Journal of Chromatography A 1685 (2022) 463602

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

High performance liquid chromatography–tandem mass spectrometry
quantification of tryptophan metabolites in human serum and stool –
Application to clinical cohorts in Inflammatory Bowel Diseases ✩,✩✩,★
Aurore Desmons a,∗ , Lydie Humbert a , Thibaut Eguether a , Pranvera Krasniqi a ,
Dominique Rainteau a , Tarek Mahdi b , Nathalie Kapel b , Antonin Lamazière a
a

Clinical metabolomic department, Sorbonne Université, INSERM, Centre de Recherche Saint-Antoine (CRSA), Saint Antoine Hospital, Assistance Publique des
Hôpitaux de Paris (AP-HP.Sorbonne Université), Paris, France
Laboratoire de Coprologie Fonctionnelle, Hôpitaux Universitaires Pitié-Salpêtrière - Charles Foix, Assistance Publique des Hôpitaux de Paris
(AP-HP.Sorbonne Université), Paris, France

b

a r t i c l e

i n f o

Article history:
Received 12 July 2022
Revised 12 October 2022
Accepted 23 October 2022
Available online 3 November 2022
Keywords:

tryptophan metabolites profile
LC-MS/MS
Inflammatory bowel diseases

a b s t r a c t
Tryptophan, an essential amino acid, and its metabolites are involved in many physiological processes including neuronal functions, immune system, and gut homeostasis. Alterations to tryptophan metabolism
are associated with various pathologies such as neurologic, psychiatric disorders, inflammatory bowel diseases (IBD), metabolic disorders, and cancer. It is consequently critical to develop a reliable, quantitative
method for the analysis of tryptophan and its downstream metabolites from the kynurenine, serotonin,
and indoles pathways. An LC-MS/MS method was designed for the analysis of tryptophan and 20 of its
metabolites, without derivatization and performed in a single run. This method was validated for both
serum and stool. The comparisons between serum and plasma, collected with several differing anticoagulants, showed significant differences only for serotonin. References values were established in sera and
stools from healthy donors. For stool samples, as a proof of concept, the developed method was applied
to a healthy control group and an IBD patient group. Results showed significant differences in the concentrations of tryptophan, xanthurenic acid, kynurenic acid, indole-3-lactic acid, and picolinic acid. This
method allowed an extensive analysis of the three tryptophan metabolic pathways in two compartments.
Beyond the application to IBD patients, the clinical use of this method is wide-ranging and may be applied to other pathological conditions involving tryptophan metabolism, such as neurological, psychiatric,
or auto-inflammatory pathologies.
© 2022 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license
( />
1. Introduction
Tryptophan (Trp) is an essential aromatic amino acid involved
in protein synthesis and is the precursor of many bioactive compounds. Trp and its metabolites are implicated in many physiological processes such as neuronal functions, immunoregulation, and
inflammation. Trp is also identified as a key marker in gut homeostasis and its metabolism is closely linked to the intestinal microbiome. Around 90% of Trp is metabolized through the kynurenine
✩

Disclosure statement: no
All authors declare no competing financial interests and consent for publication
★
Data and material are available.
∗

Corresponding author at: Hôpital Saint-Antoine, AP-HP Sorbonne Université, 27,
rue Chaligny, 75012 Paris, France.
E-mail address: (A. Desmons).
✩✩

pathway, also called the indoleamine-2,3-dioxygenase (IDO) pathway [1–3]. This pathway leads to kynurenine production as well
as other neurologically active compounds such as kynurenic acid
(KA), quinolinic acid (QA) and 3-hydroxykynurenine (3-HK). QA
and 3-HK have neurotoxic properties while KA has neuroprotective
effects [4]. The second pathway utilizing Trp, which is quantitatively less important and contributes less than 5% of Trp degradation, is the serotonin pathway. This pathway leads to the production of 80% of total serotonin by intestinal compartment and plays
an important role in neurotransmission and neurological functions
[3,5]. The last pathway, the aryl hydrocarbon receptor (AhR) pathway utilizes Trp in the synthesis of indole and indoles derivatives
by intestinal bacteria. Many of these derivatives, such as indole3-acetic acid (IAA), indole-3-propionic acid (IPA), and indole-3-

/>0021-9673/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ( />

A. Desmons, L. Humbert, T. Eguether et al.

Journal of Chromatography A 1685 (2022) 463602

Fig. 1. Tryptophan metabolism via kynurenine (blue), serotonin (green) and indoles (orange) pathways.
Chemicals structures of tryptophan and metabolites quantified by the LC-MS/MS method developed. Figure adapted from Agus et al. [3]

carboxaldehyde, can activate the AhR expressed by some immune
and intestinal cells [6,7].
Alterations to Trp metabolism are associated with many pathological states such as neurodevelopmental, neurologic and psychiatric disorders, metabolic disorders, and cancer [8,9]. Many
publications have highlighted the modifications of tryptophan
metabolism in gastrointestinal disorders including inflammatory
bowel diseases (IBD) and irritable bowel syndrome (IBS) [3,10]. In
IBD, these alterations of Trp metabolism may be involved in the

pathogenesis of the disease [3,11]. Indoles metabolites are highly of
interest in several disorders, a recent study reported that metabolic
disorders are associated with a decrease of AhR agonists produced
from trp [12]. Depletion of trp metabolites including AhR agonists
may be affect the severity of the disease [13]. It is consequently
critical to develop a reliable, quantitative method for the analysis
of Trp and its metabolites for research and clinical purposes.
To explore Trp metabolism, several quantitative methods have
been developed. The first methods consisted of liquid chromatography (LC) separation associated with detection based on UV absorbance, fluorescence, or electrochemistry [14]. Recent methods
have been developed using LC coupled to tandem mass spectrometry (LC-MS/MS) focusing on the major metabolites of kynurenine
and serotonin pathways, and quantification performed mainly on
serum or plasma matrices [15,16]. Global analysis of Trp metabolites was developed in different biological samples but quantifications were dedicated to non-human matrices [17,18]. The purpose
of our study was to explore the metabolism of Trp in the two compartments (i.e. matrices) relevant in IBD and IBS contexts [12,13].
Beyond theses pathologies, exploration of trp metabolism is highly
of interest in neurological diseases, such as multiple sclerosis and
Huntington’s disease [19,20]. This panel may be useful to follow
new therapies targeting the gut microbial metabolism [13]. Published methodology studies have mainly been performed either in
serum or stool, but not in both compartments of patients.
Here, we developed a LC-MS/MS method for the quantification
of tryptophan and 20 of its metabolites in three different human
biological matrices: feces, serum, and plasma (Fig. 1). This method
allows an extensive analysis of the three stated metabolic pathways in a single test, without derivatization. Quantification was es-

tablished and validated in human serum, plasma, and stool using
both internal standards and external calibration, which allowed the
establishment of references values in the two compartments. As a
proof of concept, the developed method was applied to a control
healthy group and a patients group diagnosed with IBD, including
ulcerative colitis (UC) and Crohn’s disease.
2. Material and methods

