A Biomimetic Buffering Hydrogel Scaffold for Long‐Term Culture of Patient‐Derived Tumor Organoids
Elizaveta Gusarova, Fatemeh Ahmadi, Jennifer Cruickshank, Zheyuan Miao, Mariia Moshkova, Iuliia Pilipenko, David W. Cescon, Eugenia Kumacheva
TL;DR
A new hydrogel scaffold helps maintain the right pH for long-term cancer organoid growth, mimicking the tumor environment better.
Contribution
A biomimetic hydrogel with built-in buffering capacity is introduced to maintain physiological pH during organoid culture.
Findings
The hydrogel maintained pH in the physiologically relevant range during long-term culture.
The hydrogel preserved its structural and mechanical properties over 21 days.
Cancer cell proliferation was enhanced in organoids cultured in the buffering hydrogel.
Abstract
Patient‐derived cancer organoids have emerged as a promising in vitro model for fundamental cancer research and drug screening for therapeutic cancer treatment. Yet, while the inherent acidification of the tumor environment in vivo is controlled at a particular level, hydrogel scaffolds used for organoid culture lack this ability and their pH falls outside the physiologically relevant range. The excessive acidification can also lead to the degradation of pH‐sensitive hydrogel scaffolds during long‐term organoid culture, thus changing the mechanical properties of the organoid microenvironment. Here, we report a biomimetic fibrous hydrogel with built‐in buffering capacity, which enables control of the local acidification of the organoid environment to maintain its mechanical and structural stability. The hydrogel is formed from aldehyde‐functionalized cellulose nanocrystals carrying…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7- —New Frontiers Research Fund
- —Natural Sciences and Engineering Research Council of Canada10.13039/501100000038
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsCancer Cells and Metastasis · 3D Printing in Biomedical Research · Nanoplatforms for cancer theranostics
Introduction
1
Patient‐derived tumor organoids (PDTOs) or tumoroids are 3D cell aggregates grown from tumor biopsies, which recapitulate many aspects of tumor heterogeneity, gene and protein expression, and metabolic activity [1, 2, 3]. Over the past decade or so, PDTOs initiated and established in hydrogel scaffolds have emerged as highly promising in vitro model for fundamental cancer research and drug screening for therapeutic cancer treatment [4, 5]. One of the major objectives of further development and advancing this model is close mimicking of tumor microenvironments in vivo.
In vivo, cells are surrounded by the extracellular matrix (ECM) composed of fibrous proteins, proteoglycans, and polysaccharides [6]. The ECM provides biophysical, mechanical, and biochemical cues regulating cell behavior [7, 8, 9]. The extracellular pH of solid tumors is inherently acidic due to a combination of elevated glycolysis‐driven production of lactic acid and hydration of CO_2_ formed in the oxidative tumor regions [10, 11]. The extent of acidification of the tumor environment is controlled at a physiologically relevant level of 6.5–6.9 via a combination of intra‐ and extracellular mechanisms [12, 13]. Extracellular buffering to pH in this range is achieved using a bicarbonate buffer, proteins, and hemoglobin in blood [14, 15, 16]. In addition, cells, e.g., cancer‐associated fibroblasts (CAFs) utilize various transporters to move protons across their membranes [10, 17]. Acidosis is recognized as one of the major hallmarks of tumors and a key regulator of their progression.
In PDTO culture, pH stabilization is generally achieved by using conventional soluble buffers, such as 4‐(2‐hydroxyethyl)‐1‐piperazineethanesulfonic acid (HEPES) or the CO_2_/HCO_3−_ system [18]. Such strategies do not provide localized or sustained buffering at the PDTO/matrix interface and are sensitive to even small fluctuations in pCO_2_ (in the case of bicarbonate buffering) [19], while HEPES is ineffective in the mildly acidic range below pH 6.8 that is characteristic of PDTO culture [20].
Furthermore, hydrogels containing pH‐sensitive dynamic covalent bonds in hydrazone, imine, or boronic ester crosslinking groups have attracted significant interest as scaffolds for cancer spheroid and PDTO growth [21, 22, 23, 24, 25]. Yet, hydrazone and imine groups undergo acid‐catalyzed hydrolysis at pH ∼6.5–7.0 [26, 27], while ester linkages hydrolyze at pH < 6.5 [28]. Such pH‐responsiveness has been utilized in hydrogels developed for the localized drug delivery [29, 30, 31, 32], however it poses a challenge for hydrogels used as scaffolds for the PDTO culture, since tumoroids acidify their environment to pH reaching the values of 6.4 and in some regions, reducing below 5.5 [33, 34]. Under these conditions, pH‐sensitive covalent bonds undergo cleavage, leading to the change in hydrogel's mechanical properties and ultimately, loss of its integrity within 5–7 days of tumoroid growth [22, 35]. As a result, such hydrogels become unsuitable for long‐term culture [21, 36, 37, 38].
This limitation is also important for drug screening applications, where prolonged PDTO culture may be needed to evaluate the therapeutic efficacy, capture the emergence of resistance mechanisms, and monitor tumoroid growth or pharmacodynamic effects over time [4, 39]. In addition, excessive acidification can influence the efficacy of antibody–drug conjugates (ADCs) formed using pH‐sensitive linkers [40, 41, 42].
