Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 197 (3), 439-55

Nonpolarized Signaling Reveals Two Distinct Modes of 3D Cell Migration

Affiliations

Nonpolarized Signaling Reveals Two Distinct Modes of 3D Cell Migration

Ryan J Petrie et al. J Cell Biol.

Abstract

We search in this paper for context-specific modes of three-dimensional (3D) cell migration using imaging for phosphatidylinositol (3,4,5)-trisphosphate (PIP3) and active Rac1 and Cdc42 in primary fibroblasts migrating within different 3D environments. In 3D collagen, PIP3 and active Rac1 and Cdc42 were targeted to the leading edge, consistent with lamellipodia-based migration. In contrast, elongated cells migrating inside dermal explants and the cell-derived matrix (CDM) formed blunt, cylindrical protrusions, termed lobopodia, and Rac1, Cdc42, and PIP3 signaling was nonpolarized. Reducing RhoA, Rho-associated protein kinase (ROCK), or myosin II activity switched the cells to lamellipodia-based 3D migration. These modes of 3D migration were regulated by matrix physical properties. Specifically, experimentally modifying the elasticity of the CDM or collagen gels established that nonlinear elasticity supported lamellipodia-based migration, whereas linear elasticity switched cells to lobopodia-based migration. Thus, the relative polarization of intracellular signaling identifies two distinct modes of 3D cell migration governed intrinsically by RhoA, ROCK, and myosin II and extrinsically by the elastic behavior of the 3D extracellular matrix.

