Antibodies interfering with the type 3 muscarinic receptor pathway inhibit gastrointestinal motility and cholinergic neurotransmission in Sjögren's syndrome

Authors


Abstract

Objective

In primary Sjögren's syndrome (SS), impairment of the gastrointestinal (GI) tract is common, and includes reduced esophageal motor function, delayed gastric emptying, and abnormalities in colonic motility; the pathogenesis is as yet unknown. We undertook this study to investigate the role of functional antibodies to the type 3 muscarinic receptor (M3R) in GI dysfunction associated with primary SS.

Methods

Muscle strip and whole-organ functional assays were used to determine whether IgG with anti-M3R activity from patients with primary SS disrupted neurotransmission in tissue from throughout the mouse GI tract. Specificity of the autoantibody for the M3R was determined using knockout mice that were deficient in the expression of muscarinic receptor subtypes.

Results

Functional antibodies to the M3R inhibited neuronally mediated contraction of smooth muscle from throughout the GI tract and disrupted complex contractile motility patterns in the colon. The autoantibodies were not active on tissue from mice that lacked the M3R, providing compelling evidence of the direct interaction of patient autoantibodies with the M3R.

Conclusion

Our results indicate that anti-M3R autoantibodies have the potential to mediate multiple dysfunctions of the GI tract in primary SS, ranging from reduced esophageal motor activity to altered colonic motility. We hypothesize that altered GI motility forms part of a broader autonomic dysfunction mediated by pathogenic anti-M3R autoantibodies in primary SS.

Primary Sjögren's syndrome (SS) is a systemic autoimmune disorder characterized by exocrine failure of salivary and lacrimal glands (sicca syndrome) and extraglandular features, including gastrointestinal (GI) dysfunction (1–4). Impairment of the GI tract in primary SS is widespread, and includes reduced esophageal motor function (5, 6), delayed gastric emptying (7), and abnormalities in colonic motility (3). To date, the pathogenesis underlying GI autonomic dysfunction in primary SS remains unclear.

It has been postulated that functional autoantibodies that target the type 3 muscarinic receptor (M3R) contribute to autonomic dysfunction in patients with primary SS (3, 7–10). Investigators in our group have previously demonstrated that autoantibodies from patients with primary SS cause acute inhibition of parasympathetic neurotransmission in bladder and colon smooth muscle strips in vitro, presumably by targeting the M3R (11, 12). In addition, passive transfer of primary SS patient IgG to mice alters M3R expression and cholinergic responses in bladder smooth muscle (13). However, these findings do not provide direct evidence that autoantibodies from patients with primary SS interact specifically with the M3R. Mice lacking specific muscarinic receptors have been generated using gene targeting techniques (14–16). Analysis of these muscarinic receptor–knockout mice has revealed clear evidence of the roles of specific muscarinic receptors in the regulation of GI tract motility (17, 18), thereby providing a model for determining the specific target of antimuscarinic antibodies that mediate GI dysfunction in primary SS.

In the present study, we used muscle strip and whole-organ contractile assays to investigate the role of antimuscarinic receptor autoantibodies in disrupting GI motility in primary SS. Our results demonstrate that functional autoantibodies from patients with primary SS inhibit neuronally mediated contraction of smooth muscle throughout the GI tract and disrupt colonic motility. The autoantibodies are not active on tissue from mice that lack the M3R, which provides convincing evidence of the presence of pathogenic anti-M3R autoantibodies in primary SS.

MATERIALS AND METHODS

Patients.

Blood samples were collected from patients who fulfilled the American–European Consensus Group Criteria for SS (19) (n = 4) and from healthy controls (n = 6), after informed consent was obtained. Purified IgG from patients with primary SS was positive for inhibitory anti-M3R activity, as determined by in vitro bladder strip assay (11), while results were negative for anti-M3R activity in controls. The study was approved by the Clinical Ethics Committee of Flinders Medical Centre.

Preparation of IgG.