2.1. Chemicals and reagents
Tryptophan (Trp) (≥ 99 %), picolinic acid (Pico) (99 %), quinolinic acid (QA) (99 %), 3-hydroxy-kynurenine (3HK) (≥ 98 %),
serotonin (5HT) (≥ 98 %), 5-hydroxy-tryptophan (5HTP)(≥ 98 %),
3-hydroxy-anthranilic acid (3-HAA) (97 %), kynurenine (KYN) (≥
98 %), xanthurenic acid (XA) (96 %), tryptamine (TA) (≥ 97 %),
kynurenic acid (KA) (≥ 98 %), 5-hydroxyindole acetic acid (5HIAA) (≥ 98 %), N-acetylserotonin (NAS) (≥ 99 %), indole-3acetamide (IAM) (98 %), indole-3- lactic acid (ILA) (99 %), indole3-carboxaldehyde (I-3CA) (97 %), melatonin (MELA) (≥ 98 %), tryptophol (TOL) (97 %), indole-3-acetic acid (IAA) (≥ 98 %), indole3-propionic acid (3-IPA) (99 %), indoxyl-sulfate (3-IS) (≥ 97 %)
and methanol (hypergrade for LC-MS LiChrosolv®) were purchased
from Sigma-Aldrich (St Louis, Missouri, USA). Chemicals purity
were provided in brackets for each standards molecules.
Internal standards (ISs), L-Tryptophan-d8 (Trp-d8) (98 %) and
anthranilic acid 15N (A15N) (98 %) were obtained from Eurisotop
(Saint-Aubin, France) and from Cambridge Isotope Laboratories, Inc.
(Andover,USA), respectively.
Formic acid (FA) was acquired from Honeywell-Fluka Fisher
scientific (IllKirch France), HiPerSolv CHROMANORM acetonitrile
(ACN) for HPLC LC-MS grade from VWR (Radnor, USA).
2.2. Analytical protocol
2.2.1. Calibration standards
The stocks solutions of unlabeled standards (1mg/mL for each
molecule) were prepared in water/methanol (50/50) (v/v) for
KYN and I3CA. TA, TOL, IAM, IAA, 3-IPA, 5-HIAA, ILA, MELA,
2


A. Desmons, L. Humbert, T. Eguether et al.

Journal of Chromatography A 1685 (2022) 463602

3-IS, Pico and NAS were dissolved in methanol (1mg/mL for
each molecule). XA, KA were dissolved in dimethylsulfoxide

(DMSO) at 1mg/ml. QA, 3-HAA were dissolved in dimethylsulfoxide (DMSO)/methanol 10/90 (v/v) at 1mg/ml. 5HT, 5HTP, were dissolved in water/methanol/acetic acid 90/9.9/0.1 (v/v/v) at 1mg/ml.
3HK was dissolved in water/methanol/NaOH0.2M 90/9.9/0.1 (v/v/v)
at 1mg/ml. Trp solution (2mg/ml) was prepared in water/sodium
hydroxide 0,2 M (75/25). Stock solutions of ISs Trp-d8 (1mg/ml)
and A15N (1mg/ml) were dissolved in methanol.
For serum, a stock calibration standards solution was prepared
by mixing equal volumes of the twenty tryptophan metabolites.
Twelve calibration standards levels were prepared from the stock
solution (ranging from 23.8 μg/ ml to 3 ng/ml) by serial dilutions
in methanol. Seven levels of calibration standards were prepared
separately for tryptophan (ranging from 125 μg/ml to 60 ng/ml) by
serial dilutions in methanol.
All stocks solutions were stored at -80°C until analysis.

2.3. HPLC-ESI-MS/MS analysis
Samples were analyzed using a LC-20ADXR (Shimadzu, Kyoto,
Japan) chromatographic system in tandem with a linear ion trap
quadrupole MS/MS spectrometer QTRAP 5500 system (SCIEX, Ontario, Canada). Chromatographic separation was performed with a
kinetex biphenyl column (100×2.1 mm; particle size 2.6μm) (Phenomenex, Torrance, USA) with a 2.1 mm C8 SecurityGuardTM ULTRA Cartridges UHPLC guard column (Phenomenex, Torrance, USA).
The mobile phases were composed of 0.4 % formic acid (FA) in
water (v/v) (mobile Phase A) and 0.4% FA in acetonitrile (mobile
phase B). The column temperature was set at 17°C, the gradient
elution was performed at 0.3 mL/min, starting at 3% of phase B;
then increasing to 20 % of phase B from 0 to 0.1 min; then increasing to 65 % B from 0.1 to 7 min; finally increasing to 95 % of phase
B from 7 to 7.5 min, and completed to 2.5 min at 95% of phase B
for wash. The column reequilibration consisted in a plateau of 3%
B for 2.5 min at 0.3 mL/min. The injection volume was 2 μL (for
plasma, serum, standards and stool).
MS detection was performed using electrospray ionisation (ESI)
in positive and negative modes, using the Multiple Reaction Monitorig (MRM) function of the analyser. For each analyte, the MS conditions were determined via direct infusion of individual standard

solutions.”
Compressed air was used as the desolvation gas and nitrogen
was used as the collision gas. The instrument parameters were set
as follows: nebuliser gas and turbo gas: 40 psi, curtain gas: 20
psi, ion spray voltage: +/- 4500 V for positive or negative ionization respectively, source temperature: 450°C. Declustering potentials (DPs) were set at 60 V, except for QA and Pico, at 33 V. A
dwell-time of 10 ms was set for all transitions at positive ionization and 20 ms for negative ionization.

2.2.2. Serum calibration standards preparation
20 μL of calibration standards were transferred to 1.5 mL Eppendorf tube, 50 ng of each internal were added from stock solution (1mg/mL) for Trp d-8 and A15N, 50 μL of “lyophilized
blank serum” (Chromsystems, Gräfelfing, Germany) and 430 μL of
methanol were added.
Samples were mixed for 15 s and placed for 30 min at 4°C. After
centrifugation (18 0 0 0 g for 15 min at +4°C), the supernatant was
transferred to a glass vial. We injected 2 μl of the mix into the
HPLC-MS/MS system for analysis.

2.2.3. Serum and plasma samples preparation
50 μL of plasma or serum sample were transferred to 1.5 mL
Eppendorf tube, 50 ng of each internal were added from stock
solution (1mg/mL) for Trp d-8 and A15N, and 450 μL of cold
methanol was added. The samples were vortex-mixed for 15 s and
placed for 30 min at 4°C. The samples were centrifuged (18 0 0 0
g for 15 min at +4°C). The total content of the supernatant was
transferred to a vial from which 2 μl were injected into the HPLCMS/MS system for analysis.

