEF1, MMP2, IL10, IGF1R, GSTM3, FGF1, EIF4E, E2F1, CDKN1B, CDKN1A, CDK6, CDK6, CCND1, APAF1, ANXA2, SLC25A5, AP2B1, FARP2, TEAD3, STARD3NL, PPP5C, RUFY3, CD74, FAM136A, RAI14, PHPT1, BCAT1, RPL18, CLEC2D, ZFYVE26, CRMP1, HSP90AA1, RPS5, TPX2, RPLP0, GANAB, RGS1, CDV3, SNAP23, MSH2, ARCN1, SYDE2, DDT, EIF3L, PMM1, ERH, PSMB5, NFKBIA, WDR59, FZD3, EYA1, URI1, GPI, LSR, EIF3B, TBC1D9, CHORDC1, UBE4A, C11orf58, PPFIBP1, CORO1C, ATP5F1B, KRR1, HSPA9, DBN1, EEF1B2, ALMS1, CEP104, CAPZA1, MTR, RBBP5, CA14, VAMP8, RBM25, DNMT3A, TMPO, CRY2, RPL21, ARL4A, MKKS, DSTN, ZBTB1, EIF2S3, UBE2G1, H3F3B, ZRANB2, GPS2, ERAL1, CNDP2, WDR74, VRTN, GNL2, DZIP1, LCP1, GRHPR, PML, RPS11, RPS8, CD53, HDGF, SMYD2, TPM3, ARL6IP5, ATXN7L1, NDUFC2, TIAM1, EEF1A1, EPB41, SON, DFFA, EIF4A1, RPL26, PPP1R21, PIGR, FSTL1, PPP4R2, ANXA5, WASHC5, OTX2, NUDT5, NDUFB8, RPL27A, B2M, TUBA1A, PROSER3, C19orf48, NDUFV1, ATP5PD, ECI1, RPSA, RAB3B, PTK2, CLIC4, PA2G4, CFL1, ZNF621, TOMM20, AGFG1, CNP, CNTNAP2, ZNF354B, TAF7, CTC1, TMEM107, NRIP1, IGSF5, SELENOF, UBE2L3, PPP1CC, ARL4C, TUBB, PCBP2, HHLA3, RPS4X, RPL23A, CALM1, HIST2H2AA3, C6orf48, ARL2, VAMP2, CKMT1A, C19orf24, SLC7A2, PAFAH1B1, SPAG9, SYT7, SLC30A9, AQR, FHL1, TPR, HEXB, POLQ, SYNE2, CCAR1, ERBB3, RHOA, LAPTM4A, DRD4, MAPK6, TCOF1, WDR1, ENO1, ACTB, BAZ2A, GPC4, SRCAP, TXLNA, DNAJA1, PIR, STRN4, WDR7, MFSD11, SEC22C, DEPDC5, RPL3, AP4S1, GMPR2, TM9SF4, NOP56, ABCC1, IL21R, IMPAD1, LAPTM4B, EIF3E, GSR, AES, TNNT1, HNRNPUL1, RPL18A, ZNF85, HOXA1, DNAJB6, EIF4H, GRPEL1, NCAPG, HSPA8, EIF4G2, SLC11A2, TIMELESS, LDHB, SRSF3, TMEM30A, HMGCS1, PPP2CA, ACTR3, STAT1, MOB4, ORC2, SET, EIF2B2, MASTL, KIAA1217, ETF1, POLR3GL, RPL5, CBX3, TUBA1B, PREX1, MED20, HP1BP3, EPS15L1, YWHAH, DOCK4, SERGEF, KIF1C, ATP5IF1, PRRG1, ACSS2, CAP1, TOP2A, TPT1, SLC38A2, YWHAQ, LDHA, FADS2, GCC2, SRSF1, RPS6, HMGA1, SEC31A, G3BP2, C12orf10, ESD, TLE3, RANBP10, GALNT1, APP, RPL13A, CAND2, RPL32, RPS3A, COMMD10, G3BP1, NONO, RPL7A, CHEK1, ARL14EP, CARHSP1, HNRNPU, DDAH1, UCHL1, TMEM237, RNF20, EIF4A2, CLDN12, EIF5B, CTBP1, H3F3A, RPL9, YEATS2, TIGD6, UQCRQ, FAT3, GTF2A1, TAF1D, RPL13, EEF2, FNTA, SNRNP48, E2F6, INO80E, LDB2, HNRNPF, UBB, POLH, TRIAP1, BSG, CORO1B, MOB1B, CHD2, CCDC14, DRAP1, BASP1, MAN1B1, UBE2N, ZFP42, GEMIN4, NPM1, ACTG1, TRIM69, RPS23, ZNF567, H1F0, FAT4, PPTC7, SPN, MT-ND2, MT-ND3, ZNF814, RACK1, C5orf51, FAM72D, ATAD3C, HSBP1L1, OST4, APOBEC3C, TUNAR, HMBS, AC104109.3, HIST1H3E, AC013394.1, AC106886.5, Z82190.2, HLA-B, TMEM189, RPL36A, ARPC4, APELA, ATP5MGL, AL133352.1, AC026954.2, AC135178.2, HIST2H2AA4, PPP4R3B, NEFL
lncRNA MIAT	Up/down	59	CDKN2A, DUSP7, HDAC4, KMT2A, NFE2L2, PTGS2, GDI2, SART3, DLG3, HUWE1, FTL, ERP29, RPL3, RHEB, CREB3, INO80B, TMEM189-UBE2V1, STAU1, PLAGL2, CDC16, MTCH1, PYCR2, LSM14B, SAP18, PTMS, LINGO1, CSRP2, IST1, YBX1, MLF2, KIF4A, TINF2, EZR, PRODH, MAZ, MEGF8, PFN1, YWHAE, CORO1C, COPS5, TUBA1B, CAPRIN1, HADHB, CEP170, NONO, SIGMAR1, RPS3, LSM12, BEX3, RPSA, CFL1, PPP1R42, PRR5, PRRC2A, FASTKD5, PRR5-ARHGAP8, UBE2V1, AL117348.2, INO80B-WBP1
lncRNA CASC2	Down	4	TGFB1, SOX4, RUNX1, BCL2
lncRNA NEAT1	Up	674	ZEB1, SRp40, MAPK6, HIV-1, GPHA2, FUS, EZH2, CALCR, ADARB2, SCIN, IL32, CSDE1, CD9, TTC19, CD74, NDUFS1, RABEP1, MRI1, PI4K2B, PHF23, MAP4, USP36, YBX3, WNK1, QSER1, CS, EIF4B, SLC9A7, FOXJ2, ARFGEF1, SIRT2, ING3, EPN2, NCBP3, CELSR1, WDR62, NRDC, PCDHGA2, ZNF586, CNOT3, FUS, RGS1, PDCD7, SETD1A, POLR2E, KLHL22, SNRPD3, PES1, HIRA, MICALL1, TTC28, DDX17, RBFOX2, MYH9, KIAA0586, GMPR2, RNF24, PLS3, SLC25A14, KLF5, NFAT5, SLC38A7, NPRL3, COTL1, MAZ, CHRAC1, NUCB1, CD37, URI1, RPS16, AKT2, CCDC94, PPP2R1A, COPE, GTPBP10, CHCHD2, RBM28, UBE2S, MED13, PFN1, EZH1, LUC7L3, TNFAIP1, RPL34, HSPA8, CHORDC1, MDK, MADD, CD81, RPS13, CHKA, MLEC, ATP5F1B, ARHGDIB, SLC38A1, CPSF6, USP5, GNPTAB, PTPN6, COX6A1, AL021546.1, DSE, NR3C1, ACVR2B, EIF4G1, MOB1A, STEAP3, TANC1, SPTBN1, HTRA2, ARID3A, QSOX1, PLEKHM2, ADGRL2, ARID1A, RAB3GAP2, DENND1A, TMEM214, PANK3, SIL1, CBX3, HNRNPA2B1, CHST3, CIT, MORF4L2, STAMBP, SNRPC, MED20, SYMPK, IFI6, AIF1L, UBR4, SLC35E1, GNAI1, TUBGCP6, CALU, CLN6, VPS13C, SUMF2, RPAIN, ILF3, KLF16, TRPM4, NUP210, UNK, ARHGEF11, IRS4, BTBD2, CNDP2, WDR74, BTG1, AGO4, GOLM1, KHDC1, CDK4, MDM2, FAM129A, LIMD2, PIM1, TMEM63B, MDC1, PARP6, HADHB, 