Supported gel slab scaffolds as a three-dimensional cell-based assay platform†

Zachary R. Sitte ^a, Elizabeth E. Karlsson ^a, Tyler S. Larson ^a, Haolin Li ^b, Haibo Zhou ^bc and Matthew R. Lockett *^ad
aDepartment of Chemistry, University of North Carolina at Chapel Hill, 125 South Road, Chapel Hill, NC 27599-3290, USA. E-mail: mlockett@unc.edu
^bDepartment of Biostatistics, University of North Carolina at Chapel Hill, 135 Dauer Drive, Chapel Hill, NC 27599-7400, USA
^cUNC Center for Environmental Health and Susceptibility, University of North Carolina at Chapel Hill, 135 Dauer Drive, Chapel Hill, NC 27599-7400, USA
^dLineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, 450 West Drive, Chapel Hill, NC 27599-7295, USA

First published on 22nd July 2024

---

---

Introduction

Cell-based assays are heavily relied on in the drug discovery pipeline to identify candidates for further characterization and evaluation. These assays rely on monolayer cultures in well plates, which are well-suited for screening but cannot provide physiologically relevant tissue structures or microenvironments. Compared to tissue slices, cells maintained as monolayers have significantly altered phenotypes and morphologies.¹ Three-dimensional (3D) structures rich in extracellular matrices (ECM) restore cellular polarization lost in monolayers and better mimic cell cycle regulation and metabolism rates found in vivo.^2,3 Culture platforms that incorporate physiologically relevant tissue microenvironments are needed to meet the commitment outlined in the 2020 FDA Modernization Act 2.0, which focuses on using cell-based assays and in vitro testing to replace, reduce, and refine the use of animal models when evaluating drug candidates.⁴ This policy presents opportunities and challenges for existing 3D culture platforms, as there are few agreed-upon best practices and a significant gap between development and widespread adoption. A single platform capable of predicting outcomes in each tissue type with reproducible measures of parameters related to pharmacokinetic and dynamic responses is unlikely. What is likely are multiple platforms that allow tissue culture laboratories to improve their current experimental workflows and methods of analysis without the need for extensive expertise or infrastructure.

Current culture platforms incorporating 3D structures can be grouped by their preparation method. Suspension cultures promote cell–cell interactions that result in free-standing cellular aggregates such as spheroids or organoids.⁵ ECM-rich cultures place cells in naturally occurring or synthetic hydrogels that can be deposited onto substrates or into molds.⁶ Scaffold-based cultures support cells in pre-formed structures,⁷ which are often porous and chemically modified to promote cell attachment. The paper-based culture platform first described by Whitesides combines aspects of ECM-rich and scaffold-based cultures, using sheets of paper to support thin slabs of cell-laden gels.⁸ The power of this approach is evident from the many basic biochemistry studies and regenerative medicine applications it has enabled.^9–11 An individual sheet of paper can be patterned or cut to meet the requirements of a particular experimental setup. Thick, tissue-like structures can be assembled by stacking paper scaffolds containing different cell types or extracellular matrices. Physically separating these stacked scaffolds after an experiment provides spatially resolved datasets without fixation or histological slicing.

One drawback of paper scaffolds is their limited compatibility with optical readouts, a consequence of the opacity and highly scattering nature of the cellulose fibers. Optical clearing methods improve fluorescence microscopy in paper scaffolds but preclude live cell imaging.¹² Colorimetric and fluorescence-based readouts also suffer from low signal-to-background ratios, resulting from dyes non-specifically adsorbing on the fibers and the high autofluorescence of pre-whitened paper.¹³ We showed optical coherence tomography can image cells in paper scaffolds without any special pretreatment or sample preparation,¹⁴ but requires specialized equipment and expertise to collect and process the images.

Here, we describe a free-standing gel slab construct that retains the modularity and utility of paper scaffolds but removes the optical limitations caused by cellulose fibers. The supported gel slab (SGS) scaffolds are open well structures where cell-laden gels are deposited. These scaffolds were prepared from two polymeric sheets—a silicone sheet that defined the thickness of the well and a porous sheet that served as a bottom support and retained the gel in the well. We performed several demonstrations based on our previously published works with paper-based scaffolds to demonstrate the SGS scaffold's ability to support tissue-like structures. These demonstrations are part of ongoing work in our laboratory to develop better cell-based models to address the current limitations of extrapolating in vivo responses from in vitro measurements.

