Urine Metabolomics Exposes Anomalous Recovery After Maximal Exertion in Female ME/CFS Patients Part 1

Why we will be tired? How can we solve the fatigue problems?

【Contact】Email: george.deng@wecistanche.com / WhatsApp:008613632399501/Wechat:13632399501

Abstract: Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a debilitating disease with unknown etiology or effective treatments. Post-exertional malaise (PEM) is a key symptom that distinguishes ME/CFS patients. Investigating changes in the urine metabolome between ME/CFS patients and healthy subjects following exertion may help us understand PEM. This pilot study aimed to comprehensively characterize the urine metabolomes of eight female healthy sedentary control subjects and ten female ME/CFS patients in response to a maximal cardiopulmonary exercise test (CPET). Each subject provided urine samples at baseline and 24 hours post-exercise. A total of 1403 metabolites were detected via LC-MS/MS by Metabolon® including amino acids, carbohydrates, lipids, nucleotides, cofactors and vitamins, xenobiotics, and unknown compounds. Using a linear mixed effects model, pathway enrichment analysis, topology analysis, and correlations between urine and plasma metabolite levels, significant differences were discovered between controls and ME/CFS patients in many lipids (steroids, acylcarnitines, and acyl glycines) and amino acid subpathways (cysteine, methionine, SAM, and taurine; leucine, isoleucine, and valine; polyamine; tryptophan; and urea cycle, arginine and proline). Our most unanticipated discovery is the lack of changes in the urine metabolome of ME/CFS patients during recovery while significant changes are induced in controls after CPET, potentially demonstrating the lack of adaptation to severe stress in ME/CFS patients.

Cistanche can act as an anti-fatigue and stamina enhancer, and experimental studies have shown that the decoction of Cistanche tubulosa could effectively protect the liver hepatocytes and endothelial cells damaged in weight-bearing swimming mice, upregulate the expression of NOS3, and promote hepatic glycogen synthesis, thus exerting anti-fatigue efficacy. Phenylethanoid glycoside-rich Cistanche tubulosa extract could significantly reduce the serum creatine kinase, lactate dehydrogenase, and lactate levels, and increase the hemoglobin (HB) and glucose levels in ICR mice, and this could play an anti-fatigue role by decreasing the muscle damage and delaying the lactic acid enrichment for energy storage in mice. Compound Cistanche Tubulosa Tablets significantly prolonged the weight-bearing swimming time, increased the hepatic glycogen reserve, and decreased the serum urea level after exercise in mice, showing its anti-fatigue effect. The decoction of Cistanchis can improve endurance and accelerate the elimination of fatigue in exercising mice, and can also reduce the elevation of serum creatine kinase after load exercise and keep the ultrastructure of skeletal muscle of mice normal after exercise, which indicates that it has the effects of enhancing physical strength and anti-fatigue. Cistanchis also significantly prolonged the survival time of nitrite-poisoned mice and enhanced the tolerance against hypoxia and fatigue.

Click on sudden tiredness during the day

Keywords: ME/CFS; metabolomics; urine; exercise 

1. Introduction

Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a debilitating disease affecting an estimated 1.5–3 million adults and children in the United States alone [1,2]. The majority of ME/CFS patients are unable to work due to their illness, leading to an estimated minimum economic impact of 35–51 billion USD per year in medical costs and lost productivity combined [1]. Symptoms of this acquired, systemic disease include a new onset of persistent physical and mental fatigue severe enough to prevent normal activities, unrefreshing sleep, pain, cognitive impairment, orthostatic intolerance, immune manifestations such as recurrent flu-like symptoms and sore throat, and neuroendocrine manifestations such as intolerance to heat and cold [3].

In addition, the hallmark symptom of ME/CFS is post-exertional malaise (PEM), which is a worsening of symptoms after any type of exertion, including both physical and cognitive exertion, beginning from immediately following the exertion to more than 24 h later [4]. PEM may last hours to months, and the duration varies extensively even for individual patients [4,5]. Therefore, unlike most chronic illnesses, in which exercise is beneficial, people with ME/CFS are exercise intolerant. This exercise intolerance can be clinically assessed using a two-day cardiopulmonary exercise test (CPET).

