Transcriptomic Profiling MicroRNA and Non-Coding RNA from Whole Blood in African Americans with MASLD

1. Introduction

Metabolic dysfunction-Associated Steatotic Liver Disease (MASLD), formerly known as non-alcoholic fatty liver disease (NAFLD), is recognized as the most prevalent chronic liver disease worldwide, affecting approximately 25–30% of the global adult population [1]. MASLD encompasses a spectrum of liver disorders, ranging from simple hepatic steatosis to non-alcoholic steatohepatitis (NASH), progressive fibrosis, cirrhosis, and hepatocellular carcinoma (HCC) [2]. MASLD is strongly associated with metabolic syndrome, obesity, insulin resistance, type 2 diabetes mellitus (T2DM), and dyslipidemia, making it a major contributor to liver-related morbidity and mortality [3]. Despite its high prevalence, MASLD remains underdiagnosed and poorly understood in diverse ethnic populations, limiting the development of targeted therapeutic strategies [4].

A limited, but increasing, number of studies have underscored the regulatory roles of non-coding RNAs, particularly microRNAs (miRNAs), and other non-coding RNAs (ncRNAs) in the development and progression of MASLD [10,11]. The miRNAs are small non-coding RNA molecules that post-transcriptionally regulate gene expression by targeting mRNAs for degradation or translational repression. Small, non-coding RNA molecules also play a crucial role in regulating gene expression by targeting mRNA transcripts [39] by binding to the 3’ untranslated region (3’UTR) of target mRNAs, leading to mRNA degradation or inhibition of translation. This post-transcriptional regulation of gene expression helps control various cellular processes, including cell growth, development, and differentiation [39]. miRNAs play crucial roles both within cells and in the extracellular environment by regulating gene expression intracellularly and functioning as intercellular messengers when secreted into extracellular fluids. They are transported via exosomes or RNA-binding protein complexes such as those involving AGO proteins to modulate gene expression in target cells [40]. Previous studies have reported the relationship of multiple miRNAs in hepatic lipid metabolism, insulin signaling, inflammation, and fibrosis [12,13]. Similarly, other ncRNAs have emerged as critical modulators of hepatic gene expression, influencing hepatocyte apoptosis, inflammation, and fibrosis [14]. However, there have been few attempts to coordinate expression patterns of non-coding RNAs with the altered gene expression patterns in MASLD subjects, especially within the same individuals.

To determine the feasibility of addressing this issue, a case-control pilot study of the expression patterns of ncRNAs was conducted using individuals from a previous [7,8] transcriptomic investigation conducted by our group. This population was comprised of African Americans (AA) to address the lack of inclusion of minority participants in prior studies of transcriptomic patterns in MASLD patients. This study utilized whole blood samples which we and others have demonstrated to have substantial utility in establishing transcriptomic patterns that largely overlap patterns observed in hepatic tissue [7,8]. The current pilot study, to the best of our knowledge, reports for the first time the transcriptomic profiling study of miRNA and other ncRNA expressions in whole blood samples from AA individuals with MASLD.

2. Results

2.1. Study Participants

All selected participants (total n=8) were AAs and were part of the previous transcriptomic study from the Washington DC area [7]. The participants were separated into two groups: a control group of individuals without MASLD, and a case group of individuals with early stage MASLD. MASLD was diagnosed based on standard criteria including confirmed hepatic steatosis (based on their imaging/biopsy records supported by the presence of hepatic steatosis on cross-sectional imaging, liver elastography, and/or histological confirmation by percutaneous liver biopsy) and exhibited one or more comorbid metabolic features, viz., type 2 diabetes, hypertension, hyperlipidemia, or obesity [7]. Table 1 displays the characteristics of each group. No significant differences in age, BMI, and Hba1c were observed. To select the participants, we checked the data of BMI, HbA1c value, dyslipidemia, and presence of steatosis (as measured by FibroScan). For the purposes of liver steatosis staging, we used the S grade (S0 - S3);the higher the grade, the higher percentage of liver affected by fatty changes.

2.2. miRNA and Noncoding-RNA Sequencing, Differential Expression, Number of Reported Studies

Out of a total of 1412 miRNA identified transcripts, 35 miRNAs in the MASLD cases were significantly differentially expressed when compared to the controls (fold change cutoff 1.5-fold and p-value < 0.05) with 24 downregulated and 11 upregulated (Figure 1A and Table 2). Out of a total of 5423 other ncRNAs transcripts, 28 were significantly differentially expressed with 17 downregulated and 11 upregulated in the MASLD cases compared to the controls (Figure 1B and Table 3).

