Exosomal mRNAs for Angiogenic–Osteogenic Coupled Bone Repair

Abstract Regenerative medicine in tissue engineering often relies on stem cells and specific growth factors at a supraphysiological dose. These approaches are costly and may cause severe side effects. Herein, therapeutic small extracellular vesicles (t‐sEVs) endogenously loaded with a cocktail of human vascular endothelial growth factor A (VEGF‐A) and human bone morphogenetic protein 2 (BMP‐2) mRNAs within a customized injectable PEGylated poly (glycerol sebacate) acrylate (PEGS‐A) hydrogel for bone regeneration in rats with challenging femur critical‐size defects are introduced. Abundant t‐sEVs are produced by a facile cellular nanoelectroporation system based on a commercially available track‐etched membrane (TM‐nanoEP) to deliver plasmid DNAs to human adipose‐derived mesenchymal stem cells (hAdMSCs). Upregulated microRNAs associated with the therapeutic mRNAs are enriched in t‐sEVs for enhanced angiogenic–osteogenic regeneration. Localized and controlled release of t‐sEVs within the PEGS‐A hydrogel leads to the retention of therapeutics in the defect site for highly efficient bone regeneration with minimal low accumulation in other organs.

This PDF file includes: Supplemental Fig. S1-7 Fig. S1.pDNA maps used in this study and hAdMSCs after TM-nanoEP process at different voltages.Fig. S2.RNA standard curves or ladders, zeta potential and total RNA amount in different sEV cohorts, and TIRFM images of blank control.Fig. S3.TSC1/2 knockdown and quantitative results of intracellular activities from Western blot analysis.Fig. S4.Correlation between the pDNA concentration ratios of BMP-2 to VEGF-A used in TM-nanoEP and the actual mRNA ratios found in t-sEVsBone RNAs.Fig. S5.Supplemental information of exosomal miRNA profiling.Fig. S6.Characterization of customized PEGS-A/sEVs hydrogel.Fig. S7.In vivo distribution of therapeutic sEVs (t-sEVs) and morphometric evaluation of bone regeneration.Fig. S8.Expression of heat shock proteins (HSPs) and stress granules (SGs) within sEVs, and their biocompatibility.
Supplemental Fig. 2| RNA standard curves or ladders, total RNA amount in different sEV cohorts and TIRFM images of blank control.A) Distribution of synthetic BMP-2 and VEGF-A mRNAs (100 ng each mRNA).The sizes of the two mRNAs are similar and around 1,500 nt.B) Zeta potential of b-sEVs, e-sEVs PBS, and t-sEVs Bone RNAs.All the sEVs exhibit a negative charge, and there is no significant difference in zeta potential between transfected sEVs and native derived sEVs.C) RNA amount of b-sEVs, e-sEVs PBS, and t-sEVs Bone RNAs with the same sEV number (1×10 12 ).D) qPCR standard curves using synthetic mRNAs as internal standard.There are ~1.3 copies and ~1.8 copies of BMP-2 and VEGF-A per therapeutic sEV, respectively.E) Representative TIRFM images by using a single-sEV biochip for exosomal mRNA detection of b-sEVs from untreated hAdMSCs.Red dots: sEVs with VEGF-A mRNA; green dots: sEVs with BMP-2 mRNA; yellow dots: sEVs with both mRNAs (Scale bar: 10 μm).Few spots with fluorescence and weak mRNA expression are observed in b-sEVs, indicating blank sEVs carry few mRNAs.*P < 0.05, ***P < 0.005, ****P < 0.0001.All data are presented as mean ± SD.Student's t-test was used for comparison.Supplemental Fig. 3| TSC1/2 knockdown and quantitative results of intracellular activities from Western blot analysis.A) Effects of tuberous sclerosis complex 1/2 (TSC1/2) knockdown on hAdMSCs.Quantitative Western blot analysis of B) mTORC1 and autophagy markers (pS6 and LC3-II), and C) intracellular sEV markers (CD9 and CD63).The results show TM-nanoEP has no discernible effect on mTORC1 and autophagic activities in TSC1/2 −/− hAdMSCs, and there is no significant variation in exosomal markers within the cells following the TM-nanoEP process.