Cardiomyocyte-specific modRNA-induced FoxM1 overexpression promotes recovery from myocardial infarction in adult mammals

Heart failure occurs due to the loss of the contractile unit of the left ventricle: cardiomyocytes. Mammalian cardiomyocytes exit the cell cycle shortly after birth. Therefore, postinfarction left ventricular remodeling often progresses to heart failure with a dilated left ventricle. Based on the single-nucleus RNA sequencing data collected from newborn neonatal mouse¹ and pig² models of cardiac injury, we identified FoxM1 uniquely upregulated in cardiomyocytes from regenerating hearts in both species. FoxM1 was known to regulate cardiomyocyte proliferation during mouse embryonic development³ and adult zebrafish⁴, but has never been investigated in cardiac repair of mammalian hearts. Therefore, we investigated whether FoxM1 promotes cardiomyocyte proliferation after cardiac injury in mammals by using a highly efficient Cardiomyocyte-specific Modified mRNA Translation Syste ⁵ to transiently overexpress FoxM1 (^FoxM1CM-SMRTs) in the cardiomyocytes of mammalian hearts. Control experiments were conducted with CM-SMRTs that coded for the expression of luciferase (^LucCM-SMRTs) buffer.

First, we demonstrated the pro-proliferative role of FoxM1 in 70-day-old, non-proliferating cardiomyocytes differentiated from human-induced pluripotent stem cells. Compared with ^LucCM-SMRTs-treated cells, those treated with ^FoxM1CM-SMRTs exhibited robust expression of FoxM1, accompanied by significant increases in the proliferation marker Ki67 and the G2/M-phase marker phosphorylated histone 3 (PH3). Notably, ^FoxM1CM-SMRTs also resulted in higher percentage of iPSC-CMs expressing symmetric and asymmetric Aurora B (sAuB and aAuB, respectively), which indicate cell division (sAuB) and multinucleation (aAuB), along with a significantly greater cell counts on Day 14 post-treatment (A).

All animal studies were approved by IACUC UAB and were consistent with the Guidelines by the NIH. We will make our data, analytic methods, and study materials available upon the requests. 8-week-old male C57BL/6J mice were used to minimize sex-related variables. MI was surgically induced via permanent ligation of the left-anterior descending (LAD) coronary artery as described previously⁵, and animals were injected with CM-SMRTs (100 μg/mouse) or vehicle at three sites (10 μL/site) in the border zone of the infarct; a third group underwent sham surgery and recovered without receiving either treatment. When injected with ^GFPCM-SMRTs, GFP expression was observed in cardiomyocytes within border zone but not remote zone (B,i). FoxM1-, Ki67-, and PH3-expressing cardiomyocytes were significantly more common in the ^FoxM1CM-SMRTs treatment group than Vehicle group (B,ii). Furthermore, echocardiographic measurements of left-ventricular ejection fraction (LVEF) and fractional shortening (LVFS) in the ^FoxM1CM-SMRTs group increased from Day 5 to Day 14 after treatment and were significantly greater than in Vehicle-treated mice on Days 14, 21, and 28 (B,iii). Infarct size (B,iv) and cardiomyocyte cross-sectional areas (B,v) were also significantly smaller from the ^FoxM1CM-SMRTs group than in Vehicle-treated mice on Day 28.

