FERM domain proteins, including talins, ERMs, FAK and certain myosins, regulate connections between the plasma membrane, cytoskeleton and extracellular matrix. Here we show that FrmA, a Dictyostelium discoideum protein containing two talin-like FERM domains, plays a major role in normal cell shape, cell-substrate adhesion and actin cytoskeleton organisation. Using total internal reflection fluorescence (TIRF) microscopy we show that FrmA-null cells are more adherent to substrate than wild-type cells because of an increased number, persistence and mislocalisation of paxillin-rich cell-substrate adhesions, which is associated with decreased motility. We show for the first time that talinA colocalises with paxillin at the distal ends of filopodia to form cell-substrate adhesions and indeed arrives prior to paxillin. After a period of colocalisation, talin leaves the adhesion site followed by paxillin. Whereas talinA-rich spots turnover prior to the arrival of the main body of the cell, paxillin-rich spots turn over as the main body of the cell passes over it. In FrmA-null cells talinA initially localises to cell-substrate adhesion sites at the distal ends of filopodia but paxillin is instead localised to stabilised adhesion sites at the periphery of the main cell body. This suggests a model for cell-substrate adhesion in Dictyostelium whereby the talin-like FERM domains of FrmA regulate the temporal and spatial control of talinA and paxillin at cell-substrate adhesion sites, which in turn controls adhesion and motility.

In mammalian cells, the four-point-one, ezrin, radixin and moesin (FERM) domain proteins have been heavily implicated in the regulation of cell-substrate adhesion and in particular the formation and turnover of focal adhesions (FAs) during cell migration (Lo, 2006). FERM domain proteins that have been shown to directly associate with FAs include talin, FAK and kindlin. Whereas talin has been shown to be a key player in the formation of FAs, FAK regulates their turnover (Mitra et al., 2005; Nayal et al., 2004). Although the role of kindlin is unclear, it has been shown to associate with FAs and is required for normal integrin-mediated cell spreading (Kloeker et al., 2004). Thus, FERM domain proteins play a critical role in regulating FAs and cell-substrate adhesion, which in turn regulates cell migration.

In mammalian cells, the study of FERM domain proteins in cell-substrate adhesion, in the context of migration, is complicated by both redundancy and signalling complexity. In humans, there are two isoforms of both talin and FAK, and three isoforms of kindlin. In addition, there are approximately 50 other FERM-domain-containing proteins present in the sequencing database and their regulation is likely to be complex and they may also have overlapping functions (Diakowski et al., 2006). To overcome these complexities, the model organism Dictyostelium discoideum was chosen to study the role of FERM-domain proteins during cell-substrate adhesion and migration.

There are only six FERM domain proteins in Dictyostelium [FrmA, FrmB, FrmC (enlazin), talinA, talinB and myosinVII (www.dictybase.org)] (supplementary material Fig. S1A) of which, knockout studies have been carried out for FrmC, talinA, talinB and myosinVII (Niewohner et al., 1997; Octtaviani et al., 2006; Tsujioka et al., 1999; Tuxworth et al., 2001). Loss of talinA greatly reduces cell-substrate adhesion and this cannot be compensated for by the expression of talinB, suggesting a unique role for talinA in cell-substrate adhesion (Niewohner et al., 1997). TalinA has also been shown to localise to the leading edge of cells undergoing directed cell migration and to discrete spots at the cell-substrate boundary (Hibi et al., 2004; Kreitmeier et al., 1995). Taken together with the requirement of talinB for migration within a multicellular environment (Tsujioka et al., 2004), it suggests that Dictyostelium talins play a role in adhesion at the leading edge of cells undergoing cell migration. MyosinVII and FrmC have also been shown to play a role in the regulation of cell-substrate adhesion as loss of either, results in reduced cell-substrate adhesion (Octtaviani et al., 2006; Tuxworth et al., 2001). Indeed, talinA and myosinVII have been shown to bind to each other (Tuxworth et al., 2005). Non-FERM domain proteins such as the adaptor molecule paxillin, and the adhesion receptor SadA, also regulate adhesion, because loss of either results in reduced cell-substrate adhesion (Bukharova et al., 2005; Fey et al., 2002).

Of the remaining FERM-domain proteins awaiting study, FrmA seemed of particular interest owing to the similarity of its FERM domains to the FERM domain of talin (supplementary material Fig. S1B). Below, we present evidence demonstrating that FrmA, a multi-talin-like FERM-domain protein, is required for correct cell-substrate adhesion. More precisely, FrmA promotes the turnover of paxillin-enriched cell-substrate adhesions during both random and directed cell migration.

