Brain, Behavior, and Immunity 31 (2013) 128–133

Contents lists available at SciVerse ScienceDirect

Brain, Behavior, and Immunity

journal homepage: www.elsevier .com/locate /ybrbi

Tryptophan metabolism and immunogenetics in major depression: A role forinterferon-c gene

Aye Mu Myint a,b,⇑, Brigitta Bondy a, Thomas C. Baghai a, Daniela Eser a, Caroline Nothdurfter a,Cornelius Schüle a, Peter Zill a, Norbert Müller a, Rainer Rupprecht a, Markus J. Schwarz a

a Department of Psychiatry, Ludwig-Maximilian University Munich, Nussbaumstr. 7, D-80336 Munich, Germanyb School for Mental Health and Neuroscience, Faculty of Medicine, Maastricht University, Universiteitssingel 40, 6229 ER Maastricht, The Netherlands

a r t i c l e i n f o

Article history:Available online 15 April 2013

Keywords:Affective disorderKynurenine pathwayInterferon-gammaSNPTryptophan breakdownGenetics

0889-1591/$ - see front matter � 2013 Elsevier Inc. Ahttp://dx.doi.org/10.1016/j.bbi.2013.04.003

⇑ Corresponding author. Address: Laboratory for Pschiatric Hospital, Ludwig-Maximilian University, NussGermany. Tel.: +49 89 5160 3433; fax: +49 89 5160 5

E-mail address: AyeMu.Myint@med.uni-muenchen

a b s t r a c t

The tryptophan metabolism and immune activation play a role in pathophysiology of major depressivedisorders. The pro-inflammatory cytokine interferon-c transcriptionally induces the indoleamine 2,3-dioxygenase enzyme that degrades the tryptophan and thus induces serotonin depletion. The polymor-phism of certain cytokine genes was reported to be associated with major depression. We investigatedthe association between interferon-c (IFNc) gene CA repeat polymorphism, the profile of serotoninand tryptophan pathway metabolites and clinical parameters in 125 depressed patients and 93 healthycontrols. Compared to controls, serum tryptophan and 5-hydroxyindoleacetic acid (5HIAA) concentra-tions in the patients were significantly lower and serum kynurenine concentrations were significantlyhigher at baseline (p < 0.0001). The presence of IFNc CA repeat allele 2 homozygous has significant asso-ciation with higher kynurenine concentrations in controls (F = 4.47, p = 0.038) as well as in patients(F = 3.79, p = 0.045). The existence of interferon-c CA repeat allele 2 (homo- or heterozygous) showed sig-nificant association with increase of tryptophan breakdown over time during the study period (F = 6.0,p = 0.019). The results indicated the association between IFNc CA repeat allele 2, tryptophan metabolismand the effect of medication.

� 2013 Elsevier Inc. All rights reserved.

1. Introduction

The importance of tryptophan metabolism in the pathophysiol-ogy of major depressive disorders has been well documented. Sev-eral studies in depressed patients who responded to the selectiveserotonin reuptake inhibitors (SSRIs) and other antidepressantsshowed that acute tryptophan depletion could reverse the antide-pressant effects (Aberg-Wistedt et al., 1998; Bremner et al., 1997;Delgado et al., 1990, 1991, 1999; Smith et al., 1999; Spillmannet al., 2001).

Tryptophan, instead of being synthesized into 5-hydroxytophanand serotonin, could be degraded into kynurenine and could leadto tryptophan depletion and then to serotonin depletion. In addi-tion, the kynurenines which are neuroactive are proposed to beinvolved in the pathophysiological mechanism in depression (La-pin and Oxenkrug, 1969; Lapin, 1973). This depletion is the resultof enhanced tryptophan catabolism by the enzymes tryptophan2,3-dioxygenase (TDO) in the liver (Hayaishi, 1980) and the

ll rights reserved.

ychoneuroimmunology, Psy-baumstr. 7, D-80336 Munich,890..de (A.M. Myint).

indoleamine 2,3-dioxygenase (IDO) in the lungs, placenta, bloodand brain (Heyes et al., 1995; Mellor and Munn, 1999). TDO2immunoreactivity was demonstrated in the frontal cortex of theschizophrenic brain (Miller et al., 2004).

The TDO specifically metabolises tryptophan only (Hayaishi,1980) whereas IDO also metabolises serotonin and melatonin(Hayaishi, 1976). The activity of TDO is enhanced by the trypto-phan concentration (Saito et al., 1990; Satyanarayana and Rao,1980; Smith et al., 1980) and by high cortisol (Salter and Pogson,1985) whereas IDO activity is transcriptionally induced by thepro-inflammatory cytokines such as interferon-c (Carlin et al.,1987, 1989; Hu et al., 1995; Taylor and Feng, 1991; Yasui et al.,1986) and inhibited by anti-inflammatory cytokine, interleukin-4(IL4) (Musso et al., 1994).

