Loss of Tgif1 reduces osteoblast size in vivo and impairs osteoblast adhesion and migration in vitro.

(A) Representative images of the femora from 12-week-old Tgif1+/+ (n=4) and Tgif1-/- (n=4) mice stained with Toluidine blue. (B) Femora of 12-week-old Dmp1-Cre-;Tgif1fl/fland Dmp1-Cre+;Tgif1fl/fl mice stained with toluidine blue. (A, B) Arrows indicate osteoblasts. Scale bars indicate 100 µM. (C, D) Osteoblasts isolated from calvariae of neonatal Tgif1+/+and Tgif1-/- mice upon adherence on Col-I coated surfaces for 1, 2 and 4 hours, fixation and staining with toluidine blue. (C) Representative images of Toluide blue-stained cells after 2 hours of adhesion. (D) Quantification of adherent Tgif1+/+and Tgif1-/- calvarial osteoblasts after 1, 2, 4 hours of adhesion on Col-I-coated surfaces. (E) Osteoblasts isolated from long bones of 8-week-old Tgif1+/+ and Tgif1-/- mice upon adherence on Col-I coated surfaces for 1, 2 and 4 hours, fixation and staining with toluidine blue. Quantification of adherent cells at indicated time points. (F-H) Migration of calvarial osteoblasts obtained from neonatal Tgif1+/+ and Tgif1-/- mice was analyzed using live cell imaging. Quantification of (F) track length, (G) migration velocity and (H) meandering index of Tgif1+/+ and Tgif1-/- calvarial osteoblasts. (I) Long bones were harvested from Tgif1+/+ and Tgif1-/- mice. Osteoblasts were isolated from bone chips and spread on Col-I matrices placed on porous membranes for 48h. Membranes were frozen, cut and stained with (J) H&E or (K) phalloidin (yellow) and DAPI (blue). Quantification of cell area (L) and cell perimeter (M) of Tgif1+/+ and Tgif1-/-long bone osteoblasts on Col-I matrices. Scale bars indicate 50 µm. n = minimum of 3 independent experiments with technical duplicates. Unpaired t-test, *p<0.05, **p<0.01, ***p<0.001 vs. Tgif1+/+.

Tgif1-deficient osteoblasts are impaired to spread and form focal adhesions

(A, B) Calvarial osteoblast obtained from Tgif1+/+ and Tgif1-/- mice adhered on Col-I coated slides for 20, 40 and 60 minutes. (A) Cell spreading was visualized by Calcein-AM. (B) The number of round (non-spread) cells was counted. Scale bar indicates 100 µm in low magnification images and 20 µm in the insets. n=6 independent experiments; Repeated Measures ANOVA, Estimated margins means test, *p<0.05 between genotypes. (C) Tgif1+/+and Tgif1-/- calvarial osteoblasts were allowed to adhere on Col-I coated slides for 60 minutes. Focal adhesion formation was visualized by paxillin staining (green) and actin cytoskeleton by phalloidin (magneta). Scale bars indicate 100 µm (lower magnification) or 10 µm (single cells). (D) Analysis of single cell sphericity after 3D reconstruction using IMARIS and (E) quantification of focal adhesions using Image J. (F-H) Tgif1 was silenced in OCY454 cells using siRNA and cells were allowed to adhere on Col-I coated slides for 60 minutes. (F) Focal adhesion formation was visualized by Paxillin staining (green) and cell protrusions by phalloidin (magenta). (G) Analysis of cell sphericity after 3D reconstruction and (H) quantification of focal adhesions. Unpaired t-test, ***p<0.001, **p<0.01 vs. Tgif1+/+ (D, E) and scr (G, H).

Silencing of PAK3 restores cell spreading and focal adhesion formation in Tgif1-deficient cells.

(A) PAK3 mRNA expression in OCY454 cells transfected with siRNA targeting Tgif1 (siTgif1) or scrambled control (scr) siRNA before (0) and after 20 and 60 minutes of spreading on Col-I coated slides. n=9, 1-way ANOVA with Tukey’s multiple comparison test, *p<0.05, ***p<0.001. (B) Representative images of immunoblots demonstrating the abundance of PAK3 expression upon silencing of Tgif1. Actin was used as a control. (C) PAK3 mRNA expression in tibiae of 12-week-old Tgif1+/+(n=6) and Tgif1-/- (n=6) mice. (D) Tgif1 binding to the predicted site of the PAK3 promoter in OCY454 cells analyzed by ChIP and quantified as fold enrichment to the relative IgG control. Negative and positive (RAR) controls were used as indicated. n=8, unpaired t-test with Welch’s correction, *p<0.05, **p<0.01 vs. respective IgG control. (E) Tgif1-deficient (siTgif1) or control (scr) OCY454 cells were transfected with renilla plasmid and a pGL3 plasmid (EV) or a pGL3 plasmid containing a 2.3 kb fragment of the rat PAK3 promoter upstream of the luciferase gene. The promoter activity was quantified using luciferase assays and presented as relative luciferase activity (luciferase/renilla). 1-way ANOVA, Tukey’s multiple comparison test, ***p<0.001. (F-H) OCY454 cells were transfected alone or in combinations with siTgif1 for 48h, siPAK3 for 24 h and scrambled (scr) control. (F) Cells were allowed to adhere on Col-I coated slides for 60 minutes. Formation of focal adhesions was visualized by paxillin staining (green) and actin cytoskeleton by phalloidin staining (magenta). Scale bars indicate 100 µm (two upper rows) or 10 µm (two lower rows). (G) Quantification of cell sphericity using IMARIS. (H) Quantification of the number of mature focal adhesions per cell using the Image J software. n=4 independent experiments in which individual cells were analyzed. 1-way ANOVA, Tukey’s multiple comparison test, **p<0.01 ***p<0.001.

Tgif1 expression is increased during cell spreading via ERK and AP1 signaling pathways.

(A) Quantification of Tgif1 mRNA expression and (B) representative images of immunoblots of Tgif1 expression in calvarial osteoblasts before (0 min) and after 20 and 60 minutes of spreading. Actin was used as control. (C) Tgif1 mRNA expression and (D) representative images of immunoblots of Tgif1 expression in OCY454 cells before (0 min) and after 20 and 60 minutes spreading on Col-I coated slides. Actin was used as control. (E) OCY454 cells were treated with DMSO as control or with 5 µM Actinomycin D for 15 minutes prior to adherence on Col-I coated slides for 20 and 60 minutes. Tgif1 mRNA was quantified before (0 min) and during spreading. (F) Representative immunoblot images demonstrating the efficiency of ERK1/2 inhibitor SCH772984 (1µg/ml) to prevent ERK1/2 phosphorylation (n=4). (G) Quantification of Tgif1 mRNA expression during spreading after 15 minutes pre-treatment with DMSO (control) or with the ERK inhibitor SCH772984. (H) Quantification of the Tgif1 promoter activity during cell spreading. OCY454 cells were transfected with a luciferase reporter plasmid encoding a 2.2 kb fragment of the Tgif1 promoter or progressive truncations thereof (1.7, 1.2 and 0.5 kb) along with a plasmid encoding renilla firefly as control. Upon spreading for 60 minutes, promoter activity was quantified using a dual luciferase reporter gene assay and presented as normalized luciferase activity (luciferase/renilla). (I) Quantification of the AP1 transcriptional activity during cell spreading. OCY454 cells were transfected with a 6X-TRE-luciferase reporter to determine AP-1 activity along with a plasmid encoding renilla firefly as control. Upon spreading for 60 minutes, promoter activity was quantified using a dual luciferase reporter gene assay and presented as normalized luciferase activity (luciferase/renilla). (J) Quantification of the activity of the wild-type (wt) Tgif1 2.2 kb promoter and of the same fragment bearing a mutant AP1 binding site (AP1 mut) during cell spreading. OCY454 cells were transfected either with a luciferase reporter plasmid encoding a wild-type (wt) 2.2 kb fragment of the Tgif1 promoter or with the same promoter in which the AP1 binding site has been mutated (mut) along with a plasmid encoding renilla firefly as control. Upon spreading for 60 minutes, the promoter activity was quantified using a dual luciferase reporter gene assay and presented as normalized luciferase activity (luciferase/renilla). 1-way ANOVA, Tukey’s multiple comparisons test, *p<0.05, **p<0.01, ***p<0.001 vs. respective control.

Tgif1 deficiency causes a reduced number and activity of osteoblasts during bone repair.

(A) Representative images of Toluidine blue-stained sections of bones from male Tgif1+/+(n=9) and Tgif1-/- mice (n=5) that received an open fracture of the midshaft tibia at 10 weeks of age followed by bone healing for 3 weeks upon intramedullary stabilization. Arrows indicate osteoblasts. Scale bars indicate 100 µm (left panels) and 50 µm (right panels). (B - E) Histological sections were used to quantify the histomorphometric parameters (B) bone volume per callus volume (BV/Cal.V), (C) osteoid surface per bone surface (OS/BS), (D) number of osteoblasts per bone surface (N.Ob/BS) and (E) osteoblast surface per bone surface (Ob.S/BS). (F) Representative µCT images of the tibiae of Tgif1+/+ and Tgif1-/- mice 21 days after fracture. Unpaired t-test, *p<0.05, **p<0.01 vs. Tgif1+/+.

PTH is impaired in activating quiescent bone surfaces in Tgif1-deficient mice.

(A) Toluidine blue staining of tibiae from 12-week-old Tgif1+/+ and Tgif1-/- mice treated with vehicle (veh) or PTH. Quiescent surfaces are indicated by red arrows. Representative images are shown. Scale bar indicates 100 µm. (B) Quantification of the percentage of quiescent surfaces per bone surface (BS). Tgif1+/+, veh n=10; Tgif1+/+, PTH n=10, Tgif1-/-, veh n=8; Tgif1-/-, PTH n=10. (C) Toluidine blue staining of tibiae from 12-week-old Dmp1-Cre-;Tgif1fl/fland Dmp1-Cre+;Tgif1fl/fl mice treated with vehicle (veh) or PTH. Quiescent surfaces are indicated by red arrows. Representative images are shown. Scale bar indicates 100 µm. (D) Quantification of the percentage of quiescent surfaces per bone surface (BS). Dmp1-Cre-;Tgif1fl/fl, veh n=8; Dmp1-Cre-;Tgif1fl/fl, PTH n=8, Dmp1-Cre+;Tgif1fl/fl, veh n=8; Dmp1-Cre+;Tgif1fl/fl, PTH n=8. 1-way ANOVA with Tukey’s multiple comparison test, *p<0.05, **p<0.01.

PTH promotes cell spreading via Tgif1-PAK3 signaling.

(A) OCY454 cells were transfected with siRNA against Tgif1 (siTgif1) or scramble control siRNA (scr) and treated with vehicle (veh) or PTH for 4h. Cells were allowed to adhere on Col-I coated slides for 60 minutes. The cytoskeleton was visualized by phalloidin staining (magenta). Scale bars indicate 100 µm (upper panel) or 10 µm (lower panel). (B) Cell sphericity was quantified using IMARIS software. 1-way ANOVA with Tukey’s multiple comparisons test was applied, **p<0.01, ***p<0.001. (C) PAK3 mRNA expression after 4 hours of PTH treatment. Unpaired t-test, **p<0.01 vs. veh. (D) Representative image of immunoblots demonstrating PAK3 and Tgif1 protein abundance upon PTH treatment in cells transfected with siTgif1 or scr. Actin was used as a loading control. (E) OCY454 cells were transfected with scr or Tgif1 siRNA for 48h and PAK3 or scr siRNA for 24 h and treated with vehicle (veh) or PTH for 4h. Cells adhered on Col-I coated slides for 60 minutes. The cytoskeleton was visualized by phalloidin staining (magenta). Scale bar indicates 50 µm. (F) Cell sphericity was quantified using IMARIS software. n=4, 1-way ANOVA, Tukey’s multiple comparison test, *p<0.05, ***p<0.001.