The review, led by researchers at the Medical University of Warsaw together with collaborators at the University of Warsaw and Penn State College of Medicine, synthesizes findings from preclinical models, human cohorts and mechanistic studies across the full spectrum of hepatic pathology. Its central message is one of context: the direction in which miR-122 moves, whether up or down, depends critically on which biological compartment is measured and at what stage of disease. Hepatic expression of the microRNA tends to decline as disease progresses, while circulating levels rise with hepatocyte injury. Far from being contradictory, the authors argue, these opposing patterns reflect distinct biological processes and must be interpreted together for miR-122 to serve as a valid biomarker.

Nowhere is this more relevant than in metabolic dysfunction-associated fatty liver disease, or MAFLD, a condition now estimated to affect between twenty-five and thirty percent of the global population. The disease begins as simple steatosis and can progress to inflammatory steatohepatitis, fibrosis, cirrhosis and, ultimately, hepatocellular carcinoma or acute liver failure. In mouse models fed a high-fat diet and in hepatic cells exposed to free fatty acids, miR-122 expression rises, suppressing the metabolic sensor Sirt1 and thereby activating the lipogenic genes SREBP1, FASN, ACC1 and SCD1. Inhibiting miR-122 restores Sirt1, reactivates the downstream LKB1/AMPK energy-sensing pathway and reduces fat accumulation in liver cells, establishing the miR-122/Sirt1 axis as a key driver of hepatic lipogenesis.

Human studies reinforce this mechanistic picture. Early clinical work from 2011 showed that circulating miR-122, together with miR-34a, was substantially elevated in MAFLD patients compared with healthy controls, with levels increasing progressively as simple steatosis gave way to steatohepatitis and correlating with liver enzymes, fibrosis stage and inflammatory markers. In a Japanese cohort of sixty-seven patients in whom paired liver and serum samples were available, hepatic and circulating miR-122 levels tracked each other closely, yet the relationship with fibrosis proved stage-dependent: expression was higher in mild than in severe fibrosis. A separate longitudinal biopsy study found that changes in circulating miR-122 were linked to steatosis, ballooning degeneration and fibrosis over a median follow-up of 4.6 years. More recently, a 2025 systematic review of 1,149 studies identified miR-122 as the most frequently investigated circulating microRNA in MASLD, appearing in roughly thirty-six percent of studies and reaching diagnostic AUROCs of 0.81 to 1.0 for steatohepatitis.

The inflammatory dimension of the story connects the gut to the liver. In an Egyptian study of fifty MAFLD patients, fifty MASH patients and fifty healthy controls, blood-exosomal miR-122 and miR-128 were upregulated alongside the bacterial product LPS, its receptor TLR-4 and the transcription factor FoxO3, while adiponectin and several other microRNAs were downregulated. This LPS/TLR-4/FoxO3 signaling axis links gut microbial imbalance to hepatic inflammation and offers a mechanistic route by which miR-122 participates in the transition from fatty liver to inflammatory disease. Adipocyte-derived exosomes add another layer: these vesicles carry high levels of miR-122 into hepatocytes, where suppression of Sirt1 increases gluconeogenic and lipogenic enzymes, raises inflammatory cytokines such as TNF-alpha and IL-1beta, and reduces fat oxidation through PPARalpha.

In acute liver injury, miR-122 behaves as a sensitive gauge of hepatocyte damage. In acetaminophen-induced injury models, the drug increased liver necrosis, inflammation, serum transaminases and miR-122-5p expression, while a miR-122-5p suppressor partially reversed these effects through upregulation of the target gene NDRG3 and favorable shifts in apoptotic markers, increasing Bcl-w and Bcl-2 while decreasing Bax. In a clinical cohort of 223 critically ill patients, serum miR-122 was substantially elevated compared with healthy controls, correlated strongly with traditional injury markers including ALT, AST and GLDH, and persisted as an independent indicator of hepatocyte injury regardless of whether the underlying condition was septic or non-septic.

Viral hepatitis reveals the molecule’s most striking paradox. In hepatitis B virus infection, miR-122 acts as a natural antiviral factor: overexpression in hepatoma cells suppressed HBsAg, HBeAg and viral DNA by fifty to seventy percent, while inhibition doubled HBV mRNA levels. Clinically, serum miR-122 was lower in HBV patients than in controls and inversely correlated with viral load. Mechanistically, miR-122 downregulates cyclin G1, disrupting its interaction with p53 and thereby removing a pro-viral loop; it also induces the antiviral enzyme heme oxygenase-1. In hepatitis C, by contrast, miR-122 is essential for viral replication, binding the HCV genome and stabilizing it. Circulating miR-122 is nonetheless elevated in chronic HCV patients, correlates with disease severity independently of viral load, and its upregulation appears independent of viral genotype, holding across both genotype 1 and genotype 3 predominant populations.

