Mass Spectrometry Imaging and Histology for the Analysis of Budding Intestinal Organoids

Abstract

Three-dimensional (3D) organoids have been at the forefront of regenerative medicine and cancer biology fields for the past decade. However, the fragile nature of organoids makes their spatial analysis challenging due to their budding structures and composition of single layer of cells. The standard sample preparation approaches can collapse the organoid morphology. Therefore, in this study, we evaluated several approaches to optimize a method compatible with both mass spectrometry imaging (MSI) and immunohistological techniques. Murine intestinal organoids were used to evaluate embedding in gelatin, carboxymethylcellulose (CMC)-gelatin-CMC-sucrose, or hydroxypropyl methylcellulose (HPMC) and polyvinylpyrrolidone (PVP) solutions. Organoids were assessed with and without aldehyde fixation and analyzed for lipid distributions by MSI coupled with hematoxylin and eosin (H&E) staining and immunofluorescence (IF) in consecutive sections from the same sample. While chemical fixation preserves morphology for better histological outcomes, it can lead to suppression of the matrix-assisted laser desorption/ionization (MALDI) lipid signal. By contrast, leaving organoid samples unfixed enhanced MALDI lipid signal. The method that performed best for both MALDI and histological analysis was embedding unfixed samples in HPMC and PVP. This approach allowed assessment of cell proliferation by Ki67 while also identifying putative phosphatidylethanolamine (PE(18:0/18:1)), which was confirmed further by tandem MS approaches. Overall, these protocols will be amenable to multiplexing imaging mass spectrometry analysis with several histological assessments and help advance our understanding of the biological processes that take place in district subsets of cells in budding organoid structures.

Graphical Abstract

INTRODUCTION

Organoids hold a promise for precision and personalized medicine.¹ In the late 1980s, cells from breast epithelium were shown to self-organize in the extracellular matrix and form 3D structures capable of producing milk proteins.² In 2009, Hans Clevers’ research group was able to establish intestinal organoids from single intestinal stem cells, and since then, these epithelial organoids have rapidly grown in popularity and utility due to their ability to capture some of the cellular and structural complexity of organs.³ Organoids can be derived from somatic stem cells, tissue units enriched for adult stem cells, or induced pluripotent stem cells (iPSC),⁴ which differentiate into multiple cell types guided by specialized factors added in growth media. Organoids have complex 3D budding structures frequently comprised of a single-cell layer, which can be difficult to maintain intact during processing for downstream applications.

Organoids derived from intestinal stem cells or whole crypt units enriched in stem cells have revolutionized epithelial biology and regenerative medicine. Budding structures containing crypt like domains enriched for highly proliferating stem and progenitor stem cells frequently develop as intestinal organoids grow and differentiate.^5-7 Additionally, intestinal organoids can produce all differentiated epithelial ligands found in the intestine. Given the analogous architecture to the tissue of origin, these organoids are ideal candidates for studying regeneration and diseases altered by lifestyle factors such as diet.^8-12

Mass spectrometry imaging (MSI) has become an increasingly popular and powerful data visualization tool to aid in answering biological questions.¹³ As an untargeted approach, MSI has the capability of imaging thousands of molecules simultaneously, providing insight into the spatial distributions of lipids, proteins, and small molecules. While first employed by Caprioli et al. to visualize proteins and peptides within tissues,¹⁴ recent innovations allow for single-cell level analysis and the ability to probe intercellular differences.¹⁵ MSI has been successfully applied to the imaging of spheroids,^16-19 which have many similar properties to organoids but are often multilayer and grown from immortalized cancer cell lines.

Utilizing MSI to study 3D organoids has previously been described in the literature. In 2018, Liu et al. first employed MSI to evaluate drug treatments within colorectal cancer organoids.²⁰ Another group presented the use of gelatin microarrays to arrange patient-derived pancreatic cancer organoids within a singular plane before sectioning to aid in finding the small, difficult-to-visualize organoids.²¹ In this study, endogenous lipid signals were detected to indicate the presence and location of the organoids within the microarray but were not used to map out spatial differences within the complicated organoid structure.

