Article Text

Download PDFPDF

Research paper
Identification of novel biomarkers for Parkinson's disease by metabolomic technologies
  1. Taku Hatano1,
  2. Shinji Saiki1,
  3. Ayami Okuzumi1,
  4. Robert P Mohney2,
  5. Nobutaka Hattori1,3
  1. 1Department of Neurology, Juntendo University School of Medicine, Tokyo, Japan
  2. 2Metabolon Inc., Durham, North Carolina, USA
  3. 3Core Research for Evolutionary Science and Technology (CREST), Japan Science and Technology Agency, Tokyo, Japan
  1. Correspondence to Dr Nobutaka Hattori, Department of Neurology, Juntendo University School of Medicine, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113-8421, Japan; nhattori{at}juntendo.ac.jp

Abstract

Objective The pathogenesis of Parkinson's disease (PD) involves complex interactions between environmental and genetic factors. Metabolomics can shed light on alterations in metabolic pathways in many diseases, including neurodegenerative diseases. In the present study, we attempted to elucidate the candidate metabolic pathway(s) associated with PD.

Methods Serum samples were collected from 35 individuals with idiopathic PD without dementia and 15 healthy age-matched control participants without PD. This analysis used a combination of three independent platforms: ultrahigh-performance liquid chromatography/tandem mass spectrometry (UPLC/MS/MS) optimised for basic species, UPLC/MS/MS optimised for acidic species and gas chromatography/MS (GC/MS).

Results The metabolomic profiles of PD were clearly different from normal controls. PD profiles had significantly lower levels of tryptophan, caffeine and its metabolites, bilirubin and ergothioneine, and significantly higher levels of levodopa metabolites and biliverdin than those of normal controls. Alterations in the bilirubin/biliverdin ratio and ergothioneine can indicate oxidative stress intensity and may suggest elevated oxidative stress and/or insufficient ability for scavenging free radicals, which could contribute to PD pathogenesis. Decreased serum tryptophan level is associated with psychiatric problems in PD. A decrease in serum caffeine levels is consistent with an inverse association of caffeine consumption with development of PD based on past epidemiological studies.

Conclusions Metabolomic analysis detected biomarkers associated with PD pathogenesis and disease progression. Since critical metabolic biomarkers need to be identified in PD, future studies should include assay validation and replication in independent cohorts.

  • PARKINSON'S DISEASE
  • MOVEMENT DISORDERS
View Full Text

Statistics from Altmetric.com

Introduction

The pathological hallmarks of Parkinson's disease (PD) are the marked loss of dopaminergic neurons in the substantia nigra pars compacta (SNc), causing the depletion of dopamine in the striatum, as well as the presence of intracytoplasmic inclusions known as Lewy bodies. Neuronal loss and the formation of Lewy bodies have been detected not only in the SNc, but also in the locus coeruleus, pedunculopontine nucleus, raphe nucleus, dorsal motor nucleus of the vagal nerve, olfactory bulb, parasympathetic and sympathetic postganglionic neurons, Meynert nucleus, amygdaloid nucleus and cerebral cortex. Thus, PD is recognised as a multicentric disorder. Widespread neurodegeneration is responsible for the various motor and non-motor symptoms of PD. Although the pathomechanisms underlying neuronal degeneration in PD remain unknown, various PD-related genetic-environmental interactions may contribute to the pathogenesis of this disease.1 ,2 Previous studies revealed the diagnostic value of measuring α-synuclein3 ,4 and DJ-15 levels in cerebrospinal fluid; however, useful biomarkers related to environmental factors have not as yet been elucidated. The recent development of several ‘omics’ technologies has allowed for the profiling of biospecimens for ageing and/or disease-specific genetic markers, proteins and small-molecular-weight (<1.5 kD) molecules.6 Serum/plasma metabolomics is a useful tool for understanding metabolic pathways and networks in neurodegenerative diseases.7 Numerous non-motor symptoms caused by the dysfunction of cardiovascular, autonomic and central nervous systems may influence some of the symptoms in PD.8 Furthermore, environmental factors and genetic susceptibility may impact not only on the whole central nervous system, but also on systemic organs. In this context, metabolomic analysis would be beneficial in determining perturbations in systemic metabolic pathways in PD. Previous studies identified PD-specific alterations in several metabolic pathways, including polyamine metabolism,9 purine pathway,10 pyruvate pathway11 and redox markers.12 ,13 In this study, we attempted to identify candidate metabolic pathway(s) associated with the pathomechanisms of PD.

