Article Text

Download PDFPDF

Motor neuropathies and serum IgM binding to NS6S heparin disaccharide or GM1 ganglioside
  1. Alan Pestronk,
  2. Miguel Chuquilin,
  3. Rati Choksi
  1. Department of Neurology, Washington University in St Louis School of Medicine, St Louis, Missouri, USA
  1. Correspondence to Dr A Pestronk, Department of Neurology, Washington University in St Louis School of Medicine, 660 South Euclid Avenue, Box 8111, St Louis, MO 63110, USA; pestronka{at}


Background Serum IgM binding to GM1 ganglioside (GM1) is often associated with chronic acquired motor neuropathies. This study compared the frequency and clinical associations of serum IgM binding to a different antigen, a disulphated heparin disaccharide (NS6S), with results of IgM binding to GM1.

Methods Serums and clinical features were retrospectively compared from 75 patients with motor neuropathies and 134 controls with amyotrophic lateral sclerosis (ALS), chronic immune demyelinating polyneuropathy (CIDP) and sensory neuropathies. Clinical correlations of positive IgM anti-GM1 testing found in 27 of 2113 unselected serums were also reviewed. Serum testing for IgM binding to NS6S and GM1 used covalent antigen linkage to ELISA plates.

Results High titre IgM binding to NS6S and GM1 each occurred in 43%, and to one of the two in 64%, of motor neuropathy patients. Motor neuropathy syndromes were present in 25 of 27 patients with high titre serum IgM binding to GM1 in the unselected serums. IgM anti-GM1 or NS6S antibody related motor neuropathy syndromes usually have asymmetric, predominantly distal, upper extremity weakness.

Conclusions IgM binding to NS6S disaccharide is associated with motor neuropathy syndromes and occurs with similar frequency to IgM binding to GM1. Testing for IgM binding to NS6S in addition to GM1 increases the frequency of finding IgM autoantibodies in motor neuropathies from 43% to 64%. High titres of serum IgM binding to GM1, tested with covalent ELISA methodology, have 93% specificity for motor neuropathy syndromes. High titres of serum IgM binding to NS6S have specificity for immune motor neuropathies compared with ALS and CIDP.

Statistics from

Distinction between potentially treatable immune motor neuropathies and other disorders of lower motor neurons depends partly on electrodiagnostic evaluations and serum antibody testing.1 2 Nerve conduction testing showing motor conduction block at non-entrapment sites along the length of nerves, especially in distal regions, is widely recognised as a marker of the treatable syndrome, multifocal motor neuropathy (MMN).3–5 Serum IgM binding to GM1 ganglioside is common in, and has specificity for, multifocal and other immune motor neuropathies.1 6–9 Testing for IgM anti-GM1 antibodies is especially helpful when electrodiagnostic findings of demyelination are equivocal or negative. As a quality control to examine the specificity of antibody testing in our clinical neuromuscular laboratory, we examine serum reactivity to multiple antigens in addition to those requested by ordering clinicians. We noted that serum IgM from motor neuropathy patients often bound in low titres to myelin associated glycoprotein (MAG) by ELISA testing but not by western blot. We hypothesised that the IgM might weakly recognise the sulphated carbohydrate moieties that comprise part of the extracellular post-translational glycosylation on the MAG protein. We tested this idea by examining the reactivity of IgM in several motor neuropathy serums to a variety of sulphated antigens. We found that IgM in some serums bound in high titre to IdoA-GlcNS-6S, a disulphated glucosamine–uronic acid heparin disaccharide (NS6S) (figure 1). We now report ELISA testing results, and clinical and laboratory correlations, of IgM binding to NS6S disaccharide and GM1 ganglioside in serums from patients with motor neuropathies and other neuromuscular disorders.

Figure 1

Structure of NS6S (α-4-deoxy-l-threohex-4-enopyranosyluronic acid-[1-4]-d-glucosamine-S-6S; IdoA-GlcNS-6S).


