Reconstruction of Par polarity in apolar cells reveals a dynamic process of cortical polarization

Cellular polarization is fundamental for various biological processes. The Par network system is conserved for cellular polarization. Its core complex consists of Par3, Par6, and aPKC. However, the dynamic processes that occur during polarization are not well understood. Here, we artificially reconstructed Par-dependent polarity using non-polarized Drosophila S2 cells expressing all three components endogenously in the cytoplasm. The results indicated that elevated Par3 expression induces cortical localization of the Par-complex at the interphase. Its asymmetric distribution goes through three steps: emergence of cortical dots, development of island-like structures with dynamic amorphous shapes, repeating fusion and fission, and polarized clustering of the islands. Our findings also showed that these islands contain a meshwork of unit-like segments. Par-complex patches resembling Par-islands exist in Drosophila mitotic neuroblasts. Thus, this reconstruction system provides an experimental paradigm to study features of the assembly process and structure of Par-dependent cell-autonomous polarity.


Abstract 1
Cellular polarization is fundamental for various biological processes. The Par network 2 system is conserved for cellular polarization. Its core complex consists of Par3, Par6, 3 and aPKC. However, the dynamic processes that occur during polarization are not well 4 understood. Here, we artificially reconstructed Par-dependent polarity using 5 non-polarized Drosophila S2 cells expressing all three components endogenously in the 6 cytoplasm. The results indicated that elevated Par3 expression induces cortical 7 localization of the Par-complex at the interphase. Its asymmetric distribution goes 8 through three steps: emergence of cortical dots, development of island-like structures 9 with dynamic amorphous shapes, repeating fusion and fission, and polarized clustering 1 0 of the islands. Our findings also showed that these islands contain a meshwork of 1 1 unit-like segments. Par-complex patches resembling Par-islands exist in Drosophila 1 2 mitotic neuroblasts. Thus, this reconstruction system provides an experimental 1 3 paradigm to study features of the assembly process and structure of Par-dependent 1 4 cell-autonomous polarity. 1 5 Polarization is a fundamental cellular property that plays a vital role in 1 8 various biological processes in multi-cellular as well as single cell organisms. 1 9 Par-complex system is a conserved mechanism that regulates cell 2 0 polarization (Kemphues et al, 1988;Suzuki & Ohno, 2006;Johnston, 2018). The core 2 1 Par-complex consists of Par6, Par3, and typical protein kinase C (aPKC) (Kemphues et 2 2 al, 1988;Tabuse et al, 1998). Domain structures of these components and their 2 3 interactions have been extensively studied (Lang & Munro, 2017). Par3 exhibits 2 4 membrane binding affinity through its C-terminal domain and the ability to 2 5 self-oligomerize via its N-terminal CR1 domain, which is essential for its localization 2 6 and function (Benton, 2003;Mizuno et al, 2003;Krahn et al, 2010;Harris, 2017) Structural studies have revealed that the CR1 domain forms helical polymers of 10 nm 2 8 diameter (Zhang et al, 2013). Par6 and aPKC, which form a stable subcomplex, interact 1 with the CR3 and PDZ domains of Par3(Izumi et al, 1998;Renschler et al, 2018). 2 Phosphorylation of this domain by aPKC inhibits this interaction(Morais- de-Sá et al, 3 2010;Soriano et al, 2016). Thus, Par-complex assembly is a dynamic process. CDC42 4 binds to the aPKC-Par6 subcomplex and anchors it to the cell membrane as a diffusible 5 cortical form (Joberty et al, 2000;Aceto et al, 2006;Rodriguez et al, 2017;Wang et al, 6 2017) On the other hand, Lgl and/or Par1 kinase act as inhibitory factors against 7 aPKC (Guo & Kemphues, 1995;Betschinger et al, 2003;Yamanaka et al, 2003;Plant et 8 al, 2003;Hurov et al, 2004), and distribute complementarily to the core Par complex . 9 Interplay between these components results in cytocortical asymmetry ( Ohno, 2006). On the other hand, cell polarization is coupled with mitosis during 1 6 asymmetric divisions, and autonomously induced or triggered by an external cue, 1 7 depending on the cell type (Yamashita et al, 2010). Because of such association between 1 8 Par-dependent polarization and other processes, the Par-complex exhibits different 1 9 behavioral characteristics in an individual context, making it difficult to determine 2 0 general features of the dynamic process taking place during cell polarization by the 2 1 Par-complex. We attempted to address this problem by establishing an artificial 2 2 polarization system induced by the Par-complex (Baas et al, 2004;Johnston et al, 2009).
We used Drosophila Schneider cells (S2 cells) of mesodermal origin as host 2 4 cells (Schneider, 1972). They are neither polarized nor adhere to the substratum and 2 5 between cells. The 3 core components of the Par-complex are endogenously expressed 2 6 in S2 cells but are distributed in the cytoplasm throughout the cell cycle. Thus, S2 cells 2 7 appear to be an ideal system for cell polarity induction. 2 8