IgG was prepared using the caprylic acid precipitation technique (20). Briefly, the pH of human serum (5 ml) was first lowered to 4.8 by adding 1M acetic acid, and the serum was then stirred vigorously while 0.37 ml of caprylic acid was added (dropwise, resulting in a final ratio of 0.74:10). The caprylic acid–serum mixture was then stirred for 30 minutes at room temperature, and centrifuged at 10,000g for 15 minutes. Supernatant was recovered and filtered through a 0.25μ filter (Millipore), and the pH was adjusted to 7.4 with 0.5M of sodium hydroxide. Following dialyzation against phosphate buffered saline (PBS) (pH 7.4) (conducted twice in 1 liter PBS each time), the IgG was concentrated to 40 mg/ml using a Millipore B15 concentrator, and dialyzed against 1 liter of Krebs-Ringer solution (containing NaCl [118 mM/liter], KCl [4.7 mM/liter], KH2PO4 [1.2 mM/liter], MgSO4 [1.5 mM/liter], NaHCO3 [25.0 mM/liter], D-glucose [11.0 mM/liter], and CaCl2 [2.5 mM/ liter]). IgG concentrations were determined by nephelometry (ICS II; Beckman).

Functional assay of smooth muscle contraction in esophagus, stomach, and ileum.

Male BALB/c, C57BL/6, and M2R- or M3R-knockout mice (ages 10–12 weeks) were used for the study. The generation of the knockout mice has previously been described in detail (21–23). Mice were killed and tissue was excised and suspended in Krebs-Ringer solution with 95% O2/5% CO2 (pH 7.4) at 37°C. Ten-millimeter strips were prepared from longitudinal smooth muscle of the esophagus and central ileum. For studies of stomach muscle, strips (10 × 2 mm) of circular smooth muscle were prepared from the fundic region. The various strips were then mounted longitudinally in 10-ml jacketed organ baths containing Krebs-Ringer solution with 95% O2/5% CO2 at 37°C. At the beginning of each experiment, preparations were desensitized to capsaicin (10 μM; Sigma) to inactivate sensory neurons, and hexamethonium (100 μM; Sigma) and guanethidine (0.3 μM; Sigma) were added to block nicotinic ganglionic neurotransmission and sympathetic neurons, respectively. Electrical field stimulation was performed with a pulse width of 0.5 msec at 70V, using a Grass 588 stimulator (Grass Technologies) for durations of 3 seconds at a frequency of 70 Hz. Smooth muscle carbachol contraction–response curves were generated by cumulative addition of 0.1–10 μM of carbachol to the preparations. Contractile responses to electrical field stimulation and to carbachol were measured using isometric transducers (Letica) connected to a PowerLab/8s data acquisition system (ADInstruments).

Following a 30-minute equilibration period, baseline smooth muscle contractions induced by electrical field stimulation or carbachol were recorded. To analyze the effects of IgG, patient or control IgG was added, bringing the final concentration to 1.0 mg/ml. After a 30-minute incubation, smooth muscle contractions induced by electrical field stimulation or carbachol were recorded at 30-minute intervals.

Functional assay of spontaneous colonic migrating motor complexes in the intact colon.

Colonic migrating motor complexes (CMMCs) result from a complicated interplay between enteric afferent neurons, interneurons, and motor neurons and comprise a major motor pattern involved in moving fecal matter along the bowel (24–26). In the current study, CMMCs were studied using previously described methods (20, 27, 28) with minor modifications. Male BALB/c, C57BL/6, and M2R- or M3R-knockout mice (ages 10–12 weeks, 20–30 gm) were killed and the colons excised and suspended in a 100-ml organ bath containing Krebs-Ringer solution with 95% O2/5% CO2 (pH 7.4) at 37°C. Any fecal matter present was gently flushed from the colon and the mesentery was dissected free. The mechanical activity of the circular smooth muscle was recorded with 3 force-displacement transducers (ADInstruments) attached at 2-cm intervals to the mesenteric border of the colon under an initial tension of 6 mN. Transducer output was fed into a Quad bridge amplifier (ADInstruments) and recorded using Chart version 4.2 software and a PowerLab/8s data acquisition system.

