The Role of Surface Immunoglobulin Isotype in Chronic Lymphocytic Leukemia Disease Biology and Clinical Outcome

Upon encountering cognate antigen, naïve B cells may undergo germinal center reaction. This results in DNA modification including somatic hypermutations (HSM) of the immunoglobulin heavy chain variable region gene (IGHV) and immunoglobulin class switch recombination (CSR). HSM and CSR are unlinked, but share initiating and mechanistic factors. Prior evidence indicates CLL B cells are antigen experienced; however, half of all CLL patients have an unmutated IGHV and follow more aggressive clinical courses. The biologic importance of the CLL immunoglobulin isotype and CSR in B cell receptor (BCR)-mediated signaling is not well understood. We hypothesized the immunoglobulin isotype and expression density of surface immunoglobulin would correlate with IGHV mutation status and other clinical parameters, and predict BCR responsiveness in vitro.

195 samples from our CLL specimen repository with detailed clinical and biologic characterization were evaluated. Surface expression of IgD, IgM and IgG was determined using multi-parameter flow cytometry and samples were classified as either IgG+ or IgM+IgD+ co-expressing. An ROC analysis was performed to identify the most discriminate value for dichotomization of IgM surface expression as a function of IGHV mutation status. IgM^hiIgD+, and IgM^lowIgD+ patients were then analyzed for associations with clinical and biologic parameters. To investigate the biological mechanisms of potential associations, BCR-mediated signaling after isotype specific crosslinking in vitro was measured by quantification of downstream phospho-proteins (SYK, AKT, ERK, MEK1, NFκB) by flow cytometry (phospho-flow). Samples were crosslinked with IgM (or IgG if IgG+), IgD, and isotype control and incubated for 0, 15, 30, 60, and 120 minutes. The extent of phosphorylation at each timepoint was expressed as mean fluorointensity (MFI) of the examined proteins. Crosslinked samples were also analyzed using oligonucleotide microarray gene analysis (GEP) to assess differential transcription.

A correlation was identified between the percentage of CLL cells expressing surface IgM and IGHV mutation status (r²=0.33); higher surface expression of IgM correlated with unmutated IGHV. The most discriminate value of IgM expression for prediction of IGHV mutation was 47%, with an area under ROC curve of 0.72. Of the 195 patient samples analyzed, 91 (47%) were IgM^hiIgD+; 85 (43%) were IgM^lowIgD+; and 19 (10%) were IgG+. IgM^hiIgD+ samples showed increased expression of CD38 (p<0.01) and ZAP70 (p<0.05) compared to IgM^lowIgD+ samples. Only 1 of 19 IgG+ samples was IGHV unmutated. Neither immunoglobulin isotype nor density correlated with gender, lymphocyte doubling time, age, or Rai stage at diagnosis. There was a trend toward shorter time to first treatment by IgM expression (IgM^hiIgD+ 7.9yrs vs. IgM^lowIgD+ 9.4yrs) but this was not significant (p=0.09). To date, we have tested the biologic relevance of surface immunoglobulin (sIg) density and isotype in 13 patients using phospho-flow analysis. Crosslinking in the IgM^low patients (sIgM <30%; n=5) demonstrated no change in MFI of tested proteins. Among IgM^mod patients (sIg 31–60%; n=2), a maximal 3 fold increase in p-SYK MFI with IgM crosslinking and 3.8 fold increase with IgD crosslinking was observed. In IgM^hi samples (sIgM >60%; n=3), there was a maximal 3.5 fold increase in MFI with IgM but no change with IgD crosslinking. In the IgG+ samples (n=3), a maximum 5 fold increase in MFI was observed. Analysis of GEP data is ongoing.

Surface expression of IgM in CLL varies considerably between patients. This difference in density of surface IgM expression correlates with markers of clinical risk. Although we do not propose immunoglobulin isotype or expression density as a novel prognostic factor, the findings that high surface expression of IgM in CLL predicts BCR mediated signaling after crosslinking provides insight into CLL immunobiology and disease responsiveness to antigens in the microenvironment. In addition, the observation of significant BCR activation within the IgG+ population (which was predominantly IGHV mutated) suggests that the responsiveness of the BCR to crosslinking may be related to the immunoglobulin isotype and density, and not solely determined by the IGHV mutation status.

No relevant conflicts of interest to declare.