S2 cells polarize by an elevated expression of Par3 2
First, we tested the effect of overexpressing each core component of the 3 Par-complex in S2 cells, which distribute these components evenly throughout the 4 Thus, these two methods essentially provided the same value for the segmental length. 1 In addition, these segments had a fairly homogeneous diameter in STED microscopic 2 images, where the mean half width of Par-segments was 0.22±0.03μm, (Supplementary 3 Figs. 4e, f). These results raise the possibility that the Par-island meshwork contains a 4 unit segment. Indeed, separate rod-or string-shape structures as well as open square-5 structures were often observed in the earlier phases of the Par-complex aggregation time 6 course ( Supplementary Fig. 4g, movie 2), supporting the notion that Par-islands are 7 assembled from these elemental structures, generating regularity in the meshwork 8 organization. 9 Roles of Par components and the cytoskeleton in polarity formation 1 0 Because the elevation of Par3 expression induced cortical polarization in S2 1 1 cells, we investigated the role of functional domains of Par3 by observing phenotypes 1 2 with Par6-GFP following the overexpression of mutant Par3 forms via the 1 3 Metallothionein promoter (Fig. 5a). First, we tested the role of the CR1 domain 1 4 responsible for self-polymerization in the polarized Par-complex assembly(Benton, and a degree of polarization with a mean ASI value of 0.54±0.14 for polarized cells. 4 This suggested strong enhancement of Par-complex clustering (Figs. 5c, f). Clustering 5 of the Par-islands was so tight under this condition that the polarized region sometimes 6 assumed a bowl-like shape, in which the island structure was hardly discernible. 7 Subsequently, this dense aggregation gradually separated into small and nested islands. 8 Dense packing of the Par complex containing Par3S980A suggested that the turnover of 9 Par3-aPKC association and dissociation played a role in the normal clustering of 1 0 Par-islands. This was similar to that of Drosophila epithelial cells, wherein Par3S980A 1 1 colocalized with aPKC-Par6 in the apical domain with disorganized adhesion 1 2 belts(Morais- de-Sá et al, 2010). 1 3 Next, we examined the effect of the membrane association region (MAR) of 1 4 Par3 by overexpressing Par3ΔMAR (Krahn et al, 2010). The Par-complex no longer 1 5 localized cortically, but formed several cytoplasmic aggregates, which coalesced into a 1 6 single large sphere (Fig. 5d). Thus, the functional domains of Par3 and the interactions 1 7 between these domains, together, play a role in the properly polarized distribution of the 1 8 Par-complex in the S2 cell system. 1 9 Lastly, we examined the effects of the actin cytoskeleton on islands. While 2 0 ROCK inhibitor, Y27632, did not significantly affect the behavior of Par-islands (data 2 1 not shown), an actin inhibitor, Latrunculin B, changed the islands into a spherical shape, 2 2 which frequently formed membrane protrusions (Fig. 5g, Supplementary movie 5), 2 3 suggesting that the actin-membrane skeleton is necessary to balance the surface tension 2 4 of Par-islands (see Discussion).

