In this issue of Clinical Science, Schueller et al. [Clin. Sci. (2017) 131, 1971-1987] evaluated the role of miR-223 across multiple etiologies of acute and chronic liver insults in murine models and clinical samples. The authors find that while miR-223 is not mechanistically involved in liver injury, its intracellular levels in hepatocytes are increased upon hepatic damage in a broad panel of mechanistically distinct injury models. Furthermore, the authors provide evidence that circulating miR-223 levels provide a promising minimally invasive biomarker for acute liver failure (ALF) that defines a distinct subset of ALF cases and correlates with clinical outcomes. Combined, the highlighted study suggests that miR-223 constitutes a promising biomarker whose clinical validity and utility warrant further investigations.
The liver fulfills crucial functions in energy homeostasis and is the principal organ for drug metabolism. As such, it is commonly exposed to diverse chemical, biological, and nutritional insults of various etiologies. Upon liver injury, hepatic cells rapidly regenerate the damaged tissue [1,2]. Multiple mechanistically diverse modes of liver regeneration have been postulated depending on the model of injury. In most injury models, regeneration occurs by self-duplication from preexisting hepatocytes [3,4]. However, under certain injury models, such as choline-deficient/ethionine-supplemented diet, treatment with D-galactosamine, or hepatectomy combined with N-2-fluorenyl-acetamide (2-AAF/PHx) exposure, the formation of atypical ductal cells (‘oval cells’) with mixed mesenchymal and epithelial expression signature is observed in the parenchyma [5,6]. Oval cells have long been assumed to be a population of facultative liver stem cells that derive from cholangiocytes and reside in the canal of Hering . Importantly though, under chronic liver damage hepatocytes can undergo reversible ductal metaplasia and dedifferentiate into bipotent progenitor cells, thus suggesting that atypical ductal cells are mistaken for cells of biliary origin, whereas they in fact arise via dedifferentiation from mature hepatocytes .
When the hepatic insult is too severe, the liver’s regenerative capacity is not sufficient to compensate for the injury, resulting in failure of the damaged organ. Acute insults, such as drug-induced liver injury (DILI) and ingestion of certain toxins, can result in acute liver failure (ALF), which is characterized by sudden onset and typically occurs within 48 h. In contrast, in chronic liver diseases (CLDs), such as alcoholic and viral hepatitis, non-alcoholic steatohepatitis (NASH), and autoimmune disorders, the injury persists over prolonged periods (months to years) and initiates a cascade of events causing hepatic inflammation, excessive deposition of extracellular matrix (hepatic scarring), and disrupted tissue architecture . Combined, CLDs affect 3.9–6.9% of the general population in the US  and account for 1.8% of all deaths in Europe . When diagnosed early, close monitoring, life style changes, and therapeutic interventions can delay disease progression and increase life expectancy . In end-stage disease, however, orthotopic liver transplantation (OLT) is commonly the only therapeutic option, irrespective of the etiology of liver failure. However, the shortage of donor organs results in 16,000 patients on the waiting list in the US alone and a mortality rate of 10% while waiting .
Due to the grave consequences of late diagnoses, it is of paramount clinical importance to detect and stratify liver diseases as early as possible in order to enable appropriate treatment. Liver biopsy is considered the gold-standard for diagnosis and staging of liver injury. However, assessing liver injury by biopsies is expensive, inconvenient, and entails risks of bleeding complications or perforation in approximately 1 in 1000 biopsies . Thus, much research is focused on the development of less invasive methods that in combination with objective clinical variables are benchmarked against data from biopsy-confirmed CLD cases. For this purpose, miRNAs have received growing attention as they are important regulators of hepatic functionality and plasticity , and deregulation of miRNAs has been implicated in onset and progression of liver diseases of various etiologies . Upon liver injury, miRNAs are released from hepatic cells either bound to protein or encapsulated in exosomes [17,18] and thus constitute promising candidates as minimally invasive biomarkers for routine clinical diagnostics that can be identified in patient serum [19,20]. miR-122 is the most abundant miRNA in liver and its serum levels correlate with the extent of hepatic injury in mouse models and patients [21–24]. Circulating levels of miR-122 are currently in clinical trials as a diagnostic biomarker for DILI due to chemotherapy (NCT03039062 and NCT03000621). In addition to miR-122, a plethora of other miRNAs, including miR-15b, miR-21, miR-155, and miR-221, have been linked to liver injury . Notably, however, there is little standardization in the field with regards to protocols for sample collection, data normalization and analysis and as a result, only few miRNA biomarkers are consistently identified across models of hepatic injury and even between studies that use the same model. The resulting discordances complicate result interpretation and impair the clinical implementation of potential biomarkers.
