CRISPR/Cas-based activation of PPARGC1A boosts endogenous mitochondria and enhances cardiac function and recovery after myocardial infarction

Abstract

Insufficient energy supply, due to impaired mitochondria, has emerged as a key pathological factor in the development of heart failure (HF) after myocardial infarction (MI). Unfortunately, no current therapeutic strategies directly augment myocardial energy production. While mitochondrial biogenesis is orchestrated by the activity of multiple genes, activation of PPARGC1A, a key regulator, can increase cellular mitochondria, however, supraphysiological levels of PPARGC1A result in adverse tissue remodeling and heart disfunction. CRISPR activation (CRISPRa) technologies present a unique opportunity to address these shortcomings, as they enable tunable control over endogenous target gene expression. Here, we demonstrate that transcriptional activation of PPARGC1A using CRISPRa increases cellular mitochondria in human cell types. This effect is mediated through activation of transcriptional programs driving mitochondrial biogenesis, mitochondrial function, and cellular bioenergetics. These activated transcriptional programs synergize to increase ATP production and reserve capacity in human cardiomyocytes. CRISPRa targeting of PPARGC1A in vivo increases cardiac mitochondria to recover heart ejection fraction in an acute-MI model. Furthermore, CRISPRa acts on adult human heart to increase PPARGC1A protein and cellular mitochondria, elevating mitochondrial function in both normal and HF-diagnosed hearts. Altogether, these results provide first proof-of-concept that endogenous gene activation via CRISPRa can improve heart function after MI.

Graphical Abstract

This study demonstrates that activation of PPARGC1A with a CRISPRa system drives transcriptional programs to increase mitochondrial quantity and ATP production, improving heart function after myocardial infarction. These findings highlight CRISPRa as a promising therapeutic strategy to boost cardiac energy production and improve heart function after injury.

INTRODUCTION

Ischemic heart disease is the leading cause of death in the industrialized world, with nearly one third of patients developing heart failure (HF) after myocardial infarction (MI).^1,2 An estimated 6.8 million Americans develop HF with a projected increase of 41% by 2050.^3,4 A major component in the development of HF is insufficient energy supply, which manifests as reduced mitochondrial capacity and altered cellular metabolism.^5–12 Due to the major role that mitochondria play in cellular bioenergetics, these organelles have been recognized as a valuable therapeutic target in HF.^6,7,13–15 While increasing the quantity and function of mitochondria has been observed to elicit positive therapeutic effects in the setting of acute-MI and HF,^16–26 current strategies to increase cellular mitochondria are technically challenging, labor intensive, and difficult to scale clinically.^27–30

Endogenous generation of mitochondria is primarily regulated by the key upstream master transcription factor PPARGC1A (also called PGC1α).^31–34 While forced overexpression of PPARGC1A transgenes can increase cellular mitochondria, these methodologies cause uncontrolled PPARGC1A transcription, which results in abnormal physiological responses, such as arrhythmia and dilated cardiomyopathy.^35–39 In contrast, CRISPR activation (CRISPRa) is an unexplored approach that permits tunable control over endogenous gene expression.^40–51 For example, CRISPRa systems consisting of protein fusions between nuclease-inactive CRISPR/Cas proteins and transcriptional activation domains can specifically bind to genomic sequences, such as enhancers or promoters, via specific guide RNAs (gRNAs) and robustly activate endogenous target gene expression. Furthermore, the small size of these systems allows for seamless incorporation into clinically applicable routes of delivery, namely, adeno-associated virus (AAV)^52,53 or lipid nanoparticle mRNA.^54,55 This transcriptional control methodology and straightforward clinical integration has led to promising therapeutic breakthroughs in diseases with direct target deficiency, such as Dravet syndrome,^52,53 inherited blindness,⁵⁶ and muscular dystrophy.⁵⁷ However, these CRISPRa tools have not yet been applied in more complex pathological processes such as mitochondrial biogenesis and enhancement of cellular energetics.

In this study, we evaluated CRISPRa-based induction of PPARGC1A as a strategy for transcriptional control of mitochondrial biogenesis and examined the mitochondrial and cardiac function enhancements caused by increased endogenous PPARGC1A. We identified a robust CRISPRa system for modulation of PPARGC1A expression across various human cell types. Transcriptomic analysis demonstrated that CRISPRa-mediated increase in PPARGC1A expression drives programs enhancing cellular bioenergetics and mitochondrial function. Using respirometry analysis, we also observed a coincident improvement in mitochondrial function in human cardiomyocytes. Using an in vivo acute-MI model also revealed that CRISPRa-based stimulation of PPARGC1A led to increased cardiac mitochondrial content and improved heart function after MI. Finally, we used ex vivo adult human heart donor tissue to establish that CRISPRa-based regulation of PPARGC1A increases mitochondrial quantity and activity in both normal and diseased human hearts. Overall, our findings demonstrate that CRISPRa can robustly enhance mitochondrial quantity and function, leading to significant functional improvements in vivo, in vitro, and ex vivo in human samples, positioning CRISPRa regulation as a potential therapeutic strategy for complex diseases such as HF after MI.

