A potential regulatory loop between Lin28B:miR‑212 in androgen-independent prostate cancer

Abstract

Lin28 is a family of RNA binding proteins and microRNA regulators. Two members of this family have been identified: Lin28A and Lin28B, which are encoded by genes localized in different chromosomes but share a high degree of sequence identity. The role of Lin28B in androgen-independent prostate cancer (AIPC) is not well understood. Lin28B is expressed in all grades of prostatic carcinomas and prostate cancer cell lines, but not in normal prostate tissue. In this study we found that Lin28B co-localized in the nucleus and cytoplasm of the DU145 AIPC. The expression of Lin28B protein positively correlated with the expression of the c-Myc protein in the prostate cancer cell lines and silencing of Lin28B also correlated with a lower expression of the c-Myc protein, but not with the downregulation of c-Myc messenger RNA (mRNA) in the DU145 AIPC cells. We hypothesized that Lin28B regul-ates the expression of c-Myc protein by altering intermediate c-Myc suppressors. Therefore, a microRNA profile of DU145 cells was performed after Lin28B siRNA silencing. Nineteen microRNAs were upregulated and eleven microRNAs were downregulated. The most upregulated microRNAs were miR-212 and miR-2278. Prior reports have found that miR-212 is suppressed in prostate cancer. We then ran TargetScan software to find potential target mRNAs of miR-212 and miR-2278, and it predicted Lin28B mRNA as a potential target of miR-212, but not miR-2278. TargetScan also predicted that c-Myc mRNA is not a potential target of miR-212 or miR-2278. These observations suggest that Lin28B:miR-212 may work as a regulatory loop in androgen-independent prostate cancer. Furthermore, we report a predictive 2-fold symmetric model generated by the superposition of the Lin28A structure onto the I-TASSER model of Lin28B. This structural model of Lin28B suggests that it shows unique microRNA binding characteristics. Thus, if Lin28B were to bind miRNAs in a manner similar to Lin28A, conformational changes would be necessary to prevent steric clashes in the C-terminal and linker regions between the CSD and ZNF domains.

Figure 1

Lin28B expression in human prostate cancer tissues and cellular localization of Lin28B in AIPC cells. (A) Normal prostate tissue. (B) Adenocarcinoma tissue. (C) Undifferentiated carcinoma. (D) Cellular localization of Lin28B in DU145 AIPC cells.

Figure 2

Lin28B and c-Myc expression in the prostate cancer cell lines LnCaP and VCaP (androgen-dependent), PC3 and DU145 (androgen-independent). (A) Determined by western blotting. (B) Determined by flow cytometer (LnCaP and VCaP, DU145 and PC3 prostate cancer cell lines).

Figure 3

Expression of Lin28B and c-Myc in DU145 AIPC cells upon silencing Lin28B using siRNA. (A) The silencing of Lin28B protein correlates with the downregulation of c-Myc protein at 48 and 72 h as detected by western blotting. (B) Quantification using flow cytometry and data analysis by FlowJo software at 42 h upon transfection with Lin28B siRNA showing background fluorescence (2.4%), Lin28B expression upon Lin28B silencing (32.7%), control expression of Lin28B in DU145 cells (76.8%), c-Myc expression upon Lin28B silencing (58.6%) and control expression of c-Myc in DU145 cells (99.9%). (C) The fold difference in the Lin28B and c-Myc messenger RNA expression relative to the control (2^−ΔΔCT) upon Lin28B silencing with c-MYC messenger RNA expression slightly decreasing at 24 h (0.72) but recovering by 48 h (1.45).

Figure 4

Fold difference (ratio of Lin28B silencing samples vs. control) of microRNAs upregulated in DU145 AIPC cells. MicroRNA profiling was determined by GeneChip-miRNA 2.0 array. A threshold of 1.5-fold change was used.

Figure 5

Fold difference (ratio of Lin28B silencing samples vs. control) of microRNAs downregulated in the DU145 AIPC cells. MicroRNA profiling was determined by GeneChip-miRNA 2.0 array. A threshold of 1.5-fold change was used.

Figure 6

(A) Crystal structure of Lin28A:let7d complex (PDB:3TRZ). The complex consists of 2-fold symmetric Lin28A subunits (magenta/cyan) that coordinate the miRNA (red/blue) via interaction with the Lin28A CSD and ZNF domains. (B) Lin28B structure predicted from I-TASSER colored as in panel A. C-terminal residues of the putative Lin28B dimer are indicated by the arrow. (C) Zoomed-in view of the C-terminal residues (Pro 222-Lys 240) downstream of the ZNF2 domain as well as the N-terminal residues (Met1 to Gly 10) that would need to adopt a different conformation.

Figure 7

Model of the potential regulatory loop between Lin28B and miR-212. (A) In androgen-independent prostate cancer. (B) Normal prostate.