2.4. Data processing
Spectral data acquisitions were processed using Analyst (v1.6.3)
software and quantifications were performed with Multiquant
(v3.0.2) software (SCIEX, Ontario, Canada). GraphPad Prism (v6.01)
(San Diego, CA, USA) and SIMCA (v16.0.2.10561) (Göttingen, Germany) softwares were used for statistical analysis.


2.2.4. Stool calibration standards preparation
20 μL of calibration standards were transferred to 1.5 mL Eppendorf tube, 50 ng of each internal were added from stock solution (1mg/mL) for Trp d-8 and A15N, and 480 μL of methanol were
added. Samples were vortex-mixed again for 15 s and transferred
into glass vials before injection of 2 μL into the HPLC-MS/MS system for analysis.

2.5. Method validation
The method validation was based on the recommendations of
NF EN ISO 15189 criteria and international guidelines dedicated to
LC-MS/MS methods [21–23].
2.5.1. Linear range
Suitable calibration ranges were determined based upon analysis of pooled sample of serum and adjusted accordingly. For each
analyte using an internal standard, peak area response ratios were
calculated and plotted against the nominal concentration. A linear
fit was employed, and a 1/x weighting factor was applied.

2.2.5. Stool samples preparation
Stool samples were frozen at -80°C then, all samples were
feeeze-dried (Freeze dryer Buchi L-200 with scroll pump 6 qm/h)
at -55.9°C, 0.300 mbar, for 24h. 10 mg of freeze-dried stool were
transferred to 1.5 mL Eppendorf tube and 500 μL of milliQ water
were added. The samples were vortex-mixed for 15 s. 500 μL of
methanol and 50 ng of each internal were added from stock solution (1mg/mL) for Trp d-8 and A15N and samples vortex-mixed
again for 15 s. After centrifugation at 18 0 0 0 g for 15 min at +4°C,
the supernatant is collected. A second extraction was made by
adding 1 ml of methanol / water (90/10) (v/v) to the pellet. Samples were vortex-mixed for 15 s. After centrifugation (18 0 0 0 g for
15 min at +4°C) the two supernatants were pooled. Evaporation
was performed under a nitrogen flow. Five hundred μl of methanol
were added to the residue. Five μl were injected onto the HPLCMS/MS system for analysis.


2.5.2. Lower limit of detection (LOD) and lower limit of
quantification (LLOQ)
To determine LOD and LLOQ, analysis of spiked samples with
decreasing concentrations of analytes were performed. The LOD
was accepted for a signal-to-noise ratio (S/N) ≥ 3 and the LLOQ
for a signal-to-noise ratio (S/N) ≥ 10.
2.5.3. Intra- and interday accuracy and precision
Accuracy and precision were estimated from the analysis of 2
levels of quality control (QC) samples prepared for each analyte.
For intraday analysis, 20 samples were prepared and assayed in
3


A. Desmons, L. Humbert, T. Eguether et al.

Journal of Chromatography A 1685 (2022) 463602

the same day for serum and stool. For interday analysis, 6 samples and 3 samples were analyzed in separate days for serum and
stool, respectively. Acceptance criteria for accuracy was determined
as a bias within ± 15% of the nominal value and within ± 20% of
the LLOQ. Acceptance criteria for precision were defined as within
15% of relative standard deviation (R.S.D) and 20% of R.S.D of the
LLOQ.

3. Results and discussion

from these common losses. For molecules with very close precursor ions (m/z 206.1 for XA and 206.2 for ILA), product ions used
for quantification were different (m/z 160.1 and 130.1 for XA and
ILA, respectively) and the chromatographic retentions times were
also different (RT: XA 2.2 min and ILA 3.3 min).

For KA and 3IPA, with the same precursor ion (m/z 190.1), the
assigned product ions were different (m/z 115.9 KA and 130.1 IPA),
and peaks were chromatographically resolved (RT: 2.4 min for KA
and 4.6 min for 3IPA). The same ion products were shared for
quantification of some indoles derivatives (m/z 130.1): for IAM, ILA,
3IPA but the peaks were chromatographically resolved (RT: 3.0 min
(IAM); 3.3 min (ILA); 4.6 min (3-IPA)). All metabolites, except for
indoxyl sulfate, were analyzed using a positive ESI and were all
metabolites were tuned for maximum sensitivity across the linear
range (Table 1).
A challenge for the development of this method was the presence of metabolites in serum, plasma, and stool at very different levels. This heterogeneity in metabolite concentrations required
an instrument with a great dynamic range and the appropriate
preparation of standard solutions at linear ranges. Thus, two solutions were prepared, one containing 20 metabolites, ranging from
3 ng/ml to 23.8 μg/ml, and another for Trp, ranging from 60
ng/ml to 125 μg/ml. Both solutions were obtained by serial dilutions in methanol. During preparation of stock solution, poor solubility and stability of some metabolites in methanol or in aqueous solutions were observed. This phenomenon of instability was
previously described for some metabolites from the kynurenines
pathway in aqueous solutions (16). For these metabolites, a water/methanol mix, with the addition of sodium hydroxide, acetic
acid, or DMSO, was required, as detailed in the materials and
methods section. For chromatographic separation, a C18 column
˚ 100×2.1 mm). C18
was tested first (Kinetex® 5 μm, C18, 100 A,
columns, which use octaldecylsilane, were widely used in other
quantitative methods dedicated to Trp and its metabolites because
of the use of non polar solvents for these methods (e.g. water,
methanol and acetonitrile)[14–16,18]. We also tested a biphenyl
column, a new phase, that have short alkyl biphenyl ligands covalently bound to the silica surface, stable under 100% aqueous
conditions and exhibited good reverse phase retention and aromatic selectivity. Eventually, separation was performed on this column because we obtained an enhanced retention for the aromatic
derivatives.
Different column temperatures were used, from 15°C to 55°C
[18,24,25], and was maintained at 17°C following assays, which

corresponded to the lowest value available for our device and environment. This temperature value, in comparison with higher values, allowed increased retention times and better separation for
the compounds eluted first. Different percentages of FA were tested
in mobile phases, ranging from 0 to 0.4%, to check for molecule
specific ionization; the percentage of 0.4 % was set. The presence
of formic acid (FA) at the highest percentage (i.e. 0.4 %) in mobile phases allowed a better ionization of molecules, improved
the shapes of the peak, and targeted analytes exhibited MS better responses.The elution gradient was tested with H2 O/Methanol
and H2 O/ACN respectively for mobile phases A/B; better shapes of
peaks were obtained when ACN was used.

