Keeping the vimentin network under control: cell-matrix adhesion-associated plectin 1f affects cell shape and polarity of fibroblasts

Abstract

Focal adhesions (FAs) located at the ends of actin/myosin-containing contractile stress fibers form tight connections between fibroblasts and their underlying extracellular matrix. We show here that mature FAs and their derivative fibronectin fibril-aligned fibrillar adhesions (FbAs) serve as docking sites for vimentin intermediate filaments (IFs) in a plectin isoform 1f (P1f)-dependent manner. Time-lapse video microscopy revealed that FA-associated P1f captures mobile vimentin filament precursors, which then serve as seeds for de novo IF network formation via end-to-end fusion with other mobile precursors. As a consequence of IF association, the turnover of FAs is reduced. P1f-mediated IF network formation at FbAs creates a resilient cage-like core structure that encases and positions the nucleus while being stably connected to the exterior of the cell. We show that the formation of this structure affects cell shape with consequences for cell polarization.

Figure 1.

Codistribution of P1f with FAs in primary mouse fibroblasts. (A and B) Confocal immunofluorescence microscopy of wild-type cells double-immunolabeled using antibodies to P1f and talin. Arrows and arrowheads in A indicate P1f-negative and P1f-positive FAs, respectively. Note, P1f is associated predominantly with FAs in more central regions of the cell but is missing from FAs at the cell periphery. (B) A magnified interior cell region. Arrows, FAs that are only partially associated with P1f. (C) Confocal image of triple-stained cells using antibodies to proteins indicated. Arrows, coalignment of P1f with centrally located elongated FbAs; arrowheads, peripheral FAs showing no association with integrin β1. Note, the predominant part of integrin β1 is found associated with central FbAs (arrows). Scale bars, (A and C) 20 μm; (B) 5 μm. (D) Statistical evaluation of P1f's codistribution with topological subpopulations (central vs. peripheral) of talin-positive FAs and FbAs. Data shown represent mean values (±SEM) from randomly chosen cells (n = 15); on average, 70 FAs or FbAs were counted per cell. ^†p < 0.001. (E) Quantitation of lamellipodial/filopodial areas invaded by vimentin filament networks in plectin^+/+ and plectin^−/− fibroblasts. Cells were double-immunolabeled using antibodies to vimentin and talin and lamellipodial/filopodial areas positive/negative for vimentin staining were measured. Data shown represent mean values (±SEM, n = 10). ^†p < 0.001. Note, most of the prostrusional areas (∼80%) in plectin^−/− fibroblasts are positive for vimentin staining.

Figure 2.

P1f codistributes with actin stress fibers and fibronectin fibrils. (A) Triple fluorescence microscopy of primary wild-type fibroblasts that had been seeded onto coverslips coated with Alexa 488–labeled fibronectin. To visualize P1f and FbAs, cells were double-immunolabeled using antibodies to P1f and actin. Images in the bottom row represent magnified view of boxed areas. Arrowheads, colocalization of P1f, actin stress fibers, and fibronectin fibrils in cell center. Scale bars, 10 μm. (B) Statistical evaluation of P1f-positive FbAs colocalizing with actin stress fibers and fibronectin fibrils. Colocalization was scored either individually after selection or as total FbAs signal within an area (see Materials and Methods). Data shown represent mean values (±SEM) from randomly chosen cells (n = 10). Note, both methods of measuring indicated ∼80 and ∼70–80% colocalization of P1f-positive FbAs with actin and fibronectin, respectively. (C) Quantitation of exterior Alexa 488–positive fibronectin fibrils (per cell) in primary cultures of plectin^+/+ and plectin^−/− fibroblasts stained as in A. Values shown were normalized for cell area. Note, plectin^+/+ and plectin^−/− fibroblasts form comparable numbers of fibronectin fibrils.

Figure 3.

P1f couples vimentin filaments to FAs. (A and B) Primary wild-type fibroblasts were subjected to triple immunofluorescence microscopy using antibodies to the proteins indicated. The image shown in B is a high-power view of a representative lamellipodial cell region of the cell shown in A. Large arrows, sites where vimentin filaments are targeted to P1f-positive FAs. Small arrows, vimentin filament intermediates not colocalizing with FAs and not, or only partially, decorated with P1f. (C) Immortal (p53^−/−) plectin^+/+ fibroblasts were cotransfected with vimentin-EGFP (green) and P1f-mCherry (red) expression plasmids and subjected to live imaging videomicroscopy. Three frames from a time-lapse movie (0′, 7′, and 11′) display a magnified view of the boxed area indicated in the image on the left. Arrowheads, P1f-positive FAs that are targeted by vimentin filaments. Arrows, parts of vimentin filaments that are not colocalizing with P1f and appear not to be fixed to FAs. Scale bars, (A) 10 μm; (B and C) 5 μm.

