Polo-like kinase 1 is involved in invasion through extracellular matrix

Abstract

Polo-like kinase 1 (PLK1) has important functions in maintaining genome stability via its role in mitosis. Because PLK1 is up-regulated in many invasive carcinomas, we asked whether it may also play a role in acquisition of invasiveness, a crucial step in transition to malignancy. In a model of metaplastic basal-like breast carcinoma progression, we found that PLK1 expression is necessary but not sufficient to induce invasiveness through laminin-rich extracellular matrix. PLK1 mediates invasion via vimentin and beta1 integrin, both of which are necessary. We observed that PLK1 phosphorylates vimentin on Ser82, which in turn regulates cell surface levels of beta1 integrin. We found PLK1 to be also highly expressed in preinvasive in situ carcinomas of the breast. These results support a role for the involvement of PLK1 in the invasion process and point to this pathway as a potential therapeutic target for preinvasive and invasive breast carcinoma treatment.

Figure 1

PLK1 expression is necessary but not sufficient for invasion. A, schematic presentation of the HMT-3522 model of metaplastic breast cancer progression (13). B, RT-PCR analysis for mRNA and Western blot for protein levels of PLK1. S1 cells in two-dimensional or three-dimensional lrECM were grown in the absence of epidermal growth factor (EGF) to completely growth arrest cells as a negative control for PLK1 signal. GAPDH, glyceraldehyde-3-phosphate dehydrogenase. C, Western blot for PLK1 in cells transfected with siRNA to PLK1 versus scrambled control siRNA (Scr.); 44% and 46% reduction in T4-2 and S3-C cells, respectively. Invasion assay for T4-2 or S3-C cells (induced by T4-2 conditioned medium) after cells were transfected with PLK1 or scrambled control siRNA; three experiments, duplicate samples. D, the percentage of Ki67-positive T4-2 cells transfected with PLK1 (◆) or scrambled control (■) siRNAs at indicated time points after release from synchrony; invasion assay for these T4-2 cells. E, the percentage of TUNEL-positive T4-2 cells after transfection with the PLK1 or scrambled control siRNAs; two experiments, duplicate samples.

Figure 2

Vimentin (VIM) and β₁ integrin are necessary for PLK1-mediated invasion. A, the mechanism proposed. B, Western blot for vimentin in T4-2 cells transfected with the indicated siRNAs. Vim2 used for subsequent experiments. Invasion assay for T4-2 cells transfected with the indicated siRNAs. Four experiments, duplicate samples. C, cell surface expression of β1 integrin. Four experiments. Values normalized to S1. P < 0.05, between T4-2 and all other cell types. Invasion assay for T4-2 cells treated with the indicated amounts of β1 integrin blocking antibody A2BII. P < 0.05, compared with untreated control, six experiments. D, invasion assay for T4-2 cells treated with siRNAs against PLK1 or vimentin, or β1 blocking antibody, in combinations indicated; four experiments, duplicate samples. P < 0.05, compared with scrambled control siRNA.

Figure 3

PLK1 affects invasion via phosphorylating vimentin and down-regulating cell surface β1 integrin. A, Western blot for Ser82 phosphorylated vimentin in T4-2 cells treated with the indicated siRNAs. B, Western blot for Ser82 phosphorylated vimentin in T4-2 cells infected with lentivirus expressing WT vimentin pVIM (WT) or mutated vimentin pVIM (S82A). Invasion assay for T4-2 samples infected with lentivirus expressing a WT vimentin pVIM (WT) or mutated vimentin pVIM (S82A). Normalized to WT values. P = 0.036, three experiments, triplicate samples. C, cell surface expression of β1 integrin on T4-2 cells treated with the indicated siRNAs. P = 0.0007 (PLK1-scrambled control siRNA) and 0.003 (vimentin-scrambled control siRNA); four experiments. D, cell surface β1 integrin levels (totaland active, as indicated), normalized to T4-2 cells expressing WT vimentin; four experiments.

Figure 4

PLK1 in vivo. A, frequency of tumor formation and mean tumor volumes in fat pad. T4-2 transfected with siRNA against PLK1 (n = 10) versus scrambled control siRNA (n = 5). B, control experiment for immunohistochemical detection of PLK1, comparing mock antibody with PLK1 antibody-treated samples. Bar, 100 µm. C, example; PLK1 immunohistochemical signal in normal, in situ, and invasive samples from the same patient. Bar, 100 µm. D, PLK1 signal intensity. Two normal (N) and two invasive (I) cases as controls, compared with eight cases each containing areas of normal as well as in situ/preinvasive (P) and invasive carcinomas.