Both increased pro-inflammatory cytokines (Anisman et al.,1999; Connor and Leonard, 1998; Kaestner et al., 2005; Kim etal., 2002; Lanquillon et al., 2000; Mikova et al., 2001; Myint etal., 2005; Thomas et al., 2005) and hypercortisolaemia (Asnis etal., 1981a,b; Cohen et al., 1984; Lin et al., 1986; Sher et al., 2005)have been well reported in patients with major depression. An in-crease in interferon-c/IL-4 ratio was observed in medication naivedepressed patients (Myint et al., 2005). Enhanced IDO activitywhich was indirectly indicated by an increased neopterin level

A.M. Myint et al. / Brain, Behavior, and Immunity 31 (2013) 128–133 129

was observed in cancer patients treated with interferon-a, a pro-inflammatory cytokine and subsequently developed depression(Capouron et al., 2002). Enhanced formation of kynurenine fromtryptophan, indicated by kynurenine/tryptophan ratio which indi-rectly indicates the IDO activity, was also observed in interferon-atreated patients who developed depression (Wichers et al., 2005)and in medication naïve patients with major depression comparedto normal controls (Myint et al., 2007). However, another study re-ported that tryptophan depletion in depressed patients occurswithout enhanced an kynurenine pathway (Hughes et al., 2012),although that conclusion was drawn without measuring concen-tration of quinolinic acid which is one of the crucial metabolitesfrom the downstream kynurenine metabolism. Taken all together,the high pro-inflammatory cytokines could induce serotonindepletion by (1) depletion of tryptophan, which is the precursorof serotonin through increased tryptophan breakdown due toenhanced IDO activity; and (2) direct breakdown of serotonin byIDO into anthraniloylalkylamine fragment (Sun, 1989) apart fromthe catabolic pathway through monoamineoxidase (MAO) activityinto the end metabolite 5-hydroxyindoleacetic acid (5HIAA), whichmight deplete serotonin and shunt the serotonin away from degra-dation into 5HIAA and result in a reduction of 5HIAA.

Some human and experimental studies had been performed onthe genes related to proinflammatory cytokines or markers andtheir relationship to depression (Jun et al., 2002, 2003a,b; Simenet al., 2006; Yu et al., 2003). The human study on IL1b (C-511T) ge-netic polymorphism showed that �511T allele of the IL1b gene hada trend of less severity of depressive symptoms and more favour-able fluoxetine therapeutic response than �511C carriers (Yu etal., 2003). An experimental study also showed that the deletionof tumour necrosis factor receptor (TNFR) 1 and 2 leads to an anti-depressant-like response in the forced swim test (Simen et al.,2006). Regarding the polymorphism of cytokine genes and IDOactivity, one study recently demonstrated that enhanced bloodIDO activity of Helicobactor pylori seropositive individuals interms of kynurenine/tryptophan ratio is related to TGFb1-509 al-lele T (Raitala et al., 2007). These findings indicated the possiblerelationship between certain cytokine genes, depression and IDOactivity which will affect the tryptophan breakdown.

Based on the above findings, we hypothesised that the trypto-phan metabolism, with emphasis on the balance between trypto-phan breakdown into kynurenine and serotonin synthesis fromtryptophan, may be related to genetic polymorphisms of certaincytokine genes. Since interferon-c (IFN-c) is the key cytokine thatenhances IDO activity, we focused on this cytokine. The CA repeatmicrosatellite box in the first intron of the IFN-c gene is well doc-umented to be associated with altered in vitro production of inter-feron-c by peripheral blood mononuclear cells of healthy subjects(Pravica et al., 1999). In addition, it was reported that the singlenucleotide polymorphism of IFN-c +874(T/A) is associated withkynurenine/tryptophan ratio in women (Raitala et al., 2005). Wetherefore placed emphasis on the association between biochemicalparameters and the IFN-c CA repeat polymorphism. In particular,we explored the relationship between tryptophan metabolism interms of tryptophan breakdown, indicated by kynurenine/trypto-phan ratio, and the balance between tryptophan breakdown andserotonin conversion from tryptophan, indicated by serotoninmetabolite 5HIAA/kynurenine ratio and interferon-c gene CArepeat polymorphism, in patients with major depression.

2. Materials and methods

2.1. Patients and controls

We investigated 125 unrelated in-patients (46 men, 79 women;age 49.75 ± 13.5 years) suffering from major depression (93.0%) or

depressed states of bipolar disorder (7.0%) according to DSM-IVcriteria. 30.3% of the patients had a first episode, 69.7% a recurrentepisode. Prior to the inclusion of the patient in the study, bloodsampling for a routine laboratory screening and a medical exami-nation were carried out to exclude severe medical disorders.Patients with any psychiatric comorbidity or any systemic chronicdiseases or acute diseases that can affect the immune system wereexcluded from the study. All patients and controls were Caucasiansand recruited in the Southern Germany/Bavaria region; 88% ofthem were of German descent. The mean sum score of the Hamil-ton Depression Rating Scale (17-item version) at admission was26.6 ± 6 and 6.7 ± 2.1 before discharge from hospital. The patientswere drug-free at baseline for at least 5 days prior to inclusion. Ma-jor medical disorders, addiction or other comorbid psychiatricdiagnoses, pregnancy, use of oral contraceptives or hormonereplacement therapy led to exclusion from the study. None of thepatients had been pre-treated with fluoxetine or depot neurolep-tics. Blood for neurochemical and genetic analyses was drawnexactly at 8:00 AM after overnight fasting to prevent an influenceof circadian fluctuations, and psychopathology was assessed.Neurochemical analyses and ratings were consecutively carriedout after 2 and 4 weeks and before discharge from hospital. Duringtheir time of admission patients received various pharmacologicaland non-pharmacological treatments, among them treatment withmirtazapine (18%), sertraline (12%) or reboxetine (13%). Forty fivepercent of the patients received a combination of one of thosemedications with benzodiazepine or lithium or valproate orlamotrigine. Five percent of the patients were treated withelectroconvulsive therapy and 7% were treated with othernon-pharmacological therapies.