RESULTS

CRISPR/Cas-based activation of PPARGC1A robustly increases mitochondria in human cells

We began by identifying an optimal CRISPR activation (CRISPRa) platform by iteratively combining the established VP64/p65/Rta tripartite activation domain (VPR) or a combination of our reported eNRF2/MRTF-A/STAT1 tripartite activation domain with a VP64 activation domain (NMS-VP64) in 3 CRISPR/dCas systems to evaluate their transcriptional modulation.^42,45 We prioritized compact deactivated Cas (dCas) proteins, Staphylococcus Aureus dCas9 (SadCas9), Acidaminococcus sp. dCas12a (AsdCas12a), and Enhanced AsdCas12a (EnAsdCas12a), to facilitate delivery and pre-clinical utility.^43,58–60 Our recently developed NMS-VP64 activation system⁴² resulted in more consistent and higher activation than the VPR activation domain⁴⁵ when combined with SadCas9 (Fig. S1). Similar to other,^42,61,62 certain gRNAs in our CRISPRa screen resulted in transcriptional repression, rather than activation (Fig. S1). While standard, this paradoxical effect has been attributed to steric hindrance, where the CRISPRa fusion protein itself blocks the assembly of native transcriptional machinery. These findings underscore the critical importance of empirically validating each guide RNA and each CRISPRa system to ensure the desired activating effect. Consequently, we corroborated the transcriptional activation capability of NMS-SadCas9-VP64 at other mitochondrial biogenesis-related genes (PPARGC1B, TFAM, NRF1; Fig. S2). These data enabled us to nominate NMS-SadCas9-VP64 as the most robust compact CRISPRa system and PPARGC1A as an optimal target for engineered mitochondrial biogenesis in human cells (Fig. 1A).

Figure 1. CRISPR/Cas-based activation of PPARGC1A induces mitochondrial biogenesis in human cells.

(A) The indicated CRISPR activation (CRISPRa) payload is delivered into cells where it binds via a guide RNA (gRNA) upstream of PPARGC1A to activate gene expression. The PPARGC1A master regulator will in turn activate mitochondrial biogenesis pathways resulting in the increase of cellular mitochondria. (B) Relative PPARGC1A expression (compared to WT) 72 hours after payload delivery in HEK293T (left), HDFa (Adult Human Dermal Fibroblast; middle), and AC16 (right) cells. (C) Mitochondrial DNA (mtDNA) quantification 72 hours after payload delivery in HEK293T (left), HDFa (middle), and AC16 (right) cells. (D) Immunofluorescence microscopy of HEK293T (left), HDFa (middle), and AC16 cells 72 hours post-delivery of CRISPRa targeting PPARGC1A compared to WT (white scale bar 25 μm). Cells were stained with mitochondrial dye TMRE (purple) and Hoechst nuclear counterstain (blue). Cells harboring the CRISPRa system co-expressed GFP (green) as a transfection control. Data presented in panels A, B, and C as mean +/− s.e.m. from 3 independent experiments. Data presented in panel D is representative of 2 independent experiments. *P < 0.05, **P < 0.01, and ****P < 0.0001 (Student’s t-test).

To further characterize the robustness of PPARGC1A activation via CRISPRa in different human cell types, we next delivered a synthetic payload expressing both a gRNA targeting PPARGC1A and our optimized CRISPRa system to HEK293T cells, adult human dermal fibroblasts (HDFa), and an immortalized human cardiomyocyte cell line (AC16) and measured the target gene expression (Fig. 1B). Similar to previous reports,⁶¹ CRISPRa-induced activation levels varied between cell types, with HEK293T having the lowest activation (~3-fold), followed by HDFa (~20-fold), and AC16 cells (~1,000-fold). We next measured changes in number of mitochondria by quantifying copies of mitochondrial DNA (mtDNA).^63,64 We found that activation of PPARGC1A using CRISPRa led to significant increases in the numbers of mitochondria at relatively similar levels across each tested cell type (HEK293T, ~80% increase; HDFa, ~70% increase; AC16, ~120% increase; Fig. 1C), despite the wider ranges in levels of increased PPARGC1A transcripts (Fig. 1B). We further explored the relationship between PPARGC1A transcript levels and numbers of mitochondria by delivering a cDNA expressing PPARGC1A to these cell types and observed that although cDNA increased the total PPARGC1A transcript levels higher than CRISPRa, this did not result in any increase in quantified mitochondria relative to CRISPRa (Fig. S3).

We validated our mitochondrial quantification using mtDNA by staining active mitochondria using TMRE. Cells expressing the CRISPRa payload additionally express GFP, as a delivery control. Transfected (GFP positive) cells had an increase in mitochondrial staining compared to WT. Even when compared to untransfected cells (GFP negative), transfected cells had stronger mitochondrial staining, indicating higher mitochondria mass and stronger mitochondrial function (Fig. 1D).^65–67 Altogether, our results indicate that CRISPRa is able to robustly activate PPARGC1A transcription, which in turn increases cellular mitochondria in various human cells.

CRISPRa-mediated induction of PPARGC1A provokes transcriptional programs linked to mitochondrial function and cellular bioenergetics.