Patient and serum samples

Subjects for evaluation of serum IgM binding were retrospectively identified from a clinical database of patient diagnoses without knowledge of results of prior antibody testing. We identified 75 patients who were diagnosed by their physicians as having acquired chronic motor neuropathies and then met several additional inclusion and exclusion criteria. All were required to have had clinical, electrodiagnostic and laboratory evaluation in our neuromuscular centre at Washington University in St Louis and have serum available for ELISA studies. All patients had at least one limb with: (1) neuropathic weakness in the distribution of two or more peripheral nerves without anatomical explanations for the weakness; and (2) normal sensory function by clinical and electrodiagnostic testing. Patients were excluded if they had a family history of a similar disorder, cranial nerve or bulbar weakness, upper motor neuron signs, rapidly progressive disease leading to inability to walk or death during the first 3 years after symptom onset or disabling sensory loss. The exclusion criteria eliminated four patients. No patients from a previous ELISA study of IgM anti-GM1 antibodies in MMN9 were included in this study. Charts were reviewed in the 75 patients with motor neuropathies without reference to antibody status. We documented clinical and electrodiagnostic features, including conduction block and serum M proteins (72 patients tested by immunofixation). Electrodiagnostic studies usually included nerve conduction studies of at least six motor nerves, including four in the upper extremity, and three sensory nerves. Fewer nerves were studied in some patients with multiple conduction blocks. Conduction block along very proximal nerve segments was not routinely tested for, and was not considered in patient subclassifications. EMG studies were available in 60 patients, with paraspinous evaluation in 25. Disease control patients were selected from our clinical database from the past 3 years based only on diagnosis and serum availability. Records were reviewed to confirm diagnoses. Diagnoses included the following: definite amyotrophic lateral sclerosis (ALS, 50 patients); chronic immune demyelinating polyneuropathy (CIDP, 28 patients); and sensory neuropathy (56 patients). Sensory neuropathy patients had distally predominant reduced sensation and normal strength. Two sensory neuropathy patients had demyelinating neuropathies and IgM binding to MAG. Six others had axonal sensory neuropathies with serum IgM binding to trisulphated heparin disaccharide (TS-HDS).10 Nine patients with sensory neuropathy also had diabetes. The other 39 were idiopathic. From 2002 to 2009, all serums submitted to the Neuromuscular Clinical Laboratory at Washington University were tested for IgM binding to GM1 ganglioside, including 2113 different patients clinically evaluated in our neuromuscular centre. To examine the specificity of IgM anti-GM1 serum testing, we reviewed the records of the 27 patients from our centre with positive results. The Washington University Human Studies Committee approved all procedures. Informed consent was not required.

Assays for IgM binding to NS6S and GM1 ganglioside

NS6S is α-4-deoxy-l-threohex-4-enopyranosyluronic acid-[1-4]-d-glucosamine-S-6S (IdoA-GlcNS-6S; Sigma H1020) (figure 1). NS6S contains two sugar moieties, L-iduronic acid linked to D-glucosamine and two sulphate groups. The sulphate groups are linked to the D-glucosamine moiety, one N-linked and another O-linked at position six. Other antigens tested in preliminary ELISA evaluations were sulphatide (purified from Sigma S-1006) and heparin disaccharides TS-HDS (IdoA-2S-GlcNS-6S; Sigma H9267),10 IdoA-2S-GlcNS (Sigma H9392), IdoA-2S-GlcN-6S (Sigma H8892), IdoA-2S-GlcN (Sigma H9142), IdoA-GlcN-6S (Sigma H9017), IdoA-GlcNS (Sigma H1145) and IdoA-GlcN (Sigma H9276). Serums were assayed, using ELISA methodology with covalent antigen linkage to plates, as previously described9 10 for antiganglioside antibodies. We used Costar microwell ELISA plates (Stripwell-Amine 2388; Corning NY USA) for NS6S and other disaccharides, CovaLink NH microwell ELISA plates (Nunc, Roskilde, Denmark) for GM1 ganglioside and Immulon 2HB (Thermo, Milford, MA, USA) for preliminary studies with sulphatide and some other antigens. To measure serum IgM binding to NS6S, we used 1.25 μg of NS6S dissolved in 50 μl of 0.05% N-hydroxysuccinamide per well. Levels of selective binding to NS6S were calculated by subtracting levels of IgM binding to GD1a ganglioside, an antigen that is used in our laboratory to calculate background IgM binding for many carbohydrate containing antigens.9 High titres of selective IgM binding to NS6S (≥7000) were greater than 4 SDs above the mean of a separate initial series of tests in serums from 10 patients with ALS and 10 control subjects. High titres of IgM binding to GM1 ganglioside were ≥2000, the criterion used for these antibodies in our Neuromuscular Clinical Laboratory.