5
Discussion 2 6 In this study, we reconstructed Par-complex-dependent cortical cell polarity induced by 2 7 Par3 overexpression in non-polar S2 cells, using the Gal4-UAS system and the 2 8 Metallothionein promoter for Par3 expression. Because this polarity requires 1 endogenous Par6, aPKC, and Lgl, the reconstruction system reproduced the 2 fundamental properties of Par-dependent polarization in vivo, at least in part. While 3 there is no firm information regarding the Par3 protein level in polarized cells in vivo, 4 the ratio of overexpressed Par3 protein level to endogenous Par3, in S2 cells, was 5 estimated to be approximately 300-fold and 20-fold for the Gal4-UAS system and 6 Metallothionein promoter, respectively ( Supplementary Fig. 5). 7 Temporal patterns of Par-complex aggregation 8 In our reconstruction system, cortical asymmetry began with the formation and growth 9 of cortical dot-like structures, which were also reportedly associated with anterior components. Thus, these dots appear to be the common initial process of Par-complex 1 3 cortical aggregation. The subsequent process of asymmetric localization proceeds in the 1 4 form of Par-islands with amorphous and dynamic behavior. To our knowledge, this 1 5 structure has not been reported in cortical Par-complex assembly in C. elegans or 1 6 Drosophila. However, island-like structures were observed during asymmetric Par 1 7 complex distribution in Drosophila neuroblasts, suggesting that Par-islands were not 1 8 specific to this artificial system that used apolar S2 cells. 1 9 In C. elegans, asymmetric segregation of the Par-complex is driven by was involved in the asymmetric clustering of the Par-complex in S2 cells. Furthermore, 2 3 no directional movement towards the pole of polarization was observed. Interestingly, 2 4 initial dot formation appeared to be biased towards the region opposite the cleavage 2 5 point, where the centrosome also appeared to be located, which was consistent with a positional cue for the initiation of Par-complex-dependent cell polarity. In this context, 1 polarization process of the S2 cell system is likely to be cell-autonomous and dependent 2 on the induction of polarity proteins, wherein the orientation of polarity appeared to be 3 dependent on internal cue(s). 4

5
The morphology and dynamics of Par-islands 6 Par-complex assembly at the cortex of S2 cells appears to stabilize the cell 7 membrane because membrane filopodia extensively formed in areas where Par-islands 8 were absent (Supplementary movie 2). Also, cell membrane curvature was higher where 9 Par-islands were attached, compared with that of the surrounding areas (Fig. 5g). Par-island and its higher membrane curvature reflects its relatively high surface tension. 1 6 This is supported by the fact that disruption of the actin cytoskeleton by Latrunculin B 1 7 treatment leads to a curled or spherical Par-island, inducing dynamic cell membrane 1 8 protrusions. This phenomenon may be explained as follows; disruption of the cortical 1 9 cytoskeleton leads to the loss of its elasticity, which had balanced the surface tension of 2 0 the Par-island. The resulting imbalance in surface tension may cause the Par-island to 2 1 shrink into a bowl or sphere shape, thereby bending the cell membrane outward and 2 2 conferring protrusive activity to the cell membrane. In contrast, when membrane 2 3 affinity is quite low, as in the case of Par3ΔMAR, Par-island shape is not affected by 2 4 either cortical cytoskeleton elasticity or membrane affinity, and its shape would be 2 5 determined only by the surface tension of Par3-islands. Under these conditions, we 2 6 found that the Par-complex forms small cytoplasmic droplets, which subsequently 2 7 coalesce into a spherical, densely packed structure, suggestinng that phase separation 1 takes place between the Par-complex aggregates and the cytoplasm(Hyman et al, 2014). 2

