Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Aug 5;13(1):4562.
doi: 10.1038/s41467-022-32335-8.

Afadin couples RAS GTPases to the polarity rheostat Scribble

Affiliations

Afadin couples RAS GTPases to the polarity rheostat Scribble

Marilyn Goudreault et al. Nat Commun. .

Abstract

AFDN/Afadin is required for establishment and maintenance of cell-cell contacts and is a unique effector of RAS GTPases. The biological consequences of RAS complex with AFDN are unknown. We used proximity-based proteomics to generate an interaction map for two isoforms of AFDN, identifying the polarity protein SCRIB/Scribble as the top hit. We reveal that the first PDZ domain of SCRIB and the AFDN FHA domain mediate a direct but non-canonical interaction between these important adhesion and polarity proteins. Further, the dual RA domains of AFDN have broad specificity for RAS and RAP GTPases, and KRAS co-localizes with AFDN and promotes AFDN-SCRIB complex formation. Knockout of AFDN or SCRIB in epithelial cells disrupts MAPK and PI3K activation kinetics and inhibits motility in a growth factor-dependent manner. These data have important implications for understanding why cells with activated RAS have reduced cell contacts and polarity defects and implicate AFDN as a genuine RAS effector.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1. BioID identification of proteins with in vivo proximity to AFDN.
a Activation of RAS subfamily small GTPases is governed by differential binding to guanine nucleotides. In the GTP-bound state RAS interacts with effector proteins, including AFDN. b Domain organization of the AFDN protein. RA = RAS Association domain; FHA = forkhead-associated domain; DIL = dilute domain; PDZ = PSD95/DLG1/ZO-1 domain. These are followed by a proline rich region and a C-terminal helical domain involved in actin binding. c High-confidence BioID hits (FDR ≤ 0.01) for both isoforms of AFDN were categorized by Gene Ontology (GO) Cellular Component (CC; top) or GO Biological Process (BP; bottom). d Top 25 BioID preys with their associated domains. Available GO CC (square) or GO BP (circle) terms are colored as in c. The relative abundance, FDR and spectral count are represented by shaded circles and denoted as per the legend (bottom).
Fig. 2
Fig. 2. A SCRIB-AFDN complex is mediated by the FHA domain of AFDN and the PDZ1 domain of SCRIB.
a Localization of AFDN and SCRIB in HeLa cells. EGFP-AFDN and immunostained FLAG-SCRIB were imaged by confocal microscopy. Scale bar represents 10 µm. b Endogenous SCRIB and AFDN co-localize at cell contacts in MCF7 cells. Antibody specificity to endogenous AFDN/SCRIB is verified in Fig. 6a. Projections of z-stacks are at right and top of the merged images, with position shown by dashed yellow line. Scale bar represents 10 µm. c Full-length AFDN and SCRIB co-immunoprecipitate (IP). EGFP-tagged SCRIB and FLAG-tagged AFDN were co-expressed in HEK 293T cells. IP with anti-FLAG followed by Western blotting with anti-GFP revealed the complex. Vector alone was a control. d Endogenous AFDN and SCRIB co-precipitate from MDCKII cells. Following AFDN IP, an anti-SCRIB blot confirmed the complex. No AFDN or SCRIB were observed in an anti-Myc IP (control). e Seven fragments comprising the whole of AFDN were generated to map the interaction, generally encompassing the AFDN modular domains. f SCRIB is precipitated by the FHA domain of AFDN. GST-fusions of the 7 AFDN fragments were mixed with lysates expressing KRAS, activated KRAS-G12V, SCRIB, or the C/N-terminal regions of SCRIB. Full-length recombinant and endogenous SCRIB and its N-terminal half associated with AFDN FHA. Activated KRAS bound the RA domains. g Fragments of SCRIB used for mapping its interaction with AFDN. h AFDN was precipitated by the PDZ1 domain of SCRIB. GST-fusions of the SCRIB LRR or PDZ1 domains were mixed with lysates expressing AFDN. Both isoforms and endogenous AFDN were bound by PDZ1. i AFDN-SCRIB complex is dependent on the FHA domain of AFDN. Full-length AFDN or an FHA deletion mutant were co-transfected with SCRIB. Following anti-GFP IP, an anti-FLAG blot revealed SCRIB association is disrupted by AFDNΔFHA. j Complex between AFDN and SCRIB is dependent on the PDZ1 of SCRIB. Full-length or a PDZ1 deletion mutant of SCRIB were co-transfected with AFDN. Following anti-GFP IP, an anti-FLAG blot revealed SCRIBΔPDZ1 does not precipitate AFDN. All source data are provided in the Source Data files.
Fig. 3
Fig. 3. AFDN FHA and SCRIB PDZ1 interact directly in a phospho-independent manner.