Methods

Participants

The target population was patients with PD without dementia. The participants consisted of 35 individuals with idiopathic PD: 18 men, with a mean age of 69.1±10.8 years, average disease duration 8.82±4.3 years (1–18 years) and mean Hoehn-Yahr (HY) 2.9±1.1; healthy age-matched and control participants without PD: 7 men, aged 70.7±9.7 years. PD and control participants’ characteristics are shown in online supplementary table S1. This includes the type of drug treatments received, levodopa dose, levodopa equivalent dose, autonomic symptoms, psychiatric and cognitive symptoms, body weight, consumption of caffeine and smoking status. The levodopa equivalent dose was calculated based on a previously reported systematic review.14 All participants with PD had been treated as outpatients at the Juntendo University in Tokyo, Japan. Movement disorder specialists performed clinical assessments and all patients with PD met the UK PD Society Brain Bank criteria. Control participants were the patients’ spouses and volunteers who were free of neurological and psychiatric illnesses. Data were collected between February 2013 and May 2013. This study was approved by the Ethics Committee of the Juntendo University School of Medicine and all participants provided informed consent.

Sample preparation and metabolite analysis

Venous blood samples for laboratory analysis were collected between 10:00 and 12:00, following approximately 4 h of fasting and the non-consumption of any medication. Participants were categorised into three groups according to disease severity: mild, HY 1 and 2 (n=14); moderate, HY 3 (n=11); severe, HY 4 and 5 (n=10). Serum samples were collected and stored at –80°C until processing at Metabolon Inc, using a standard methanol extraction as previously described.15 The resulting extract was divided into fractions for analysis by ultrahigh-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) optimised for basic species, UPLC-MS/MS optimised for acidic species and gas chromatography-MS (GC-MS).15 ,16 For UPLC-MS analysis, extracts were gradient eluted from a C18 column using water and methanol containing either 0.1% formic acid (acidic conditions) or 6.5 mM ammonium bicarbonate (basic conditions). The MS analysis alternated between MS and data-dependent MS/MS scans using dynamic exclusion, and the scan range was from 80 to 1000 m/z. For GC/MS, samples were derivatised under nitrogen using bistrimethyl-silyltrifluoroacetamide and separated on a 5% diphenyl/95% dimethyl polysiloxane fused silica column. The scan range was from 50 to 750 m/z. Ion peaks were matched to standards in a reference library for structural identification, and relative levels were quantitated.

Statistics

The significance of changes was analysed by Welch's t test (control vs PD) or analysis of variance (PD progression groups), with p<0.05 deemed to be significant. Multiple comparisons for this data set were also accounted for with the false discovery rate method, and each false discovery rate was estimated using q values, where a low q value (q<0.10) was an indication of high confidence in a result. Although a higher q value indicated diminished confidence, it did not necessarily rule out the significance of the result. To determine whether differences in metabolite abundance between two study groups could be efficiently exploited to build a sensitive biosignature of the PD status, we applied random forest (RF) classification. RF created a set of classification trees based on continual sampling of the experimental units and compounds. RF using named metabolites in the serum of healthy individuals compared with patients with PD gave a predictive accuracy of 98%.

Results

In this cohort, online supplementary table S1 shows the characteristics of the participants. Compared with control participants, patients with PD were less likely to smoke. Patients with PD tended to have more polypharmacy than control participants. There were no significant differences in age and body weight between patients with PD and control participants. The average height of the ‘HY 1 and 2’ groups of patients with PD was significantly more than those of the ‘HY 3’, and ‘HY 4 and 5’ groups. Compared with the ‘HY 1 and 2’ groups, the ‘HY 4 and 5’ groups were significantly older and received higher doses of antiparkinsonian drugs (see online supplementary table S1).

