Strategy for Biobanking of Ovarian Cancer Organoids: Addressing the Interpatient Heterogeneity Across Histological Subtypes and Disease Stages

Summary

This protocol offers a systematic framework for the establishment of ovarian cancer organoids from different disease stages and addresses the challenges of patient-specific variability to increase yield and enable robust long-term expansion for subsequent applications. It includes detailed steps for tissue processing, seeding, adjusting media requirements, and immunofluorescence staining.

Abstract

While the establishment of an ovarian cancer biobank from patient-derived organoids along with their clinical background information promises advances in research and patient care, standardization remains a challenge due to the heterogeneity of this lethal malignancy, combined with the inherent complexity of organoid technology. This adaptable protocol provides a systematic framework to realize the full potential of ovarian cancer organoids considering a patient-specific variability of progenitors. By implementing a structured experimental workflow to select optimal culture conditions and seeding methods, with parallel testing of direct 3D seeding versus a 2D/3D route, we obtain, in most cases, robust long-term expanding lines suitable for a broad range of downstream applications.

Notably, the protocol has been tested and proven efficient in a great number of cases (N = 120) of highly heterogeneous starting material, including high-grade and low-grade ovarian cancer and stages of the disease with primary debulking, recurrent disease, and post-neoadjuvant surgical specimens. Within a low Wnt, high BMP exogenous signaling environment, we observed progenitors being differently susceptible to the activation of the Heregulin 1 ß (HERß-1)-pathway, with HERß-1 promoting organoid formation in some while inhibiting it in others. For a subset of the patient's samples, optimal organoid formation and long-term growth necessitate the addition of fibroblast growth factor 10 and R-Spondin 1 to the medium.

Further, we highlight the critical steps of tissue digestion and progenitor isolation and point to examples where brief cultivation in 2D on plastic is beneficial for subsequent organoid formation in the Basement Membrane Extract type 2 matrix. Overall, optimal biobanking requires systematic testing of all main conditions in parallel to identify an adequate growth environment for individual lines. The protocol also describes the handling procedure for efficient embedding, sectioning, and staining to obtain high-resolution images of organoids, which is required for comprehensive phenotyping.

Introduction

Clinical management of patients with epithelial ovarian cancer remains challenging due to its heterogeneous clinical presentation at advanced stages and high recurrence rates¹. Improving our understanding of ovarian cancer development and biological behavior requires research approaches that address the patient-specific variability during the course of the disease, treatment response, and histopathological as well as molecular features².

Biobanking, characterized by the systematic collection and long-term preservation of tumor samples derived from ovarian cancer patients along with their clinical information offers the preservation of a large patient cohort in different disease stages, including tumor samples from primary debulking surgeries, after neoadjuvant chemotherapy and from recurrent disease. It holds valuable potential for advancing cancer research serving as a resource of promising prognostic biomarkers and therapeutic targets³. However, conventional biobanking methods, such as formalin fixation and freezing, are not amenable to conducting functional studies on the original tumor samples due to the loss of viability and the disruption of the native three-dimensional tissue architecture^4,5.

Studies of molecular mechanisms, in oncology and beyond, crucially depend on the use of appropriate experimental models that faithfully reflect the biology of the disease and maintain in vitro properties of the tissue observed in vivo. Patient-derived organoids, based on the preservation of the renewal potential, reproduce in the lab the original structure and function of the epithelium and allow testing in a patient-specific context. Therefore, they have emerged as highly promising tools for cancer research and personalized medicine, bridging the gap between clinical diversity and laboratory research^6,7,8,9. Tailored therapeutic strategies based on individual drug responses of organoid lines and testing of the functional relevance of molecular profiles, can potentially be directly applied to patient care^10,11. The possibility of long-term cultivation including patient-specific characteristics and the collection of relevant prospective clinical data over time holds great promise to identify novel prognostic and predictive factors involved in disease progression and resistance mechanisms^3,9.

Nonetheless, building a biobank that includes organoids from different tumor samples requires a combination of strict adherence to complex methodology and setting up protocols for easy maintenance¹². Process standardization ensures that the biobank can be established and maintained efficiently by trained staff even at high turnover, while at the same time adhering to the highest quality standards¹³. Several studies reported the successful generation of stable ovarian cancer organoid lines corresponding to the mutational and phenotypical profile of the original tumor with varying efficiency rates. Still, routine bio banking remains challenging in practice, particularly for long-term stable growth of lines, which is a prerequisite for large-scale expansion or successful genomic editing.

In particular, the issue of expandability remains vaguely defined in the field as organoids that show slow and limited growth potential are occasionally counted as established lines. As initially demonstrated by Hoffmann et al., a study whose principal findings provided the basis for this further developed protocol, optimal handling of ovarian cancer tissue requires a unique strategy to accommodate heterogeneity¹⁴. Phenotypic characterization of the organoids obtained by this method and close similarity with parental tumor tissue were confirmed by panel DNA sequencing and transcriptomics analysis of mature cultures (4-10 months of cultivation) demonstrating the stability of the model^8,9,12,14.