3.1. LC-MS/MS method development

3.2. Method validation

For a specific detection, MRM transitions were selected for each
metabolite and internal standard, as shown in Table 1. A representative chromatogram of all targeted metabolites was shown in Fig.
2. Ion products resulting from the losses of 18 or 44 m/z, corresponding to the loss of stable neutral molecules H2 O and CO2 respectively, were excluded where possible. This allowed us to obtain more specific transitions, and to avoid cross contamination

The method validation was performed based on different international guidelines including Food and Drug Administration (FDA)
and European Medicines Agency (EMA) recommendations [21–23].
The results for intraday and interday precision and accuracy are
reported in Table 2 and 3 for serum and stool. All metabolites
demonstrated a CV < 20% for the lowest values and a CV < 15% for
intraday and interday precisions. For the intraday assay, analytical

2.5.4. Stability
Stability was evaluated using serum calibration standards obtained as described in part 2.2.2.. Calibration standards were prepared and quantified immediately, which served as a reference.
Other aliquots from the same solution were stored at -80°C. These
samples were subjected to three freeze-thaw cycles, 10 days, one
month, and 2 months after initial freezing. Stability was acceptable
if the mean concentration was between 85% and 115% of the reference mean concentration.

2.5.5. Type of collection tube
The impact of the type of collection tube was determined by
analyzing serum and plasmas collected in citrate, heparin, and
EDTA anticoagulants from 9 healthy donors at fed state.
2.5.6. Matrix effects
Peaks areas were determined in 3 different sets of samples: a
blank matrix sample of stool (set 1), a blank matrix sample of stool
spiked with standard solutions at different levels after extraction
(set 2), and one prepared from blank matrix from the same stool
but spiked just before extraction (set 3). The matrix effect was determined as follows and expressed as a percentage: (set 2-set 1) /
(spiked concentration in methanol). An extraction yield was evaluated as follows and expressed as a percentage: (set 3)/(set 2)[21].
2.6. Biological sample testing
2.6.1. Application of the assay to clinical samples
The assay protocol was applied to clinical samples from a control group and a group of patients diagnosed with IBD (including
UC and Crohn’s disease).
2.6.2. Sample collection information
Plasma and serum samples were obtained from the French
blood establishment (EFS, Saint Antoine Hospital). Stool samples
were received from the routine workload of the laboratory for
the follow-up of patients. The study using stool residues was approved by the French public health organization (CSP-article L11213, amended by the law n°2011-2012, December 29, 2011-article 5).
Blood, from non-fasting or fasting subjects, was collected in
tubes containing coagulant, a clot activator, or anticoagulant (heparin, sodium citrate, EDTA). After collection, tubes were centrifuged
(3,0 0 0 g for 10 min at 8°C) and serum or plasma aliquoted in Eppendorf tubes, ensuring that no red blood cells or clots were carried over, and stored at -80°C before analysis. Stool samples were
collected and immediately stored at -80°C; samples were freezedried before analysis.

4


A. Desmons, L. Humbert, T. Eguether et al.


Journal of Chromatography A 1685 (2022) 463602

Table 1
MS conditions and retention times for metabolites and labeled internal standards (ordered by retention time). ∗ quantifier ion.
Component Names
Positive Mode
Piconilic acid
Quinolinic acid
3-Hydroxy Kynurenine
Serotonin
5-Hydroxytryptophan
3-Hydroxyanthranilic acid
Kynurenine
Xanthurenic acid
Tryptophan
Tryptamine
Kynurenic acid
5-Hydroxyindole acetic acid
N-Acetylserotonin
Indole-3-acetamide
Indole-3- lactic acid
Indole-3-carboxaldehyde
Melatonin
Tryptophol
Indole-3- acetic acid
Indole-3-propionic acid
Internal Standards
L Tryptophan D8
Anthranilic acid 15N
Negative Mode

Indoxyl sulfate

Retention Time (min)

Precursor ion (m/z)

Products ions (m/z)

Collision energy (V)

Internal Standard

1.2
1.3
1.4
2.0
2.0
2.1
2.1
2.2
2.3
2.4
2.4
2.6
2.6
3.0
3.3
3.7
3.7
3.8

3.9
4.6

[M+H]
124.2
168.1
225.1
177.2
221.1
154.1∗ ; 136.1
209.2
206.1
205.2
161.2
190.1
192.1
219.1
175.1
206.2
146.1
233.2
162.1
176.1
190.1

[M+H]
77.9 ∗ ; 96.0
78.0∗ ; 124.0
162.1∗ ; 208.1
115.1∗ ; 132.2

204.2; 162.2
136.1∗ ; 108.2
146.1∗ ; 94.1
160.1∗ ; 132.0
118.1∗ ; 146.1
117.1∗ ; 144.2
115.9∗ ; 144.1
146.1∗ ; 118.0
160.0∗
130.1∗ ; 103.1
130.1∗ ; 118.1
91.1∗ ; 118.1
174.2∗
115.1∗ ; 144.1
103.1∗ ; 130.1
130.1∗

30∗ ;
30∗ ;
25∗ ;
30∗ ;
25∗ ;
25∗ ;
25∗ ;
25∗ ;
55∗ ;
30∗ ;
40∗ ;
30∗ ;
25∗

20∗ ;
30∗ ;
25∗ ;
22∗
40∗ ;
50∗ ;
20∗

L-Tryptophan-d8
L-Tryptophan-d8
L-Tryptophan-d8
L-Tryptophan-d8
L-Tryptophan-d8
L-Tryptophan-d8
L-Tryptophan-d8
L-Tryptophan-d8
L-Tryptophan-d8
Anthranilic acid 15N
Anthranilic acid 15N
Anthranilic acid 15N
Anthranilic acid 15N
Anthranilic acid 15N
Anthranilic acid 15N
Anthranilic acid 15N
Anthranilic acid 15N
Anthranilic acid 15N
Anthranilic acid 15N
Anthranilic acid 15N

213.3

139.2
[M-H]
212.0

151.3∗ ; 195.3
65.1∗ ; 121.1
[M-H]
132.4∗ ; 80.4

2.2
3.0
2.6

20
13
13
25
20
25
25
30
45
30
25
45
55
30
25
30
25


25∗ ; 25
30∗ ; 30
27∗ ; 27

Anthranilic acid 15N

Fig. 2. Chromatogram of the 21 standards of analysis ordered by retention times (in brackets, transitions of precursor and quantifier ions (m/z)).

accuracy was < 15% (recovery comprised between 85 % and 115%),
and for the interday assay some recoveries were higher than 15%,
particularly for the lowest values in serum. Accuracies in serum
ranged between 85 and 115% except for 3 compounds for which
percentages moderately under 85%: NAS (84%), IAA (84%) and 3IPA(83%) (Table 2). For intraday and interday assays in stool, analytical accuracy and CV were < 15% for all molecules quantified
(Table 3). Dynamic ranges were determined according to the physiological and pathological values previously described, and results
are reported in Table 4 for LOD and LOQ, and Table 2 for ULOQ,
which corresponded to the highest calibration standards quantified.