Following a 60-minute equilibration period, ≥5 consecutive CMMCs were recorded in order to provide a baseline for comparison with later CMMCs. To analyze IgG effects on CMMC activity, patient or control IgG was added to the bath to yield a final concentration of 0.2 mg/ml. After a 30-minute incubation, consecutive CMMCs were recorded for a period equivalent to the duration of the control period in the continued presence of IgG.

Statistical analysis.

The amplitude of pre- and post-IgG smooth muscle contractions and the frequency of pre- and post-IgG CMMCs were calculated using the peak parameters–peak amplitude function of DataPad on Chart version 4.2 (ADInstruments) and compared by two-way analysis of variance (ANOVA) (amplitude of smooth muscle contractions) and by Student's paired t-test (frequency of CMMCs), using GraphPad Prism version 3.0a for Macintosh (GraphPad Software). P values less than 0.05 were considered significant.

RESULTS

IgG from patients with primary SS inhibits neuronally mediated contraction of smooth muscle from throughout the GI tract.

IgG from patients with primary SS has been shown to inhibit parasympathetic cholinergic neurotransmission in strips of colonic smooth muscle (29). In the current study, we used electric field stimulation on isolated smooth muscle strips to test whether IgG with anti-M3R activity from patients with primary SS (11) inhibited cholinergic neurotransmission in smooth muscle from other regions of the GI tract. Electric field stimulation of smooth muscle strips from the esophagus, ileum, and gastric fundus resulted in a single high-amplitude phasic contraction, which was followed by a sustained lower-amplitude tonic response in the esophagus and ileum (Figure 1A). The responses were abolished by tetrodotoxin (0.3 μM), indicating that the electric field stimulation–induced contractions were mediated by neurotransmitter release from enteric neurons.

Figure 1.

Inhibition of neuronally mediated contraction of smooth muscle from throughout the gastrointestinal tract by IgG from patients with primary Sjögren's syndrome (pSS), consistent with activity at the type 3 muscarinic receptor (M3R). A, Representative tracings of contraction following electric field stimulation of smooth muscle strips from the esophagus, stomach, and ileum both before (Pre) and at 30-minute increments after incubation with 1 mg/ml IgG from controls or patients with primary SS. Electric field stimulation induced an initial rapid, phasic contraction (∧) in all tissue, and a sustained tonic contraction (–—) in esophagus and ileum. The phasic response in the various types of tissue was reduced following incubation with IgG from patients with primary SS, while the tonic response was affected only in the esophageal tissue. B, Pooled mean maximum amplitudes of electric field stimulation–induced contractions in esophagus, stomach, and ileum, presented as percentage of control response (0 minutes). Contractions in all types of tissue were significantly reduced following incubation with IgG from patients with primary SS (solid lines) (n = 4) as compared to that from controls (broken lines) (n = 4). Values are the mean ± SD. ∗ = P < 0.05; ∗∗ = P < 0.001 by two-way analysis of variance. C, Representative tracings of electric field stimulation–induced smooth muscle contractions before (control) and after incubation (30 minutes) with 30 nM of 4-diphenylacetoxy-N-methylpiperidine (4-DAMP). Pharmacologic targeting of the M3R reduced or abolished the phasic response in all types of tissue, as well as the tonic response in esophageal tissue, thereby mimicking the effect of IgG from patients with primary SS.

In the presence of 1 mg/ml IgG from patients with primary SS (n = 4), the amplitude of the electric field stimulation–mediated phasic contractile responses in individual esophageal, ileal, and gastric smooth muscle preparations was significantly reduced at various increments of time up to 90 minutes, when compared to control IgG (P < 0.05, P < 0.001, and P < 0.001, respectively, by two-way ANOVA) (Figures 1A and B). Incubation with IgG from patients with primary SS also reduced the electric field stimulation–mediated tonic contraction in esophageal preparations, suggesting that tonic responses in this tissue are partly mediated by the M3R. However, the tonic response in ileal smooth muscle was unaffected by IgG from patients with primary SS (Figure 1A). These data indicate that IgG from patients with primary SS inhibits neuronally mediated contraction in smooth muscle from multiple sites throughout the GI tract.