In contrast to the paracrine environment that regulates the homeostasis in the healthy fallopian tubes, the epithelial layer, which likely yields high-grade serous ovarian cancer (HGSOC), cancer regeneration potential, and organoid formation capacity, is less dependent on exogenous Wnt supplementation. Moreover, active Bone Morphogenetic Protein (BMP) signaling, characterized by the absence of Noggin in organoid medium, proved to be beneficial for the establishment of long-term cultures from ovarian cancer solid tissue deposits^14,15. During systematic biobanking of solid deposits of ovarian cancer, we have confirmed these findings and set up the pipeline, with details outlined in this protocol that ensures sustained long-term expansion in the majority of cases. We find that parallel testing of different media compositions and seeding modalities when working with primary isolates are essential to improve the establishment of long-term stable organoid lines and to increase yields enabling robust propagation and expansion to multi-well formats required for downstream experiments¹⁶.

Furthermore, the purity and quality of the samples collected during surgery are of crucial importance for the translational potential of ovarian cancer organoids in basic research and molecular diagnostics. The complexity of the clinical presentation of HGSOC requires close cooperation between the surgeons, oncologists, and the scientists in the lab to ensure that relevant material is correctly identified, transport conditions are kept constant, and organoid lines are generated with high efficiency representing the most important characteristics of the disease of each patient. This protocol provides a standardized but adaptable framework to capture the full potential of ovarian cancer organoids, considering the heterogeneity that characterizes ovarian cancer^16,17. Notably, this protocol enables reliable biobanking of the broad spectrum of ovarian cancer clinical presentation, including different histological types (high-grade and low-grade ovarian cancer, LGSOC), different deposits from the same patients who exhibit differences in stemness regulation, tissues from surgeries in post neoadjuvant setting, biopsy material, and samples from surgeries in the recurrent phase of disease progression.

Protocol

Tumor tissue specimens from ovarian cancer surgeries were collected and patient-derived organoids were generated in compliance with the Ethics Committee of LMU University (17-471), adhering to the existing applicable EU, national, and local regulations. Each patient involved in the study has consented in written form. When working with fresh tissue samples, Biosafety Level 2 safety permission and Laminar Flow cabinets are required. Given the potentially infectious nature of the tissue samples, which cannot be ruled out due to the lack of routine testing of relevant infectious diseases, it is necessary to ensure that institutional bio-safety regulations are strictly adhered to and that adequate personal protective equipment is available for the personnel conducting the experiments.

1. Preparations

1. Medium preparation
   1. Prepare the 2D and 3D media freshly once per week and store them at 4 °C.
      NOTE: The exact composition of the ingredients for each medium is listed in Table 1. Both ovarian cancer medium 1 (OCM1) and OCM2 can be additionally supplemented by HER1ß (final concentration: 50 ng/mL), resulting in four different conditions: OCM1, OCM1+HER1ß, OCM2, OCM2+HER1ß.
   2. Create stock solutions of the growth factor reagents and keep them at 20 °C and of A83-01 and nicotinamide (storage at +4 °C). Due to the temperature sensitivity of the growth factors, use thawed stock solutions immediately for medium preparation and avoid repeated thawing cycles by creating aliquots.
      NOTE: Media preparation is a critical step that needs to be strictly controlled.
   3. Preparation of RSPO-1 conditioned medium
      1. Thaw HA-R-Spondin1-Fc 293 T cells (storage at -80 °C) quickly at 37 °C. Wash them with 10 mL of basal culture medium, supplemented with 10% fetal calf serum (FCS), henceforth referred to as basal culture medium++.
      2. Resuspend the cell pellet in 12 mL of basal culture medium++ and seed it in a T75 flask.
      3. Split the cells in a ratio of 1:6 with a dissociation reagent containing trypsin (4 min at 37 °C) when the cells reach confluence. Seed them into a new T75 flask.
      4. The next day, add phleomycin D1 (1,25 µL/mL) to the medium.
      5. Split the cells in a ratio of 1:20 with trypsin (4 min at 37 °C) when the cells reach confluence. Seed the cells into multiple T75 flasks (number of flasks according to the aimed final volume) in 30 mL of basal culture medium++ (containing 10 % FCS) supplemented with phleomycin D1.
      6. Start the conditioned media production when the cells reach approximately 50% of confluence: replace the medium with 40 mL of basal culture medium supplemented with 5% FCS without phleomycin D1.
      7. Collect the first supernatant after 3 days. To remove the debris, centrifuge for 10 min at 1,200 × g. Keep the supernatant in a sterile container at 4 °C.
      8. Add new basal culture medium supplemented with 5% FCS to the cells. Collect the second supernatant after 3-4 days. Centrifuge for 10 min at 1,200 × g. Add the second supernatant to the first supernatant.
      9. Strain the conditioned medium by using a 0.2 µm bottle top filter.
      10. For quality control and RSPO1-quantification, add the produced medium to a 293T cell line (293T WntR, 7xTcf-eGFP)¹⁴, stably transduced with GFP-plasmid carrying Wnt reporter¹⁸.
          NOTE: Depending on the quantity and intensity of the GFP signal, the activity of RSPO1 as an agonist of the Wnt pathway is tested by the quantification of the Wnt reporter cell line.
      11. Keep aliquots of 15 mL for immediate use (within 1 week) at 4 °C. For longer storage (6 months) keep the conditioned medium at -20 °C.

2. Initiation of an ovarian cancer organoid culture

1. Primary tissue isolation
   1. Transport fresh and viable tissue preferably immediately after surgery and perform isolation within a maximum of 20 h after collection. Ensure proper transport condition in the tissue collection medium with a transport box equipped with cold packs at approximately 4 °C.
   2. In the lab, prepare the necessary reagents and equipment to facilitate timely sample processing:
      1. Thaw Basement Membrane Extract Type II matrix (BME 2 matrix) on ice (approximately 2 h).
      2. Prepare disposable scalpels, sterilized dissection tools (scissors and tweezers), and 400 µm-sized filter inserts for 50 mL tubes.
      3. Prepare a container with a small amount of liquid nitrogen for snap freezing and a prewarmed water bath at 37 °C.
         NOTE: Work within a cell culture laminar flow hood.
   3. Wash the fresh tissue in a Petri dish thoroughly in Phosphate-Buffered Saline (PBS) (without Ca⁺⁺ and Mg⁺⁺). Inside the Petri dish, fragment the tissue using the disposable scalpels and/or scissors into small, 3-5 mm sized pieces.
   4. Separate the obtained tissue into three sections for the following purposes:
      1. Collect tissue in cryogenic tubes. Transfer the cryogenic tubes into the prepared container filled with liquid nitrogen for shock freezing.
      2. Collect tissue (2-3 mm) into a histological specimen container filled with formalin for fixation (24 h) and embed them in paraffin for staining purposes^19,20.
      3. Further, homogenize the tissue before enzymatic digestion. Maximize the mechanical dissociation with a scalpel and transfer it into a 50 mL tissue tube.
         NOTE: Make sure that the tissue is always soaked in PBS during homogenization to avoid drying out of the tissue.
   5. Combine the homogenized tissue with a digestion mixture containing PBS (without Ca⁺⁺ and Mg⁺⁺), Collagenase I (stock solution 5 U/µL, final concentration 1 U/µL), and a selective ROCK1 and 2 inhibitor (3 µM final concentration) (components listed in Table 1). Use 15 mL of digestion mixture in a 50 mL tube for each 2 cm³ of the tumor.
      NOTE: Limited material (e.g., biopsy samples) requires only small amounts (approximately 2-3 mL) of digestion mixture to ensure enzymatic dissociation.
   6. For enzymatic dissociation, incubate the tube for 1.5 h in a water bath at 37 °C. Conduct sporadic and vigorous vortexing (approximately every 20 min for 10-15 s) to support mechanic dissociation.
   7. After incubation, add 15 mL of cold basal culture medium++ to the tube. Centrifuge 5 min at 300 × g.
   8. After removing the supernatant, add 5 mL of new basal culture medium++. Strain the cell suspension by using a 400 µm filter into a fresh 50 mL tube.
   9. Centrifuge for 5 min at 300 × g. Remove the supernatant and keep the cell pellet.
   10. For removal of erythrocytes, add 5 mL of Red Blood Cell Lysing (RBC) Buffer to the cell pellet. Incubate in a water bath for 5 min at 37 °C.
   11. Add 5 mL of basal culture medium++ for lysis inactivation. Centrifuge for 5 min at 300 × g. Add 3 mL of basal culture medium++ to the cell pellet.
   12. Count viable cells after 1:1 dilution with trypan blue stain solution (e.g., 10 µL of the sample + 10 µL of trypan blue stain solution) in an automatic cell counter or manually in a Neubauer Chamber.
   13. Calculate the number of cells needed for direct 3D seeding. For each tumor deposit, seed into a 48-well format plate at least 2-3 wells per ovarian cancer medium, which corresponds to a total of 8-12 wells in a seeding matrix (OCM1, OCM1+HER1ß, OCM2, OCM2+HER1ß). Calculate 30,000 cells for each well in a droplet of 25 µL of BME Type 2 matrix. Optionally, choose a 24-well format with 50 µL of BME 2 matrix and 50,000 seeded cells per well.
   14. Seed the remaining cells in 2D (see step 2.1.13).
   15. Pipette the amount of basal culture medium++ suspension that contains the total number of needed cells into a new tube and centrifuge for 5 min at 300 × g. Remove the supernatant and keep the cell pellet.
   16. On the ice, add the total amount of cold BME 2 matrix (25 µL for each well in a 48-well format plate; hence, 100 µL for 4 wells) to the pellet and mix well to achieve an even distribution of cells per well by gently pipetting the pellet up and down without creating bubbles.
   17. Seed the cells in droplets of BME 2 matrix (25 µL) to the prewarmed, empty 48-well plate. To achieve even distribution between the wells, gently and slowly move the pipette up and down (3-4x) inside the tube before transferring the BME 2 matrix droplet into each well to prevent cell precipitation during distribution.
   18. After incubating the plate at 37 °C for at least 30 min to consolidate the droplets of the BME 2 matrix, pipette 250 µL of each of the four ovarian cancer media (OCM1, OCM1+HER1ß, OCM2, OCM2+HER1ß) into the corresponding wells.
   19. In parallel to the direct 3D seeding procedure, preserve the remaining isolated cells by starting a 2D culture.
       1. After centrifugation for 5 min at 300 × g, add the pellet to the 2D medium. Select the format (T75 flask or T25 flask), depending on the number of cells available after 3D seeding.
       2. Incubate the flask for 3-5 days at 37 °C for cell adhesion. Keep the same medium for the first 72 h.
       3. When the cells reach confluence, transfer the culture into the BME 2 matrix. Do not expand primary isolates in 2D culture by splitting as long-term propagation on 2D is detrimental to stemness potential.
2. Transfer from 2D culture to 3D culture
   1. In the T75 or T25 flask, wash the monolayer 2x with 5 mL of PBS to detach the cells from the bottom surface. Incubate with 1 mL of trypsin for 10 min at 37 °C.
   2. Add 5 mL of cold basal culture medium++ for inactivation of the dissociation reagent. Centrifuge for 5 min at 300 × g.
   3. Add 3 mL of basal culture medium++ to the pellet. Count the cells as described in steps 2.1.12-2.1.13. Seed for the different media compositions (see Figure 1) as described in steps 2.1.16-2.1.18.
3. Evaluation of organoid growth
   1. Image the wells at least once per week to document organoid growth. Make high-quality comparable images of 2D/3D culture and direct seeding plates by phase contrast microscope. Take care of consistency in magnification and use 4x for an overview of the wells (quantification) and 10x and 20x to document the morphology of organoids.
   2. Organize storage of pictures in folders dedicated to individual lines.
   3. Select the best organoid line for long-term cultivation by comparative assessment of organoid formation at the different seeding modalities (2D followed by 3D seeding versus direct 3D seeding) and different media compositions (OCM1, OCM1+ HER1ß, OCM2, and OCM2+ HER1ß).
      1. On average, 14-21 days after isolation, evaluate the organoid-forming potential (quantity) as well as the size and cellular phenotype of organoids grown under different conditions. Look for the following features: absence of vacuoles, membrane blebbing or loss of adhesion, and rounding up of cells.