FrmA regulates cell shape and cell-substrate adhesion

Homologous recombination was used to disrupt FrmA in wild-type Ax2 (wt) cells to produce FrmA null (frmA) cells. PCR amplification of genomic DNA from single-cell-derived clones was then used to identify frmA clones (Fig. 1A). PCR of genomic DNA from wt cells resulted in a 2.0 kbp product, whereas PCR of genomic DNA from disrupted clones resulted in a 2.4 kbp product (Fig. 1A). Real-time RT-PCR was used to confirm the absence of the frmA mRNA from total RNA isolated from wt and frmA clones at 0 and 6 hours of starvation (Fig. 1B). Expression in wt and frmA cells of EF1α, whose expression remains constant following starvation, and talinB, whose expression is upregulated following starvation, were similar (Fig. 1B) (http://dictybase.org/db/cgi-bin/gene_page.pl?feature_id=159674) (Tsujioka et al., 1999). Interestingly, expression of FrmA was significantly upregulated after 6 hours of starvation (Fig. 1B).

Non-starved frmA cells were more rounded and adhered more strongly to substrate than wt cells. To analyse cell shape, images of non-starved cells were collected on three separate days (Fig. 1C). Although frmA cells were able to project filopodia (yellow arrows), they rarely produced broad membrane protrusions (white arrows). The average circularity index of wt cells was 0.56±0.06 whereas that of frmA cells was 0.82±0.02 (supplementary material Fig. S2A), with a circularity index of 1 representing a perfect circle. Thus, frmA cells were significantly more circular or rounded than wt cells. Reconstitution of full-length FrmA (HA tagged) in frmA cells (frmA/FrmAHA), restored the circularity index (0.62±0.01) of the mutants.

To determine the effect of FrmA loss on cell adhesion, adhesion assays were carried out on wt and frmA cells (Fig. 1D). Constant rotation at 200 r.p.m. for 1 hour led to 23.33±3.47% of wt cells detaching, while only 8.5±2.84% of frmA cells detached under the same conditions. This was completely rescued by re-expression of FrmAHA in the mutant cells (Fig. 1D). Furthermore, overexpression of the first FERM domain of FrmA, FERM(1), fused to GFP in wt cells, led to a loss of cell-substrate adhesion (Fig. 1E). Whereas 29.66±6.93% of wt cells expressing GFP detached after constant rotation at 200 r.p.m. for 1 hour, 89.65±4.57% and 54.85±0.8% of wt cells expressing GFP fused to either FERM(1) or FERM(2) detached, respectively. Expression of either FERM(1) or FERM(2) in frmA cells had no effect on adhesion of frmA cells, suggesting that the overexpression of FERM(1) acts synergistically with FrmA in wt cells, resulting in impaired adhesion.

FrmA regulates cell-substrate adhesion sites

Cell-substrate adhesion sites have been identified in Dictyostelium as paxillin-rich spots on the cell-substrate boundary (Bukharova et al., 2005). In mammalian cells, adhesion sites are also regulated by talins, which are thought to initiate their formation (Nayal et al., 2004). Dictyostelium cells contain two talins (talinA and talinB) and although their functions overlap, only the loss of talinA results in a major cell-substrate adhesion defect (Niewohner et al., 1997; Tsujioka et al., 1999; Tsujioka et al., 2004). Interestingly, like paxillin, talinA localises to discrete spots at cell-substrate boundaries, which, also initially form at the distal ends of filopodia (Hibi et al., 2004). To visualise these spots, paxillin fused to GFP and talinA fused to GFP were stably expressed in wt and frmA cells. Total internal reflection fluorescence (TIRF) microscopy, which allows visualisation of structures in close proximity (∼150 nm) to the surface of the cell in contact with the substrate (Gerisch et al., 2004), was used to determine their localisation (Fig. 2A, supplementary material Movies 1-4). In agreement with previous reports, both paxillin- and talinA-rich spots initially formed at the distal ends of filopodia and did not form under the cell-substrate contact area of the main body of the cell. Furthermore, filopodia predominantly formed in the direction of cell movement of non-starved wt cells (Fig. 2A) (Bukharova et al., 2005; Hibi et al., 2004). However, whereas talinA spots dissipated upon arrival of the main body of the cell (supplementary material Movie 3), paxillin spots dissipated as the main body of the cell travelled over them (supplementary material Movie 1), indicating that talinA and paxillin play distinct roles in the dynamic regulation of cell-substrate adhesions.

Fig. 1.

Loss of FrmA impairs both cell shape and cell-substrate adhesion. (A) Single-cell-derived frmA clones were identified by PCR of genomic DNA. PCR of wild-type genomic DNA yielded a 2 kbp product (*), whereas PCR of frmA genomic DNA yielded a 2.4 kbp product (**). A primer set that amplified a 1 kbp fragment of eF1α was also included in the PCR reaction for FrmA as a control to show the specificity of FrmA disruption (+). At least three single-cell-derived frmA clones were isolated and the analysis of a representative frmA clone is shown. (B) Real-time RT-PCR was used to determine the expression of FrmA, talinB (example of a protein that is upregulated upon starvation) and eF1α (example of a protein that is not regulated by starvation) at 0 and 6 hours of starvation. Total RNA was isolated on two separate occasions and real-time RT-PCR reactions were carried out in triplicate. The average ± s.d. is shown. (C) Phase images of non-starved cells (40× objective used for all). Broad membrane protrusions (white arrows) and filopodia (yellow arrows) are indicated. (D) Cell adhesion to substrate was determined under conditions of increased shear stress. Experiments were carried out on at least four separate occasions in triplicate and the average ± s.e.m. is shown. (E) Adhesion of wild-type and frmA cells expressing GFP, GFP:FERM(1) or GFP:FERM(2).

