PCP and Septins govern the polarized organization of the actin cytoskeleton during convergent extension

Abstract

Convergent extension (CE) requires the coordinated action of the planar cell polarity (PCP) proteins^1,2 and the actin cytoskeleton,^3,4,5,6 but this relationship remains incompletely understood. For example, PCP signaling orients actomyosin contractions, yet actomyosin is also required for the polarized localization of PCP proteins.^7,8 Moreover, the actin-regulating Septins play key roles in actin organization⁹ and are implicated in PCP and CE in frogs, mice, and fish^5,6,10,11,12 but execute only a subset of PCP-dependent cell behaviors. Septin loss recapitulates the severe tissue-level CE defects seen after core PCP disruption yet leaves overt cell polarity intact.⁵ Together, these results highlight the general fact that cell movement requires coordinated action by distinct but integrated actin populations, such as lamella and lamellipodia in migrating cells¹³ or medial and junctional actin populations in cells engaged in apical constriction.^14,15 In the context of Xenopus mesoderm CE, three such actin populations are important, a superficial meshwork known as the "node-and-cable" system,^4,16,17,18 a contractile network at deep cell-cell junctions,^6,19 and mediolaterally oriented actin-rich protrusions, which are present both superficially and deeply.^4,19,20,21 Here, we exploited the amenability of the uniquely "two-dimensional" node and cable system to probe the relationship between PCP proteins, Septins, and the polarization of this actin network. We find that the PCP proteins Vangl2 and Prickle2 and Septins co-localize at nodes, and that the node and cable system displays a cryptic, PCP- and Septin-dependent anteroposterior (AP) polarity in its organization and dynamics.

Keywords: Septin; Xenopus; actin; convergent extension; planar cell polarity.

Figure 1:. PCP protein and septin localization in actin rich nodes.

A-C. Still images of FP fusions to indicated PCP proteins (green) and Lifeact (magenta) in polarized cells, insets show magnified view of nodes. D-F. Still images of indicated PCP proteins (green) and Lifeact (magenta) in pre-polarized cells, insets show magnified view of nodes. G, H. Quantification of the position of FP-labeled PCP foci in polarized and pre-polarized cells. I-L. Images of indicated Septin subunit s(green) and Lifeact (Magenta) at nodes; Scalebar = 10 um. M. Schematic interpretation of data in A-L. N, O. Vangl2-GFP localization in control and Sept7MO cells. P. Quantification of AP positioning of Vangl2 puncta. N values and statistics are provided in Supplemental Information. See also Figures S1-S3.

Figure 2:. Progressive polarization of the actin node and cable system requires PCP and Septins.

A-D. Individual cells mosaically labeled with Lifeact-RFP at times indicated at left (maximum intensity projections of cortical actin 10um into cell). a’-d’. Cyan box and inlay highlight individual nodes. Heatmap showing orientation of actin fibers quantified with OrientationJ (Legend in a’; cyan = mediolateral, red = anteroposterior). a”-d” Quantification of actin filament orientation. E. Quantification of cell length-to-width ratios over time. F. Upper panel, schematic of AP localization, arrow representing cellular anterior-posterior axis. Lower panel, quantification of Lifeact-RFP labelled nodes over time. G-K. Representative images of cells with indicated manipulations. (Maximum intensity projection of cortical actin 10um into cell). Inlay shows actin node and cable, where present, indicated by magenta box). g’, g”. Schematics of quantification. h’-k’: Quantification of actin fiber alignment in cells. h”-k” Quantification of actin cable positioning. Scale= 10um. N values and statistics are provided in Supplemental Information.

Figure 3:. PCP- and Septin-dependent anteroposterior pattern in the node and cable system.

A. Representative TIRF image of a control cell, outlined in pink. Boxes highlighting anterior (cyan) and posterior (yellow) regions of the cell. B, C. Representative anterior and posterior cropped ROIs from A. b’-c’. tSOAX segmentation. b”-c”. representative single actin filament traces used for quantification. D. Quantification of persistence length (Lp) of actin filaments in anterior (A) or posterior (P) cellular regions of control, Sept7MO, and Vangl2MO cells. E, F. Lifeact-RFP labeling the actin cortex in control and Sept7 MO cells. Boxes indicate regions used for kymograph, below. e’, f’. Kymograph of representative actin fiber coalescence over time. G. Quantification of actin filament coalescence over time, 0 sec indicates the full coalescence to a single fiber. N values and statistics are provided in Supplemental Information. See also Figure S4.

Figure 4:. Sept7 controls local dynamics of polarized actin flows.

A. TIRF image of a single cell labeled with Lifeact-RFP, Pk2-GFP, and actin-647 speckles. B-D. Zoomed, split channels of the boxed region in A. Orange and purple boxes indicate regions quantified in Panel I, below). E. Heatmap and vector diagram of PIV data for the region shown in A. Arrows indicate the vector and direction of flow; legend shows heatmap indicates positive divergence of flow (blue) or negative divergent (i.e. convergence, yellow). F. Zoomed view of the boxed region of the heatmap in E. G. Line plot of Pk2-GFP pixel intensities in multiple cells. H. Line plot showing vector divergence along the same line from which Pk2 intensity was derived, above. Note maximal vector convergence is coincident with the peak of Pk2 intensity (i.e., in nodes). I. Quantification of actin speckle lifetimes in regions indicated by boxes in panels B and D, above. N values and statistics are provided in Supplemental Information.