Figures

Figure 1.
Figure 1.
Lobopodia-based 3D migration occurs in the mammalian dermis. (A) A 3D reconstruction of a mouse ear dermal explant labeled with Alexa Fluor 633 (grayscale). Stratum corneum (SC), basal keratinocytes (BK), papillary dermis (PD), and reticular dermis (RD) are indicated. Sebaceous gland (SG) and hair follicle (HF) are outlined in gray. (B) Examples of ECM structures proximal (left) and distal (right) to the basal surface of a dermal explant labeled with Alexa Fluor 633. Images are from the same confocal stack, 9 µm (left) and 30 µm (right) from the basal surface. AC, adipocyte. (C) 3D reconstructions of lobopodia-bearing HFFs migrating in proximal and distal collagen; GFP-actin is shown in green, and second harmonic imaging of collagen appears in grayscale. Arrowheads indicate lateral blebs. (D) Active Rac1 is not targeted to the leading edge of HFFs migrating in the mammalian dermis. Rac1 activity was imaged in HFFs migrating in proximal or distal ECMs; active Rac1, representing the Fc image, was pseudocolored according to the 16-color scale shown to the right of the figure, and the explant was labeled with Alexa Fluor 633 (grayscale). All cells are oriented with the leading edge toward the right of the figure. Bars, 5 µm.
Figure 2.
Figure 2.
CDM and type I collagen support lobopodia- and lamellipodia-based 3D migration, respectively. (A) HFF-generated CDM has an aligned, fibrillar structure (top left), whereas polymerized 1.7 mg/ml type I collagen forms a random meshwork (bottom left). Both images are maximum projections of 30-µm confocal stacks. Collagen is remodeled by migrating HFFs (GFP, green) along the axis of migration (bottom right), whereas the organization of the CDM is unaffected by migrating HFFs (top right). The CDM was labeled with Alexa Fluor 633, and collagen was visualized by reflection microscopy. (B) Matrix stiffness (Young’s modulus [E]) of the indicated 3D matrices. (C) Strain-stiffening (Ehigh/Emed) behavior of the indicated 3D matrices. Ehigh/Emed > 1 indicates nonlinear elasticity, whereas Ehigh/Emed = 1 indicates linear elasticity (dashed red line). (D) Collagen and 2D CDM support lamellipodia-based migration, whereas 3D CDM triggers lobopodia-based motility. (top) Maximum projections of HFFs expressing GFP migrating inside the 3D CDM, on top of the 2D CDM, or inside type I collagen. LM, lamellipodium; LB, lobopodium. (bottom) The orthogonal views of the corresponding panel above, with the CDM (Alexa Fluor 633) and type I collagen (reflection microscopy) in red. Arrowheads indicate lateral blebs. (E) Cortactin is not enriched at the leading edge during lobopodia-based migration. HFFs migrating in the indicated ECM were fixed and immunostained for cortactin. Arrows indicate the local accumulation of cortactin at the leading edge. Bottom graphs correspond with their respective top images and represent the mean cortactin intensity measured from the leading edge (0 µm) toward the cell center. Each cortactin intensity profile was averaged from 13 cells, with three measurements per cell. Bars: (A and E) 5 µm; (D, top left and middle) 10 µm; (D, top right) 20 µm. All cells are oriented with the leading edge toward the right of the figure. Error bars show means ± SEM. *, P < 0.001 versus the dermal explant. a.u., arbitrary unit.
Figure 3.
Figure 3.
Nonpolarized PIP3, Rac1, and Cdc42 signaling during lobopodia-based 3D migration. (A) Maximally projected confocal stacks of HFFs expressing GFP-AktPH to detect PIP3, migrating on the 2D CDM, in the 3D CDM, or inside 1.7 mg/ml of 3D collagen. Images were pseudocolored according to the 16-color scale. Bars, 10 µm. (B) Mean PI of GFP-AktPH in HFFs migrating in the indicated ECM environments. (C–F) Localization of active Rac1 and Cdc42 away from the leading edge during lobopodia-based 3D migration inside the CDM. (C and E) Maximally projected confocal stacks of HFFs expressing Rac1 (C) or Cdc42 (E) biosensors migrating on the 2D CDM (left), in the 3D CDM (middle), or in 3D collagen (right). The Fc images, representing the total activity of each GTPase, were pseudocolored according to the 16-color scale. Arrowheads indicate regions of intracellular signaling. Bars: (3D CDM) 5 µm; (collagen and 2D CDM) 10 µm. (D and F) Mean PI of active Rac1 (D) and active Cdc42 (F) in HFFs migrating in the indicated ECM environments. All cells are oriented with the leading edge toward the top of the figure. Error bars show means ± SEM. *, P < 0.05 versus the 2D CDM; **, P < 0.05 versus the 3D CDM.
Figure 4.
Figure 4.
RhoA, ROCK, and myosin II are required for lobopodia-based 3D migration inside CDM. (A) A representative Western blot demonstrating the specificity of siRNA-mediated knockdown of Rac1, Cdc42, or RhoA. HFFs were transfected with the indicated siRNAs and lysed 72 h after transfection, and the lysates were blotted with the indicated antibodies. (B) Quantification of Western blots represented in A. (C) RhoA siRNA treatment switches HFFs to lamellipodia-based 3D migration in the CDM. The percentage of lobopodia-bearing HFFs migrating inside the CDM after the indicated treatments. (A and B) *, P < 0.001 versus the siGLO control. 48 h after siRNA treatment, HFFs were transfected with GFP-actin and imaged migrating in the 3D CDM. (D) Quantification of the velocity of siRNA-treated HFFs in the CDM 72–84 h after transfection. *, P < 0.05 versus the siGLO control. (E) ROCK dependence distinguishes HFF migration in collagen from the CDM. Quantification of the velocity of HFFs migrating in the CDM or 1.7 mg/ml collagen treated with FBS or FBS + 10 µM Y-27632. *, P < 0.03 versus the FBS control. (F) Quantification of HFF velocity in the CDM when treated with FBS or FBS + 25 µM blebbistatin. Blebbistatin treatment resulted in two subsets of HFFs, rapidly moving spread cells on top of the CDM (2D), and slowly moving elongated cells inside the CDM (3D). *, P < 0.001 versus the FBS control. (G and H) Myosin II is required for 3D lobopodia-based migration. (G) Cortactin localization in HFFs in the 3D CDM, either untreated or treated with 25 µM blebbistatin. The arrow indicates the local accumulation of cortactin at the leading edge. Bars, 5 µm. (H) The mean cortactin intensity profile, measured from the leading edge (0 µm) toward the cell center, of cells treated with FBS or FBS + 25 µM blebbistatin. Each cortactin intensity profile was averaged from 13 cells, with three measurements per cell. All cells are oriented with the leading edge toward the right of the figure. Error bars show means ± SEM. GAPDH, glyceraldehyde 3-phosphate dehydrogenase; a.u. arbitrary unit.
Figure 5.
Figure 5.