Fig. 1.

Loss of FrmA impairs both cell shape and cell-substrate adhesion. (A) Single-cell-derived frmA clones were identified by PCR of genomic DNA. PCR of wild-type genomic DNA yielded a 2 kbp product (*), whereas PCR of frmA genomic DNA yielded a 2.4 kbp product (**). A primer set that amplified a 1 kbp fragment of eF1α was also included in the PCR reaction for FrmA as a control to show the specificity of FrmA disruption (+). At least three single-cell-derived frmA clones were isolated and the analysis of a representative frmA clone is shown. (B) Real-time RT-PCR was used to determine the expression of FrmA, talinB (example of a protein that is upregulated upon starvation) and eF1α (example of a protein that is not regulated by starvation) at 0 and 6 hours of starvation. Total RNA was isolated on two separate occasions and real-time RT-PCR reactions were carried out in triplicate. The average ± s.d. is shown. (C) Phase images of non-starved cells (40× objective used for all). Broad membrane protrusions (white arrows) and filopodia (yellow arrows) are indicated. (D) Cell adhesion to substrate was determined under conditions of increased shear stress. Experiments were carried out on at least four separate occasions in triplicate and the average ± s.e.m. is shown. (E) Adhesion of wild-type and frmA cells expressing GFP, GFP:FERM(1) or GFP:FERM(2).

In frmA cells, paxillin spots were not seen to form at the distal ends of filopodia but instead formed uniformly around the circumference of the cell-substrate contact area of the main body of the cell and appeared to stabilise that area (supplementary material Movie 2). The number of adhesions containing paxillin spots (Fig. 2B) and their duration (Fig. 2C), were increased in frmA cells compared with wt cells. Whereas wt cells contained up to six paxillin-rich spots on average, frmA cells contained more than 25. Analysis of the persistence of paxillin spots showed that in wt cells, the spots lasted for 1 to 2 minutes, whereas in frmA cells, the spots lasted for more than 5 minutes, with many lasting for more than 10 minutes. The rate of appearance of paxillin spots in frmA cells, although confined to the cell-substrate contact area of the main body of the cell (2.23±0.41/minute/cell), was similar to the rate of appearance in wt cells (2.88±0.59/minute/cell), suggesting that this is unaffected and thus does not involve FrmA.

Fig. 2.

Regulation of paxillin and talinA adhesion sites is impaired in frmA cells. (A) TIRF images showing GFP fused paxillin (top panel) and talinA (bottom panel) localisation in wild-type (left panels) and frmA cells (right panels). (B) The number of paxillin and talinA spots observed per cell. TIRF images of wild-type and frmA cells expressing paxillin or talinA fused to GFP were captured 100 seconds apart over 600 seconds and the average number of paxillin- and talinA-rich spots determined. 10 or more cells were analysed for each strain in total over three separate occasions and the average ± s.e.m. is shown. (C) Using TIRF microscopy, the duration of paxillin-rich spots was followed by measuring the fluorescence intensity (Image J software) of an area where a spot would form. The fluorescence intensity values were plotted against time for spots in frmA cells (various coloured lines) and a typical wild-type cell (black line). More than 10 cells from each strain were analysed in total over three separate occasions. (D) TIRF images of a wild-type cell expressing paxillin fused to GFP (green) and talinA fused to RFP (red), with the merged image on the right. (E) Graphical representation of the appearance and disappearance of a paxillin (green) and talinA (red) spot over time, observed using TIRF and quantified using Image J software. (F) Sequential and merged TIRF images of the appearance of talinA (red) followed by paxillin (green) at an adhesion site. White arrow highlights the spot in question.

Fig. 2.

Regulation of paxillin and talinA adhesion sites is impaired in frmA cells. (A) TIRF images showing GFP fused paxillin (top panel) and talinA (bottom panel) localisation in wild-type (left panels) and frmA cells (right panels). (B) The number of paxillin and talinA spots observed per cell. TIRF images of wild-type and frmA cells expressing paxillin or talinA fused to GFP were captured 100 seconds apart over 600 seconds and the average number of paxillin- and talinA-rich spots determined. 10 or more cells were analysed for each strain in total over three separate occasions and the average ± s.e.m. is shown. (C) Using TIRF microscopy, the duration of paxillin-rich spots was followed by measuring the fluorescence intensity (Image J software) of an area where a spot would form. The fluorescence intensity values were plotted against time for spots in frmA cells (various coloured lines) and a typical wild-type cell (black line). More than 10 cells from each strain were analysed in total over three separate occasions. (D) TIRF images of a wild-type cell expressing paxillin fused to GFP (green) and talinA fused to RFP (red), with the merged image on the right. (E) Graphical representation of the appearance and disappearance of a paxillin (green) and talinA (red) spot over time, observed using TIRF and quantified using Image J software. (F) Sequential and merged TIRF images of the appearance of talinA (red) followed by paxillin (green) at an adhesion site. White arrow highlights the spot in question.

