Fig 2: Identification of the androgen-independent elements responsible for upregulation of L-plastin. (a) the × 400 photographs of LNCaP and LNCaP-AI cells. (b) MTT assays showed the proliferation of LNCaP and LNCaP-AI cells in steroid-depleted medium (cs-FBS). (c) The dose–response curve of LNCaP-AI and LNCaP cells proliferation over a range of DHT concentration. (d) The expressions of AR, PSA and p21 were examined by western blot analyses in LNCaP and LNCaP-AI cells. (e,f) Transcriptional activity of the L-plastin promoter was evaluated by the sequentially deletions of the androgen and oestrogen elements into the LNCaP cell line by examining the L-plastin promoter linked to Renilla luciferase activity. (g) Deletion analysis of pGL3-ΔARE1,2,3-ΔERE which is the construct with deletions of both AREs and EREs. Schematic presentation of pGL3-L-plastin-P 0.2 which encompassed nucleotides −216 to +118 in L-plastin promoter. The pGL3-basic and pGL-3-CMV were transfected as negative control and positive control. (h,i) EMSA analyses were applied to identify androgen-independent response elements in the −216 bp fragment at 3′end of L-plastin promoter. Ten pairs of distinctive, overlapping oligoes in this area for synthesized and labelled with 32P. The arrows indicated positive band shift sites

Fig 3: AP4 regulates PCa via activation of PI3K/AKT pathway. (a,b) LNCaP-AI and PC-3 cells were treated with PI3K inhibitor LY294002, AP4 and L-plastin mRNA and protein levels were determined by qRT-PCR and western blot analysis. GAPDH was used as loading control. (c,d) LNCaP-AI and PC-3 cells were treated with AKT inhibitor Perifosine, AP4 and L-plastin mRNA and protein levels were determined by qRT-PCR and western blot analysis. (e–f) The expressions of AP4, L-plastin, GSK3β, β-catenin and Bad after transfection with si-NC, si-AP4#1 and si-AP4#2 were examined by qRT-PCR and western blot analyses in LNCaP-AI cells. (g) Relative density of AP4, L-plastin, GSK3β, β-catenin and Bad protein expressions after normalization to GAPDH in LNCaP-AI cells. (h–j) The levels of AP4, β-catenin and GSK3β were examined with the PI3K inhibitor LY294002, and overexpression of AP4 could partly rescue the inhibitory effects in the changes of AP4 and β-catenin in western blotting analysis, MTT assay and transwell assays. Error bars indicate S.D.s (n=3), *P<0.05. **P<0.01

Fig 4: Identification of AP4 binding site in L-plastin promoter in PCa cells. (a,b) EMSA and supershift assays consensus AP4 probe containing the consensus AP4 binding site (CAGCTG) was used to validate the presence of the AP4 factor using the NE from LNCaP cells, LNCaP-AI cells and PC-3 cells. Both supershifted complexes and DNA/protein complexes were shown as an arrow head to the right of the panel. (c,d) The chromatin immunoprecipitation (ChIP) assay revealed that AP4 binds the L-plastin promoter. IgG was used as a negative control. (e,f) The diagram shows L-plastin promoter–reporter constructs used in transfection assays indicating the putative AP4 site in the proximal promoter; relative luciferase activity was calculated as firefly luciferase activity compared to the pRL-TK Renilla transfection control plasmid. Three independent experiments were performed. Error bars indicated S.D.s (n=3), *P<0.05. **P<0.01

Fig 5: AP4 promotes tumorigenicity and metastatic potential in vivo. (a–c) The images of animals and tumors were shown. Tumor weights were shown as the means±S.D. when the tumors were collected. (d,e) qRT-PCR and western blot analyzed the expression of L-plastin in tumor tissues from sh-AP4 in PC-3 cells compared with sh-NC cells. (f) Representative images of HE and IHC staining of the tumor. The IHC staining showed that AP4 knockdown decreased the proliferation index Ki67. (g,h) Representative images of lung metastasis of subcutaneous xenografts assay. Histological analysis of lung wet weight is presented as the mean±S.D. (n=4). *P<0.05, **P<0.01