(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



Fos regulates macrophage infiltration against surrounding tissue resistance by a cortical actin-based mechanism in Drosophila

['Vera Belyaeva', 'Institute Of Science', 'Technology Austria', 'Klosterneuburg', 'Stephanie Wachner', 'Attila Gyoergy', 'Shamsi Emtenani', 'Igor Gridchyn', 'Maria Akhmanova', 'Markus Linder']

Date: 2022-02 

The infiltration of immune cells into tissues underlies the establishment of tissue-resident macrophages and responses to infections and tumors. Yet the mechanisms immune cells utilize to negotiate tissue barriers in living organisms are not well understood, and a role for cortical actin has not been examined. Here, we find that the tissue invasion of Drosophila macrophages, also known as plasmatocytes or hemocytes, utilizes enhanced cortical F-actin levels stimulated by the Drosophila member of the fos proto oncogene transcription factor family (Dfos, Kayak). RNA sequencing analysis and live imaging show that Dfos enhances F-actin levels around the entire macrophage surface by increasing mRNA levels of the membrane spanning molecular scaffold tetraspanin TM4SF, and the actin cross-linking filamin Cheerio, which are themselves required for invasion. Both the filamin and the tetraspanin enhance the cortical activity of Rho1 and the formin Diaphanous and thus the assembly of cortical actin, which is a critical function since expressing a dominant active form of Diaphanous can rescue the Dfos macrophage invasion defect. In vivo imaging shows that Dfos enhances the efficiency of the initial phases of macrophage tissue entry. Genetic evidence argues that this Dfos-induced program in macrophages counteracts the constraint produced by the tension of surrounding tissues and buffers the properties of the macrophage nucleus from affecting tissue entry. We thus identify strengthening the cortical actin cytoskeleton through Dfos as a key process allowing efficient forward movement of an immune cell into surrounding tissues.

Our lab examines Drosophila macrophage migration into the embryonic germband (gb) to investigate mechanisms of immune cell tissue invasion. Macrophages, also called plasmatocytes or hemocytes, are the primary phagocytic cell in Drosophila and share striking similarities with vertebrate macrophages [ 9 – 13 ]. They are specified in the head mesoderm at embryonic stages 4 to 6 and by stage 10 start spreading along predetermined routes guided by platelet-derived growth factor-related and vascular endothelial growth factor-related factors (Pvf) 2 and 3 [ 9 , 14 , 15 ] to populate the whole embryo. One of these paths, the movement into the gb, requires macrophages to invade confined between the ectoderm and mesoderm [ 16 , 17 ]. The level of tension and thus apparent stiffness of the flanking ectoderm is a key parameter defining the efficiency of macrophage passage into and within the gb [ 16 ]. Penetration of macrophages into the gb utilizes Integrin, occurs normally without matrix metalloproteinases (MMPs) [ 17 ], and is even enhanced by ECM deposition [ 18 , 19 ] likely because the basement membrane has not yet formed at this stage [ 16 , 20 ]. Thus, Drosophila macrophage gb invasion represents an ideal system to explore the mechanisms by which immune cells and surrounding tissues interact with one another to aid the invasion process.

Migration in 2D and 3D environments requires actin polymerization to power forward progress. The assembly of actin at the leading edge, when coupled to Integrin adhesion to anchor points in the surrounding extracellular matrix (ECM), can allow the front of the cell to progress [ 8 ]. This anchoring also allows the contraction of cortical actin at the rear plasma membrane to bring the body of the cell forwards. But a role for cross-linked actin at the cell surface in assisting forward progress by helping to counteract the resistance of surrounding tissues and in buffering the nucleus has not been previously identified.