a GST-FHA domains with distinct C-termini precipitate full-length SCRIB from lysate. b Lack of positively charged residues on the surface of the AFDN FHA domain. The electrostatic surface of MDC1 FHA domain (left) shows the position of K73 and R58, necessary for binding pThr (PDBid 3UNN). Electrostatic surface of the AFDN FHA domain (PDBid 1WLN) reveals no positive charges (right; dashed circle). Electrostatic potential was calculated using APBS and PyMol. c AFDN and SCRIB co-IP following incubation with phosphatase. EGFP-SCRIB and FLAG-AFDN were co-expressed and precipitated with anti-FLAG following treatment with phosphatase. d Purified SCRIB PDZ1 can interact directly with GST-tagged AFDN FHA. Fragments end at FHA domain residues 501-590 and could only bind SCRIB PDZ1 if extended past residue 580. GST alone was a control. e 1H/15N-HSQC spectra of the AFDN FHA domain. The extended FHA domain (371-580, black) overlays with the core FHA domain (371-514, red) and exhibits several additional peaks. The same peaks are broadened upon addition of unlabeled SCRIB PDZ1 (right, blue), verifying importance of the extended C-terminal region to SCRIB binding. f The extended C-terminal FHA region alone does not bind SCRIB. Mixing PDZ1 domain of SCRIB with GST-tagged FHA domain fragments revealed neither the core FHA domain (371-531) nor C-terminal extension (531-580) in isolation complex with PDZ1. g ITC established AFDN FHA binds SCRIB PDZ1 with a Kd of 14.8 µM, comparable to most PDZ domain interactions. h Plot of peak intensity ratios from 1H/15N-HSQC spectra of PDZ1 alone vs PDZ1 in the presence of 2-fold molar excess AFDN FHA domain. Widespread broadening was observed, most impacted are labelled by residue number (over 6X difference, dashed red line). i Surface representation of the amino acids in SCRIB PDZ1 involved in binding the AFDN FHA domain (PDBid 5VWK). Residues demonstrating the most broadening are coloured red, moderate broadening in orange and no broadening in grey. Unassigned residues are white. Position of the β-PIX peptide is characteristic of PDZ interactions with C-terminal motifs. All source data are provided in the Source Data files.
Fig. 4
Fig. 4. RAS GTPases complex with AFDN-SCRIB.
a Activated KRAS is co-precipitated with AFDN and SCRIB. HEK 293T cells were co-transfected with vectors expressing FLAG-AFDN and EGFP-KRAS-G12V (constitutively active). Both endogenous SCRIB and KRAS-G12V were detected following anti-FLAG IP. FLAG vector alone was a control. b Activated KRAS and AFDN co-IP with SCRIB. HEK 293T cells were co-transfected with vectors expressing FLAG-SCRIB and EGFP-KRAS-G12V. Endogenous AFDN and KRAS-G12V were detected following anti-FLAG IP. c The RA1/2 domains of AFDN interact with multiple RAS small GTPases. EGFP-tagged GTPases with activating mutations were expressed in HEK 293T cells, and GST-tagged RBD domains from BRAF or AFDN purified from bacteria. GST-RBDs bound to glutathione beads were used to pull down (PD) GTPases, a Western blot with anti-GFP revealed interacting proteins. d Specificity of individual AFDN RA domains for RAS GTPases. GST-tagged RA1, RA2 or RA1/2 were purified and mixed with lysates expressing the indicated GTPases. An anti-GFP immunoblot following precipitation on glutathione beads revealed bound GTPases. Each blot was exposed for 30 seconds. e Activated GTPases co-precipitate with AFDN and augment interaction with SCRIB. Full-length, EGFP-tagged SCRIB or activated RAS family GTPases were co-transfected with FLAG-tagged AFDN. Following immunoprecipitation with anti-FLAG, Western blot with anti-GFP revealed activated KRAS, RAP2C or RAP1B complex with AFDN and SCRIB. f SCRIB lacking PDZ1 does not co-precipitate with AFDN and does not prevent the association between AFDN and KRAS. Full-length wild-type or ΔPDZ1 SCRIB were co-transfected with AFDN and KRAS. Association with SCRIB and KRAS was detected by Western blot following anti-FLAG immunoprecipitation of AFDN. g Quantitation of SCRIB binding to AFDN when co-expressed with activated GTPases. The amount of EGFP-SCRIB co-precipitating with AFDN in anti-FLAG Western blots was determined by densitometry (n = 4 for RAP1B/RAP2C and n = 7 for KRAS, from distinct experiments). The ratio is the amount of SCRIB detected when co-expressed with RAP1B-G12V (P = 0.018), RAP2C-G12V (P = 0.020) or KRAS-G12V (P = 0.007), versus wild-type KRAS. Line represents the median, box the interquartile range (IQR) and whiskers the min/max. **P < 0.01, *P < 0.05 as measured by paired, two-tailed t-test. All source data are provided in the Source Data files.
Fig. 5
Fig. 5. RAS GTPases co-localize with AFDN.
a Expression of wild-type KRAS does not alter AFDN or SCRIB localization. EGFP-KRAS was expressed in MCF7 cells, and localization of endogenous AFDN and SCRIB was detected by immunostaining and confocal microscopy. Projections of z-stacks are at right and top. Position of associated z projections are shown with dashed yellow lines in the merged image. Scale bars represent 10 µm. b KRAS-G12V co-localizes with AFDN and SCRIB in MCF7 cells. Following expression of EGFP-KRAS-G12V, endogenous AFDN and SCRIB were detected by immunostaining. Cells expressing KRAS are found in the monolayer (1), detaching from the monolayer (2) or on top of the monolayer (3). Projections of z-stacks are at right and bottom of the merged images and their position is marked with a dashed yellow line. Scale bars represent 10 µm. c Activated RAS GTPases recruit AFDN to the cell membrane. Cherry-AFDN was co-expressed with EGFP-tagged RAS or RAP GTPases in HeLa cells. EGFP alone or EGFP-tagged wild-type KRAS with AFDN were controls, as was Cherry alone with activated KRAS-G12V. Scale bars represent 10 µm.
Fig. 6
Fig. 6. KO of AFDN and SCRIB in MCF7 cells.
a Individual clones of MCF7 cells were selected following CRISPR KO of AFDN, SCRIB or both AFDN/SCRIB with two different guide sequences. Western blot with antibodies recognizing endogenous AFDN or SCRIB confirmed the proteins are not expressed. Source data are provided in the Source Data files. b Model of the SCRIB-AFDN-RAS GTPase complex based on our data. c KRAS does not complex with SCRIB in the absence of AFDN. FLAG-SCRIB was expressed in AFDN KO MCF7 cells together with EGFP-KRAS-G12V. Following IP with anti-FLAG, activated KRAS did not co-precipitate with SCRIB unless cells were also expressing EGFP-AFDN. Source data are provided in the Source Data files. d SCRIB is internalized and demonstrates a punctate pattern in AFDN KO MCF7 cells, which do not form cell-cell contacts (top). Expression of EGFP-KRAS-G12V in these cells does not alter SCRIB localization (bottom). Scale bars represent 10 µm. e AFDN remains predominantly at sites of cell contact in SCRIB KO MCF7 cells, which grow in multi-cell layers (top). EGFP-KRAS-G12V expression results in AFDN internalization and breakdown of cell-cell contacts (bottom). Scale bars represent 10 µm.
Fig. 7
Fig. 7. Co-localization of SCRIB and AFDN is dependent on their PDZ1 and FHA domains, respectively.
a Expression of EGFP-AFDN in AFDN KO MCF7 cells restores endogenous SCRIB localization at cell contacts. SCRIB is also recruited to cell contacts in AFDN KO cells expressing EGFP-AFDNΔFHA but appears poorly co-localized with AFDN. Projections of z-stacks are at right and top and their position is marked with a dashed yellow line in the merged image. Scale bars represent 10 µm. b EGFP-SCRIB co-localizes with endogenous AFDN in SCRIB KO MCF7 cells. EGFP-SCRIBΔPDZ1 is also enriched at cell contacts in SCRIB KO cells. Projections of z-stacks are at right and top and their position is marked with a dashed yellow line in the merged image. Scale bars represent 10 µm. c z-stack projections along a 10 µm length of cell contacts show exogenous, wild-type AFDN and endogenous SCRIB localization overlaps in AFDN KO cells (top, 3 independent cells). Conversely, expressed AFDNΔFHA remains apical to endogenous SCRIB (bottom). d z-stack projections show exogenous, wild-type SCRIB and endogenous AFDN co-localize in SCRIB KO cells (top) while SCRIBΔPDZ1 remains basal to endogenous AFDN (bottom).
Fig. 8
Fig. 8. Loss of AFDN or SCRIB disrupts ERK and AKT activation kinetics and cell motility in a growth factor-dependent manner.
Activation of the MAPK and PI3K pathways following EGF stimulation of MCF7 parental (a), AFDN KO (b) or SCRIB KO (c) cells. pERK or pAKT Western blots measure pathway activity 2.5, 5, 15, 30 or 45 minutes after addition of EGF. Total ERK and AKT Western blots confirmed expression and serve as loading controls. Quantification of MAPK (d; ERK activation) and PI3K (e; AKT activation) activity following addition of EGF over time. Center is the mean and error bars SD as derived from n = 3 independent replicates. f Effect of AFDN or SCRIB KO on cell proliferation compared to parental MCF7 cells over a 72-hour time course from n = 4 independent replicates. Center represents mean and error bars SD. **P < 0.005 (P = 0.003), ns = not significant (P = 0.298) as measured by two-way ANOVA. g KO of SCRIB or AFDN results in motility defects in response to EGF. Wound closure (%) was measured at 24, 48, 72 or 96 hours for MCF7 (P = 0.0001 within group), AFDN KO (P = 0.0145), SCRIB KO (P = 0.0001), and AFDN/SCRIB KO (P = 0.0001) cells. 100 ng/ml EGF was supplemented to the media at time 0 and every 24-hour interval thereafter. n values denote independent replicates. Line represents the median, box the IQR and whiskers min/max. ***P < 0.001, *P < 0.01 as measured by RM one-way ANOVA (within group) or two-way ANOVA (between group). P = < 0.0001 for all between group comparisons. h Phase contrast images of MCF7, SCRIB KO and AFDN KO cells during wound closure. 100 ng/ml EGF was supplemented to the media every 24 hours. Inset are enlarged images of cells at the wound edge. Scale bar represents 100 µm. i AFDN and SCRIB KO cells do not polarize towards the wound. Cells were stained for GM130 (Golgi), which typically orients towards the wound at the leading edge (see parental MCF7 cells; top). Cells were fixed 24 hours after wounding and EGF stimulation. Arrows indicate direction of motility. Scale bars represent 20 µm. All source data are provided in the Source Data files.

References

    1. Wohlgemuth S, et al. Recognizing and defining true Ras binding domains I: biochemical analysis. J. Mol. Biol. 2005;348:741–758. doi: 10.1016/j.jmb.2005.02.048. - DOI - PubMed
    1. Kiel C, et al. Recognizing and defining true Ras binding domains II: in silico prediction based on homology modelling and energy calculations. J. Mol. Biol. 2005;348:759–775. doi: 10.1016/j.jmb.2005.02.046. - DOI - PubMed
    1. Katz ME, McCormick F. Signal transduction from multiple Ras effectors. Curr. Opin. Genet. Dev. 1997;7:75–79. doi: 10.1016/S0959-437X(97)80112-8. - DOI - PubMed
    1. Rodriguez-viciana P, Sabatier C, Mccormick F. Signaling Specificity by Ras Family GTPases Is Determined by the Full Spectrum of Effectors They Regulate. Mol. Cell. Biol. 2004;24:4943–4954. doi: 10.1128/MCB.24.11.4943-4954.2004. - DOI - PMC - PubMed
    1. Colicelli J. Human RAS superfamily proteins and related GTPases. Sci. STKE. 2004;2004:RE13. doi: 10.1126/stke.2502004re13. - DOI - PMC - PubMed

Publication types

Grants and funding