The inhibition of IDO can aggravate or alleviate the severity of liver damage, which depends on the immunomodulatory effect of Trp metabolism and various experimental models. An experiment on the hepatitis mice model induced by α-galactosylceramide (α-GalCer) indicated that IDO could prevent excessive immune response to weaken liver injury in this model (51). This is the first research concentrated on the effects of IDO on acute liver injury. It is worth noting that α-GalCer-induced liver injury is thought to have the potential to partially mimic autoimmune hepatitis because it is caused by the activation and apoptosis of Vα14 NKT cells. These findings indicated that the expression of IDO can down-regulate the level of the proliferation of macrophages and natural killer (NK) cells and the TNF-αproducedby these immune cells inα-GalCer-induced liver injury. The authors inferred that IDO could prevent excessive immune response in α-GalCer-induced hepatitis model based on the above results.

The CCl₄-induced rodent models were intensively used to simulate liver fibrosis patients (52). The mechanism is primarily via trichloromethyl radicals metabolized, which destroys the biomembrane structure and triggers the inflammatory responses (53). It is reported that the liver injury in IDO-knockout (KO) mice treated with CCl₄ was exacerbated compared with WT mice (54). In the inflammatory condition caused by the injection of CCl₄, the authors observed IDO deficiency caused upregulation of pro-inflammatory cytokines and fibrogenic factors, contributing to the induction of activated hepatic stellate cells (HSCs) and the progression of liver fibrosis (54). Similarly, IDO inhibition with 1-D-MT after CCl₄ injection elevated the serum alanine aminotransferase (ALT) level and increased the severity of liver injury at 16 h after disease onset (55). 1-D-MT-treated rats produced higher TNF-α levels in the liver and IL-6 levels in the serum compared with those in mock and control groups, although no differences were measured in serum MIP-2 and keratinocyte chemoattractant (KC) levels (55). Both articles pointed to IDO activity upregulated after CCl₄ administration, while IDO inhibition exacerbated CCl₄-induced hepatitis with enhanced cytokine and chemokines. Interestingly, it is reported by Zhong et al. that IDO deficiency attenuated CCl4-induced cirrhosis (56). In addition, a more recent article (57) observed that IDO2^−/− mice and administration of 1-D-MT or Kyn also prevented severe liver cell damage and liver fibrosis, which suggested that the mechanism is not attributable to the contribution of IDO1. The authors believed that Kyn produced by IDO2 in the liver might play a crucial role in CCl₄-induced acute liver injury through a mechanism involving AHR signal transduction (57). We found that the Ogiso's group (54) and the Li's group (55) only have taken IDO into account, whereas the Hoshi's group (57) and Zhong's group (56) considered IDO2 or TDO besides IDO1. Indeed, the different endings are possibly due to a compensatory mechanism among IDO1, IDO 2 and TDO. Previous evidence suggested TDO inhibition not only increased plasma and brain Trp but also the major Trp metabolite Kyn, which suggested a compensatory mechanism by extrahepatic IDO in the absence of intrahepatic TDO (58, 59). Moreover, a study reported TDO activity was also inhibited strongly by CCl₄ treatment (60), thus the incorporation of IDO1, IDO2 and TDO into liver dysfunction will be crucial revelation enlightenment.

Non-alcoholic steatohepatitis (NASH) is an inflamed fatty liver model subsequent to liver inflammation, fibrosis or even HCC. The animal model usually uses a high-fat diet (HFD)method. Nagano et al. reported that IDO gene silencing in the HFD-induced model aggravated hepatic inflammation and the progression of liver fibrosis. After being given pelleted HFD for 26 weeks, the IDO deficiency mice detected mixed inflammatory cell infiltration, especially T lymphocytes and macrophages in the liver. The authors speculated that IDO deficiency increased the number of lymphocytes that migrated to the liver, thereby further exacerbating liver damage (61). However, the use of HFD has been demonstrated to regulate constant brain Trp by inhibiting TDO activity, as well as NEFA was increased caused by the HFD treatment (62). Therefore, only IDO-KO treated by HFD are not free from the interference of “external” modulating factors, such as the compensatory effect of the flux of free Trp and TDO activity caused by HFD. Moreover, this article calculated the L-Kyn/L-Trp ratio as the activity of IDO enzyme, which was controversial. What is more, liver TDO activity is another determinant of the ratio (63). We found the value of Kyn/Trp ratio in HFD-mice reported by Nagano et al. was ~0.009, which was lower than a control ratio of 0.025 was reported by Ogawa et al. with the same procedure (64). It is likely contributed to the inhibition of TDO activity rather than the sole activity of IDO. Furthermore, the very low Kyn/Trp ratio in HFD mice further argues against IDO involvement. It is a pity that the authors did not report data on IDO in control mice.