To account for matrix effect for serum, calibration standards
were prepared using lyophilized blank serum. For stool samples,
matrix effects were evaluated as described in material and methods section. Matrix effects calculated at 2 levels were in accordance with published studies, ranging from 50 % to 125%, but few
components exhibited highest matrix effects (34%, 196%, 41 %, 36
%, for picolinic acid, tryptophan, melatonin, N acetyl serotonin, respectively) (Table 5). These resuts were obtained for the lowest
level of spike for all of the 4 metabolites, but the levels of spike
tested in our method validation were lower than other published
methods [18]. Recoveries ranging from 62% to 134% (Table 5). Despite high variance, these results were consistent with those previ-

5



A. Desmons, L. Humbert, T. Eguether et al.

Journal of Chromatography A 1685 (2022) 463602

Table 2
Intraday and interday precision in serum.
Means, coefficients of variation (CVs) and recoveries were calculated from internal quality controls (number of assays:
20 for intra- and 6 inter-assays precisions, respectively) for two levels.
Intra-Day

Inter-Day

Component Names

Nominal concentration (nM)

CV(%)

Recovery (%)

CV(%)

Recovery (%)

Piconilic acid

177.2
5669.1
261.0
4180.0

195.0
779.2
31.0
3966.1
99.0
3173.0
143.0
4562.0
105.0
3356.0
27.0
3405.0
595.0
152364.7
36.0
4362.5
115.0
3693.0
29.0
3655.0
25.0
3202.0
62.0
4012.0
106.0
3405.0
150.0
4813.0
23.0
3008.0

34.0
4335.0
31.0
3989.0
29.0
3693.0
103.0
3309.0

13.7
9.6
15.7
9.3
8.1
8.1
11.1
10.0
16.2
6.4
8.8
6.5
7.2
5.9
9.3
4.0
10.7
3.5
8.5
5.9
7.0

8.2
10.5
6.7
13.9
4.6
6.8
6.7
11.7
7.8
8.1
5.6
11.7
3.3
11.7
6.9
10.9
7.8
13.3
5.5
9.8
5.4

89.7
88.1
88.2
109.1
98.6
107.8
92.1
92.5

99.1
102.0
90.0
107.2
104.8
105.3
92.0
114.4
95.4
88.0
88.2
102.5
104.8
101.8
102.5
89.1
85.7
85.0
108.5
92.3
91.9
95.8
107.5
87.8
93.8
103.3
94.0
94.6
95.1
109.8

93.8
98.5
93.0
88.1

17.3
11.4
14.8
11.5
11.7
15.4
12.0
12.6
17.7
9.7
11.9
6.4
8.7
6.1
10.5
4.3
6.7
2.8
11.1
6.6
8.9
9.7
10.4
6.9
13.0

6.9
6.8
9.0
13.2
7.3
11.5
6.7
12.2
12.4
12.7
8.3
11.3
8.5
13.9
7.5
10.0
11.7

99.9
89.5
91.8
103.4
113.7
102.4
91.5
90.4
90.2
101.7
88.1
99.9

102.3
99.3
97.4
108.2
94.8
85.5
97.6
98.6
104.7
102.6
106.8
94.4
91.6
84.3
95.0
93.8
99.5
92.7
87.9
98.5
96.5
100.6
96.0
92.0
84.0
89.1
83.1
104.9
96.1
112.3


Quinolinic acid
3-Hydroxy Kynurenine
Serotonin
5-Hydroxytryptophan
3-Hydroxyanthranilic acid
Kynurenine
Xanthurenic acid
Tryptophan
Tryptamine
Kynurenic acid
5-Hydroxyindole acetic acid
N-Acetylserotonin
Indole-3-acetamide
Indole-3- lactic acid
Indole-3-carboxaldehyde
Melatonin
Tryptophol
Indole-3- acetic acid
Indole-3-propionic acid
Indoxyl sulfate

ously described in a similar matrix [18]. Matrix effects could be
minimized by adding a larger number of stable isotope-labeled
analogue to the targeted molecule. Some preanalyticals steps based
on on phospholipid removal and interferent proteins may be performed to minimize matrix effect. In blood, method validation was
performed in serum samples, but a comparison was done between
blood collected on different anticoagulants (EDTA, heparin, citrate).
No major difference was found regarding the presence and type of
anticoagulant except for with serotonin, where levels were higher

in serum than in plasma (supplemental data, Fig. 1). These results
for serotonin were already described [15], however, it was necessary to evaluate the percentage of difference when developing our
method. Residue stabilities are shown in supplemental data, Table
1. After the first thawing, at 10 days, 7 compounds (3-HK, XA, KA,
5-HIAA, I3A, ILA, IS) were under acceptable values.

and highlights changes in both host metabolism (serum) and microbial metabolism (stool)[26]. Following analytical validation, the
method was tested on 24 human serum samples and 22 human
stool samples from healthy donors (Table 6). Reference values were
in agreement with previously published reports [26]. The few differences between our data and published results could be explained by the non-fasting state of our donors, which was imposed by the composition of the available patient cohort. Subsequently, we tested our method on IBD patient samples to investigate its potential in clinical practice. Samples were obtained from
two groups: patients diagnosed with IBD (27 patients) and a control group of non-IBD patients (13 patients) (Fig. 3A). Fecal calprotectin was used to classify patients into three different groups
among IBD patients according to the established cut-offs used in
IBD management and diagnosis [27]: 1) Patients with fecal calprotectin < 50 μg/g, 2) patients with fecal calprotectin between 50
and 200 μg/g, which corresponds to an intermediate state, and 3)
patients with fecal calprotectin > 200 μg/g which corresponds to
an acute phase of the disease. Fecal calprotectin was also assayed
in patients from the control group, who all had a calprotectin level
< 50 μg/g.
The multivariate analysis identified the most discriminant variables (VIP > 1) to stratify non IBD patients from IBD patients:

3.3. Application of the method: human serum and stool
concentrations
The main objective of our work was the development of a reliable method for the exploration of Trp metabolism in two relevant compartments, serum and stool, in cases of IBD and IBS. This
method allows extensive metabolite profiling of Trp metabolism
6


A. Desmons, L. Humbert, T. Eguether et al.

Journal of Chromatography A 1685 (2022) 463602


Table 3
Intraday and interday precision in stool.
A. Intraday precison in stool. B. Interday precision in stool. Means, coefficients of variation (CVs) and recoveries were
calculated from internal quality controls (number of assays: 20 for intra- and 3 inter-assays precisions, respectively) for
two levels.
Sample-1

Sample-2

Sample-3

Component Names

Mean (nmol/g)

CV (%)

Mean (nmol/g)

CV (%)