In order to better understand how CRISPRa-mediated activation of PPARGC1A drives mitochondrial biogenesis, we quantified the transcriptome of immortalized cardiomyocytes (AC16 cells) following delivery of our CRISPRa system targeting PPARGC1A. As above (Figs. 1B and 2A), our CRISPRa system increased PPARGC1A transcription nearly ~1,000-fold (Fig. 2A). Further, our CRISPRa upregulation of the PPARGC1A master transcription factor also led to differential expression of ~4,000 genes (Fig. 2A). From these differentially expressed genes, we identified ~300 genes strongly associated with mitochondria function using the mitocarta 3.0 database.⁶⁸ We found that CRISPRa-mediated induction of PPARGC1A resulted in increased activation in genes across multiple mitochondrial pathways, including the mitochondrial central dogma (e.g. mtDNA maintenance, mtDNA transcription, mitochondrial translation), mitochondrial metabolism (e.g. lipid metabolism, amino acid metabolism, vitamin metabolism), and oxidative phosphorylation (OXPHOS; Fig. 2B). We also performed KEGG enrichment analysis of all upregulated genes and identified gene enrichment of OXPHOS and ROS pathways, as well as disease pathways that are known to be directly impacted by mitochondrial activities, such as prion disease, Alzheimer’s, Parkinson’s, Huntington’s, and non-alcoholic fatty liver (NAFL) diseases (Fig. 2C).⁶⁹ We further extrapolated changes in cellular function through network enrichment analysis and detected enrichment in bioenergetics (i.e. ATP synthesis, TCA cycle, respiration, metabolism), mitochondrial activities (i.e. cristae formation, transport of small molecules), and other cellular processes (Fig. 2D). Interestingly, KEGG and pathways analysis of the downregulated genes identified repression of pathways associated to cell cycle and DNA replication following CRISPRa treatment (Fig S4). These results are highly in line with the expected profile of post-natal cardiomyocytes. We and others have previously demonstrated that adult non-dividing cardiomyocytes, and not replicative pre-natal cardiomyocytes, rely on oxidative phosphorylation,^70–72 suggesting our treatment further strengthens the established adult cardiomyocyte cellular profile. Altogether, these results demonstrate that CRISPRa-driven expression of PPARGC1A results in the subsequent activation of transcriptional programs linked to increased mitochondrial and bioenergetic function.

Figure 2. CRISPRa-induced activation of PPARGC1A drives transcriptional programs linked to mitochondrial function and cellular bioenergetics.

(A) Differentially expressed genes in AC16 cells 72 hours post-delivery of CRISPRa targeting PPARGC1A compared to WT AC16 cells (blue box, downregulated genes; red box, upregulated genes; yellow dots, mitocarta 3.0 genes). Data were analyzed using the Wald test and the adjusted P value (P_adj) was calculated using the Benjamini and Hochberg method. (B) Mitochondria-associated pathways with differentially upregulated genes 72 hours after delivery of CRISPRa targeting PPARGC1A to AC16 cells. (C-D) Top 10 KEGG pathways (panel C) and network analysis (panel D) of differentially upregulated genes 72 hours after delivery of CRISPRa targeting PPARGC1A to AC16 cells. Data presented is representative of 2 independent experiments.

CRISPRa-based stimulation of PPARGC1A improves mitochondrial function and ATP production in human cells.

To further explore the functional changes resulting from CRISPRa-based stimulation of PPARGC1A, we quantified changes in mitochondrial function using respirometry (Fig. 3A). CRISPRa led to an overall increase in oxygen consumption rate (OCR) in comparison to WT AC16 cells (Fig. 3B), indicating an increase in mitochondrial activity. Additionally, we detected an increase in initial extracellular acidification rate (ECAR), which suggests an increase in glycolysis as a result of PPARGC1A activation via CRISPRa (Fig. 3C). Using various electron transport chain inhibitors, the activity of different mitochondrial functions can also be probed using OCR measurements. Basal respiration, the mitochondrial oxygen consumption necessary to meet normal ATP demand, was increased (~2-fold) after CRISPRa delivery. This increase in basal respiration was caused by an increase in oxygen consumption from both ATP production (~90% increase) and proton leakage (~3-fold). While the relative increase in proton leakage was higher than the ATP production increase, the percentage of basal respiration dedicated to proton leakage was more similar between CRISPRa (~38% of basal respiration) and WT (~30% of basal respiration).

Figure 3. CRISPRa-induced mitochondria provide enhanced bioenergetic capacity to human cells.

(A) Experimental workflow used to quantify mitochondrial function following CRISPRa-based activation of PPARGC1A. (B) Oxygen consumption rate (OCR) kinetic graph with compound delivery to AC16 cells 72 hours after delivery of CRISPRa targeting PPARGC1A or WT. (C) Average extracellular acidification rate (ECAR) throughout OCR measurement time. (D) Mitochondrial functions (basal respiration, maximal respiration, proton leak, ATP production, and reserve capacity), and non-mitochondrial oxygen consumption measured and calculated from panel B. Data presented as mean +/− s.e.m. from 3 independent experiments. **P < 0.01, and ***P < 0.001 (Student’s t-test).

Interestingly, the maximal respiration OCR, in which the maximal possible energy demand of the cell is evaluated, increased by ~3-fold compared to WT cells (Fig. 3D). Although this increase is partially mediated by the increase in basal respiration, there was also an increase in the reserve capacity (~4-fold), collectively resulting in an increased capacity of the cell to respond to metabolic needs. In line with these results, there was also an increase in non-mitochondrial oxygen consumption (~2.5-fold; Fig. 3D). This increase may be attributable to the additional non-mitochondrial pathways activated by PPARGC1A detected during our transcriptomic analysis (Fig. 2A). Nevertheless, CRISPRa improved key mitochondrial functions, in particular ATP production and reserve capacity, which are critical elements in cell homeostasis and survival. Altogether, these results demonstrate that CRISPRa-driven activation of PPARGC1A results in enhanced mitochondrial function.

Ppargc1a activation increases mitochondria quantity and heart function in rats after myocardial infarction.

