Identification of candidate microRNAs (miRs) during cardiac development. (A) Nkx2.5 cardiac enhancer eGFP (NkxCE-GFP) vector construct: (B) The cardiac specific NkxCE (~9.5kb upstream of Nkx2.5-ATG site) exclusively marks cardiac progenitor cells (CPCs) in embryonic hearts by GFP expression, as shown in an E9.5 mouse embryo. (C) Generation of cells for microRNA (miR) Array: GFP-positive cardiac progenitor cells (CPCs) and correspondent GFP negative cell fractions were sorted by flow cytometry (FACS) from E9.5 mouse embryos and in vitro differentiations on day 7 (D7) (upper images, scale bars: 500 µm). RNA/miR was isolated from murine tail tip fibroblasts (TTFs) and cardiac fibroblasts (CFs) (lower images, scale bars: 100 µm) for comparative analysis of miR expression profiles. (D) Venn diagram of upregulated miR candidates from miR Array analysis. 16 miRs were up-regulated (>1.5-fold, p < 0.05) in GFP-positive CPCs compared to their GFP-negative counterparts and 31 miRs were up-regulated in comparison to fibroblast populations (TTFs, CFs). Six of these miRs (e.g., miR-1, -133a, -30b, miR-218) were enriched in GFP-positive CPCs versus both GFP-negative cells and fibroblasts. (E) Experimental setup of in vitro differentiation (hanging drop method) of NkxCE-GFP ESCs for verification of miR candidates (scale bars: 500 µm). At differentiation day 7 (D7), GFP-positive CPCs and GFP-negative cell fractions were sorted by FACS, and total RNA was purified. (F) CPC marker miRs miR-1 (n = 6, Mann–Whitney test, p = 0.0022) and miR-133a (n = 3, t-test, p = 0.0465) were significantly upregulated at D7 in NkxCE-GFP CPCs in comparison to stage-matched negative cells, whereas only miR-128a (n = 6, Mann–Whitney-test, p = 0.0022) was significantly enriched of candidate miRs in NkxCE GFP-positive CPCs. No significance was found for miR-20b, -30a, and -30b (n = 3 each, t-test). Three assays were performed, whereas only for miR-1 and miR-128, samples were measured in duplicates. (G) To evaluate miR kinetics during in vitro differentiation, non-transgenic murine ESCs (V6.5, scale bar: 500 µm) were differentiated (hanging drop method) until day 10. RNA was isolated every other day for qRT-PCR. (H) Expression of miR-1 and -133a followed a typical course by rising at the beginning of cardiomyogenesis around D4 to D6. MiR-30a, -30b, and -128a also rose around D4 to D8. However, expression of miR-20b appears to be biphasic with a peak on D4 and D8. Three assays were performed in triplicate (n = 9 per timepoint; each sample was measured in duplicates n = 18 per timepoint; ANOVA followed by Dunn’s Method). All data are represented as means ±SEM. * p ≤ 0.05, ** p ≤ 0.01.

Evaluation of candidate miR function during zebrafish development. (A) MiR kinetics were evaluated during early zebrafish development. Total RNA was isolated on respective timepoints (hours post fertilization; hpf) for qRT-PCR. (B) MiR-20b, -30a, -30b, and -128a were upregulated in developing zebrafish larvae from 24 hpf increasing until 72 hpf. (C) Morpholino oligos (MOs) were injected in 1–2 cell-stage zebrafish embryos and miR knockdown, morphological changes, fractional shortening, and beating rate was analyzed (24, 48, 72 hpf). (D) MiR expression (evaluated by qRT-PCR) in zebrafish larvae was sufficiently reduced by MOs at 24 hpf by more than 85% for all candidate miRs in comparison to controls without MOs (t-test). (E) Morphological changes of MO-morphants at 48 hpf. MO-20b morphants showed cerebral hemorrhage (red arrow) and edema (edema in the eye region indicated by a white arrow). MO-30b morphants developed malformations and edemas (red arrow highlights a visible edema and enlarged hydrocephalus). MO-128a larvae exhibited a robust pericardial edema (red arrow) and blood congestion at the right outflow tract of the heart (white arrow). Scale bars: 250 µm. (F) Cardiac phenotype of MO-128a treated Tg(myl7:ras-GFP) zebrafish larvae at 48 hpf including smaller ventricles (white arrow) and abnormalities in heart looping. Scale bars: 5 µm. (G) MO-128a larvae (n = 8) showed a significantly reduced fractional shortening at 72 hpf compared to non-treated larvae (w/o MO, n = 10) (t-test) (H) MO-128a morphants (n = 11) appeared to have a significantly slower heart rate at 72 hpf compared to the controls (w/o MO, n = 11) (t-test). All data are represented as means ± SEM. * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001.

