Collective Cell Migration in 3D Epithelial Wound Healing - ACS Nano

Feb 13, 2019 - ... The Pennsylvania State University , University Park , Pennsylvania 16802 ... Harvard−MIT Division of Health Sciences and Technolo...
0 downloads 0 Views 9MB Size
Download PDF
Collective Cell Migration in 3D Epithelial Wound Healing Yuan Xiao,†,‡ Reza Riahi,‡,§ Peter Torab,∥ Donna D. Zhang,⊥ and Pak Kin Wong*,†,‡,∥,#

Downloaded via WASHINGTON UNIV on February 14, 2019 at 05:41:27 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

Department of Biomedical Engineering and ∥Department of Mechanical Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802, United States ‡ Department of Aerospace and Mechanical Engineering and ⊥Department of Pharmacology and Toxicology, The University of Arizona, Tucson, Arizona 85721, United States § Harvard−MIT Division of Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States # Department of Surgery, The Pennsylvania State University, Hershey, Pennsylvania 17033, United States S Supporting Information *

ABSTRACT: Collective cell migration plays a pivotal role in development, wound healing, and metastasis, but little is known about the mechanisms and coordination of cell migration in 3D microenvironments. Here, we demonstrate a 3D wound healing assay by photothermal ablation for investigating collective cell migration in epithelial tissue structures. The nanoparticle-mediated photothermal technique creates local hyperthermia for selective cell ablation and induces collective cell migration of 3D tissue structures. By incorporating dynamic single cell gene expression analysis, live cell actin staining, and particle image velocimetry, we show that the wound healing response consists of 3D vortex motion moving toward the wound followed by the formation of multicellular actin bundles and leader cells with active actin-based protrusions. Inhibition of ROCK signaling disrupts the multicellular actin bundle and enhances the formation of leader cells at the leading edge. Furthermore, single cell gene expression analysis, pharmacological perturbation, and RNA interference reveal that Notch1-Dll4 signaling negatively regulates the formation of multicellular actin bundles and leader cells. Taken together, our study demonstrates a platform for investigating 3D collective cell migration and underscores the essential roles of ROCK and Notch1-Dll4 signaling in regulating 3D epithelial wound healing. KEYWORDS: 3D, wound healing, nanorod, spheroid, Notch signaling, leader cell

C

Collective cell migration and the formation of leader cells are often studied by a “wound healing” assay, in which a cellfree region is created by physically scraping the cell monolayer or blocking the region before cell seeding.16,17 The release of contact inhibition is sufficient to trigger collective cell migration and the formation of leader cells.18,19 Nevertheless, 2D monolayer culture does not provide a physiologically relevant context for determining the wound healing response of 3D epithelial tissues. The tissue architecture, extracellular matrix, and damage tissues can all contribute to the 3D wound healing process. The wound closure behaviors in 3D microenvironments were primarily investigated by laser ablation of in vivo models, such as zebra fish and Drosophila embryos.20−22 In these assays, cells are ablated at the outer surface of the embryos with a pulsed laser, and the wound healing process finishes typically in 2−4 h. Recently,

ollective cell migration plays important roles in various pathological and physiological processes.1,2 For instance, epithelial cells migrate collectively to restore the tissue integrity after injury. Mounting evidence suggests that the invasion and metastasis of malignant tumors also utilize the same cooperative mechanisms.3−5 In contrast to single cell migration, collective migration of epithelial tissues involves a cohesive group of cells with E-cadherin-mediated mechanical feedback, which coordinates the motion of the multicellular assembly.6 In the purse-string mechanism, an intercellular actomyosin ring is formed among cells at the leading edge, which draws the wound together.7 Rhodependent leader cells, which display active actin lamellipodia, also coordinate epithelial and endothelial wound healing.8−10 The leader cells are typically defined by their distinct morphologies and ruffling actin-based protrusion. Molecularly distinct leader cells with an invasive phenotype have also been shown to lead collective invasion in cancer.11−15 Nevertheless, the mechanisms and regulation of collective cell migration in 3D microenvironments remain poorly understood. © XXXX American Chemical Society

Received: August 19, 2018 Accepted: February 8, 2019

A

DOI: 10.1021/acsnano.8b06305 ACS Nano XXXX, XXX, XXX−XXX

Article

www.acsnano.org

Cite This: ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 1. Photothermal ablation induces 3D wound healing response. (A) Schematics of the 3D wound healing assay. Gold nanorods are internalized by cells to facilitate laser ablation. A near-infrared laser is irradiated to induce local hyperthermia for ablating desired regions of the spheroid. (B) Bright-field images of a 3D spheroid before and after laser irradiation (white arrow). The lesion (yellow asterisk) appeared dark after laser irradiation. Scale bars, 50 μm. (C) Photothermal ablation induces a wound healing response with 3D collective cell migration. Blue arrows indicate the direction of motion determined by particle image velocimetry. Yellow asterisk indicates the ablated area. Scale bars, 50 μm. (D) Cell trajectories of individual cells in controlled and ablated spheroids for 8 h. (E) Overlay images illustrate dynamic gene expression analysis in 3D spheroids. Fluorescence (red) indicates β-actin mRNA detected by a double-stranded locked nucleic acid (dsLNA) biosensor. Scale bars, 50 μm. (F) Schematics of the 3D cell motion to expel the lesion from the spheroid. (G) 3D reconstruction of the actin structure in an ablated spheroid. XY, XZ, and YZ cross sections of the 3D spheroid are shown. Scale bar, 50 μm. Images are representative of five independent experiments.