Figure 4.

Reexpression of P1f in plectin-deficient cells rescues FA anchorage of vimentin filaments. (A) Triple immunofluorescence microscopy of primary plectin^−/− mouse fibroblasts using antibodies to proteins indicated. Note vimentin filaments extending to the cell periphery (arrows), without being anchored at FAs (arrowheads); the cell edge is delineated by the talin signal. (B) Single frame of a time-lapse movie of immortal plectin^−/− mouse fibroblasts cotransfected with plasmids encoding full-length P1f-EGFP and mCherry-vimentin. Arrows, P1f-positive FAs targeted by vimentin filaments. Arrowheads, elongated patches of P1f (resembling FbAs) colocalizing with thin threads of bundled vimentin filaments. The boxed area corresponds to the cropped and magnified view shown in ROI. Note, expression of full-length P1f-EGFP suffices to reattach vimentin bundles to FAs. (C) Immunofluorescence microscopy (talin and vimentin) after forced expression of P1f-8-EGFP in immortal plectin^−/− fibroblasts. Note vimentin filaments overshooting FAs and extending toward the plasma membrane (arrows), in spite of P1f-8-EGFP being associated with FAs (arrowheads). Scale bars, 10 μm.

Figure 5.

First exon-encoded sequence of P1f controls FA-targeting and actin-binding. (A) Schematic representation of P1f-8's domain composition and first exon-encoded isoform-specific amino acid sequence. Arrowhead, tyrosine residue at position 20. (B and C) Fluorescence microscopy of immortal (p53^−/−) plectin-deficient fibroblasts transfected with cDNA constructs encoding EGFP-tagged plectin fragments as indicated. Scale bars, 10 μm. Note, only the wild-type P1f-8-EGFP fragment and the P1f-8F-EGFP mutant protein are targeted to FAs. (D) Cosedimentation of P1f mutant proteins with F-actin. Recombinant plectin fragments indicated were incubated with preassembled F-actin. F-actin and proteins bound were sedimented by centrifugation, and equal amounts of supernatant (S) and pellet (P) fractions were subjected to SDS-PAGE; separated proteins were visualized by Comassie Brilliant Blue staining. (E) Densitometric quantitation of gel bands. Data shown represent mean values (±SEM) of three independent experiments, including the one shown in D.

Figure 6.

P1f-induced formation of vimentin filament intermediates and their association with different plectin isoforms. (A) Immunofluorescence microscopy (vimentin) after forced expression of full-length P1f-EGFP in immortal plectin^−/− fibroblasts. A transfected cell (green) and a nontransfected cell (red only) are shown side by side. The boxed area corresponds to the cropped and magnified view shown in ROI. Arrows, vimentin filament intermediates colocalizing with P1f; arrowheads, vimentin filament intermediates bare of P1f. Note, that the nontransfected cell still exhibits wavy vimentin bundles extending to the cell periphery typical of plectin^−/− fibroblasts. Scale bars, 20 μm, and (ROI) 5 μm. (B) Immunoblotting of cell lysate from primary mouse fibroblasts using plectin isoform-specific (1f, 1c, and 1) and anti-pan plectin (no. 9) antibodies. (C) Triple immunofluorescence microscopy (P1c, talin, and vimentin) of primary plectin^+/+ fibroblasts. The bottom panels display cropped and magnified views of the boxed area in the top panel. Arrowheads, filament intermediates that are positive for P1c, but are not associated with FAs; arrows, P1c-decorated filament intermediates colocalizing with FAs. Scale bars, 20 μm, and (ROI) 5 μm. (D and E) Statistical evaluation of P1f and P1c distribution with vimentin filament intermediates at FAs. Data shown represent mean values (± the SEM) from randomly chosen cells (n = 6); on average, 150 FAs were counted per cell. *p < 0.05 and **p < 0.01, respectively. Note, P1c was found at FAs only in combination with filament intermediates, contrary to P1f, which also showed association with vimentin-negative FAs. (F) Immunofluorescence microscopy (vimentin) after forced expression of full-length P1c-EGFP in immortal plectin^−/− fibroblasts. The boxed area corresponds to the cropped and magnified view shown in the ROI. Arrows, P1c-positive vimentin filament intermediates; arrowheads, vimentin filament intermediates bare of P1c. Scale bars, 20 μm, and (ROI) 5 μm. (G) Statistical evaluation of vimentin filament intermediates forming in lamellipodia/filopodia upon forced expression of full-length isoform versions P1f-EGFP and P1c-EGFP in immortal plectin^−/− fibroblasts. Cells were immunolabeled as in A and F. Vimentin-positive filamentous structures, displaying free ends and residing in lamellipodial/filopodial areas devoid of vimentin network arrays, were scored as filament intermediates. Data shown represent mean values (±SEM) per area from randomly chosen cells (n = 15); *p < 0.05. (H) Immunofluorescence microscopy (talin and vimentin) after forced expression of P1c-EGFP in immortal plectin^−/− fibroblasts. Note partially bundled vimentin filaments extending toward the plasma membrane (arrowheads), without obvious attachment to FAs (arrows), Scale bar, 10 μm.