Mean (nmol/g)

CV (%)

Piconilic acid
Quinolinic acid
3-Hydroxy Kynurenine
Serotonin

5-Hydroxytryptophan
3-Hydroxyanthranilic acid
Kynurenine
Xanthurenic acid
Tryptophan
Tryptamine
Kynurenic acid
5-Hydroxyindole acetic acid
N-Acetylserotonin
Indole-3-acetamide
Indole-3- lactic acid
Indole-3-carboxaldehyde
Melatonin
Tryptophol
Indole-3- acetic acid
Indole-3-propionic acid
Indoxyl sulfate
A
B

393.7
< LOQ
1.4
16.8
3.8
4.0
< LOQ
0.7
106.0
1.2

6.6
1.1
0.3
< LOQ
< LOQ
8.4
< LOQ
0.6
29.2
29.1
2.6

5.8

116.1
3.2
< LOQ
3.2
1.2
7.2
< LOQ
4.6
34.4
2.2
16.6
0.28
0.2
< LOQ
< LOQ
9.4

< LOQ
1.3
20.9
11.4
2.7

8.5
8.5

524.4
2.6
< LOQ
4.3
2.6
4.5
1.7
9.3
1492.0
49.6
5.5
< LOQ
0.6
< LOQ
9.1
6.9
< LOQ
1.7
32.6
54.6
1.0


5.6
9.0

7.4
6.7
3.0
6.7
6.5
4.4
4.9
4.4
3.9
4.1

5.6
10.3
5.1
6.2
5.5

Sample-1

2.4
11.0
5.2
6.5
7.9
9.4
5.2

4.72
4.7

3.9
7.5
7.6
5.3
6.9

Sample-2

3.3
11.3
6.2
7.1
3.7
3.1
5.6
5.9
4.9
5.7
6.2
7.6
5.5
5.7
10.7

Sample-3

Component Names


Mean (nmol/g)

CV (%)

Mean (nmol/g)

CV (%)

Mean (nmol/g)

CV (%)

Piconilic acid
Quinolinic acid
3-Hydroxy Kynurenine
Serotonin
5-Hydroxytryptophan
3-Hydroxyanthranilic acid
Kynurenine
Xanthurenic acid
Tryptophan
Tryptamine
Kynurenic acid
5-Hydroxyindole acetic acid
N-Acetylserotonin
Indole-3-acetamide
Indole-3- lactic acid
Indole-3-carboxaldehyde
Melatonin

Tryptophol
Indole-3- acetic acid
Indole-3-propionic acid
Indoxyl sulfate

383.1
< LOQ
1.4
16.2
3.5
3.6
< LOQ
0.7
110.2
1.6
6.1
1.0
0.3
< LOQ
< LOQ
8.4
< LOQ
0.6
26.4
27.7
2.2

10.3

106.7

3.1
< LOQ
3.2
1.1
6.7
< LOQ
4.8
34.5
2.7
15.5
0.3
0.2
< LOQ
< LOQ
8.8
< LOQ
1.2
19.7
12.0
2.5

4.3
9.6

466.9
2.7
< LOQ
4.2
2.5
4.2

1.6
9.8
1502.0
45.4
5.6
< LOQ
0.6
< LOQ
8.6
6.7
< LOQ
1.7
31.7
55.3
1.0

11.1
9.7

4.6
6.5
5.0
6.6
4.5
10.3
5.5
10.5
2.5
2.0


6.6
7.6
8.4
10.7
6.2

3.7
8.3
10.5
3.5
3.7
2.6
1.8
7.6
8.5

3.5
6.0
5.4
7.3
7.7

6.0
5.8
7.4
7.2
5.2
1.5
4.6
7.8

2.7
10.7
7.7
2.1
9.1
8.7
5.1

Fig. 3. Analysis of clinical samples.
Stool samples from non IBD and IBD patients. A. Score plot of patients obtained from OPLS-DA. B. Variable importance in projection (VIP) exhibited significant differences
for XA, KA, ILA, calprotectin, trp, Pico. XA, KA, ILA, calprotectin and trp are increased (red bar plot) and pico is decreased (blue bar plot) in IBD patients compared to control
group.

7


A. Desmons, L. Humbert, T. Eguether et al.

Journal of Chromatography A 1685 (2022) 463602

Table 4
Limit of detection (LOD) and limit of quantification (LOQ).
LOD was determined by a signal-to-noise ratio (S/N) of 3:1 and LOQ by a signal-to-noise ratio (S/N) of 10:1

Piconilic acid
Quinolinic acid
3-Hydroxy Kynurenine
Serotonin
5-Hydroxytryptophan
3-Hydroxyanthranilic acid

Kynurenine
Xanthurenic acid
Tryptophan
Tryptamine
Kynurenic acid
5-Hydroxyindole acetic acid
N-Acetylserotonin
Indole-3-acetamide
Indole-3- lactic acid
Indole-3-carboxaldehyde
Melatonin
Tryptophol
Indole-3- acetic acid
Indole-3-propionic acid
Indoxyl sulfate

Lower limit of Detection (nmol/L)

Lower limit of Quantification (nmol/L)

1.4
65.3
48.7
31
3.1
35.6
3.3
0.8
28.3
34

3.3
3.6
3.1
3.9
0.8
4.7
0.7
16.9
15.5
0.9
25.8

88.5
130
97.5
124
49.6
71
105
26.6
37.2
136
105
14.3
6.2
62.5
26.5
37.6
5.9
67.5

62.3
14.5
103

Table 5
Matrix effect and yield extraction in stool at 3 levels (low, medium, high).
Component Names

Spiked Concentration μg/ml

Matrix effect %

Recovery %

Piconilic acid

1.6
12.2
1.2
9.0
0.9
6.7
1.1
8.5
0.9
6.8
1.3
9.8
1.0
7.2

1.0
7.3
1.0
7.3
1.2
9.4
1.1
7.9
1.0
7.8
0.9
6.9
1.1
8.6
1.0
7.3
1.4
10.3
0.9
6.5
1.2
9.3
1.1
8.6
1.1
7.9
0.9
7.0

34

57
125
93
81
82
57
50
71
55
89
94
95
74
107
86
196
68
60
76
108
118
76
84
36
57
48
55
89
74
93

82
41
60
108
93
80
84
106
79
100
77

90
123
80
103
96
80
82
72
75
86
87
84
78
86
94
96
99
101

76
62
94
93
89
80
86
83
95
93
89
85
73
65
91
85
79
82
97
80
93
91
123
134

Quinolinic acid
3-Hydroxy Kynurenine
Serotonin
5-Hydroxytryptophan
3-Hydroxyanthranilic acid

Kynurenine
Xanthurenic acid
Tryptophan
Tryptamine
Kynurenic acid
5-Hydroxyindole acetic acid
N-Acetylserotonin
Indole-3-acetamide
Indole-3- lactic acid
Indole-3-carboxaldehyde
Melatonin
Tryptophol
Indole-3- acetic acid
Indole-3-propionic acid
Indoxyl sulfate

8


A. Desmons, L. Humbert, T. Eguether et al.

Journal of Chromatography A 1685 (2022) 463602

Table 6
Metabolites concentrations in serum and feces. References values
were determined from healthy individuals in serum (n=24) and stool
(n=22), results were expressed as mean ± standard deviation (SD).