χ2 and t tests were generally used to calculate the significance of differences between diagnostic groups. Fisher's exact tests were used to calculate the significance of differences within the motor neuropathy group. Patients with missing data were excluded from that analysis only. AP performed the statistical analysis. Results are expressed as mean±SE.


Clinical and electrodiagnostic features in motor neuropathy patients

In the series of motor neuropathy patients ascertained from our clinical database, most had asymmetric, distal and arm predominant weakness (table 1). Seventy-two motor neuropathy patients had weakness in two or more limbs that had normal sensation on clinical examination. Onset ages ranged from 23 to 76 years with a mean of 46±2 years. Clinical sensory testing was abnormal in the arms in two patients. Both had distal predominant reduction in pin sensation in an arm with weakness but normal electrodiagnostic sensory testing. Clinical testing was abnormal for pin or vibration sensation in the legs in 28 patients, 26 of whom had distal predominant, bilateral, symmetric loss in the legs with normal strength. Distal sensory loss on clinical examination of the legs was present more often (p=0.0003) in patients aged more than 50 years (68%) than in younger patients (22%). Motor conduction block was present in 56% (42 of 75), and full clinical and electrodiagnostic criteria for MMN4 in 31% (23 of 75), of motor neuropathy patients. Among the 33 patients with no conduction block, three had other clear features of demyelination (one each with slowed nerve conduction velocities, prolonged distal latencies and temporal dispersion) while 30 had motor axon loss without evidence of demyelination on distal nerves. Sural sensory nerve action potentials were absent, or had reduced amplitude, in 16 patients. On EMG, acute or chronic denervation changes in muscles in weak limbs were present in 55 of 60 patients tested. The five patients with no denervation on EMG all had focal conduction block in at least one location. Thoracic paraspinous denervation was found in 11 of the 25 patients tested.

Table 1

Motor neuropathy patient characteristics

Preliminary ELISA testing

Comparison of 18 motor neuropathy serums for IgM binding to disaccharides and sulphatide showed that the highest titres were obtained using the NS6S disaccharide. Two other antigens with at least two sulphate moieties, IdoA-2S-GlcNS and TS-HDS, gave titres that averaged approximately 70% of NS6S. All other antigens tested gave titres that averaged less than 50% of NS6S. The lowest titres, within control ranges, were found with sulphatide and non-sulphated disaccharides. IgM binding to NS6S was not higher than controls when tested with the antigen added to Immulon or Covalink ELISA plates.

IgM binding to NS6S in motor neuropathies and controls

High titres of serum IgM binding to NS6S (≥7000) were found in 43% of motor neuropathy patients and ranged from 7000 to 180 000 (figure 2). High titre IgM binding to NS6S occurred more commonly in motor neuropathy patients than in the overall disease control group (p<10−6). No patient in the control groups with ALS or CIDP had high titre IgM binding to NS6S. High titres of IgM binding to NS6S were less frequent (p=0.01) in sensory neuropathy patients than in the motor neuropathy group. Levels of IgM binding to NS6S were higher (p=0.003) in the motor neuropathy group (mean 20 180±3854) than in the sensory neuropathies (mean 6230±2393). The sensory neuropathy serum with the highest titre of IgM binding to NS6S also had very high titre IgM binding to MAG (90 000).