In frmA cells, although talinA spots formed at distal ends of filopodia, they formed randomly around the circumference of the cell, instead of being orientated in the direction of cell movement, as in wt cells (supplementary material Movie 4). Although the average number of talinA spots per cell was fewer than that in wt cells (wt 5.29±0.89 spots/cell; frmA cells 2.23±0.65 spots/cell), their persistence was similar (Fig. 2B, supplementary material Fig. S2B). TalinA spots tended to last around 60 seconds in both wt and frmA cells.

To determine the spatiotemporal relationship between talinA and paxillin in Dictyostelium cells, talinA fused to RFP (talinA, red) and paxillin fused to GFP (paxillin, green) were coexpressed in wt cells and imaged using TIRF microscopy (Fig. 2D, supplementary material Movie 5). TalinA arrived first at a cell-substrate adhesion site and was then followed by paxillin, some 6.57±2.98 seconds (8 cells and 14 spots) later. These adhesion sites formed at the distal ends of filopodia, predominantly in the direction of cell movement. There was then a period of colocalisation, after which talinA left first and was then followed by paxillin. Plotting the fluorescence intensity over time, of the individual RFP and GFP signals over a particular region where spots appears, confirmed this (Fig. 2E). The sequential appearance of talinA and paxillin at a cell-substrate adhesion site, is shown in the series of images in Fig. 2F (white arrows indicate the spot in question).

F-actin regulation is impaired by the loss of FrmA, and F-actin and FrmA colocalise

Another key player in the regulation of adhesion is F-actin and therefore F-actin localisation was examined in the wt and frmA cells. Staining with TRITC-phalloidin (Fig. 3A) showed that F-actin was mainly localised in broad membrane protrusions as well as in the filopodia of wt cells. By contrast, frmA cells rarely had broad membrane protrusions and the F-actin was instead seen in discreet `patches' throughout the cell cortex. Cross-sectional analysis of the mutant cells showed that some of these actin-rich patches were at the basal surface of the cells where they were in contact with the substrate (Fig. 3A, right image). In Dictyostelium, F-actin-rich patches have been reported at cell-substrate boundaries and probably represent the polymerisation of localised actin driven by the Arp2/3 complex (Bretschneider et al., 2004). LimE is a LIM-domain-containing protein that associates with F-actin. A GFP fusion construct with a fragment of Dictyostelium LimE (LimEΔcoil:GFP) has been shown to mark out the network of F-actin within the cell, as well as short-lived patches at the cell-substrate boundary (Bretschneider et al., 2004). Use of TIRF microscopy showed that in wt cells LimEΔcoil:GFP was localised in discrete patches at the cell-surface boundary and also to broad membrane protrusions (Fig. 3B, supplementary material Movie 6). In frmA cells, LimEΔcoil:GFP was also localised in discrete patches at the cell-surface boundary, but unlike wt cells, it also localised uniformly around the circumference of the cell-substrate contact area (Fig. 3B, supplementary material Movie 7). The number, and persistence, of these patches was also affected by the loss of FrmA. On average, wt cells had three LimEΔcoil:GFP patches whereas frmA cells had seven. Analysis of the duration of LimEΔcoil:GFP patches showed that in wt cells the patches persisted for ∼30 seconds, whereas in frmA cells they persisted for ∼20 seconds (supplementary material Fig. S2C,D).

Fig. 3.

F-actin regulation is impaired by the loss of FrmA, and F-actin and FrmA colocalise. (A) Confocal images of wild-type (left) and frmA (middle) cells fixed and stained with TRITC-phalloidin (red) and DAPI (blue) to highlight the actin cytoskeleton and the nucleus, respectively. Cross-sectional images were captured and the maximum projections shown. F-actin-rich patches were located at the cortex of frmA cells and in particular at the cell-substrate boundary. The frmA image closest to the cell-substrate boundary, is shown on the right with white arrows highlighting patches and yellow lines indicating the cross section being shown above and beside the layer. (B) TIRF images showing LimEΔcoil:GFP localisation in wild-type (left panels) and frmA cells (right panels). (C) Confocal images of frmA/FrmAHA cells, fixed and stained with TRITC-phalloidin and an anti-HA antibody conjugated to FITC to highlight the actin cytoskeleton and localisation of FrmAHA, respectively. Images closest to the cell-substrate boundary are shown. Specific areas (rectangles 1, 2 and 3) were further magnified and shown immediately below with arrows highlighting F-actin patches and FrmAHA colocalisation.

Fig. 3.

