Human Induced Pluripotent Stem Cell-Derived Cardioids as a Model to Assess Oligonucleotide Delivery

Summary

This protocol enables the generation of self-organizing cardiomyocyte-based cardioids from human induced pluripotent stem cells for microscopic assessment of fluorescently labeled oligonucleotide uptake.

Abstract

Oligonucleotide-based therapeutics represent a rapidly advancing class of drugs with significant potential for treating cardiovascular diseases; however, achieving efficient delivery to cardiac tissue remains a critical and unresolved challenge. A key obstacle is the limited availability of robust, physiologically relevant human in vitro models capable of supporting quantitative assessment of oligonucleotide cellular uptake and intracellular distribution. A detailed, step-by-step protocol is presented for generating self-organizing, 3D cardioids from human induced pluripotent stem cells (iPSCs) and applying them as a platform to evaluate the uptake of fluorescently labeled oligonucleotides. The protocol guides users through directed cardiac differentiation in suspension culture by temporally modulating Wnt/β-catenin signaling, enabling sequential specification of iPSCs through the mesoderm, cardiac mesoderm, and cardiomyocyte progenitor stages. Under these conditions, cells spontaneously self-assemble into beating, cavity-containing three-dimensional structures that express canonical cardiomyocyte markers. The resulting cardioids provide a scalable, experimentally tractable platform for imaging-based assessment of oligonucleotide uptake efficiency, supporting the development and optimization of delivery strategies for cardiac applications.

Introduction

Oligonucleotide-based therapeutics, primarily antisense oligonucleotides (ASOs) and small interfering RNAs (siRNAs), are rapidly emerging as a transformative and highly specific approach for the treatment of a broad range of diseases, including both hereditary and acquired conditions^1,2,3. These modalities enable direct modulation of gene expression at the mRNA level through sequence-specific binding to target transcripts, allowing gene silencing or splicing correction of disease-associated genes that are often inaccessible to conventional small-molecule therapies^4,5. As such, oligonucleotide-based gene therapeutics provide a versatile platform for targeting diverse pathogenic mechanisms across multiple disease areas, including cardiovascular disorders⁶. To date, several oligonucleotide-based therapeutics, including ASOs, siRNAs, and splice-switching agents, have been approved for clinical use⁷. Despite this progress, their therapeutic efficacy in both preclinical and clinical settings remains constrained by inefficient intracellular delivery⁸. Current estimates indicate that less than 1% of internalized oligonucleotides escape endosomal compartments to reach the cytosol and engage their target RNA, thereby limiting functional activity^9,10. This limitation is particularly pronounced in cardiac tissue, where uptake and intracellular trafficking are poorly understood.

Historically, preclinical evaluation of cardiovascular therapeutics has relied heavily on animal models, which remain indispensable for assessing pharmacokinetics, biodistribution, efficacy, and safety in vivo. However, important interspecies differences in cardiac physiology, gene regulation, and cellular responses often limit their ability to accurately predict therapeutic efficacy and toxicity in humans. These limitations have driven the development of human-based in vitro models with improved physiological relevance. Among these, human induced pluripotent stem cell (iPSC) technology has revolutionized cardiovascular disease modeling by providing a virtually unlimited source of cells that can be differentiated into cardiovascular lineages. Human iPSC-derived cardiomyocytes (iPSC-CMs) recapitulate many of the molecular, structural, and functional properties of the human myocardium and have become a widely adopted platform for investigating disease mechanisms, drug responses, and emerging therapeutic modalities¹¹. However, conventional two-dimensional cultures of human iPSC-CMs have important limitations. Most differentiation protocols generate cardiomyocytes that resemble fetal rather than adult cells, displaying immature structural, electrophysiological, metabolic, and contractile properties¹².

Moreover, monolayer cultures lack the three-dimensional (3D) tissue architecture, multicellular interactions, and extracellular matrix composition that influence cardiomyocyte function in vivo. To address these challenges, increasingly sophisticated 3D cardiac models have been developed¹³. Among these, engineered heart tissues (EHTs), generated by embedding iPSC-CMs within biomaterial scaffolds and applying mechanical and electrical stimulation, promote structural and functional maturation while more closely reproducing the biomechanical environment of the native myocardium¹³. Although EHTs provide highly physiologically relevant platforms for disease modeling and drug testing, their generation typically requires specialized bioengineering expertise and dedicated equipment. At the other end of the spectrum, ex vivo cardiac tissue slices most faithfully preserve native tissue architecture and cellular composition but depend on access to primary cardiac tissue, have limited viability in culture, and are not readily scalable.

