Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 64 of the 64 references it cites were found in a registry or an index; 18 of the 33 sources its claims cite were checked for support.

Inactive β1-integrin acts as a junctional scaffold for angiopoietin/TIE2/FOXO1 signaling

Open access

Abstract

The study provides structural and mechanistic evidence that inactive α5β1-integrin can act as a junctional scaffold for angiopoietin/TIE2/FOXO1 signaling.

Overview

CoChat Summary

TL;DR

The paper provides structural and functional evidence that ANGPT2 binds simultaneously through its asymmetric FLDs to TIE2 and α5β1-integrin, with inactive β1-integrin at endothelial junctions acting as a scaffold for ANGPT/TIE2/FOXO1 signaling. It proposes that inactive α5β1-integrin promotes junctional TIE2 accumulation and FOXO1 nuclear exclusion, whereas ANGPT2 can engage active β1-integrin via its SCD when TIE2 is absent.

Key findings

  • ANGPT2 engages α5β1-integrin through both N- and C-terminal domains: the FLD binds α5Calf1 (Kd 50–90 nM for α5Calf1-2; MST WT α5Calf1 4 nM, mutants undetectable), while the SCD binds the β1-integrin PSI domain (Kd 2–5 nM for β1-ECD; ANGPT2 19–76 Kd 80 nM, ANGPT2 19–202 Kd 10 nM).
  • The α5Calf1-2 crystal structure was solved alone at 1.92 Å and in complex with an ANGPT2 peptide (E358–Q366); MD simulations and mutagenesis validated the interface, and SEC-SAXS supported a 2:1 ANGPT2–α5Calf1-2 complex and a trimeric ANGPT2–α5Calf1-2–TIE2 LBD complex (measured 165.6 kDa vs theoretical 177.5 kDa; favored model χ² = 1.9).
  • Inactive β1-integrin colocalized with ANGPT-induced TIE2 at EC-EC junctions; PLA detected TIE2 with inactive β1-integrin (mAb13) but not active β1-integrin (9EG7), and stabilizing inactive β1-integrin with AIIB2 increased junctional TIE2 and decreased nuclear FOXO1 in an ANGPT2- and TIE2-dependent manner.
  • In vivo, endothelial β1-integrin deletion in adult mice reduced venous total and pY992-TIE2, and LPS endotoxemia reduced junctional β1-integrin along with pY992-TIE2; oligomeric ANGPT2 (141 kDa dimer, 383 kDa oligomer) activated TIE2/AKT and decreased nuclear FOXO1, blocked by BAY-826 or LY294002.

Methods

  • Cell-based assays used HUVECs, human dermal BECs, LECs, CHO, HEK293FT, and insect cells; mouse experiments used endothelial-specific Itgb1 deletion in adult mice and LPS-induced endotoxemia, with both male and female mice examined.
  • Structural and biophysical approaches included SPR, MST, SEC-MALS, X-ray crystallography, SEC-SAXS, atomistic MD simulations (n = 3 repeats of 1 μs each), mutagenesis, proximity ligation assay, bulk RNA-Seq, VIPER, and GSEA.
  • Statistics used GraphPad Prism 9 with 2-tailed unpaired t tests or 1-way/2-way ANOVA followed by Tukey’s post hoc testing; P < 0.05 was considered significant, and data are presented as mean ± SEM or as indicated.

Limitations

  • Stated by the paper: the study primarily focuses on cell-based assays and protein–protein interactions and does not comprehensively capture all in vivo contexts of β1-integrin function in ANGPT/TIE signaling; ANGPT2 signaling modes may overlap depending on relative bioavailability and ANGPT1 occupancy of TIE2; ANGPT2 oligomerization was not assessed in vivo; α5β1-integrin was used as a tractable model because α9 integrin structure remains unresolved.
  • Evident from the text: in vivo quantitative analysis of junctional β1-integrin in the lymphatic vasculature was limited because β1-integrin expression was low along dermal capillaries and enriched at lymphatic valves.
  • Evident from the text: the full inactive α5β1-integrin–ANGPT2–TIE2 ternary model is based on SAXS, MD, mutagenesis, and previously published structures rather than a co-crystal structure of the complete ternary complex.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1172/jci190552