F-actin regulation is impaired by the loss of FrmA, and F-actin and FrmA colocalise. (A) Confocal images of wild-type (left) and frmA (middle) cells fixed and stained with TRITC-phalloidin (red) and DAPI (blue) to highlight the actin cytoskeleton and the nucleus, respectively. Cross-sectional images were captured and the maximum projections shown. F-actin-rich patches were located at the cortex of frmA cells and in particular at the cell-substrate boundary. The frmA image closest to the cell-substrate boundary, is shown on the right with white arrows highlighting patches and yellow lines indicating the cross section being shown above and beside the layer. (B) TIRF images showing LimEΔcoil:GFP localisation in wild-type (left panels) and frmA cells (right panels). (C) Confocal images of frmA/FrmAHA cells, fixed and stained with TRITC-phalloidin and an anti-HA antibody conjugated to FITC to highlight the actin cytoskeleton and localisation of FrmAHA, respectively. Images closest to the cell-substrate boundary are shown. Specific areas (rectangles 1, 2 and 3) were further magnified and shown immediately below with arrows highlighting F-actin patches and FrmAHA colocalisation.

Loss of FrmA had different effects on the persistence of paxillin spots, talinA spots and F-actin patches. Whereas the persistence of paxillin spots was markedly prolonged, the duration of talinA spots and F-actin rich patches were not seriously affected. However, the localisation of all three was affected by the loss of FrmA. To further analyse these data, localisation of FrmA was determined by fixing and staining frmA/FrmAHA cells with anti-HA antibody conjugated with FITC and phalloidin conjugated to TRITC. Using confocal microscopy, FrmAHA was predominantly diffusely localised throughout cells, although at the cell-substrate boundary it was enriched within patches (Fig. 3C, white arrows). Occasionally, FrmAHA was also seen at the distal ends of filopodia (Fig. 3C, yellow arrows). Thus, like the talinA and paxillin spots and F-actin patches, which transiently form at the cell-substrate boundary, FrmA also localised to similar structures. Taken together these data suggest that talinA, paxillin, F-actin and FrmA localise at the same adhesion structures at one point or another in their lifetime. However, whereas both talinA and paxillin promote formation of adhesion sites, FrmA promotes their turnover.

Cell migration is impaired in frmA cells

Data presented above show that FrmA is required for the correct regulation of cell-substrate adhesion, the actin cytoskeleton and cell shape – all of which are critical for cell migration. To determine the role of FrmA in cell migration, wt and frmA cells were starved and their ability to undergo directed cell migration determined. Starvation of Dictyostelium cells induces a survival response that induces cells to aggregate and requires directed cell migration along a cAMP gradient. This results in the formation of larger structures such as mounds and streams and the time required to form them gives a rough indication of the ability of a cell to undergo directed cell migration. The formation of mounds and streams was significantly delayed in frmA cells compared with that in wt cells (Fig. 4A). Whereas wt cells formed mounds and streams within 10 hours of starvation, frmA cells eventually did form mounds and streams after 18 hours of starvation, but they were visibly smaller than those formed by wt cells at 10 hours. Re-expression of full-length FrmAHA in the mutant cells restored the ability of the cells to form streams and mounds by 10 hours (Fig. 4A).

To further characterise the defect in cell migration, the speed of randomly moving non-starved cells and 6-hour-starved cells undergoing directed cell migration towards cAMP, were determined. The speed of randomly moving frmA cells was ∼30% that of wt cells (2.4 μm/minute), whereas movement of frmA cells towards cAMP was ∼50% of that in wt cells (10.7 μm/minute) (Fig. 4B). Although frmA cells were able to move towards the source of cAMP, they did so without adopting the highly elongated and polarised shape that wt cells did. Finally, to exclude the possibility that frmA cells were unable to move efficiently towards cAMP because of an inability to express starvation-induced genes required for cAMP signalling, RT-PCR was carried for cAMP receptor 1 (cAR1) and G alpha 2 (gα2) (Fig. 4C). cAR1and gα2 are both known to be induced by starvation and are required for directed cell migration towards cAMP in Dictyostelium (Kumagai et al., 1989; Saxe et al., 1991). Both these genes, as well as talinB (Fig. 1B), another starvation-induced gene (Tsujioka et al., 1999), are expressed to the same extent by both wt and frmA cells at 0 and 6 hours of starvation. Starvation also induced a greater expression of FrmA in wt cells (Fig. 1B and Fig. 4C), suggesting that FrmA may be required for efficient directional cell migration.

As cell-substrate adhesion was increased in frmA cells, we predicted that the contact area between frmA cells and the substrate would also be increased. This measurement was achieved by the parallel use of TIRF, to visualise the contact area of the cell to the substrate, and transmitted light to visualise the surface area of the entire cell. In non-starved cells (Fig. 4D) expressing GFP, the ratio of these areas showed that for frmA cells, the TIRF/transmitted light ratio was 1.0±0.18 (mean ± s.d., n=20 cells) whereas that of wt cells was 0.6± 0.17 (n=18 cells). Images showed that the area occupied by the transmitted light image of wt and frmA cells was similar. Thus, an increase in adhesion correlated with an increase in the contact area between cells and the substrate of randomly moving cells. As seen with non-starved cells, the `footprint' of GFP-expressing frmA cells undergoing directed cell migration was also greater than that of GFP-expressing wt cells (Fig. 4E and supplementary material Movie 8). The ratio of the areas of starved cells imaged using TIRF and transmitted light supports this, with wt cells having a ratio of 0.43± 0.19 (n=20 cells) whereas frmA cells had a ratio of 0.80±0.19 (n=18 cells). This increase in the cell-substrate contact area of frmA cells undergoing directed cell migration implies that frmA cell migration is impaired owing to increased cell-substrate adhesion as a result of stabilised cell-substrate adhesions.

