SEARCH

SEARCH BY CITATION

Introduction

  1. Top of page
  2. Introduction
  3. Muscarinic receptors
  4. Fluid secretion by salivary acinar cells
  5. Antimuscarinic antibodies and SS
  6. Antimuscarinic antibodies: where are we?
  7. Where we are going
  8. REFERENCES

Sjögren's syndrome (SS) has been described as “an autoimmune disease of the exocrine glands, which particularly involves the salivary and lacrimal glands. The secretory tissues in the affected glands are progressively destroyed and replaced by a lymphoreticular cell infiltrate. This process may occur as an isolated phenomenon, in which case it is termed primary Sjögren's syndrome, or in conjunction with a connective tissue or collagen disease, in which case it is referred to as secondary Sjögren's syndrome” (1). Implicit within this description is the concept that the characteristic salivary gland hypofunction of SS is the direct consequence of immune-mediated destruction of the secretory acinar tissue. The assumption that lack of glandular function is the direct consequence of tissue loss has become so ingrained that it has directed SS research for more than 60 years.

More recently, our understanding of the pathology underlying the glandular hypofunction associated with SS has undergone a dramatic change. Two key observations have led to this change: 1) many patients with SS have within their salivary glands large amounts of acinar tissue that is unable to function in vivo, as demonstrated by the lack of salivary flow (2–4), and 2) data from work on salivary acinar cells isolated from patients with SS demonstrates that the remaining tissue is functional in vitro (5, 6), but with a reduced sensitivity to threshold levels of muscarinic stimulation (5). In addition, it is well established that glandular atrophy is the long-term consequence of diminished function and that, in the elderly, it is possible to lose significant amounts of glandular tissue without affecting salivary flow (7–9). Overall, these findings strongly suggest that the lack of glandular function in many patients with SS is the result of a perturbation of acinar function, ultimately followed by atrophy (2–4, 10).

In other autoimmune diseases in which there are clearly defined organ targets of the autoimmune response, such as myasthenia gravis and Graves' disease, pathogenic autoantibodies have been identified (11, 12). Although SS is classified as an autoimmune disease, no specific pathologic autoantibody has been found (3, 13). However, data from recent studies have suggested that patients with primary SS and patients with secondary SS may have inhibitory autoantibodies directed against muscarinic receptors (14–16). The presence of inhibitory autoantibodies directed against salivary gland muscarinic receptors would serve to unite the pathologies underlying the glandular hypofunction of both primary and secondary SS, explaining why the degree of glandular hypofunction is equivalent in the two (17). Furthermore, perturbation of muscarinic receptor function by the presence of antimuscarinic antibodies would account in large part for some of the reported extraglandular features of SS, such as bladder irritability (15, 18, 19), impairment of esophageal motor function (20) and microvascular responses to cholinergic stimulation (21), Adie pupil (22), and variable heart rate (23).

Identification of a pathogenic autoantibody would revolutionize the management of SS, in that 1) antibody identification could form the basis of a diagnostic test, 2) removal of the antibody could form the basis of a therapy akin to that for myasthenia gravis, and 3) better understanding of the pathology underlying the glandular hypofunction could lead to the development of disease-modifying treatments. In the present report, we critically evaluate the evidence for the existence of pathogenic antimuscarinic autoantibodies in SS.

Muscarinic receptors

  1. Top of page
  2. Introduction
  3. Muscarinic receptors
  4. Fluid secretion by salivary acinar cells
  5. Antimuscarinic antibodies and SS
  6. Antimuscarinic antibodies: where are we?
  7. Where we are going
  8. REFERENCES

Early pharmacologic data demonstrated that acetylcholine (ACh) could mediate physiologic responses, such as smooth muscle contraction, glandular secretion, and cardiac rate, through a family of muscarinic receptor subtypes (24). However, it was not until the cloning of the first muscarinic receptors, as reported by Kubo and colleagues in 1986 (25, 26), that the full extent of the muscarinic receptor gene family was realized (27). It is now clear that the muscarinic receptor family is encoded by 5 separate genes (28, 29). The fact that the coding region of these genes lacks an intron, a feature characteristic of a number of other G protein–coupled receptor families, means that there are just 5 muscarinic gene products, designated M1–M5 (30).

Because the muscarinic receptor gene family was cloned within a short period of time by investigators at 3 different laboratories, there was initially some confusion regarding the designation of the sequences (compare refs. 27 and28). This confusion has now been resolved, and the accession numbers shown in Table 1 refer to the human muscarinic receptor subtypes held in GenBank and conform to the designations given by Bonner (27).