A total of 93 healthy controls (35 men, 58 women) wererecruited. All controls were asked and investigated for past or pres-ent psychiatric and rheumatic illness through questionnairesincluding MMPI-2 testing and were medically examined for thepresence of infectious illness through standard laboratory testsincluding CRP and differential blood count. This was done withthe intention that an infectious status according to the actual lab-oratory parameters would be considered an exclusion criterion.Additionally, control individuals were asked for any history of psy-chosis in a first-degree relative, which was also considered anexclusion criterion.

The study was carried out according to the Declaration of Hel-sinki and local ethics committee approval. Patients were includedin the study after they gave written informed consent. A writteninformed consent was taken from each patient and control.

2.2. Analyses of genetic polymorphism

Genomic analysis was done on the blood samples taken at thetime of recruitment for both patients and controls. Genomic DNAwas isolated from whole blood according to standard procedures.

The CA-repeat polymorphism in intron 1 of the interferon-cgene was genotyped by fluorescence detection using the PISM310 automated sequencer (Applied Biosystems, Foster City, CA,USA). A DNA fragment containing the repeat sequence was ampli-fied with the following conditions: Forward primer: 50-6-FAM-gCTgTC ATA ATA ATA TTC AgA C-30; reverse primer: 30-CgA gCT TTAAAA gAT AgT TCC-50. PCR was performed with 50 ng of genomicDNA in a total volume of 25 ll containing 0.5 lM each primer,200 lM dNTPs, 2.5 ll 10 � PCR buffer and 0.5 Units Taq (AmpliTaq Gold; Perkin Elmer). PCR conditions were as follows: 95 �Cfor 5 min; 30 cycles of denaturation 95 �C/60 s; annealing 57 �C/45 s; elongation 72 �C/45 s; and final 72 �C/7 min. Prior to loadingonto the PRISM 310 10 ll formamide was added to 1 ll of the reac-tion mixture and samples were denaturated at 95 �C for 5 min.Reactions were electrophoresed on a 41 cm capillary array at

Fig. 1. Comparison of basal serum tryptophan levels between patients and controlsamong matched pairs and unmatched diagnostic groups. This figure shows thecomparison between 30 matched depressed patient-control pairs using student ‘t’test (t = 8.000, p < 0.0001) and the comparison between 125 depressed patients and93 healthy controls using multivariate analysis (F = 92.5, p < 0.0001).

130 A.M. Myint et al. / Brain, Behavior, and Immunity 31 (2013) 128–133

60 �C using Pop 4 polymer. Detection of the alleles (11–17 repeats;123–137 bp) and data analysis was processed by using GenscanAnalysis version 3.7 and Genotyper version 3.7 (AppliedBiosystems)

All laboratory procedures were carried out blind to case controlstatus.

2.3. Analyses of tryptophan metabolism

The serum tryptophan (lg/ml), kynurenine (ng/ml) and 5HIAA(ng/ml) concentrations were analysed in the sera from early morn-ing fasting samples using gradient high performance liquid chro-matography (HPLC) with UV and fluorescent detectors.

In brief, analytes were extracted from samples and calibrators/controls using Waters Oasis MCX 1 cc (30 mg) extraction cartridges(Waters, Milford, MS) as follows (all extractions were conductedwith a manual vacuum-manifold system): (1) the cartridge waspreconditioned by rinsing with 1 ml of methanol followed by1 ml of water; (2) 1 ml of sample and 100 ll 1 M H3PO4 wereapplied to the cartridge and pulled through under a light vacuum(2 min); (3) the cartridge was washed with 1 ml of 0.1 M HClfollowed by 1 ml 100% methanol; and finally, (4) the analytes wereeluted by rinsing the cartridge with 1.5 ml of acetonitrile contain-ing 6% NaOH. The eluent was then evaporated under nitrogen todryness and reconstituted with 150 ll 0.1 M PBS. The reconstitutedsample or calibrator or control was then transferred to a microin-jection vial (Waters). The calibrations were done for each metabo-lite (tryptophan or 5HIAA or kynurenine) with pure correspondingstandard (Sigma) of five concentration points.

Analyses were carried out on a Waters 2695 chromatographconnected to a Waters Model 2487 dual-k UV detector and a2475 fluorescence detector. For determination of kynurenine and5HIAA, 100 ll of the samples were loaded onto a 250 mm � 4 mmSupersphere 60 RP-select B, C8 column (Merck, Darmstadt, Ger-many). Due to the relatively higher concentration, a second injec-tion with a volume of 10 ll was performed for determination ofTRP. In order to ensure optimal peak resolution in the chromato-grams, and hence efficient separation of the analytes in a reason-ably short time (30 min), elution was carried out in the gradientmode using a mobile phase consisting of a mixture of 0.050 M so-dium acetate (solvent A: pH 4.80; solvent B: pH 3.65), acetonitrile(solvent C), and methanol (solvent D) at distinct proportions. Flowrate was set at 0.80 ml/min, column temperature was set at35.0 �C, while the samples were cooled at 4.0 �C. TRP was mea-sured by fluorescence detection (kex: 300 nm; kem: 350 nm),and kynurenine, at 365 nm and 5HIAA at 340 nm by UV detection.Approximate run time after injection until detection of thecompounds was about 20.4 min for TRP, 25 min for 5HIAA andfor kynurenine, 22.5 min. Data were processed using EMPOWERfor Windows 2000 software (Waters). The concentrations wereestablished through comparison of peak heights of the single ana-lytes with the peak heights of the respective calibration curves.

The tryptophan breakdown index was calculated by the ratiobetween serum kynurenine (ng/ml) and serum tryptophan (lg/ml). The serum 5HIAA (ng/ml)/serum kynurenine (ng/ml) ratio(5HIAA/Kyn) was also calculated to indicate the balance betweentryptophan breakdown into kynurenine and synthesis of trypto-phan into serotonin in the form of its stable metabolite, 5HIAA.