Materials and methods

Reagents

All reagents were used as received unless otherwise noted. Calcein-AM was purchased from Cayman Chemical, Prolong Gold anti-fade from Cell Signaling Technologies, and paraformaldehyde from Electron Microscopy Services. Resazurin and SN-38 were from Millipore Sigma. Dimethyl sulfoxide (DMSO) was from MP Biomedical, and doxorubicin hydrochloride was from Tocris. The Beta-Glo and CellTiter-Glo 2.0 (CGT) assays were purchased from Promega. 7-AAD, CellTracker Green CMFDA, the Click-iT EdU proliferation kit, and DAPI were purchased from ThermoFisher.

Cell culture maintenance

All cell culture medium and supplements were purchased from Gibco except for cell basement membrane extracellular matrix (ECM, ATTC), McCoy's 5A medium (Corning), and fetal bovine serum (FBS, VWR). The MDA-MB-231 (M231) and HCT116 cell lines were purchased from ATCC and validated with short-tandem repeat sequencing. Both cell lines were maintained as monolayers under standard culture conditions: 37 °C, 5% CO₂, and ambient oxygen tensions. The HCT116 cells were maintained in McCoy's 5A medium supplemented with L-glutamine, 10% FBS, 25 mM HEPES, and 1% PenStrep. The M231 cells were maintained in high glucose DMEM supplemented with L-glutamine, 10% FBS, 25 mM HEPES, 1% PenStrep, and sodium pyruvate. The culture medium was exchanged every 48 h, checked periodically for mycoplasma (MycoAlert, Lonza), and passed at ≥1:20 dilution when the cells reached 80% confluency with TrypLE. Each frozen stock of cells was cultured for no more than 20 passages.

SGS scaffold preparation

Fig. 1A outlines the scaffold preparation process. First, a PETG-silicone sheet (Silex Ltd) was laser cut to contain a series of 3.0 mm circles. Next, the PETG side of the sheet was affixed to a nylon mesh (Component Supply Company) or a porous PET film (CellQART, Sabeu GmbH) with a spray adhesive (Gorilla Waterproof Patch and Seal Spray). The affixed sheets were cured at room temperature (15 min), baked at 100 °C (10 min), and cooled between two sheets to cast acrylic to prevent deformation (overnight). The individual scaffolds were laser cut from the composite sheet, soaked in 70% ethanol (2 min), and dried at room temperature (overnight). The scaffolds were sterilized in an ethylene oxide chamber (Andersen Sterilizers) and stored until needed. Each scaffold was deposited with 0.78 μL of cell-laden or cell-free ECM. Cell suspensions were prepared by resuspending a cell pellet in 8.5 mg mL⁻¹ of ECM. This concentration was chosen based on previous studies on the stability of the ECM in paper-based scaffolds.¹⁵ When needed, the cells were pre-labeled with Cell-Tracker Green CMFDA (4.3 μM, 15 min, room temperature) before resuspension in ECM. Unless otherwise specified, the SGS scaffolds were 5.5 mm in diameter, composed of a 1.25 mm silicone ring surrounding a 3.0 mm culture well.

Cell viability, proliferation, and senescence measures

Cellular viability was determined with the CTG or resazurin assay, whose conditions were pre-optimized for the SGS scaffolds as detailed in the ESI.† DNA replication was measured with the Click-iT EdU proliferation kit. Each scaffold was incubated in fresh culture medium (100 μL) containing 10 μM EdU reagent on an XY shaker (60 rpm) for 4 h at 37 °C. The scaffolds were then processed following the manufacturer's protocol. Senescence measurements were determined with the Beta-Glo assay. Each scaffold was rinsed with 1× PBS and transferred to a 96-well plate containing 100 μL of fresh culture medium and 100 μL of the Beta-Glo reagent. After 30 min, the cell lysate solution (150 μL) was transferred to a fresh opaque 96-well plate. Luminescence measurements (350–850 nm) were performed on a SpectraMax i3x Multi-Mode microplate reader. Fluorescence intensity measures of the Click-iT EdU labeled scaffolds were determined from images collected on a Sapphire Biomolecular Imager (Azure Biosystems): 100 μm imaging resolution, λ_exc = 488 nm, λ_em = 518 nm BP22 filter, and a PMT setting of 3.

Microscopy

Cell-containing scaffolds were washed with 1X DPBS before mounting with Prolong Gold. Images were collected on a Zeiss LSM710 or LSM900 laser scanning confocal microscope with a 20×/0.80 Plan Apo objective. For nuclear stains, the SGS scaffolds were fixed in a 3.2% paraformaldehyde solution (20 min, room temperature) before incubating in 1× DPBS, 0.5% Triton X-100, and either propidium iodide (PI, 7.5 μM) or DAPI (7.2 μM) for 20 min.

Invasion assays

For the invasion assays, 2.0 × 10⁴ CellTracker Green-labeled M231 cells suspended in ECM were deposited in SGS scaffolds with nylon mesh bottom pieces. The average pore size of these meshes was 7 μm. For the single scaffold invasion assays, the cell-laden scaffolds were transferred to a 96-well plate containing 200 μL of serum-containing or serum-free medium. Cells at the bottom of the well plate after a 24 h incubation were counted with a MiniMax 300 Imaging Cytometer (λ_ex = 456 and λ_em = 541, BP108 nm, Molecular Devices). For the invasion stack assays, a cell-laden scaffold was sandwiched between six cell-free scaffolds. The stacks were placed in a customized holder that restricted the exchange of nutrients and oxygen to the top of the stack and incubated for 24 h. The separated layers were placed in a 25-well plate containing fresh medium (1 mL) and incubated under standard culture conditions for at least 4 h before analysis with the CTG assay. Schematics of the holder and a description of its fabrication are detailed in the ESI.†

Library screening

SGS scaffolds deposited with 1.0 × 10⁴ M231 cells suspended in ECM were incubated in standard culture conditions overnight and then transferred to a 96-well plate containing 3.125 μM of doxorubicin (positive control), 0.1% DMSO (v/v, negative control), or 10 μM of an unknown in 200 μL of medium. The unknown compounds were from the Approved Oncology Drugs Set library, generously donated by the National Cancer Institute. Stock solutions of each compound were prepared at 1000× the dosing concentration in DMSO and stored at −20 °C until needed. The cells were dosed for 72 h before analysis with the CTG assay. To determine the Z′ factor associated with the SGS scaffolds and the monolayer cultures,¹⁶ 30 positive and 30 vehicle control wells from the same 96-well plate were analyzed. The position of each well was determined with a random number generator. For the library screen, the position of each unknown compound and the accompanying controls were also placed in the same 96-well plate according to the results from a random number generator. The same experimental setup and conditions were used for 2D cultures of 500 M231 cells per well.

Spheroid-on-demand stack preparation and analysis

Seven scaffolds, each containing 6 × 10⁴ M231 cells, were stacked and placed in a customized holder that restricted the exchange of nutrients and oxygen to the top of the stack. The assembled holders were placed in a 12-well plate containing culture medium (2 mL) containing 0.625 μM doxorubicin or 0.1% (v/v) DMSO and maintained under standard culture conditions for 72 h on an XY orbital shaker (60 rpm). For the characterization studies, culture medium was used.

Statistical analyses

Unless otherwise stated, all values are the average and error-propagated standard error of the mean (SEM) of six separate setups prepared from at least two cell passages. All micrographs were analyzed with Imaris Viewer or FIJI.¹⁷ All datasets were analyzed with GraphPad Prism v9.0.0, using unpaired t-tests that were corrected for unequal standard deviations with Welch's correction. The R code used to compare viability profiles in the spheroid-on-demand stacks is included in the ESI.† These distribution comparisons were tested using a modified paired t-test and the Holm-Bonferroni method to sequentially reject hypotheses in multiple layer-wise comparisons.¹⁸ A p-value <0.05 was considered significant for all statistical comparisons, and the null hypothesis was rejected.

Results and discussion

SGS scaffolds support cell viability and proliferation in open well structures

We designed the SGS scaffolds to retain the most attractive features of the paper-based culture platform while overcoming its limited compatibility with optical readouts. These attractive features include generating 3D culture environments with commercially available materials, rapidly prototyping scaffolds of different shapes and sizes with tools commonly found in tissue culture labs, and easily generating thick tissue- or tumor-like structures on demand by simply stacking cell-containing scaffolds.

The SGS scaffolds were prepared using the cut-and-paste process outlined in Fig. 1A. We note that similar structures can be prepared with scissors or a hole punch and do not require a laser cutter. Fig. 1B is a cross-sectional schematic of a scaffold, highlighting its well-like structure. Detailed engineering schematics can be found in the ESI.† We chose sheets of polyethylene terephthalate glycol (PETG)-silicone to form the walls of the SGS scaffolds because both materials are cell culture-compatible and easily sterilized. Once cut with a series of holes that defined the cell-containing region of the SGS scaffolds, the PETG side of the sheet was coated with a spray-on adhesive and affixed to a nylon mesh or a sheet of porous polyethylene terephthalate (PET). These materials ensured the cell-laden gel was retained in the scaffold and enabled free-exchange of soluble factors with culture medium. Nylon mesh comes with various pore sizes, and allowed us to limit cell movement from the scaffolds as needed. The porous PET films were optically clear and highly compatible with imaging. The photographs in Fig. 1C highlight the differences in the optical transparency of the nylon mesh and porous PET bottom pieces, despite both having an average pore size of 7 μm.

