SEARCH

SEARCH BY CITATION

Keywords:

  • Actin;
  • bleb;
  • contex;
  • myosin

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

Blebs are protrusions of the cell membrane. They are the result of actomyosin contractions of the cortex, which cause either transient detachment of the cell membrane from the actin cortex or a rupture in the actin cortex. Then, cytosol streams out of the cell body and inflates the newly formed bleb. During expansion, which lasts ∼30 s, the bleb is devoid of actin and the surface area increases through further tearing of membrane from the cortex and convective flows of lipids in the plane of the membrane through the bleb neck. Once expansion slows, an actin cortex is reconstituted. First actin-membrane linker proteins, such as ezrin, are recruited to the bleb, then actin, actin-bundling proteins and finally myosin motor proteins. Retraction lasts ∼2 min and is powered by myosin motor proteins. Though it has been less studied than other actin-based membrane protrusions such as lamellipodia or filopodia, blebbing is a common feature of cell physiology during cell movement, cytokinesis, cell spreading and apoptosis. This review will succinctly attempt to summarize what we know about the mechanisms involved in blebbing, when it appears in cell physiology and what open questions remain.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

Blebs are cellular membrane protrusions that appear and disappear on a timescale of minutes. They manifest themselves in a number of instances during cell physiology such as during cytokinesis, where they appear at the poles of dividing cells (Boss, 1955; Porter et al., 1973; Fishkind et al., 1991; Burton & Taylor, 1997; Boucrot & Kirchhausen, 2007), during cell spreading in many cell lines (Bereiter-Hahn et al., 1990; Pletjushkina et al., 2001), or locomotion of tumour and embryonic cells (where they are known as lobopodia) (Trinkaus, 1973; Kubota, 1981; Sroka et al., 2002; Friedl & Wolf, 2003; Sahai & Marshall, 2003; Blaser et al., 2006). One particularly striking form of blebbing, known as ‘circus movement’, appears in dissociated blastomeres from amphibian and fish embryos, where one single bleb travels around the cell periphery relentlessly (Fujinami & Kageyama, 1975; Fujinami, 1976; Johnson, 1976; Olson, 1996). Blebs are commonly mentioned in the same breath as apoptosis. During cell death, two types of blebs appear: dynamic blebs and larger stationary blebs. Dynamic blebbing is associated with the execution phase of apoptosis and appears closely related to blebbing in ‘healthy’ cells (Mills et al., 1999). Larger stationary blebs appear during cell necrosis and are a common feature of cells exposed to noxious stimuli such as hypoxia, oxidants, or ATP depletion. These blebs are, to the trained eye, different: they are larger and more transparent and their growth is independent of actomyosin contractions (Barros et al., 2003). This review will restrict itself to dynamic blebs. Although blebs seem to have been preserved through evolution in a variety of organisms, we still do not fully understand why cells bleb. In some of the cases mentioned in this paragraph, the raison d'être of blebbing is fairly obvious (e.g. during cell movement – though exactly how polarized blebbing is translated into cell movement remains to be understood); in others, it is far less clear (apoptosis, cytokinesis, cell spreading) and purely speculative.

In addition to being an interesting cellular phenomenon, blebs also offer us a window into cytomechanics as a number of interesting biophysical phenomena happen during the bleb life cycle. First, bleb nucleation results from either the detachment of the cell membrane from the actin cortex (Charras et al., 2005) or a localized rupture of the actin cortex (Paluch et al., 2005) (Figs 1(A1) and (A2)). In the former, blebs can be used to investigate changes in adhesion energy between the cortex and the membrane (Charras et al., 2008); in the latter, blebbing could be used to study stress build-up in the actin cortex and the mechanical properties of the cell cortex (Paluch et al., 2005). Second, during growth, the plasma membrane is transiently devoid of an actin cytoskeleton and this enables the study of the mechanics of somatic cell membrane (Figs 1(C1) and (C2)). Furthermore, for bleb growth to occur, lipids must flow into the bleb through its neck, making the examination of the dynamics of lipid flows within the membrane possible. Third, after growth has stalled, an actin cortex is reconstituted under the plasma membrane and this phenomenon enables the mechanics of a two-dimensional contractile actin cortex to be investigated (Charras et al., 2008) (Figs 1(D) and (E)). Finally, in many cells, incubation with actin depolymerizers such as cytochalasins or latrunculin gives rise to blebs; assuming that membrane–cortex attachment is homogeneously fragilized throughout the cell, the appearance of blebs may be reported on higher local intra-cellular pressure.

image

Figure 1. Schematic diagrams of the life time of a bleb resulting from either detachment of the membrane from the actin cortex due to a transient increase in intra-cellular pressure (1) or to local rupture of the actin cortex (2). The actin cytoskeleton is shown in red, the membrane in mauve, the cytoplasm in green, membrane-actin linker proteins are shown as blue ellipses, actin-bundling proteins are depicted as green ellipses and myosin motor proteins are represented as circles with a squiggly tail. (A1) A local shortening of the actin cortex caused by myosin contraction gives rises to a centripetal compression of the cytoplasm and, because the cytosol cannot move instantaneously out of the zone being compressed, this causes a localized increase in intra-cellular pressure (P). (A2) Myosin movement locally increases tension in the actin cortex. (B1) High local intra-cellular pressure tears the membrane from the actin cortex and cytosol is expelled from the cell body by the pressure gradient (blue arrows). (B2) The actin cortex ruptures and cytosol is expelled from the cell body, inflating the cell membrane (blue arrows). (C1 and C2) Cytosol rushes into the bleb and the growth in volume is accommodated by tearing of the membrane from the actin cortex (which increases the diameter of the bleb neck, double arrow) and by flow of lipids into the bleb membrane through the bleb neck (mauve arrows). (D) Once growth has stalled, an actin cortex re-forms and catex at the base of the bleb is disassembled. Membrane-actin linker proteins (blue ellipses) are recruited first, then actin, then actin-bundling proteins (green ellipses). (E) Later, myosin is recruited to foci along the newly formed cortex and retraction of the bleb starts forcing cytosol back into the cell body (blue arrows). During retraction, the actin cortex and the membrane crumple.