The classical model of cell migration on a surface postulated in the 1980s by Abercrombie has been extended [ 1 ] by studies showing that migrating cells utilize diverse strategies depending on the architecture and physical properties of their three-dimensional (3D) surroundings [ 2 ]. Much of this work has been conducted in vitro, where variations in the environment can be strictly controlled. However, most 3D migration occurs within the body, and much less research has elucidated the mechanisms used to efficiently move in these diverse environments, particularly into and through tissues. Such migration is crucial for the influence of the immune system on health and disease. Vertebrate macrophages migrate into tissues during development where they take up residence, regulating organ formation and homeostasis and organizing tissue repair upon injury [ 3 , 4 ]. A variety of types of immune cells infiltrate into tumors and can both promote or impede cancer progression [ 5 , 6 ]. Responses to infection require immune cells to traverse through the vascular wall, into the lymph node, and through tissues [ 7 ]. Yet the mechanisms utilized by immune cells to allow migration into such challenging cellular environments in vivo are not well understood.

Results

Dfos modulates Filamin and Tetraspanin to aid gb tissue invasion To identify Dfos targets that promote macrophage invasion, we FACS isolated macrophages from wild-type and mac>DfosDN embryos during the time when invasion has just begun and conducted RNA sequencing of the corresponding transcriptomes (Fig 3A and S2 Data). We first assessed reads that map to Dfos, which can correspond to both endogenous and DfosDN mRNA; we found a 1.6-fold increase in the presence of the one copy of DfosDN in this line, arguing that this transgene is expressed at levels similar to each endogenous copy of Dfos and is unlikely to produce extraneous effects. We then examined genes that displayed a log 2 fold change of at least 1.5 with an adjusted p-value less than 0.05 in the presence of DfosDN. Ten genes were down-regulated (Fig 3B and S3A and S3B Fig) and 9 up-regulated by DfosDN (S2 Table). Up-regulated genes in DfosDN encoded mostly stress response proteins, consistent with the role previously demonstrated for fos in C. elegans in suppressing stress responses [34]. We concentrated on the down-regulated class. Of these, we focused on the actin cross-linking filamin Cher and the tetraspanin TM4SF from a group that can form membrane microdomains that affect signaling and migration [35,36]. No known role for TM4SF had been previously identified in Drosophila. To determine if these Dfos targets were themselves required for invasion, we knocked down Cher and TM4SF through RNAi individually or simultaneously and observed significantly reduced macrophage numbers in the gb, particularly upon the knockdown of both targets simultaneously (Fig 3C–3G) while not affecting macrophage numbers in the pre-gb zone (S3D Fig) or on the vnc (S3E Fig). Overexpression of Cher or TM4SF along with DfosDN in macrophages increased the mean macrophage numbers in the gb, and overexpression of TM4SF rescued the DfosDN macrophage invasion defect (Fig 3H–3L). Expression of a GFP control did not restore macrophage invasion indicating that the rescue we observed through Cher or TM4SF expression was not due to promoter competition leading to reductions in DfosDN expression. We conclude that Dfos aids macrophage gb invasion by increasing the mRNA levels of the filamin actin cross-linker Cher and the tetraspanin TM4SF. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Dfos regulates macrophage gb invasion through cytoskeletal regulators: The filamin Cher and the tetraspanin TM4SF. (A) Schematic representing the pipeline for analyzing mRNA levels in FACS-sorted macrophages. (B) Table of genes down-regulated in macrophages expressing DfosDN. Genes are ordered according to the normalized p-value from the RNA sequencing. The closest mouse protein orthologs were found using UniProt BLAST; the hit with the top score is shown in the table. (C-F) Lateral views of representative St 12 embryos in which the 2 targets with links to actin organization, (D) the tetraspanin TM4SF and (E) the filamin Cher, have been knocked down individually or (F) together, along with the control (C). Scale bar: 50 μm. (G) Quantification shows that the number of macrophages in the gb is reduced in embryos expressing RNAi against either cher (KK 107451) or TM4SF (KK 102206) in macrophages, and even more strongly affected in the double RNAi of both. Control vs. cher RNAi p = 0.0005 (46% reduction). Control vs. TM4SF RNAi p = 0.009 (37% reduction), Control vs. cher/TM4SF RNAi p > 0.0001 (61% reduction). cher RNAi vs. TM4SF RNAi p = 0.15. SD: 29, 23, 17, 12. (H-K) Lateral views of a representative St 12 embryo from (H) the control, as well as embryos expressing DfosDN in macrophages along with either (I) GFP, (J) Cher, or (K) TM4SF. (L) Quantification shows that overexpression of TM4SF in DfosDN expressing macrophages restores their normal numbers in the gb. Overexpression of Cher in this background shows a strong trend toward rescue but did not reach statistical significance. Control vs. DfosDN p = 0.015 (28% reduction); Control vs. cher p = 0.74; Control vs. TM4SF p > 0.99; DfosDN vs. DfosDN cher p = 0.14; DfosDN vs. DfosDN, TM4SF p < 0.0001; Control vs. DfosDN cher p = 0.97; Control vs. DfosDN TM4SF p = 0.35. SD: 22, 16, 16, 21, 22, 13. (M-O) q-PCR analysis of mRNA extracted from the bones of mice that are wt, tg for Fos controlled by a Major Histocompatibility promoter and viral 3′ UTR elements, and those in which such c-Fos transgenesis has led to an OS. Analysis of mRNA expression shows that higher levels of (M) Fos correlate with higher levels of (N-N”) FlnA-C, and (O) Tspan6 in OS. p-values = 0.86, 0.001, 0.003, SD: 0.7, 0.6, 0.3 in M, 0.98, 0.009, 0.007 and 0.4, 0.2, 1.5 in N, 0.39, <0.0001, <0.0001 and 0.2, 0,3, 1.1 in N’, 0.76, 0.005, 0.002 and 0.8, 2.3, 2.4 in N”, 0.99, 0.004, 0.003 and 0.1, 0.2, 0.2 in O. Scale bar: 50 μm. Macrophages are labeled using either (C-F) srpHemo-H2A::3xmCherry or (H-K) srpHemo-Gal4 (“mac>”) driving UAS-mCherry::nls. ***p < 0.005, **p < 0.01, *p < 0.05. Unpaired t test or one-way ANOVA with Tukey post hoc were used for statistics. Each column contains the number of analyzed embryos. The data underlying the graphs can be found in S1 Data. Cher, Cheerio; gb, germband; ns, not significant; RNAi, RNA interference; OS, osteosarcoma; tg, transgenic; wt, wild type. https://doi.org/10.1371/journal.pbio.3001494.g003

