Modifying miRs for effective reprogramming of fibroblasts to cardiomyocytes

Abstract

Reprogramming scar fibroblasts into cardiomyocytes has been proposed to reverse the damage associated with myocardial infarction. However, the limited improvement in cardiac function calls for enhanced strategies. We reported enhanced efficacy of our miR reprogramming cocktail miR combo (miR-1, miR-133a, miR-208a, and miR-499) via RNA-sensing receptor stimulation. We hypothesized that we could combine RNA-sensing receptor activation with fibroblast reprogramming by chemically modifying miR combo. To test the hypothesis, miR combo was modified to enhance interaction with the RNA-sensing receptor Rig1 via the addition of a 5'-triphosphate (5'ppp) group. Importantly, when compared with unmodified miR combo, 5'ppp-modified miR combo markedly improved reprogramming efficacy in vitro. Enhanced reprogramming efficacy correlated with a type-I interferon immune response with strong and selective secretion of interferon β (IFNβ). Antibody blocking studies and media replacement experiments indicated that 5'ppp-miR combo utilized IFNβ to enhance fibroblast reprogramming efficacy. In conclusion, miRs can acquire powerful additional roles through chemical modification that potentially increases their clinical applications.

Keywords: 5′-triphosphorylation; IFNβ; MT: Non-coding RNAs; RNA modification; cardiomyocytes; fibroblasts; innate immune signaling; miRs; reprogramming.

Graphical abstract

Figure 1

Optimizing 5′ppp modification of miR combo (A) Schematic describing the methodology to generate 5′ppp-modified miRs. (B) A table listing the miR combo variants used in (C)–(E). (C–E) Cardiac fibroblasts were transfected with unmodified miR combo, miR combo variants 1–11 or the appropriate non-targeting control miRs. After 7 days, expression of the indicated cardiomyocyte-specific genes was determined by qPCR. Expression levels were normalized to the housekeeping gene Gapdh and are represented as a fold change to the unmodified non-targeting control. N = 10. ANOVA (1-way) with post hoc tests were used to determine significances: ∗∗p < 0.01, ∗p < 0.05 comparisons with unmodified miR combo.

Figure 2

5′ppp-miR combo enhances fibroblast to cardiomyocyte reprogramming efficacy (A) Cardiac fibroblasts were transfected with unmodified or 5′ppp-miR combo and appropriate non-targeting control miRs. Fourteen days after transfection days, expression of the indicated cardiomyocyte-specific genes was determined by qPCR. Expression levels were normalized to the housekeeping gene Gapdh and are represented as a fold change to the unmodified non-targeting control. N = 3–12. ANOVA (1-way) with post hoc tests were used to determine significance: ∗∗∗p < 0.001 comparisons with unmodified miR combo. (B–D) Cardiac fibroblasts were transfected with unmodified or 5′ppp-miR combo and appropriate non-targeting control miRs. Fourteen days after transfection, the number of cardiomyocyte-like cells was determined by staining for the cardiomyocyte-specific protein Actn2. The number of Actn2+ cells (cardiomyocyte-like cells) per field was calculated from five independent fields (×20 magnification) from four to six independent experiments. ANOVA (1-way) with post hoc tests was used to determine significances: ∗p < 0.05 comparisons with unmodified miR combo. Representative images of unmodified and 5′ppp-modified miR combo are shown in (C) (scale bar, 20 μm). An enlarged image showing sarcomere structure in cardiomyocyte-like cells generated via 5′ppp-miR combo is shown in (D) (scale bar, 20 μm). (E) Cardiac fibroblasts were transfected with unmodified or 5′ppp-miR combo or appropriate non-targeting control miRs. Fourteen days after transfection, the number of beating cells per field (×20 magnification) was counted. N = 4. ANOVA (1-way) with post hoc tests were used to determine significances: ∗p < 0.05 comparisons with unmodified miR combo. (F) Cardiac fibroblasts were transfected with 5′ppp-miR combo. Fourteen days after transfection, spontaneous calcium oscillations were measured and compared with those derived from freshly isolated neonatal cardiomyocytes. Representative traces are shown from 10 cells per group.

Figure 3

5′ppp-miR combo promotes expression of IFNβ pathway components Cardiac fibroblasts were transfected with unmodified or 5′ppp-miR combo or appropriate non-targeting control miRs. After 4 days, RNA-seq was performed. N = 3. (A) Heatmap showing genes whose expression was significantly affected by 5′ppp-miR combo (>3-fold compared with the unmodified non-targeting control). (B) Volcano plots showing comparisons between 5′ppp-negmiR and 5′ppp-miR combo and between unmodified miR combo and 5′ppp-miR combo. Red indicates a significant increase/decrease in expression (≥3-fold). (C–F) Two gene lists were produced. One gene list comprised those genes whose expression was increased by unmodified miR combo by ≥ 3-fold when compared with the unmodified control negmiR. The second list was composed of genes whose expression was increased by 5′ppp-miR combo by ≥3-fold when compared with the 5′ppp-modified control negmiR. The Venn diagram reports the number of genes that were common and unique in both gene lists (C). Gene ontology (GO) was then performed on those common and unique genes. (D) The GO results on the common set of genes. (E) The GO results on the genes that were unique to 5′ppp-miR combo. (F) The GO results on the genes that were unique to unmodified miR combo. For (D)–(F), only the top 10 GO terms are shown. (G) Cardiac fibroblasts were transfected with unmodified or 5′ppp-modified miR combo or appropriate non-targeting control miRs. After 4 days, expression of the indicated genes was determined by qPCR. Expression levels were normalized to the housekeeping gene Gapdh and are represented as a fold change to the unmodified non-targeting control. ANOVA (1-way) with post hoc tests were used to determine significances: ∗∗∗p < 0.001 comparisons between 5′ppp-miR combo and unmodified miR combo.