3. Long-term organoid cultivation

1. Organoid passaging
   1. Avoid splitting organoids too frequently and during the proliferative phase. Conduct mechanical and enzymatic digestion procedures with intervals of more than 10 days.
   2. To disrupt each BME 2 matrix droplet, add 250 µL (48-well format) or 500 µL (24-well format) of ice-cold basal culture medium++. Ensure complete dissolution of the droplet and pipette intermittently up and down and scrape the bottom of the plate. Transfer the cell suspension into a 15 mL tube and place it on ice. Add 1 mL of ice-cold basal culture medium++.
      NOTE: The removal efficiency of the BME 2 matrix depends strongly on the medium temperature since the best dissolution is achieved below 4 °C.
   3. Pool 2-3 technical replicate wells for even expansion . Centrifuge for 5 min at 300 × g. Inspect the content of the suspension against the source of light to see if the BME 2 gel is still visible. Remove the supernatant and repeat washing with an ice-cold medium if the BME 2 matrix is still visible.
   4. Incubate with 1 mL of trypsin (stored at 20-25 °C) in a water bath at 37 °C for 7-10 min. Vortex sporadically for 10 s to enhance dissociation. Inspect the tube periodically visually to see if dissociation is progressing.
   5. Pull the cell suspension up and down with a syringe through a needle with a size ranging from 23 G to 27 G. Check if large cell clumps are still present and repeat the procedure if necessary.
      NOTE: Fragmentation through a syringe is optional, and needle size depends on the required dissociation efficacy. A homogeneous cell suspension leads to even distribution among the wells.
   6. Add 1 mL of cold basal culture medium++ to inactivate the reaction.
   7. OPTIONAL: Count the cells as described in steps 2.1.12-2.1.13 and decide the splitting ratio to the parental passage (e.g., 1:3 wells).
   8. Centrifuge for 5 min at 300 × g. Remove the supernatant.
   9. Conduct seeding as described in steps 2.1.16-2.1.18 and apply the respective optimal ovarian cancer organoid medium according to the already established long-term culture. Change the ovarian cancer medium every 3-4 days.
   10. In case of disruption of the BME 2 matrix droplet during medium change, consider reseeding without enzymatic digestion. To avoid disruption of the droplet, carefully conduct the medium change procedure without direct pipetting onto the BME 2 matrix droplet. Additionally, avoid leaving an excessive amount of residual basal culture medium++ on the pellet before dilution with the BME 2 matrix at step 2.1.16 as it might affect the fragility of the droplets.
2. Cryopreservation of organoids
   NOTE: Conduct cryopreservation of the organoids during the proliferation phase in the first week after passage.
   1. Disrupt the BME 2 matrix droplet with ice-cold basal culture medium++ according to step 3.1.2. After centrifugation and discarding the supernatant as described in step 3.1.3., work on ice and resuspend the cell pellet in 1 mL of ice-cold cryopreservation medium. Transfer the cell suspension to prelabeled 1.8 ml cryogenic tubes.
   2. Store the tubes in a precooled isopropyl alcohol-containing freezing container and place them at -80 °C overnight.
   3. Keep the stock tubes at -80 °C for a maximum of 2 months. Transfer to liquid nitrogen for long-term stable storage.
3. Thawing of organoid stocks
   1. Prewarm 9 mL of basal culture medium++ in a 37 °C water bath in a labeled 15 mL tube.
   2. Transfer the desired stock vial from the storage at -80 °C or to liquid nitrogen into a transport container filled with liquid nitrogen.
   3. Transfer the cryotube to a water bath at 37 °C and gently agitate it until the frozen material near the walls of the tube begins to thaw.
   4. Distribute the organoid suspension slowly into the labeled tube filled with 9 mL of medium. Gently shake the tube. Centrifuge for 5 min at 300 × g and remove the supernatant.
   5. Conduct seeding as described in steps 2.1.16-2.1.18 and apply the respective optimal ovarian cancer organoid medium according to the already determined long-term culture conditions for the individual line.
4. Fixation of organoids
   1. Perform organoid collection as described in steps 3.1.2-3.1.3.
   2. Add 3 mL of 4% paraformaldehyde (PFA) in PBS (pH 7.4) and incubate for 1 h at room temperature.
