Abstract

   Despite its promise, cardiac regenerative therapy remains clinically
   elusive due to the difficulty of spatio‐temporal control of
   proliferative induction, and the need to coordinately reprogram
   multiple regulatory pathways to overcome the strict post‐mitotic state
   of human adult cardiomyocytes. To address this unmet therapeutic need,
   a combinatorial miRNA interference screen is performed specifically
   targeting cardiac‐predominant miRNAs regulating key aspects of
   cardiomyocyte mitotic induction to cell‐cycle completion in neonatal
   rat cardiomyocytes. In doing so combinatorial interference of miRNA‐1a
   and miRNA‐15b (LNA‐1a/15b) is identified as drivers of adult
   cardiomyocyte proliferation. Due to miRNA‐1a/15b function on multiple
   processes modulating adult cardiomyocyte mitosis, its inhibition
   augmented adult cardiomyocyte cell‐cycle completion and daughter cell
   formation, and improved contractility in 3D human cardiac organoids,
   and in a mouse model of ST‐segment elevation myocardial infarction. Due
   to the cardiac‐restricted pattern of miRNA‐1a/15b expression, this
   strategy provides a feasible means for specific cardiomyocyte
   proliferative induction with minimal risk of neoplasm formation and
   off‐target toxicity. The approach further highlights an underutilized
   therapeutic strategy for simultaneous co‐regulation of multiple disease
   pathways through combinatorial interference of miRNAs.

   Keywords: cardiac organoids, cardiomyocyte proliferation, microRNAs,
   myocardial infarction
     __________________________________________________________________

   The article explores a novel therapeutic strategy for cardiac
   regeneration by targeting miRNA‐1a and miRNA‐15b. Combinatorial
   inhibition of miR‐1a and miR‐15b enhances cardiomyocyte proliferation,
   improves heart function, and reduces fibrosis in myocardial infarction
   models. The study demonstrates minimal off‐target effects and suggests
   a clinically feasible approach using LNA‐based inhibitors to stimulate
   cardiac regeneration for the treatment of heart failure.

   graphic file with name ADVS-12-2414455-g005.jpg

1. Introduction

   Despite improvements in interventional and pharmacological therapies,
   heart failure remains a major cause of death globally. It affects 64
   million people worldwide and leads to cardiovascular morbidity and
   mortality.^[ [56]^1 ^] Patients with heart failure suffer from major
   impairments in quality of life and poor long‐term prognosis.^[ [57]^2
   ^] Loss of cardiomyocytes is a major hallmark of heart failure, in both
   congenital, age‐ and pathologic stress‐related heart failure, and is
   linked to an ≈25% decrease in cardiomyocytes due to cell death.^[
   [58]^3 ^] However, unlike other tissues that can compensate cell death
   through proliferative induction, cardiomyocytes are post‐mitotic and
   have negligible capacity for proliferation, regeneration or repair to
   restore damaged tissue and overcome remodeling and fibrotic processes
   leading to heart failure and death.^[ [59]^4 ^] Thus, there remains a
   high unmet need for novel therapeutics to enhance cardiomyocyte
   regeneration and thereby reduce morbidity and mortality from heart
   failure.

   In mammals, cardiomyocytes exit the cell cycle and cease proliferation
   in the early post‐natal period.^[ [60]^5 ^] Whilst evidence suggests
   some capacity for adult cardiomyocytes to re‐enter the cell cycle,
   particularly in response to myocardial insult or injury, proliferation
   rates are minimal with negligible physiologic or clinical benefits.^[
   [61]^5 , [62]^6 , [63]^7 ^] Gene therapy approaches have in recent
   years been utilized to overcome this proliferative block with varying
   degrees of success.^[ [64]^8 , [65]^9 ^] Problematically, these
   attempts have entailed catheter‐ or intracardial injection for
   orthotropic delivery of therapeutics into the myocardium, which
   potentially limits the accessibility of such therapeutics for patients
   due to the complex and invasive nature of delivery.^[ [66]^8 ^] These
   invasive measures are necessary to direct therapeutic delivery to the
   heart and to restrict leakage of proliferation drivers to non‐cardiac
   tissues in order to prevent ectopic neoplasm formation. Adding to these
   concerns, studies indicate that proliferative induction of adult
   cardiomyocytes requires coordinated regulation of multiple genes to
   overcome the many innate barriers to adult cardiomyocyte cell‐cycle
   re‐entry and mitotic completion, including sarcomeric structural
   rigidity, stability, and density, its anti‐proliferative metabolic
   state and the fact that adult cardiomyocytes are prone to undergo
   endomitosis, rather than mitosis.^[ [67]^10 ^] Thus, the need to
   simultaneously modulate multiple pathways to induce physiologically
   relevant proliferative rates has raised additional technical challenges
   and clinical concerns.

   microRNAs (miRNAs) are small non‐coding RNAs that promote mRNA
   degradation and translational inhibition and have been shown to play
   key roles in heart disease development and progression to heart
   failure. This is evidenced by large animal preclinical, and clinical
   trial data from recent and ongoing trials modulating disease‐linked
   miRNAs in heart disease.^[ [68]^11 , [69]^12 , [70]^13 ^] miRNAs
   function by regulating key genes involved in proliferation,
   hypertrophy, sarcomerogenesis, inflammation, metabolic reprogramming,
   and contractility.^[ [71]^14 , [72]^15 , [73]^16 ^] Whilst miRNAs can
   regulate the expression of multiple genes by binding to target
   transcripts to remodel the cellular transcriptome,^[ [74]^17 ^] it is
   less clear if modulation of a single miRNA alone is sufficient to
   re‐program all pathways necessary to effectively induce cell‐cycle
   re‐entry and completion of adult cardiomyocytes. Given that we were not
   able to identify single miRNAs capable of robustly modulating all
   relevant pro‐proliferative pathways in cardiomyocytes, we sought to
   determine if a combinatorial RNA interference (co‐RNAi) strategy would
   provide the necessary reprogramming impetus for adult cardiomyocyte
   cell‐cycle re‐entry and mitotic completion. Although largely unexplored
   in heart disease therapeutics, co‐RNAi strategies are widely and
   successfully applied in anti‐viral and cancer therapy, wherein
   combinatorial targeting of multiple key genes serves to better overcome
   phenotypic and genomic evolution to increase therapeutic robustness and
   efficacy.^[ [75]^18 , [76]^19 , [77]^20 ^] Considering precedent for
   successful clinical safety and efficacy of miRNA targeting therapies,
   we rationalized that the discovery of combinatorial targeting of
   cardiac‐predominant miRNAs would serve to confer a degree of
   cardiac‐specific therapeutic targeting and provide a feasible route for
   clinical translation–but also as means to coordinately reprogram the
   multiple pathways necessary for clinically effective adult
   cardiomyocyte regeneration.

   To this end, we performed in silico expression and pathway connectivity
   analysis to identify miRNAs predominantly expressed in the heart with
   the capacity to modulate pro‐proliferative pathways in cardiomyocytes.
   These efforts led to identification of 17 miRNAs, including miRNA‐1
   (miRNA‐1a in rodents), miRNA‐15a/b, miRNA‐16, miRNA‐27b, miRNA‐29a,
   miRNA‐34a/b/c, miRNA‐92a, miRNA‐132, miRNA‐133a, miRNA‐155, miRNA‐195a,
   miRNA‐208a, miRNA‐497a and miRNA‐873a as potential regulators.^[
   [78]^21 ^] Utilizing a novel combinatorial RNAi screening methodology
   wherein the effects of single miRNA interference were compared on the
   background of interference of all other miRNAs, we identified a common
   core combination of miRNAs whose inhibition induced cardiomyocyte
   proliferation in mature human 2D and 3D in vitro cardiac ischemic
   models. We tested the physiologic relevance and efficacy of
   combinatorial inhibition of miRNA‐1a and miRNA‐15b in driving cardiac
   regeneration in mice subjected to myocardial infarction (MI).
   miRNA‐1a/15b interference with locked nucleic acid (LNA) anti‐miRs
   significantly augmented cardiomyocyte proliferation, as determined by
   EdU incorporation, phospho‐Histone H3 staining, and Aurora B
   localization in newly proliferating adult daughter cardiomyocytes. The
   capacity of miRNA‐1a/15b to augment cardiac regeneration was further
   reflected by the decrease in fibrosis, improved cardiac contractility,
   and reduced mortality in MI mice treated with miRNA‐1a/15b LNA, but not
   in placebo controls. Critically, despite its pro‐proliferative effects
   on adult cardiomyocytes, LNA‐anti‐miR‐1a/15b (LNA‐1a/15b) did not
   induce neoplasm formation or neoplastic metabolic reprogramming in
   non‐cardiac tissue. Taken together, our data may provide a novel and
   clinically feasible strategy to induce adult cardiac regeneration.

2. Results

2.1. Combinatorial miRNA Interference (co‐miRNAi) Screen Identifies
Co‐Operative Cardiomyocyte Pro‐Proliferative Drivers

   Utilizing the microRNA Disease Database to identify miRNAs linked to
   myocardial infarction (MI) and associated with cardiac
   anti‐proliferative pathways,^[ [79]^21 ^] we identified 17 miRNAs as
   potential regulators of adult cardiomyocyte proliferation. These miRNAs
   are elevated in MI and linked to proliferative pathway suppression.^[
   [80]^16 , [81]^21 , [82]^22 ^] All 17 miRNA were expressed at
   detectable levels in healthy and infarcted human heart biopsies, albeit
   elevated in the MI patient biopsies (Figure [83]S1A, Supporting
   Information). To explore the individual and potential cooperative or
   synergistic consequence of combined miRNA suppression, we developed an
   LNA combinatorial screening assay connecting specific LNAs anti‐miRs
   with their capacity to induce cardiomyocyte proliferation. Starting
   with an initial set of LNA anti‐miRs targeting all 17 miRNAs, we
   systematically removed one LNA (targeting a specific miRNA) from the
   pool of 17 LNAs, to determine how the exclusion of a single LNA
   (targeting one miRNA) would affect cardiomyocyte proliferation. By
   comparing proliferation rates in samples lacking the single LNA
   anti‐miR against the reference control containing all 17 LNA anti‐miRs,
   this assay identifies individual LNAs that function to drive
   cardiomyocyte proliferation. This strategy and the downstream steps of
   the workflow are depicted in Figure [84]1 .

