A375 cells were cultured in DMEM containing 2 mM Glutamine, 10% Fetal Bovine Serum (FBS), 100 IU penicillin, and 100 μg/ml streptomycin. HEK 293T cells were maintained in DMEM medium containing 10% FBS, 100 IU penicillin, and 100 μg/ml streptomycin.

Cells were transfected with HiFugene transfection reagent (Promega) according to the manufacturer’s protocol. Three days after transfection, cells were collected and cell lysates were subjected to immunoprecipitation (IP) using the GFP-Nanotrap Kit (Chromotek) according to the manufacturer’s instructions. Cells were lysed in IP lysis buffer (25mMTris-HCl pH 7.4, 150 mM NaCl, 1% NP-40, 1 mM EDTA, 5% glycerol) on ice for 30 min. The supernatants obtained were used in Western blotting and subsequent IP. Western blotting was conducted as previously described (Gui et al., 2016). All biochemical data shown are representative of no less than three independent assays.

Cells were cultured on coverslips for 3 days, then fixed with 4% paraformaldehyde for 20 min at room temperature. Cells were permeabilized in PBT for 15 min at room temperature, then stained with antibodies or proximity ligation assay reagents. For the Duolink proximity ligation assay, the blocking buffer, PLA probe and detection reagent kit were obtained from Sigma-Aldrich. Cells were blocked in Duolink Blocking buffer for 1 h at 37 °C. Primary antibodies were diluted in Duolink Antibody Diluent buffer, and cells were incubated overnight at 4°C. Experimental samples were incubated with mouse anti-human α-Cat and rabbit anti-human Scrib, whereas negative controls were incubated with mouse anti-human α-Cat and rabbit anti-Myc as control. The following morning, cells were incubated with PLUS PLA probe anti-rabbit and MINUS PLA probe anti-mouse for 2 h at 37°C. Ligation and amplification steps of the PLA were performed using the Duolink in situ Detection Reagents Red kit according to the manufacturer’s instructions. Images were then obtained by confocal microscopy.

The following shRNAs were synthesized by Sigma-Aldrich:

Lentivirus-mediated shRNA: Based on the protocols provided by Addgene, lentivirus constructs carrying shRNA against human Scrib or α-Cat were generated and delivered to cultured cells upon packaging into replication-deficient lentiviral particles to generate knockdown clones. The packaging and envelope plasmids, including psPAX2 and pMD2. G, were obtained from Addgene. The shRNA transfer vector plasmid against Scrib (TRCN0000004458) and the shRNA transfer vector plasmid against α-Cat (TRCN0000234533) were obtained from Sigma. Control cells were transduced with lentiviral particles carrying pLKO.1 Non-Target shRNA Control Plasmid (Sigma).

Ex-lacZ (#44248), nub-GAL4 (#25754), tubP-GAL80 ^ts (#7017) and scrib ² (#41775) were obtained from the Bloomington Drosophila Stock Center (BDSC). Scrib:GFP (CA07683) was obtained from Fly Trap projects (Morin et al., 2001). αCat:GFP (#115921) was obtained from Kyoto Stock Center. Fly stocks were maintained at 25°C unless otherwise mentioned. To induce MARCM clones, larvae were heat-shocked for 2 h at 37°C 3 days after egg laying, and samples were collected 2 days after heat shock. Larvae and adults for conditional knockdown mediated by nub-GAL4 were maintained at 27°C after egg laying.

Figure 3A, B: Ex-lacZ.

Figure 3C: Scrib:GFP.

Figure 3F αCat::GFP.