These data show that FrmA is critically required for the regulation of cell-substrate adhesions during cell migration. More specifically FrmA promotes the turnover of adhesions and not their formation. The data presented above indicate that in Dictyostelium cells, adhesion sites are initiated at the distal ends of filopodia at the cell-substrate boundary and can be recognised by the presence of talinA. Paxillin is then recruited to the same sites, after which they are rapidly turned over. Loss of either paxillin or talinA leads to cells forming fewer cell-substrate adhesions and becoming less adherent to substrate (Bukharova et al., 2005; Niewohner et al., 1997). Furthermore, loss of myosinVII, which has been shown to bind talinA, also results in cells forming fewer cell-substrate contacts (Tuxworth et al., 2005; Tuxworth et al., 2001). All of these proteins are required for cell-substrate contacts and/or adhesions and loss of any one impairs the ability of cells to adhere to substrate and/or form cell-substrate contacts. FrmA is the first protein that has been shown to be anti-adhesive and can also localise to the cell-substrate boundary and tips of filopodia. These data are consistent with a model for cell-substrate adhesion whereby talinA, and possibly also myosinVII, initiate the formation of cell-substrate adhesions, after which paxillin is recruited to the same sites. Loss of FrmA prevents formation of these transient adhesions at the tips of filopodia and stabilises paxillin adhesion sites around the periphery of the cell body, resulting in increased adhesion and impaired motility. This is the first direct visualisation of the temporal and spatial sequence of events associated with cell-substrate adhesion in Dictyostelium and identifies FrmA as a key regulator of adhesion dynamics.

Strains and developmental conditions

All strains used in this study were derived from Ax2 cells and maintained in HL-5 medium at pH 6.8 (ForMedium Ltd., HLG0102). Wild-type cells null for FrmA (frmA) were generated as described below and clones selected for and maintained in 10 μg/ml of blasticidin. Three single-cell-derived frmA clones were analysed and showed identical phenotypes and the results are presented from one clone. The following proteins were expressed in the Ax2 (wt) and frmA cells: GFP, talinA:GFP, paxillin:GFP, (a kind gift from C. J. Weijer, WTB, University of Dundee, Scotland), LimEΔcoil:GFP (a kind gift from G. Gerisch, Max-Planck-Institut für Biochemie, Martinsried, Germany). HA epitope-tagged FrmA was also expressed in frmA cells. All strains were maintained in 15 μg/ml G418.

Fig. 4.

Starvation-induced development and directed cell migration is impaired in frmA cells. (A) Transmitted light images of cells taken at 0 and 10 hours of starvation. (B) The speed of non-starved, randomly migrating cells and starved cells undergoing cell migration towards a micropipette containing 2 μM cAMP was determined and expressed as a percentage of the wild type. 15 cells were tracked (using Image J software) from each strain on three separate occasions and the average speed ± s.e.m. is shown. (C) RT-PCR of total RNA from non-starved and 6-hour-starved cells was used to determine the expression of cAR1, gα2 and FrmA. Expression of eF1α was used as a template loading control. (D) Consecutive and merged TIRF (green) and transmitted light (red) images of randomly moving non-starved cells. (E) Consecutive and merged TIRF (green) and transmitted light (red) images of 6-hour-starved cells migrating towards a micropipette containing 2 μM cAMP (white asterisks).

Fig. 4.

Starvation-induced development and directed cell migration is impaired in frmA cells. (A) Transmitted light images of cells taken at 0 and 10 hours of starvation. (B) The speed of non-starved, randomly migrating cells and starved cells undergoing cell migration towards a micropipette containing 2 μM cAMP was determined and expressed as a percentage of the wild type. 15 cells were tracked (using Image J software) from each strain on three separate occasions and the average speed ± s.e.m. is shown. (C) RT-PCR of total RNA from non-starved and 6-hour-starved cells was used to determine the expression of cAR1, gα2 and FrmA. Expression of eF1α was used as a template loading control. (D) Consecutive and merged TIRF (green) and transmitted light (red) images of randomly moving non-starved cells. (E) Consecutive and merged TIRF (green) and transmitted light (red) images of 6-hour-starved cells migrating towards a micropipette containing 2 μM cAMP (white asterisks).