Indeed, IDO deficiency does have beneficial effects in several other liver injury animal models, even though IDO expression was still increased in the model group. Using the concanavalin A (ConA)-induced liver injury mice model, Ting et al. observed that the inhibition or deficiency of IDO could alleviate murine liver damage by the reduction of inducible nitric oxide synthase (NOS) and 3-nitrotyrosine (65). As the T-cell mitotic plant lectin, ConA is recognized as a classic inducer for animal models of acute hepatitis, which can simulate the pattern of fulminant immunological liver injury (66, 67). ConA-induced hepatitis was recognized to be mainly related to ferroptosis, which was IDO-dependent. Noteworthy, in vivo and in vitro studies, IDO deficiency promoted the ferroptosis resistance by activating the expression of solute carrier family 7 member 11 (SLC7A11, also known as xCT), while reducing murine liver lesions and reactive nitrogen species (RNS) (65). In addition, another common bile duct ligation mice model also indicated IDO overexpression accelerated liver fibrosis and IDO-deficient fibrotic mice exhibited milder liver fibrosis than WT fibrotic mice by altered hepatic inflammatory cells (68). Researchers proved that IDO1 overexpression inhibited the maturation of CD11c + DCs in the liver and spleen, inhibited T cell proliferation mediated by mature DCs and worsened liver fibrosis, whereas in IDO1^−/− mice the above pathological phenomena were reversed (69). Similarly, in diethylnitrosamine-induced HCC, IDO overexpression and higher Kyn levels were detected in IDO-wild-type mice compared to surrounding normal tissue. Furthermore, IDO-KO mice prevented the development of HCC, which was caused by the increasing the mRNA expression levels of CD8, perforin and granzyme B (70).

Based on these previous studies (Table 1), we found that a single discussion of the contribution of IDO on the liver is obviously over simplify the problem since Trp metabolism is a considerably complex pathway in liver injury. We speculated there might be existed a compensation mechanism besides IDO in KP, which brought to light a novel view in liver injuries (78).

Bacterial and viral infections are critical risk factors for liver injury. Epidemiological data demonstrated that chronic hepatitis C virus (HCV) and hepatitis B virus (HBV) are the most common forms of infectious liver disease. It is estimated that the total number of deaths due to liver disease caused by HBV is about 800,000, while about 500,000 deaths are caused by HCV (84). The HCV infection progression is possibly mediated by innate and adaptive immune responses in infected patients (85). Impairment of HCV-specific CD4⁺ and CD8⁺ T-cell responses (86–88) and abnormal DCs function has been observed in HCV infection (89) as well as the upregulation of IDO expression in HCV infection (25). Persistent IDO expression within the liver microenvironment may play a crucial role in reducing HCV-specific T-cell responses (6). Yang et al. (90) revealed that plasma level of IDO was associated with TGF-β expression and severity of chronic HCV infection. In the acute infection stage, the expression of IDO1 is upregulated to promote the process of anti-viral defense, as IFN-γ did so in HCV-infected liver cells to inhibit anti-HCV T cells in the chronic infection phase.