   3. Wash 2x with 5 mL of PBS, centrifuge for 3 min at 300 × g, and remove the supernatant. Add 4 mL of fresh PBS to the pellet and keep it at 4 °C until embedding.
      NOTE: Fixed organoids are stable and storage at 4 °C is possible for 1 month before embedding.
   4. Heat the histological gel (stored at -20 °C) at 65 °C until liquefied.
   5. Detect the fixed organoids settled and visible at the bottom of the tube. Remove the supernatant. In case of a disrupted cell pellet, conduct centrifugation for 3 min at 300 × g and discard the supernatant PBS.
   6. Suspend the pellet in 100 µL of warm gel by gently pipetting up and down. Transfer the droplet to a piece of sealing film and allow solidification to occur after ~15 min at room temperature.
   7. Move the fixed gel droplet to the tissue paraffin cassette. Conduct standard tissue embedding protocols^19,20.
   8. Cut 5-10 µm thick slices with a microsectioning cutting tool. Transfer the cuts onto histology slides. Dry the slides for 1 h at 65 °C; keep them in a dry place.
5. Immunofluorescence staining of organoid sections
   1. Move the prepared slides through glass trays with the following dilution series: 2 x 15 min clearing agent for histology; 15 min 100% ethanol; 1 min 100% ethanol; 10 min 96% ethanol; 5 min 70% ethanol; 5 min 50% ethanol.
   2. Add PBS and keep 5 min on the shaker at 100 rpm. Repeat the washing 2x.
   3. Add antigen retrieval solution (Tris(hydroxymethyl)aminomethane-Ethylenediaminetetraacetic acid (EDTA)-buffer [pH 9.0] or Citrate [pH 6.0]) to the slides that are placed in the thermostable chamber. In a steamer, keep the container with the slides for 30 min, followed by cool down to room temperature.
   4. Repeat step 3.5.2.
   5. Transfer the chamber containing the slides for 15 min into 1% permeabilization detergent solution (in PBS).
   6. Repeat step 3.5.2.
   7. Contour the location of organoids on the slides with an immunostaining wax pen to maintain the droplet during the following steps.
   8. Add on each slide 100 µL of 10% serum in a dilution medium, depending on the species of the secondary antibody.
   9. Repeat step 3.5.2.
   10. Dilute the primary antibodies in a medium, leading to a final volume of 100 µL. Keep at 4 °C for at least 16 h in an incubation tray/ humidity chamber.
   11. Repeat step 3.5.2.
   12. Dilute the secondary antibodies in a medium, leading to 100 µL droplets, and keep for 2 h at room temperature in an incubation tray.
   13. Repeat step 3.5.2.
   14. Add nuclear counterstain solution DAPI (4',6-Diamidin-2-phenylindol, Dihydrochloride) diluted in a dilution medium 1:1,000. Keep for 10 min at room temperature for incubation.
   15. Repeat step 3.5.2.
   16. Add mounting medium and cover with the coverslip. Secure the slides with transparent nail polish once dry (approximately after 8-12 h).
   17. Image with a confocal microscope or other fluorescence microscope. Make overview pictures as well as detailed images of subcellular structures that capture the main characteristics of organoid morphology.

Representative Results

After initial tissue dissociation, filtration, and counting, cells are seeded in parallel directly in 3D format, as explained above, as well as the suspension in the flask for brief 2D expansion. In some cases, the transient 2D expansion positively influences the organoid formation, and the long-term line is successfully established via this route while comparative parallel 3D seeding can result in growth arrest (Figure 1). For each donor tissue that is processed, the cells are tested according to the media matrix. Following this strategy, our biobank now contains lines representative of each standard growth condition as shown in Figure 2. By stringent implementation of this mini screening platform of testing different media and modes of seeding, we have successfully generated organoid lines from different histological types and stages of disease development of ovarian cancer (Figure 3) of primary high-grade serous, post neoadjuvant interval surgeries and from recurrent disease. Phenotypic characterization of the organoid lines by immunofluorescence staining of main markers in comparison to parental tissue convincingly demonstrates that hallmarks of epithelial tumor compartment are preserved in the organoid model: epithelial architecture and adhesion marked by (EpCAM), lineage identity (PAX8), and typical TP53 point mutation characteristic of HGSOC leading to accumulation in the nucleus (Figure 4).