Knockdown of miR-128a during in vitro differentiation of murine NkxCE-GFP embryonic stem cells (ESCs). (A) NkxCE-GFP ESCs (scale bar: 500 µm) were differentiated using the hanging drop method. Transfections of locked nucleic acid (LNA) probes for knockdown of miR-128a (LNA-128) and corresponding non-targeting controls (LNA-Ctr) were performed on day 3.5 (D3.5) and after one week (wk) of differentiation. Analysis was performed on respective timepoints between 0.75 and 2.25 wks. Abbreviations: EB: Embryoid bodies; FACS: Flow cytometry; (B) MiR-128a was significantly downregulated (p < 0.001, Mann–Whitney test) by LNA-128 from 0.75 wks until 2 wks when compared to respective LNA-Ctrs. At least 3 assays in triplicates were performed per time-point. (C) NkxCE-GFP-positive CPCs after 1 wk (upper panel, scale bars: 500 µm) and 1.5 wks (lower panel, scale bars: 500 µm) after transfection with either LNA-Ctr or LNA-128. The images are an overlay between phase contrast and fluorescent pictures. (D) The frequency of NkxCE-GFP-positive CPCs with miR-128 knockdown (n = 9) was slightly reduced after 1 wk, becoming significantly downregulated after 1.5 wks (n = 5, 9; p = 0.001, Mann–Whitney test). The CPC frequency after 2 wks was not affected by LNA-128 compared to LNA-Ctr (n = 6). 3 assays in triplicates for 1 wks and 1.5 wks, 2 assays in triplicates for 2 wks. (E–I). Gene expression panels during NkxCE-GFP ESC differentiation after miR-128 knockdown. The panels show the expression of early CPC markers after 0.75 wks (n = 11, 4 assays in triplicates, t-test or Mann–Whitney test) (E) and 1 wk (n = 9, 3 assays in triplicates, t-test or Mann–Whitney test) (F) as well as neuroectodermal (t-test) (G), proliferation (H), smooth muscle (Mann–Whitney test), endothelial, and angiogenesis markers (t-test) (I) after 0.75 wks. (J) Beating frequency (beats per minute, bpm) of early cardiomyocytes was downregulated from 1 wk to 2.25 wks on mir-128a knockdown compared to LNA-Ctr (n > 23, t-tests) (at least 4 assays, for 2.25 wks 2 assays, 8 videos were evaluated per condition from 3 independent observers). Data are represented as means ± SEM. * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001.

