Advertisement

Inactivation of platelet-derived TGF-β1 attenuates aortic stenosis progression in a robust murine model

Rohan Varshney, Brennah Murphy, Sean Woolington, Shahrouz Ghafoory, Sixia Chen, Tyler Robison and Jasimuddin Ahamed

Data supplements

Article Figures & Data

Figures

  • Figure 1.

    Development of aortic stenosis in LDLR mice. (A) Stenosis parameters (fractional valve opening, WSS, and AV peak velocity) in LDLR mice on HFD (n = 10-48) and WT mice on HFD (n = 5-13) at 0, 3, and 6 months after initiation of HFD, as measured by echocardiography. (B, top) Ultrasound images of AV from LDLR mice at indicated time on HFD. (B, bottom) Ultrasound B-mode images with color Doppler of blood flow through aortic valves in LDLR mice at indicated timepoints. Yellow lines outline the aortic valves and blue lines indicate cusp separation at systole. Black line represents the Doppler line in the color Doppler images. (C) Fractional valve opening (right y-axis) and wall shear stress (left y-axis) in LDLR mice fed HFD at indicated timepoints (n = 7-31). Two-tailed unpaired Student t test with Welch’s correction was used to compare 2 groups with different sample sizes. Data represented as mean ± SEM throughout.

  • Figure 2.

    A robust and aggressive mouse model of aortic stenosis. (A) Ultrasound B-mode images (upper panels) and histological pictures (lower panels) of aortic valves in WT and LDLR mice on HFD. Yellow lines and black lines outline the AVs; arrows indicate AV leaflets. (B) Total valve area measured from echocardiographic images (n = 14 WT; n = 9 LDLR) and from histological images (n = 14 WT; n = 19 LDLR) of AV from mice at 6 months on HFD. (C) Pearson correlation of AV areas measured from echocardiography and histology (r = 0.8626; P < .0001; n = 16). (D-E) Immunohistochemistry for α-1 type I collagen (n = 4 WT; n = 5 LDLR) and picrosirius red staining for collagen (n = 8 WT; n = 17 LDLR) in AV from WT and LDLR mice after 6 months on HFD. Quantification of collagen staining in the valves by immunohistochemical (left) and picrosirius red (right) staining. (F) Pearson correlation of collagen positive area measured from immunohistochemical and picrosirius red-stained images of AV in LDLR mice after 6 months on HFD (r = 0.87; P = .055; n = 5). (G) Total TGF-β1 levels in the plasma of LDLR and WT mice at 0, 3, and 6 months on HFD, as measured by ELISA. (H) Pearson correlation of WSS across aortic valves and plasma TGF-β1 levels in LDLR mice at 6 months on HFD (r = 0.63; P = .002; n = 20). (I) Representative images of immunofluorescent staining for p-Smad2 in aortic valves from WT and LDLR mice after 6 months on HFD and quantification of p-Smad2–positive nuclei in the aortic valves of WT (n = 5) and LDLR (n = 10) mice after 6 months of HFD (see supplemental Figure 6 for individual valves with p-Smad2 staining). Two-tailed unpaired Student t test with Welch’s correction was used to compare 2 groups with different sample sizes. Data represented as mean ± SEM throughout.

  • Figure 3.

    Platelets are associated with valvular cells. (A) Whole-mount confocal staining (CD41 [green], CD62P [red], and 4′,6-diamidino-2-phenylindole [DAPI; blue]; upper-left and upper-middle panels) for activated platelets and corresponding SEM (valve surface) images of the same area (lower-left and lower-middle panels) in LDLR mice 6 months after HFD. Yellow arrows indicate activated platelets. High-magnification whole-mount staining (upper-right panel) and SEM (lower-right panel) images of aortic valve (from LDLR mice fed with HFD) show activated platelets are physically attached to valvular cells (yellow and green arrows). (B) Representative SEM images of aortic valve leaflets from WT and LDLR mice 6 months after HFD showing fewer activated platelets in WT than LDLR valves, indicated by yellow arrows (smaller platelet aggregates) and blue arrows (larger platelet aggregates). Quantification of the total number of platelet aggregates on the surface of valve leaflets in WT (n = 5) or LDLR (n = 6) mice analyzed after 6 months of HFD. Two-tailed unpaired Student t test was used to compare the 2 groups. Data represented as mean ± SEM throughout.

