WP 13. Technologies - proteomics and antibody profiling

This area of work will be lead by Partner 7 (London) with input from Partners 1 (KI), 2 (Leiden), Partner 8 and 14 (Manchester), who will provide valuable patient material from their Biobanks, as well as from Partners 3 (Berlin) and 4 (Vienna) who also have extensive experience in serum autoantibody profiling and Partner 25 (BMD)

Detailed plan of investigation

Acquisition of serum samples from patients (first 6 months)

Initially, sera from clinically homogeneous patients with established seropositive (anti-CCP+/RF+, or anti-CCP-/RF+) erosive RA and from age and sex matched healthy controls will be studied. These samples will be used for screening of potential cell substrates. A second set of control samples from patients with OA and from healthy older individuals (> 60yrs) will be used to distinguish between autoantibody profiles reflecting the rheumatoid versus the osteoarthritic process as well as the effects of ageing. These samples will be accessed locally as well as through consortia biobanks provided by Partners 1, 2 and 15. For immunoblotting, samples will be pooled to provide sufficient antibody serum to use as a probe for the identification of antigenic; panels of individual well characterised sera will also be used. For genotype-serotype correlations serum samples will be available from consortia biobanks from RA patients who have also been genotyped at various loci, including the MHC (especially class II and III regions), as defined in WP1, 2, 3 and 11.

Preparation and selection of cell substrates as reference standard (Year 1)

Protein substrates, which will be screened for the presence of putative autoantigens using patient sera, will be generated from whole cell lysates prepared from human cell lines available in the laboratory. Cell lines will include EBV transformed B cells, Jurkat T cells, THP-1 macrophages, human umbilical vein endothelial cells, HeLa epithelial cells, synovial joint fibroblasts, a chrondrosarcoma cell line, and an osteosarcoma cell line. Cells transfected with PAD2 or PAD4 will also be studied to examine for the presence of antibodies to citrullinated cellular proteins. Alternatively, the effects of endogenous PAD can be investigated using HL60 monocytic cells differentiated in vitro with vitamin D3. Deimination of cell lysates in vitro with recombinant PADs has also been used successfully at the Kennedy Institute to identify novel citrullinated autoantigens. Finally, purified nuclear extracts will be tested as they provide a subcellular fraction potentially rich in autoantigens.

Gel electrophoresis and immunoblotting (Years 1 and 2)

Relevant substrates will be resolved by 1D gradient SDS polyacrylamide gel electrophoresis in the first instance. After electroblotting, nitrocellulose membranes will be probed with a 1:100-1:200 dilution of pooled serum from control subjects or RA patients. The bands detected with anti-human IgG-HRP and recognised by control or RA sera will be quantitatively compared by computerised image analysis. Each substrate will be scored according to the number of detectable antigenic specificities recognised by blotting with each set of sera, so as to define cells which express the most autoantigens. Once a suitable and reproducible reference standard substrate has been defined, lysates will then be resolved by isoelectric focussing (IEF) for the first dimension and by apparent mass in the second. While the resolving power of 2D gel electrophoresis (2DE) is extremely high, many proteins are resistant to 2D analysis on physicochemical grounds. To facilitate resolution of such proteins we will also undertake IEF in liquid phase, which improves recovery and detection of hydrophobic, acidic and basic proteins (include nuclear antigens), when compared to conventional 2DE facilitate resolution of relatively insoluble cellular proteins (Peirce et al , Mol Cell Proteomics 2004; 3:56-65) . Resolution of subcellular fractions will also increase the representation of low abundance proteins. After transfer to nitrocellulose, membranes will be probed with pooled sera, and the pattern of spots detected quantified using Phoretix imaging software used routinely at the Kennedy Institute.

Subsequent studies

Identification of autoantigens by mass spectrometry

Replicate gels to those used for immunoblotting will be silver stained, and the stained spot patterns correlated to immunoblots. Identification of silver stained gel features matching the positions of immunoblotted proteins will be facilitated by the high resolution attainable by use of large format (24 cm) gels and narrow range (single pH unit) IPG strips. Silver stained spots will be excised, digested in situ with trypsin, and identified by high-sensitivity nanospray tandem mass spectrometry, using procedures now routine at the Institute (Wait et al , Electrophoresis 2001 ; 22 :3043-3052; Wait et al , Electrophoresis 2002; 23 :3418 -27; Hermansson et al , J Biol Chem 2004, in press ). We will also employ an MS strategy to determine the presence and location of deiminated arginine residues. Once reproducible patterns of autoantibody reactivity have been achieved individual serum samples from genotyped patients can be tested, to address specifically whether the MHC genotype influences the profile of autoantibodies in inflammatory arthritis. In the longer term, a major objective will be to define sufficient numbers of autoantigens that will allow us to discriminate between subsets of RA patients, patients with undifferentiated arthritis and other forms of inflammatory arthritis, such as psoriatic or reactive arthritis. This information will then be used to design platforms for high throughput multi-analyte assays, including the development of protein arrays or the application of Luminex technology. This phase of the programme will be undertaken in association with consortia SME partners, in particular BMD, who has extensive expertise in the field of diagnostics.