mechanical wounding with a microdissection knife is shown to study fibrous tissue repair in mammalian cells.23 Wounding 3D tissues in a controllable manner and monitoring cell dynamics, such as gene expression, cytoskeletal organization, and migratory behaviors, remain challenging and hinder investigations seeking to understand collective cell migration in 3D epithelial wound healing. Consequently, the mechanisms of collective cell migration, such as the formation of intercellular actomyosin rings and leader cells, in 3D wound healing remain largely unknown despite its physiological significance. In this study, we present a 3D wound healing assay by ablating multicellular spheroids to study collective cell migration of mammalian cells. We develop the 3D wound healing assay by adapting the photothermal ablation technique with nanoparticle-mediated localized hyperthermia, which is applied in photothermal therapy of cancer,24 to induce collective cell migration in epithelial tissues. We monitor the 3D wound healing response, including actin dynamics and leader cell formation, using single cell biosensors, live cell actin

labeling, and time-lapse confocal imaging. The modes of 3D motion were studied using particle image velocimetry (PIV). Dynamic gene expression analysis, RNA interference, and pharmacological perturbations were applied to study the roles of ROCK and Notch1-Dll4 signaling in the regulation of 3D collective cell migration.

RESULTS Photothermal Ablation Induces 3D Wound Healing Response. We developed a 3D wound healing assay by selective photothermal ablation of epithelial cells in multicellular spheroids (Figure 1A,B). Photothermal ablation of cells has previously been demonstrated in 2D monolayer cultures,15 lung tissues ex vivo,25 and microvascular networks.26 In this assay, gold nanorods were first internalized into the cells before self-assembly of 3D spheroids on extracellular matrices.27 The absorption of near-infrared laser by the gold nanorods created a local hyperthermia for selective cell B

DOI: 10.1021/acsnano.8b06305 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 2. Actin dynamics and collective cell migration in 3D wound healing. (A) Bright-field and fluorescence images of a 3D spheroid. Live cell fluorescence actin staining for characterizing multicellular actin bundles and cortical actin. Scale bar, 50 μm. (B) Actin dynamics in a 3D spheroid. The spheroid was irradiated at the beginning of the experiment for five seconds. Yellow asterisk indicates the ablated area. Scale bar, 50 μm. (C) Evolution of actin structures at the leading edge (area immediately adjacent to the wound). The cross-sectional views are shown to illustrate the actin structures. Yellow arrows indicate the ends of the multicellular actin bundle at 3 h after ablation. Leader cells with active actin-based protrusions (yellow arrowheads) appeared in 6 and 9 h after ablation. Scale bar, 20 μm. (D) Actin-based protrusion structures in leader cells after 6 h of ablation. Yellow dashed lines outline the leading front of the cell, and yellow arrowheads indicate actinbased protrusions. Scale bars, 10 μm. (E) Velocity distributions in XY, XZ, and YZ cross sections of an ablated spheroid. Insets on the left indicate the cross-sectional planes in the spheroid. Velocity vectors are superimposed on the mean displacement heat maps. Scale bars, 50 μm. (F) Schematics of the expulsion and retraction process. Upon ablation, cells migrate collectively to create a vortex motion to push away the lesion. Then, multicellular actin bundles and leader cells are formed near the leading edge. (G) Fluorescence images illustrate expulsion and retraction of the lesion (yellow circles) by the spheroid. Scale bar, 50 μm. Data and images are representative of five independent experiments.

Accurately monitoring cell dynamics, such as the cytoskeletal organization, migratory behaviors, and gene expression profiles, often makes studying 3D collective cell migration difficult. Therefore, we evaluated the applicability of the wound healing assay for studying the cell dynamics during collective cell migration in 3D spheroids. Using confocal microscopy with 3D reconstruction, cell motion in the 3D spheroid was characterized by PIV analysis and real-time tracking of cell trajectories (Figure 1C,D). Similar to 2D wound healing assays, the ablation process was sufficient to enhance the mobility and displacement of cells. PIV analysis suggested that neighboring cells exhibited coherent rates and direction of motion (Supporting Information Figure S2). A control experiment without gold nanorods was performed to evaluate the effect of gold nanorods on cell migration (Supporting Information Figure S3). To map the gene expression profile of the 3D spheroid, a double-stranded locked nucleic acid (dsLNA) biosensor was transfected into the cells before the formation of the spheroid.28−30 For instance, Figure 1E,F shows detection of β-actin mRNA during 3D collective cell

ablation. In the experiment, a lesion was created by focusing the laser at the surface of the spheroid (Figure 1A,B). The beam size of the infrared laser was approximately 10 μm, and the wound size was controlled by the irradiation duration. A short irradiation time resulted in localized heating, whereas a long duration increased the area of ablation due to heat diffusion. Heat transfer analysis illustrates the effects of the irradiation duration on wound size (Supporting Information Figure S1A). The effects of the irradiation duration on wound size of the 3D spheroid was also tested experimentally (Supporting Information Figure S1B−D). Laser exposure for 1 s created a lesion about 10 μm in size, which was similar to the size of a single cell. For a long exposure time (e.g., 15 s), a large wound over 100 μm in diameter was created, which was comparable to the size of a spheroid. The local hyperthermia induced by laser irradiation agreed with the heat transfer analysis and our previous single cell gene expression mapping study.25 Unless specified otherwise, we applied 5 s laser irradiation to create lesions of ∼50 μm in this study. C