Table 1. Accession numbers for members of the human muscarinic receptor gene family
ReceptorNucleotide accession nos.Protein accession nos.No. of amino acids
M1NM_000738NP_000729460
 X15263CAA33334 
M2AF498916AAM18939466
 X15264CAA33335 
M3NM_000740NP_000731590
 X15266CAA33337 
M4NM_000741NP_000732479
 X15265CAA33336 
M5NM_012125NP_036257532

The cloning of the muscarinic receptor genes has allowed for their isolated expression in immortalized cells followed by analysis of the pharmacologic and signaling properties of each of the receptor subtypes. These cloning studies have highlighted the inadequacy of the pharmacologic tools available to study muscarinic receptors. Table 2 shows the binding affinity constants of some of the more commonly used antagonists. The data illustrate the lack of specificity of these ligands; for example, pirenzepine, the widely used “selective” antagonist for the M1 receptor (M1R), has a respectable 45-fold selectivity compared with that for the M2R, but only 17.9-fold and 3.5-fold selectivity compared with that for M3R and M4R, respectively (30). Similarly, darifenacin, which is considered to be M3R selective, has only 10-fold selectivity for the M3R receptor over the M1R.

Table 2. Antagonist affinities (log affinity values) for the various muscarinic receptor subtypes*
AntagonistMuscarinic receptor
M1M2M3M4M5
  • *

    Values in boldface are the highest affinity value for the particular antagonist. Values in parentheses are the fold decrease in affinity compared with the value in boldface. 4-DAMP = 4-diphenylacetoxy-N-methylpiperidine methiodide. Adapted, with permission, from ref. 30.

Pirenzepine8.156.5 (44.6)6.9 (17.9)7.6 (3.54)6.65 (31.6)
AF-DX 3847.4 (15.8)8.67.5 (12.6)8.35 (2.0)6.3 (200)
Darifenacin7.65 (10.0)7.2 (6.3)8.657.85 (6.4)8.05 (4.0)
Methoctramine7.45 (4.0)8.056.6 (28.2)7.75 (2.0)7.05 (10.0)
4-DAMP8.9 (1.6)8.1 (10)9.18.9 (1.6)8.95 (1.4)

The poor selectivity of the muscarinic receptor ligands makes the identity of a muscarinic receptor subtype in a tissue or cell line difficult to establish. To accurately define the pharmacologic identity of a muscarinic receptor response, the antagonist affinity constants ideally should be determined for a number of antagonists and compared against the affinity constants reported for the individual receptors analyzed in cloned expression systems (see ref. 31). This can often be practically difficult due to limited availability of tissue or cells; certainly the process is time-consuming and expensive, and it may ultimately yield ambiguous results due to experimental variation or vagaries associated with a particular biologic sample.

There are, however, good examples in which this principle has been applied, for instance, in the determination of the muscarinic receptor population expressed in rat parotid gland and the parotid cell line PAR-5 (32). The affinities for a number of muscarinic selective antagonists were determined and compared against the affinities obtained for the expressed recombinant receptors. In those experiments the affinity of the “specific” M3 antagonist 4-diphenylacetoxy-N-methylpiperidine methiodide (4-DAMP) was found to be in the nM range (32). However, it can be seen from Table 2 (and in ref. 32) that 4-DAMP has poor selectivity for the M3R. It was therefore important to also include other muscarinic antagonists to develop a picture of a range of antagonist affinities. In the study by Bockman et al (32), the affinities of 4-DAMP, pirenzepine, methoctramine, and darifenacin (among others) were determined in parotid gland and PAR-5 cells and found to be consistent with that for the M3R.

To circumvent some of these pharmacologic limitations, a series of antibodies have been produced to allow for the detection of specific receptor subtypes using immunologic methods such as Western blotting, immunoprecipitation, and immunocytochemistry. The first of these antibodies were raised by Li, Wall, Wolfe, and colleagues, who, in a series of elegant studies, combined the selectivity of antimuscarinic receptor antibodies with the high-affinity binding of a radiolabeled muscarinic antagonist, N-methylscopolamine (33–35). This approach led to a highly sensitive and quantitative assay for muscarinic receptors in tissues and cell lines (35). Using this technique, the preponderance of M2R over M3R and M1R subtypes in lung, ileum, and bladder was confirmed (35). Similar studies on rabbit submaxillary glands demonstrated that in this species the M1R and M3R were expressed in equal proportions, whereas in the rat parotid gland the M3R was found to represent 93% of the muscarinic receptor population (31).

