13C-tryptophan breath test detects increased catabolic turnover of tryptophan along the kynurenine pathway in patients with major depressive disorder

Altered tryptophan–kynurenine (KYN) metabolism has been implicated in major depressive disorder (MDD). The l-[1-13C]tryptophan breath test (13C-TBT) is a noninvasive, stable-isotope tracer method in which exhaled 13CO2 is attributable to tryptophan catabolism via the KYN pathway. We included 18 patients with MDD (DSM-IV) and 24 age- and sex-matched controls. 13C-tryptophan (150 mg) was orally administered and the 13CO2/12CO2 ratio in the breath was monitored for 180 min. The cumulative recovery rate during the 180-min test (CRR0–180; %), area under the Δ13CO2-time curve (AUC; %*min), and the maximal Δ13CO2 (Cmax; %) were significantly higher in patients with MDD than in the controls (p = 0.004, p = 0.008, and p = 0.002, respectively). Plasma tryptophan concentrations correlated negatively with Cmax in both the patients and controls (p = 0.020 and p = 0.034, respectively). Our results suggest that the 13C-TBT could be a novel biomarker for detecting a subgroup of MDD with increased tryptophan–KYN metabolism.

. Schematic view of 13 C-tryptophan metabolism . After utilization for protein synthesis, over 90% of dietary tryptophan is converted to KYN by one of two rate-limiting enzymes: by TDO in the liver under healthy conditions and by IDO in various human tissues (e.g., intestines, lungs, placenta, kidneys, spleen, blood, and brain) under inflammatory conditions. The 13 CO2 derived from 13 C-tryptophan is produced by the conversion of 3-OH-KYN into 3HAA via the KYN-NAD pathway (box with solid line) as the post-KYN metabolism. In addition, although 13 CO2 is produced to a minor degree in the tryptophan-serotonin and anthranilic acid pathways, it is not produced by the KA (box with dotted line) or xanthurenic acid pathways. Bold arrows indicate quantitatively major pathways. Minor pathways and pathways without release of 13 CO2 are omitted. Red circles indicate the 13 C-labeled carbon. Abbreviations: IDO, indoleamine 2,3-dioxygenase; KA, kynurenic acid; KYN, kynurenine; NAD, nicotinamide adenine dinucleotide; TDO, tryptophan 2,3-dioxygenase; 3HAA, 3-hydroxyanthranilic acid; 3-OH-KYN, 3-hydroxy-l-kynurenine. blood-brain barrier (BBB) [33][34][35] , fluctuations in peripheral blood levels can directly affect metabolism of KYN pathway metabolites, including the synthesis of KA, 3-hydroxyanthranilic acid (3HAA), and QUIN in the brain, although these three metabolites do not readily cross the BBB 36,37 . IDO is distributed in various human tissues, including the intestine, lungs, placenta, kidneys, spleen, brain, and blood 8 , and when peripherally activated, IDO can affect the concentration of QUIN in the brain and CSF 38 . Raison et al. reported that there were significant positive correlations between peripheral and central measures of KYN and QUIN, but not tryptophan, and that peripheral tryptophan decreased while CSF tryptophan remained unchanged in patients receiving IFN-α therapy 16 . These factors point to the merit of monitoring KYN pathway activation in the whole body.
Tracer tests are advantageous for assessing in vivo metabolism of biological compounds as a more direct approach. Using a 13 C-phenylalanine breath test, our group reported that altered phenylalanine metabolism could be identified in patients with schizophrenia 39 . Hankes et al. administered a 14 C-tryptophan breath test to patients with scleroderma and reported that expiratory 14 CO 2 was decreased and labeled QUIN in the urine was increased 40,41 . Therefore, they suggested that metabolism after 3HAA may be altered in these patients. They also reported that the production of 14 CO 2 from 14 C-labeled d-tryptophan was different from that of 14 C-labeled l-tryptophan, and suggested that labeled l-tryptophan should be used for studies of tryptophan metabolism. In the present study, we used a stable isotope (i.e., non-radioactive tracer) in an l-[1-13 C]tryptophan breath test ( 13 C-TBT) to examine the whole-body kinetics and metabolism of orally administered tryptophan. In this method, the 13 C label of l-[1-13 C] tryptophan is released as 13 CO 2 when 3HAA, an immediate precursor for QUIN, is produced from tryptophan via the KYN-NAD pathway both peripherally and in the brain. This method also selectively observes the KYN-NAD pathway, but not the KYN-KA pathway, because 13 CO 2 is not produced during the synthesis of KA or xanthurenic acid. Therefore, 13 CO 2 formation measured by the 13 C-TBT is an index of whole-body tryptophan metabolism via the KYN pathway and excludes the pathways for KA and xanthurenic acid production.
The aim of this study was to assess KYN pathway activity in patients with MDD. We hypothesized that the 13 C-TBT would detect altered tryptophan metabolism via the KYN pathway in patients with MDD, and that the 13 C-TBT indices would correlate negatively with plasma tryptophan concentrations.

