The subcellular localization of IGFBP5 affects its cell growth and migration functions in breast cancer

Abstract

Background: Insulin-like growth factor binding protein 5 (IGFBP5) has been shown to be associated with breast cancer metastasis in clinical marker studies. However, a major difficulty in understanding how IGFBP5 functions in this capacity is the paradoxical observation that ectopic overexpression of IGFBP5 in breast cancer cell lines results in suppressed cellular proliferation. In cancer tissues, IGFBP5 resides mainly in the cytoplasm; however, in transfected cells, IGFBP5 is mainly located in the nucleus. We hypothesized that subcellular localization of IGFBP5 affects its functions in host cells.

Methods: To test this hypothesis, we generated wild-type and mutant IGFBP5 expression constructs. The mutation occurs within the nuclear localization sequence (NLS) of the protein and is generated by site-directed mutagenesis using the wild-type IGFBP5 expression construct as a template. Next, we transfected each expression construct into MDA-MB-435 breast cancer cells to establish stable clones overexpressing either wild-type or mutant IGFBP5.

Results: Functional analysis revealed that cells overexpressing wild-type IGFBP5 had significantly lower cell growth rate and motility than the vector-transfected cells, whereas cells overexpressing mutant IGFBP5 demonstrated a significantly higher ability to proliferate and migrate. To illustrate the subcellular localization of the proteins, we generated wild-type and mutant IGFBP5-pDsRed fluorescence fusion constructs. Fluorescence microscopy imaging revealed that mutation of the NLS in IGFBP5 switched the accumulation of IGFBP5 from the nucleus to the cytoplasm of the protein.

Conclusion: Together, these findings imply that the mutant form of IGFBP5 increases proliferation and motility of breast cancer cells and that mutation of the NLS in IGFBP5 results in localization of IGFBP5 in the cytoplasm, suggesting that subcellular localization of IGFBP5 affects its cell growth and migration functions in the breast cancer cells.

Figure 1

Establishment of the wild-type and mutant forms of IGFBP5 stably expressing cell lines. A. Schematic diagram of the IGFBP5 cDNA fragments used to generate the IGFBP5 expression constructs. Wild-type IGFBP5 contains a NLS located at the C-terminal domain of the protein from amino acid 201 to 218 and indicated by a gray solid square. For the mutant form of IGFBP5, 5 amino acids in the NLS from 214–218 (black characters) were deleted. B. Western blot analysis of IGFBP5 expression. MBA-MD-435 breast cancer cells were stably transfected with either the wild-type or the mutant forms of IGFBP5 construct and the whole cell lysate from parental MBA-435 cells (P) and different stable lines were subjected to western blot analysis (40 μg/lane) using an anti-human IGFBP5 antibody. The blot was re-probed with actin to normalize for protein loading. Neither parental nor vector-transfectant (V) cells had detectable levels of IGFBP5. 10 ng of purchased human recombinant IGFBP5 protein was served as a positive control (+). wt, wild-type and d, deletion form IGFBP5.

Figure 2

Deletion of amino acids from 214 to 218 of IGFBP5 protein eliminates its nuclear localization. The wild-type (IGFBP5wt/pDsRed2-N1) or mutant (IGFBP5mt/pDsRed2-N1) fusion construct was transiently transfected into MBA-MD-435 breast cancer cells. The cells were fixed 48 h post-transfection and counterstained with DAPI. The images were taken by a fluorescence microscope at 63× magnification.

Figure 3

Wild-type IGFBP5 mainly localized at the cellular nucleolus. IGFBP5 wt/pDsRed2-N1 and B23/pGFP constructs were co-transfected into MBA-MD-435 cells. After 48 h post-transfection, the cells were fixed and counterstained with DAPI. The IGFBP5 fusion protein showed co-localization with B23, which is a marker for the nucleolus.

Figure 4

Western blot analysis of IGFBP5 expression in both cell lysate and conditioned medium. One million cells were cultured in 100-mm diameter tissue culture dish in the complete medium for 24 h followed by an additional 24-h incubation in the serum-free medium. Both cell lysate and conditioned medium were collected. The conditioned medium was concentrated 3 times using YM-10 column (Millipore) and 40 μg of protein from the cell lysate and 45 μl of the concentrated conditioned medium were subject for western blot analysis. The blot was re-probed with anti-actin to normalize for protein loading. wt, wild-type and mt, mutant form IGFBP5.

Figure 5

Changes of IGFBP5 cellular distribution promotes breast cancer cell proliferation. The cell growth rates for the vector and IGFBP5 transfectants were measured by BrdU incorporation using the BrdU flow kits per the manufacturer's instruction. A. The cells were incubated in 10% FBS medium containing 10 μM BrdU for 2 h followed by immunofluorescence staining with FITC conjugated BrdU antibody. The nuclei were counterstained with DAPI and images were taken at a magnification of 63×. Non-BrdU-treated parental cells were served as a negative control. B. Quantification data from 3 fields of each group. The total cells and the BrdU-positive cells were counted from 3 fields of each group, and the data were graphed as a ratio of BrdU positive cells vs. the total cells. The error bars represents standard deviation from three counts of samples. *, p < 0.05 and **, p < 0.005.C. The geometric mean fluorescence intensity from the same cell population as in A was recorded by FACS. D. The fluorescence intensity shown in B is shown as a bar graph. The fluorescence intensity from the negative control is graphed as zero, and the fluorescence intensity from the rest of the samples is the value after the value of the fluorescence intensity is subtracted from the negative control.

Figure 6

Deletion of 5 amino acids from 214 to 218 of IGFBP5 protein promotes MDA-MB-435 breast cancer cell motility. A. Representative light-microscopy images from the migration assay. B. Quantification data of in vitro migration assay from vector transfectants (vector 1 and vector 4), wild-type IGFBP5 overexpression clones (wt-c11 and wt-c13), and mutant IGFBP5 overexpression clones (mt-c2, mt-c6, and mt-c29). The error bars represent standard deviations from 2 separate experiments with triplicate samples for each (n = 6). *, p < 0.001. C. Quantification data of in vitro migration assay from transient transfection. The error bars represent standard deviations from triplicate samples. *, p < 0.05 and **, p < 0.001.