In murine osteosarcoma, c-fos mRNA level increases correlate with those of Filamins and Tetraspanin-6 To determine if these Dfos targets in Drosophila could also be Fos targets in vertebrate cells, we utilized a well-established murine transgenic model that overexpresses c-fos. In these mice, transgenic c-fos expression from viral 3′ UTR elements in osteoblasts (the bone forming cells) leads to osteosarcoma (OS) development accompanied by a 5-fold increase in c-fos mRNA expression (Fig 3M) [37]. We examined by qPCR the mRNA levels of our identified Dfos targets’ orthologs, comparing their levels in OS (Fos tg OS) to neighboring, osteoblast-containing healthy bones from Fos tg mice (Fos tg bone) and control bones from wild-type mice (wt bone). We saw 2.5- to 8-fold higher mRNA levels of the 3 murine Filamin orthologs (Fig 3N–3N”) and a 15-fold increase in Tetraspanin-6 (Fig 3O) in OS cells. mRNA levels of several of the orthologs of other Dfos targets we had identified showed less strong inductions or even decreases; the Glutathione S transferase Gstt3 and the Slit receptor Eva1c increased 4- and 2.8-fold, respectively, while the mitochondrial translocator Tspo was 25% lower (S3F–S3I Fig). These results suggest that Dfos’s ability to increase mRNA levels of 2 key functional targets for migration, a Filamin and a Tetraspanin, is maintained by at least one vertebrate fos family member.