To evaluate the therapeutic potential of CRISPRa-mediated activation of PPARGC1A, we synthesized lentivirus payloads expressing the CRISPRa protein (NMS-SadCas9-VP64) and either a gRNA targeting Ppargc1a (i.e. CRISPRa) or a non-targeting gRNA (i.e. Scramble; Fig. 4A). We validated lentiviral delivery, PPARGC1A expression, and mitochondria increases by delivering these payloads to healthy rat hindlegs (Fig. S5). We continued to further characterize therapeutically beneficial effects of CRISPRa upregulation of Ppargc1a using a rat myocardial infarction (MI) model. In this model myocardial infarction (MI) is induced through permanent ligation of the left anterior descending artery, which we and others have shown produces a pathological state resembling heart failure with reduced ejection fraction (HFrEF).^73,74 Cardiac function was measured utilizing echocardiographic measurements of left ventricle ejection fraction (EF) before the myocardial infarction (i.e. Pre-MI), 10 days after the infarction (i.e. Post-MI), and 21 days after the infarction (i.e. Treatment). On day 21, hearts were collected and processed for immunostaining (Fig. 4B). CRISPRa resulted in an increase of both PPARGC1A protein levels and the mitochondrial marker TOMM20 in comparison to Scramble control (Fig. 4C). Specifically, we detected a 55% increase in PPARGC1A protein leading to the downstream 84% increase in mitochondria, measured through TOMM20 (Fig. 4D). In line with the expected CRISPRa-driven mitochondrial biogenesis mechanism of action, treatment did not result in noticeable differences in cardiac fibrosis or inflammation (Fig. S6), reinforcing that the observed functional benefits are driven by enhanced cardiac mitochondria in the surviving cardiomyocytes, rather than by reduction on the scar tissue or modulation of the inflammatory response. Nonetheless, these results mirror the increases in mitochondria measured in vitro, highlighting the broad applicability of using this CRISPRa system to stimulate Ppargc1a for enhanced mitochondrial biogenesis and cellular functions.

Figure 4. CRISPRa-based activation of Ppargc1a increases mitochondria in vivo and improves cardiac function in an acute-MI disease model.

(A) Components from injected lentivirus are schematically depicted with their associated condition and transcriptional effects. (B) Experimental workflow used to induce and treat acute-MI model. (C-D) Immunofluorescence microscopy (panel C) and quantification (panel D) of CRISPRa-based treatment in rat heart (white scale bar 50 μm). Cells were probed with antibodies against the FLAG epitope on CRISPRa molecules (green), PPARGC1A target protein (red), mitochondrial marker TOMM20 (purple), and nuclear counterstain DAPI (blue). (E) Left ventricle ejection fraction before CRISPRa injection and surgery (Pre-MI), 10 days following CRISPRa injection and LAD ligation (Post-MI), and 21 days after CRISPRa injection and surgery (Treatment). Data presented in panel C and D represent 3 independent experiments. Data in panel E presented as mean +/− s.e.m. from 6 independent experiments (excluding 1 premature mortality of subject in the CRISPRa group). In panel D, ***P < 0.001 (Student’s t-test). In panel E, **q < 0.01 (Student’s t-test with FDR correction).

Our assessment of cardiac function demonstrated that CRISPRa significantly increased cardiac function over Scramble control. At 10 days post-MI, EF was similar between groups, but at 21 days post-treatment, CRISPRa treatment led to a 39% increase in EF compared to Scramble control (Fig. 4E). Altogether, these results demonstrate that CRISPRa-induced upregulation of Ppargc1a increases PPARGC1A protein levels and in turn increased mitochondria and cardiac function in a therapeutically relevant HF disease model.

CRISPRa-mediated PPARGC1A activation enhances mitochondria in human hearts ex vivo

To further test the translational potential of our CRISPRa strategy to enhance mitochondrial biogenesis and function, we evaluated efficacy in human hearts using a biomimetic human heart culture system. We have demonstrated that this ex vivo system allows for robust and controlled testing in therapeutically relevant contexts.^74–76 In this platform, myocardium from the left ventricle of human donor hearts was excised and sectioned into individual slices. Each slice was treated with CRISPRa targeting PPARGC1A (i.e. CRISPRa) or CRISPRa with non-targeting gRNA (i.e. Scramble). 3 days post-transduction, samples are collected and analyzed (Fig. 5A). We first confirmed that none of the conditions negatively impacted tissue viability (Fig. S7). CRISPRa treatment increased PPARGC1A protein levels as expected and also increased the levels of mitochondrial marker TOMM20 (Figs. 5B and 5C). Furthermore, CRISPRa increased the number of copies of mtDNA by ~32% (Fig. 5D), in line with our in vitro results (Fig. 1C). Finally, we evaluated general mitochondrial function by measuring the basal levels of oxygen consumption, as well as glycolysis by measuring extracellular acidification. In normal hearts, we detected an increase in mitochondrial function (~101% increase) and no changes in glycolysis relative to Scramble condition (Figs. 5E left and S8). Interestingly, hearts from donors with reduced EF (EF < 40%), a clinical indicator for heart failure with reduced ejection fraction (HFrEF),^77–79 were also responsive to the CRISPRa treatment. After delivery of CRISPRa targeting PPARGC1A, we measured an increase in mitochondrial function (~63% increase) along with no changes in glycolysis (Fig. 5E right and S8). Together, these indicate CRISPRa is able to activate transcription of PPARGC1A, leading to an increase in mitochondrial quantity and function in adult human heart slices, positioning CRISPRa as a potential therapeutic path for addressing heart failure, as well as other pathologies wherein improved mitochondrial and cellular bioenergetics would be advantageous.

Figure 5. PPARGC1A activation enhances mitochondria numbers and cellular function in healthy and diseased adult human heart tissue.