Generation of gene disruption mutant

Full-length FrmA was amplified and subcloned into the cloning vector pTOPO2.1 (Invitrogen) to create pTOPO2.1-FrmA. pTOPO2.1-FrmA was digested with SspI to remove a ∼800 bp fragment and generate pTOPO2.1-FrmAΔSspI, into which a blasticidin-resistance cassette was inserted (pTOPO2.1-FrmAΔSspI-Blastr). Using pTOPO2.1-FrmAΔSspI-Blastr as a template, PCR amplification with a mutated forward primer (5′-GAGGCCTGTTTCATGAGTATTTAGGAGAGGAAGGAATA-3′, with the point mutations in italics and an added StuI restriction site underlined) and a reverse primer (5′-GCGGCCGCTTAGCTTCCAACCAATTTAG-3′, with new NotI site underlined), was carried out to generate the FrmA-knockout construct. The PCR-amplified FrmA knockout construct was transformed into Ax2 cells and 10 μg/ml of blasticidin was used to select for FrmA-disrupted clones. Clones derived from single cells were isolated and the nature of the disruption confirmed by PCR of genomic DNA using 5′-GTAGTACATCCATTAC-3′ and 5′-CAGGTTTATGGAT-3′ as primers.

Determination of gene expression levels

Total RNA was isolated (Qiagen, RNeasy Mini kit) from non-starved and cells starved for 6 hours. Real-time RT-PCR (Qiagen, QuantiTect Sybr Green RT-PCR kit) was used to determine the expression levels of FrmA, talinB and eF1α and the primers used were 5′-ATTGGTGCAGCTGATGATTG-3′ and 5′-GACCCCACAATTCCAAACTG-3′; 5′-ACCAGGTGAAGGTGAAGGTG-3′ and 5′-TTTTAGCGGCCTGAGTGAGT-3′; and 5′-ATGGGTAAAGAGAAAACTC-3′ and 5′-GGTTGATTTTTCATCC-3′, respectively. RT-PCR amplification was used to determine the presence of starvation-induced proteins cAR1 and gα2 with the primers 5′-ATGGGTCTTTTAGATG-3′ and 5′-CATTAATGTGGCGT-3′; and 5′-GCATCATCAATG-3′ and 5′-CTGCTGTAACATC-3′, respectively.

Measurement of cell-substrate adhesion

1-3 × 105 cells growing in HL-5 medium were added to a well of a 12-well plate and allowed to settle for at least 5 hours. The plates were then shaken at 80, 120, 160 or 200 r.p.m. for 1 hour and the number of adhered and non-adhered cells determined using a cell haemocytometer.

Immunofluorescence

Cells were washed twice in phosphate buffer (10 mM KH2PO4 and K2HPO4, pH 6.8) and fixed (1% paraformaldehyde, 15% H2O-saturated picric acid (v/v), 10 mM PIPES, pH 6.0 for 15 minutes at RT). Cells were then postfixed with 70% ethanol (20 minutes at RT), blocked in 10% FCS (1 hour) and incubated with 0.5 μg/ml TRITC-phalloidin (1 hour) at RT. For FrmAHA staining, 5 μg/ml of anti-HA antibody conjugated to FITC (QED Bioscience Inc., 18848F) was added with the TRITC-phalloidin. Images were captured on an Olympus FV1000 confocal microscope using a 60× objective.

Development assays

Non-starved cells were washed twice in phosphate buffer and plated on 1% phosphate agar. Phase images were captured at 0 and 10 hours after plating using a standard Olympus CKX41 inverted microscope with a 5× objective.

Directed cell migration

Non-starved cells were washed twice in phosphate buffer and starved by shaking at 115 r.p.m. for 6 hours in phosphate buffer. Cells were then shaken in 5 mM caffeine for 20 minutes before adhesion to glass bottom dishes. The response of cells to a micropipette containing 2 μM cAMP was recorded using either TIRF microscopy or a standard Olympus CKX41 inverted microscope with a 60× and 40× objective, respectively. Cells were tracked using Image J software and the speed determined by dividing the total distance travelled over time.

TIRF microscopy

TIRF was performed on a Nikon Eclipse TE 2000-U microscope equipped with 60× and 100× 1.45 NA Nikon TIRF oil-immersion objectives. The Nikon Epi-fluorescence condenser was replaced with a custom condenser in which laser light was introduced into the illumination pathway directly from the output of a 3.5 μm optical fibre oriented parallel to the optical axis of the microscope. GFP and RFP excitation were performed at 473 nm and 561 nm respectively using individually coupled diode lasers (Omicron) controlled by a DAC 2000 card running MetaMorph (Molecular Devices). All live-cell imaging was performed with a Cascade 512F EMCCD camera (Photometrics UK). For GFP imaging a filter block consisting of a Z 473/10 excitation filter, a 488 RDC dichroic mirror and a ET 525/50M emission filter was used. For dual-colour imaging a dual band-pass filter block (Chroma 59022) was used, combined with an Optical Insights DualView containing 595 nm dichroic, and 525/50 and 630/60 emission filters.

We wish to thank C. J. Weijer, G. Gerisch for the gift of various reagents and R. H. Insall and Juliane P. Schwarz for useful discussions during the preparation of the manuscript. This work was supported by AICR.