More recently, cardiac organoids have emerged as accessible 3D model systems that recapitulate key aspects of cardiac tissue organization and function¹⁴. Cardioids are self-assembling cardiac organoids composed predominantly of cardiomyocytes that undergo intrinsic specification, spatial patterning, and morphogenesis to form chamber-like, cavity-containing structures^15,16. Methods to form cardioids typically rely on an initial activation of Wnt signaling to induce mesoderm specification, followed by subsequent inhibition to promote cardiac lineage commitment. Cardioids can be generated in either static or stirred suspension cultures, which support scalable production and enhance tissue self-organization^{15,16,17,18,19}. More complex multi-lineage cardiac organoids incorporating endothelial, epicardial, or other cell types further increase physiological relevance but require more elaborate differentiation protocols, longer culture periods, and extensive optimization^13,14.

A simple, reproducible, and efficient protocol is presented for generating self-organizing cardiomyocyte-based cardioids from human iPSCs and applying them to evaluate the uptake of fluorescently labeled oligonucleotides. The protocol should prove valuable for researchers seeking to optimize oligonucleotide chemistry and delivery strategies for cardiovascular applications. All the materials used in this study are listed in the Table of Materials.

Protocol

1. Culture and preparation of iPSCs for differentiation

CAUTION: Perform all open handling of cells and media in a certified Class II biosafety cabinet. Wear a lab coat and gloves. Decontaminate work surfaces and liquid waste with an appropriate disinfectant.

1. Maintain iPSCs.
   1. Coat culture plates with an appropriate basement-membrane matrix according to the manufacturer’s instructions.
   2. Feed iPSCs with a defined, feeder-free pluripotent stem cell medium and replace medium every 24 h.
   3. Maintain cultures below over-confluency and passage cells when they reach 70–80% confluency.
      ​NOTE: Ensure that iPSCs retain a high-quality, undifferentiated morphology and are confirmed to be mycoplasma-free before initiating differentiation.
2. Pre-differentiation passaging.
   1. Thaw iPSCs and expand cells for at least two passages before initiating cardiac differentiation.
   2. Passage cells using an enzyme-free dissociation method, PBS-EDTA, compatible with iPSC maintenance.
      PAUSE POINT: Maintain iPSCs in culture until they reach 70–80% confluency and retain typical morphology without signs of spontaneous differentiation.

2. Preparation of iPSC aggregates in microwell plates

1. Dissociate iPSCs to a single-cell suspension.
   1. Aspirate medium and wash cells once with 1 mL of 1× PBS.
   2. Add cell detachment solution (Accutase) at 500 µL per well of a 6-well plate and incubate at 37 °C for 5–7 min.
   3. Add 1.5 mL of washing medium per well, and resuspend gently to generate a single-cell suspension.
   4. Transfer the suspension to a 15 mL conical tube and centrifuge at 200 × g for 3 min at room temperature.
   5. Aspirate the supernatant and resuspend the cell pellet in feeder-free iPSC medium supplemented with ROCK inhibitor (final 10 µM).
      NOTE: Adjust dissociation incubation time for each iPSC line by checking detachment at 5 min and extending in short increments if required.
2. Count cells and prepare the plating density.
   1. Count viable cells using a standard cell counting method.
   2. Prepare a single-cell suspension at 0.9 × 10⁶ cells/mL for each well of an 800 µm microwell aggregation plate (24-well format, ~300 microwells/well).
      NOTE: Optimize the cell number per well, as aggregate size or differentiation efficiency varies across iPSC lines.
3. Prime the microwell plate and remove bubbles.
   1. Add 500 µL of feeder-free iPSC medium containing ROCK inhibitor (final 10 µM) to each well.
   2. Centrifuge the microwell plate at 300 × g for 3 min at room temperature to remove air bubbles from the microwells.
      ​CRITICAL: Remove all bubbles completely to allow even cell deposition at the bottom of each microwell and support efficient aggregate formation.
4. Seed iPSCs into microwells.
   1. Add 1.0 mL of the prepared single-cell suspension (0.9 × 10⁶ cells/mL) to each well, corresponding to 0.9 × 10⁶ cells per well.
   2. Gently resuspend within the well 3–5 times using a P1000 and avoid introducing air bubbles.
   3. Centrifuge the plate at 200 × g for 3 min at room temperature to pellet cells into the microwells.
   4. Verify under a microscope that the cells have settled into the microwells before placing the plate at 37 °C with 5% CO₂ for 24 h.
      CRITICAL: Handle the plate carefully after centrifugation. Do not agitate or shake the plate, as this may cause cells to lift out of the microwells.