Some important methodological points are also to be considered for efficient decision-making during organoid biobanking. Organoid growth potential is not only determined by the increase in organoid diameter. Additionally, phenotypic characteristics determine expansion potentials such as color, darkness, and contour integrity. The formation of cytoplasmic stress vacuoles indicates suboptimal conditions. Initial issues with respect to growth and organoid-forming potential might occur due to inappropriate transportation conditions and delay of tissue procession. As a high interindividual variability in expansion potential is observed within the tumor lines, we recommend waiting at least 14 days before making a final decision about the growth potential. If similar growth patterns are initially observed in different media, multiple conditions should be expanded. From our experience, a clear distinction of long-term stable growth potential is often possible only after cultivation of several weeks or months.

Figure 1: Benefits of brief seeding in 2D of primary isolates for subsequent organoid generation. (A) Scheme of the experimental layout showing two-way, parallel seeding strategy: 2D/3D, and direct 3D seeding in four different media. (B) Image of adherent primary isolates before trypsinization and 3D transfer. (C) An example of the primary deposit where parallel seeding revealed the clear advantage of the 2D/3D route as long-term organoid expansion was possible only from progenitors that were initially isolated on plastic. The top left image shows a 3D culture 7 days after isolation at passage 0 (referred to as P0) with insufficient organoid forming after the first passage (referred to as P1) on the bottom left picture. After 2D seeding on plastic (refer to Figure 1B) followed by transfer to a 3D culture, a better organoid forming is already apparent at P0, while long-term expansion potential is confirmed at passage 4 (P4). Scale bar = 200 µm. Please click here to view a larger version of this figure.

Figure 2: Patient-specific medium requirement. Examples of four different long-term stable expanded lines each growing in a different medium. Scale bar = 500 µm. Abbreviations: OCM = ovarian cancer medium; her = heregulin 1β; HGSO = high-grade serous ovarian cancer. Please click here to view a larger version of this figure.

Figure 3: Organoid generation from all stages of the disease. Images of long-term stable expanding lines, from high-grade serous and low-grade serous ovarian cancer in primary disease presentation, from interval surgery (post neoadjuvant chemotherapy), and from recurrent cancer tissue. Scale bar = 500 µm. Abbreviations: HGSOC = high-grade serous ovarian cancer; LGSOC = low-grade serous ovarian cancer; NACT = neoadjuvant chemotherapy. Please click here to view a larger version of this figure.

Figure 4: Close match of phenotype of organoids and parental cancer tissue. Confocal images of immunofluorescence staining of (A) cancer tissue and (B) paired organoid line show very high similarity in staining pattern and level of expression of all markers, EpCAM (green), PAX8 (red), TP53 (magenta). Scale bar = 20 µm. Please click here to view a larger version of this figure.

Table 1: Composition of the media and digestion mixture used in this protocol. Please click here to download this Table.

Discussion

The designed protocol addresses previous challenges of ovarian cancer organoid biobanking with regard to organoid formation and long-term passage potential and ensures the generation of fully expandable lines from the majority of solid tumor deposits. The surgical collection process of tumor samples to be used for organoid generation significantly impacts yield and expansion potential. Tumor tissue samples can be obtained during various procedures, including multi-visceral surgery, diagnostic laparoscopy, or biopsy. The experienced gyneco-oncological surgeon should prioritize obtaining clean tumor samples from peritoneal locations. Notably, it has been observed that macroscopically necrotic areas within large tumor deposits are less suitable for the successful generation of organoid cultures. As sample purity is crucial to reflect molecular and phenotypical characteristics for downstream applications, synchronized efforts of both the clinical and the laboratory teams are required for optimal biobanking, ensuring that the lines are created from the most relevant regions of the tumor.

While this protocol covers variations in growth factor dependencies that we have observed during 2 years of biobanking, it is clear that additional refinement is warranted to further increase the efficacy as we still did not observe any organoid growth in 20% of the cases and the number of lines that demonstrated limited expansion potential suggested suboptimal maintenance of stemness. Although our organoid biobank currently includes only ovarian cancer samples of serous histology (HGSOC and LGSOC), our experience with samples from different epithelial ovarian cancer histologies prior to the structured biobanking program exhibited comparable success rates without specific differences. Notably, the majority of epithelial ovarian cancer subtypes share a similar columnar pseudostratified morphology with tissues that developed from the Mullerian tract (fallopian tube, uterus, cervix) while in contrast, the ovarian epithelium itself originates developmentally from the genital ridge and mesonephros, being covered with a single layer of cuboidal epithelium. Since developmental pathways regulating epithelial homeostasis have a central role in the control of adult stem cell potential, the differences in embryonic origin should be taken into account when developing protocols for bio-banking of organoids. Thus, we believe that progenitors from the surface of the ovary likely require organ-specific culture conditions to sustain long-term stable growth.