Figure 4A–C: nub-GAL4/Ex-lacZ; tub-GAL80 ^ts /+, or nub-GAL4/Ex-lacZ, UAS-scrib. FL:Myc; tub-GAL80 ^ts /+, or nub-GAL4/Ex-lacZ, UAS-LRR:Myc; tub-GAL80 ^ts /+, or nub-GAL4/Ex-lacZ, UAS-LRR:α-Cat:Myc; tub-GAL80 ^ts /+

Figure 5B–D: hsFLP/+; tub-GAL4, UAS-mCD8-GFP/+; FRT82B, tub-GAL80/FRT82B scrib ² , or hsFLP/+; tub-GAL4, UAS-mCD8-GFP/UAS-scribFL:Myc; FRT82B, tub-GAL80/FRT82B scrib ² , or hsFLP/+; tub-GAL4, UAS-mCD8-GFP/UAS-LRR:Myc; FRT82B, tub-GAL80/FRT82B scrib ² , or hsFLP/+; tub-GAL4, UAS-mCD8-GFP/+; FRT82B, tub-GAL80/FRT82B, or hsFLP/+; tub-GAL4, UAS-mCD8-GFP/UAS-dLRR:Myc; FRT82B, tub-GAL80/FRT82B scrib ² , or hsFLP/+; tub-GAL4, UAS-mCD8-GFP/UAS-LRR:αCat:Myc; FRT82B, tub-GAL80/FRT82B scrib ² , or hsFLP/+; tub-GAL4, UAS-mCD8-GFP/UAS-α-Cat:Myc; FRT82B, tub-GAL80/FRT82B scrib ²

Larvae were fixed in 3.7% formaldehyde at room temperature for 20 min, after which wing imaginal discs were dissected. The following primary antibodies were used: rat anti-DE-Cad (1:50) (Developmental Studies Hybridoma Bank (DSHB)), mouse anti-GFP (1: 5,000 for Western blotting, Millipore), rabbit anti-Myc (1:100, Santa Cruz Biotechnology), mouse anti-β-Galaxtosidase (1:500, Promega), mouse anti-human YAP (1:50), rabbit anti-cleaved-Dcp-1 (1:200, Cell Signaling Technology), mouse anti-human α-Cat (1:50, Santa Cruz Biotechnology), rabbit anti-human Scrib (1:50, Abcam), and rabbit anti-human YAP (1:50, Cell Signaling Technologies).

Secondary antibodies (1:200) were as follows: goat anti-mouse IgG Alexa 488, goat anti-mouse IgG Alexa 568, goat anti-mouse IgG Alexa 647, goat anti-rabbit IgG Alexa 488, goat anti-rabbit IgG Alexa 568, goat anti-rat IgG Alexa 488 and Alexa488-conjugated phalloidin (all from Thermo Fisher Scientific).

Fluorescent images were obtained with Zeiss LSM700 and Leica SP8 upright confocal microscopes. All confocal immunofluorescent images were processed and analyzed with ImageJ/FIJI. The images were the composite of a stack with projection of max intensity unless otherwise specified. Quantification of nuclear/cytoplasmic signal (ratio N/C) was obtained from YAP intensity in the nuclei and cytoplasm segmented in ImageJ/FIJI (Schindelin et al., 2012).

To generate human scrib cDNA construct, we obtained full-length (FL) and truncated human scrib cDNA from PLK45 (#37250, Addgene). Human α-Cat cDNA was from α-Cat1-906 (#24194, Addgene). α-Cat, FL, truncated Scrib or LRR:αCat with C-terminal Myc tag were cloned into the PEF1-His-myc vector. The constructs encode the following human Scrib protein amino acids: FL, 1-1630; LRR, 1-701; PDZ1/2, 702-976; PDZ3/4, 977-1216; CT, 978-1630. All Scrib fragments were cloned into EcoRI and Xbal sites; α-Cat was cloned into EcoRI and NotI sites. The chimeric construct was generated by inserting α-Cat CDS into the C-terminus of LRR before the stop codon. α-Cat with C-terminal GFP was cloned into pEGFP-N1 at EcoRI and KpnI sites.

All experiments were carried out independently at least three times. Statistical analyses were performed using GraphPad Prism software (v.9.0.2, GraphPad). The number for all quantified data is indicated in the figure legends. Data are means ±95% confidence intervals (CIs). Statistical significance was calculated by the two tailed t-test method and Dunnett’s test after one-way ANOVA.

