The Effect of Submaximal Doses of Semaglutide on the Expression of Microribonucleic Acid in Patients with Obesity: A Pilot Study

Introduction

Obesity is considered a chronic metabolic multisystemic disease, with excessive accumulation and/or improper distribution of body fat [1]. Obesity is a multifactorial disease caused by a combination of endogenous (genetic) and exogenous factors such as excessive energy intake and lack of exercise [1]. In clinical practice, it is usually defined by BMI over 30 kg/m² [2]. The prevalence and incidence of obesity are increasing in both developed and developing countries. In 2022, the World Health Organization (WHO) reported that 890 million (16%) of adults over the age of 18 were obese [2]. The number of obese individuals has more than doubled since 1990, and it is predicted that the prevalence of obesity will continue to grow in the following decade [3]. Overweight and obesity cause nearly 2.8 million deaths worldwide each year while being responsible for 35.8 million (2.3%) of all disability-adjusted life years [4, 5]. Those affected require multidisciplinary treatment including non-pharmacological methods, pharmacotherapy, surgical and endoscopic procedures [1].

Glucagon-like peptide 1 receptor agonists (GLP-1RAs), which were initially developed for the treatment of type 2 diabetes mellitus (T2DM), have been identified as one of the most promising therapeutic agents in the treatment of obesity [6]. GLP-1RAs have several effects on various organ systems, combining central and peripheral actions that promote satiety, decrease hunger, decrease appetite, and ultimately lead to reduced food intake [7]. Long-acting GLP-1RA semaglutide has proven to be the most effective GLP-1RA regarding weight loss [8]. A meta-analysis of randomized controlled trials showed that the use of once-weekly 2.4 mg semaglutide, compared with placebo, was associated with a substantial long-term relative (−12.1%) and absolute (−12.3 kg) weight loss [9]. A third of participants randomized to semaglutide achieved at least a 20% reduction in their body weight by the end of the treatment period compared with 2.2% of participants randomized to placebo [9]. However, high (maximal) doses of semaglutide might be connected with increased risk of side effects and treatment intolerance, which might lead to premature treatment discontinuation and treatment failure [9].

Furthermore, studies have also shown other beneficial properties of GLP-1RA such as their cardioprotective and nephroprotective effects, which seem to be independent of weight loss [10, 11]. Research is being conducted in several other areas such as metabolic dysfunction-associated steatotic liver disease (MASLD), polycystic ovary syndrome, addictions, and neurodegenerative diseases [12, 13]. However, the exact molecular mechanisms for the listed GLP-1RA effects remain unclear. It has been speculated that weight loss independent effects of GLP-1RA are mediated through the regulation of microribonucleic acid (miRNA) expression [14].

miRNAs (miRs) are small non-coding ribonucleic acids (RNAs) consisting of 21–25 nucleotides (most commonly 22), which serve as key modulators of protein expression [15]. It is estimated that more than 45,000 miRNA target sites are present in the human deoxyribonucleic acid, affecting nearly 60% of coding genes [15]. Research suggests that miRNAs play a role in numerous metabolic diseases such as T2DM and lipid spectrum disorders [16]. It has been found that miR-126 (miRNA) expression is significantly reduced in both impaired fasting glucose and T2DM individuals, miR-375 has been proposed as an important regulator of pancreatic β-cell function, and miRNA miR-1249, miR-320b, and miR-527 have been identified as biomarkers of T2DM susceptibility [16–18]. Furthermore, obesity increases the expression of miR-132 and miR-212 in the islets of diabetes-resistant and diabetes-susceptible mice suggesting that the mentioned miRNAs may have a role in the pathogenesis of T2DM [19]. It has been suggested that miR-122, which is most abundantly expressed in the liver-regulating triacylglycerol (TAG) and cholesterol metabolism, plays a pivotal role in nonalcoholic fatty liver disease (NAFLD) progression providing, perhaps, a novel therapeutic target [20]. Higher levels of miR-34a were found in patients with NAFLD and mice models, while its inhibition in mice and tissue cultures significantly improved steatosis and TAG plasma levels suggesting that miR-34a may be a novel pharmaceutical target for NAFLD [21, 22]. It has been found that naturally occurring GLP-1 and also synthetic GLP-1RA can modulate the expression of different miRNAs. In the pancreas, GLP-1 agonists increase the expression of miRNA-212 and miRNA-132, stimulating insulin secretion [23]. GLP-1RA also decreases miRNA-338 levels leading to an increase of promoter factor 1 insulin (PDX1) gene expression, followed by increased insulin secretion and an improvement of pancreatic β-cell function [24]. A short-acting GLP-1RA liraglutide has been shown to have protective effects against pancreatic β-cell apoptosis by downregulating miR-139-5p [25]. GLP-1RAs also enhance β-cell responsiveness by promoting GLP-1 receptor expression through the suppression of miR-204 [25]. With respect to MASLD, GLP-1RAs are able to downregulate miRNA-34a and miRNA-21 and upregulate miRNA-200b and miRNA-200c expression in the liver, reducing intrahepatic lipid accumulation and liver steatosis [24]. It has been suggested that cardiovascular (CV) protective effect of GLP-1RA in coronary artery disease is mediated via modulation of miRNA expression, since miR-203a-3p and miR-429 have shown significantly decreased expression after treatment with GLP-1RA [14]. Semaglutide, liraglutide, and dulaglutide have been reported to reduce systolic and diastolic blood pressure, and exenatide-induced miR-29b-3p has demonstrated a protective effect against diabetic cardiomyopathy as well [25–27]. Moreover, liraglutide modulates critical regulators of endothelial cell function and atherosclerosis, including miR-93-5p, miR-26a-5p, and miR-181a-5p [25]. GLP-1RAs also participate in the regulation of exosomal miRNAs, such as miR-192, which take part in the development of fibrosis and inflammation in T2DM microvascular complications such as diabetic kidney disease and diabetic retinopathy [25]. MiRNAs have also been shown to predict response to GLP-1RA therapy in patients with T2DM. Higher miR-378-3p and miR-126-3p levels are related to better glycemic control and lower glycosylated hemoglobin after 1 year of treatment. MiR-375-5p was also reported to be a predictor of glycosylated hemoglobin levels. The highly expressed miR-375 in pancreatic islets can be considered a biomarker for assessing the cytoprotective action of GLP-1RAs on β-cells. Lastly, baseline miR-15a-5p can predict weight loss using GLP-1RA [25, 28]. However, in spite of all the aforementioned findings, there is still a lack of studies examining the effect of semaglutide on the expression levels of miRNA in nondiabetic individuals with obesity. Therefore, in this study we aimed to investigate the effect of semaglutide therapy administered at submaximal doses on changes in the expression of plasma miRNA in patients with obesity.

Based on the data described above, we hypothesized that the administration of submaximal doses of semaglutide (together with standard non-pharmacological care) would lead to changes in expression of miRNA. We wanted to describe the changes in known metabolism-related miRNA, as well as the changes of those miRNAs which were not previously described, and therefore performed the descriptive analysis using a novel microarray assay for miRNA analysis. The specific objectives of our study were set as follows:

• 1.

  To identify and describe the changed miRNA after the course of the submaximal semaglutide administration;

• 2.

  To quantify the changes in miRNA;

• 3.

  To try to perform pathway analysis of the identified miRNA which levels changed after the therapy.

Materials and Methods

Results

Discussion

Our study observed trends in the expression of 8 miRNAs 6 months after the initiation of semaglutide therapy dosed sub-maximally (with standard non-pharmacological therapy of obesity) compared to pretreatment levels. Three miRNAs (hsa-miR-1273f, hsa-miR-4644, and hsa-miR-6071) had nominally decreased levels of expression (downregulation), while 5 miRNAs (hsa-miR-151b, hsa-miR-223-3p, hsa-miR-27b-3p, hsa-miR-29c-3p, and hsa-miR-4284) showed nominally increased expression (upregulation). The lack of FDR-significant findings presumably reflects the limited sample size and the stringent correction required for high-dimensional data. However, the primary objective of the present study was not biomarker discovery but to provide an initial, hypothesis-generating characterization of miRNA expression changes in this clinical context. The 6-month duration of semaglutide treatment in this study was deemed sufficient to show any potential significant changes based on the experience of investigators and improvements observed in the anthropometric parameters of the study group. As already mentioned, submaximal doses of semaglutide were used. The rationale behind this relied on the fact that greater doses of semaglutide were difficult to obtain due to the country’s drug administration policy as only 0.25 mg, 0.5 mg, and 1.0 mg injection pens were at disposal while a 2.4-mg injection pen was not available at all. Nevertheless, the patients did achieve significant weight reduction (defined as a loss of at least 5% of the initial body weight over the first 3 months of treatment) [1], represented by an 11-kg difference between the mean baseline weight and the mean weight after 6 months of GLP-1RA therapy as shown in Table 1.

With respect to the basic demographic, anthropometric, laboratory parameters and comorbidities of the study group, the mean BMI of the investigated patients corresponded to class 2 obesity according to the WHO. In fact, the anthropometric parameters were all improved after 6 months of GLP-1RA treatment as expected (see Table 1). Regarding the baseline laboratory parameters, the patients had, on average, normal kidney function, normal blood count, minerals, and no significant elevation in hepatic enzymes despite the fact that nearly a third of all participants presented with MASLD. A mild elevation in serum total cholesterol and serum TAG was detected as well. With respect to the posttreatment parameters, a statistically significant reduction was observed in AST, ALT, total cholesterol, and LDL levels during 6 months of semaglutide treatment at submaximal doses suggesting a potential effect of semaglutide on hepatic function as well as on the lipid profile even at lower doses than those reported in major clinical trials [10]. However, it is still unclear whether this metabolic effect is a function of the GLP-1RA itself or the weight reduction achieved due to the complex treatment of obesity.

Upon reviewing data in the current literature, to the best of our knowledge, currently, there are no studies analyzing the effect of semaglutide on expression levels of miRNA in individuals with obesity, and therefore the results of this study cannot be compared or validated with any other study. When analyzing the available data, we identified some studies, which attempted to determine the effect of GLP-1RA treatment on the expression of miRNA, but predominantly in patients with T2DM and additional CV or metabolic diseases [24, 25, 28]. However, no study examined such an effect on miRNA expression levels in patients with obesity in particular, in whom GLP-1RA semaglutide was the cornerstone of pharmacotherapy. The rationale for performing such a study stems from the fact that semaglutide showed a reduction in CV risk and mortality in obese patients even in the absence of T2DM [10]. But the exact mechanism, which can explain this GLP-1RA-induced effect, has yet to be elucidated. Additionally, despite the well-known positive effect of GLP-1RA on weight reduction and CV benefits in obese patients, there are still some patients, who do not respond to treatment. Identifying such individuals prior to or shortly after the initiation of GLP-1RA therapy could minimize the cost of treatment and enable physicians to tailor obesity treatment more effectively. As a matter of fact, certain studies showed that GLP-1 and GLP-1RA can modulate the expression of different miRNAs, which might potentially mediate cardioprotective effects of GLP-1RA in obese patients [24]. Our study also showed that changes in the expression of several miRNAs significantly correlated with changes in certain laboratory parameters during the 6-month semaglutide therapy. However, there are no other studies for comparison, and these findings should lay out the foundation for further research into this area providing potential pathways, in which semaglutide acts on metabolism and important metabolic sites via miRNA modulation. Our study showed a direct correlation between hsa-miR-1273f and AST, ALT as well as a number of miRNAs correlating with TAG as mentioned above. Furthermore, miRNAs have been extensively studied as novel biomarkers of disease progression and/or of therapeutic response to various drugs as well. However, the role of these miRNAs as predictors of treatment responsiveness to GLP-1RA in patients with obesity has not yet been investigated, which makes our study unique and a first of its kind in this regard.

Liu et al. [14] studied the effect of GLP-1RA on miRNA expression in patients with T2DM and coronary artery disease with an aim to determine whether specific miRNA mediates cardioprotective effects of GLP-1RA. Two miRNAs were identified with significantly decreased expression, miR-203a-3p and miR-429, with the study concluding that the cardioprotective effects of GLP-1RA are mediated via the suppression of the aforementioned miRNA. However, the study did not focus specifically on patients with obesity, and the miRNAs identified were not found to have significant changes of expression in our study. Although semaglutide was reported to upregulate miR-5120 in the blood as well as the liver of mice while suppressing the expression of ABHD6’s mRNA and protein related to miR-5120, which was observed in several metabolic diseases including obesity [32], our study did not find any significant changes in the expression of this particular miRNA. Lastly, Formichi et al. [28] identified miR-223-3p as a potential predictor of treatment effectiveness of GLP-1RA therapy in 26 patients with T2DM and demonstrated that patients with higher baseline levels of miR-223-3p had a more favorable glycemic outcome after a year-long GLP-1RA therapy. miR-223-3p has also been associated with adipocyte tissue dysfunction [33], which often develops prior to the onset of T2DM and is exacerbated by systemic inflammation, commonly observed in obesity. Indeed, our study found this particular miRNA to be upregulated by GLP-1RA after 6 months, which might open the discussion whether, to some extent, the metabolic and cardioprotective effects of GLP-1RA might be mediated via this biomarker. On the other hand, these two studies used different GLP-1RA agents, because while we administered semaglutide alone at submaximal doses, Formichi et al. [28] utilized liraglutide or dulaglutide in combination with metformin. Even so, it could be argued that there is a potential class effect of GLP-1RA on the expression levels of miR-223-3p in patients with T2DM and obesity, which could be explored further by studying the effects of GLP-1RA alone on the expression changes in T2DM and obesity in a larger sample of patients. Regarding the origin of this particular miRNA, it was found in hepatic and cardiac tissues in animal models of hyperlipidemia in addition to adipose tissue, particularly in preadipocytes exerting the abovementioned effects [33]. This observation is further supported by the KEGG pathway analysis of the miRNAs examined in this study (Fig. 3), which identifies the AGE-RAGE signaling pathway in diabetic complications as one of the most significantly enriched pathways [31]. These findings suggest that GLP-1RAs may exert favorable cardiometabolic effects, at least in part, by modulating the expression of miRNAs implicated in microvascular complications of T2DM, including miR-223-3p. In addition, the hepatic miR-223-3p was shown to be involved in cholesterol synthesis and release [34]. miR-27b-3p was shown to be present across various human tissues such as liver, adipocytes, or kidney; additionally, numerous studies have described its role as a tumor suppressor in the pathogenesis of hepatocellular carcinoma by inhibiting the NF-кB activity [35], suggesting a potential protective effect of GLP-1RA against carcinogenesis. miR-29c-3p is another miRNA, which showed increased expression with semaglutide therapy in our study and was shown to possess a critical role in oncogenesis, particularly as a potent tumor suppressor gene as well as having antifibrotic properties, since miR-29c-3p was found to be downregulated in conditions such as cardiac hypertrophy, congestive heart failure, diabetic nephropathy, and chronic hepatic injury [36, 37]. Interestingly, miR-29c-3p also regulates insulin signaling, thereby interacting with lipid and glucose metabolism and potentially T2DM [37]. miR-4284 has not been attributed to a particular human tissue or organ system so far; instead, it has been found in various cancer types; for example, it was upregulated in non-small cell lung carcinoma [38]. One study focused on miR-151b as a potential biomarker in acute ischemic stroke, and it was found to have significantly greater expression in the blood of the affected patients compared to healthy controls [39]. However, current literature does not offer evidence for any other specific human tissues to express miR-151b. Likewise, much less is known about the origin of other significantly altered miRNAs in our study such as miR-1273f or miR-6071, though the latter has been detected in human glioblastoma tissues [40, 41]. Furthermore, miR-4644 has been reported in plasma exosomes obtained from patients with bladder cancer with increased levels of expression compared to control subjects, providing yet another example of a potential use of miRNAs as biomarkers of cancer formation and progression, which could theoretically be impacted by semaglutide treatment [42]. Indeed, as far as we know, we have conducted the first study to demonstrate a possible function of miRNAs as biomarkers of treatment response to semaglutide in individuals with obesity; however, our findings need to be validated in further studies with larger samples of patients.

Limitations

There were a number of limitations in our study that need to be addressed. First, the sample size was relatively small. This can be explained to some extent by the fact that our study had to respect its financial limitations with regard to materials such as diagnostic kits for miRNA analysis and processing, which were expensive, and frankly, did not allow for the examination of a larger sample size. The limited sample size likely led to the aforementioned lack of FDR-significant findings. On the other hand, the aim of this study was to provide an initial, hypothesis-generating characterization of miRNA expression changes in the clinical context of possible weight-independent GLP-1RA effects and not to discover any new biomarkers. Second, some patients followed at our clinic had to be excluded due to the history of bariatric/metabolic surgery or endoscopic weight loss procedures in the past. Third, submaximal doses of semaglutide were used, and the titration protocol could not achieve the target doses of semaglutide as recommended in the literature due to the aforementioned reasons, the lack of a 2.4-mg dose in the country above all. Nonetheless, the real-world data as well as our clinical practice and experience suggest that the submaximal doses are effective at reducing weight and additional anthropometric parameters, which is certainly underlined by the results achieved in this study. Fourth, in our study, standard non-pharmacological therapy of obesity (reduced daily caloric intake by 500 kcal and increased physical activity) was recommended to all enrolled patients, which is in line with the current practice for the treatment of obesity. Unfortunately, we were unable to create a control sample of individuals with non-pharmacological approaches alone, as the patients enrolled in our study were unable to reduce their weight with non-pharmacological approaches alone and wanted to start the pharmacological therapy. As previously published studies showed possible effect of caloric restriction and exercise on both metabolic parameters and circulating miRNAs [43–46], one cannot determine whether the described changes in miRNA expression were achieved by semaglutide therapy alone, by non-pharmacological interventions alone, or by a combination of these. Although it is not likely that the observed changes were not determined by semaglutide administration, due to the lack of the control group, the effect of non-pharmacological intervention cannot be excluded. To answer this question, a study specifically examining the changes in miRNA expression (specifically those which tended to be changed) in these three subgroups needs to be performed.

Fifth, this study did not account for the impact of concomitant medications on possible changes in the expression of miRNA. But as already mentioned, patients with other pharmacological therapies for obesity were not included, and no patients with active oncological disease requiring cytostatic or any other gene-based therapy were studied either. In addition, most patients did not present with polypharmacy, generally taking treatment for common CV diseases such as arterial hypertension, hyperlipidemia, ischemic heart disease. Sixth, one could question the biological relevance of our observation, as none of the identified microRNAs correlate with anthropometric parameters, and only weak associations were observed with selected hepatic markers. As mentioned, our study was designed as a preliminary descriptive one. Thus, the real biological relevance of our observation remains, at this time, unexplained. Right now it remains unclear, whether the identified miRNA would be useful as diagnostic markers of optimal/in-optimal response on submaximal semaglutide administration. Further studies with larger patient samples and larger treatment duration would be needed for the clarification of this issue. Therefore, our study and our results should be still interpreted as primarily descriptive and associative, and all upper mentioned facts/limitations should be taken into consideration when the results of our study are interpreted.

Conclusion

This study demonstrated a nominally significant increase in the expression of hsa-miR-151b, hsa-miR-223-3p, hsa-miR-27b-3p, hsa-miR-29c-3p, and hsa-miR-4284 and a nominally significant decrease in the expression of hsa-miR-1273f, hsa-miR-4644, hsa-miR-6071 in patients with obesity after 6 months of semaglutide therapy at submaximal doses and standard non-pharmacological therapy (reduced daily caloric intake by 500 kcal and increased physical activity). However, these observations come from a pilot study, and the implications of these findings remain unclear. Nevertheless, the results advocate further research to elucidate the role of miRNA in predicting treatment responsiveness to GLP-1RA in patients with obesity.

Acknowledgments

The first two authors (Jakub Jurica and Zuzana Miertová) share their first authorship as they contributed equally to this work and manuscript preparation.

Statement of Ethics

This research was conducted according to ethical standards. The study protocol was formally reviewed and approved by the Ethics Committee of Jessenius Faculty of Medicine in Martin, ethical approval code EK 17/2019, date of approval April 24, 2019. All enrolled participants agreed with the study participation, and a written informed consent was obtained from participants to participate in the study.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

The study was funded and conducted as part of the long-term strategic research of prevention, intervention, and mechanisms of obesity and its comorbidities (project code in ITMS 2014+: 313011V344); sources from this project were used for study design and for laboratory equipment, laboratory reagents, and researcher salaries. The study was also supported by the Scientific Grant Agency of the Ministry of Education, Research, Development and Youth of the Slovak Republic and the Slovak Academy of Sciences (VEGA) (Grant Nos. 1/0549/19; 1/0090/20; 1/0145/22); sources from these projects were used for study indirect costs, statistical evaluation, data interpretation, and manuscript preparation.

Author Contributions

J.J., Z.M., M.S., T.B., P.G., M.Š., and E.H. designed the study; J.J., Z.M., and D.B. drafted the manuscript; J.J., Z.M., M.J.P., B.F., A.M., P.T., N.N., I.S., and M.P. collected, analyzed, and interpreted the clinical data, and performed the search of the literature; M.Š., D.D., D.B., I.B., A.K., and Z.D. performed the miRNA analysis, and collected, analyzed, and interpreted the laboratory data; T.B., P.G., M.M., E.H., and M.S. supervised the study and acquired the funding; T.B., M.M., E.H., and M.S. revised the manuscript critically. J.J. and Z.M. share first authorship, and their contribution to this work was equal. All the authors have read and approved the final version of the manuscript.

Funding Statement

The study was funded and conducted as part of the long-term strategic research of prevention, intervention, and mechanisms of obesity and its comorbidities (project code in ITMS 2014+: 313011V344); sources from this project were used for study design and for laboratory equipment, laboratory reagents, and researcher salaries. The study was also supported by the Scientific Grant Agency of the Ministry of Education, Research, Development and Youth of the Slovak Republic and the Slovak Academy of Sciences (VEGA) (Grant Nos. 1/0549/19; 1/0090/20; 1/0145/22); sources from these projects were used for study indirect costs, statistical evaluation, data interpretation, and manuscript preparation.

Data Availability Statement

The data that support the findings of this study are not publicly available due to legal and privacy restrictions; but these data are available from the corresponding author upon reasonable request.