In this issue of Clinical Science, Schueller et al.  provide significant advances in this direction as they systematically analyze the regulation of one miRNA species, miR-223, across multiple etiologies of acute and chronic liver injury using a broad panel of mechanistically distinct injury models. The authors analyze mouse models of ischemia/reperfusion (I/R) and hypoxia/reoxygenation injury as well as models of acetaminophen- and concanavalin A-induced liver injury and detect significant up-regulation of miR-223 in liver in all these acute models. Interestingly in I/R, increases in serum miR-223 levels were already apparent 3 h after injury, whereas increased tissue expression of miR-223 was only detectable after 24 h. These findings are similar to changes observed in miR-122 and miR-29b profiles upon acetaminophen-induced liver injury and is suggestive of a controlled secretion process that precedes the passive release due to hepatocyte rupture upon apoptosis or necrosis ; however, the mechanisms that regulate whether miRNAs are retained within the cell or secreted remain poorly understood and pose an interesting target for further investigations.
Importantly, the authors could confirm their mouse data in clinical samples of ALF patients. They show that miR-223 expression was increased in liver biopsies from ALF patients and levels significantly correlated with clinical outcome. Interestingly, hepatic levels of miR-223 correlated with miR-21 levels but not with other miRNAs implicated in liver injury, such as miR-122, miR-192, or miR-29b, implying that miR-223 elevations might define a distinct subset of ALF cases. Consequently, miR-223 might be a promising addition to preexisting serum biomarker panels of acute liver injury that warrants further investigation in larger prospective trials.
Moreover, Schüller et al. carefully investigate changes of miR-223 levels in various murine models of CLD . miR-223 was up-regulated in hepatocytes of mice after repeated carbon tetrachloride injections or bile duct ligation, two murine fibrotic models. Strikingly, miR-223 expression levels drastically decreased once injections were stopped, correlating with a decrease in expression of other markers of hepatic fibrogenesis. These findings suggest that up-regulation of miR-223 might represent a direct readout of ongoing liver injury. In liver biopsies from fibrotic patients miR-223 expression was increased, consistent with murine data. In contrast, serum levels of miR-223 in a cohort of CLD patients showed discordant results: in fibrotic or cirrhotic patients circulating levels were significantly decreased whereas no change of miR-223 expression was detected in sera of hepatocellular carcinoma (HCC) patients. Interestingly, this discrepant data aligns with previous conflicting studies: while some studies showed increased miR-223 serum levels in patients with hepatocellular carcinoma as well as chronic type B hepatitis [28–30], others detected inverse correlations [31,32].
While miR-223 was increased in models of acute and chronic liver injury, no significant elevations were detected in the partial hepatectomy model of liver regeneration, implying that increased intracellular levels of miR-223 in hepatocytes are a direct indication of active hepatic stress in acute and chronic liver injury. In future studies, it will be interesting to see whether miR-223 levels remain unaffected also in models in which hepatic regeneration involved ductular reactions.
Importantly, upon deletion of miR-223 mice were protected from FasL-induced ALF in agreement with previous work . In contrast, the authors did not detect alterations in response to I/R injury or acute carbon tetrachloride injection nor was the magnitude or kinetics in the pathophysiology of hepatic fibrosis affected. Furthermore, regeneration kinetics upon partial hepatectomy is not detectably altered in miR-223−/− mice, indicating that miR-223 is not involved in the regulation of hepatocyte proliferation. Taken together, these findings imply that miR-223 sensitizes hepatocytes to extrinsic apoptosis whereas it plays only a minor role in other types of hepatic injury.
In summary, Schueller et al.  identify a promising minimally-invasive biomarker for ALF that significantly correlated with clinical outcomes. While it is too early to advocate its implementation into clinical diagnostics, this study reveals important avenues for future research. Specifically, this work incentivizes
monitoring of miR-223 expression in a prospective setting using larger cohorts with hepatic insults of diverse etiology alone or in combination with other established markers.
further research into the regulatory logic underlying miR-223 expression in acute and chronic liver injury.
the comprehensive identification of hepatic miR-223 target genes in health and disease.
analysis of miR-223 expression under injury conditions that involve the formation of atypical ductal cells.
As such, the presented study represents an important step forward toward the improvement of early diagnosis of ALF and paves the way for more mechanistic analyses of changes of non-coding RNAs in liver injury.
The research in the author’s laboratory is supported by the Swedish Research Council [grant numbers 2016-01153 and 2016-01154]; and the Malin och Lennart Philipson Stiftelse [Award 2017].
The authors declare that there are no competing interests associated with the manuscript.