B-Cell Receptor Recognition of Multiple Epitopes Correlates with An Aggressive Clinical Course of Chronic Lymphocytic Leukemia.

The malignant B-cells in chronic lymphocytic leukemia (CLL) express membrane immunoglobulins (B-cell receptors; BCR) which are specific for the leukemic clone of each individual patient. Emerging evidence suggests that the development and course of CLL may be driven by antigenic stimulation through the BCR. Here we set up a model system of epitope recognition in CLL to explore how diverse epitope recognition in CLL is and whether the epitope recognition pattern has clinical relevance. METHODS: BCRs from six randomly chosen CLL patients were cloned and recombinantly expressed as IgG1 Fab fragments. Combinatorial phage-displayed peptide libraries with five different insert designs were constructed and used for the selection of epitope-mimicking peptides on the Fab fragments. We tested the binding of phage displayed epitope mimics to the respective Fab fragment by ELISA as well as to the native BCR on the cells of CLL patients. Therefore, cell-bound phage displayed epitope mimics were separated from unbound phage by differential centrifugation and bound phage were quantified by bacterial infection. The binding of six ‘index‘ epitope mimics representative for each BCR was evaluated in a set of 100 unrelated CLL cell samples. Epitope recognition patterns of CLL BCRs were correlated with the clinical course of the disease by standard biostatistical analysis including Kaplan-Meier estimator, log-rank test, cox regression analysis and Chi-square test. RESULTS: We selected epitope-mimicking peptides from phage display libraries on six CLL BCRs from randomly chosen patients. The selected peptides bound to the recombinant BCRs as well as to the native BCRs on the respective CLL cells. To model epitope recognition in a larger cohort of CLL patients we chose six representative index epitope mimics and evaluated their binding in a large set of 100 unrelated CLL cases. Surprisingly, all CLL samples recognized one or several index epitopes. Some of the CLL samples showed marked polyreactivity whereas other samples were mono- or oligoreactive. We determined whether the degree of BCR polyreactivity correlates with the clinical course of the disease using time to first treatment (TTFT) as surrogate marker of disease progression. We found that CLL patients expressing BCRs reactive with each of the epitope mimics had a significantly worse clinical course than less reactive control patients (median TTFT 27 months versus 87 months). Moreover, CLL patients whose cells express BCRs reactive with five or more epitope mimics were also characterized by an aggressive clinical course as compared to patients reacting with less than five epitopes (median TTFT 24 months versus 97 months). These outcomes were unrelated to known prognostic markers such as BCR mutational status and high risk receptor configurations. CONCLUSIONS: We introduce a system for modelling and monitoring of BCR epitope reactivity in CLL. Our findings indicate that a polyreactive epitope recognition pattern may be a determinant of an aggressive clinical course in this disease. These findings further emphasize the functional and prognostic relevance of BCR epitope recognition patterns in CLL.

No relevant conflicts of interest to declare.