Participants.
We enrolled 18 patients with MDD (age range: 25-62 years; 10 women) and 24 healthy volunteers (age range: 21-61 years; 12 women), matched for age and sex, and who were biologically unrelated. Participants were recruited through advertisements in free local information magazines, by word of mouth, and via a website announcement; all were Japanese residents from the western part of Tokyo. Patients were also recruited from the National Center of Neurology and Psychiatry (NCNP) Hospital, Tokyo, Japan.
Consensus diagnoses were made for MDD by at least two experienced psychiatrists according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) 42 , based on the Japanese version of the Mini-International Neuropsychiatric Interview (MINI) 43 , careful examination of medical records, observations, and additional unstructured interviews.
Healthy volunteers were screened by psychiatrists using unstructured interviews and the MINI. We confirmed that they had no history of psychiatric illness or contact with psychiatric services. Participants were excluded from both the patient and control groups if they met the DSM-IV criteria for intellectual disability, had a history of substance dependence or abuse, or had a history of severe head injury. We also excluded participants if they had a current serious physical disorder, particularly an inflammatory or liver disease (where tryptophan is primarily metabolized) or if they received any immunotherapy. All participants underwent blood tests for aspartate aminotransferase (AST), alanine aminotransferase (ALT), total protein, albumin, total bilirubin, blood urea nitrogen (BUN), and creatinine levels to exclude liver and kidney dysfunction. Individuals with abnormal blood test results were excluded.
The present study was conducted in accordance with the principles of the Declaration of Helsinki. The capacity of patients to understand the nature and purpose of the study and to provide consent was confirmed by an experienced psychiatrist. After explaining the study aim and protocol, signed informed consent was obtained from each participant. The study was approved by the ethics committee of the NCNP, Japan.
Principle of the 13 C-TBT. The kinetic values for l-[1-13 C]tryptophan were assumed to be the same as for unlabeled tryptophan 44 . The following theory underpinned the 13 C-TBT. After protein synthesis, dietary tryptophan, an essential amino acid for humans, is primarily metabolized along the KYN pathway in the liver, which accounts for over 90% of tryptophan catabolism 35 ; in contrast, only 1%-3% of dietary tryptophan is metabolized via the serotonin pathway 34 . Under healthy conditions, the conversion of tryptophan to KYN via N-formyl-l-kynurenine is initiated by TDO, which is activated by stress and high glucocorticoid levels 1,2,45 . Thus, this is the product of elevated hypothalamic pituitary-adrenal (HPA) axis activity, which is often implicated in the development of depressive disorders 46,47 . In conditions where the immune system is stimulated, IDO is induced by proinflammatory cytokines 2,3,7,48 . Crucial to the test is that the 13 CO 2 derived from orally administered 13 C-tryptophan can be produced by the conversion of 3-OH-KYN into 3HAA via the KYN-NAD pathway, but not by the KA or xanthurenic acid pathways (Fig. 1). The 13 CO 2 is also produced to much lesser degrees by minor pathways, such as the tryptophanserotonin and anthranilic acid pathways. The 13 CO 2 measured by the 13 C-TBT was, therefore, expected to reflect whole-body tryptophan-KYN-3HAA metabolism initiated by IDO or TDO.

