Generating Free-floating Normal Human Epithelial-Fibroblast Spheroid Co-Cultures

Summary

Here, we present a simple and very practical but reliable protocol for direct spheroidal co-cultivation of human lung epithelial cells and fibroblasts without relying on a matrix.

Abstract

Three-dimensional (3D) culture systems have become increasingly popular due to their ability to reproduce natural cell properties and architectures, thus mimicking tissue-like structures in vitro. Among these models, the culture of spherical cell aggregates (so-called spheroids) embedded in a semi-solid extracellular matrix (ECM) represents an advanced near-in vivo cell culture model, as it allows for different functional cell states as a result of cell-cell and cell-ECM interactions, as well as oxygen and nutrient gradients. Because spheroids are technically less demanding and relatively inexpensive to obtain, they are frequently used in drug screening and toxicity testing, enabling high-content screening for testing new drugs in the preclinical phase. For this purpose, 3D structures of various tumors are predominantly recreated. At the same time, Matrigel is currently one of the most widely used ECMs and is considered the gold standard in spheroid and organoid cultivation, although this is a 3D matrix based on mouse experiments, the extraction of which- like animal testing in general-is associated with major ethical concerns. Here, we present a matrix-free protocol for direct spheroidal co-cultivation of human bronchial epithelial cells and fibroblasts, which can be considered as an optimized co-cultivation method, especially concerning epithelial-fibroblast communication and interactions within the respective spheroids, without relying on an (animal-based) carrier matrix. By establishing these free-floating epithelial-fibroblastoid spheroid co-cultures as a patient-oriented in vitro platform to model normal (lung) tissue toxicities of cancer therapeutic approaches, not only a reduction in the number of experimental animals but also an adequate and meaningful replacement of corresponding animal experiments can be achieved.

Introduction

Two-dimensional (2D) cell cultures, in which flat monolayer cells are cultured, expanded, and then plated for cell-based research assays, remain an easy, convenient, cost-effective, and widely used method for standard preclinical screening procedures^1,2. To study the interactive (paracrine) crosstalk between various types of cells in vitro, a conditioned medium from one cell type for stimulation of another cell type can be used. Additionally, various co-culture methods emerged, e.g., indirect methods, when co-cultured cells are physically separated using a microporous membrane. These 'Transwell systems' include an insert (the microporous membrane) that is placed in a traditional 2D cell culture plate. One cell type is then seeded and grown on the membrane, sharing the same media as the other type growing on the companion well, thus allowing bidirectional paracrine communications². The two cell types might even be seeded on the two sides of the membrane, forming more of a direct cell-to-cell contact co-culture, which is suitable for investigating barrier-structured tissues^3,4.

However, the cellular responses of these even advanced 2D cell culture systems do not always reflect the results of the corresponding animal experiments and even clinical studies, which is due, among other things, to the lack of direct signaling between the different cell types and thus the lack of in vivo-like cellular organizations. Therefore, three-dimensional (3D) culture systems have become increasingly popular because of the capabilities of natural-like cell characteristics and architectures mimicking tissue-like structures in vitro^5,6. Among these models, the culture of spherical cell aggregates, termed spheroids, embedded in semi-solid extracellular matrices provide a more advanced cell co-culture model where different functional cellular states following cell-cell and cell-ECM interactions as well as gradients of oxygen and nutrition can be achieved⁷. These desired more complex in vivo-like in vitro cultures 3D cultures should be generated and maintained homogeneous concerning cell assembly sizes while promoting cell development and differentiation, while still being effective, rapid, consistent, and convenient^7,8.

Using the natural characteristics of self-aggregation and adhesion, spheroids (mainly tumor spheroids) are usually generated using the hanging drop technique^9,10. As one of the first techniques used to generate embryonic bodies, the hanging drop assay has become a widely used method to generate numerous homogenous spheroids based on the principles of gravity and surface tension prior ECM embedding using Matrigel, the 'gold standard' of extracellular matrices, although being mouse-based. Substantial effort was made to generate spheroids in scaffold-free approaches^8,11. For example, spheroids can be generated using a cell culture bottle with an agitator, thus preventing cell seeding, the simplest method to produce spheroids on a large scale¹².