11-Sep, TMBIM6, WARS, SERF2, SRP14, POLG, TICRR, RNF157, MIEN1, FKBP10, APP, RPL13A, RFX5, SLC39A1, INTS3, ARF1, HSPD1, LIMD1, TRPC1, ABRACL, MEPCE, ZMYM3, NONO, RPL7, NFIB, FBXO10, MTG1, RPS3, LSM14B, HMGA2, ITGB1, C18orf25, CARHSP1, UCHL1, MIA3, ELOC, TMSB4Y, VOPP1, APOOL, FBRS, B3GNT7, EIF4A2, DHRS4, SLC35B2, RER1, DUSP2, MRPL10, ZYX, CALM3, U2AF1, RPL8, TEDC2, LBHD1, RPL29, LAPTM5, FZD5, CIP2A, STT3B, NEPRO, LZTFL1, APEH, DNAAF5, ZNF22, ARHGAP42, FRS2, TRIM66, RPL27A, NDEL1, VPS39, YWHAB, C15orf39, CRK, IGF2, CDK12, ENGASE, MYO5B, LENG8, EEF2, KIFC2, AC010323.1, SF1, RBPJ, BMI1, PTPN9, FASN, PCDH7, PFKFB3, RNF34, PRKCE, ZNF274, PIK3CD, MALT1, RPL38, VANGL1, ZNF622, PC, AGFG1, STAT5B, CCS, GCSAM, SLC29A2, UNC119B, COX8A, SPRYD4, FAM210A, TBL1XR1, SUMO4, ZNF354B, FAM219B, TMEM107, CYC1, CALR, LDLRAD3, GEMIN4, MYADM, CSNK1A1L, CCDC66, MRPL14, MUC16, NOP10, EXOC7, GAS6, DAZAP2, PSMG1, HMCES, DENND5A, C6orf120, SP1, DMWD, C15orf41, EDARADD, KPNA4, SMYD4, COL4A5, HES4, PPIA, TRRAP, NACA, ZNF512B, STRN3, SPN, RPL12, RPS4X, RPS6KL1, BAZ1A, NCOA6, MT-CO2, MT-CO1, MT-ND1, TOP1, MT-CO3, LIME1, PRRC2A, HLA-C, RACK1, PCDHGA1, LIN52, C17orf51, SFT2D2, AC239799.1, SUPT4H1, ZSWIM8, KIFC1, TXNDC5, PCDHGC3, PCDHGC5, C22orf39, PCDHGC4, GGCT, PKD1, GPRC5A, YAF2, CXorf56, TOMM34, SARS, FAM136A, RFC1, DSG2, NEDD4L, FOXC1, ZFR, POLD1, U2AF2, SLC12A2, AP3D1, ELAVL1, TRAM1, ACSL4, PABPC1, SLC6A15, LMAN1, ENO1, ZNF37A, FNDC3B, ACTB, BAZ2A, TRAF4, TOP2B, CAPZB, CDC14B, NFE2L1, FAT1, TXLNG, HUWE1, KHSRP, MAVS, RPL6, ANAPC5, RPLP0, BIRC5, GOLGA3, TMEM101, HNRNPC, TYRO3, GABRP, SUCO, HSP90AB1, ABL1, GGA1, SBF1, LMF2, RPL3, TNRC6B, GNPNAT1, ACIN1, TPD52L2, MRGBP, NOP56, VAPA, AMMECR1, NAA10, USP11, DNAJC3, PIEZO1, KNOP1, MTFMT, IMPAD1, STK3, KLHDC4, SARS2, PPP1R13L, PLD3, FKBP8, AVL9, CDK6, ABHD11, EIF3A, ACTA2, SHOC2, TSPAN14, GIT1, CPD, SNX25, SNX15, LPXN, HIPK3, SOX6, BCL7A, CDK2AP1, GAPDH, CHD4, CCNC, SENP6, MSH3, COMMD2, TTC31, RTN4, BIRC6, MTR, ETV3, APH1A, NENF, KLF7, CCND2, KIAA1217, ZNF706, PILRB, PDS5A, MRPS2, DDX39A, PFDN5, STIL, VAPB, XPO5, AL365205.1, CDKN1A, SOX4, RPL23, UBA2, FGFRL1, SMARCA4, AAMP, MYO5C, KRI1, DOCK6, AFDN, PRRC2B, DKC1, TOP2A, ANKRD17, WBP2, EIF5A, ZNF414, HSPBP1, DPF2, PRPF38B, YWHAQ, LDHA, BTF3L4, NREP, ACVR1B, WNT10A, HUS1, RPL35, IL11RA, TMEM14B, NUMA1, YAP1, SLC37A4, SCARB2, WDR61, MAN2C1, RPS2, PELP1, NBPF3, CD53, UBAP2L, GOLGA4, RPS3A, PHIP, CASK, TACC1, MKI67, PPRC1, HSD17B12, TNKS1BP1, CELF1, YTHDF1, PPP4C, ALDOA, CTAGE5, UBE3B, VIPAS39, DST, RANBP2, MSI2, TBCEL, CXADR, RPL30, IGF2BP1, C21orf2, NACC1, PKDCC, REL, DPY30, PTPN13, CDCP1, TMEM41A, UBXN7, PGRMC2, ZNF589, HMGB2, GALNT10, SPTSSA, RPL10L, E2F7, ARL5B, MCM7, B2M, PLK1, STIM1, RPL13, KMT2D, C19orf48, PRDX2, SRRM2, NT5DC2, HIST1H1E, ADAM9, AXIN2, C2orf68, MAP2K1, SLC25A6, SIN3A, CLIC4, PCBP1, NIPA1, SIK2, ARL6IP1, JAGN1, ZNF692, CLSTN1, RRM2, PPIH, ZMAT3, CFL1, MAP3K11, EIF1, VCPIP1, RPS6KB2, SLC35A4, SLC38A9, ZBTB38, GRB2, DPY19L3, CTC1, SAMD4B, EDC3, CLK3, AEN, C8orf33, PLCXD1, ACTG1, PLA2G6, UBALD2, IFIT1, TEAD1, PTMA, ZNF177, TMEM201, TRMT2B, ARID2, S100A13, PLEKHG4, TUBB, ZNF775, SIAH1, PRPF40A, MKL1, HDAC2, ARHGEF12, SYNGAP1, MCMBP, FKBP1C, HLA-DRB5, DDX39B, MT-ND5, MT-ATP6, TBKBP1, BMPR2, HLA-DOA, BRD2, HSPA1A, KM-PA-2, E2F4, TMSB4X, HLA-A, MT-ND4L, ZNF580, ZBTB9, ZNF90, TOMM6, PHB2, POTEJ, TTC4, P2RY11, ATP6V1G2-DDX39B, MTRNR2L1, AL162231.3, EPPK1, OTUD7B, MTRNR2L12, AC011455.2, ZNF559-ZNF177, MROH7-TTC4, KMT2B, ARL2-SNX15, HIST1H3I, NEFL, AC005154.6, ALDOA, CHCHD10, PCDHGA12, PCDHGB6, PCDHGA5, PCDHGA7, ATXN7L3B, PCDHGA6, PCDHGA8, PCDHGA10, PCDHGA11, PCDHGB2, PCDHGB4, PCDHGB7, PCDHGB1, PCDHGA3, AL360181.3, MTRNR2L8, BLOC1S5-TXNDC5, AC107871.1, PCDHGA9, PCDHGB3, PCDHGA4, NOTCH2NL, NDUFA7, AC010522.1, ZNF587B, MTRNR2L6, AL121845.2, AL121845.3, TAF9, NOL12, U2AF1L5, PIP4K2B, PCDHGB5, CCL3, HSPA14, IGF2

