Combining multiple cell lines and migration assays, this study is the first to show inter-dependent polarization of caveolin-1 and PTRF/cavin-1 in migrating cells, and to demonstrate that PTRF/cavin-1 expression regulates cell migration. It further unveils that caveolin-1 plays a different role in cell migration depending on the presence or absence of PTRF/cavin-1 (and thus of caveolae).
Caveolin-1 was previously known to regulate cell migration and polarization. We now show that caveolin-1 polarization is accompanied by PTRF/cavin-1 polarization. In two dimensional migrating cells, both proteins accumulate in the cell rear, where caveolae are also known to accumulate 
. Furthermore, caveolae seem to be required for the polarization of both PTRF/cavin-1 and caveolin-1, since conditions where the cells lack caveolae result in polarization loss. In 3D migrating cells, we have been able to visualize non caveolar caveolin-1 (throughout cytoplasmic protrusions of trans-migrating cells) but also non caveolar PTRF/cavin-1 (as a ring where the protrusion contacts the filter pore). The mechanism by which PTRF/cavin-1 can localize to non-caveolar membrane is currently unknown. PTRF/cavin-1 binds phosphatidylserine in vitro
, thus one possibility for caveola-independent PTRF/cavin-1 membrane recruitment is an enrichment, or distinct organization, of phosphatidylserine at the plasma membrane-filter (and possibly extracellular matrix) points. Polarization of PTRF/cavin-1, caveolin and caveolae in transmigrating cells is summarized in .
Summarizing what we currently know about the effect of PTRF/cavin-1 expression on cell migration, a mechanism emerges through which by allowing caveola formation and recruitment of caveolar proteins to plasma membrane microdomains, PTRF/cavin-1 allows caveola recruitment of PKCα, modulates cell polarization, decreases MMP-9 production and ultimately cell migration. Multiple pathways are likely to be involved in mediating the effect of PTRF/cavin-1 on cell migration. It is interesting to note that like the phenotype of caveolin-1-null mice, the phenotype of PTRF/cavin-1-null mice does not seem to encompass major defects in cell migration, and regulation must therefore be either compensated for, subtle, or apparent only upon challenge. PKCα targeting to caveolae was previously reported to be mediated through interaction with SDPR/cavin-2 
. However, PC3 cells do not express SDPR/cavin-2, or PRKCDBP/cavin-3 
, both of which are known to bind to PKC isoforms. Hence PTRF/cavin-1 induced DRM-recruitment of PKCα likely occurs via a novel mechanism.
Studies on caveolin-1 and caveolae function commonly employ caveolin-1 over-expression or knockdown approaches. The present data suggest that interpretation of such experiments should to be revisited with knowledge of the PTRF/cavin-1 expression status. Indeed, expression and subcellular localization of these caveolar adapter proteins, are co-regulated with caveolin-1. Furthermore it seems that both caveolin-1 and PTRF/cavin-1 can function from non-caveola locations, as reported here. As illustrated in , the loss of caveolin-1, (presumably via loss of caveolae) releases PTRF/cavin-1 which can function at extra-caveolar locations. Conversely, ectopic expression of caveolin-1 in PTRF (and caveola)-null cells may lead to extra-caveolar caveolin-1 functions. Over-expression of caveolin-1 in PTRF/cavin-1-positive cells would not only increase caveola density, but could also deplete extra-caveolar cavins, and/or increase extra-caveolar caveolin-1 function. Notably, cavins show tissue-specific expression patterns that partially parallel caveolins 
. Furthermore, caveolin-1 expression in some caveolin-1-negative cell lines does not induce PTRF/cavin-1 expression 
, suggesting that both caveolin-1 and PTRF/cavin-1 expression are limiting for caveola formation in a cell- and tissue- specific manner. Clearly, these previously unrecognized parameters confound the interpretation of experiments. Thus distinct extra-caveolar functions of caveolin-1 and PTRF/cavin-1 reveal a new paradigm in caveolin biology and calls for re-interpretation of the numerous studies on caveolin-1 and caveola function.
The manipulation of PTRF/cavin-1 and caveolin-1 levels in different cell lines allowed us to identify distinct effects of these proteins on cell migration and polarization. Heterologous expression of PTRF in caveolin-1-positive, PTRF/cavin-1-negative PC3 cells results in caveola formation, and reduced cell migration. Down-regulating caveolin-1 in (PTRF/cavin-1-negative) PC3 cells to reduce non-caveolar caveolin led to an increase in cell migration. In contrast, knockdown of caveolin-1 in PTRF/cavin-1-expressing PC3 cells reduced cell migration. These results, summarized in , indicate that caveolin-1 can regulate cell migration without being part of caveolae and functionally build on previous data obtained via imaging 
. Future studies will establish whether other functions previously ascribed to caveolae 
may be due to non-caveolar caveolin. Overexpression of caveolin-1 in a PTRF/cavin-1 null breast cancer cell line, SK-BR-3 was recently reported to elicit formation of long tubules 
that may mediate signaling events different from caveolae.
Distinct non-caveolar functions of caveolin-1 will also have implications for our understanding of cancer development and progression since caveolin-1 mutation or overexpression is suggested to be involved in several cancers 
. It is now clear that not only the levels of caveolin-1, but also the relative levels of PTRF/cavin-1, must be taken into account when interpreting the role of caveolin-1 in specific disease contexts.