1. Tryptophan Metabolism via the Kynurenine Pathway
In tryptophan catabolism, one of the pathways of particular biological interest is that leading to the formation of niacin, the kynurenine pathway [
1], (see
Figure 1). Tryptophan catabolism along this pathway has a remarkable function, namely the synthesis of neuroactive metabolites, 3-hydroxykynurenine with neurotoxic [
2], quinolinic acid with excitatory [
3] and kynurenic acid (KYNA) with inhibitory properties [
4]. L-kynurenine, the first major metabolite of tryptophan catabolism along the kynurenine pathway, crosses the blood-brain barrier well [
5], is actively accumulated in various tissues (brain, heart, liver) [
5], and is taken up by various cells (astrocytes, neurons, macrophages) including their organelles (mitochondria) [
6,
7].
In the kynurenine pathway, L-kynurenine is catabolised by kynurenine-3-hydroxylase, followed by kynureninase and 3-hydroxyanthranilic acid oxygenase to quinolinic acid [
4]. And quinolinic acid is catabolised by quinolinic acid phosphoribosyltransferase to nicotinic acid mononucleotide and finally to the active coenzyme nicotinamide adenine dinucleotide (NAD) [
1,
4].
The other possibility of L-kynurenine degradation is the formation of KYNA by irreversible transamination of L-kynurenine [
8], or the formation of anthranilic acid by kynureninase [
1], followed by hydroxylation of anthranilic acid to 3-hydroxyanthranilic acid [
9]. KYNA synthesis from tryptophan using indole-3-pyruvic acid has also been described [
10,
11,
12], but its significance remains to be elucidated.
The capacity for KYNA formation from L-kynurenine by kynurenine aminotransferase (KAT) is very high and several KATs have been discovered in different mammalian organs [
13,
14].
In rat and human peripheral tissues, four types of proteins are capable of catalysing the kynurenine-2-oxoacid transamination reaction to form KYNA [
13,
14,
15,
16,
17]. Very low activity of KYNA formation has been reported in peripheral tissues from pigs [
17] or the snail
Helix pomatia [
18], suggesting species differences. The central nervous system (CNS) is also significantly involved in KYNA formation and three proteins, KAT I, KAT II and KAT III, have been found in rat, mouse, piglet, snail and human brain tissue [
19,
20,
21,
22,
23,
24,
25].
They are characterised by specific enzymatic properties and different biochemical activities, and their different physiological roles have been suggested. KAT I, characterised by a high pH optimum of 9.6, may be particularly important in pathological conditions [
19,
20,
21], whereas KAT II, with a neutral pH of 7.4, may operate essentially under physiological conditions [
21,
22], and KAT III, with a pH around 8.0-9.0 [
14,
23,
24,
25], may share its action between physiological and pathological conditions. There are data suggesting that human KAT I is a protein with multifunctional activities and may also be an important protein in KYNA synthesis under physiological conditions [
20]. Interestingly, KATs appear to have the ability to change their chemical properties and presumably their actions under physiological and pathological conditions [
26,
27]. KYNA formation has been shown to occur preferentially in glia, astrocytes and to a lesser extent in neurons [
7,
16,
26,
27,
28].
2. NMDARs and α-7nAChRs, and Neuroprotection
In the last decades of the last century, important information about the action of KYNA was revealed. KYNA acts as an endogenous antagonist of the glutamate ionotropic excitatory amino acid receptors N-methyl-D-aspartate (NMDAR), alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid, kainate [
4,
29,
30], and the nicotinic acetylcholinergic subtype alpha-7 receptor (α-7nAChR) [
31], but is also an agonist at the orphan G-protein coupled receptor GPR-35 [
30] and has anticonvulsant and neuroprotective activities [
32,
33]. In the brain, KYNA levels range from low nanomolar to low micromolar concentrations [
4,
10] and there is evidence that these physiologically relevant concentrations of KYNA are likely to block α-7nACh more effectively than NMDA receptors [
31,
34,
35]. With regard to KYNA inhibition of glutamatergic neurotransmission, accumulated data suggest that the glycine likely co-agonist site of the NMDA receptor is not saturated [
4,
36,
37,
38], so physiological concentrations of KYNA may be sufficient to inhibit NMDA receptor activity.
