Регулирование Kynurenin Pathway в Relapping Rising Multial Sclerosis и его корреляции с прогрессивным нейродегенерацией.

Introduction

Multiple sclerosis (MS) is a chronic neurodegenerative disease 1 with no effective treatment. 2 Axonal degeneration is the major determinant of progressive neurologic disability in MS and can be attributed to the activation of adaptive and innate immune cells. 3 Although immune cells play a key role in the neurodegeneration in MS, current disease-modifying therapies (DMTs) effective in modulating T-cell and B-cell response showed limited efficacy. This underscores the intricate nature of mechanisms contributing to the disease's pathophysiology, and the kynurenine pathway (KP) has been proposed as a potential mechanism connecting inflammation with neurodegeneration in MS.

The KP is the major pathway in the breakdown of the essential amino acid tryptophan (TRP), producing various metabolites. Indoleamine 2,3 dioxygenase (IDO) is the rate-limiting enzyme that catalyzes TRP into kynurenine (KYN) in extrahepatic tissues. 4 Notably, IDO activity and/or gene expression is induced by proinflammatory factors such as interferon interferon gamma (IFN-γ) 5 , 6 to produce a variety of bioactive metabolites that are immunomodulatory, neurotoxic, or neuroprotective. 7 Thus, the KP has been implicated in various neurodegenerative conditions, including Alzheimer diseases. 7

Key consequences of chronic KP activation include the overproduction of metabolites, such as quinolinic acid (QUIN) and 3-hydroxykynurenine (3HK), that are typically considered neurotoxic and gliotoxic because of their adverse effects on glutamatergic neurotransmission and/or direct toxicity towards neurons, astrocytes, and oligodendrocytes cells. 8 On the other hand, kynurenic acid (KYNA) is an antagonist of the NMDA receptor, blocks the excitotoxic effects of QUIN, and has been, therefore, considered a neuroprotective molecule. 9 In addition, KP metabolites such as KYN, KYNA and 3-hydroxyanthranilic acid (3HAA) can activate the aryl hydrocarbon receptor to expand Th17 and regulatory T cells, which plays a significant role in autoimmune disorders. 10

The balance and regulation of KP compounds are crucial because imbalances can lead to neurodegeneration and have been reported in blood, urine, and CSF of patients with MS and mice with experimental autoimmune encephalitis (EAE), one of the preclinical models of MS. 11 On the basis of the significant data that KP is dysregulated in MS 4 and that levels of KP can predict disease course, 12 this points toward the notion that profiling of the KP may potentially be a useful marker for changes to axonal and functional loss in the disease.

Advanced noninvasive methods such as visual pathway assessments have emerged as valuable tools for neurodegeneration evaluation. These assessments include the optical coherence tomography (OCT), measurement of the thickness of the retinal nerve fiber layer (RNFL), 13 the ganglion cell inner plexiform layer (GCIPL), 14 and multifocal visual evoked potential. 14 , 15 In this study, we used the MRI in conjunction with these visual pathway assessments established within our group 6 to evaluate neurodegeneration in relapsing remitting MS (RRMS). We aim to correlate these changes with KP changes to establish whether the KP could serve as a useful blood-based biomarker for clinical assessment of MS. Importantly, this could shed light on the role of KP in continued neurodegeneration in MS despite the absence of acute phrase or flare.

Methods

Participants

Participants in this study were selected from 5 specialized hospitals and tertiary neurology clinics in Sydney between 2018 and 2019. The patient group included individuals with a confirmed clinical diagnosis of MS, as verified by their treating neurologists. These patients were matched by age and sex with healthy controls. All MS patients met the criteria for RRMS. 16 The exclusion criteria for the study were other neurologic conditions affecting the visual system or cases of optic neuropathy, autoimmune disorders other than MS, major illnesses such as cancer, immunodeficiency syndrome, or hepatitis, any prior chemotherapy or similar treatments, current pregnancy or breastfeeding, and recent episodes of acute optic neuritis or flare-ups within the past 6 months. The control group consisted of healthy, non–blood-related family members of the patients with MS who did not meet any of the exclusion criteria and had no history of autoimmune diseases. The study size, i.e., MS group (n = 98) and healthy controls (n = 39), derived from all the participants who met the inclusion criteria during the period of recruitment and selection.

Standard Protocol Approvals, Registrations, and Patient Consents

Ethical approval for this study was granted by the Macquarie University Human Ethics Committee (HREC 520211087035353). Participation was entirely voluntary, with all participants providing informed consent. To ensure confidentiality, all demographic, medical data, and collected samples were anonymized and deidentified by the research team.

Sample Collection Protocol and Assessment of Clinical Progressive Neurodegeneration

Blood samples were obtained at baseline from patients with MS and healthy controls, following a nonfasting protocol. These samples were collected into Vacutainer EDTA tubes (Becton Dickinson), followed by centrifugation at 2,000 × g for 10 minutes. The plasma was then removed for storage at −80°C for subsequent analysis.

Patients with MS also underwent annual clinical evaluations, which included MRI and OCT scans. 17 , 18 These assessments focused on indicators such as T2 lesions, brain volume, cortical thickness, and retinal thickness, providing insights into the extent of brain atrophy and neurodegeneration. All neurologic examinations were conducted by experienced neurologists. Baseline clinical data synchronized within 30 days of blood sampling, then the successive clinical evaluations were conducted annually.

