Normal and tumor-derived organoids as a drug screening platform for tumor-specific drug vulnerabilities

Summary Patient-derived tumor organoids can be predictive of patient’s treatment responses, and normal tissue-derived organoids allow for drug toxicity testing. Combining both types of organoids therefore enables screening for tumor-specific drug vulnerabilities. Here, we provide a detailed protocol for organoid drug screening using, as proof-of-principle, patient-derived malignant rhabdoid tumor organoids. The protocol can be adapted for drug testing on any tumor and/or normal tissue-derived organoid culture. For complete details on the use and execution of this protocol, please refer to Calandrini et al. (2021).


SUMMARY
Patient-derived tumor organoids can be predictive of patient's treatment responses, and normal tissue-derived organoids allow for drug toxicity testing. Combining both types of organoids therefore enables screening for tumor-specific drug vulnerabilities. Here, we provide a detailed protocol for organoid drug screening using, as proof-of-principle, patient-derived malignant rhabdoid tumor organoids. The protocol can be adapted for drug testing on any tumor and/or normal tissue-derived organoid culture. For complete details on the use and execution of this protocol, please refer to Calandrini et al. (2021).

BEFORE YOU BEGIN
The protocol below describes in detail how to perform drug screens on organoid cultures ( Figure 1). For example, we use patient-derived malignant rhabdoid tumor (MRT) organoid cultures Custers et al., 2021). Yet, the protocol can be easily adapted to any type of tumor and/or normal tissue-derived organoid culture model.

Genetic characterization of tumor organoids before drug screening
Timing: 1-4 weeks 1. Verification tumor organoid cultures: It is crucial to verify the cancerous nature of the established organoid cultures before performing any drug screening experiments. This can be done in different ways, including: a. Targeted PCR amplification and subsequent Sanger sequencing of any known tumor-specific somatic mutations identified in the parental tissue; b. SNP array to confirm the presence of the same copy number alterations in organoid and parental tissue; c. Whole exome/genome sequencing of the organoid culture and subsequent comparison to the genetic profiles of parental tissue.
Note: In case no genetic information is available for the tumor tissue, immunohistochemistry analysis of known tumor markers could be performed on tissue and organoid samples. For instance, in case of MRT, INI1 staining should be performed to confirm lack of expression in tumor cells (Sigauke et al., 2006).
Note: This protocol describes the procedure for drug testing in organoids. As such, we plate intact organoid structures and not single cells. In the specific case of MRT cultures, we plate 500 organoids with an approximate size of 3-6 cells per organoid per well.
5. Calibration tubing: When using an automated system for organoid dispensing, it is crucial to calibrate the equipment before every drug screen to achieve an accurate plating. We refer to the manufacturer's instructions for further details about calibration. 6. Preparation Milli Q, Ethanol 70%, and PBS: Aliquot Milli Q water, Ethanol 70% (vol/vol) and PBS into 50 mL tubes. For each line that will be plated, prepare a total of $500 mL Milli Q, 250 mL Ethanol 70% vol/vol and 50 mL PBS. Store the aliquots at 4 C for a minimum of 12 h.

MATERIALS AND EQUIPMENT
Organoid medium for MRT growth The organoid medium for MRT growth was established by Schutgens et al. (2019), adapted from the medium originally developed for normal kidney tubuloids growth (Schutgens et al., 2019).

STEP-BY-STEP METHOD DETAILS
Single cell dissociation and plating of organoids Day 1 Timing: 1 h (per organoid line) This section describes the experimental steps to expand and plate organoids for drug screening. We use MRT organoids as example, but in principle the same steps can be applied to any type of organoid model.
1. Collect the organoids embedded in a BME droplet with a p1000 pipette using the culture medium in the well and transfer to a 15 mL tube. 2. Pipette up and down to disrupt the BME droplets. 3. Fill up the tube with ice-cold adDMEM/F12 +++.