In hepatocellular carcinoma, loss of miR-122 emerges as a key oncogenic event. Tumor tissue from 142 HBV-related liver cancers showed significantly reduced miR-122, and low expression correlated with larger tumors, venous invasion, poor differentiation and reduced overall survival. Multiple target axes explain the tumor-suppressive effect: derepression of NDRG3, PTTG1-binding factor, PEG10 and GALNT10 each contributes to tumor growth, invasion and progression, while the Hnf4alpha/miR-122/GALNT10 pathway links transcriptional dysregulation to oncogenic protein activity. Mice engineered to lack miR-122 develop dysregulated lipid metabolism, steatosis, fibrosis and spontaneous liver cancer with epithelial-mesenchymal features; reintroducing the microRNA reverses these defects. Notably, the regulation of miR-122 differs by etiology: in HBV-associated HCC, epigenetic silencing through a PPARgamma/RXRalpha complex, histone methylation by SUV39H1 and viral protein effects drive suppression, whereas in chronic HCV the reduction appears linked to the host interferon response.

The therapeutic history of anti-miR-122 drugs is a cautionary tale. Miravirsen, a locked nucleic acid-modified antisense oligonucleotide and the first anti-microRNA drug to enter human trials, produced dose-dependent and sustained reductions in HCV RNA in phase 2a studies. Yet development was discontinued after viral resistance emerged through mutations in the viral 5-prime untranslated region that permitted miR-122-independent replication, and after direct-acting antivirals achieved cure rates above ninety-five percent, rendering a host-targeted strategy largely redundant. The GalNAc-conjugated antagomir RG-101 achieved marked viral-load reductions, including sustained virological responses after a single dose, but was placed on FDA clinical hold in 2016 following cases of severe hyperbilirubinemia attributed to inhibition of the MRP2 bilirubin transporter combined with preferential uptake into hepatocytes. Beyond these specific liabilities lies a more fundamental concern: because miR-122 is a tumor suppressor whose loss drives steatosis, fibrosis and cancer, sustained systemic suppression carries a theoretical long-term oncogenic risk.

The review’s authors conclude that the future of miR-122 lies in restoration rather than inhibition, particularly for metabolic and neoplastic liver disease, and in its use as a minimally invasive biomarker within multi-marker panels rather than as a standalone diagnostic. Given its high sensitivity to hepatocellular injury but limited disease specificity, serial measurement of circulating miR-122, ideally alongside other microRNAs and conventional markers, is likely to outperform any single-timepoint reading. A longitudinal study of eighty-one Japanese MAFLD patients illustrates the prognostic potential: among patients with persistent severe fibrosis, those who later developed liver cancer had significantly lower miR-122 levels, and a miR-122 ratio below 0.5 predicted higher cancer incidence and poorer survival. Standardized assays, longitudinal multi-omics studies and well-designed trials of miR-122-restoring strategies, the authors argue, are the necessary next steps to translate two decades of mechanistic insight into clinical benefit for the hundreds of millions of people living with liver disease worldwide.

Subject of Research: The role of microRNA-122 as a regulator and biomarker of liver disease

Article Title: MicroRNA-122 as a regulator and biomarker of liver disease

Article References: Ahmadova, S., Wicik, Z., Mucha, J., Palatini, J., Ziętal, K., Mirowska-Guzel, D., Przybylkowski, A., & Eyileten, C. (2026). MicroRNA-122 as a regulator and biomarker of liver disease. Journal of Molecular Medicine, 104(1), Article 110. https://doi.org/10.1007/s00109-026-02718-1

Image Credits: AI Generated

DOI: 10.1007/s00109-026-02718-1

Keywords: microRNA-122, liver disease, MAFLD, MASH, hepatitis B, hepatitis C, hepatocellular carcinoma, biomarker, fibrosis, lipid metabolism, antiviral therapy, miravirsen

That suspicion has been fueled by a striking clinical observation. Reduced GLO1 expression has been reported both in experimental models of metabolic dysfunction-associated fatty liver disease, or MAFLD, and in liver biopsies from patients with the condition. MAFLD, driven by excessive fat storage in the liver, is estimated to affect roughly 24 percent of the United States population and is present in more than 70 percent of patients with type 2 diabetes, making it one of the most common and consequential comorbidities of the modern obesity pandemic. Sustained consumption of high-carbohydrate diets, particularly those rich in fructose, is a major driving factor in the pathogenesis of both obesity and MAFLD, and fructose-containing foods are capable of inducing insulin resistance in humans and metabolic syndrome in mice. Yet despite the clear association between diminished GLO1 and fatty liver disease, a fundamental question remained unanswered: is the loss of GLO1 a cause of the disease, or merely a compensatory response to it?