DOI: 10.1021/acsnano.8b06305 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 3. ROCK inhibition enhances leader cell formation and modulates the modes of migration. (A) Live cell actin staining showing the evolution of actin dynamics in a Y-27632-treated spheroid. Yellow asterisk indicates the ablated area. Arrowheads indicate leader cells with active protrusions at both the leading edge and the outer edge of the spheroid. (B) Number of active protrusions per spheroid with and without Y-27632. Y-27632 increases cells with active actin-based protrusion (n = 3, unpaired, two-tailed Student’s t-test, *p < 0.05). (C) Wound boundary exhibited an irregular, jagged edge without the formation of multicellular actin bundles at the initial phase of collective cell migration (1−3 h). Active protrusions (arrowheads) were observed at the leading edge after 6 h of ablation. Scale bar, 25 μm. (D) Displacement heat maps of a Y-27632-treated spheroid. Velocity vectors are superimposed on the displacement heat maps. A decrease in displacement compared to control was observed in all phases of migration. (E) Mean velocity, (F) persistence, and (G) acceleration of migration for the whole spheroid and cells near the leading edge. The leading edge is defined as the cells immediately adjacent to the wound. The persistence of migration is the ratio between the net displacement and total path traveled, which indicates the level of consistence in migration direction. The acceleration represents the convective acceleration of the motion (n = 3, one-way ANOVA followed by Tukey’s post hoc test, *p < 0.05, ***p < 0.001; ns, not significant).

migration, which appeared to reorganize the cell structure and push the necrotic tissue away from the spheroid. Live cell actin labeling was also applied to monitor actin-rich structures in real time. Confocal z-stack reconstruction revealed 3D actin architecture in the spheroid (Figure 1G and Supporting Information Figure S1C,D). These results support the use of the photothermal wound healing assay for studying collective cell migration in 3D microenvironments. Formation of Multicellular Actin Cables and Leader Cell Formation. We studied the actin dynamics of the

spheroid during 3D wound healing because actin cytoskeletal structures are essential components of intercellular actomyosin rings and leader cells. Without wounding, the outer surface of the spheroid exhibited circumferential actin filaments (Figure 2A). Real-time monitoring of the actin dynamics revealed a time-dependent transformation of the actin structures in the leading edge near the wound (Figure 2B and Supporting Information Movies 1, 2, and 3). In this study, the leading edge is defined as cells immediately adjacent to the wound. At the beginning of the experiment (1−3 h), the integrity of the outer D

DOI: 10.1021/acsnano.8b06305 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 4. Dll4 mRNA is dynamically regulated in 3D spheroids. (A) Schematic of the dsLNA biosensor for dynamic single cell gene expression analysis. With a target mRNA, the quencher probe is physically displaced from the donor probe, allowing the fluorophore to fluoresce. The dsLNA probes were transfected into cells before spheroid self-assembly and photothermal ablation. (B) Fluorescence and bright-field images showing analysis of β-actin mRNA in a 3D spheroid. Scale bar, 50 μm. Expression profiles of (C) β-actin and (D) Dll4 mRNA in the leading edge and rest of the spheroid (nonleading edge) after ablation. (E) Spatial distribution of β-actin and Dll4 mRNA expression 0 and 2 h or 0 and 4 h after ablation. Yellow dashed lines indicate the wound boundaries, and yellow asterisks illustrate the ablated area. Scale bars, 50 μm. Images are representative of three independent experiments.

the cell motion was attenuated. The velocity of the whole spheroid decreased significantly with a 2-fold reduction when compared to the first phase (Figure 3). In particular, the velocity decreased from 0.85 to 0.38 μm/h in the first phase of wound closure. A 4-fold reduction in velocity was observed near the leading edge (Figure 3). The velocity decreased from 1.17 to 0.32 μm/h (Figure 3). The velocity increased again during the third phase (7−9 h) when leader cells formed. Unlike the coordinated 3D vortex motion observed in the first phase, the velocity fields in the third phase were less coherent and displayed a dispersion pattern that moved in various directions (Supporting Information Figure S4). Interestingly, the leader cells also appeared to retract the lesion into the spheroid (Figure 2F,G). ROCK Inhibition Enhances Leader Cell Formation and Modulates the Modes of Migration. In 2D wound healing, Rho-mediated actomyosin-based contractility regulates both purse-string closure and leader cell mechanisms.8−10 We, therefore, studied the roles of a Rho kinase, ROCK, in the formation of multicellular actin cables and leader cells. We applied a ROCK inhibitor, Y-27632, which specifically inhibits myosin II phosphorylation and actin filament levels.31 The overall level of F-actin was reduced by the Y-27632 treatment. Leader cells with actin-rich protrusions increased significantly at the leading edge of Y-27632-treated spheroids (Figure 3A,B and Supporting Information Movie 4). A large number of actin-based protrusions were also observed in the outer edge of the spheroid, which were not observed in the control case. Furthermore, the multicellular actin bundle was attenuated by Y-27632, and the wound of the spheroid displayed a rough boundary throughout the experiment (Figure 3C). We examined the velocity profiles of cells in Y-27632-treated spheroids (Figure 3D). Overall, the velocity decreased in all

