Circulating mRNA and miRNA Signatures Used as Predictive Markers of Lifestyle Modification Outcomes in Pediatric...

The response to lifestyle modification (LSM) in children with obesity is variable and difficult to predict. The aim of the study was to explore the relationship between circulating miRNA and mRNA expression patterns and the individual response to simple LSM in pediatric obesity: (i) identification of specific miRNAs/mRNA associated with successful weight reduction following LSM and (ii) use of these associations to inform the development of more targeted and effective intervention strategies.

Out of 240 children with obesity (BMI >97%) recruited to a prospective "multi-OMICS" study granted by ESPE Research Unit, 159 subjects (age 8-17 years, median 12.8 years; 45% females) finished 12 months (between visits V0-V12) of LSM obesity management at three clinical centers in Poland, Turkey, and Italy. From >180 baseline clinical and laboratory variables, four prespecified key features at V0/before LSM (sex, age, severity of obesity, and steroid metabolome) and delta z-score BMI V12-V0 were used to stratify and select 50 representative samples for RNA-seq. We searched for circulating mRNA and miRNA capable of establishing the link with positive response to LSM defined by a decrease in z-score BMI (delta z score BMI V12-V0<0, responders/RS). First, the candidate gene approach with previously documented STAT3, CORO1C, SERPINH1, MVP, ITGB5 (upregulated in obesity), PCM1, SIRT1, EEF1G, PTEN, and RPS2 (downregulated in obesity) hub genes was applied. The pairwise ratios of transcripts per million (TPMs) of these genes were tested for statistical significance between RS (n = 31) and non-RS (n = 19) groups. Second, the TPM expression values of 2,129 miRNAs and 33,246 RPL13A-normalized mRNAs were compared between RS and non-RS to get a list of putative significant genes (p < 0.05; the Welch t test). Enrichr, a Web-based tool (https://amp.pharm.mssm.edu/Enrichr), was used for analyzing gene sets.

Expression values for 909 mRNA and 24 miRNA genes were significantly different between RS vs. non-RS. For example, STAT3/PTEN, CORO1C/PTEN, MVP/PTEN; miRNA-10a, miRNA-122, miRNA-127, miRNA-502, miRNA-210 were significantly higher and SIRT1/PTEN, EEF1G/PTEN, RPS2/PTEN, miRNA-27a were lower comparing RS vs. non-RS (p = 0.009, 0.022, 0.030, 0.048, 0.029, 0.045, 0.035, 0.020, 0.047, 0.04, 0.039, 0.029, respectively). Several pathways, including "vitamin B6 metabolism" and "insulin receptor signaling," were enriched in our differentially expressed mRNA (p < 0.001): PNPO, PHOSPHO2, PDXP, PSAT1 were upregulated in RS vs. non-RS. "Leptin signaling pathway WP2034" was enriched in STAT3 and PTEN (p < 0.0001).

The role of the balance between STAT3 and PTEN in LSM was discovered. The importance of previously described miRNA markers (miRNA-122, miRNA-127, miRNA-27a) related to adipogenesis risk, insulin resistance, hepatic fat reduction, PPAR-γ expression was confirmed. A combination of the previously reported leptin feature together with STAT3/PTEN and miRNA-27a levels showed promising predictive results. Our findings not only showed the predictive molecular markers but also pointed to their role in personalization of therapeutic management.