A new study published in Physiological Reports by a team at the University of Arizona set out to resolve this question directly. Rather than merely observing correlations, the researchers generated whole-body Glo1 knockout mice using CRISPR-SpCas9 genome editing in the C57Bl/6NN strain, targeting exon 3 of the Glo1 gene with guide RNAs to produce frameshift deletions through non-homologous end joining. They then challenged these mice, alongside wild-type controls, with sixteen weeks of a high-fat, high-sucrose diet containing 36 percent fat and 30 percent sucrose, a regimen well documented to induce fatty liver disease and impair glucose tolerance. The team’s initial hypothesis was straightforward: if reduced GLO1 expression contributes to MAFLD, then deleting the gene should exacerbate metabolic dysfunction by increasing MGO-mediated stress. What they found instead was a surprise that reshapes how the field should think about the glyoxalase system.

Contrary to expectations, the knockout mice were substantially protected from hepatic fat accumulation. When wild-type mice consumed the obesogenic diet, they developed substantial hepatic steatosis, with triglycerides building up in the liver as expected. In the Glo1-deficient mice fed the same diet, this triglyceride accumulation was significantly blunted. The protection was specific to the liver: adipose tissue biology was largely unaffected, with no significant differences in adipocyte size, epididymal white adipose tissue mass, or markers of fat tissue injury between genotypes under the high-fat, high-sucrose conditions. Serum triglycerides and beta-hydroxybutyrate, a readout of fatty acid oxidation, varied only with diet and not with genotype, suggesting that the hepatic phenotype was not secondary to altered fat export or whole-body fat burning. The findings point toward a previously unrecognized role for GLO1 in directly regulating hepatic lipid metabolism, positioning the enzyme as an unexpected participant in the biology of fatty liver rather than a passive bystander.

But the metabolic ledger did not balance cleanly. While the knockout mice enjoyed relative protection from fatty liver, their systemic glucose handling deteriorated in a diet-dependent manner. Fasting blood glucose was significantly elevated in high-fat, high-sucrose-fed knockout mice compared with their wild-type counterparts. Oral glucose tolerance testing revealed a significant reduction in the ability of knockout mice to clear systemic glucose when compared with chow-fed controls, and although the difference between the two genotypes on the obesogenic diet did not reach statistical significance in the raw tolerance curves, a deeper analysis told a more troubling story. The constant of glucose decay, calculated from insulin tolerance testing as the rate of glucose disappearance, revealed a significant reduction in insulin responsiveness specifically in the high-fat, high-sucrose-fed knockout mice. Notably, serum insulin levels and hepatic insulin signaling, assessed through phosphorylation of the insulin receptor and AKT, showed no significant differences, indicating that the glucose defect operates independently of measurable changes in insulin activity.

To understand the biochemistry underlying these divergent phenotypes, the researchers turned to sensitive mass spectrometry-based quantification of MGO and its downstream molecular footprints. Free hepatic MGO was not significantly elevated in any treatment group, and the product of GLO1 activity, lactoylglutathione, was significantly reduced in the knockout mice, consistent with the loss of enzyme function. A complicating factor emerged, however: the high-fat, high-sucrose diet produced a marked reduction in hepatic glutathione regardless of genotype, and because glutathione is required for GLO1 activity, the reduction in lactoylglutathione may partly reflect this glutathione depletion. When the team examined MGO-derived post-translational modifications on proteins, the results were equally nuanced. Levels of MGO-hydroimidazolone 1, a signature MGO-derived arginine modification, were not significantly affected by diet or genotype, while carboxyethylarginine was elevated in chow-fed knockout mice. These data contradict previous reports indicating dramatic elevations in MGO-derived modifications under diet-induced metabolic stress, and they suggest that steady-state MGO biology in vivo is more buffered than cell culture experiments would predict.