Although these antibodies have been extremely useful in defining muscarinic receptor populations, researchers need to be cautious. Over the past 5 years many of the muscarinic receptor–specific antibodies have become commercially available, but unfortunately, in our experience the majority of these antibodies have not been selective for the receptors to which they have been raised. It is imperative that researchers obtain receptors cloned by recombinant expression so that the selectivity of the antibody can be tested. We have successfully raised a number of polyclonal (36, 37) and monoclonal antibodies (Tobin AB, et al: unpublished observations) to the M3R and have rigorously tested the specificity of these antibodies on cloned receptors before using them on tissue samples (Figure 1). Similar studies carried out with commercially available antibodies have (with only a few exceptions) led us to reject the antibodies.

thumbnail image

Figure 1. Western blot identification of the human M3 muscarinic receptor (M3R). Chinese hamster ovary cells, nontransfected (NT) or transfected with human M3, were grown in a 6-well culture dish and solubilized with a detergent-based Tris buffer (36). Samples from the lysate (1 μg protein/lane) were resolved by sodium dodecyl sulfate–8% polyacrylamide gel electrophoresis. The proteins were then transferred to a nitrocellulose membrane, which was probed with an M3R-specific polyclonal antibody (36). The receptor reacted as a broad band at 100–110 kd. Note the nonreceptor bands at 96 kd and 212 kd, which could be mistaken for the receptor.

Download figure to PowerPoint

Early cloning studies revealed a number of consensus glycosylation sites in the N-terminal extracellular domain of the muscarinic receptors (27). When the receptors are resolved by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE), the glycosylation status of the receptor causes the receptor to run at a higher apparent molecular weight than that predicted from the amino acid backbone. Hence, the human M3R has a predicted molecular weight of 66 kd but runs on SDS-PAGE as a broad band at a molecular weight of 100–110 kd (Figure 1). Similar shifts in molecular weight have been reported for nearly all G protein–coupled receptors studied to date, including the β2-adrenergic receptor (38), angiotensin AT1A receptor (39), cholecystokinin receptor (40), substance P receptors (41), and vasopressin V1a receptor (42). Interestingly, mutation of the glycosylation sites in the M2R had no effect on receptor expression or function (43). Hence, the role of glycosylation in relation to the muscarinic receptor family is uncertain.

Fluid secretion by salivary acinar cells

  1. Top of page
  2. Introduction
  3. Muscarinic receptors
  4. Fluid secretion by salivary acinar cells
  5. Antimuscarinic antibodies and SS
  6. Antimuscarinic antibodies: where are we?
  7. Where we are going
  8. REFERENCES

The textbook description (44) of the control of secretion outlines a linear sequence of events, as follows: ACh binds to and activates M3R, the G protein–coupled muscarinic type 3 ACh receptor (45, 46) (action 1 in Figure 2). The activated G protein stimulates phospholipase C to generate inositol 1,4,5-trisphosphate (IP3) (action 2 in Figure 2), which binds to IP3 receptors on intracellular Ca2+ stores, causing them to release Ca2+ (action 3 in Figure 2). The rise in intracellular Ca2+ activates apical membrane Cl channels (47) (action 5 in Figure 2). Na+ follows Cl across the cell in order to maintain electrochemical neutrality, and the osmotic effect of moving Na+ and Cl across the acinus drags water into the lumen (action 6 in Figure 2).

thumbnail image

Figure 2. Control of fluid secretion in salivary acinar cells. Binding of acetylcholine (ACh) to the G protein–linked M3 muscarinic ACh receptor (action 1) stimulates phospholipase C to generate inositol 1,4,5-trisphosphate (IP3) (action 2). IP3 binds to and opens the IP3 receptor on the endoplasmic reticulum at the apical pole of the cell, causing the release of Ca2+ (action 3). Release of Ca2+ stimulates Ca2+-induced Ca release via the IP3 receptor and the ryanodine receptor, which both amplifies and propagates the Ca2+ signal (action 4). Increased [Ca2+]i activates the apical membrane Cl channel (action 5) and the basolateral K+ channel. Efflux of Cl into the acinus lumen draws Na+ across the cells to maintain electroneutrality, and the resulting osmotic gradient generates fluid secretion (action 6). cADPr = cyclic ADP ribose.

