Biology Reference
In-Depth Information
possible that other molecules may serve this function, it is clear that a5 is the
major integrin in CHO cells and is required for adhesion and migration on
fibronectin. In addition, antibody staining against various integrins and other
putative adhesion molecules did not show readily detectable amounts of these
proteins associated with the nascent adhesions. Recent data using time-lapse
confocal microscopy suggests that the integrin serves as the nucleation site for
the new adhesions (Claire Brown and Donna Webb, unpublished observa-
tion). We detect diffuse a5 integrin near the leading edge of the cell prior to the
recruitment of paxillin, which overlays the integrin. Thus, the a5 integrin-
containing nucleation sites are not organized as large visible complexes, but
present in a concentration sucient to stimulate the recruitment of other
molecules, such as paxillin.
At this stage our studies support a regulated, sequential model for adhesion
formation. We have clear evidence that some components enter the adhesions
serially and with quite distinct kinetics, e.g., a-actinin and paxillin. The distinct
kinetics suggest a role for key, uncharacterized regulatory events that mediate
the entry of some components. It is also likely that some components enter
adhesions simultaneously, either individually or as preformed complexes. For
example, GIT1 appears to target to adhesions in a complex with signalling
components including PIX and PAK (Manabe et al., 2002). Higher resolution
kinetics of other adhesion components, such as paxillin and zyxin, will help
determine whether their entry is also coordinated. The next step in our studies
of adhesion assembly is to use the ratio imaging approach with mutant
molecules to determine what targets various molecules into nascent adhesions.
Adhesion turnover
In both fibroblasts and CHO cells, the paxillin clusters at the base of a
protrusion disappear while adhesions form near the new leading edge of the
cell (Figure 19.1). This turnover appears to differ from the breakdown of
adhesions at the cell rear (see next section) and thus, to distinguish them, we
refer to the former as turnover and the latter as disassembly. Since adhesion
turnover has not been previously described, we examined it in more detail.
Figure 19.1 (opposite) Paxillin adhesions localize near the leading edge of a protrusion
and turn over as new adhesions form. WI38 cells (human fibroblasts) expressing paxillin-
GFP were plated on 1 mg/ml fibronectin for 1 h at 378C and adhesion dynamics were
visualized by time-lapse fluorescent microscopy. The paxillin clusters at the base of the
protrusion disappear while adhesions form near the new leading edge of the cell. We refer
to this as adhesion turnover. The arrow indicates a group of paxillin adhesions that turn
over as the cell protrudes (compare t
¼
0 and t
¼
20 min). Scale bar
¼
20 mm
