Article Text

Download PDFPDF

NKT cells and the hedgehog pathway: an unhealthy marriage in non-alcoholic fatty liver disease?
  1. Mariana Verdelho Machado,
  2. Helena Cortez-Pinto
  1. Departamento de Gastrenterologia, Unidade de Nutrição e Metabolismo, Hospital Santa Maria, Faculdade de Medicina de Lisboa, IMM, Lisbon, Portugal
  1. Correspondence to Professor Helena Cortez-Pinto, University Hospital of Santa Maria, Lisbon, Av. Prof. Egas Moniz, Lisboa 1649-028, Portugal; hlcortezpinto{at}

Statistics from

In the progression of non-alcoholic fatty liver disease (NAFLD) to steatohepatitis (NASH), the hedgehog (Hh) pathway has recently been implicated, with a new target gene, osteopontin (OPN), being a candidate in the coordination of inflammation and fibrosis development.1 This pathway, which is critical in embryogenesis, can be activated in adult life in the context of tissue regeneration. It has been suggested that dying and injured hepatocytes, such as ballooned hepatocytes,2 against a background of lipotoxicity, can produce Hh that will then act on inflammatory cells such as natural killer T (NKT) cells3 and on progenitor cells in such a way as to promote growth and hepatocyte differentiation, maintaining the liver mass.1 Unfortunately, it also acts on stellate cells inducing a profibrogenic phenotype.1

Syn et al4 further showed that NKT cells are responsive to Hh, and also are themselves a major source of Hh, inducing OPN production in an autocrine and paracrine fashion, thus being crucial in promoting fibrogenesis in NASH. The authors evaluated wild-type and knockout mice with NKT cell depletion, either knockout for the specific junction J α segment of the invariant T cell receptor (Jα18−/−mice), or for the major histocompatibility complex-like molecule CD1d to which NKT cells are responsive (CD1d−/− mice). The mice were submitted to a methionine and choline deficient diet versus a chow diet. NKT cell-deficient mice developed significantly less fibrosis, which was paralleled by less Hh and OPN expression. Also, in vitro data showed that activated NKT cells can induce stellate cell activation of a profibrogenic phenotype through Hh and subsequent OPN production. Lastly, in human liver samples from patients with NASH, the severity of fibrosis was associated with increased OPN expression and an increase in NKT cells expressing OPN.

NKT cells are a unique subgroup of lymphocytes that express features of both T cells and natural killer cells, representing a bridge between innate and adaptive immunity. They co-express T cell receptors and markers associated with natural killer cells, such as CD56 and/or CD161 (NK1.1).5 NKT cells depend on the presentation of glycolipid antigens through CD1d, a non-polymorphic class I antigen-presenting molecule, expressed in many immune cells, as well as in hepatocytes.6 The most widely studied NKT cells, the invariant NKT or class I, express a highly restricted T cell receptor with an invariant Vα24-Jα18, preferentially paired with Vβ11.5 NKT cells rapidly release cytokines, T-helper-1, -2 or -17 subsets, and display antigen-specific and cytolytic activities, being able to both promote and suppress immune responses.5 The role of these immune cells in hepatic steatosis and NASH is not completely understood.

Animal models of hepatic steatosis, with genetic or diet-induced obesity, showed a reduction in liver NKT cells, which was associated with endoplasmic reticulum stress with deregulation of CD1d trafficking to the plasma membrane of hepatocytes.7 Also, adoptive transfer of NKT cells improved steatosis and insulin sensitivity in those mice.8 The results of NKT cell reduction in human simple steatosis differed between studies.9 ,10 This apparent paradox may result from NKT cells being beneficial in the early stages of NAFLD, by preventing liver fat accumulation and improving glucose metabolism, while later on becoming highly deleterious in steatohepatitis and fibrosis progression. Previous studies had already shown that CD1d−/− mice, with no NKT cells, on a methionine and choline deficient diet were protected from fibrosis, and that mice partially deficient in the Hh receptor patched (Ptc+/−) mice, which have an overactive Hh signalling pathway, have NKT cell enrichment in the liver and more severe fibrosis.11 Hh induces proliferation and activation of NKT cells, as well as expression of interleukin 13, a profibrogenic cytokine.3 The article in Gut adds a new layer of complexity.4 NKT cells activated by Hh start an amplification loop, by further producing Hh, that can act on several other cell types, concurring in the fibrogenesis process.

OPN was recently identified as a final target in the Hh signalling pathway.12 In animal models of NASH, OPN is produced in hepatocytes, progenitor and stellate cells.1 ,13 Syn et al, have now shown us that it is also expressed in NKT cells in human NASH, correlating with hepatic fibrosis. In a small set of patients, blood levels mirrored hepatic levels. Indeed, patients with advanced fibrosis presented higher plasma levels of OPN than patients with early fibrosis (6.5 ng/ml vs 5 ng/ml).4 One issue that must be dealt with in a larger population is whether OPN plasma levels could be used as a biomarker for advanced fibrosis. The value of OPN as a surrogate marker has been suggested in other liver diseases such as acute liver dysfunction, viral hepatitis and hepatocellular carcinoma.14–16 It is worth evaluating the range of levels in those diseases, although a head-to-head comparison has not been done. It seems that OPN plasma levels in benign conditions, such as NASH, viral hepatitis or liver cirrhosis are much lower (in the range of 4–6 ng/ml),4 ,14 ,15 than those found in malignant disease, such as hepatocellular carcinoma.16 In a recent paper by Shang et al OPN performed better than α-fetoprotein, in which a cut-off point of 91 ng/ml presented a 74% sensitivity and 66% specificity for early hepatocellular carcinoma.16

In conclusion, these new data on the role of NKT cells and OPN in the fibrosis progression in NASH could be a hint to prompt further investigation of potential new diagnostic tools and therapeutic targets.


View Abstract


  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles