Fig 1.
Role of LIM domain and IDR regions of LIMD1 in binding to LATS1/2 and recruiting LATS1 to adherens junctions.
(A) Schematic diagram of LIMD1 protein, with amino terminal helical region (Helix), intrinsically disordered region (IDR), and LIM domains (1, 2, and 3) indicated. Conserved residues required for binding to strained F-actin (F512, F575 and Y646) are indicated on their respective LIM domains. (B-C) The ability of LIM domain and IDR regions of LIMD1 to bind with LATS1/2 were tested by co-immunoprecipitation. (B) LATS2 was tested for binding to full-length (LIMD1), the amino-terminal half (1-467), or the carboxy-terminal half (468-676) of LIMD1 by co-immunoprecipitation. GFP (Control plasmid) was used as a control. GFP, GFP-LIMD1 variants, and LATS2-FLAG were separately transfected in HEK293A cells. HEK293A cell lysates transfected with GFP/GFP-LIMD1 variants were combined with LATS2-FLAG lysates. Anti-FLAG or control (IgG) antibodies were used to isolate immune complexes. Immune complexes and lysates were probed by Western blotting for GFP/GFP-LIMD1 variants (GFP) and LATS2-FLAG (LATS2). (C) LATS1 was tested for binding to full-length (LIMD1), the amino-terminal half (1-467), or the carboxy-terminal half (468-676) of LIMD1 by co-immunoprecipitation. GFP (Control plasmid) was used as a control. GFP, GFP-LIMD1 variants, and LATS1-3xMyc were separately transfected in HEK293A cells. HEK293A cell lysates transfected with GFP/GFP-LIMD1 variants were combined with LATS1-3xMyc lysates. Anti-Myc or control (IgG) antibodies were used to isolate immune complexes. Immune complexes and lysates were probed by Western blotting for GFP/GFP-LIMD1 variants (GFP) and LATS1-3Myc (LATS1). (D-E) LIMD1 knockout (LIMD1-KO) MCF10A cells stably expressing GFP tagged wild-type LIMD1 and LIMD1 variants were established by lentiviral transduction and imaged using fixed and live-cell imaging. (D) The indicated cell lines were stained using anti-LATS1 and anti-TRIP6 antibodies. Merged images show LATS1 (red), TRIP6 (green) and DNA (blue). (E) The GFP-LIMD1 fusion expressing cell lines from part (D) were imaged live for GFP fluorescence. (F) Quantification of the LATS1 junction to cytoplasm mean intensity ratio for cells in (D). Error bars represent the standard deviation. ANOVA statistical comparisons between wild-type LIMD1 and LIMD1 variants are indicated above the plot (mean ± SD; n = 5; ****P < 0.0001, ns non-significant).
Fig 2.
Localization of LIM domain deletion variants of LIMD1 and their ability to recruit LATS1 to adherens junctions.
(A) Schematic diagrams of the different LIM domain deletion variants of LIMD1 used. Each LIM domain is designated by a number (1, 2, or 3). Single horizontal lines denote regions deleted. (B-C) LIMD1 knockout MCF10A cells stably expressing GFP tagged LIMD1 deletion mutants shown in (A) were established by lentiviral transduction and imaged using fixed and live-cell imaging. (B) The indicated cell lines were stained using anti-LATS1 and anti-TRIP6 antibodies. Merged images show LATS1 (red), TRIP6 (green) and DNA (blue). (C) The indicated cell lines from part (B) were imaged live for GFP fluorescence. (D) Quantification of the LATS1 junction to cytoplasm mean intensity ratio. Error bars represent the standard deviation. ANOVA statistical comparisons between wild-type LIMD1 and LIM domain deletion variants of LIMD1 are indicated above the plot using different letters (mean ± SD; n ≥ 4; **P = 0.01 (a), ***P = 0.001 (c), ****P < 0.0001 (b), ns non-significant (d)).
Fig 3.
Effect of mechanical strain insensitive LIMD1 mutants on LATS1 recruitment to adherens junctions and LATS1/2 binding.
(A-B) LIMD1-KO MCF10A cells stably expressing GFP tagged wild-type LIMD1 (WT) or LIMD1 strain insensitive mutants (F512A, F575A, and Y646A, see also Fig 1A.) were established by lentiviral transduction and imaged using fixed and live-cell imaging. (A) The indicated cell lines were stained using anti-LATS1 and anti-TRIP6 antibodies. Merged images show LATS1 (red), TRIP6 (green) and DNA (blue). (B) The indicated cell lines from part (A) were imaged live for GFP fluorescence. (C) Quantification of the LATS1 junction to cytoplasm mean intensity ratio. Error bars represent the standard deviation. ANOVA statistical comparisons between wild-type LIMD1 and mechanical strain insensitive LIMD1 mutants are indicated above the plot (mean ± SD; n = 5; ****P < 0.0001). (D-E) LATS1/2 were tested for their ability to bind to strain insensitive mutants of LIMD1 by co-immunoprecipitation. (D) LATS2 was tested for its ability to bind to wild-type (WT) and mechanical strain insensitive mutants (F512A, F575A or Y646A) of LIMD1 by co-immunoprecipitation. V5-tagged WT and mutants of LIMD1 and LATS2-FLAG were separately transfected in HEK293A cells. HEK293A lysates from cells transfected with V5-tagged WT and mutants of LIMD1 were combined with LATS2-FLAG lysates. Anti-LIMD1 or control (IgG) antibodies were used to isolate immune complexes. Immune complexes and lysates were probed by Western blotting for V5-tagged WT and mutants of LIMD1 (V5) and LATS2-FLAG (LATS2). (E) LATS1 was tested for binding to wild-type (WT) and mechanical strain insensitive mutants (F512A, F575A or Y646A) of LIMD1 by co-immunoprecipitation. V5-tagged WT and mutants of LIMD1, and LATS1-3xMyc were separately transfected in HEK293A cells. HEK293A lysates from cells transfected with V5-tagged LIMD1 variants were combined with LATS1-3xMyc lysates. Anti-Myc or control (IgG) antibodies were used to isolate immune complexes. Immune complexes and lysates were probed by Western blotting for V5-tagged WT and mutants of LIMD1 (LIMD1) and LATS1-3xMyc (LATS1). (F) Quantification of relative amounts of LATS2 (normalized to wild-type LIMD1 in the IP fraction) immunoprecipitated by wild-type LIMD1 and mechanical strain insensitive LIMD1 mutants in part (D). ANOVA statistical comparisons between wild-type LIMD1 and mechanical strain insensitive LIMD1 mutants are indicated above the graph (mean ± SD; n = 3; ***P = 0.001, ****P < 0.0001). (G) Quantification of relative amounts of wild-type LIMD1 and mechanical strain insensitive LIMD1 mutants (normalized to wild-type LIMD1 in the input fraction) immunoprecipitated by LATS1-3xMyc from part (E). ANOVA statistical comparisons between wild-type LIMD1 and mechanical strain insensitive LIMD1 mutants are indicated above the graph (mean ± SD; n = 3; *P < 0.05, ***P = 0.001).
Fig 4.
LATS-LATCH is necessary for LATS1/2 recruitment to AJs and for LATS1/2 binding to LIMD1.
(A) Live-cell imaging of MCF10A cells stably expressing GFP tagged LATS1 or LATS2 N-terminal regions (aa 1-635 in LATS1, aa 1-598 in LATS2) with and without (∆LATCH) the LATCH regions (aa 472-520 in LATS1, aa 418-466 in LATS2) as indicated. (B) AlphaFold2 model showing the three tandem LIM domains of LIMD1 (orange) and two regions of LATS2 (green) that are predicted to interact (LATS-LATCH and the Helical Hairpin). (C) Multiple sequence alignments of LATS1/2 from the indicated species showing the conserved LATCH region. (D-E) The requirement of the conserved LATCH sequence of LATS1/2 to bind with LIMD1 was tested by co-immunoprecipitation. (D) Full-length LATS2 (LATS2) or LATS2 with either the LATCH deleted (LATS2-∆LATCH, aa 418-466 deleted) or with the helical hairpin region deleted (LATS2-∆HH, aa 599-667 deleted) tagged to FLAG and (E) Full-length LATS1 (LATS1) or LATS1 with the LATCH deleted (LATS1-∆LATCH, aa 472-520 deleted) tagged with 3xMyc were tested for binding to LIMD1 by co-immunoprecipitation. FLAG-LATS2 variants, LATS1-3xMyc variants, and LIMD1-V5 were separately transfected in HEK293 cells. HEK293A cell lysates transfected with either FLAG-LATS2 variants (D) or LATS1-3xMyc variants (E) were combined with V5-tagged LIMD1. Anti-LIMD1 or control (IgG) antibodies were used to isolate immune complexes. Immune complexes and lysates were probed by Western blotting for LATS2 variants (FLAG) (D), LATS1 variants (Myc) (E) and LIMD1 (V5).
Fig 5.
The LATS-LATCH is sufficient to bind to LIMD1 and localize to AJs in a tension-dependent manner.
(A) Full-length LATS2, LATS2-LATCH and LATS2-LATCH-4mut, were tested for binding to LIMD1 by co-immunoprecipitation. Lysates were prepared from HEK293A cells separately transfected with GFP-tagged LATS2, LATS2-LATCH, LATS2-LATCH-4mut, or LIMD1-V5. Lysates from cells expressing LIMD1-V5 were mixed separately with those expressing the different GFP-LATS2/LATCH constructs, and anti-LIMD1 or control (IgG) antibodies were used to isolate immune complexes. Immune complexes and lysates were probed by Western blotting for LATS2/LATCH (GFP) and LIMD1 (V5). (B) Live-cell imaging of MCF10A cells stably expressing GFP tagged LATS2-LATCH or LATS2-LATCH-4mut. (C) Live-cell imaging of MCF10A cells stably expressing GFP tagged LATS2-LATCH treated with (Blebbistatin) or without (DMSO) Blebbistatin.