The Heeren group was the first to successfully use MSI to study the spatial distribution of endogenous molecules within epithelial organoids.²² They used a gelatin embedding protocol supplemented with cresyl violet staining to embed multiple isolated pancreatic cancer (PDAC) organoids and showed the successful imaging of endogenous molecular species within these organoids. Recently, they followed up this study with a spatial analysis of lipid distributions in small intestine organoids by MALDI-MSI made possible using a cellulose-based M-1 matrix for embedding, which we also found in parallel to be particularly suitable for histological analysis. Using MALDI IHC probes, they were also able to visualize protein distributions in organoids by MALDI-MSI for the first time.²³ Cappuccio et al. detailed MSI of human brain organoids from induced PSCs and compared the utilization of fish and porcine gelatin or water as an embedding material.²⁴ While these studies substantially moved the field of organoid MSI analysis further, the aforementioned protocols require further adaptations to protect and preserve the single cell layer of branched crypt domains in these nontransformed intestinal organoids, while also obtaining reliable measurements for putative metabolites across consecutive organoid cross sections and domains.

In this present study, we have developed a workflow (Supplemental Figure 1) that is amenable to MSI to analyze endogenous lipid species in organoids, while maintaining the structural integrity of the organoid for parallel histological analyses, including H&E staining, and immunofluorescence on consecutive sections of the organoids. Using a combination of accurate mass measurements and tandem mass spectrometry, we identify multiple lipid species with distinct spatial localization. There are multiple steps to this workflow, and each was evaluated and optimized to generate the highest quality data for both MALDI and histology approaches. In all experiments, the organoids were grown in Matrigel as previously described⁸ until they developed differentiated crypt domains (Supplemental Figure 2). Subsequently, they were removed from the Matrigel, and we optimized several approaches, embedding materials, and tested the effects of fixation on MALDI imaging. Embedding unfixed organoids in HPMC-PVP generated the most optimal results for MALDI analysis, histology, and immunofluorescence.

Our goal was to develop a workflow that utilized compatible reagents and solvents that would allow for the analysis and integration of complementary molecular spatial information. While, fixing organoids preserves morphology and enhances the histological images, it leads to dampened signal intensity and spatial resolution for the MALDI analysis. Although the focus of this paper is to describe the optimal workflow best suited for MSI and histology, we have also provided details on the protocols that worked best for one imaging modality over another. By optimizing compatible protocols for MSI and histological analysis from consecutive cross sections of organoids, we hope to investigate metabolic signatures and differences during differentiation and proliferation in distinct domains in the future, and these methods can be applied to organoids from many tissue types. It is worth noting that the organoids used in this study were derived from noncancerous, primary murine intestinal epithelial tissues. As a result, they are more fragile and present greater challenges to work with compared to tumor tissue-derived organoids but possess greater cellularity diversity. Thus, the methods described in this manuscript can also be efficiently applied to cancer-derived or iPSC organoids.

MATERIALS AND METHODS

Organoid Generation.

Organoids were generated from 6 month-old male wild type C57BL/6 mice by previously described isolation methods.²⁵ Briefly, mice were sacrificed, and intestines were flushed with ice-cold 1× PBS. Crypt domains were chemically and mechanically dissociated and embedded in Matrigel for 5–8 days (Supplemental Figure 2). Following sufficient organoid growth and differentiation, organoids were removed from the Matrigel and either fixed with 4% paraformaldehyde (PFA) or left unfixed.

Organoid Harvest and Pre-Embedding Preparation.

Fixed organoids.

Organoids were grown for 5–8 days to allow adequate differentiation (Supplemental Figure 2) with media changes every 3 days. Next, organoids were fixed with 200 μL of 4% paraformaldehyde (PFA) in PBS for 20 min or removed from Matrigel with cell recovery solution for 30 min on ice. PFA dissolves Matrigel; therefore, this rapid fixation method allows for simultaneous fixation and liberation of organoids from Matrigel with minimal morphological destruction. Organoids were collected and spun down gently at 250g for 5 min at 4 °C and washed with ice-cold 1× PBS twice. After the last wash, PBS was replaced by 30% sucrose solution in PBS and organoids and left at 4 °C overnight to prevent the formation of intracellular ice crystals and cell bursting.

Unfixed Organoids.

Unfixed organoids were similarly liberated from Matrigel and washed with ice-cold 1× PBS twice and spun down at 250 g for 5 min at 4 °C. After discarding PBS, unfixed organoids were immediately processed for downstream analysis while maintained at 4 °C as much as possible.

Embedding Methods.

