Exercise-induce hyperalgesia, complement system and elastase activation in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome – a secondary analysis of experimental comparative studies

1 Introduction

The immune system has been extensively shown as capable of modulating nociceptive inputs and central nervous system processes (reviewed in [1] and [2]). Immune cells and glia interact with peripheral and central neurons through the release of inflammatory mediators, that in turn alter nociceptive processing and lead to hyperalgesia in animals [3], [4], [5], [6]. Patients with chronic pain might show altered levels of immune system markers such as pro-inflammatory cytokines and other inflammatory markers [7], [8]. Inflammatory markers might rapidly decrease after a single bout of aerobic, sub-maximal exercise [9]. However, it is unclear whether the reduction of immune markers determines an improvement in symptoms in humans. In addition, this might be different for other conditions, like myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). Two systematic reviews showed that cytokine expression do not differ between people with ME/CFS and healthy controls at baseline [10], [11] or following exercise [11], [12]. However, exercised-induced changes in other immune markers such as toll-like receptor 4 (TLR4), and products of the complement system have been reported [12]. A better understanding of the relation between the immune system and nociceptive modulation would allow the development of new analgesic treatments, improving the management of different chronic pain conditions.

The complement system (CS) is part of both innate and adaptive immunity [13], [14]. It includes several proteins that interact with each other through complex regulatory mechanisms at both RNA and protein levels [15]. The CS plays a role in immune response enhancement, immune complexes clearance and even synaptic maturation and angiogenesis [14], and it has been proposed as key mediator in pain states. After an injury, complement fragments C3a and C5a are produced and they are able to recruit and activate immune cells [14] – in turn enhancing inflammatory response. C3a and C5a injection in an otherwise healthy tissue induced nociceptive sensitization, hyperalgesia to heat, and neuroinflammation, in animals [16], [17].

Similarly, other pathways can potentially play a role in inducing nociceptive sensitization and inflammation. A relevant one involves the activity the enzyme elastase. This enzyme exerts proteolytic cleavage of a number of structurally and functionally vital proteins, including protease-activated receptor-2 (PAR₂) [18] and Ribonuclease L (RNase L) [19]. PAR₂ in turn activates transient receptor potential channels of nociceptive fibers, leasing to neurogenic inflammation and hyperalgesia [18]. RNase L activation is associated with protein kinase R (PKR) activation and nucleus transcription factor kB (NF-kB) expression [20], [21]. NF-kB has a recognized role in both up-regulating genes responsible for immune responses and in nervous system plasticity [22], [23]. Taken together, these findings support the role of elastase in initiating a downstream cascade of events that ultimately leads to neuronal sensitisation, immune activation, and hyperalgesia. Elastase has indeed been shown to cover a role in the pathophysiology of chronic inflammation and rheumatoid arthritis [24], [25].

In humans, CS and elastase activations in relation to pain have been studied in people with ME/CFS. ME/CFS is a condition characterized by many symptoms, being post-exertional malaise, fatigue and pain the most disabling ones [26]. In addition, patients with ME/CFS show a malfunctioning descending modulatory system for nociceptive stimuli [27]. Following aerobic exercise, healthy people normally show an increase in pressure pain thresholds (PPTs) [28] – a phenomenon known as exercise-induced hypoalgesia (EIH). On the contrary, the same exercise fails to induce hypoalgesia in patients with ME/CFS, and PPTs might even decrease following exercise [27], [28], [29], [30]. This accounts for dysfunctional EIH in patients with ME/CFS. Exercise or physical activity indeed induce symptoms worsening in ME/CFS patients that can last up to 24–48 h [26].

Evidence from animal and human studies suggest that immune system activation might contribute to both the pathophysiology of ME/CFS and the etiology of post-exertional malaise [7], [12], [31], [32], [33]. Sorensen et al. [34] studied three key products of the CS, namely C3a, C4a, and C5a, and found that only C4a increased in ME/CFS patients following exercise. This result is in line two previous studies from our group, showing a change in C4a levels and a decrease in PPTs in ME/CFS patients following exercise [27], [35]. Together with C4a changes, exercise-induced elastase alterations in ME/CFS patients have been reported, too [36], [37]. However, to the best of our knowledge no studies have investigated whether this exercise-induced immune system alteration can be linked to dysfunctional EIH in any chronic pain population.

