Cell migration and controlled assembly and disassembly of focal adhesions are highly integrated multistep processes and a central feature in the molecular pathology of cancer (Ridley et al, 2003
). To date, more than 50 different adhesion proteins that regulate the rate and organisation of actin polymerisation and focal adhesion turnover in protrusion have been identified.
In earlier publications, overexpression of LASP-1 mRNA in metastatic lymph nodes derived from breast cancer patients, as well as the co-amplification of the gene together with HER-2/neu (c-erbB2) were demonstrated (Chew et al, 1998
; Legge et al, 2005
). Two additional observations underscore the importance of LASP-1 in cancer. First, altered expression of LASP-1 is associated with the MLL gene in acute myeloid leukaemia (Strehl et al, 2003
). Second, recent studies have shown LASP-1 to be transcriptionally upregulated in response to the morphogen Sonic Hedgehog (Ingram et al, 2002
Consistent with these data, we just recently described the overexpression of LASP-1 to very high levels in breast carcinomas and lymph node metastases (Grunewald et al, 2006
). The functional significance of LASP-1 in cancer metastasis is further supported by the presented data showing high LASP-1 expression in ovarian cancer tissue and reduced cell migration in ovarian cancer cells depleted of LASP-1. The absence of LASP-1 in cultures of primary ovarian cancer cells in contrast to established cell lines may reflect a downregulation of LASP-1 in the nonmigratory floating ascites cells which will be reverted after several passages of adherent cell culture. Comparable observations are published for LASP-1 in human mesenchymal stem cells showing an upregulation of the protein during later passages (Sun et al, 2006
During LASP-1 silencing we observed reduced cell cycle progression and an induced G2
/M phase accumulation of the cells without disrupted normal mitotic microtubule polymerisation. This was accompanied by the upregulation and downregulation of several proteins. The differentially expressed proteins pyruvate kinase, enolase-1 and glucose dehydrogenase are part of the glycolytic metabolism and their regulation correlates well with the cell cycle arrest in G2
/M after LASP-1 silencing. Furthermore, pyruvate kinase and glucose dehydrogenase have been suspected to be highly important for tumour cell metabolism (Altenberg and Greulich, 2004
). Pyruvat kinase is one of the proteins to be upregulated in cancer cells gaining energy by means of aerobic glycolysis, which is a characteristic of a number of cancer entities (Gatenby and Gillies, 2004
). In addition, pyruvat kinase has been identified as a proteomic marker of cancer progression in breast cancer (Isidoro et al, 2005
). The glycolytic enzyme enolase-1 as well as HSP27, two additional proteins identified in the 2D-gel experiments, are associated with high metastatic activity in breast cancer cells (Espana et al, 2005
; Zhang et al, 2005
14-3-3, found to be upregulated after LASP-1 depletion in ovarian cancer cells, has been implicated in cell cycle deregulation. The 14-3-3 proteins are a family of highly conserved DNA-binding proteins, which associate with the centrosomes during mitosis and are inhibitors of G2
/M progression at the mitotic and G2
cell cycle checkpoint (Pietromonaco et al, 1996
; Wang and Shakes, 1996
; Hermeking et al, 1997
; Peng et al, 1997
; Alvarez et al, 2002
). Overexpression of 14-3-3 led to cell cycle arrest in cell culture models (Tzivion et al, 2006
) and, therefore, might contribute to the observed G2
arrest in ovarian cancer cells lacking LASP-1.
Heat-shock proteins are molecular chaperons and are induced during cellular stress. Upregulation of HSP27 after LASP-1 silencing correlates well with increased survival by inhibiting key effectors of the apoptotic pathway (Concannon et al, 2003
So far, the identified proteins are regulated in response to cell cycle arrest, but do not substitute for LASP-1 after silencing.
