Substrate adhesion regulates sealing zone architecture and dynamics in cultured osteoclasts

Abstract

The bone-degrading activity of osteoclasts depends on the formation of a cytoskeletal-adhesive super-structure known as the sealing zone (SZ). The SZ is a dynamic structure, consisting of a condensed array of podosomes, the elementary adhesion-mediating structures of osteoclasts, interconnected by F-actin filaments. The molecular composition and structure of the SZ were extensively investigated, yet despite its major importance for bone formation and remodelling, the mechanisms underlying its assembly and dynamics are still poorly understood. Here we determine the relations between matrix adhesiveness and the formation, stability and expansion of the SZ. By growing differentiated osteoclasts on micro-patterned glass substrates, where adhesive areas are separated by non-adhesive PLL-g-PEG barriers, we show that SZ growth and fusion strictly depend on the continuity of substrate adhesiveness, at the micrometer scale. We present a possible model for the role of mechanical forces in SZ formation and reorganization, inspired by the current data.

Figure 1. Osteoclasts and sealing zones on vitronectin (VN)/PLL-g-PEG micro-patterns.

(a, b) SEM micrographs of differentiated osteoclasts spreading over non-adhesive PLL-g-PEG areas on square and striped VN-coated micro-patterns: a) 15×15 µm, marked with double arrows, 10 µm-wide barrier; b) 6 µm-wide stripes, marked with double arrows, 1.8 µm-wide barrier, marked by arrowheads). The cells were critical point dried after partial removal of the cell body, to enable better detection of the adhesion pattern on the substrate. (c, e) Osteoclasts growing on 20×20 µm VN patterns separated by (c) 8.5 µm- or (e) 1 µm-wide PLL-g-PEG barriers. Panel c shows a single frame from a time- lapse movie of a GFP-actin expressing osteoclast, while in Panels d-f, actin was stained with phalloidin-FITC. (d, f) Osteoclasts growing on 11 µm-wide VN coated stripes separated by (d) 4.5 µm- or (f) 900 nm-wide PLL-g-PEG barriers. Green: GFP-actin/Phalloidin-FITC, Blue: PLL-g-PEG-TRITC. Scale bars, 20 µm.

Figure 2. Sealing zone dynamics on micro-patterned surfaces.

Time-lapse images of an expanding SZ on a 6 µm-wide adhesive VN-coated stripe flanked by 3 µm-wide PLL-g-PEG barriers. Green: GFP-actin; Blue: PLL-g-PEG-TRITC. Inserts show overlays of the same ring at three time points, represented by different colours: t = 0 s blue; t = 120 s green; t = 270 s red. Scale bars, 5 µm. Note that as soon as the SZ reaches a passivated area, its outward movement is arrested.

Figure 3. Sealing zone architecture at the VN/PLL-g-PEG interface.

(a–c) A ventral membrane (shown at 3 levels of magnification) displaying a sealing zone at the VN/PLL-g-PEG interface. (b) Arrows point to podosome cores, while the asterisk indicates the border between VN and PLL-g-PEG. (c) Arrows point to lateral and interconnecting actin fibers anchoring the podosomes to the membrane and connecting them to neighboring podosomes, respectively. (d–f) A ventral membrane (shown at 3 levels of magnification) displaying a row of podosomes formed at the VN/PLL-g-PEG interface. (d) 10 µm-wide VN stripes, marked with double arrows; 1 µm-wide barriers, marked by arrowheads. Scale bar, 10 µm. (e, f) Magnifications of indicated selections. Scale bars, 1 µm. (f) Arrows point to podosome cores formed at the VN/PLL-g-PEG interface.

Figure 4. Loss of one vinculin band upon blocking sealing zone expansion.

(a, b) SZ arrested at a VN/PLL-g-PEG interface on striped and square micro-patterns: a) 5.35 µm-wide VN stripes, separated by 3.3 µm-wide barriers. Inserts: display of line scans of actin and vinculin fluorescence across the SZ at the VN/PLL-g-PEG border (left insert) and when on a uniform VN-coated surface (right insert). b) 20×20 µm VN squares, separated by 1 µm-wide barriers; (c, d) SZ arrest by an intercepting SZ within the same cell or a neighbouring cell, respectively. c) a SZ on square micro-patterns of 40×40 µm, barrier 1 µm-wide; d) a peripheral SZ on non-patterned, VN-coated glass slides. Inserts (bottom) show higher magnification of selected areas, with arrowheads indicating the loss of the vinculin ring in arrested SZ areas, green: phalloidin-FITC; red: anti-vinculin 546 nm; blue: PLL-g-PEG-Atto-633. Scale bars, 10 µm.

Figure 5. Interconnecting actin fibers bridging over a 1 µm-wide PLL-g-PEG barrier.

(a) SZ formed on striped micro-patterns, bridging a PLL-g-PEG barrier (2.8 µm-wide adhesive stripes, barriers 1 µm wide). Note that only the actin component of the SZ bridges the barrier, while the vinculin domains remain restricted to the adhesive areas. Inserts display magnifications of respective selections. Green: Phalloidin-FITC; Red: anti-vinculin 546; Blue: PLL-g-PEG-Atto-633. Scale bars, 10 µm. (b–e) SEM micrograph of a ventral membrane formed on 40×40 µm VN coated micro-patterns displaying a SZ spanning a 1 µm-wide PLL-g-PEG barrier by interconnecting actin fibers (shown at 4 levels of magnification). (c–e) Magnifications of the respective selections in b and d. Scale bars: (b) 3 µm; (c, d) 1 µm; (e) 200 nm. Images in b, d and e were taken after tilting the stage of the SEM by 30°, to facilitate visualization of the area underneath the fibers.

Figure 6. Fusion of sealing zones on VN/PLL-g-PEG micro-patterns.

Time sequence depicting fusion of two sealing zones on a 10 µm-wide VN stripe separated by a 3.5 µm-wide PLL-g-PEG barrier, over 1 min. SZs fuse at the site of contact, followed by a widening of the fusion site toward the edges of the micro-pattern. Green: GFP-actin. Blue: PLL-g-PEG-TRITC. Scale bars 10 µm.

Figure 7. Schematic drawing depicting a possible model for force associated regulation of SZ formation, expansion and fusion.

Podosomes are displayed in cross-section, with vertical stripes representing the branched actin core bundle, and horizontal lines indicating the interconnecting actin fibers; top view shows podosome circular actin cores and centrifugally expanding actin fibers either anchoring the cores to the membrane, or interconnecting them. Adhesive plaque areas are represented in red, while black bars represent the adhesive surface. Arrows display forces (Fi  =  force acting on the inner plaque band and Fo  =  force acting on the outer plaque band). (a) Cross-section and top view of a SZ. (b) A possible two-force regime balancing SZ expansion and integrity, responsible for the force-dependent formation of the inner and outer adhesive plaque domains. (c) Model indicating the force-dependent loss of the inner plaque band upon arrest of SZ expansion. The blue bar represents the PLL-g-PEG barrier. (d) Force- dependent model of SZ fusion, where fusion of the two outer plaque bands is followed by the interlinking of the podosome cores of the two rings, leading to local cancellation of the net force. This, in turn, finally leads to the local disassembly of the podosomes, and the fusion of the two rings.