3. Expansion of iPSC aggregates in microwells

1. Carefully aspirate and replace the medium without disturbing or dislodging the aggregates.
   1. Tilt the plate and slowly aspirate approximately 1.0 mL of medium from each well using a P1000 pipette, ensuring that the aggregates remain seated in the microwells.
   2. Slowly aspirate the remaining ~500 µL of medium using a P1000 pipette.
   3. Add 1.5 mL of feeder-free iPSC medium to each well by dispensing it slowly down the side of the well using a serological pipette.
   4. Incubate at 37 °C, 5% CO₂ and repeat the medium change daily.
      CRITICAL: Dispense medium onto the wall of the well and avoid direct flow onto aggregates.

4. Cardiac differentiation of iPSC aggregates (Day 0 to Day 12)

1. Mesoderm induction (Day 0).
   1. Remove medium using the gentle aspiration method described in Steps 3.1.1–3.1.2.
   2. Add 1.5 mL per well of basal medium supplemented with 2% B27 minus insulin and the Wnt/β-catenin pathway activator (CHIR99021, final concentration: 11 µM).
      NOTE: Prepare a 10 mM stock solution of CHIR99021, a small-molecule inhibitor of glycogen synthase kinase-3 (GSK-3). To achieve a final concentration of 11 µM, add 1.65 µL of the stock solution to 1.5 mL of medium.
2. Medium change after induction (Day 1).
   1. At 24 h following addition of CHIR99021, gently aspirate the medium from each well.
   2. Add 1.5 mL per well of fresh basal medium supplemented with 2% B27 minus insulin, then incubate for 48 h.
      CRITICAL: Strictly adhere to the timing of the Wnt modulation steps, as deviations may compromise differentiation efficiency.
      NOTE: Some cell death is expected following exposure to the Wnt/β-catenin pathway activator. If widespread aggregate loss is observed, consider reducing the activator concentration in future experiments.
3. Cardiac commitment via WNT inhibition (Day 3).
   1. Prepare conditioned medium by transferring 760 µL of spent medium from each well to a sterile 50 mL conical tube.
   2. For each well, add 760 µL of fresh basal medium supplemented with 2% B27 minus insulin to the tube containing the pooled conditioned medium, and mix gently.
   3. Add the Wnt pathway inhibitor, IWP4, to the mixed medium to a final concentration of 5 µM and gently invert the tube several times to mix.
   4. Carefully aspirate the remaining medium from each well, then add 1.5 mL per well of conditioned medium supplemented with IWP4.
   5. Incubate at 37 °C with 5% CO₂ for 48 h.
      ​NOTE: Prepare a 1 mM stock solution of the Wnt pathway inhibitor IWP4. To achieve a final concentration of 5 µM, add 7.6 µL of the stock solution per 1.52 mL of total medium (760 µL conditioned medium + 760 µL fresh basal medium) for each well.
4. Medium change in microwells (Day 5).
   1. Replace medium using the gentle aspiration approach described in Steps 3.1.1–3.1.2.
   2. Add 1.5 mL of basal medium supplemented with 2% B27 minus insulin to each well, then incubate at 37 °C with 5% CO₂ for 48 h.
5. Transfer aggregates from microwells to 6-well plates (Day 7).
   1. Carefully aspirate the medium from each well using the gentle aspiration method described in Steps 3.1.1–3.1.2.
   2. Add 1.0 mL of basal medium supplemented with 2% B27 with insulin to each well of the microwell plate. Using a P1000 pipette, gently pipette the medium up and down five times while moving the pipette tip in a circular motion across the well surface to dislodge the aggregates in the microwells and transfer them into suspension.
   3. Transfer the aggregate suspension to a single well of a non-treated 6-well culture plate.
   4. Rinse each well of the microwell plate with an additional 1.0 mL of fresh medium two to three times to recover any remaining aggregates, transferring each rinse to the same well of the 6-well plate.
   5. Incubate at 37 °C with 5% CO₂ for 72 h.
      NOTE: The presence of dead cells and cellular debris after aggregate transfer is expected. Most debris will be removed during the subsequent medium change.
6. Medium change during suspension culture (Day 10).
   1. Gently agitate the plate to dislodge any loosely attached aggregates.
   2. Tilt the plate to allow the aggregates to settle at the bottom of the well by gravity.
   3. Carefully aspirate the spent medium without disturbing or aspirating the aggregates.
   4. Add 2.0 mL of fresh basal medium supplemented with 2% B27 with insulin to each well and incubate at 37 °C with 5% CO₂ for 48–72 h.
      NOTE: Spontaneous beating is typically observed from approximately Day 10 onward.
      PAUSE POINT: Beating cardioids can be maintained in suspension culture for extended maturation by replacing the medium every 2–3 days.

