Tryptophan–Kynurenine Metabolism as a Common Mediator of Genetic and Environmental Impacts in Major Depressive Disorder: The Serotonin Hypothesis Revisited 40 Years Later
Abstract
The original 1969 Lancet paper proposed, “in depression the activity of liver tryptophan-pyrrolase is stimulated by raised blood corticosteroids levels, and metabolism of tryptophan is shunted away from serotonin production, and towards kynurenine production.” Discovery of neurotropic activity of kynurenines suggested that up-regulation of the tryptophan-kynurenine pathway not only augmented serotonin deficiency but also underlined depression-associated anxiety, psychosis and cognitive decline.
The present review of genetic and hormonal factors regulating kynurenine pathway of tryptophan metabolism suggests that this pathway mediates both genetic and environmental mechanisms of depression. Rate-limiting enzymes of kynurenine formation, tryptophan 2,3-dioxygenase (TDO) and indoleamine 2,3-dioxygenase (IDO) are activated by stress hormones (TDO) and/or by pro-inflammatory cytokines (IDO). Simultaneous presence of high producers alleles of proinflammatory cytokines genes (e.g., interferon-gamma and tumor necrosis factor-alpha) determines the genetic predisposition to depression via up-regulation of IDO while impact of environmental stresses is mediated via hormonal activation of TDO. Tryptophan-kynurenine pathway represents a major meeting point of gene-environment interaction in depression and a new target for pharmacological intervention.
Although often referred to as “serotonin hypothesis,” the 1969 Lancet paper proposed the disturbances of tryptophan (TRY) metabolism, i.e., the shunt of TRY from serotonin (5-HT) synthesis to kynurenine (KYN) formation, as a major etiological factor of depression (1). It suggested the formation of “vicious cycle” perpetuating the increase of KYN and decrease of 5-HT production in depression due to a) stress hormones – induced activation of tryptophan 2,3-dioxygenase (TDO), the rate-limiting enzyme of TRY – KYN pathway; b) diminished availability of TRY as an initial substrate of 5-HT biosynthesis due to increased formation of KYN from TRY; and c) increased production of cortisol due to weakening of 5-HT inhibitory effect on amygdaloidal complex (2) (Fig. 1).
5-HT deficiency was thought as a major consequence of the shift of TRY metabolism to KYN formation, and “intensification of the central 5-HT-ergic processes” was suggested as “a possible determinant of the thymoleptic (mood-elevating) effect” (1). Introduction and wide use of selective 5-HT uptake inhibitors as antidepressant drugs contributed to almost 40 years of continued interest in the “serotonin hypothesis.”
Another important consequence of so-called “serotonin hypothesis” was stimulation of research of biological and neurotropic activity of KYN and its derivatives (summarily called “kynurenines”) (3–7) and of factors regulating KYN pathway of TRY metabolism (8). This review offers analysis of the current status of the serotonin hypothesis with special consideration of the discovery of indoleamine 2,3-dioxygenase (IDO) (9), the other rate-limiting enzyme of TRY - KYN pathway, different from TDO in substrate specificity, localization and regulatory mechanisms.
TRYPTOPHAN METABOLISM
In humans TRY is an essential amino acid with two non-protein metabolic pathways: methoxyindoles and KYN (Fig. 2).
Methoxyindoles and kynurenine pathways of tryptophan metabolism
Abbreviations: TRY – tryptophan; IDO – indoleamine 2,3-dioxygenase; TDO- tryptophan 2,3-dioxygenase; TPH – tryptophan hydroxylase; 5-HT – serotonin; NAT – N-acetyltransferase; NAS – N-acetylserotonin; NAD - nicotinamide adenine dinucleotide
THE METHOXYINDOLES PATHWAY
Availability of TRY as a substrate is one of the rate-limiting factors of methoxyindoles pathway of 5-HT biosynthesis since less than 5% of TRY metabolized along this pathway (10). The other rate-limiting step is hydroxylation of TRY catalyzed by TRY-hydroxylase with the formation of 5-hydroxytryptophan. The subsequent decarboxylation results in the formation of 5-HT, a substrate for melatonin synthesis. The rate-limiting step of melatonin synthesis is 5-HT-N-acetylation resulting in the formation of N-acetyl-serotonin (NAS) with subsequent O-methylation into 5-methoxy-N-acetyltryptamine (melatonin) (11) (Fig. 2).
