The full-length IGFBP5 cDNA was generated by reverse transcriptase-polymerase chain reaction (RT-PCR) using total RNA from MCF7 breast cancer cell line. The primer sequence was: forward primer, 5'-GCCACCATGGTGTTGCTCACCGCGGTCCTCCTGC-3'; and reverse primer, 5'-TCACTCAA-CGTTGCTGCTGTCGAAGGTGTG-3'. A Kozak consensus GCCACC is inserted in front of the IGFBP5 translation initiation site to increase the translation efficiency. The PCR product was inserted into pCR^®2.1TOPO plasmid (Invitrogen, Carlsbad, CA), and several clones were chosen for DNA sequencing. The IGFBP5 cDNA with the wild-type sequence was digested from the TA-cloning plasmid using EcoRI restriction enzyme and was inserted into the EcoRI site of the pcDNA3.1(+) expression vector (Invitrogen) to obtain the IGFBP5wt/pcDNA3.1(+) construct. Again, several clones were sequenced, and the clone with right orientation and the cDNA sequence was chosen to perform cell transfection to generate wild-type IGFBP5-overexpressing cell lines. The mutant IGFBP5 expression construct [IGFBP5mt/pcDNA3(+)] was generated using the QuickChange Kit (Stratagene, La Jolla, CA) using the IGFBP5wt/pcDNA3.1(+) expression plasmid as a template. As shown in Figure 1A, 5 amino acids in the NLS (214 to 218; black characters) were deleted. Several clones were sequenced, and the clone with right orientation and IGFBP5 cDNA with the mutant sequence was chosen to perform the transfection and to generate mutant IGFBP5-overexpressing cell lines.

To generate wild-type and mutant IGFBP5 fluorescence fusion proteins, we used the same upstream primer used to generate the wild-type IGFBP5 expression construct. To make the fusion protein in the reading frame, a new downstream primer was generated using the following sequence: 5'-CCGGAATTCCGGACTCAACGTTGCTGCTGTCGAAGG-TGTG-3'. The PCR products were generated using the IGFBP5wt/pcDNA3.1(+) and the IGFBP5mt/pcDNA3.1(+) plasmids to generate wild-type and mutant IGFBP5-fluorecence fusion proteins (respectively). The PCR products were inserted into the pDsRed2-N1 vector (BD Biosciences, San Jose, CA) at the EcoRI restriction enzyme site after digestion with the same enzyme to obtain the IGFBP5wt/pDsRed2-N1 and the IGFBP5mt/pDsRed2-N1 constructs.

MDA-MB-435 cells were grown at 70% confluence and transfected with the pcDNA3.1(+) vector alone, with IGFBP5wt/pcDNA3.1(+), and with IGFBP5mt/pcDNA3.1(+) using Nucleofector (Amaxa, Gaithersburg, MD). The transfection was performed with Solution T and program T₂₀. Selection of stable cell clones was carried out in the presence of 600 μg/ml G418 (Invitrogen) in culture medium at 48 h post-transfection, and the cells were re-fed every 3 days. Several single clones were selected and expanded for 2–3 weeks, and the IGFBP5 expression level was detected by western blot analysis.

The cell growth rates of the vector and of the wild-type and the mutant IGFBP5 transfectants were measured by bromodeoxyuridine (BrdU) incorporation using BrdU flow kits (BD Biosciences) per the manufacturer's instructions. Briefly, the cells were cultured in complete medium for 24 h and then incubated in fresh complete medium with 10 μM BrdU for 2 h. The incorporated BrdU was visualized by immunofluorescence staining using an anti-BrdU antibody conjugated with fluorescein isothiocyanate. The nuclei were counterstained with DAPI, and images were taken at a magnification of 63×. Cells not treated with anti-BrdU antibody served as negative controls. The number of total cells and the BrdU-positive cells were counted from 3 fields of each group under a fluorescence microscope, and the quantification data were graphed as a ratio of the number of BrdU-positive cells to the number of total cells. The geometric mean fluorescence intensity from the same cell population was recorded by fluorescence-activity cell sorter (FACS) analysis. The fluorescence intensity from the negative control was graphed as zero, and the fluorescence intensity from the rest of the samples became the value remaining after the fluorescence intensity was subtracted from the negative control.

