Podoplanin-positive cell-derived small extracellular vesicles contribute to cardiac amyloidosis after myocardial infarction

Abstract

Cardiac amyloidosis is a secondary phenomenon of an already pre-existing chronic condition. Whether cardiac amyloidosis represents one of the complications post myocardial infarction (MI) has yet to be fully understood. Here, we show that amyloidosis occurs after MI and that amyloid fibers are composed of macrophage-derived serum amyloid A 3 (SAA3) monomers. SAA3 overproduction in macrophages is triggered by exosomal communication from cardiac stromal cells (CSCs), which, in response to MI, activate the expression of a platelet aggregation-inducing type I transmembrane glycoprotein, Podoplanin (PDPN). CSC^PDPN+-derived small extracellular vesicles (sEVs) are enriched in SAA3, and exosomal SAA3 engages with macrophage by Toll-like receptor 2, triggering overproduction with consequent impaired clearance and aggregation of SAA3 monomers into rigid fibers. SAA3 amyloid deposits reduce cardiac contractility and increase scar stiffness. Inhibition of SAA3 aggregation by retro-inverso D-peptide, specifically designed to bind SAA3 monomers, prevents the deposition of SAA3 amyloid fibrils and improves heart function post MI.

Keywords: CP: Cell biology; CP: Metabolism; Podoplanin; Toll-like receptor 2; cardiac amyloidosis; myocardial infarction; retro-inverso D-peptide; serum amyloid A 3.

Figure 1.. CSC^PDPN+ sEVs impair cardiac function in healthy mouse hearts

(A) Echocardiography analysis showed a reduction of percent ejection fraction (%EF) and fractional shorting (FS) and an increase in end-systolic volume (ESV) and end-diastolic volume (EDV) in healthy mouse hearts injected with sEVs derived from pretreated CSC^PDPN+ or mouse lymphatic endothelial cells (mLECs) compared to mouse hearts injected with control sEVs (pretreated mouse cardiac endothelial cells [mCECs]) or saline. Data are presented as mean ± SEM. *p < 0.05, **p < 0.002, ***p < 0.0005, and ****p < 0.0001. N = 5–10. Ordinary two-way ANOVA and Tukey’s post hoc test were performed among the groups. (B and C) Histological characterization by Masson’s trichrome staining of mouse hearts injected with pretreated CSC^PDPN+ sEVs or mLECs sEVs showed infiltrative epicardial fibrosis (top) when compared to animals injected with control sEVs or saline (bottom). Scale bar: 0.25 mm. (D) Fibrotic tissue was characterized by fibronectin deposition (labeled in green; scale bar: 100 μm) and recruited CD45⁺ cells infiltrating the fibrotic tissue (D, bottom, labeled in red and quantified in E; scale bar: 10 μm). Data are presented as mean ± SEM. *p < 0.05. N = 5–10. Ordinary one-way ANOVA and Tukey’s post hoc test were performed among the groups.

Figure 2.. CSC^PDPN+ sEVs exclusively express SAA3

(A–C) sEVs derived from TNF-α- and AngII-pretreated CSC^PDPN+, mLECs, and control mCECs were characterized for size (A) and by sEV protein marker array (B) and transmission electron microscopy (C). Scale bar: 100 nm. (D) BMDM were treated with PKH 26-labeled (red) CSC^PDPN+ sEVs to show cell internalization. Scale bar: 25 μm. (E and F) Mass spectrometry analysis of protein content in sEVs derived from TNF-α- and AngII-pretreated mLECs and CSC^PDPN+ cells revealed that ~1,000 proteins are commonly expressed by the two groups of sEVs and that ~2,000 proteins were exclusively expressed in the pretreated CSC^PDPN+ sEVs (E). Within these 2,000 proteins, ~1,200 were upregulated after TNF-α and AngII treatment, as shown by the volcano plot (F, left) and the heatmap (F, right). (G) Western blot analysis of isolated sEVs derived from TNF-α- and AngII-pretreated CSC^PDPN+, mLECs, mCECs, and SAA3 knockout (KO)-CSC^PDPN+. (H) SAA3 is highly expressed in CSC^PDPN+ cell lysate after TNF-α and AngII pretreatment. Treatment of mCECs with TNF-α and AngII slightly increased SAA3 expression, which was undetectable in the cells’ lysate of CSC^PDPN+ isolated from global SAA3 KO mice following the same treatment. Data are presented as mean ± SEM. **p < 0.002 and ****p < 0.0001. N = 3–5. Ordinary one-way ANOVA and Tukey’s post hoc test were performed among the groups.

Figure 3.. SAA3-KO-CSC^PDPN+ sEVs did not impair cardiac function in healthy mouse hearts

(A and B) Healthy mouse hearts injected with sEVs isolated from TNF-α- and AngII-pretreated SAA3 KO-CSC^PDPN+ (A) did not show impairment in heart function during echocardiography analysis (B). Data are presented as mean ± SEM. N = 5. Ordinary two-way ANOVA and Tukey’s post hoc test were performed among the groups. (C and D) Treatment of healthy mouse hearts with sEVs isolated from pretreated SAA3 KO-CSC^PDPN+ failed to induce epicardial and infiltrative fibrosis (C, left) compared to sEVs derived from CSC^PDPN+ treatment (C, right; quantified in D; scale bar: 50 μm). Data are presented as mean ± SEM. **p < 0.002. N = 5–10. Student’s t test analysis and Tukey’s post hoc test were performed between the groups.

