Fatty acid binding protein 5 promotes metastatic potential of triple negative breast cancer cells through enhancing epidermal growth factor receptor stability

Abstract

Fatty acid binding protein 5 (FABP5), an intracellular lipid binding protein, has been shown to play a role in various cancers, including breast cancer. However, FABP5 and its role in triple negative breast cancer (TNBC) have not been studied. We show FABP5 protein expression correlates with TNBC, high grade tumors, and worse disease-free survival in a tissue microarray containing 423 breast cancer patient samples. High FABP5 expression significantly correlates with epidermal growth factor receptor (EGFR) expression in these samples. Decreased tumor growth and lung metastasis were observed in FABP5-/- mice othotopically injected with murine breast cancer cells. FABP5 loss in TNBC tumor cells inhibited motility and invasion. Mechanistic studies revealed that FABP5 knockdown in TNBC cells results in decreased EGFR expression and FABP5 is important for EGF-induced metastatic signaling. Loss of FABP5 leads to proteasomal targeting of EGFR. Our studies show that FABP5 has a role in both host and tumor cell during breast cancer progression. These findings suggest that FABP5 mediates its enhanced effect on TNBC metastasis, in part, through EGFR, by inhibiting EGFR proteasomal degradation. These studies show, for the first time, a correlation between FABP5 and EGFR in enhancing TNBC metastasis through a novel mechanism.

Keywords: epidermal growth factor receptor; fatty acid binding protein 5; metastasis; protein degradation; triple negative breast cancer.

Figure 1. FABP5 expression correlates with worse prognosis in patient samples

(A) Representative images from SFBCAS tissue microarray immunohistochemically stained for FABP5, scale bar is 80 μm. (B) Graphical representation of FABP5 expression in patient samples with low and high grade tumors. (C) Graphical representation of FABP5 expression in patient samples with TNBC status. (D) Disease-free survival of patients expressing low or high FABP5. (E) Western blot analysis of TNBC cell lines MDA-MB-231, MDA-MB-436, and SUM159PT, ER⁺ cell lines, T47D and MCF7, and an epithelial cell line, MCF10A, for FABP5 expression (15 kDa) with GAPDH (37 kDa) as a loading control. (F) Recurrence-free survival analysis from 3,455 patients expressing high or low FABP5 (202345_s_at) dichotomized by median FABP5 expression. (G) Recurrence-free survival in 1,000 systemically untreated patients expressing high or low FABP5 (202345_s_at) dichotomized by median FABP5 expression. (H) Recurrence-free survival analysis from 936 patients with lymph node positive (left panel) or lymph node negative status (right panel) expressing high or low FABP5 (202345_s_at) dichotomized by median FABP5 expression. (I) Graphical representation of the correlation between FABP5 and EGFR expression in the SFBCAS TMA.

Figure 2. Host FABP5 modulates primary tumor growth and metastasis

(A) Tumor growth of PyMT tumors in wild-type (WT) or FABP5 knockout (FABP5^−/−) mice. (B) Representative images of excised tumors in wild-type (right panel) and FABP5^−/− (left panel) mice (n=8/group). (C) Endpoint weights of wild-type and FABP5^−/− tumors 6 weeks post injection. (D) Tumor growth of E0771 tumors in wild-type of FABP5^−/− mice. (E) Representative pictures of E0771 tumors in wild-type and FABP5^−/− mice (lower panel) (n=4/group). (F) Number of metastatic nodules in wild-type and FABP5^−/− mice lungs (G) Representative images of whole lungs fixed in Bouin's stain of wild-type and FABP5^−/− mice (left panel) and H&E sections of wild-type and FABP5^−/− lungs with micro-metastasis (black arrow), scale bar is 80 μm (right panel). Data is represented as mean ± SD.

Figure 3. Functional consequences of FABP5 knockdown in TNBC cancer cells

(A) FABP5 western blot analysis of shRNA mediated FABP5 knockdown (FABP5 KD) in MDA-MB-231 cells and GAPDH as loading control. (B) Representative image of scrambled control (scrambled) and FABP5 KD cells attached to fibronectin coated wells. (C) Number of scrambled and FABP5 KD cells attached to fibronectin coated wells. (D) Representative images of wound healing of scrambled and FABP5 KD cells in 5% FBS/DMEM over 24 hours. (E) Percentage wound closure of scrambled and FABP5 KD cells over 24 hours. (F) Number of migrated scrambled and FABP5 KD cells after 4 hours in the presence of 5% FBS in DMEM. (G) Number of invaded scrambled and FABP5 KD cells after 24 hours in the presence of 5% FBS in DMEM. (H) Scrambled and FABP5 KD cells were serum starved for 36 hours and media was collected and run on a gelatin zymogram. Gel was stained and imaged for MMP9 and MMP2. (I) Graphical representation of zymogram band intensity of MMP9 and MMP2. Data is represented as mean ± SD.

Figure 4. FABP5 mediates metastatic function of TNBC cell line through EGFR

(A) Mean fluorescence intensity of EGFR on scrambled and FABP5 KD cells by FACS analysis. (B & C) Confocal microscopy of scrambled and FABP5 KD cells for EGFR (red) expression; DAPI (blue) in serum-free medium (SFM) (B) or EGF (100 ng/mL) stimulation (C). (D) Representative images of scrambled and FABP5 KD wound healing in the absence or presence of EGF (50 ng/mL) at 24 hours. (E) Percentage wound healing of scrambled and FABP5 KD in the absence or presence of EGF (50 ng/mL) at 24 hours. Data is represented as mean ± SD.

Figure 5. FABP5 KD dampens EGFR mediated downstream signaling of metastatic targets

(A) Scrambled and FABP5 KD cells were serum starved overnight and stimulated with EGF (50 ng/mL) for indicated time points. Cell lysates were harvested and subjected to western blot analysis for phospho-EGFR (Y1173), EGFR, phospho-Pyk2, Pyk2, and GAPDH as a loading control. (B) Quantification of relative phospho-EGFR expression in scrambled cells and FABP5 KD cells treated with EGF (50 ng/mL) for varying time pointscalculated with respect to total EGFR expression (upper panel) and relative phospho-Pyk2 expression calculated with respect to total Pyk2 expression (lower panel). C) FABP5 KD cells were treated with varying concentrations of MG-132 or veh for 4 hours. Cells were lysed and analyzed by western blot for EGFR and GAPDH as a loading control. (D) Quantification of relative EGFR expression calculated with respect to GAPDH. (E) Ubiquitin assay of EGFR in scrambled and FABP5 KD cells treated for 4 hours with veh or MG-132 (MG) (15 μM).