With regard to KYNA inhibition of cholinergic neurotransmission, KYNA inhibits α-7nAChRs non-competitively (IC 50 approximately 7 µM) [
31,
34,
35], the inhibition of α-7nAChRs by KYNA is likely to be voltage independent [
34,
35]. Speculatively, it has been suggested that KYNA may act via intracellular second messengers that affect α-7nAChR function [
34,
35]. Interestingly, KYNA also increases the expression of non-α-7nAChRs, α-4-ß-2nAChRs [
30,
34,
35], suggesting a notable interaction between KYNA and the receptor functions of cholinergic neurons. Importantly, α-7nAChRs play multiple roles in modulating the glutamatergic system in both normal and diseased nervous systems [
38].
On the other hand, a biphasic change in KYNA formation has been reported after prolonged nicotine application [
39], suggesting a down- and/or up-regulation by nicotine.
The neuroprotective and anticonvulsant activities of KYNA have been documented [
32], and this role is most likely related to the modulation of both glutamatergic and acetylcholinergic neurotransmission [
3], since the overall effects of NMDA receptor antagonists and the effects of α-7nAChR antagonists on neuronal plasticity and also on viability are similar and close to those of KYNA, as demonstrated by different pharmacological approaches. For example, KYNA and α-7nAChR antagonists can block neurite outgrowth [
40,
41] or both can reduce apoptotic neuronal death [
40,
42] or both have anticonvulsant activity [
30,
43,
44]. Accumulating data suggest that both NMDA and nACh receptors are involved in the regulation of neuronal plasticity and survival in the brain. An antioxidant activity of KYNA may also play a role [
33].
3. Kynurenic Acid and Significant Abnormalities
Dysfunction of KYNA synthesis in the brain has been suggested as an important factor contributing to neuronal degeneration [
4,
30,
32]. Indeed, in vivo experimental studies have shown that reducing KYNA synthesis in the rat brain using non-specific inhibitors can lead to neurotoxic effects [
45,
46,
47]. Importantly, a preferential loss of layer III of the entorhinal cortex after local injection of a non-specific inhibitor of KYNA synthesis has been shown to be an important factor in the pathophysiology of human temporal lobe epilepsy [
46,
48]. On the other hand, increases in brain and serum KYNA levels have been observed in rats exposed to kainic acid-induced stereotypic behaviour and seizures [
49], or during oxygen deprivation [
50], or in a genetic model of dystonia [
51], or in chronic kainic acid rats with spontaneous seizures [
52], or after encephalomyocarditis infection in piglets [
53], or after HIV-1 infection in humans [
54]. Respiratory distress, bronchopneumonia, lobar pneumonia, pulmonary oedema or even tuberculosis occur after various infections such as EMCV, HIV-1, influenza A or Corona virus [
53,
54]. In particular, high mortality has been reported in infected mammals, including humans [
53,
54,
55,
56,
57,
58,
59]. It would therefore be reasonable to consider the importance of activation of kynurenine metabolism during events associated with impaired oxygen consumption capacity. Indeed, a marked increase in brain KYNA has been found during asphyxia due to oxygen deprivation [
50]. Of particular interest in this work was the observation that the longer the period of oxygen deprivation, the higher the observed peak in brain KYNA and the higher the lethality [
50]. The decrease in ATP synthesis also correlates significantly with the duration of asphyxia [
50,
60]. Of particular note is the dramatic increase in brain KYNA levels during the 15-20 min period of asphyxia, characterised by almost 100% lethality [
50], which may be relevant to ischaemic events in stroke or even sudden death [
53,
61]. Experimental work in vivo has shown that administration of KYNA (icv) to rats resulted in ataxia and stereotypy in a dose-dependent manner (0.025-1.6 µmol) [
62,
63]. Importantly, the authors demonstrated that administration of 0.8 µmol KYNA resulted in drowsiness and approximately 25% animal mortality, and at a dose of 1.6 µmol, all animals died of cardiorespiratory failure within 2-5 minutes [
62,
63].