Download figure to PowerPoint

A phenomenological definition of blebbing

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

Before getting into the particulars of blebbing, it is useful to agree on a functional definition of blebbing. I will call bleb any dynamic cellular protrusion in which the membrane is transiently decoupled from the actin cortex and that is dependent upon actinomyosin contraction for nucleation. In all cases, expansion is rapid and generally outpaces typical velocities encountered in protrusions dependent on actin polymerization (Keller & Eggli, 1998). Experimental criteria for this definition are inhibition of blebbing by (1) depolymerization of actin, (2) inhibition of myosin contractility, or (3) hyperomostic shock. This definition is general enough that it covers blebbing from filamin-deficient melanoma cells (Cunningham, 1995; Charras et al., 2005), cells in cytokinesis (GC unpublished observations), apoptosing cells where caspase proteolysis is inhibited (Mills et al., 1998), circus movements in dissociated blastomeres (Fujinami, 1976; Charras et al., 2008) and locomotion of walker-carcinosarcoma and embryonic cells (Keller & Eggli, 1998; Blaser et al., 2006), while still grasping the essence of the phenomenon.

This functional description purposely does not make any mention of morphology because blebs come in different shapes and forms. When imaged using light or electron microscopy, blebs from non-motile, dividing, or apoptotic cells look like quasi-spherical membrane protrusions (Fig. 2(A); Charras et al., 2006). However, in motile cells bleb shapes vary from spherical to bean-shaped (Fedier et al., 1999; Sroka et al., 2002; Fink, 2003; Blaser et al., 2006; Yoshida & Soldati, 2006) and in circus movements they are shaped like tear drops (Fujinami, 1976).

image

Figure 2. Scanning Electron Micrographs (SEM) of filamin-deficient blebbing cells. (A) SEM of a blebbing filamin-deficient cell. Blebs appear as quasi-spherical surface protrusions up to 5 μm in diameter. Blebs are distributed over the cell surface with no preferential region. Newly formed blebs appear taut, whereas older blebs are more crumpled. (B) A fully retracted bleb. Enlargement of area boxed in A. The cell membrane appears very wrinkled, reflecting buckling of the actin cortex during bleb retraction. (C) SEM of the actin cortex of a retracting bleb. The cell membrane was removed by treatment with 2% Triton-X and 1% CHAPS in the presence of phalloidin to stabilize F-actin before fixation with formaldehyde, tannic acid and uranil acetate (Svitkina & Borisy, 1998). F-actin filaments are clearly visible and form a cage-like structure under the bleb membrane with a ∼100 nm mesh size.

Download figure to PowerPoint

Motive force for bleb nucleation

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

Blebs arise from an actomyosin-driven increase in intra-cellular pressure that causes the membrane to delaminate from the actin cortex or ruptures the actin cortex (Fig. 1). Experimentally, incubation with the myosin II-ATPase inhibitor blebbistatin inhibits blebbing (Cheung et al., 2002) and myosin-null Dictyostelium cells cannot bleb (Traynor & Kay, 2007). Upstream of myosin, the small GTPase Rho and its downstream effector Rho-kinase (ROCK) play an essential role in governing cortical contractility (Mills et al., 1999; Coleman et al., 2001; Yarrow et al., 2005; Charras et al., 2006). One particularly convincing proof of RhoA contractility as the source of blebbing comes from experiments on spreading 3T3 cells (Pletjushkina et al., 2001). In control cells, spreading occurs without blebs; however, if microtubules are depolymerized by nocodazole treatment [which upregulates rhoA (Waterman-Storer & Salmon, 1999)], contractions of the cortex occur and give rise to blebs. In this system, blebbing can be prevented by inclusion of ROCK or myosin light-chain kinase (MLCK) inhibitors in the medium. Blebbing cells display vigorous contractions of the actin cortex that can be detected by plating the cells on wrinklable substrates or soft collagen gels (Pletjushkina et al., 2001; G.C., unpublished observations). Accompanying cortical contractility are cyclic intra-cellular calcium transients (Pletjushkina et al., 2001; Blaser et al., 2006; G.C., unpublished data), but whether these calcium oscillations are the cause or a consequence of the enhanced contractility is unclear. Indeed, chelation of extracellular calcium inhibits blebbing (G.C., unpublished observations) and therefore elevated intra-cellular calcium could cause the myosin contractions that give rise to blebs. Conversely, calcium may enter the cells as a consequence of increased membrane tension in blebs.

There are two main hypotheses to explain how blebs result from cortical contractions:

  • 1) 
    delamination of the cell membrane from the actin cortex,
  • 2) 
    fracture of the cellular actin cortex.

