Circular RNA circZFPM2 regulates cardiomyocyte hypertrophy and survival

Hypertrophic cardiomyopathy (HCM) constitutes the most common genetic cardiac disorder. However, current pharmacotherapeutics are mainly symptomatic and only partially address underlying molecular mechanisms. Circular RNAs (circRNAs) are a recently discovered class of non-coding RNAs and emerged as specific and powerful regulators of cellular functions. By performing global circRNA-specific next generation sequencing in cardiac tissue of patients with hypertrophic cardiomyopathy compared to healthy donors, we identified circZFPM2 (hsa_circ_0003380). CircZFPM2, which derives from the ZFPM2 gene locus, is a highly conserved regulatory circRNA that is strongly induced in HCM tissue. In vitro loss-of-function experiments were performed in neonatal rat cardiomyocytes, human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs), and HCM-patient-derived hiPSC-CMs. A knockdown of circZFPM2 was found to induce cardiomyocyte hypertrophy and compromise mitochondrial respiration, leading to an increased production of reactive oxygen species and apoptosis. In contrast, delivery of recombinant circZFPM2, packaged in lipid-nanoparticles or using AAV-based overexpression, rescued cardiomyocyte hypertrophic gene expression and promoted cell survival. Additionally, HCM-derived cardiac organoids exhibited improved contractility upon CM-specific overexpression of circZFPM2. Multi-Omics analysis further promoted our hypothesis, showing beneficial effects of circZFPM2 on cardiac contractility and mitochondrial function. Collectively, our data highlight that circZFPM2 serves as a promising target for the treatment of cardiac hypertrophy including HCM. Supplementary Information The online version contains supplementary material available at 10.1007/s00395-024-01048-y.


Human tissue sampling
Heart tissue samples for RNA sequencing were obtained from 5 HCM patients after myectomy and 5 healthy donors.For qPCR validation, heart tissue samples from 18 HCM patients and 18 healthy donors were utilized.This study was conducted with the approval of the international ethics committee of the Hannover Medical School, Germany, and in accordance with the guidelines from the declaration of Helsinki and its amendments or comparable ethical standards.

Circular RNA sequencing and in silico validation
Total RNA isolated from cardiac tissue from 5 healthy donors and 5 HCM patients was tested for RNA integrity using the Agilent Bioanalyzer, after RNase R treatment and removal of ribosomal RNA.Then, according to the manufacturer's instructions of the Illumina Truseq kit, the RNA library was prepared.
With the help of the Illumina Hiseq2000 at a read length of 100 bp, the paired-end sequencing was performed.After quality trimming, short reads were aligned to circRNA reference sequences from circBase (www.circbase.org)utilizing bowtie2 in sensitive mode.Raw read counts were then obtained by feature counts.To quantifiy the data and differentiate significant levels between HCM and healthy group the R package DESeq2 was utilized.The integrative genomics viewer [11] was used to display data.
Supplemental Table 3

Animal experiments
Male C57BL/6J WT mice at the age of 3-6 months were kept in a light-controlled room with a 12:12 lightdark cycle and had free access to standard mouse chow and water.Left coronary artery ligation was conducted as previously described to induce myocardial infarction (MI) [5].In brief, the mice were anesthetized with 2% isoflurane, then intubated and artificially ventilated.A 1.5 cm incision was made into the skin at the mid-thorax parallel to the ribs.The left pectoralis muscle was cut off, the muscle layers were pulled aside and thoracotomy was conducted between the second and third rib.Then, the left coronary artery was ligated with 8-0 suture, 1mm distal to the left atrium.When the chest was closed again, the animals received 0.

Neonatal rat cardiomyocyte isolation and culture
Following the manufacturer's instructions of the Neonatal Heart Dissociation Kit (Miltenyi), cardiomyocytes were isolated from neonatal rat pups at the age of 0.5 to 3 days.Minimum Essential Medium (MEM, Bioconcept), supplemented with 1% penicillin/streptomycin (Gibco), 2 µg/mL vitamin B12 (Sigma), 100 nM bromdeoxyuridine (BrdU, Sigma) and 5% fetal bovine serum (FBS, Gibco) were used for cell maintenance for three days until experiments were conducted.