A. “Single” Organoid Pipetting Method.

Gelatin arrays were prepared fresh using 20% w/v porcine gelatin in distilled water and set at 4 °C until ready to use. To recognize the location of embedded organoids, green food dye was utilized to stain the bottom layer of gelatin to aid in finding the layer containing embedded organoids (approximately one drop for 10 mL of prepared gelatin solution). Fixed and unfixed organoids were resuspended in ice-cold PBS in a 24-well plate. Utilizing a wide-orifice 200 μL pipet, single organoid was pipetted per well, and excess PBS was removed. Next, 50 μL of gelatin was placed on top of the organoids and cooled at 4 °C for 30 min. Once solidified, additional 150 μL of gelatin was used to fill the remaining mold. Using a secondary color of gelatin in this step aided in determining where organoids are located within the embedded block. The block was allowed to cool at 4 °C for 30 min before moving to the −20 °C freezer to continue solidifying for an hour prior to moving it to the −80 °C freezer until sectioning.

B. Modified Mold Protocol (Gelatin or HPMC-PVP).

In a secondary method, 50 μL of 20% w/v porcine gelatin solution with green food dye or 7.5% HPMC–2.5% PVP was transferred to the bottom of a 7 mm × 7 mm × 5 mm paraffin base mold and allowed to harden at 4 °C until ready to embed samples. Fixed and unfixed organoids were resuspended in approximately 100 μL of ice-cold 1× PBS and mixed 1:1 with either gelatin warmed at 37 °C or room-temperature HPMC-PVP. The polymer-PBS mixture was then pipetted into the paraffin base mold and allowed to solidify slowly at 4 °C. Once solidified, another 50 μL of colored gelatin or HPMC-PVP was utilized to fill the remaining mold creating a colored sandwich of polymer material. The block was cooled gradually as previously mentioned and stored at −80 °C freezer until sectioning.

For the Gelatin-CMC-sucrose Embedding method, see the Supporting Information and Methods.

Preparation of Poly-d-lysine Slides.

To prepare poly d-lysine solution, a mixture of 750 μL of LC-MS grade water, 750 μL of poly d-lysine, and 1 μL of IGEPAL CO-630 detergent was vortexed thoroughly. To each slide, 20 μL of the prepared mixture was spotted and spread homogeneously across the slide. Samples were dried using a hot plate set to 80 °C for 5 min. Freshly coated slides were rinsed twice in LC-MS grade water and dried in a speed-vac for 10 min. Slides were stored at 4 °C until used for sectioning.

Sectioning of Embedded Samples.

Samples embedded in gelatin, HPMC-PVP, or Gelatin-CMC-Sucrose were sectioned at a chamber and sample head temperature of −30 °C, −20 °C, or −14 °C, respectively. To determine whether organoids were present, test sections were sectioned onto microscope slides and stained with 0.1% cresyl violet solution for 30 s and analyzed using brightfield microscopy. Subsequent samples were thaw-mounted on indium tin oxide (ITO) slides for MALDI analysis or microscopy slides coated with poly D-lysine for further immunohistological staining at 8–10 μm thickness.

Histological Evaluation.

a. Preparation of Previously Unfixed Organoids.

Sections of HPMC-PVP embedded organoids on microscope slides were taken from 80 °C and put on dry ice to prevent thawing. Consecutive slides were saved for MALDI. Organoids were encircled with a hydrophobic pen, and slides were placed on a metal rack on ice. Immediately, 200 μL of 4% PFA in 1× PBS was released on top of organoids and fixed for 10 min. Next, slides were washed twice with chilled 1× PBS and air-dried at RT for 15–30 min.

b. Preparation of Prefixed Organoids.

Microscope slides with either HPMC-PVP or gelatin-CMC-sucrose embedded organoid sections were taken from −80 °C and left at RT for 30 min to air-dry completely. Then, the outside of the organoids was marked with a hydrophobic pen.

Hematoxylin and Eosin Staining.

H&E staining protocol in Table S1 was followed to stain organoids. After the last xylene wash, slides were sealed with 1–2 drops of Cytoseal XYL solution and a coverslip. Slides were left to dry overnight. Images were taken by using a Nikon Eclipse Ti2 inverted microscope.

Immunofluorescent Staining.