Knockdown of miR-128a during in vitro differentiation of murine Isl1-Cre/ROSA26^mTmG induced pluripotent stem cells (iPSCs) (iITG-iPSCs). (A) iITG-iPSCs (scale bar: 500 µm) were differentiated using the hanging drop method. Transfections of LNA probes for knockdown of miR-128a (LNA-128) and corresponding controls (LNA-Ctr) were performed after 0.75 weeks (wks) and after 2 wks of differentiation. Analysis was performed on respective timepoints between 1 and 3 wks. Abbreviations: EB: Embryoid bodies; FACS: Flow cytometry; (B) MiR-128a was significantly downregulated by LNA-128 from 1 wk until 3 wks compared to respective LNA-Ctr (n = 8,7, Mann–Whitney tests, 3 assays in triplicates). (C) Isl1-GFP-positive CPCs after 2 wks (upper panel, scale bars: 500 µm) and after 3 wks (lower panel, scale bars: 500 µm) of in vitro differentiation after transfection with either LNA-Ctr or LNA-128. Images are an overlay between phase contrast and fluorescent microscopic pictures. (D) The frequency of Isl1-GFP-positive CPCs after miR-128a knockdown was significantly increased to the double amount compared to LNA-Ctr (n = 9, p = 0.024, t-test; 3 assays in triplicates) after 2 wks. At later timepoints, the Isl1-CPC frequency was not significantly affected by miR-128 knockdown (3 assays in triplicates at 2.25 wks, 2 assays in triplicates at 3 wks). (E) Isl1 expression was tendentially upregulated during iITG-iPSC differentiation after miR-128 knockdown (n = 6, 2 assays in triplicates). (F) Beating frequency (beats per minute, bpm) of differentiated iITG-iPSC EBs was reduced after miR-128 knockdown (2 and 2.25 wks: 3 assays in triplicates, 2.75 (t-test) and 3 wks (Mann–Whitney test): Two assays in triplicates, 4–9 videos per condition, 3 independent observers). Data are represented as means ± SEM. * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001.

Overexpression (OE) of miR-128a during in vitro differentiation of murine NkxCE-GFP OE-128 and OE-Ctr ESCs. (A) OE-128 and OE-Ctr ESCs (scale bars: 500 µm) were differentiated for 2 weeks (wks) by the hanging drop method (without doxycycline). Analysis was performed on respective timepoints between 0.75 and 2 wks of differentiation. Abbreviations: EB: Embryoid bodies; FACS: Flow cytometry; (B) MiR-128a was overexpressed in OE-128 EBs compared to OE-Ctr EBs from early onset of cardiogenesis (0.75 wks), becoming significant after 1 wk (p < 0.01, Mann–Whitney test); 0.75 and 1 wk: 3 assays in triplicates, different clones; 1.5 and 2 wks: 2 assays in triplicates, different clones. (C) NkxCE-GFP-positive CPCs in OE-Ctr EBs or OE-128 EBs after 1.5 wks (upper panel, scale bars: 500 µm) and after 2 wks (lower panel, scale bars: 500 µm). Images are an overlay between phase contrast and fluorescent microscopic pictures. (D) NkxCE GFP-positive CPC frequency after miR-128 OE appeared to be reduced after 1 wk, whereas the frequency of NkxCE-GFP-positive CPCs slightly increased after 1.5 wks upon miR-128 OE becoming significant after 2 wks (p = 0.0088, t-test). 1 wk: 3 assays in triplicates (different clones), 1.5 and 2 wks: 2 assays in triplicates (different clones). (E,F) Gene expression panels during NkxCE-GFP ESC OE-128 and OE-Ctr differentiation. The panels show the expression of early CPC markers after 1 wk (3 assays in triplicates, different clones) (t-tests) (E) and 1.5 wks (2 assays in triplicates, different clones) (t-tests). (F,G) The beating frequency (beats per minute, bpm) of early cardiomyocytes was significantly upregulated from 1 wk to 2 wks on mir-128a OE compared to OE-Ctr EBs (2 assays in triplicates, different clones, 5–9 videos per timepoint) (t-tests or Mann–Whitney test). Data are represented as means ± SEM. * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001.

Role of miR-128a in cardiac development. During murine in vitro differentiation, LNA-mediated miR-128a knockdown in differentiating ES/iPSCs (1) increased cardiac transcription factors such as Isl1, Sfrp5, and Hcn4 but reduced Irx4 at the onset of cardiogenesis; (2) upregulated Isl1-positive CPCs, whereas NkxCE-GFP-positive CPCs were downregulated; and (3) increased the expression of the ventricular cardiomyocyte marker Myl2 accompanied by a reduced beating frequency of early cardiomyocytes. Overexpression of miR-128a (4) diminished the expression of Isl1, Sfrp5, Nkx2.5, and Mef2c, but increased Irx4, (5) enhanced the NkxCE-GFP-positive CPC population, and (6) favored nodal-type-like cardiomyocytes marked by Tnnt2, Myh6, and Shox2 accompanied by increased beating rates. Parts of the figure from BioRender.com