In the two-day CPET protocol, the first CPET is used to measure “baseline functional capacity” while inducing PEM [6]. The second CPET, 24 hours later, measures the impaired performance at a time when most patients will already be experiencing PEM symptoms from the first CPET. Subjects with other various illnesses can perform similarly on a CPET two days in a row, whereas ME/CFS patients are unable to perform as well on the second day [7,8]. This reduced performance ability can be documented with objective measures including reduced maximal oxygen consumption and peak workload greater than the typical variability of repeated testing, despite subjects maintaining a respiratory exchange ratio (RER) above 1.1, which corresponds to maximum effort [6].

Although there is a growing body of knowledge describing the pathophysiology of ME/CFS, the etiology of the disease remains unknown and there are currently no diagnostic laboratory tests nor FDA-approved treatments. Despite the lack of diagnostic biomarkers, there are many documented molecular pathophysiological changes occurring in ME/CFS,  including in metabolomics [9,10].

The plasma metabolome of ME/CFS patients has received a substantial amount of attention for over a decade, although often on limited cohort sizes, with an increasing number of metabolites measured (from just over 20 to about 1200 metabolites more recently) [11–22]. On the contrary, previous studies of urine metabolomics in ME/CFS patients are very limited. The few published studies have measured only 28–42 metabolites, which have primarily been amino acids [19,20,23,24]. While most findings have not been consistent between studies, two studies did find phenylalanine at lower concentrations in ME/CFS  patients than healthy controls [23,24]. Although one study by McGregor et al. examined urine metabolites in the context of self-reported PEM [19], and our group recently published a thorough investigation of the plasma metabolome (1157 metabolites) before and after exercise [25], no studies of ME/CFS patients have measured metabolites in urine after a  deliberate exercise challenge.

Measuring compounds in urine is advantageous due to non-invasive and easy sample collection which makes it ideal for diagnostics. Additionally, altered excretion of metabolites in ME/CFS patients after an exercise challenge may yield insights into the pathophysiology of PEM that complement the changes documented in the plasma metabolome. This pilot study aimed to comprehensively investigate changes in the urine metabolomes of eight female healthy sedentary control subjects and ten female ME/CFS patients in response to a maximal cardiopulmonary exercise test (CPET). This study represents an approximately 30-fold increase in the number of metabolites measured in the urine of ME/CFS patients, from less than 50 in previous studies to 1403 in the current study.

Our extensive analysis reveals numerous and significant differences in the urine metabolomes of ME/CFS and control groups in response to exercise, despite the small number of subjects studied. Such changes are predominantly present in the lipid and amino acid metabolic superpathways. We found a large number of metabolites with increased levels in the urine of controls 24 h post-exercise. The post-exercise increase in urinary excretion did not occur in the ME/CFS patients, which is evidence of metabolic dysregulation during exercise recovery.

2. Results 

2.1. Study Design and Subject Characteristics

Eight female healthy sedentary control subjects and ten female ME/CFS patients provided a baseline urine sample in the morning before exercise testing (Figure 1). All subjects performed the CPET on a stationary bicycle and were monitored to ensure that they used maximal effort (RER > 1.1). A post-exercise urine sample was collected from all subjects 24 hours later. Metabolites were measured in all urine samples by Metabolon® using their Precision Metabolomics™ LC-MS/MS global metabolomics platform.

As expected, the ME/CFS patients scored significantly lower on multiple measures of physical function, including the Bell disability scale and the SF-36 physical component. On the Bell scale, the median for the controls was 90, which corresponds to “No symptoms at rest; mild symptoms with activity; normal overall activity level; able to work full-time without difficulty” [26]. The median for the ME/CFS patients was 30, which corresponds to “Moderate to severe symptoms at rest. Severe symptoms with any exercise; overall activity level reduced to 50% of expected. Usually confined to the house. Unable to perform any strenuous tasks. Able to perform desk work 2–3 hours a day, but requires rest periods” [26]. The ME/CFS patients had a wide range of disease duration of 2–27 years (median 7.5 years).