The most difficult thing in complexing cell cultures and thus in establishing a physiologically valuable system as in vivo-like in vitro alternatives in one's laboratory are complex protocols with many different steps (which still need to be established). For example, generating cellular structures first using a method, harvesting them, and then embedding them in a matrix, especially the use of matrices that are not easy to handle, are critical steps that require more convincing before a method can be used routinely and serially. Here we used an ultra-low attachment plate in a 96 well format, which allows virtually no cell attachment to the culture ware for establishing complex 3D spheroids of lung epithelial cells and fibroblasts. The resulting spheroids are formed highly uniform and rather fast within 24 h, which shortens the experimental timeline. Prolonged cultivation and, thus, reliable cell growth and differentiation are carried out in the same well over the entire period, so that after initial plating only the medium needs to be replaced over the entire cultivation period. Therefore, we present a very simple, rapid, and thus highly practical protocol for generating human lung spheroids that morphologically and functionally replicate the airway epithelium at near-physiological levels. Another advantage is that this protocol can be easily adapted to other cells.

Protocol

All steps are performed under sterile conditions (Biosafety cabinet). For the media compositions used in this study, refer to Table 1.

1. Preparation of experimental material

1. Culture and expand human normal lung epithelial cells (HBEC3-KT) on standard plastic cell culture dishes (T75 flasks) in normal epithelial growth medium using a standard incubator¹³ (see Table 1).
   NOTE: For improved growth, the medium can be additionally supplemented with 10% fetal calf serum (FCS).
2. Culture human fibroblasts (HS-5) in RPMI Medium supplemented with 10% fetal calf serum and penicillin/streptomycin^10,14 (see Table 1).
3. Prepare branching lung organoid medium (BLO-Medium) by supplementing DMEM/F12 with N-2, B27, penicillin-streptomycin, bovine serum albumin, monothioglycerol, ascorbic acid, KGF/FGF7, FGF10, ATRA, and CHIR99021. (see Table 1 and Table of Materials).
   NOTE: It is recommended to prepare the medium freshly, at least with the addition of KGF/FGF7, FGF10, ATRA, and CHIR99021. If more medium is needed within a short period of time, only a short storage period of a few days (up to 1 week) at 4 °C is recommended. Alternatively, the commercially available PneumaCult Airway Organoid Differentiation Medium can be used.

2. Preparation of epithelial cells and fibroblasts

NOTE: Use both cell types when cells reach around 80% confluency. Prepare each cell type separately.

1. Remove and discard the culture medium.
2. Wash cells by briefly rinsing the cell layer with 5 mL of PBS.
3. Trypsinize epithelial cells by adding 2 mL of 0.5% Trypsin-EDTA, followed by a short incubation at 37 °C for 4-8 min.
4. Harvest rounded and almost detached cells by adding normal growth media or PBS supplemented with 10% FCS using a 5 mL serological pipette.
5. Pipett the appropriate volume up and down several times in the T75 flask (along the growth surface) to further separate cell clumps and transfer the cell suspension to a 15 mL tube.
6. Harvest HS5 fibroblasts respectively using TE trypsinization at 37 °C for 3-5 min.
7. Pellet each cell type (and thus free from trypsin) by centrifugation (e.g., 300-400 × g for 5 min at room temperature [RT]) and resuspend (both epithelial cells and fibroblasts) in 1-2 mL of normal epithelial growth medium prior counting.
8. Determine viable cell numbers using an appropriate volume of resuspended cells; each diluted 1/5 using 0.4% Trypan Blue Stain in a Neubauer counting chamber under a microscope.

3. Generation of epithelial-fibroblast 3D spheroid cultures

1. Dilute the desired number of cells in normal epithelial growth medium using a 15 mL tube. Calculate 2000 epithelial cells with 50 fibroblasts in 100 µL of medium for each well of the 96 well plates. The resulting cell numbers adequate for one 96 well plate are 200,000 HBEC cells and 5,000 HS5 fibroblasts in 10 mL of normal epithelial growth medium.
2. Mix gently using 10 mL sterile serological pipettes.
3. Transfer the cell suspension to a sterile reagent reservoir for a multichannel micropipette.
4. Seed the cells into an ultra-low attachment plate by transferring 100 µL of the cell suspension to each well using a (12 channel) multichannel pipette.
5. Incubate under standard cell culture conditions at 37 °C and 5% CO₂ in a humidified atmosphere.
   NOTE: Spheroid formation is assessed via regular microscopic examination using an inverted microscope and by photographing spheroids for each desired time point.
6. After 4 days, add 100 µL of of BLO-Medium per well to generated spheroids.
7. Incubate further under standard cell culture conditions at 37 °C and 5% CO₂ in a humidified atmosphere.

