Contribution of Caspr2 autoantibodies to the induction of pain phenotypes in patients with neuropathic pain

Chronic pain affects about 20% of people in the western civilization and thus, represents a major threat for the quality of life. Neuropathic pain, a form of chronic pain, is a frequent symptom in patients with autoantibodies against Caspr2, a cell adhesion molecule of the voltage-gated potassium channel complex (VGKCC) in the central (CNS) and peripheral nervous system (PNS). In general, the number of identified autoantibodies related to neurological diseases increased strongly in the last three decades. Caspr2 interacts intracellularly with scaffold proteins post-synaptic density protein 95 (PSD95) and erythrocyte membrane protein band 4.1-like 3 (4.1b), which connect Caspr2 to type 1 voltage-gated potassium channel subunits Kv1.1 and Kv1.2. The role of Caspr2 lies in the adequate incorporation and maintenance of the VGKCC in the membrane of dorsal root ganglion (DRG) neurons, particularly at the juxtaparanode (JXP) of myelinated sensory fibers. Kv channels within this complex mediate the efflux of positively charged potassium ions, which lowers the membrane potential. Due to dysfunctions of the VGKCC, the membrane potential can remain elevated facilitating action potentials. As a result, increased neuronal activity in pain-transmitting DRG neurons of the PNS can lead to amplified pain sensation in patients. At the molecular level, we aim to investigate structural and functional consequences of autoantibodies from different IgG subclasses in correlation with the patients pain phenotype by examining a cohort of 54 seropositive patients. First, the spatial proximity of PSD95, 4.1b, Kv1.1, and Kv1.2 to Caspr2 was demonstrated by immunocytostainings of the axonal and somatic membrane of embryonic and adult DRG neurons from mice. Additionally, these proteins were subsequently detected in neuronal membrane fractions of the DRG and the brain by Western blots. To examine VGKCC proteins in an overexpression system such as HEK-293 cells, a plasmid for the transfection of mammalian cells with protein 4.1b was missing. Therefore, the gene for 4.1b was transferred from a commercially available bacterial clone into a mammalian expression vector by subcloning. Frequency and distribution of IgG subclasses (IgG1, IgG2, IgG3, IgG4) in patient sera was assessed by HEK-293 cell-based assays. For a comparative assessment of effects on the structural arrangement of VGKCC proteins, the cohort was divided into four groups according to the presence or absence of pain and the detection of only IgG4 or IgG4 and additional subclasses (IgG4+X). This subclassification allowed distance measurements between Caspr2 and Kv1 subsequent to serum pool presentation. Microfluidic chambers allowed a discrimination between effects of sera applied to the somatic or the axonal compartment. At last, the pathogenicity of anti-Caspr2 autoantibodies was tested through a passive transfer model into rats. To evaluate the motor performance of rats, Catwalk and Rotarod testing were conducted, while Hargreaves and Von Frey testing served to assess sensory responsiveness indicating pain-like symptoms. 93% of anti-Caspr2 patients were seropositive for IgG4 and about half of the cohort experienced pain. However, the presence of IgG4 compared to IgG4+X did not correlate with the occurrence of pain symptoms. Affected patients were divided into two groups based on quality and localization of pain: 64% experienced mostly burning or tingling pain in the extremities, similar to neuropathic pain in small fiber neuropathy (SFN-like), while 36% of patients described widespread pain with myalgia, muscle cramps, sensations similar to aching muscles, and even arthralgia. Follow-up sera and clinical data from two patients suggested that improvement of symptoms by treatment with intravenous methylprednisolone (IVMP) or intravenous immunoglobulins (IVIG) was insufficient, and the subclass distribution remained unchanged. Looking at the mechanism of IgG at the molecular level, an increased distance between Kv1.2 and Caspr2 at axons of embryonic DRG neurons after axonal treatment with sera from patients with pain suggests a serum effect exclusive for patients with pain. No change in expression levels of Kv1.2 and a reduced expression of Caspr2 in axons after application of sera containing IgG4+X argue for partial internalization of Caspr2 by IgG1-3. The passive transfer experiments resulted in rats hypersensitive to thermal stimuli upon injection with patient IgG. Motor impairments and mechanical allodynia were not observed in rats during 14 days of the experiment. The injected human IgG reached the blood circulation system of rats, but no IgG binding to neurons in the DRG and the JXP region of peripheral nerves was observed. In addition, demyelination of the sciatic nerve was not detected. These data support IgG4 as a major player in the development of neuropathic pain, but IgG1, 2, and 3 may modulate the outcome via partial internalization and mediators of pain could shape structure and function of the VGKCC. In addition, the involvement of thermoreceptive neurons in the induction of pain needs further investigation.