Clinical status and antidepressant medication. Current depression severity in patients was
assessed using a 21-item version of the Hamilton Depression Rating Scale (HAMD-21) administered by a single experienced psychiatrist. Antidepressant doses were converted to imipramine-equivalent (IMIeq) doses according to published guidelines 51,52 . Statistical analysis. The Statistical Package for the Social Sciences (SPSS) version 11.0 (SPSS Japan, Tokyo, Japan) was used to conduct all statistical tests. Averages were expressed as mean ± standard deviation (SD). Statistical differences in the demographic data between groups were compared using the χ 2 test and t-test for categorical variables and continuous variables, respectively. We then performed stepwise multiple regression analyses with two models, using the 13 C-TBT indices (C max , AUC, and CRR 0-180 ) as the dependent variable: in model 1, diagnosis, sex, age, and body weight were entered as possible predictors; and in model 2, diagnosis, sex, age, body weight, and plasma tryptophan concentration were entered as possible predictors.
A 2 × 10 (group × sampling point) repeated measures analysis of covariance (ANCOVA) was conducted to examine group differences in Δ 13 CO 2 or CRR over time, controlling for sex, age, and body weight. Greenhouse-Geisser corrections were used for lack of sphericity. We used ANCOVA (controlling for sex, age, and body weight) and t-tests to compare 13 C-TBT indices between the two groups. The T max , a discrete variable, was compared between the two groups using the Mann-Whitney U test. The correlations of each 13 C-TBT index (AUC, CRR 0-180 , and C max ) with the liver parameters and the plasma tryptophan concentrations were examined by Spearman's and Pearson's correlation analyses, respectively. Cramer's V, Cohen's d, η 2 , and partial η 2 were calculated as measures of effect size for χ 2 test, t-tests, ANOVAs, and ANCOVAs, respectively. In the patients, the correlations of each 13 C-TBT index (AUC, CRR 0-180 , and C max ) with IMIeq and HAMD-21 total score were examined using Spearman's correlation analysis. A stepwise multiple regression was employed to assess independent predictors of C max , AUC, and CRR 0-180 , using sex, IMIeq, HAMD-21 total score, and plasma tryptophan concentration as candidate predictors. Statistical significance was two-tailed and set at p < 0.05.

Results
Demographic and clinical characteristics. The demographic and clinical characteristics of the patients and controls are presented in Table 1. There was no significant difference in sex, age, body weight, body mass index, education length, liver or kidney blood test results, or plasma tryptophan concentrations between the two groups.

Predictors for 13 C-TBT indices.
Stepwise regression analyses were performed (Table 2). In model 1, diagnosis and sex were identified as significant predictors for C max , AUC, and CRR 0-180 . In model 2, diagnosis and sex remained significant predictors for C max , AUC, and CRR 0-180 , but the plasma tryptophan concentration was not a significant predictor in the model. Figure 2 shows the time course of mean values of Δ 13 CO 2 ( Fig. 2a) and CRR (Fig. 2b) in patients and controls. The repeated measures ANCOVA, controlling for sex, age, and body weight, indicated significant differences between the two groups in both Δ 13 CO 2 and CRR over time. The repeated measures ANCOVA with Greenhouse-Geisser adjustment revealed significant interactions between the diagnosis and sampling point for both the Δ 13 CO 2 and the CRR, indicating different patterns of change over time between the two groups. The Δ 13 CO 2 and CRR values were higher in the patients than in the controls at all sampling points during the 180-min test ( Fig. 2 and Table S1).

Increased C max , AUC, and CRR in MDD.
Using t-tests, the C max , AUC, and CRR 0-180 were all significantly higher in the patients than in the controls, and these differences remained significant after ANCOVA controlling for sex, age, and body weight ( Table 3). The T max was reduced in the patients compared to the controls at the trend level.