SGS scaffolds are compatible with live and fixed sample imaging

Fig. 2A contains brightfield micrographs of the letter N printed on an overhead transparency viewed through the nylon mesh and porous PET films that served as the bottom piece of the SGS scaffolds. Adding glycerol to the nylon mesh significantly improved image clarity but could not remove the scattering effects from the fiber edges. The PET films were optically clear (%T > 90%) across the visible spectrum,¹⁹ and have been used in live cell imaging applications. Fig. 2B contains representative, single-plane confocal micrographs of cell-containing scaffolds with PET bottom pieces. Each scaffold was deposited with 1.0 × 10⁴ M231 cells suspended in ECM and stained with PI or DAPI after 24 h. These two nuclear stains have proven problematic in the paper scaffolds as they non-specifically adsorb onto the cellulose fibers. Fig. S1† contains representative micrographs of identical experimental setups in the paper and SGS scaffolds with a nylon bottom piece. Fig. 2C shows the cells are evenly distributed and retained in the culture region of the SGS scaffolds when deposited with 1.0 × 10⁴ or 6.0 × 10⁴ M231 cells suspended in ECM. The ESI† contains additional confocal and widefield micrographs of these scaffolds.

Cells proliferate in the SGS scaffolds

The viability of M231 cells suspended in ECM 72 h after deposition was statistically indistinguishable from the same number of cells in the presence of each component of the SGS scaffolds (Fig. S5†). Fig. 2D contains representative single-plane confocal micrographs of SGS scaffolds with a PET film bottom deposited with 1.0 × 10⁴ HCT116 or M231 cells suspended in ECM, stained with the Click-iT EdU reagent 72 h after deposition. This alkyne-containing thymidine derivative is incorporated into newly synthesized DNA and then reacted with an azide-containing fluorophore once the cells are fixed and permeabilized. By co-staining with DAPI, we quantified the percentage of replicating cells—46% of HCT116 and 37% of M231 cells. The average doubling time for scaffolds deposited with 1.0 × 10⁴ cells after 96 h was 28 h for the HCT116 cells (a 10.6-fold increase in cellular density) and 35 h for the M231 cells (a 6.7-fold increase). Fig. S6† contains representative single-plane confocal micrographs collected during this experiment and details of the experimental setup.

Quantify cellular invasion with SGS scaffolds

We developed paper-based invasion assays capable of quantifying cells by the extent, distance, and direction traveled.^20–24 We emulated two paper-based invasion assay formats with the SGS scaffolds to highlight their functionality beyond cellular imaging.

Fig. 3A is a schematic of the single scaffold invasion setup, akin to the Transwell assay used to identify activators or inhibitors of cellular movement.^25,26 These assays count the number of cells that degrade ECM and cross a porous membrane. The datasets in Fig. 3 highlight that SGS scaffolds are an attractive alternative to commercially available Transwells, which are costly and limited to well plate formats. Fig. 3B is a plot of M231 cells that invaded from SGS scaffolds with nylon mesh bottom pieces with an average pore size of 7 μm. The scaffolds were deposited with 2.0 × 10⁴ cells suspended in ECM and placed in a 96-well plate pre-filled with serum-starved (0% FBS) or serum-containing (10%) medium. After 24 h, the scaffolds were moved to a second well in the plate and the number of cells attached to the bottom counted. The serum-starved cells were 3.8-fold more invasive over the first 24 h period and 2.3-fold more invasive over the second 24 h period. These results agree with previous findings, which showed a lack of serum-activated phospholipase D increased the invasiveness of M231 cells.²⁷Fig. 3C is a plot of cellular invasion from scaffolds with nylon mesh bottoms of increasing pore size. Over a 48 h period, the fewest number of cells invaded through the mesh with the 0.1 μm pores. The number of cells invading through the meshes with 7 and 18 μm pores were equivalent, suggesting the cells could easily pass through both. These results agree with the guidelines provided by commercial vendors, which state that pores ≥5 μm are suitable for invasion studies. The ability to tune the pore size of the SGS scaffolds during fabrication offers increased experimental control that is not possible with paper scaffolds.²²