2.4. Statistical analyses

Univariate analysis was used to test the diagnostic group effecton the tryptophan and serotonin metabolites between 30 age- andgender-matched patient-control pairs (18 men, 12 women; age19–66 years) while BMI (patients vs controls �24.2 ± 3.2 vs23.7 ± 5) was considered as covariate. Multivariate regression anal-

ysis with age, gender and body mass index (BMI) as covariates wasapplied to analyse the effect of diagnostic group and genotype andgroup and genotype interaction on biochemical parameters. Theresults of regression analyses were shown in terms of effect size‘F’ and ‘p’ values. The v2 test was applied to find the associationbetween the genotype and the diagnostic group. The ANOVA (fornormally distributed data) with post hoc Benferroni and Friedmantest (for skewed data), followed by Wilcoxon signed rank test, wereapplied to test the change of each biochemical parameter duringthe course of antidepressant treatment. A repeated measure anal-ysis was applied to find the effect of genetic polymorphism onchanges of biochemical parameters during the course of antide-pressant treatment. Since only five patients were smokers and theywere excluded from data analyses. SPSS version 20 was used toperform data analyses.

3. Results

3.1. Tryptophan metabolites between patients and controls

The analyses were performed in exact age-and gender-matchedpairs (n = 30 in each group) and overall diagnostic groups (n = 112in depressed group, n = 87 in controls).

The basal serum concentrations of tryptophan in depressed pa-tients were significantly lower than those of normal controls inboth matched pairs (8.99 ± 0.31 vs 13.06 ± 0.40; F = 66.349,p < 0.0001) (Fig. 1) and among overall group comparison(8.31 ± 0.14 vs 13.51 ± 0.26; F = 147.91, p < 0.0001) when age, gen-der and BMI were controlled (Fig. 1).

The basal serum concentrations of 5HIAA in depressed patientswere significantly lower than those of normal controls in bothmatched pairs (5.43 ± 0.41 vs 12.62 ± 2.07; F = 11.586, p = 0.001)(Fig. 2) and among overall group comparison (7.45 ± 0.61 vs13.51 ± 1.52; F = 11.09, p < 0.0001) (Fig. 2).

The basal serum concentrations of kynurenine in depressed pa-tients were significantly higher than those of normal controls inboth matched pairs (373.17 ± 16.09 vs 266.63 ± 16.38; F = 21.748,p < 0.0001) (Fig. 3) and among overall group comparison(381.54 ± 9.03 vs 265.58 ± 9.92; F = 31.65, p < 0.0001) (Fig. 3).

Fig. 2. Comparison of basal serum 5HIAA levels between patients and controlsamong matched pairs and unmatched diagnostic groups. This figure shows thecomparison between 30 matched depressed patient-control pairs using student ‘t’test (t = 3.40, p < 0.002) and the comparison between 125 depressed patients and 93healthy controls using multivariate analysis (F = 3.4, p = 0.065).

A.M. Myint et al. / Brain, Behavior, and Immunity 31 (2013) 128–133 131

The basal tryptophan breakdown indices (kynurenine/trypto-phan) in depressed patients were significantly higher than thoseof healthy controls in both match pairs (42.01 ± 1.55 vs20.49 ± 1.17; F = 122.482, p < 0.0001) and among overall groupcomparison (46.47 ± 1.09 vs 19.86 ± 0.72; F = 142.183, p < 0.0001).

The basal 5HIAA/kynurenine ratios of the depressed patientswere significantly lower than those of healthy controls in both30 matched pairs (0.015 ± 0.011 vs 0.055 ± 0.010; F = 14.006,p < 0.0001) and among overall group comparison (0.02 ± 0.001 vs0.057 ± 0.005; F = 16.516, p < 0.0001).

Fig. 3. Comparison of basal serum kynurenine levels between patients and controlsamong matched pairs and unmatched diagnostic groups. This figure shows thecomparison between 30 matched depressed patient-control pairs using student ‘t’test (t = 11.10, p < 0.0001) and the comparison between 125 depressed patients and93 healthy controls using multivariate analysis (F = 19.7, p < 0.0001).

3.2. Effect of gene polymorphism on tryptophan metabolites

In the overall depressed group, 31 patients (27.6%) carriedinterferon-c CA repeat allele 2 homozygous, 47 patients (42%), car-ried interferon-c CA repeat allele 2 heterozygous, and 34 patients(30.4%) carried other allele. In overall controls, 18 (20.7%) carriedCA repeat allele 2 homozygous, 43 (49.4%) carried heterozygous,and 26 (29.9%) carried other allele. In age- and gender-matchedpairs, there are no control subjects who carried allele 2 heterozy-gous. Therefore, paired group analyses were not done for the effectof gene polymorphism on metabolites.

In overall patients and healthy controls combined group, whenage, gender and BMI were controlled in regression analysis, thepresence of interferon-c CA repeat allele 2 homozygous is associ-ated with higher serum kynurenine concentrations(311.02 ± 12.01 vs 335.56 ± 11.78; N = 49, 150; F = 4.07, p = 0.042).

In regression analyses, the interaction between the above geno-types and presence of depression showed significant associationwith lower concentrations of tryptophan (F = 6.5, p = 0.012), and5HIAA (F = 4.4, p = 0.038), and higher concentration of kynurenine(F = 3.7, p = 0.057) at baseline. These indicate that the presence ofinterferon-c CA repeat allele 2 homozygous and the presence ofdepression is associated with lower serum tryptophan, lower5HIAA, and higher kynurenine.