OPEN ACCESS
4. Centrifuge at 300 3 g for 5 min at 4 C. 5. Remove supernatant. 6. Add pre-warmed (37 C) TrypLE with addition of 10 mM Y-27632 to the pellet. Use 1 mL solution per 200 mL of harvested BME. 7. Incubate for 3-4 min at 37 C. 8. To mechanically dissociate the organoids, pipette up and down for 10-15 times using a fire polished glass pipette. 9. Visually inspect the organoid mixture under a brightfield microscope. At this point, most of the suspension should composed out of single cells.
Optional: if this is not the case, repeat steps 7-9 once more. Troubleshooting 1 10. Fill up the tube with ice-cold adDMEM/F12 +++. 11. Centrifuge at 300 3 g for 5 min at 4 C. 12. Remove supernatant. 13. Measure the pellet size using a p100/p1000 pipette set to a known volume, or make use of falcon tubes filled with standard volumes to infer the volume of the pellet. Add 70% volume of ice-cold BME to the pellet.
Optional: if organoids tend to stick to the plastic surface of the tip when pipetting, precoat the tip by pipetting up and down with adDMEM/F12 +++.
14. Carefully resuspend the pellet in the leftover supernatant without creating air bubbles and plate the mixture with droplets of 15-20 mL volume into prewarmed 12 well or 6 well plates. 15. Place the plate in the 37 C incubator upside down for 20-25 min, to prevent adherence of the cells to the plate bottom. 16. Add pre-warmed (37 C) organoid culture medium to the wells and inspect the culture under a brightfield microscope ( Figure 2A).
Preparation of organoids for dispensing into multi-well drug screening plates Day 4 Timing: 1-2 h (per organoid line) This section describes the harvesting, counting and preparation of the organoid solutions for dispensing into multi-well drug screening plates. We use MRT organoids as example, but the basic principles can be generally applied to any type of organoid model. 17. Visually inspect the organoid cultures using a brightfield microscope to assess their viability and size. As example, MRT organoids should be at the 3-6 cells stage. (Figure 2A) Troubleshooting 2 18. Collect the organoids embedded in a BME droplet with a p1000 pipette using the culture medium in the well and transfer to a 15 mL tube. 19. Fill up the tube with ice-cold adDMEM/F12 +++. 20. Centrifuge at 300 3 g for 5 min at 4 C. 21. Remove supernatant. 22. Repeat washing step by adding 10 mL of ice-cold adDMEM/F12 +++ and resuspend the pellet.
Troubleshooting 3 23. Centrifuge at 300 3 g for 5 min at 4 C. Troubleshooting 4 26. Transfer the filtered solution to a 15 mL tube. 27. Centrifuge at 300 3 g for 5 min at 4 C. 28. Remove supernatant and resuspend pellet in a total volume of 1 mL, making use of ice-cold adD-MEM/F12 +++ to achieve the correct volume.
CRITICAL: The organoid suspension must be thoroughly mixed to obtain a homogeneous solution that is devoid of cell clumps. This step is critical for an as accurate as possible estimation of the number of organoids in solution.
29. Take 1-5 mL of the organoid mix and count the number of organoids using a hemocytometer.
Note: each organoid is count as a single unit.
30. Calculate how many organoids are present in 1 mL. If the number of organoids is sufficient, transfer the required volume to a new tube and proceed with protocol. Otherwise, re-plate the organoids as described in steps 10-16 to further expand the culture.
Note: The total number of organoids needed depends on the number of wells required for the experiment. In the case of a 384-well plate set-up, plate 500 organoids per well (in total $160,000 organoids per 384-well plate).
31. Prepare the organoid solution by mixing organoids, organoid culture medium and 5% vol/vol BME matrix in a 50 mL tube or a sterile flask. Mix by inversion -preventing air bubbles -and put on ice.
Note: The total volume required depends on the number of wells that are used for the experiment. In a 384-well plate set-up, dispense a total volume of 40 mL per well, for a total of $ 13 mL volume per screening plate. Calculate extra volume to compensate for plating inaccuracy/ dead volume needed for dispensing. We recommend calculating at least 10% extra volume/ organoids per screen.
Dispensing organoid mixture into 384-well drug screening plates Day 4 Timing: 30 min (per organoid line) The following steps describe the plating of the organoid solution into 384-well screening plates. In order to plate the organoids in the most accurate manner make use of a reagent dispenser machine (e.g., Multidrop Combi Reagent dispenser (ThermoFisher)). The reagent dispenser should be placed on a cleaned bench or in a flow hood to ensure a sterile environment. Note that all solutions should be ice-cold to prevent polymerization and solidification of the BME in the tubing. 35. Before dispensing, gently invert the organoid-containing solution to ensure that the mixture is as homogenous as possible. 36. Empty the tubing and prime the organoid mixture until the solution has filled every capillary of the tubing.
CRITICAL: Make sure that the tubing is ice-cold before priming the organoid mixture. Proceed promptly to avoid the tubing from reaching 15 C-18 C and the BME to solidify.
37. Proceed with the plating of the organoid mixture to the drug screening plate by dispensing 40 mL of organoid mixture to the selected columns/wells.
CRITICAL: To ensure a homogeneous distribution of the organoids during dispensing, gently swirl the solution while plating. Make sure not to generate air bubbles by swirling too vigorously.
38. In a separate screening plate, dispense the organoid mixture in several (at least 3) wells to measure cell viability at timepoint 0 (T 0 ). ( Figure 2B) This is described in more detail in steps 47-53. 39. Directly after finishing with dispensing the organoid mixture, empty the tubing and prime Milli Q water to start the cleaning procedure.
CRITICAL: do not leave the tubing primed with organoid mixture to avoid clogging due to BME polymerization and/or precipitation of reagents.
40. Continue with the cleaning procedure by flushing the tubing with 100 mL (23 50 mL tubes) Milli Q water, followed by 100 mL of 70% vol/vol Ethanol and again 100 mL of Milli Q water. 41. Visually inspect all plates to confirm that the organoids are plated homogeneously throughout the different wells and plates. 42. Place the screening plates in an incubator at 37 C. Wait at least 4 h before proceeding with the addition of the drugs.
Optional: Use leftover organoids for genomic DNA extraction and subsequent genotyping applications.