These data suggest that spontaneous synthesis of KYNA in the brain due to infection and/or unknown cause could lead to cardiorespiratory dysfunction followed by acute death. Our data indicate that kynurenine metabolism is also activated in the brain after EMCV infection [
64,
65], and it is reasonable to assume that this increase may be responsible for the occurrence of cardiorespiratory dysfunction and sudden death. However, this proposed mechanism requires further investigation. We have previously demonstrated differences in the effect of tryptophan metabolites on cardiac and brain mitochondrial respiratory parameters. KYNA increases the oxygen consumption of rat heart mitochondria [
66,
67] and this observation suggests an essential role for KYNA in cellular mitochondrial function, at least in cardiac myocytes. While cardiac mitochondria are sensitive to KYNA, brain mitochondria are only slightly affected by this kynurenine metabolite [
67].
3-hydroxykynurenine and 3-hydroxyanthranilic acid have been shown to significantly affect mitochondrial function in the brain and heart. No effect was reported in the presence of the potent neurotoxin quinolinic acid [
67]. Furthermore, the sensitivity of brain and heart mitochondria to these metabolites was insignificantly affected by ageing, at least in healthy rats [
68], indicating a high availability of mitochondrial energy supply throughout life.
4. Kynurenic Acid Metabolism throughout Life
KYNA metabolism in the CNS shows a characteristic pattern of changes throughout the lifespan in different animal species and in humans [
69,
70,
71,
72,
73,
74]. A marked increase in KYNA levels in the CNS occurs before birth, accompanied by a dramatic decrease on the day of birth [
69,
73]. Low KAT activity was found during the first week after birth [
70] and a slow and progressive increase was observed during maturation [
70,
71] and ageing [
71,
72]. The data published at that time showed the changes in KAT II activity [
70], as the changes in KAT I or KAT III during ontogeny and ageing were not elucidated at that time.
The remarkable change in KYNA metabolism during ontogeny and maturation in the mammalian brain has been suggested to be a consequence of the development of neuronal connectivity organisation, synaptic plasticity and receptor recognition [
4,
30,
34,
35,
40,
48]. It is likely that KYNA has different functions throughout life, depending on the neuronal network currently available. Furthermore, the dramatic decrease in KYNA on the day of birth [
69,
73] suggests that KYNA is involved in biochemical processes/events that regulate lung function.
Consistent with an increase in KYNA throughout life, an increase in L-kynurenine, the bioprecursor for KYNA synthesis, has been reported in aged rats [
72] and also in the cerebrospinal fluid of elderly humans [
74]. Tryptophan has also been found to increase with age [
75]. The high levels of tryptophan in the elderly may be due to uncontrolled consumption, e.g. of chocolate, but also to a reduced capacity for biochemical degradation during ageing.
6. Kynurenic Acid and Neuropsychiatric Disorders and Dementia
The suggestion that elevated levels of KYNA in the CNS may be involved in cognitive decline is supported by the fact that elevated levels of KYNA have been found in patients with Alzheimer's disease [
84,
85,
86], patients with DOWN syndrome [
87] patients with hydrocephalus [
75], patients with subcortical sclerotic encephalopathy [
88], patients infected with the HIV-1 virus [
54,
89], patients with early stage Huntington's disease [
90], patients with schizophrenia [
91,
92], and also in normal elderly people [
74].
Interestingly, KYNA levels are significantly increased in demented PD patients but not in non-demented PD patients [
93], and these observations raise the question of whether an increase in KYNA levels in the brain of demented PD patients is an important neurochemical marker of dementia. Nevertheless, the development of pathological events depends not only on the degree of dopamine depletion [
94,
95] but also on the degree of KYNA enhancement, processes that may be interdependent or separate.
It is known from pharmacological studies that KYNA can improve cognition and memory [
96], but it has also been shown to impair working memory [
97], and increasing endogenous KYNA levels induces spatial memory deficits [
98].