MRI scans were performed annually using a 3T General Electric Discovery MR750 scanner (GE Medical Systems, Milwaukee, WI), incorporating both precontrast and postcontrast (gadolinium) sagittal 3D T1 and fluid-attenuated inversion recovery sequences. The choroid plexus (CP) within the lateral ventricles was manually delineated on coregistered T1 postcontrast images using Java Image Manipulation 9 software (Xinapse Systems, Essex, United Kingdom) by a trained analyst (AK), at both baseline and final follow-up visits. To adjust for interindividual head size variability, CP volume was normalized (CPn) using SienaX-based scaling. The average annual change in CPn volume was calculated for each patient. Central brain atrophy (CBA) was quantified by measuring ventricular enlargement over time on T1 images, using the AI-based brain segmentation tool AssemblyNet (Coupé et al., 2020). A volume change per year metric was developed to account for variations in follow-up durations among participants. Further details on these methods are available. 19 Annual OCT scans were also performed to monitor changes in the RNFL and GCIPL thicknesses throughout the follow-up period as previously described.

Sample Preparation for Analytical Chemistry

Plasma samples were deproteinized by adding an equal volume of 10% (w/v) trichloroacetic acid (TCA; Sigma Aldrich, MA), followed by centrifuged at 12,000 ×  g for 10 minutes at 4 °C. The resulting plasma supernatants were filtering using a 0.22 PTFE syringe filter (Merck-Millipore, CA) and transferred into vials for liquid chromatography analysis. A portion of the deproteinized plasma was combined with a fixed amount of picolinic acid (PIC) and QUIN deuterated internal standards, then dried under vacuum. The samples were derivatized by adding 120 μL of trifluoroacetic anhydride and 1,1,1,3,3,3 hexafluoroisopropanol and incubating at 60°C for 1 hours. The fluorinated esters were extracted into toluene and washed with 5% sodium bicarbonate solution. The upper organic layer was collected, washed with ultra pure water type I, dried using sodium sulphate-packed pipette tips, and finally transferred to vials for gas chromatography analysis.

KP Metabolites Quantification by High Performance Liquid Chromatography

To quantify the plasma levels of TRP, KYN, 3HK, 3HAA, and anthranilic acid (AA), we used the Agilent 1,290 Infinity ultra-high performance liquid chromatography (HPLC) (Agilent, CA) equipped with both a diode array (G4212A; Agilent) and a fluorescence (G1321B; xenon flash lamp; Agilent) detectors. The procedure followed the method previously published. 20 A 20 μL sample was injected into a ZORBAX Rapid Resolution High Definition C18 column (2.1 mm × 150 mm, 1.8 μm particle size; Agilent, CA) with the separation achieved through reverse phase gradient elution at 38 °C. The mobile phase consisted of s 0.1 M sodium acetate (pH 4.6) and an isocratic flow rate was maintained at 0.75 mL/minute for a total run time of 12 minutes. Fluorescence detection was conducted at excitation/emission wavelengths of 280 nm/438 nm for TRP and 320 nm/438 nm for 3HAA and AA. KYN and 3HK were detected using ultraviolet absorbance at 365 nm (without reference signal). Data output from the chromatograms was processed using Agilent OpenLAB Chromatography Data System (CDS) ChemStation (Edition C.01.04).

For the measurement of KYNA levels in plasma, we used the Agilent 1,260 Infinity HPLC (Agilent, CA) with the fluorescence detector (G1321B xenon flash lamp, Agilent, CA).This analysis, also performed according to previously described methods26, involved injecting a 10 μL sample into a ZORBAX Rapid Resolution High Definition C18 column (4.6 mm × 100 mm, 3.5 μm particle size; Agilent, CA) under reverse-phase gradient conditions at 30°C. This analysis was performed as previouslyd 20 and involved injecting a 10 μL sample ZORBAX Rapid Resolution High Definition C18 column (4.6 × 100 mm, 3.5 μm particle size, Agilent Technologies, CA) under reverse-phase gradient at 30°C. The mobile phase comprised 95% of 0.05 M sodium acetate and 0.05 M zinc acetate (pH 5.2) and 5% HPLC-grade acetonitrile, with an isocratic flow rate set at 1 mL/minute for 8-minute runtime. The fluorescence detector was tuned to an excitation/emission wavelength of 344 nm/388 nm for KYNA. KYNA chromatograms were analyzed with Agilent OpenLAB CDS ChemStation (Edition C.01.04).

The concentrations of these metabolites were determined by comparing sample data with calibration curve, with the results expressed as μmol/L or nmol/L. Regression analysis of the obtained data was performed using Excel 2016.

KP Metabolites Quantification by Gas Chromatography Mass Spectrometry

The plasma concentrations of PIC and QUIN were measured using an Agilent 7890 gas chromatography mass spectrometry (GCMS) (Agilent, CA) with an Agilent 5,975 mass spectrometer. The procedure followed was consistent with previously established methodsd. 20 A 1 μL sample was introduced in spitless mode onto a HP-5MS gas chromatography (GC) capillary column (Agilent, CA). The analysis was performed under negative chemical ionization conditions, with simultaneous monitoring of selected ions (m/z 273 for PIC, m/z 277 for d4-PIC, m/z 467 for QUIN, and m/z 470 for d3-QUIN). The gas chromatography oven was initially held at 75°C for 3 minutes, then increased to 300°C at a rate of 25°C/minute, where it was maintained for an additional 4 minutes, resulting in a total runtime of 15.6 minutes.