surface of the spheroid was interrupted, leaving an irregular wound with a jagged edge (Figure 2C). The leading edge then displayed active remodeling and formed a smooth boundary. At 3−6 h, a multicellular actin bundle was observed at the leading edge. Leader cells with actin-rich protrusions then emerged between 6 and 9 h while the multicellular actin bundles near the leading edge broke into multiple punctate foci. The actin-rich protrusions at the wound edge actively probed into the open area. Close examination of the protrusion structures revealed diverse actin dynamics of leader cells. Cells in the leading edge exhibited large, cone-shaped structures or small, finger-like protrusions (Figure 2D). The length scale of the structures spanned from over 10 μm (the size of a cell) to submicron scale. These observations reveal that multicellular actin cables and leader cells are involved in the 3D wound healing response. Cell Migratory Behaviors in 3D Wound Healing. We also investigated the migratory behaviors of cells during the 3D wound healing process. Time-lapse confocal reconstruction along with PIV analysis was performed to extract the 3D velocity fields. Velocity maps were generated for visualizing the 3D cell motion in different cross-sectional views and Z planes (Figure 2E and Supporting Information Figure S4). In agreement with the actin dynamics, which correlated with the formation of multicellular actin bundles and leader cells, time-dependent velocity profiles were observed in the 3D spheroids. In the first phase (1−3 h), cells moved coherently to form a 3D vortex motion that moved toward the ablated area (Figure 2E). The vortex motion was observed in both XY and XZ planes, and the velocity was highest in the XY plane near the ablated area. The cells collectively expelled the necrotic tissue away from the spheroid (Figure 2F,G). In the second phase (4−6 h), when the multicellular actin bundle emerged, E

DOI: 10.1021/acsnano.8b06305 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 5. Notch1-Dll4 signaling negatively regulates the formation of multicellular actin bundles and leader cells during 3D wound healing. (A) Immunostaining of Dll4 and activated Notch1 in ablated spheroids. The spheroids were treated with Jag1 peptide, Notch1 siRNA, or Dll4 siRNA. Images are representative of five different experiments. Scale bars, 25 μm. (B) Confocal cross-sectional images for illustrating the multicellular actin bundles near the ablated area. The smooth borders lined with the multicellular actin bundles were disrupted in Jag1treated spheroids. Images were taken 3 h after ablation. Yellow arrows indicate the ends of the bundles. Scale bars, 25 μm. (C) Representative live cell actin staining showing F-actin based protrusions (arrowheads) in the leading edge and the outer region of the spheroids at 6 h after ablation. The lower panels show the zoomed-in view of the leading edge of the spheroids. Scale bars, 50 μm (top panel) and 25 μm (bottom panel). Images are representative of five independent experiments. (D) Heat maps of cell velocity in spheroids with different treatment groups during 3D collective cell migration. Velocity vectors are superimposed on the heat maps. Scale bar, 50 μm. Data are representative of five independent experiments. (E) Trajectories of individual cells in spheroids treated with Jag1 peptide, Notch1 siRNA, and Dll4 siRNA. (F) Relative speed of cells in the leading edge (n = 10 for each case, one-way ANOVA followed by Tukey’s post hoc test, *p < 0.05, ***p < 0.001; ns, not significant).

three phases compared to the control. Y-27632 attenuated the 3D vortex motion in the first phase and generated a scatter velocity profile compared to control. Despite the decrease in velocity in the whole spheroid, the Y-27632-treated spheroid maintained mobility near the leading edge during the second phase (Figure 3E). We also estimated the persistence of migration of the spheroid (Figure 3F and Supporting Information Figure S5). In this study, the persistence of migration was defined as the ratio between the net displacement and total path traveled, indicating the level of directional persistence (i.e., moving in a consistent direction in contrast to random motion).32 Although the overall velocity was reduced, the persistence of migration at the leading edge was mostly preserved in Y-27632-treated spheroids. Furthermore, we calculated the acceleration of the velocity field, which was significantly reduced by the Y-27632 treatment (Figure 3G). The effects of ROCK inhibition on leader cell formation and

the cell mobility were also confirmed by siRNA (Supporting Information Figure S6). Dll4 mRNA Is Dynamically Regulated in 3D Spheroids. We then investigated the regulation of leader cells in 3D collective cell migration. Dll4 expression has been shown to upregulate in leader cells during 2D wound healing.15 Thus, we monitored the distributions of Dll4 mRNA using the dsLNA probe, which allows dynamic gene expression analysis in 3D spheroids (Figure 4A and Supporting Information Figure S7). In the dsLNA assay, the probes are transfected into the cells before the formation of 3D spheroids and photothermal ablation. Using the dsLNA, we quantified the level of β-actin mRNA in the 3D spheroid (Figure 4B). An increase in β-actin mRNA was observed near the leading edge, whereas the rest of the cells (i.e., nonleading edge) maintained a stable level of βactin mRNA (Figure 4C). The β-actin mRNA expression of cells near the leading edge increased transiently between 2 and 4 h after injury. The peak time occurred during the transition F