Slides were equilibrated at RT and washed twice with 1X PBS for 15 s to dissolve residual mounting material. Next, the slides were baked at 60 °C for 15–30 min. This baking step was skipped when previously unfixed organoids. After baking, slides were washed with 1X PBS and incubated in 0.2% Triton X-100 in 1X PBS (PBST) plus 5% normal donkey serum solution for 1 h in a humidifying chamber at RT. Then, slides were incubated in primary antibody solution (1:400) in PBST with 1% donkey serum for 2 h at RT in a humidifying chamber. Slides were then washed with PBST for 5 min twice and incubated with a secondary antibody at 1:400 concentration in PBST with 1% donkey serum for 2 h at RT in the humidifying chamber at dark. After 2 h, Hoechst solution with a 1:1000 dilution in 1X PBS from a 1 mg/mL stock was prepared and slides were incubated for 5 min at RT at dark. Coverslips were mounted using Vectashield Vibrance mounting medium, and slides were left in the cold room in the dark overnight. Confocal images were taken with an Olympus FV3000 Multiconfocal microscope on the next day.

Sample Preparation for MALDI-MSI Analysis.

Organoids embedded and sectioned utilizing one of the aforementioned techniques were brought to RT under a vacuum before being washed twice with 50 mM ammonium formate and twice with LC-MS grade water prior to drying under a vacuum. Washed slides were coated with 1,5-diaminonapthalene (DAN) matrix at a concentration of 10 mg/mL in 90% acetonitrile using an HTX imaging M5 TM-Sprayer (Chapel Hill, NC). The matrix was applied at 70 °C for two passes over the samples. The flow rate of the matrix was 0.12 mL/min at a velocity of 800 mm/min, track spacing of 2 mm, pressure of 10 psi, gas flow rate of 3 L/min, and nozzle height of 40 mm. Samples were analyzed immediately after application of the MALDI matrix.

MALDI-MSI.

MALDI-MSI spectra were acquired using a timsTOF fleX MALDI-2 mass spectrometer (Bruker Daltonics, Bremen, Germany) in the negative ion mode from 100–1600 Da. Transfer settings were 200 V peak-to-peak (Vpp; funnel 1 RF), 200 Vpp (funnel 2 RF), and 200 Vpp (multipole RF). Focus pretime-of-flight (TOF) transfer time was set at 90 μs and prepulse storage at 10 μs. The quadrupole ion energy was 5 eV with a low mass of m/z 100. Collision cell energy was 10 eV with the collision RF set to 2100 Vpp. All the spectra were recorded using a 1 kHz laser repetition rate with 500 laser shots accumulated at each pixel with MALDI-2 on and an interlaser pulse delay of 10 μs. The smart beam was set to a single focused beam at 55% power with a 0% power boost at a scan range of 5 μm × 5 μm. External calibration was completed using an Agilent Tuning Mix for ESI-TOF.

Tandem Mass Spectrometry Analysis.

Organoids were washed in 1× PBS, resuspended in LC-MS grade water, and centrifuged at 5000g for 5 min. The supernatant was removed, and the resulting organoid pellet was combined 1:1 by volume with the MALDI matrix (10 mg/mL DAN, 90% ACN in water). Five microliters of this lysate-matrix mixture was spotted on a MALDI target plate. MS/MS analysis was carried out on a Bruker timsTOF fleX MALDI-2 mass spectrometer in negative mode, set to acquire spectra from m/z 100 to 1000 for MS1. The laser power was set at 35% with a diameter of 20 μm. Electronic parameters were set as follows: 50 V MALDI plate offset, 200 Vpp Funnel 2 RF, 200 Vpp Multipole RF, 1000 Vpp Collision RF, 100 μs transfer time, 10 μs prepulse storage. Precursor ions of interest were isolated using a collision energy (CE) of 10 eV in the collision cell and subsequently fragmented by ramping the CE by increments of 5 eV. The CE needed to fragment each precursor ion varied by species, but values ranged from 35 to 70 eV.

Data Analysis.

After the acquisition, data were imported and analyzed within the SCiLS laboratory (version 2023b). Data was normalized utilizing the total ion chromatogram. SCiLS lab was utilized to convert data files to.imzML and.ibd using no normalization and the complete spectrum for Metaspace analysis (https://metaspace2020.eu).²⁶

RESULTS

While the primary goal of this study was to find one embedding approach that worked well for multiple imaging applications, there were notable results from the different media that warrant discussion. During sectioning, it was imperative to utilize cresyl-violet staining for quick analysis to detect the overall presence of organoid structures and determine if they were intact or burst following the freezing. While fixation of the organoids prior to embedding improved organoid morphology and increased the number of intact organoids within sections, methodology for the use of unfixed organoids was also completed for the HPMC-PVP embedding method to evaluate the feasibility for other approaches like proteomics. It is important to note that preferred embedding solutions for organoids, such as O.C.T., are incompatible with MSI applications due to signal interference.

