基于WGCNA转录组分析构建区分牙周炎与种植体周围炎的lncRNA-miRNA-mRNA竞争性内源RNA网络

doi:10.3969/j.issn.1674-8115.2026.05.006

摘要/Abstract

摘要：

目的·系统比较牙周炎（periodontitis，PD）与种植体周围炎（peri-implantitis，PI）状态下长链非编码RNA（long noncoding RNA，lncRNA）及其介导的竞争性内源RNA（competing endogenous RNA，ceRNA）调控网络特征。方法·将24只雄性SD大鼠随机分为对照组（CON）、PD组和PI组，每组8只。PD组采用双侧上颌第一磨牙丝线结扎法建立实验性PD模型，PI组于双侧上颌第一磨牙植入定制钛种植体后行丝线结扎诱导PI。采集各组大鼠牙龈组织进行转录组测序，获得lncRNAs、mRNAs表达谱数据。采用DESeq2软件筛选差异表达lncRNAs与mRNAs。通过加权基因共表达网络分析（weighted gene co-expression network analysis，WGCNA）筛选与PD、PI相关的模块基因。运用KOBAS软件对模块基因进行基因本体论（Gene Ontology，GO）功能富集分析。基于miRanda软件（score≥150）预测微RNA（microRNA，miRNA）结合位点，分别构建与PD、PI模块基因相关的lncRNA-miRNA-mRNA ceRNA调控网络，并筛选与PD及PI相关模块的ceRNA调控轴。结果·与CON组相比，PD组共鉴定出124个上调、406个下调的差异表达lncRNAs，PI组共筛选出43个上调、95个下调的差异表达lncRNAs，且共有22个lncRNAs在PD、PI组中均发生差异表达；同时，PD组特异性差异表达lncRNAs为508个，PI组为116个。WGCNA识别出与PD显著正相关的lightcyan模块和与PI显著正相关的midnightblue模块。GO功能富集分析显示，lightcyan模块基因显著富集于内质网相关蛋白降解（endoplasmic reticulum-associated degradation，ERAD）信号、对内质网应激（endoplasmic reticulum stress，ERS）的反应等相关的生物学过程，midnightblue模块基因则显著富集于皮肤屏障建立、角质形成细胞分化、表皮发育等生物学过程。基于上述模块构建ceRNA调控网络，筛选出与PD组ERS相关的ENSRNOG00000063579-miR-1249-Calr（钙网蛋白，calreticulin）、ENSRNOG00000063579-miR-1249-Man1b1（甘露糖苷酶α1B类成员1，mannosidase α class 1B member 1）、ENSRNOG00000063488-miR-328a-5p-Rnf183（环指蛋白183，ring finger protein 183）、ENSRNOG00000063230-miR-3558-5p-Fbxo2（F盒蛋白2，F-box protein 2）等调控轴，以及与PI组皮肤屏障功能相关的ENSRNOG00000067449-miR-238a-5p-Krt1（角蛋白1，keratin 1）、ENSRNOG00000067449-miR-238a-5p-Krt16、ENSRNOG00000067625-miR-3541-Klf4（Krüppel样因子4，Krüppel-like factor 4）、ENSRNOG00000067625-miR-667-5p-Tgm3（转谷氨酰胺酶3，transglutaminase 3）等调控轴。结论·该研究揭示了PD与PI在lncRNAs和mRNAs表达谱上的差异，提示两者分别涉及ERS与皮肤屏障功能相关通路；上述发现为理解2种疾病的分子病理特征提供了依据。

关键词: 种植体周围炎, 牙周炎, 加权基因共表达网络分析, lncRNA-miRNA-mRNA ceRNA调控网络

Abstract:

Objective ·To systematically compare the characteristics of long non-coding RNAs (lncRNAs) and their mediated competing endogenous RNA (ceRNA) regulatory networks in periodontitis (PD) and peri-implantitis (PI). Methods ·Twenty-four male SD rats were randomly divided into the control (CON), PD, and PI groups, with eight rats in each group. In the PD group, an experimental PD model was established by ligating the bilateral maxillary first molars with silk sutures. In the PI group, custom titanium implants were placed in the bilateral maxillary first molar regions, followed by silk ligation to induce PI. Gingival tissues from each group were collected for transcriptome sequencing to obtain expression profiles of lncRNAs and mRNAs. DESeq2 was used to screen differentially expressed lncRNAs and mRNAs. Weighted gene co-expression network analysis (WGCNA) was performed to identify module genes associated with PD and PI. Gene Ontology (GO) functional enrichment analysis of the module genes was conducted using KOBAS software. miRNA binding sites were predicted using miRanda software (score≥150), and lncRNA-miRNA-mRNA ceRNA regulatory networks associated with PD- and PI-related module genes were constructed. ceRNA regulatory axes related to PD and PI modules were screened. Results ·Compared with the CON group, 124 up-regulated and 406 down-regulated differentially expressed lncRNAs were identified in the PD group, while 43 up-regulated and 95 down-regulated differentially expressed lncRNAs were identified in the PI group. A total of 22 lncRNAs were differentially expressed in both PD and PI groups. Additionally, 508 lncRNAs were specifically differentially expressed in the PD group, and 116 in the PI group. WGCNA identified the lightcyan module significantly positively correlated with PD and the midnightblue module significantly positively correlated with PI. GO functional enrichment analysis showed that genes in the lightcyan module were significantly enriched in biological processes related to endoplasmic reticulum-associated degradation (ERAD) signaling and response to endoplasmic reticulum stress (ERS), whereas genes in the midnightblue module were significantly enriched in biological processes such as skin barrier formation, keratinocyte differentiation, and epithelial development. Based on these modules, ceRNA regulatory networks were constructed, revealing regulatory axes associated with ERS in the PD group, including ENSRNOG00000063579-miR-1249-Calr (calreticulin), ENSRNOG00000063579-miR-1249-Man1b1 (mannosidase α class 1B member 1), ENSRNOG00000063488-miR-328a-5p-Rnf183 (ring finger protein 183), and ENSRNOG00000063230-miR-3558-5p-Fbxo2 (F-box protein 2), as well as regulatory axes associated with skin barrier function in the PI group, including ENSRNOG00000067449-miR-238a-5p-Krt1 (keratin 1), ENSRNOG00000067449-miR-238a-5p-Krt16, ENSRNOG00000067625-miR-3541-Klf4 (Krüppel-like factor 4), and ENSRNOG00000067625-miR-667-5p-Tgm3 (transglutaminase 3). Conclusion ·This study reveals differences in lncRNA and mRNA expression profiles between PD and PI, suggesting that the two diseases are respectively associated with pathways related to ERS and skin barrier function. These findings provide a basis for understanding the molecular pathological characteristics of the two diseases.

Key words: peri-implantitis (PI), periodontitis (PD), weighted gene co-expression network analysis (WGCNA), lncRNA-miRNA-mRNA ceRNA regulatory network

中图分类号:

R780.2

王艳, 郭涛, 薄雨佳, 余颖, 谢鑫涛, 黄旭. 基于WGCNA转录组分析构建区分牙周炎与种植体周围炎的lncRNA-miRNA-mRNA竞争性内源RNA网络[J]. 上海交通大学学报（医学版）, 2026, 46(5): 602-611.

Wang Yan, Guo Tao, Bo Yujia, Yu Ying, Xie Xintao, Huang Xu. Construction of a lncRNA-miRNA-mRNA competing endogenous RNA network for distinguishing periodontitis from peri-implantitis based on WGCNA transcriptome analysis[J]. Journal of Shanghai Jiao Tong University (Medical Science), 2026, 46(5): 602-611.

图/表 4

图1 3组大鼠牙龈组织样本的差异表达lncRNAs分析Note： A. PCA of lncRNA expression profiles in the three groups. B. Statistical chart of the numbers of DElncRNAs in the PD group vs CON group and the PI group vs CON group. C. Venn diagram of DElncRNAs in the PD group vs CON group and the PI group vs CON group.

Fig 1 Analysis of DElncRNAs in gingival tissue samples of rats among the three groups

图2 3组大鼠牙龈组织样本的基因共表达网络的构建及其模块识别Note： A. Gene clustering tree and module partitioning results of the three groups. Different colors represent different modules, and the grey module represents genes not assigned to any module. B. Network topology analysis under different soft-thresholding powers. C. Clustering heatmap of module eigengenes in the three groups. Red indicates high similarity, and blue indicates low similarity. D. Correlation heatmap between different modules and the CON, PD, and PI groups. Red indicates positive correlation, and green indicates negative correlation.

