Generation and Single-Cell Transcriptomic Analysis of Hepatocellular Carcinoma Organoids following Drug Treatment

Summary

This protocol outlines a streamlined method for generating hepatocellular carcinoma organoids, applying drug treatment, and performing single-cell RNA sequencing before and after treatment to characterize treatment-associated transcriptional changes.

Abstract

Hepatocellular carcinoma is a major cause of cancer-related mortality worldwide and is characterized by marked intratumoral heterogeneity, which contributes to variable treatment responses, tumor recurrence, and disease progression. A deeper understanding of how tumor cells change at the transcriptional level before and after drug treatment is therefore essential for improving therapeutic strategies. However, practical experimental workflows that connect organoid-based drug treatment with downstream single-cell transcriptomic profiling remain limited. In this study, a standardized workflow for generating hepatocellular carcinoma organoids, applying defined drug treatment, and performing single-cell RNA sequencing on samples collected before and after treatment is developed. The protocol includes organoid revival and expansion, pre-treatment quality assessment, drug exposure, organoid preparation for single-cell dissociation, library construction, and basic comparative analysis of single-cell transcriptomic data. Critical technical precautions are provided to improve reproducibility and sample quality. This workflow enables side-by-side characterization of cellular composition and transcriptional changes associated with drug treatment in hepatocellular carcinoma organoids. The protocol is robust, scalable, and adaptable across different organoid systems, providing a practical platform for investigating treatment-associated gene expression changes at single-cell resolution.

Introduction

Hepatocellular carcinoma (HCC) is the most common primary liver cancer in adults and a major cause of cancer-related deaths worldwide^1,2,3. Its poor clinical outcome is due to significant intratumoral heterogeneity, leading to treatment resistance and recurrence⁴. A recent study indicates that developmental diversity and cell-state plasticity driven by key transcriptional programs such as the FOXM1/CEBPB axis contribute to therapeutic resistance in HCC⁵. These findings highlight the need to understand how tumor cell populations adapt and transition between states under sustained treatment pressure. However, most preclinical models do not adequately capture the dynamic, continuous nature of these treatment-associated state transitions, limiting mechanistic investigation of tumor adaptation and therapeutic failure.

Patient-derived organoids (PDOs) provide a useful platform for this purpose because they preserve important genomic alterations and cellular heterogeneity of the original tumor⁶. HCC organoids can be maintained in long-term culture while retaining key tissue features⁷, making them particularly suitable for treatment-response studies. In addition, organoid-based studies have demonstrated the feasibility of using liver cancer organoids for drug testing and for capturing interpatient and intratumoral differences in therapeutic sensitivity, supporting their value as preclinical models for treatment-response research^8,9. However, most organoid–based drug studies still rely primarily on endpoint viability measurements or focus mainly on bulk phenotypic responses, thereby providing only limited information about heterogeneous cellular responses and treatment-associated transcriptional changes. In parallel, the rapid development of RNA-based therapeutic strategies for HCC further underscores the need for preclinical systems that can accurately predict and mechanistically dissect treatment responses at single-cell resolution¹⁰.

Single-cell RNA sequencing (scRNA-seq) enables high-resolution characterization of cellular composition and transcriptional states within organoid models¹¹. When applied to organoids collected before and after treatment, scRNA-seq can reveal gene-expression changes, shifts in cell populations, and heterogeneous responses that are not captured by bulk assays. This approach therefore represents a practical and powerful strategy for investigating treatment-associated transcriptional remodeling in HCC organoids. However, a practical, reproducible workflow integrating HCC organoid culture, drug treatment, and pre- and post-treatment single-cell RNA sequencing remains lacking. Such a workflow is most effective when organoid integrity is maintained during drug exposure and sufficient viable cells are recovered for downstream single-cell profiling. As with other organoid-based approaches, important limitations include incomplete representation of stromal, vascular, and immune components of the tumor microenvironment, as well as potential selection bias during prolonged in vitro culture.