Cites

64 of 64 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 ANGPT2 and TIE2 differentially interact with active and inactive β1-integrin at endothelial junctions. (A) Surface staining of cultured BECs for inactive β1-integrin (mAb13) and VE-cadherin (CDH5). (B and C) Mander’s overlap coefficient (MOC) for mAb13/CDH5 colocalization in junctions (J) versus membrane (M) in BECs (B) and LEC (C) (image shown in Supplemental Figure 1B). (D) Surface staining of rhAng2-stimulated and control (Ctrl) (image shown in Supplemental Figure 1G) BECs for mAb13 and TIE2. (E) MOC for mAb13/TIE2 colocalization in junctions versus membrane in Ctrl and rhAng2-stimulated BECs, relative to Ctrl-treated BEC junctions. (F and G) Quantification of TIE2 (F) and mAb13 (G) in junctions in Ctrl and rhAng2-stimulated BECs. (H) MOC for TIE2/mAb13 colocalization in LECs. Image shown in Supplemental Figure 1F. (I) PLA design using antibodies for inactive β1-integrin (mAb13), active β1-integrin (9EG7), TIE2, and ANGPT2. Isotype-IgG as control. (J–O) PLA in HUVECs using indicated antibodies and quantification normalized to nuclei (J and K: TIE2/mAb13/9EG7; L and M: ANGPT2/mAb13; N and O: ANGPT2/9EG7 ± rhAng2 stimulation). Scale bars: 10 μm (A and D), 50 μm (J, L, and N). Confocal Z-stack projections are shown. n = 3 independent experiments (except in O, n = 4); 150 regions of interest per sample (B, C, and E–H). Statistical tests were a 2-tailed t test (B, C, F–H, and M), 1-way ANOVA with Tukey’s post hoc test (K and O), 2-way ANOVA with multiple correction (E). Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.
Figure 1 ANGPT2 and TIE2 differentially interact with active and inactive β1-integrin at endothelial junctions. (A) Surface staining of cultured BECs for inactive β1-integrin (mAb13) and VE-cadherin (CDH5). (B and C) Mander’s overlap coefficient (MOC) for mAb13/CDH5 colocalization in junctions (J) versus membrane (M) in BECs (B) and LEC (C) (image shown in Supplemental Figure 1B). (D) Surface staining of rhAng2-stimulated and control (Ctrl) (image shown in Supplemental Figure 1G) BECs for mAb13 and TIE2. (E) MOC for mAb13/TIE2 colocalization in junctions versus membrane in Ctrl and rhAng2-stimulated BECs, relative to Ctrl-treated BEC junctions. (F and G) Quantification of TIE2 (F) and mAb13 (G) in junctions in Ctrl and rhAng2-stimulated BECs. (H) MOC for TIE2/mAb13 colocalization in LECs. Image shown in Supplemental Figure 1F. (I) PLA design using antibodies for inactive β1-integrin (mAb13), active β1-integrin (9EG7), TIE2, and ANGPT2. Isotype-IgG as control. (J–O) PLA in HUVECs using indicated antibodies and quantification normalized to nuclei (J and K: TIE2/mAb13/9EG7; L and M: ANGPT2/mAb13; N and O: ANGPT2/9EG7 ± rhAng2 stimulation). Scale bars: 10 μm (A and D), 50 μm (J, L, and N). Confocal Z-stack projections are shown. n = 3 independent experiments (except in O, n = 4); 150 regions of interest per sample (B, C, and E–H). Statistical tests were a 2-tailed t test (B, C, F–H, and M), 1-way ANOVA with Tukey’s post hoc test (K and O), 2-way ANOVA with multiple correction (E). Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.CC BY · PubMed Central
Figure 2 Regulation of phosphorylated TIE2 and β1-integrin in cell junctions in Itgb1iECKO mice and in endotoxemia. (A and B) Vena cava from Itgb1iECKO and control (Ctrl) mice were stained for β1-integrin (HMb1), pY992-TIE2 (pTIE2), and VE-cadherin (CDH5). Larger fields of view are shown in Supplemental Figure 3A. (C and D) Quantification of junctional HMb1 (C) and junctional pTIE2 (D) normalized to CDH5. n = 3 mice per group. (E) Correlation of junctional pTIE2 with HMb1 in Ctrl and Itgb1iECKO mice per image. There was a positive correlation between pTIE2 and HMb1 (Pearson’s r = 0.590 [95% CI 0.303–0.778]; n = 32 images; 2-tailed P = 0.0004). (F and G) Vena cava from control mice and after LPS administration for 16 hours were stained for inactive β1-integrin (HMb1), pTIE2, and CDH5. Larger fields of view are shown in Supplemental