Proteases and Abdominal Pain—Old Dog, New (Microbial) Tricks

Abstract

Chronic abdominal pain is a debilitating symptom associated with various gastrointestinal (GI) conditions, such as inflammatory bowel disease (IBD) and irritable bowel syndrome (IBS). Unfortunately, the pathophysiology of chronic abdominal pain is poorly understood, leading to inadequate clinical treatments. Proteases have emerged as critical modulators of nociception through the activation or inactivation of Protease-Activated Receptors (PARs) on pain sensing neurons (nociceptors). Traditionally viewed as exclusively host-derived, recent work highlights the gut microbiota as a significant source of proteases and demonstrates that a delicate equilibrium exists between pro-nociceptive and anti-nociceptive signalling in IBS and IBD. Consequently, dysbiosis of the gut microbiota may contribute to the pathogenesis of chronic pain in these disorders. This review outlines the mechanisms by which host and microbial proteases contribute to the pathogenesis of chronic abdominal pain, emphasizing recent advancements in microbial involvement.

Chronic abdominal pain, in the setting of gastrointestinal (GI) conditions such as inflammatory bowel disease (IBD) or irritable bowel syndrome (IBS), is a challenging clinical problem. Up to 60% of patients with IBD experience chronic pain in the absence of active inflammation, whereas IBS affects 10% of the North American population.^1,2 Thus, a greater understanding of the pathogenesis of chronic pain and mechanisms underlying the increased excitability of pain-sensing nerves innervating the gut (nociceptor sensitization) in these disorders is critical. In recent years, the gut microbiota has emerged as a key regulator in the bidirectional communication between the GI tract and the nervous system. Clinical studies have reported altered microbiota composition and/or metabolic capacity in patients suffering from chronic and/or recurrent visceral pain.^3,4 As bacteria are located at barrier sites throughout the GI tract, they can potentially stimulate nociceptors through both direct and indirect mechanisms.⁵

Proteases are important mediators of nociceptor sensitization. Previously, proteases were thought to be exclusively made by the host. However, the gut microbiota has recently been shown to act as a source of proteases, with commensal microbes producing serine proteases that have been shown to play an antinociceptive role.⁵ In this review, we highlight a recent study by Baker et al, which demonstrates that antibiotic treatment causes a shift in microbial protease production, from antinociceptive bacterial serine proteases to pronociceptive bacterial cysteine proteases produced after antibiotic treatment.⁶ Bacterial cysteine proteases not only increased visceral hypersensitivity but also caused somatic hypersensitivity. These exciting data suggest that dysbiosis, or a change in microbial composition and metabolism, which is a feature of both IBD and IBS, could lead to an imbalance between host- and microbial-derived pro- and antinociceptive proteases. We will discuss the mechanisms of protease activation of nociceptors and how these contribute to the pathogenesis of chronic abdominal pain.

Neurobiology of Nociception

The International Association for the Study of Pain defines pain as “an unpleasant sensory and emotional experience associated with actual or potential tissue damage.”⁷ Damage (or potential damage) is sensed at the level of the gut via a process termed nociception. Nociception is the transduction, transmission, modulation, and perception of a noxious stimulus.⁸ Primary afferent (sensory) neurons are activated by mechanical, thermal, or chemical stimuli.⁸ These stimuli are then transduced into an electrical signal, which, if sufficient, will trigger action potentials. Seven subclasses of pain-sensing primary afferent neurons innervating the GI tract have their cell bodies in the dorsal root ganglia (DRG), and synapse with second-order neurons in the dorsal horn of the spinal cord.⁹ These, in turn, relay information through ascending pathways to the brain where pain is perceived.¹⁰

Chronic pain is often the result of peripheral sensitization of primary afferent nerves and is characterized by both a reduced threshold and an increased magnitude of responsiveness at the nerve endings.¹¹ As a result of altered nociceptive signaling, individuals may experience excessive sensitivity giving rise to pain (ie, hyperalgesia) and pain with typically nonpainful stimuli (ie, allodynia).¹² Either in the case of chronic macroscopic inflammation, as seen in IBD, or low-grade inflammation associated with disorders of gut–brain interaction (such as IBS),¹³ there is an accumulation of endogenous inflammatory factors that can be released by both immune cells (ie, mast cells, macrophages, and neutrophils) and neighboring activated nociceptors.¹⁴ These mediators can include neurotransmitters (eg, substance P, calcitonin gene–related peptide), bradykinin, chemokines, cytokines, and extracellular proteases.¹³ During peripheral sensitization, inflammatory mediators bind to various receptors on the peripheral endings of nerves, leading to the modification of ion channels, such as transient receptor potential vanilloid-1.¹⁵ These modifications at peripheral nerve endings enhance the excitability of the peripheral nerve fibers, resulting in hyperalgesia that persists despite the resolution of inflammation.¹⁴