An integrated experimental workflow is described for generating HCC organoids, applying defined drug treatment, and processing organoid samples for single-cell RNA sequencing before and after treatment. This workflow encompasses organoid revival and expansion, pre-treatment quality assessment, controlled drug exposure, and downstream single-cell RNA sequencing of matched samples. The approach provides a practical framework for comparing, at single-cell resolution, treatment-associated changes in cellular composition and gene-expression programs in HCC organoids.

Protocol

Studies using human tissues were reviewed and approved by the Committees for Ethical Review of Research of Guangzhou Medical University (approval no. GYZL-2024-KY31). All procedures involving human specimens were conducted in accordance with the institutional guidelines for human research. Written informed consent was obtained from all patients for the use of their clinical specimens for medical research. The reagents and equipment used are listed in the Table of Materials.

This workflow includes the recovery and expansion of cryopreserved hepatocellular carcinoma organoids, with minor adaptations for downstream therapeutic perturbation and single-cell RNA sequencing. The aim is to generate organoid cultures of sufficient quality and consistency for subsequent drug treatment experiments¹².

1. Establishment and expansion of HCC patient-derived organoids

1. Thawing and Recovery of Cryopreserved Organoids
   1. Preparation before thawing
      1. Pre-warm a water bath to 37 °C. Pre-equilibrate a 24-well plate in a 37 °C incubator for at least 1 h before plating.
      2. Prepare a complete organoid culture medium in advance and keep the basement membrane extract (BME) on ice until use. The composition of the complete organoid culture medium is provided in Supplementary Table 1.
   2. Thawing of organoids
      1. Remove cryovials containing frozen HCC organoids from liquid nitrogen storage and immediately place them in a 37 °C water bath.
      2. Gently agitate the vial and stop thawing when only a small ice crystal remains.
   3. Removal of cryoprotectant
      1. Gently pipette the thawed suspension once to resuspend the contents.
      2. Transfer the suspension into a 15 mL conical tube containing 3–5 mL of complete organoid culture medium.
      3. Centrifuge at 300 × g for 5 min at 4 °C. Carefully aspirate and discard the supernatant without disturbing the pellet.
2. Embedding and Replating of Organoids
   1. Preparation of recovery mixture
      1. Resuspend the organoid pellet in complete organoid culture medium. The composition of the complete organoid culture medium is provided in Supplementary Table 1.
      2. Mix gently with an equal volume of ice-cold BME to generate the plating mixture.
         NOTE: Maintain the mixture on ice to prevent premature polymerization.
   2. Plating of organoid domes
      1. Dispense 50 µL of the organoid-matrix mixture into each well of a pre-warmed 24-well plate.
      2. Incubate at 37 °C for 30 min to allow matrix solidification.
      3. Add 500 µL of complete organoid culture medium supplemented with 20 µM Y-27632 to each well along the wall of the well.
         NOTE: Limit Y-27632 use to the first 2–3 days after thawing or passaging. Ensure that each dome remains compact and attached to the bottom of the well, without spreading outward.
3. Organoid culture and expansion
   1. Recovery culture conditions
      1. Maintain organoids at 37 °C with 5% CO₂.
      2. Use medium containing 20 µM Y-27632 during the first 2–3 days after thawing.
   2. Medium replacement
      1. Replace culture medium every 3–4 days using gentle aspiration.
      2. Add fresh medium along the wall of each well to avoid disrupting matrix domes.
      3. Maintain the same medium replacement schedule across all wells within the same experiment.
   3. Growth monitoring
      1. Monitor organoid recovery and expansion using a bright-field microscope with consistent imaging settings.
      2. Record representative images at defined time points during recovery and expansion.
      3. Confirm successful recovery when the majority of structures remain intact, organoids exhibit round to oval morphology with smooth borders, progressive enlargement is observed over time, and minimal debris or collapsed structures are present.
      4. Exclude wells containing fewer than 10 intact organoids from downstream quantitative analysis.
   4. Passage and expansion criteria
      1. Passage organoids after stable expansion is observed.
      2. Confirm stable expansion by reproducible increases in organoid size or total area with preserved morphology.
      3. Use consistent imaging interval, magnification, and inclusion criteria across experiments.
      4. Generate representative growth curves based on organoid area or counts to confirm stability.
   5. Criteria for downstream experiments
      1. Use organoid lines that exhibit stable three-dimensional growth, show no contamination or extensive collapse, and provide sufficient yield for parallel experimental conditions.
      2. Maintain cultures for at least 7–10 days after thawing or until completion of recovery and one expansion cycle.
      3. Advance only organoid lines showing consistent morphology across replicate wells.