2.2. Many Differences in the Urine Metabolomes of ME/CFS Patients and Controls Emerge  through Analysis of Changes between Pre- and Post-Exercise Samples

Metabolon® detected a total of 1403 metabolites in these samples using their Precision Metabolomics™ LC-MS/MS global metabolomics platform. Out of the 1403 metabolites measured, 886 are known metabolites, 64 are partially characterized molecules, and 453 are unknown compounds (Supplementary File S1—Raw Data). All data were osmolality normalized to account for differences in the overall concentration of each urine sample. Osmolality data for each experimental group are shown in Supplementary Figure S1 and were not significantly different between ME/CFS patients and controls at either time point. There was a trend toward increased osmolality in the controls 24 h post-exercise (p < 0.1, linear mixed effects model, followed by pairwise comparisons with Tukey’s posthoc test). By normalizing to the osmolality, we ensured that any differences detected in metabolite levels both between ME/CFS and control groups and from baseline to post-exercise within cohorts are not simply a reflection of changes in osmolality.

Missing values were imputed with the minimum as recommended by Metabolon®,  except in the case of drugs and tobacco, where the missing values were imputed with 0. The data for each metabolite were median centered to 1, and all data were log10 transformed using MetaboAnalystR (available at www.metaboanalyst.ca, accessed on 14 December 2022). Filtering was applied to eliminate compounds with a large amount of missing data from the analysis according to the modified 80% rule: a metabolite is included if it is detected in at least 80% of the samples in either the ME/CFS patients, the controls, or both groups [27]. In total, 1154 metabolites met the criteria and were included in the analysis.

A linear mixed effects model (LMM) for each metabolite was utilized to determine the differences between ME/CFS and control groups at each time point (baseline or postexercise), the change over time within the ME/CFS and control groups, and which metabolites were changing differently after exercise in the ME/CFS and control groups. The model  formula is as follows:

Metabolite ~ Disease status ∗ Time Point + Age + BMI + (1|Subject)

The p-values for each metabolite were adjusted for multiple comparisons using the Benjamini–Hochberg (BH) procedure with a significance threshold of q < 0.1 (Supplementary File S2—Linear Mixed Model Results) [28]. Because we adjusted for the confounder's age and BMI, the significant differences we detected with this model are not due to the difference in BMI in the two groups.

No significant differences were detected between controls and ME/CFS patients at baseline (Figure 2A). At 24 h post-exercise, four compounds were significantly different, all of which were at lower concentrations in the ME/CFS patients compared to the controls (Figure 2B). The four compounds included three acyl glycines and one unknown compound.

The control group exhibited large-scale changes in the urine metabolome when comparing the baseline and post-exercise urine samples, with 255 compounds significantly altered at the q < 0.1 threshold (Figure 2C). All except five compounds showed increased concentrations after exercise. This is in stark contrast to the ME/CFS group, in which we did not detect any compounds with significant changes in concentration due to exercise (Figure 2D).

A significant interaction between disease status (ME/CFS vs. control) and time point (baseline vs. post-exercise) in the linear mixed effects model shows which metabolites are changing differently in the ME/CFS and control groups during exercise recovery (i.e., over time). Figure 2E displays 110 significantly different metabolites at the q < 0.1 thresholds (red dots), and 35 metabolites are also below q < 0.05 (1.3 on the -Log10q y-axis of Figure 2E). In this volcano plot, the log2 fold change is a ratio of ratios; the ratio of the mean postexercise/baseline ratios in the ME/CFS patients to the mean post-exercise/baseline ratios in the controls. The post-exercise/baseline ratio for each subject shows whether the metabolite is increased or decreased in urine in each subject during exercise recovery. The mean normalized concentrations for the controls and ME/CFS patients at both time points for the 56 known compounds included in these 110 compounds show that for most compounds, there is a post-exercise increase in the controls that is not seen in the ME/CFS patients (Supplementary Figure S2). Therefore, the compounds that are changing differently during recovery in the control and ME/CFS groups are predominantly increased postexercise in the control group and not significantly altered in the ME/CFS group, leading to a negative log2 fold change for the ME/CFS vs. control post-exercise/baseline ratios. In total, 102 of the 110 compounds that are changing differently over time in patients vs. controls are also significantly increased in the controls post-exercise.