Proteases

Proteases are enzymes that are responsible for the degradation of proteins. Proteases are classified by their site of action as either endopeptidases or exopeptidases. Endopeptidases break peptide bonds of nonterminal amino acids, whereas exopeptidases cleave peptide bonds at the terminal ends of proteins.¹⁶ Host-derived proteases are responsible for the digestion of ingested proteins but also a variety of other processes, including immune cell signaling, cell differentiation, intracellular signaling, and pathogen defense.¹⁷ Proteases encompass 2% of the human genome and are found in high levels throughout the GI tract. Proteases in the GI tract can be found in the lumen (digestive, dietary, and microbial) and the mucosa (epithelial cells, immune cells). Although physiological sources of proteases in the GI tract include the pancreas and the intestinal microbiota, during inflammation, proteases from the circulation, the epithelium, and immune cells can cause pathologic damage by activating downstream receptors and channels.¹⁸

Proteases are classified by their catalytic mechanisms into 6 major classes: serine proteases, cysteine proteases, aspartic proteases, metalloproteases, threonine proteases, and glutamic proteases.¹⁷ Serine proteases, cysteine proteases, and some metalloproteases bind to a specialized family of G-protein–coupled receptors known as protease-activated receptors (PARs).^19,20 PARs are irreversibly activated when the extracellular N-terminus of the receptor is cleaved by proteases, exposing a tethered ligand, which acts on the receptor itself, prompting downstream intracellular signaling.²¹ Some proteases can also inactivate PARs by cleavage of an alternative site, leading to a disabled receptor that cannot activate.^22,23

To date, 4 PARs (PAR1–4) have been identified, with each receptor having a specific site, function, and distribution among cells.²⁴ PAR1 is widely expressed in various cell types such as smooth muscle cells, platelets, endothelial cells, and neurons. PAR1 is primarily activated by thrombin, a serine protease, and matrix metalloproteinase, a metalloprotease, and inactivated by cathepsin G and elastases, both serine proteases.^19,20,25 Asfaha et al found that selective activation of PAR1 reduced mechanical and thermal hyperalgesia in both noninflammatory and inflammatory conditions using mouse models.²⁶

PAR2 is expressed by epithelial cells, immune cells, and sensory neurons. PAR2 can be activated by various serine proteases, of which key enzymes are trypsin and tryptase.²⁷ Alternatively, PAR2 can be inactivated by cathepsin G and elastase (serine proteases).²⁰ A 2007 study by Cenac et al found that supernatants from IBS colonic biopsies contained elevated levels of tryptase and trypsin compared with control patients.²⁸ Incubation of cultured mouse sensory neurons with these supernatants resulted in increased calcium mobilization (a measure of activation) compared with control supernatants; this effect was absent in the presence of the serine protease inhibitor, FUT-175, and in neurons derived from PAR2^−/− mice. Intracolonic administration of supernatant from colonic biopsies of patients with IBS resulted in increased visceral sensitivity to colorectal distention compared with control supernatants; again, the effect was lost in PAR2^−/− mice. This seminal study suggested that host-derived serine proteases played a pronociceptive role in visceral pain via PAR2. Work from De Winter’s group has shown that administration of serine protease inhibitors in acute colitis and postcolitis rats (with full mucosal healing) reduces visceral hypersensitivity.^29,30 PAR2 can also be activated by cysteine proteases, such as cathepsin S. A study by Cattaruzza et al found that PAR2 activation by cathepsin S in a mouse model of colitis led to visceral hyperalgesia.³¹ In nociceptors, PAR2 downstream signaling occurs both at the level of the cell membrane and continues despite receptor endocytosis, resulting in persistent cell excitability.³² Therefore, proteases that are released during intestinal inflammation can directly stimulate nociceptors through PAR2 and cause persistent activation, resulting in increased pain neurotransmission (Figure 1A).

Figure 1.

(A) Proteases are released during intestinal inflammation by the host (epithelial cells, immune cells) or the gut microbiota. These proteases can directly stimulate nociceptors through PAR2. PAR2 downstream signaling occurs both at the level of the cell membrane and continues despite receptor endocytosis, resulting in persistent cell excitability.³² (B) PAR3^−/− mice exhibit a pronociceptive phenotype compared with wild-type mice.³³ PAR1 signaling is potentiated in the absence of PAR3. Furthermore, hyperalgesic priming through PAR2 was lost in PAR3^−/− mice. These findings suggest that PAR3 signaling mediates PAR1 and PAR2 activity. Figure created using Biorender.com.