Figure 2

Frequency of IgM binding to NS6S and GM1 ganglioside in different patient groups. High titre IgM binding to NS6S (≥7000) and GM1 ganglioside (≥2000) was more common in motor neuropathies than controls. Horizontal lines represent titres of 7000 for IgM binding to NS6S and 2000 for IgM binding to GM1 ganglioside. Controls for IgM binding to GM1 ganglioside include all 134 CIDP, ALS and sensory neuropathy control patients. ALS, amyotrophic lateral sclerosis; CB+, patients with motor conduction block found in at least one peripheral nerve; CB–, patients with no conduction block; CIDP, chronic immune demyelinating polyneuropathy; Controls, all ALS, CIDP and sensory polyneuropathy controls; Motor PN, motor neuropathy; Sensory PN, sensory polyneuropathy; #Serums, total number of serums tested in each clinical category; #Titres= 0, number of serums tested in each clinical category with no reactivity to antigen (NS6S or GM1 ganglioside) and no data points shown on the logarithmic plot.

IgM binding to GM1 ganglioside or NS6S in motor neuropathies and controls

High titres of IgM binding to GM1 were present in 43% of motor neuropathy patients. No patient with ALS, CIDP or sensory neuropathy had high titres of serum IgM binding to GM1 ganglioside. High titres of IgM binding to at least one of NS6S or GM1 ganglioside were present in 64% (48 of 75) of motor neuropathy patients. High titres of IgM binding to both NS6S and GM1 were present in 21% (16 of 75) of motor neuropathy patients, and to NS6S only or GM1 only in 21% (16 of 75) each. High titres of IgM binding to NS6S were not more frequent (p=0.25) in serums with IgM binding to GM1 (52%) than without (36%).

IgM antibodies in relation to other motor neuropathy patient features

Either motor conduction block or IgM antibody binding to NS6S or GM1 ganglioside was found in 83% (62 of 75) of our motor neuropathy patients. The frequency and titres of IgM binding to NS6S and GM1 were similar (p=0.5) in subgroups of motor neuropathy patients with and without motor conduction block (table 2). Titres of IgM binding to NS6S and GM1 averaged 19 042±4945 and 20 730±9696 in motor neuropathy patients with conduction block and 21 664±6184 and 32 703±14 036 without conduction block. Frequencies and titres of IgM binding to NS6S and GM1 were not different in patients with and without criteria for MMN. There was a trend (p=0.08) towards a reduced frequency of high titre IgM binding to GM1 ganglioside in full criteria MMN compared with other motor neuropathy patients. Frequencies of high titre IgM binding to NS6S and GM1 ganglioside were similar in patients with and without paraspinous denervation, distal sensory loss or abnormal sensory nerve action potentials (SNAPs). The frequency of high titre IgM binding to NS6S or GM1 was higher (p=0.0001) in patients with serum IgM M-proteins than in those with no IgM M-protein.

Table 2

NS6S and GM1 antibody associations

Clinical features in patients with positive testing serum testing for IgM binding to GM1

Review of the 2113 serums from our clinic patients obtained over a 7 year period revealed 27 patients with high titres of IgM binding to GM1. Of the positive results, 24 were from patients with motor neuropathies who are included in this series, one had slowly progressive, unilateral hand weakness with motor axon loss but no demyelination and a diagnosis of monomelic amyotrophy, and two had mild idiopathic axonal, symmetric sensory polyneuropathies with predominant involvement of the legs. These two patients were the only ones of the 27 GM1 positive serums or of the collection of 75 motor neuropathy patients who also had high titres of IgM binding to sulphatide.