We also examined the differential miRNA expression of each individual case compared to the control group’s expression status (Figure 2A). We observed miR-1299, miR-193a-5p, miR-185-3p, miR-3960, miR-1343-5p, and miR-224-5p, were significantly downregulated and miR-206 significantly upregulated in all the MASLD subjects.

3.4. Top Biofunctions, Canonical Pathways, and Network Analysis

The IPA analysis revealed significant biofunctions, including “fibrosis of liver” and “cirrhosis of liver,” with significant overlap percentage (p-value <0.05). These biofunctions were associated with 10 distinct miRNAs: miR-100-5p, miR-1273h-5p, miR-130a-3p, miR-133a-3p, miR-135a-5p, miR-143-3p, miR-16-5p, miR-199a-5p, miR-27a-3p, and miR-526a-5p (Figure 2A, 2B).

In addition to the miRNA findings, other ncRNAs were prominently featured in the network analysis (Figure 3). Several long non-coding RNAs (lncRNAs), such as LINC00963, SNHG7, CYTOR, and HORMAD2-AS1, were identified as key regulators in the hepatic fibrosis and cirrhosis pathways. These lncRNAs interacted with critical molecular hubs, including EZH2, AKT, and YAP1, highlighting their regulatory roles in fibrosis-related processes. The network further identifies ncRNAs, such as RP11 and LINC01234, contribut to the modulation of gene expression, emphasizing their potential involvement in liver disease pathogenesis. The canonical pathway analysis (Figure 4) highlighted the “hepatic fibrosis signaling pathway” as a central mechanism linking these ncRNAs and miRNAs to critical molecular and cellular functions. The interactions between miRNAs, lncRNAs, and target genes suggest a tightly regulated network underlying MASLD.

3. Discussion

In this pilot study, we present the first transcriptomic analysis of miRNA and other ncRNA expression profiles in whole blood samples from AA individuals with MASLD. By identifying differentially expressed ncRNA species, our work aims to generate hypotheses on regulatory pathways involved in MASLD development.

We observed that, among the most commonly dysregulated miRNAs , miR-206, miR-1299, miR-224-5p, and miR-193a-5p emerged as key candidates modulating hepatic lipid metabolism, insulin signaling, and fibrogenesis. miR-206, which plays a dual role in lipid metabolism and insulin sensitivity, has been shown to inhibit hepatic lipogenesis and gluconeogenesis, thereby promoting insulin responsiveness [23]. Consistent with these findings, our data reveal a 2.5-fold upregulation of miR-206 in all four MASLD patients, suggesting a protective response to hepatic insulin resistance [24].

miR-224-5p is known to be involved in multiple regulatory processes such as lipid accumulation, endoplasmic reticulum stress, mitochondrial damage, inflammatory response, autophagy, and hepatic stellate cell activation, potentially influencing the progression of MASLD [20,25]. Multiple studies have reported an upregulation trend of miR-224-5p, in serum, plasma, or liver biopsy samples [18,19,20,21]. Upregulated miR-224-5p targets the leptin (LEP) gene and leads to its suppression through dysregulation of the AMPK pathway, which is associated with MASLD progression [18]. Interestingly, in the current study we observed a significant downregulation of miR-224-5p in all four cases, with an average -2.65-fold change. These findings are consistent with the observation that AAs experience a slower progression of MASLD to cirrhosis and liver-related events compared to other racial groups [26].

miR-1343-5p is known to act as a potent repressor of TGF-β signaling and fibrosis through the direct attenuation of both canonical TGF-β receptors [28]. The downregulation of miR-1343-5p in all four MASLD patients in our study supports our previously published findings, which noted an upregulation of TGF-β and activation of the hepatic fibrosis signaling pathway in these subjects [7]. Another previous study reported that upregulation of circulating miR-1343-5p is a potential biomarker in MASLD in adolescents with severe obesity, [27]. Although our study included only adults’ patients with MASLD, it is well reported that the dysregulation of TGF-β (Transforming Growth Factor-beta) signaling is associated with obesity and its related metabolic disorders [84].