  • Figure 4.

    Platelet TGF-β1 contributes to AS progression. (A) ELISA and immunoblot for total platelet TGF-β1 (left and middle panel) in TGF-β1platelet-KO-LDLR mice (n = 5) and TGF-β1flox-LDLR controls (n = 5). Immunoblot with anti-TGF-β1 antibody showing 25 kD TGF-β1 bands (lanes 1, 2, 3, and 4 represent platelet releasates from TGF-β1flox-LDLR controls, and lanes 5, 6, 7, and 8 represent platelet releasates from TGF-β1platelet-KO-LDLR mice). Equal number of platelets (5 × 108) were used to prepare platelet releasates from each mouse. Total plasma TGF-β1 (right panel) in TGF-β1platelet-KO-LDLR mice (n = 45) and TGF-β1flox-LDLR controls (n = 14). (B) AS parameters: fractional valve opening and WSS in TGF-β1platelet-KO- LDLR mice (n = 15) and TGF-β1flox-LDLR controls (n = 7) at 6 months on HFD, as measured by echocardiography. (C) Representative images (left panel) of α-1 type I collagen stained aortic valves from TGF-β1platelet-KO- LDLR mice and TGF-β1flox-LDLR control mice after 6 months on HFD. Quantification of total valve area (middle panel) and collagen-positive area (right panel) in aortic valves of TGF-β1platelet-KO- LDLR mice (n = 4) and TGF-β1flox-LDLR controls (n = 4) at 6 months on HFD from immunohistochemical staining images. (D) Whole-mount confocal (maximum intensity projection) images showing activated platelets adjacent to valvular cells in a TGF-β1platelet-KO-LDLR mouse and a TGF-β1flox-LDLR control. Images also show intracellular p-Smad2 localization in the valvular cells. Whole-mount staining done for CD41 (green), CD62P (gray), p-Smad2 (red), and DAPI (blue). Quantification of nuclear or cytoplasmic p-Smad2 in aortic valves of TGF-β1platelet-KO-LDLR and littermate TGF-β1flox-LDLR control mice at 6 months on HFD (right panel). Two-tailed unpaired Student t test with Welch’s correction was used to compare 2 groups with different sample sizes. Data represented as mean ± SEM throughout.

  • Figure 5.

    VEC undergo mesenchymal transition. Aortic valve sections from WT and LDLR mice fed with HFD for 6 months were stained with vimentin (A), p-Smad2 (B), α-SMA (C), isolectinB4 (D), and α-1 type I collagen (E); nuclei were stained with DAPI (blue). Red arrows indicate a subset of cells expressing high levels of these markers in LDLR mice, but not WT. (F) Valve sections from WT and LDLR mice fed with HFD for 6 months were stained with vimentin and isolectinB4. (G) Quantification of double-positive (isolectinB4+vimentin), vimentin-positive cells, and total nucleated cells by counting DAPI. (H) Real-time polymerase chain reaction for Col1α1 in human umbilical vein endothelial cells stimulated with platelet-derived TGF-β1 for 3 hours, with or without anti-TGF-β1 antibody (AF-101-NA).

  • Figure 6.

    Platelet TGF-β1 contributes to VEC-mesenchymal transition. (A) Whole-mount aortic valve staining with isolectinB4 (green), vimentin (red), α-SMA (gray), and DAPI (blue) in littermate TGF-β1flox-LDLR control mice (top) and TGF-β1platelet-KO-LDLR mice (lower) at 6 months on HFD. (B) Higher-magnification images showing colocalization (yellow arrows) of isolectinB4, vimentin, and α-SMA in aortic valves from the 2 groups of mice. Quantification of α-SMA–positive area fraction in whole-mount images of aortic valves from TGF-β1flox-LDLR control mice and TGF-β1platelet-KO-LDLR mice after 6 months on HFD (right).