Coding and non-coding circulating RNAs associated with the presence and rapid expansion of abdominal aortic aneurysms

Abstract

IntroductionAbdominal aortic aneurysms (AAA) are pathological dilatations of the abdominal aorta. AAA may progressively enlarge and eventually rupture; ruptured AAA has a high mortality and morbidity even with timely surgical intervention. AAA diagnosis is usually made using imaging modalities such as duplex ultrasonography (USS) or computed tomography (CT). While there has been increasing interest in medical management of AAA, no treatment other than surgery has been found to adequately treat the condition. Non-coding RNAs such as microRNAs, which regulate gene expression at the post-transcriptional level, are increasingly being investigated for their use as diagnostic and prognostic biomarkers as well as therapeutic targets in AAA. One barrier to their clinical utility is a lack of specificity for AAA compared to other cardiovascular diseases such as ischaemic heart disease (IHD) and peripheral artery disease (PAD). This study aimed to identify circulating non-coding RNAs associated with AAA presence and growth, as well as to investigate their downstream targets to provide insights into their pathophysiological roles in AAA development and progression.MethodsA two-stage prospective case-control study was conducted using participants recruited to the Vascular Database and Peripheral Vascular Biobank at James Cook University.To identify circulating non-coding and coding RNAs associated with AAA presence, serum and whole blood samples from 36 age- and sex-matched participants, comprising 12 cases with AAA, 12 cases with PAD, and 12 healthy controls, were used. AAA was excluded using USS or CT imaging for the PAD and healthy control groups, and PAD was excluded using ankle-brachial pressure indices (ABPI) for the AAA and healthy control groups. Demographic details, comorbidities and medications were recorded for all participants. Using a commercially-available and validated RNA extraction kit, total RNA was extracted from the serum and blood samples and analysed for expression of 800 non-coding RNAs using the Human miRNA v3 Assay Panel on nanoString Technologies’ nCounter Analysis System for serum miRNA, and the Illumina HiSeq 2500 platform for coding RNA from blood cells. Results were analysed using nanoString Technologies’ nSolver Analysis Tool and DESeq2 Bioconductor v3.5 tool, with statistical analyses performed using the appropriate categorical, parametric and non-parametric tests in SPSS v23. A larger cohort of 107 cases with AAA was then recruited to validate the findings from the initial serum analysis using the same methods. Downstream coding RNA (mRNA) target and pathways analysis was conducted for differentially-expressed miRNAs using online prediction tools.To identify circulating non-coding RNAs associated with AAA growth, serum samples were identified from 106 AAA cases who underwent 12-monthly serial CT assessments over a 12 – 24-month period. Maximum orthogonal AAA diameter was recorded in millimetres (mm) from each CT using a semi-automated protocol whose inter- and intra-observer reproducibility has been previously established. Annual AAA growth was calculated from these measurements using a linear mixed effects model. Clinical detail recording, RNA extraction, analysis, and statistical methods were performed as described previously.ResultsSix miRNAs from serum, and two coding RNAs from blood cells were identified as significantly differentially-expressed in the AAA group compared to the PAD and healthy control groups in the discovery phase. After validation in a larger AAA cohort, only one miRNA in serum (let-7b-5p) remained significantly differentially-expressed. let-7b-5p was 1.388-fold upregulated in the serum of the AAA cohort compared to the healthy control group, and 0.761-fold downregulated in the PAD cohort (p<0.001). Receiver operator characteristic (ROC) curve analysis demonstrated an area under the curve (AUC) of 0.918 for let-7b-5p for diagnosing AAA compared to the other two groups, marking it as a potential biomarker for AAA diagnosis. Downstream target and pathways analysis revealed biologically plausible mRNA targets for let-7b-5p, including one of the two identified differentially-expressed coding RNAs (SUB1) in whole blood of AAA patients, functionally implicating let-7b-5p in AAA development.The AAA growth cohort was divided by median annual AAA growth rate in mm into two equal groups. Serum miR-1268a was significantly downregulated in the fast-growing AAA group compared to the slow-growing AAA group (0.707-fold, p = 0.043). ROC curve analysis demonstrated an AUC of 0.618 for miR-1268a for diagnosing fast-growing AAA, indicating this miRNA is not suitable in isolation as a biomarker for AAA growth. However, downstream targets and pathways analysis revealed biologically plausible mRNA targets of miR-1268a, functionally implicating it in AAA progression.ConclusionIn this study, one circulating miRNA in serum (let-7b-5p) was associated with AAA diagnosis, while another miRNA (miR-1268a) was associated with fast-growing AAA. let-7b-5p demonstrates promise as a biomarker for AAA diagnosis, while miR-1268a was not suitably sensitive or specific for independent biomarker use to predict AAA progression. The downstream targets and pathways associated with these miRNAs potentially implicate them in AAA pathogenesis and progression. Further larger focussed studies are required to validate these findings.