(A) Representative example of flow cytometry staining and (B) summary showing frequency Ki67+ CD56dim NK cells before (light blue) and after (dark blue) treatment in tumor lymph node and peripheral blood. of killer cell immunoglobulin-like receptors (KIR), CD57 and CD16. Rituximab therapy induced a rapid drop in NK cell numbers coinciding with a relative increase in the frequency of Ki67+ NK cells both in the lymph node and peripheral blood. The Ki67+ NK cells had slightly increased expression of CD16, CD57 and higher levels of granzyme A and perforin. The activation of NK cells was LX-4211 paralleled by a temporary loss of functionality, primarily manifested as decreased IFN production in response to rituximab-coated targets. However, patients with pre-existing NKG2C+ adaptive NK cell subsets showed less Ki67 upregulation and were refractory to the loss of functionality. These data reveal variable imprints of rituximab monotherapy on the NK cell repertoire, which may depend on pre-existing repertoire diversity. studies have shown that rituximab activates a broad range of NK cell subsets, independently of their expression of self-HLA binding inhibitory KIRs, thus overriding the need for education (12). On the other hand, it has been reported that tumor cells can increase HLA class I expression in response to IFN stimulation and thereby escape NK cell killing. However, the dynamics of the NK cell repertoire, both systemically and in LX-4211 the lymph node, during monotherapy with rituximab is largely unexplored. In this study, we examined the immune repertoire in sequential biopsies of the affected lymph node and in peripheral blood in FL patients receiving monotherapy with rituximab. Our results point to a diversified immunological response where some patients display a pronounced up-regulation of Ki67 associated with a temporary Mouse monoclonal to IHOG drop in NK cell function. The kinetics of the response was linked to the presence of adaptive NK cell subsets in the patient and may hold clues to clinical responsiveness to antibody therapy. Results NK Cell Frequency and Phenotype in Lymph Node and Peripheral Blood Eight patients diagnosed with follicular lymphoma were included in the study (Table 1). All patients were previously untreated and received in total four doses of rituximab (Figure 1A). We first established multi-color flow cytometry panels to monitor the immune subset composition in fine needle biopsies from tumor LX-4211 lymph nodes (LN) and LX-4211 peripheral blood (PB) before each treatment cycle at a weekly interval. The biopsy sample collection continued until the tumor lymph nodes were too small to access. The NK cell frequency in LN samples were consistently low compared to frequencies seen in PB (Figures 1B,C), with patients showing both increasing and decreasing trends over time. However, the relative LN-NK frequency of total CD45+ and CD19? CD20? cells were similar to what we found in tonsil samples from healthy donors. In agreement with earlier studies (13), we found a decrease of NK cells in peripheral blood 7 days after rituximab treatment started manifested as lower frequencies and lower absolute counts (Figures 1D,E). Table 1 Patient characteristics. = 8, healthy controls = 10. Differences were assessed using the Wilcoxon signed rank test for comparisons of matched samples within patients or Mann-Whitney 0.05. Next, we determined the expression of activating and inhibitory receptors, including killer cell immunoglobulin-like receptors (KIR), NKG2A and NKG2C, effector molecules and maturation markers on intra-nodal and peripheral blood NK cells (Figure 2). In line with previous findings (14, 15), we observed a dominance of CD56brigh NK cells in tonsils from healthy donors (average 56%, range 37C71%). Tonsils are widely used as a control in FL (16, 17), albeit they represent a more inflamed tissue compared to normal lymph nodes from healthy individuals. Tonsils contain more differentiated T cells and are more similar to FL tumors in terms of immune cell composition and differentiation states (18). Indeed, compared to normal tonsils, LN-NK cells in FL patients showed an intermediate phenotype, with an average of 71% CD56dim cells. This intermediate state was also reflected in the relative expression of CD57, KIRs and CD16 on CD56dim NK cells when compared to the same subset in tonsil-derived NK cells and PB-NK cells (Figure 2B). Although we cannot formally exclude that ILCs contributed to the relative composition in CD56+CD3- cells in LN and tonsils, ILCS typically lack CD16, KIR LX-4211 and NKG2A. Furthermore, LN-NK cells expressed lower levels of the effector molecules Granzyme A/B.