Figure 1.

   Figure 1
   [85]Open in a new tab

   Screening of co‐miRNAi that can promote cardiomyocyte proliferation in
   the P1 and P7 rat cardiomyocytes. A) Schematic diagram of the research.
   A set of 17 LNAs have been screened and then all combinations of the
   top hits will be tested in the neonatal P1 and P7 rat cardiomyocytes.
   The effects of the best combination will then be validated in vitro and
   in vivo. B,C) Effect of the removal of individual LNA from the pool of
   17 LNAs in the neonatal P1 rat cardiomyocytes. Neonatal P1 rat
   cardiomyocytes were treated with different combinations of LNAs and
   stained for EdU (green) and cardiomyocyte‐specific α‐actinin (red). (B)
   Heat map shows the fold change of EdU^+ cells in α‐actinin^+
   cardiomyocytes. Data are normalized to All (17 LNAs). The red and green
   colors indicate high and low values, respectively. n = 3. (C)
   Representative images of EdU and α‐actinin in cardiomyocytes. Scale bar
   is 50 µM. D,E) Neonatal P1 rat cardiomyocytes were treated with
   different combinations of LNAs for 2 days, and then cardiomyocytes were
   stained for Aurora B (green), cardiomyocyte‐specific α‐actinin (red),
   and DAPI (blue). (D) Heat map shows the quantification of the
   percentage of Aurora B^+ cells in telophase in α‐actinin^+
   cardiomyocytes. The red and green colors indicate high and low values,
   respectively. (E) Representative images of Aurora B, α‐actinin, and
   DAPI in cardiomyocytes. Means of n = 3 biological replicates per group.
   Scale bar is 100 µm. F,G) P7 rat cardiomyocytes were treated with
   different combinations of LNAs for 48 h and stained for Aurora B
   (green) and α‐actinin (red), and DAPI (blue). (F) Heat map shows the
   quantification of rat α‐actinin^+ cardiomyocytes that were Aurora B^+
   in telophase as a percentage of total α‐actinin^+ cells. Data are
   normalized to LNA‐NC. The red and green colors indicate high and low
   expression values, respectively. n = 3. G) Representative images of
   Aurora B, α‐actinin, and DAPI.

   We first confirmed that simultaneous interference of all 17 miRNAs by
   LNA anti‐miRs can lead to effective inactivation of all 17 target
   miRNAs in neonatal P1 rat cardiomyocytes (Figure [86]S1B, Supporting
   Information). Next, we initiated the combinatorial LNA anti‐miR screen
   utilizing 5‐Ethynyl‐2′‐deoxyuridine (EdU)‐incorporation as an initial
   readout for proliferation, with co‐staining for the
   cardiomyocyte‐specific marker, sarcomeric α‐actinin to facilitate
   identification of cycling cardiomyocytes (Figure [87]1B,C). Strikingly,
   the exclusion of LNA anti‐miRs targeting miRNA‐1a, miRNA‐15a,
   miRNA‐15b, miRNA‐27b, and miRNA‐34c led to a reduction in cardiomyocyte
   proliferation of more than 15% – indicative of potential cooperative
   function with other LNA miRNAs in driving cardiomyocyte proliferation.

   To confirm these findings, we reverted to screening LNAs targeting
   miRNA‐1a, miRNA‐15a, miRNA‐15b, miRNA‐27b, and miRNA‐34c both
   individually, and in all possible permutations in P1 rat
   cardiomyocytes. Whilst EdU serves as an important maker for S‐phase DNA
   replication, it precludes the determination of cell‐cycle completion
   and daughter cell formation.^[ [88]^23 ^] Hence, we specifically
   quantified Aurora B staining at telophase as a readout for completed
   cardiomyocyte proliferation (Figure [89]S2A, Supporting Information).
   Telophase is the final phase of mitosis, during which duplicated
   genetic material carried in the nucleus of a parent cell is separated
   between two daughter cells. We assessed Aurora kinase B localization at
   telophase in 31 individual and combinatorial conditions corresponding
   to the 5 candidate LNA anti‐miRs targeting miRNA‐1a, miRNA‐15a,
   miRNA‐15b, miRNA‐27b, and miRNA‐34c. Cardiomyocyte proliferation was
   assessed by co‐localization of the contractile ring and midbody Aurora
   kinase B staining, with sarcomeric α‐actinin and DAPI counterstains to
   mark cardiomyocytes and their nuclei, respectively (Figure [90]1D,E;
   Figure [91]S2B, Supporting Information). These analyses revealed that
   treatment with 3 specific LNA combinations, comprising LNA anti‐miRs
   targeting miRNA‐1a and miRNA‐15b (LNA‐1a/15b), miRNA‐1a, miRNA‐15a and
   miRNA‐15b (LNA‐1a/15a/15b), and miRNA‐1a, miR15a and miRNA‐27b
   (LNA‐1a/15a/27b), resulted in consistently increased numbers of
   cardiomyocytes in telophase compared to control scrambled LNA (LNA‐NC)
   treated cells (Figure [92]1D,E; Figures [93]S2B and [94]S3A–C,
   Supporting Information). In accord, all miRNAs targeted within the
   respective combinations were effectively repressed (Figure [95]S3D–F,
   Supporting Information).

   The early neonatal mouse heart has some proliferative and regenerative
   response, but this capacity is lost from P7.^[ [96]^24 ^] Thus, to
   determine the impact of the identified LNA miRNA combinations in
   post‐mitotic cardiomyocytes, all combinations of LNAs targeting
   miRNA‐1a, miRNA‐15a, miRNA‐15b, and miRNA‐27b were screened in P7 rat
   cardiomyocytes. Utilizing Aurora B staining at telophase as readout,
   the specific combination of LNAs targeting miRNA‐1a and miRNA‐15b
   (LNA‐1a/15b) proved most effective at inducing proliferation in
   post‐mitotic cardiomyocytes (Figure [97]1F,G; Figure [98]S2C,
   Supporting Information). Consistent with previous findings, LNAs
   targeting miRNA‐1a and miRNA‐15b led to efficient target knockdown
   (Figure [99]S4A, Supporting Information). Thus, combinatorial
   interference of miRNA‐1a/15b drives post‐mitotic cardiomyocyte
   proliferation and cell‐cycle completion in rat cardiomyocytes in vitro.
   To further confirm that the proliferative responses detected as a
   result of combinatorial inhibition of miRNA‐1a/15b were indeed
   originating from cardiomyocytes, and from not a hyper proliferative
   cell‐type such as cardiac fibroblasts, we assessed the effects of
   miRNA‐1a/15b LNA in cardiac fibroblasts. As shown in Figure [100]S5A
   (Supporting Information), B, LNA‐anti‐miR1a/15b had negligible impact
   on fibroblasts proliferation. Taken together, these data suggest a
   degree of cell‐type specificity in LNA ‐miR1a/15b function in
   proliferative induction.

2.2. Combinatorial miRNA‐1a/15b Interference Drives Adult Cardiomyocyte
Proliferation and Maintenance of Cardiac Function in Pathology

   The pro‐proliferative impact of miRNA‐1a/15b function in vitro led us
   to explore its potential therapeutic utility in vivo. To that end, we
   assessed the effects of LNA‐1a/15b treatment in a mouse model of
   ST‐Elevation Myocardial Infarction (STEMI). STEMI is characterized by
   the development of acute myocardial infarction as a result of decreased
   coronary blood flow, accounting for ≈40% of all myocardial infarction
   presentations in hospitals.^[ [101]^25 , [102]^26 ^]

   We first assessed the efficiency of LNAs targeting miRNA‐1a and
   miRNA‐15 both individually and in combination in adult mice, as
   depicted in Figure [103]S6A (Supporting Information). Both individual
   and combinatorial delivery of LNAs targeting miRNA‐1a and miRNA‐15b led
   to effective target suppression (Figure [104]S6B–D, Supporting
   Information). Whilst depletion of miRNA‐1a or miRNA‐15b in mouse hearts
   led to a slight but statistically insignificant increase in
   proliferation, combinatorial depletion of both miRNA‐1a and miRNA‐15b
   significantly induced cardiomyocyte proliferation as shown by increased
   pHH3 (Figure [105]S6E,F, Supporting Information).

   We next examined if inhibition of miRNA‐1a/15b could similarly increase
   cardiomyocyte proliferation in response to a mouse STEMI model (Figure
   [106]2A). MI was induced by permanent ligation of left anterior
   descending coronary artery, resulting in elevated expression of
   miRNA‐1a and miRNA‐15b 8 h post‐MI in the border zone (Figure [107]2B).
   Mice were treated with LNA‐1a/15b four h post‐surgery to mimic the
   average duration that post‐STEMI patients receive treatment.^[ [108]^25
   , [109]^26 ^] At 28 days post‐MI surgery the experiment was terminated
   due to the decline in cardiac function and increased mortality in mice
   subjected to MI and treated with control non‐silencing LNAs (LNA‐NC)
   (Figure [110]2C). LNA‐1a/15b treatment effectively suppressed RNA
   levels of miRNA‐1a and miRNA‐15b in remote zone biopsies of the
   respective mice (Figure [111]2D,E). Echocardiographic measure of
   ejection fraction (EF), as a readout for cardiac contractile function,
   was rapidly reduced in LNA‐NC treated mice subjected to MI concomitant
   to increased mortality (Figure [112]2F,G). In contrast, mice treated
   with LNA‐1a/15b largely maintained cardiac function and organismal
   viability despite MI intervention (Figure [113]2F,G). To ascertain the
   effects of LNA‐1a/15b treatment in mice subjected to MI, the heart was
   first divided into remote and border zones (Figure [114]2H). As the
   detected rescue and maintenance of healthy cardiac in LNA‐1a/15b
   treated mice would predict reduced expression of pathologic hypertrophy
   markers, we examined the expression of atrial natriuretic peptide
   (Nppa), beta‐myosin heavy chain 6 (Myh6) and beta‐myosin heavy chain 7
   (Myh7) (Figure [115]2I). Nppa and Myh7 RNA levels were reduced after
   LNA‐1a/15b injection in the border zone compared to the control LNA‐NC
   injection (Figure [116]2I). Similarly, MI‐induced Myh7:Myh6 ratio was
   reduced by combinatorial inhibition of miRNA‐1a and miRNA‐15b,
   indicative of suppression of the pathologic hypertrophic gene program
   induction in LNA‐1a/15b treated hearts (Figure [117]2I). The remote
   zone serves as an internal tissue control due to its distance from the
   infarct site and is thus less affected by tissue damage caused by MI
   (Figure [118]2H). Consistent with this, LNA‐1a/15b treatment primarily
   impacted hypertrophic marker expression in border zone biopsies, but
   not in the largely unaffected remote zone (Figure [119]2I).