DOI: 10.1021/acsnano.8b06305 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano between the first and second phase of collective cell migration when the multicellular actin bundle formed. The expression returned to the basal level in the third phase when leader cells emerged. In contrast, the expression of Dll4 mRNA displayed a distinct pattern (Figure 4D). Unlike the β-actin mRNA, cells near the leading edge maintained a relatively low level of Dll4 mRNA expression throughout the wound healing process. In contrast, cells in the inner region (i.e., nonleading edge) of the spheroid upregulated Dll4 mRNA after the first hour of ablation and maintained a high level of expression during the wound healing response (Figure 4E). Notch1-Dll4 Signaling Negatively Regulates the Formation of Multicellular Actin Bundles and Leader Cells during 3D Wound Healing. Our results suggest Dll4 mRNA is dynamically regulated in spheroids during 3D collective cell migration. As the Notch1 receptor is activated by Dll4,33,34 we questioned the function of Notch1-Dll4 signaling in 3D wound healing. RNA interference was applied to knockdown Notch1 and Dll4 expressions, and Jag1 peptides were applied to enhance the Notch activity. Immunostaining characterized the expressions of Dll4 and Notch1 in 3D spheroids (Figure 5A and Supporting Information Figures S8 and S9). Jag1 treatment increased the expression of both Notch1 and Dll4 in the spheroids. In contrast, Notch1 and Dll4 siRNA individually were able to reduce the levels of both genes, suggesting transcriptional feedback (i.e., the expressions reinforce each other) of the ligand−receptor pair.33,34 We then examined the effects of Notch perturbation on the formation of multicellular actin bundles and leader cells (Supporting Information Movies 3, 4, and 5 and Figure S10). The enhancement of Notch1 expression by Jag1 inhibited the formation of multicellular actin bundles, and a rough boundary was observed at the leading edge (Figure 5B). Multicellular actin bundles, in contrast, were observed in spheroids treated with Notch1 siRNA or Dll4 siRNA. Furthermore, Jag1 inhibited the formation of leader cells in the third phase (Figure 5C), whereas the inhibition of Notch1-Dll4 signaling by siRNA led to an increase in leader cells with actin-rich protrusions and lesion retraction. These results indicate a negative role of Notch1-Dll4 signaling in the formation of multicellular bundles and leader cells. Examination of the velocity profiles suggests Notch1 activation by Jag1 peptide reduced the velocity and displacement significantly, whereas Notch1/Dll4 siRNA treatment enhanced the cell motion near the leading region (Figure 5D−F). These results collectively support that Notch1-Dll4 signaling negatively regulates the formation of multicellular actin bundles and leader cells during 3D wound healing.

dynamic gene expression analysis in 3D wound healing. We also characterized multicellular actin bundles and leader cells with distinct protrusion structures during 3D wound healing. These results demonstrate the feasibility of investigating 3D collective cell migration with the photothermal ablation assay. Our study highlights several characteristics of 3D epithelial wound healing that are not observed in 2D wound healing. First, we identified large 3D vortex motion immediately after laser ablation. Cell migration in 3D extracellular matrix typically involves diverse protrusion structures, such as lamellipodia enriched in F-actin and actin-binding proteins, lobopodia driven by intracellular pressure, and amoeboid migration with blebs and actin-enriched leading edges.35 Cells can also switch between these migratory mechanisms, depending on the properties of the extracellular matrix.36 In contrast, we observed coherent 3D vortex motion within epithelial tissues with neither apparent actin-based protrusion nor interaction with the extracellular matrix. Second, 2D wound healing is often dominated by either purse-string or lamellipodia-based cell crawling mechanisms near the wound edge. Traction force microscopy suggests that 2D wound closure is initially dominated by cell crawling followed by actomyosin-based contraction.37 In the 3D wound healing assay, the formation of multicellular actin bundle unexpectedly occurs before the formation of leader cells. Third, we observed distinct characteristics of leader cells in 3D spheroids compared to 2D monolayers. The leader cells exhibited diverse actin-based structures, such as cone-shaped structures and finger-like structures, which are different from the lamellipodial structures in 2D cell migration. Furthermore, Dll4 expression was also downregulated in the leading edge and showed a negative effect on the formation of leader cells. This is in contrast to the Dll4-rich leader cells observed in the 2D wound healing assay,15 suggesting additional signal pathways and mechanisms may be involved in the regulation of 3D collective cell migration. Our results shed light on the mechanistic regulation of 3D collective cell migration. Actomyosin-based contractility is previously shown to inhibit the formation of leader cells by pharmacological perturbation and two-photon photoablation.8−10 Our recent study has revealed the mechanosensitivity of the Notch signaling in mouse retina and embryoid body models, suggesting a complex interaction between ROCK and Notch1-Dll4 signaling.15 In agreement, our data show ROCK inhibition by Y-27632 or siRNA disrupts the formation of multicellular actin bundles and promotes the formation of leader cells. We also demonstrate Notch1-Dll4 signaling negatively regulates 3D collective migration by inhibiting multicellular actin bundles and leader cell formation. Inhibition of ROCK and Notch1-Dll4 signaling by Notch1 and Dll4 siRNA enhances the formation of leader cells in 3D spheroids. In contrast, Jag1 attenuates the formation of leader cells, suggesting a negative role of Notch1-Dll4 signaling in leader cell formation. This is in contrast to leader cells in 2D wound healing, which has an elevated level of Dll4 mRNA and can be enhanced by Dll4 upregulation.15 These results collectively underscore that different or additional molecular regulation processes are involved in 3D collective cell migration. Future investigations should identify the regulation of these migratory mechanisms and decipher the differences between 2D and 3D collective cell migration.