In some cells such as filamin-deficient cells, the actin cortex appears intact during bleb expansion and no crack is apparent in light microscopy images. Contractions of the actin cortex force cytosol out of the cell body into an area of membrane that has detached from the actin cytoskeleton. What causes the initial separation of the membrane from the cytoskeleton is unknown. Phosphatidylinositol-4,5-diphosphate (PIP2) is an important regulator of proteins involved in actin–cortex attachment, such as the ERM proteins (which it activates) or MARCKS (myristoylated alanine-rich C kinase substrate, to which it binds), and loss of PIP2 may lead to downregulation of membrane–cortex adhesions (Sheetz, 2001). Dephosphorylated MARCKS associates with the cell membrane, where it sequesters PIP2 and binds F-actin; upon phosphorylation by PKC its affinity for F-actin decreases and it dissociates from the cell membrane (Hartwig et al., 1992; Glaser et al., 1996). Cells that express a mutant version of MARCKS that cannot leave the membrane display large surface blebs during spreading, whereas wild-type cells do not (Myat et al., 1997). One possible explanation of these results is that mutant MARCKS plays the role of a PIP2 chelator at the membrane, preventing it from binding or activating membrane–cytoskeleton linker proteins (Sheetz et al., 2006). However, in filamin-deficient cells, PIP2 does not appear to play a role in tethering the actin cytoskeleton to the cell membrane or activating linker proteins because treatment of cells with chelators of PIP2 such as neomycin sulfate or PBP-10 (Cunningham et al., 2001) does not prevent retraction of blebs by a contractile actin cortex (Charras et al., 2006). Another possible mechanism for bleb formation is that membrane could just be torn from the cytoskeleton owing to an increase in intra-cellular pressure created by myosin contraction (Albrecht-Buehler, 1982; Cunningham et al., 1992). In support of this hypothesis, detachment of the membrane from the cytoskeleton can be induced by suction of part of the cell into a micropipette (Merkel et al., 2000; Sheetz et al., 2006) or by pulling the membrane with a magnetic bead (Vonna et al., 2003). Filamin-deficient melanoma cells bleb constitutively because of decreased adhesion energy between the cortex and the cell membrane owing to lack of filamin (Cunningham et al., 1992; Dai & Sheetz, 1999). In these cells, blebbing can be decreased by increasing the adhesion energy between the membrane and the actin cortex by stable transfection with either filamin (Cunningham et al., 1992) or a constitutively active mutant of the ERM protein ezrin (Charras et al., 2006), both proteins that link the actin cortex to the cell membrane (Stossel et al., 2001; Bretscher et al., 2002). Conversely, blebbing can be increased by decreasing adhesion energy between the cell and the membrane by microinjection or transfection with the FERM domain (which acts as a dominant negative) (Amieva et al., 1999; Charras et al., 2006) or by increasing cell contractility by incubation with the phosphatase inhibitor calyculin A (Charras et al., 2008).

In other cells, myosin contractions lead to fracture of the actin cortex and cytosol streams out of the cell body to fill the bleb (Paluch et al., 2006). Cracks in the actin cortex have been observed experimentally in Walker carcinosarcoma cells or L929 fibroblast fragments (Keller & Eggli, 1998; Paluch et al., 2005). Lobopodia in Walker carcinosarcoma cells have actin walls between the newest lobopodium and the previous one, known as constrictions. When confocal stacks of these constrictions are examined, cracks in the cortex are clearly apparent. In L929 fibroblast fragments, cracks in the actin cortex are noticeable shortly before the emergence of a new bleb (Paluch et al., 2005).

Blebbing is a locally controlled phenomenon

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

The first indication that blebbing is a locally controlled phenomenon came from experiments in which cells were broken into fragments by fluid shear stresses after exposure to cytochalasin D (Albrecht-Buehler, 1980). These fragments, some as small as 2% of the cell volume, still displayed motile behaviours such as lamellipodial movement, ruffling, or blebbing (Fig. 3, arrow). More recently, studies using local treatment of cells showed that blebs result from myosin contraction in the direct vicinity of the bleb (Charras et al., 2005). In these experiments, only one half of the cell was exposed to myosin or ROCK inhibitors. Blebbing was inhibited on that side but, surprisingly, continued unperturbed in the other half of the cell. Similarly, if the actin cortex was disrupted by treatment with actin depolymerizers, blebbing was inhibited on the disrupted side but not on the other side. If blebs are due to tearing of membrane from the actin cortex in response to hydrostatic pressure, this signifies that blebs are the result of locally generated hydrostatic pressure and that pressure does not have time to equilibrate over the whole cell before being used to form a bleb (Mitchison et al., 2008). If blebs are nucleated by rupture of the actin cortex, the experiments where contractility was locally inhibited imply that fractures in the actin cortex are caused by myosin motors acting in the vicinity of the bleb and therefore that tangential stresses within the cortex are not uniform, something that has also been shown in L929 cells (Paluch et al., 2005). However, the second set of experiments in which the actin cortex was disrupted cannot be so readily explained in that framework but may be attributed to the absence of significant intra-cellular pressure build-up in filamin-deficient cells. Indeed, when the cortex of HeLa cells blocked in metaphase is locally disrupted by local treatment with latrunculin, a bleb forms in that region. Experimental estimates of the intra-cellular pressure in mitotic HeLa cells show that this is ∼3-fold higher than in filamin-deficient cells (Dai & Sheetz, 1999; Charras et al., 2008).

image

Figure 3. Differential interference contrast image of a blebbing cell fragment (arrow). Fragments were obtained by incubation of cells with 5 μM cytochalasin D for 45 min and subsequent mechanical disruption by pipetting (Albrecht-Buehler, 1980). Cell fragments (arrow) as small as 2% of original cell volume can bleb autonomously until they run out of ATP. Blebbing is independent of attachment of the fragments to the glass cover slip. Scale bar = 10 μm.