Download figure to PowerPoint

While accurate in every particular, this description is too simplistic and, above all, too static to be a useful model in any meaningful exploration of the potential role of antimuscarinic antibodies in salivary secretion. The reasons for this are described below:

  • 1
    The size and shape of the Ca2+ signal depend also on a multiplicity of factors that are not directly regulated by activation of the muscarinic receptors, including cyclic ADP ribose and cyclic AMP (48) (action 4 in Figure 2).
  • 2
    Salivary acinar cells are structurally (49) and functionally (50) polarized, and secretagogue-evoked Ca2+ signals originate in and may be restricted to the apical pole of the cell (50) (see Figure 2). A small Ca2+ signal at the right location will be a much more effective trigger for secretion than a large Ca2+ signal that is distributed across the cell (50). Factors that affect cellular polarization, e.g., anything that affects the cytoskeleton, will therefore also alter the functionality of the Ca2+ signal (51, 52).
  • 3
    Ca2+ signals do not simply diffuse across the cell, but are instead actively amplified and propagated by a process of Ca2+-induced Ca2+ release (CICR) (53) (action 4 in Figure 2). CICR, as the name suggests, is the process whereby Ca2+ causes its own release. The Ca2+ release channels, the IP3 receptor and the ryanodine receptor, both gate the endoplasmic reticulum Ca2+ store, and both are Ca2+ sensitive (54). Ca2+ mobilization is therefore an explosive example of positive feedback. Factors that affect the propagation of the Ca2+ signal, e.g., anything that modifies Ca2+ removal from the cytoplasm (44) including anything that alters mitochondrial metabolism (55), will also alter the Ca2+ signal. Therefore, in light of what is known about the process of Ca2+ mobilization, it is not sufficient to demonstrate that cells are capable of generating a change in enzyme activity or an increase in a second messenger, but rather it must be shown that they are capable of generating the appropriate Ca2+ signal.

There are few data available on the secretory physiology of salivary acinar cells from patients with SS. It is known that the degree of nervous innervation is unaffected in the human disease (6, 56). However, an increase in acinar cell expression of M3R has been demonstrated (57). Such an increase in M3R would be expected to result in acinar cell hyperfunction, a phenomenon that has been observed in the early stages of disease both in the MRL/lpr mouse model of SS (58) and in a very small number of human patients (59). Conversely, detailed in vitro experiments using labial glands isolated from patients with SS have demonstrated that with maximal doses of ACh, the [Ca2+]i response was not significantly different from that of labial gland acinar cells isolated from control subjects (5, 6). However, with submaximal concentrations of ACh, there was a significant reduction in the 50% maximum response concentration, consistent with the notion of a significant change in the number or function of muscarinic receptors in the labial glands (5) that could be related to the activity of antimuscarinic antibodies.

Antimuscarinic antibodies and SS

  1. Top of page
  2. Introduction
  3. Muscarinic receptors
  4. Fluid secretion by salivary acinar cells
  5. Antimuscarinic antibodies and SS
  6. Antimuscarinic antibodies: where are we?
  7. Where we are going
  8. REFERENCES

Effects on smooth muscle function.

It is well established from radioligand binding studies that, in both the bladder and the colon, the muscarinic receptor population primarily comprises the M2R and M3R subtypes, in a mixture of ∼80%:20% (60). Although M2R is the predominant receptor subtype present, work with knockout mice has confirmed that M3R is the receptor subtype that mediates the majority of the contractile response in smooth muscle (46, 61). Accordingly, if present, antimuscarinic antibodies should be able to disrupt the function of the smooth muscle within the bladder and the colon and could therefore account for the extraglandular symptoms of SS.

Data from in vitro studies indicate that the majority of patients with primary or secondary SS have IgG antibodies that are capable of inhibiting both agonist- and nerve-evoked contractions in isolated mouse bladder or colon muscle strips, in a noncompetitive manner (15, 62). However, these findings do not provide conclusive proof that the SS antibodies specifically interact with the M3R, since contraction in smooth muscle is known to be mediated by a number of mechanisms. A range of pharmacologic inhibitors were used to isolate the role of the muscarinic receptors in muscle contraction in those experiments. Unfortunately, pharmacologic inhibitors are rarely totally specific, and this is particularly true for muscarinic receptor antagonists, as discussed above. Therefore, although the results support the notion that antibodies against muscarinic receptors have a role in the impaired smooth muscle function observed in SS, it cannot be definitively concluded that 1) M3R is the only muscarinic subtype involved (although this is likely to be the case), 2) in vivo, the inhibition is noncompetitive, or 3) the antibody interaction is with the receptor itself rather than with a related molecule. In spite of these limitations, the sensitivity of bioassays using muscle strips to detect antimuscarinic antibodies (63) has allowed investigators to make some potentially important observations about the nature and disease specificity of antimuscarinic antibodies, as discussed below.

Neutralization of antimuscarinic antibodies by antiidiotypic antibodies.