In regression analyses, after controlling age, gender and BMI inthe controls group, the presence of interferon-c CA repeat allele 2homozygous significantly associated with lower baseline serumtryptophan and higher kynurenine concentrations (F = 5.57,p = 0.021 and F = 4.47, p = 0.038).

In regression analyses in the patients group, the presence ofinterferon-c CA repeat allele 2 homozygous showed significantassociation with higher serum kynurenine concentrations(F = 3.79, p = 0.045). In repeated measure analyses, the existenceof interferon-c CA repeat allele 2 (either homo or heterozygous)is associated with an increase of tryptophan breakdown through-out the study period in the patients (F = 6.0, p = 0.019).

3.3. Tryptophan breakdown and response to treatment

In repeated measure analysis, the tryptophan breakdown indexincreased significantly during the course of antidepressant treat-ment (F = 9.7, p < 0.0001). That was mainly due to the differencebetween the basal level and (1) the time point just before end oftreatment (47.34 ± 12.01 vs 51.75 ± 12.26; N = 112, 108; t = 4.104,p < 0.0001), and (2) at the end of antidepressant treatment(47.34 ± 12.01 vs 56.67 ± 12.56; N = 112, 108, t = 5.739,p < 0.0001). The serum 5HIAA concentrations also showed a trendof increase during the course of treatment (v2 = 15.513, p = 0.001)which was mainly due to the difference between the basal 5HIAAconcentration and 5HIAA just before end of treatment(7.46 ± 3.45 vs 10.05 ± 7.32; N = 112, 108; Z = 2.141, p = 0.032).The responders who showed a reduction in HAMD score of 50%or more showed a significantly smaller increase in tryptophanbreakdown at the time of discharge from baseline (11.87 ± 4.50vs 34.36 ± 7.48; N = 53, 55; t = 2.649, p < 0.011).

4. Discussion

Our findings indicate lower serum concentrations of tryptophanand 5HIAA and higher serum kynurenine in depressed patientscompared to healthy controls. Due to the results of thosemetabolite changes, we also demonstrate the higher tryptophanbreakdown and lower 5HIAA/kynurenine ratio in depressed pa-tients, which indicate the enhanced tryptophan breakdown intokynurenine and reduction of the synthesis of serotonin which

132 A.M. Myint et al. / Brain, Behavior, and Immunity 31 (2013) 128–133

would be degraded into 5HIAA. These findings indirectly indicatedthe enhanced IDO activity. These findings also support the previousobservation in which the cerebrospinal fluid 5HIAA of a subgroupof depressed patients was lower than that of normal control indi-viduals (Traskman-Bendz et al., 1984).

The higher tryptophan breakdown indices in depressed patientscould be due to the enhanced activities of both IDO and TDO oreither of them. The enhanced IDO activity is associated with in-creased secretion of pro-inflammatory cytokine interferon-c. Itwas reported that a six-week antidepressant treatment could sig-nificantly reduce the level of interferon-c (Myint et al., 2005).Therefore, when the patients were treated with antidepressantsand had recovered from clinical depression, it was expected thatthe tryptophan breakdown index, which is the ratio betweenkynurenine/tryptophan, would have decreased. However, we ob-served that the tryptophan breakdown index increased withoutreducing the tryptophan level. This is also in agreement with oneprevious study in which the mean tryptophan breakdown indexshowed a trend to increase after 6 weeks of antidepressant treat-ment in Korean depressed patients (Myint et al., 2007).

This raises the question of whether the increase of kynurenine/tryptophan ratio is due to increased tryptophan degradation be-cause of further enhanced IDO activity or due to the relative in-crease in kynurenine concentration because of a reduction infurther downstream degradation. One could hypothesise that thisincrease in kynurenine/tryptophan ratio after treatment is due toa reduction of further downstream metabolism since the increaseof kynurenine occurs without a reduction in tryptophan. In thetryptophan breakdown pathway, the proinflammatory cytokinesenhance not only the activity of IDO but also the activity ofkynurenine-3-monooxygenase (KMO) and increase the furtherdownstream metabolism (Hayaishi, 1976, 1980; Taylor and Feng,1991). Depression is shown to be associated with increased pro-inflammatory cytokines, and antidepressant treatment was dem-onstrated to have an anti-inflammatory effect. Therefore, beforethe treatment, because of the high pro-inflammatory activity, bothIDO and KMO activities would be enhanced. After the antidepres-sant treatment, the enhanced downstream metabolism might becorrected. This could lead to a relative increase of kynurenine inthe serum, although the actual formation of kynurenine did notincrease. This might result in the increased tryptophan breakdownindex which was calculated from the kynurenine/tryptophan ratio.Therefore, the tryptophan breakdown index would increase with-out associated decrease in tryptophan concentration and thatwould induce neither tryptophan depletion nor serotonindepletion.

In addition we observed the trend of increase in 5HIAA aftertreatment. IDO enzyme degrades not only tryptophan but alsoserotonin and its direct precursor 5-hydroxytryptophan (Hayaishi,1976). When IDO activity enhancement was corrected after theantidepressant treatment, more serotonin is available to be metab-olised by MAO into 5HIAA, and therefore resulted in an increasedlevel of 5HIAA.

The reasons mentioned above could also explain the fact that,after the antidepressant treatment, both 5HIAA and kynurenineincreased, resulting in no significant change of the 5HIAA/kynuren-ine ratio.

In addition, we found significant association between thetryptophan breakdown index and response to treatment. This find-ing indicates the involvement of tryptophan metabolism in thepathophysiology of major depression.