To understand whether the interaction between Scrib and α-Cat is also observed in vertebrates, we employed a mammalian cell culture system. HEK293T cells, derived from human embryonic kidney epithelium, were transfected with plasmids harboring tagged cDNAs of human Scrib (Scrib, Myc-tagged) and human α-Cat (α-Cat, GFP-tagged), or GFP as a negative control. Gene products were utilized in immunoprecipitations (IP) with antibodies directed against the Myc and GFP epitopes. To dissect functionally the domains of Scrib, various fragments containing different domains were co-expressed with α-Cat:GFP or GFP (Figure 1A). GFP or α-Cat:GFP interactors were precipitated using beads conjugated to a GFP-specific nanobody and subjected to Western blot assays. No Scrib or fragments thereof were detected in the GFP interactome, whereas LRR or PDZ3/4 domains of Scrib were immunoprecipitated by α-Cat:GFP, suggesting that the interaction between Scrib and α-Cat is conserved in vertebrates and invertebrates (Figure 1B). We barely detected full-length Scrib in the α-Cat:GFP interactome, suggesting that physical interactions with full length Scrib are transient or weak.

To further test protein interactions within the HEK293T cells, we employed proximity-ligation assay (PLA) to detect the endogenous interaction between Scrib and α-Cat in situ in cell culture (Alam, 2022). Rabbit anti-human Scrib and mouse anti-human α-Cat antibodies were combined to detect their interaction, and rabbit anti-Myc antibodies were used as a negative control when combined with anti-α-Cat antibodies. Only when two proteins are very proximal (40 nm or closer) to each other can they be ligated. Remarkably, multiple foci were observed in the cells incubated with antibodies against Scrib and α-Cat, in contrast to almost zero in the control (Figures 1C,D). Notably, these foci were mainly detected at the cell periphery, consistent with their cellular localization. These data further validate the interaction between Scrib and α-Cat in HEK293T cells.

Next, we investigated the roles of Scrib and α-Cat interaction. In A375 cells the nuclear translocation of YAP is significantly reduced when cells become confluent (Fig. S1A, B). Both α-Cat and Scrib have been shown to be involved in YAP regulation in mammalian cells (Silvis et al., 2011; Totaro et al., 2018). To interrogate whether Scrib can regulate YAP through α-Cat, we performed RNAi-mediated knockdown (KD) of Scrib or α-Cat in confluent A375 cells. Both Scrib and α-Cat KD significantly increase the nuclear translocation of YAP (Fig. S1C-E). Furthermore, YAP protein signal is higher upon reduction of α-Cat or Scrib (Fig. S1C, D). This can be interpreted as cytosolic retention of YAP, ultimately leading to its degradation, which can be avoided when released upon knockdown of α-Cat or Scrib. To further understand how Scrib and α-Cat interact in A375 cells, we employed the Duolink PLA assay. Multiple foci were observed in the cells incubated with antibodies against Scrib and α-Cat (Fig.S1F, G). These data further support that the interactions between Scrib and α-Cat are conserved in mammalian epithelial cells.

We next address how Scrib is involved in YAP localization while HEK293T cells are maintained at low density. Ectopic expression of Scrib is sufficient to suppress the nuclear translocation of YAP in these cells (Figures 2A,B). To further understand the mechanism by which Scrib regulates YAP, the LRR domain of Scrib, α-Cat or Scrib-LRR:α-Cat, a chimera comprising the Scrib LRR domain and α-Cat, were expressed. Expression of LRR domain or α-Cat barely affects ratio of nuclear/cytoplasmic (N/C) YAP localization. Remarkably, cells with Scrib-LRR:α-Cat expression show lower nuclear translocation of YAP (Figures 2A,B). We then wondered whether α-Cat is required for Scrib-mediated reduced nuclear localization of YAP. To test this idea, ectopic expression of Scrib and KD of α-Cat were combined. The ratio of N/C of YAP is largely reversed upon KD of α-Cat, suggesting that α-Cat is indispensable for Scrib-mediated YAP suppression (Figures 2C,D). Taken together, these data demonstrate that association with α-Cat is important for Scrib’s role in YAP suppression, and Scrib can regulate YAP through α-Cat.

