P-Rex1 is an exchange factor for the RhoGTPase molecule Rac1, which has been implicated in progression to metastasis in a number of cancer models 
. Through ectopic expression of P-Rex1 in cell culture models, we have identified a close physical and functional relationship between P-Rex1 and the receptor tyrosine kinase PDGFRβ during the acquisition of invasive cellular migration in vitro
P-Rex1 is a well-established upstream regulator of RhoGTPase signalling 
. In our hands, as in prior reports, ectopic overexpression of the RacGEF P-Rex1 or the related molecule, Tiam1 results in morphological changes in immortalised human fibroblasts 
. These phenotypic changes are characteristic of increased Rac1 activity 
, and suggest that expression of P-Rex1 or Tiam1 influences the balance of RhoGTPase signalling. Importantly, these alterations to cell shape could not be driven by expression of a catalytically inactive mutant P-Rex1 molecule (E56A, N238A), suggesting dependence upon inherent GEF activity. The morphological changes reported in this study are truly robust, as more than 7,500 cells per condition were assayed, utilising a high throughput imaging system and two independent morphological criteria. Thus, we can say objectively and with confidence that active P-Rex1 and Tiam1 drive morphological adaptation in human fibroblasts. Importantly, even when subjected to such detailed quantitative analysis, it is clear that both P-Rex1 and Tiam1 induce alterations to cell shape and ruffling that are indistinguishable from one an another, and entirely consistent with the ability of these two molecules to activate Rac.
When one analyses cell migration, the phenotypes induces by P-Rex1 and Tiam1 are quite distinct. In scratch-wound assays, P-Rex1 drove a significant, GEF-dependent increase in mean velocity, persistence and forward migratory index of cells at the migratory front of the wound 
. However, no change in the migratory behaviour of Tiam1 expressing cells was observed, despite clear alterations in the morphology of sparsely cultured non-migrating cells. This implies that a generalised increase in RacGEF activity is not sufficient in itself to influence migratory behaviour and that P-Rex1 has a particular role in this regard. Consistent with previously described Tiam1 driven formation of Rac1 dependent cell-cell contacts 
, we observe loss of refractile morphology of Tiam1 expressing fibroblasts under brightfield microscopy at high cell densities which hints at altered cell to cell adhesion (). Taken together, these studies indicate that different functions can be ascribed to the two Rac GEFs of the present study - with Tiam1 primarily influencing cell:cell adhesion, whilst P-Rex1 is a powerful promoter of cellular migration.
The earliest characterisation of P-Rex1 described a synergistic regulation through association with the phospholipid PI(3,4,5)P3
and βγ subunits of the heterotrimeric G-proteins 
, indicating that it is potentially capable of responding to signalling inputs generated downstream of both G-protein coupled receptors and RTKs. Moreover, these studies indicate that P-Rex1 may be recruited to the plasma membrane by association with the PI(3,4,5)P3
produced by growth factor receptor signalling. This said, our observation of a physical association between P-Rex1 and an RTK (PDGFRβ) occurs in serum-starved cells indicates that regulated production of PI(3,4,5)P3
is not a prerequisite for P-Rex1 recruitment cellular membrane compartments. Additionally, P-Rex1-GD co-immunoprecipitates efficiently with PDGFRβ indicating that Rac activation and the cytoskeletal changes occurring downstream of this are not required for recruitment of P-Rex1 to PDGFRβ.
In addition to the physical association between P-Rex1 and PDGFRβ, there is also a close functional relationship that is not shared by Tiam1. Here we identify PDGF to be a serum component that is both necessary and sufficient for P-Rex1 driven migration and chemotaxis in a 3D microenvironment. Furthermore, we have used siRNA to show that this is mediated primarily via PDGFRβ and does not require PDGFRα. Furthermore, in fibroblasts from mice in which the genes for PDGFRα and β have been disrupted, expression of PDGFRβ is sufficient to restore P-Rex1-driven migration towards a serum gradient. Taken together, these results indicate that PDGFRβ, and not PDGFRα is a major contributor to P-Rex1 “activation” in regard of invasive cell migration in the context of our culture model. Subsequent chemical inhibition of the PDGFR family in P-Rex1 expressing fibroblasts also demonstrates that an active signalling axis exists involving PDGFR and P-Rex1, whereby specific loss of PDGFR activity while maintaining protein expression, negatively affects P-Rex1-dependent migratory behaviour both on a 2D substrate and in 3D. Furthermore, similar observations made following inhibition of PI3K serve to highlight the importance of phosphatidyl inositide signalling in regulation of P-Rex1 activity.
There is a strong correlation between P-Rex1 expression and progression to metastasis in a number of cancer models
. The observation therefore, that expression of this single RacGEF was sufficient to drive a dramatic increase in invasive migration in 3D matrices in vitro
was particularly striking. Furthermore, evidence exists to suggest that a functional relationship between P-Rex1 and PDGFRβ as described here may be a contributing factor to cancer progression in vivo
. The first suggestion of an association arose in a genetic screen to identify key factors in glioma progression in vivo
, which highlighted PREX1 as a gene which cooperates with PDGF signalling to promote metastasis 
. Moreover, in light of our recent description of P-Rex1 as a key driver of melanoma progression in vivo
, it is noteworthy that a common mechanism of acquired resistance to novel therapeutic kinase inhibitors in BRAFV600E
driven melanoma is upregulation of PDGFRβ 
. Indeed, we present data which demonstrates that in vitro
invasive migration of a human melanoma derived cell line, WM852, is dependent upon expression of both P-Rex1 and PDGFRβ.
In addition to their contribution to malignant growth in certain cancer cell types, the PDGF receptors and autocrine PDGF signalling are critical drivers of recruitment and growth of non-transformed cells, such as pericytes, vascular smooth muscle cells and stromal fibroblasts to the tumour site. These can in turn contribute both to tumour growth and vascularity. It is therefore important to note that through the in vitro
experimentation detailed here, we examine tumour cell autonomous effects of PDGFR stimulation, in particular in the context of P-Rex1 mediated cellular migration, and do not consider the role of the PDGF signalling axis on survival or growth of stromal cells in an extant tumour in vivo
. By way of example, while silencing of PDGFRβ negatively affects in vitro
migratory behaviour in individual cell lines, a number of publications have indicated that mono- or combination therapies incorporating inhibition of PDGFRs in vivo
can result in a more aggressive invasive or metastatic disease 
. This appears at least in part due to a stromal response, where negative effects upon both pericyte growth and migration, can lead to poor pericyte coverage on nascent blood vessels, resulting in a “leaky” tumour vasculature, more susceptible to intravasation 
Moreover, such therapies may have a similar systemic effect, resulting in a more conducive vascular environment for extravasation and subsequent metastasis at a distant site 
It seems that the relationship between P-Rex1 and PDGFRβ may have direct relevance to tumour progression, and further studies to reveal the details of the complex formed between these two important signalling proteins and how they collaborate to direct cell migration will be necessary to determine whether components of this pathway can be targetted to oppose cancer progression and metastasis.