4. Performing a regular media change

1. After 7-8 days, perform a regular media change. Therefore, position the plate at a slight angle so the spheroids sink in the tilting direction according to gravity.
2. Aspirate the supernatant carefully and as best as possible with a 100-200 µL pipette tip.
3. Re-position the plate horizontally and fill with 100-150 µL of BLO medium per well (using a multichannel pipette)
4. Incubate further under standard cell culture conditions at 37 °C and 5% CO₂ in a humidified atmosphere or start an experiment/treatment.
   NOTE: The spheroids can be cultured for longer periods (up to 21 days). Then, a complete medium change must be carried out at least twice a week.

5. Downstream analyses (harvesting spheroids)

1. Perform regular microscopic examination and documentation by taking photos in combination with spheroid size measurements and volume calculations, finally allowing quantification of spheroid growth and potential growth alterations following treatments.
2. For live cell imaging at desired time points, incubate the spheroids for an additional 15 min with 1 µg/mL Hoechst 33342 (or DAPI) for nuclei staining and analyze by fluorescence microscopy.
3. Harvest generated spheroids at the desired time point using a 200 µL pipette with the outlet cut off (keep sterile) to increase the diameter. Therefore, aspirate the media containing the spheroid as well. Combine and collect the spheroids in a 2-5 mL tube. After 2-5 min, when spheroids have sunk due to gravity, remove the supernatant and proceed to downstream analysis.
4. For generation of whole spheroid protein lysates, add directly ice-cold lysis buffer (150 mmol/L NaCl, 1% NP40, 0.5% sodium-deoxycholate, 0.1% sodium-dodecylsulfate, 50 mmol/L Tris/HCl, pH 8, supplemented with protease inhibitor cocktail) to the tube containing the harvested spheroids. After two to three freeze-and-thaw cycles, determine the protein content of the lysates and proceed to downstream analysis, e.g., Western blot analysis¹⁵.
5. For generation of whole spheroid RNA lysates, add directly RNA lysis buffer to the tube containing the harvested spheroids and perform RNA isolation according to the manufacturer instructions of the respective RNA isolation kit used.
6. For immunohistochemistry (IHC) and immunofluorescence staining, follow steps 5.6.1-5.6.3.
   1. Treat the harvested spheroids with 1-2 mL of 4% paraformaldehyde/phosphate buffered saline (PBS) for 30 min and subsequently subject to paraffin-embedding and sectioning (3-5 µm).
   2. Before immunofluorescent staining, prepare samples using a descending series of alcohol and incubation with a target retrieval solution. Afterward, block slides with a blocking solution (2% normal goat serum/PBS) to reduce unspecific interactions and incubate primary antibodies overnight at 4 °C.
   3. Detect antigens by fluorescently labeled secondary antibodies. Counterstain nuclei with DAPI. Perform additional staining (e.g., PAS staining kit) according to the manufacturer's protocols for histological evaluation.
7. For single-cell suspensions, follow steps 5.7.1 and 5.7.2.
   1. Resuspend harvested spheroids in 1-2 mL of TrypLE containing 100 U/mL DNase I followed by a 15-min incubation at 37 °C. Stop digestion by adding PBS containing 2-5% FCS, 2 mM EDTA, and DNAse I.
   2. Pass cellular solution through a 70 µm cell strainer into a fresh conical tube prior to downstream analysis (e.g., flow cytometry or single cell analysis).
8. For single-cell RNA sequencing analysis, follow steps 5.8.1-5.8.3.
   1. Label the different single-cell suspensions of interest representing different time points or treatments using a (human) single-cell multiplexing kit according to the manufacturer's instructions.
   2. Combine samples afterward and load on a lane on a single-cell cartridge aiming for 10,000-40,000 cells.
   3. Retrieve single cells, perform reverse transcription and library prep according to the manufacturer's instructions, and sequence isolated RNA.