We next sought to determine whether the primary SS IgG–mediated inhibition of smooth muscle contraction was consistent with activity at the M3R. Responses to electric field stimulation in the various types of tissue were assessed both before and after incubation with the M3R-selective antagonist, 4-diphenylacetoxy- N-methylpiperidine (4-DAMP). Incubation with 30 nM 4-DAMP for 30 minutes significantly reduced the amplitude of electric field stimulation–induced phasic contractions in esophageal (n = 4) and gastric (n = 11) smooth muscle (by a mean ± SD of 20.0 ± 1.5% and 47.2 ± 6.4%, respectively, compared to pretreatment), while in ileal smooth muscle (n = 4), the phasic response was abolished completely (Figure 1C), indicating regional variation in M3R mediation of phasic contractions in GI smooth muscle. Treatment with 4-DAMP also resulted in a partial reduction in tonic response in esophageal tissue, but not in ileal smooth muscle. Thus, these data confirm that pharmacologic blockade of the M3R inhibits neuronally mediated GI tract smooth muscle contraction, similar to the effect of IgG from patients with primary SS.

Primary SS IgG–mediated inhibition of gastric smooth muscle contraction is dependent upon expression of the M3R.

There is considerable pharmacologic evidence indicating that IgG from patients with primary SS contains autoantibodies to the M3R (9, 11, 30, 31). In the current study, we used sequential addition of increasing concentrations of carbachol to gastric smooth muscle from muscarinic receptor–knockout mice to demonstrate a specific interaction between patient autoantibodies and the target receptor. The specific responses of muscarinic receptor–knockout mice to carbachol have been described previously (18). In individual preparations of gastric smooth muscle strips from mice lacking the M3R, carbachol (0.1–30 μM) elicited a concentration-dependent contraction response, with a maximum of 6.6 ± 1.1 mN (n = 5). The carbachol concentration–response curves for the preparations were not affected by incubation with 1 mg/ml IgG from patients with primary SS (P > 0.5 by two-way ANOVA) (Figure 2A).

Figure 2.

Primary Sjögren's syndrome (SS) IgG–mediated inhibition of smooth muscle contraction and its dependency upon type 3 muscarinic receptor (M3R) expression. Averaged carbachol (CCh) concentration–response curves of gastric smooth muscle strips from M3R-knockout mice (n = 5) (A) and M2R-knockout mice (n = 5) (B) before (broken lines with open circles) and after (solid lines with solid circles) incubation with 1 mg/ml IgG from patients with primary SS were determined. Incubation with IgG did not alter the carbachol concentration responses in strips from M3R-knockout mice (n = 6), but significantly inhibited responses in strips from M2R-knockout mice (n = 5) (P < 0.01 by two-way analysis of variance). Values are the mean ± SD percentage of maximum control response.

In preparations of gastric smooth muscle from mice lacking the M2R, carbachol (0.1–30 μM) elicited a concentration-dependent contraction response, with a maximum mean of 13.5 ± 1.5 mN (n = 5). In contrast to the effect of IgG preparations in M3R-knockout mice, IgG from patients with primary SS elicited a significant decrease across the carbachol concentration–response curve in individual gastric smooth muscle strips (n = 5) from mice lacking the M2R (P < 0.01 by two-way ANOVA) (Figure 2B). IgG from controls had no effect on the carbachol concentration–response curves of gastric smooth muscle from either M2R- or M3R-knockout mice (n = 4 for both types) (Figure 3). These data indicate that expression of the M3R is required for primary SS IgG–mediated inhibition of gastric smooth muscle contraction.

Figure 3.

IgG from healthy controls does not alter carbachol responses in gastric strips from muscarinic receptor–knockout mice. Averaged carbachol concentration–response curves of gastric smooth muscle strips (broken lines with open circles) from M2R-knockout mice (n = 4) (A) and from M3R-knockout mice (n = 4) (B) were not altered by incubation with 1 mg/ml control IgG (solid lines with solid circles) (P > 0.05 by two-way analysis of variance). Values are the mean ± SD percentage of maximum control response.

IgG from patients with primary SS acts via the M3R to inhibit the frequency of migrating motor complex activity in isolated mouse colon.