Studies have indicated that the inhibition of smooth muscle contraction can be reproduced by the monovalent Fab fragment of IgG in patients with SS, indicating that in smooth muscle the antimuscarinic antibody activity does not require receptor crosslinking (64). Furthermore, antimuscarinic antibody activity was neutralized in vitro by antiidiotypic antibodies in both pooled intravenous immunoglobulin (IVIG) and IgG from healthy individuals. However, IVIG did not alter the levels of anti-Ro/La antibodies (64), raising the possibility that naturally occurring antiidiotypic antibodies may prevent the emergence of potentially pathogenic antimuscarinic autoantibodies. Moreover, administration of IVIG to patients with circulating antimuscarinic antibodies specifically neutralized the activity of these autoantibodies in vivo and was associated with an improvement in bladder and bowel symptoms in a few patients (65). These preliminary findings further support the notion of a pathogenic role for antimuscarinic antibodies and offer a rationale for the use of IVIG as a treatment of autonomic dysfunction in patients with SS. The neutralizing antiidiotypic antibodies are likely to be germline encoded, and if isolated may prove to be a more specific agent than whole IVIG for dysautonomia in SS.

Induction of cholinergic hyperresponsiveness by passive transfer of antimuscarinic antibodies.

Although the commonly observed effect of antimuscarinic antibodies on smooth muscle in vivo is inhibition of contraction, early work also indicated that addition of SS patient IgG itself occasionally resulted in an acute episode of muscle contraction that preceded the inhibition (15). Data from more recent work in which mice were injected with IgG from antimuscarinic antibody–positive SS patients or rabbit anti-M3R antibodies (63) showed that passive transfer of antimuscarinic antibodies produced a marked up-regulation of postsynaptic M3R expression in bladder smooth muscle, which was associated with both muscle hyperresponsiveness to stimulation by either parasympathetic nerves or exogenous muscarinic agonist and enhanced parasympathetic responses to bladder distention. In addition, an increase in M3R immunolabeling was demonstrated in the bronchial smooth muscle and salivary glands of these mice (66).

As a result of these data, it has been suggested that the in vivo effects of antimuscarinic antibodies on smooth muscle are an initial antibody-mediated blockade of cholinergic neurotransmission, followed by a reciprocal overexpression of M3R, leading to overactivity. However, up-regulation of M3R is likely to be an early effect and, in the long term, a chronic down-regulation in receptor responsiveness would be expected. This mechanism could provide an explanation for the bronchial hyperresponsiveness and early sialorrhea observed in patients with SS (59, 67). Overall, these results not only reinforce the previous findings that SS patient IgG affects the muscarinic receptors of smooth muscle, but also fulfill an important criterion for demonstrating that antimuscarinic antibodies are potentially pathogenic (68), by showing that passive transfer of SS IgG can perturb organ function.

Specificity for SS.

The data from smooth muscle bioassays demonstrate that antimuscarinic antibodies are not exclusive to SS. Functional antimuscarinic autoantibodies have been detected by smooth muscle bioassay in 15 of 24 patients with primary SS (63%), 9 of 11 patients with rheumatoid arthritis (RA) and secondary SS (82%), 0 of 6 patients with RA, 8 of 10 patients with scleroderma (80%), and 3 of 3 patients with dermatomyositis (100%) (Gordon T: unpublished observations). Therefore, they may represent a common pathogenic link between primary and secondary SS, as well as contribute to parasympathetic dysfunction in other systemic rheumatic diseases.

In summary, the pathophysiologic consequences of antimuscarinic antibodies in vivo are likely to be complex and may reflect a combination of direct antibody-mediated effects and longer-term counterregulatory mechanisms. While functional smooth muscle assays can be regarded as the current gold standard for the sensitive detection of autoantibodies that inhibit cholinergic neurotransmission, they cannot be used to identify the precise target(s) of the autoantibody, and caution must be used in extrapolating results of these assays to the multifactorial pathophysiology of glandular hypofunction in patients with SS.

Effects on salivary and lacrimal acinar cells.

Animal experiments.

The NOD mouse model is a well-established animal model for study of the salivary gland hypofunction associated with human SS (69). Antibodies capable of immunoprecipitating muscarinic receptors have been detected in the NOD mouse (70), suggesting that antimuscarinic antibody activity could underlie the pathology responsible for the salivary gland hypofunction. This theory is further supported by the findings that 1) NOD Igμ-null mice that lack B cells do not develop salivary gland hypofunction, 2) passive transfer of NOD mouse IgG or of SS patient IgG or IgG F(ab′)2 fragments to NOD Igμ-null mice results in the development of salivary gland hypofunction in the majority of recipient mice (although a few SS IgG fractions did cause hypersecretion) (see below) (71), and 3) infusion of an anti-M3R monoclonal antibody into NOD/SCID mice causes salivary gland hypofunction (72). However, a problem with data from animal whole-body experiments is that they do not demonstrate the precise site or mechanism of action of the infused antibody.