Results

This simple protocol enables the matrix-free spheroidal co-cultivation of human bronchial epithelial cells and fibroblasts in combination with ultra-low attachment plates (Figure 1A). During the cultivation process, more complex lung spheroids gradually form from the initially generated spheroids, with the efficient generation of different epithelial and fibroblastic structures representing direct cellular interactions as revealed by morphological analyses using live cell imaging (Figure 1B), immunohistochemistry (Figure 1C) and immunofluorescence (Figure 1D). In order to distinguish respective cells from co-cultured epithelial cells, fluorescently labeled human HS5 fibroblasts that were transduced using a ('empty') lentiviral SIN vector expressing the gene encoding Turquoise2 (cyan) fluorescent protein were used¹⁶. While a kind of main spheroid appears where the fibroblasts are intercalated within the epithelial cells (Figure 1B-D), the fibroblasts also appear to continue to proliferate in the form of associated small spheroids. Viability measurements (Figure 1E) confirmed the viable and proliferative state of the cells inside the formed structures, which also grew steadily at the same time (Figure 1F). Surprisingly, despite the number of donor cells used (2000 epithelial cells and 50 fibroblasts), after a short 14-day culture, slightly less than half of the cells present are fibroblasts (identified by the CFP signal; Figure 1G), which indicates that the fibroblasts proliferate significantly more than the epithelial cells, which in turn differentiate more. A cell cycle analysis of the spheroids confirms that the fibroblasts are increasingly found in the G2/M phase (Figure 1G).

Continuous live cell imaging containing plated HBEC cells (2000 cells/well) together with HS5 fibroblasts (50 cells/well) labeled in green (cyan-fluorescent protein, CFP) was performed starting 48 h after cell plating and thus spheroid generation for an additional 72 h. Movies of three independent spheroids are provided (Supplementary Video 1, Supplementary Video 2, Supplementary Video 3, Supplementary Video 4, Supplementary Video 5, and Supplementary Video 6). The videos depict either the continuous phase contrast recording together with the CFP signal (overlay) or the CFP signal derived from the fibroblasts alone (CFP-FIB).

Single cell RNA sequencing analysis (scRNAseq) of these epithelial-fibroblast-spheroids (after an 11-day culture) was performed to finally prove the presence of the different epithelial and mesodermal lung cell subsets and characterize respective phenotypes in more detail (Figure 2). Clustering of the single cells revealed the presence of 11 different main cellular subsets, including epithelial and mesodermal lung cells (Figure 2A-D). Among the different subsets, basal cells expressing TP63 and different KRT (KRT7, KRT17, KRT23 and KRT80) were identified, together with bronchioalveolar stem cells (BASCs), which are generally characterized by the additional co-expression of bronchiolar (CLDN6, CLDN7, FOXJ1, RSPH1) and alveolar (KLF5, GABRP and LAMP3) epithelial genes. Two TUBA1B/1C and TUBB4B expressing ciliated (#1, #2) cell clusters were identified, with cluster ciliated #1 (CEP78, FOXN4, E2F5) comprising (more deuterosomal, differentiating) luminal epithelial cells and with cluster ciliated #2 comprising more mature ciliated cells (DNAH5, CEP55). Secretory cells and particular goblet cells were identified by the expression of ARG2, CEACAM6, SSH, and RTKN2, close to designated Clara (club) cells expressing MUC3A, SCGB2A1, and SFTPD. Two additional airway and four mesenchymal clusters were further identified. The designated airway #1 cluster comprises cells expressing both epithelial (TUBA1A, TUBA1B, and TUBB2B) and mesenchymal genes (PDGFRA, PDGFRB, ACTA2) in close vicinity to the ciliated cell clusters, and was therefore designated as myoepithelial (-like) cells including sub-mucosal gland epithelial cells. In turn, close by, a small pericyte/mural cell cluster (ACTA2, CDH11, POSTN) became prominent. The neighboring airway #2 cluster characterized by TUBA1A, TUBB2B together with SOX2, ACL1, and CHGA represents (proximal) epithelial cells in which (rare) NEBs were also present. All three fibroblast clusters were characterized by a general expression of mesodermal genes (VIM, FN1, FOXD1, and THY1). The designated adventitial fibroblast cluster FIB#1 (MMP2, CD248, WNT5A) showed the highest fibroblast cell numbers, followed by the (distal) alveolar FIB# 2 cluster (SOX9, SERPINE1, CAV1) and a subgroup of more secretory potentially immune conferring fibroblasts (CCL5, IFIT1/2, PDPN) comprising cluster FIB #3. Collectively, scRNAseq revealed the presence of classical epithelial and mesenchymal subpopulations of lung cells in more complex lung structures following direct spheroidal cultivation of both commercially available normal bronchial epithelial cells together with fibroblasts. Thus, we provide here a quite simple and fast method to generate scaffold-free bronchial epithelial cell-fibroblast spheroids as a physiologically relevant 3D lung model.

Figure 1: Experimental design and morphological spheroid analysis. (A) Scheme for the experimental design, including the timeline. Depicted wells represent each well of a 96 well ultra-low attachment plate. (B) Continuous live cell imaging containing plated HBEC cells (2000 cells/well) together with HS5 fibroblasts (50 cells/well) labeled in green (cyan-fluorescent protein, CFP) at indicated time points was used to track the spheroids forming using the 5x of an inverted phase contrast microscope. Scale bar represents 100 µm. (C) Periodic acid-Schiff (PAS) staining on paraffin-embedded spheroids was used for histological evaluation (11-day time point). Scale bar represents 150 µm. (D) Immunofluorescent analysis was performed in order to visualize epithelial cells (red, pan-cytokeratin, CK) interacting with fibroblasts (green, CFP). Representative pictures are shown. Hoechst was used for nuclei staining. Magnification 20x. Scale bar represents 150 µm. # indicate different spheroids (different biological replicates) (E) Cell viability was determined using the WST-1 reagent, a kind of water-soluble tetrazolium salt that induces the intracellular mitochondrial dehydrogenase to conduct NADH-dependent enzyme digestion reaction, releasing the water-soluble methyl benzene product (measured as. light absorption/ optical density, OD). Individual symbols represent different biological replicates. (F) Spheroid growth was measured, and the respective volumes were calculated^9,10 for indicated time frames from 3-4 individual experiments. (G) The number of CFP-expressing cells (fibroblasts, FIB) was determined by flow cytometry as well as the distribution cell cycle phases and apoptotic cells (subG1) following DNA staining using Nicoletti staining solution (50 µg/mL 7-Aminoactinomycin D [7-AAD], 0.1 % sodium citrate [w/v], and 0.05 % Triton X-100 [v/v] in PBS) (30-min incubation at room temperature prior analyses). Respective signals were additionally related to the CFP-expressing fibroblasts. Please click here to view a larger version of this figure.

Figure 2: Composition of generated lung spheroids. For single-cell RNA sequencing analysis, harvested epithelial-fibroblast-spheroids (after an 11-day culture) were dissociated, and single cells were subjected to bead-based mRNA isolation, bead retrieval, and subsequent whole transcriptome analysis. All further analyses were conducted with R packages. (A) Uniform Manifold Approximation and Projection for Dimension Reduction (UMAP) of all single cells (2581 in total) colored by identified clusters following single-cell RNA sequencing analysis. (B) Row-normalized heatmap of top ten differentially expressed genes per cluster. Important genes are indicated by name. Cell-type labels for each cluster are based on the expression of canonical cell-type markers (C) and the top 5 expressed genes per cluster (D) displayed in the dot plot. The size of the dots indicates the cell percentage in which this gene was found, and the color specifies the normalized value of the expression. Please click here to view a larger version of this figure.

Figure 3: Insights into spheroid co-cultures of other cell combinations. (A) Live cell imaging of spheroids consisting of HBEC cells (4000 cells/well) together with WI-38 fibroblasts (50 cells/well) was used forming (14-day time point) using inverted phase contrast microscopy. Scale bar represents 100 µm. Hematoxylin and eosin stain (HE) staining on paraffin-embedded spheroids was used for histological evaluation. Immunofluorescent analysis was performed in order to visualize indicated marker proteins. Hoechst was used for nuclei staining. Scale bars represent 50 µm. (B) Continuous live cell imaging containing plated HBEC cells together with HS5 fibroblasts labeled in green (as described in the protocol) and human microvascular endothelial cells (HMEC-1; 50 cells/well) labeled in red are exemplarily shown for spheroids consisting of three different cell types. Scale bar represents 100 µm. Please click here to view a larger version of this figure.

Table 1: Media composition. Please click here to download this Table.

Supplementary Video 1: Spheroid_1_48-120 h_Overlay. Please click here to download this File.

Supplementary Video 2: Spheroid_1_48-120 h_CFP-FIB. Please click here to download this File.