Several studies support the notion that M3R has a role in regulating complex motor patterns within the colon. For example, the muscarinic receptor antagonist, atropine, and the M3R selective antagonist, 4-DAMP, reduce the frequency of the myoelectric complexes underlying CMMC activity (26, 32–34). In the current study, we used a validated assay of CMMC activity (20, 27) to determine whether the anti-M3R autoantibodies present in IgG from patients with primary SS also altered CMMC activity in the gut.

In this assay, spontaneous, regular CMMCs were recorded as contractions migrating from the proximal to the distal colon at intervals of 2–3 minutes, separated by periods of quiescence, during which inhibitory motor activity occurs (28, 35) (Figures 1A and 4A). Following 30 minutes of incubation with 5 IgG preparations from 4 individual patients with primary SS, all 5 preparations caused a significant reduction (53–77%) in the frequency of CMMCs (Figures 4B and C) (P < 0.01 by Student's t-test). In contrast, addition of 7 IgG preparations from 6 healthy controls to individual colon preparations did not alter the frequency of CMMCs (Figure 4C).

Figure 4.

IgG from patients with primary Sjögren's syndrome (pSS) induces reduction of the frequency of spontaneous, neuronally mediated migrating complex activity in isolated mouse colon. A, Representative tracing showing normal pattern of colonic migrating motor complex (CMMC) activity measured at 3 recording points along the mesenteric border of the colon from proximal (top trace) to distal (bottom trace). The spontaneous, contractile CMMC occurs every 2–3 minutes, separated by periods of relative quiescence. B, Reduced frequency of CMMCs following addition of IgG (0.2 mg/ml) from a patient with primary SS. C, Pooled data on CMMC frequency before and following addition of control or patient IgG. Points represent the average pre- and post-IgG frequency of 5 consecutive CMMCs. Post-IgG frequency was determined after a 30-minute incubation period. Patient IgG (n = 5) significantly reduced CMMC frequency compared to IgG from controls (n = 7) (CMMC frequency in patient IgG–treated colon significantly [P < 0.01] lower than pretreatment frequency, by paired t-test).

We then incubated patient IgG with colon preparations from mice lacking M2 or M3 receptors. Regular CMMC activity was observed in preparations from both M2R- and M3R-knockout mice (Figures 5A and B). In colons from M2R-knockout mice, the addition of IgG from patients with primary SS (n = 2), but not IgG from controls (n = 2), resulted in a significant reduction in the frequency of CMMCs (P < 0.01) (Figures 5A and C). In contrast, in preparations from M3R-knockout mice, addition of IgG from patients with primary SS (n = 2) did not alter the frequency of CMMCs (Figures 5B and C). These data indicate that anti-M3R autoantibodies present in primary SS IgG reduce the frequency of CMMCs in mouse colon.

Figure 5.

IgG from patients with primary Sjögren's syndrome (SS) disrupts colonic motility via activity at the type 3 muscarinic receptor (M3R). A, Normal colonic migrating motor complex (CMMC) activity in colon from an M2R-knockout (M2R-KO) mouse and reduced frequency of CMMCs following addition of IgG (0.2 mg/ml) from a patient with primary SS. B, Unaltered frequency of CMMCs in colon from an M3R-knockout mouse after addition of IgG from a patient with primary SS. C, Pooled data on CMMC frequency following addition of IgG from a patient with primary SS to isolated mouse colon. Points represent the average pre- and post-IgG frequency of 5 consecutive CMMCs. Post-IgG frequency was determined after a 30-minute incubation period. Patient IgG did not alter CMMC activity in M3R-knockout mice but significantly reduced CMMC frequency in M2R-knockout mice (CMMC frequency in patient IgG–treated colon significantly [P < 0.01] lower than pretreatment frequency, by paired t-test).

DISCUSSION

Autoantibodies targeting the M3R have been suggested as a potential pathogenic mechanism for GI tract dysfunction in patients with primary SS (3, 11). In the current study, we used validated ex vivo assays of smooth muscle strip contraction (11) and whole-colon migrating motor activity (20, 27) to demonstrate that IgG from patients with primary SS that had anti-M3R activity inhibited neurotransmission in smooth muscle taken from throughout the GI tract, and reduced the frequency of CMMC activity. The effect of the IgG on GI tissue was dependent upon expression of the M3R, showing for the first time that autoantibodies specific for this receptor mediate autonomic dysfunction in primary SS.

