A peptide-based immunoassay for antibodies against botulinum neurotoxin A

Cervical dystonia (CD) is due to neck-muscle spasms that cause pain and involuntary contractions resulting in abnormal neck movements and posture. Symptoms can be relieved by injecting the affected muscle with a botulinum neurotoxin (BoNT, usually type A or type B). The therapeutic benefits are impermanent and toxin injections need to be repeated every 3-6 months. In a very small percentage of patients (less with BoNT/A than with BoNT/B) the treatment elicits blocking anti-toxin antibodies (Abs), which reduce or terminate the patient's responsiveness to further treatment. We have recently mapped (Dolimbek et al., 2006) the CD sera Ab-binding profile using a panel of 60, 19-residue peptides that encompassed the entire H chain sequence 449-1296 and overlapped consecutively by 5 residues. Abs in CD sera bound to one or more of the peptides N25, C10, C15, C20, and C31. This suggested the possibility that binding to these peptides could be used for assay of Abs in CD sera. Data analysis reported here found that Ab binding to these regions showed very significant deviations from the control responses. Of these four peptides, C10 showed the most significant level of separation between patient and control groups (p = 5 X 10(-7)) and the theoretical resolution (i.e., ability to distinguish CD patients from control, see full definition under 'Statistical analysis' in Methods), 84%, was about 4% higher than the least resolved response, C31 (p = 6 X 10(-6), resolution 80%). Since the amounts of Abs bound to a given peptide varied with the patient and not all the patients necessarily recognized all four peptides, there was the possibility that binding to combinations of two or more peptides might give a better discriminatory capability. Using two peptides, C10 plus C31, the resolution improved to 87% (p = 4 X 10(-8)). These two peptides appeared to compliment each other and negate the lower resolution of C31. Combination of three peptides gave resolutions that ranged from 85 (N25 + C15 + C31; p = 2 x 10(-7)) to 88% (C10 + C15 + C31; P = 1 X 10(-8)). Finally, using the data of all four peptides, N25 + C10 + C15 + C31, gave a resolution of 86% (p = 1 X 10(-7)). Although these levels of resolution are somewhat lower than that obtained with whole BoNT/A (resolution 97%; p = 6 X 10(-12)), it may be concluded that the two-peptide combination C10 + C31, or the three-peptide combination C10 + C15 + C31 (affording resolutions of 87 and 88%, respectively) provide a good diagnostic, toxin-free procedure for assay of total specific anti-toxin Abs in BoNT/A-treated CD patients. Copyright (c) 2006 John Wiley & Sons, Ltd.