Bretschneider, T., Diez, S., Anderson, K., Heuser, J., Clarke, M., Muller-Taubenberger, A., Kohler, J. and Gerisch, G. (
2004
). Dynamic actin patterns and Arp2/3 assembly at the substrate-attached surface of motile cells.
Curr. Biol.
14
,
1
-10.
Bukharova, T., Weijer, G., Bosgraaf, L., Dormann, D., van Haastert, P. J. and Weijer, C. J. (
2005
). Paxillin is required for cell-substrate adhesion, cell sorting and slug migration during Dictyostelium development. [erratum appears in J. Cell Sci. 2005 Oct 15; 118 (Pt 20): 4913 Note: Bukahrova, Tanya [corrected to Bukhavora, Tanya]].
J. Cell Sci.
118
,
4295
-4310.
Diakowski, W., Grzybek, M. and Sikorski, A. F. (
2006
). Protein 4.1, a component of the erythrocyte membrane skeleton and its related homologue proteins forming the protein 4.1/FERM superfamily.
Folia Histochem. Cytobiol.
44
,
231
-248.
Fey, P., Stephens, S., Titus, M. A. and Chisholm, R. L. (
2002
). SadA, a novel adhesion receptor in Dictyostelium.
J. Cell Biol.
159
,
1109
-1119.
Gerisch, G., Bretschneider, T., Muller-Taubenberger, A., Simmeth, E., Ecke, M., Diez, S. and Anderson, K. (
2004
). Mobile actin clusters and traveling waves in cells recovering from actin depolymerization.
Biophys. J.
87
,
3493
-3503.
Hibi, M., Nagasaki, A., Takahashi, M., Yamagishi, A. and Uyeda, T. Q. (
2004
). Dictyostelium discoideum talin A is crucial for myosin II-independent and adhesion-dependent cytokinesis.
J. Muscle Res. Cell Motil.
25
,
127
-140.
Kloeker, S., Major, M. B., Calderwood, D. A., Ginsberg, M. H., Jones, D. A. and Beckerle, M. C. (
2004
). The Kindler syndrome protein is regulated by transforming growth factor-beta and involved in integrin-mediated adhesion.
J. Biol. Chem.
279
,
6824
-6833.
Kreitmeier, M., Gerisch, G., Heizer, C. and Muller-Taubenberger, A. (
1995
). A talin homologue of Dictyostelium rapidly assembles at the leading edge of cells in response to chemoattractant.
J. Cell Biol.
129
,
179
-188.
Kumagai, A., Pupillo, M., Gundersen, R., Miake-Lye, R., Devreotes, P. N. and Firtel, R. A. (
1989
). Regulation and function of G alpha protein subunits in Dictyostelium.
Cell
57
,
265
-275.
Lo, S. H. (
2006
). Focal adhesions: what's new inside.
Dev. Biol.
294
,
280
-291.
Mitra, S. K., Hanson, D. A. and Schlaepfer, D. D. (
2005
). Focal adhesion kinase: in command and control of cell motility.
Nat. Rev. Mol. Cell Biol.
6
,
56
-68.
Nayal, A., Webb, D. J. and Horwitz, A. F. (
2004
). Talin: an emerging focal point of adhesion dynamics.
Curr. Opin. Cell Biol.
16
,
94
-98.
Niewohner, J., Weber, I., Maniak, M., Muller-Taubenberger, A. and Gerisch, G. (
1997
). Talin-null cells of Dictyostelium are strongly defective in adhesion to particle and substrate surfaces and slightly impaired in cytokinesis.
J. Cell Biol.
138
,
349
-361.
Octtaviani, E., Effler, J. C. and Robinson, D. N. (
2006
). Enlazin, a natural fusion of two classes of canonical cytoskeletal proteins, contributes to cytokinesis dynamics.
Mol. Biol. Cell
17
,
5275
-5286.
Saxe, C. L., 3rd, Johnson, R., Devreotes, P. N. and Kimmel, A. R. (
1991
). Multiple genes for cell surface cAMP receptors in Dictyostelium discoideum.
Dev. Genet.
12
,
6
-13.
Tsujioka, M., Machesky, L. M., Cole, S. L., Yahata, K. and Inouye, K. (
1999
). A unique talin homologue with a villin headpiece-like domain is required for multicellular morphogenesis in Dictyostelium.
Curr. Biol.
9
,
389
-392.
Tsujioka, M., Yoshida, K. and Inouye, K. (
2004
). Talin B is required for force transmission in morphogenesis of Dictyostelium.
EMBO J.
23
,
2216
-2225.
Tuxworth, R. I., Weber, I., Wessels, D., Addicks, G. C., Soll, D. R., Gerisch, G. and Titus, M. A. (
2001
). A role for myosin VII in dynamic cell adhesion.
Curr. Biol.
11
,
318
-329.
Tuxworth, R. I., Stephens, S., Ryan, Z. C. and Titus, M. A. (
2005
). Identification of a myosin VII-talin complex.
J. Biol. Chem.
280
,
26557
-26564.

Supplementary information