Our finding that IgG from patients with primary SS inhibits neuronally mediated contraction in smooth muscle from the esophagus, gastric fundus, and ileum expands upon previous findings of inhibition of colonic smooth muscle contraction by primary SS IgG (12, 29). Motor dysfunction manifesting as reduced peristaltic velocity and contraction pressure in the esophagus has been described as a feature of primary SS (5, 6), and delayed gastric emptying is recorded in the majority of patients assessed by scintigraphy (7).

Our study is the first to demonstrate that IgG with anti-M3R activity is sufficient to cause motor dysfunction in tissue from throughout the GI tract, and provides a pathologic mechanism for motor abnormalities in SS. However, the observation that patient IgG also mediates a reduction in the frequency of CMMCs suggests a broader effect of anti-M3R autoantibodies on GI motility. Interestingly, the M3R-selective antagonist, 4-DAMP, prolongs the duration of colonic motor complexes in canines (34) and reduces their frequency in isolated rabbit colon (33), which suggests that M3Rs participate in CMMC cycling.

Taken together, our results indicate that anti-M3R autoantibodies have the potential to mediate multiple GI tract dysfunctions in primary SS, ranging from reduced esophageal motor activity to altered colonic motility. While longitudinal studies correlating results from the functional assays with manometry, GI imaging, and clinical features in patients are needed to clarify the relationship between anti-M3R autoantibodies and GI symptoms in primary SS, these autoantibodies have also been implicated in primary SS–associated dysfunction of bladder detrusor and secretomotor activity in lacrimal and salivary glands (3, 11, 13, 31, 36, 37). Thus, we hypothesize that GI transit abnormalities form part of a broad spectrum of autonomic dysfunction that is mediated by anti-M3R autoantibodies in primary SS.

Criteria for demonstrating autoantibody-mediated pathogenesis in autoimmune disease include interaction of the putative antibody with the target (8, 38). To date, attempts to identify specific anti-M3R autoantibodies in the sera of patients using conventional approaches have proved problematic (9, 39–42), presumably based on the low abundance of these autoantibodies in serum and preference for conformational epitopes (for review, see ref.42). In the present study, IgG from patients with primary SS had no effect on smooth muscle contraction or CMMC activity in tissue from mice lacking the M3R. While it is impossible to determine in the current study whether the effects of the autoantibodies are direct or indirect, these results provide the most compelling evidence to date that the anti-M3R activity in IgG from patients with primary SS results specifically from antibody interactions with the M3R signaling pathway. In addition, the inability of IgG from patients with primary SS to alter spontaneous, neuronally mediated CMMC activity in M3R-knockout mice provides evidence against the presence of other pathogenic autoantibodies against targets that regulate CMMC activity in these animals. Presumably, the use of muscarinic receptor–knockout mice would also be a suitable strategy for determining the specificity of autoantibody-mediated pathogenesis in other autonomic features of primary SS, including secretory function in salivary acinar cells (31, 43, 44).

In conclusion, the current study has demonstrated that IgG from patients with primary SS inhibits neuronally mediated contraction of smooth muscle from multiple sites throughout the GI tract, and reduces the frequency of colonic migrating motor complexes. Using muscarinic receptor–knockout mice, we confirmed the M3R signaling pathway as the specific target of the pathogenic autoantibodies, and ruled out the presence of other autoantibodies capable of disrupting neurotransmission in our assays. While further work is needed to clarify the role of these antibodies in bladder and sicca symptoms, we hypothesize that altered GI transit forms part of a broader autonomic dysfunction mediated by pathogenic anti-M3R autoantibodies in primary SS.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Jackson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Park, Haberberger, Gordon, Jackson.

Acquisition of data. Park, Jackson.

Analysis and interpretation of data. Park, Gordon, Jackson.

Acknowledgements

Muscarinic receptor–knockout mice were a kind gift from Dr. Jürgen Wess (National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, Maryland).

Ancillary