OPEN ACCESS
Here, we describe the procedure for cell viability measurements using Cell Titer Glo 3D (Promega). In this protocol, the readout is performed at 2 timepoints, T 0 (day of plating) and T 120h (5 days, endpoint experiment).
47. Make sure the Cell Titer Glo 3D (Promega) and the multi-well screening plates are at 21 C before starting. 48. Visually inspect plates to ensure that the organoids in the wells containing the vehicle control have grown as expected. (Figures 2B and 2C) Troubleshooting 5 49. Prepare and clean the tubing of the multidrop Combi as illustrated in step 32.
CRITICAL: Use solutions stored at 16 C-24 C since temperature fluctuations can interfere with Cell Titer Glo 3D luminescence readout.
50. Use the multidrop combi system to dispense 40 mL of Cell Titer Glo 3D (1:1) in each well.
CRITICAL: During dispensing, protect the tubing from light exposure by covering with aluminum foil. 51. Use a horizontal shaker to agitate the multi-well plate(s) for 5 min at 600 rpm.
CRITICAL: Prior to shaking, wipe the lid of the plate from any condensed water, as this may cause bubble formation and interference with the accuracy of the readout.
52. Incubate the multi-well plate(s) covered with aluminum foil for 25 min at 16 C-24 C. 53. Measure the luminescence signal per well using a suitable reader (e.g., FLUOstar Omega, BMG Labtech).

EXPECTED OUTCOMES
This protocol allows for drug testing of any patient-derived organoid culture in a low-to mediumthroughput manner. If effective, a visible reduction in cell growth compared to vehicle control should be detectable after 120 h incubation with the compounds of interest ( Figure 2C).
After the readout, dose response curve can be generated and parameters such as Area Under the dose response Curve (AUC) and inhibitory half concentration (IC 50 ) values can be calculated ( Figure 2D,E). This set-up can be used to identify tumor-specific and/or patient-specific therapeutic vulnerabilities, as described Custers et al., 2021;Driehuis et al., 2019;Kim et al., 2019).