Download figure to PowerPoint

Bleb growth

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

Once a bleb has been nucleated, there are several mechanisms through which growth in volume could proceed: (1) tearing of the cell membrane from the actin cytoskeleton, (2) unfolding of membrane wrinkles, (3) flow of lipids into the bleb through the bleb neck, (4) fusion of vesicles to the membrane in the vicinity of a growing bleb.

Tearing of the cell membrane from the actin cortex has been experimentally observed in filamin-deficient blebbing cells and is particularly prevalent during circus movements (Charras et al., 2008). In filamin-deficient blebbing cells, the base of a growing bleb increases over time, a clear indication of detachment of the membrane from the actin cortex (Fig. 4(A), double arrow). In most cases, tearing proceeds smoothly but, in some cases, distinct periods of detachment and pause can be distinguished (Charras et al., 2008). During pauses in bleb base growth, bleb volume still increases and this causes the angle between the bleb membrane and the cell body to increase until it reaches a point where tearing of the membrane from the cortex becomes favourable: Tm (1 − cos  (α)) ≥J, with Tm the membrane tension, α the angle between the bleb membrane and the cell body (shown on Fig. 4(A)) and J the adhesion energy between the membrane and the cortex. Assuming that Tm stays constant because the cell possesses excess membrane and J stays constant because the nature of the membrane–actin linkers does not change, tearing continues until the angle becomes unfavourable to delamination: Tm (1 − cos  (α)) ≤J. An exact analogy of this is peeling a piece of tape from a flat surface: when the angle is shallow, the tape is difficult to detach; when the tape is folded back onto itself (at a 180° angle), peeling is easy. One particularly spectacular manifestation of tearing of the membrane from the cortex takes place during circus movements in dissociated blastomeres from frog embryos (Fig. 5). There, the leading edge of the bleb alternates between pauses during which the front of the bleb bulges owing to influx of cytosol from the bleb rear and the angle between the bleb membrane and the cell body increases, and periods of rapid growth when the membrane delaminates from the cortex and the angle between the cell body and the bleb membrane decreases.

image

Figure 4. Shape of the bleb membrane during expansion and retraction. Filamin-deficient blebbing cells were transfected with PH-PLCδ-GFP, which binds PIP2 and localizes to the cell membrane, and imaged with a spinning disk confocal microscope at 0.5-s intervals. In both images, multiple time points were projected onto the same image to reflect growth or retraction. Time intervals were chosen such that successive bleb outlines are easily distinguishable from one another. The evolution of time is shown by the white arrow. (A) Time series of an expanding bleb. During expansion, the bleb membrane is taut. The base of the bleb (double arrow) increases in diameter over time and the angle α between the bleb membrane and the cell body cycles during growth. Once expansion has slowed, a secondary bleb appears to the right side of the primary bleb. Scale bar = 2 μm. (B) Time series of a retracting bleb. During retraction, the membrane buckles and appears progressively thicker and wavier.

Download figure to PowerPoint

image

Figure 5. Circus movement in a Xenopus blastomere expressing PH-PLCδ-GFP. mRNA encoding PH-PLCδ-GFP (which localizes to the cell membrane) was injected into xenopus embryos at the two-cell stage. The animal cap was dissected from stage 10 embryos and dissociated in Ca2+ free Mg2+ free buffer. The isolated blastomeres display a form of blebbing in which one single bleb travels around the cell periphery over time. In this image, three different time points depicting the movement of the bleb membrane during circus movements are shown chronologically in red, green and blue. In the first time point (red), the membrane has started to delaminate from the cell body; this continues (in green) until it slows and the bleb bulges owing to travel of cytosol from the rear of the bleb (in blue). The white arrow shows the direction of travel. Scale bar = 10 μm.

Download figure to PowerPoint

Cells possess a large excess of membrane area and can increase their volume 10-fold (and hence their apparent area ∼4.6-fold) in response to exposure to dilute media (Hamill & Martinac, 2001). This excess membrane is stored in the form of folds and microvilli or endosomes (Hamill & Martinac, 2001; Morris et al., 2003; Boucrot & Kirchhausen, 2007). Therefore, bleb growth could simply be the result of unfolding of membrane. If this was the case, there should be a constant ratio between the area of unfolded membrane (the membrane in the bleb) and the area of folded membrane that gave rise to it (the area of the base of the bleb). Experimentally, the ratio of the bleb perimeter to the base diameter increases over time rather than staying constant (Charras et al., 2008), contradicting this hypothesis. Furthermore, if growth were due to membrane unfolding, the angle between the cell body and the bleb membrane should stay constant over time, which is also not the case (Charras et al., 2008). Therefore, unfolding cannot be the only source of bleb growth, though it may contribute to increase in surface area.

