Major 5′terminally deleted enterovirus populations modulate type I IFN response in acute myocarditis patients and in human cultured cardiomyocytes

Figure 1. Enterovirus-RNA genome levels and phylogenetic viral VP1 gene sequences obtained from heart tissues or peripheral blood samples from acute myocarditis patients. (A) Viral RNA genomic replication activity was assessed using a RT-qPCR assay on heart samples (copies/ng) (n = 4) or blood and plasma samples (copies/mL) (n = 6). Data represent the median with range. Non-significant (n.s) according to Mann–Whitney U test. (B) Molecular phylogenetic analysis (nucleotide comparative analysis) of the enterovirus VP1 gene was inferred by using the maximum likelihood method with MEGA version 7.0. A total of 1,000 bootstrap iteration has been calculated to assess the toughness of the method, percentage of trees in which the associated taxa clustered together is shown next to the branches. The scale refers to the distance between sequences. Black dots represent the eleven identified strains isolated during the present study. GenBank accession nos. of the eleven identified strains are the following: MK281350-MK281351-MK281352-MK281348-MK281345-MK561365-MK561366-MK561367-MK561371- MK561372).

Figure 2. Quantitative detection of undeleted and 5′terminally deleted EV-B populations in clinical samples. (A) Electrophoregrams of RACE-PCR FL synthetic cDNA (135 bp ± 10%). (B) Quantification of EV-B full-length (FL) form after serial tenfold dilutions of synthetic FL form diluted in human cardiac (○) or plasmatic (∆) total RNA extracts of EV-negative human samples (n = 3). Linear regression curve between cDNA copies detected per µL and concentration of full-length synthetic cDNA. (C) Electrophoregrams of TD50 synthetic cDNA (55 bp ± 10%) after RACE-PCR. Image obtained from “Agilent 2,100 Bioanalyzer” after analysis with High sensitivity DNA (Agilent). (D) Quantification of EV-B 5′terminally deleted of 50 nt (TD50) form after serial tenfold dilutions of synthetic TD50 form diluted in human cardiac (○) or plasmatic (∆) total RNA extracts of EV-negative human samples (n = 3). Linear regression curve between cDNA copies detected per µL and concentration of TD50 form of synthetic cDNA. (E) Electrophoregrams of FL and TD50 mix (95% of TD50 form with 5% of FL form) synthetic cDNA after RACE-PCR. Image obtained from “Agilent 2,100 Bioanalyzer” after analysis with High sensitivity DNA…

Figure 3. Detection of full-length and 5′terminally deleted EV-B population levels in heart and peripheral blood samples. (A) Linear regression curve between cDNA copies detected per µL and concentration of FL and 5′TD forms of detected cDNA in heart and peripheral blood samples (n = 10). Linear regression was performed, and slopes were compared using Spearman test. (B) Identification of FL and 5′TD EV-B and proportion of each 5′TD forms of study samples. (C) Two-dimensional representation of the EV-B 5′non-coding region (NCR) RNA sequences inducing a cloverleaf replant with a stem “a”, and three stem-loop “b” to “d” essential for EV-B replication. In position 1, the position of the EV-B 5′ terminally deleted forms with 15 nucleotides. In position 2, the position of the EV-B 5′TD forms with 37 nucleotides. (D) Proportion of 5′TD forms (15–36 nt and 37–50 nt) and FL form in heart and peripheral blood samples. Data represent the mean ± SD. Viral populations deleted from less than 8 nucleotides were considered as full- length viral populations. FL, full-length; TD, terminally deleted, bp, base paired, nt, nucleotides.

Figure 4. Relationships between EV-B full-length or 5′terminally deleted RNA population proportions and total viral RNA levels in peripheral blood of acute myocarditis patients. (A) FL form proportion (%) correlation between EV-B viral load (cp/mL) (n = 6). (B) 5′terminally deleted 15–36 nt RNA genome proportion (%) correlation between EV-B viral load (cp/mL) (n = 6). (C) Full-length associated with 5′terminally deleted 15–36 nt RNA genome proportion (%) correlation between EV-B viral load (cp/mL) (n = 6). (D) 5′terminally deleted 37–50 nt RNA genome proportion (%) correlation between EV-B viral load (cp/mL) (n = 6). (E) Full-length associated with 5′terminally deleted 37–50 nt RNA genome proportion (%) correlation between EV-B viral load (cp/mL) (n = 6). Linear regression was performed, and slopes were compared using Spearman test. FL, full-length; TD, terminally deleted.

Figure 5. Relationships between EV-B full-length or identified EV-B 5′terminally deleted population proportions and type 1 IFN levels. (A) IFN-α and IFN-β semi-quantitative RT-qPCR analysis in peripheral blood samples expressed as fold change over control patients after normalization to housekeeping mRNA expression. Data represent the mean ± SD. **P < 0.01 according to Mann–Whitney U test. (B) FL form proportion (%) correlation between relative fold change of IFN-β (A.U.) (n = 6). (C) 5′terminally deleted 15–36 nt RNA genome proportion (%) correlation between relative fold change of IFN-β (A.U.) (n = 6). (D) Full-length and 5′terminally deleted 15–36 nt RNA genome proportion (%) correlation between relative fold change of IFN-β (A.U.) (n = 6). (E) 5′terminally deleted 37–50 nt RNA genome proportion (%) correlation between relative fold change of IFN-β (A.U.) (n = 6). (F) Full-length and 5′terminally deleted 37–50 nt RNA genome proportion (%) correlation between relative fold change of IFN-β (A.U.) (n = 6). Linear regression were performed, and slopes were compared using Spearman test. FL, full-length; TD, terminally deleted.

Figure 6. IFN-β mRNA synthesis and secretion levels by HeLa229 or human primary cardiac cells transfected by CVB3/28 full-length or 5′terminally deleted RNA forms. (A) Synthetic CVB3/28 RNAs 5′NC, CL, FL and TD100 were untreated or treated by Antarctic Phosphatase (AnP). Result of synthetic RNA forms treatment by AnP was visualized using 1% agarose gel electrophoresis. Arrows indicated the RNA band sizes on the gel picture. (B) and (D) IFN-β mRNA were expressed as fold change over mock transfected cells after normalization to housekeeping mRNA expression respectively in immortalized HeLa229 cells and in human primary cardiomyocytes. Data represent mean ± SD (n = 3) (Mann–Whitney U test; **P < 0.01). (C) and (E) IFN-β levels were quantified by ELISA in supernatants of cardiomyocytes cells at 8 h’ post-transfection of various EV-B RNA forms. Data represent mean ± SD (n = 3) (Mann–Whitney U test; * P < 0.05 and **P < 0.01). CVB3/28 RNA, full-length form (FL) and deleted CVB3/28 forms of 100 nucleotides (TD100); Positive controls: CL (Cloverleaf, Domain I), 5′NC: 5′non-coding region; LMW (poly (I:C) low molecular weight), HMW (poly (I:C) high molecular weight).