T cell fitness in the liver: How can T cells keep it up? EditorialJnl of Hepatology June 2016
Emerging Infectious Diseases (EID) Program, Duke-NUS Medical School, Singapore; Viral Hepatitis Laboratory, Singapore Institute for
Clinical Sciences, Agency of Science Technology and Research (A⁄ STAR), Singapore
See Article, pages 1327–1338
T cells play an important role in the immunological control of pathogens and tumors. CD8+ T cells scan cell surfaces with their T cell receptors for the presence of non-self antigens (originated from viruses, intracellular bacterial or from modified self antigens present in malignancies) bound to MHC-class I molecules. Specific recognition triggers T cell activation and results in the lysis of infected or tumor cells. The efficiency at which this occurs is extremely high such that a functional effector CD8+ T cell can be activated upon recognition of very few MHC-class I/non-self peptides present on the surface of the cells and can kill multiple targets [[url=]1[/url]]. The ability of T cells to access tissues typically requires multiple steps of T cell extravasation/migration that are regulated by the expression of specific adhesion and homing molecules [[url=]2[/url]]. However, such a process is not strictly necessary in the liver where the lack of a continuum basal membrane separating the blood from the hepatic parenchyma allows direct T cell scanning of target cells through the fenestrated liver sinusoidal endothelial cells [[url=]2[/url]]. Since the liver is an organ indispensable for life, the easy accessibility of CD8+ T cells to hepatocytes has been balanced by additional levels of control of intrahepatic CD8+ T cell function to avoid that immunological control of pathogens result in the self-destruction of the organ.
The elegant paper from the group of Bowen and Bertolino published in this issue of the Journal of Hepatology [[url=]3[/url]] directly investigates this problem and analyzes the mechanisms underlying the control of CD8+ T cell mediated liver damage. The authors used the Met-Kb transgenic mouse model in which liver damage is induced by the transfer of transgenic T cells that recognize a self antigen presented by MHC-class I H-2 Kb, expressed only in the liver and in lymph-nodes and they ask whether the most important parameter controlling liver damage is the survival or the cytokine-producing and lytic activity of T cells. To address this question they adoptively transfer in mic,e two types of alloreactive T cells that lack either a gene that regulates T cell survival (BIM-deficient- thus resistant to apoptosis) or one that regulates T cell function (SOCS-1, suppressor of cytokine signaling 1 deficient). As expected, BIM-deficient T cells display increased survival capacity and thus are present in increased numbers in the intrahepatic environment in comparison to normal T cells. However, the increased numbers of alloreactive T cells in the liver have a negligible effect on acute liver damage and do not lead to chronic liver inflammation. In contrast, the transfer of SOCS-1-deficient T cells causes a more severe and slightly prolonged hepatitis but again does not result in the development of chronic liver inflammation. Indeed, wild-type, BIM−/− or SOCS-1−/− T cells became functionally exhausted upon interaction with high levels of their cognate antigen in the liver within 5–7 days after adoptive transfer.
Thus, this work provides a demonstration that an important parameter that determines the extent of liver cell damage during acute hepatitis is the initial functional property of the CD8+ T cells. This finding is not unexpected: it seems logical that CD8+ T cells (SOCS-1 deficient) that possess a higher intrinsic ability to directly kill hepatocytes and produce inflammatory cytokines (such as TNF-α) will cause higher levels of liver damage as compared to equal number of less fit T cells. These findings also confirm the observation that liver damage in acute or chronically virus-infected livers is not directly proportional to the frequency of virus-specific T cells but depends on the functional capacity of these cells [[url=]4[/url]]. However, what is striking and somehow surprising is the similar and rapid kinetics at which intrahepatic T cells with different effector capacities are functionally exhausted and the extent of their functional silencing in a liver that constitutively presents high dose of the T cell cognate antigen. T cells that enter the liver appear to lose their functionality within a few days upon adoptive transfer (5–7 days). The functional exhaustion of the transgenic T cells is observed preferentially in the intrahepatic environment as T cells present in the lymphonodes maintain a degree of function ability.