The fact that there are metabolite changes in the controls and not in the patients is not due to increased variation in metabolite levels in the ME/CFS patients compared to the controls. There is no trend toward higher standard deviation in the ME/CFS group when comparing the standard deviations for ME/CFS to controls for each compound (Supplementary Figure S3). Additionally, these changes are detected despite normalizing urine osmolality, which is trending toward an increase in the controls from baseline to post-exercise (Supplementary Figure S1). To increase significantly after normalization, levels of a metabolite have to increase even higher than any increase in overall urine concentration.

Together, these results demonstrate that exercise induces a significant increase in many metabolites in urine 24 h post-exercise in healthy sedentary controls, and the lack of these changes in the ME/CFS patients is a key component of their disease state and could be related to exercise intolerance.

2.3. Two Approaches to Enrichment Analysis Reveal Metabolic Subpathways with the Most Significant Changes 

We then performed enrichment analysis using Metabolon®’s pathway annotations and ChemRICH (in R) [29,30]. This analysis does not rely on any background or reference metabolome and contains non-overlapping pathway assignments for each metabolite. Unlike the standard ChemRICH analysis which assigns metabolites to clusters by chemical similarity,  this analysis used non-overlapping pathway assignments by Metabolon®. By looking at the pathways as opposed to single metabolites, we can identify metabolite sets that are significantly altered as a group, whereas individual metabolites in that set may not have achieved significance in the LMM on their own. Additionally, we can holistically describe the changes in the urine metabolome occurring post-exercise and how they are significantly altered in ME/CFS patients. The input for this analysis is the fold change of the mean normalized concentration for each comparison and the p-values from the LMM contrasts for the 734 known compounds out of 1154 total analyzed (assigned to 84 subpathways out of 92 subpathways for all detected metabolites). Enrichment is assessed using the Kolmogorov– Smirnov test, with BH FDR correction (significance threshold q < 0.05).

Figure 3 shows the significantly altered subpathways for the same comparisons as in Figure 2: ME/CFS vs. controls at baseline, post-exercise, and for the post-exercise/baseline ratios, and the change post-exercise in the control group alone. We also analyzed the change post-exercise in the ME/CFS patients, but there were no significantly altered pathways.

In this figure, the subpathways are organized by Metabolon® by chemical similarity and then superpathways are organized alphabetically, with amino acids first, followed by carbohydrates, lipids, nucleotides, vitamins and cofactors, and finally xenobiotics. An altered ratio of 1 means every compound we measured in a pathway was significantly altered (p < 0.05 in the LMM). Red bubbles indicate that all significantly altered compounds increased in that comparison, and blue bubbles indicate that all significantly altered compounds decreased in that comparison. We also performed ChemRICH analysis using the Medical Subject Headings (MeSH) ontology and Simplified Molecular Input Line Entry System (SMILES) codes to assign compounds to clusters based on chemical similarity (Supplementary Figure S4). Only 516 compounds could be matched to SMILES codes for this analysis. We chose a significance threshold of q < 0.15 for the MeSH ontology enrichment,  because we did not want to exclude potentially interesting findings for this pilot study, and at q < 0.15 all clusters originally had p < 0.025. For the Metabolon® subpathway enrichment,  only at q < 0.05 were all significant clusters also at p < 0.05.

At baseline, only xanthine metabolism and the xanthines’ chemical cluster were significantly different between ME/CFS patients and controls, with lower concentrations of all compounds found in the urine of ME/CFS patients (Figure 3, q < 0.05 and Supplementary Figure S4, q < 0.15). Xanthines include caffeine and theophylline, so this chemical cluster is affected by diet.

Twelve Metabolon® subpathways were significantly altered post-exercise in the ME/CFS  patients vs. controls, again with predominantly lower metabolite concentrations in the ME/CFS patients (Figure 3). Five of these subpathways belong to the lipid super-pathway and are involved in fatty acid metabolism (including acyl glutamine, acyl glycine, and acylcarnitine). Within the amino acid metabolism super pathway, tyrosine metabolism was significantly different 24 h post-exercise and within nucleotide metabolism, uracil-containing pyrimidine metabolism was significantly altered. 24 h post-exercise, only three chemical clusters were significantly altered in the ME/CFS patients using the MeSH ontology, including fatty acids (10:1), xanthines, and sugar acids (Supplementary Figure S4).