5. Microscopic characterization of cardioids

NOTE: This method provides a workflow for the histological characterization of cardioids using cryosections. The resulting sections can be used for a variety of histological analyses, including hematoxylin and eosin (H&E) staining to assess overall tissue architecture and immunofluorescence labeling to evaluate cardiomyocyte identity.

1. Fixation and cryosectioning.
   1. Fix cardioids in 4% paraformaldehyde overnight at 4 °C.
      CAUTION: Paraformaldehyde is toxic and a probable carcinogen; handle it in a chemical fume hood while wearing appropriate personal protective equipment. Collect all paraformaldehyde-containing waste (fixative and washes) in a clearly labeled container for aldehyde/chemical hazardous waste, and dispose of it in accordance with institutional hazardous-waste procedures.
      SAFETY: Perform all paraformaldehyde fixation and handling in a chemical fume hood while wearing appropriate personal protective equipment.
   2. Wash three times with PBS and cryoprotect the fixed cardioids in 30% (w/v) sucrose in PBS overnight at 4 °C.
   3. Embed the cardioids in gelatin, freeze, and cut 10 µm-thick cryosections onto microscope slides. Store the sections at −20 °C until use.
2. Immunolabeling.
   1. Thaw the sections for 15 min at room temperature.
   2. Wash the sections in PBS at 37 °C for 10 min, twice.
   3. Circle the area of interest on each slide with a hydrophobic pen to create a well that retains the blocking and antibody solutions.
   4. Rehydrate the sections for 2–3 min in permeabilization buffer (PB; PBS containing 0.1% Tween-20) in a Coplin jar.
   5. Remove the slides from PB, drain off the excess solution, and place the slides in a humid chamber.
      CRITICAL: From this point onward, it is critical not to let the sections dry.
   6. Block the sections with 200–300 µL of PB containing 5% fetal bovine serum (FBS) for 90 min at room temperature in a humid chamber.
   7. Remove the blocking solution and incubate the sections overnight at 4 °C with 100 µL of primary antibody (anti-cMYBP-C; 1:50) diluted in blocking solution (PB containing 1% FBS).
   8. Wash the slides three times with PB and incubate for 2 h at room temperature with a green fluorescent anti-mouse secondary antibody (1:200), far-red fluorescent phalloidin (1:400), and DAPI (1 µg/mL).
   9. Wash three times in PB and mount with an antifade mounting medium.
      ​CAUTION: Collect all fluorescent-dye-containing solutions (secondary antibody, phalloidin, and DAPI staining solutions and the corresponding washes) in a container and dispose of them as chemical waste according to institutional guidelines.
3. Image acquisition.
   1. Acquire confocal images using a laser-scanning confocal microscope equipped with an array detector for super-resolution imaging using a 10× and a 63×/1.40 oil-immersion objective.
   2. Excite the far-red fluorophore, green fluorophore, and nuclear stain using 639 nm, 488 nm, and 405 nm laser lines at 4.0%, 4.0%, and 1.0% laser transmission, respectively.
   3. Acquire single optical sections with a pixel size of 0.0353 × 0.0353 µm², unidirectional scanning, scan speed 6, zoom 1.7, line averaging 1, and a pixel dwell time of 0.935 µs per pixel.
   4. Apply array-detector super-resolution image processing using automatic Wiener filtering.

6. Fluorescent oligonucleotide uptake assay

1. Prepare a stock solution of the fluorescently labeled oligonucleotide conjugated to palmitic acid to increase lipophilicity.
   1. Reconstitute the lyophilized fluorescently labeled oligonucleotide in nuclease-free water to a final concentration of 1 mM. Vortex briefly and centrifuge at 3,000 × g for 30 s to collect the solution at the bottom of the tube.
   2. Prepare single-use aliquots of the 1 mM stock solution and store at −20 °C protected from light.
   3. On the day of the experiment, add 1 µL of the 1 mM stock to 1 mL of cardioid culture medium to obtain a final oligonucleotide concentration of 1 µM. Mix by gentle plate agitation.
2. Incubate cardioids with the fluorescently labeled oligonucleotide (1 µM final concentration) in serum-free medium at 37 °C with 5% CO₂ for 24 h.
3. Wash the cardioids three times with phosphate-buffered saline (PBS) to remove extracellular oligonucleotide.
4. Fix the cardioids in 4% paraformaldehyde for 20 min at room temperature, wash three times with PBS, and continue with fluorescent labeling before imaging.
5. Incubate intact fixed cardioids overnight at 4 °C, protected from light, in PBS containing 5 µg/mL wheat germ agglutinin (WGA) conjugated with a green fluorophore and 1 µg/mL DAPI. Wash three times in PBS and mount with an antifade mounting medium.
   CAUTION: Collect all oligonucleotide- and fluorescent-dye-containing media and washes, including the labeled-oligonucleotide incubation medium from step 6.2 and the WGA/DAPI staining solution, in a separate container and dispose of them as chemical/biohazardous waste according to institutional guidelines.
6. Acquire confocal images of oligonucleotide uptake and distribution.
   1. Acquire confocal images using a laser-scanning confocal microscope equipped with an array detector for super-resolution imaging and a 40× water-immersion objective. Acquire images as 8-section z-stacks with a z-step size of 2 µm, covering a total depth of 16 µm.
   2. Excite the nuclear stain, membrane marker, and far-red fluorescently labeled oligonucleotide using 405 nm, 488 nm, and 639 nm laser lines at 0.9%, 0.2%, and 3.0% laser transmission, respectively.
   3. Acquire images with a voxel size of 0.0449 × 0.0449 × 2 µm³, bidirectional scanning, scan speed 5, line averaging 1, and a pixel dwell time of 1.51 µs per pixel. Apply Airyscan/super-resolution image processing using automatic Wiener filtering.
   4. Keep all acquisition settings (laser power, detector gain, z-step, and voxel size) constant across all z-positions and across all samples to allow valid comparison of oligonucleotide signal between conditions.

Results

Figure 1 summarizes the differentiation workflow and illustrates the characteristic morphological changes expected at each stage of the protocol, from iPSC aggregation to the formation of self-organizing cardioids. Suspension-based microwell aggregation provides a simple and standardized approach for the parallel generation of large numbers of size-controlled cardioids. Successful differentiation is characterized by the generation of uniformly sized aggregates, followed by the development of internal cavities and the onset of spontaneous beating, typically from around Day 10 onward. Together, these morphological features provide practical quality-control checkpoints for assessing the efficiency and reproducibility of cardioid generation.

Cardioid architecture and tissue organization can be assessed by histological analysis of cryosections. As shown in Figure 2, hematoxylin and eosin (H&E) staining of successfully differentiated cardioids reveals well-organized three-dimensional tissues containing one or more internal cavities surrounded by compact cellular layers, consistent with chamber-like organization. Successful cardioids typically exhibit no necrotic regions and minimal structural fragmentation. In contrast, suboptimal differentiations frequently produce irregular or collapsed structures lacking discernible cavities, indicative of impaired tissue self-organization.

Cardiomyocyte differentiation can be further evaluated by immunofluorescence staining of cryosections. As illustrated in Figure 3A–B, successful cardioids display widespread expression of the sarcomeric marker cMYBP-C, organized into highly ordered, periodic striations characteristic of sarcomeres. These striated structures are broadly distributed throughout the tissue and are often aligned along the longitudinal axis of the cardiomyocytes, reflecting efficient cardiac differentiation. By contrast, suboptimal cardioids exhibit weak, patchy, or regionally restricted sarcomeric staining, frequently confined to peripheral regions, with extensive areas lacking detectable signal, indicating inefficient cardiomyocyte differentiation.

Uptake of fluorescently labeled oligonucleotides can be assessed by either live-cell imaging or confocal microscopy of fixed cardioids. As shown in Figure 4, successful uptake is characterized by readily detectable intracellular oligonucleotide-associated fluorescence distributed as discrete puncta, consistent with endosomal localization. WGA delineates cell membranes and DAPI counterstains nuclei, enabling confirmation of intracellular oligonucleotide localization. In addition to providing qualitative visualization of oligonucleotide uptake, the acquired image stacks can be used for quantitative analyses, including fluorescence intensity per punctum, puncta number per cell, and the percentage of oligonucleotide-positive cells. As shown in Figure 5A–C, Z-stack imaging further enables assessment of oligonucleotide distribution throughout the three-dimensional tissue, with fluorescence detected across multiple optical sections extending from the surface into the interior of the cardioid, indicating effective tissue penetration.

Figure 1: Workflow for the generation of human iPSC-derived cardioids. Human induced pluripotent stem cells (iPSCs) are aggregated in microwell plates (Day−2) and subjected to directed cardiac differentiation through temporal modulation of Wnt signaling. Mesoderm induction is initiated on Day 0 by treatment with the Wnt pathway activator CHIR99021 in basal medium supplemented with B27 minus insulin, followed by removal of the activator on Day 1. Cardiac specification is induced on Day 3 by treatment with the Wnt pathway inhibitor IWP4, which is removed on Day 5. On Day 7, the aggregates are transferred to 6-well plates and maintained in basal medium supplemented with B27 containing insulin. By Day 14, self-organizing cardioids are established and are suitable for downstream applications, including oligonucleotide uptake assays. This image is created by the authors using Biorender, and the required copyright license is attached. Representative brightfield images acquired at the indicated time points illustrate the expected morphological progression from uniformly sized iPSC aggregates to cavitated, self-organizing cardioids. Scale bar = 500 µm. Please click here to view a larger version of this figure.

Figure 2: Histological characterization of human iPSC-derived cardioids. Representative hematoxylin and eosin (H&E)-stained cryosections of Day 14 cardioids showing internal cavities surrounded by compact cellular layers, consistent with chamber-like tissue organization. Images were acquired using an imaging Microscope. Quantification of differentiation outcomes showed that 94.0 ± 2.6% of cardioids exhibited spontaneous beating and 60.3 ± 20.6% contained one or more visible internal cavities. Data are presented as mean ± SD from three independent differentiations, with approximately 115 cardioids analyzed per differentiation. Scale bar = 100 µm. Please click here to view a larger version of this figure.

Figure 3: Immunofluorescence characterization of cardiomyocyte differentiation. Representative cryosections of Day 14 cardioids. (A) DAPI staining showing the distribution of cell nuclei. (B) Immunofluorescence image showing nuclei labeled with DAPI (blue) and sarcomeres labeled with an antibody against cardiac myosin-binding protein C (cMYBP-C, green), demonstrating organized sarcomeric structures in differentiated cardiomyocytes. Images were acquired by laser-scanning confocal fluorescence microscopy using a 10×/0.30 dry objective (left) and a 63×/1.40 oil-immersion objective (right). Scale bars = 10 µm. Please click here to view a larger version of this figure.

Figure 4: Imaging of oligonucleotide uptake by cardioids. Confocal images showing the intracellular localization of fluorescently labeled antisense oligonucleotides (ASOs, yellow) in Day 14 cardioids. Cell membranes are labeled with fluorescent wheat germ agglutinin (WGA, magenta), and nuclei are stained with DAPI (blue). ASO-associated fluorescence is detected as discrete intracellular puncta (arrows). The bottom panels show negative control cardioids that were not incubated with fluorescent ASOs. Images in the right panels correspond to higher-magnification views of the boxed regions in the left panels. Representative images were acquired by laser-scanning confocal microscopy with array-detector super-resolution imaging using a 40× water-immersion objective. Scale bars = 10 µm (left) and 5 µm (right). Please click here to view a larger version of this figure.

Figure 5: Assessment of oligonucleotide penetration throughout cardioids. (A) Schematic illustrating the acquisition of optical sections at increasing depths within the cardioid. This image is created by the authors using Biorender, and the required copyright license is attached. (B) Representative confocal images showing the distribution of fluorescently labeled ASOs at z-positions of 0 µm, 6 µm, and 10 µm relative to the tissue surface. Imaging settings were kept constant across all z-positions. Scale bar = 10 µm. (C) Quantification of the percentage of cells containing at least one ASO-associated punctate signal at each imaging depth showed comparable oligonucleotide uptake throughout the tissue, with 83.1 ± 2.5%, 84.9 ± 4.5%, and 84.0 ± 4.5% ASO-positive cells at 0 µm, 6 µm, and 10 µm, respectively. Data are presented as mean ± SD from three cardioids, with three optical sections analyzed per cardioid and a total of 1,305 cells quantified. Representative images were acquired by laser-scanning confocal microscopy with array-detector super-resolution imaging using a 40× water-immersion objective. Please click here to view a larger version of this figure.

Video 1: Time-lapse sequence. A video acquired at 40 frames per second demonstrates spontaneous contractile activity in cardioids. Imaging was performed using a widefield microscope equipped with an sCMOS camera. Please click here to download this Video.

Discussion

Oligonucleotide-based therapeutics have emerged as a powerful class of gene-targeting modalities, supported by decades of advances in oligonucleotide chemistry that have improved target specificity, metabolic stability, and immunogenic tolerability^7,8. Despite the increasing number of clinically approved RNA therapeutics, their broader application remains limited by inefficient intracellular delivery^7,8. Due to their size, charge, and hydrophilicity, oligonucleotides cannot passively diffuse across the plasma membrane and are instead internalized via endocytic pathways^9,10. Following uptake, the vast majority (~99%) of internalized oligonucleotides remain sequestered within endosomal compartments^9,10. Endosomal trafficking proceeds through a maturation process from early to late endosomes and ultimately lysosomes; however, the precise compartments from which oligonucleotides escape to the cytosol remain incompletely understood and may involve multiple endosomal populations⁹. This limited cytosolic release represents a critical rate-limiting barrier and underscores the need for physiologically relevant experimental systems that enable quantitative assessment of oligonucleotide uptake, intracellular trafficking, and endosomal escape²⁰.

The protocol described here was developed to address this need by combining human iPSC-derived cardioids with imaging-based readouts of oligonucleotide uptake. The robustness of this system relies on several critical steps that directly influence reproducibility and experimental outcomes. First, the quality of the starting iPSC cultures is a key determinant of successful cardiac differentiation. The authors recommend periodic karyotype analysis to ensure long-term genomic stability and routine assessment of colony morphology, integrity, and growth dynamics. High-quality cultures are characterized by compact, well-defined colonies with smooth borders, minimal spontaneous differentiation, and consistent proliferation. Cultures displaying irregular colony edges, heterogeneous cell morphology, or areas of differentiation should be discarded and replaced by low-passage frozen stocks prior to initiating differentiation.

A second critical step is the controlled formation of iPSC aggregates using microwell plates, which promotes the generation of uniform, size-defined three-dimensional structures that are essential for reproducible cardiac differentiation²¹. Microwell-based aggregation generates embryoid bodies with defined diameters (optimally 200–230 µm), thereby standardizing cell–cell interactions and exposure to morphogen gradients during temporal modulation of Wnt signaling. In our experience, this size is consistently achieved by seeding approximately 3,000 cells per individual microwell (Step 2.2), a condition that was empirically optimized and validated across multiple human iPSC lines, including commercially available lines and lines reprogrammed in-house. However, because iPSC proliferation rates can vary, the optimal seeding density may require adjustment to maintain the desired aggregate size. Failure to control these parameters can lead to heterogeneous structures, reduced viability, and impaired differentiation efficiency.

Histological analysis, as shown in Figure 2, provides a convenient quality-control assessment of cardioid formation. In particular, the absence of internal cavities or the presence of collapsed or fragmented structures may indicate suboptimal aggregate formation or inadequate temporal control of Wnt signaling during the differentiation steps (Steps 4.1–4.3). In some unsuccessful differentiations, non-contractile cell populations emerge within or around the cardioids. These cells typically appear yellowish by brightfield microscopy and tend to attach to low-attachment culture plates, whereas cardiomyocyte-rich cardioids are generally gray, remain as compact three-dimensional structures, and do not readily attach. If a substantial proportion of the aggregates exhibit this adherent, non-contractile phenotype, the differentiation should be considered suboptimal. In that case, troubleshooting should begin by reviewing iPSC colony morphology before differentiation, confirming aggregate size between Day −1 and Day 0, and, if necessary, re-optimizing the seeding density and the concentrations of CHIR99021 and IWP4 for the specific iPSC line.

Low or absent spontaneous beating by Day 10–12 usually reflects inefficient mesoderm induction or cardiac specification, most commonly due to inappropriate Wnt-modulator concentration or timing, or to excessive cell death following CHIR99021 treatment (Step 4.2). In our experience, titrating CHIR99021 within a range of 9–13 µM for each iPSC line and verifying the concentration and activity of stock solutions can substantially improve differentiation efficiency. Immunofluorescence analysis of sarcomeric protein organization, as shown in Figure 3, provides a convenient quality-control assessment of cardiomyocyte differentiation. Weak or patchy sarcomeric protein staining, particularly when restricted to peripheral regions of the cardioid, generally indicates incomplete or heterogeneous cardiomyocyte differentiation. In these cases, extending the suspension-culture maturation period (Step 4.6 and Pause Point) before fixation and verifying fixation and antibody staining conditions are recommended.

While this cardioid model provides a three-dimensional human cardiac microenvironment that substantially improves upon conventional two-dimensional cardiomyocyte cultures, it does not fully recapitulate the complexity of the adult human heart. Cardioids represent an early stage of human heart development, with cardiomyocytes retaining an immature, fetal-like phenotype¹⁴. Moreover, the tissues lack vascular perfusion, innervation, resident immune cells, and the full chamber-specific architecture of the mature myocardium. Consequently, while this platform is well-suited for investigating cellular uptake and intracellular distribution of oligonucleotides under controlled experimental conditions, caution should be exercised when extrapolating uptake kinetics and delivery mechanisms directly to the in vivo setting. Future incorporation of additional cardiac cell types, vascular networks, and tissue maturation strategies should further enhance the physiological relevance of this platform.

Reliable assessment of oligonucleotide uptake by the cardioids described in this protocol requires optimization and standardization of ASO concentration, incubation time, and imaging acquisition settings, as shown in Figure 4 and Figure 5. The choice of fluorophore is particularly important, as dyes differ in brightness, photostability, and sensitivity to environmental effects. In this study, we used Janelia Fluor 646 (JF646)²², a far-red silicon-rhodamine fluorophore with high brightness and photostability that provides robust signal detection while minimizing tissue autofluorescence in three-dimensional specimens. These properties make it well-suited for imaging oligonucleotide uptake in live or fixed cardioids. Notably, fluorophore conjugation can influence oligonucleotide behavior, including uptake and trafficking, and should therefore be validated for each experimental system.

In its current form, this protocol provides both qualitative and quantitative assessment of oligonucleotide uptake. Confocal imaging of intracellular ASO-associated puncta enables visualization of oligonucleotide localization within cardioids, whereas uptake efficiency is quantified by manually determining the percentage of ASO-positive cells using the Cell Counter plugin in Fiji/ImageJ. Although this approach is well-suited for the moderate-throughput analyses presented here, manual image analysis is labor-intensive and may be subject to observer bias, particularly when applied to large datasets. Recent advances in self-supervised deep learning for three-dimensional microscopy offer promising opportunities to overcome these limitations. AI-based image analysis can accurately segment, identify, and track heterogeneous subcellular structures in complex three-dimensional environments, enabling automated quantification of oligonucleotide uptake and intracellular trafficking, including endosomal dynamics²³. Incorporation of these approaches into future implementations of the protocol should improve throughput, reproducibility, and analytical precision.

Looking forward, the integration of human iPSC-derived cardioids with quantitative imaging, machine learning-based image analysis, and single-particle methodologies offers a powerful framework for investigating the mechanisms that govern oligonucleotide delivery in human cardiac tissue. Combining physiologically relevant three-dimensional cardiac models with advanced analytical tools capable of resolving subcellular organization and dynamics will enable comprehensive quantitative characterization of oligonucleotide uptake, intracellular trafficking, and, with appropriate complementary assays, endosomal escape. Such integrated approaches are expected to facilitate the systematic identification of factors that determine delivery efficiency, support the rational optimization of oligonucleotide chemistries and delivery vehicles, and provide a scalable platform for evaluating emerging oligonucleotide therapeutics in the context of cardiovascular disease.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
4', 6-diamidino-2'-phenylindole, dihydrochloride	Thermo Scientific	62248	DAPI
10× dry objective	Zeiss		EC Plan-Neofluar 10×/0.30
40× water objective	Zeiss		LD C-Apochromat 40×/1.10 Corr
63× oil-immersion objective	Zeiss		Plan-Apochromat 63×/1.40 Oil DIC M27
800-µm microwell aggregation plate	STEMCELL Technologies	34815	AggreWell 800
Andor Neo 5.5 sCMOS camera	Andor		Andor Neo 5.5
Antifade mounting medium	Vector Laboratories	H-1000-10	VECTASHIELD Antifade Mounting Medium
B27 supplement minus insulin, 50×	Gibco	A18956-01	B27 Minus Insulin 50×
B27 supplement with insulin, 50×	Gibco	17504-044	B27 Supplement 50×
Basal Medium	Gibco	21875-034	RPMI 1640
Basement-membrane matrix	Gibco / Corning	A14132-02 / 354230	Geltrex LDEV-Free / Matrigel LDEV-Free
brightfield microscope	EVOS		EVOS XL Core Imaging Microscope
Cell detachment solution	STEMCELL Technologies	07922	ACCUTASE
confocal microscope	ZEISS		ZEISS LSM 980 with Airyscan 2 Microscope
Defined feeder-free iPSC maintenance medium	STEMCELL Technologies	100-0276	mTeSR Plus
DMEM/F12	Gibco	31331-028	DMEM/F12
Non-treated 6-well plate	Avantor (VWR)	734-2777	Untreated 6-well Plate
Phalloidin	Thermo Scientific	A22287	Phalloidin Labeling Probes ALexa 647
Phosphate-buffered saline (1× PBS)	Gibco	14190-144	DPBS 1×
primary antibody c-MYBPC3	Santa Cruz	sc-137181	MYBPC3 Antibody (F-1)
ROCK inhibitor	STEMCELL Technologies	72304	ROCK Inhibitor (Y-27632)
Secondary antibody (green)	Thermo Scientific	A-11017	F(ab')2-Goat anti-Mouse IgG (H+L) Cross-Adsorbed Secondary Antibody, Alexa Fluor 488
Washing medium supplement	Gibco	10828028	KnockOut Serum Replacement
WGA, Alexa Fluor™ 488 conjugate	Invitrogen	W11261	WGA
widefield microscope	Nikon		Nikon Eclipse Ti
WNT pathway inhibitor	STEMCELL Technologies	72552	IWP4
Wnt/β-catenin pathway activator	Sigma	SML1046	CHIR99021