Calibration curves for QUIN and PIC were developed by plotting the peak area ratios (QUIN/d3-QUIN and PIC/d4-PIC) against the concentrations of the respective standards. These curves were then used to determine the concentrations of QUIN and PIC in the samples. All spectra were analyzed, and peak areas were integrated using Agilent GC/mass selective detector (MSD) ChemStation software (Edition 02.02.1431). The data was further processed using Excel 2016 for regression analysis, with the final concentrations of QUIN and PIC reported in nmol/L.

Statistical Analysis

A χ 2 test assessed sex distribution differences, and a Wilcoxon test evaluated age distribution consistency across groups. Multiple linear regression analyzed the effects of age, sex, and treatment on KP metabolites, metabolite ratios, and brain structure. Linear regression examined the relationship between metabolite levels and age in both patients with MS and controls. Finally, a Mann–Whitney U test compared sex differences within the subdivided male and female MS group. A multiple linear regression model adjusted for age and sex following by a false discovery rate correction method were applied to control for confounding factors and multiple comparisons between the 2 groups. The data were analyzed for the presence of multiple outliers using robust regression and outlier removal (ROUT) test, and all identified outliers were excluded. Descriptive statistics, including measures of central tendency and dispersion, were reported both before and after outlier removal (eTables 1 and 2). The Kolmogorov-Smirnov test was used to assess the normality of the data distribution. The Mann-Whitney U test (for non-normally distributed data) was used to test for statistically significant differences between the 2 groups. Unpaired and two-tailed tests were used for these group comparisons. A generalized estimating equation was used to investigate the relationship between the clinical progression of MS and the levels of KP metabolites. To ensure a comprehensive analysis, we included patient age, sex, disease duration, and treatment intake as covariates in the model. The statistical analyses were performed with GraphPad Prism Software (Prism 9, GraphPad Software Inc.) and R (The R Stats Package version 4.3.3). The p < 0.05 was considered significant for all the analyses, and all the significant values were reported.

Data Availability Statement

The datasets generated for this study are available on request to the corresponding or senior authors.

Results

Demographics and Clinical Measures

The demographics and clinical measures in the participants from MS and healthy control groups are summarized in Table . The proportion of female participants was greater in both groups, represented by 71.8% and 73.5% of female participants in the healthy control and MS groups, respectively. The mean age was 36 ± 6.3 in the healthy control and 42.3 ± 10.8 years in the MS group, with a significant difference in age distribution between the groups ( p = 0.01; Table ). The average time of the disease in the MS group was 7 years, ranging from a minimum of 1 to a maximal of 28 years of disease time. The average age at the time of diagnosis for the MS-affected participants was 35.2 ± 10 years. Among the participants from the MS group, 55.1% (n = 54) had the presence of optic neuritis. The overall median EDSS for disability in the participants with MS was 1, ranging from a minimum of 0 to a maximal of 6 points. Most of the participants from the MS group (n = 87%; 88.8%) were taking medication to manage their disease with DMTs, and distributions of treatment type among the patients were listed in Table .

Table Demographics and Clinical Measures in Multiple Sclerosis (MS) and Healthy Control Groups Healthy control (n = 39) MS (n = 98) p Value Sex (female), (%) (n = 28) 71.8 (n = 72) 73.5 1 Age (y, mean [SD])ac 36 (6.3) 42.3 (10.8) 0.01 Disease time (y, median [range]) NA 7 (1–28) Age at the diagnostic (y, mean [SD]) NA 35.2 (10.0) Presence of optic neuritis (yes), (%) NA (n = 54) 55.1 EDSS score (median [range]) NA 1 (0–6) Disease-modifying treatment at baseline, (%) NA  Fingolimod (n = 21) 21.4  Natalizumab (n = 17) 17.3  Interferon beta (n = 12) 12.2  Dimethyl fumarate (n = 9) 9.2  Glatiramer acetate (n = 8) 8.2  Ocrelizumab (n = 7) 7.1  Alemtuzumab (n = 6) 6.1  Teriflunomide (n = 3) 3.1  Cladribine (n = 3) 3.1  Daclizumab (n = 1) 1  Nil (n = 11) 11.2 Abbreviations: NA = not applicable. Sex, presence of optic neuritis, and disease-modifying medications are expressed in percentage (%); age and age at the diagnostic are expressed as mean (range). Disease time and EDSS score are expressed as median (range). p values refer to the χ 2 test for sex to determine whether the male-female proportions are the same across the groups and to the Wilcoxon test for age to assess whether age distributions are consistent across the groups. The analysis was conducted using GraphPad Prism Software version 9.

Demographics and Clinical Measures in Multiple Sclerosis (MS) and Healthy Control Groups

Abbreviations: NA = not applicable.

