FIGURE 4. Metastatic progression correlates with nuclear activation of IKK, inflammatory cell infiltration and massive upregulation of RANKL.

a, CaP cells from TRAMP mice of indicated ages (1 and 2 are individual mice), fractionated into cytoplasmic (C) and nuclear (N) extracts, were analysed for IKK, phospho-IKK, IKK, IKK, histone H3 and Maspin by immunoblotting. b, Normal human prostate, benign prostatic hyperplasia and prostate tumours of different clinical stages (TNM stage) were divided into nuclear (N) and cytoplasmic (C) fractions and analysed for the indicated proteins by immunoblotting. c, Prostate tumours from 4–5- and 7–9-month-old TRAMP mice of the indicated genotype were stained for the T cell marker CD3. FITC, fluorescein isothiocyanate. d, Prostate tumours from mice of the indicated age and genotype were analysed by qRT–PCR for expression of RANKL and Lta mRNAs, (results are averages  s.d.; n  =  4). e, WT/TRAMP and Ikk^AA/AA/TRAMP prostate epithelial cells were treated with lipopolysaccharide-free RANKL (200 g ml^-1). At the indicated times cell extracts were prepared and examined for Maspin and p38 content by immunoblotting. f, A model explaining how RANK signalling leads to repression of Maspin transcription. After transient IKK-mediated repression, Maspin transcription is likely to be permanently silenced through DNA methylation.