Figure 3E. (H and I) Quantification of junctional HMb1 (H) and junctional pTIE2 (I) normalized to CDH5. n = 4 mice per group. (J) Correlation of junctional pTIE2 vs. HMb1. There was a strong positive correlation between pTIE2 and HMb1 (Pearson’s r = 0.877 [95% CI 0.763–0.938]; n = 33; 2-tailed P < 0.0001). (K and L) Western blot for phosphotyrosine and TIE2 after TIE2 immunoprecipitation from lung lysates from control and Itgb1iECKO mice and from control mice after LPS administration for 16 hours (K). Quantification of phosphorylated TIE2/total TIE2, normalized to control (Ctrl) (L). n = 9, 10 (Itgb1iECKO), and 4 (LPS) mice per group. Scale bars: 20 μm (B and G). Confocal Z-stack projections are shown. Statistical tests were as follows: 2-tailed t test (C, D, H, and I); 1-way ANOVA with Tukey’s post hoc test (L); and Pearson’s correlation (E and J). Data are presented as mean ± SD (C, D, H, and I)and ± SEM (L). *P < 0.05, **P < 0.01, ***P < 0.001. Tamoxifen, TAM.
Figure 2 Regulation of phosphorylated TIE2 and β1-integrin in cell junctions in Itgb1iECKO mice and in endotoxemia. (A and B) Vena cava from Itgb1iECKO and control (Ctrl) mice were stained for β1-integrin (HMb1), pY992-TIE2 (pTIE2), and VE-cadherin (CDH5). Larger fields of view are shown in Supplemental Figure 3A. (C and D) Quantification of junctional HMb1 (C) and junctional pTIE2 (D) normalized to CDH5. n = 3 mice per group. (E) Correlation of junctional pTIE2 with HMb1 in Ctrl and Itgb1iECKO mice per image. There was a positive correlation between pTIE2 and HMb1 (Pearson’s r = 0.590 [95% CI 0.303–0.778]; n = 32 images; 2-tailed P = 0.0004). (F and G) Vena cava from control mice and after LPS administration for 16 hours were stained for inactive β1-integrin (HMb1), pTIE2, and CDH5. Larger fields of view are shown in Supplemental Figure 3E. (H and I) Quantification of junctional HMb1 (H) and junctional pTIE2 (I) normalized to CDH5. n = 4 mice per group. (J) Correlation of junctional pTIE2 vs. HMb1. There was a strong positive correlation between pTIE2 and HMb1 (Pearson’s r = 0.877 [95% CI 0.763–0.938]; n = 33; 2-tailed P < 0.0001). (K and L) Western blot for phosphotyrosine and TIE2 after TIE2 immunoprecipitation from lung lysates from control and Itgb1iECKO mice and from control mice after LPS administration for 16 hours (K). Quantification of phosphorylated TIE2/total TIE2, normalized to control (Ctrl) (L). n = 9, 10 (Itgb1iECKO), and 4 (LPS) mice per group. Scale bars: 20 μm (B and G). Confocal Z-stack projections are shown. Statistical tests were as follows: 2-tailed t test (C, D, H, and I); 1-way ANOVA with Tukey’s post hoc test (L); and Pearson’s correlation (E and J). Data are presented as mean ± SD (C, D, H, and I)and ± SEM (L). *P < 0.05, **P < 0.01, ***P < 0.001. Tamoxifen, TAM.CC BY · PubMed Central
Figure 3 Oligomeric ANGPT2 induces TIE2/FOXO1 signaling in ECs. (A) Schematic representation of analysis of ANGPT2 dimers and oligomers in cell culture medium. (B–D) Representative immunoblots and quantification of oligomeric and dimeric ANGPT2 in 10× concentrated LEC- (B–D) and HUVEC-conditioned medium (C and D) under nonreducing and reducing conditions; rhAng2 was used as a control. n = 7 independent experiments. (E–G) Representative immunoblots (E) and quantification of phosphorylated (PY) and total TIE2 following immunoprecipitation from LECs stimulated with ANGPT2 (A2) oligomer ± TIE2 inhibitor (TIE2i, BAY-826) (n = 3 independent experiments) (F), and phospho-Ser473 AKT in whole-cell lysates from the same conditions (n = 4 independent experiments) (G). (H and I) Staining of HUVECs stimulated with oligomeric or dimeric ANGPT2 purified from CHO cells, or rhAng2, for CDH5 and FOXO1 (H). Quantification of nuclear/cytoplasmic FOXO1 ratio (I). n = 3 independent experiments. Asterisks mark cells with decreased nuclear FOXO1. (J and K) Staining of FOXO1 and VE-cadherin (CDH5) in LECs treated with A2 oligomer in the presence of PI3Ki (LY294002) or control (Ctrl) (J). Quantification of nuclear vs cytoplasmic FOXO1 ratio (K). (L) Bulk mRNA sequencing of LECs stimulated with A2 oligomer for 3 or 6 hours. Inferred transcription factor activity is shown as normalized VIPER enrichment scores (NES) for FOXO1, MYC, and E2F1 across treatments. Scale bars: 10 μm (H), 25 μm (J). Two-tailed unpaired t test (D), 1-way ANOVA with Tukey’s post hoc test (F, G, I, and K). NES values were compared across conditions by 1-way ANOVA with Tukey’s multiple comparison, and with Benjamini-Hochberg FDR correction across transcription factors (L). Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Figure 3 Oligomeric ANGPT2 induces TIE2/FOXO1 signaling in ECs. (A) Schematic representation of analysis of ANGPT2 dimers and oligomers in cell culture medium. (B–D) Representative immunoblots and quantification of oligomeric and dimeric ANGPT2 in 10× concentrated LEC- (B–D) and HUVEC-conditioned medium (C and D) under nonreducing and reducing conditions; rhAng2 was used as a control. n = 7 independent experiments. (E–G) Representative immunoblots (E) and quantification of phosphorylated (PY) and total TIE2 following immunoprecipitation from LECs stimulated with ANGPT2 (A2) oligomer ± TIE2 inhibitor (TIE2i, BAY-826) (n = 3 independent experiments) (F), and phospho-Ser473 AKT in whole-cell lysates from the same conditions (n = 4 independent experiments) (G). (H and I) Staining of HUVECs stimulated with oligomeric or dimeric ANGPT2 purified from CHO cells, or rhAng2, for CDH5 and FOXO1 (H). Quantification of nuclear/cytoplasmic FOXO1 ratio (I). n = 3 independent experiments. Asterisks mark cells with decreased nuclear FOXO1. (J and K) Staining of FOXO1 and VE-cadherin (CDH5) in LECs treated with A2 oligomer in the presence of PI3Ki (LY294002) or control (Ctrl) (J). Quantification of nuclear vs cytoplasmic FOXO1 ratio (K). (L) Bulk mRNA sequencing of LECs stimulated with A2 oligomer for 3 or 6 hours. Inferred transcription factor activity is shown as normalized VIPER enrichment scores (NES) for FOXO1, MYC, and E2F1 across treatments. Scale bars: 10 μm (H), 25 μm (J). Two-tailed unpaired t test (D), 1-way ANOVA with Tukey’s post hoc test (F, G, I, and K). NES values were compared across conditions by 1-way ANOVA with Tukey’s multiple comparison, and with Benjamini-Hochberg FDR correction across transcription factors (L). Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.CC BY · PubMed Central
Figure 4 Inactive β1-integrin facilitates ANGPT2/TIE2/FOXO1 signaling in LECs. (A) Schematic representation of ANGPT2/TIE2/FOXO1 pathway and AIIB2 antibody recognizing non–ECM-bound β1-integrin. (B and C) LECs stimulated with control or AIIB2 antibody for 60 minutes were stained for surface TIE2, CDH5, and AIIB2 (using secondary antibodies) (C). Quantification of junctional TIE2 normalized to number of nuclei (B). (D and E) LECs were stimulated with AIIB2 antibody for 30 and 60 minutes and stained for FOXO1 and AIIB2 (see Supplemental Figure 6A for staining at 60 minutes) (D). Quantification of nuclear/cytoplasmic FOXO1 ratio normalized to nuclei (E). (F and G) LECs were stimulated with ANGPT2 (A2) oligomer, rhAng2, or AIIB2 antibody and FOXO1 were analyzed using immunoblotting from nuclear and cytoplasmic fractions, using lamin A/C and α-tubulin markers, respectively. (H and I) shANGPT2- and shScr-treated LECs were stimulated with AIIB2 antibody for 60 minutes and stained for FOXO1, TIE2, and AIIB2. Nuclear/cytoplasmic FOXO1 ratio was quantified normalized to nuclei. n = 3 (B, E, G [n = 5 for control, n = 4 for A2 oligomer], and H) or n = 5 (E) independent experiments. Statistical tests included 2-tailed t test (B) and 1-way ANOVA with Tukey’s post hoc test (E, G, and H). Data are presented as mean ± SEM. Scale bars: 50 μm (C), 25 μm (D and I). Confocal Z-stack projections are shown. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Figure 4 Inactive β1-integrin facilitates ANGPT2/TIE2/FOXO1 signaling in LECs. (A) Schematic representation of ANGPT2/TIE2/FOXO1 pathway and AIIB2 antibody recognizing non–ECM-bound β1-integrin. (B and C) LECs stimulated with control or AIIB2 antibody for 60 minutes were stained for surface TIE2, CDH5, and AIIB2 (using secondary antibodies) (C). Quantification of junctional TIE2 normalized to number of nuclei (B). (D and E) LECs were stimulated with AIIB2 antibody for 30 and 60 minutes and stained for FOXO1 and AIIB2 (see Supplemental Figure 6A for staining at 60 minutes) (D). Quantification of nuclear/cytoplasmic FOXO1 ratio normalized to nuclei (E). (F and G) LECs were stimulated with ANGPT2 (A2) oligomer, rhAng2, or AIIB2 antibody and FOXO1 were analyzed using immunoblotting from nuclear and cytoplasmic fractions, using lamin A/C and α-tubulin markers, respectively. (H and I) shANGPT2- and shScr-treated LECs were stimulated with AIIB2 antibody for 60 minutes and stained for FOXO1, TIE2, and AIIB2. Nuclear/cytoplasmic FOXO1 ratio was quantified normalized to nuclei. n = 3 (B, E, G [n = 5 for control, n = 4 for A2 oligomer], and H) or n = 5 (E) independent experiments. Statistical tests included 2-tailed t test (B) and 1-way ANOVA with Tukey’s post hoc test (E, G, and H). Data are presented as mean ± SEM. Scale bars: 50 μm (C), 25 μm (D and I). Confocal Z-stack projections are shown. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.CC BY · PubMed Central
Figure 5 Direct interactions of α5β1-integrin with N- and C-terminal ANGPT2 domains. (A and B) Schematic representation of ANGPT2 (A) and α5β1-ECD (B) recombinant proteins used. (C) Binding of ANGPT2-FLD-Fc to immobilized α5β1-ECD in SPR. Two-state reaction kinetics were fitted to the original data. (D) Binding of α5β1-ECD (1.3 μM) to immobilized ANGPT2-N-Fc in SPR in the presence and absence of Mn2+ and Mg2+. (E) Binding of ANGPT2-FLD-Fc to immobilized α5Calf1-2 in SPR. (F) Binding of α5Calf1 to immobilized ANGPT2-FLD-Fc in SPR. (G) Schematic representation of N-terminal ANGPT2 and β1-ECD constructs. (H) Binding of β1-ECD to immobilized ANGPT2-N-Fc in SPR. (I) Binding of β1-head and -leg piece (β1-leg) to immobilized ANGPT2-N-Fc in SPR. (J and K) Binding of ANGPT219–202–SUMO (J) and ANGPT219–76-SUMO (K) to fluorescently labelled β1-ECD (2 nM) in MST. (L) Binding of ANGPT2-N-Fc to fluorescently labelled β1-PSI (20 nM) in MST. (M) Schematic representation of ANGPT2 binding through FLD and SCD domains to Calf1 and PSI domains of α5 and β1-integrins, respectively. Representative experiments with average KD are shown. n = 3 independent experiments (C–F and H–L). Nd, not determined.
Figure 5 Direct interactions of α5β1-integrin with N- and C-terminal ANGPT2 domains. (A and B) Schematic representation of ANGPT2 (A) and α5β1-ECD (B) recombinant proteins used. (C) Binding of ANGPT2-FLD-Fc to immobilized α5β1-ECD in SPR. Two-state reaction kinetics were fitted to the original data. (D) Binding of α5β1-ECD (1.3 μM) to immobilized ANGPT2-N-Fc in SPR in the presence and absence of Mn2+ and Mg2+. (E) Binding of ANGPT2-FLD-Fc to immobilized α5Calf1-2 in SPR. (F) Binding of α5Calf1 to immobilized ANGPT2-FLD-Fc in SPR. (G) Schematic representation of N-terminal ANGPT2 and β1-ECD constructs. (H) Binding of β1-ECD to immobilized ANGPT2-N-Fc in SPR. (I) Binding of β1-head and -leg piece (β1-leg) to immobilized ANGPT2-N-Fc in SPR. (J and K) Binding of ANGPT219–202–SUMO (J) and ANGPT219–76-SUMO (K) to fluorescently labelled β1-ECD (2 nM) in MST. (L) Binding of ANGPT2-N-Fc to fluorescently labelled β1-PSI (20 nM) in MST. (M) Schematic representation of ANGPT2 binding through FLD and SCD domains to Calf1 and PSI domains of α5 and β1-integrins, respectively. Representative experiments with average KD are shown. n = 3 independent experiments (C–F and H–L). Nd, not determined.CC BY · PubMed Central