Dfos increases assembly of cortical actin through Cheerio and TM4SF to aid macrophage invasion We wished to determine what cellular properties Dfos could affect through such targets to facilitate Drosophila macrophage invasion. Given Cher’s known role as an actin cross-linker, we stained embryos with phalloidin to detect F-actin. Line scan analysis revealed reduced intensity at the macrophage cortex in fixed Dfos1 mutant embryos in the pre-gb and gb entry zone (Fig 4A). To examine this in invading mac>DfosDN macrophages within live embryos, we utilized a srpHemo-moe::3xmCherry reporter, which marks cortical F-actin [38,39] only in macrophages and observed a reduction of 53% (Fig 4B–4D) in its signal. We saw no change by western analysis in the levels of the Moe::3xmCherry protein itself upon DfosDN expression (S4A and S4A’ Fig). We hypothesized that the changes in macrophage cortical actin we observed in the mac>DfosDN could be due to the lower levels of Cher and/or TM4SF mRNA. Indeed, we observed reductions in Moe::3xmCherry all around the edge of invading macrophages in live embryos expressing RNAi against Cher or TM4SF in macrophages (Fig 4E–4H). To test if a decrease in actin assembly could underlie the reduced tissue invasion of mac>DfosDN macrophages, we forced cortical actin polymerization by expressing a constitutively active version of the formin Diaphanous (DiaCA), in which Dia’s inhibitory autoregulatory domain has been deleted, allowing active Dia to localize to the macrophage cortex [40]. Indeed, expressing DiaCA in macrophages completely rescued the Dfos1, Dfos2 (S4B Fig), and mac>DfosDN invasion defect (Fig 4I and 4J). Given that Dia, like Dfos, does not affect general macrophage migratory capacities along the vnc [41], we examined if Dia might normally play a role in invasion. We utilized 2 RNAis against Dia and observed decreased macrophage numbers in the gb in each (Fig 4K and 4L) with no effect on numbers in the pre-gb (S4C Fig) or on the vnc (S4D Fig). These results argue that Dfos aids invasion by increasing levels of TM4SF and Cher to enhance assembly of actin around the surface of the macrophage. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Dfos increases Rho1-GTP, the formin Diaphanous and actin at the cortex through Cher and TM4SF. (A) Quantification of phalloidin intensity to detect F actin at the macrophage–macrophage contacts in Stage 11/12 Dfos1 embryos. F-actin is strongly reduced at these homotypic contacts. (B-C) Representative confocal images of live embryos expressing in invading macrophages the F-actin binding and homodimerizing portion of Moesin (srpHemo-moe::3xmCherry) to label F-actin, presented as a maximum z-projection. Relative Moe-3xmCherry intensity is indicated with a pseudo-color heat map as indicated on the left, with yellow as the highest levels and dark blue as the lowest as indicated in the calibration bar to the left. Insets in the bottom left corner of each panel show a grayscale single z-plane corresponding to the white box in the main image. Embryo genotype indicated below. Strong reductions in cortical actin are observed in macrophages expressing DfosDN compared to the control. (D-E) Quantification of the macrophage Moe:3xmCherry intensity as a measure of cortical F-actin, normalized to the average fluorescence intensity of the control per batch. (D) Quantification shows that macrophages expressing DfosDN display a 53% reduction in Moe::3xmCherry intensity compared to the control when the 2 outliers shown as single dots are excluded, 37% if they are included. Outliers identified by 10% ROUT. n of ROIs analyzed = 650 for control, 687 for DfosDN. p = 0.0007 for analysis including outliers (Kolmogorov–Smirnov) and p < 0.0001 for analysis excluding outliers (Welch’s t test). SD: 0.2, 0.4. (E) Quantification reveals that macrophage expression of an RNAi against either cher or TM4SF, the 2 genes whose