We analysed a total of 434 metabolites whose analysis methods had already been well established in patients with PD and controls (see online supplementary tables S3 and S4). We also determined whether changes in serum levels of the target biochemicals were dependent on disease severity (see online supplementary table S2). Welch's two-sample t test and RF analysis (a supervised classification technique that can aid in the identification of biomarkers that differentiate classification groups) suggested key differences in monoamine metabolism, caffeine metabolism and redox homoeostasis between controls and patients with PD.

Monoamine metabolism

Multiple metabolites in the dopamine degradation pathway were among the top 30 biochemical species in the RF importance plot (see online supplementary table S2). Serum concentrations of 3-methoxytyrosine, 3-methoxytyramine sulfate, homovanillate (HVA) and HVA sulfate were significantly higher in patients with PD than in controls (figure 1A). In addition, the levels of 3-methoxytyrosine tended to correlate with the dose of L-dopa in PD (r=0.33, p=0.051; figure 1C). Whereas the levels of tyrosine, which is the levodopa and dopamine precursor amino acid, were similar between the groups, the dopamine pathway metabolites were significantly higher in patients with PD than in controls (figure 1A). Serum levels of tryptophan were slightly lower in patients with PD with psychiatric issues, including depression, panic attack and anxiety, than in controls (p=0.056; figure 1A, B, D). Tryptophan is metabolised by at least two major pathways, the serotonin and kynurenine pathways. Serum kynurenine levels were normal in patients with PD, whereas serotonin levels were significantly lower in the moderate and severe groups than in the mild group (figure 1A, B).

Figure 1

Alterations in tyrosine and tryptophan metabolites in patients with PD. (A) Data are expressed as ratios with controls or disease severity; HY 1/2, HY 3, HY 4/5, as the denominator. Significance (p<0.05) is indicated by the colour of the cell with yellow text (a red shaded cell indicates that the mean values are significantly higher for that comparison and green signifies a significant decrease). Trends in metabolite levels are indicated by colours (light green indicates lower than the control and light red indicates higher than the control). Significantly higher serum concentrations of 3-methoxytyrosine, 3-methoxytyramine sulfate, HVA and HVA sulfate were detected in patients with PD than in controls. Whereas the levels of the levodopa/dopamine precursor amino acid tyrosine were similar between the groups, the dopamine pathway metabolites were significantly higher in patients with PD than in controls. The serum levels of tryptophan were lower in patients with PD than in controls (p=0.0562). Serum kynurenine levels were normal, whereas serotonin levels were significantly lower in the moderate and severe groups than in the mild group. (B) Tryptophan is metabolised by at least two major pathways, the serotonin and kynurenine pathways. Metabolic alterations were detected in the serotonin pathway in patients with PD. (C) Linear correlation tests revealed a significant positive correlation between the levels of 3-methoxytyrosine and the dose of L-dopa in PD (r=0.33, p=0.051). (D) Serum tryptophan levels tended to be lower in patients with PD with depression, panic attack and anxiety than in control subjects (p=0.056). Significance (p<0.05) is indicated by an asterisk (green indicates lower than the control and red indicates higher than the control). HY, Hoehn-Yahr; HAV, homovanillate; FC, fold of control; CTRL, control; PD, Parkinson's disease; 1, 2, HY 1 and 2; 3, HY 3; 4, 5, HY 4 and 5; PSY, psychosis including depression, panic attack and anxiety.

Hypoxanthine metabolism

Caffeine and caffeine metabolites were significantly lower in patients with PD than in controls (figure 2A, B). Caffeine is metabolised into three primary metabolites, theophylline (1,3-dimethylxanthine), theobromine (3,7-dimethylxanthine) and paraxanthine (1,7-dimethylxanthine). Caffeine, paraxanthine and paraxanthine metabolite (1,7-dimethylurate and 5-acetylamino-6-amino-3-methyluracil) levels were significantly lower in patients with PD than in controls (figure 2A, B). Theophylline levels were significantly higher in patients with PD (fold of control; 1.78, p=0.01, q=0.228). Although markedly elevated levels of theophylline were detected in one male patient with PD, it was discovered that he had been treated for asthma with theophylline (see online supplementary table S1, figure 2B). Therefore, the outlier patient was removed from the statistical test for theophylline. We found that the levels of theophylline were significantly lower in patients with PD without the outlier (fold of control; 0.51, p=0.001, q=0.086). However, removal of the outlier had little impact on both the association between serum theophylline levels and disease progression and other metabolic pathways (data not shown).

Figure 2

Alterations in xanthine metabolites in patients with PD. (A) Data are expressed as ratios with controls or disease severity; HY 1/2, HY 3, HY 4/5, as the denominator. Significance (p<0.05) is indicated by the colour of the cell with yellow text (a red shaded cell indicates that the mean values are significantly higher for that comparison and green signifies a significant decrease). Trends in metabolite levels are indicated by colours (light green indicates lower than the control and light red indicates higher than the control). Caffeine and caffeine metabolites were significantly lower in patients with PD than in controls. (B) Caffeine is metabolised into three primary metabolites, theophylline, theobromine and paraxanthine. Significantly lower levels of caffeine, paraxanthine and paraxanthine metabolites (1,7-dimethylurate and 5-acetylamino-6-amino-3-methyluracil) were detected in patients with PD than in controls. Significance (p<0.05) is indicated by an asterisk (green indicates lower than the control and red indicates higher than the control). HY, Hoehn-Yahr; FC, fold of control; CTRL, control; PD, Parkinson's disease; AAMU, 5-acetylamino-6-amino-3-methyluracil.

Redox homoeostasis

Among the metabolites involved in redox homoeostasis, ergothioneine ranked high in the RF results (see online supplementary table S2). Serum levels of ergothioneine were significantly lower in patients with PD than in controls. In contrast, the level of biliverdin, the heme metabolite, was significantly higher, while its metabolite, the antioxidant bilirubin (the Z,Z isomer), was significantly lower in patients with PD than in controls (figure 3A, B). Serum levels of biliverdin tended to be higher in the severe PD group than in the mild and moderate PD groups (p=0.07 and p=0.055, respectively). Serum levels of bilirubin (the Z,Z isomer) decreased significantly based on the progression of disease severity (figure 3A, B). Given that biliverdin is regenerated by reactive oxygen species (ROS) reactions with bilirubin (figure 3B), a bilirubin/biliverdin ratio may be a measure of the ROS level. The bilirubin/biliverdin ratio was significantly lower in patients with PD than in controls, and correlated significantly with disease severity (figure 3B).

Figure 3

Alterations in redox metabolites in patients with PD. (A) Data are expressed as ratios with controls or disease severity; HY 1/2, HY 3, HY 4/5, as the denominator. Significance (p<0.05) is indicated by the colour of the cell with yellow text (a red shaded cell indicates that the mean values are significantly higher for that comparison and green signifies a significant decrease). Trends in metabolite levels are indicated by colours (light green indicates lower than the control and light red indicates higher than the control). Serum levels of ergothioneine were significantly lower in patients with PD than in controls. In contrast, biliverdin, the heme metabolite, was significantly higher, while its metabolite, the antioxidant bilirubin (the Z,Z isomer), was slightly lower in patients with PD than in controls. (B) Biliverdin is regenerated by reactive oxygen species reactions with bilirubin. Significance (p<0.05) is indicated by an asterisk (green indicates lower than the control and red indicates higher than the control). The bilirubin/biliverdin ratio was lower in patients with PD than in controls and significantly increased based on the progression of disease severity. HY, Hoehn-Yahr; FC, fold of control; CTRL, control; PD, Parkinson's disease; 1, 2, HY 1 and 2; 3, HY 3; 4, 5, HY 4 and 5.

Other metabolites

Among the metabolites involved in the urea cycle, ornithine ranked high in the RF analysis. Serum levels of ornithine were significantly higher in patients with PD than in controls (see online supplementary tables S3 and S4). Increased levels were also specifically detected in the advanced PD group. Although the arginases catalyse the divalent cation-dependent hydrolysis of arginine to produce ornithine and urea, the levels of arginine were not significantly different between patients with PD and controls. Ornithine is converted to polyamines, including putrescine, spermidine and spermine via ornithine decarboxylase. Serum levels of N-acetylisoputreanine-γ-lactam, which may potentially be formed from N1-acetylspermidine, were significantly decreased in patients with PD. We also identified significant alterations in lysolipids, eicosapentaenoate, catechol sulfate (an end product of the aromatic compound metabolism) and sugar metabolism, such as glucose, mannose, mannitol and sucrose (see online supplementary tables S2, S3 and S4).

Discussion

In this study, the metabolomic analysis of serum samples revealed marked differences in several metabolic pathways between patients with PD and healthy control participants. We identified several metabolites specific for PD as the top 30 ranking biochemical markers in the RF importance plot (see online supplementary table S2), which is consistent with previous reports suggesting that catecholamine metabolism,13 caffeine metabolism10 and redox metabolism13 are affected in PD.

Catecholamine metabolism

Significantly elevated levels of monoamine metabolites were detected in PD samples, but all study participants with PD had received levodopa with a dopamine carboxylase inhibitor. Although 3-methoxytyrosine, 3-methoxytyramine sulfate, HVA and HVA sulfate, which are all located downstream of the levodopa metabolites, were significantly increased in patients with PD, the levels of tyrosine, which is a levodopa and dopamine precursor amino acid, were similar between the groups. These results are consistent with the effects of levodopa medication. Thus, monoamine metabolite analysis among patients with PD with and without L-dopa medication should be performed to assess the precise difference in the metabolic activity.

Serum levels of tryptophan were also found to tend to be lower in patients with PD with depression, panic attack and anxiety than in controls (p=0.056, figure 1). Molina et al17 reported a decrease in plasma tryptophan levels in PD, which is consistent with the results of this study. Previous studies have suggested that serum levels of tryptophan may be associated with depression.18–20 In this context, decreased serum tryptophan levels may be related to the psychiatric symptoms of PD. We also showed that the decrease in tryptophan levels may be linked to differences in the serotonin pathway, but not in the kynurenine pathway. Serum serotonin levels decreased according to disease severity (figure 1). A previous report indicated that lower levels of cerebrospinal fluid 5-hydroxyindolacetic acid, a main metabolite of serotonin, may contribute to suicide attempts and depression.21 However, it currently remains unclear whether low plasma serotonin levels are actually associated with psychotic symptoms in PD. In the body, serotonin plays a role in gastrointestinal regulation and is a modulator of blood vessel tone.20 Taken together, these findings suggest that there is some degree of autonomic dysfunction in advanced PD.

Caffeine and xanthine metabolism

Serum levels of caffeine and caffeine metabolites were lower in PD than in controls (figure 2). Considering the low level of caffeine itself, this pattern may reflect lower intake and/or lower absorption of substances with caffeine. Consistent with these results, several studies have described an inverse association between caffeine consumption and the risk of PD.22–24 In our cohort, the level of caffeine consumption in patients with PD did not differ significantly from that in control participants (see online supplementary table S1). Therefore, PD might be associated not so much with the level of consumption of caffeine as lower serum levels due to differences in absorption. One potential explanation for the apparently protective effects of caffeine is that it acts as an inhibitor of adenosine A2A receptors. In a 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine-treated PD animal model, caffeine attenuated dopaminergic neuronal degeneration owing to the blockade of adenosine A2A receptors.25 However, no significant differences were observed in caffeine metabolite levels with disease staging (figure 2A). Therefore, serum levels of caffeine may be associated with disease onset, but not disease progression. Interestingly, plasma caffeine levels have also been found to be reduced in patients with amyotrophic lateral sclerosis compared with healthy controls.26 In this context, serum levels of caffeine may be associated with the pathomechanisms of neurodegenerative disease. Johansen et al10 reported an alteration in serum levels of hypoxanthine in PD, and described an association between the uric acid metabolic pathway and the development of PD. Our results support the hypothesis that xanthine and caffeine metabolites may reduce the risk of developing PD.

Redox metabolism

We also found that the bilirubin/biliverdin ratio was significantly lower in patients with PD than in controls (figure 3). Since biliverdin is generated by ROS reactions with bilirubin, it serves as an indicator of ROS levels.27 Roede et al9 suggested that serum levels of bilirubin may be related to disease progression in sporadic PD, which is consistent with our results. Interestingly, heme oxygenases (HO), which are responsible for the degradation of heme to bilirubin/biliverdin, free iron and carbon monoxide, may contribute to the pathomechanisms of neurodegenerative diseases.27 In PD, HO-1 was found to be markedly upregulated in the astrocytes of the SNc and was immunohistochemically identified as a component of Lewy bodies.28 Therefore, a decrement in the serum bilirubin/biliverdin ratio may imply the upregulation of HO due to systemic oxidative stress. We also showed that ergothioneine levels were significantly lower in patients with PD. Ergothioneine is a diet-derived, slowly metabolised antioxidant. It tends to accumulate in tissues over time and protects against tissue damage by decreasing lipid peroxidation levels via the catabolism of s-nitrosoglutathione.29 Yang et al30 recently reported the protective effects of ergothioneine against the accumulation of amyloid-β protein in a mouse model. These findings suggest that the decrease in serum ergothioneine levels may be associated with the development of neurodegenerative disorders. Alterations in redox markers indicate that patients with PD may be exposed to systemic oxidative stress by several environmental factors, such as agricultural chemicals, pesticides and chemical agents, which is consistent with findings from previous epidemiological studies.31 Moreover, the alteration of redox metabolites may indicate insufficient abilities of antioxidant mechanisms in patients with PD to scavenge ROS. Dopamine is a highly reactive molecule that is prone to oxidation and contributes to the generation of ROS. Many studies describe the association between mitochondrial dysfunction due to ROS and the development of PD.32 ROS has also been linked to the aggregation of α-synuclein, which contributes to the formation of Lewy bodies in brains with PD.1 ,2 Our results support the concept that perturbation of systemic redox homoeostasis is evident in dopaminergic neurodegeneration in PD.

Other metabolites

Significantly elevated serum levels of ornithine were observed in patients with PD. Hare et al33 also reported elevations in plasma ornithine levels in PD that could be reversed with levodopa. Manyam et al34 showed that cerebrospinal fluid ornithine levels were significantly increased in both de novo and treated PD. The main biosynthetic route of ornithine is by arginase activity on arginine, which also results in the formation of urea. Although it remains unclear how elevated serum levels of ornithine are associated with the pathomechanisms of PD, urea pathways may be perturbed in PD.33 ,34 We also identified decreased serum levels of N-acetylisoputreanine-γ-lactam, derived from ornithine via ornithine decarboxylase. Roede et al9 reported elevations in serum levels of N8-acetylspermidine, which is also a product that is derived from ornithine, with the rapid progression of PD. Both N8-acetylspermidine and N-acetylisoputreanine-γ-lactam are polyamines that play several essential roles in many cellular functions. Polyamines influence gene expression, proteins and membranes, and have been shown to function as antioxidants and scavengers of ROS.35–37 Therefore, ornithine and its metabolites may participate in the pathomechanisms of PD. Consistent with our results, previous studies suggested that the polyamine pathway may participate in the aetiology and pathology of neuropsychiatric diseases, such as schizophrenia, mood disorders, anxiety and Alzheimer disease.37 ,38

Catechol sulfate was among the RF importance plot metabolites and its serum levels were lower in patients with PD than in controls. Catechol sulfate is an end product of the metabolism of aromatic compounds, which involves the combined activity of gut microbial metabolism and liver and kidney functions. Gastrointestinal dysfunctions, such as constipation and dysphagia, are commonly reported in PD. Gabrielli et al39 recently showed that small intestinal bacterial overgrowth was highly prevalent in PD. Therefore, decreases in serum catechol sulfate levels in patients with PD may be attributed to abnormal gut microbiota.

Eicosapentaenoate, a polyunsaturated fatty acid, was among the top RF importance plot metabolites, and its serum levels were lower in patients with PD than in controls. Several investigations described the association between polyunsaturated fatty acids and the aggregation of α-synuclein.40 The alteration of eicosapentaenoate may be associated with the aggregation propensity of α-synuclein in PD.

Conclusions

We attempted to confirm and establish promising PD-specific metabolites. However, we have to consider some limitations in our study. First, patients with PD were administered antiparkinsonian drugs and no samples from non-medicated patients were available. In this study, the regimen of patients with PD was more complex and different from that of control participants (see online supplementary table S1). Therefore, we should consider the influence of drugs on the metabolites and drug-induced changes in metabolism. However, previous studies suggested that the effects of medication for PD may not significantly impact the profiling of metabolism, with the exception of the dopamine pathway.13 Further studies may need to be conducted to see if dopaminergic treatments are potentially confounding factors. To minimise the effect of medications, the next step of exploring the metabolic biomarkers of PD would be to investigate differences between drug-naïve patients with PD and non-medicated control participants. The second limitation of our study was the relatively small sample size. However, our results are in line with previous findings on metabolomic analyses or epidemiological investigations of PD. Therefore, despite these limitations, we believe that the results reflect alterations in metabolic pathways that are related to the pathomechanisms of PD.

This study has made an important contribution because it has revealed that several metabolic pathways, including the catecholamine metabolism, the caffeine and xanthine pathways, the ornithine pathway and redox homoeostasis, are altered in patients with PD compared with controls. We also found altered levels of biliverdin and ergothioneine in PD, which are redox markers. Moreover, the bilirubin/biliverdin ratio was significantly decreased and correlated with disease severity. Therefore, this ratio could be a potential prognostic marker for PD. These results suggest that oxidative stress may be involved in the pathomechanisms of PD, consistent with previous findings. Our results also suggest that several environmental factors and genetic susceptibility may perturb several metabolic pathways, and that metabolic alterations may promote dopaminergic neuronal degeneration in PD. Future studies will be needed to determine pathways that contribute to neurodegeneration in PD.

Acknowledgments

The authors thank Dr Yuuki Yonekawa (Summit Pharmaceuticals International Corporation), Ms. Yoko Imamichi (Juntendo University, School of Medicine) and all the participants in this study.

References

View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Contributors TH and SS were involved in study concept and design. TH, SS and AO were involved in acquisition of data. TH and RPM were involved in analysis and interpretation, and drafting of the manuscript. TH, SS and NH were involved in critical revision of the manuscript for important intellectual content. TH, AO and RPM were involved in administrative, technical and material support. NH was involved in study supervision.

  • Funding This work was supported by a Strategic Research Foundation Grant-in-Aid Project for Private Universities, and Grants-in-Aid for Scientific Research on Priority Areas (KAKENHI) (to TH, 25461290; to SS, 25111007; and to NH, 24390224), Ministry of Education, Culture, Sports, Science and Technology Grant-in Aid for Scientific Research on Innovative Areas (Brain Environment: Molecular mechanism of aggregate formation) (to NH), and Japan Science and Technology Agency, CREST; Creation of innovative technology for medical applications based on the global analyses and regulation of disease-related metabolites (to NH).

  • Competing interests RPM is an employee of Metabolon Inc.

  • Patient consent Obtained.

  • Ethics approval The Ethics Committee of the Juntendo University, School of Medicine.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.