Flow of lipids into the bleb through the bleb neck is plausible because the cell membrane is a two-dimensional fluid. In non-blebbing cells, flows of lipids in the plane of the membrane have been observed and their dynamics quantified using tether extraction from the cell membrane with optical tweezers (Dai & Sheetz, 1999; Hochmuth & Marcus, 2002). In blebbing cells, the evidence for convective flows of lipids is circumstantial and based on the impossibility of unfolding being the only source of increase in bleb membrane area. However, lipid flows reaching 2 μm s−1 would be needed to accommodate bleb growth, and there is at present no definitive experimental evidence that this happens. Further experiments using fluorescence recovery after photobleaching (FRAP) will be needed to confirm this.

Finally, fusion of exocytotic vesicles with the membrane could give rise to a local excess of membrane area that could bulge out, creating a bleb. Three lines of evidence argue against this scenario: (1) no fusion of vesicles with the membrane in the vicinity of blebs could be observed with lipid markers such as DiI in filamin-deficient blebbing cells or Dictyostelium (G.C., unpublished data; Traynor & Kay, 2007)), (2) no accumulation of exocytotic markers could be directly correlated with blebs in dividing cells (Boucrot & Kirchhausen, 2007), (3) the PH domain of PLCδ localizes to the plasma membrane of filamin-deficient cells but not to intra-cellular membranes. If vesicles fused with the membrane prior to the emergence of blebs, a local decrease in PH-PLCδ-GFP fluorescence intensity should be observed in the region where a bleb is being nucleated. In filamin-deficient cells expressing PH-PLCδ-GFP, no such decrease in fluorescence was observed experimentally (G.C., unpublished data).

In conclusion, tearing of the membrane from the cell cortex, flow of lipids in the plane of the membrane and unfolding of membrane wrinkles may all contribute to bleb growth, but their relative contributions are unknown at present.

Growing blebs may have an erythrocytic cytoskeleton

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

During expansion, the bleb membrane appears devoid of an actin cytoskeleton (Cunningham, 1995; Charras et al., 2005; Paluch et al., 2005). However, all experimental observations were effected using fluorescence microscopy, and therefore it is possible that quantities of actin below the detectable threshold are still present at the membrane during growth. However, the absence of a well-defined actin cortex raises the question of whether the bleb membrane is an unsupported bilayer or if it retains some submembranous support such as an erythrocytic cytoskeleton. Some protein components of the erythrocytic cytoskeleton (protein 4.1 and ankyrin) localize to the membrane during bleb growth and, therefore, it is likely that the cell membrane is not entirely devoid of a cytoskeleton during blebbing. To date, the presence of spectrin (the main component of the erythrocytic cytoskeleton) in blebs during growth has not been confirmed owing to the challenges in expressing a full-length GFP tagged spectrin construct, although it is present in retracting blebs. Another independent clue of the existence of an erythrocytic cytoskeleton in growing blebs comes from measurements of the bleb membrane bending rigidity. Indeed, during growth, the bending rigidity of the bleb membrane is close to that of red blood cells (Zilker et al., 1992; Charras et al., 2008).

Other cytoskeletal proteins also localize to growing blebs such as isoforms of myosin I, but their exact role during growth (or retraction) is unknown (Charras et al., 2006).

Signal for regrowth of an actin cortex

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

What triggers regrowth of an actin cortex under the membrane of a newly formed bleb and whether such a trigger exists remain unclear. There are two possible scenarios: (1) cortex assembly is constitutive and blebs are devoid of actin because their growth outpaces constitutive cortical assembly and (2) there exists a signal that detects separation of membrane from the cortex. This signal could be, for example, the presence of PIP2 freed from its interaction partners in the bleb membrane (Sheetz et al., 2006), the opening of mechanosensitive ion channels in response to bleb growth, or the presence of active rhoGTPases under the membrane of growing blebs. Although the presence of rhoA under the bleb membrane has been observed and antibodies to rhoA-GTP stain expanding blebs (Charras et al., 2006), rhoA activity has not been monitored in live blebbing cells by FRET and it is therefore difficult to know if rhoA is continuously GTP-bound or if GTP-binding occurs upon dissociation of the membrane from the cortex. In general, data supporting either hypothesis is scant and this remains an open question. Separation of membrane from the cortex can be induced by mechanical disruption in well-controlled experimental systems (Merkel et al., 2000; Vonna et al., 2003) and regrowth of a cortex has been observed. These systems may be better suited to determining the existence and nature of a cortex regrowth signal than constitutively blebbing cells. Concomitantly with regrowth of an actin cortex under the bleb membrane, actin cortex at the base of the bleb disappears and what signals this disassembly is unknown. This coordinated appearance of cortex in one location under the bleb membrane and disappearance at the other raises the possibility that the protein components of the old cortex get locally recycled to create the new cortex.

Regrowth of an actin cortex and retraction

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

Regulatory proteins of the actin cytoskeleton are present at the cell membrane during expansion and cortex regrowth. Of particular interest amongst these are the small GTPase rhoA (Charras et al., 2006), its downstream effector ROCK (Wyckoff et al., 2006; Blaser et al., 2006; Sahai, 2007) and the rhoA activators rhoGEFs (Charras et al., 2006). However, when these proteins get activated and whether they participate in the reassembly of the cortex remains unclear.