Figure 4

5′ppp-miR combo improves fibroblast to cardiomyocyte reprogramming efficacy via IFNβ (A) Cardiac fibroblasts were transfected with 5′ppp-miR combo, unmodified miR combo, or their respective non-targeting controls. Five (top panels) or 24 (bottom panels) hours after transfection an isotype control or IFNβ blocking antibody was added to the media. Cardiomyocyte-specific gene expression was then measured 14 days after transfection by qPCR. Expression levels were normalized to the housekeeping gene Gapdh and are shown as a fold change to the unmodified non-targeting control miR. N = 5–6 per group. ANOVA (1-way) with post hoc tests were used to determine significances: ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant, comparisons to unmodified miR combo + isotype control group. (B) Cardiac fibroblasts were transfected with 5′ppp-miR combo. Media was collected from 5′ppp-miR combo transfected cells 24 (top panels) or 72 (bottom panels) hours after transfection and placed onto cells 24 hours after transfection with unmodified miR combo. Non-targeting miRs were used as a control. Cardiomyocyte-specific gene expression measured by qPCR 14 days after transfection. Expression levels were normalized to the housekeeping gene Gapdh and are represented as a fold change to non-targeting control miR. N = 4. ANOVA (1-way) with post hoc tests were used to determine significances: ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns, not significant, comparisons to the unmodified miR combo group. (C) Six hours prior to transfection, cardiac fibroblasts were incubated with either an isotype control or IFNβ blocking antibody. After the 6 h incubation with antibodies, the cells were transfected with 5′ppp-miR combo, unmodified miR combo, or the unmodified non-targeting control. Cardiomyocyte-specific gene expression was then measured 14 days after transfection by qPCR. Expression levels were normalized to the housekeeping gene Gapdh and are shown as a fold change to the unmodified non-targeting control miR. N = 6 per group. ANOVA (1-way) with post hoc tests were used to determine significances: ∗∗∗p < 0.001, ns, not significant, comparisons to unmodified miR combo + isotype control group.

Figure 5

5′ppp-miR combo is selective for the IFNβ pathway (A) Cardiac fibroblasts were transfected with unmodified or 5′ppp-miR combo or the appropriate non-targeting control miRs. One and 4 days after transfection, the media was collected. IFNβ and IFNγ amounts were measured by ELISA. N = 7. ANOVA (1-way) with post hoc tests were used to determine significances: ∗∗p < 0.01, ∗p < 0.05 comparisons to the unmodified non-targeting miR group. (B) Cardiac fibroblasts were transfected with unmodified or 5′ppp-miR combo or the appropriate non-targeting control miRs. One day after transfection, expression of the indicated RNA-sensing receptor pathway components was determined by qPCR. Expression levels were normalized to the housekeeping gene Gapdh and are represented as a fold change to the unmodified non-targeting control. N = 3–7. ANOVA (1-way) with post hoc tests were used to determine significances: ∗p < 0.05, ns, not significant, comparisons to the unmodified non-targeting miR group. (C) Cardiac fibroblasts were transfected with unmodified or 5′ppp-modified miR combo and appropriate non-targeting control miRs. After 6 h, NF-κB, IRF3, and IRF-7 activity was determined by dual-luciferase assay. N = 5–6. ANOVA (1-way) with post hoc tests was used to determine significances: ∗∗∗p < 0.001; ∗∗p < 0.01, ∗p < 0.05, ns, not significant, comparisons with unmodified non-targeting miR group.

Figure 6

Rig1 mediates the effects of 5′ppp-miR combo on the IFNβ pathway and reprogramming (A) Cardiac fibroblasts were transfected with 5′ppp-miR combo and siRNAs (Rig1, TLR3 targeting, or non-targeting control siRNA). One day after transfection, expression of Rig1, Tlr3 and other RNA-sensing receptor pathway components was determined by qPCR. Expression levels were normalized to the housekeeping gene Gapdh and are represented as a fold change to the non-targeting control siRNA group. N = 4. ANOVA (1-way) with post hoc tests were used to determine significances: ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05; ns, not significant, comparisons to the 5′ppp-miR combo + non-targeting siRNA group. (B) Cardiac fibroblasts were transfected with 5′ppp-miR combo and siRNAs (Rig1, TLR3 targeting, or non-targeting control siRNA). Transfection with unmodified non-targeting miRs served as an additional control. After 14 days, expression of the indicated cardiomyocyte-specific genes was determined by qPCR. Expression levels were normalized to the housekeeping gene Gapdh and are represented as a fold change to the unmodified non-targeting control miR. N = 4. ANOVA (1-way) with post hoc tests was used to determine significances: ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05; ns, not significant, comparisons to the 5′ppp-miR combo + non-targeting siRNA group.