Figure 6 Crystal structure and MD simulations reveal the interface of ANGPT2 FLD–α5Calf1-2 complex. (A) Crystal structure of apo α5Calf1-2. (B) Crystal structure of α5Calf1-2 in complex with the ANGPT2 peptide (blue) [residue E358–Q366, P(1)–P(9)]. (C) Schematic representation of ANGPT2 peptide (blue) bound to α5Calf1. (D) Electron density of the ANGPT2 peptide bound to α5Calf1-2 visible in the co-crystal structure in (B). The peptide is seen interacting with several residues in Calf1; see text for details. (E and F) Binding of WT (E) and mutant α5Calf1 (Arg694Trp, Asn724Val, Gln766Glu) (F, no binding is detected) to fluorescently labelled ANGPT2-FLD-Fc in MST. (G) An optimized docked model of an ANGPT2-FLD dimer in complex with α5Calf1-2. Residue E358–Q366, corresponding to the ANGPT2 peptide used in the crystal structure in (B), is indicated in bright blue. (H and I) Atom-scale MD simulation of ANGPT2-FLD dimer in complex with α5Calf12 (n = 3). Thickness of the line indicates frequency of interaction (H). Heatmap of interaction frequencies of amino acid residues of α5Calf1-2 (x axis) and ANGPT2-FLD dimer (y axis) from MD simulation (I). (J and K) Binding of mutant α5Calf1 (Arg755Glu, Gln762Ala, Asp764Asn (J) and Arg694Trp, Arg755Glu, Gln762Ala, Gln766Glu) (K) to fluorescently labelled ANGPT2-FLD-Fc in MST. No binding is detected in (K), and binding is decreased to a micromolar scale in (J) as compared with wt in (E). (L) Substituted amino acid residues shown in a close-up of the docked model in G. n = 3 independent experiments (E, F, J, and K).
Figure 6 Crystal structure and MD simulations reveal the interface of ANGPT2 FLD–α5Calf1-2 complex. (A) Crystal structure of apo α5Calf1-2. (B) Crystal structure of α5Calf1-2 in complex with the ANGPT2 peptide (blue) [residue E358–Q366, P(1)–P(9)]. (C) Schematic representation of ANGPT2 peptide (blue) bound to α5Calf1. (D) Electron density of the ANGPT2 peptide bound to α5Calf1-2 visible in the co-crystal structure in (B). The peptide is seen interacting with several residues in Calf1; see text for details. (E and F) Binding of WT (E) and mutant α5Calf1 (Arg694Trp, Asn724Val, Gln766Glu) (F, no binding is detected) to fluorescently labelled ANGPT2-FLD-Fc in MST. (G) An optimized docked model of an ANGPT2-FLD dimer in complex with α5Calf1-2. Residue E358–Q366, corresponding to the ANGPT2 peptide used in the crystal structure in (B), is indicated in bright blue. (H and I) Atom-scale MD simulation of ANGPT2-FLD dimer in complex with α5Calf12 (n = 3). Thickness of the line indicates frequency of interaction (H). Heatmap of interaction frequencies of amino acid residues of α5Calf1-2 (x axis) and ANGPT2-FLD dimer (y axis) from MD simulation (I). (J and K) Binding of mutant α5Calf1 (Arg755Glu, Gln762Ala, Asp764Asn (J) and Arg694Trp, Arg755Glu, Gln762Ala, Gln766Glu) (K) to fluorescently labelled ANGPT2-FLD-Fc in MST. No binding is detected in (K), and binding is decreased to a micromolar scale in (J) as compared with wt in (E). (L) Substituted amino acid residues shown in a close-up of the docked model in G. n = 3 independent experiments (E, F, J, and K).CC BY · PubMed Central
Figure 7 SAXS models of N-terminal and C-terminal ANGPT2 complexes with α5 and β1-integrin. (A) Schematic representation of ANGPT2-SCD and β1-integrin recombinant proteins used for SAXS. (B–D) SEC-SAXS of ANGPT219–76–SUMO in complex with β1-head. The SEC elution profile (B); fit of ab initio using DAMMIF (green) and homology model, rigid body refined using SASREF (black), with SAXS data (blue) (C); and overlay of ab initio and homology models (D). (E) Schematic representation of the N-terminal truncated dimeric ANGPT2147–496, α5Calf1-2 and TIE2 (LBD, red) used for SPR and SAXS. (F) Binding of ANGPT2147–496 to α5Calf1-2 in SPR. (G–I) SEC-SAXS analysis of ANGPT2147–496 in complex with α5Calf1-2. The SEC elution profile (G), fit of ab initio (green) and homology model (black) with SAXS data (blue) (H), ab initio (DAMMIF) and homology models (SREFLEX) (I). (J–L) SEC-SAXS analysis of ANGPT2147–496 in complex with α5Calf1-2 and TIE2 LBD. The SEC elution profile (J), fit of ab initio using DAMMIF (green) and homology model rigid body refined using SREFLEX (black) with SAXS data (blue) (K), and overlay of ab initio and homology models (L). The χ² value is presented as a measure of the model’s goodness of fit and statistical confidence (C, H, and K). For homology model, ANGPT219–76 (Supplemental Figure 6A), ANGPT2147–496 (Supplemental Figure 7A), β1-head (PDB ID: 7NXD), TIE2 LBD (PDB ID: 2GY7), and our crystal structure of α5Calf1-2 were used.