Unlike the binary readout of the single scaffold invasion assay, the stacked assay format (Fig. 3D) allowed us to isolate populations of cells based on their invasiveness. The dataset in Fig. 3E was 24 h after a stack of seven SGS scaffolds was assembled and placed in a holder that limited exchange medium exchange to only the topmost scaffold. The stack contained a cell-containing scaffold (6.0 × 10⁴ M231 cells) sandwiched between six scaffolds containing only ECM. Each scaffold had a nylon mesh bottom piece with an average pore size of 7 μm. After separating the scaffolds, the number of cells in each was determined with the CTG assay. The presence of cells in Layers 3 and 5 confirms the cell-laden gels were in conformal contact throughout the assay, as control setups that evaluated cells within an hour of placing in the holder did not contain cells in those layers. This dataset also matches previous reports by our group and others, which showed bi-directional invasion in similar setups with a greater number invading toward the source of oxygen and nutrients.^23,24,28 Fig. S11† contains a plot of pericellular hypoxia for given numbers of M231 cells deposited in the SGS and paper scaffolds. These datasets relied on a luciferase-containing gene construct that was stably incorporated into M231 cells.²⁴

SGS scaffolds support spheroid-like constructs

Stacking cell-containing SGS scaffolds results in tissue-like structures with user-defined thickness and composition. When placed in a holder that limits the exchange of fresh culture medium, oxygen gradients form across the structures within hours of assembly.^29,30 These mass transport-limited environments also occur in spheroids, with radial gradients and only the outermost 100 μm region of cells receiving an adequate oxygen supply.^31,32 These stacked structures have several advantages over spheroids. First, the directionality and steepness of the oxygen gradient are not fixed but rather defined by the holder and the density of cells in each layer. Second, the stacked structures are on demand and do not require prolonged culture periods to reach a desired diameter. Third, physically peeling apart the layers provides spatially resolved datasets without the fixation or histological slicing steps needed to analyze spheroids. Finally, the scaffolds increase the number of cancer types that can be evaluated, as several cell lines in the National Cancer Institute Human Tumor Line Screen (NCI 60) do not form aggregates in vitro.³³

The schematics in Fig. 4A compare a spheroid and an SGS spheroid-on-demand structure. Fig. 4B is a plot of the relative cellular viability, proliferation, and senescence throughout a seven-layer spheroid-on-demand stack. Each scaffold was deposited with 6.0 × 10⁴ M231 cells suspended in ECM, stacked, and the scaffolds separated after 120 h. Cellular viability was measured with the CTG assay, proliferation with the Click-iT EdU kit, and senescence with the Beta-Glo assay. Viability and the number of proliferating cells decreased monotonically with increasing distance from the source of fresh medium. We and others observed similar trends in paper-based tumor stacks.^29,30 The senescence data provides indirect evidence of an oxygen gradient. Layer 2 has the largest number of senescent cells and corresponds to the theoretical onset of hypoxia. Propidium iodide staining (Fig. S7†) suggests the cells at the bottom of the stack were not necrotic at 120 h.

Fig. 4C plots the relative viability in the spheroid-on-demand stacks after a 72 h dose with 0.625 μM doxorubicin or a vehicle control. The drug decreased the number of cells in the top two layers of the stack, which contained the largest number of proliferating cells. A layer-by-layer comparison is not appropriate between these two experimental conditions as the response in one layer impacts the others. To evaluate differences in distribution with and without the drug, we first determined the viability measurements were normally distributed with the Shapiro–Wilk test.³⁴ Next, we performed paired t-tests, using a Holm-Bonferroni method to adjust p-values to account for the dependency between layers.¹⁸ This procedure sequentially rejects hypotheses in multiple layer-wise comparisons while controlling the overall Type I error rate across all comparisons. This approach rejected the null hypothesis, confirming that doxorubicin caused significant changes in the distribution of cells in the stack compared to the no-drug control.

SGS scaffolds are compatible with drug library screening and identify different potential chemotherapies than monolayer cultures

Screening assays with high reproducibility require fewer replicates, increasing throughput while decreasing cost. Reproducibility is an important analytical metric but is meaningless if the assay cannot predict in vivo outcomes or the success of a potential drug candidate. Cells maintained in ECM have more in vivo-relevant phenotypes than monolayer cultures, increasing the potential of SGS scaffolds for screening purposes.