Figure 7.

P1f restrains the movement of vimentin filament intermediates at FAs. Immortal plectin^+/+ mouse fibroblasts were cotransfected with full-length P1f-EGFP and vimentin-mCherry expression plasmids and subjected to live imaging 16 h after seeding. (A) A selected frame from a time-lapse movie, and contrast-inverted images depicting P1f-EGFP and vimentin-mCherry alone, are shown. Arrowheads, P1f-positive FAs associated with (already immobilized) vimentin filament intermediates. (B) Time series from movie showing a cropped and magnified view of the boxed area indicated in A. The series shows a vimentin filament intermediate (shown in green and marked by short arrow in every time frame) moving toward one of the distal P1f/vimentin-positive FAs (direction of movement and immobile P1f/vimentin-positive FAs are indicated by long arrow and arrowheads, respectively). Note, the vimentin filament intermediate moves toward the cell periphery and finally fuses with the already existing short vimentin filament associated with P1f. Scale bars, (A) 20 μm, (B) 5 μm.

Figure 8.

Effects of P1f-mediated IF–FA association on FA turnover, cell shape, and polarization. (A) Statistical evaluation of the life times of IF-associated compared with IF-unassociated (control) FAs. Plectin^+/+ fibroblasts were cotransfected with zyxin-EGFP and vimentin-mCherry expression plasmids, and FA lifetimes were evaluated as described in the text; 25 FAs of each cell type (2–3 FAs per cell) were measured. Mean values ± SEM are shown. n = 25; ^†p < 0.001. (B) Statistical evaluation of cell area (i), cell perimeter (ii), shape factor (iii), and aspect ratio (iv) of primary plectin^+/+ and plectin^−/− fibroblasts. Cells were inspected by phase contrast microscopy 24–28 h after seeding. More than 50 randomly chosen cells were measured in three independent experiments (14–20 cells per experiment). Mean values ± SEM are shown. **p < 0.01 and †p < 0.001. (C) Statistical evaluation of parameters i–iv (see B) of immortal plectin^+/+ and plectin^−/− fibroblasts, or of immortal plectin^−/− fibroblasts transfected with cDNAs encoding EGFP-tagged versions of P1f, P1c, and P1. Cells were inspected and measured as in B. Mean values ± SEM are shown. **p < 0.01 and ^†p < 0.001. Note, values of both shape factor and aspect ratio in P1f-transfected cells were comparable to those of nontransfected controls.

Figure 9.

(A) Working model depicting how P1f mediates de novo vimentin filament assembly. α5β1 and αvβ3, corresponding integrins; FN, fibronectin; VN, vitronectin. Note, FA-associated P1f inhibits movement of vimentin intermediates along MT tracks, thus providing a docking site and assembly platform for the de novo formation of filaments by end-to-end fusion of filament intermediates. Translocation of P1f with actin bound to integrin α5β1 and tensin transfers vimentin filaments to FbAs, spatially restricting the vimentin IF network to the central core region of the cell. (B) Schematics of a polarized cell depicting the central localization of the vimentin IF network (encapsulating the nucleus) through its attachment to FbAs and centrally located FAs. The attachment of IFs to FbAs and FAs is mediated by P1f.