Wild type hiPSC culture
Wild type human induced pluripotent stem cells (hiPSCs) (hHSC_Iso4_ADCF_SeV-iPS2, alternative name: MHHi001-A) were maintained in supplemented StemMACS full medium (Miltenyi) at 37°C and 5% CO2 on cell culture plates coated with Geltrex (Gibco) [4,4].Every 5 days the cells reached confluence and were passaged with Versene (Gibco) in StemMACS full medium (Miltenyi) supplemented with 2 µM Thiazovivin (Selleckchem) in a ratio of 1:6.By modulation of the Wnt pathway as previously described, the hiPSCs were differentiated into cardiomyocytes and then purified and maintained for 40-60 days in culture until experiments were conducted [7].

HCM patient derived iPSC culture
The HCM-iPSC line was cultured on cell culture plates coated with Geltrex (Gibco) and supplemented mTeSR1 maintenance medium (StemCell) [8].Every 5 days, the cells reached confluence and were passaged using Accutase (StemCell).HCM-iPSCs were differentiated into cardiomyocytes by modulation of the Wnt pathway as previously described, and then purified and maintained for 40-60 days in culture until experiments were conducted [7].

Human cardiac fibroblast maintenance
Human cardiac fibroblasts (HCFs) were cultured in Fibroblast Growth Medium 3 (FGM3, PromoCell).The cells were passaged using pre-warmed Trypsin and kept in a humidified incubator at 37°C with 5% CO2.
HCFs were used from passage four to eight for experiments.

HEK cell maintenance
Human embryonic kidney cells (HEK293) were cultured in DMEM media (Gibco) supplemented with 10% FBS and 1% penicillin/streptomycin and used to produce AAV6 viral particles.The cells were passaged using pre-warmed Trypsin/EDTA and were seeded with desired cell numbers onto the desired plates and kept in a humidified incubator at 37°C with 5% CO2.

Actinomycin D
NRCMs were treated with 2 µg/mL Actinomycin D (Sigma Aldrich) in MEM medium from 0 to 48 h.

Hypoxia
For hypoxic conditions, NRCMs were kept for 72h in the incubator with 0.01% O2.The control plate was incubated for 72h under normoxic condition (21% O2).

Transforming growth factor β
HCFs were treated with 5 ng/mL of the recombinant human transforming growth factor β (TGFβ) or the vehicle (0.1% bovine serum albumin + 4 mM HCl) for 72 h.

Phenylephrine
NRCMs were treated with 100 µM phenylephrine in MEM medium with reduced FBS concentration (1% FBS) for 72 h.

Leukemia inhibitory factor
NRCMs were treated with 5 nM leukemia inhibitory factor in MEM medium with reduced FBS concentrations (1% FBS) for 48 h.

Knockdown of circZFPM2
The knockdown of circZFPM2 was achieved by using species-specific siRNA specifically binding the backsplice site without targeting the linear host gene, while scrambled siRNA was used as negative control.The HPLC purified siRNAs were purchased from Eurofins Genomics.

Design of the overexpression plasmid
The overexpression plasmid carried the rodent exon 2 and exon 3 of the ZFPM2 locus, flanked by splice sites and circularization elements consisting of 100 bp of the endogenous intron 1 and the endogenous upstream Alu element and downstream by 100 bp of the endogenous intron 3 and the endogenous downstream Alu element.The plasmid additionally contained a 50 bp long sequence with several restriction enzyme sites upstream and downstream of the overexpression cassette.This construct was purchased from Invitrogen and delivered in a vector backbone, the delivery vector (pMA-RQ (AmpR)).

Bacterial transformation
100 µL of Stbl3 E.coli was gently thawed on ice.Then, 1 ng of the overexpression cassette containing the delivery vector was added and the mix was incubated for an additional 30 min on ice.The heat-shock approach was utilized for plasmid uptake, meaning that the mix was heated for exactly 90 sec at 42°C and then cooled down for a further 2 min on ice.Then, 100 µL of SOC outgrowth medium was added, and the sample was incubated at 37°C and 300 rpm for 60 min, spread on pre-warmed LB/ampicillin agar plates and incubated up-side-down overnight at 37°C.