The study also delivered a decisive verdict on a long-standing controversy in diabetes research. Earlier work using short hairpin RNA to knock down Glo1 reported that reduced GLO1 activity could spontaneously generate pathologies resembling diabetic nephropathy in non-diabetic mice, fueling the idea that GLO1 loss is a primary driver of diabetic kidney disease. The Arizona team therefore reasoned that sixteen weeks of high-fat, high-sucrose feeding would exacerbate kidney injury in their knockout animals. Instead, they found no evidence of renal pathology attributable to GLO1 loss. Kidney glycogen accumulated with the obesogenic diet but was unaffected by genotype, serum urea and creatinine were unchanged across all cohorts, and renal levels of MGO-hydroimidazolone 1, carboxyethylarginine, and 3-nitrotyrosine, a modification associated with oxidative stress, showed no significant differences. These findings independently confirm earlier reports from a separate group that complete genetic deletion of Glo1 fails to reproduce the diabetic kidney phenotype seen with knockdown approaches, and they collectively indicate that loss of GLO1 alone is insufficient to drive diabetic nephropathy.

Why might deleting a detoxification enzyme protect the liver while harming glucose control? The authors offer several mechanistic possibilities grounded in their own prior work. In cultured fibroblasts, they previously showed that loss of GLO1 reduces glucose uptake and glycolytic flux, and that Glo1-deficient cells fail to differentiate into mature adipocytes. If a similar reduction in glycolytic flux occurs in the livers of knockout mice in vivo, it would limit the substrate available for MGO generation, potentially explaining why free MGO and MGO-derived modifications remain largely unchanged despite the absence of the primary detoxification enzyme. Reduced glycolytic flux could also directly limit de novo lipogenesis, the pathway by which the liver converts excess carbohydrate into fat, providing a plausible mechanism for the blunted hepatic triglyceride accumulation. Meanwhile, the concept that MGO is not simply a toxin but a concentration-dependent metabolic signal is gaining traction: modest elevations of MGO have been reported to be protective in cardiac ischemia-reperfusion injury and even stimulatory for tumor growth, while only cytotoxic concentrations far exceeding those measured in vivo cause cell death.

The authors are careful to note the limitations of their work. The dietary intervention was terminated at sixteen weeks, a timepoint at which significant hepatic steatosis is evident but more advanced features of MAFLD, such as frank inflammation and fibrosis, have not yet developed. Whether GLO1 influences disease progression at later stages, including the transition to metabolic dysfunction-associated steatohepatitis, remains unknown and is a focus of ongoing investigation. The study also focused on male mice, leaving potential sex differences unexplored, and it did not evaluate alternative MGO detoxification pathways, such as the aldehyde dehydrogenases and aldose reductase, which are thought to play secondary roles but could become important under chronic metabolic stress. Samples for insulin signaling analysis were collected from fed rather than fasted mice, so the effect of GLO1 on glucose-stimulated insulin secretion could not be assessed.

Even with these caveats, the study carries a clear and provocative message: hepatic lipid accumulation and systemic glycemic control are mechanistically distinct in the absence of GLO1, and the glyoxalase system sits at an unexpected crossroads between the two. For a field that has largely treated MGO as a toxic metabolic accident and GLO1 as a straightforward protective enzyme, the demonstration that complete GLO1 loss limits fatty liver while impairing glucose handling in obese male mice demands a more sophisticated view. Future investigations aimed at deciphering the tissue-specific roles of GLO1 in whole-body glucose tolerance and lipid metabolism may reveal whether the glyoxalase cycle, long relegated to the footnotes of biochemistry textbooks, holds therapeutic potential for one of the most common liver diseases of our time.

Subject of Research: The role of glyoxalase 1 in obesity-associated fatty liver disease, glucose homeostasis, and kidney health in mice.

Article Title: Glyoxalase 1 loss reduces fatty liver but impairs glucose handling in male mice

Article References: Hoffman, E. A., Phoebe, A. M., Trujillo, M. N., Zhang, W. C., Jennings, E. Q., Farrera, D. O., Orlicky, D. J., Rutt, L. N., McCullough, R. L., Huacachino, A. A., Marcinkiewicz, M. M., Snyder, N. W., Bruner, K. R., Payan, K. B., Martinez, D. J. F., Stern, J. H., & Galligan, J. J. (2026). Glyoxalase 1 loss reduces fatty liver but impairs glucose handling in male mice. Physiological Reports, 14(17), Article e71106. https://doi.org/10.14814/phy2.71106

Image Credits: AI Generated

DOI: 10.14814/phy2.71106

Keywords: glyoxalase 1, methylglyoxal, fatty liver disease, MAFLD, glucose tolerance, hepatic triglycerides, glyoxalase cycle, high-fat high-sucrose diet, diabetic nephropathy, glycolysis, post-translational modifications, knockout mice