In Alzheimer's disease, increased KYNA levels have been observed in several regions, the frontal cortex, hippocampus, putamen, caudate nucleus and cerebellum, ranging from 123 to 192% of CO [
86], suggesting a general increase in KYNA levels in the brain. However, a study of the enzyme activity of KATs showed a significant increase in KAT I in the putamen and caudate nucleus, but not in KAT II, which was only moderately affected,
Figure 2, adapted from [
86].
One might ask whether increased KYNA metabolism in the basal ganglia compensates for hyperactivity in the striato-frontal lobe, as we have previously suggested in Alzheimer's disease [
86]. The interaction between high KYNA levels and dopamine reduction in the striatum has been demonstrated [
99]. We therefore suggest that the decrease in dopaminergic neurotransmission of the nigrostriatal loop, which is characteristic not only of PD but also of ageing [
94,
95], is associated with an increase in KYNA metabolism in basal ganglia regions, as has been observed not only in PD patients [
85,
93] but also in people with Alzheimer's disease, vascular encephalopathy, cerebral infarction [
85] and ageing [
74,
85].
It is therefore reasonable to assume that dopaminergic neurotransmission in the caudate nucleus is upregulated by astrocyte-derived KYNA, especially with advancing age [
85,
94]. Some interesting findings and correlations can also be drawn from experimental data: the reduction of dopamine levels in various rat brain regions, such as the limbic and basal ganglia regions of rats in the kainic acid epileptic model [
100], in the acute phase is also associated with a significant increase in KYNA [
49].
Levels of the neurotransmitter serotonin were not altered during the acute phase, although there was a marked increase in the turnover of both catecholamine and serotoninergic neurons. The lack of serotonin lowering in the kainic acid model was one of the miracles of my study and work," commented Prof Oleh Hornykiewicz, my PhD supervisor at the time.
A pronounced increase in KYNA levels (
Figure 3, adapted from [
87]), decreased KAT I activity and normal KAT II (
Figure 4, adapted from [
87]) in the frontal and temporal cortex were measured in Down's syndrome subjects, indicating a remarkable discrepancy between high KYNA levels and low KAT activities in the frontal and temporal cortex of Down's syndrome subjects [
87]. It is likely that substrate availability, i.e. L-kynurenine or cellular KYNA, or reduced KYNA clearance may be responsible for the remarkably high KYNA levels. The question arises as to whether the genetic background of Down's syndrome plays a mitigated role in Alzheimer's disease, as they share similar neurochemical events related to the deficit in nACh receptor expression/function and the increase in KYNA in the brain (frontal cortex) [
82,
83,
86,
87].
It is also important to note the pathological state of the patients used in the Down's syndrome study. According to clinical findings, the control patients and especially the Down's syndrome patients had pathology related to infarction and bronchopneumonia, respectively [
87], processes characterised by increased KYNA levels [
86,
110]. This could be the reason for such a strong KYNA enhancement, indicating a very good correlation between oxygen impairment and early lethality [
53,
60,
62,
63].
Microglial activation is an important pathological feature of a variety of neurological disorders, including ageing, and correlates with increased KYNA metabolism [
101,
102]. There is a predominant astrocytic localisation of KAT in the rat brain [
16,
28] and the major intermediate filament component of astrocytes, glial fibrillary acidic protein (GFAP), increases with age [
102,
103,
104].
In addition, the interaction between glial and immune cells and the secretion of various cytokines such as interferon or interleukin-1 and 6, tumour necrosis factor [
105] in response to injury or infection or during the ageing process play a prominent role in the initiation and propagation of CNS damage [
104].
Activation of tryptophan degradation by tryptophan 2,3-dioxygenase by interferon has also been observed in human monocytes/macrophages and a variety of human cells and cell lines in vitro [
107,
108]. The positive correlation between an increase in CSF ß2-microglobulin and KYNA levels during the ageing process [
74] suggests an activation of immune cells. Increased CSF ß2-microglobulin levels have been reported in Alzheimer's disease [
108], cerebral infarction and meningitis [
53], and HIV-1 infection [
89,
109].
Neuropsychiatric symptoms, as often seen in patients infected with HIV-1 virus [
89] or encephalomyocarditis virus [
53], are also associated with a marked increase in KYNA metabolism in the brain [
49,
51,
75,
90,
91,
92].
A marked increase in KYNA levels in the brain after HIV-1 infection (
Figure 5, adapted from [
54]) correlated significantly with high KAT I activities, whereas KAT II was only moderately changed (
Figure 6, adapted from [
54]).
7. Different Types of Pathology after HIV-1
The assessment of kynurenine metabolism, e.g. L-kynurenine and KYNA levels and the activity of KYNA, KAT I and KAT II synthesising enzymes in the frontal cortex and cerebellum of HIV-1 infected patients in relation to different types of pathology classified as HIV in the brain (HIV), opportunistic infection (OPP), cerebral infarction (INF), malignant lymphoma of the brain (LY) and glial dystrophy (GD) and in controls (CO) showed significant alterations [
110]. Importantly, of all the pathologies studied, OPP was the most common (65%), followed by HIV (26%), LY, INF and GD (22% each). In addition, 68% of HIV-1 patients had bronchopneumonia, the highest incidence of which was 60% in the OPP and LY groups.
KYNA was significantly increased in the frontal cortex in LY (392% of CO), HIV (231% of CO) and GD (193% of CO) and in the cerebellum in GD (261% of CO). Concerning the enzyme KAT I, the activity was significantly increased in the frontal cortex of all pathological subgroups, i.e. OPP = 420% > INF > LY > HIV > GD = 192% of CO [
110]. Similar changes were found in the cerebellum, where KAT I activity was significantly increased in all pathological subgroups (OPP = 320% > LY, HIV > GD > INF = 176% of CO). In contrast, KAT II activity was only moderately but significantly higher in the frontal cortex of INF and OPP; in the cerebellum of HIV, OPP and LY it was comparable to that of controls, while in INF and GD it was even slightly reduced [
110].
On the other hand, L-kynurenine, a bioprecursor of KYNA, was increased in the frontal cortex of LY (385% of CO) and INF (206% of CO) and in the cerebellum of GD, LY, OPP and HIV (between 177 and 147% of CO). Correlation analyses between kynurenine parameters showed an association between a high KAT I/KAT II ratio and increased KYNA levels and lower L-kynurenine in the frontal cortex and cerebellum of the HIV and LY subgroups. The data revealed dramatic biochemical variability in the brain between the pathological groups [
110].
Interestingly, among the normal subjects we used as controls, we found some who had been diagnosed with bronchopneumonia [
110], and they were indeed characterised by high KYNA metabolism in the brain as well. KAT I activity in the frontal cortex and cerebellum was markedly increased by about 877% and 479% of CO, respectively. This finding suggests a remarkable correlation between impaired conditions of oxygen availability and increased KYNA formation in the human brain. These observations may also have implications for understanding the pathological processes in the brain following HIV-1 infection, as well as other infections associated with the development of neuropsychiatric and neurological symptoms, including memory and cognitive impairment, as well as lethality [
110].
Acknowledgments
I would like to thank everyone who contributed to my scientific work, especially my supervisors, Univ.-Prof. MD Hornykiewicz, O. Institute of Biochemical Pharmacology, Medical University of Vienna, Austria, and Univ.-Prof. Dr. Schwarcz, R. University of Maryland, USA, and Univ.-Prof. MD Kido, R. Wakayama Medical College, Japan, and especially Prim. MD Kepplinger, B. MSc, Clinical Director of the Neuropsychiatric Clinic Mauer, Austria, for his research interest. I would like to thank Mag. Carina Kronsteiner for her technical assistance.
This work was supported in part by the Austrian FWF (project P15371 to HB); the Austrian National Bank (Jubilee Fund, grant 12316 to HB); the Austrian NFB Life Science Project Nr. LS10-032 (to BK and HB); and by NÖ Forschung und Bildung, Austria (to HB).