Identification of Circulating Autoantibodies in Patients with Lymphoma Using Antigen Arrays.

Over thirty types of lymphoma are now recognized and correct diagnosis is therefore essential. We have developed a novel approach, using arrays containing thousands of recombinant proteins (Bussow K, Cahill D, Nietfeld W et al. Nucleic Acids Res 1998 26:5007–8; Gutjahr C, Murphy D, Lueking A et al. Genomics 2005 85:285–96), to identify new potential disease markers in a number of different lymphomas. Specifically, by screening patient serum samples against human protein/antigen arrays containing over 10,000 different human proteins we are able to identify disease-associated auto-IgG antibody interactions (Horn S, Lueking A, Murphy D et al. Proteomics 2006 6:605–13). Our approach exploits the autoimmune response which has been observed in many cancer types. For example, patients with anaplastic large cell lymphoma (ALCL) produce circulating autoantibodies against the aberrantly expressed ALK tyrosine kinase (Pulford K, Falini B, Banham A et al. Blood 2000 96:1605–7; Ait-Tahar K, Hatton CSR, Banham A et al. Int J Cancer 2006 118:688–95). Also, patients suffering from follicular lymphoma (FL) have been shown to produce antibodies reactive with BCL-2 (Pulford K, Roberton H, Banham A et al. Br J Haematol 2002 116:135–41). In this study we are screening serum samples from patients with ALK-positive ALCL, ALK-negative ALCL, germinal center and non-germinal center subtypes of diffuse large B cell lymphoma (DLBCL), transformed DLBCL, peripheral T-cell lymphoma, mantle cell lymphoma and FL. Initially, a pool of 5 patient samples from each lymphoma type and a sex- and age-matched control group are screened against the antigen arrays. These experiments yield lists of putative disease-associated autoantigens. Screening larger numbers of individual serum samples from lymphoma patients and control cohorts against these autoantigens will enable an estimate of the frequency and disease specificity associated with circulating antibodies against each of the autoantigens identified. An initial study of 5 patients with ALK-positive ALCL, 5 patients with ALK-positive ALCL and 5 age- and sex-matched controls revealed the identification of 140 putative ALK-positive ALCL-associated autoantigens, 125 putative ALK-negative ALCL-associated autoantigens, and 27 autoantigens common to both ALK-positive and ALK-negative ALCL. We are currently ranking these results through individual screens of ALCL patient samples. We will also present the results obtained from the other 6 lymphoma types being screened, identifying autoantigens which show a disease-specific pattern, as well as those that are common to one or more of these lymphoma types. The approach described here should contribute not only towards the discovery of novel lymphoma markers, the discovery of potentially novel proteins involved in the development of the disease but also to the identification of therapeutic targets.