Lobopodia-based migration is distinct from cancer cell motility. (A) Amoeboid and mesenchymal HT1080 cells in the CDM. Phase-contrast image showing amoeboid (rounded) and mesenchymal (elongated) HT1080 cells in the CDM. (B) The percentage of amoeboid and mesenchymal HT1080 cells in the CDM (n = 1,001). (C and D) HT1080 cells do not form lobopodia in the 3D CDM. Round amoeboid cells lack matrix adhesions, whereas elongated mesenchymal cells have prominent lamellipodia and matrix adhesions. Images show maximally projected confocal stacks of amoeboid (C) and mesenchymal (D) HT1080 cells, expressing GFP-actin or YFP-paxillin, migrating inside the CDM (Alexa Fluor 633, grayscale). Arrowheads indicate matrix adhesions. (E) Matrix adhesions are present during both lamellipodia- and lobopodia-based 3D migration of elongated normal fibroblasts. Maximally projected confocal stacks of HFFs expressing YFP-paxillin or vinculin–tension sensor (TS) migrating inside 3D collagen or the 3D CDM. (F) Matrix adhesions contribute to both lamellipodia- and lobopodia-based migration of normal fibroblasts. Blocking αvβ3 and β1 integrins significantly decreased the velocity and directionality of HFFs migrating in 3D collagen and CDM, indicating that integrin-mediated adhesion contributed to the efficient directional migration of both cell populations. Quantification of the velocity (top) and directionality (bottom) of HFFs migrating on glass, CDM, or 3D collagen, either in media or in media with 100 µM cyclic RGD (cRGD; an αVβ3-blocking peptide) plus 500 µg/ml β1 integrin–blocking antibody (mAb13). *, P < 0.05 versus the untreated control. All cells are oriented with the leading edge toward the right of the figure. Bars: (A) 50 µm; (C–E) 5 µm.
Figure 6.
Figure 6.
Matrix elastic behavior governs the mode of normal 3D migration. (A) HFFs can remodel trypsinized CDM (middle), whereas the structure of the untreated (left) or the trypsinized (tryp) and cross-linked CDM (XL; right) is unaffected by HFF migration. CFP-Rac1 is shown in green, and Alexa Fluor 633–labeled matrix is in gray. Bars, 40 µm. (B) Cross-linked 1.7 mg/ml collagen (middle) and 8.6 mg/ml collagen (right) are not remodeled during HFF migration. GFP-actin is shown in green, and collagen is shown in gray (reflection). Bars, 10 µm. (C) Matrix stiffness (Young’s modulus [E]) of the indicated modified matrices. (D) Strain-stiffening (Ehigh/Emed) behavior of the indicated native and modified matrices. The dashed red line indicates a value of Ehigh/Emed corresponding to 1 (linear elasticity). (C and D) *, P < 0.05 versus trypsinized CDM; **, P < 0.05 versus cross-linked collagen. (E–G) Trypsinization and chemical cross-linking redistribute Rac1 activity and switch the mode of cell migration. (E) Maximally projected confocal stacks of HFFs expressing the Rac1 biosensor migrating in the trypsinized CDM (top), trypsinized and cross-linked CDMs (middle), or cross-linked collagen (bottom). The Fc images, representing the total activity of each GTPase, were pseudocolored according to the 16-color scale. Arrowheads indicate regions of intracellular signaling. Bars, 5 µm. (F) Mean PI of active Rac1 in HFFs migrating in the indicated ECM environments. *, P < 0.05 versus 3D trypsinized CDM. (G) The percentage of lobopodia-bearing HFFs migrating inside the CDM or collagen treated as indicated. *, P < 0.001 versus untreated CDM; **, P < 0.001 versus trypsinized CDM; ***, P < 0.007 versus untreated collagen. (H) 8.6 mg/ml collagen supports lamellipodia-based 3D cell migration. (top) A representative image of cortactin enrichment at the leading edge of cells migrating in 8.6 mg/ml collagen (arrow). (bottom) The mean cortactin intensity profile measured from 13 cells, with three measurements per cell. Cells are oriented with their leading edge toward the top right (A and B) or the right (E and H) of the figure. a.u., arbitrary unit.
Figure 7.
Figure 7.
Regulation of the mode and efficiency of normal 3D cell migration by extracellular soluble factors. (A) 10 ng/ml PDGF in glucose-deficient media specifically reduces cell velocity in the CDM versus collagen. Quantification of cell velocity in the CDM or collagen in response to DME with 25 mM glucose (control), 10% FBS in DME with 25 mM glucose (FBS), 10 ng/ml PDGF in DME with 25 mM glucose (PDGF), or 10 ng/ml PDGF in glucose-deficient DME (PDGF − glucose). *, P < 0.05 versus FBS. (B) Representative images of cortactin localization in HFFs in the 3D CDM treated as indicated and quantified in C. Arrows indicate the local accumulation of cortactin at the leading edge. Bottom graphs correspond with their respective top images and represent the mean cortactin intensity measured from the leading edge (0 µm) toward the cell center. Each cortactin intensity profile was averaged from 13 cells, with three measurements per cell. (C) Quantification of the percentage of cells without enrichment of cortactin at the leading edge in B. *, P < 0.001 versus FBS. (D) Treatment of HFFs with 10 ng/ml PDGF in glucose-deficient media reduces cellular RhoA activity by ∼50%. RhoA activities in HFFs treated on tissue-culture plastic as indicated were measured using G-LISA activation assays. Absorbance values were normalized to the relative amount of actin in each sample before comparison with the FBS treatment. *, P < 0.01 versus FBS; **, P < 0.05 versus PDGF. Cells are oriented with their leading edge toward the right of the figure. Bars, 5 µm. a.u., arbitrary unit.
Figure 8.
Figure 8.
Dimensionality, matrix elastic behavior, and RhoA–ROCK–myosin II govern the mode of normal cell migration. (A) During cell migration on flat or fibrillar 2D surfaces, lamellipodia form the leading edge of adherent fibroblasts. The presence of lamellipodia was confirmed by a prominent rim of F-actin and cortactin, along with active Rac1 and Cdc42 with PIP3 at the leading edge. In a 3D ECM, HFFs can use either lobopodia- or lamellipodia-based 3D migration. (B) The choice to migrate using lobopodia- or lamellipodia-based migration can be represented by a decision tree consisting of three questions: what is the dimensionality of the matrix, what is the level of RhoA activity, and is the 3D matrix linearly elastic? Lobopodia-based 3D migration predominates in linear elastic ECM and may use high actomyosin contraction downstream of RhoA–ROCK–myosin II to increase intracellular pressure and push the leading edge forward in combination with integrin-mediated adhesion. When RhoA–ROCK–myosin II activity is diminished, either in nonlinear ECM or through treatment of cells with RhoA siRNA or specific inhibitors, cells form lamellipodia with actin polymerization to advance the leading edge. Both modes of migration involve elongated cells that form 3D matrix adhesions, but the distribution of active Rac1, Cdc42, and PIP3 distinguishes the two modes.