The first protein to be recruited to blebs is the ERM protein ezrin, which links the actin cytoskeleton to the cell membrane. It is recruited to blebs independently of actin and appears earlier than actin, but is not essential for regrowth of cortex (Charras et al., 2006). Perhaps unsurprisingly, ezrin's primary role appears to be tethering the actin cortex to the cell membrane (Bretscher et al., 2002). Indeed, cells expressing the dominant negative FERM domain of ezrin do not manage to fully retract blebs though they do regrow actin cortices (Charras et al., 2006). Moesin, another ERM protein, is also recruited to the bleb membrane, but its time of arrival relative to actin is unknown.

Shortly after the recruitment of ezrin, actin appeared at the bleb membrane forming a cage-like structure underneath the bleb membrane (Fig. 2(C); Charras et al., 2006). The mechanism through which actin is nucleated during cortex regrowth remains unclear as the best characterized actin nucleators, mDia1 or Arp2/3, did not localize to blebs (Charras et al., 2006), though it has recently been suggested that the actin nucleator mDia2 may play a role in cortical actin nucleation (Eisenmann et al., 2007). Another possibility is that cortex regrowth may simply result from the elongation of small fragments of F-actin that remain attached to the membrane but that are too scarce to be detected above background by fluorescence microscopy. In this case fragments are distributed homogenously throughout the bleb membrane and the time for regrowth of an actin cortex should be independent of the size of the bleb. Once a continuous cortex has been assembled, actin turnover becomes very slow: no increase in total actin fluorescence can be detected during retraction and treatment with cytochalasin D, which caps the barbed end of actin filaments, does not prevent retraction of blebs that have already assembled an actin cortex (Charras et al., 2006). However, further experiments with FRAP or photoactivation will be needed to thoroughly assess actin turnover in retracting blebs. From a mechanical point of view, the polymerization of an actin cortex enables the bleb to better resist external forces and leads to a 5-fold increase in the apparent membrane rigidity (Charras et al., 2008).

After actin, actin-bundling proteins (α-actinin and coronin first and, later, fimbrin) and finally proteins of the contractile system (tropomyosin followed by myosin) were recruited. All of these proteins appeared as a continuous rim close to the cell membrane, except for myosin, which was localized to distinct foci (Fig. 6, arrow). Whereas ezrin appeared colocalized with the cell membrane, actin and actin-bundling proteins were displaced towards the cell centre compared to the cell membrane, and myosin appeared even further displaced to the cell centre (Charras et al., 2006). What exact role is played by actin-bundling proteins in cortex formation or mechanics remains to be determined. This will necessitate ultrastructural studies to determine whether they bundle filaments linearly as in stress fibres or link filaments to one another at the apices of the actin mesh.

image

Figure 6. Myosin powers contraction of the cell cortex and retraction of blebs. Filamin-deficient cells were transfected with Myosin regulatory light-chain GFP and with PH- PLCδ-mRFP and imaged with a spinning disk confocal microscope. Myosin regulatory light chain (in green) is localized to the cell cortex and present in distinct foci (arrow) in retracting blebs. The cell membrane is shown in red using the PH domain of PLCδ. Scale bar = 5 μm.

Download figure to PowerPoint

Myosin II motors power bleb retraction (Charras et al., 2006). An estimated 50 myosins are needed to start retraction and 5-fold more motors are recruited during the course of retraction (Charras et al., 2008). Contrary to expansion, during which the membrane is taut (Fig. 4(A)), during retraction, the actin cortex buckles (Figs 2(B) and 4(B)). The force needed to buckle it can be estimated from the wavelength of wrinkles and the cortex bending rigidity (Cerda & Mahadevan, 2003). Surprisingly, if the cortex is a static structure, too few myosins are present to buckle it (Charras et al., 2008). Fragilization of the cortex through depolymerization of actin cannot be an explanation as cytochalasin treatment experiments and GFP-actin fluorescence intensity measurements during retraction show that actin turnover is slow in retracting blebs (Charras et al., 2006; Charras et al., 2008). One possible explanation is that strands of actin within the cortex delaminate from one another because of stochastic detachment of actin-binding proteins from one filament increasing the effective length of the filaments and facilitating buckling. Retraction could then proceed through series of local crumpling events. Subsequent reattachment of the bundling protein would stabilize the buckled configuration. Another possibility is that actin is severed during retraction and one possible mechanism could be through aip1-coronin-cofilin mediated severing as coronin is present in retracting blebs (Brieher et al., 2006; Charras et al., 2006). Whether the bleb cortex integrates into the cell cortex after retraction is over or whether it is rapidly depolymerized and replaced by cortex that seamlessly links to the remainder of the cell is unknown at present.

Occasionally, secondary blebs appeared on top of other blebs or to the side of existing blebs (Figs. 4, 7). These appeared at two different phases during the bleb lifetime: during creation of the actin cortex and during retraction. What causes top secondary blebs is a source of speculation. They are quite common during cytokinesis when cortex contractility is highly upregulated (Fig. 7, white arrow); therefore, one possibility is that they could be the result of a second contractile event at the base of the existing bleb that forces more cytosol into the bleb and inflates regions of the membrane where the cortex has not fully re-formed or where membrane–cortex tethering is fragile. Side secondary blebs are presumably nucleated in a similar manner to primary blebs. First, secondary blebs could arise because retraction pressurizes the bleb cytosol (Charras et al., 2005) and this may be sufficient to detach the bleb membrane from the newly formed cortex. In cells where adhesion between the cortex and the cell membrane is compromised, for example, by microinjection of recombinant FERM domain, these types of secondary blebs are particularly prevalent (Charras et al., 2006). Second, secondary blebs could arise because of high stresses exerted by the myosin foci on the immature actin cortex that cause it to rupture locally.

image

Figure 7. Phase-contrast image of cytokinesis in a HeLa cell. The chromosomes (black arrows) have been segregated into the two daughter cells and a cytokinetic furrow has been assembled. Blebs appear at the pole of the cell. These have varied shapes and secondary, tertiary or quaternary blebs are common giving rise to very elongated bleb shapes (white arrow). Scale bar = 10 μm.

