Fig 1.
Nuclear co-localization of p65 and RelB by a quadruple multiple staining approach.
A. Epithelial staining with CK18, CK19 and PSA defines the epithelial mask using orange fluorochromes (A546, Cy3). B. Identification of epithelial area (presented in grey for optimal contrast in subsequent analyses) as ROI #1 (region of interest #1). Identification of nuclei (grey surrounded by red tracing) as ROI#2 from pre-defined ROI#1. P65 and RelB fluorescence were subsequently evaluated separately in ROI#1 and #2. C. RelB staining with green fluorescent dye (A488) and p65 staining with red fluorescent dye (Cy5). Steps 1, 2 and 3 correspond to the fluorescence analysis process followed using Visiomorph DP software.
Fig 2.
Example of immunofluorescence of p65 and RelB.
A. Magnification (40X) highlighting glandular structures in prostate cancer tissues. Merge: superimposed images of RelB (A488) in green, p65 (Cy5) in red and nuclei (DAPI) in blue, in normal adjacent (top) or cancer tissues (bottom). Arrows show stained nuclei. B. Comparative quantification from IHC (left) and IF (right) staining of p65. Graphs show the frequency distribution of nuclear p65 as evaluated visually (IHC) or automatically (IF). C. Frequency of nuclear p65 (left), RelB (middle) and double stained nuclei (right) in normal adjacent and tumor tissues. The comparison between adjacent normal and tumor cores was conducted using a Wilcoxon‘s test.
Table 1.
Spearman correlation test between NF-κB and clinic-pathological parameters.
Fig 3.
Association between nuclear p65 and/or RelB and biochemical recurrence in prostate cancerpatients.
Kaplan–Meier biochemical recurrence-free survival curves A. High (>20%) and low (<20%) frequency of nuclear p65 in the epithelia of cancer tissues. B. High (>5%) and low (<5%) frequency of nuclear RelB in the epithelia of cancer tissues. Significance (p) is indicated by log rank. C. Double nuclear epithelial staining of p65 and RelB. D. Epithelial and nuclear staining of p65 and RelB. The κB negative variable represents patients without positive tissue cores for p65 and RelB. The variable p65 alone or RelB alone represents patients positive for only nuclear p65 or nuclear RelB. The variable p65 + RelB are patients with presence of both nuclear subunits in the same core. E. Analysis of the epithelial proportion of nuclear p65 to RelB in tissue cores within both p65 and RelB staining. The ratio p65/RelB represents the frequency of p65 to the frequency of RelB. The variable p65 > RelB represents a ratio of at least 2 fold more p65 than RelB; p65-RelB represents a ratio between 0.50 and 2 fold, and RelB>p65 represents a ratio of at least 2 fold more RelB than p65.
Table 2.
Univariate and multivariate Cox regression analysis.