expression is reduced in DfosDN, also results in a decrease of Moe::3xmCherry intensity (by 40% each). n of ROIs analyzed = 549 for control, 423 for cher RNAi, 306 for TM4SF RNAi. Control vs. cher RNAi p = 0.006. Control vs. TM4SF p = 0.003. SD: 0.2, 0.3, 0.2. (F-H) Images and representation as in B-C. Strong reductions in cortical actin are observed in macrophages expressing cher RNAi or TM4SF RNAi compared to the control. (I-I’) Representative confocal images of St 12 embryos from the control and a line in which macrophages express DfosDN and a CA form of the formin Dia to restore cortical actin polymerization. (J) Quantification shows that while macrophage expression of DiaCA does not significantly affect the number of macrophages in the gb, expressing it in a DfosDN background rescues macrophage gb invasion. Control vs. DfosDN p = 0.017 (28% reduction), Control vs. diaCA p = 0.18, Control vs. DfosDN, diaCA p = 0.010, DfosDN vs. DfosDN, diaCA p < 0.0001. SD: 22, 16, 16, 24. (K–K’) Representative confocal images of St 12 embryos from the control and from a line expressing an RNAi against dia in macrophages. (L) Quantification of 2 RNAi lines against dia expressed in macrophages shows a 37% and 21% reduction in macrophage numbers in the gb compared to control. Control vs. dia RNAi1 (TRiP HMS05027) p < 0.0001; control vs. dia RNAi2 (TRiP HMS00308) p = 0.0008. SD: 13, 20, 22. (M, O) Examples of line profiles used for the determination of the membrane-to-cytoplasmic ratio of Dia in panel N and the Rho1 activity sensor DiaRBD in panel P. Line intensity profiles from fixed Stage 11 embryos of (M) Dia::GFP or (O) DiaRBD::GFP (green) and membrane Myr::Tomato (magenta) across the outward facing edge of groups of macrophages sitting within approximately 40 μm of the gb that expressed either lacZ (Control), Rho1DN, DfosDN, cher RNAi, or TM4SF RNAi as shown in the schematic in M. Line length approximately 8 μm. Blue lines indicate mean GFP intensity on the membrane and in cytoplasm. (N, P) Quantification of membrane-to-cytoplasmic intensity ratio of (N) Dia::GFP or (P) the Rho1 activity sensor DiaRBD::GFP expressed in macrophages under UAS control along with either lacZ (control, n = 233 from 15 or n = 158 line scans from 11 embryos), Rho1DN (n = 212 from 14 or n = 123 from 7), DfosDN (n = 237 from 12 or n = 135 from 8), cher RNAi (n = 252 from 13 or n = 128 from 8), TM4SF RNAi (n = 279 from 17 or n = 205 from 11). Control vs. Rho1DN ****p < 0.0001 (29% (N), 34% (P) reduction), Control vs. DfosDN **p = 0.0037 (23% (N), 21% (P) reduction), Control vs. cher RNAi ***p = 0.0007, 24% reduction (N) or ****p < 0.0001, 28% reduction (P), Control vs. TM4SF RNAi *p = 0.024 or 0.026 (20% reduction). SD: 1.9, 0.9, 1.0, 0.9, 1.0 in N; 0.7, 0.5, 0.5, 0.5, 0.4 in P. Macrophages are labeled using either srpHemo-Gal4 driving UAS-mCherry::nls (I-I’), srpHemo-H2A::3xmCherry (K-K’). srpHemo-moe::3xmCherry, srpHemo-Gal4 (mac>) crossed to (B) UAS-GFP as a Control, (C) UAS-DfosDN, (F) w− Control, (G) UAS-cher RNAi (KK 107451), (H) UAS-TM4SF RNAi (KK 102206). srpHemo-GAL4 UAS-Myr::tdTomato UAS-dia::GFP (M, O) or UAS-diaRBD::GFP (N, P) crossed to UAS-lacZ as a Ctrl, UAS-Rho1DN or the lines indicated above. ****p < 0.0001, ***p < 0.005, **p < 0.01, *p < 0.05. Unpaired t test used for A. Welch’s t test of normalized average mean intensity per embryo for D with the 2 indicated outliers excluded, for statistical assessment. One-way ANOVA with Tukey post hoc for E, J, L. Kruskal–Wallis for N, P. The number of analyzed (A) macrophage–macrophage junctions, or (D-E, J, L, N, P) embryos is shown in each column. Scale bar 10 μm in (B-C, F-H), 50 μm in (I, K). The data underlying the graphs can be found in S1 Data. CA, constitutively active; Cher, Cheerio; ctrl, control; gb, germband; ns, not significant; RNAi, RNA interference; ROI, region of interest. https://doi.org/10.1371/journal.pbio.3001494.g004