Similar to HCV, IDO overexpression was associated with HBV. The expression and activity of IDO in patients with chronic HBV infection were significantly higher than those in healthy controls (91). Further studies explicated that IDO expression was correlated with HBV viral load and responsible for immunotolerance against HBV (91). HBV infection facilitated IDO induction, mainly through response to the hepatocyte pro-inflammatory cytokine IFN-γ in mice model studies. The up-regulation of IDO was accountable for transduction of cytotoxic T lymphocytes and ultimately down-regulated responsiveness of T cells (15). Similarly, the increase of IDO expression in the livers of murine fulminant hepatitis model induced by HBV-specific cytotoxic T lymphocytes (CTLs) was also examined in another study. Moreover, IDO inhibition could reduce liver injury, with Kyn and IFN-γ cooperatively involved in the progression of liver injury (72). IDO was demonstrated as a potential and novel favorable therapeutic target for chronic HBV infection (37, 91).

IDO exhibited an essential role in HBV and HCV infections by constructing an immunotolerogenic microenvironment. Moreover, the prevailing theory is that the molecular mechanisms of the tolerogenic state are accompanied by chronic HBV and HCV infection, with only a weak response of CTLs against the HBV surface antigen (HBsAg). Nevertheless, another mechanism of IDO activity was discovered in a murine hepatitis model. Ito et al. immunized with a combination of α-GalCer, a specific agonist for NK inducing IDO and HBsAg on wild-type and IDO-KO mice. Only IDO-KO mice showed an increased expression of the cytokines IL-2 and IL-12b leading to the induction of HBsAg-specific CTLs (92). Besides, α-GalCer induced IDO production in CD11b⁺ cells, which inhibited the proliferation of HBsAg-specific CTLs (92). These data suggested that inhibition of IDO activity enhanced the induction of HBsAg-specific CTLs after immunization with HBsAg and α-GalCer. Thus, IDO played the role of the central mediator in the IFN-γ-induced anti-viral response since it mediated Trp depletion followed by suppressing HBV or HCV replication. Additionally, published literature has shown that the inhibitory effect of IDO1 can be used to treat chronic HCV patients, and its mechanism may be attributed to the reduced inducible nitric oxide synthase and 3-nitrotyrosine expression (65, 93).

As described above, IDO suppressed the degree of the immune response in viral hepatitis. In this case, IDO inhibition could be expected to enhance the immune response and shield mice from viral infections. Nevertheless, Duhalde Vega et al. got contrary conclusions from a hepatitis virus A59 (MHV-A59) infected mice model that 1-L-MT aggravated liver injury, and the survival rate of MHV-infected animals treated with 1-L-MT was severely decreased compared to the control group without treatment of 1-L-MT (94). Excitingly, they continued their follow-up exploration and demonstrated TDO inhibition by LM10 resulted in decrease levels of the Ab to MHV induced by the same virus infection (78, 95). The two contrary responses might be clarified by the compensation mechanism controlling Trp metabolism (96). It is confirmed by Too et al. that deletion of the IDO1 or IDO2 gene does not alter brain Trp, Kyn or the Kyn/Trp ratio, whereas TDO2 deletion increases brain Trp without altering brain Kyn (78). The above evidence indicated that there might exist compensatory effects of TDO when IDO is inhibited. Therefore, inhibition of IDO alone cannot alter the progression of liver injury.

It is known that pro-inflammatory cytokines elevated caused by viral infections can induce IDO instead of TDO. Further, another study about the contribution between IDO and TDO explained that the Trp metabolism that occurred during infection was related to several KP metabolites involved in immune response mediated by IDO rather than the amount of Trp depletion (45, 96). Thus the pro-inflammatory Kyn metabolites mediated by IDO may be the real pathogenic factor in viral infections.

Notably, it was reported in this article by Duhalde Vega et al. that liver (Trp) also increased in MHV-infected mice, instead of the expected decrease compared with that of control (95). Badawy et al. supported that the flux of plasma free (non-albumin-bound) Trp was another vital factor (63). In MHV-induced hepatitis, the plasma free Trp may be upregulated, which is supported by evidence of decreasing albumin and increasing non-esterified fatty acids (NEFA) in hepatitis patients (81, 95, 97, 98). And the increase of the plasma free Trp promotes the flux of Trp through TDO.