Association of 13 C-TBT indices with tryptophan concentrations and clinical variables in patients. Both C max (Pearson's correlation coefficient [r]
= − 0.54, p = 0.020) and CRR 0-180 (r = − 0.48, p = 0.045) showed significantly negative correlations with plasma tryptophan concentrations in the patient group. In the controls, a significantly negative correlation was also observed for the plasma tryptophan concentrations with the C max of the 13 C-TBT (r = − 0.43, p = 0.034). Further, stepwise multiple regression analysis demonstrated that plasma tryptophan concentrations were significant predictors of C max , while sex predicted the AUC and CRR 0-180 (Table S2). In addition, both the AUC [Spearman's rank correlation coefficient (ρ ) = − 0.53, p = 0.023] and CRR 0-180 (ρ = − 0.50, p = 0.033) were significantly and negatively correlated with the IMIeq dose. There were no significant correlations between 13 C-TBT indices and depression severity (HAMD-21 score; data not shown). There were also no significant correlations between the 13 C-TBT indices and the liver tests in the participants (Table S3).

Discussion
We examined, for the first time, whole-body tryptophan metabolism using a 13 C-TBT in patients with MDD and found that levels of exhaled 13 CO 2 were significantly increased in patients compared with healthy controls. Although plasma tryptophan concentrations were not significantly different between the two groups, tryptophan showed significant and negative correlations with the C max in both the patient and control groups. In our 13 C-TBT, exhaled 13 CO 2 is mainly attributable to the conversion of tryptophan via KYN into 3HAA. Therefore, our results support the study hypotheses and corroborate previous findings of KYN-pathway activation in MDD (i.e., the serotonin-KYN hypothesis) 48,53 . The 13 CO 2 measured by our 13 C-TBT can be produced by either the KYN pathway or the serotonin pathway (Fig. 1). Although just 1%-3% of dietary tryptophan is metabolized via the serotonin pathway 34 , expiration of 14 CO 2 (produced by the conversion of 14 C-labeled 5-hydroxytryptophan to serotonin) can be measurably decreased in depression 54 . Therefore, in our patients with MDD, the observed increase in expired 13 CO 2 could be attributed to increased metabolism in the KYN pathway rather than the serotonin pathway. Specifically, the observed increase in 13 CO 2 could relate to increased conversion of 3-OH-KYN to 3HAA (Fig. 1), suggesting an increased production of KYN and 3-OH-KYN (precursors of 3HAA), as well as QUIN (the immediate catabolite of 3HAA). Importantly, 3-OH-KYN and 3HAA have neurotoxic effects that are mediated by oxidative stress 2,55 . In addition, QUIN not only has neurotoxic effects mediated by the activation of NMDA-type glutamate receptors 2,56 but also induces lipid peroxidation and oxidative stress 57 . Therefore, our results are consistent with the neurodegeneration hypothesis of MDD 48 . Although tryptophan, KYN, and 3-OH-KYN are able to cross the BBB 33-35 , KA, 3HAA, and QUIN cannot do so readily 36,37 . It has also been reported that there are significant positive correlations between peripheral and central measures of the KYN and QUIN, but not tryptophan 16 . Further studies are warranted to elucidate the relationship between concentrations of peripheral and central KYN pathway metabolites and indices of 13 C-TBT. These neurotoxic metabolites in the KYN pathway are also implicated in several neurodegenerative disorders 11,12,58 . It may, therefore, be interesting to examine the possible clinical use of the 13 C-TBT in these disorders.
We observed significant increases in 13 C-TBT indices in the patient group, but found no significant difference in the plasma tryptophan concentrations between patients and controls. This result is inconsistent with previous data, including the results of our own meta-analysis [22][23][24][25][26][27] . The inconsistency may have occurred due to the small sample size and the modest severity of depression in our patients (mean HAMD-21 score of 12.6), as well as the modest effect size. Indeed, our meta-regression analysis found a negative correlation between plasma tryptophan concentrations and depression severity 22 . Moreover, some previous studies have reported no significant difference in plasma tryptophan levels between patients with MDD and healthy controls [59][60][61] . Our results may suggest that 13 C-TBT is more sensitive than plasma tryptophan levels at detecting increased metabolism in the KYN pathway.
