The pathogenicity and origin of auto-antibodies in chronic inflammatory demyelinating polyradiculoneuropathy and the identification of cutaneous biomarkers in Charcot-Marie-Tooth 1A patients

Peripheral neuropathies can severely affect patients. Causes for the disease are diverse but can be classified into two main groups, acquired and hereditary. Examples for these two types are chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) and Charcot-Marie-Tooth disease type 1A (CMT1A). CIDP has an estimated prevalence of about 1-9:100 000. In this pathogenetically hetereo- geneous patient group about 5-10% show auto-antibodies against the node of Ranvier and present with distinct symptoms. Treatment with rituximab - a monoclonal antibody that deletes CD20 + B cells - has been shown to be effective in a majority of auto-antibody as- sociated CIDP cases. This suggests that B cells and the produced auto-antibodies might be pathogenic. Previous studies delivered evidence that auto-antibodies alone can induce nerve damage. In this study, the aim was to investigate the pathomechanism of auto-antibodies in vivo and their exact origin: For the analysis of the pathogenicity of auto-antibodies, passive transfer experiments on Lewis rats were performed with whole IgG from a patient with anti-contactin-1 (CNTN1) IgG4 auto-antibodies. IgG was infused through an intrathe- cal catheter targeting the thoracic/lumbar region of the spine over a long-term, 3-week period. In a previous study of our group, the IgG from the same patient has been re- ported to have mild pathogenic effects when applied intraneurally into the sciatic nerve of Lewis rats. In this study however, binding of auto-antibodies to nerve roots could not be detected. Neither evaluation of electrophysiological properties after the injection period nor motor and sensory skills tested throughout the injection period showed differences when compared to animals infused with control IgG. This suggests that in the chronic intrathecal protocol anti-CNTN1 auto-antibodies did not have a pathogenic effect. In peripheral blood, four B cell subsets capable to produce antibodies were previously described: memory B cells, plasmablasts (PBs), B1 cells and CD20 + CD38 hi cells. For the identification of the B cell subsets that produce auto-antibodies, purification and sort protocols as well as an enzyme-linked immuno spot (ELISpot) assay for IgG and IgM were established successfully. Since unstimulated B cell subsets produced very small amounts of IgG and IgM, peripheral blood mononuclear cells (PBMCs) were stimulated with IL-2 and R848 for 72 h prior to sorting. While the memory B cell frequency decreased after stimulation, the frequency of CD20 + CD38 hi cells increased and the overall number of antibody-secreting cells was increased. When stimulating patient PBMCs for 10 days though, detection of anti-neurofascin-155 (NF155) auto-antibodies in supernatants by enzyme-linked immunosorbent assay (ELISA) was possible in two out of three patient samples. Even though cell sorting was feasible after 10 days of stimulation, detection of auto-antibodies could not be accomplished using antigen-specific ELISpot. Although the implementation of the cell sorting and purification protocol was successful, further adjustments of the antigen-specific ELISpot need to be performed. However, we could show that after 10 days of stimulation auto-antibody detection is possible by ELISA which helps to pre-screen if patient PBMC contain auto-reactive B cells. CMT1A has an estimated prevalence of 1:5000 and is caused by a duplication of the peripheral myelin protein 22 kDa (PMP22) gene. Patients suffer from distal weakness and muscle wasting leading even to wheelchair-dependency in some cases. Although different treatment options for CMT1A have been tested in previous clinical trials, none of them have been successful. In this study, the aim was to identify objective and reproducible outcome measures that assess the actual nerve damage in a large cohort of CMT1A patients by analyzing a series of parameters. Glabrous skin samples were collected from 48 CMT1A, 7 CIDP and 16 small fiber neuropathy patients and 45 healthy controls. 40-µm cryosections from the lateral part of the index finger were double-labeled using immunoflu- orescence to investigate cutaneous innervation. The disease severity which was assessed using the Charcot-Marie-Tooth Neuropathy Score version 2 (CMTNSv2) and ranged between mild to severe (3-27) correlated with age in CMT1A patients. Furthermore, the intraepidermal nerve fiber density (IENFD) was reduced in CMT1A patients in comparison to controls and correlated negatively with the disease severity. In controls however, the IENFD correlated inversely with age. Meissner corpuscle density tended to be reduced and correlated inversely with age in CMT1A patients. This was not observed in healthy controls though. Compared to controls, Merkel cell density was also reduced in CMT1A, while the fraction of denervated Merkel cell was increased and correlated with age. Further differences were revealed concerning the node of Ranvier. Paranodes were shortened and the fraction of long nodes was decreased in CMT1A patients compared to controls. These data suggest that the IENFD, the Meissner corpuscle and Merkel cell densities are possible candidates for outcome measures as they are associated with disease severity or age of patients. However, a reliable statement about the suitability as a marker for disease progression can not be made in this study since only six CMT1A patients agreed to give a follow-up biopsy two years later.