DISCUSSION In this study, we demonstrate a wound healing assay for investigating 3D collective cell migration. Collective cell migration is typically studied in 2D monolayers, and existing 3D migration assays often focus on the behaviors of individual cells, such as fibroblasts and cancer cells, in 3D matrices. In contrast, the photothermal ablation assay allows us to study 3D collective cell migration within epithelial tissue structures of mammalian cells and investigate the migratory mechanisms that are not captured in individual cell migration. When compared to mechanical wounding,23 the photothermal ablation technique precisely controls the location and size of the wound in 3D spheroids. By incorporating the dsLNA biosensor and time-lapse confocal imaging, we demonstrate G

DOI: 10.1021/acsnano.8b06305 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Jag1 (188−204) peptide (Genscript) was used at 20 μM. Y-27632 (Calbiochem) was used at 20 μM. The gold nanorods (10−1064) were purchased from Nanopartz. dsLNA probes were synthesized by Integrated DNA Technologies. Matrigel and Lipofectamine 3000 was purchased from Thermo Fisher Scientific. All other chemicals were obtained from Fisher Scientific, unless otherwise noted. Immunostaining. Immunofluorescence staining was carried out as detailed previously. The spheroids were stained with goat anti-Dll4 antibodies (1:50, Santa Cruz Biotechnology) and rabbit antiactivated Notch antibodies (1:100, Abcam) followed by Alexa Fluor 555 and 647 labeled secondary antibodies (Thermo Fisher Scientific). Nuclei were counterstained with SYTOX Green (Life Technology). Immunofluorescence was detected by confocal microscopy. Statistical Analysis. Statistical analyses were performed with GraphPad Prism 5 software. One-way analysis of variance and Tukey’s post hoc test were used to compare between groups. Data represent mean ± SEM. Statistical significance was assigned as follows: ns, not significant, *p < 0.05, **p < 0.01, or ***p < 0.001.

CONCLUSION In conclusion, this study demonstrates a photothermal ablation technique for investigating 3D collective cell migration. The technique identifies characteristics in 3D collective cell migration that are distinct from 2D migration and reveals the regulatory role of ROCK and Notch1-Dll4 signaling in 3D epithelial wound healing. MATERIALS AND METHODS Formation and Dynamic Monitoring of Multicellular Spheroids. The 3D spheroid model is based on an established protocol with modification.27 Briefly, a glass-bottomed polydimethylsiloxane microwell with a diameter of 15 mm was coated with a thin layer of Matrigel (30 μL). Epithelial MCF-7 cells were cultured in Dulbecco’s modified Eagle’s medium and incubated with 1010/mL gold nanorods with mercaptoundecyltrimethylammonnium bromide coating for 12 h to facilitate photothermal ablation. The concentration was optimized to minimize any potential toxicity of the gold nanoparticles. The cells were washed with PBS three times to remove excess gold nanorods that were not internalized. To form multicellular spheroids, 1.2 × 105 cells were plated on top of the Matrigel and cultured at 37 °C and 5% CO2 for 24 h. To study the actin dynamics in 3D spheroids and identify the wound boundary after ablation, cells were stained with 1 μM live cell actin staining probe SiR-actin (SC001; Cytoskeleton) for 4 h. To perform dynamic single cell analysis, dsLNA probes targeting β-actin or Dll4 mRNA were synthesized as described previously.18,20 The dsLNA probes were assembled by mixing the fluorophore and quencher sequences (1:2) followed by denaturing at 95 °C and annealing at room temperature. The probes were then transfected into cells by Lipofectamine 3000 24 h before the formation of 3D spheroids. 3D Wound Healing Assay and 3D Live Cell Imaging. To perform 3D wound healing, the spheroids were exposed to laser irradiation (0.85 mW/μm2). The effects of irradiation duration were characterized experimentally and computationally in our previous study.25 The ablation effectiveness in the 3D spheroid could be observed visually by the cell morphology and bubble generation. Modification of the fluorescence signal from the live cell actin label and dsLNA biosensor could also be observed near the wound site. The ablated spheroids were mounted inside a microscope stage incubator (Okolabs) to maintain 5% CO2 and 37 °C. The sample was heated to a stable temperature for 15 min before live imaging. 3D confocal z-stack images of the spheroids were collected every 10 min for 9 h with a Leica SP8 confocal microscope equipped with a 25× water immersion objective and a Leica HyD hybrid detector. Cell Migratory Behavior Analysis. Image analysis was performed using ImageJ and MATLAB. 3D images were registered using the ImageJ plugin to correct drift. Unless specified otherwise, the middle section of the spheroid, in the XY plane, was extracted and analyzed. Particle image velocimetry was performed using the PIVlab software package in MATLAB. The interrogation window size was chosen as 64 pixels by 64 pixels with 50% overlap to satisfy the Nyquist sampling criterion. For single cell tracking, cells were tracked based on the nuclear fluorescence using the ImageJ plugin TrackMate. The persistence of the velocity field was calculated as the net displacement divided by the length of the trajectory. The divergence was calculated using the Matlab command. The convective acceleration of the tissue was computed as the product of velocity and divergence of velocity field. The local/unsteady acceleration was negligible considering the integrity of 3D structure. To reflect the magnitude of the mechanical loads, the mean tissue acceleration was computed as the absolute value of convective acceleration. Reagents. Antiactivated Notch1 antibody (ab8925) was purchased from Abcam. Goat anti-Dll4 antibodies (C-20) were purchased from Santa Cruz Biotechnology. Goat and rabbit IgG labeling secondary antibodies were purchased from Sigma-Aldrich and Thermo Fisher Scientific. The control siRNA and Dll4 siRNA were purchased from Santa Cruz Biotechnology. Notch pathway activator

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.8b06305. Additional experimental methods, numerical analysis, and supplementary figures (PDF) Movie 1: Actic dynamics in a 3D spheroid (AVI) Movie 2: 3D reconstruction of a spheroid before and after ablation (AVI) Movie 3: 3D reconstruction of a spheroid 3 h after ablation (AVI) Movie 4: Actin dynamics in a Y-27632-treated 3D spheroid (AVI) Movie 5: Actin dynamics in a Jag1 peptide-treated 3D spheroid (AVI) Movie 6: Actin dynamics in a Dll4 siRNA-treated 3D spheroid (AVI) Movie 7: Actin dynamics in a Notch1 siRNA-treated 3D spheroid (AVI)