Fig 2 Construction of gene co-expression network and identification of modules in gingival tissue samples of rats among the three groups

图3 PD、PI相关模块基因的表达特征及GO功能富集分析Note： A/B. Expression characteristics of genes in the lightcyan module (A) and midnightblue module (B) in the CON, PD, and PI groups. C/D. GO functional enrichment analysis of genes in the lightcyan module (C) and midnightblue module (D). ER—endoplasmic reticulum.

Fig 3 Expression characteristics and GO functional enrichment analysis of PD- and PI-related module genes

图4 PD与PI相关lncRNA-miRNA-mRNA ceRNA调控网络Note： A. lncRNA-miRNA-mRNA ceRNA network closely associated with ERS in the PD network. B. Expression levels of lncRNAs and mRNAs in the core regulatory axes of the PD network. C. lncRNA-miRNA-mRNA ceRNA network closely associated with epithelial barrier function in the PI network. D. Expression levels of lncRNAs and mRNAs in the core regulatory axes of the PI network.

Fig 4 lncRNA-miRNA-mRNA ceRNA regulatory networks associated with PD and PI

参考文献 30

[1]	Chen J K, Chang Y M. Dental implant outcomes in Fibula Mandible reconstruction with or without vertical distraction for symphysis cross-arch segmental defects[J]. Plast Reconstr Surg, 2022, 150(4): 863e-868e.
[2]	Roccuzzo A, Imber J C, Salvi G E, et al. Peri-implantitis as the consequence of errors in implant therapy[J]. Periodontol 2000, 2023, 92(1): 350-361.
[3]	Heitz-Mayfield L J A. Peri-implant mucositis and peri-implantitis: key features and differences[J]. Br Dent J, 2024, 236(10): 791-794.
[4]	Derks J, Schaller D, Håkansson J, et al. Peri-implantitis:onset and pattern of progression[J]. J Clin Periodontol, 2016, 43(4): 383-388.
[5]	巫佩瑶, 周陶, 曹志炜, 等. 种植体周围炎与牙周炎的比较[J]. 医学综述, 2019, 25(17): 3333-3337.
	Wu P Y, Zhou T, Cao Z W, et al. Comparison of peri-implantitis and periodontitis[J]. Medical Recapitulate, 2019, 25(17): 3333-3337.
[6]	Wei H X, Yi T, Li Q, et al. Application of lncRNA-miRNA-mRNA ceRNA network analysis in the treatment of androgenic alopecia[J]. J Clin Lab Anal, 2023, 37(1): e24791.
[7]	Bazrgar M, Mirmotalebisohi S A, Ahmadi M, et al. Comprehensive analysis of lncRNA-associated ceRNA network reveals novel potential prognostic regulatory axes in glioblastoma multiforme[J]. J Cell Mol Med, 2024, 28(11): e18392.
[8]	Zhu Z X, Liu Y, Wang J H, et al. A novel lncRNA-mediated epigenetic regulatory mechanism in periodontitis[J]. Int J Biol Sci, 2023, 19(16): 5187-5203.
[9]	Li J, Jin F, Cai M, et al. LncRNA nron inhibits bone resorption in periodontitis[J]. J Dent Res, 2022, 101(2): 187-195.
[10]	Lin C, Liu M M, Guo J H, et al. Cryotherapy attenuates inflammation via the lncRNA SNHG1/miR-9-5p/NFKB1 regulatory axis in periodontal ligament cells[J]. Int J Mol Sci, 2023, 24(15): 12097.
[11]	Jia Q, Jiang W K, Ni L X. Down-regulated non-coding RNA (lncRNA-ANCR) promotes osteogenic differentiation of periodontal ligament stem cells[J]. Arch Oral Biol, 2015, 60(2): 234-241.
[12]	Wang L, Wu F, Song Y, et al. Long noncoding RNA related to periodontitis interacts with miR-182 to upregulate osteogenic differentiation in periodontal mesenchymal stem cells of periodontitis patients[J]. Cell Death Dis, 2016, 7(8): e2327.
[13]	Chai M, Zhang J, Meng Q, et al. Diagnostic value of lncRNA XIST in saliva for early peri-implantitis[J]. Oral Health Prev Dent, 2024, 22: 381-388.
[14]	Liu Y D, Liu Q F, Li Z P, et al. Long non-coding RNA and mRNA expression profiles in peri-implantitis vs periodontitis[J]. J Periodontal Res, 2020, 55(3): 342-353.
[15]	Kim D, Langmead B, Salzberg S L. HISAT: a fast spliced aligner with low memory requirements[J]. Nat Methods, 2015, 12(4): 357-360.
[16]	Love M I, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2[J]. Genome Biol, 2014, 15(12): 550.
[17]	Xie C, Mao X Z, Huang J J, et al. KOBAS 2.0: a web server for annotation and identification of enriched pathways and diseases[J]. Nucleic Acids Res, 2011, 39(Web Server issue): W316-W322.
[18]	Galarraga-Vinueza M E, Pagni S, Finkelman M, et al. Prevalence, incidence, systemic, behavioral, and patient-related risk factors and indicators for peri-implant diseases: an AO/AAP systematic review and meta-analysis[J]. J Periodontol, 2025, 96(6): 587-633.
[19]	Jiang M, Li Z N, Zhu G X. The role of endoplasmic reticulum stress in the pathophysiology of periodontal disease[J]. J Periodontal Res, 2022, 57(5): 915-932.
[20]	李鑫, 李艳, 丁旭, 等. 内质网应激在牙周炎影响全身疾病过程中的作用[J]. 国际口腔医学杂志, 2021, 48(1): 12-17.
	Li X, Li Y, Ding X, et al. Role of endoplasmic reticulum stress in periodontitis affecting systemic diseases[J]. International Journal of Stomatology, 2021, 48(1): 12-17.
[21]	Feng Y, Zhang R, Wang Y R, et al. Inhibition of endoplasmic reticulum stress by 4-phenyl butyric acid presents therapeutic effects on periodontitis: experimental studies in vitro and in rats[J]. Stem Cells Int, 2021, 2021: 6618943.
[22]	Alves C H, Russi K L, Rocha N C, et al. Host-microbiome interactions regarding peri-implantitis and dental implant loss[J]. J Transl Med, 2022, 20(1): 425.
[23]	Heitz-Mayfield L J A, Salvi G E. Peri-implant mucositis[J]. J Clin Periodontol, 2018, 45 : S237-S245.
[24]	朱星宇, 唐菡, 陈陶, 等. 基于生物信息学分析探索种植体周围炎的免疫特征基因及其对免疫细胞的调控机制[J]. 重庆医科大学学报, 2024, 49(4): 436-443.
	Zhu X Y, Tang H, Chen T, et al. Immunogenetic features and their regulatory mechanisms on immune cells in peri-implantitis: a bioinformatics analysis[J]. Journal of Chongqing Medical University, 2024, 49(4): 436-443.
[25]	Hämmerle C H, Giannobile W V, Working Group 1 of the European Workshop on Periodontology. Biology of soft tissue wound healing and regeneration: consensus report of Group 1 of the 10th European Workshop on Periodontology[J]. J Clin Periodontol, 2014, 41(Suppl 15): S1-S5.
[26]	Fucikova J, Spisek R, Kroemer G, et al. Calreticulin and cancer[J]. Cell Res, 2021, 31(1): 5-16.
[27]	Bellei E, Bertoldi C, Monari E, et al. Proteomics disclose the potential of gingival crevicular fluid (GCF) as a source of biomarkers for severe periodontitis[J]. Materials, 2022, 15(6): 2161.
[28]	Ji J, Jing A X, Ding Y Y, et al. FBXO5-mediated RNF183 degradation prevents endoplasmic reticulum stress-induced apoptosis and promotes colon cancer progression[J]. Cell Death Dis, 2024, 15(1): 33.
[29]	Yu Q, Zhang S H, Chao K, et al. E3 ubiquitin ligase RNF183 is a novel regulator in inflammatory bowel disease[J]. J Crohns Colitis, 2016, 10(6): 713-725.
[30]	Cui Y Y, Yang Y H, Zheng J Y, et al. Elevated neutrophil extracellular trap levels in periodontitis: implications for keratinization and barrier function in gingival epithelium[J]. J Clin Periodontol, 2024, 51(9): 1210-1221.