Molecular network of the Par-complex in the island state.
3 In this study, we revealed that a Par-island is a meshwork of various 4 polygonal shapes, which appear to be unit-like segments with an average length of 5 approximately 0.4 μm. Isolated fragments such as single fragments and structures made 6 up of a few connected fragments were observed during the development of Par-islands 7 via live-imaging. These observations suggested that these isolated fragments assembled 8 into a meshwork to form islands. These islands change shape rapidly during their 9 movement along the cortex, and sometimes fuse to release pieces of different sizes, 1 0 raising the possibility that Par-islands and small free fragments are mutually 1 1 exchangeable. The factors that determine the size of these unit segments need further Par3 is known to polymerize in vitro via the CR1 domain at its N-terminus to 1 4 form a helical polymer of 8-fold symmetry (Zhang et al, 2013;Feng et al, 2007). 1 5 Whether Par3 polymers are involved in the cortical cluster of the Par complex remain 1 6 unclear. Our super-resolution microscopic observations and the ability of Par3 to form 1 7 filaments lead to the simple hypothesis that the unit segment of a Par-island is formed 1 8 by Par3 polymers as the core structure. Cell phenotypes expressing Par3 lacking its 1 9 oligomerization domain, Par3ΔCR1, is compatible with this hypothesis. While there are 2 0 many possibilities via which Par3 filaments may form a unit segment, a single Par3 2 1 polymer may form a single segment. Another possibility is for Par3 polymers to be 2 2 aligned along the long axis of the segment. Since Par6 and aPKC bind the PDZ and 2 3 CR3 domains of Par3, respectively, Par6 and aPKC may act as cross linkers between 2 4 segments (Feng et al, 2007). Given the phenotype of ParS980A overexpression, the 2 5 association of Par3 and aPKC by aPKC phosphorylation may confer flexibility and 2 6 dynamism to the structure and/or assembly of the segmental elements. These 2 7 hypotheses need to be tested in future studies.

8
The two states of the Par-island distribution at steady state 1 An interesting property of Par-islands is that they are not unified into one 2 large island under the cell membrane, even when polarized. Overexpression of 3 Par3ΔMAR or Par3S980A is an exception. In the latter case, rapid and stable formation 4 of the cortical Par-complex does not seem to permit separate island formation, and a 5 large, transient dome is formed instead. In the former case, the Par-complex aggregates 6 to form one large sphere. This cytoplasmic phenomenon is likely to be due to a phase 7 separation between the Par-complex and the cytoplasm. Considering this property of the 8 Par-complex, the unique feature of Par-islands associated with the cell membrane may 9 reflect phase separation in 2 dimensions. where a negative regulator Lgl is involved (Betschinger et al, 2003). The initial 2 2 condition, which is possibly determined by a stochastic distribution of islands, may 2 3 select one of the two stable patterns in a cell. We propose that such cell-scale patterning 2 4 is coupled with local phase separation of Par-islands as previously described for the 2 5 membrane lipid domain (John & Bär, 2005).
In summary, we have developed a potential Par complex-polarization system 2 7 upon induction of Par3 in non-polar S2 cells, which provides a useful model for 2 8 cell-autonomous cell polarization, allowing us to easily manipulate gene expression and 1 image at the super-resolution level. One intriguing challenge will be the coupling of 2 mitosis with cell polarization in this system to induce asymmetric division. 3

Materials and Methods 1
Cell culture 2 S2 cell culture and transfection were performed as previously described (Ogawa et al., 3 2009). Expression vectors were transfected at two days prior to microscopic or Western 4 blot analysis. For induction of the Metallothionein promoter, 100 mM CuSO 4 solution 5 was added to a medium at a final concentration of 1 mM. 6 7 Live cell imaging 8 Cells were mounted on a 35 mm glass-bottom dish coated with 15 µg/ml 9 poly-L-ornithine and incubated at 25ºC for 30 min, followed by microscopic analysis. 1 0 Images were taken at a 1 μm z-interval with a spinning disk confocal microscopy 1 1 CSU-W1 (Yokogawa, Tokyo, Japan) equipped with a sCMOS camera Neo (Andor, 1 2 Belfast, Northern Ireland) and MetaMorph software (Molecular Devices, San Jose, CA, For immunostaining of S2 cells, transfected cells were mounted on a 1 7 poly-L-ornithine-coated cover slip and fixed with 4% paraformaldehyde in PBS for 15 1 8 min at room temperature. Cells were washed with PBS, followed by treatment with 1 9 0.1% Triton-X100 in PBS for 15 min. After washed with PBS, cells were treated with a 2 0 blocking buffer containing 2% BSA in PBS for 30 min and incubated with primary 2 1 antibodies in the blocking buffer for 30 min, followed by incubation with secondary 2 2 antibodies for 30 min. Immunostained cells were embedded in mounting medium 2 3 PermaFluor (Thermo Fisher Scientific) and analyzed with a confocal microscopes 2 4 LSM510 (Zeiss, Oberkochen, Germany). For super-resolution microscopy, samples 2 5 were embedded in ProLong Glass Andifade Mountant (Thermo Fisher Scientific).