Radioligand binding.

Findings of radioligand binding studies have suggested that serum from a large proportion of patients with SS, but not from controls, contains IgG antibodies that can noncompetitively bind to the M3R of rat parotid gland (14) or rat exorbital lacrimal gland membranes (73), as well as salivary IgA antibodies that can noncompetitively bind to the M3R of rat parotid salivary gland membranes (74). While these data suggest that antimuscarinic antibodies may have a pathologic role in SS, they do not conclusively demonstrate that the interaction is exclusively with M3R. This is because experimental differentiation between the muscarinic receptor subtypes present in a tissue is difficult since there are no specific selective antagonists available and, in rodent salivary glands, M3R, M4R (75), and M1R (76) are all present.

As noted above, experimental differentiation between the muscarinic receptor subtypes present in a tissue is difficult since there are no specific selective antagonists available. In only one of the radioligand binding studies examining the ability of SS patient IgG or IgA to bind to muscarinic receptors (14, 73, 74) did the investigators attempt to use a rank order of antagonists to characterize the site of SS IgG interaction with muscarinic receptors (14). Unfortunately, only 4 antagonists were used, and these included AF-DX 116, a compound known to produce variable results with M3R (77). The remainder of the studies used 4-DAMP as an M3-“selective” muscarinic antagonist (see Table 2). Therefore, it is impossible to differentiate the M3 subtype from M1 or M4. In addition, data from studies in which radioligand binding has been used to detect SS IgG recognition of muscarinic receptors present on parotid gland membranes isolated from Igμ-null mice suggest that the inhibition is competitive rather than noncompetitive (71).

Overall, the results of radioligand binding experiments have confirmed that lacrimal and salivary muscarinic receptors are recognized by SS IgG antibodies. However, the precise muscarinic receptor targets and the nature of the interaction have not been confirmed using this approach.

Immunologic approaches.

Conventional immunologic approaches offer the ability to determine the site specificity of an antigen–antibody reaction. Multiple techniques have been used to investigate the interaction between both salivary and lacrimal gland muscarinic receptors and SS IgG. Data from experiments using rat lacrimal glands have suggested that SS IgG recognition of M3R can be detected by immunofluorescence, since the immunofluorescent signal could be quenched by preincubation of the SS IgG with a synthetic peptide corresponding to the second extracellular loop of M3R (78). Further support for this is provided by the demonstration that the same M3R peptide could be used to detect anti-M3R antibodies in SS IgG (79) and SS salivary IgA by enzyme-linked immunosorbent assay (ELISA) (74).

The above data would appear to strongly support the notion that the antimuscarinic antibodies in SS IgG recognize M3R. However, as we have noted previously (16, 80), the published sequence for the second extracellular loop of M3R differs from the peptide used (74, 78, 79). The peptide used actually corresponds to the second extracellular loop of M4R (Table 1). The reason for this confusion has been discussed above. The data from immunohistochemistry and ELISA experiments thus inadvertently imply that SS IgG recognizes M4R, a muscarinic receptor subtype not associated with secretion by salivary acinar cells. Furthermore, in a subsequent ELISA study using the correct sequence of the second extracellular loop of M3R, it was concluded that these peptides were not an appropriate tool for detection of anti-M3R antibodies (80).

Western blotting using crude lacrimal membrane fractions as a source of M3R has also been reported as a suitable method for the detection of anti-M3R antibodies in SS IgG (73). However, using membranes obtained from Chinese hamster ovary (CHO) cells that had been stably transfected with functional human M3R, we were unable to detect anti-M3R activity in SS IgG by Western blotting (16). This fundamental discrepancy between the findings of two studies can be explained by considering the molecular weight of the protein recognized by SS IgG. In the earlier report (73) it was stated that SS IgG recognized a 70-kd lacrimal membrane protein. While this is in good accordance with the molecular weight of M3R predicted on the basis of the amino acid sequence (∼65 kd), it is difficult to reconcile with the molecular weight of human M3R (102 kd) expressed in CHO cells following posttranscriptional modification (16, 36). Given that the homology shared between human and rat M3R is >91% and they differ in size by only 1 amino acid, the rat protein is likely to be the same size as the human protein, and the rat lacrimal membrane protein recognized by SS IgG (73) may therefore not be M3R.