Fig. 5A is a plot of the luminescence signal from the CTG assay from two 96-well plates containing a randomized pattern of two culture conditions: a 72 h dose of doxorubicin (3.125 μM) or vehicle control (0.1% DMSO). One of the 96-well plates contained SGS scaffolds deposited with 1.0 × 10⁴ M231 cells suspended in ECM. The second well plate contained monolayer cultures of 500 M231 cells. These datasets were used to calculate each setup's Z′-factor (eqn (1)). This value compares the average signal of the positive and negative controls ( and ), their associated standard deviations (s_pos and s_neg), and the number of replicates (n_pos and n_neg). The separation between the positive and negative controls determines the robustness of the screen to identify a statistically significant cellular response from the negative control (a “hit”) from a limited number of replicates.³⁵

	(1)

Doxorubicin served as the positive control, and the DMSO vehicle as the negative control. The number of replicates for each control was 30, each arranged in a checkerboard pattern across the plate. The average signals for the positive and negative controls were statistically indistinguishable between the two assay formats. However, the variance in those signals was significantly greater in the SGS scaffolds compared to the monolayer cultures. The Z′-factor for this particular setup was 0.83 for the SGS scaffolds and 0.93 for the monolayer layer cultures—with both acceptable according to the Z′ ≥ 0.4 threshold set by the National Center for Advancing Translational Science.³⁵ Assay variance decreases with sample size but determines the overall suitability of the approach for single data point measures. What is not captured in this value is the inter-screen variation. Therefore, each well plate must incorporate a fixed number of positive and negative controls to compare against this master dataset. The variation in the 3D format to sample preparation that currently limits this format compared to the monolayers is the difficulty in ensuring the cell suspension remains homogeneous throughout the deposition process; this homogeneity influences the distribution of cells between scaffolds. This variable can be mitigated with automation or other means that remove the time it takes an individual to deposit cells in a large number of scaffolds.

Next, we screened 80 compounds from the Approved Oncology Drug Set provided by the National Cancer Institute's Division of Cancer Treatment and Diagnosis. Fig. 5B is a plot of cellular viability for a single replicate culture dosed with the designated library compound, 72 h after dosing. A compound that decreased cellular viability by relative luminescence units was considered a hit, as determined with the CTG assay. This analysis assumes these signals were directly proportional to cellular viability and not a consequence of the drug depleting ATP pools or decreasing cellular metabolism. Fig. 5C summarizes the results from this screen. A total of 62 hits were identified: 17 shared between the two monolayer and 3D culture formats, 44 unique to the monolayer cultures, and 1 unique to the 3D cultures. Table S1† identifies each library compound by their Cancer Chemotherapy National Service Center number, Chemical Abstracts Service Registry number, and if the compound was identified as a hit in the 2D and 3D culture formats during our screen.

The different hits between the two formats are unsurprising and highlight the importance of context when evaluating cellular responses to drug candidates. Fig. 5D groups the hits by their proposed mechanism of action. Many monolayer-only hits were anti-neoplastic agents targeting proliferating cells; this mechanism of action is less effective in 3D formats as proliferation rates are decreased. The lone compound that was an SGS-only hit is associated with protoporphyrin IX (pPIX) accumulation. Ogura reported that cells in prostate cancer spheroids of various sizes contained increased levels of pPIX compared to the same cells maintained as monolayer cultures.³⁶ These increased levels made the cells more susceptible to photodynamic therapy in the presence of 5-aminolevulinic acid. The remainder of the hits targeted the cell cycle (39 in total from both culture formats), disrupting DNA replication or microtubule formation. Previous studies support our finding that two compounds targeting the Akt/mTOR pathway and two compounds targeting the epidermal growth factor receptor (EGFR) signaling pathways were only effective in the monolayer cultures. Hiscox showed that breast cancer cell lines in 3D culture environments significantly lose Akt/mTOR pathway activity and an increased MAPK pathway.³⁷ O'Driscoll found that breast cancer cell lines in 3D environments have increased expression of EGFR and drug transport proteins,³⁸ likely causing therapeutic resistance not found in the monolayers.