Figure 2.

   Figure 2
   [120]Open in a new tab

   The effects of combinatorial miRNA‐1a/15b interference on myocardial
   infarction‐induced cardiac damage in vivo. A) Schematic representation
   of the experimental timeline of C57BL/6J mice subjected to MI surgery.
   The LNA‐1a/15b or LNA‐NC was injected at 4 h after MI surgery, and
   echocardiography measurement was performed on days 0, 7, 14, 21, and
   28. The mouse hearts were harvested on day 28 after MI surgery. A
   subset of heart biopsies were harvested at 8 hours post‐MI, serving as
   an internal control for the pathologic stress‐induced upregulation of
   miR‐1a and miR‐15b. B) Relative miRNA expression in heart border zone
   biopsies from mice from mice 8 hours post‐MI and healthy controls. Data
   are expressed as means ± SEM; n = 3 mice per group; ^** P < 0.01 versus
   control. Two‐tailed unpaired t‐test. C) Whole hearts harvested from
   mice. D) Relative expression of miRNA‐1a in remote zone of mice
   subjected to control or MI surgery. Data are expressed as means ± SEM;
   n = 4 mice per group. ^** P < 0.01 versus Control LNA‐NC, %P < 0.05
   versus MI LNA‐NC. Two‐tailed unpaired t‐test. E) Relative expression of
   miRNA‐15b in remote zone of mice subjected to control or MI surgery.
   Data are expressed as means ± SEM; n = 4 mice per group; ^** P < 0.01
   versus Control LNA‐NC, %P < 0.05 versus MI LNA‐NC. Two‐tailed unpaired
   t‐test. F) Longitudinal monitoring of cardiac % Ejection fraction
   (%EF). Data are expressed as means ± SEM; N = 4/5 mice per group; ^** P
   < 0.01 versus MI LNA‐NC. Two‐tailed unpaired t‐test. G) Survival curve
   in MI group. N = 4/5 mice per group. H) The mouse heart was divided
   into the Remote and Border zone, as depicted. I) Heat map shows the
   mRNA expression levels of pathological remodeling marker genes (Nppa,
   Myh7, and Myh7/Myh6 ratio) and fibrotic remodeling marker genes (Tgfβ2,
   Col3a1, and Thbs1) after LNA‐NC or LNA‐1a/15b injection and MI injury
   in the Remote and Border zone. Data was normalized to Control LNA‐NC.
   The red and green colors indicate high and low expression values,
   respectively. n = 4 mice per group. Two‐tailed unpaired t‐test. J)
   Representative hematoxylin and eosin (HE) staining and picrosirius red
   stained heart sections at 28‐days post MI surgery. K) Quantitative
   cardiac fibrosis analysis. Data are expressed as means ± SEM; n = 4
   mice per group; ^** P < 0.01 versus Control LNA‐NC, %P < 0.05 versus MI
   LNA‐NC. Two‐tailed unpaired t‐test. MI, Myocardial Infarction.

   Maintenance of cardiac function and the absence of mortality in MI mice
   treated with LNA‐1a/15b indicates a possible attenuation of cardiac
   fibrosis and stiffness in these mice. Thus, we profiled key fibrotic
   marker genes including transforming growth factor β2 (Tgfβ2), Collagen
   type III alpha 1 chain (Col3a1), and thrombospondin 1 (Thbs1). As noted
   in Figure [121]2I, fibrotic marker gene expression was significantly
   reduced in MI mice treated with LNA‐1a/15b compared to LNA‐NC controls.
   Consistent with the negligible impact of the infarct on the remote
   zone, fibrotic marker gene expression in the remote zone was
   indistinguishable between control and MI mice, regardless of the
   treatment regime (Figure [122]2I). Finally, we stained sections of the
   respective hearts with Picrosirius Red, an established marker for
   tissue fibrosis.^[ [123]^27 ^] As noted in Figure [124]2J,K, LNA‐1a/15b
   treated MI mice exhibited considerably less cardiac fibrosis compared
   to LNA‐NC treated controls. Consistent with this observation, staining
   for the apoptosis marker cleaved‐Caspase 3, revealed increased cell
   death in mouse LNA‐NC‐treated hearts subjected to MI, but not in hearts
   treated with LNA‐1a/15b (Figure [125]S7A, Supporting Information). In
   sum, LNA‐1a/15b treatment significantly attenuated the development of
   cardiac fibrosis, and its progression to heart failure and death in a
   mouse model of STEMI.

2.3. Combinatorial miRNA‐1a/15b Interference Specifically Drives
Cardiomyocyte Proliferation In Vivo with Minimal Off‐Target Effects

   To ascertain if the beneficial effects of combinatorial miRNA‐1a and
   miRNA‐15b depletion are linked to adult cardiomyocyte proliferation, we
   examined the distribution of the mitotic S‐phase marker Ki67 and the
   metaphase marker phosphorylated Histone H3 (pHH3) in heart biopsies of
   the respective groups. Increased Ki67 and pHH3 signal was detected in
   cardiomyocytes of both sham and MI LNA‐1a/15b treated mice (Figure
   [126]3A–D; Figure [127]S8A,B, Supporting Information), with greater
   levels of Ki67 and pHH3 induction observed in MI operated mice.
   Strikingly, when the respective biopsies were assessed for expression
   of the cytokinesis marker Aurora B, as a readout for cell‐cycle
   completion and daughter cell formation, we only detected a significant
   increase in the fraction of Aurora B^+ cardiomyocytes in hearts of mice
   subjected to MI and treated with LNA‐1a/15b (Figure [128]3B,D).

Figure 3.

   Figure 3
   [129]Open in a new tab

   The effects of combinatorial miRNA‐1a/15b interference on proliferation
   in cardiac tissue and in non‐cardiac tissues in vivo. A) Representative
   immunofluorescence images of Ki67 on heart sections of adult hearts
   injected with LNA‐1a/15b or control 28‐days post MI injury. Ki67 labels
   proliferating cells (green); cardiomyocyte‐specific α‐actinin (red) and
   DAPI (Blue). B) Representative immunofluorescence images of Aurora B on
   heart sections of adult hearts injected with LNA‐1a/15b or control
   28‐days post MI injury. Aurora B labels proliferating cells (green);
   cardiomyocyte‐specific α‐actinin (red) and DAPI (Blue). C)
   Quantification of percentages of Ki67^+ cardiomyocytes. Data are
   expressed as means ± SEM; N = 3 mice per Control group, n = 5 mice per
   MI group; ^* P < 0.05 versus Control LNA‐NC, %P < 0.05 versus MI
   LNA‐NC. Two‐tailed unpaired t‐test. Scale bar is 100 µm. D)
   Quantification of percentages of Aurora B^+ cardiomyocytes. Data are
   expressed as means ± SEM; N = 3 mice per Control group, N = 5 mice per
   MI group; %P < 0.05 versus MI LNA‐NC. Two‐tailed unpaired t‐test. Scale
   bar is 100 µm. E) Representative images of Ki67 (green), Phalloidin
   (red), and DAPI (blue) in liver. Scale bar is 100 µm. F) Representative
   images of Ki67 (green), Phalloidin (red), and DAPI (blue) in SKM. Scale
   bar is 100 µm. G) Quantification of percentages of Ki67^+ liver cells.
   Data are expressed as means ± SEM; n = 3 mice per group. Two‐tailed
   unpaired t‐test. H) Quantification of percentages of Ki67^+ SKM cells.
   Data are expressed as means ± SEM; n = 3 mice per group. Two‐tailed
   unpaired t‐test. I) Representative images of Ki67 (green), Phalloidin
   (red), and DAPI (blue) in kidney. Scale bar is 100 µm. J)
   Representative images of Ki67 (green), Phalloidin (red), and DAPI
   (blue) in WAT. Scale bar is 100 µm. K) Quantification of percentages of
   Ki67^+ kidney cells. Data are expressed as means ± SEM; n = 3 mice per
   group. Two‐tailed unpaired t‐test. L) Quantification of percentages of
   Ki67^+ WAT cells. Data are expressed as means ± SEM; n = 3 mice per
   group. Two‐tailed unpaired t‐test. SKM, Skeletal Muscle; WAT, White
   Adipose Tissue. M) ASAT and N) ALAT levels in plasma of C57BL/6J mice
   injected with wither LNA‐1a/15b or Control LNA‐NC 28‐days post MI
   injury. Data are expressed as means ± SEM; n = 5‐6 mice per group. MI,
   Myocardial Infarction; ASAT, Aspartate aminotransferase; ALAT, Alanine
   aminotransferase.

   LNAs are delivered systemically and can detrimentally affect non‐target
   tissues. This is particularly relevant in regenerative‐ or
   proliferative‐ type therapies where proliferation activators targeting
   specific cell‐types, may in parallel induce neoplasm formation due to
   mitotic activation in non‐target cells. Thus, a degree of cell‐type
   specificality in therapeutic response would be prudent. In Figure
   [130]S5A,B (Supporting Information), we demonstrated a lack of
   LNA‐1a/15b impact on cardiac fibroblasts proliferation. To extend these
   findings, we asked how LNA‐1a/15b inactivation would impact tissue
   proliferation of key non‐cardiac cell‐types in mice. As noted in
   Figures [131]3E–L and [132]S9A–H (Supporting Information), negligible
   changes in cell proliferation, as determined by Ki67 and pHH3 staining,
   was detected in tissues including liver, skeletal muscle (SKM), kidney
   and white adipose tissue (WAT) of mice subjected to sham or MI surgery,
   in the absence or presence of LNA‐1a/15b treatment.