In other autoimmune diseases such as Graves' disease and myasthenia gravis, recognition of the epitope by an autoantibody is highly dependent on the conformation of the epitope (11, 12, 81, 82). This finding could provide an explanation of why, in the above studies, anti-M3R activity in SS IgG could not be reliably detected by either ELISA or Western blotting. In an attempt to reconcile these issues we used human M3R–transfected CHO cells as the basis of a whole cell ELISA. Under our experimental conditions, the conformation of the expressed M3R was maintained; nevertheless, we were unable to definitively detect anti-M3R activity in the SS IgG (16). In a study using an analogous approach, with M3R-transfected CHO cells as the basis of a flow cytometric assay, the results indicated that anti-M3R may be present in SS IgG (83). However, the reported molecular weight of the transfected receptor expressed by the CHO cells used in that study was only 70 kd, and only a portion of the transfected CHO population was recognized by SS IgG. Therefore, further studies are needed to confirm these findings.

Overall, the findings from studies using immunologic approaches serve to confuse the issue of the site specificity of the SS IgG antigen–antibody interaction, rather than clarify it. However, the data do support the concept that epitope confirmation is important. A possible reason for the lack of ability to detect anti-M3R in SS IgG by immunologic approaches could be that the antibodies are present in the circulation only at low levels, as is well established in myasthenia gravis (11). This may explain why bioassay has proved to be the most reliable method to date for the detection of putative anti-M3R activity in SS IgG.

Bioassay approaches with salivary acinar cells.

Compared with the volume of data obtained by muscle bioassay, information from salivary gland bioassays is in its infancy. Nevertheless, SS is a disease characterized by salivary gland hypofunction, and data demonstrating that SS IgG can interfere with salivary cell function are vital if SS IgG is to be linked with the pathology of salivary gland hypofunction in the disease.

Data from a study in which HSG cells (human salivary ductal cell line) were continuously exposed to SS IgG demonstrated that SS IgG was able to partially inhibit the carbachol-evoked, but not the ATP-evoked, [Ca2+]i response (84), suggesting the presence of antimuscarinic antibody activity. Although these results are highly supportive of those obtained using smooth muscle bioassay, HSG cells are a ductal cell line while the salivary gland cells that are responsible for fluid secretion and involved in SS are the salivary acinar cells (5). Using mouse submandibular acinar cells, we were able to demonstrate that, in vitro, SS IgG partially inhibited carbachol-evoked [Ca2+]i in a reversible manner (16), confirming that, at least in vitro, SS IgG may also have a functional role in salivary gland hypofunction. Unfortunately these data do not unequivocally pinpoint the acinar cell target recognized by SS IgG. However, preliminary experiments analogous to those conducted with smooth muscle (63) indicate that human submandibular acinar cells exposed to antibodies raised against the second extracellular loop of human M3R precisely mimic the effects of SS IgG, i.e., they reversibly inhibit the agonist-evoked [Ca2+]i signal (Dawson L, et al: unpublished observations).

The reversibility of the binding observed when mouse or human submandibular acinar cells are exposed to SS IgG or rabbit polyclonal anti-M3R is intriguing since it confirms the previous findings in Igμ-null mice (71) but is contradictory with data obtained from smooth muscle bioassay (15, 62, 64) and radioligand binding studies (14, 73). In myasthenia gravis it is established that, in vitro, the reversibility of the interaction between the autoantibodies and the nicotinic cholinergic receptors is highly dependent on the length of time of exposure and the concentration of the autoantibody (85). Therefore, the differences in the nature of the binding of SS IgG with muscarinic receptors could be related to the experimental conditions and may also contribute to the equivocal findings with immunologic approaches.

Alternatively it has been suggested that SS IgG may recognize an as-yet-unidentified molecule that is closely related to the muscarinic receptor (13, 16). A precedent for this exists in myasthenia gravis, in which antibodies that affect nicotinic ACh receptors, but are directed against the closely associated molecule muscle-specific receptor tyrosine kinase (68, 86), have been identified in 70% of patients who are seronegative for autoantibodies against nicotinic ACh receptor.

The experimental findings to date relating to the interaction of SS IgG with glandular acinar cells must be interpreted with caution, but the available data do demonstrate that SS IgG can affect the function of salivary acinar cells both in vivo and in vitro. Recognition of the salivary acinar cell epitope(s) by the autoantibody is likely to be conformationally dependent. However, the precise epitope involved remains to be confirmed, although the second extracellular loop of M3R is currently the most attractive candidate.

Antimuscarinic antibodies: where are we?