Notably, in our lab, the protocol has proven to be successful also for post-neoadjuvant ovarian cancer samples, although the neoadjuvant chemotherapy impacts the cell viability and tissue phenotype. These samples exhibit more debris and extracellular tissue aggregates compared to samples deriving from chemotherapy-naïve tissue of primary debulking surgeries which could affect organoid formation potential. In these cases, we experienced that an appropriate amount of surgical tissue and strict adherence to the recommended transport conditions are crucial for successful organoid generation, as the post-neoadjuvant samples might be more sensitive.

A very interesting phenomenon of the beneficial effect of brief expansion on plastic in 2D of freshly isolated progenitors on subsequent organoid growth, which we have repeatedly observed and therefore included in the standardized experimental procedure, is an important addition to the methodology of cancer organoids research field, which largely relies on direct seeding strategies. It is tempting to speculate that the sensitivity of progenitors after enzymatic digestion and the ability of growth factors to adequately prime them could be the underlying mechanisms behind this difference, which is in some cases a decisive factor if the line can be established. However, more research is needed to establish clear dependencies.

In meeting the challenges posed by high cell adhesion and functional junctions in 3D ovarian cancer organoids, which are difficult to digest, a combination of mechanical and enzymatic dissociation with needle and syringe may help when large clumps remain after trypsinization. Experimenting with different enzymatic conditions could lead to further improvements.

After long-term storage, organoid lines can be thawed and brought into culture according to the experimental design and used for subsequent applications. This enables access at any time to already generated organoid lines for specific research purposes and is particularly interesting for investigating the long-term behavior of certain lines in light of the patient's disease progression. However, cryopreservation of ovarian cancer organoids remains a challenge. Temperature variations during freezing and thawing cycles can negatively affect organoid viability, functionality, and expansion potential. Long-term storage is more stable in liquid nitrogen, where very low temperatures minimize the risk of degradation so that the integrity of the organoids can be maintained. However, ongoing optimization is warranted to establish consistent procedures for freezing, storage, and thawing, to further improve these processes.

Despite the mentioned outstanding issues, this protocol demonstrates the capacity to consistently generate stable organoid lines from solid tumor samples of patients with ovarian cancer. In our laboratory, we processed to date 120 primary ovarian cancer tissue samples achieving success in approximately 50% of cases including a wide range of histological subtypes and stages of the disease with primary debulking, recurrent disease, and postneoadjuvant surgical specimens. By providing a structured framework for the generation of organoids derived from various patient samples, parallel seeding in different media, and implementation of different seeding strategies, this protocol provides an opportunity to assess individual differences in stemness potential, thus providing additional information about tumor biology. To our knowledge, the vast majority of studies usually follow the reverse approach of testing the medium components on a small number of samples and choosing for simplicity mostly one optimal medium. Our systematic testing of the effect of HER1ß, the effect of RSPO1 and FGF10 supplementation, and 2D/3D seeding demonstrates conclusively that ovarian cancer tissue has a degree of interpatient variability in stemness properties, and the optimal medium is indeed patient-specific. Therefore, systematic parallel testing of different media compositions providing different exogenous paracrine signaling environments is essential.

Establishing a living biobank with an extensive panel of organoids derived from various ovarian cancer patients together with their prospectively collected clinical information, serves as a valuable resource for a wide range of research applications and reflects the heterogeneous clinical landscape that is potentially applicable to a larger patient population¹⁷. Long-term cultivation and cryopreservation enable experimental consistency and the repeated accessibility of the same organoid line over time, serving as the base for longitudinal experimental designs with unaltered tumor-line cellular properties.

By retaining epithelial architecture and polarization, the organoid lines are an adequate model to study cell-cell communication and context-dependent cell fate decision-making mediated by paracrine signaling pathways. Embedding of the organoids in the histological gel is a practical and efficient intermediate step to avoid loss of material post-fixation and ensures that downstream processing (embedding in paraffin and sectioning) can be performed in parallel with tissue samples. The loss of organoids during the fixation procedure is a critical issue when the number of organoids is very limited. However, this loss can be counteracted by thoroughly removing the organoids from the extracellular matrix by washing in a cold basal culture medium and by pooling at least two full-grown wells of a 24-well format plate before fixation. The immunofluorescence staining protocol allows high-resolution confocal imaging to correspond molecular and phenotypical characteristics of the ovarian cancer organoids with the parental tumor tissue but is also a valuable tool to study the cellular response to chemical compounds or characterization of genetically modified lines. The method is also suitable for histology staining without any specific modifications. Depending on the purpose of the study, thinner sections cut with a microtome (2-3 µm) should be considered.

Importantly, stable long-term culture is a prerequisite for gene editing experiments and functional assays about drug response and resistance to novel and standard therapies^17,21. In particular, organoids generated from re-biopsies that are performed during disease progression and recurrence, offer the possibility for direct comparative analyses between the therapy-naive original tumor and the newly acquired characteristics observed during relapse. Identification of patient-specific treatment responses and the formation of therapy resistance in vitro advance the field of precision medicine in ovarian cancer research²¹.