(A) Left ventricle myocardium was resected from a recently deceased donor heart and sectioned into slices. Each slice was transduced with lentivirus encoding either CRISPRa or Scramble payloads as indicated. (B-C) Immunofluorescence microscopy (panel B) and quantification (panel C) of CRISPRa-based treatment in human heart slices or Scramble lentivirus delivery 3 days post-transduction to human heart slice (white scale bar 50 μm). Cells were probed with antibodies against the FLAG epitope on CRISPRa molecules (green), PPARGC1A target protein (red), mitochondrial marker TOMM20 (purple), and nuclear counterstain DAPI (blue). (D) Mitochondrial DNA (mtDNA) quantification 3 days after lentivirus transduction. (E) Initial oxygen consumption rate (OCR) from healthy (left) or diagnosed heart failure with reduced ejection fraction (HFrEF) adult human heart (right). Data presented in panel B represents 3 independent experiments. Data presented in panels C, D, and E as mean +/− s.e.m. from 3 independent experiments. *P < 0.05, **P < 0.01, and ****P < 0.0001 (Student’s t-test).

DISCUSSION

Loss of mitochondria and reduction in mitochondrial function are major drivers in the development of heart failure after acute-MI. Here, we have identified that the activation of endogenous PPARGC1A using CRISPRa stimulates mitochondrial biogenesis and is a potential therapeutic opportunity to enable energetic recovery in damaged cardiac cells. While related approaches have displayed limited success at specifically and safely upregulating PPARGC1A,^35–39 our study demonstrates that CRISPRa can precisely and effectively activate this master regulator of endogenous mitochondrial biogenesis. In human cells, we found that gRNA-targeted CRISPRa increases the transcription of PPARGC1A, resulting in increased cellular mitochondria numbers. Furthermore, our results indicate that CRISPRa-upregulated PPARGC1A results in the activation of a diverse array of mitochondrial and bioenergetic cellular pathways, while also strengthening transcriptomic profiles characteristic of adult cardiomyocytes, such as repression of cell cycle and DNA replication.

Our findings also demonstrate that CRISPRa can effectively increase mitochondrial biogenesis in both skeletal muscle and cardiac muscle. Notably, CRISPRa-driven increases in mitochondria led to an improvement in ejection fraction through enhancement of cardiac bioenergetics, rather than repression of cardiac fibrosis or modulation of inflammatory response, opening the possibility of further recovery through combination with therapies targeting said pathways. Nonetheless, our current outcomes are comparable to previous reports using intramyocardial injection of mitochondria,^21,22 suggesting that CRISPRa can achieve similar bioenergetic benefits as existing mitochondrial transplantation therapies. One key advantage of CRISPRa lies in its potential to be a less invasive and more clinically scalable procedure for treating heart failure. Unlike direct mitochondrial injections, which can be technically challenging and invasive,^27–30 CRISPRa presents the opportunity for a more scalable and robust gene therapy-based approach for the enhancement of cardiac mitochondria. While a promising and exciting proof-of-concept, our current CRISPRa system has the opportunity to be further refined before clinical usage: delivery through adeno-associated viruses could increase tissue-specific targeting of the therapy;^80,81 alternatively, non-viral delivery strategies, such as mRNA delivery through lipid nanoparticles,^82–84 open the possibility of therapeutic redosing. Similarly, the therapeutic effect can be enhanced by employing multiple gRNAs for synergistic activation of a single target, or by multiplexing gRNAs to activate several beneficial genes simultaneously.^42,48 However, any changes in delivery modality or system targeting would necessitate long-term studies to establish both the efficacy and safety of any future formulation. Nonetheless, our results still position CRISPRa as a promising therapeutic strategy for heart failure, with the added benefit of being adaptable to other tissues and conditions where mitochondrial dysfunction plays a critical role, such as cerebral stroke, Alzheimer’s disease, and wound healing.^69,85

The results from our human donor heart experiments demonstrate that our CRISPRa system can be effectively delivered to adult cardiac cells and activate target gene expression in these tissues. In this study, we used this capability to improve mitochondrial function in both normal and diseased human heart tissue. These findings are a crucial addition to the increasing therapeutic promise of CRISPRa.^86–88 The outcomes observed in our experiments underscore the growing utility of CRISPRa as a powerful tool in regenerative medicine, offering a less invasive and more scalable approach to addressing mitochondrial dysfunction and related diseases.

MATERIALS AND METHODS

Molecular Cloning

All plasmids encoding dCas protein variants and transactivation domains utilized in this study were developed from previously reported constructs. SadCas9-VPR, the hU6 promoter, and P2A-eGFP were polymerase chain reaction (PCR)-amplified from SadCas9 VPR (Addgene, 373494), pSPgRNA (Addgene, 47108), and pCMV-T7-SpRY-P2A-EGFP (RTW4830; Addgene, 139989), respectively. The NMS domain was PCR-amplified from a previously described plasmid.⁴² The SadCas9 CRISPRa plasmid was constructed by cloning these PCR-amplified fragments and a PCR-amplified EFS promoter and SadCas9-gRNA scaffold (Addgene, 373494) into the Lenti-EFS-dCas9-dMSK1 plasmid (Addgene, 165602) digested with KpnI and BsrGI via NEBuilder HiFi DNA assembly (NEB, E5520). These steps were repeated to construct the AsdCas12a CRISPRa plasmid using a PCR-Amplified AsdCas12a-VPR (Addgene, 188512) and EFS-AsdCas12a gRNA scaffold (Addgene, 373994). The modified EnAsdCas12a-VPR variant was constructed similarly via PCR amplification of the AsdCas12a scaffold cassette and the AsdCas12a-VPR fragment with specific primers to induce the previously described⁶⁰ enhancing mutations E174R/S542R/K548R. The gRNAs for each variant were cloned into the compatible gRNA backbone following BbsI or Esp3I digestion and NEB T4 DNA Ligase (NEB, M0202M). cDNA plasmids were constructed by cloning the PCR-amplified EFS promoter and P2A-eGFP fragments into the same backbone along with PCR amplified fragments from a cDNA library from HEK293T cells. All gRNA protospacer sequences are provided in Table S1.