Gelatin Embedding.

The first methodology aimed to adapt previous methods utilizing porcine gelatin (20% w/v) to allow for multimodal imaging within a single plane. On first attempts, it was observed that fast freezing of the organoid cultures tended to result in more burst structures potentially due to intracellular ice crystal formation. To alleviate this, we attempted to mimic the freezing down of cell stocks, which maintain single cells for future analyses. Therefore, gelatin was allowed to set slowly at 4 °C to maintain the cellular architecture and budding domains. Furthermore, gentle pipetting of the organoids into wells using wide-bore pipettes maintained the organoid integrity while aligning organoids within a single plane for both fixed and unfixed samples. Care must be taken when pipetting, as over time, the gelatin will begin to degrade if the PBS warms up. As can be seen in Figure 1, the putative lipid signal at 887.55 Da demonstrates the maintenance of budding crypt domains within MSI and matching serial sections of H&E staining. Tandem mass spectrometry experiments were used to identify this species in unfixed organoid lysates as phosphatidylinositol (PI(18:0_20:3) or PI(20:3_18:0), based on detection of the RCOO⁻ ion (Figure 1D).

Figure 1.

Identification of lipid species in intestinal organoids. (A) H&E image of PFA fixed gelatin organoids. (B) Ion image showing the distribution of PI(18:0_20:3), m/z 887.55, in fixed gelatin organoids (scale bar is 100 μm). (C) Lipid portion of the MALDI mass spectrum of unfixed organoid lysate. The lipid at m/z 887.55 was selected for identification by MS/MS analysis and its tandem mass spectrum is shown in panel (D) with identifying fragments. m/z 283.261, (18:0) RCOO⁻ ion; m/z 305.246, (20:3) RCOO⁻ ion; m/z 419.252, neutral loss of (20:3) acyl chain and inositol; m/z 581.308, neutral loss of (20:3) acyl chain; m/z 887.557, parent ion.

While receptive to H&E staining, gelatin produces a strong autofluorescent signal, which interferes with IF staining. Gelatin can be removed from sections by placing slides in 37 °C water or PBS for 5 min, but the structure of unfixed organoids may become compromised during this process. Fixing the gelatin samples postsectioning without prior removal of the gelatin caused it to form an irreversible gel that was difficult to remove. Due to this, organoids fixed prior to embedding in gelatin were the most manageable for downstream histological procedures. While fixed organoids can be utilized to analyze some biomolecular species by MSI, the fixation process resulted in a decreased observable signal in the mass spectrometer compared to the unfixed samples. Within unfixed samples in gelatin (Supplemental Figure 3), masses specific to the organoid single-cell layer along with the crypt domains (m/z 744.54) and the central region of intestinal organoids (m/z 536.50) could be ascertained. We performed tandem MS analyses from organoid lysates to complement the accurate mass measurements and identify these analytes (Figure 2). We determined that the m/z 744.54 species is PE(18:0_18:1) based on the detection of the (18:1)RCOO⁻ fragment ion (m/z 281.246), the (18:0)RCOO⁻ ion (m/z 283.262) and a fragment corresponding to loss of (18:1) acyl chain as ketene (m/z 480.305) and was detected in the organoid (Figure 2). We were unable to identify the molecular species in the center of the organoids with an m/z value of 536.50. However, other abundant lipid species were identified with this approach, including several phosphatidylinositols (Supplemental Figure 4).

Figure 2.

Tandem mass spectra of selected lipids in organoid lysate with identifying fragments. (A) PE(18:0_18:1) at m/z 744.5460; (B) PI(16:0_18:1) at m/z 835.5347.

Another concern when embedding with gelatin is the need to work with the organoids at warmer temperatures (37 °C or above), which may introduce other complications such as damage to the delicate structures and risk changes to the biological or metabolic states observed at the time of harvest. While gentle gelatin procedures proved successful in obtaining MSI data, the further development of techniques more easily amenable to other imaging modalities pushed forward the development of alternate embedding methods. A Gelatin-CMC-sucrose embedding procedure was also investigated and is detailed in the Supplemental section (Supplemental Figure S5).