Thrombin activates PAR3, whereas cathepsin G inactivates the receptor.²⁰ Mwirigi et al confirmed that PAR3 is coexpressed on murine nociceptors with PAR1 and PAR2.³³ These authors also demonstrated a pronociceptive phenotype due to PAR1 and PAR2 activation, which was potentiated in PAR3^−/− mice, suggesting that PAR3 may regulate the activation of PAR1 and PAR2 (Figure 1B). Several serine proteases have been identified as agonists for PAR4, such as thrombin, trypsin, and cathepsin G.²⁰ Asfaha et al found that PAR4 was expressed in rat DRGs, and PAR4 colocalized with calcitonin gene-related peptide and substance P.³⁴ They showed that intraplantar injection of a PAR4 agonist increased the nociceptive threshold to mechanical and thermal noxious stimuli. Likewise, Auge et al found that intracolonic administration of PAR4 agonist was able to reduce the visceral motor response to colorectal distention in mice.³⁵ Pretreatment with a PAR4 agonist was able to significantly reduce the calcium response to PAR2 agonist in vitro, indicating a possible mechanism of PAR4-mediated antinociception. These findings demonstrate that PAR signaling constitutes a critical regulatory axis in sensory transduction and provides a mechanistic framework for how dysregulated proteolytic activity may contribute to the pathogenesis of chronic pain.

Dysregulated proteolytic activity and PAR signaling have been implicated in disorders characterized by chronic visceral pain, including IBD and IBS. In both IBD and IBS, there is an increase in host serine protease production by colonic epithelial cells and mucosal immune cells.^28,36,37 Patients with IBD display increased cathepsin G and thrombin (serine proteases) activity in supernatants derived from the colonic mucosa, compared with healthy individuals.³⁷ Elastase (a serine protease) has also gained attention in recent years, as increased elastase-like activity has been observed in patients with IBD.³⁸ A recent landmark study demonstrated that β-glucuronidase derived from a commensal bacterium, Alistipes putredinis, regulated host serine protease production and proteolytic activity.³⁹ In patients with postinfectious IBS, the gut microbial composition was persistently altered. This resulted in decreased A putredinis, microbial β-glucuronidase activity. and elevated host luminal serine protease activity. Germ-free mice colonized with stool derived from hosts with elevated proteolytic activity displayed increased intestinal permeability,³⁹ a frequent finding associated with the development of disorders of gut-brain interaction.⁴⁰ These data suggest that regulation of host proteases by commensal microbes is essential to the proteolytic balance in health.

Microbial Proteases: the Good and the Bad

Bacterial proteases are mechanistically, structurally, and functionally diverse from host proteases.⁴¹ Notably, they possess a broader substrate specificity when compared with host proteases. The most abundant bacterial proteases are serine proteases, metalloproteases, and cysteine proteases.⁴² Studies looking at fecal samples have found that patients with IBD display both increased protease levels and protease activity, suggesting a possible defect in protease inhibition in these individuals.⁴³ However, fecal samples contain proteases derived both from host and the microbiota; thus, it is difficult to determine the origin of protease activity in these samples. As previously outlined in this review, protease signaling can induce both hyperalgesia and hypoalgesia, which are dependent on downstream PAR signaling. Recent studies into bacterial proteases indicate that a delicate equilibrium exists between pronociceptive and antinociceptive signaling; consequently, dysbiosis of the gut microbiota may contribute to the pathogenesis of chronic pain.^5,6,44 For the remainder of this review, we will highlight pivotal research characterizing how serine and cysteine bacterial proteases modulate nociceptive pathways through the activation of PAR4 and PAR2.

A 2017 study from the Lomax group focused on the protease-mediated suppression by commensal microbes on DRG excitability. Sessenwein et al utilized a defined community of 33 commensal microbes, isolated from a healthy human donor, termed microbial ecosystems therapeutics-1 (MET-1) to investigate whether secretory microbial mediators could alter DRG excitability.⁵ The authors found that serine proteases secreted by MET-1 were able to decrease the excitability of colonic-projecting DRG neurons in vitro. Inhibition of PAR4 but not PAR2 blocked the effects of MET-1 (Figure 2). Interestingly, the authors were able to replicate the antinociceptive effects of MET-1 with Faecalibacterium prausnitzii 16-6-I 40 FAA alone. Although this study highlights the translational potential of utilizing commensal microbes as therapeutics, it is crucial to investigate the effect of antinociceptive bacterial serine proteases using an in vivo model. Previous research has shown that proteases increase barrier permeability, thus facilitating translocation across the intestinal epithelium. However, commensal microbes are known to strengthen barrier function;⁴⁵ thus, it is essential to explore the routes and mechanisms by which these proteases produced by commensal microbes can activate PAR4 on nociceptors, and whether this effect is direct or indirect.