Sex, presence of optic neuritis, and disease-modifying medications are expressed in percentage (%); age and age at the diagnostic are expressed as mean (range). Disease time and EDSS score are expressed as median (range). p values refer to the χ 2 test for sex to determine whether the male-female proportions are the same across the groups and to the Wilcoxon test for age to assess whether age distributions are consistent across the groups. The analysis was conducted using GraphPad Prism Software version 9.

Imbalanced KP Metabolites Production in MS

The differences in plasma levels of KP metabolites between MS and healthy control participants are highlighted in Figure 1 . KYNA (MS 29 ± 11.5 nM vs healthy controls 35.9 ± 17.4 nM; U = 1,281, p < 0.05, Figure 1A ) and 3HK (MS 10.6 ± 6 nM vs healthy controls 15.6 ± 9.8 nM; U = 1,222, p < 0.05, Figure 1B ) plasma levels were lower in the participants with MS compared with healthy controls. While AA (MS 81.6 ± 38.3 nM vs healthy controls 26.1 ± 13.3 nM; U = 130.5, p < 0.0001, Figure 1C ) and 3HAA (MS 34.6 ± 15.4 nM vs healthy controls 28.6 ± 21.8 nM; U = 1,326, p < 0.05, Figure 1D ), plasma levels were higher in the participants with MS compared with healthy controls. These comparisons were further adjusted for confounding factors, and multiple comparisons between the 2 groups and both, exact and adjusted p values, were reported (eTable 3). Among the variables AA ( p < 0.0001), KYN ( p < 0.0001), 3HK ( p = 0.0008), and KYNA (0.0394) were different between the groups. The remaining KP metabolites showed no significant differences between the groups and are visually represented in a schematic diagram of the KP metabolites' production (eFigure 1).

Kynurenine Pathway Metabolites Levels in the Plasma of Participants With Multiple Sclerosis (MS) and Healthy Controls

(A) Kynurenic acid, (B) 3-hydroxykynurenine (C) anthranilic acid, and (D) 3-hydroxyanthranilic acid in the plasma of MS and healthy control groups. * p < 0.05, **** p < 0.0001, unpaired 2-tailed Mann-Whitney U test. Data are median (IQR), and each dot represents individual values. This analysis was carried out using GraphPad Prism Software version 9. 3HAA = 3-hydroxyanthranilic acid; 3HK = 3-hydroxykynurenine; AA = anthranilic acid; IQR = interquartile range; KYNA = kynurenic acid; MS = multiple sclerosis.

Ratios from KP metabolites measured in the plasma of participants with MS and healthy controls were calculated ( Figure 2 ). Indirect measure of enzyme activities is inferred in the literature as ratios of products to substrates. 21 The ratio of KYN:TRP, which indirectly reflects the activity of the enzymes IDO/tryptophan 2,3 dioxygenase (TDO), was higher in the MS compared with the healthy control group (MS 4.6 ± 1 vs healthy controls 4 ± 0.8; U = 1,246, p < 0.01, Figure 2A ). The ratio of 3HK:KYN, which indirectly reflects the activity of the enzyme kynurenine 3-monooxygenase (KMO), was lower in the MS compared with the healthy control group (MS 0.6 ± 0.3 vs healthy controls 0.9 ± 0.5; U = 1,054, p < 0.001, Figure 2B ).

Ratios From Kynurenine Pathway (KP) Metabolites Measured in the Plasma of Participants With Multiple Sclerosis (MS) and Healthy Controls

(A) The KYN:TRP ratio reflects the activity of the enzymes IDO/TDO, while (B) the 3HK:KYN ratio reflects the activity of the enzyme KMO. (C) Higher QUIN:KYNA ratio has been proposed as indicator of neurotoxicity status, while (D) lower 3HAA:AA ratio has been proposed as a marker of inflammation in disease conditions. * p < 0.05, ** p < 0.01, **** p < 0.0001, unpaired two-tailed Mann-Whitney U test. Data are median (IQR), and each dot represents individual values. This analysis was performed using GraphPad Prism Software version 9. 3HAA = 3-hydroxyanthranilic acid; 3HK = 3-hydroxykynurenine; AA = anthranilic acid; IDO = indoleamine 2,3-dioxygenase; IQR = interquartile range; KMO = kynurenine 3-monooxygenase; KYN = kynurenine; KYNA = kynurenic acid; QUIN = quinolinic acid; TDO = tryptophan 2,3-dioxygenase; TRP = tryptophan.

The direction of KP metabolite production into specific pathway branches is expressed in the literature as ratios of KP intermediates and can provide insights into the balance between the metabolites and its potential implications in various physiologic and pathologic processes. Higher ratio of QUIN:KYNA has been proposed as an indicator of neurotoxicity status. 22 In this study, participants with MS presented a higher QUIN:KYNA ratio compared with the healthy controls (MS 15.6 ± 7.4 vs healthy controls 12.3 ± 5.8; U = 1,148, p < 0.05, Figure 2C ). Lower ratio of 3HAA:AA has been proposed as a marker of inflammation in disease conditions. 23 The 3HAA:AA ratio was lower in the MS group compared with the healthy control (MS 0.5 ± 0.3 vs healthy controls 0.9 ± 0.4; U = 830.5, p < 0.0001, Figure 2D ). After adjusting for confounding factors and multiple comparisons IDO/TDO (0.0187), KMO ( p < 0.0001) and 3HAA/AA ( p < 0.0001) were different between the groups (eTable 3).

Considering the observed changes in KP levels, but also the age distribution differences between groups, adjustments for age and other confounders were made. No significant correlations were found between KP metabolite levels and age, sex, or treatment type, except for a positive correlation between PIC and age in healthy controls (R 2 = 0.2; p < 0.01; eFigures 2–5). The MS group was categorized as untreated, receiving platform therapies (interferon beta, glatiramer acetate), or on high-efficacy DMTs. Multiple linear regression analysis evaluated the effects of age, sex, and treatment on KP metabolites, metabolite ratios, and brain structure, revealing no significant effects overall, except for a sex-related impact on RNFL ( p < 0.01) measurement (eTable 4).

Association Between KP Levels and Disease Progression/Neurodegeneration Parameters

Our findings indicate an association between CBA and the levels of TRP ( B = 113.97, p = 0.007; Figure 3A ), KYN ( B = −224.90, p = 0.01; Figure 3B ), as well as IDO/TDO (KYN:TRP ratio; B = −6.97, p = 0.03; Figure 3C ). Moreover, changes in CP volume, potentially indicative of neuroinflammation, 19 are linked to KYNA levels at the baseline ( B = −1.50, p = 0.006; Figure 3D ).

Correlation Between Kynurenine Pathway (KP) Metabolites With Disease Progression and Neurodegeneration Parameters in Multiple Sclerosis (MS)

A generalized linear model assessed the relationship between plasma KP metabolites and MS clinical parameters, with adjustments for patient age, sex, disease duration, and treatment intake using a generalized estimating equation. The middle line represents the regression slope, while the 2 outer lines represent the 95% CIs, providing insight into the strength and precision of the associations. The central brain atrophy (i.e., increased ventricle volume) was associated with (A) TRP, (B) KYN, and (C) KYN/TRP ratio plasma levels, while the choroid plexus volume was associated with the (D) KYNA plasma levels in the participants with MS. This analysis was performed using GraphPad Prism Software version 9. KP = kynurenine pathway; KYN = kynurenine; KYNA = kynurenic acid; MS = multiple sclerosis; TRP = tryptophan.

In addition, while the correlation observed did not reach statistical significance, the data suggest a potential relationship between retinal ganglion cell and axonal loss—considered indicative of general neurodegeneration in the disease—and various components of the KP pathway (eTable 5). Specifically, GCIPL thinning rate exhibited a borderline association with 3HK ( B = 0.06, p = 0.06; eTable 5) and KMO ( B = −0.85, p = 0.06; eTable 5), while RNFL thinning rate demonstrated a linear association with 3HAA:AA ( B = −1.38, p = 0.07; eTable 5). Therefore, those observations warrant further investigation in larger or more targeted studies. Within the clinical parameters (eTable 6), plexus change showed a positive association with ventricle volume ( B = 0.758; p < 0.001) but was negatively associated with brain volume ( B = −0.432; p = 0.04), GCIPL ( B = −0.452; p = 0.03), RNFL ( B = −0.551; p = 0.008), and tRNFL ( B = −0.587; p = 0.004).

Discussion

The existing body of research on MS pathophysiology suggests that the immune cells infiltrate into the CNS and targets the myelin sheath surrounding axons, causing progressive demyelination, which ultimately leads to axon atrophy and neuronal death. Numerous proinflammatory factors and cytokines are dysregulated in the blood, brain tissues, and CSF of patients with MS. 24 In MS, the KP is induced by proinflammatory cytokine cascades, leading to changes in the levels of neuroactive KP metabolites. Indeed, alterations in the activity of the KP and the abnormal levels of KP metabolites have been detected in MS. 25 The first study to suggest a possible involvement of the KP in MS reported that patients with MS had lower concentrations of the amino acid TRP compared with healthy controls. 26 An increasing body of evidence highlights that the KP metabolites, alongside with antigen-presenting cells, including microglia and macrophages expressing IDO1, play a key role in neuroinflammatory and neurodegenerative diseases such as MS. 25 Our results point toward an abnormal KP metabolism in stable RRMS in the absence of acute phrase or flare. Our data points toward an imbalance between the production of immunomodulatory, neuroprotective, and neurotoxic KP metabolites potentially contributing to the progression of the disease or simply reflecting it. We observed nominally lower levels of KYNA and 3HK and higher levels of AA and 3HAA in our MS cohort. Another important finding is that KP metabolites were correlated with measures of neuroinflammation, chronic demyelination, and neurodegeneration in patients with MS.

In this study, we uncovered preliminary evidence suggesting that an altered KP profile may be associated with hallmarks of MS pathophysiology. This aligns with existing evidence highlighting the key role of KP metabolism in regulating immune activation and production of neuroactive metabolites. Specifically, we found that RRMS is characterized by a higher KYN:TRP ratio than healthy controls, confirming previous studies. 25 In addition, our analysis using generalized estimating equation identified a correlation between plasma KYN:TRP ratio and ventricle volume, an indirect measure of brain atrophy and neurodegeneration. When brain tissue deteriorates because of conditions like neurodegenerative diseases or aging, it can lead to the enlargement of the brain's ventricles. Previous results showed increased levels of IDO expression in peripheral blood mononuclear cells as well as serum concentrations of neopterin in RRMS concomitant with the appearance of clinical signs and symptoms, indicating that an ongoing inflammatory response is indeed taking place in these patients. 25 Noteworthy, the serum KYN:TRP ratio decreased in patients with MS after corticosteroid treatment given the known immunosuppressive and anti-inflammatory effect of corticosteroid. 25 In addition, increased levels of KYN have previously been associated with a higher rate of signs and symptoms in RRMS patients. 27 Taken together, these findings seem to suggest an activation of KP toward the production of bioactive metabolites can be found in continued neurodegeneration in MS.