5-HT (not competitively) and NAS and melatonin (competitively) inhibit liver TDO (12). Deficient production of 5-HT, NAS and melatonin contribute to depressed mood (13), and disturbances of sleep (14) and circadian rhythms (15).
THE KYNURENINE PATHWAY
About 95% of TRY is metabolized via the KYN pathway (10, 16). There are two steps of the TRY–KYN pathway: a) formation of KYN from TRY, and b) post-KYN metabolism via two routes competing for KYN as their initial substrate.
a. Tryptophan conversion into Kynurenine
Unlike the methoxyindoles pathway that does not affect the indole ring of TRY, the KYN pathway begins by the cleavage of the indole ring of TRY which results in the formation of N-formylkynurenine followed by kynurenine in an ensuing step (10). The rate-limiting enzymes of KYN formation from TRY are IDO (9) in astrocytes, microglia, microvascular endothelial cells and macrophages and TDO in liver, kidney and brain (10).
KYN inhibits TRY transport via the blood-brain barrier (4), stimulates IDO activity (10), and exerts anxiogenic activity in animal models of anxiety (4).
b. Post-Kynurenine metabolism
Kynurenine is further metabolized along the two distinct routes competing for KYN as a substrate: KYN–kynurenic acid (KYNA) pathway, and KYN–nicotinamide adenine dinucleotide (NAD) pathway
b1. The KYN –KYNA pathway
The KYN-KYNA pathway is regulated by KYN aminotransferases, the major biosynthetic enzymes of KYNA formation in the brain (17).
KYNA, the only known endogenous antagonist to N-methyl-D-aspartate (NMDA) receptors, might, similarly to the exogenous NMDA antagonists, ketamine and MK-801, exert antidepressant (18) and psychotomimetic (19) effects. KYNA may contribute to cross talk between the melatonin and kynurenine pathways by inhibiting 5-HT-N-acetylation, the rate-limiting step of melatonin biosynthesis (20).
KYNA has higher affinity to alpha-7-nicotinic acetylcholine than to NMDA receptors, and as such might contribute to cognitive impairment observed in depression, schizophrenia, dementia, and Down’s and Crohn’s syndromes (21, 22).
b2. The KYN-NAD pathway
The KYN–NAD pathway produces NMDA agonists (quinolinic and picolinic acids) and free radical generators (3-hydroxykynurenine and 3-hydroxyanthranilic acid) (16). Increased formation of NMDA agonists might result in hyperglutamatergic status suggested to be associated with depression (21).
Quinolinic and picolinic acids exerted an anxiogenic effect in experimental models (4). Quinolinic and picolinic acids stimulate inducible nitric oxide synthase (iNOS) and together with 3-hydroxykynurenine and 3-hydroxyanthranilic acids might increase lipid peroxidation, and trigger arachidonic acid cascade resulting in the increased production of inflammatory factors: prostaglandines, via activation of cycloxygenase (COX) and leucotrienes, via activation of arachidonate 5-lipoxygenase (5-LO) (16, 23, 24). COX-2 is of particular interest since its inhibitors blocked KYNA production (25) and exerted antidepressant and antipsychotics effects (21) while 5-LO was suggested as a link between depression and atherosclerosis (26–28) (Fig. 3).
Kynurenines and psychiatric and vascular complications
Abbreviations: TRY – tryptophan; IDO – indoleamine 2,3-dioxygenase; TDO – tryptophan 2,3-dioxygenase; BBB – blood brain barrier; KYN - kynurenine, 3HKYN - 3-hydroxyKYN; KYNA – kynurenic acid; QUIN – quinolinic acid; iNOS – inducible nitric oxide synthase; NO - nitric oxide; PLA – phospholipase; AA – arachidonic acid; COX - cycloxygenase; 5-LO - arachidonate 5-lipoxygenase; PGE – prostaglandines.