Plasmid amplification and validation
Single E.coli colonies were picked from the LB/ampicillin agar plates after bacterial transformation and inoculated in 5 mL LB/ampicillin media overnight at 37°C and 225 rpm.Then, according to the manufacturer's instruction of the NucleoSpin Plasmid (NoLid) Mini kit, the plasmid DNA was isolated and Sanger sequencing verified the correct delivery vector.

Cloning of overexpression vector
For the AAV6 overexpression system, the overexpression cassette was cloned into the AAV MCS 1.3 vector.Therefore, both the delivery vector and the AAV vector were cut with the restriction enzymes EcoRI and XbaI.Gel electrophoresis (1% agarose) was conducted to detect bands of the expected size, which were cut out and extracted from the gel, using the QIAquick gel extraction kit.Correct vectors and inserts (verified by Sanger sequencing) were ligated with a molar ratio of 1:3 with the T4 DNA ligase and transformed into Stbl3 E.coli.Then they were amplified, isolated and validated as previously described.

Plasmid amplification
Stbl3 E.coli containing the empty AAV MCS plasmid (cell biolabs, #VPK-410) as negative control, the circZFPM2 overexpression plasmid or the helper plasmid PDP6rs with rep and cap genes necessary for AAV production, were each added to 5 mL LB media supplemented with Ampicillin.These bacterial suspensions were incubated for 16 h at 37°C and 225 rpm and then added to 600 mL LB media supplemented with Ampicillin (100 µg/mL).Using the NucleoBond Xtra Maxi EF Kit (Macherey-Nagel) according to the manufacturer's protocol, the plasmids were isolated after an additional 16 h (37°C, 225 rpm).

Plasmid validation
After isolation, the plasmids were validated by restriction enzyme digestions.The PDP6rs plasmid was cut with Clal and KpnI (incubation for 2 h at 37°C), while AhdI (incubation for 2 h at 37°C) and SmaI (incubation for 2 h at 25°C) were used for the overexpression plasmid.The correct digest pattern was verified by agarose gel electrophoresis before proceeding with viral production.

AAV6 production
HEK293T cells were seeded on cell culture dishes.After reaching 80% confluence, the cells were transfected with the empty AAV MCS 1.3 vector or the overexpression vector along with the PDP6rs helper plasmid and polythylenimine (PEI) as shown in the following table : Transfection mix for AAV production.After medium change, the cells were incubated for an additional 72 h.The supernatant was then collected and the cells were scraped from the dishes.The collected cells were centrifuged twice for 10 min at 1000 x g, pellets were collected, resuspended in lysis buffer and stored at -80°C.To collect viral particles secreted into the supernatant or contained in detached cells, the collected supernatant was also centrifuged for 10 min at 1000 x g. 25 mL 40% PEG8000 solution were added per 100 mL supernatant.The mix was stirred for 1 h at 4°C and then kept at 4°C overnight without stirring.The next day, the sample was centrifuged for 15 min at 4°C and 2,800 x g.The pellets were resuspended in lysis buffer.Then, the frozen cells were subjected to 3 thaw-freeze cycles and the resulting cell lysate was added to the lysate from the supernatant of the dishes and incubated with 1 M MgCl2 (final 1 mM) and benzonase (final 250 U/mL).This mixture was lysed mechanically and centrifuged.AAV particles from the supernatant were isolated using an iodixanol (OptiPrep, Progen) gradient.After ultracentrifugation (rotor type 70Ti) for 70 min at 16°C and 63,000 rpm, the 40% iodixanol fraction containing the mature virus was collected and the viral particles were concentrated using Amacon100K columns (Millipore).5 µL of the viral suspension were collected for gDNA isolation (DNeasy Blood and Tissue Kit) to calculate the viral titer via qPCR.The remaining viral suspension was aliquoted and stored at -80°C.

In vitro transcription of circRNA
To produce the artificial human circZFPM2, the linearized human circZFPM2 sequence (exon 2 and 3 of the human ZFPM2 locus) was purchased from Invitrogen.The DNA template containing the T7 promoter for in vitro transcription was generated by PCR.The entire linearized circZFPM2 sequence was amplified using a primer pair of which the forward primer exhibited a T7 promoter sequence.According to the manufacturer's protocol of the QIAquick PCR Purification Kit (Qiagen), the amplicon was purified and then used for in vitro transcription with the RiboMax Large Scale RNA Production System (Promega) with a modified protocol.In brief, a mixture of 5-10 µg DNA template, 10 µL T7 RNA polymerase, 25 mM ATP/CTP/UTP, 5 mM GTP and 20 mM GMP was incubated for 2 h at 37°C while shaking.DNA templates were removed by DNase treatment with the ratio of 1U DNase/1µg DNA template.For RNA isolation, 700 µL Qiazol and 140 µL Chloroform were added to the samples, mixed well and centrifuged for 5 min at 4°C and 17,000 x g.After transferring the supernatants to new Eppendorf tubes, sodium acetate in a 1:10 ratio and the same volume of isopropanol were added and then centrifuged for 5 min at 4°C and 17,000 x g.
Finally, precipitated RNA was washed with 100% ethanol and resuspended in 40 µL double-distilled, RNase-free water.In order to facilitate the circularization, DNA splints (20 bp in length) binding 10 bp upstream and 10 bp downstream to the backsplice site were incubated in a molar ratio of 1:3 with the produced RNA and the T4 DNA ligase for 1 h at 37°C.Then the circularized RNA was isolated as described above and validated using the 2100 Bioanalyzer (Agilent Technologies) and by Sanger sequencing of the PCR products generated by using divergent primers.

IVT transfection
HiPSC-CMs and NRCMs were transfected with the in vitro transcribed (IVT) circZFPM2 using Lipofectamine RNAiMAX. 2 µL RNAiMAX per 1 mL OptiMEM and 100-500 ng IVT circRNA per 1 mL OptiMEM were incubated separately for 5 min at room temperature, then mixed together and incubated for further 15 min.20 µL/40 µL/200 µL of the transfection mix were added drop-wise per well of a 96-/48-/12-well plate and experiments were performed without additional medium change after 24 h, 48 h or 72 h.

RNA isolation
RNA was isolated with QIAzol (Qiagen) according to the manufacturer's protocol.RNA concentrations were measured using Take3 Plates and the Bio-Tek plate reader (Synergy HT).

cDNA synthesis
Using the Biozym cDNA Synthesis Kit with random hexamer primers according to the manufacturer's protocol, 500-1000 ng RNA were reverse transcribed.
Divergent primer pairs were designed to bind upstream and downstream of the backsplice site and thereby amplifying the characteristic backsplice region of the circRNA.
Primer list.

PCR and qPCR
Polymerase chain reaction (PCR) was conducted using the HotStarTaq Master Mix (Qiagen) and the Thermocycler Biometra Trio (Analytik Jena).Real-time qPCR was performed with the Absolute Blue qPCR SYBR Green Mix (Thermo Scientific), utilizing the QuantsStudio7 (ABI).For both methods, target specific primers were used and all experiments were performed according to the manufacturer's instructions.The target gene expression levels were normalized to the house keeping genes hypoxanthine-guanine phosphoribosyl transferase (HPRT), TATA-binding protein (TBP) or 18S ribosomal RNA (18S) and analyzed via the ∆∆CT method.
the endogenous upstream and downstream Alu elements.Black regions are 100 bp of the endogenous intron 1 and 3. Bold regions are splice sites, the grey region is exon 2 of circZFPM2 and the green region is exon 3 of circZFPM2.

Table 1 .
Overview on human heart samples used for the RNA sequencing.

Table 2 .
Overview on human heart samples used for the qPCR validation.
: Differentially regulated circRNAs in HCM tissues and their level of regulation