Interestingly, our result showed low KYN:TRP ratio related to brain atrophy in patients with MS with continuous neurodegeneration and no signs of acute flare. This suggests that IDO/TDO (indirectly measured by the KYN:TRP ratio) may have another role in addition to its neurodegeneration activity. Of note, increased IDO1 activity is known to suppress the T-cell–mediated response in MS through depletion of TRP 25 or activation of aryl hydrocarbon receptor. 10 Accordingly, inhibition of IDO1 in EAE mice exacerbated the disease progression. 28 In the early stage of the disease, upregulation of IDO1 may have a beneficial effect because this enzyme mediates an immunomodulatory effect in MS, which may partially explain the therapeutic effect exerted by interferons in early-mild stages of MS. 29 Of interest, our results reveal that KYN plasma levels were inversely correlated with the ventricle volume while TRP was positively correlated with the ventricle volume at the time of plasma sampling. In the context of ventricle volume and brain health, an increased IDO/TDO activity (as reflected by higher KYN production along the decrease of the substrate TRP, i.e., high KYN:TRP ratio) correlated with lower CBA, therefore, could reflect the protective immunomodulation exerted by the high IDO1 activity in the early stage of MS. The potential dual role of IDO1 being both protective and deleterious depending on the stage of the disease, highlights the complexity of its effect. Simultaneously, the correlation between these KP measures and CBA may suggest alterations in the KP resulting from brain atrophy. This necessitates further research to better understand IDO1 effect on MS pathophysiology, aiming to optimize potential therapeutic strategies that leverage its immunomodulatory capabilities effectively.

Activation of the KP leads to a production of a variety of compounds with neurotoxic or neuroprotective properties that can affect MS pathology and contribute to the severity of neurologic impairment. Among the human plasma KP downstream metabolites that were measured in our study, KYNA was significantly reduced in MS. KYNA can reduce overexcitation of the neurons and overproduction of nitric oxide free radical by antagonizing QUIN-mediated excitotoxicity at NMDA receptor. 9 In MS, decreased neuroprotective metabolites such as KYNA and PIC have been associated with the loss of oligodendrocytes, astrocytes, and neurons. 30 In our recent in vivo study, we showed that the KP was skewed toward the production of QUIN as indicated by 4 times increased concentration of QUIN in the EAE cohort (2.625 ± 0.406 nM/g) as compared with healthy animals (0.662 ± 0.11 nM/g) and correlates with the disease severity. 8 Although QUIN levels were not significantly different between the groups in this study, an increased QUIN:KYNA ratio was observed in the MS cohort, suggesting an excess QUIN concentration as compared with the neuroprotective KYNA. Hence, the ratio of QUIN to KYNA defines the overall glutamatergic activity at the NMDA receptor, determining whether QUIN-mediated excitotoxicity leads to neurodegeneration. 31 This is further supported by our previous study showing a decline in KYNA to a subnormal concentration while synthesis of QUIN elevated as the disease progresses in MS patients. 12

In addition, KYNA exhibit immunomodulatory properties through activation of the aryl hydrocarbon receptor, the expansion of regulatory T cells 32 and, therefore, can regulate the activity of immune cells and influence the balance between proinflammatory and antiinflammatory responses in the CNS. 10 Animal studies suggest that KYNA can mitigate autoimmunity by suppressing the activity of Th1 and Th17 cells, reducing proinflammatory cytokines production and enhancing regulatory T-cell response in EAE models. 33 Previous analyses of KYNA levels in patients with MS reveal conflicting results. 11 Lower blood KYNA levels have been found in RRMS patients in remission, 34 whereas increased levels have been found in acute relapse MS. 35 Rejdak and coworkers reported elevated CSF levels of KYNA during relapse and lower levels during remission in patients with RRMS. 34 , 35 In this study, we demonstrated decreased plasma levels of KYNA in patients with RRMS in continued neurodegeneration despite the absence of acute phrase or flare. In general, activation of the KP seems to provide a short-term benefit in response to an increased inflammatory state by establishing an immunosuppressant feedback loop.

Another finding that stands out from the results reported earlier is that in the present study plasma KYNA was negatively correlated with the change in the CP volume, potentially indicative of neuroinflammation. Recent research have shown that the CPs are enlarged and inflamed in patients with MS, particularly in those with RRMS characterized by inflammatory profiles. 36 Interestingly, the authors identified in the same cohort a correlation between enlargement of CPs with white matter lesions and with annualized relapse rate in RRMS patients. 36 Our findings provide support for neuroprotective and antiinflammatory properties of KYNA. Most striking our data point out blood KYNA levels as an indication of change in the CP volume and, therefore, as an indirect measure of neuroinflammation.