3.4.2. PPI Network

To distinguish the links between the 1,921 target genes, the study plotted the PPI using data from the STRING database (https://stringg.embl.de/). The 1,125 candidate genes were linked using Cytoscape 3.8.1 to build an initial PPI network, and the top 50 linked target genes were identified by degree criteria. The network consisted of 50 nodes and 1,086 edges with an average local clustering coefficient of 0.722. The study imported the PPI network into Cytoscape for further analysis. Finally, two key subnetworks consisting of 12 target genes were obtained using CytoNca and CytoHubba, respectively, and all edges were discriminated according to the connection score (Table 4, Figure 6).

Table 4

Top 10 genes in the network sorted by degree.

Rank	Symbol	Ensembl ID	Score
1	UBA52	ENSG00000221983	114
2	TP53	ENSG00000141510	113
3	RPS11	ENSG00000142534	90
4	RPS6	ENSG00000137154	89
5	RPS13	ENSG00000110700	88
6	RPS9	ENSG00000170889	87
7	RPS16	ENSG00000105193	87
8	GNB2L1	ENSG00000204628	86
9	RPS5	ENSG00000083845	86
10	RPS3A	ENSG00000145425	85

Figure 6

PPI network of target genes. (a) Complete protein-protein network. (b) PPI network was first filtered through Cytohubba, and all circles were proteins encoded by the top 50 positions. (c) The key subnets of the top 12 nodes revealed by CytoHubba analysis. (d) Top 12 key subnetworks screened after two filters using CytoNca. Red represents the highest-ranking genes, and yellow circles represent the lowest-ranking genes.

3.4.3. Biological Analysis of Hub Genes

The ClueGO tool was used to perform GO and KEGG enrichment analysis on the above target genes. The highly connected proteins in the network are the hub proteins of regulation. The hub proteins in this study include UBA52, TP53, RPS11, RPS6, RPS13, RPS9, RPS16, GNB2L1, RPS5, and RPS3A. These involve 12 nodes and 56 edges. In addition, KEGG analysis was performed on potential hub genes and the top 10 enriched pathways were identified. The study investigated the role of core targets in gene function and obtained 996, 120, and 84 GO entries, respectively, and the top 10 entries were selected according to the p value (p < 0.01) (Figure 7).

Figure 7

Biological analysis of hub genes. (a) KEGG enrichment analysis of hub genes by Cytoscape; (b) GO biological process (BP) analysis of hub genes; (c) GO cellular component (CC) analysis of hub genes; (d) GO molecular function (MF) analysis of hub genes.

4. Discussion

At present, some studies have found that the production of “metabolic memory” in kidney cells has led to diabetic nephropathy. The change of epigenetics is the core event in the progression of DKD [43]. In this case, there is a bidirectional regulation between ncRNA and epigenetic modification. lncRNAs are a type of ncRNAs that seem to be involved in the pathogenesis of DKD [44]. Therefore, the study conducted a systematic review to further study which lncRNAs are mainly related to DKD. The results indicated that eight lncRNAs were persistently dysregulated in DKD patients. Four kinds of lncRNA disorders (MALAT1, NEAT1, NR_033515, and ARAP1-AS2) were upregulated, while CASC2 and ARAP1-AS1 were downregulated in diabetes cases compared with the control group. Among them, lncRNA MIAT and lncRNA GAS5 showed different trends of regulation.