6
For immunostaining of Drosophila neuroblasts, brains isolated from third 2 7 instar larvae were fixed with 4% paraformaldehyde in PBS for 20 min at room 2 8 temperature. Samples were treated with the blocking buffer for 1 h, followed by 2 9 incubation with primary antibodies and secondary antibodies for 1 h each. Samples 3 0 were then embedded in ProLong Gold Antifade Mountant (Thermo Fisher Scientific) 3 1 and analyzed with a confocal microscope LSM880 (Zeiss).

2
Primary antibodies used were anti-aPKC (rabbit polyclonal, used at 1:1000, 1 Santa Cruz), anti-Par-3 (rabbit polyclonal, used at 1:1000 or mouse monoclonal, used at Using SRRF-processed images, Par3 contour lengths along the meshwork 2 2 were manually traced with Fiji. Each image was overlayed by an edge-enhanced image 2 3 generated with the Sobel filter, to highlight Par3 contour shapes. Lengths between their 2 4 terminal ends and/or branching points were measured. A histogram and a density plot 2 5 were generated from all contour lengths, and the shape of the density plot was fitted 2 6 with a linear combination of 7 Gaussian curves by a fitting function implemented in R 2 7 with the non-linear least square method. Power spectral density of the second derivative 2 8 of the density plot was calculated using fast Fourier transform method with R.

9
Stimulated emission depletion (STED) imaging was performed using TCS 3 0 SP8 STED 3X microscope (Leica, Wetzlar, Germany) with an objective lens HC PL 3 1 APO 93X/1.30 GLYC (Leica). Deconvolution was performed with a deconvolution 3 2 software package Huygens Professional (version 17.10, Scientific Volume Imaging, 1 Hilversum, Netherlands). 2 Deconvoluted STED images were used for the analyses of Par3 segment 3 lengths and widths. The segment length was defined as a shortest length between 4 terminal ends, corners and/or branching points of Par3 contours, and manually traced 5 with Fiji. The segment width was given by the full width at half maximum (FWHM) of 6 a Gaussian-fitted signal distribution orthogonal to each Par3 segment. 7 8 Quantification of asymmetry and statistics 9 The equatorial z-plane of each cell was analyzed for the estimation of asymmetric index 1 0 (ASI) (see also Supplementary Fig. 1b). The cell perimeter was traced by a 0.5 1 1 µm-width line and the signal intensity along the line was measured with Fiji. The signal 1 2 intensities were summed up along the half (L) of the total perimeter length (2L). The 1 3 difference between this value and that of the other half was calculated and normalized 1 4 by the total signal intensity along the perimeter. This measurement was done starting 1 5 from every pixel along the perimeter (1 pixel = 0.108 µm), The maximum value of them 1 6 was defined as ASI. ASI larger than 0.35 was defined as polarized cell, and the 1 7 statistical significance of polarized cell population was analyzed by Fisher's exact 1 8 test with post-hoc Bonferroni correction for multiple comparisons (Fig. 3a, and Fig. 5e, 1 9 -f). Statistical analyses were performed with R software. 2 0 2 1 Western blot analysis 2 2 Whole cell extracts of the untransfected S2 cells and the transfected S2 cells were 2 3 subjected to SDS-polyacrylamide gel electrophoresis. Primary antibodies used were 2 4 anti-Par3 antibody (rabbit polyclonal, used at 1:1000), anti-alpha-tubulin (rat 2 5 monoclonal, Santa Cruz). Secondary antibodies used were horseradish peroxidase 2 6 (HRP)-conjugated anti-mouse antibody (sheep polyclonal, used at 1:3000, GE 2 7 Healthcare), HRP-conjugated anti-rabbit antibody (sheep polyclonal, used at 1:3000, 2 8 GE Healthcare) and HRP-conjugated anti-chicken antibody (donkey polyclonal, used at 2 9 1:250, SA1-300, Thermo Fisher Scientific). Protein level was analyzed by 3 0 chemiluminescence with Chemi-Lumi One L (Nacalai tesque, Kyoto, Japan) and 3 1 quantified with an image analyzer LAS-3000 system (Fujifilm, Tokyo, Japan). To 3 2 compare the expression level of the overexpressed fluorescent protein per cell between 3 3 two different transfectants or with the endogenous Par3 proteins, transfection efficiency 1 for each sample was calculated by counting fluorescence-positive cells and negative 2 cells. The ratio of the expression level per cell was calculated by dividing the measured 3 staining intensity on the Western blot by the transfection efficiency. 4 5 Knock-down experiment 6 Long double-stranded RNAs (dsRNAs) were used for RNA interference (RNAi) in S2 7 cells as previously described(Bettencourt-Dias & Goshima, 2009). dsRNA for 8 knocking-down Par-6 or aPKC was synthesized with MEGAscript T7 Transcription Kit 9 (Ambion, Thermo Fisher Scientific) according to the manufacurer's instructions, by 1 0 using pBS-T7/Par-6/T7 or pBS-T7/aPKC/T7 plasmid, which contains a full-length ORF 1 1 of Par-6 or aPKC flanked by two T7 promoters, as a template, respectively. 1 2 dsRNA for knocking-down Lgl was by using pUAS-Flag-Lgl, and primers in below. promoter and poly(A) addition signals (and a small region of hsp70 promoter) was 3 0 cloned between BamHⅠ and SalⅠ site of pUC19 plasmid. Then, a synthetic 3 1 double-strand oligonucleotide containing multiple cloning sites was cloned into the 3 2 BamHⅠsite to produce pDAMCS expression plasmid. Par-3 conjugated with Myc or 1 Flag epitope and mKate2 at the N-terminus and C-terminus, respectively, and Lgl 3A 2 with Flag epitope was inserted into pUAST plasmid (Brand & Perrimon, 1993). To 3 construct expression vectors for Par-3 under the control of the induction system, Par-3 4 that had been conjugated with Myc epitope and mKate2 at the N-terminus and 5 C-terminus, respectively, was inserted into pMT plasmid (Invitrogen, Thermo Fisher 6 Scientific). pUbq-Spd2-GFP is kindly provided by Jordan Raff (University of Oxford, 7 UK). Plasmids used for transfection were purified with Wizard Plus SV Minipreps