Serum IgM and IgG anti-GM1 ganglioside autoantibodies are associated with motor predominant neuropathy syndromes.1 11 The association of IgM anti-GM1 antibodies with chronic immune motor neuropathies like MMN has been confirmed in two large studies and a meta-analysis.1 7 8 The IgM anti-GM1 testing results in this study, with a frequency of 43% in our motor neuropathy patients, reflect our experience for all serums from acquired motor neuropathy patients over the past decade and are similar to other recent studies of MMN and motor neuropathies with sensitivities of IgM anti-GM1 antibody testing ranging from 25% to 50%.1 7 8 11–13 Testing for additional antibodies, including IgM binding to GM2 and GalNAc-GD1a gangliosides, has been reported to add only small degrees of sensitivity.1 Many patients with MMN and immune motor neuropathies have had no identified associated serum autoantibody.

Our study documents that serum IgM binding to NS6S disaccharide is common in chronic acquired motor neuropathies and MMN. Frequencies of high titre serum IgM binding to NS6S and to GM1 ganglioside are similar in motor neuropathy syndromes and MMN but very low in ALS and CIDP groups. IgM binding to either NS6S or GM1 occurs in high titre in 64% of our motor neuropathy patients, a 21% absolute and 49% relative increase in frequency compared with testing for IgM anti-GM1 antibodies alone. The presence of a serum IgM M-protein increases the likelihood of finding IgM binding to NS6S or GM1 in motor neuropathy patients but these antibodies were also common (51%) in motor neuropathy patients without serum IgM M-proteins.

High levels of IgM binding to GM1 were not found in any control group patients and were false positive results in only 0.1% of the serum series. The specificity of finding high titre IgM anti-GM1 antibodies for acquired motor neuropathies in all serums with the testing method used in our clinical laboratory is 93% (25 of 27 positives in our serum series) and has been reported by others at 90% to 96%.1 The specificity figure would be higher in the context of only testing patients who have motor or demyelinating neuropathy syndromes. Covalent linkage of antigens to ELISA wells, and subtraction of background levels of GD1a ganglioside binding, rather than using blank well controls, greatly reduces the frequency of false positive results in our laboratory.9 IgM binding to NS6S is less specific for motor neuropathies than IgM binding to GM1. While high titre IgM binding to NS6S (≥7000) was not found in the CIDP and ALS groups, some patients with sensory neuropathies, especially with concurrent binding to TS-HDS or MAG, had high levels. Overall, a distinction between immune motor neuropathies and their main differential diagnoses of ALS and CIDP seems to be the strongest clinical inference in IgM anti-NS6S positive motor neuropathy patients.

Most clinical features in our motor neuropathy patients are similar to those reported in series of patients with MMN.3 14 Weakness was usually asymmetric and predominantly located in the distal upper extremities. Some of our patients had unexpected findings such as sensory changes, mild distal sensory loss or abnormal sensory nerve action potentials in the legs,15 or very proximal denervation with thoracic paraspinous denervation. None of these ‘atypical’ features altered the frequency of occurrence of IgM binding to GM1 or NS6S (table 2), suggesting that such findings should not be exclusion criteria for identifying motor neuropathy syndromes. The similarity of antibody frequency in chronic motor neuropathies with and without conduction block, and with only features of axonal loss, and with mild sensory abnormalities (table 2), suggests that IgM binding to NS6S or GM1 is associated with motor, rather than demyelinating, features of these syndromes. The presence of conduction block in motor neuropathy patients did not change the likelihood of finding, or titres of, IgM binding to either NS6S or GM1. A lack of association with motor conduction block, and other features of demyelination, is also seen for IgG anti-GM1 antibodies in patients with acute motor axonal neuropathies.2 The mechanisms underlying predominant involvement of motor axons in anti-GM1 and anti-NS6S antibody related motor disorders are uncertain. Several other IgM antibodies that bind to sulphated epitopes on sulphatide,16 MAG17 and TS-HDS10 are associated with predominantly sensory neuropathies. The IgM antibodies to the sulphated disaccharide NS6S described in the current study are often related to a predominantly motor disorder. There is no information regarding the distribution of the NS6S disaccharide epitope on peripheral nerves. There is no epitope that is clearly shared between NS6S and GM1. One possible mechanism underlying IgM antibody specificity for different types of axons involves variable binding according to local glycolipid environments.18–20