The MDA-MB-435 breast cancer cell line was obtained from the American Type Culture Company (ATCC; Menassas, VA) and was routinely maintained in Dulbecco's modified Eagle's medium/F-12 supplemented with 10% fetal bovine serum (FBS) and 1% antibiotic cocktail under standard conditions. For microscopy imaging, 2 × 10⁴ IGFBP5wt/pDsRed2-N1 or IGFBP5mt/pDsRed2-N1 fusion construct-transfected cells were seeded per cell into glass slide chambers (BD Pharmingen) and allowed to attach for 40 h. The cells were then fixed with 4% paraformaldehyde in phosphate-buffered saline. After washing, the cells were counterstained with DAPI. Images were taken using an Axioplan 2 fluorescence microscope (ZEISS, Oberkochen, Germany) at 63× magnification.

Falcon cell-culture inserts containing an 8-μ pore PET membrane (BD Biosciences) were used to perform the in vitro migration assay. The vector-alone transfectants and the IGFBP5wt/pcDNA3.1(+) and IGFBP5mt/pcDNA3.1(+) stable cells were seeded into the upper chambers at a density of 2 × 10⁴ cells/well in 500 μl of serum-free medium. To the lower chambers, 1 ml of 10% FBS-containing medium was added. The plates were then incubated for 20 h at 37°C with 5% CO₂. The unmigrated cells in the upper chambers were gently removed using a cotton swab. Cells that had migrated through the membrane pore into the lower sides of the filters were fixed, stained with HEMA-DIFF solution (Fisher Scientific, Philadelphia, PA), and counted under a regular light microscope.

IGFBP5 in the whole-cell lysate and conditioned media (CM) was detected according to a previously describe standard protocol 21. Briefly, cell lysates from different cell populations were collected, and protein concentration was measured using the Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA). Equal amounts of proteins were then loaded into 10% sodium dodecyl sulfate-polyacrylimide gels and resolved by electrophoresis. After transfer to nitrocellulose membranes, it was probed with anti-IGFBP5 antibody (diluted 1:1000 in TBST, 5% non-fat dry milk; Santa Cruz Biotechnology, Santa Cruz, CA). Bands were detected using the chemiluminescence detection method (Amersham Biosciences, Piscataway, NJ) and exposed on x-ray film. The IGFBP5 level in the same volume of conditioned media from the same cell populations of similar densities for the cell lysate was measured simultaneously.

Data are expressed as the mean ± standard error. Statistical analysis was performed using the Student's t test. Differences in means were evaluated by a 2-tailed t test assuming unequal variances. A p value of ≤ 0.05 was considered statistically significant.

To determine whether the nuclear localization sequence is required and sufficient for transporting IGFBP5 into the nucleus, we constructed a mutant IGFBP5 expression vector in which 5 amino acids in the NLS were deleted (Figure 1A). We transfected both wild-type and mutant IGFBP5 expression vectors into the MDA-MB-435 breast cancer cells, and established several stable cell lines overexpressing these vectors. IGFBP5 expression was detected in the cells by western blot analysis using whole-cell lysates (Figure 1B). The parental cells and the vector-alone transfectants were used as controls. In order to visualize the subcellular localization of wild-type and mutant IGFBP5 in the cells, we generated constructs that fused fluorescent pDesRed on the C-terminal region of either wild-type IGFBP5 or mutant IGFBP5. After transfection of the expressing vectors, we observed that wild-type IGFBP5 was mostly localized in the nucleolus. The nucleolus localization of wild-type IGFBP5 was confirmed by the observation of colocalization of IGFBP5 and the known nucleolus protein B23 22 (Figure 2). Mutated IGFBP5 was not detected in the nucleolus but was detected at a low level in the cytoplasm (Figure 3). To determine whether the low cytoplasmic level of mutant IGFBP5 was the result of increased secretion of the IGFBP5 into the media, we collected both cell lysate and conditional media from the stable cell lines and performed western blot analysis. Again, we could not detect IGFBP5 from either the cell lysates or the conditioned media in the parental and vector-transfected cells (Figure 4). We did observe overexpression of IGFBP5 from the IGFBP5wt and IGFBP5mt stable clones, and the cells overexpressing the mutant form IGFBP5 showed a much higher level of IGFBP5 in the conditioned medium compared with the wild-type IGFBP5 transfectants (Figure 4). Interestingly, we observed that the majority of the secreted wild-type IGFBP5 was cleaved whereas that majority of the mutant IGFBP5 remained intact in the conditioned media (Figure 4).