AUTHOR INFORMATION Corresponding Author

*Tel: +1-814-863-5267. Fax: +1-814-863-0490. E-mail: pak@ engr.psu.edu. ORCID

Pak Kin Wong: 0000-0001-7388-2110 Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS The authors thank Y. Lu and C. Protchko for technical discussion and assistance in the experiment. This work was funded by National Institutes of Health Director’s New Innovator Award (DP2OD007161) and National Science Foundation Biophotonic program. REFERENCES (1) Hu, B.; Leow, W. R.; Cai, P.; Li, Y. Q.; Wu, Y. L.; Chen, X. Nanomechanical Force Mapping of Restricted Cell-To-Cell Collisions Oscillating between Contraction and Relaxation. ACS Nano 2017, 11, 12302−12310. (2) Friedl, P.; Gilmour, D. Collective Cell Migration in Morphogenesis, Regeneration and Cancer. Nat. Rev. Mol. Cell Biol. 2009, 10, 445−457. H

DOI: 10.1021/acsnano.8b06305 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano (3) Friedl, P.; Locker, J.; Sahai, E.; Segall, J. E. Classifying Collective Cancer Cell Invasion. Nat. Cell Biol. 2012, 14, 777−783. (4) Clark, A. G.; Vignjevic, D. M. Modes of Cancer Cell Invasion and the Role of the Microenvironment. Curr. Opin. Cell Biol. 2015, 36, 13−22. (5) Wu, Y.; Ali, M. R. K.; Dong, B.; Han, T.; Chen, K.; Chen, J.; Tang, Y.; Fang, N.; Wang, F.; El-Sayed, M. A. Gold NanorodPhotothermal Therapy Alters Cell Junctions and Actin Network in Inhibiting Cancer Cell Collective Migration. ACS Nano 2018, 12, 9279−9290. (6) Cai, D. F.; Chen, S. C.; Prasad, M.; He, L.; Wang, X. B.; Choesmel-Cadamuro, V.; Sawyer, J. K.; Danuser, G.; Montell, D. J. Mechanical Feedback through E-Cadherin Promotes Direction Sensing during Collective Cell Migration. Cell 2014, 157, 1146−1159. (7) Martin, P.; Lewis, J. Actin Cables and Epidermal Movement in Embryonic Wound Healing. Nature 1992, 360, 179−183. (8) Omelchenko, T.; Vasiliev, J. M.; Gelfand, I. M.; Feder, H. H.; Bonder, E. M. Rho-Dependent Formation of Epithelial ″Leader″ Cells during Wound Healing. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 10788−10793. (9) Yang, Y.; Jamilpour, N.; Yao, B.; Dean, Z. S.; Riahi, R.; Wong, P. K. Probing Leader Cells in Endothelial Collective Migration by Plasma Lithography Geometric Confinement. Sci. Rep. 2016, 6, 22707. (10) Reffay, M.; Parrini, M. C.; Cochet-Escartin, O.; Ladoux, B.; Buguin, A.; Coscoy, S.; Amblard, F.; Camonis, J.; Silberzan, P. Interplay of RhoA and Mechanical Forces in Collective Cell Migration Driven by Leader Cells. Nat. Cell Biol. 2014, 16, 217−223. (11) Cheung, K. J.; Gabrielson, E.; Werb, Z.; Ewald, A. J. Collective Invasion in Breast Cancer Requires a Conserved Basal Epithelial Program. Cell 2013, 155, 1639−1651. (12) Dean, Z. S.; Elias, P.; Jamilpour, N.; Utzinger, U.; Wong, P. K. Probing 3D Collective Cancer Invasion Using Double-Stranded Locked Nucleic Acid Biosensors. Anal. Chem. 2016, 88, 8902−8907. (13) Tse, J. M.; Cheng, G.; Tyrrell, J. A.; Wilcox-Adelman, S. A.; Boucher, Y.; Jain, R. K.; Munn, L. L. Mechanical Compression Drives Cancer Cells toward Invasive Phenotype. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 911−916. (14) Cheung, K. J.; Padmanaban, V.; Silvestri, V.; Schipper, K.; Cohen, J. D.; Fairchild, A. N.; Gorin, M. A.; Verdone, J. E.; Pienta, K. J.; Bader, J. S.; Ewald, A. J. Polyclonal Breast Cancer Metastases Arise from Collective Dissemination of Keratin 14-Expressing Tumor Cell Clusters. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, E854−863. (15) Riahi, R.; Sun, J.; Wang, S.; Long, M.; Zhang, D. D.; Wong, P. K. Notch1-Dll4 Signalling and Mechanical Force Regulate Leader cell Formation during Collective Cell Migration. Nat. Commun. 2015, 6, 6556. (16) Riahi, R.; Yang, Y. L.; Zhang, D. D.; Wong, P. K. Advances in Wound-Healing Assays for Probing Collective Cell Migration. J. Lab. Autom. 2012, 17, 59−65. (17) Kramer, N.; Walzl, A.; Unger, C.; Rosner, M.; Krupitza, G.; Hengstschlager, M.; Dolznig, H. In Vitro Cell Migration and Invasion Assays. Mutat. Res., Rev. Mutat. Res. 2013, 752, 10−24. (18) Poujade, M.; Grasland-Mongrain, E.; Hertzog, A.; Jouanneau, J.; Chavrier, P.; Ladoux, B.; Buguin, A.; Silberzan, P. Collective Migration of an Epithelial Monolayer in Response to a Model Wound. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 15988−15993. (19) Riahi, R.; Long, M.; Yang, Y.; Dean, Z.; Zhang, D. D.; Slepian, M. J.; Wong, P. K. Single Cell Gene Expression Analysis in InjuryInduced Collective Cell Migration. Integr. Biol. 2014, 6, 192−202. (20) Wood, W.; Jacinto, A.; Grose, R.; Woolner, S.; Gale, J.; Wilson, C.; Martin, P. Wound Healing Recapitulates Morphogenesis in Drosophila Embryos. Nat. Cell Biol. 2002, 4, 907−912. (21) Richardson, R.; Metzger, M.; Knyphausen, P.; Ramezani, T.; Slanchev, K.; Kraus, C.; Schmelzer, E.; Hammerschmidt, M. ReEpithelialization of Cutaneous Wounds in Adult Zebrafish Combines Mechanisms of Wound Closure in Embryonic and Adult Mammals. Development 2016, 143, 2077−2088.