  1. Top of page
  2. Introduction
  3. Muscarinic receptors
  4. Fluid secretion by salivary acinar cells
  5. Antimuscarinic antibodies and SS
  6. Antimuscarinic antibodies: where are we?
  7. Where we are going
  8. REFERENCES

There have been several attempts to define criteria for autoimmune diseases (87–89), and recently, Drachman suggested a set of 5 criteria that a putative autoantibody must fulfill in order to be considered pathogenic: 1) “Autoantibodies are present in patients with the disease,” 2) “Antibody reacts with the target antigen,” 3) “Passive transfer of antibody reproduces features of disease,” 4) “Immunization with antigen produces a model disease,” and 5) “Reduction in antibody levels ameliorates the disease” (68). It seems appropriate to consider the case for antimuscarinic antibodies in SS in relation to these criteria, and the individual criteria are addressed below.

“Autoantibodies are present in patients with the disease.”

The available data suggest that a significant number of patients with primary SS and secondary SS have serum IgG antibodies that are capable of binding to and influencing the function of muscarinic receptors on salivary acinar cells and the smooth muscle of bladder and colon, in vitro. However, due to the lack of an effective screening assay, only a small number of subjects have been tested. Furthermore, early data indicate that antibodies with similar actions also occur in RA, scleroderma, and dermatomyositis. Further work is needed to establish the prevalence of anti-M3R in SS.

“Antibody reacts with the target antigen.”

Although the data strongly indicate that the second extracellular loop of M3R is the target antigen, this has not been demonstrated conclusively, and the precise epitopes are currently unknown.

“Passive transfer of antibody reproduces features of disease.”

Passive transfer is arguably the most important evidence to highlight the pathogenic role of an antibody. In the case of SS IgG, results obtained following the transfer of SS IgG to mice have indicated that the recipient mice develop glandular hypofunction (71) and exhibit up-regulated M3R expression in bronchioles and marked hyperresponsiveness of bladder smooth muscle (66). However, these data should be considered to be preliminary and must be confirmed in substantially larger trials.

“Immunization with antigen produces a model disease.”

The target antigen is currently unknown, so these crucial experiments are lacking.

“Reduction in antibody levels ameliorates the disease.”

The first stages in demonstrating this criterion have recently been undertaken, and it was demonstrated that 1) antimuscarinic antibody activity was neutralized in vitro by antiidiotypic antibodies in both pooled IVIG and IgG from healthy individuals (65) and 2) administration of IVIG to patients with circulating antimuscarinic antibodies specifically neutralized the activity of these autoantibodies in vivo and was associated with an improvement in bladder and bowel symptoms in a few patients (65). Although the data from the studies with IVIG are supportive rather than substantive, they currently are the best evidence available to address this criterion.

Although a substantial amount of further information is needed before it can be definitively stated that antimuscarinic antibodies have a pathogenic role in SS, the available data strongly support this suggestion.

Where we are going

  1. Top of page
  2. Introduction
  3. Muscarinic receptors
  4. Fluid secretion by salivary acinar cells
  5. Antimuscarinic antibodies and SS
  6. Antimuscarinic antibodies: where are we?
  7. Where we are going
  8. REFERENCES

We believe future research on antimuscarinic antibodies should focus on clearly identifying the epitopes recognized by SS IgG. This information is pivotal if we are to 1) understand the antigen–antibody interactions of this antibody; 2) develop a simple screening test to allow determination of antibody incidence within the SS population; 3) induce the disease in experimental models through immunization with the antigen; and 4) neutralize antibody activity in vivo, thus fulfilling key elements of the criteria proposed by Drachman (68). A starting point for determining possible antigens may be derived from in vitro data showing that cleavage of M3R by granzyme B results in novel fragments (90).

Diversification of the research into salivary gland hypofunction in SS is also important. There has been a tendency for investigators to concentrate on a limited number of pathogenic mechanisms that could culminate in loss of salivary gland tissue; to this end, the roles of lymphocytes, cytokines, apoptosis, and various autoantibodies have all been investigated. Although this review has concentrated on the evidence supporting a role for antimuscarinic antibodies in the glandular and extraglandular symptoms of SS, it would be a gross oversimplification to assume that these antibodies are the sole etiologic factor in salivary gland hypofunction. While it is likely that antimuscarinic antibodies have a key role in the development of salivary gland hypofunction, other factors have been proposed, including cytokines, cholinesterase (91, 92), and aquaporins (93, 94). The research challenge now is to firmly establish the role of antimuscarinic antibodies in the pathology of SS and to determine which of the many additional mechanisms under consideration are pivotal and which are simply epiphenomena.

REFERENCES

  1. Top of page
  2. Introduction
  3. Muscarinic receptors
  4. Fluid secretion by salivary acinar cells
  5. Antimuscarinic antibodies and SS
  6. Antimuscarinic antibodies: where are we?
  7. Where we are going
  8. REFERENCES