To determine any proliferative effects of the mutant form of IGFBP5, we performed BrdU incorporation analyses by flow cytometry and fluorescence microscopy. The cells overexpressing wild-type or mutant IGFBP5 were cultured in the complete media for 24 h and then incubated in the fresh complete media with 10 μM of BrdU for 2 h. The vector alone transfectant was used as the control. The incorporated BrdU was visualized by immunofluorescence staining using an anti-BrdU antibody conjugated with fluorescein isothiocyanate. Cells with no anti-BrdU antibody treatment served as the negative controls for the immunostaining. More cells overexpressing the mutant form of IGFBP5 were positively stained compared with the vector and the wild-type IGFBP5-overexpressing cells (Figure 5A). To obtain quantification data from the assay, we randomly chose 3 fields to count the total cells and the BrdU positive cells, and the data were graphed as a ratio of BrdU positive cells to total cells. The result revealed that overexpression of wild-type IGFBP5 significantly inhibited BrdU incorporation compared with the vector transfectants (*, p < 0.05) (Figure 5B). In contrast, overexpression of the mutant IGFBP5 significantly promoted BrdU incorporation (**, p < 0.005) (Figure 5B). To confirm our findings, the geometric mean fluorescence intensity from the same cell populations after the immunofluorescence staining was recorded by FACS. The fluorescence intensity from the negative control was graphed as zero, and fluorescence intensity from the rest of samples was the value after the fluorescence intensity value was subtracted from the negative control. The results from FACS analysis showed that wild-type IGFBP5 inhibited cell growth compared with the vector transfectant and that the mutated form of IGFBP5 promoted cell growth compared with the vector transfectant (Figure 5C and 5D).

To determine whether mutation of NLS of IGFBP5 affects cell motility, we performed an in vitro migration assay using the stable cells overexpressing either wild-type or mutant IGFBP5. The vector-alone transfected cells were used as the control. As shown in Figure 6, the mutant IGFBP5-overexpressing cells had a significantly higher number of migrated cells than the vector-alone control cells (p < 0.01) (Figure 6A and 6B). In contrast, the wild-type IGFBP5-overexpressing cells had significantly lower numbers of migrated cells than the vector-alone control cells (p < 0.001) (Figure 6A and 6B). To confirm that deletion of amino acids 214 to 218 within the NLS of IGFBP5 promotes breast cancer cell motility, we transiently transfected the MDA-MB-435 breast cancer cells with IGFBP5wt/pcDNA3.1(+) and IGFBP5mt/pcDNA3.1(+) constructs and performed the in vitro migration assay 48-h post-transfection. The vector-alone transfected cells were used as the control. As shown in Figure 6C, we observed results similar to those from the in vitro migration experiments carried out using the stable cells. IGFBP5wt/pcDNA3.1(+)-transfected cells had a significantly lower number of migrated cells than the vector-alone control cells (* p < 0.05). In contrast, IGFBP5mt/pcDNA3.1(+)-transfected cells had a significantly higher number of migrated cells than the control cells (** p < 0.001) (Figure 6C).