(22) Cai, D.; Dai, W.; Prasad, M.; Luo, J.; Gov, N. S.; Montell, D. J. Modeling and Analysis of Collective Cell Migration in an in Vivo Three-Dimensional Environment. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, E2134−2141. (23) Sakar, M. S.; Eyckmans, J.; Pieters, R.; Eberli, D.; Nelson, B. J.; Chen, C. S. Cellular Forces and Matrix Assembly Coordinate Fibrous Tissue Repair. Nat. Commun. 2016, 7, 11036. (24) Zou, L.; Wang, H.; He, B.; Zeng, L.; Tan, T.; Cao, H.; He, X.; Zhang, Z.; Guo, S.; Li, Y. Current Approaches of Photothermal Therapy in Treating Cancer Metastasis with Nanotherapeutics. Theranostics 2016, 6, 762−772. (25) Riahi, R.; Wang, S.; Long, M.; Li, N.; Chiou, P. Y.; Zhang, D. D.; Wong, P. K. Mapping Photothermally Induced Gene Expression in Living Cells and Tissues by Nanorod-Locked Nucleic Acid Complexes. ACS Nano 2014, 8, 3597−3605. (26) Wang, S.; Sun, J.; Xiao, Y.; Lu, Y.; Zhang, D. D.; Wong, P. K. Intercellular Tension Negatively Regulates Angiogenic Sprouting of Endothelial Tip Cells via Notch1-Dll4 Signaling. Adv. Biosyst. 2017, 1, 1600019. (27) Lee, G. Y.; Kenny, P. A.; Lee, E. H.; Bissell, M. J. ThreeDimensional Culture Models of Normal and Malignant Breast Epithelial Cells. Nat. Methods 2007, 4, 359−365. (28) Dean, Z. S.; Riahi, R.; Wong, P. K. Spatiotemporal Dynamics of MicroRNA during Epithelial Collective Cell Migration. Biomaterials 2015, 37, 156−163. (29) Riahi, R.; Dean, Z.; Wu, T. H.; Teitell, M. A.; Chiou, P. Y.; Zhang, D. D.; Wong, P. K. Detection of mRNA in Living Cells by Double-Stranded Locked Nucleic Acid Probes. Analyst 2013, 138, 4777−4785. (30) Gidwani, V.; Riahi, R.; Zhang, D. D.; Wong, P. K. Hybridization Kinetics of Double-Stranded DNA Probes for Rapid Molecular Analysis. Analyst 2009, 134, 1675−1681. (31) Sun, J.; Xiao, Y.; Wang, S.; Slepian, M. J.; Wong, P. K. Advances in Techniques for Probing Mechanoregulation of Tissue Morphogenesis. J. Lab. Autom. 2015, 20, 127−137. (32) Ng, M. R.; Besser, A.; Danuser, G.; Brugge, J. S. Substrate Stiffness Regulates Cadherin-Dependent Collective Migration through Myosin-II Contractility. J. Cell Biol. 2012, 199, 545−563. (33) Kovall, R. A.; Gebelein, B.; Sprinzak, D.; Kopan, R. The Canonical Notch Signaling Pathway: Structural and Biochemical Insights into Shape, Sugar, and Force. Dev. Cell 2017, 41, 228−241. (34) Fortini, M. E. Notch Signaling: The Core Pathway and Its Posttranslational Regulation. Dev. Cell 2009, 16, 633−647. (35) Petrie, R. J.; Yamada, K. M. Fibroblasts Lead the Way: A Unified View of 3D Cell Motility. Trends Cell Biol. 2015, 25, 666− 674. (36) Petrie, R. J.; Gavara, N.; Chadwick, R. S.; Yamada, K. M. Nonpolarized Signaling Reveals Two Distinct Modes of 3D Cell Migration. J. Cell Biol. 2012, 197, 439−455. (37) Brugues, A.; Anon, E.; Conte, V.; Veldhuis, J. H.; Gupta, M.; Colombelli, J.; Munoz, J. J.; Brodland, G. W.; Ladoux, B.; Trepat, X. Forces Driving Epithelial Wound Healing. Nat. Phys. 2014, 10, 683− 690.

I

DOI: 10.1021/acsnano.8b06305 ACS Nano XXXX, XXX, XXX−XXX