Supplementary Video 3: Spheroid_2_48-120 h_Overlay. Please click here to download this File.

Supplementary Video 4: Spheroid_2_48-120 h_CFP-FIB. Please click here to download this File.

Supplementary Video 5: Spheroid_3_48-120 h_Overlay. Please click here to download this File.

Supplementary Video 6: Spheroid_3_48-120 h_CFP-FIB. Please click here to download this File.

Discussion

3D cell culture models have been proposed to bridge the gap between 2D cultures and in vivo studies, as this type of (co-) culture better mimics the physiological, morphological, and pathological properties present in vivo because more in vivo-like cell-cell interactions, cell-ECM interactions, responses to stimuli, gene expression, protein expression, and differentiation are present^17,18. In addition to the numerous and sometimes very complex tumor spheroids^10,15,19,20, these spheroids are also increasingly used for normal tissue examinations^11,21,22. In particular, multicellular spheroid models, in which human bronchial epithelial cells are generated in combination with human lung fibroblasts, represent a valuable tool for studying respiratory toxicities. Further, these co-cultures might be suitable for high-throughput testing and enable the (rapid) investigation of drug effects on cell-cell interactions. 3D primary human airway epithelial cultures were already described as well suited for lung modeling, although obtaining well-differentiated 3D cultures remains challenging^22,23. For certain populations, an improvement in cultivation success is required.

Here, we have established a very simple, fast but reliable protocol where commercially available human lung epithelial cells are co-cultivated together with stromal fibroblasts as spheroids in ultra-low attachment plates without having to rely on a carrier matrix. The absence of an ECM not only simplifies the handling of this protocol and the processing of the spheroids for numerous possible downstream analyses, but it also represents a step toward xeno-free in vitro science. Avoiding animal-derived materials such as serum or basement membrane extracts, whose exact composition is often unknown, can ultimately reduce unwanted variability between batches and thus ensure optimal and standardized cultivation conditions for improved reproducibility and the overall scientific validity of experimental approaches. The high practicability of this increases the likelihood that others will be able to establish this method relatively quickly in the laboratory.

In combination with the differentiation medium used, different direct spheroidal co-cultures are generated, which consist of different epithelial and mesenchymal lung cell types that can be classically found in human lungs. Bronchial epithelial cells, which form the inner lining of the bronchial tubes, can crudely be classified into three principal types: basal, ciliated, and secretory cells^24,25,26. Basal epithelial cells in the bronchi are the principal stem cells of the airway capable to self-renew and to differentiate into other epithelial cell types, and exhibiting an important role in the production of various bioactive molecules. The bronchial secretory cells (goblet and Clara cells) play an important role in producing mucus that traps foreign particles, including pathogens and toxins, in the airway lumen and the secretion of anti-inflammatory proteins (e.g., uteroglobin). Ciliated epithelial cells are the most abundant cell types within the bronchi and are characterized by a high number of energy-producing mitochondria adjacent to their apical surface. Ciliated cells clear particles that are trapped in mucus out from the airways utilizing bidirectional cilia beating. More rare pulmonary neuroendocrine cells secrete a range of neuropeptides, although the exact role of these molecules in a respiratory context remains unclear. The bronchial epithelium thus constructed rests on a thick basement membrane²⁶. Underneath the basement membrane is the lamina propria, which contains connective tissue and blood vessels. Underneath this is the submucosa, which contains a large number of sero-mucinous glands. The lung cell types formed here predominantly represent the bronchial epithelium (of the conducting airways), while alveolar cell types are not formed (or only sporadically).

With regard to the morphology of the desired target tissue, the human lung, it should, of course, be noted that the structures formed do not adequately reflect the morphology of the human lung. Differential plating can be performed to increase the homogeneity of the spheroids formed by efficiently mixing epithelial cells and fibroblasts. This means that the fibroblasts could also be seeded to the well first (e.g., 50 fibroblasts in 50 µL of medium per well), and then 4-24 h afterward, the epithelial cells are added (e.g., 2000 HBEC cells in 50 µL of medium per well). It also works very well for increased epithelial cell counts (4000 cells per well), as an increased number of epithelial cells might reflect a more appropriate epithelial/fibroblast cell ratio since the epithelial cells predominantly differentiate and proliferate less, while this is exactly the opposite for the fibroblasts. However, for a complete reconstruction of the lung structure, the use of biological lung scaffolds is required to mimic the complex ECM and scaffold structure in such a way that functional properties of the lung can also be imitated²⁷. A further development of this method offers promising possibilities in the sense of lung bioengineering. Thus, by this type of co-cultivation, the different epithelial and mesenchymal lung cell types can be efficiently generated, but the mixing of the spheroids with respect to the fibroblasts and epithelial cell arrangement or the formation of a 'real' bronchial vascular structure is not (yet) given.

