Differential CRABP-II and FABP5 expression patterns and implications for medulloblastoma retinoic acid sensitivity

Abstract

Medulloblastoma (MB) cells exhibit different responses to retinoid acid (RA) for reasons that are poorly understood. RA signaling can be transduced by two approaches that are mediated by cellular retinoic acid-binding protein 2 (CRABP-II) as a tumor-suppressive pathway, and by fatty acid-binding protein 5 (FABP5) as a tumor-promoting pathway. The biological effects of RA on cancer cells are largely determined by the patterns of CRABP-II and FABP5 expression. This study aims to profile the statuses of CRABP-II and FABP5 expression in MB and to evaluate their correlation with RA sensitivities using RA-sensitive (Med-3) and RA-insensitive (UW228-2, UW228-3) MB cells. Our results show that CRABP-II is distinctly expressed and the level of FABP5 is extremely low in Med-3 cells, while the patterns of CRABP-II and FABP5 expression are reversed in UW228-2 and UW228-3 cells. RA up-regulates CRABP-II expression in Med-3 cells, whereas it up-regulates FABP5 expression in the other two cell lines. The FABP5-specific inhibitor BMS309403 increases the RA sensitivity of UW228-2 cells (p < 0.01). Tissue microarray-based immunohistochemical staining showed CRABP-II/FABP5 expression patterns in MB that were variable (CRABP-II-/FABP5-, CRABP-II-/FABP5+, CRABP-II+/FABP5- and CRABP-II+/FABP5+) and imbalanced (CRABP-II↑/FABP5↓ and CRABP-II↓/FABP5↑). MB cases exhibited patterns ofCRABP-II-/FABP5- (12.24%, 6/49), CRABP-II-/FABP5+ (30.61%, 15/49) or CRABP-II↓/FABP5↑ (12.24%, 6/49), implicating unresponsiveness or insensitivity to RA. In conclusion, the ratios of CRABP-II/FABP5 levels are closely related to the RA sensitivities of MB cells. The differential CRABP-II and FABP5 expression patterns are prospective parameters, and of potential value in personalized RA therapy for MB.

Fig. 1. Expression of RA receptors in MB cells and in vitro responsiveness to RA. (A) HE morphological staining was performed on UW228-2 and Med-3 cells that were either untreated (control) or treated with 10 μM RA for 72 h. (B) Western blotting profiling of RARα and PPARβ/δ expression in UW228-2, UW228-3 and Med-3 before and after RA treatment. Grayscale quantitative analysis was performed on the western blot results. *p < 0.05, compared with untreated UW228-2 cells; ☆p < 0.05, compared with untreated UW228-3 cells; #p < 0.05, compared with untreated Med-3 cells. (C, D and E) Evaluation of the responses of UW228-2, UW228-3 and Med-3 to 10 μM RA for 72 h using an MTT assay. *p < 0.05, compared with control.

Fig. 2. CRABP-II and FABP5 expression patterns in MB cells with and without RA treatment. (A) CRABP-II and FABP5 mRNA expressions were measured using real-time RT-PCR. **p < 0.01. (B) Immunofluorescence staining for CRABP-II and FABP5 in RA-treated MB cells. Magnified regions of each image are shown in the insets. (C) Western blot analysis of UW228-2 and Med-3 cells treated with and without RA. Grayscale quantitative analysis was performed on the Western blot results. *p < 0.05. Scale bar = 100 μm.

Fig. 3. Effect of FABP5 inhibitor BMS309403 on the growth and survival of MB cells treated with RA. (A) HE morphological staining performed on UW228-2 cells under the following conditions: normal culture (ctrl); 25 μM BMS309403 treatment; 10 μM RA treatment; 25 μM BMS309403 pretreatment for 6 h followed by 10 μM RA treatment (RA + BMS309403). (B and C) Evaluation of RA responses by MTT and cell counting. **p < 0.01.

Fig. 4. Expression patterns of CRABP-II and FABP5 in MB tissues. Immunohistochemical profiling of CRABP-II and FABP5 expression patterns in classic and large-cell MBs. The staining patterns were scored as CRABP-II↑/FABP5↓ (FABP5 down-regulated) and CRABP-II↓/FABP5↑ (FABP5 up-regulated) based on the labeling intensity of these two proteins in individual cases. The insets illustrate regions at higher magnification. Scale bar = 50 μm.