These findings nicely complement the recent demonstration of the same group that the quantity of antigen expressed by the hepatocytes in the liver is the major determinant of the functional fate of CD8+ T cells [[url=]5[/url]] even though what we are still missing is the exact mechanisms of the intrahepatic T cell functional silencing.
The authors argued that T cell exhaustion is likely mediated by the upregulation of different co-inhibitory molecules like PD-1 and TIM-3. Functionally exhausted T cells present in the liver of these mice overexpressed these two co-inhibithory molecules. However, T cell functional exhaustion was not only mediated by PD-1 engagement since treatment with anti-PD-L1 antibody was not sufficient to restore a hepatic immune response. It is possible that the rapid silencing of T cell functionality in the liver observed few days after adoptive transfer could have be at least partially caused by metabolic changes in the inflamed intrahepatic environment. Arginase which is known to be released by dying hepatocytes [[url=]6[/url]], has been shown to dampen T cell function during acute [[url=]7[/url]] and chronic [[url=]8[/url]] hepatitis through the removal of the amino acid (arginine) essential for T cell function. However, such mechanism which is associated with the presence of liver inflammation, does not explain the prolonged suppression of T cell function observed in these animals after the resolution of liver damage. Despite the absence of any signs of chronic liver inflammation, the T cells that persist in the liver of the treated mice (present in very large numbers when the authors adoptively transfer BIM-deficient T cells) maintain a profound functionally exhausted phenotype as they lack any lytic or cytokine-producing abilities and they are refractory to treatment with anti-PD-L1. It will be interesting to test whether the observed intrahepatic T cell exhaustion is caused by other T cell-intrinsic functional parameters, such as expression of other inhibitory molecules (CTLA4, TGIT) or T cell receptor downregulation, mechanisms which are likely to occur in the presence of the high doses of antigen observed in the liver of these mice. Exhausted T cells in chronically infected livers express more than a single co-inhibitory molecule [[url=][9][/url], [url=][10][/url], [url=][11][/url]]. Nevertheless, a different, although not mutually exclusive possibility is that the maintenance of T cell silencing is mediated by other immune regulatory networks, like regulatory T cells [[url=]12[/url]] or granulocytic myeloid suppressor cells that were recently shown to be present in high numbers in the intrahepatic environment of patients with chronic hepatitis B and low/absence level of liver pathology [[url=]13[/url]]. Evaluation of the contribution of these different factors will be important to understand whether a recovery of intrahepatic T cell function in liver pathologies characterized by persistent and high expression of non-self antigens, (chronic viral hepatitis, hepatocellular carcinoma) would always necessitate a reduction in the number of antigen expressing hepatocytes or whether strategies designed only to alter the liver microenvironment (for example with Toll-like receptors mediated stimulation) might have an effect.
Certainly as the authors pointed out in their discussion, the data presented in their study is relevant for immunological therapies that attempt to restore T cell effector function both in patients chronically infected with hepatotropic viruses or in liver cancer. Despite the success of anti-PD-1 treatment for the restoration of intra-heaptic T cell function in vitro [[url=]14[/url]] and in chimps [[url=]15[/url]], the ineffectiveness of such treatment in this mouse model questions its therapeutic potential in patients with HCC or chronic viral hepatitis, a concept further supported by the modest results of anti-PD-1 clinical trial in HCV-infected subjects [[url=]16[/url]].
The prospect of immunotherapy on these liver pathologies appears even more daunting if we consider that T cell exhaustion was observed here in mice with relatively normal livers, where T cell access and function is not constrained by anatomical barriers [[url=]17[/url]] and by the suppressive immune environment present in livers with chronic inflammation (reviewed in [[url=]18[/url]]). On the other hand, the demonstration that hepatocytes are so well protected against T cell mediated immunological damage could be seen as an incentive to test new approaches of immunotherapy like transfer of gene-modified T cells or that combine check point inhibitors with agents that modify antigen expression in liver cells and T cell induction [[url=]19[/url]]. An approach combining therapeutic vaccination with anti-PD-L1 and antiviral therapies was shown to be effective in woodchucks chronically infected with woodchucks hepatitis virus [[url=]20[/url]]. This study shows that T cells struggle to maintain their fitness in the liver environment and help on multiple fronts may be needed for them to do so.
------------------------