A general limitation of lung spheroids and even lung organoids is the lack of additional cell types, e.g., endothelial and infiltrating immune cells, and thus the inability to fully recapitulate the naïve lung cellular composition¹¹. We are currently working on increasing the complexity of the spheroids, e.g., by intercalating additional cell types, namely endothelial cells. Increasing the complexity of the system, together with differential plating, might result in an optimized arrangement of the different lung cell types. Presumably, this could reflect a more natural environment and thus also cause sufficient differentiation of the epithelial cells, which is now at least partly brought about by the differentiation medium, which we adapted from previous studies for the efficient generation of lung organoids from stem cells^11,28. If sufficient differentiation can be efficiently achieved by further complexation (i.e., the presence of additional cell types) of the spheroids, the protocol would be further simplified, and costs would also be reduced as the differentiation medium would be simplified. In addition to the advantage over primary cell cultures from lung biopsy material with relatively high donor variability and limited availability, a possible disadvantage of the spheroids generated here is that they may not be easily expanded, as has already been described for lung organoids²⁹, but must be generated anew each time.

The protocol can even be easily adapted for other starting cells. Similar results were also achieved using human HBEC cells in combination with WI-38 fibroblasts (Figure 3A). Even more cell types can be co-cultured; for example, the here investigated HBEC-fibroblasts (HS-5) combination together with endothelial cells (HMEC-1) (Figure 3B). Even numerous tumor cell-fibroblast combinations worked out very well in that system (not shown). The most critical point here is the estimation of optimal starting cell numbers and ratios, as well as the cultivation media used. When combining several different cell types, it is always recommended to use the most complex cultivation medium as a starting point, which is usually - at least for normal tissue - the epithelial cell medium.