When comparing the exercise recovery (the change over time) in ME/CFS patients vs. controls using the post-exercise/baseline ratios, even more pathways are found to be significantly different, namely seven amino acid subpathways, seven lipid subpathways, and two carbohydrate subpathways. For this comparison, 13 chemical clusters are significantly different between patients and controls (Supplementary Figure S4).

The highest number of significantly altered Metabolon® subpathways and chemical clusters was found in the control group when comparing the post-exercise and baseline time points (Figures 3 and S4), with the large majority of compounds increased post-exercise (red color). Only the adenine-containing purine metabolism subpathway, the histidine metabolism subpathway, and the methylhistamine chemical cluster had a similar amount of increased and decreased compounds.

Most of the Metabolon® subpathways and chemical clusters increased in the controls after exercise are the same ones that were significantly altered in the ME/CFS vs. controls when comparing the post-exercise/baseline ratio, showing that they were changing differently during exercise recovery in the ME/CFS patients vs. healthy sedentary controls. For both of these comparisons, the subpathway with the lowest q-value for subpathway enrichment was corticosteroids. Overall, 8/13 compounds were significantly altered in the ME/CFS patients vs. controls (and all had lower post-exercise/baseline ratios in patients), and 11/13 compounds significantly increased in controls after exercise (p < 0.05 in the LMM).

Four subpathways, all belonging to the lipid superpathway, had significantly altered concentration in urine in all three comparisons: ME/CFS vs. control at the post-exercise time point, ME/CFS vs. control post-exercise/baseline ratios, and post-exercise vs. baseline in the control group. These subpathways include two acylcarnitine fatty acid metabolism subpathways (medium chain and dicarboxylate), androgenic steroids, and secondary bile acid metabolism.

There are also several amino acid subpathways in which most altered compounds are significantly increased in the urine in controls post-exercise and decreased when comparing the ME/CFS to control post-exercise/baseline ratios. Of these subpathways, the altered ratio is the highest for polyamine metabolism (and it has the lowest q value for the ME/CFS vs. controls): 6/9 metabolites are significantly increased in controls, and 5/9 are significantly decreased in the ME/CFS vs. controls post-exercise/baseline ratio. These compounds include acisoga, (N(1)+N(8))-acetylspermidine, diacetylspermidine, N1,N12- diacetylspermine, and N-acetyl-isoputreanine (for this subpathway, all compounds also  have q < 0.1 in the LMM).

The arginine and proline metabolism subpathway of the urea cycle in the amino acid superpathway is the only one with a metabolite that is increased in ME/CFS vs. controls for the post-exercise/baseline ratios (besides food component/plant), with 8/10 metabolites significantly altered and methylurea being the only increased metabolite (p < 0.05 in LMM). Methylurea is also significantly decreasing in the controls over time (p < 0.05 in LMM) and is the only one of 11 altered compounds in this subpathway that is not increasing. However, the changes in methyl urea were not significant when considering the univariate LMM  analysis with the q < 0.1 thresholds. Nevertheless, four compounds in this subpathway do have q < 0.1 in the LMM for the ME/CFS vs. controls post-exercise/baseline comparison: carboxy-methyl-arginine, proline, symmetric dimethylarginine, and asymmetric dimethylarginine.

The leucine, isoleucine, and valine metabolism subpathway has the lowest q-value in the controls of the subpathways in the amino acid superpathway. This subpathway has 36 compounds, with six changing differently over time in ME/CFS vs. controls: 3-methylglutarylcarnitine, tiglylcarnitine(C5), valine, beta-hydroxy isovalerate, beta-hydroxyisovaleroylcarnitine, and methylsuccinoylcarnitine. Sixteen out of thirty-six compounds are significantly increased in the controls, including all six changing differently over time in the ME/CFS patients vs. controls.

【Contact】Email: george.deng@wecistanche.com / WhatsApp:008613632399501/Wechat:13632399501