Regarding the association between biochemical changes andgenotyping, we concentrated on the existence of the CA repeatallele 2, since this allele was reported to be associated with highin vitro production of interferon-c (Pravica et al., 1999) and

tryptophan degradation in women (Raitala et al., 2005). The allele2 homozygosity was observed to have effects on the differences inbasal serum tryptophan, kynurenine and 5HIAA between the de-pressed patients and controls. The significant association of inter-feron-c CA repeat allele 2 homozygosity with lower serumtryptophan and higher kynurenine concentrations was alsoobserved when regression analyses were performed in the groupof only controls, which indirectly indicate the functional polymor-phism of interferon-c. This effect was still observed on the serumbasal kynurenine levels when the analyses were performed inthe group with only depressed patients. In the depressed group,both homozygosity and heterozygosity of interferon-c CA repeatallele 2 showed significant effects on the change of the tryptophanbreakdown index over time during the course of treatment. Thesefindings indirectly indicate the involvement of interferon-c CArepeat allele 2 in biological changes in depression. All our findingsare not restricted to any gender group although Raitala et al., 2005has reported the restriction of the effect of interferon-c genepolymorphism to the female group.

Our findings are in agreement with the proposal of Oxenkrugand group (Oxenkrug et al., 2011) who studied the associationbetween IFNc (+874) T/A genotypes and the risk of developmentof depression in IFNa-treated patients. They have demonstratedthat carriers of one T allele were more frequent in a depressedIFNa-treated patient group than in a non-depressed IFNa-treatedpatient group. They have suggested that high producer T allelewould be the cause of higher tryptophan breakdown and thatmight be associated with development of depression in IFNa-trea-ted patients. Their finding was also not restricted only to thewomen.

Our study has some limitations. Though the total sample size ofoverall groups was not small, the groups were not perfectlymatched with respect to age and gender and BMI. However, age,gender and BMI were controlled in regression analyses. Whenanalyses were carried out only in age- and gender-matched pairs,the sample size became too small to find a risk polymorphismand there is no carrier of CA repeat allele 2 heterozygous in con-trols. Another limitation of this study is that, although the majorityof the subjects were German, there were some non-German Cauca-sians and people of Turkish origin in the patient population. How-ever, the ethnic group showed no effect on biochemical parametersin statistical analyses. Since tryptophan is an essential amino acid,it could be argued that differences in the dietary patterns of the pa-tients and control persons might have resulted in different trypto-phan levels (Smith et al., 1980). However, if low tryptophan intakeis the cause of lower tryptophan levels in patients, the kynureninelevels should not be higher, as we observed in this study; we there-fore conclude that, in our study, possible differences in individualdietary intake had no effect on the determined biochemical mea-sures. Another point is that we could not analyse the other metab-olites from the tryptophan breakdown pathway, especiallykynurenic acid and 3-hydroxykynurenine, to show the effect ofinterferon-c on downstream metabolism in depressed group be-cause of lack of availability of serum sample volume. Nevertheless,we could clearly demonstrate the role of tryptophan degradationto kynurenine in depression and its association with interferon-cCA repeat allele polymorphism.

Acknowledgements

The work of the first and the last authors was partially fundedby EU FP7 collaborative project: MOODINFLAME (Grant No.22963). The authors thank Mr. Jim Burgess for language editionas a native speaker.

A.M. Myint et al. / Brain, Behavior, and Immunity 31 (2013) 128–133 133

References

Aberg-Wistedt, A., Hasselmark, L., Stain-Malmgren, R., Aperia, B., Kjellman, B.F.,Mathe, A.A., 1998. Serotonergic ‘vulnerability’ in affective disorder: a study ofthe tryptophan depletion test and relationships between peripheral and centralserotonin indexes in citalopram-responders. Acta Psychiatr. Scand. 97, 374–380.

Anisman, H., Ravindran, A.V., Griffiths, J., Merali, Z., 1999. Endocrine and cytokinecorrelates of major depression and dysthymia with typical or atypical features.Mol. Psychiatry 4, 182–188.

Asnis, G.M., Sachar, E.J., Halbreich, U., Nathan, R.S., Novacenko, H., Ostrow, L.C.,1981a. Cortisol secretion in relation to age in major depression. Psychosom.Med. 43, 235–242.

Asnis, G.M., Sachar, E.J., Halbreich, U., Nathan, R.S., Ostrow, L., Halpern, F.S., 1981b.Cortisol secretion and dexamethasone response in depression. Am. J. Psychiatry138, 1218–1221.

Bremner, J.D., Innis, R.B., Salomon, R.M., Staib, L.H., Ng, C.K., Miller, H.L., et al., 1997.Positron emission tomography measurement of cerebral metabolic correlates oftryptophan depletion-induced depressive relapse. Arch. Gen. Psychiatry 54,364–374.

Capouron, L., Ravaud, A., Neveu, P.J., Miller, A.H., Maes, M., Dantzer, R., 2002.Association between decreased serum tryptophan concentrations anddepressive symptoms in cancer patients undergoing cytokine therapy. Mol.Psychiatry 7, 468–473.

Carlin, J.M., Borden, E.C., Sondel, P.M., Byrne, G.I., 1987. Biologic-response-modifier-induced indoleamine 2,3-dioxygenase activity in human peripheral bloodmononuclear cell cultures. J. Immunol. 139, 2414–2418.

Carlin, J.M., Borden, E.C., Sondel, P.M., Byrne, G.I., 1989. Interferon-inducedindoleamine 2,3-dioxygenase activity in human mononuclear phagocytes. J.Leukoc. Biol. 45, 29–34.