Dfos stimulates the cortical activity of Rho1 and Diaphanous through its targets TM4SF and Cheerio We hypothesized that Dfos and its targets enhance cortical actin assembly by affecting Dia. We had observed no change in Dia’s mRNA levels (S3C Fig) upon DfosDN expression and thus examined localization of Dia protein. We expressed Dia::GFP [42] in macrophages along with myristoylated Tomato (Myr::Tomato) to mark the membrane and quantified intensity profiles of linescans across the membrane in various genetic backgrounds, assessing the ratio of membrane/cytoplasmic Dia (Fig 4M and S4E Fig). Dia’s autoinhibition negatively regulates its cortical localization and activity in Drosophila macrophages [40,43]. For mDia, binding to activated Rho GTPases as well as to other unknown membrane associated proteins can release this autoinhibition [44]. Drosophila Rho1 has been shown to directly bind Dia lacking its autoinhibitory domain [45]. As predicted by these prior results, upon the expression of Rho1DN, we observed a significant reduction, by 29%, in the enrichment of Dia at the cortex compared to the control (mem/cyto = 2.46 in control, 1.76 for Rho1DN) (Fig 4N). We found that expressing either DfosDN or RNAis against Cher or TM4SF resulted in a significant reduction of cortical Dia, 80%, 83%, and 70%, respectively, as strong as that seen upon Rho1DN expression (mem/cyto = 1.9, 1.88, 1.97). To assess if this effect of the Dfos pathway on Dia could be due to an effect on Rho activity itself, we expressed a sensor of active Rho1, the Rho1 binding domain of Dia (DiaRBD::GFP) [46], in macrophages along with Myr::Tomato to delineate the plasma membrane and quantified intensity profiles of linescans across the membrane in various genetic backgrounds as above (Fig 4O and S4F Fig). To validate the assay, we expressed Rho1DN and found, as expected, a significant reduction, by 34%, in the enrichment of the Rho1 sensor DiaRBD at the cortex compared to the control (mem/cyto = 1.15 in control, 0.76 for Rho1DN) (Fig 4P). Expressing either DfosDN or RNAis against the filamin Cher or the tetraspanin TM4SF also resulted in a significant reduction of cortical DiaRBD, by 62%, 82%, and 59%, respectively, as much as that seen upon Rho1DN expression (mem/cyto = 0.91, 0.83, 0.92, respectively). The lower Rho1 activity we observed in the absence of the Dfos pathway could be a result of reduced Rho1 GEF recruitment, as Filamin has been shown to bind the Rho GTPase GEFs Trio and Vav2 [47,48] and a tetraspanin can recruit a filamin [49,50]. Our data argue that higher levels of the Dfos targets TM4SF and Cher increase Dia localization at the cortex and thus stimulate cortical actin assembly, at least partially through increased Rho1 activity. We examined what consequence these lower cortical F-actin levels had on the cellular behavior of macrophages during entry. Quantitation showed that the actin protrusion that macrophages initially insert between the ectoderm and mesoderm during invasion was actually longer in the mac>DfosDN>LifeAct::GFP macrophages than in the control (Fig 5A and S5A Fig and S4 Movie). We then performed live imaging of macrophages labeled with CLIP::GFP to visualize microtubules and thus cell outlines in both genotypes; we determined the aspect ratio (maximal length over width) that the first entering cell displays as it enters into the gb. Unlike the control, the first DfosDN-expressing macrophage was extended even before it had fully moved its rear into the gb (S5B Fig). We carried out measurements, taking only the first cells that had entered the gb to be able to clearly distinguish the rear of the first macrophage from the tips of following cells (Fig 5B). We also avoided including in this measurement the forward protrusion and determined that the first DfosDN-expressing macrophage inside the gb displays an average increase of 23% in the maximal length (L) of the cell body and a 12% reduction in the maximal width (W) (Fig 5D and S5C Fig). Interestingly, in the pre-gb zone, the aspect ratio (max L/W) of mac>DfosDN macrophages was not different from control macrophages (Fig 5C and 5D), although the mac>DfosDN cells were 9% smaller in both their length and width (S5D Fig). This suggested that the gb could impose resistance on the entering macrophage, an effect that mac>DfosDN macrophages have trouble overcoming due to their compromised cortical actin cytoskeleton. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. Dfos aids macrophage gb invasion against the resistance of surrounding tissues and buffers the nucleus. (A) Quantification from live embryos shows that the length of the F-actin protrusion of the first entering macrophage is longer in macrophages expressing DfosDN. p = 0.011. The F-actin protrusion labeled with srpHemo-Gal4 driving UAS-LifeAct::GFP was measured in the direction of forward migration (see schematic). SD: 2.4, 3.7. (B-C) Stills from 2-photon movies of St 11 embryos showing (B) the first macrophages entering the gb and (C) macrophages in the pre-gb zone in the control and in a line expressing DfosDN in macrophages. Microtubules are labeled with srpHemo-Gal4 driving UAS-CLIP::GFP. A blue arrow indicates the front and a yellow arrow indicates the rear of the macrophage. Schematics above indicate where images were acquired. (D) Schematic at left shows macrophage measurements: vertical line for the maximum length and horizontal line for the maximum width. Histograms show the probability density distributions of the aspect ratios (maximum length over maximum width) of the first macrophage entering the gb (left) and macrophages in the pre-gb (right). Macrophages expressing DfosDN are more elongated than the controls. Control vs. mac>DfosDN aspect ratios at gb entry p = 0.0011, in pre-gb p = 0.53. SD: in gb 1.0, 1.6; in pre-gb 0.5, 0.5. (E-F”’) Confocal images of St 12 embryos expressing RNAi against Lamin or LaminC in macrophages in (E-E”’) the control, or (F-F”’) in embryos also expressing DfosDN in macrophages. srpHemo-GAL4 used as driver. Lam RNAi1: GD45636. Lam RNAi2: KK107419. LamC RNAi: TRiP JF01406. (G) Macrophage RNAi knockdown of Lamins, which can increase nuclear deformability did not affect macrophages numbers in the gb in the control. In embryos in which macrophages expressed DfosDN, Lamin knockdown rescued their reduced numbers in the gb. Control vs. DfosDN p < 0.0001. Control vs. Lam RNAi1 p > 0.99, vs. Lam RNAi2 p = 0.83, vs. LamC RNAi p > 0.99. Control vs. DfosDN, Lam RNAi1 p = 0.024, vs. DfosDN, Lam RNAi2 p > 0.99, vs. DfosDN, LamC RNAi p > 0.99. DfosDN vs. DfosDN, Lam RNAi1 p < 0.0001, vs. DfosDN, Lam RNAi2 p = 0.0049, vs. DfosDN, LamC RNAi p < 0.0001. SD: 22, 10, 19, 11, 21, 23, 16, 20. (H) Expressing DfosDN in macrophages reduces their number in the gb. Concomitantly reducing tissue tension in the ectoderm (light blue in schematic) through Rho1DN substantially rescues invasion. srpHemo-QF QUAS control (mac<>) governed macrophage expression and e22c-GAL4 ectodermal (ecto>). Control vs. mac<>DfosDN p < 0.0001 (56% reduction), vs. mac<>DfosDN; ecto>Rho1DN p > 0.99, vs. ecto>Rho1DN p = 0.11. mac<>DfosDN vs. mac<>DfosDN; ecto>Rho1DN p < 0.0001, vs. ecto>Rho1DN p = 0.0044. mac<>DfosDN; ecto>Rho1DN vs. ecto>Rho1DN p > 0.99. SD: 23, 16, 21, 18. Macrophages are labeled in B-C by srpHemo-Gal4 driving UAS-CLIP::GFP, and in E-F’” by srpHemo-Gal4 UAS-mCherry::nls. ****p < 0.0001, ***p < 0.005, **p < 0.01, *p < 0.05. Unpaired t test was used for A, one-way ANOVA with Tukey post hoc for G-H. The number shown within the column corresponds to measurements in A, and analyzed embryos in G-H. Scale bar 5 μm in B-C, and 50 μm in E-F”’. The data underlying the graphs can be found in S1 Data. ctrl, control; gb, germband; ns, not significant; RNAi, RNA interference. https://doi.org/10.1371/journal.pbio.3001494.g005

[END]

[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001494

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/