2. Preparation of organoids for therapeutic perturbation

1. Pre-treatment quality assessment
   1. Morphology-based assessment
      1. Confirm that organoids are round to oval in morphology.
      2. Confirm that organoids display smooth and well-defined borders.
      3. Confirm minimal cellular debris within the dome.
      4. Exclude cultures with extensive fragmentation, darkened lumens, widespread collapse, or obvious contamination.
   2. Expansion-state assessment
      1. Use only organoid cultures that have completed recovery and entered stable expansion.
      2. Confirm progressive enlargement over time.
      3. Confirm consistent morphology across replicate wells.
2. Replating of organoids for drug treatment
   1. Recovery of organoids from matrix
      1. Dissolve organoid domes on ice using 1 mL of ice-cold cell recovery solution per well for 15–20 min.
      2. Pipette gently every 5 min during incubation.
      3. Centrifuge the recovered suspension at 300 × g for 5 min at 4 °C.
      4. Resuspend the pellet in complete organoid culture medium.
   2. Standardization before replating
      1. Standardize the starting organoid burden across wells before drug treatment.
      2. Determine viable cell number after partial dissociation when feasible.
      3. Normalize samples to the same number of input cells per well.
      4. Plate approximately 1,000–5,000 viable cells per well in a 24-well format.
      5. If direct cell counting is not performed, normalize by ensuring a comparable number of morphologically intact organoids per well and a similar size distribution across groups.
      6. Record initial organoid number and baseline morphology for each well before treatment.
      7. Use identical inclusion criteria and plating strategy across all wells within the same experiment.
   3. Replating for drug treatment
      1. Mix recovered organoids with ice-cold basement membrane matrix at a final concentration of 30%–50%.
      2. Dispense 50 µL domes into each well of a 24-well plate.
      3. Incubate at 37 °C for 30 min to allow polymerization.
      4. Add 500 µL of complete organoid culture per well along the wall of the well.
         NOTE: Ensure each dome remains compact and attached to the bottom of the well, without spreading outward. Use identical dome volume and matrix concentration across all wells within the same experiment.
3. Recovery before drug treatment
   1. Recovery after replating
      1. Allow organoids to recover overnight under standard culture conditions.
      2. Initiate treatment only after intact morphology is re-established within replated domes.
   2. Entry criteria for drug treatment
      1. Initiate drug treatment only when organoids remain intact after replating, no widespread collapse or excessive debris is present, and organoid burden is comparable across wells.
      2. Record baseline bright-field images for all experimental groups before treatment.
      3. Maintain identical starting organoid burden, recovery time after replating, and treatment initiation time across all control and treated wells.
4. Safety considerations
   1. Perform all procedures involving human-derived organoids using appropriate personal protective equipment and biosafety practices.
   2. Handle enzymatic dissociation reagents carefully to avoid skin or eye exposure.
   3. Dispose of liquid and solid biological waste according to institutional biosafety regulations.