Our results raise issues regarding current ideas of sensitivity and specificity of clinical diagnosis and laboratory testing in motor neuropathies. The term MMN was originally used to describe two patients who had both motor conduction block and serum IgM binding to GM1 ganglioside.3 Recent formal criteria4 have recharacterised ‘MMN’ as a disease, and defined it restrictively21 22 on the basis of clinical and electrodiagnostic disease features. However, in this study, patients with full criteria for MMN comprised only 37% (23 of 62) of motor neuropathy patients with electrodiagnostic features of demyelination or positive antibody testing. Immune motor neuropathies with and without conduction block or other demyelinating features can be similar clinical entities.21 22 Our data suggest that the spectrum of immune motor neuropathies with an MMN clinical phenotype,3 14 who may respond to treatment,23 includes patients with no features of demyelination on electrodiagnostic studies of distal nerves. Antibody evaluation is especially important for identifying these, likely immune, axonal motor neuropathies with a clinical phenotype similar to MMN as their electrodiagnostic studies can have features, such as thoracic paraspinous denervation, that may suggest untreatable motor neuron disease.24 Antibody measurement may also help to clarify the status of patients with a motor neuropathy phenotype who also have sensory abnormalities on clinical or nerve conduction testing, an exclusion criterion for MMN.4 The frequency of serum IgM binding to NS6S and GM1 in our patients with abnormal sensory nerve action potentials was similar to that seen in the whole group of patients diagnosed clinically with motor neuropathies.

Limitations of this study include its retrospective design. Patient data and estimates of sensitivity of antibody testing may reflect referral bias to a large tertiary academic centre with an interest in motor neuropathies. The inclusion of patients with acquired motor neuropathies in our database may have been influenced by physician knowledge of the presence of laboratory features or treatment responses, leading to an overestimate of the frequencies of conduction block or positive antibody testing in the group of motor neuropathies. We may have underestimated the frequency of conduction block in our patients. Our electrodiagnostic studies did not include special evaluations to identify demyelination, or axonal dysfunction, at more proximal sites along nerves. Proximal stimulation techniques can detect static or activity dependent conduction block in roots or the brachial plexus in motor neuropathies.25 26 In practice, interpretation of proximal conduction studies can be uncertain and needs further clinical validation.27 Further study of acquired motor neuropathies is necessary to define whether serum IgM anti-GM1 or anti-NS6S antibodies are associated with any different, or specific, clinical or laboratory features, or treatment responses, or have any pathogenic role in producing the disorders. However, it is unlikely that these limitations would alter the main conclusions of our study. High titre serum IgM binding to GM1 and NS6S have specificity for motor neuropathy syndromes. Serum IgM binding to NS6S disaccharide is associated with motor neuropathy syndromes, occurs with similar frequency to IgM binding to GM1 and adds to the sensitivity of antibody testing in acquired motor neuropathy syndromes.



  • Linked articles 205310.

  • Funding This study was supported by the Washington University Neuromuscular Research Fund.

  • Competing interests AP serves on the scientific advisory board of the Myositis Association. He receives revenue related to patents and speaker honoraria from Athena. He owns stock in Johnson & Johnson. He is a director of the Washington University Neuromuscular Clinical Laboratory which performs antibody testing and muscle and nerve pathology analysis. The Washington University Neurology Department bills for these procedures. AP receives research support from NIH, the Muscular Dystrophy Association, Genzyme, Insmed, Knopp, Prosensa, ISIS, Cytokinetics and Sanofi.

  • Ethics approval This study was conducted with the approval of the Washington University HPRO.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles