The production of KYNA is directly correlated to inflammation as KYNA acts as an important immune regulated during inflammation response. KYNA inhibits TNF-α at transcriptional level and suppresses the secretion of TNF-α in mononuclear cells and in CD14⁺ monocytes (55). Oral administration of KYNA decreases the activity of the peripheral blood leukocytes in mice (56).

The immune response signaling pathway and metabolic regulation signaling pathway, especially insulin signaling pathway, are highly integrated, because organism would need to redistribute its energy resources during the activation of immune response (57). Chronic inflammation is activated in overweight individuals as a consequence of adipose expansion. Recent insights suggest that it may play an indispensable role in the over-nutrition induced insulin resistance (58). During the past two decades, it became clear that nutrient excess and activation of the innate immune system are highly associated in most organs such as adipose tissues, liver, gut, muscle, and islets (58). Low-grade chronic inflammation or metabolically triggered inflammation is considered as a fundamental characteristic of metabolic diseases particularly in the context of obesity and type 2 diabetes. Targeting inflammation has been suggested as an important strategy to prevent and control these diseases (58).

Numbers of studies suggest that KYNA is produced during inflammation and it has been shown to mediate various immunomodulatory effects under inflammatory conditions (2). Since inflammation is one of the main factors in many metabolic diseases, it can be foreseen that KYNA may also play an important regulatory role in the metabolic diseases.

Increased tryptophan/kynurenine metabolite levels are frequently observed in overweight individuals (3). An increased serum KYNA level can be found in Zucker fatty rats (59), and in HFD fed LDL receptor knockout mice (60). Clinically, serum KYNA has been found positively correlated with BMI in overweight individuals (3, 61). The elevation of serum KYNA concentrations is closely associated with the activation of immune cells as increased IDO1 activity in macrophages and increased serum KYNA levels have been reported in obese animal models. The increased KYNA levels may result from an up-regulated biosynthesis in the omental adipose tissue (but not in subcutaneous adipose tissue), as the expression levels of IDO1, KYAT1 and KYAT3 were significantly higher in overweight individuals than in lean individuals (3). Furthermore, the activation of KYNA production was not restrained in resident immune cells of adipose tissue as the increased expression of IDO1, KYAT2 and KYAT3 can also be found in adipocytes. This up-regulation may be due to the increased production of pro-inflammatory cytokines from resident immune cells since adipocyte does not express KMO (3). However, in another study, Pyun et al. found a negative correlation between serum KYNA levels and BMI (6). These controversial observations may be due to the different criteria. Another possible explanation to the controversy is that the KYNA determination method used in these studies are different: the last one used an ELISA kit to determinate serum KYNA levels, while the others used the HPLC-MS/MS method. It needs to be further verified whether these divergent results are due to some immeasurable confounding factors.

Liver seems to be another important source of serum KYNA in overweight individuals as TDO and KYATs are highly expressed in this organ (17, 18). Moreover, overweight is frequently associated with a low-grade chronic inflammation with an induction of IDO1 in liver (62). Furthermore, both of the liver and adipose tissue have a closed crosstalk between resident immune cells (Kupffer cells or macrophages) and metabolic cells (adipocytes or hepatocytes). However, to the best of our knowledge, neither the production of KYNA nor the regulation of IDO, TDO and KYATs in hepatocytes under metabolic challenge have been reported.

Physical exercise has been described as a promising non-pharmacological treatment for overweight and some other metabolic diseases (63). In addition to its role in energy expenditure regulation, skeletal muscle is increasingly considered as one of the largest endocrine organs in our body. It secretes a variety of myokines and bioactive metabolites, which exerts important effects on the regulation of metabolism and inflammation. KYNA can also be synthesized by skeletal muscle and its production is closely correlated to the physical exercise in both human and mouse (11, 64).

All the four KYATs are expressed in skeletal muscle, but KYAT isoforms display fiber-type specific expression. KYAT1, KYAT3 and KYAT4 are more abundant in oxidative type I than glycolytic type II fiber (65). Accordingly, an increased serum KYNA level has been found in individuals after endurance exercise (64). Within the first hour after aerobic exercise, there is an increase in plasmatic KYNA and this effect lasts for 2 hours after exercise. In contrast, high-intensity eccentric exercise did not lead to increased plasmatic KYNA concentration (64). Regarding the effect of long-term exercise on KYNA, a recent study of 4-week physical exercises on human found that physical exercises promote an increase in the amount of KYNA in sweat on day 14. The KYNA level returned to baseline on day 28 (66). Additionally, inhibition of KYATs reduces myotube oxidative capacity and exercise performance in mice (67).

Physical exercise induces peroxisome-proliferator activated receptor γ coactivator 1α (PGC-1α) expression in skeletal muscle (11). PGC1-1α coordinates the expression of several genes involved in the adaptive energy metabolism and fatigue-resistance such as mitochondrial biogenesis and fatty acid oxidation. Recently, the canonical and longest transcript variant of PGC-1α, PGC-1α1, is reported to up-regulate KYAT2 and KYAT4 expression (11). Such mechanism in skeletal muscle during physical exercise may be primarily aimed at enhancing the malate-aspartate shuttle as both KYATs are important enzymes in the malate-aspartate shuttle (67). Consequently, this mechanism shifts the kynurenine metabolism to KYNA production.

This exercise-induced KYNA production is originally described as the crosstalk between skeletal muscle and the brain to elucidate the effectiveness of exercise in reducing depressive symptoms. A recent study by Agudelo et al. demonstrated that KYNA increases energy utilization by activating G-protein-coupled receptor 35 (GPR35), which stimulates lipid metabolism, thermogenic, and anti-inflammatory gene expression in adipose tissue (68). Also, GPR35 agonists was reported to suppress high fat diet-induced fatty liver development (5). These data uncovered that skeletal muscle derived from KYNA may be a potential regulator of energy homeostasis and a coordinator of exercise-induced adaptations in other organs including liver, adipose tissue and brain.

However, it should be noted that physical exercise induces strong and transit increases in KYNA levels while inflammation leads to mild and sustained increases in KYNA levels.

Epidemiological and experimental data have suggested that perinatal nutrition has a significant role in the development of lifelong metabolic disorders (69). KYNA may also act as a link between perinatal offspring and mother. KYNA can pass through the placenta into the fetus (70). However, placental and fetal KYNA were not affected by placental infusion of L-KYN in mice (70). Also, under physiological conditions, KYNA was higher in the liver and brain of mouse’s fetuses than in the placenta, and KYNA in the fetus was not affected during oral maternal administration of KYNA. It can be hypothesized that maternal KYNA cannot affect fetus through placenta (70, 71).

Although maternal KYNA cannot directly affect the fetus, some studies found the KYNA content in breast milk gradually increases in different lactation periods (10). Epidemiological studies showed a slower body weight gain in naturally fed newborns compared to artificially fed ones (72). Although the formula milk powder for infants in different periods are different, studies have found that KYNA content in formula milk powder is much lower than that in breast milk (10). Rats postnatally exposed to KYNA supplementation were observed to have a significant reduction of body weight gain, but no changes in total body surface and bone mineral density. The rat offspring supplemented with KYNA presents a lower mass gain during the first 21 days of life, which indicates that KYNA may act as an anti-obese agent (10).

Another potential mechanism is that perinatal KYNA may be protective against overweight by modulating the gut microbiota. Formula feeding appears to promote the microbiota associated with overweight (73), while KYNA stimulates the growth of certain probiotics (74). It still needs to be further explored whether the presence of KYNA in breast milk acts as a modulator of gut microbiota.

KYNA is well-known for its role as an endogenous N-methyl-D-aspartate receptor (NMDAR) antagonist in the brain. In the periphery, the expression of functional NMDA receptor is reported in the pancreatic β-cell. Activation of NMDA receptor reduces the glucose-stimulated insulin secretion. Likewise, NMDA receptor knockout in mouse islets increases glucose-stimulated insulin secretion. NMDA activation in β-cells also promotes cell death under stress. In microphages, activation of NMDA receptor induces ABCA1 degradation which promotes cholesterol accumulation and foam cell formation (75). In liver, NMDA receptor is present on the surface of Kupffer cells, and its activation has been reported to limit inflammasome activation (76).

KYNA acts as a low affinity competitive antagonist of AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid) receptors. It directly acts on the glutamate binding domain. Meanwhile, low concentrations (0.03–30μM) of KYNA potentiate AMPA receptor responses (77). Therefore, KYNA has a dual action on AMPA receptor responses.

KYNA also directly interacts with the glutamate-binding domain of kainate receptors. Its antagonistic effect on this type of receptor is the least potent (IC50 500µM) among the 3 types of glutamate receptors (78).

However, serum KYNA level hardly reaches to micromolar levels. It is unclear whether KYNA in serum or in the periphery is sufficient to antagonize these glutamate receptors.

The G-protein-coupled receptor 35 (GPR35) is an orphan receptor that was identified in 1998 (79). It was originally described as a receptor for zaprinast, a phosphodiesterase (PDE) inhibitor. Recently, the KYNA was identified as an endogenous ligand for GPR35 with an EC₅₀ of 39µM in human and 7.9µM in rat (80). Although the plasmatic concentration of KYNA is often in the nanomolar range in humans, it can become micromolar under inflammatory conditions.

GPR35 is associated with Gi/G0 and G13 proteins (81). Thus, activation of GPR35 reduces the activity of adenylate cyclase (Gi/G0) and/or increases that of the RhoA pathway (G13). GPR35 is expressed in central nervous system and in many peripheral tissues. In humans, significant expression of GPR35 has been detected in the colon, pancreas, small intestine, spleen and immune cells (monocytes, neutrophils, T cells and dendritic cells). The level of its expression is lower in the stomach, skeletal muscle, adipose tissue, pancreatic islets, kidney, liver, and thymus (82). Activation of GPR35 by KYNA has anti-inflammatory effect (83) by inducing autophagy-dependent degradation of NLRP3 in macrophage (84). It also plays anti-nociceptive (85) and anti-asthmatic (86) roles. Furthermore, KYNA enhances Pgc-1α1 and UCP1 expression GPR35 signaling in adipocytes, which suggests KYNA is a signaling molecule which directly controls energy homeostasis (68).

More recently, KYNA has been discovered to significantly increase AMP-activated protein kinase (AMPK) phosphorylation and to ameliorate palmitate-induced inflammation and insulin resistance. It potentially alleviates inflammation and insulin resistance in skeletal muscle and adipose tissues through GPR35/AMPK and SIRT6-mediated pathways (87). It may also ameliorate hepatic steatosis via the AMPK/autophagy‐ and AMPK/ORP150‐mediated suppression of endoplasmic reticulum stress (6).

KYNA is also identified as an endogenous ligand for Aryl hydrocarbon receptor (AhR). AhR was originally described as a xenobiotic receptor, also known as the dioxin receptor. It is activated by exogenous ligands, such as flavonoids, natural plant polyphenols, indoles and dioxins. AhR plays multiple roles in xenobiotic metabolism, the regulation of inflammation, development, and the homeostasis of several organs (88). DiNatale and colleagues (89) showed that KYNA is a potent endogenous agonist of AHR with an EC₂₅ around 100nM. Activation of AHR by KYNA may lead to IL6 expression in tumor cells.

Moreover, studies have shown that KYNA can also act at nicotinic receptors as a potent noncompetitive antagonist, particularly at the α7 subunit of the nicotinic receptor (90). KYNA inhibits CHRNA7 in a non-competitive manner at physiological concentrations. CHRNA7 was found to be expressed in glutamatergic axon terminals. Activation of CHRNA7 enhances glutamate release. Thus, KYNA may also be involved in the repression of glutamate release at the presynaptic level. This represents another mechanism by which KYNA exerts its anti-glutamatergic effect (90). The characteristics of KYNA receptors are summarized in Table 2 .

In addition to its receptor-dependent effects, KYNA at high concentrations (100-300 µM) also acts as a potent endogenous antioxidant, as it is a scavenger of free radicals such as hydroxyl radicals (OH▪), superoxide anion (O₂ ^-) and peroxynitrite (ONOO^-) (92). Since oxidative stress is also critical for the pathogenesis of metabolic diseases (93), the antioxidative properties of KYNA represent an important mechanism in preventing the onset of metabolic diseases.

Interestingly, another study showed that KYNA is not a guaranteed protector against oxidative stress. It exhibits a strong pro-oxidative effect combined with δ-aminolaevulinic acid (ALA), an endogenous precursor of heme and source of hydroxyl radical, and elevates deoxyribose deterioration by 9 times compared to ALA alone (94).

The mitochondrial localization of KYAT suggests a direct release of KYNA into the mitochondria. KYNA plays a key role in the redox balance in the mitochondria. The expression and function of KYATs has been shown to be diminished in rat model with mitochondrial dysfunction (95, 96).

Firstly, the formation of KYNA diverts the pathway from de novo synthesis of NAD+/NADH, which regulates the mitochondrial TCA cycle, oxidative state and mitochondrial dynamics, suggesting the involvement of KYNA in the mitochondria energy metabolism regulation. MPTP and 3-nitropropionic acid (3-NA) have inhibitory effects on mitochondrial respiratory chain complexes and on KYAT1 and KYAT2, thus compromise the ATP and KYNA production in the mitochondria (96). Experimentally, FK506, a neuroimmunophilin drug, not only enhanced the formation of KYNA, but abolished the inhibition of KYNA synthesis caused by MPTP and 3-NA. This result suggested that the restoration of respiratory chain function may activate the KYNA synthesis pathway (97). In the case of monogenic form of Leigh Syndrome, the loss-of-function mutation in LRPPRC causes mitochondrial RNA metabolism disorder. The metabolic signature demonstrated a decrease in kynurenine, the precursor of KYNA (98). In patients with Schizophrenia, the prescription of N-acetylcysteine has shown inhibitory effect on KYAT, decreasing the deleterious effect of elevated KYNA on glutamate and dopamine signaling (99). Furthermore, the plasmatic KYNA is positively correlated with fatty acid oxidation and mitochondrial proliferation in the liver of rat (100).

Secondly, KYNA has shown scavenging property of OH▪, O₂ ^-, ONOO^- (92, 101). In a preparation of oocytes, KYNA significantly reduced ROS and lipid peroxidation induced by FeSO₄. For 3-Methylglutaric acid (3MGA) which accumulates in the brains of children coursing with metabolic acidurias, experiments showed that 3MGA induced an increase in ROS production and lipid peroxidation and a decrease in mitochondrial function. Addition of KYNA showed antagonist effects (102).

Thirdly, as has been discussed in 4.1, KYNA has potent antagonist effects over NMDAR (103), which may reduce de notorious effect of excitotoxicity on mitochondria, via decreasing excessive intracellular Ca²⁺ as example (104).

Finally, KYNA was shown to directly impair respiratory parameters of heart mitochondria. Moreover, the effect is selective for complex I (105, 106). However, this respiratory chain-modulating property was only observed in heart mitochondria, and is absent for brain and liver mitochondria, suggesting profound differences between tissular mitochondria content and helping to explain the tissue-specific effect of KYNA. It should be pointed out that these in vivo experiments were carried out with high concentration of KYNA (125-1000 μM) which is rarely achieved under physiological conditions (92).