Notably, the significant downregulation of miR-193a-5p in the current study, which has been previously reported as a biomarker for liver fibrosis and cirrhosis [29,30,31,32,33,34], reinforces its potential role as an early indicator of MASLD. The downregulation of miR-193a-5p is thought to inhibit pro-fibrotic gene targets, thereby promoting hepatic stellate cell activation, a central event in the fibrotic cascade [35]. Our prior research identified dysregulation of the TGFB1 and E2F1 genes and associated pathways in peripheral blood samples in the cohort of AA patients with early-stage MASLD [7] used for the current study. We observed the activation of hepatic fibrosis signaling pathways and their potential role in the development of hepatocellular carcinoma, particularly when TGFB1 was upregulated and E2F1 was downregulated [7]. However, the study did not establish a definitive role for TGFB1 and E2F1 regulation in the development of hepatic steatosis or the lower prevalence of MASLD in AAs. Some previous studies have reported an upregulation trend of miR-193a-5p in blood serum and plasma samples, mainly in Caucasian populations [29,30,31].

In addition to miRNAs, our pathway and network analyses identified several long non-coding RNAs (lncRNAs) involved in fibrosis-related signaling networks. LncRNAs such as LINC00963, SNHG7, CYTOR, and HORMAD2-AS1 interact with pivotal regulatory molecules including EZH2, AKT, and YAP1. EZH2, a histone methyltransferase, has been implicated in the progression of fibrosis, while YAP1, a core component of the Hippo signaling pathway, plays a critical role in liver regeneration and fibrogenesis [37]. Our canonical pathway analysis indicates strong associations between several ncRNAs (LINC00963, HCG18, ST7-AS1, RP11_25D31, CYTOR and LINC01234) and hepatic fibrosis pathways. The notable upregulation of WDFY3-AS2 and LINC02767 suggests that these lncRNAs may contribute to hepatic inflammation and fibrotic remodeling, although further validation is required in larger populations [38].

Although our study offers new hypotheses on the transcriptomic landscape of MASLD in African Americans and demonstrates the utility of whole blood samples for such investigations, it also has some limitations. The small study size restricts the generalizability of our findings and is vulnerable to selection bias. We focused on patients with early stage MASLD, for whom future interventions might prevent the progression of the disease. Additionally, we were unable to segregate our data to analyze other ncRNAs individually for each patient, as the data was analyzed and provided in group-wise format by the manufacturer of the analysis tools used, which needs to be addressed for future investigations. Future studies should incorporate larger, multi-ethnic cohorts and additional analytical tools to enhance statistical power and validate these preliminary observations. As our analysis was based on whole blood RNA profiles, the findings may not fully recapitulate hepatic transcriptomic alterations. Future research therefore should attempt to include matched analyses of liver tissue with the circulating plasma miRNAs to provide a more comprehensive understanding of MASLD pathogenesis.

4. Materials and Methods

4.1. Study Particepants and Blood Sample Collection.

In this pilot study, the study participants consisted of eight individuals (control n=4; MASLD n=4), with equal numbers of males and females, who self-identified as AA and were born in the USA. All participants responded to an advertisement through Howard University and Georgetown University Community Newsletter via email and/or flyers and public announcements and were recruited with their informed consent. The protocol was approved by Georgetown-MedStar IRB (MODCR00002260). Participants with MASLD were recruited from the MedStar Georgetown Transplant Institute. We selected only those patients who were at early stages of development MASLD (confirmed hepatic steatosis and exhibited one or more comorbid metabolic features, viz., type 2 diabetes, hypertension, hyperlipidemia, or obesity). Individuals with progression to severe fibrosis or cirrhosis were not included because of the small size of the participant groups and the wide spectrum of tissue features present during different stages of liver disease; we chose to limit subject inclusion to earlier stages of liver disease to increase the homogeneity of disease presentation in the different subjects. Individuals with severe fibrosis or cirrhosis were not included; we chose to limit subject inclusion to earlier stages of liver disease to increase the homogeneity of disease presentation in the different subjects (Table 1). Patients with other potential causes of liver disease, including viral, immunological, iron storage disease, Wilson disease, or alpha 1 antitrypsin deficiency, were excluded from the study. Participants with heavy alcohol use were also excluded from the study. Control participants were those who responded to the same flyers and advertisements described above, but did not have MASLD. These individuals were negative for self-reported HCV and HBV and had normal liver enzyme profiles. A questionnaire was provided to all participants to collect demographic and clinical information.