3. Drug treatment of HCC organoids

1. Drug preparation and treatment
   1. Drug preparation
      1. Prepare lenvatinib stock solution in DMSO according to the manufacturer’s instructions.
      2. Dilute the stock solution in pre-warmed culture medium immediately before use to the predefined working concentration.
      3. Maintain identical final DMSO concentration across all wells (≤ 0.1%).
      4. Prepare sufficient volume to ensure equal final treatment volume in each well.
   2. Drug treatment
      1. Add lenvatinib-containing medium (20 µM) along the wall of each well to avoid disturbing matrix domes.
      2. Include a vehicle control with the same final DMSO concentration.
      3. Incubate organoids under standard culture conditions for the predefined treatment period.
      4. Replace drug-containing medium every 48–72 h for treatment durations longer than 72 h.
      5. Use the same medium replacement schedule across all wells within the experiment.
      6. Record baseline bright-field images before treatment.
      7. Acquire images at fixed intervals during treatment using identical microscope settings, magnification, and field positions.
2. Assessment of treatment response
   1. Morphology-based evaluation
      1. Evaluate treatment response using the following image-based readouts: organoid area growth curve, mean organoid diameter, and surviving organoid count.
      2. Use identical exposure settings, magnification, and analysis thresholds for all images.
      3. Measure organoid area growth curve by calculating total organoid area per well or field at each time point and normalizing to baseline.
      4. Measure mean organoid diameter by determining the diameter of individual intact organoids and calculating the mean value per well.
      5. Determine surviving organoid count by counting morphologically identifiable intact organoids and excluding debris and collapsed fragments.
   2. Optional viability measurement
      1. Perform a three-dimensional luminescence viability assay at the treatment endpoint if additional bulk viability assessment is required.
      2. Follow the manufacturer’s instructions for assay execution.
3. Data analysis
   1. Plot organoid area growth curves over time.
   2. Compare mean organoid diameter between vehicle-treated and lenvatinib-treated groups.
   3. Compare surviving organoid counts between treatment groups.
   4. Apply identical image acquisition schedules, organoid inclusion criteria, and analysis parameters across all groups to ensure reproducibility.

4. Preparation of organoids for single-cell RNA sequencing

1. Selection and recovery of organoids
   1. Pre-dissociation assessment
      1. Select organoid wells that exhibit intact three-dimensional morphology without widespread collapse, contain sufficient material for downstream single-cell capture, and show no obvious contamination.
      2. Collect organoids at matched experimental time points across control and treated samples.
      3. Maintain identical plating density, treatment schedule, and medium replacement conditions across samples.
      4. Record bright-field images of each selected well before dissociation using identical microscope settings, magnification, and field selection criteria.
   2. Removal of matrix
      1. Aspirate culture medium completely from each well.
      2. Wash each well twice with ice-cold PBS to remove residual medium and debris.
      3. Add 1 mL of ice-cold cell recovery solution to each well.
      4. Incubate on ice for 20–30 min.
      5. Pipette gently every 5–7 min during incubation to facilitate matrix dissolution.
      6. Transfer dissolved material into pre-chilled tubes using a wide-bore or cut pipette tip to minimize shear stress.
      7. Proceed to enzymatic dissociation only after the matrix is largely dissolved and minimal visible gel residue remains.
2. Enzymatic dissociation and single-cell suspension preparation
   1. Enzymatic dissociation
      1. Centrifuge the recovered organoid suspension at 300 × g for 5 min at 4 °C.
      2. Remove the supernatant carefully without disturbing the pellet.
      3. Resuspend the pellet in 1 mL of recombinant cell dissociation enzyme.
      4. Incubate at 37 °C for 5–10 min.
      5. Pipette gently 8–10 times every 2–3 min using a wide-bore P1000 tip to promote dissociation.
      6. Stop digestion when most organoid fragments have dispersed into single cells and only small residual clusters remain.
      7. Avoid prolonged digestion to prevent reduced cell viability, stress-associated transcriptional artifacts, and increased ambient RNA contamination.
   2. Quenching and filtration
      1. Add 4 mL of ice-cold PBS containing 2% fetal bovine serum to stop digestion.
      2. Filter the suspension through a 40 µm cell strainer.
      3. Wash once with PBS to remove residual enzyme, aggregates, and debris.
      4. Perform red blood cell lysis if erythrocyte contamination is present, following the manufacturer’s instructions.
   3. Cell preparation and quality control
      1. Count cells using trypan blue exclusion.
      2. Determine cell viability before library preparation.
      3. Use only samples with viability ≥ 85%.
      4. Adjust final cell concentration to 700–1,200 cells/µL.
      5. Exclude samples with excessive debris, abundant dead cells, large visible aggregates, or incomplete dissociation.
      6. Repeat filtration before loading if aggregates are present.
      7. Process control and treated samples under identical conditions, including the dissociation enzyme, digestion time, pipetting frequency, filtration method, cell concentration, and loading strategy.
3. Single-cell library preparation and sequencing
   1. Single-cell capture and library preparation
      1. Load the prepared single-cell suspension onto a droplet-based single-cell capture platform according to the manufacturer’s instructions.
      2. Perform reverse transcription, cDNA amplification, and library construction using the corresponding single-cell 3′ workflow.
   2. Sequencing parameters
      1. Sequence libraries at a depth of at least 50,000 reads per cell.
      2. Load cells to achieve recovery of approximately 6,000–10,000 cells per library.
      3. Record sequencing chemistry, read structure, and instrument settings.
4. Quality criteria for downstream analysis
   1. Ensure that the final preparation has high cell viability, minimal debris, and low aggregate burden.
   2. Confirm suitability for downstream single-cell transcriptomic analysis of cellular heterogeneity and cell-state transitions.

5. Basic processing and quality control of single-cell RNA sequencing data

1. Primary processing
   1. Generation of count matrix
      1. Process raw sequencing data using the recommended pipeline for the selected platform.
      2. Generate a gene-by-cell count matrix.
      3. Record software version, reference genome build, and transcript annotation release.
   2. Initial library-level quality assessment
      1. Summarize sequencing metrics for each library, including total reads, estimated number of recovered cells, median genes per cell, median UMIs per cell, and sequencing saturation (if applicable).
      2. Use identical reference genome and annotation settings across all samples.
2. Cell-level quality control
   1. Filtering of low-quality cells
      1. Evaluate for each cell the number of detected genes, total UMI count, proportion of mitochondrial transcripts, and evidence of doublets or multiplets.
      2. Exclude cells with extremely low gene counts, extremely low UMI counts, or abnormally high mitochondrial transcript fractions.
      3. Define filtering thresholds based on dataset distribution.
      4. Apply identical thresholds across matched treatment groups.
   2. Removal of probable doublets
      1. Identify probable doublets using an appropriate computational method.
      2. Remove doublets using a platform-compatible detection workflow.
      3. Apply the same doublet detection strategy across all samples.
3. Normalization, clustering, and annotation
   1. Data normalization
      1. Normalize the filtered count matrix using a standard single-cell RNA-seq workflow.
      2. Apply identical normalization procedures across all samples.
   2. Dimensionality reduction and clustering
      1. Perform dimensionality reduction.
      2. Conduct unsupervised clustering to identify major transcriptional populations.
      3. Visualize cell states using a low-dimensional embedding method such as UMAP.
   3. Marker gene identification
      1. Identify marker genes using differential expression analysis between clusters.
      2. Assign biological interpretation to major epithelial cell states based on marker expression.
4. Comparative analysis of treatment conditions
   1. Metadata integration
      1. Annotate each cell with treatment condition, patient-derived organoid line, and processing batch.
      2. Maintain consistent metadata structure across all samples.
   2. Analysis of treatment-associated changes
      1. Compare control and treated samples.
      2. Identify treatment-associated transcriptional remodeling.
      3. Identify shifts in cell-state composition.
      4. Identify candidate cell populations associated with therapeutic adaptation.
      5. Interpret results together with morphology-based and viability readouts.
5. Data analysis output
   1. Generate a quality-controlled gene-by-cell count matrix.
   2. Generate low-dimensional visualizations of transcriptionally distinct populations.
   3. Identify marker genes for major clusters.
   4. Maintain consistency across all samples in the reference genome, preprocessing workflow, filtering criteria, clustering strategy, and annotation criteria.
   5. Document any sample-specific deviations from the standard workflow.

Results

Figure 1 outlines the generation and characterization of hepatocellular carcinoma (HCC) organoids. In Figure 1A, patient-derived tumor tissues were used to establish organoid cultures for histological analysis and treatment studies. Figure 1B shows bright-field images of organoids at days 1 and 6 post-recovery, demonstrating survival and expansion under standard three-dimensional culture conditions. At day 1, organoids appear as sma...

Discussion

To ensure reproducibility, efficiency, and high-quality single-cell transcriptomic analysis, several technical steps in this workflow require careful control. Consistency during revival and expansion of patient-derived HCC organoids is essential, because variations in seeding density, basement membrane matrix composition, and culture timing can influence organoid morphology, growth, and downstream treatment responses^13,14. In addition, high cell viability during ...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
[Leu15]-gastrin I human	Merck	G9145
1.5 mL Microtubes	Merck	AXYMCT150LC
1X RBC Lysis Buffer	Thermo Fisher Scientific	00-4333-57
2.0ml Internal Thread Cryogenic Vials	Nest	607001
A8301 (TGFb inhibitor)	Tocris Bioscience	2939
AFP Polyclonal antibody	Proteintech	14550-1-AP
Anti-Albumin antibody [ALB/2144]	abcam	ab236492
Anti-HNF-4-alpha antibody [EPR16885] - ChIP Grade	abcam	ab181604
Anti-Prealbumin antibody [EPR3219]	abcam	ab92469
Anti-SOX9 antibody [EPR14335-78]	abcam	ab185966
B27 Supplement (503), minus vitamin A	Thermo Fisher Scientific	12587010
Cell Recovery Solution, 100 mL	Corning	354253	Reagent used to dissolve basement membrane matrix and recover organoids before replating or downstream processing
CellTiter-Glo® 3D Cell Viability Assay	Promega	G9681
CHIR99021	Merck	SML1046
Chromium Next GEM Single Cell 3' GEM, Library & Gel Bead Kit v3.1	10x Genomics	CG000227
Corning Cell Strainer, 40 μm, Blue, S, IND, 1/50	Corning	431750
Costar 24-well Clear Flat Bottom Ultra-Low Attachment Multiple Well Plates, Individually Wrapped, Sterile	Corning	3473
Cultrex Reduced Growth Factor BME, Type 2 PathClear (BME)	Merck	3533-005-02	Extracellular matrix used to support three-dimensional organoid culture and dome formation
DMSO	Merck	C6164	Solvent used to prepare drug stock solutions and vehicle control
Dulbecco's Modified Eagle Medium/Ham's F-12	Thermo Fisher Scientific	12634028	Advanced DMEM/F-12
E7080 (Lenvatinib)	Selleck	S1164	Tyrosine kinase inhibitor used for organoid treatment and therapeutic perturbation
Fetal Bovine Serum Value FBS	Gibco	A5256701
Forskolin	Tocris Bioscience	1099
GlutaMAX supplement	Thermo Fisher Scientific	35050061
Hematoxylin and Eosin Staining Kit	Beyotime	C0105S
HEPES, 1 M	Thermo Fisher Scientific	15630080
Leica DM6 B Fluorescence Motorized Microscope	Leica	N/A
N2 supplement (1003)	Thermo Fisher Scientific	17502048
N-acetylcysteine	Merck	A0737-5MG
Nicotinamide	Merck	N0636
Nunc 15 mL Conical Sterile Polypropylene Centrifuge Tubes	Thermo Fisher Scientific	339651
Penicillin/streptomycin (10,000 U/mL)	Thermo Fisher Scientific	15140122
Phosphate-Buffered Saline	GENOM BIO	GNM20012	Buffer used for washing organoids and preparing samples for downstream processing
Recombinant human EGF	Peprotech	AF-100-15
Recombinant human FGF10	Peprotech	100-26
Recombinant human HGF	Peprotech	100-39
Recombinant human Noggin	Peprotech	120-10C
Rho kinase inhibitor Y-27632 dihydrochloride	Merck	Y0503	ROCK inhibitor used to enhance organoid survival after thawing or passaging
R-spodin1-conditioned medium	(Broutier et al.)	N/A	Secretion of cell lines
Trypan Blue Stain, 0.4% membrane filtered Prepared in 0.85% saline	Gibco	15250061
TrypLE Express Enzyme (1x), no phenol red	Thermo Fisher Scientific	12604013	Trypsin substitute
Wnt-3a-conditioned medium	(Broutier et al.)	N/A	Secretion of cell lines