Conclusion

Supported gel slab (SGS) scaffolds deposited with cell-laden gels result in a 3D structure with tissue-like microenvironments. The cut-and-paste manner of fabrication enables rapid prototyping of scaffolds with user-defined diameters, thicknesses, and abilities to retain the cells throughout an experiment. These open well constructs are more compatible with optical microscopies than their paper-based scaffolds, as the light-scattering cellulose fibers are removed. The demonstrations presented highlight aspects of the SGS scaffolds that we feel will improve cell-based assays that rely on cellular monolayers. (1) Single scaffolds whose dimensions are compatible with commercial 96 well plates support viability, invasion, and imaging-based readouts commonly used when evaluating cellular responses to drug candidates or potential toxicants. The compatibility with commercially available viability assays enabled the screening of a drug library with the same reagents used in monolayer cultures. The optically transparent bottoms enable live cell imaging and eliminate the need for fixation or optical clearing. The pore size of the scaffold bottom introduces a selectivity that is needed to prevent or promote cellular invasion and could be used to control the nutrients exchanged between cells and the culture medium. (2) Stacking the scaffolds generates a spheroid-on-demand structure, enabling studies of cellular invasion and response in a spheroid structure that provides spatially resolved datasets with minimal sample processing (simply destacking the scaffolds). Such experimental control over cell placement and density is not possible with spheroid or organoid structures. The spatially resolved datasets of cellular responses to a chemotherapeutic could be further mined for drug metabolism profiles or omics-level responses. SGS scaffolds are promising for teasing out biochemical mechanisms of cellular responses, translational applications such as tissue generation, and developing in vitro-to-in vivo extrapolation models.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by funds provided by the National Institute of General Medicine through Grant Award Number R35 GM128697 and the National Institute of Environmental Health Sciences (NIEHS) through Grant Award Numbers R01 ES032730. H. Li is supported by the grant T32 ES007018 from the NIEHS. We thank the Biostatistics Core of the UNC Center of Environmental Health for help with statistical analyses. This core is supported in part by the NIEHS through Grant Number P30 ES010126. We thank the Microscopy Services Laboratory and its director, Dr Pablo Ariel, for confocal microscope access. This laboratory is supported in part by the National Cancer Institute through Grant Award Number P30 CA016086, a Cancer Center Support Grant to the UNC Lineberger Comprehensive Cancer Center. We also thank Dr Thomas DiProspero for his thoughtful input and discussions on this project.