Recently, several LASP-1-binding partners have been identified. Along with zyxin (Li et al, 2004
) and actin (Schreiber et al, 1998
), LASP-1 interacts with Krp1 (Spence et al, 2006
), palladin (Rachlin and Otey 2006
), lipoma-preferred partner (LPP) and VASP (Keicher et al, 2004
), which all can influence actin filament dynamics and pseudopodial elongation. In the case of palladin, LPP and zyxin, the binding occurs between the C-terminal SH3 domain of LASP-1 and the N-terminal proline-rich domains of these proteins, whereas in the case of Krp1, binding is observed between the nebulin-like repeats of LASP-1 and the N-terminal BTB/POZ domain of Krp1. The interaction of LASP-1 and Krp1 is crucial for pseudopodial elongation in fibroblasts in absence of fibronectin and results in their colocalisation with F-actin at the tips of extending pseudopodia (Spence et al, 2006
Zyxin is localised primarily at focal adhesion plaques and plays a central role in actin filament polymerisation in mammalian cells (Beckerle, 1997
Silencing of zyxin in HeLa cells resulted in significantly reduced actin stress fibres (Griffith et al, 2005
), whereas under cyclic stretch zyxin only dissociated from focal contacts and accumulated in the nucleus, without affecting vinculin or actin filaments (Cattaruzza et al, 2004
). Recent data show that in genetically zyxin-deficient fibroblasts, focal adhesions are depleted from Mena and VASP, and that cells lacking zyxin display deficits in actin cytoskeleton remodelling (Hoffman et al, 2006
). In our immunflourescence experiments with LASP-1-deficient SKOV-3 cells, we observed a diffuse cytoplasmic localiation of zyxin without protein loss and without changes in neither vinculin distribution nor actin stress fibre organisation, emphasising the importance of LASP-1 for binding and recruiting zyxin to focal adhesions.
The loss of zyxin at the sites of focal contacts without changing cellular zyxin protein levels is not restricted to cancer cells, but was also observed in human umbilical vein endothelial cells (Grunewald et al, 2006
). Interestingly, in these cells, zyxin could still be detected along the actin stress fibres, indicating the potential existence of another zyxin-recruiting protein along actin stress fibres since earlier results detected LASP-1 only in the focal adhesion plaques (Chew et al, 2002
; Butt et al, 2003
In our zyxin knock-down experiments, neither changes in LASP-1 localisation, actin cytoskeleton, microtubule polymerisation nor vinculin distribution were detectable suggesting that zyxin does not change focal adhesion morphology. This is concordant with the fact that genetically zyxin-deficient fibroblasts show even enhanced adhesion to surface and increased integrin expression (Hoffman et al, 2006
). In synopsis, our LASP-1 and zyxin silencing studies have demonstrated that LASP-1 is necessary for recruiting zyxin to focal contacts.
The decreased cell motility after LASP-1 silencing can be explained by the functional loss of zyxin as a scaffolding protein that facilitates the formation of molecular complexes to promote site-specific actin assembly required for cell migration. This is in agreement with previous findings using a nongenetic approach and injecting a peptide derived from the N-terminus of zyxin to displace zyxin from its normal subcellular location thus leading to reduced cell migration (Drees et al, 1999
). On the other hand, the knock-down of zyxin in SKOV-3 cells had no influence on cell migration while genetically zyxin-deficient fibroblasts display enhanced migration (Hoffman et al, 2006
). These contrary effects have not been fully elucidated yet.
Recent work has shown that zyxin also shuttles through the nucleus – most likely by association with other LIM proteins – and may regulate gene transcription (Nix et al, 2001
; Wang and Gilmore, 2003
; Kadrmas and Beckerle, 2004
). During mitosis, a fraction of zyxin becomes associated with the tumour suppressor h-warts at the mitotic apparatus (Hirota et al, 2000
). h-warts is a key player in mitosis in mammalian cells and loss of its function disrupts normal cell cycle regulation, possibly leading to tumour development (Iida et al, 2004
). In SKOV-3 cells transfected with LASP-1 siRNA, zyxin has been shown to dissociate from focal adhesion plaques and to distribute diffusely into the cytoplasm. It is, therefore, likely that part of zyxin enters the nucleus, binds to h-warts and leads to G2
cell cycle arrest and inhibition of proliferation as observed after LASP-1 silencing.
Interestingly, in Ewing tumour cells, zyxin is only expressed at very low levels and remains diffusely distributed throughout the cytoplasm instead of concentrating in actin-rich dynamic structures. Zyxin gene transfer into EWS-FLI1-transformed fibroblasts elicits reconstitution of zyxin-rich focal adhesions and leads to decreased cell motility and inhibition of anchorage independent tumour growth, indicating that zyxin has tumour suppressor activity in these cells (Amsellem et al, 2005
Similar to findings in human breast cancer (Grunewald et al, 2006
), our immunofluorescence experiments have shown that absence of LASP-1 in focal contacts dramatically influences zyxin distribution. In reverse, tumour cells, that are overexpressing LASP-1, could functionally inhibit zyxin from shuttling into the nucleus and acting as a tumour suppressor through increased recruiting of zyxin to focal contacts by LASP-1. In summary, our observations suggest an expanded role for LASP-1 in proliferation and cancer cell migration. Further studies will define the potential of LASP-1 as an independent marker for diagnosis of cancer, as well as a marker for prognosis of this disease.