4.2. RNA Extraction and miRNA Library Preparation

Whole blood was collected in a DNA/RNA Shield™ Blood Collection Tube (Manufacturer: Zymo Research, Cat # R1150) during recruitment by experienced phlebotomists. Blood collection tubes were prefilled with 6 ml DNA/RNA Shield™ for direct collection of up to 3 ml whole human blood. DNA/RNA Shield lyses cells, inactivate nucleases and infectious agents (e.g., viruses and pathogens), and is ideal for safe sample storage and transport at ambient temperatures. RNA was extracted from DNA/RNA Shield tubes using the Quick-DNA/RNA™ Blood Tube Kit (Zymo Research, Cat. # R1151) according to the manufacturer’s instructions. DNA contamination was removed using an Applied Biosystems Inc. (ABI) DNA-free kit (ThermoFisher, CA, Cat # AM 1906). RNA was quantified using a NanoDrop™ One spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA). The ratio of absorbance at 260 and 280 nm was used to assess the purity of DNA and RNA.

We used Zymo-Seq™ miRNA Library Kit (Catalog Numbers: R3006, R3007) to generate small RNA libraries. To acquire miRNAs, present in total RNA or cell-free RNA extracted from biofluids we followed manufacturer’s instructions and protocol. Briefly the protocol was with five major steps, which are adapter ligation and blocking, circularization and dimer removal, reverse transcription, index PCR, and library purification. At the time of index PCR pre-mixed forward and reverse primers with the following sequence:

Forward Primer Sequence:

5′-AATGATACGGCGACCACCGAGATCTACACNNNNNNNN(NN)ACAC TCTTTCCCTACACGACGCTCTTCCGATCT-3′

Reverse Primer Sequence:

5′-CAAGCAGAAGACGGCATACGAGATNNNNNNNN(NN)GTGACTGGA GTTCCTTGGCACCCGAGAATTCCA-3′

4.3. Sequencing and Data Analysis

The Sequencing was performed by Zymo Research (Irvine, California, US). Data analysis was performed according to instruction mentioned in the Zymo-Seq™ miRNA Library Kit (Catalog Numbers: R3006, R3007). To read the Zymo-Seq™ miRNA libraries we used bioinformatics tools (QIAGEN CLC Genomics Workbench) designed for Illumina’s TruSeq Small RNA libraries. Prior to sequence alignment, sequenced reads processed with adapter trimming. For the trimming we used sequence of TGGAATTCTCGGGTGCCAAGG. In the final analysis data was extracted in the form of fold change and p-values keeping false discovery rate (FDR) at the level of p<0.05.

4.4. Ingenuity Pathway Analysis (IPA)

IPA was utilized to explore complex biofunctions within a biological system, identifying functional roles, molecular processes, and key networks associated with significantly differentially expressed genes in participants with MASLD. From the differential expression of miRNAs and other ncRNAs datasets described above, the identification of cellular processes and pathways by IPA (Qiagen, USA) was performed according to the methods described in our earlier study [15,16,17]. Briefly, datasets comprising miRNAs and other ncRNAs identifiers and corresponding expression values (fold-change) from the sequencing data were imported into IPA. Differentially expressed identifiers (miRNAs and other ncRNAs) were mapped to related changes in biofunctions [16]. The networks were generated algorithmically based on their connectivity. Using IPA, we identified the top network by amalgamating a large set of differentially expressed miRNAs and other ncRNAs with the goal of uncovering the most extensive array of relationships among the focus genes [13]. A score (P-score = − log10 (p-value)) according to the fit of the set of supplied genes and a list of biological functions stored in the Ingenuity Knowledge Base are generated [15] Networks were “named” on the most prevalent functional group(s) present. Canonical pathway analysis identified function-specific genes that were significantly present within the networks.

5. Conclusions

Our study is among the first to generate hypotheses on the significance of miRNA and ncRNA expression patterns in MASLD among AAs. The pathway analyses reveal correlations between these ncRNA patterns and the transcriptomic patterns of coding genes, particularly in pathways involving TGFB1 and E2F signaling. The findings also suggest that miR-206 upregulation may represent a protective response to insulin resistance. miR-206, miR-370-3p, and miR-193a-5p downregulation could contribute to MASLD pathogenesis via impaired lipid metabolism and fibrosis promotion. Other ncRNAs such as LINC00963, SNHG7, and CYTOR are implicated in hepatic fibrosis signaling. These findings highlight promising candidates for future biomarker investigation and therapeutic targeting. Further large-scale studies, including longitudinal transcriptomic profiling, will be essential to elucidate the role of ncRNAs in MASLD progression and ethnic disparities.