QUANTIFICATION AND STATISTICAL ANALYSIS
In order to convert raw luminescence data to dose-response curves, one can make use of Microsoft Excel and/or GraphPad Prism software. 6. Transfer the relative cell viability values calculated and corresponding drug concentration -converted in logarithm values -to GraphPad Prism, under the XY table option. 7. Draw dose response fitting curves by selecting New analysis / Nonlinear regression curve / log(inhibitor) vs response -variable slope. Troubleshooting 7 8. In the resulting table, you will find the calculated IC 50 value. 9. Calculate AUC values making use of the option Area under curve under the tab Analyze. 10. To evaluate growth ratios of the organoid lines, compare luminescence values measured at timepoint T 0 with those measured at T 120h , using the following equation: Growh ratio = luminescence value untreated well T120 = luminescence value untreated well T0

LIMITATIONS
In order to perform drug screening in a low to medium-throughput fashion, a significant number of organoids are required. Depending on the type of organoid model, this expansion step can range from several weeks to sometimes even months after establishment of the culture from primary material. As a consequence, drug screening in some cases is only possible several months after culture establishment (Drost and Clevers, 2018;Veninga and Voest, 2021). Although it remains unclear whether prolonged organoid culturing results in the enrichment of specific tumor subclones, we recommend performing any type of drug testing assay on early passage cultures.
Cultures conditions used for growing tumor organoids were often optimized for the growth of normal tissues (Drost and Clevers, 2018). As a consequence, the presence of normal cells in the tumor resection specimen can result in overgrowth by normal organoids (Dijkstra et al., 2020). This can be prevented by using culture conditions that selectively support tumor cells growth. For instance, in the case of TP53mutated tumors, the P53 stabilizing agent Nutlin-3a can be added to the culture medium (Drost et al., 2015;Calandrini et al., 2020). When considering the case of MRT of the kidney -P53 wild type tumors of mesenchymal origin, characterized by a lack of epithelial differentiation Judkins et al., 2004) -MRT organoids can be separated from contaminating normal kidney epithelial cells based on the negativity for epithelial markers, such as EPCAM . When these strategies are not possible, it is important to ensure that the tissue specimen is not contaminated with normal tissue. For instance, part of the obtained specimen can be used for histological examination. In any case, it should be verified that the organoid culture is still representative of the parental (tumor) tissue prior to drug screening (see characterization of tumor organoid cultures). Furthermore, although organoids have been shown to be valid models that recapitulate features of the tissue they were derived from, the lack of vasculature, tumor microenvironment and immune cells are still important limitations to be considered (Ooms et al., 2020;Veninga and Voest, 2021;Drost and Clevers, 2018), which currently still limits the use of organoids for immunotherapy testing.
Here, we describe the use of patient-derived organoids to test drug efficacy in a 120 h (5 day) time window. The here-described timelines and operations can be adjusted based on (1) the growth speed of the cultures and, as a consequence, the potential need for medium refreshing and (2) the stability of the tested compounds. If longer treatment periods are required, adjustments to the protocol -such as medium refreshing, organoid passaging, and compound re-additionhave to be considered.
The Cell Titer Glo 3D assay is used in this protocol to infer the relative cell viability from the metabolic activity of cells via the quantification of ATP molecules. This readout cannot therefore distinguish between a cytostatic (reduction in growth) and a cytotoxic (cell death) effect of the tested compounds, since both would cause a reduction in ATP production compared to an untreated control. Therefore, to further characterize the drug-induced effects, additional validation experiments such as apoptosis and cell proliferation assays should be performed .

TROUBLESHOOTING Problem 1
Intact organoids or pieces of organoids remain after the digestion procedure (step 9).