Figure 4.. CSC^PDPN+ sEVs are required to activate and induce SAA3 overexpression in BMDMs via TLR2 activation and p38-MAPK phosphorylation

(A) qPCR analysis of SAA3 and cytokine mRNA in BMDMs (MΦ) stimulated with sEV-depleted CSC^PDPN+ conditioned medium and with CSC^PDPN+ sEVs. Data are presented as mean ± SEM. *p < 0.05 and ***p < 0.0005. N = 3. Ordinary one-way ANOVA and Tukey’s post hoc test were performed among the groups. (B) Western blots showing total cell lysate of BMDMs treated either with recombinant SAA3 (rSAA3), sEVs derived from TNF-α- and AngII-pretreated CSC^PDPN+, control sEVs derived from similarly treated mCECs, or SAA3-null CSC^PDPN+. BMDMs treated with rSAA3 or CSC^PDPN+ sEVs specifically expressed SAA3 in cell lysates compared to BMDMs treated with control sEVs. (C) Western blot showing that BMDMs from TLR2 KO mice did not synthesize SAA3 upon any stimulation. (D) Quantification of SAA3 expression and release via qPCR, western blot, and ELISA analysis of wild-type and TLR2 KO BMDMs after different treatments. Data are presented as mean ± SEM. *p < 0.05, ***p < 0.0005, and ****p < 0.0001. N = 3–5. Ordinary one-way ANOVA and Tukey’s post hoc test were performed among the groups. (E) Western blot analysis and quantification (right) of p38-MAPK phosphorylation (T180/Y182) in BMDMs after treatment with rSAA3. (F) Western blot analysis and quantification (right) of p38-MAPK phosphorylation (T180/Y182) in BMDMs after treatment with sEVs isolated from TNF-α- and AngII-pretreated CSC^PDPN+, sEVs from similarly treated CSC^PDPN+ from SAA3 KO mice or mCECs. sEVs derived from treated CSC^PDPN+ from SAA3 KO mice or mCECs only, did not induce p38-MAPK phosphorylation in BMDMs compared to sEVs derived from pretreated CSC^PDPN+. Data are presented as mean ± SEM. *p < 0.05 and **p < 0.002. N = 3–6. Student’s t test analysis or ordinary one-way ANOVA and Tukey’s post hoc test were performed among the groups. (G) qPCR analysis of SAA3 mRNA in BMDMs shows an important reduction of SAA3 expression after specific inhibition of p38-MAPK with the SB 203580 compound. (H) Western blot analysis of SAA3 protein expression shows an important reduction of SAA3 synthesis after specific inhibition of p38-MAPK with the SB 203580 compound. Data are presented as mean ± SEM. *p < 0.05 and ***p < 0.0005. N = 3–6. Student’s t test analysis or ordinary one-way ANOVA and Tukey’s post hoc test were performed among the groups.

Figure 5.. SAA3 aggregates as extracellular amyloid deposits, leading to cardiac amyloidosis

(A and B) Mouse heart sections 30 days after myocardial infarction (MI) (left) and healthy hearts injected with TNF-α- and AngII-pretreated CSC^PDPN+ sEVs (center) were stained with thioflavin S (scale bar: 50 μm) and Congo red staining (B) (scale bar: 25 μm). Both showed the presence of amyloid deposits when compared with healthy hearts injected with similarly treated mCECs (a right). Congo red staining also showed specific birefringence (B) of amyloid deposits, observed with a polarized light microscope in mouse hearts 30 days after MI or after treatment with CSC^PDPN+ sEVs. (C) Immunohistochemistry analysis further showed that SAA, labeled in green, aggregated in ischemic tissue of mouse hearts 30 days after MI along with ECM, labeled with fibronectin in red (left), and sEVs isolated from pretreated CSC^PDPN+ were able to initiate SAA amyloidosis (center) when injected into healthy mouse hearts (right). Conversely, sEVs derived from pretreated SAA3-null CSC^PDPN+ failed to induce SAA amyloidosis. Scale bar: 50 μm. N = 7–10. (D and E) Cardiac sections obtained from failing human hearts showed SAA aggregation, labeled in green, in the fibrotic tissue (E) with the same pattern as SAA aggregation in mouse hearts 30 days after MI (D). Scale bar: 50 μm. N = 7–10.

Figure 6.. D-Peptide DRI-R5S reduces aggregation of SAA3 in vivo and in vitro

(A) D-Peptide DRI-R5S docking with the mouse SAA3 motif. (B) DRI-R5S reduced the aggregation of SAA3 in vitro in BMDMs conditioned medium after treatment with rSAA3. (C and D) In vivo, treatment with DRI-R5S reduced the aggregation of SAA3 in the scar border zone area of mouse hearts after MI. (E) Specific immune-histological staining for SAA showed reduced deposition of SAA in the ischemic area of mouse hearts 30 days after MI when compared with untreated mouse hearts 30 days after MI. Scale bar: 25 μm. Data are presented as mean ± SEM. **p < 0.002, ***p < 0.0005, and ****p < 0.0001. N = 3–5. Ordinary one-way ANOVA and Tukey’s post hoc test were performed among the groups.

Figure 7.. D-Peptide DRI-R5R improves cardiac function after MI

(A) Echocardiographic analysis of wild-type animals that underwent MI and were either treated or not with D-Peptide DRI-R5S. Treated animals showed improved cardiac function. (B and C) Masson’s trichrome staining (B) of wild-type animals that underwent MI and were either treated or not with D-Peptide DRI-R5S showed reduced cardiac scar size (C; scale bar: 100 μm) and better left ventricle wall composition (B, magnification; scale bar: 25 μm). Data are presented as mean ± SEM. *p < 0.05, **p < 0.002, and ***p < 0.0005. N = 7–10. Ordinary two-way ANOVA (echocardiography data) or Student’s t test analysis and Tukey’s post hoc test were performed among the groups.