The aim of the present work was therefore to investigate the link between dysfunctional EIH and immune system responses. To answer to our research question, we performed a secondary analysis of original data from our previous works [27], [35]. C4a products and elastase activity were used as measures of the immune function. PPTs are a reliable [38] and widely use tool for assessing local or widespread hyperalgesia [39], [40], [41]. We hypothesized that changes in C4a products and elastase activity following exercise are associated to hyperalgesia in ME/CFS patients.

2 Methods

3 Results

Twenty-two women with ME/CFS and 22 healthy sedentary women participated in the study. Mean age (±SD) of the CFS group was 34.3±8.8 years and of the healthy controls was 38.9±15 years. Mean BMI in the CFS group was 24.1±4.7 kg*m² and in the healthy control group it was 24.5±4.8 km*m². Between group analysis at baseline showed no differences for age and BMI (p=0.341 and 0.814, respectively). In baseline (visit 1), 10 patients in the ME/CFS group reported the use of anti-depressants, and 12 used analgesics. One subject in the control group reported using anti-depressants.

The results of the between group comparison and/or the between exercise tests comparison regarding accelerometry, exercise response, exercise capacity, PPTs, elastase activity, and complement C4a levels have been reported elsewhere [5], [10]. Here, we report within-group and between-group comparisons regarding PPTs, elastase activity, and complement C4a levels during either sub-maximal exercise test or self-paced/physiologically-limited one, in Table 1 and Table 2, respectively.

The current analysis showed that PPTs measured prior to and following the exercise tests were not significantly correlated to elastase activity or C4a products in either group (ME/CFS patients or healthy controls). In addition, changes in elastase activity were not associated to changes in PPTs in neither of the exercise bouts. However, when analysing possible associations between changes in PPTs and complement products, significant correlations were observed after both exercise bouts. After the submaximal exercise test, the reduction of PPTs at the thumb showed a very strong correlation with the change in C4a (r=0.669; p=0.001 – see Table 3 and Fig. 1 for details). In addition, after the paced exercise, the decrease in PPTs at the calf was correlated with the change in C4a (r=0.429; p=0.047, Table 3). The remaining correlations were non-significant. Regression analysis showed that the change in C4a following experiment 1 significantly predicted the change of PPTs at the thumb (R²=0.238; p=0.02. Model fit: F=6.25; p=0.2). WLS regression analysing the relation between changes in C4a and PPTs at the calf, was not significant.

Fig. 1:

Associations between exercise-induced complement system changes and pain threshold changes in healthy controls and patients with chronic fatigue syndrome. No significant correlations have been found in healthy controls. Details of the correlations can be found in Table 3.

On the contrary, we found no correlations in healthy controls (Fig. 1 and Table 3).

4 Discussion

We investigated whether immune system changes following exercise can be associated to a common feature of ME/CFS patients – that is the dysfunction in the endogenous pain inhibition. To the best of our knowledge, this is the first attempt to explore this association in ME/CFS or any other chronic pain conditions.

Our group previously showed that in healthy inactive people pain pressure thresholds increase in response to exercise [27]. On the contrary, PPTs in people with ME/CFS decreased following exercise [27]. These responses were observed following a submaximal exercise, as well as a self-paced and physiologically-limited exercise. Furthermore, the decrease in PPTs was shown to be associated with post-exertional malaise and symptom exacerbation in ME/CFS [27]. In addition, post-exertional increase in pain and fatigue were related to two immune factors, namely C4a products and elastase, respectively [35]. However, it had not previously been explored whether immune system changes and descending pain inhibitory activity in response to exercise are related.

While no association between elastase activation and pain was found, both ME/CFS patients and healthy subjects showed a similar decrease of C4a products in response to exercise. However, no relation was found between the change in complement products and the changes in pain thresholds in healthy controls. On the contrary, exercise-induced C4a changes seen in patients were associated with pressure pain thresholds changes. Interestingly, complement activity was not associated with pain thresholds at rest, or after the exercises, but it is rather the magnitude of exercise-induced changes that showed association. This is important, as post-exertional symptoms worsening is a characteristic peculiar of people with ME/CFS [52]. Fatigue and pain do not worsen after sub-maximal exercise in other conditions affected by fatigue such as depression and multiple sclerosis [53], [54], [55].