REGULATION OF RATE-LIMITING ENZYMES OF KYN PATHWAY
REGULATION OF TDO
a. Substrate activation
TDO is activated by its substrate (TRY) (10). Because KYN competes for cerebral transport and cellular uptake of TRY, and because of substrate inhibition on TRY hydroxylase, the rate-limiting enzyme of 5-HT biosynthesis, excessive TRY doses may decrease 5-HT production (29).
b. Hormonal activation
Cortisol activates TDO and increased KYN production (30). Literature data regarding the effect of estrogens and testosterone on TDO are controversial: both adrenalectomy and ovariectomy reduced TDO activity in homogenates of liver from mature rats. However, administration of estrogens and testosterone had no effect on TDO (31, 32).
Hormonal activation of TDO and consequent shift of TRY metabolism from 5-HT to KYN formation was suggested as an etiological factor in depression (1) (Fig.1).
REGULATION OF IDO
a. Cytokines and IDO
a1. IFN-gamma
Pro-inflammatory cytokines (including, most notably, interferons) transcriptionally induce IDO in a variety of immune cells (e.g., monocyte-derived macrophages and microglia) (33). IFNG is the strongest known inducer of IDO (34). Therefore, the shift of TRY metabolism from 5-HT to KYN formation might be caused not only by TDO activation by cortisol but by IFNG-induced IDO activation as well. The shift of TRY towards formation of kynurenines might be further augmented by IFNG-induced stimulation of the enzymes of KYN – NAD route: 3-hydroxylase and kynureninase (35).
a2. IFN-alpha
Systemically administrated IFN-alpha passes the blood-brain barrier and reaches effective concentrations acting on microglial cells as well as macrophage receptors (36). IFN-alpha has much weaker direct IDO stimulating effect than IFNG but might increase IDO activity by stimulating the production of IFNG and TNF-alpha (37) (Fig. 4). Administration of IFN-alpha to patients with hepatitis C is associated with depression attributed to increased formation of kynurenines (38–40).
a3. TNF-alpha
TNF-alpha stimulates IDO activity and enhances (up to 300%) IFNG-induced IDO expression (41). The induction of IDO by the bacterial endotoxin lipopolysaccharide was IFNG-independent and might be mediated by toll-like receptors (42).
a4. Other proinflammatory molecules
such as IL-1, IL-12, Il-18, PGE2 synergistically with IFNG induce IDO activity (43, 44).
Experimental and clinical data demonstrated that IFNG and TNF-alpha trigger depression (and depressive-like symptoms) via stimulation of IDO and consequent increase of kynurenines formation from TRY (39, 45).
CYTOKINE GENE POLYMORPHISM AND IDO
Cytokine genes are polymorphic and certain SNPs located within coding/regulatory regions affect the overall expression and secretion of cytokines (46).
IFNG production is encoded by polymorphic IFNG (+874) gene with high (T) and low (A) producer alleles (46). Mean concentration of IFNG cytokine released by stimulated peripheral blood mononuclear cells was higher in healthy carriers of T than in carriers A allele (47). High producer T allele was associated with increased IDO activity (i.e., elevated plasma kynurenine levels and kynurenine/tryptophan ratios) in healthy females (48). These results suggest that IFNG genotype influences TRY catabolism via regulation of IDO activity.
TNF-alpha production is encoded by the TNF-alpha (−308 A/G) polymorphic gene. Since TNF-alpha stimulates IDO and potentiates IFNG-induced stimulation of IDO (see above), a (TNF-alpha (−308) high producer (A) allele might strength the association between IFNG (+874) high producer (T) and IDO up-regulation (16). Increased frequency of the TNF-alpha −308A allele (high producer) was reported in Korean subjects with major depression (49), in Korean bipolar I patients (50), and in Polish subjects with bipolar affective disorder and positive family history (51). It suggested that (−308) TNF-alpha gene polymorphism might be involved in genetic susceptibility to mood disorders.
IDO–TDO INTERACTION
a. Hormonal activation of IDO
Although IDO is mainly induced by cytokines, experimental data suggested that expression of IFNG gene may be subject to direct hormonal control since receptors of prolactin and IFNG share their structure and signal transduction pathway. Prolactin by itself has little or no effect on IDO but potentiates INFG-induced IDO activation in CD14-positive cells (52). 17beta-estradiol increased the activity of the IFNG promoter in lymphoid cells (53).