Based on these, once TDO activity is inhibited by LM10 treatment on MHV-infected mice, the degradation of intrahepatic Trp mediated by TDO failed and the flux of Trp through TDO is prevented, leading to the decrease of pro-inflammatory Kyn metabolites.Consequently, the flux of Trp through TDO together with IDO induction enhanced pro-inflammatory Kyn metabolites by two mechanisms in the MHV-infected mice.

Obviously, the immune response is very complicated, and the various effects on viral infection should not be simply interpreted as the sole IDO inhibition or Kyn decrease. Overall, the expression of IDO might have a significant effect on the prognosis of viral-induced patients. Trp metabolism may provide new strategies as an adjuvant therapy intervention for viral hepatitis.

Compared with immune responses after a chemical- or viral-induced liver injury, post-hepatectomy and partial liver transplantation are also closely related to IDO activity by suppressing liver regeneration and protecting against immune rejection. This vital role of IDO in transplantation originated from a landmark study published in science magazine in 1998. Mellor et al. found that IDO suppressed the maternal T-cell response against fetus and inhibition of IDO with 1-MT caused abortion of allogeneic concepti but not syngeneic concepti (33). This work showed the most inspiring results of late years, which has completely changed the statement of IDO, from what was initially thought to be an innate defense mechanism to a potent immunomodulatory enzyme of which can down-regulate immune activation and inhibit T cell response and promote tolerance induction eventually. Since then, several studies on the effect of IDO in transplantation have been published. To determine whether IDO is the cause of liver tolerogenicity, Madeleine et al. studied IDO expression in liver-induced acceptance of kidney grafts in sensitized patients undergoing combined auxiliary liver-kidney transplantation. Combined auxiliary graft transplantation showed an increase of tryptophan degradation in peripheral blood and expression of IDO mRNA compared to regular renal transplantations (99). These results were proved by Xing et al. in a comparable rat experimental model where they demonstrated that IDO could be significantly overexpressed in rat livers after syngeneic and allograft liver transplantation compared to the sham-operated group (100). In addition, the results of immunohistochemistry showed that the number of IDO-positive cells was positively correlated with the exacerbation of rejection, and the level of inflammatory cell infiltration in the portal area was different in vitro, the expression of IDO gene and enzyme activity was increased in the IFN-γ-treated DC group within 7 days after transplantation compared to the untreated DC group and their survival rates were also significantly improved (100). Therefore, these results suggested that IDO might be involved in the spontaneous tolerance of liver allograft. IDO-positive DC induced by IFN-γ might relieve acute rejection and accelerate local tryptophan metabolism through IDO enzyme expression, resulting in post-liver transplantation immune tolerance (101, 102).

Besides living donor's liver transplantation, liver regeneration is another highly orchestrated process influenced by various factors. Liver regeneration is initiated immediately after the loss of hepatocytes for partial hepatectomy or partial liver transplantation, while its failure leads to hepatic injury and death. Hepatocyte proliferation is regulated by several growth factors such as TGF-α. Then hepatocytes need priming by TNF-α and IL-6, which are secreted by Kupffer cells, leading to the transition of growth factors from resting phase (G0) to interstitial phase (G1) (103, 104). Hideyuki et al. demonstrated that the liver-body weight ratio and the number of Ki-67-positive cells per field were significantly higher in IDO1-KO mice than WT mice. Likewise, the hepatic mRNA expression of cell cycle genes (cyclin D1, cyclin E) and pro-inflammatory cytokines (IL-1β, TNF-α and IL-6) were significantly higher in IDO1-KO mice than in WT mice (105). Moreover, the administration of the 1-DL-MT at the time of transplantation resulted in promoting liver regeneration (105). Therefore, these results indicated that IDO1 suppressed the production of inflammatory cytokines and subsequently inhibited hepatocyte proliferation during liver regeneration.

As previous studies described above, tryptophan catabolism via the IDO pathway is required for immune tolerance in allogeneic concepti rather than syngeneic concepti (33). Increasing evidence suggested that IDO might be a potential target to prevent acute rejection in spontaneously tolerant mouse liver allografts (106) and delayed rejection (107). Jerome et al. applied recombinant adeno-associated virus 2/8 (rAAV2/8) to deliver the transgene to allograft prior to transplantation (108). However, the median survival of recipients of allografts pretreated with recombinant adeno-associated virus 2/8–liver-specific promoter 1–IDO (rAAV2/8-LSP1-rIDO) vectors was 11–15 days, which was not significantly greater than that of the control group pretreated with recombinant adeno-associated virus 2/8–liver-specific promoter 1–enhanced green fluorescent protein (rAAV2/8-LSP1-eGFP) or untreated Piebald Virol Glaxo-to-Lewis liver allografts (108). These results indicated that transfecting single adenovirus-mediated IDO in rat liver allografts could not prevent acute rejection. Interestingly, Yakun et al. challenged to explore the effects of adenovirus-mediated combined genes of CTLA4Ig and IDO in the immune tolerance after orthotopic liver transplantation in rat models. Combined transfection of CTLA4Ig-IDO exhibited milder acute rejection and a higher survival rate in inducing immune tolerance after liver transplantation than the groups that using single adenovirus-mediated genes (CTLA4Ig or IDO) alone. These results were in line with previous reports indicating the therapeutic value of using DCs with IFN-γ-induced IDO expression to treat acute rejection after rat liver transplantation (100). The data exhibited a positive correlation between increased IDO-positive cells in the portal area and the severity of rejection. Meanwhile, experimental results from other in vivo models also pointed out that although IDO expression is upregulated in rejected grafts, merely IDO does not play a crucial role in the process of acute rejection (108–112). This result may be caused by a strong rejection process that IDO cannot reverse (113). Therefore, the combined application of IDO and other targets may become a feasible and valid method for immunological tolerance in liver transplantation and may become a promising clinical method for treating immune liver disease.

HCC is the third most common cause of cancer-associated deaths worldwide and confers a poor prognosis (114). Tumor-immune escape mechanisms have been currently proposed as emerging topics that were involved in HCC progression. Therefore, it has begun to focus on tumor immunology to reveal the immunosuppressive effect of IDO recently.

The prevailing theory suggested IDO has a tumor-promoting outcome in the occurrence and development of some solid tumors through immunosuppressive effects (115–117). IDO expression in various histologic cancer types seemed to build an immune-suppressive microenvironment (118, 119) by regulating immune cells including T effector cells (120), regulatory T cells (121), and Myeloid-derived suppressors cells (MDSC) (122, 123). In HCC pathogenesis, IDO was expressed in HCC cells after IFN-γ stimulation, which was a prognostic factor for poor survival of HCC patients (115, 124). One new subset of human CD14⁺ CTLA-4⁺ regulatory dendritic cells (CD14⁺DCs) was identified in HCC patients, which significantly suppressed T cell response in vitro via IDO (125). This finding added new insights into HCC induced immunosuppression mechanism related to IDO. Another study indicated that expression of PGE2 and IDO by HCC-associated fibroblasts might represent a novel mechanism by inducing NK cell dysfunction and creating favorable conditions for tumor progression (126).

Interestingly, some contradictory data demonstrated that induction of the IDO enzyme by IFN-γ exhibited opposite anticancer immune reactions in tumor-infiltrating cells of HCC. The recurrence-free survival rate of IDO-positive HCC patients was significantly higher than that of IDO-negative HCC patients, which was the first report to suggest that IDO expression at the molecular level might be essential for TIL to suppress tumor proliferation in HCC (127). Notably, IFN-γ is known to antitumor, namely a low, rather than a high IFN-γ level predicts HCC recurrence after therapy (128, 129). Indeed, the elevation of IFN-γ in HCC may reflect liver dysfunction rather than inflammation (130). Thus, besides IDO, the role of IFN-γ is not exclusive to the present experiment by Ishio et al. In addition, IDO was strongly induced in HCC cells following IFN-γ does not simply mean that IDO is expressed (strongly or otherwise) in HCC. Likewise, mRNA expression of IDO in HCC patients was performed in this article, which is also not always synonymous with IDO enzyme functional activity (129, 130). The mechanisms of primary and secondary resistant tumors were complex, including factors related to both the tumor and the host itself. Among them, immune metabolism dysregulation played an eventful role in the disease development of HCC patients.