Another KP metabolite, 3HK, is known to potentiate QUIN-induced excitotoxicity 37 and increased levels of 3HK in tissues from EAE rat model of MS was also observed, 38 which might be relevant to the neurodegenerative process in MS. Our results did not confirm the increased 3HK levels and KMO activity in patients with MS. By contrast, our analysis showed that the proportions of plasma levels of 3HK and 3HK:KYN ratio (indirect measure of KMO activity) tended to be lower in patients with MS compared with healthy controls, and this may be related to the phase of MS of our cohort because all participants had absence of recent episodes of flare-ups within the past 6 months. Nevertheless, our previous in vivo study provided strong evidence that inhibiting KMO, can prevent the accumulation of QUIN, and improve functional disability in the experimental autoimmune encephalomyelitis mouse model. 39 In support, this cross-sectional study revealed a borderline correlation between KMO activity and 3HK with neuroaxonal and functional loss in the visual system, that is GCIPL thinning rate, suggesting that the increased neurotoxicity branch metabolization that has previously been described in patients with MS and EAE 8 , 31 , 40 models are linked to neurodegeneration in the disease.

The observed reduction of 3HK in MS can be also reminiscent of reduced KMO activity because of vitamin B2 and/or B6 deficiency 41 or negative regulation by IDO1 and cytokines. 42 Although vitamin B2 and B6 status were not measured in this study, the high KYN:TRP ratio (as a proxy of IDO activity) in patients with MS supports the later hypothesis that IDO in an inflammatory context could reduce the KMO activity. We further hypothesized that 3HK reduction might be also related to a deviation toward increased AA and 3HAA production in our MS cohort, that is, a spillover that drains KYN toward AA and 3HAA production. In agreement, low KMO activity have been correlated with elevated AA serum levels, while reduced level of KMO activity and vitamin B6 were correlated with low 3HAA and a reduced ratio of 3HAA:AA in a cohort of Sclerodermatous/fibrosing chronic graft vs host disease, an immunologic complication after allogeneic stem cell transplantation. 41

Interestingly, AA levels has been observed to elevate in the blood across various human diseases, potentially reflecting a compensatory “clean-up” mechanism after an acute injury. 23 AA has also been associated with antagonizing neurotoxic KP metabolites, such as QUIN and contributing to the reduction of oxidative stress and inflammatory responses. 23 , 43 There is accumulating evidence supporting the overall protective effect of immunomodulatory KP metabolites, e . g ., AA and 3HAA, in EAE models and patients with MS, which has somewhat shifted diverted attention from their potential detrimental roles in MS. 44 , 45 Studies have shown that the synthetic analog of AA, tranilast (N-[30,40 -dimethoxycinnamoyl]-AA), can suppress CD4 + and CD8 + T-cell differentiation and, therefore, possess a protective effect on MS progression/symptoms/lesions. 44 Darlington et al. 23 reported a reduced plasma 3HAA:AA ratio in various inflammatory diseases, suggesting that this ratio may either indicate an inflammatory state or represent an antiinflammatory response. However, it remains unclear whether the increased production of AA and 3HAA is causally linked to MS progression or represents a compensatory mechanism aimed at mitigating the detrimental inflammatory response associated with MS. The RNFL thinning rate inversely correlated with 3HAA:AA ratio in patients does not support the former; furthermore, there was no increased axonal or ganglion loss associated with AA or 3HAA levels. However, because correlation does not imply causality, it cannot be ruled out that the observed KP modifications are a consequence of the neurodegeneration inherent to MS pathology. While neuroinflammation in MS may cause changes in the KP that contribute to neurodegeneration, alterations in the KP may also result from structural changes in the brain.

This study included only patients with relapsing remitting phase of the disease without flare-ups. This prevents us from excluding the possibility that the regulation of KP metabolism may evolve throughout the disease course and can be related to the stage and severity of the disease. Indeed, it is well known that KP profile in MS is associated with the disease phase and/or clinical subtype. 12 , 40 In addition, the inclusion of primary and secondary progressive MS with a more severe disease condition would allow a better understanding of the participation of KP in the pathophysiology of MS. Nevertheless, confounding by treatment in the group of patients receiving DMT and corticosteroids cannot be entirely ruled out, 25 , 29 considering that significant changes in blood KP metabolites have been reported between patients with MS without treatment compared with patients with MS treated with IFN 46 or corticosteroids. 25 Another limitation of this study is the absence of direct measurements of proinflammatory cytokines, which would provide a clearer understanding of the peripheral inflammatory status. Future research should aim to include cytokine assays to better elucidate the inflammatory processes at play. In addition, the participants were not fasting, which can potentially affect the KP metabolites levels. 47 Finally, 1 potential limitation of using blood-based biomarker for neurologic disease is that blood parameters may not always accurately reflect conditions within the CNS. However, a previous study demonstrated that the changes in the blood KP profile closely correspond to alterations in CNS KP profile. 48 In addition, our previous research in Alzheimer disease found a positive correlation between CNS and plasma KP metabolites, 49 supporting plasma KP metabolite measurement as an indirect indicator of CNS KP activity. Because TRP, KYN, and 3HK cross the blood-brain barrier (BBB), changes in their blood levels can affect brain concentrations and influence the metabolism of other KP metabolites, such as KYNA, 3HAA, and QUIN, which do not easily cross the BBB. 50 Further studies should explore these and other peripheral metabolites to improve the accuracy of blood-based biomarkers for MS. Altogether, these limitations should be taken into account and addressed in further studies.

In this study, we report the first characterization of blood KP metabolites in patients with MS in relation to measures of neuroinflammation and neurodegeneration, including chronic neuroaxonal damage, functional loss. In conclusion, although preliminary, our findings suggest that an imbalanced KP metabolism may be associated with MS pathology, potentially reflecting disease activity and severity. Further studies in larger and more diverse cohorts are required to validate these observations. Our study reports specific deviations of the KP in MS, whereof lower levels of KYNA and 3HK and higher levels of AA and 3HAA were identified in the plasma from patients with MS. It was identified a negative correlation between plasma KYN:TRP ratio and ventricule volume, an indirect measure of brain atrophy and neurodegeneration. In addition, KYN plasma levels were negatively correlated with the ventricle volume while TRP was positively correlated with the ventricle volume. Finally, KYNA plasma levels were negatively correlated with the CP volume, an indicative measure of neuroinflammation. The KP metabolite signature identified in this study represents a potential biomarker for neuroinflammation and neurodegeneration in MS, offering a valuable tool for monitoring ( Figure 4 ) MS progression in clinical trials. From a translational perspective, these findings support the rationale for evaluating vitamin B2/B6 supplementation in future clinical trials. Our findings suggest that KP changes in the blood may serve as a useful biomarker for assessing chronic neurodegeneration and monitoring disease progression in MS.

Imbalance of Kynurenine Pathway Metabolites Production Can Contribute to Chronic Neurodegeneration in Multiple Sclerosis

(1) Simplified representation of the kynurenine pathway highlighting the metabolites that were increased (red) or decreased (green) in the plasma of participants with MS of the present study. (2) Possible mechanisms by which prolonged dysregulation of circulatory kynurenine pathway metabolites can trigger progressive demyelination of axons ultimately resulting in axon atrophy and death of the neurons; (2a) direct effects of neuroactive metabolites that can cross the BBB and exert direct toxicity toward neurons and oligodendrocytes cells; (2b) indirect effect through immunomodulatory properties of circulatory kynurenine pathway metabolites that can regulate the activity of immune cells and influence the balance between proinflammatory and antiinflammatory responses in the CNS and ultimately trigger demyelination and axon loss. (3) Alteration of BBB integrity in MS allow the trafficking and infiltrating of immune cells, particularly T lymphocytes autoreactive against CNS antigens. (4) These cells become activated and enter the CNS to attack the myelin sheath, releasing proinflammatory cytokines. In response, infiltrated macrophages and resident microglial cells are activated, amplifying neuroinflammation. (5) This orchestrated immune attack drives the production of toxic mediators, ultimately resulting in damage to the myelin sheath and the formation of inflammatory plaques in the CNS. Over time, this damage can lead to the loss of axons, contributing to neurodegeneration and the progressive neurological disability in MS. 3HAA = 3-hydroxyanthranilic acid; 3HK = 3-hydroxykynurenine; AA = anthranilic acid; BBB = blood-brain barrier; CNS = central nervous system: IDO = indoleamine 2,3-dioxygenase; KATs = kynurenine aminotransferase; KMO = kynurenine 3-monooxygenase; KYN = kynurenine; KYNA = kynurenic acid; KYNU = kynureninase; PIC = picolinic acid; QUIN = quinolinic acid; TRP = tryptophan.

Collectively, our studies, and studies by others, provide compelling evidence that the KP is dysregulated during MS progression, and this pathway may represent a therapeutic target to reduce MS progression and symptomology, while blood levels of KP metabolites can potentially be used as a less invasive indicators of MS severity and progression such as chronic neuroaxonal damage, functional loss, neuroinflammation, and neurodegeneration status. It is important to note that the KP is highly complex, with multiple metabolites and interactions with other pathways. Its exact role in MS likely involves a combination of factors and interactions with the immune system and other pathways, more research is needed to fully understand its role and profile.

Glossary

anthranilic acid

blood-brain barrier

central brain atrophy

choroid plexus

CP volume was normalised

disease-modifying therapy

encephalitis

ganglion cell inner plexiform layer

gas chromatography mass spectrometry

high performance liquid chromatography

indoleamine 2,3 dioxygenase

interferon gamma

kynurenine pathway

kynurenine

kynurenic acid

multiple sclerosis

optical coherence tomography

picolinic acid

quinolinic acid

retinal nerve fiber layer

relapsing remitting MS

tryptophan 2,3 dioxygenase

tryptophan

Author Contributions

A. Staats Pires: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. S. Krishnamurthy: major role in the acquisition of data; analysis or interpretation of data. S. Sharma: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. S. Chow: major role in the acquisition of data; analysis or interpretation of data. S. Klistorner: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. G.J. Guillemin: drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data. A. Klistorner: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. Y. You: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. B. Heng: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data.

Study Funding

This study was supported by the National Multiple Sclerosis Society.

Disclosure

The authors report no relevant disclosures. Go to Neurology.org/N for full disclosures

References

Associated Data

Data Availability Statement

The datasets generated for this study are available on request to the corresponding or senior authors.