Metastasis-associated lung adenocarcinoma transcript 1, also known as Neat2 (Malat1) is one of the most analyzed lncRNAs in DKD samples. The qualitative analysis showed that lncRNAs were expressed in serum, urine, peripheral blood mononuclear cells, and kidney tissue [29, 30, 36, 37, 39]. There were increased expressions of lncRNA MALAT1 in DKD, translocation of β-catenin to the nucleus, enhanced expressions of MALAT1 connexin serine/arginine splicing factor 1 (SRSF1), and expressions of p-cadherin and tight junction protein ZO-1. The study also found that β-catenin participated in MALAT1 transcription by binding to the promoter region of MALAT1, and β-catenin knockdown also reduced the level of MALAT1, indicating that there is a new feedback between MALAT1 and β-catenin [45]. Increased levels of MALAT1 are related to the upregulation of serum amyloid A3, TNF, and IL-6 genes [46]. Some studies have reported that the lncRNA signal is related to the pathogenesis of DM, such as NRF2 signal FoxO1, MAPK/ERK, and Wnt/β-catenin signal pathways [47-49]. Therefore, computer analysis showed that Malat1 was involved in many pathways related to DM and its complications. In addition to PI3K/Akt, MAPK and TNF, apoptosis, insulin, cell cycle, AMPK, FoxO, ErbB, HIF-1, and AGE/RAGE were also covered. Increased expression of MALAT1 was also found in the kidneys of patients with DKD and STZ-induced rats [30, 50]. lncRNA MALAT1 regulated cell pyrolysis and inflammation. Therefore, human proximal tubular cells (HK-2 cells) cultured under HG conditions also express high levels of Malat1. MALAT1 in diabetic-related complications is both pro-inflammatory and apoptosis in different cell types.

Growth arrest-specific 5 (GAS5) is differentially expressed in the serum of DKD patients [17, 42]. It has been found that GAS5 is differentially expressed in the plasma of diabetic people and mice. The expression of GAS5 is significantly upregulated in the renal cortex of a mouse model of diabetic nephropathy induced by HFD/STZ [51]. GAS5 is located on chromosome lq25 and has 13 exons, producing a series of long noncoding RNA. Part of RNA secondary structure, the encoded transcript, mimics the glucocorticoid response element (GRE), which means that GAS5 can bind to the DNA binding domain of the glucocorticoid receptor. The combination of the two blocks of the glucocorticoid receptor cannot further regulate the transcription of the target gene. GAS5 is also considered to regulate its transcriptional activity through GRE-like regions bound to related hormone receptors (such as androgen, progesterone, and mineralocorticoid receptors) [52]. GAS5 plays an important role in cell stagnation, proliferation, and apoptosis, autophagy and many other biological processes [53], which is also related to the TGFβ/Smad pathway [54].

Myocardial infarction associated translation (MIAT) is differentially expressed in the serum of DKD patients [30, 55]. MIAT is an IncRNA produced by transcription of a gene located on chromosome 22ql2.1, which is involved in the regulation process of diabetic retinopathy (DR). Knockout of MIAT can alleviate the dysfunction of endothelial cells and reduce the formation of new blood vessels, blood vessel leakage, and inflammation during DR [56]. In addition, there may be an interaction between IncRNA MIAT and inflammation and apoptosis. Knockdown of the expression of IncRNA MIAT can reduce podocyte damage induced by high glucose and protect podocytes [57]. However, in the review, it was found that the expressions of GAS5 and MIAT were inconsistent, which may be caused by differences in different samples and human samples.

Cancer susceptibility candidate 2 (CASC2) is also continuously downregulated in the serum of DKD patients [20, 21]. CASC2 is located at 10q26 of chromosome and spans D10S190. It is involved in cell proliferation, apoptosis inhibition, fibrosis, ECM accumulation, and oxidation stress. In vitro experiments showed that the expression of CASC2 was downregulated in podocytes cultured with high glucose. After CASC2 overexpression, the phosphorylation level of JNK1 and the apoptosis rate of podocytes were significantly reduced, and the JNK1 activator could significantly antagonize the apoptosis of podocytes caused by CASC2 overexpression. The inhibitory effect of DN increased the rate of podocyte apoptosis. It can be seen that overexpression of CASC2 inhibits podocyte apoptosis by blocking the JNK pathway, thereby delaying the progression of DN [21]. Long noncoding RNA CASC2 inhibits human mesangial cell proliferation, inflammation, and fibrosis.

Nuclear-enriched abundant transcript 1 (NEAT1) is continuously downregulated in the serum and urine of patients with DKD [31, 55]. NEAT1 is a lncRNA located in the nucleus, which is involved in the transcription of many genes, and participates in cell proliferation, inhibition of apoptosis, fibrosis and ECM accumulation, and oxidative stress [58]. It has been found that NEAT1 expression is upregulated in rat kidney tissues and mouse GMCs cultured with high glucose, NEAT1 can promote ECM accumulation and EMT processes and accelerate the process of renal fibrosis [59]. Computer analysis showed that NEAT1 was involved in many pathways related to DM and its complications, including cell cycle and HIF-1. lncRNA NEAT1 accelerates the development and progression of diabetic nephropathy.

The expression of NR_033515 in the kidney, peripheral blood, urine, and serum is increased compared with the control group [19, 28], which is related to renal fibrosis, and enhance expression levels of fibrogenesis-related gene proteins (P38, ASK1, and ASK1), fibronectin and α-SMA [19]. lncRNA-NR_033515 promotes proliferation, fibrosis, and epithelial-mesenchymal transition in DKD.

lncRNA-ARAP1-AS2 and lncRNA-ARAP1-AS1 are natural antisense lncRNAs located on chromosome 11. In the review, under high glucose conditions, lncRNA ARAP and lncRNA ARAP1-AS2 (antisense RNA 2) are in HK-2 cells. In addition, the overexpression of lncRNAARAP1-AS2 enhances the EMT process, and ARAP1 gene knockout can reduce the occurrence of EMT and fibrosis in HK-2 cells induced by high glucose [60]. In addition, ARAP1-AS1 is related to signaling pathways such as PGF, PLAGL2, EZH2, HDAC2, Wnt/β-catenin, and is involved in the pathogenesis of colon cancer, gastric cancer, and breast cancer [61-63].

The bioinformatics analysis also showed that the pathways regulated by Malat1 related target genes, such as PI3K/Akt, TNF, HIF-1, AGE/RAGE, insulin resistance, apoptosis, and FoxO, were important pathogenic mechanisms of DKD.

In summary, the systematic review showed that eight lncRNAs were continuously dysregulated in DM and DKD patients. The study also clarified the pathways regulated by these lncRNAs and involved in the pathogenesis of DM, such as PI3K/Akt, TNF, HIF-1, AGE/RAGE, insulin resistance, apoptosis, and FoxO. Although this systematic review included 28 studies that analyzed the expression of lncRNA in DKD-related tissues, the involvement of lncRNAs in the pathogenesis of this complex disease remains to be further investigated. Although lncRNAs seem to be a good candidate for DKD biomarkers and therapeutic targets, more studies on the different tissues and cell distribution of these regulatory molecules may further clarify their role in DKD.

5. Additional Points

Although this systematic review showed that lncRNA always regulated pathways related to DKD, it still has some limitations. First of all, the lncRNAs in some studies did not adopt formal names; therefore, the study may have lost some of the details in the study. Second, different techniques were used to quantify the expression of lncRNA in different studies, and only the expression pattern of lncRNA was provided. Therefore, it is impossible to perform a reliable quantitative meta-analysis based on the expression level of lncRNA. Finally, compared with the control group, the eight types of lncRNAs were always dysregulated in DKD patients, but it is not possible to perform hierarchical analysis by tissue type, because the number of studies using different tissues to evaluate the same lncRNA is very small. However, most of the studies included in this systematic review did not report the stages of kidney disease in patients with these DKDs. Therefore, here, it is impossible to assess whether different kidney function stages affect results. The findings remain to be further verified.