Paraneoplastic cerebellar degeneration: Characterizing the Yo antibody antigens

Abstract

Background: The pathogenesis of paraneoplastic cerebellar degeneration (PCD) with Yo-antibodies is unclear. The disease is generally accepted as immune-mediated, but whether the Yo antibodies themselves are pathogenic or if T cells are responsible for the neurodegeneration is not known. Yo antibodies are, nevertheless, good biomarkers for the disease.

The primary target of Yo antibodies was until recently thought to be CDR2. We showed, however, that these antibodies bind to CDR2L and not CDR2. CDR2L is present on both bound and free ribosomes in the cytoplasm of cerebellar Purkinje neurons as well as other cells types, but the cellular function and spatial conformation remains unknown.

Objective: Paper I: To determine the major antigen of Yo antibodies in PCD patients. Paper II: To define the subcellular location of both CDR2 and CDR2L and potential interaction partners. Paper III: To generate a protocol for Purkinje neuron culture from both embryonic and postnatal rats that can be used for further characterization of PCD pathogenesis.

Methods: Paper I: Patient samples (serum and CSF), cerebellar tissue (human and rat), cancer cell cultures (OvCar3 and HepG2), immunostaining, immunoprecipitation, fluorescent immunoblotting and recombinant DNA transfection. Paper II: Patient samples (serum and CSF), cerebellar tissue (human), cancer cell cultures (OvCar3 and HepG2), Purkinje neuron cultures (rat), mass spectrometry-based proteomics, immunostaining, proximity ligation assay, super-resolution microscopy and immunoprecipitation. Paper III: Culturing of dissociated rat cerebellar tissue, immunostaining, Purkinje neuron counting, dendritic branch analysis, lentiviral transfection and micro-electrode array recordings.

Results: Paper I: We demonstrated that CDR2L, and not CDR2, is the major antigen target of Yo antibodies. These antibodies do, however, bind recombinant CDR2. Paper II: We found that CDR2L is predicted to interact with several ribosomal proteins and that it indeed does interact with the ribosomal protein rpS6. Interaction partners of CDR2 included the nuclear speckle proteins eIF4A3, SON and SRSF2. Paper III: We found that a support layer, pH stability and co-factor supplements were essential to generate rat cerebellar cell cultures with high yield of mature Purkinje neurons.

Conclusions: Paper I: The finding that Yo antibodies bind endogenous CDR2L, and not CDR2, allows us to rethink the mechanisms involved in Yo-mediated PCD. The binding of recombinant CDR2 suggests that these proteins have common epitopes which is not surprising considering their 45% amino acid sequence identity. Furthermore, test assays using CDR2L instead of CDR2 could be more sensitive, reducing the large amounts of false-positive results obtained today.

Paper II: Previous studies suggested that Yo antibodies bind a ribosomal target, but the locations of CDR2 and CDR2L were unknown. Our finding that CDR2L interacts specifically with ribosomal proteins, while CDR2 interacts with nuclear speckle proteins, adds further support for CDR2L being the primary Yo antibody target. Since one of the interaction partners of CDR2, eIF4A3, translocates from the nucleus to the ribosome, where it interacts with rpS6, this also adds an indirect link between CDR2L and CDR2. Whether CDR2L and CDR2 have similar roles or are involved in related processes in protein transcription and translation remains to be resolved.

Paper III: We established a robust primary culture protocol that gave high yields of mature Purkinje neurons from both embryonic and postnatal rats. These cultures were well suited to high-throughput screening, genetic manipulation and electrophysiological recordings and will be useful for exploring both neurodegenerative and regenerative mechanisms.