MiRNAs Action and Impact on Mitochondria Function, Metabolic Reprogramming and Chemoresistance of Cancer Cells: A Systematic Review

MicroRNAs (miRNAs) are involved in the regulation of mitochondrial function and homeostasis, and in the modulation of cell metabolism, by targeting known oncogenes and tumor suppressor genes of metabolic-related signaling pathways involved in the hallmarks of cancer. This systematic review focuses on articles describing the role, association, and/or involvement of miRNAs in regulating the mitochondrial function and metabolic reprogramming of cancer cells. Following the PRISMA guidelines, the articles reviewed were published from January 2010 to September 2022, with the search terms “mitochondrial microRNA” and its synonyms (mitochondrial microRNA, mitochondrial miRNA, mito microRNA, or mitomiR), “reprogramming metabolism,” and “cancer” in the title or abstract). Thirty-six original research articles were selected, revealing 51 miRNAs with altered expression in 12 cancers: bladder, breast, cervical, colon, colorectal, liver, lung, melanoma, osteosarcoma, pancreatic, prostate, and tongue. The actions of miRNAs and their corresponding target genes have been reported mainly in cell metabolic processes, mitochondrial dynamics, mitophagy, apoptosis, redox signaling, and resistance to chemotherapeutic agents. Altogether, these studies support the role of miRNAs in the metabolic reprogramming hallmark of cancer cells and highlight their potential as predictive molecular markers of treatment response and/or targets that can be used for therapeutic intervention.


Introduction
MicroRNAs (miRNAs) are a class of small, highly conserved endogenous non-coding RNAs (19)(20)(21)(22)(23)(24)(25) nucleotides) that regulate gene expression post-transcriptionally. This regulation occurs mainly through binding to mRNA targets' complementary sequences in the 3 UTR region, blocking translation, and/or leading to mRNA degradation or destabilization. Although less common, miRNAs can activate gene expression by interacting with the 5 UTR and gene promoter regions [1][2][3]. MiRNAs are involved in several mechanisms of tumor development and progression, acting as both oncogenes and tumor suppressors, depending on the type of cell and tissue [4,5]. In addition, a given miRNA can regulate multiple targets, and a single target can be regulated by multiple miRNAs, showing the intricate and complex interaction between miRNA and mRNA pairings [6].
MiRNAs regulate mitochondrial functions and homeostasis, such as metabolic reprogramming, redox signaling, mitochondrial membrane potential, calcium transport, mitochondrial fusion, fission dynamics, mitophagy, and apoptosis [7][8][9]. Metabolic reprogramming, also known as "deregulating cellular metabolism," is one of the emerging hallmarks of cancer that occurs as a result of the metabolic plasticity of cancer cells [10,11]. metabolic hallmark of cancer. The main goal was to focus on searching for articles that described the role, association, and/or involvement of miRNAs (and mitomiRs) and their corresponding mRNA targets in regulating mitochondrial function, homeostasis, and metabolic reprogramming. The selected thirty-six articles reported on miRNAs regulating target genes involved in the aforementioned metabolic processes. In addition, some of these miRNAs have been reported to affect tumor resistance by mediating metabolic reprogramming and mitochondria-associated functions, which can point to a new perspective on cancer treatment based on cell metabolism.

Method
This review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [41,42]. The review protocol was registered at the International Prospective Register of Systematic Reviews (PROSPERO) database under the identifier CRD42022319233.

Data Sources and Search Strategy
The databases Pubmed, Scielo, Lilacs, EMBASE, and Scopus, were searched using the terms "mitochondrial microRNA" and its synonyms (mitochondrial microRNA, mitochondrial miRNA, mito microRNA, or mitomiR), "reprogramming metabolism," and "cancer" in the title or abstract. The articles searched were published from January 2010 to January 2022. A new search was conducted from January to September 2022 and included one article. Two reviewers independently performed the searches similarly for all of the databases. The online tool Rayyan (https://www.rayyan.ai/ (last accessed on 3 October 2022)) was used to analyze the selected studies. Duplicate articles were excluded, and two reviewers screened the remaining articles based on their title and abstract. Conflicting articles were evaluated by a third reviewer, followed by assessing the full text for relevance and eligibility.

Study Selection and Eligibility Criteria
Two reviewers independently evaluated and selected the studies according to the following inclusion and exclusion criteria. Inclusion criteria: (1) articles reporting the action of miRNAs in mitochondrial activities and the potential role of miRNAs on cellular mechanisms and pathways associated with mitochondrial functioning in cancer, including chemoresistance; (2) peer-reviewed articles written in English. Exclusion criteria: (1) articles on miRNA that did not report on the role, function, association, and/or involvement of mitochondria and cancer; (2) articles on miRNA analysis performed in animals; (3) nonoriginal articles (reviews), editorials, letters from editors, book chapters, unpublished or non-peer-reviewed studies; (4) articles for which the full text was not available.

Data Extraction
After the selection and eligibility assessment of the studies, two reviewers extracted the following information independently: year of publication, name of first author, country, title, study aim, methodology type (in miRNA and/or mitochondria-related activity or function), sample source (experimental data on patients' samples and/or cell lines, and/or data procured from online databases) used, miRNAs description, type of cancer analyzed, description of the main results, and conclusions.

Quality and Bias Evaluation
The Quality in Prognosis Studies (QUIPS) tool was used to assess the quality of the studies and the risk of bias, evaluating the studies in the following six categories: study participation, study attrition, prognostic factor measurement, outcome measurement, study confounding, and statistical analysis and reporting [39]. The articles were evaluated for quality according to the following classification-high quality (+): with little or no risk of bias; acceptable (+/−): moderate risk of bias; low quality (−): with a high risk of bias; and

Most of the Articles Selected Presented a Low Risk of Bias
The risk of bias was determined for all of the studies using the six categories of the QUIPS tool [43]. The overall assessment of the six categories resulted in 27 studies with a low risk of bias and nine with a moderate risk of bias. Following this evaluation, all 36 studies were retained for further analysis.

Most of the Articles Selected Presented a Low Risk of Bias
The risk of bias was determined for all of the studies using the six categories of the QUIPS tool [43]. The overall assessment of the six categories resulted in 27 studies with a low risk of bias and nine with a moderate risk of bias. Following this evaluation, all 36 studies were retained for further analysis.
The miRNAs that acted as oncogenes were associated with different functions and regulated specific target genes involved in several cancer phenotypes (except the studies [25,54,69]). The most frequent alterations observed in the gene expression deregulation of these miRNAs were drug resistance, cell metabolism (lactic acid secretion and OXPHOS), apoptosis, colony formation, cell growth and cell cycle, and development of metastasis [51].
Breast cancer was the most frequently studied cancer, with reported alterations in the expressions of miR-27a [51], miR-137 [61], miR-155 [66], and miR-210-3p [67]. In a study by Zhou et al. (2015), overexpression of miR-27-a was associated with downregulated expression of the BCL2 antagonist/killer 1 (BAK) gene and the second mitochondria-derived activator of caspase/DIABLO-IAP binding mitochondria protein/X-linked inhibitor of apoptosis (SMAC/DIABLO/XIAP) axis, resulting in a reduction in the apoptosis and chemosensitivity of cancer cells. It also increased tumorigenicity, as observed by the increase in colony formation and metastasis development [51]. Hu et al. (2020) also demonstrated a reduction in apoptosis with miR-137 overexpression by downregulating the FUN14 domain containing one (FUNDC1) gene. In this study, miR-137 overexpression led to a decrease in ROS levels [61]. Overexpression of miR-155 was shown to distinctly affect the FOXO3a/c-MYC axis and promote tumor growth by increasing glucose uptake and glycolysis [66]. Finally, in a breast cancer study by Du et al. (2020), overexpression of miR-210-5p led to the downregulation of glycerol-3-phosphate dehydrogenase 1-like (GPD1L) and cytoglobin (CYGB) target genes, causing metabolic alterations in the cells, with an increase in glucose and lactate uptake and a reduction in apoptosis [67]. Alterations in the expression of miR-210 were reported in colon [52] and colorectal cancer [54] studies. In colon cancer, its overexpression led to the downregulation of iron-sulfur cluster assembly enzyme (ISCU) and cytochrome C oxidase assembly factor heme A (COX10) genes, and increased cell survival in hypoxic microenvironment [52].
In other tumor types, the most cited miRNAs in the selected studies were those from the miR-181 (a, b, and c) family, reported in cervical [69], colon [62], colorectal [54], and liver [68] cancers. In cervical cancer, miR-181b was observed with high expression in the mitochondria of the HeLa cells [69]. In the nuclear factor (erythroid-derived 2)-like 2 (NFE2L2/NRF2) knockdown colon cancer study by Jung et al. (2017), significant mitochondrial dysfunction was reported with miR-181c overexpression. This altered expression led to the downregulation of mitochondria-encoded cytochrome c oxidase subunit-1 (mt-CO1), and these changes induced adenosine monophosphate (AMP)-activated protein kinase-a (AMPKa) activation and its subsequent metabolic adaptation signaling, including a reduction in OXPHOS. In a colorectal cancer study, miR-181 overexpression led to the transformation of precancerous cells in adenocarcinomas [54]. Finally, in a liver cancer study [68], overexpression of miR-181a-5p was shown to cause electron transport chain (ETC) remodeling, which reduced OXPHOS and increased cell survival in a hypoxic microenvironment, as well as glucose consumption and lactic secretion.
Altogether, these studies demonstrate the prominent role of miRNAs in the cell metabolism and reprogramming of cancer cells by regulating critical mRNA targets of both glycolysis-and mitochondrial-mediated pathways (Figure 2). They also demonstrate that miRNAs with the same mode of action can affect these pathways by regulating distinct targets, which highlights their versatile regulation of gene expression. Overview of the involvement of the identified miRNAs in the apoptosis and metabolic processes of the cancer cells. Fifteen miRNAs were identified which suppress or induce apoptosis by regulating the expression of pro-and anti-apoptotic gene-targets. Eleven miRNAs were observed affecting specific steps of glycolysis, from the glucose uptake to pyruvate synthesis, by regulating several glycolytic enzymes. Finally, eight miRNAs directly affected OXPHOS by regulating genetargets involved in mitochondria function and homeostasis.  Overview of the involvement of the identified miRNAs in the apoptosis and metabolic processes of the cancer cells. Fifteen miRNAs were identified which suppress or induce apoptosis by regulating the expression of pro-and anti-apoptotic gene-targets. Eleven miRNAs were observed affecting specific steps of glycolysis, from the glucose uptake to pyruvate synthesis, by regulating several glycolytic enzymes. Finally, eight miRNAs directly affected OXPHOS by regulating genetargets involved in mitochondria function and homeostasis.