We then examined whether the beneficial effects of ^FoxM1CM-SMRTs can be observed in large mammals: young (60-day) Yorkshire pigs (~30 kg) and adult (12 month-old) Sinclair Minipigs (25 kg). AMI was induced via temporary ligation of the LAD coronary artery as described previously⁵, and animals were injected with ^FoxM1CM-SMRTs (6 mg/animal) or vehicle at five sites (200 μL/site) in the border zone of the infarct. FoxM1, sAuB, and aAuB expression on Day 3 (corresponding to the period of confirmed modRNA activity⁵), was significantly more common in cardiomyocytes from ^FoxM1CM-SMRT-treated hearts than in cardiomyocytes from Vehicle-treated hearts in Yorkshire swine (C,i). More importantly, co-immunofluorescence staining for FoxM1 together with Ki67 or PH3 revealed that proliferative marker expression was observed exclusively in FoxM1-positive cardiomyocytes (C,i). The echocardiographic assessments of LVEF and LVFS decreased to the similar level at Day 3 after AMI in subgroups of ^FoxM1CM-SMRTs and Vehicle-treated group. However, at Days 10 and 28 LV function were greater in ^FoxM1CM-SMRTs–treated than in Vehicle-treated pigs (C,ii). Cardiac magnetic resonance images (cMRI) collected on Day 28 also indicated that ^FoxM1CM-SMRTs treatment was associated with significantly improved LVEFs (C,iii), and smaller infarct sizes (C,iv). In adult minipigs, measurements of LVEF and LVFS on Day 7 and 28, as well as stroke volume on Day 7, were significantly greater when the animals were treated with ^FoxM1CM-SMRTs than vehicle group (D,i). The ^FoxM1CM-SMRTs administration was associated with significant decreases in fibrosis (D,ii) and cardiomyocyte cross-sectional area (D,iii) on Day 28. The proportion of cardiomyocytes that expressed Ki67 or PH3 did not differ significantly between ^FoxM1CM-SMRTs– and Vehicle-treated minipigs on Day 28 (D,iv), indicating that ^FoxM1CM-SMRTs did not lead to persistent increases in cardiomyocyte proliferation.

In conclusion, our results demonstrated that CM-SMRTs–mediated FoxM1 overexpression activates cardiomyocyte proliferation and promotes myocardial regeneration in the infarcted hearts of both adult mice and pigs. As the first direct report in mammalian hearts, this study highlights the beneficial role of FoxM1 in cardiomyocyte proliferation and heart repair, despite the limitation of a relatively small animal size and missed opportunity to explore FoxM1’s functions other than its pro-proliferative effect. Collectively, these findings and the well-established safety of modRNA-based therapeutics support the continuing development of this strategy for further mechanistic understanding of myocardial regeneration in AMI hearts.

Figure. ^FoxM11CM-SMRTs administration promoted cardiomyocyte proliferation and improved recovery from MI in mice and pigs.

(A) hiPSC-CMs were maintained for 70 days after differentiation to minimize their intrinsic cell-cycle activity and proliferation; then, the cells were transfected with ^FoxM1CM-SMRTs or ^LucCM-SMRTs for two days, and the proportion of cells that co-expressed cardiac troponin T (cTnT) and FoxM1, Ki67, PH3, sAuB, or aAuB was evaluated via immunofluorescence staining (IF); nuclei were counterstained with DAPI. Representative images are displayed. The proportions of cardiomyocytes that expressed each marker were reported as a percentage, and hiPSC-CM cell counts were quantified before (Day 0) and 14 days after (Day 14) treatment. n=5 biological replicates for immunofluorescence analyses, and n=9 for cell counts. P values were determined with the student’s t-test. (B) Male C57BL/6J mice (stock #000664; The Jackson Laboratory) were used for mouse MI study. Echocardiography was performed one day before MI induction (Day –1) and 3, 5, 14, 21, and 28 days afterward (Day 3, Day 5, Day 14, Day 21, and Day 28, respectively); histological and IF analyses were performed on Day 3 and Day 28 subgroup of hearts. (B,i) Sections from ^GFPCM-SMRTs–treated hearts on Day 3 were IF stained for the expression of GFP (green) and cTnT (red); nuclei were counterstained with DAPI (blue). The multichannel merged GFP-positive cardiomyocytes (CMs) appear yellow. The proportion of cardiomyocytes that expressed GFP was quantified as a percentage of total cardiomyocytes within border zone or remote zone. (B,ii) Sections from the border zones of ^FoxM1CM-SMRTs– and Vehicle-treated hearts on Day 3 were IF stained for the expression of cTnT (green) and FoxM1 (red), Ki67 (red), or PH3; nuclei were counterstained with DAPI (blue). Representative images are displayed. The proportions of cardiomyocytes that expressed each marker were reported as a percentage. (B,iii) LVEF and LVFS were determined via echocardiographic imaging; representative M mode images are displayed for Day 3 and Day 28. (B,iv) Infarct size (Infarct size % = 100 % x infarct area / total LV area) was quantified by cutting the LV into eighty serial sections from the base to the apex, staining the sections with Masson’s trichrome, and then evaluating 1/10 of the total ~80 sections to calculate the percentage of the area of the LV scar to the total area of the LV. (B,v) Sections from the border zone were stained with wheat germ agglutinin (WGA, white) to visualize cell borders, cardiomyocytes were visualized by staining for cTnT (red), and nuclei were counterstained with DAPI (blue); then, cardiomyocyte cross-sectional areas were measured. n=5 mice per group, and P values were determined with the student’s t-test for Bii and one-way ANOVA with the Tukey multiple comparisons test for Biii-v. aP<0.001 vs. Sham; bP<0.001 vs. Vehicle; cP<0.05 vs. Sham; dP<0.05 vs. Vehicle. (C) For experiments with Yorkshire pigs (Valley Brook Research, Inc., Madison, GA), echocardiography was performed before MI induction (Day 0, Baseline) and on Days 10 and 28 afterward, cMRI was performed on Day 28, and histological and immunofluorescence analyses were performed on Day 3 and Day 28. (C,i) Sections from the border zones of ^FoxM1CM-SMRTs– and Vehicle-treated hearts on Day 3 were IF stained for the expression of cTnT and FoxM1, sAuB, or aAuB; nuclei were counterstained with DAPI. The proportions of cardiomyocytes expressing each marker were reported either as a percentage or as density per square millimeter. Co-localization of FoxM1 (white) and Ki67 (red) or PH3 (red) was visualized by IF co-staining. The ratios of Ki67- or PH3-positive cells were quantified separately for FoxM1-positive and FoxM1-negative cardiomyocytes. (C,ii) LVEF and LVFS were determined via echocardiographic imaging; representative images are displayed for Day 0 and Day 28. (C,iii) Cine cMRI was performed on Day 28, and the images were used to calculate LVEF. (C,iv) Late-gadolinium–enhanced cMRI was performed on Day 28, and the images were used to calculate infarct size. n=7 for MI+Vehicle and n=5 for MI+^FoxM1CM-SMRTs in C ii; n=7 for Vehicle and n=2 for ^FoxM1CM-SMRTs in C iii-iv. P values were determined with the student’s t-test. (D) In minipigs, echocardiography was performed shortly before MI induction (Day 0) and on Days 7 and 28 afterward; histological and IF analyses were performed on Day 28. (D,i) LVEF, LVFS, left ventricle end of diastolic volume, end of systolic volume, and stroke volume were determined from echocardiographic images; representative images are displayed for Day 0 and Day 28. (D,ii) Heart sections obtained 28 days after MI were stained with Masson’s trichrome to identify normal (red) and fibrotic (blue) tissues (bar=10 mm); then, infarct size was quantified as the percentage ratio of the area of the LV scar to the total area of the LV. Images shown are from similar cutting regions. (D,iii) Sections from the border zone were stained with WGA (white) to visualize cell borders, cardiomyocytes were visualized by staining for cTnT (red), and nuclei were counterstained with DAPI (blue); then, cardiomyocyte cross-sectional areas were measured. (D,iv) Sections from the border zones of ^FoxM1CM-SMRTs– and Vehicle-treated hearts on Day 28 were IF stained for the expression of cTnT (green) and Ki67 (red) or PH3 (red); nuclei were counterstained with DAPI (blue). Representative images are displayed for both makers; Ki67-positive cardiomyocytes appear pink. The proportions of cardiomyocytes that expressed each marker were reported as a percentage. n=5 for each group, and P values were determined with the two-tailed Student’s t-test. All quantified data in A-D are presented as mean±SEM, and all scale bars are 50 μm in length unless stated otherwise.

Non-standard Abbreviations and Acronyms

CM-SMRTs

Cardiomyocyte Specific Modified mRNA Translation system

PH3

phosphorylated histone H3

sAuB

symmetric Aurora B

aAuB

asymmetric Aurora B

LAD

left anterior descending coronary artery

LVEF

left ventricular ejection fraction

LVFS

left ventricular fractional shortening

cMRI

Cardiac magnetic resonance imaging