The purpose of this study was to analyse the early events triggered by PlGF in the responsive breast cancer cell, and elucidate how the peptide, BP-1, works in inhibiting cellular motility in vitro
, and in preventing metastasis formation in xenograft models of breast cancer (Taylor and Goldenberg, 2007
). To carry out this, the effect of inhibitors on cellular motility, a pro-metastatic behaviour, was analysed. The results show that PlGF-driven breast tumour cell motility include the activation of ERK1/2 kinases, stabilisation of IF proteins, and re-organisation of the actin cytoskeleton. The PlGF specificity of these results was shown by their ablation with anti-Flt-1 antibody or BP-1.
The ERKs are activated by a number of different factors, including c-Src, the G-protein-linked kinase Raf, and MEK (Barros and Marshall, 2005
; Chen et al, 2006
). Increased pERK stimulated by PlGF was reversed by treatment with BP-1 or a MEK-specific inhibitor. Ablation of MEK not only prevented pERK formation, but it also prevented PlGF-stimulated cellular migration. Thus, as the only known substrates of MEK are the ERK1/2 kinases, it can be concluded that PlGF-stimulated motility depends on activation of the ERKs by MEK. This does not eliminate other potential ERK activators, and preliminary data (not shown) suggest that Raf participates in PlGF effects as well.
While this study was in progress, another report investigated PlGF-mediated migration of leukaemia cells (Casalou et al, 2007
). Our findings are in agreement with this report, in that we found no evidence of Akt involvement in PlGF-mediated breast cancer cell movement. However, unlike Casalou et al
, little evidence of p38 MAPK involvement in PlGF-mediated breast cancer cell movement was found because movement was abrogated by inhibition of MEK, which does not activate p38 MAPK. Therefore, it is unlikely that p38 MAPK contributes substantially to the movement of breast cancer cells in the presence of PlGF. Another point of difference was the finding that VEGF had little to no effect on breast cancer cell migration (Taylor and Goldenberg, 2007
). This contrasts with the results obtained by Casalou et al
for leukaemia cells. These differences may be due to different effects of VEGF and PlGF on cancers of epithelial origin, such as breast cancer, in contrast to cancers of hematopoietic origin, where the function of VEGF and PlGF may be redundant.
When activated, the ERK kinases localise to the nucleus or to focal adhesions where they are associated with motility by interaction with cytoskeletal substrates, including IF proteins, the actin cytoskeleton, and myosin light chain kinase (Zheng and Guan, 1994
; Schlaepfer et al, 1998
; Arthur and Burridge, 2001
; Pawlak and Helfman, 2002
; Yin et al, 2005
). This study presented evidence that PlGF promoted translocation of pERK1/2 to the cellular periphery, and so there it may function to increase motility.
Although CK18 and CK19 are expressed by normal breast epithelium, their expression is often dysregulated in breast cancers (Ferrero et al, 1990
; Boecker et al, 2002
). Aggressive MDA-MB-231 displays a pattern of predominant CK19 with little or no CK18, whereas this is not the case for MCF-7 or MDA-MB-468, which both produce abundant CK18 (and CK19). Human breast cancers expressing high levels of CK18 (without vimentin), similar to MCF-7 or MDA-MB-468, are less likely to recur. In addition, MDA-MB-231 transfected with CK18 obtains an epithelial phenotype, reduces aggressiveness in tumour growth, and loses vimentin expression (Buhler and Schaller, 2005
). On the basis of the results of this study, in which CK19 and vimentin expression rose on PlGF stimulation, it is possible that RTK activation, such as Flt-1, stimulates co-expression of CK19 and vimentin.
Likewise, vimentin, a mesenchymal IF protein not expressed by normal breast cells, is associated with poorer clinical outcomes and resistance to apoptosis and drugs (Yuan et al, 1997
; Stathopoulou et al, 2002
; Willipinski-Stapelfeldt et al, 2005
), as well as increased metastasis and invasion (Hartig et al, 1998
; Tolstonog et al, 2001
; Singh et al, 2003
; Hu et al, 2004
; Schoumacher et al, 2010
). McInroy and Maatta (2007)
showed that the aggressiveness of MDA-MB-231 depends on vimentin expression, because its ablation impaired migration and invasion. Our data show that disrupting PlGF or Flt-1 signalling with antibody or BP-1 depressed vimentin, CK19 expression, and motility in MDA-MB-231. Thus, in the case of MDA-MB-231 and perhaps other breast cancers, PlGF, or other abnormally expressed growth factors, may maintain the high CK19 and vimentin levels necessary for motility and invasion. The results also suggest that PlGF does this by stabilising the proteins independently of ongoing transcription and translation, and that, because of their concurrent increase, both factors are controlled by intersecting or common pathways, some of which are regulated by PlGF.
Flt-1 has been considered a cell surface-bound receptor (Wang et al, 2000
; Liu et al, 2005
); however, nuclear expression that delivered intracrine signals promoting breast cancer cell survival through VEGF was reported by Lee et al (2007)
. Our previous report indicated that exogenous PlGF, rather than VEGF, promoted cellular movement, and that soluble Flt-1 and anti-PlGF antibody blocked migration. In this study, we show that exogenous anti-Flt-1 attenuates PlGF-driven phosphorylation of ERK also. Except for the non-significant increase in nuclear translocation of pERK with PlGF treatment, there is no indication that the PlGF interactions leading to activation of ERK reported herein occurred in the nucleus. In almost all cases, ERKs are activated by factors associated with the inner leaflet of the plasma membrane and cell surface receptors. Therefore, it can be concluded that PlGF stimulates motility by interaction with cell surface Flt-1, rather than nuclear Flt-1. The intracrine and cell surface functions of Flt-1, one conferring survival and the other motility through PlGF, are not mutually exclusive, and both promote metastasis.
In summary, this study describes for the first time, the effect of the growth factor, PlGF, on the activation of key kinases, which participate in increased cellular motility, a behaviour associated with metastasis rather than tumour growth and tumour cell proliferation. In other human cancers, ectopically expressed growth factors may also drive metastasis by mechanisms similar to those shown here for PlGF. For the breast cancer cell lines studied, the specificity of the kinase activation and increased migration was verified with antibody against both PlGF and Flt-1. The peptide, BP-1, abrogates PlGF-driven cellular motility and several of the PlGF-mediated changes in intracellular signalling, and this may, in part, explain its anti-tumour activity.