   Changes in cell and tissue metabolism is another important feature of
   neoplastic transformation and are broadly characterized by a shift from
   glucose and fatty acid oxidation as the primary mode of ATP generation
   to glycolysis, a process less efficient in ATP generation but
   advantageous for the generation of nucleotides, amino acids and
   phospholipids necessary building‐blocks of cell proliferation.^[
   [133]^28 , [134]^29 ^] Given this, we performed metabolomic analysis on
   non‐cardiac tissue namely the liver, lung, kidney, spleen and white
   adipose tissue (WAT) (Figure [135]S10A–E, Supporting Information).
   Principal Component Analysis (PCA) revealed broad and overlapping
   signals distribution across the respective groups in liver, lung,
   kidney, and spleen biopsies, suggestive of insignificant deviation from
   Control LNA‐NC treated mice. In contrast, analysis of WAT biopsies from
   the respective groups revealed unique signal clustering of biopsies
   derived from LNA‐NC mice subjected to MI compared to the other groups.
   Importantly, WAT biopsies harvested from mice subjected to MI and
   treated with LNA‐1a/15b exhibited pronounced signal overlap with
   Control LNA‐NC treated tissue (Figure [136]S10A–E, Supporting
   Information). Taken together, combinatorial inhibition of miRNA‐1a and
   miRNA‐15b caused negligible off‐target effects on non‐cardiac tissue
   metabolism and proliferation rates.

   Finally, to understand any possible effect of LNA‐1a/15b therapy on
   systemic toxicity, we assessed circulating levels of biomarker enzymes
   Aspartate aminotransferase (ASAT) and Alanine aminotransferase
   (ALAT).^[ [137]^30 ^] While ASAT is commonly associated with liver and
   heart damage, it can also be found in other tissues, including skeletal
   muscle and kidneys. ALAT is primarily found in hepatocytes. Upon tissue
   damage due to drug toxicity, these enzymes are released into the blood.
   Upon tissue damage due to drug toxicity, release these enzymes into the
   blood. Thus, circulating ASAT and ALAT levels in the blood directly
   correlates with tissue damage and toxicity.^[ [138]^31 ^] As noted in
   Figure [139]3M,N, LNA‐1a/15b had minimal impact on circulating ASAT and
   ALAT levels compared to controls, indicative of an acceptable safety
   profile. In sum, these data indicate that LNA‐1a/15b treatment is
   non‐toxic and provokes minimal tissue off‐target effects.

2.4. Combinatorial miRNA‐1/15b Interference Enhances Cardiomyocyte
Proliferation and Function in Human Cardiac Organoids

   In order to understand the potential of LNA‐1a/15b therapy for clinical
   translation, we first assessed the effects of LNA‐1/15b in the human
   induced pluripotent stem cells‐derived cardiomyocytes (hiPSC‐CMs)
   cultured under control normoxia or in hypoxia (1%O[2]) as an in vitro
   model for myocardial infarction.^[ [140]^32 ^] LNA‐1/15b treatment
   effectively suppressed RNA levels of miRNA‐1 and miRNA‐15b in the
   hiPSC‐CMs (Figure [141]S11A,B, Supporting Information) in normoxia and
   hypoxia. As observed in Figure [142]S11C,D (Supporting Information),
   LNA‐1/15b treatment led to increased Aurora B signal, indicative of a
   conserved function for LNA‐1/15b in driving mature human cardiomyocyte
   proliferation and cell‐cycle completion. Next, we utilized human
   iPSC‐derived 3D cardiac tissue mimetics to determine the proliferative
   impact of LNA‐1/15b treatment in tissue. Utilizing a variation of a
   previously established 3D cardiac mimetic model with the inclusion of a
   maturation phase,^[ [143]^33 , [144]^34 , [145]^35 ^] we were able to
   generate adult cardiac mimetics exhibiting spontaneous beating
   characteristics and gene expression profiles characteristic of adult
   tissue^[ [146]^36 ^] (Figure [147]S12A–H and Video [148]S1, Supporting
   Information). In this setting, cardiac mimetics were cultured under
   control normoxia or in hypoxia, treated with LNA‐NC or LNA‐1/15b, and
   assessed for proliferation by EdU‐incorporation (as readout for S‐Phase
   entry and DNA replication) and pHH3 staining (Figure [149]4A–D). As
   observed in the mouse STEMI model, a similar pattern of proliferative
   induction by LNA‐1/15b was detected in both control normoxia and
   hypoxic conditions, with elevated levels of proliferation observed in
   hypoxia. Consistent with previous data, LNA‐1/15b treatment effectively
   suppressed RNA levels of miRNA‐1 and miRNA‐15b in human cardiac
   mimetics (Figure [150]S13A,B, Supporting Information). Furthermore, as
   observed in the mouse studies, LNA‐1/15b treatment of human organoids
   subjected to severe hypoxic stress prevented the onset of contractile
   dysfunction (Figure [151]4sE,FZ; Figure [152]S13C,D, Supporting
   Information). Contractility was determined by determining the peak
   amplitude of contracting cardiac organoids (Figure [153]4E) and calcium
   flux measurements (Figure [154]S13C, Supporting Information) from the
   respective groups, and quantified as shown in Figures [155]4F and
   [156]S13D (Supporting Information). Cardiac organoids cultured in
   hypoxia revealed the expected decrease in contractile amplitude
   concomitant to increased cycling frequency (Figure [157]4E,F; Figure
   [158]S13C,D, Supporting Information). However, LNA‐1/15b treatment in
   human cardiac mimetics rescued the loss in contractility induced by
   severe hypoxic stimulation (Figure [159]4; Figure [160]S13C,D,
   Supporting Information). In sum, our data indicates that LNA‐1/15b
   drives mitotic entry and cell‐cycle completion in to maintain cardiac
   function in mature human tissue in the face of hypoxic insult.

Figure 4.

   Figure 4
   [161]Open in a new tab

   The effects of combinatorial miRNA‐1a/15b interference on cardiomyocyte
   proliferation and cardiac function in human cardiac tissue mimetics. A)
   Representative immunofluorescence images of EdU incorporation on
   sections of 40‐day‐old human cardiac mimetics treated with LNA‐1a/15b
   or LNA‐NC in both normoxia and hypoxia. EdU labels proliferating cells
   (Magenta); Wheat‐Germ‐Agglutinin (WGA) marks cell surfaces (green);
   cardiomyocyte‐specific α‐actinin (red) and DAPI (Blue). B)
   Quantification of percentages of EdU^+ cardiomyocytes. Data are
   expressed as means ± SEM; n = 3 per group; ^** P < 0.01 versus Normoxia
   LNA‐NC, %P < 0.05 versus Hypoxia LNA‐NC. Two‐tailed unpaired t‐test.
   Scale bar is 100 µm. C) Representative immunofluorescence images of
   pHH3 on sections of 40‐day‐old human cardiac mimetics treated with
   LNA‐1a/15b or LNA‐NC in both normoxia and hypoxia. pHH3 labels
   proliferating cells (Magenta); cardiomyocyte‐specific α‐actinin (red)
   and DAPI (Blue). Scale bar is 100 µm. D) Quantification of percentages
   of pHH3^+ cardiomyocytes. Data are expressed as means ± SEM; n = 3 per
   group; ^** P < 0.01 versus Normoxia LNA‐NC, %P < 0.05 versus Hypoxia
   LNA‐NC. Two‐tailed unpaired t‐test. E) Representative traces of
   contractile human cardiac mimetics. Human cardiac mimetics were treated
   with LNA‐1a/15b or LNA‐NC for 3 days in both normoxia and hypoxia, and
   then the contractility assays were performed by determining the
   amplitude peak of contracting human cardiac mimetics. F) Quantification
   of the amplitude peak of contracting human cardiac mimetics. Data are
   expressed as means ± SEM; 4 – 5 organoids per group; *P < 0.05 versus
   Normoxia LNA‐NC, %P < 0.05 versus Hypoxia LNA‐NC. Two‐tailed unpaired
   t‐test.