Cell Culture

HEK293T (ATCC, CRL-3216), HDFa (ATCC, PCS-201-012), and AC16 (Sigma, SCC109) cells were obtained from commercial suppliers. HEK293T cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% FBS (Sigma, F2442) and 1% Pen/Strep (100 units/mL penicillin, 100 μg/mL streptomycin; Fisher, 15-140-163). HDFa cells were cultured in Fibroblast Basal Medium (ATCC, PCS-201-030) supplemented with Fibroblast Growth Kit-Low serum (ATCC, PCS-201-041) and 1% Pen/Strep. AC16 cells were cultured in DMEM/F12 (Sigma, D6434) supplemented with 2 mM L-Glutamine (Fisher, 25030081), 12.5% FBS, and 1% Pen/Step. All cells were incubated at 37°C in 5% CO₂. For HEK293T, cells were seeded in a 24-well plate with 1×10⁵ cells per well. The next day, transfection was performed with 375 ng of a CRISPRa expression vector and 125 ng of a corresponding gRNA vector or 500 ng of cDNA plasmid. All transfections were performed using Lipofectamine 3000 (ThermoFisher, L3000001) per manufacturer’s instructions. For AC16 and HDFa, cells were seeded in a 24-well plate with 1×10⁵ cells per well and corresponding lentivirus at a multiplicity of infection (MOI) of ~10, as previously reported.⁴³ The next day, media with lentivirus was replaced with new media.

Lentiviral Production

HEK293T cells were seeded in a 10 cm plate to result in ~80–90% confluency the following day. The next day, cells at ~80–90% confluency were transfected. For each transfection, 10 μg of plasmid containing the vector of interest, 10 μg of pMD2.G (Addgene, 12259), and 15 μg of psPAX2 (Addgene, 12260) were transfected using calcium phosphate. 5 hrs post-transfection the media was changed. The supernatant was harvested 24 and 48 hrs post-transfection and filtered with a 0.45 μm PVDF filter (Millipore, SLGVM33RS). For in vitro and ex vivo experiments, the virus was concentrated using a Lenti-X Concentrator (Takara, 631232), aliquoted, and stored at −80°C until use. For in vivo experiments, the virus was concentrated and purified using a Lenti-X Maxi Purification Kit (Takara, 631234), aliquoted, and stored at −80°C until use. Lentiviral titers were measured using a Lenti-X qRT-PCR Titration Kit (Takara, 631232).

Tissue Culture

As previously described,⁷⁵ left ventricles from commercially procured, de-identified donor heart tissue were cut into 1–2 cm³ cubes after being placed in cardioplegia solution in a sterile glass container. Briefly, a 4% agar bed was prepared on a specimen holder for each cube, and Histoacryl Blue tissue glue was used to attach the epicardium of the cube to the bed such that the endocardium was oriented upwards. The sample structure was then placed in a Vibrating Microtome 700SMZ filled with oxygenated modified Tyrode’s cutting solution (140 mM NaCl, 6 mM KCl, 10 mM glucose, 10 mM HEPES, 1 mM MgCl₂, 1.8 mM CaCl₂, and 10 mM 2,3-butanedione monoxime (BDM), at pH 7.4 and 4°C) for cutting. The vibrating microtome was used under the following settings: calibration of z-axis vibration with a ceramic cutting blade < 0.5 μm; 300 μm slice thickness; 0.03 mm/s advance speed; 80 Hz vibration frequency; and 2 mm horizontal vibration amplitude. Tissue slices were immediately placed in 100 μm nylon mesh cell strainers placed in oxygenated washout Tyrode’s solution (140 mM NaCl, 4.5 mM KCl, 10 mM glucose, 10 mM HEPES, 1 mM MgCl₂, 1.8 mM CaCl₂, and 2x Antibiotic-Antimycotic, at pH 7.4 and room temperature) for at least 20 min under a metal washer. Heart slices were then placed in stimulated culture. Histoacryl Blue tissue glue was used to attach the ends of the tissue to a sterilized polyurethane printer 6 mm wide printer timing belt with embedded metal wires (Uxcell). Each attached slice was placed in a 6 well plate with 6 mL culture medium (Medium 199, 1x ITS Supplement, 10% FBS, 5 ng/mL VEGF, 10 ng/mL FGF-basic, and 2x Antibiotic-Antimycotic). The 6 well plate was covered with a graphite electrode-containing C-Dish top (Ionoptix) and stimulated at 10 V, 1.2 Hz using the C-Pace-EM system (Ionoptix). Stimulated tissues were incubated at 37°C in 5% CO₂. Media was replaced with preoxygenated media every 8 hrs, and the C-Dish top was changed daily.

Surgery / MI Model and Hindleg

Male Sprague Dawley rats (Envigo) underwent acute-MI via ligation of the left anterior descending (LAD) coronary artery at 70 – 90 days old. Animals were anesthetized with 2 – 3.5% isoflurane delivered via oral intubation with a ventilator. A volume-cycled Rat Ventilator Minivent (Hugo Sachs Electronik Harvard Apparatus GmbH) was connected to the endotracheal tube and supplemental 100% oxygen (70 – 80 breaths/min, 1–3 mL/breath) was provided. Animals were placed supine prior to local hair removal, subcutaneous injection of 1% lidocaine (0.50 cc), and three rounds of skin cleansing, alternating betadine and ethanol.

Intramuscular injection was performed in the caudal thigh muscle of the hind limb using the backpack technique. The area of interest was shaved and swabbed with alcohol. Injection was performed with CRISPRa, control lentivirus, or control saline. Euthanization was performed at 21 days post-injection, followed by histological analysis.

Anesthetized rats underwent lateral thoracotomy. A 1 – 2 cm incision was made through the ventral chest skin to mid-thorax, and the thorax was opened to mid-sternum with blood vessel coagulation induced by mechanical pressure via a cotton tip applicator. Permanent ligation was performed upon subsequent exposure of the heart with a 7–0 or 8–0 suture placed underneath the LAD coronary artery.

Lentivirus was injected into the myocardium immediately after MI with a Hamilton syringe (34-gauge needle). Five-20 μL injections were delivered along the border of the ischemic area. A 20-gauge vascular catheter was inserted post-injection, and the chest wall was closed using embedded technique with 4–0 Vicryl sutures. A 10-cc empty syringe was connected to the chest tube after skin closure, and the chest tube was removed under negative pressure. After procedure completion animals were recovered from anesthesia. Animals received strict follow up for 5 days post-surgery survival. Heart function was assessed at days 1, 14, and 21 post-MI via echocardiography. Animals were euthanized at 21 days post-MI for histological analysis.

All animals were maintained under standard housing conditions throughout the study. Postoperative pain and welfare were assessed daily using the Grimace Scale, with all animals consistently scoring 0 (No Signs Present) throughout the study. All treatment was performed in accordance with the Guide for Care and Use of Laboratory Animals (NIH publication 86–23, 1996). The study was approved by the Baylor College of Medicine Institutional Animal Care and Use Committee (AN-8952).

RT-qPCR and qPCR

RNA isolation was performed on cells with the RNeasy Plus Mini Kit (Qiagen, 74134) per manufacturer’s instructions. Isolated RNA was used as a template for cDNA synthesis with the iScript cDNA Synthesis Kit (Bio-Rad, 1708890). A 1:10 dilution of the cDNA was used in Real-time quantitative PCR (RT-qPCR) using the Luna Universal qPCR Master Mix (NEB, M3003) and a CFX96 Real-Time PCR Detection System with a C1000 Thermal Cycler (Bio-Rad, 1845096). Results are provided relative to untransfected control cells following normalization to GAPDH using the ΔΔCt method. Endogenous expression primers were designed targeting the 5’ UTR of the target transcript, similar to previous reports.^63,89,90

Mitochondrial DNA (mtDNA) copy number quantification was performed with purified DNA from cells. DNA was isolated using the DNeasy Blood & Tissue Kit (Qiagen, 69504) per manufacturer’s instructions. Purified DNA was quantified using the Real-time qPCR. All qPCR primer sets used for RT-qPCR underwent the following cycling conditions: 95°C for 30 s, then 40 cycles of 95°C for 10 s and 60°C for 30s. All qPCR primers are provided in Table S2.

RNA sequencing

RNA was isolated from transfected AC16 cells with the RNeasy Plus mini kit (Qiagen, 74136), per manufacturer’s instructions, and sent to Novogene to create and sequence RNA sequencing (RNA-seq) libraries with Illumina NovaSeq X Plus. Reads were trimmed using Trim Galore (0.6.10), after which they were aligned to the hg38 transcriptome⁹¹ using STAR (2.5.2b), and finally quantified using RSEM (1.3.1). DESeq2 (v1.28.1, R studio v1.2.13) with default settings was used to determine differential expression.⁹² DAVID Functional Annotation Bioinformatics Microarray Analysis was used for gene ontology.⁹³

Live Microscopy

Live imaging was performed on mitochondria using the TMRE Mitochondrial Membrane Potential Assay Kit (Abcam, ab113852). AC16 cells were seeded in a microplate to be at 80–90% confluency by the day of assay and allowed to adhere to the microplate in preparation for staining. TMRE (100 nM; Abcam, ab274305) was added to the cells in media. Following incubation for 15 min, TMRE media was removed and replaced twice with 100 μL PBS/0.2% BSA. The plate was imaged with a fluorescent plate reader with the excitation and emission wavelengths set to 549 and 575 nm, respectively.

Immunofluorescence

Rat hearts and hindlegs muscles were isolated, and rat hearts were trimmed at the apex and the area above the suture site. Samples were then washed in PBS and incubated in 4% paraformaldehyde (ThermoFisher, 28906) for 24 hrs. After the incubation, each sample was washed in PBS 3 times for an hour each and then moved to 30% sucrose for 24 hrs. The tissues were then embedded in OCT (Sakura, 4583) to produce frozen OCT blocks, in which they remained at −80°C for 24 hrs before they were sectioned. Frozen tissues were cut into 8 μm thick sections with a cryostat. Sections were placed on slides and stored at −20°C prior to staining. At the time of staining, the section was placed at room temperature for 30 min in PBS to dissolve the OCT. The tissues were then placed in Citrate Buffer (Novus, NB900–62075) heated to 95°C for 30 min. After antigen retrieval was complete, the sections were washed in TBST 3 times for 5 min each. The tissues were then placed in permeabilization solution (Biotium, 23012) for 30 min to block and permeabilize the tissue. The sections were washed in TBST 3 times for 5 min each. The sections were incubated with primary antibody against PPARGC1A (1:10; Novus, NBP1–04676) overnight, a fluorescent antibody against rabbit (1:200; ThermoFisher, A-11035) for 30 min, a primary antibody against Troponin T (1:200; ThermoFisher, MA5–12960) for 90 min, a fluorescent antibody against mouse (1:50; Jackson ImmunoResearch Laboratories, 115–685-166) for 60 min, and fluorescent antibodies against TOMM20 (1:200; Abcam, ab209606), FLAG (10 μg/mL; Sigma, F4049), α-SMA (1:50; BioLegend, 614855), Collagen I (1:200; Bioss, BS-10423R-BF488), CD4 (1:50; BioLegend, 201511), and CD8 (1:50; BioLegend, 201710) overnight, with 3 TBST washes in between steps. Troponin T antibody was omitted for hindleg samples. Tissue sections were then stained with DAPI-containing mounting medium (Biotium, 23004).

Masson’s Trichrome

Cryosections were placed in 1X PBS (pH 7.4) for 30 minutes to remove OCT. Slides were then fixed in preheated (56–64°C) Bouin’s Fluid (Epredia, 57211) for 60 min, followed by a 10 min cooldown period. Sections were rinsed in water until visibly clear, after which nuclei were stained with Weigert’s Iron Hematoxylin (Electron Microscopy Sciences, 26044–15 and 26044–06) and rinsed in water. The sections were then stained for 15 min in Biebrich Scarlet/Acid Fuchsin Solution (Electron Microscopy Sciences, 26033–25) and rinsed in water. Differentiation was then performed by immersion in Phosphomolybdic/Phosphotungstic Acid Solution (Electron Microscopy Sciences, 26367–05) for 15 minutes. Without rinsing, the slides were immediately stained in Aniline Blue Solution (Electron Microscopy Sciences, 26027–10) for 10 minutes, rinsed in water, and treated with 1% Acetic Acid Solution (Electron Microscopy Sciences, 26367–07) for 3 minutes to set the stain. Finally, slides were rapidly dehydrated using two changes each of 95% and 100% ethanol, cleared in Xylene, and permanently mounted in synthetic resin.

Respirometry

Prior to assessment, the sensor cartridge from the XFe96 Extracellular Flux Assay Kit was placed upside down adjacent to the utility plate. Each well of the utility plate was filled with 200 μL Seahorse XF Calibrant (Agilent 100840–000), then topped with the sensor cartridge such that the sensors would be fully submerged in water. The structure was incubated overnight in a non-CO₂ incubator at 37°C. AC16 cardiomyocytes were seeded at 1×10⁴ cells per well in an 80 μL suspension. Background wells received 80 μL of media. Plated cells were incubated at room temperature in a tissue culture hood for 1 hr to allow for even cell distribution and minimized edge effects, followed by overnight incubation at 37°C in 5% CO₂. 40 mL of frozen Seahorse XF base medium (Agilent, 103335–100) and L-glutamine was thawed overnight at 4°C. The next day, adequate cell growth was confirmed via microscopy, and the thawed XF base medium was supplemented with 400 μL of 100 mM pyruvate (100X, 1 mM final concentration; Agilent, 103578–100), 400 μL of 200 mM glutamine (100X, 2 mM final concentration; Agilent, 103579–100), and 400 μL of 10000 mM glucose (100X, 10 mM final concentration). Media in each well was replaced with 180 μL of supplemented XF media, and cells were incubated at 37°C in a non-CO₂ incubator for 45 min to 1 hr. Mito stress compounds from the XF Cell Mito Stress Test Kit (Agilent, 103015–100) were reconstituted as follows: Oligomycin in 630 μL (100 μM final concentration), FCCP in 720 μL (100 μM final concentration), and Rotenone/Antimycin A in 540 μL (50 μM final concentration). Reconstituted compounds were diluted as follows: 450 μL Oligomycin into 2550 μL supplemented XF medium or 600 μL Oligomycin into 3400 μL supplemented XF medium, 600 μL FCCP into 2400 μL supplemented XF medium or 700 μL FCCP into 2800 μL supplemented XF medium, and 300 μL Rotenone/Antimycin A into 2700 μL supplemented XF medium or 500 μL Rotenone/Antimycin A into 4500 μL supplemented XF medium. With the utility plate submerged in calibration buffer, the mito stress compounds were added to the XFe96 Extracellular Flux Assay Kit sensor using the loading guides. Port A was filled with 20 μL Oligomycin; port B was filled with 22 μL FCCP; and port C was filled with 25 μL Rotenone/Antimycin A. Software was calibrated with the XFe96 Extracellular Flux Assay Kit sensor and calibration buffer plate, and cells were subsequently loaded to measure OCR and ECAR. Raw data was then normalized to the number of cells, which was determined using the CyQUANT Direct Red Cell Proliferation Assay (ThermoFisher, C35013), per manufacturer’s instructions.

Viability

Tissue viability was assessed via MTT assay with the CyQUANT MTT Cell Proliferation Assay Kit (ThermoFisher, V13154) per manufacturer’s instructions. A 6 mm biopsy punch of heart slice tissue was incubated in MTT (5 mg/mL) at 37°C for 3 hrs. The MTT solution in each well was then replaced with 1 mL DMSO before incubation at 37°C for 15 min. At 15 min, once the purple color had disappeared, a Cytation plate reader was used to read the 570 nm absorbance of each tissue sample. The tissue was removed from the DMSO solution and dried and pressed for weighing. The absorbance was calculated relative to the blank and normalized to sample weight to yield values in OD/mg tissue.

Data and materials availability:

All data associated with this study are present in the source data (Table S3) and under GEO accession number GSE298270. Materials are available by request to the authors.