Comment in

  • Switching to 3D.
    Baumann K. Baumann K. Nat Rev Mol Cell Biol. 2012 May 23;13(6):338. doi: 10.1038/nrm3365. Nat Rev Mol Cell Biol. 2012. PMID: 22617463 No abstract available.

Similar articles

See all similar articles

Cited by 138 articles

See all "Cited by" articles

References

    1. Ahlfors J.-E.W., Billiar K.L. 2007. Biomechanical and biochemical characteristics of a human fibroblast-produced and remodeled matrix. Biomaterials. 28:2183–2191 10.1016/j.biomaterials.2006.12.030 - DOI - PubMed
    1. Allio A.E., McKeown-Longo P.J. 1988. Extracellular matrix assembly of cell-derived and plasma-derived fibronectins by substrate-attached fibroblasts. J. Cell. Physiol. 135:459–466 10.1002/jcp.1041350313 - DOI - PubMed
    1. Bard J.B., Hay E.D. 1975. The behavior of fibroblasts from the developing avian cornea. Morphology and movement in situ and in vitro. J. Cell Biol. 67:400–418 10.1083/jcb.67.2.400 - DOI - PMC - PubMed
    1. Bass M.D., Roach K.A., Morgan M.R., Mostafavi-Pour Z., Schoen T., Muramatsu T., Mayer U., Ballestrem C., Spatz J.P., Humphries M.J. 2007. Syndecan-4–dependent Rac1 regulation determines directional migration in response to the extracellular matrix. J. Cell Biol. 177:527–538 10.1083/jcb.200610076 - DOI - PMC - PubMed
    1. Bear J.E., Svitkina T.M., Krause M., Schafer D.A., Loureiro J.J., Strasser G.A., Maly I.V., Chaga O.Y., Cooper J.A., Borisy G.G., Gertler F.B. 2002. Antagonism between Ena/VASP proteins and actin filament capping regulates fibroblast motility. Cell. 109:509–521 10.1016/S0092-8674(02)00731-6 - DOI - PubMed

Publication types

MeSH terms

Feedback