HPMC-PVP Embedding.

As an alternative to gelatin and CMC, a mixture of HPMC and PVP produces a hydrogel matrix amenable to lower temperature embedding of samples and has been previously identified to be compatible with a wide range of mass spectrometry imaging modalities.²⁷ Due to HPMC-PVP being viscous at 4 °C and hardening only when on dry ice, the material was not able to easily produce arrays like the gelatin method to ensure samples were within a singular plane. While under extreme care, microarray molds of HPMC-PVP could be generated, when the organoid-HPMC-PVP mixture was added to the molds, they tended to lose their shape. To adapt to this property, aliquots of HPMC-PVP with food dye can be utilized to create a colored “sandwich” in which organoids are within a colorless section as opposed to the colored buffer layers. This configuration assists with the sectioning process to locate the organoids in combination with cresyl violet stains.

The technique is less labor-intensive than working with the single-pipetting gelatin method as a mixture of HPMC-PVP and organoids within PBS can be gently and quickly mixed within an Eppendorf tube prior to placing it into the mold similar to those used by Bakker et al.²² In comparison, while our technique slowly freezes the organoids within HPMC-PVP at 4 °C, the aforementioned technique quickly freezes samples within gelatin using isopentane over dry ice. The slower freezing at 4 °C with this medium provided a larger quantity of intact organoids for subsequent analyses. Both fixed and unfixed organoids could be sectioned and analyzed by MSI while, importantly, maintaining intact budding structures (Figures 3 and 4). As opposed to gelatin, HPMC-PVP is quickly and easily removed from sections during the fixation process prior to subsequent analysis by histological procedures. As observed in Figure 4, unfixed organoids embedded within HPMC-PVP are amenable to stains by Hoechst, E-cadherin, and Ki67 allowing for the identification of highly proliferative regions and general organoid structure. Through colocalization of the images, in future studies, molecular maps generated by MSI can be correlated to morphological features and proteins of interest in specific organoid regions.

Figure 3.

HPMC-PVP embedded budding organoids used for multipronged imaging approaches. Consecutive PFA fixed organoid cryosections were generated and analyzed by complementary imaging methods (A) MALDI-MSI, (B) H&E, and (C) immunofluorescence (IF). Ki67 staining (red) was used to detect proliferating stem and progenitor cell populations in organoid crypt domains; cell structure was shown by e-cadherin staining (green), and nuclei were stained with Hoechst 33342. M/z value of 744.54 Da, identified as PE(18:0/18:1) through MS/MS, was detected in crypt domains.

Figure 4.

HPMC-PVP is a suitable embedding medium for the applications of multipronged imaging approaches with unfixed budding intestinal organoids. Consecutive unfixed organoid cryosections were generated and analyzed by complementary imaging methods (A) MALDI-MSI and (B) immunofluorescence (IF). Ki67 staining was used to detect proliferating stem and progenitor cells, e-cadherin was used to stain cell adhesions and nuclei were stained with Hoechst 33342. (C) Representative mass spectrum in the lipid mass range. m/z value of 744.55 Da, identified as PE(18:0/18:1) through MS/MS, was detected in crypt domains.

CONCLUSIONS

Three-dimensional organoid models provide a unique opportunity for personalized medicine and a high-throughput platform that recapitulates experimentally derived phenotypes in culture. In this study, we describe methodologies that extend spatial analytical pipelines to more complex organoid structures, including the budding and crypt domains of single-cell layer intestinal organoids by MSI, H&E, and IF, expanding upon previous work within the MSI community. After evaluating several combinations of embedding and fixing protocols, we have determined that HPMC-PVP embedding of unfixed organoids preserves both the 3D budding structures of the crypt domains and enables analysis by both MALDI-MSI and histological methods. While some of the protocol steps, for example, fixation, work well for histology, it resulted in a degradation of the MALDI signal. Furthermore, the development of methodologies with and without fixation allows for downstream workflow adaptation for a wide range of biomolecules. While this study primarily focused on the feasibility and signal detection of lipids in organoids with different fixation and embedding methods, we have also detected smaller molecular weight analytes. In future studies, we will expand comparative analysis for a wide range of metabolites of interest following drug treatments or to perform further phenotypic characterization of complex 3D organoid models.