Hydrocortisone and dexamethasone induced IDO in human astrocytoma cells and in native human astrocytes (54).
b. Cytokines and the HPA axis
Proinflammatory cytokines may cause hypothalamic-pituitary-axis (HPA) hyperactivity (that is frequently observed in depression) by disturbing the negative feedback inhibition of circulating corticosteroids on the HPA axis (55, 56).
c. Aging as a merging point of IDO-TDO interaction
Aging is characterized by elevated cortisol production due to disinhibition of the HPA axis (57–59), and by increased IFNG and TNF-alpha production (60). It is noteworthy that high producer allele (T) of the IFNG +874 gene (61) and high IDO activity predicted high lethality in elderly subjects (62). In the same vein, Drosophila Melanogaster mutants with impaired KYN production have longer life span than wild type flies (63). These results suggest that activation of both IDO and TDO might contribute to high risk of depression in the elderly.
HYPOTHESIS
The original 1969 hypothesis of up-regulation of the KYN pathway of TRY metabolism as an etiological factor in depression suggested that “in depression the activity of liver TRY-pyrrolase (TDO) is stimulated by raised blood corticosteroids levels” that resulted in 5-HT deficiency due to the shift of TRY metabolism from 5-HT to KYN formation (1). The discovery of neurotropic activity of kynurenines (37) emphasized the increased formation of kynurenines as etiological factor in depression (in addition to 5-HT deficiency) (39). IDO, the other enzyme catalyzing TRY conversion into KYN, is transcriptionally activated by pro-inflammatory cytokines (mainly, IFNG and TNF-alpha). The polymorphisms of genes impacting the production of pro-inflammatory cytokines provides further insight in the etiological role of upregulated TRY – KYN metabolism in depression.
This review suggests that genetic and/or environmental (life stresses) factors trigger depression by upregulation of TRY-KYN metabolism. Effect of genetic factors (such as high producer alleles of pro-inflammatory genes) is mediated by cytokine-induced up-regulation of IDO. Combination of high producer alleles of IFNG (+874) and TNF-alpha (−308) genes might result in high production of these cytokines, and trigger the “super-induction” of IDO. The effect of life stressors might be mediated by hormonal activation of TDO. Cytokine–induced stimulation of cortisol production and augmentation of IFNG-induced activation of IDO by stress hormones suggest that the TRY-KYN pathway might be the converging point of gene-environmental interaction (e.g., like in aging) (Fig. 5).
LIMITATIONS
Some major limitations should be mentioned with the hope to stimulate further evaluation of the proposed hypothesis.
The present hypothesis predicts higher frequency of carriers of high producer alleles of pro-inflammatory genes in subjects with mood disorders. While some studies of TNF-alpha genotypes support this suggestion (49–51), no publications regarding IFNG (or other pro-inflammatory cytokines) were found in the available sources.
The present hypothesis is based on the assumption that carriers of high producer alleles of cytokine genes have higher production of cytokines than carriers of low promoter alleles. These relationships were reported for healthy volunteers but were not studied in depressed patients. Similarly, the association between high producer alleles for the IFNG (+874) gene and high IDO activity was observed in healthy volunteers but no studies done in depressed patients.
TRY–KYN METABOLISM AS A NEW TARGET FOR PREVENTION AND TREATMENT OF MDD AND PSYCHIATRIC COMPLICATIONS OF IFN-ALPHA TREATMENTS
Genotype assessment might help identify subjects-at-risk of developing depression in response to environmental stressors and/or to IFN-alpha therapy of hepatitis C, cancer, amyotrophic lateral sclerosis and multiple sclerosis.
Potential pharmacological interventions in identified subjects may include:
- a)inhibition of cytokine production by antibodies to TNF-alpha (e.g., etanercept, infiximab) and IFNG (64), and/or more careful selection of antidepressants. While both selective 5-HT uptake inhibitor, fluoxetine (65) and the dopamine enhancer, wellbutrin (66) inhibit cytokine production, the latter might be of advantage considering the impaired 5-HT synthesis as a result of IDO activation.
- b)
- d)administration of methoxyindoles that might modulate the TRY-KYN pathway due to their inhibitory effect on cortisol (72) and proinflammatory cytokines (73–75) production. Methoxyindoles (melatonin, in particular) might attenuate excitatory, glutamate-mediated responses triggered by KYN pathway metabolites (76, 77).
Acknowledgement
This paper is supported by NIH R21MH083225.