Figure 7 SAXS models of N-terminal and C-terminal ANGPT2 complexes with α5 and β1-integrin. (A) Schematic representation of ANGPT2-SCD and β1-integrin recombinant proteins used for SAXS. (B–D) SEC-SAXS of ANGPT219–76–SUMO in complex with β1-head. The SEC elution profile (B); fit of ab initio using DAMMIF (green) and homology model, rigid body refined using SASREF (black), with SAXS data (blue) (C); and overlay of ab initio and homology models (D). (E) Schematic representation of the N-terminal truncated dimeric ANGPT2147–496, α5Calf1-2 and TIE2 (LBD, red) used for SPR and SAXS. (F) Binding of ANGPT2147–496 to α5Calf1-2 in SPR. (G–I) SEC-SAXS analysis of ANGPT2147–496 in complex with α5Calf1-2. The SEC elution profile (G), fit of ab initio (green) and homology model (black) with SAXS data (blue) (H), ab initio (DAMMIF) and homology models (SREFLEX) (I). (J–L) SEC-SAXS analysis of ANGPT2147–496 in complex with α5Calf1-2 and TIE2 LBD. The SEC elution profile (J), fit of ab initio using DAMMIF (green) and homology model rigid body refined using SREFLEX (black) with SAXS data (blue) (K), and overlay of ab initio and homology models (L). The χ² value is presented as a measure of the model’s goodness of fit and statistical confidence (C, H, and K). For homology model, ANGPT219–76 (Supplemental Figure 6A), ANGPT2147–496 (Supplemental Figure 7A), β1-head (PDB ID: 7NXD), TIE2 LBD (PDB ID: 2GY7), and our crystal structure of α5Calf1-2 were used.CC BY · PubMed Central
Figure 8 Proposed models of ANGPT2 complexes with α5β1-integrin and TIE2. (A and B) Surface model of dimeric ANGPT2147–496 in complex with TIE2 ECD and inactive conformation of α5β1-integrin ECD in 2 orientations with 1 (A) or 2 (B) integrin heterodimers. ANGPT2147–496 dimer in complex with the TIE2 LBD and Calf1-2 domains of α5 integrin are derived from the SAXS data in this report. The TIE2 LBD–ANGPT2-FLD interaction is consistent with the report of Barton et al. (16). An array of TIE2 ECD dimers (blue and gray) is based on interactions of fibronectin type III domain 2 (Fn2) and Fn3 of TIE2, according to Moore et al. (54) and Leppänen et al. (13, 55) and the inactive conformation of the α5β1-ECD according to Schumacher et al. (56). (C and D) Proposed mode of multimeric ANGPT2 signaling as a TIE2 agonist, in complex with inactive α5β1-integrin, leading to inactivation of FOXO1 (C). ANGPT2 acts as an α5β1-integrin agonist in the absence of TIE2, involving its unique N-terminal SCD interacting with β1-integrin PSI (D). This may involve clustering of integrin heterodimers. Image is not to scale.
Figure 8 Proposed models of ANGPT2 complexes with α5β1-integrin and TIE2. (A and B) Surface model of dimeric ANGPT2147–496 in complex with TIE2 ECD and inactive conformation of α5β1-integrin ECD in 2 orientations with 1 (A) or 2 (B) integrin heterodimers. ANGPT2147–496 dimer in complex with the TIE2 LBD and Calf1-2 domains of α5 integrin are derived from the SAXS data in this report. The TIE2 LBD–ANGPT2-FLD interaction is consistent with the report of Barton et al. (16). An array of TIE2 ECD dimers (blue and gray) is based on interactions of fibronectin type III domain 2 (Fn2) and Fn3 of TIE2, according to Moore et al. (54) and Leppänen et al. (13, 55) and the inactive conformation of the α5β1-ECD according to Schumacher et al. (56). (C and D) Proposed mode of multimeric ANGPT2 signaling as a TIE2 agonist, in complex with inactive α5β1-integrin, leading to inactivation of FOXO1 (C). ANGPT2 acts as an α5β1-integrin agonist in the absence of TIE2, involving its unique N-terminal SCD interacting with β1-integrin PSI (D). This may involve clustering of integrin heterodimers. Image is not to scale.CC BY · PubMed Central