Finally, we will investigate respective spheroids concerning the potential to serve as patient-specific platforms, e.g., for the identification of involved cell type-dependent signaling pathways and/or potentially new biomarkers within cancer therapeutic (radiation) treatment. Similar studies have already been successfully performed using a 'similar' matrix-free protocol to generate lung organoids from induced pluripotent stem cell-derived to model radiation-induced lung injury¹¹. Thus, the use of alternative in vitro methods, particularly the use of organotypic complex spheroids and specific organoids, has already and will continue to facilitate the analysis of the underlying molecular and cellular mechanisms of both lung homeostasis and disease states.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
10 mL sterile serological pipettes	ThermoFischer	170356N	Nunc sterile serological pipettes, or equivalent
15 mL conical tube	Corning	430052	or equivalent
37 °C/ 5% CO2 humidified  incubator
5 mL sterile serological pipettes	ThermoFischer	170355N	Nunc sterile serological pipettes, or equivalent
50 mL conical tube	Corning	CLS430829	or equivalent
ATRA (All-Trans Retinoic Acid)	StemCell Technologies	Cat#72262	50 nM
B-27 Supplement (50x),	ThermoFischer/Gibco	Cat#17504044	1x
BD Rhapsody WTA Amplification Kit	BD Bioscience	Cat#633801	use according to the manufacturer's instructions
Biosafety cabinet
Bovine Serum Albumin	Roche	Cat#BSAVHS-RO	0.4% (v/v)
Cell Strainer	Corning	352350	Falcon 70 µm Cell Strainer, White, Sterile, Individually Packaged
Centrifuge	Eppendorf	https://www.fishersci.com/shop/products/eppendorf-5804-series-centrifuge-rotor-packages-9/p-4119001	e.g., Eppendorf Centrifuge 5804 - Benchtop Centrifuge
DMEM/F12	Invitrogen	Cat#11330-032	500 mL
DNase I, RNase-free (1 U/μL)	ThermoFischer	Cat#EN0521	100 U/mL
EDTA (0,5 M), pH 8,0, RNase-free	ThermoFischer	AM9260G	2 mM
Eppendorf Research plus 12 channel multichannel pipette	Eppendorf		e.g., #3125000044 (10–100 µL) or #3125000060 (30–300 µL)
Fetal calf serum	ThermoFischer/Gibco	Cat#A5256701	10% (v/v)
HBEC3-KT	ATCC	CRL-4051	human normal lung epithelial cell line
Heparin Solution (0.2%)	StemCell Technologies	Cat#7980	0.8 mL
Hoechst 33342	ThermoFischer	Cat#H1399	1 µg/mL
HS-5	ATCC	CRL-3611	human stromal fibroblasts
Hu Recom FGF-10 (KGF-2) ACF	StemCell Technologies	Cat#78173.1	10 ng/mL
Hu Recom FGF-7 (KGF) ACF	StemCell Technologies	Cat#78186.1	10 ng/mL
Human Single-Cell Multiplexing Kit	BD Bioscience	Cat#633781	use according to the manufacturer's instructions
Hydrocortisin Stock Solution	StemCell Technologies	Cat#7926	0.5 mL per 500 mL
Laduviglusib (CHIR-99021)	Biozol/Selleckchem	Cat#S1263	3 µM
L-Ascorbic acid	Sigma-Aldrich	Cat#A92902	50 µg/mL
Microscope	ZEISS
Monothioglycerol	Sigma-Aldrich	Cat#M6145	0.4 µM
N-2 Supplement (100x)	ThermoFischer/Gibco	Cat#17502048	1x
Neubauer counting chamber	Merck	BR718605	e.g., BRAND counting chamber BLAUBRAND Neubauer pattern
Normal goat serum	ThermoFischer	Cat#31873	2 % (v/v) in PBS
NP-40	Merck/Millipore	Cat#492016	NP-40-Alternative, 1 %
PAS Staining Kit	Sigma Aldrich	Cat#1016460001	use according to the manufacturer's instructions
PBS	ThermoFischer	Cat#10010023	1x
Penicillin-Streptomycin (10.000 U/mL), 100x	ThermoFischer/Gibco	Cat#15140122	1x
PFA	ThermoFischer	Cat#J61899.AK	Paraformaldehyd, 4 % in PBS
PneumaCult Airway Organoid Differentiation Medium (ODM)	StemCell Technologies	Cat#05060	PneumaCult Airway Organoid Basal Medium (#05061, 360 mL) and PneumaCult Airway Organoid Differentiation Supplement* (#05063, 40 mL)
PneumaCult Ex Plus Medium	StemCell Technologies	Cat#5040	Kit consisting of PneumaCult-Ex Plus Basal Medium (#05041, 490 mL ) and PneumaCult-Ex Plus 50X Supplement*(#05042, 10 mL)
Protease inhibitor cocktail	Merck/Roche	Cat#4693132001	1 tab/10 mL
Rhapsody cDNA Kit	BD Bioscience	Cat#633773	use according to the manufacturer's instructions
RPMI 1640 Medium	ThermoFischer/Gibco	Cat#11875093	500 mL
Sodium Chloride	Merck/Sigma-Aldrich	S9625	150 mmol/L
Sodium Citrate	Merck/Sigma-Aldrich	Cat#1613859-1G	0.10%
Sodium Deoxycholate	Merck/Sigma-Aldrich	Cat#30970	0.50%
Sodium Dodecyl Sulfate	Merck/Sigma-Aldrich	Cat#11667289001	0.10%
Sterile reagent reservoir for multichannel micropipette (e.g., 60 ml)	ThermoFischer	9510037	or equivalent
TC-Platte 96well BIOFLOAT a 4 Stück	Sarstedt (or facellitate)	83.3925.400 (or F202003)	ultra-low attachment plate
Tris/HCl	ThermoFischer	Cat#AM9855G	50 mmol/L , pH 8
Triton X-100	Sigma-Aldrich	Cat#X100-5ML	0.05 %
Trypan Blue Stain (0.4%)	ThermoFischer	Cat#15250061	1/5 (v/v) with NGM
TrypLE	ThermoFischer	Cat#12604013	1x
Trypsin-EDTA (0.5 %)	ThermoFischer	Cat#15400054	0.05 % Trypsin-EDTA (TE) solution aseptically diluted to 1X using phosphate buffered saline (without calcium and magnesium)
WST-1 reagent	CELLPRO-RO, Roche	Cat#5015944001	cell proliferation reagent