Download figure to PowerPoint

In circus movements, flow of cytosol from the rear to the front of the bleb is evident from the movement of granules within the bleb and may be due to the formation of a contractile cortex at the rear of the bleb that attempts to retract it. Though cortex formation has not been observed in Xenopus blastomeres, HeLa cells blocked in metaphase occasionally generate travelling blebs akin to circus movements and in these blebs, an asymmetric actomyosin cortex can clearly be observed: the leading edge of the bleb is devoid of a cytoskeleton, whereas the rear possesses an actomyosin cortex that retracts towards the cell body, forcing the cytosol to the leading edge (Charras et al., 2008). The bleb travels over long periods of time because one side of the bleb is always older than the other. However, what causes the initial asymmetry in cortex regrowth is unknown. Bleb volume does not vary during periods of travel presumably because pushing cytosol back into the cell body is more energetically costly than delaminating the membrane from the cortex.

Physiological functions of blebbing

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

Cell movement

Contrary to common belief, cells do not move exclusively through lamellipodial protrusion, especially in three-dimensional matrices. Polarized blebbing is utilized during motility in a number of cell types such as newt blastomeres (Kubota, 1981), Walker carcinosarcoma cells (Keller, 2000; Sroka et al., 2002), Dictyostelium (Langridge & Kay, 2006; Yoshida & Soldati, 2006; Traynor & Kay, 2007), Fundulus deep cells (Trinkaus, 1973; Fink, 2003), Zebrafish germ cells (Blaser et al., 2006) and neutrophils and tumour cells migrating within three-dimensional collagen matrices (Haston & Shields, 1984; Friedl et al., 2001; Sahai & Marshall, 2003). In some of these cell types, experiments clearly show that the cell membrane transiently separates from the actin cortex and bulges out (Fink, 2003; Blaser et al., 2006; Yoshida & Soldati, 2006), whereas in other cell types, the leading edge morphologically resembles a bleb but membrane delamination from the cortex has not been ascertained (Haston & Shields, 1984; Friedl & Wolf, 2003; Sahai & Marshall, 2003). Once bleb growth stalls, an actin cortex is re-formed and rather than the bleb retracting, the cell body contracts. Bleb formation is dependent upon the presence of a functional actomyosin cortex, and a number of the proteins involved in blebbing have been identified in these cell types.

For movement to arise, blebs must be polarized to the leading edge and we do not know how this occurs. Cells may become polarized following stimulation by a gradient of chemoattractant, as is the case in the migration of zebrafish germ cells (Blaser et al., 2006), and this could lead to an increase in contractility and fragilization of the cortex–membrane attachment at the leading edge. How the membrane–cortex attachment is downregulated at the leading edge is unknown and one may only speculate – based on the crucial role played by phospholipids and effector proteins in two-dimensional migration (Franca-Koh & Devreotes, 2004) – that mechanisms such as loss of PIP2 could be involved. In Walker carcinosarcoma cells, ezrin is polarized to the rear of the cell and this suggests a weakening of actin–membrane attachment at the front of the cell (Rossy et al., 2007). In migrating zebrafish germ cells, a localized increase in intra-cellular calcium was observed shortly before the emergence of a bleb (Blaser et al., 2006) and this may serve to coordinate detachment of the membrane from the cortex with generation of the contractile force needed for blebbing. In these cells, myosin is concentrated in two arcs, one at the rear of the cell and one at the front. This localization of myosin in cells is particularly well suited to create blebs in the context of rupture of the actin cortex and forward streaming of the cell mass. In Walker carcinosarcoma cells, myosin light chain concentrates at the front of the cell under the newly formed bleb and ezrin concentrates at the rear of the cell (Rossy et al., 2007), leading to increased contractility and fragilized membrane–cortex attachments at the leading edge.

During neutrophil migration, the rhoGTPases rac and rho play antagonistic roles and are active respectively at the lamellipodium and the uropod (Van Keymeulen et al., 2006). Since both blebbing and uropod contraction are rho-controlled phenomena, polarization in cells that utilize blebbing as a means of locomotion cannot be governed only by rhoGTPases. One possibility is that polarization may only be necessary for the initial movement forward. The initial polarization event might give rise to a bipolar concentration of myosin at the front and rear of the cell (as observed in zebrafish germ cells) and a concentration of membrane-actin linker proteins to the rear of the cell (as observed in Walker carcinosarcoma). This would cause a crack in the actin cortex to appear at the leading edge owing to increased myosin contractility and weakened membrane–cortex attachment (Paluch et al., 2006). The phenomenon could then persist because the new contractile cortex at the leading edge would always be younger, more fragile and less strongly tethered to the membrane than the uropod and therefore blebbing would preferentially occur at the front in response to each subsequent contraction of the cortex.