Cohen, M.R., Pickar, D., Extein, I., Gold, M.S., Sweeney, D.R., 1984. Plasma cortisoland beta-endorphin immunoreactivity in nonmajor and major depression. Am.J. Psychiatry 141, 628–632.

Connor, T.J., Leonard, B.E., 1998. Depression, stress and immunological activation:the role of cytokines in depressive disorders. Life Sci. 62, 583–606.

Delgado, P.L., Charney, D.S., Price, L.H., Aghajanian, G.K., Landis, H., Heninger, G.R.,1990. Serotonin function and the mechanism of antidepressant action. Reversalof antidepressant-induced remission by rapid depletion of plasma tryptophan.Arch. Gen. Psychiatry 47, 411–418.

Delgado, P.L., Price, L.H., Miller, H.L., Salomon, R.M., Licinio, J., Krystal, J.H., et al.,1991. Rapid serotonin depletion as a provocative challenge test for patientswith major depression: relevance to antidepressant action and theneurobiology of depression. Psychopharmacol. Bull. 27, 321–330.

Delgado, P.L., Miller, H.L., Salomon, R.M., Licinio, J., Krystal, J.H., Moreno, F.A., et al.,1999. Tryptophan-depletion challenge in depressed patients treated withdesipramine or fluoxetine: implications for the role of serotonin in themechanism of antidepressant action. Biol. Psychiatry 46, 212–220.

Hayaishi, O., 1976. Properties and function of indoleamine 2,3-dioxygenase. J.Biochem. 79, 13P–21P (Tokyo).

Hayaishi, O., 1980. Biochemical and Medical Aspect of Tryptophan Metabolism.Elsevier/North-Holland Biomedical, Amsterdam.

Heyes, M.P., Saito, K., Milstien, S., Schiff, S.J., 1995. Quinolinic acid in tumors,hemorrhage and bacterial infections of the central nervous system in children. J.Neurol. Sci. 133, 112–118.

Hu, B., Hissong, B.D., Carlin, J.M., 1995. Interleukin-1 enhances indoleamine 2,3-dioxygenase activity by increasing specific mRNA expression in humanmononuclear phagocytes. J. Interferon Cytokine Res. 15, 617–624.

Hughes, M.M., Carballedo, A., McLoughlin, D.M., Amico, F., Harkin, A., Frodl, T.,Connor, T.J., 2012. Tryptophan depletion in depressed patients occursindependent of kynurenine pathway activation. Brain Behav. Immun. 26 (6),979–987.

Jun, T.Y., Pae, C.U., Chae, J.H., Bahk, W.M., Kim, K.S., Han, H., 2002. Report on IL-10gene polymorphism at position �819 for major depression and schizophrenia inKorean population. Psychiatry Clin. Neurosci. 56, 177–180.

Jun, T.Y., Pae, C.U., Chae, J.H., Bahk, W.M., Kim, K.S., Pyo, C.W., et al., 2003a. Tumornecrosis factor-beta gene polymorphism may not be associated with majordepressive disorder in the Korean population. Psychiatry Clin. Neurosci. 57, 31–35.

Jun, T.Y., Pae, C.U., Hoon, H., Chae, J.H., Bahk, W.M., Kim, K.S., et al., 2003b. Possibleassociation between �G308A tumour necrosis factor-alpha gene polymorphismand major depressive disorder in the Korean population. Psychiatr. Genet. 13,179–181.

Kaestner, F., Hettich, M., Peters, M., Sibrowski, W., Hetzel, G., Ponath, G., et al., 2005.Different activation patterns of proinflammatory cytokines in melancholic andnon-melancholic major depression are associated with HPA axis activity. J.Affect. Disord. 87, 305–311.

Kim, Y.K., Suh, I.B., Kim, H., Han, C.S., Lim, C.S., Choi, S.H., et al., 2002. The plasmalevels of interleukin-12 in schizophrenia, major depression, and bipolar mania:effects of psychotropic drugs. Mol. Psychiatry 7, 1107–1114.

Lanquillon, S., Krieg, J.C., Ing-Abu-Shach, U., Vedder, H., 2000. Cytokine productionand treatment response in major depressive disorder.Neuropsychopharmacology 22, 370–379.

Lapin, I.P., 1973. Kynurenines as possible participants of depression.Pharmapsychiatry. Neuropsychopharmakol. 6, 273–279.

Lapin, I.P., Oxenkrug, G.F., 1969. Intensification of the central serotoninergicprocesses as a possible determinant of the thymoleptic effect. Lancet 1(7586), 132–136.

Lin, S.C., Maruta, T., Newman, D.C., Kao, P.C., 1986. Plasma levels of cortisol,corticotropin, and beta-endorphin in patients with major depression. J. Clin.Psychiatry 47, 413–414.

Mellor, A.L., Munn, D.H., 1999. Tryptophan catabolism and T-cell tolerance.immunosuppression by starvation? Immunol. Today 20, 469–473.

Mikova, O., Yakimova, R., Bosmans, E., Kenis, G., Maes, M., 2001. Increased serumtumor necrosis factor alpha concentrations in major depression and multiplesclerosis. Eur. Neuropsychopharmacol. 11, 203–208.

Miller, C.L., Llenos, I.C., Dulay, J.R., Barillo, M.M., Yolken, R.H., Weis, S., 2004.Expression of the kynurenine pathway enzyme tryptophan 2,3-dioxygenase isincreased in the frontal cortex of individuals with schizophrenia. Neurobiol. Dis.15, 618–629.