Piconilic acid
Quinolinic acid

3-Hydroxy Kynurenine
Serotonin
5-Hydroxytryptophan
3-Hydroxyanthranilic acid
Kynurenine
Xanthurenic acid
Tryptophan
Tryptamine
Kynurenic acid
5-Hydroxyindole acetic acid
N-Acetylserotonin
Indole-3-acetamide
Indole-3- lactic acid
Indole-3-carboxaldehyde
Melatonin
Tryptophol
Indole-3- acetic acid
Indole-3-propionic acid
Indoxyl sulfate

Serum (nmol/L)

Feces (nmol/g)

< LOQ
231 ± 139
12 ± 14
377 ± 177
< LOQ
33 ± 47

1652 ± 272
332 ± 52
51460 ± 9187
< LOQ
< LOQ
57 ± 49
< LOQ
< LOQ
823 ± 228
< LOQ
< LOQ
< LOQ
1489 ± 475
1337 ± 1039
3130 ± 1451

69390 ± 8730
190 ± 120
< LOQ
910 ± 790
170 ± 180
1730 ± 1000
70 ± 50
510 ± 550
10320 ± 6400
770 ± 1120
1790 ± 1900
230 ± 240
50 ± 50
< LOQ

450 ± 450
1150 ± 900
< LOQ
340 ± 300
5360 ± 4760
6500 ± 3380
70 ± 320

serum of IBD patients [16,31] but no studies have determined PA
levels in feces for these patients.
Eventually, despite a rather small number of patients in the
cohort, the results obtained through our developed method evidenced significative differences of tryptophan derivatives patterns
between IBD and non IBD patients.
4. Conclusions
A HPLC-ESI-MS/MS method has been developed for the analysis
of Trp and 20 metabolites for application to human serum, plasma,
and stool samples. This method allows for an extensive analysis of
the three Trp metabolic pathways in two compartments and was
applied to a clinical study of IBD patients. Increased concentrations
of Trp, XA, KA, ILA, and a decrease of PA were observed in IBD patients compared to healthy controls using this approach. The clinical use of this method is wide-ranging and may be applied to other
pathological conditions involving Trp metabolism such as neurological, psychiatric, or auto-inflammatory pathologies.
Authors contribution
A.D. and A.L. performed the design of research. L.H., A.D. and
D.R. performed experiments and data acquisition. A.D., N.K. and
A.L. performed data analysis and interpretation. A.D. wrote the paper. All authors critically revised the article. All authors approved
the final version to be published.

calprotectin, Trp, and four Trp metabolites, xanthurenic acid (XA),
kynurenic acid (KA), indole-3-lactic acid (ILA) and picolinic acid
(Pico) (Fig. 3 B). These results were confirmed by univariate statistical analysis as shown in supplemental data, Fig. 2. The results obtained showed that Trp derivatives are important markers for study

in clinical settings with our method.
Three patients (numbers 30, 31, and 37) tested very close to
the control group in the score plot. Patient 37 was suffering from
Crohn’s disease and presented 3 consecutive normal levels of calprotectin over 6 months, which was probably due to a deep remission. Patient 31 also presented Crohn’s disease in deep remission
and did not receive drug treatment. Patient 30 was suffering from
UC under treatment and presented two consecutive normal levels
of calprotectin over 6 months, which was probably due to a steady
state of the disease.
Most published studies of Trp derivatives in IBD patients have
been performed in serum or plasma samples. Trp metabolites, such
as PA and XA, were reported at a lower concentration in serum or
plasma than control groups [16,26]. The increase we observed for
Trp in stool samples supports previous results, which showed an
increase in fecal amino acids [26]. Trp increase in feces could be
explained by inflammation and mucosal damage in IBD due to a
compromised gut epithelial barrier [26,28]. An extensive quantification of amino-acid profiles in blood and stool may be useful to
highlight the impaired absorption of all amino acids in the gut in
patients suffering from IBD.
Regarding the indoles pathway, which corresponds to indoles
derivatives produced by commensal bacteria, an increase of ILA
in feces was observed. Some studies have shown how indoles
molecules derived from bacteria are implicated in intestinal inflammation, IPA, an indole molecule, have been reported to be decreased in serum [29]. This result is consistent with the increase
of ILA found in feces, a downstream product of IPA. Other results
showed an increase for two downstream products of the kynurenine pathway, KA and XA, and a decrease in PA. The increase of
KA and XA, which are derived from kynurenine metabolism by indolemanine 2,3 dioxygenase (IDO), could be supported by the previously reported increase of IDO activity in IBD patients [30]. A decrease for PA in feces may be explained by an increase of the production of QA, however, in our study, no significance was found for
the increase in QA. A decrease in PA was previously reported in the

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Data Availability
Data will be made available on request.
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.chroma.2022.463602.
References
[1] PJ Kennedy, JF Cryan, TG Dinan, G. Clarke, Kynurenine pathway metabolism
and the microbiota-gut-brain axis, Neuropharmacology 112 (2017) 399–412.
[2] I Cervenka, LZ Agudelo, JL. Ruas, Kynurenines: Tryptophan’s metabolites in exercise, inflammation, and mental health, Science 357 (6349) (2017) eaaf9794.
[3] A Agus, J Planchais, H. Sokol, Gut microbiota regulation of tryptophan
metabolism in health and disease, Cell Host Microbe 23 (6) (2018) 716–724.
[4] L Boulet, G Besson, L Van Noolen, P Faure, ECOPHEN Study Group, F Maillot,
et al., Tryptophan metabolism in phenylketonuria: A French adult cohort study,
J. Inherit. Metab. Dis. 43 (5) (2020) 944–951.
[5] GM Mawe, JM. Hoffman, Serotonin signalling in the gut–functions, dysfunctions and therapeutic targets, Nat. Rev. Gastroenterol. Hepatol. 10 (8) (2013)
473–486.
[6] EE Alexeev, JM Lanis, DJ Kao, EL Campbell, CJ Kelly, KD Battista, et al., Microbiota-derived indole metabolites promote human and murine intestinal homeostasis through regulation of interleukin-10 receptor, Am. J. Pathol. 188 (5)
(2018) 1183–1194.
[7] TD Hubbard, IA Murray, GH. Perdew, Indole and tryptophan metabolism: endogenous and dietary routes to ah receptor activation, Drug Metab. Dispos.
Biol. Fate Chem. 43 (10) (2015) 1522–1535.
[8] M Platten, EAA Nollen, UF Röhrig, F Fallarino, CA. Opitz, Tryptophan
metabolism as a common therapeutic target in cancer, neurodegeneration and
beyond, Nat. Rev. Drug Discov. 18 (5) (2019 May) 379–401.
[9] W Roth, K Zadeh, R Vekariya, Y Ge, M. Mohamadzadeh, Tryptophan
Metabolism and Gut-Brain Homeostasis, Int. J. Mol. Sci. 22 (6) (2021) 2973.
[10] RL Burr, H Gu, K Cain, D Djukovic, X Zhang, C Han, et al., Tryptophan metabolites in irritable bowel syndrome: an overnight time-course study, J. Neurogastroenterol. Motil. 25 (4) (2019) 551–562.
9