We then hypothesized that spatiotemporal control of Scrib expression plays an important role in epithelial tissue development. To understand how Scrib expression is regulated in developing tissues, an in vivo model is crucial. By employing the Drosophila larval wing imaginal disc, an exhaustively studied in vivo system for understanding epithelial homeostasis and growth, we investigated protein distributions of Scrib and E-Cad, a key component of AJs. Both Scrib and E-Cad show similar spatial distributions in the late third instar wing imaginal disc, when mitosis gradually decreases, with lower peripheral and higher medial expression (Figure 3A) (Pan et al., 2018). Intriguingly, these patterns are complementary to Yki signal, for which Ex-lacZ expression is a readout (Figure 3A) (Pan et al., 2018). We then analyzed Scrib and E-Cad expression in the second instar wing disc, in which mitosis is very active and ubiquitously observed in the wing pouch (Pan et al., 2018). Yki signal is ubiquitously expressed adjacent to the compartment boundary, and expression of both Scrib and E-Cad remains ubiquitous (Figure 3B). To quantify the Scrib expression level at different stages, we used GFP trap lines (Scrib:GFP) in which scrib expression is under the endogenous scrib enhancer (Kelso et al., 2004; Pan et al., 2018). GFP signal was collected under identical conditions to compare signaling intensities. Our results reveal that Scrib expression in the medial of the wing pouch increases at the late third instar in correspondence with down-regulation of Yki activity (Figures 3C–E). We then wondered how α-Cat expression is regulated. By using GFP trap lines (α-Cat:GFP) in which α-cat expression is under the endogenous α-cat enhancer, GFP signal was quantified in different positions. α-Cat expression is higher in the center than the periphery of the wing pouch at the late third instar (Figures 3F,G). Taken together, these data reveal that spatiotemporal expression of Scrib and α-Cat is tightly controlled in developing epithelial tissue.

We next hypothesized that the amount of available Scrib is crucial for growth control in wing development. If this is the case, ectopic expression of Scrib alone may change the tissue size through modulating Yki activity. By employing an UAS-GAL4 system in Drosophila, Scrib or related proteins were overexpressed in the developing wing. Ectopic expression of Scrib or Scrib-LRR:α-Cat chimera, but not Scrib-LRR, resulted in a significantly smaller wing (Figures 4A,B). These results suggest that amount of Scrib functions as a growth signal. To understand the molecular mechanisms underlying this, we studied the wing imaginal disc. Our data reveal that ectopic expression of Scrib or Scrib-LRR:α-Cat chimera, but not Scrib-LRR, significantly suppresses Yki acvitity (Figure 4C). Since all the UAS transgenic flies were generated by the site-specific insertion system, the transcriptional level among different constructs are ensured to be equivalent (Gui et al., 2021). Accordingly, protein levels of different transgenic lines reflect protein stability (Figure 4C). These results suggest that Yki signal is regulated by amounts and quality of Scrib-related proteins.

To further address how the Scrib/α-Cat complex contributes to tissue growth and homeostasis, we employed mosaic analysis with a repressible cell marker (MARCM) clone analysis in the wing imaginal disc (Lee and Luo, 1999). We applied MARCM analysis by expressing various forms of Scrib in scrib null mutant (scrib ² ) clones (Figure 5A). It has been shown that cell competition takes place when scrib ² clones were generated in the wing imaginal disc (Chen et al., 2012). Since JNK signal was up-regulated in scrib ² clones to suppress Yki activity, apoptosis (marked by cDcp1) but not over proliferation was observed in scrib ² clones, thus the sizes of scrib ² clones are significantly smaller than control cells, (Figures 5B–D). These phenotypes are largely restored by expression of full-length Scrib (Figures 5B–D). Scrib-LRR partially restores clone sizes and reduces numbers of apoptotic cells, but with lower efficiency than full-length Scrib (Figures 5B–D). Intriguingly, clone sizes and number of apoptotic cells are fully rescued by expression of Scrib-LRR:α-Cat chimera protein (Figures 5B–D). Expression of deleted LRR-Scrib or α-Cat alone shows similar phenotypes to scrib ² clones (Figures 5B–D). Taken together, these results indicate that the LRR-Scrib and α-Cat complex sufficiently replaces full-length Scrib signaling for tissue growth and homeostasis when those cells are surrounded by wild type cells.