Conflict of interest 2 3
The authors declare no conflict interests. 2 4 Data availability 2 5 All raw data are available upon the request. Gaussian curves are shown in Supplementary Fig. 3b. The averaged mean of individual 1 2 Gaussian curves was 0.38±0.062 μm for 754 contours from 28 cells. 1 3 e. Power spectral density for the second derivative of the contour distribution plot 1 4 shown in (c). The major frequency was 2.4 μm -1 . 1 5 f and g. Deconvoluted super-resolution (STED) images of cells expressing 1 6 myc-Par3-GFP (e) and Par6-GFP (f), stained for GFP. S2-cells were transfected with 1 7 pMT-myc-Par3-GFP and pAct-Par6 (e) or with pMT-myc-Par3 and pAct-Par6-GFP (fb), 1 8 followed by CuSO 4 addition for induction two days following transfection. Cells were 1 9 fixed for immune-staining for GFP at 8 h post-induction. Scale bar, 10 μm. Par3 S980A -mKate2 (mean ASI value = 0.54±0.14). In both cases, the polarized cell 1 7 population (ASI >0.35) is significantly altered (p=0.0006 for Par-3 Δ CR1 -mKate2 and 1 8 p=0.0088 for Par-3 S980A -mKate2) compared with that of wild type Par3-mKate2 ( Fig.  1  9  3a). Quantification was performed 8 h after CuSO 4 addition onwards. In all images, 2 0 CuSO 4 was added at two days post-transfection. Par-islands 2 a. Temporal pattern of the fluorescence intensity of Par6-GFP expressed in S2 cells 3 transfected with pAct-Par6-GFP. The fluorescence intensity of Par6-GFP was measured 4 for 6 cells every 1 h from two days following transfection. Expression levels did not 5 drastically change from 6 h onward following transfection, indicating that Par6-GFP is 6 an appropriate marker for the Par complex distribution in cells, when Par3-mKate2 was 7 induced.