NF-κB activation in bone marrow samples from MDS patients. (A) Evidence for NF-κB activation in high-risk MDS, as determined by EMSAperformed on total bone marrow samples. Bone marrow aspirates from the indicated low- or high-risk patients (same numbers as in Table 1) were analyzed by EMSA, as described in “Patients, materials, and methods.” Cord blood CD34⁺ cells before and after stimulation with G-CSF served as negative and positive control, respectively. The intensity of bands was quantified (right panel). (B) Evidence of NF-κB activation in purified CD34⁺ or CD33⁺ cells from high-risk MDS bone marrow samples. Cells from patients 24 and 54 (Table 1), representing low- and high-risk MDS, respectively, were subjected to confocal immunofluorescence determination of p65 activation and EMSAdetection of NF-κB alone or with the p65 antibody (supershift in right panel), as well as Oct-1 as a loading control. Images were acquired as described for Figure 1C. (C) Percentage of purified CD34⁺ or CD33⁺ cells with nuclear p65 staining among different MDS patient groups classified according to the percentage of bone marrow blast infiltration. Samples from 3 different control individuals, as well as different patients (nos. 1, 24, 34, 35, and 37 representing MDS patients with <5% blast; nos. 9, 11, 27, 28, 38, and 47 for patients with 5%-9% blasts; nos. 16, 17, 20, 54, and 57 for patients with >10% blasts; nos. 21 and 32 for >20% blasts), were subjected to p65 staining as in panel B, and the frequency of cells exhibiting nuclear p65 staining was determined. Values are expressed as means ± SD; n = 3. (D) Combined detection of nuclear p65 and cytogenetic alterations. Bone marrow aspirates from untreated MDS patients were subjected to the simultaneous immunohistochemical detection of p65 and FISH, using probes specific for the centromeres of chromosomes 7 and 8. Note that cells with manifest cytogenetic alterations (trisomy 8 in left panel or monosomy 7 in right panel) can exhibit the presence of p65 (gray) in the nucleus whereas euploid cells have p65 in the cytoplasm (not shown). Fluorescence micrographs were acquired as described for Figure 2I. (E) Correlation between blast counts and the frequency of cells with nuclear p65 (same patients as in Table 2 and patients 36, 39, and 42-46). The coefficient of correlation was calculated using linear regression. (F) Longitudinal study of blast counts and nuclear p65 in MDS patients. Bone marrow aspirates from patients 2, 7, 8, and 25 obtained at the beginning of clinical monitoring (values in Table 1) or several months later (as indicated in the figure) were subjected to immunofluorescence analyses, showing an increase in p65 translocation to the nucleus. Blast counts and p65 translocation were correlated using the Spearman correlation model.