*P < 0.05, **P< 0.01.All data are presented as mean ± SD.Student's t-test was used for comparison.Supplemental Fig. 4| Correlation between the pDNA concentration ratios of BMP-2 to VEGF-A used in TM-nanoEP and the actual mRNA ratios found in t-sEVsBone RNAs (n = 5).When the pDNA ratios used are 4:1, 2:1 and 1:1, the actual mRNA ratios of BMP-2 to VEGF-A are approximately 2.96:1, 1.60:1, and 0.62:1, respectively.The intrinsic VEGF mRNAs in sEVs from hAdMSCs could lead to decreased mRNA ratios in the sEVs.
Supplemental Fig. 6| Characterization of PEGS-A/sEVs hydrogel.A) Synthesis schematics of PEGylated PGS (PEGS) and PEGS-A hydrogel.B) 1 H NMR spectra of PEGS and PESG-A pre-polymers.C) Gelation time of PEGS-A hydrogels with different PEGS-A pre-polymer concentrations when the ratio of DTT to Borax is 3:1.Among them, 10% PEGS-A pre-polymer is unable to crosslink, while gelation of 50% PEGS-A pre-polymer occurs in 10 s.D) Stress-strain curves of PEGS-A hydrogels with different PEGS-A pre-polymer concentrations when the ratio of DTT to Borax is 3:1.E) SEM images of PEGS-A hydrogel and PEGS-A/t-sEVs with a ratio of DTT to Borax at 2:1.Yellow arrows indicate aggregation of sEVs.F) IL-6, G) TNF-α, (M1 markers), and H) IL-10 (M2 marker) secretion in conditioned media of RAW264.7 macrophages (Mφs) assessed by ELISA after incubation with 100 ng/mL lipopolysaccharides (LPS) and sEVs (b-sEVs, e-sEVsPBS, and t-sEVsBone RNAs).*P < 0.05, **P < 0.01 and ****P < 0.0001.All data are presented as mean ± SD.Student's t-test was used for comparison.Supplemental Fig. 7| In vivo distribution of therapeutic sEVs (t-sEVs) and morphometric evaluation of bone regeneration.A) Quantification of fluorescence intensity over a 14-day time period.The results are normalized to the fluorescence intensity at day 1.B) SEV distribution in main organs and femur defects.The analyses show that sEVs encapsulated in PEGS-A hydrogels can be locally delivered with low hepatic, splenic, and renal accumulation.Morphometric analyses of C) bone mineral density (BMD) and D) trabecular thickness (Tb.Th.) for pure and PEGS-A/sEVs hydrogel cohorts at 4-and 8-week after injection.*P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.0001, ####P < 0.0001.All data are presented as mean ± SD.Student's t-test or one-way ANOVA was used for comparison.Supplemental Fig. 8| Expression of heat shock proteins (HSPs) and stress granules (SGs) within sEVs, and their biocompatibility.A) Western blot analyses of HSPs (HSP70 and HSP90) and SGs expression (using G3BP Stress Granule Assembly Factor 1 (G3BP1) as a marker) in different sEVs.Both e-sEVsPBS and t-sEVsBone RNAs showed increased HSP expression when compared to the b-sEVs, suggesting the role of the thermal shock mechanism in TM-nanoEP-induced sEV biogenesis.Furthermore, these TM-nanoEP-induced sEVs demonstrated a significantly increased G3BP1 expression, indicating the presence of SG components in sEVs produced by TM-nanoEP.B) Cell viability of hBMSCs exposed to different sEVs at a dosage of 1×10⁶/cell over a 3-day period.The sEVs produced by TM-nanoEP (e-sEVsPBS and t-sEVsBone RNAs) showed no significant difference from the native sEVs (b-sEVs).

Table S4 . | KEGG pathway enrichment of miRNA target genes (t-sEVsBone RNAs vs. e-sEVsPBS).
Two signaling pathways (TGF-β signaling pathway and VEGF signaling pathway) are directly associated with the introduced bone plasmid cocktail.

Table . S6. | Selected miRNAs related to TGF-β and VEGF signaling pathways.
Based on the fold change, expression level, and predicted targeted gene nodes within the TGF-β (BMP-2)/VEGF pathways, 10 of the 29 miRNAs were selected and labeled in blue.