Materials

Name	Company	Catalog Number	Comments
100 Sterican 26 G	Braun, Melsungen, Germany	4657683
100 Sterican 27 G	Braun, Melsungen, Germany	4657705
293T HA Rspo1-Fc	R&D systems, Minneapolis, USA	3710-001-01	Alternative: R-Spondin1 expressing Cell line, Sigma-Aldrich, SC111
A-83-01 (TGF-b RI Kinase inhibitor IV)	Merck, Darmstadt, Germany	616454
Advanced DMEM/F-12 Medium	Gibco, Thermo Scientific, Waltham, USA	12634028
Anti-p53 antibody (DO1)	Santa Cruz Biotechnology, Texas, USA	sc-126
Anti-PAX8 antibody	Proteintech, Manchester, UK	10336-1-AP
B-27 Supplement (50x)	Gibco, Thermo Scientific, Waltham, USA	17504-044
Bottle-top vacuum filter 0.2 µm	Corning, Berlin, Germany	430049
CELLSTAR cell culture flask, 175 cm2	Greiner Bio-one, Kremsmünster, Austria	661175
CELLSTAR cell culture flask, 25 cm2	Greiner Bio-one, Kremsmünster, Austria	690160
CELLSTAR cell culture flask, 75 cm2	Greiner Bio-one, Kremsmünster, Austria	658175
Collagenase I	Thermo Scientific, Waltham, USA	17018029
Costar 48-well Clear TC-treated	Corning, Berlin, Germany	3548
Cryo SFM	PromoCell – Human Centered Science, Heidelberg, Germany	C-29912
Cultrex Reduced Growth Factor Basement Membrane Extract, Type 2, Pathclear	R&D systems, Minneapolis, USA	3533-005-02	Alternative: Matrigel, Growth Factor Reduced Basement membrane matrix  Corning, 356231
Cy5 AffiniPure Donkey Anti-Mouse IgG	Jackson Immuno	715-175-151
DAKO  Citrate Buffer, pH 6.0, 10x Antigen Retriever	Sigma-Aldrich, Merck, Darmstadt, Germany	C9999-1000ML
DAPI	Thermo Scientific, Waltham, USA	62248
Donkey anti rabbit Alexa Fluor Plus 555	Thermo Scientific, Waltham, USA	A32794
Donkey anti-Goat IgG Alexa Fluor Plus 488	Thermo Scientific, Waltham, USA	A32814
Dulbecco´s Phosphate-Buffered Saline	Gibco, Thermo Scientific, Waltham, USA	14190-094
Epredia Richard-Allan Scientific HistoGel	Thermo Scientific, Waltham, USA	Epredia HG-4000-012
Falcon 24-well Polystyrene	Corning, Berlin, Germany	351447
Feather scalpel	Pfm medical, Cologne, Germany	200130010
Fetal Bovine Serum	Gibco, Thermo Scientific, Waltham, USA	10270106
Formalin 37% acid free, stabilized	Morphisto, Offenbach am Main, Germany	1019205000
GlutaMAX	Gibco, Thermo Scientific, Waltham, USA	35050038
HEPES (1 M)	Gibco, Thermo Scientific, Waltham, USA	156630080
Human EpCAM/TROP-1 Antibody	R&D systems, Minneapolis, USA	AF960
Human FGF10	Peprotech, NJ, USA	100-26
Human recombinant BMP2	Gibco, Thermo Scientific, Waltham, USA	PHC7146
Human recombinant EGF	Gibco, Thermo Scientific, Waltham, USA	PHG0311L
Human recombinant Heregulin beta-1	Peprotech, NJ, USA	100-03
LAS X core Software	Leica Microsystems		https://webshare.leica-microsystems.com/latest/core/widefield/
Leica TCS SP8 X White Light Laser Confocal Microscope	Leica Microsystems
N-2 Supplement (100x)	Gibco, Thermo Scientific, Waltham, USA	17502-048
Nicotinamide	Sigma-Aldrich, Merck, Darmstadt, Germany	N0636
Omnifix 1 mL	Braun, Melsungen, Germany	3570519
Paraffin
Parafilm	Omnilab, Munich, Germany	5170002
Paraformaldehyd	Morphisto, Offenbach am Main, Germany	1176201000
Pen Strep	Gibco, Thermo Scientific, Waltham, USA	15140-122
Penicillin-Streptomycin (10,000 U/mL)	Sigma-Aldrich, Merck, Darmstadt, Germany	P4333-100
PluriStrainer 400 µm	PluriSelect, Leipzig, Germany	43-50400-01
Primocin	InvivoGen, Toulouse, France	ant-pm-05
Red Blood Cell Lysing Buffer	Sigma-Aldrich, Merck, Darmstadt, Germany	11814389001
Roticlear	Carl Roth, Karlsruhe, Germany	A538.5
Surgipath Paraplast	Leica, Wetzlar, Germany	39602012
Thermo Scientific Nunc Cryovials	Thermo Scientific, Waltham, USA	375418PK
Triton X-100	Sigma-Aldrich, Merck, Darmstadt, Germany	T8787
Trypan Blue Stain	Sigma-Aldrich, Merck, Darmstadt, Germany	T8154
TrypLE Express Enzyme	Gibco, Thermo Scientific, Waltham, USA	12604-013
Tween-20	PanReac AppliChem, Darmstadt, Germany	A4974-0100
Y-27632	TOCRIS biotechne, Wiesbaden, Germany	1254
Zeocin	Invitrogen, Thermo Scientific, Waltham, USA	R25001