Verification

Claims the paper makes

13 Valid · 0 Not valid · 7 Not checked · 119 Own results and interpretations

96 more claims · 3 with their cited sources checked

Create a free account to see all 139 claims and to verify the 4 references still open.

See all claims — free account
  • Probably not a substantial claim. 9
  • It may state two separate claims as one. 30
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Citations

Cites

Show all 64 references

Metadata

Authors
Sipilä, T., Ponna, S. K., Venkatesha Murthy, A., Pink, A., Enkavi, G., Bohra, S. K., Lewna, K., Ganesh, K., Liu, Q., Korhonen, M., Kajander, T., Potente, M., Ivaska, J., Vattulainen, I., Leppänen, V.-M., & Saharinen, P.
Affiliations
  • Translational Cancer Medicine Program, Research Programs Unit,
  • Department of Physics, and
  • University of Helsinki (FI)
  • Max Delbrück Center (DE)
  • Åbo Akademi University (FI)
Corresponding author
Saharinen, Pipsa
Venue
Journal of Clinical Investigation
Published
2026-06-15
Volume
136
Issue
12
ISSN
  • 0021-9738 (ISSN-L)
  • 1558-8238
OpenAlex
W7164738136
Cited by
0
References
64
FWCI
0.00
Type
journal article
Language
en
Open access
Locations
  • Journal of Clinical Investigation (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • MDC Repository (Max-Delbrueck-Center for Molecular Medicine) (submittedVersion)
  • Työväentutkimus Vuosikirja (submittedVersion)
  • UTUPub (University of Turku) (submittedVersion)
Funders
  • European Research Council (773076)
  • Research Council Finland (346134, 310075,347771,331349,336234)
  • Human Frontier Science (RGP0059/2019)
Topics
  • Health Sciences › Medicine › Immunology and Allergy › Cell Adhesion Molecules Research
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Angiogenesis and VEGF in Cancer
  • Health Sciences › Medicine › Oncology › Lymphatic System and Diseases
Keywords
beta1 integrin, alpha5beta1 integrin, angiopoietin-2, Tie2, FOXO1, endothelial cells, cell-cell contact, receptor cross-talk, fibrinogen-like domain, nuclear exclusion, endotoxemia
MeSH terms
Forkhead Box Protein O1, Animals, Humans, Intercellular Junctions, Signal Transduction, Mice, Knockout, Integrin beta1, Integrin alpha5beta1, Angiopoietin-2, Receptor, TIE-2, Mice

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Inactive β1-integrin acts as a junctional scaffold for angiopoietin/TIE2/FOXO1 signaling. https://app.cochat.ai/papers/10.1172/jci190552
The paper
Sipilä, T., Ponna, S. K., Venkatesha Murthy, A., Pink, A., Enkavi, G., Bohra, S. K., Lewna, K., Ganesh, K., Liu, Q., Korhonen, M., Kajander, T., Potente, M., Ivaska, J., Vattulainen, I., Leppänen, V.-M., & Saharinen, P. (2026). Inactive β1-integrin acts as a junctional scaffold for angiopoietin/TIE2/FOXO1 signaling. Journal of Clinical Investigation. https://doi.org/10.1172/jci190552

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.