Figure 2.

Commensal microbes produce anti-nociceptive serine proteases that activate PAR4 on DRG neurons, resulting in increased voltage-gated K⁺ currents and decreased activation of PAR2.⁵ This PAR4 signaling cascade results in decreased neuronal excitability. Alternatively, a dysbiotic microbiota results in a shift to pronociceptive cysteine proteases. These cysteine proteases activate PAR2 on DRG neurons, leading to an increase in increased voltage gated Na⁺ current density. Activation of the PAR2 signaling cascade leads to increased DRG neuronal excitability and hyperalgesia.⁶ Figure created using Biorender.com

In a 2024 issue of Cellular and Molecular Gastroenterology and Hepatology, Baker et al investigated the cellular mechanisms of visceral and somatic hypersensitivity caused by antibiotic-induced microbial perturbations.⁶ In this study, the authors used vancomycin treatment to disrupt the gut microbiota. Using a behavioral assay, they found that vancomycin induced visceral hypersensitivity to colorectal distention and thermal hypersensitivity to noxious heat. Through a series of electrophysiological studies, the group established that DRG neurons from vancomycin-treated mice were hyperexcitable due to increased voltage-gated Na⁺ current density (Figure 2). Naïve DRG neurons incubated with serum derived from vancomycin-treated mice became hyperexcitable, suggesting that a soluble factor produced after antibiotic treatment was responsible. To confirm that the dysbiotic microbiota was the source of the excitatory proteases, the authors compared the potential of bacterial and host protease to excite DRG neurons. They found that only the bacterial proteases had the capacity to excite naïve DRG neurons in vitro. Interestingly, they also found that stool from vancomycin-treated mice possessed decreased proteolytic capacity.⁶ These data support the notion that a disruption to the microbiota causes a shift in the type of bacterial proteases, as opposed to an increase in bacterial proteases, which leads to visceral hypersensitivity. Blocking either cysteine proteases or the PAR2 receptor abolished the effects of the vancomycin-treated serum (Figure 2). The characterization and identification of these cysteine proteases as well as the microbe(s) responsible for their production require further investigation. More recently, Lakemeyer et al identified a previously uncharacterized Bacteroides fragilis serine protease 1 that cleaves the N terminal of PAR2, leading to the sensitization of nociceptors in vivo and in vitro.⁴⁴ These are the first studies to date focusing on the activation of neuronal PAR2 by bacterial proteases, thus providing novel fundamental evidence for the potential use of bacterial protease inhibitors as therapeutics in the management of chronic abdominal pain in GI disease.

Conclusion and Outlook

Chronic abdominal pain is a significant burden on the health care system.⁴⁶ Currently, there is a gap in knowledge of the exact mechanisms that lead to chronic visceral pain, which prevents the development of safe and effective therapies. Both proteases and PARs have been implicated as critical mediators in nociception, and there is a growing interest in targeting these pathways in the treatment of chronic pain. It is important to note that the effect of proteases on nociception is dependent on the PAR pathway that is activated. Moreover, the colon is innervated by distinct subpopulations of DRG neurons;⁹ thus, characterizing the differential expression of PARs across these functional classes is warranted. Furthermore, it would be important to delineate whether host or microbial proteases are the primary mediators of nociception, as an altered microbiota is commonly found in chronic pain disorders.^3,4 As highlighted in this review, extensive work has shown that PAR2 activation results in neuronal hyperexcitability in vitro and leads to hyperalgesia in vivo.^28,31,47,48 Alternatively, studies focused on PAR4 signaling show that PAR4 activation reduces DRG neuronal hyperexcitability in vitro and induces hypoalgesia in vivo.^34,49 Thus, cysteine and serine proteases can lead to either hyperalgesia or hypoalgesia, depending on the PAR pathway that is activated. This subtlety is critical to consider in future work conducted on the development of therapies targeting proteases (bacterial or host) and PARs. The work from both Baker et al and Sessenwein et al highlights the importance of gut microbes as key regulators of PAR-mediated nociceptor excitability and underscores this exciting avenue for further investigation.