A translation to Dutch of this article is available here.
During the Community Symposium on the molecular basis of ME/CFS (R) two different groups of researchers reported on an increased level of antibodies to beta-adrenergic and muscarinic receptors in sera from ME/CFS patients vs healthy controls (Figure 1). These new data have been collected independently by Alan Light (University of Utah) and Jonas Bergquist (Uppsala Universitet). Bergquist also reported that these autoantibodies can’t be found in cerebrospinal fluid from ME/CFS patients.
What was already known on these autoantibodies
The presence of a higher than normal reactivity of sera from ME/CFS patients to muscarinic receptors was reported for the first time by a Japanese group, more than a decade ago (Tanaka S et al. 2003) and it has been confirmed recently in a work by Osaka City University Graduate School of Medicine (Yamamoto S et al. 2012) and in another paper by University of Bergen (Norway) and Charité University (Germany) (Loebel M et al. 2016). In particular, while Tanaka and colleagues measured an increased level of autoantibodies against muscarinic cholinergic receptor 1 (CHRM1) in about half of patients, the European group described an increase in the reactivity of sera to subtypes M3 and M4, in a subset of patients (Figure 2). They used two completely different assays, as we will see later, and this might be the reason for the different results.
As you can see from figure 2, the study by Loebel et al. also indicated an increase in antibodies to beta-adrenergic receptors (subtype 2), in agreement with the latest data from Light and Bergquist. In this regard, it is worth noting that autoantibodies to muscarinic receptors M2 and M3, and to beta-adrenergic receptors (subtype 1 and 2) have been already reported in orthostatic hypotension (OH) (Yu X et al. 2012), (Li H et al. 2012) and that antibodies to beta 2 adrenergic receptors have been identified in patients with postural-orthostatic tachycardia syndrome (POTS) (Li et al. 2014). This means that this group of autoantibodies is associated with orthostatic intolerance (POTS and/or OH), but orthostatic intolerance is part of the clinical picture of ME/CFS (IOM 2015) and those patients who have a diagnosis of POTS often have many features in common with ME/CFS patients, see for instance (Okamoto L et al. 2012), (Wise S et al. 2015). So, it might be conceivable that these autoantibodies play a role in the pathogenesis of some symptoms in a subgroup of patients, although this has not been proven, so far.
We don’t know the reason why the immune system of some ME/CFS patients reacts to these receptors, but Alan Light suggested, during the symposium, that a possible source for these antibodies might be a mechanism known with the name of molecular mimicry (MM). The basic idea behind MM is that B cells can erroneously produce antibodies to human proteins when epitopes of an infectious agent closely resemble epitopes found in the host (Rose NR 1998). MM is currently believed to explain the pathogenesis of Guillain-Barré syndrome, where lipo-oligosaccharides on the Campylobacter jejunii outer membrane seems to elicit (in predisposed individuals) immune response to human gangliosides, due to the similarity between these antigens (Van den Berg B et al. 2014). Now, if molecular mimicry was involved in the origin of antibodies to beta 2 adrenergic receptors, which could be the epitope on the receptor? And which the pathogen-borne antigen? In order to provide a possible answer to this question we have to consider that the regions of a receptor that can be involved in B cell autoimmunity are only those that have extracellular exposure; the other regions are immersed in the plasma membrane and in the cytoplasm, so they can’t interact with antibodies. As you can see from Figure 3, beta 2 adrenergic receptor (ADRB2) has four extracellular regions, in particular peptides 1-34, 96-106, 175-196, 299-305. In general, epitopes are mainly conformational and that means that they are regions of the protein surface, produced by the folding of the protein itself. Nevertheless, in our example, we will search only for linear epitopes.
I have used QuickBLASTP provided by NCBI, with default settings (E=100, a word of 6 letters, BLOSUM62 as substitution matrix) and I have considered for each of the four extracellular peptides both the sequence of residues from the N-terminus to the C-terminus and the inverted sequence. We obtain as the only match the protein sensor histidine kinase MtrB belonging to Pseudonocardia sp. Ae331_Ps2 (R) (Figure 4). I can’t find this particular protein in UniProt, but if it was a membrane protein and if peptide 67-77 was exposed to the extracellular space, this peptide could perhaps be a candidate as a trigger for anti-ADRB2, according to the MM theory. It is important to note here that although molecular mimicry is a popular theory (perhaps because of its simplicity) it has been proven to be a cause of autoimmunity only in Guillain-Barré syndrome.
Members of the genus Pseudonocardia have been recently proposed as human pathogens (N. Asdamongkol et al. 2012), (Amalia Navarro-Martínez et al. 2017) and are known to be antifungal commensal microorganism (Sen R. eta al. 2009) which means that an immunological memory for this genus might reflect exposure to fungi too. It is also worth mentioning that sensor histidine kinase MtrB from other bacteria is known to be present on the cytoplasmic membrane, thus it has possible antigenicity potential during infection. This genus has been found – in a 2016 metagenomic study – both in patients and controls, but it is interesting that the lowest number of readings is the one of ME/CFS patients (see row number 80 of this .xlsx file). Might this indicate immunization against this genus of bacteria in patients, and thus previous exposure?
So, what about a test for these autoantibodies?
If antibodies to adrenergic and muscarinic receptors were involved in the pathogenesis of some cases of ME/CFS, it would be interesting for patients to test for them. In this regard, it is worth noting that the measure of antibodies to membrane receptors should be done using an assay in which these receptors are expressed by living cells in their physiological position (CBA, cell-based assay). In fact, with assays in which receptors are coated on plates, we may have both false positives (due to the fact that sera react with peptides that are not in the extracellular domain) and false negatives (due to protein denaturation, which leads to the formation of epitopes that would not be present if the protein were correctly folded). The superiority of CBA over the other kind of test is well accepted in the case of anti-MOG antibodies (Ramanathan S et al 2016). It is worth noting that both the study by Loebel et al. and the previous one (Tanaka et al.) used recombinant proteins coated on plates. As far as I know, there are no commercial CBA assays for anti-muscarinic cholinergic receptors and beta adrenergic receptors, at present. The only assay available does not seem to be a CBA, from the provided documentation (R).
In silico experiment
We will now try to simulate what could happen with a test for the search of anti-ADB2R antibodies if the protein was coated on a plate. We will use the prediction of DiscoTope 2.0, which is a software that calculates all possible B cell epitopes of a given protein, using both the geometry of the protein (in particular a parameter called protrusion index, calculated from the protein’s ellipsoid of inertia) and statistical data on known B cell epitopes (Kringelum, et al., 2012). If we use the 3D structure of ADB2R experimentally determined in (Rasmussen et al. 2007) with standard settings, DiscoTope predicts peptide 231-242 as the only possible epitope (consider that the experimental 3D structure of ADB2R is incomplete). As you can see from figure 5 this peptide belongs to the intracellular domain of the receptor and so it by no means could be a B cell epitope, in vivo. In conclusion, according to this simulation, there is a risk of false-positive results with any test that uses recombinant ADB2R coated on a plate.