2.5. Mechanistic Basis of LNA‐1a/15b Function in Cardiomyocyte Proliferative
Induction

   Whilst the vast majority of therapeutic interventions involving miRNA
   inhibition rely on the targeted inactivation of a single miRNA, our
   screening efforts reveal that combinatorial approaches can potentially
   provide greater efficacy and robustness in driving adult cardiomyocyte
   proliferation. Further, this combinatorial effect is not restricted to
   rodents, rather it is conserved across species as noted by the effects
   of miRNA‐1a/15b inactivation in mouse, rat, and human cardiomyocytes
   and cardiac tissue (Figure [162]1D–G; Figure [163]S6E,F, Supporting
   Information; Figure [164]3A–D; Figure [165]S8A,B, Supporting
   Information; Figure [166]4A–D). Strikingly, due to the enrichment of
   miRNA‐1a and miRNA‐15b in cardiomyocytes, we detected a negligible
   effect of LNA‐1a/15b treatment on proliferative induction of
   fibroblasts, and on the heterogenous population of cells contained
   within non‐cardiac tissue (Figure [167]S5A,B, Supporting Information;
   Figure [168]3E–L; Figure [169]S9A–H, Supporting Information). The
   specificity of LNA‐1a/15b function is also reflected at the level of
   the respective non‐cardiac tissue metabolomes (Figure [170]S10A–E,
   Supporting Information). Given this, we sought to understand the
   downstream consequence of miRNA‐1a/15b inhibition on coding gene
   regulators, as mediators of proliferative induction, specifically in
   cardiomyocytes. Thus, NRCs were treated with LNA‐NC, LNA‐1a, LNA‐15b or
   LNA‐1a/15b and hypoxia stimulated (to mimic the MI setting). As shown
   in the heatmap in Figure [171]5A, differential gene expression analysis
   revealed specific changes in gene expression unique to cardiomyocytes
   stimulated with the combination of LNAs targeting both miR1a/15b, but
   not in control LNA‐NC, or in cardiomyocytes treated with LNAs targeting
   only miRNA‐1a or miRNA‐15b individually. Next, Kyoto Encyclopedia of
   Genes and Genomes (KEGG) clustering analysis revealed enrichment of
   genes linked to cardiac development, cell/tissue homeostasis, and
   cardiac growth and morphogenesis (Figure [172]5B; Figure [173]S14A–G,
   Supporting Information). Consistent with this data, Gene Set Enrichment
   Analysis (GSEA) enrichment analysis of all differentially expressed
   genes between LNA‐NC, LNA‐1a, LNA‐15b, and LNA‐1a/15b treated groups
   revealed distribution into cardiac linked pathways (Figure [174]5C).
   Given that the LNA‐1a/15b combination proved most relevant by virtue of
   its ability to maintain cardiac function and organismal survival in the
   face of MI, we sub‐classified genes within the respective Gene Ontology
   (GO) pathways into 4 groups comprising of genes regulated specifically
   by miRNA‐1a or miRNA‐15b depletion, respectively (miRNA‐1a and
   miRNA‐15b only), genes regulated in‐common by both miRNA‐1a and
   miRNA‐15b depletion (Common), and differentially expressed genes
   specific to combinatorial inhibition of miRNA‐1a and miRNA‐15b (Unique)
   (Figure [175]5C). As expected, regulation of the vast majority of genes
   can be linked directly to miRNA‐1a and/or miRNA‐15b function
   (Figure [176]5C; Figure [177]S15A–H, Supporting Information).
   Surprisingly, a fairly large number of differentially expressed genes
   could neither be linked to direct miRNA‐1a or miRNA‐15b function, based
   on known and predicted targets of the respective miRNAs and our
   expression data (Figure [178]S15D,H, Supporting Information).^[
   [179]^21 ^] Genes depicted in the Figure [180]S15D (Supporting
   Information) heatmap encompass miRNA‐1a/15b regulated genes that are
   unaffected or only mildly altered (and below the significance
   threshold) upon treatment with miRNA‐1a LNAs or miRNA‐15b LNAs
   individually. Many of the genes within this Unique subgroup are
   characterized by a role in inducing cardiomyocyte de‐differentiation,
   early cardiac development and remodeling comprising of genes such as
   T‐Box Transcription Factor 20 (TBX20), PR‐Domain Zinc Finger Protein 16
   (PRDM16), ADP‐Ribosylhydrolase Like 1 (ADPRHL1),^[ [181]^37 ^] Glycogen
   Synthase Kinase 3 Beta (GSK3B), Fibroblast Growth Factor Receptor
   Substrate 2 (FRS2), Sorbin And SH3 Domain Containing 2 (SORBS2), WW
   Domain Containing Transcription Regulator 1 (WWTR1) and SRY‐Box
   Transcription Factor 9 (SOX9) (Figure [182]S15D, Supporting
   Information). Connected to these are modulators of sarcomerogenesis,
   cardiac calcium handling and contractility including Cysteine And
   Glycine Rich Protein 3 (CSRP3), Synaptopodin 2 Like (SYNPO2L), Myosin
   XVIIIB (MYO18B), Xin Actin Binding Repeat Containing 1 (XIRP1), NADPH
   Oxidase 4 (NOX4), Calcium Voltage‐Gated Channel Subunit Alpha1 C
   (CACNA1C), Mitochondrial Dynamin Like GTPase (OPA1), Ankyrin 2 (ANK2),
   PDZ And LIM Domain 7 (PDRM17) and Ryanodine Receptor 2 (RYR2) (Figure
   [183]S15D, Supporting Information). Further, we detected upregulation
   of drivers of angiogenesis, lymphangiogenesis, neurogenesis, and
   extracellular matrix and stromal cell activators consisting of Prospero
   Homeobox 1 (PROX1), Neuropilin 2 (NRP2) and Transforming Growth Factor
   Beta 3 (TGFB3) (Figure [184]S15D, Supporting Information). Finally,
   this Unique group contained general transcriptional and translational
   modulators, and epigenetic regulators including key factors such as
   AT‐Rich Interaction Domain 1A (ARID1A), Polybromo 1 (PBRM1), Mediator
   Complex Subunit 1 (MED1), Lysine Demethylase 6B (KDM6B), and lesser
   understood components including Zinc Finger MIZ‐Type Containing 1
   (ZMIZ1) and Dishevelled Segment Polarity Protein 3 (DVL3) (Figure
   [185]S15D, Supporting Information). Taken together, these data suggest
   a possible molecular synergism wherein simultaneous depletion of
   miRNA‐1a and miRNA‐15b facilitates the specific activation of genes,
   which in this context serves to provide a pro‐proliferative environment
   in post‐mitotic adult cardiomyocytes by co‐opting key accessory
   pathways driving adult cardiomyocyte de‐differentiation and plasticity,
   remodeling of the cardiomyocyte sarcomere and contractile machinery,
   concomitant to the induction of cardiac angiogenesis, neurogenesis and
   stromal activation to support survival of new cardiomyocytes, to enable
   effective adult cardiac regeneration.

Figure 5.

   Figure 5
   [186]Open in a new tab

   Changes in gene expression and metabolites with inhibition of miRNA‐1a
   and miRNA‐15b. A) The heatmap shows the variant genes. X‐axis: samples
   of LNA‐NC, LNA‐1a, LNA‐15b, LNA‐1a/15b; The red and blue colors
   indicate upregulation and downregulation, respectively. Means of n = 3
   biological replicates per group. B) Gene Ontology (GO) biological
   process terms enriched in LNA‐1a/15b‐regulated genes. C) Top 7 GO
   biological process terms related to the heart development enriched in
   LNA‐1a, LNA‐15b, and LNA‐1a/15b. Red, genes regulated specifically by
   miRNA‐1a inhibition; Blue, genes regulated specifically by miRNA‐15b
   inhibition; grey, genes regulated in‐common by both miRNA‐1a and
   miRNA‐15b depletion; yellow, genes regulated unique to combinatorial
   inhibition of miRNA‐1a and miRNA‐15b. D) Pathway analysis bubble plot
   in KEGG shows the enriched metabolic pathways in P1 rat cardiomyocytes
   treated with LNA‐1a/15b under hypoxia. The horizontal axis is the
   pathway impact, which represents the importance of differential
   metabolites in metabolic pathways. The vertical axis represents the
   results of pathway enrichment analysis. The dot size represents the
   centrality of the metabolites involved in the corresponding metabolic
   pathway. The color of the dot represents the impact factor; large sizes
   and dark colors represent the central metabolic pathway enrichment and
   pathway impact values, respectively. E) Heat map of relative metabolite
   abundance in Neonatal P1 rat cardiomyocytes treated with either control
   LNA‐NC or LNA‐1a/15b, and then incubated in the hypoxia and hypoxia for
   2 days. Depicted are metabolites with log2(fold change) > 0.5 compared
   to Normoxia LNA‐NC treatment and adjusted p value < 0.01 in at least
   one treatment group compared to the corresponding control LNA‐NC. n = 4
   biological replicates per group.

   Given the crucial role of the cell metabolome in enabling and
   facilitating cell proliferation,^[ [187]^38 ^] we analyzed changes in
   cellular metabolite composition as a result of LNA‐1a/15b treatment
   under basal and hypoxic stress conditions. As noted in
   Figure [188]5D,E, depletion of miRNA‐1a/15b in hypoxic cardiomyocytes
   led to an elevation of metabolites necessary for proliferative
   induction and cell‐cycle completion, including metabolites of purine
   and pyrimidine biosynthesis,^[ [189]^39 , [190]^40 ^] branched‐chain
   amino acid (BCAA) metabolites,^[ [191]^28 , [192]^41 ^] nicotinamide
   and riboflavin components,^[ [193]^28 , [194]^41 ^] panthotenate and
   co‐enzyme A metabolites necessary for acylation of metabolic
   intermediates,^[ [195]^39 , [196]^42 ^] folate and 1‐carbon pathway
   metabolites,^[ [197]^41 , [198]^42 ^] phosphatidylcholines^[ [199]^43
   ^] and glycerol hydroxybutyrates.^[ [200]^28 , [201]^44 , [202]^45 ^]
   Strikingly, there was little evidence for elevation of these
   metabolites neither in response to hypoxia stimulation alone nor upon
   LNA‐NC treatment (Figure [203]5D,E). Thus, these data suggest that
   LNA‐1a/15b serves to very specifically reprogram the cardiac
   transcriptome and metabolome to enable and support cardiac tissue
   regeneration.

2.6. Discussion and Conclusion

   miRNAs are known to play critical roles in cardiac pathophysiology,
   including MI and heart failure. Most studies have primarily focused on
   the role of individual miRNAs, although combinatorial activities of
   miRNAs are being increasingly investigated and recognized in recent
   years.^[ [204]^46 , [205]^47 ^] Although the impact of individual miRNA
   modulation cannot be underestimated, we speculate that more complex
   processes might necessitate the simultaneous combinatorial modulation
   of two or more regulatory moieties. Given the complexity and robustness
   engrained within adult cardiomyocytes to maintain terminal
   differentiation, structural integrity within the tissue complex, and
   interaction with surrounding cardiomyocytes, stromal and mural cells in
   order to facilitate cardiac contractility and function, our
   observations indicate that the induction and re‐entry of post‐mitotic
   adult cardiomyocytes could be one such example to benefit from complex
   and combinatorial miRNA modulation. This is supported by recent data
   where combinatorial modulation of coding genes through simultaneous
   co‐expression of four genes (comprising transcription factors and
   cell‐cycle regulator combinations) have been shown necessary to drive
   adult cardiomyocyte mitotic re‐entry.^[ [206]^48 , [207]^49 , [208]^50
   ^] Although these findings are of key importance in understanding
   mechanisms underlying the induction of adult cardiomyocyte mitosis,
   translating these findings into a clinically feasible strategy is
   challenging due to the need for cardiac‐directed gene therapy
   (necessitating the need for invasive delivery procedures) and the clear
   risk of therapy leakiness which could lead to neoplasm formation in
   non‐cardiac tissue (due to the ubiquitous nature and function of the
   transcription factors and cell‐cycle regulator combinations ectopically
   expressed in these strategies^[ [209]^48 , [210]^49 , [211]^50 ^]).
   Thus, developing clinically feasible and accessible therapeutic
   strategies for cardiac regeneration remains an unmet need.
   Combinatorial miRNAs therapy targeting cardiac enriched miRNAs may
   offer such an avenue by virtue of the ability to inactivate these
   miRNAs through systemic delivery of stabilized anti‐sense moieties,
   thus removing the need for invasive procedures such as catheter‐based
   delivery or direct cardiac injection.^[ [212]^51 , [213]^52 ^]

   According to the Human microRNA Disease Database (HMDD)
   microRNA–disease associations, the 17 miRNAs we selected are
   upregulated in the patients with MI and heart failure.^[ [214]^21 ^]
   This is consistent with our data demonstrating the elevation of all 17
   miRNAs heart biopsies of patients with myocardial infarction (Figure
   [215]S1A, Supporting Information). Among the proliferation‐regulating
   miRNAs, we have used a combinatory approach to identify a combination
   that could further increase cardiomyocyte proliferation compared to
   individual miRNA. Our study identified that simultaneous inhibition of
   miRNA‐1a and miRNA‐15b increased cardiomyocyte proliferation and
   improved cardiac function in vitro and in vivo. Both of miRNA‐1a and
   miRNA‐15b are highly expressed in hearts, and they have been documented
   to regulate cardiomyocyte proliferation.^[ [216]^53 , [217]^54 ^]
   Overexpression of miRNA‐1 in the developing hearts leads to decreased
   cardiomyocyte proliferation.^[ [218]^55 , [219]^56 ^] Conversely,
   miRNA‐1a deletion increases cardiomyocyte proliferation in the adult
   hearts,^[ [220]^54 ^] and attenuates cardiac I/R injury in mice and
   rats.^[ [221]^57 , [222]^58 ^] miRNA‐15b is upregulated in the
   infarcted zone of porcine cardiac tissue in response to ischemic
   injury, and miRNA‐15b inhibition increases cardiac regeneration and
   protects against ischemic‐induced injury.^[ [223]^59 ^] Our data showed
   that combinatory inhibition of miRNA‐1a and miRNA‐15b could further
   increase cardiomyocyte proliferation in P1 and P7 rat cardiomyocytes,
   and could enhance cardiomyocyte proliferation and improve cardiac
   function in mice in vivo. As access to cardiac tissue, in particular,
   is challenging due to ethical concerns, human cardiac organoids serve
   as proxy material in which the functionality of this therapeutic
   approach can be assessed. In utilizing iPSC‐derived human cardiac
   mimics, we were able to demonstrate a conservation of function of
   inhibition of miRNA‐1a and miRNA‐15b in driving adult human
   cardiomyocyte proliferation, leading to maintenance of cardiac
   contractility in the face of hypoxic insult. Therefore, combinatorial
   miRNAs therapy may provide therapeutic benefits by simultaneously
   interfering with the many dysregulated pathways in myocardial
   infarction and heart failure. In sum, the present finding: i)
   LNA‐1a/15b synergistically enhanced cardiomyocyte proliferation in
   vitro and in vivo; and ii) LNA‐1a/15b synergistically reduced
   MI‐induced EF and fibrosis, and improved cardiac function in mice in
   vivo; and iii) LNA‐1a/15b synergistically increased cardiomyocyte
   proliferation and improved contractility in human cardiac mimetics. Our
   study thus reveals a novel mode of two‐miRNA combination that offers
   synergistic regulation on cardiac regeneration and enhanced cardiac
   function in mice following MI.

   miRNA‐1 and miRNA‐15b are expressed in a highly tissue‐restricted
   manner. miRNA‐1 and miRNA‐15b are predominantly expressed in mouse and
   human heart ventricle, with low levels detected in other tissues.^[
   [224]^60 , [225]^61 ^] In the present study, the use of LNA‐modified
   anti‐miRs in targeting miRNA‐1a and miRNA‐15b, facilitates indirect
   tissue‐targeting due to the predominant expression of miRNA‐1a and
   miRNA‐15b in heart tissue. In accord, macroscopic analysis of
   morphology and structure of non‐cardiac tissue, including the lung,
   spleen, SKM, liver, and WAT revealed negligible differences between
   LNA‐1a/15b and LNA‐NC treated mice. Taken together, the synergism
   offered through simultaneous inhibition miRNA‐1a and miRNA‐15b,
   combined with the restricted expression of miRNA‐1a and miRNA‐15b, may
   facilitate tissue‐specific precision therapy, to lower side effects and
   the possibility of neoplasm formation in other tissues. Thus,
   inhibiting miRNA‐1a and miRNA‐15b is a promising strategy to enhance
   cardiac regeneration in patients with heart failure.

   Currently, the LNA‐based anti‐miRs have been widely used to suppress
   the function of miRNAs, and have also been shown to be safe and
   effective in human studies.^[ [226]^11 , [227]^62 , [228]^63 ^] Various
   studies have convincingly shown that targeting miRNAs may improve cell
   therapy or enhance endogenous cardiac repair processes, and LNA‐based
   anti‐miRs have already been used in large animal models and a first
   clinical trial. For example, LNA‐anti‐miR‐21 treatment reduces cardiac
   fibrosis and hypotrophy and improves cardiac function in a pig model of
   Ischemia/reperfusion (I/R) injury.^[ [229]^64 ^] In addition, LNA‐based
   anti‐miR‐132 treatment improves EF and ameliorates cardiac dysfunction
   in response to MI in a porcine model.^[ [230]^62 ^] Furthermore, a
   recent clinical trial involving the delivery of a synthetic LNA‐based
   antisense oligonucleotide against miRNA‐132 (anti‐miR‐132) in heart
   failure patients ([231]NCT04045405) revealed a potential benefit of
   CDR132L in attenuating heart failure.^[ [232]^62 ^] MRG‐110, an
   LNA‐based antisense oligonucleotide targeting miRNA‐92a, has been also
   tested in healthy human adults and MRG‐110 treatment reduces miRNA‐92a
   levels and de‐represses the target genes in human peripheral blood
   cells.^[ [233]^11 ^] These studies indicate that further development of
   LNA‐based anti‐miR therapies could provide a route to developing
   efficacious therapies, of which LNA‐anti‐miR‐1a/15b would be one such
   strategy. Moreover, unlike approaches that target individual miRNAs,
   this dual inhibition addresses multiple regulatory pathways
   simultaneously, leading to a synergistic enhancement of cardiomyocyte
   proliferation and improved cardiac function post‐myocardial infarction.
   This approach not only enhances the cardiac regenerative capacity but
   also minimizes potential off‐target effects due to the heart‐specific
   expression patterns of these miRNAs. Furthermore, by using two LNAs,
   the dose of each LNA is lowered, thus reducing off‐target effects and
   improving robustness by simultaneously modulating multiple key
   pathways. This is in line with strategies used in cancer therapy, where
   combination treatments help overcome resistance, increase efficacy, and
   provide more comprehensive pathway inhibition. Given the parallels to
   combination therapies in oncology, this strategy may serve as a model
   for future regenerative medicine approaches beyond cardiology. Its
   success in myocardial infarction treatment could pave the way for
   similar interventions in other tissues where regenerative capacity is
   limited, such as the central nervous system.

   However, several limitations must be addressed before translating this
   strategy into clinical applications. Most current studies are limited
   to animal models and human cardiac organoids, and the safety and
   efficacy of miR‐1 and miR‐15b inhibition in humans remain uncertain.
   Potential off‐target effects and the long‐term consequences of miRNA
   inhibition are not fully understood. Future research should focus on
   conducting comprehensive clinical trials to evaluate the therapeutic
   potential and safety profile of this approach in humans. Furthermore,
   variability in patient responses must be considered. Differences in
   genetic backgrounds, comorbidities, and disease progression could
   influence the effectiveness of miRNA inhibition, necessitating
   personalized therapeutic strategies. Biomarker‐based patient selection
   and stratification may improve treatment outcomes and reduce the risk
   of adverse effects. Additionally, exploring the underlying molecular
   mechanisms of miR‐1a and miR‐15b in cardiac biology will be crucial for
   optimizing therapeutic strategies and minimizing adverse effects.

   Taken together, combinatorial LNA‐anti‐miR‐1a/15b treatment results in
   pro‐proliferative effects on adult cardiomyocytes and improved cardiac
   function in response to MI. Our data provides a novel and clinically
   feasible LNA‐based anti‐miR‐1a/15b strategy for the treatment of STEMI,
   and potentially heart failure through the re‐activation of adult
   cardiomyocyte mitotic re‐entry and cardiac regeneration.

3. Experimental Section

   All data produced in the present study are available upon reasonable
   request to the authors.

Animals

   C57BL/6J mice were obtained from Janvier (Le Genest Saint‐Isle,
   France). All animals used in this experiment were housed in a
   temperature‐controlled room with a 12‐h light‐dark cycle and were
   allowed free access to food and water in agreement with NIH animal care
   guidelines, §8 German animal protection law, German animal welfare
   legislation, and with the guidelines of the Society of Laboratory
   Animals (GV‐SOLAS) and the Federation of Laboratory Animal Science
   Associations (FELASA). Twelve‐week‐old C57BL/6J mice were subjected to
   MI surgery, and LNA delivery was performed 4 h after MI. Mice were
   killed and hearts isolated at the end of the experiment. All protocols
   were approved by the ethics committee of the regional council (License
   number FU1201, Regierungspräsidium Darmstadt, Hesse, Germany).

Human Heart Biopsies

   Human heart biopsies were provided by the Institute of Legal Medicine
   (Goethe University Hospital, Frankfurt am Main, Germany). The human
   heart biopsies were conducted in compliance with the local ethics
   committee (license number 116/14 from Goethe University). Samples from
   healthy hearts as control and from hearts with macroscopically visible
   signs of acute cardiac infarction as MI hearts and total miRNA were
   extracted.

Myocardial Infarction (MI)

   MI was performed in 12‐week‐old male C57BL/6J mice and was induced by
   permanent ligation of the left anterior descending coronary artery
   (LAD). The mice were monitored up to 28 days after surgery, and the
   heart dimensions and cardiac function were determined by
   echocardiography on days 0, 7, 14, 21, and 28. 4 h after MI, in vivo
   miRCURY LNA Power Inhibitors were injected intraperitoneally (i.p.)
   into mice at a dose 10 mg kg^−1. The following in vivo miRCURY LNA
   Power Inhibitors were purchased from QIAGEN: LNA‐miR‐1a‐3p (339203
   YCI0200659‐FZA), LNA‐miR‐15b‐5p (339203 YCI0200641‐FZA), and negative
   control LNA (339203 YCI0200578‐FZA).

Isolation and Maintenance of Primary Neonatal Rat Cardiomyocytes

   Mated female Sprague Dawley rats were obtained from Janvier (Le Genest
   Saint‐Isle, France). Primary neonatal rat cardiomyocytes (NRCs) were
   isolated from post‐natal day 1 (P1) and P7 rat pups by using the
   neonatal heart dissociation kit (Miltenyi Biotec) and the gentleMACS
   Dissociator according to the manufacturer's instruction. Isolated cells
   were pre‐plated in a plating medium (DMEM high glucose, M199 EBS
   (BioConcept), 10% horse serum, 5% fetal calf serum, 2% L‐glutamine and
   1% penicillin/streptomycin) for 75 min in 10 cm cell culture dishes
   (Nunc) at 37 °C and 5% CO[2] in a humidified atmosphere to get rid of
   fibroblasts as described previously.^[ [234]^32 ^] NRCs were plated on
   Type‐I bovine Collagen‐coated (Advanced Biomatrix) plates or dishes in
   the plating medium. 24 h after isolation of NRCs, the plating medium
   was changed to a maintenance medium (DMEM high glucose, M199 EBS, 1%
   horse serum, 2% L‐glutamine, and 1% penicillin/streptomycin).

Isolation and Maintenance of Primary Rat Fibroblasts

   Primary rat fibroblasts were isolated from P1 rat hearts. After
   pre‐plating the cells in 10 cm cell culture dishes for 75 min, the
   suspension cells were harvested as cardiomyocytes, the attached cells
   will be digested and harvested as fibroblasts. The fibroblasts were
   cultured in DMEM medium with 10% FBS, 2% L‐glutamine, and 1%
   penicillin/streptomycin at 37 °C and 5% CO[2] in a humidified
   atmosphere.

Preparation and Maintenance of Human iPSC‐CMs

   Human induced pluripotent stem cells (hiPSCs) were purchased from
   Cellular Dynamics International (CMC‐100‐010‐001) and cultured as
   recommended by the manufacturer. The human iPSC‐derived cardiomyocytes
   (hiPSC‐CMs) were reprogrammed using the STEMdiff Cardiomyocyte
   Differentiation Kit (STEMCELL Technologies) according to the
   manufacture's protocol. Briefly, human iPS cells were plated at a cell
   density of 3.5 × 10^5 cells/well on Matrigel‐coated 12‐well‐plates
   using mTeSR medium supplemented with 5 µM ROCK inhibitor (Y‐27632,
   STEMCELL Technologies) for 24 h. After 1 day (‐1), the medium was
   replaced with fresh TeSR medium. To induce cardiac differentiation, the
   TeSR medium was replaced with Medium A (STEMdiff Cardiomyocyte
   Differentiation Basal Medium containing Supplement A) at day 0, Medium
   B (STEMdiff Cardiomyocyte Differentiation Basal Medium containing
   Supplement B) at day 2, Medium C (STEMdiff Cardiomyocyte
   Differentiation Basal Medium containing Supplement C) at day 4 and day
   6. On day 8, the medium was switched to STEMdiff Cardiomyocyte
   Maintenance Medium with full medium changes every 2 days, to promote
   further differentiation into mature cardiomyocyte cells. All
   experiments were performed in the hiPSC‐CMs at day 40. Hypoxic
   condition was achieved by using the Hypoxia chamber and the hiPSC‐CMs
   were cultured at 1% O[2] for 2 days.

Human Cardiac Organoids Formation Technique

   Human cardiac organoids were created by hiPSC‐CMs. Aggrewell 800
   microwell culture plates were used to create the human cardiac
   organoids in STEMdiff Cardiomyocyte Support Medium (STEMCELL
   Technologies). Approximately 900 000 wells in 2 mL cell suspension were
   pipetted into each well. After 2 days of culture, the medium was
   switched to STEMdiff Cardiomyocyte Maintenance Medium for long‐term
   culture.

EdU Labeling In Vitro

   The Click‐iT EdU Cell Proliferation Kit for Imaging, Alexa Fluor 488
   dye ([235]C10337, Invitrogen, CA, USA), Alexa Fluor 647 dye
   ([236]C10640, Invitrogen, CA, USA), and Click‐iT Plus EdU Cell
   Proliferation Kit for Imaging were used in the study according to the
   manufacturer's protocol. The NRCs were labeled by 2 µM EdU for 2 days,
   the rat fibroblasts were labeled by 2 µM EdU for 4 h, and the hiPSC‐CMs
   were labeled by 2 µM EdU for 6 h.

Contractility Measurement

   Every single human cardiac mimetic was transferred into 96‐well‐plate
   and treated with either LNA‐1/15b or LNA‐NC. Hypoxic condition was
   achieved by using the Hypoxia chamber and cardiac organoids were
   cultured at 1% O[2] for 3 days, then the contractility will be analyzed
   by IonOptix system. Units/pixels were determined by calibrating the
   system with a micrometer.

Calcium Transient Measurements

   Human cardiac organoids were cultured in a 96‐well plate and incubated
   with 2 µM Cal‐520 AM (AAT bioquest) with 0.04% Pluronic F‐127 (AAT
   bioquest) as previously described.^[ [237]^33 ^] After incubation in
   the incubator at 37 °C for 90 min, the cardiac organoids were rinsed
   with PBS and medium, then the cardiac organoids were transferred to the
   384 well U‐bottom plate and incubated at 37 °C for 1 h. Relative
   fluorescence units (RFU) were measured using a fluorescence plate
   reader.

In Vitro Administration of Locked Nucleic Acids (LNA)

   miRCURY LNA Power Inhibitors were directly added to the cell culture
   medium at a final concentration of 50 nM. All miRCURY LNA Power
   Inhibitors and negative control LNA were purchased from QIAGEN:
   LNA‐miR‐1a‐3p (1999990‐801), LNA‐miR‐1‐3P (YI04100840‐AFA),
   LNA‐miR‐15a‐5p (339146 YCI0200640‐FDA), LNA‐miR‐15b‐5p (339146
   YCI0200641‐FDA), LNA‐miR‐16 (339146 YCI0200647‐FDA), LNA‐miR‐27b‐5p
   (339146 YCI0200648‐FDA), LNA‐miR‐29a‐3p (339146 YCI0200644‐FDA),
   LNA‐miR‐34a‐5p (339146 (YCI0200649‐FDA), LNA‐miR‐34b‐5p (339146
   YCI0200645‐FDA), LNA‐miR‐34c‐5p (339146 YCI0200650‐FDA), LNA‐miR‐92a‐3p
   (339146 YCI0200646‐FDA), LNA‐miR‐132‐3p (1999990‐801), LNA‐miR‐133a‐3p
   (1999990‐801), LNA‐miR‐155‐3p (1999990‐801), LNA‐miR‐195a‐5p (339146
   YCI0200642‐FDA), LNA‐miR‐208a‐3p (1999990‐801), LNA‐miR‐497a‐5p (339146
   YCI0200643‐FDA), LNA‐miR‐873a‐5p (1999990‐801), and negative control
   LNA CEL‐CONTROL‐INH (339147 YCI0199066‐FFA).

RNA Isolation, Reverse Transcription, and qRT‐PCR

   Samples were harvested in QIAzol Lysis Reagent (QIAGEN), total miRNAs
   and total mRNAs were isolated with miRNA mini Kit (QIAGEN) according to
   the manufacturer's protocol. To assess mature miRNA levels, 10 ng total
   miRNAs were reverse transcript into cDNA using Taqman Advanced miRNA
   cDNA synthesis kit (Thermo Fisher Scientific) following the
   manufacturer's instructions. The Applied Biosystems StepOnePlus
   Real‐Time PCR system (Applied Biosystems, CA, USA) with TaqMan fast
   Advanced PCR Master Mix (Thermo Fisher Scientific) were used for
   analysis. TaqManTM Advanced miRNA Assays for mmu‐miR‐1a‐3p (ID:
   mmu482914_mir), has‐miR‐1‐3p (ID: mmu482914_mir), rno‐miR‐15a‐5p (ID:
   rno483022_mir), mmu‐miR‐15b‐5p (ID: mmu482957_mir), rno‐miR‐16‐5p (ID:
   rno481312_mir), rno‐miR‐27b‐5p (ID: rno481016_mir), hsa‐miR‐27b‐sp (ID:
   478 789_mir), hsa‐miR‐29a‐3p (ID: 478 587_mir), hsa‐miR‐34a‐5p (ID:
   rno481304_mir), miR‐34b‐5p (ID: 0 02617), hsa‐miR_34c‐5p (ID:
   478 052_mir), miR‐92a‐3p (ID: 000430), rno‐miR‐132‐3p (ID:
   rno480919_mir), rno‐miR‐133a‐3p (ID: rno481491_mir), mmu‐miR‐155‐3p
   (ID: mmu481328_mir), mmu‐miR‐195a‐5p (ID: mmu482953_mir), miR‐208a‐3p
   (ID: 463567‐mat), mmu‐miR‐497a‐5p (ID: mmu482607_mir), miR‐873a‐5p (ID:
   rno481425_mir), hsa‐miR‐26a‐5p (ID: 477 995_mir), and hsa‐26a‐5p (ID:
   000405).

   Total RNAs (200 ng) were reverse transcript into cDNA using QuantiTect
   Reverse Transcription Kit (QIAGEN) according to the manufacturer's
   protocol. The Applied Biosystems StepOnePlus Real‐Time PCR system
   (Applied Biosystems, CA, USA) with Fast SYBR Green Master Mix (Thermo
   Fisher Scientific) were used for analysis. Gene expression levels were
   normalized against the housekeeping gene Hprt1. The following qRT‐PCR
   primers were used: Mouse: Nppa: 5′ GGAGGTCAACCCACCTCTGA 3′ and 5′
   CGAAGCAGCTGGATCTTCGT 3′; Myh6: 5′ CGGTGACAGTGGTAAAGGCA 3′ and 5′
   GGAATGATGCAGCGCACAAA 3′; Myh7: 5′ AGAAGAAGGTGCGCATGGAC 3′ and 5′
   ATCCTGGCATTGAGTGCATTT 3′; Tgf‐β2: 5′TTGTTCCACAGGGGTTAAGG 3′ and 5′
   AGCTCGGTCCTTCA GATCCT 3′; Col3a1: 5′ CTGTAACATGGAAACTGGGGAAA 3′ and 5′
   CCATAGCTGAACTGAAAACCACC 3′; Thbs1: 5′ CACCTCTCCGGGTTACTG AG 3′ and 5′
   GCAACAGGAACAGGACACCTA 3′; Hprt: 5′ CCACTTGTGACGA AAGCACC 3′ and 5′
   GTTGTCTACGCTCTGGCAGT 3′.

Immunofluorescence staining

   Immunofluorescence staining was performed as described previously.^[
   [238]^65 ^] After fixation of cells with 4% paraformaldehyde (PFA)/PBS,
   the cells were permeabilized and incubated overnight at 4 °C with
   primary antibodies against sarcomeric α‐actinin (Sigma Aldrich), Aurora
   B (Abcam), α‐SMA (Abcam), Ki67 (Abcam), pHH3 (Abcam), cleaved (Cl.)
   Caspase 3 (Abcam) or Tomato lectin (TL) (Sigma Aldrich) diluted in 2%
   HS/PBS (v/v). After 3 washes with PBS for 5 min, cells were incubated
   with 4′,6‐diamidino‐2‐phenylindole (DAPI; Thermo Fisher Scientific),
   AlexaFluor 555 anti‐mouse (Thermo Fisher Scientific) and AlexaFluor 488
   anti‐rabbit (Thermo Fisher Scientific) secondary antibody for 1 h at
   room temperature. Dishes were mounted onto glass slides (Fisher
   Scientific) with a drop of ProLong Antifade (Thermo Fisher Scientific).
   Fluorescent images were acquired with the SP5 confocal microscopy
   (Leica) using a 40x magnification.

   Following the α‐SMA, α‐actinin, Wheat–Germ–Agglutinin (WGA), the cells
   or sections were stained with EdU. EdU staining was conducted using
   Click‐iT EdU imaging kit (Invitrogen, CA), according to the
   manufacturer's protocol. Briefly, after incubation with a secondary
   antibody for 1 h, cells were washed with 3XPBS for 5 min. The cells
   were then incubated with a Click‐iT reaction cocktail for 30 min for
   EdU staining, then the cells were washed with 3XPBS for 5min. Finally,
   fluorescent images were acquired with the SP8 confocal microscopy
   (Leica) using a 40x magnification.

ASAT and ALAT Measurement

   The blood was collected in tubes prefilled with EDTA for determination
   of Aspartate aminotransferase (ASAT) and Alanine aminotransferase
   (ALAT) in sera from animals in different groups. Sera samples were
   measured in a Cobas 8000 Analyzer (Roche Diagnostics, Germany) using
   Roche reagents following the manufacturer's instructions.

Picro Sirius Red Staining

   Picro Sirius red staining was used to determine collagen deposition and
   fibrosis in heart cryosections. 0.1% Picrosirius Red solution was
   prepared by solving 0.5 g Sirius Red (Waldeck GmbH) in 500 mL picric
   acid (Sigma‐Aldrich). The cryosections were washed with water and PBS,
   and then fixed with 4% paraformaldehyde (PFA) for 30min. Then the
   cryosections were incubated in the 0.1% Picrosirius Red solution for
   1 h. After washing two times with acidified water, the sections were
   dehydrated with 100% ethanol, cleared with xylene, and mounted with
   Pertex (Medite) as previously described.^[ [239]^32 ^]

Metabolomics Sample Preparation

   P1 rat cardiomyocytes (5 × 10^5) were cultured per 3 cm dish. 24 h
   after isolation miRCURY LNA Power Inhibitors were added to the
   maintenance medium at a final concentration of 50 nM. The cell culture
   plates were snap frozen in liquid nitrogen 48 h after miRCURY LNA Power
   Inhibitors incubation, right after the cells were washed with 75 mM
   ammonium carbonate (Sigma), adjusted to pH 7.4 with acetic acid. In
   addition, 20 mg of organ tissues (liver, kidneys, lungs, spleen, and
   kidneys) were homogenized at full speed. Metabolite extraction from NRC
   and homogenized tissues was done twice with cold extraction buffer (−20
   °C) containing acetonitrile: methanol: water in a 40:40:20 ratio. After
   centrifugation at full speed, sample extract supernatants were dried by
   vacuum‐assisted centrifugation.

Non‐Targeted Metabolomics

   Untargeted metabolomics was conducted through metabolite profiling
   utilizing liquid chromatography high‐resolution tandem mass
   spectrometry (LC‐HR‐MS/MS). The experimental setup comprised an
   ultra‐performance liquid chromatography system (Aquity I‐class; Waters
   GmbH, Eschborn, Germany) coupled with a quadrupole – time‐of‐flight
   mass spectrometer (QToF) equipped with an additional ion mobility
   spectrometer (IMS) (VION IMS QToF; Waters GmbH, Eschborn, Germany).
   Sample extract residues were reconstituted in 200 µL aqueous
   acetonitrile (10% water) containing 0.1% formic acid and transferred to
   auto‐sampler glass vials for analysis. A pooled sample, consisting of
   20 µL from each individual sample, was prepared separately and analyzed
   randomly throughout the experiment. Chromatographic separation utilized
   hydrophilic interaction chromatography (HILIC) with a BEH Amide column
   (2.1 × 100 mm, 1.7 µm; Waters GmbH, Eschborn, Germany) at 45 °C,
   employing a gradient of mobile phase A (water with 0.1% formic acid)
   and mobile phase B (acetonitrile with 0.1% formic acid). The mass range
   covered mass‐to‐charge ratios between 50 and 1000 Da, with a total scan
   time set to 0.3 s. Positive and negative electrospray ionization (ESI)
   in combination with high‐definition data acquisition scan mode (HDMSE)
   were employed, including ion mobility screening, determination of
   metabolite‐specific collision cross‐section (CCS) values, accurate
   mass, and fragment ion mass screenings.

   Ion source parameters in positive ESI mode were set to 1 kV and 40 V
   for capillary voltage and sample cone voltage, and 120 °C and 550 °C
   for source and desolvation temperature. In negative ESI mode,
   parameters were adjusted to 1.5 kV and 40 V, and 120 and 450 °C,
   respectively. Nitrogen was used as cone gas and desolvation gas with
   flow rates of 50 and 1000 L h^−1 in positive and negative ESI.
   Leucine–Enkephalin solution injections were performed for mass
   correction. Data acquisition and raw data processing were carried out
   using the Waters Unifi software package 2.1.2. Further data processing
   involved importing Unifi export files into Progenesis QI software
   (Nonlinear Dynamics, Newcastle upon Tyne, UK) for peak picking,
   chromatographic alignment, signal deconvolution, data normalization,
   and experimental design setup.

   For metabolite identification, features obtained from untargeted
   metabolomics screenings were compared with the Human Metabolome
   Database (HMDB, [240]www.hmdb.ca) using the Progenesis Metascope
   plugin, applying a precursor mass accuracy of ≤ 10 ppm and a
   theoretical fragment mass accuracy of ≤ 10 ppm. Additionally, features
   were compared to in‐house data obtained by injecting metabolite
   standards in neat solution, allowing deviations of retention time and
   CCS of 0.3 min and 3%, respectively. Metabolic features without a
   metabolite annotation were excluded, and the remaining features, along
   with their signal intensities and identifications, were exported to
   Microsoft Excel for further data processing.

Bioinformatic Analysis

   The P1 rat cardiomyocytes were treated with either control LNA‐NC or
   LNA‐1a/15b in hypoxia with 1% O[2], and then RNAs were extracted for
   next‐generation RNA sequencing (RNA‐seq) (Novogene). The integrity and
   purity of the RNA were assessed, and qualified RNA samples were
   utilized for PCR amplification in order to construct a cDNA library.
   The cDNA library was sequenced using an Illumina HiSeq 150 platform.
   For RNA sequencing analysis, raw RNA‐Seq data (fastq files) were first
   cleaned using Trimmomatic v0.39 (Trimmomatic: a flexible trimmer for
   Illumina sequence data). The cleaned reads were then accurately mapped
   to the Ensembl rat genome (mRatBN7.2) using HISAT 2.2.1, and sequence
   comparison/mapping format (sam) files were generated.^[ [241]^66 ^] The
   Sam files were then sorted by Samtools v1.19.2 to generate binary
   comparison/mapping format (bam) files, which HTSeq‐count v2.0.3 uses to
   summarize gene level counts (HTSeq‐a Python framework for working with
   high‐throughput sequencing data). DESeq2^[ [242]^67 ^] was used to
   detect differentially expressed genes (DEGs), and the DEGs with the
   cut‐off criterion (P value < 0.05 and Log|FC|>1) were considered for
   further analysis. The BioSample Database (BioProject ID: PRJNA000000)
   was selected for the data analysis.

Statistical Analysis

   Data are represented as mean and error bars indicate the standard error
   of the mean (SEM). Two‐tailed unpaired Student's t‐test (Excel) or
   Mann–Whitney test was used as indicated in the respective figure
   legends. P values were determined with Prism 9.0 (GraphPad) and P
   < 0.05 was considered statistically significant.

Conflict of Interest

   T.Y., S.D., and J.K. are inventors on a patent application pertaining
   to the inhibition of miR‐1a and miR‐15b for the treatment of heart
   disease. The other authors declare that they have no conflict of
   interest.

Author Contributions

   T.Y. and M.W. equally contributed to this work. T.Y., C.Z., P.M.D.,
   K.B., Y.M., Y.W., A.P.D., and P.M. conducted experiments and analyzed
   results. H.Y. performed the bioinformatics analyses. S.K. provided the
   human heart biopsies. Experiments were designed by A.Z., S.D., and J.K.
   T.Y. and J.K. wrote the manuscript with support from J.W.

Supporting information

   Supporting Information
   [243]ADVS-12-2414455-s001.pdf^ (3MB, pdf)

   Supplemental Movie 1
   [244]Download video file^ (530.5KB, avi)

Acknowledgements