References

1. M. Kapalczynska, T. Kolenda, W. Przybyla, M. Zajaczkowska, A. Teresiak, V. Filas, M. Ibbs, R. Blizniak, L. Luczewski and K. Lamperska, Arch. Med. Sci., 2018, 14, 910–919 Search PubMed .
2. R. Bhat and M. J. Bissell, Wiley Interdiscip. Rev.: Dev. Biol., 2014, 3, 147–163 CrossRef CAS PubMed .
3. K. M. Yamada and E. Cukierman, Cell, 2007, 130, 601–610 CrossRef CAS PubMed .
4. J. J. Han, Artif. Organs, 2023, 47, 449–450 CrossRef PubMed .
5. F. Hirschhaeuser, H. Menne, C. Dittfeld, J. West, W. Mueller-Klieser and L. A. Kunz-Schughart, J. Biotechnol., 2010, 148, 3–15 CrossRef CAS PubMed .
6. J. Nicolas, S. Magli, L. Rabbachin, S. Sampaolesi, F. Nicotra and L. Russo, Biomacromolecules, 2020, 21, 1968–1994 CrossRef CAS PubMed .
7. S. Pina, V. P. Ribeiro, C. F. Marques, F. R. Maia, T. H. Silva, R. L. Reis and J. M. Oliveira, Materials, 2019, 12, 1824 CrossRef CAS PubMed .
8. R. Derda, A. Laromaine, A. Mammoto, S. K. Tang, T. Mammoto, D. E. Ingber and G. M. Whitesides, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 18457–18462 CrossRef CAS PubMed .
9. S. M. Cramer, T. S. Larson and M. R. Lockett, Anal. Chem., 2019, 91, 10916–10926 CrossRef CAS PubMed .
10. K. Ng, B. Gao, K. W. Yong, Y. H. Li, M. Shi, X. Zhao, Z. D. Li, X. H. Zhang, B. Pingguan-Murphy, H. Yang and F. Xu, Mater. Today, 2017, 20, 32–44 CrossRef CAS .
11. D. Lantigua, Y. N. Kelly, B. Unal and G. Camci-Unal, Adv. Healthcare Mater., 2017, 6, 1700619 CrossRef PubMed .
12. H. Dermutz, G. Thompson-Steckel, C. Forro, V. de Lange, L. Dorwling-Carter, J. Voros and L. Demko, RSC Adv., 2017, 7, 39359–39371 RSC .
13. K. G. Shah and P. Yager, Anal. Chem., 2017, 89, 12023–12029 CrossRef CAS PubMed .
14. J. C. McIntosh, L. Yang, T. Wang, H. Zhou, M. R. Lockett and A. L. Oldenburg, Biomed. Opt. Express, 2020, 11, 3181–3194 CrossRef CAS PubMed .
15. Unpublished datasets.
16. P. W. Iversen, B. J. Eastwood, G. S. Sittampalam and K. L. Cox, J. Biomol. Screening, 2006, 11, 247–252 CrossRef CAS PubMed .
17. J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, J. Y. Tinevez, D. J. White, V. Hartenstein, K. Eliceiri, P. Tomancak and A. Cardona, Nat. Methods, 2012, 9, 676–682 CrossRef CAS PubMed .
18. S. Holm, Scand. J. Stat., 1979, 6, 65–70 Search PubMed .
19. D. A. Gordon, A. Gok, C. W. Meyer, C. L. Fagerholm, N. W. Sweet, L. DeNoyer, L. S. Bruckham and R. H. French, Characterizing the weathering induced haze formation and gloss loss of poly(ethylene-terephthalate) via MaPd:RTS spectroscopy, Reliability of Photovoltaic Cells, Modules, Components, and Systems IX, SPIE Optics + Photonics for Sustainable Energy, San Diego, California, United States, 2016, vol. 9938,  DOI:10.1117/12.2238295 .
20. R. M. Kenney, M. W. Boyce, A. S. Truong, C. R. Bagnell and M. R. Lockett, Analyst, 2016, 141, 661–668 RSC .
21. R. M. Kenney, M. C. Lee, M. W. Boyce, Z. R. Sitte and M. R. Lockett, Anal. Chem., 2023, 95, 3054–3061 CrossRef CAS PubMed .
22. R. M. Kenney, A. Loeser, N. A. Whitman and M. R. Lockett, Analyst, 2019, 144, 206–211 RSC .
23. A. S. Truong, C. A. Lochbaum, M. W. Boyce and M. R. Lockett, Anal. Chem., 2015, 87, 11263–11270 CrossRef CAS PubMed .
24. A. S. Truong and M. R. Lockett, Analyst, 2016, 141, 3874–3882 RSC .
25. S. SenGupta, C. A. Parent and J. E. Bear, Nat. Rev. Mol. Cell Biol., 2021, 22, 529–547 CrossRef CAS PubMed .
26. E. Sahai, Curr. Opin. Genet. Dev., 2005, 15, 87–96 CrossRef CAS PubMed .
27. Q. Ye, S. Kantonen and J. Gomez-Cambronero, J. Mol. Biol., 2013, 454, 755–766 CrossRef PubMed .
28. B. Mosadegh, M. R. Lockett, K. T. Minn, K. A. Simon, K. Gilber, S. Hillier, D. Newsome, H. Li, A. B. Hall, D. M. Boucher, B. K. Eustace and G. M. Whitesides, Biomaterials, 2015, 52, 262–271 CrossRef CAS PubMed .
29. M. W. Boyce, G. J. LaBonia, A. B. Hummon and M. R. Lockett, Analyst, 2017, 142, 2819–2827 RSC .
30. D. Rodenhizer, E. Gaude, D. Cojocari, R. Mahadevan, C. Frezza, B. G. Wouters and A. P. McGuigan, Nat. Mater., 2016, 15, 227–236 CrossRef CAS PubMed .
31. A. S. Nunes, A. S. Barros, E. C. Costa, A. F. Moriera and I. J. Correia, Biotechnol. Bioeng., 2019, 116, 206–226 CrossRef CAS PubMed .
32. D. R. Grimes, C. Kelly, K. Bloch and M. Partridge, J. R. Soc., Interface, 2015, 11, 20131124 CrossRef PubMed .
33. M. Selby, R. Delosh, J. Laudeman, C. Ogle, R. Reinhart, T. Silvers, S. Lawrence, R. Kinders, R. Parchment, B. A. Teicher and D. A. Evans, SLAS Discovery, 2017, 22, 473–483 CrossRef CAS PubMed .
34. S. S. Shapiro and M. B. Wilk, Biometrika, 1965, 52, 591–611 CrossRef .
35. P. W. Iversen, B. Beck and Y. F. Chen, et al., HTS Assay Validation, In Assay Guidance Manual [Internet], ed. S. Markossian, A. Grossman and M. Arkin et al., Bethesda (MD): Eli Lilly & Company and the National Center for Advancing Translational Sciences, 2004, Available from: https://www.ncbi.nlm.nih.gov/books/NBK83783/ Search PubMed.
36. T. Nakayama, S. Otsuka, T. Kobayashi, H. Okajima, K. Matsumoto, Y. Hagiya, K. Inoue, T. Shuin, M. Nakajima, T. Tanaka and S. I. Ogura, Sci. Rep., 2016, 6, 36478 CrossRef CAS PubMed .
37. S. Gangadhara, C. Smith, P. Barrett-Lee and S. Hiscox, BMC Cancer, 2016, 16, 345 CrossRef PubMed .
38. S. Breslin and L. O'Driscoll, Oncotarget, 2016, 7, 45745–45756 CrossRef PubMed .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4an00691g

This journal is © The Royal Society of Chemistry 2024