Locomotion using blebbing bears some striking similarities with amoeboid locomotion. In amoeba proteus, an actin cortex is attached to the entirety of the membrane except at the growing pseudo-podium. Growing pseudo-podia alternate periods of rapid growth (7 μm s−1) during which the membrane is delaminated from the actin cortex owing to the apparition of a crack and periods of stalling when an actin cortex is reassembled (Pomorski et al., 2007). In amoeba motility, uroid contraction gives rise to hydrostatic pressure, which powers cytosolic streaming and motility (Grebecka & Grebecki, 1981; Yanai et al., 1996). In higher organisms, the importance of streaming and hydrostatic pressure for amoeboid locomotion has neither been firmly established nor invalidated and this subject merits more attention.

During tumour cell movement in three-dimensional matrices, cells may employ amoeboid locomotion when they no longer express metalloproteinases (Friedl & Wolf, 2003; Sahai & Marshall, 2003). In this case, rather than burrowing through the extracellular matrix, the cells squeeze through it by utilizing blebs, though delamination of the membrane from the actin cortex has not been firmly established. The rapidly inflating bleb forces its way through gaps in the extracellular matrix and after reassembly of an actin cortex, the cell body contracts and moves forward. This gives rise to constriction rings in the actin cortex at sites where the bleb squeezed through the extracellular matrix. Constriction rings have been observed experimentally in neutrophils migrating through collagen gels (Haston & Shields, 1984; Friedl et al., 2001) and during Walker carcinosarcoma cell movement (Keller & Eggli, 1998). Blebs may also be used by metastatic tumour cells during endothelial transmigration (Voura et al., 1998). In this case, the cells exhibit polarized blebbing towards the endothelium, and the hydrostatic pressure in blebs is harnessed to push endothelial cells aside to allow the invading cell into the connective tissue. A similar mechanism has been reported in transmigrating neutrophils where these extend lobopodia, which appear morphologically similar to blebs, towards the endothelial layer (Wolosewick, 1984).

The physical phenomena at play during cell migration using polarized blebbing remain completely unstudied. Careful examination of actin and myosin dynamics will be required to understand whether the polarized blebs that appear during migration are the result of detachment of membrane from an intact cortex or due to a localized rupture of the actin cortex in response to bipolar distribution of myosin at the leading edge and the rear of the cell.

Cytokinesis

In dividing cells, blebs tend to form at the poles away from the cleavage furrow (Fishkind et al., 1991; Burton & Taylor, 1997; Boucrot & Kirchhausen, 2007). The role of blebbing during cytokinesis is unknown and it may just be an epiphenomenon of a weakening of the cortex or membrane–cortex attachment at the poles owing to depletion by recruitment of actin and membrane–actin linker proteins to the furrow region. However, some experimental data point towards a possible role for blebbing during cytokinesis. Depolymerization of the actin cortex in the pole region of dividing cells using cytochalasin D was found to inhibit cytokinesis (O'Connell et al., 2001). One possible explanation is that during cytokinesis the cell needs to increase its cortical surface area and that blebbing serves as a rapid way to create new cortex into the cell. This cortex may then flow towards the cleavage furrow over time (Bray & White, 1988; Wang et al., 1994) and subsequent blebs serve to make more cortex to drive this flow. If this were the case, depolymerization of actin at the poles would impede cortex creation and block cytokinesis. Speculating further, another possible role of blebbing in the context of cell division in a three-dimensional matrix is the generation of motive force to separate cells from one another via polarized blebbing movement.

Apoptosis

Blebs present during the execution phase of apoptosis are identical to those occurring in filamin-deficient cells or in dividing cells (Mills et al., 1999; Coleman et al., 2001). As during cytokinesis, the actual role of blebs during apoptosis is unclear. One possibility is that the convective flows of cytosol that give rise to blebs serve to fragment the nucleus and organelles in the apoptosing cell (Mills et al., 1999). However, blebs do not form a necessary part of cell death as cells exposed to staurosporine, a broad-spectrum blocker of blebbing (Charras et al., 2005), still undergo apoptosis (Nicotera et al., 1999). A more intriguing and systemic function that has some experimental support is that membrane proteins present at the surface of blebs liberated from apoptosing cells may function as chemoattractants to recruit macrophages to the site of cell death (Segundo et al., 1999).

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References

To date, blebbing has been the poor cousin of the actin-based protrusion family, perhaps because, contrary to lamellipodia or filopodia, its study necessitates the integration of molecular biology and biophysics to be fully understood. However, with the current focus on interdisciplinary science and the rediscovery of blebbing as a form of motility worthy of study, many of the remaining open questions in the bleb life cycle and its use in locomotion will hopefully get more well-deserved attention. Two aspects of blebbing are particularly deserving of attention. First, the role of blebbing in cell movement is being rediscovered and its full understanding will shed light on an alternate mode of locomotion used by cells during physiology. Second, blebbing offers us a window into cytomechanics and has the potential to enhance our understanding of membrane mechanics, membrane–cortex adhesion and contractile gels.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. A phenomenological definition of blebbing
  5. Motive force for bleb nucleation
  6. Blebbing is a locally controlled phenomenon
  7. Bleb growth
  8. Growing blebs may have an erythrocytic cytoskeleton
  9. Signal for regrowth of an actin cortex
  10. Regrowth of an actin cortex and retraction
  11. Physiological functions of blebbing
  12. Conclusion
  13. Acknowledgement
  14. References