Potential solution
Carefully increase the incubation time with TrypLE, followed by mechanical disruption (up and down pipetting). Repeat the process for a maximum of 3 times. Stop the procedure as soon as substantial cell lysis (>30%) is detected in the solution. Cell death quantification should be performed via staining with markers/dyes discriminating live and dead cells (e.g., Trypan blue).Note that it is not always possible to fully digest organoids into a single cell suspension without causing significant cell death in the cultures. In such a case, aim for obtaining a suspension that is as homogeneous as possible, performing mechanical and enzymatic disruption until organoids are dissociated into fragments of equal size (< 70 mm). At this point, big organoid fragments still present in solution can be filtered out using a 70 mm strainer. Notably, take the size of the disrupted organoids into consideration when determining the correct timing for organoid harvesting for drug screening (step 17).

Problem 2
Organoids did not reach optimal size for drug screening plating (step 17) because (1) the culture is still composed out of single cells or (2) the structures are too big (diameter > 70 mm).

Potential solution
Wait (>3 days) until the culture has recovered and organoids have reached a sufficient size before harvesting. A longer period of recovery upon single cell digestion could be a sign of a too harsh/ not careful handling of the organoids. Optimize the timing of and method for organoid dissociation.
Reschedule the harvesting of the organoids. It is critical to follow the optimized digestion protocol of the organoid lines strictly and consistently, as it will ensure a predictable growth rate of the cultures and make the results reproducible.

Problem 3
Large amount of BME is present after washing steps (step 22).

Potential solution
When dealing with large amounts of BME (> 800mL), at the moment of harvesting divide the content in to several falcon tubes, so that a maximum of 400 mL BME is present in each tube. If the issue persists, before proceeding with the filtering step, it is recommended to perform a short enzymatic digestion with TrypLE or Cell Recovery Solution.

Potential solution
Loss of organoids during the filtering step can be caused by (1) insufficient digestion of the culture when plating (see problem 2); (2) high content of BME remaining in the organoid solution. When a considerable amount of BME is still visible at step 24 (recognizable as a thick white ring on top of the organoid pellet), wash the organoid pellet once more with ice-cold adDMEM/F12 +++. If after filtering a layer of BME is still present on top of the filter, add 5/10 mL of ice-cold adDMEM/ F12 +++ to the strainer and use the tip of a 10mL pipette to scrape the surface of the strainer. At this point, if part of the solution cannot pass through the filter, use a new strainer.
Problem 5 Insufficient growth in untreated control wells at T 120h (step 48). Potential solution Possible causes include: (1) inadequate growth factor activity in the culture medium which can be solved by making a new batch of medium; (2) limited viability of the organoids at the moment of plating, which can be solved by using less harsh conditions during digestion and plating of the cultures; (3) the growth of the organoids was reduced due to the different plating set-up (5% BME slurry instead of 70% BME), which can be solved by performing a growth curve replicating the screening conditions prior to the screen, to verify changes in growth rate.

Problem 6
Large variation in cell viability values between technical replicates (step 4).

Potential solution
This can be caused by an insufficient number of plated organoids per well and/or by a non-homogeneous dispensing of the organoids. Repeat the screen and dispense more organoids per well. If the organoid cultures show a change in growth speed during the expansion phase, we recommend to replicate the growth curve experiment to assess the best organoid seeding density. To avoid a heterogeneous dispensing of the organoids upon plating, ensure that a homogeneous population of organoids is present at the moment of harvesting. Furthermore, be mindful that the organoid solution should be regularly inverted/mixed prior and during dispensing to avoid sinking of the organoids to the bottom of the tube.

Problem 7
Drug screening results are not reproducible (step 7).

Potential solution
This could be caused by (1) overgrowth of normal cells/ cross-contamination with other lines; (2) low quality/deterioration of drug aliquots. To verify that no cross-contamination has taken place in the organoid cultures, harvest genomic DNA and verify via genotyping/SNP array that the genomic profile is matching those of the early passage cultures and parental tumor tissue. Check the storing conditions and the lifetime/stability of the used compounds; make use of fresh aliquots when repeating the screen.

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Jarno Drost (J.Drost@prinsesmaximacentrum.nl)

Materials availability
This study did not generate new unique reagents.
Data and code availability This paper did not generate new datasets.