As shown in Table 3, differences existed between the 2 groups in most of the pharmacokinetic parameters, including the AUC, CRR 0-180 , and C max , but not the T max . All these parameters are affected by gastric emptying, digestive absorption, metabolism rates, and respiratory status 62 . The AUC reflects the global amount of exhaled 13 CO 2 during the test session. However, although the CRR 0-180 also reflects the global process, it may be a better index because it takes into account the body surface area. The CRR may therefore be the best parameter for biomarker research in depression, particularly given the high    statistical significance. Finally, while T max is particularly subject to the influence of digestion, C max could be more relevant because it more closely reflects the metabolic rate. There are several limitations to this study. First, the sample size was small and the patients with MDD had only modestly severe disease. In addition, depression severity was not measured for the control sample. Despite conflicting results 31 , there is normal variability in mild depression symptoms in the healthy population together with evidence of a relationship between depressive symptoms and the KYN/tryptophan ratio 30 . Such methodological limitation might have minimized, rather than exaggerated the difference between the MDD and control groups in 13 C-TBT indices. Further studies are needed to address these issues, and should include larger samples, patients with more severe forms of MDD, and assessment for depressive symptoms among healthy controls. A second limitation is that we included 6 patients who had co-morbid diagnoses of dysthymia or an anxiety disorder. There is some evidence suggesting that there is a different KYN metabolism pathway between anxiety and depression 63 . To examine the possible influence of the comorbidity on the current results, we re-analyzed the data, excluding the 6 patients with MDD comorbidities. However, the results were essentially unchanged and the significant differences in 13 C-TBT indices remained between the 2 groups (data not shown). In addition, some studies have shown positive associations of suicidality with KYN and QUIN 29,[64][65][66] . Almost all of the patients in our study (17 of 18) had no history of having attempted suicide. However, because the MDD group in our study might be the heterogeneous group with different biological mechanisms, further studies will be needed with a sample of MDD patients without comorbidity or suicidal behavior. Third, only 10 patients were drug-free, with a further 8 being antidepressant-treated. The AUC and CRR 0-180 correlated negatively with the IMIeq dose. Therefore, because some antidepressants have anti-inflammatory effects 67 , inhibit TDO 68 , or can affect KYN/tryptophan 31 , it is possible that antidepressant treatment minimized the differences in the indices of the 13 C-TBT and concentrations of plasma tryptophan between the patients and controls. Fourth, although we measured plasma tryptophan, we did not measure any KYN pathway metabolites. Unfortunately, no conclusive relationship can be determined between the measures in this study and the 13 C-TBT indices, without knowing the data for other KYN pathway metabolites. In addition, we did not obtain data on inflammatory markers such as C-reactive protein, proinflammatory cytokines, or cortisol levels, which may affect depressive symptoms through the KYN pathway 31 . Thus, additional studies are required to examine the relationship between these data and the 13 C-TBT indices. Lastly, we did not assess the effects of vitamin B6 deficiency, which can interfere with the breakdown of KYN 69 . Pyridoxal phosphate, the active form of vitamin B6, is needed for the conversions of KYN to anthranilic acid and KA, and for that of 3-OH-KYN to 3HAA and xanthurenic acid.
In conclusion, the 13 C-TBT, a noninvasive tracer test, provided evidence of increased tryptophan metabolism along the KYN-NAD pathway in patients with MDD. Consistent with this, 13 C-TBT indices correlated negatively with plasma tryptophan concentrations. Therefore, whole-body tryptophan metabolism detected by our 13 C-TBT could be a novel and sensitive biomarker for the diagnosis and subtyping of MDD, with the potential to enhance the development of new drugs for the treatment of MDD. Moreover, when 13 C-TBT is combined with the 13 C-phenylalanine breath test 39 , discrimination between MDD and schizophrenia may be possible.