The CS has a key role in innate immune responses as well as in humoral immunity, and its activation ultimately promotes inflammatory and neuro-inflammatory responses [56]. CS activation has been related to hyperalgesia in both inflammatory and neuropathic animal pain models [57], [58], [59]. These studies showed that the increased amount of complement proteins C3a and C5a can induce hyperalgesia via both peripheral and central (i.e. microglia-mediated) mechanisms. However, only C4a, and not C3a or C5a, changed after exercise in ME/CFS patients [34].

To the best of our knowledge, C4a has not been previously related to pain. However, a number of autoimmune diseases, like Systemic Lupus Erythematosus, Rheumatoid Arthritis, or autoimmune hepatitis, have been associated with C4a suppression or C4 gene depletion [60], [61], [62], [63]. C4a deficiencies was associated with elevated level of immune complexes [64]. These findings corroborate the hypothesis that a suppression in the CS might be associated with over-reaction of autoimmune pathogens, which in turn is capable to explain ME/CFS patients’ symptoms after exercise or physical activity [65], [66], [67]. This hypothesis is in line with our current understanding of the great complexity of the immune system, which shows a great amount of interplay within the immune system itself, but also with other systems namely the endocrine and nervous system [68]. Complex, reciprocal, and far from being understood interactions show that the immune response is not straightforward, but it often includes a combination of enhanced and suppressed responses.

Interestingly, the majority of immune factors described in the scientific literature increases immediately post-exercise in ME/CFS [12]. In two similar studies, C4a was found to increase after sub-maximal exercise [32], [34], but this was 6 h after exercise, and no actual changes were found immediately after exercise. Here we have found a quick suppression of C4a products, suggesting that C4a is initially suppressed, and then increases over time to eventually peak a few hours after exercise. The change in C4a can explain 24% of the change in pain thresholds.

Notably, we found interesting results although we used indirect measures to assess a systemic immune response and pain modulatory mechanisms. In addition, exercises were tailored to the patient’s ability, in a safe and supervised environment, and both tests were stopped as soon as patients reached the target heart rate, or when they chose to stop. Both types of exercise test – especially the self-paced one – were chosen in order to induce a limited amount of stress or no stress at all and avoid severe symptom exacerbations. Despite this, associations between C4a changes and pain changes following exercise were found even after a low-demanding, self-paced exercise. A quick suppression of C4a might be part of the immune response associated with exercise-induced hyperalgesia in ME/CFS patients.

However, our results need to be interpreted with caution. First, sample size was small, and associations between changes in complement system activation were associated to changes in pain thresholds in only one out of three body areas in both exercises. For these reasons, it cannot be completely excluded that our findings are just coincidental. Alternatively, C4a might be a confounding factor, and its change after a single bout of exercise might be mediated by other mechanisms. Other immune factors, inflammatory markers or oxidative stress might be more important contributors to the dysfunctional endogenous pain inhibition that ME/CFS patients display during exercise and are worth exploring. Anti-oxidant response after exercise is delayed and reduced [10], [12]. Changes in gene expression following exercise can also account for ME/CFS symptoms. In particular, interleukin-10 and Toll-like receptor 4 gene expression increase in ME/CFS patients but not in healthy people [33], [69].

5 Conclusion

Not all the changes in PPTs in response to exercise were related to changes in immune factors. However, strong and moderate associations between the exercise-induced changes in PPTs measured at the hand and the leg and complement protein C4a were found in CFS, but not in the healthy inactive people. We acknowledge that complement protein C4a decreased in both ME/CFS patients as in healthy people, but this decrease was associated with the changes in the PPTs only in the ME/CFS group. Despite some limitations, our findings support the hypothesis that immune system responses can play a role in, at least partially, unravelling those mechanisms that are involved in the dysfunctional endogenous modulation seen in ME/CFS. Further investigation is warranted and should include CS products but also other immune or inflammatory mediators. Finally, we highlight the importance of assessing exercise-induced changes, and not only to absolute values, when dealing with ME/CFS patients.