Musso, T., Gusella, G.L., Brooks, A., Longo, D.L., Varesio, L., 1994. Interleukin-4inhibits indoleamine 2,3-dioxygenase expression in human monocytes. Blood83, 1408–1411.

Myint, A.M., Leonard, B.E., Steinbusch, H.W., Kim, Y.K., 2005. Th1, Th2, and Th3cytokine alterations in major depression. J. Affect. Disord. 88, 167–173.

Myint, A.M., Kim, Y.K., Verkerk, R., Scharpe, S., Steinbusch, H., Leonard, B., 2007.Kynurenine pathway in major depression: evidence of impairedneuroprotection. J. Affect. Disord. 98, 143–151.

Oxenkrug, G., Perianayagam, M., Mikolich, D., Requintina, P., Shick, L., Ruthazer, R.,Zucker, D., Summergrad, P., 2011. Interferon-gamma (+874) T/A genotypes andrisk of IFN-alpha induced depression. J. Neural Transm. 118 (2), 271–274.

Pravica, V., Asderakis, A., Perrey, C., Hajeer, A., Sinnott, P.J., Hutchinson, I.V., 1999. Invitro production of IFN-gamma correlates with CA repeat polymorphism in thehuman IFN-gamma gene. Eur. J. Immunogenet. 26, 1–3.

Raitala, A., Pertovaara, M., Karjalainen, J., Oja, S.S., Hurme, M., 2005. Association ofinterferon-gamma +874(T/A) single nucleotide polymorphism with the rate oftryptophan catabolism in healthy individuals. Scand. J. Immunol. 61, 387–390.

Raitala, A., Karjalainen, J., Oja, S.S., Kosunen, T.U., Hurme, M., 2007. Helicobacterpylori-induced indoleamine 2,3-dioxygenase activity in vivo is regulated byTGFB1 and CTLA4 polymorphisms. Mol. Immunol. 44, 1011–1014.

Saito, K., Ohta, Y., Nagamura, Y., Sasaki, E., Ishiguro, I., 1990. Relationship between L-tryptophan uptake and L-tryptophan 2,3-dioxygenase activity in rathepatocytes. Biochem. Int. 20, 71–80.

Salter, M., Pogson, C.I., 1985. The role of tryptophan 2,3-dioxygenase in thehormonal control of tryptophan metabolism in isolated rat liver cells. Effects ofglucocorticoids and experimental diabetes. Biochem. J. 229, 499–504.

Satyanarayana, U., Rao, B.S., 1980. Dietary tryptophan level and the enzymes oftryptophan NAD pathway. Br. J. Nutr. 43, 107–113.

Sher, L., Oquendo, M.A., Galfalvy, H.C., Zalsman, G., Cooper, T.B., Mann, J.J., 2005.Higher cortisol levels in spring and fall in patients with major depression. Prog.Neuropsychopharmacol. Biol. Psychiatry 29, 529–534.

Simen, B.B., Duman, C.H., Simen, A.A., Duman, R.S., 2006. TNFalpha signaling indepression and anxiety: behavioral consequences of individual receptortargeting. Biol. Psychiatry 59, 775–785.

Smith, S.A., Carr, F.P., Pogson, C.I., 1980. The metabolism of L-tryptophan by isolatedrat liver cells. Quantification of the relative importance of, and the effect ofnutritional status on, the individual pathways of tryptophan metabolism.Biochem. J. 192, 673–686.

Smith, K.A., Morris, J.S., Friston, K.J., Cowen, P.J., Dolan, R.J., 1999. Brain mechanismsassociated with depressive relapse and associated cognitive impairmentfollowing acute tryptophan depletion. Br. J. Psychiatry 174, 525–529.

Spillmann, M.K., Van der Does, A.J., Rankin, M.A., Vuolo, R.D., Alpert, J.E., Nierenberg,A.A., et al., 2001. Tryptophan depletion in SSRI-recovered depressedoutpatients. Psychopharmacology 155, 123–127.

Sun, Y., 1989. Indoleamine 2,3-dioxygenase–a new antioxidant enzyme. Mater.Med. Pol. 21, 244–250.

Taylor, M.W., Feng, G.S., 1991. Relationship between interferon-gamma,indoleamine 2,3-dioxygenase, and tryptophan catabolism. FASEB J. 5, 2516–2522.

Thomas, A.J., Davis, S., Morris, C., Jackson, E., Harrison, R., O’Brien, J.T., 2005. Increasein interleukin-1beta in late-life depression. Am. J. Psychiatry 162, 175–177.

Traskman-Bendz, L., Asberg, M., Bertilsson, L., Thoren, P., 1984. CSF monoaminemetabolites of depressed patients during illness and after recovery. ActaPsychiatr. Scand. 69, 333–342.

Wichers, M.C., Koek, G.H., Robaeys, G., Verkerk, R., Scharpe, S., Maes, M., 2005. IDOand interferon-alpha-induced depressive symptoms: a shift in hypothesis fromtryptophan depletion to neurotoxicity. Mol. Psychiatry 10 (6), 538–544.

Yasui, H., Takai, K., Yoshida, R., Hayaishi, O., 1986. Interferon enhances tryptophanmetabolism by inducing pulmonary indoleamine 2,3-dioxygenase: its possibleoccurrence in cancer patients. Proc. Natl. Acad. Sci. USA 83, 6622–6626.

Yu, Y.W., Chen, T.J., Hong, C.J., Chen, H.M., Tsai, S.J., 2003. Association study of theinterleukin-1 beta (C-511T) genetic polymorphism with major depressivedisorder, associated symptomatology, and antidepressant response.Neuropsychopharmacology 28, 1182–1185.