A. Desmons, L. Humbert, T. Eguether et al.


Journal of Chromatography A 1685 (2022) 463602

[11] M Sun, N Ma, T He, LJ Johnston, Ma X. Tryptophan, Trp) modulates gut homeostasis via aryl hydrocarbon receptor (AhR), Crit. Rev. Food Sci. Nutr. 60 (10)
(2020) 1760–1768.
[12] L Xiao, Q Liu, M Luo, L. Xiong, Gut microbiota-derived metabolites in irritable
bowel syndrome, Front. Cell Infect Microbiol. [Internet] (2021) [cited 2022 Apr
19];11. Available from: />729346 .
[13] HM Roager, TR. Licht, Microbial tryptophan catabolites in health and disease,
Nat. Commun. 9 (1) (2018) 3294.
´
[14] I Sadok, K Jedruchniewicz,
˛
K Rawicz-Pruszynski,
M. Staniszewska, UHPLC-ESI-MS/MS quantification of relevant substrates and metabolites of the kynurenine pathway present in serum and peritoneal fluid from gastric cancer patients-method development and validation, Int. J. Mol. Sci. 22 (13) (2021) 6972.
[15] L Boulet, P Faure, P Flore, J Montérémal, V. Ducros, Simultaneous determination of tryptophan and 8 metabolites in human plasma by liquid chromatography/tandem mass spectrometry, J. Chromatogr. B 1054 (2017) 36–43.
[16] L Whiley, LC Nye, I Grant, N Andreas, KE Chappell, MH Sarafian, et al., Ultrahigh-performance liquid chromatography tandem mass spectrometry with
electrospray ionization quantification of tryptophan metabolites and markers
of gut health in serum and plasma-application to clinical and epidemiology
cohorts, Anal. Chem. 91 (8) (2019) 5207–5216.
[17] W Hou, D Zhong, P Zhang, Y Li, M Lin, G Liu, et al., A strategy for the targeted metabolomics analysis of 11 gut microbiota-host co-metabolites in rat
serum, urine and feces by ultra high performance liquid chromatography-tandem mass spectrometry, J. Chromatogr. A 1429 (2016) 207–217.
[18] A Lefèvre, S Mavel, L Nadal-Desbarats, L Galineau, S Attucci, D Dufour,
et al., Validation of a global quantitative analysis methodology of tryptophan
metabolites in mice using LC-MS, Talanta 195 (2019) 593–598.
[19] V Rothhammer, ID Mascanfroni, L Bunse, MC Takenaka, JE Kenison, L Mayo,
et al., Type I interferons and microbial metabolites of tryptophan modulate
astrocyte activity and central nervous system inflammation via the aryl hydrocarbon receptor, Nat. Med. 22 (6) (2016) 586–597.
[20] J Kaye, V Piryatinsky, T Birnberg, T Hingaly, E Raymond, R Kashi, et al.,
Laquinimod arrests experimental autoimmune encephalomyelitis by activating

the aryl hydrocarbon receptor, Proc. Natl. Acad. Sci. U. S. A. 113 (41) (2016)
E6145–E6152.

[21] FT Peters, OH Drummer, F. Musshoff, Validation of new methods, Forensic Sci.
Int. 165 (2–3) (2007) 216–224.
[22] A Mochizuki, K Ieki, H Kamimori, A Nagao, K Nakai, A Nakayama, et al.,
Proposal for risk-based scientific approach on full and partial validation for
general changes in bioanalytical method, Bioanalysis 10 (8) (2018) 577–
586.
[23] H Blume, E Brendel, M Brudny-Klöppel, S Grebe, B Lausecker, G Rohde, et al.,
Workshop/conference report on EMA draft guideline on validation of bioanalytical methods, Eur. J. Pharm. Sci. 42 (3) (2011) 300–305.
[24] L Schwieler, A Trepci, S Krzyzanowski, S Hermansson, M Granqvist, F Piehl,
et al., A novel, robust method for quantification of multiple kynurenine pathway metabolites in the cerebrospinal fluid, Bioanalysis 12 (6) (2020) 379–
392.
[25] Z Galla, C Rajda, G Rácz, N Grecsó, Á Baráth, L Vécsei, et al., Simultaneous
determination of 30 neurologically and metabolically important molecules:
A sensitive and selective way to measure tyrosine and tryptophan pathway
metabolites and other biomarkers in human serum and cerebrospinal fluid, J.
Chromatogr. A 1635 (2021) 461775.
[26] L Aldars-García, JP Gisbert, M. Chaparro, Metabolomics insights into inflammatory bowel disease: a comprehensive review, Pharm. Basel Switz. 14 (11)
(2021) 1190.
[27] A Dhaliwal, Z Zeino, C Tomkins, M Cheung, C Nwokolo, S Smith, et al., Utility of
faecal calprotectin in inflammatory bowel disease (IBD): what cut-offs should
we apply? Frontl. Gastroenterol. 6 (1) (2015) 14–19.
[28] A Murgia, C Hinz, S Liggi, J Denes, Z Hall, J West, et al., Italian cohort of patients affected by inflammatory bowel disease is characterised by variation in
glycerophospholipid, free fatty acids and amino acid levels, Metabolomics 14
(10) (2018) 140.
[29] Y Lai, J Xue, CW Liu, B Gao, L Chi, P Tu, et al., Serum metabolomics identifies
altered bioenergetics, signaling cascades in parallel with exposome markers in
Crohn’s disease, Molecules 24 (3) (2019) 449.

[30] MA Ciorba, Indoleamine 2,3 dioxygenase (IDO) in intestinal disease, Curr. Opin.
Gastroenterol. 29 (2) (2013) 146–152.
[31] LM Chen, CH Bao, Y Wu, SH Liang, D Wang, LY Wu, et al., Tryptophan-kynurenine metabolism: a link between the gut and brain for depression in inflammatory bowel disease, J. Neuroinflammation. 18 (1) (2021) 135.

10