A summary of the studies above, distributed per tumor type with the identified miRNAs, their corresponding target gene mechanism(s) of action, and the impact on cancer cell phenotypes, is presented in Table 1.
The expression of the altered miRNAs varied based on the main mRNA targets and tumor type. The most reported impact of tumor suppressive action on cancer phenotypes was on cell proliferation, apoptosis, and cytotoxicity to chemotherapeutic agents. Others affected the cell metabolism processes, such as glycolysis and mitochondrial organization, structure, and function [56,72,77].
MiR-125b interacted with two targets, HCLS-1-associated protein X-1 (HAX-1) and myeloid-cell leukemia 1 (MCL-1). In a study by Hu et al. (2018), overexpression of miR-125b was associated with reduced HAX-1 expression in breast cancer cells exposed to doxorubicin. This expression increased caspase 1 and ROS activity, resulting in increased cell death (apoptosis), chemosensitivity, and mitochondrial damage [46]. Other breast cancer studies have demonstrated a similar impact via the MCL-1 gene. The overexpression of miR-125b reduced the expression of MCL-1, increasing caspase-3 and apoptosis and reducing doxorubicin resistance [47].
In a study on breast cancer and melanoma by Zhang et al. (2019), miR-1 was upregulated, affecting the expression of ATP synthase membrane subunit 6 (ATP6), cytochrome C oxidase subunit 1 (COX1), glycerol-3-phosphate dehydrogenase 2 (GPD2), mitochondrial inner membrane organizing system 1 (MINOS1), NADH dehydrogenase subunit 1 (ND1), and ND4. These alterations decreased tumorigenicity and caused disorganization of the mitochondrial crest [56]. Yi et al. (2022) showed that the overexpression of miR-34a-5p led to downregulation of the mitochondrial inner membrane protein MPV17-like 2 (MPV17L2) in lung cancer and osteosarcoma cell lines. The miR-34a-5p suppressed the expression of MPV17L2, resulting in lower levels of respiratory chain complex I activities and intracellular ATP, a significant decrease in mitochondrial NADH dehydrogenase 1 (MT-ND1) protein levels, and an increase in oxidative stress, resulting in elevated apoptotic cell death. [63].
Two prostate cancer studies [73,74] reported alterations in the expression of miR-17* and miR-17-3p. In a study by Xu et al. (2010), the high expression of miR-17* increased mitochondrial ROS, which resulted in increased cytotoxicity to disulfiran in the cells and, consequently, cell death [73]. Xu et al. (2018) reported that the overexpression of miR-17-3p was positively associated with ionizing radiation, increasing the radiosensitivity and cell death of prostate tumor cells. In both studies, changes in miR-17 expression occurred via glutathione-dependent peroxidase (GPX2), manganese superoxide dismutase (MnSOD), and thioredoxin reductase 2 (TRXR2) targets expression [74].
Finally, a study by Chen et al. (2020) in pancreatic and breast cancer showed the overexpression of miR-1291, which acts in the estrogen-related receptor alpha (ERRα) and carnitine palmitoyl transferase 1C-CPT1C (ERRα-CPT1C axis). This miRNA alteration led to mitochondrial dysfunction and decreased cell metabolism, proliferation, invasion, and tumorigenesis [72].
These studies indicated the putative tumor suppressive action of the described miR-NAs on metabolic reprogramming and mitochondria-related functions, highlighting the need for further evidence for their potential application in cancer pharmacological therapy. Interestingly, the mitochondrial action of the same described miRNAs can also occur in other human diseases, such as cardiac diseases, supporting the discovery of new treatments based on epigenetic targets [78]. Nonetheless, these results show the diverse and complex regulatory action of miRNAs in the metabolic processes, by regulating interactions among multiple enzymes and complex metabolic components, which are among the major challenges for their clinical application.
The summary of these studies per tumor type with the identified tumor suppressor miRNAs and their corresponding target genes, mechanisms of action, and impacts on cancer cell phenotypes are presented in Table 2. The involvement of these miRNAs in the distinct cell metabolic process is shown in Figure 2, and the specific metabolic pathways and corresponding enzyme precursors that are targeted by these miRNAs (identified using https://www.proteinatlas.org (last accessed on 25 January 2023) (metabolic search)) are presented in Table 3.

Nine miRNAs Were Identified Acting on Tumor Chemoresistance Mediating Metabolic Reprogramming and Mitochondria Related Functions
Chemoresistance is one of the main problems in cancer treatment and can cause a lack of treatment response, tumor recurrence, and high mortality rates [39]. MiRNAs play a key role in chemoresistance by regulating target genes involved in diverse cellular mechanisms, including metabolic reprogramming [43,61,79].
For breast cancer, two studies showed an association between miR-27a and cytotoxicity to cisplatin, doxorubicin, and/or paclitaxel [45,51]. In a study by Ueda et al. (2020), miR-27a was described as having a tumor suppressive action, considering that its overexpression increased mitochondrial ROS and rendered MCF-7 and MDA-MB-231 cells more sensitive to doxorubicin and paclitaxel by inhibiting CTH (Cystathionine gamma-lyase), xCT (Cystine/glutamate transporter), and NRF2 (Nuclear factor erythroid-derived 2-like) expression [45]. In a study conducted by Zhou et al. (2015), miR-27a presented an oncogenic function, with overexpression associated with the inhibition of the BAK (BCL2 family member) and SMAC/DIABLO/XIAP pathways, increasing resistance to cisplatin in T-47D breast cancer cells [51].
MiR-125b was described in two breast cancer articles to be associated with sensitivity to doxorubicin [46,47]. Hu et al. (2018) showed that the overexpression of miR-125b increased sensitivity to doxorubicin in MCF-7 cells resistant to doxorubicin (MCF-7/DOX R), which was mediated by the downregulation of the HAX-1 gene and increased Caspase 9 and ROS levels [46]. Using a different in vitro model, Xie et al. (2015) demonstrated that inhibition of miR-125b decreased the sensitivity of tumor cells to doxorubicin by increasing the expression of its target MCL-1 [47]. However, Yuan et al. (2015), using the same MCF-7/DOX R model above, reported the involvement of a different miRNA and target gene modulating the doxorubicin cytotoxicity; the decreased expression of miR-133a increased doxorubicin sensitivity by increasing the expression of the target Uncoupling Protein 2 (UCP-2) [48].
MiR-let-7a [44], miR-223 [75], and miR-519d [76] have also been reported to impact the resistance to chemotherapy of breast cancer cells. A study conducted by Serguienko et al. (2015) showed that the overexpression of miR-let7a inhibited the expression of BACH1, G6PD, IMPDH2, FASN, SCD, AASDHPPT, and ND4, and, consequently, increased the mitochondrial ROS and chemosensitivity of MDA-MB-231 triple negative breast cancer (TNBC) cells to doxorubicin. The same was observed in WM239 metastatic melanoma cells [44]. In a study by Sun et al. (2016), the induction of miR-223 expression in TNBC stem cells increased their sensitivity and cytotoxicity to doxorubicin or cisplatin, mediated by the decrease in HAX expression and increase in mitochondrial ROS [75]. Another study in breast cancer stem cells reported that overexpression of miR-519d in cells incubated with cisplatin decreased MCL-1 expression and increased cytochrome C, activating the SMAC/DIABLO pathway and leading to apoptosis [76].
In the tongue cancer studies of Fan et al. (2015;, miR-593-5p and miR-2392 were associated with cisplatin resistance [38,53]. In a study by Fan et al. (2019), conducted in CAL-27 and SCC-9 oral squamous carcinoma cells, overexpressed miR-2392 co-immunoprecipitated AGO2 which, in turn, decreased OXPHOS and increased glycolysis, making the cells more resistant to cisplatin [38]. In another study conducted by Fan et al. (2015) in the same cell models, the overexpression of miR-593-5p with the overexpression of breast cancer gene 1 (BRCA1) decreased MFF expression, conferring cisplatin resistance to the cells [53].
These studies (summarized in Table 4 and Figure 3) highlight the essential role of miR-NAs in conferring tumor resistance by modulating mitochondria-mediated cell processes. They also point to miRNAs as having potential use as predictive molecular markers of treatment response and/or as molecular targets for therapeutic intervention. Nonetheless, additional in vitro studies in well-established drug resistant cell models and/or tumor cells that are directly immortalized from patients' tumors are required.

Conclusions
In conclusion, based on the 36 studies identified, this systematic review compiles evidence of the involvement of miRNAs and their corresponding mechanisms of action and biological impact in the metabolic reprogramming of cancer cells. By regulating target genes of diverse cancer-associated signaling pathways, miRNAs have been reported to be involved in cell metabolic processes, mitochondrial dynamics, mitophagy, apoptosis, redox signaling, and resistance to chemotherapeutic agents. As increasing evidence has emerged regarding the role of miRNAs in metabolic reprogramming and other associated hallmarks of cancer, their potential as predictive molecular markers of treatment response and/or druggable targets can be determined.