We hypothesized that the introduction of buffering moieties into an otherwise pH‐sensitive hydrogel would enable control of the pH of the environment in the vicinity of the encapsulated PDTOs at the pH level optimal for cancer cell growth. Such control would also lead to enhanced stability and evaluation of ADC performance. In contrast to soluble buffering systems, covalent grafting of the buffering moieties directly into the fibrous ECM‐mimetic hydrogel scaffold would provide localized and sustained regulation of the extracellular pH value, while preserving the structural and mechanical stability of the hydrogel under acidic conditions. We validated this hypothesis for the imine group‐crosslinked hydrogel derived from aldehyde‐modified cellulose nanocrystals (a‐CNCs) that were functionalized with histidine molecules and crosslinked with gelatin. In these hydrogels, the a‐CNCs rendered a biomimetic fibrous gel structure, the histidine moieties led to the buffering in the range of 5.12≤pH≤ 7.12 [43, 44, 45], and the Arg‐Gly‐Asp (RGD) sequences in gelatin provided bioadhesiveness [46, 47].
We show that this hydrogel has a high buffering capacity leading to the invariant density of imine crosslinks, structural hydrogel integrity, and its close‐to‐constant mechanical properties during 21‐day PDTO growth. The hydrogel enabled long‐term culture of proliferative PDTOs with a preserved tumorigenic phenotype. After 21 days, the fraction of proliferating cells in the PDTOs grown in the buffering hydrogel was significantly higher than in the non‐buffering conditions, indicating enhanced support for tumor cell proliferation. Although the mechanism of buffering in the hydrogel scaffold was different than complex extracellular buffering in vivo, this work shows promising applications of pH‐buffered hydrogels as a viable platform for PDTO growth.
Results
2
Hydrogel Design
2.1
Figure 1A illustrates the mechanism of buffering by the histidine molecules, with an imidazole group accepting and donating protons to counteract pH changes. With pKa ≈ 6.1 it provides the buffering ability of histidine to maintain pH of the environment close to the physiological range [48].
Design of the buffering hydrogel. (A) Protonation–deprotonation equilibrium of the imidazole group in histidine molecules with pKa = 6.1. (B) Schematic of the formation of the pH‐buffering hydrogel (EKGel‐H).
Recently, an a‐CNC/gelatin non‐buffering hydrogel (termed as EKGel) with a range of mechanical properties and reduced batch‐to‐batch variability has been developed for PDTO initiation and growth [23, 49, 50]. Here, a buffering capability of this gel (EKGel‐H) was achieved by covalently attaching histidine molecules to the hydrogel's network. Figure 1B illustrates the formation of the EKGel‐H. We hypothesized that a fraction of aldehyde groups of a‐CNCs will react with amine groups of histidine via Schiff base reaction, leading to the histidine‐modified a‐CNCs (referred to as aH‐CNCs), while the remaining aldehyde groups will react with primary amine groups of gelatin to form a hydrogel network. By varying the histidine content in the gel, control can be achieved over its buffering properties. Importantly, since both histidine and gelatin react with the aldehyde groups of a‐CNCs, a delicate balance should be achieved between the buffering capacity of the gel (controlled by the histidine content) and its mechanical properties (controlled by the reaction of gelatin with a‐CNCs). Increasing the density of aldehyde groups by CNC oxidation would favor their reaction with histidine and gelatin, however this should not compromise the structural integrity of a‐CNCs and the excess of aldehyde groups beyond reaction stoichiometry should not lead to hydrogel's cytotoxicity [51].
Hydrogel Preparation
2.2
The modification of a‐CNCs with histidine was achieved via the reaction between the aldehyde groups of a‐CNCs and the amine groups of histidine. By varying the degree of oxidation of CNCs (see details in Supporting Information), the density of aldehyde groups on their surface was changed in the range of 6.0–14.2 mmol/g (Table S1). By varying the ratio of mass concentrations of a‐CNCs‐to‐histidine from 4.0 to 1.0, the histidine loading on aH‐CNCs was changed from 1.45 to 11.4 mmol/g, respectively. Based on the analysis of TEM images of aH‐CNCs at histidine concentration of 11.4 mmol/g, the average aH‐CNC diameter decreased by 17%, in comparison with a‐CNCs. The aH‐CNCs exhibited a weak tendency for association, due to the π–π stacking interactions between the histidine residues. The details of a‐CNC and aH‐CNC preparation and characterization are provided in Figures S1–S4 and Table S1.
An EKGel‐H was formed by mixing an aqueous aH‐CNC suspension and gelatin solution at 37 °C and incubating the mixture for 3 h. Unless otherwise specified, the water‐swollen hydrogels used for the physicochemical characterization were prepared using the sample aH‐CNC‐8.1 containing 11.4 mmol/g of histidine (Table S1). The pH of the aH‐CNC/gelatin suspension was 5.5, favorable for the formation of a weak Schiff base by the reaction of aldehyde and primary amine groups of gelatin [52]. Upon mixing, a notable reduction in the intensities of the aldehyde band at 1738 cm^−1^ (corresponding to the aldehyde groups of aH‐CNCs) and the amine group at 1556 cm^−1^ of gelatin was observed (Figure S5A). These changes, along with the increase in the intensity of the imine stretching band at 1670 cm^−1^ indicated aH‐CNC crosslinking with gelatin molecules with the formation of Schiff base. The appearance of the bands at 1562, 1546, and 3080 cm^−1^, corresponding to N─H and C═O stretching of amide groups [53] (Figure 2A) was ascribed to the formation of hydrogen bonds between gelatin and histidine, thus, enhancing aH‐CNC and gelatin crosslinking. In contrast, no amide stretching bands were detected in the IR spectra of a‐CNCs mixed with gelatin. The interaction of gelatin with histidine moieties was also confirmed by the decrease in intensity of the 1556 cm^−1^ band of the amine groups and the narrowing of the 3300 cm^−1^ band (reflecting O─H/N─H vibrations) in the spectrum of the mixed solution of gelatin and histidine (Figure S5B).
(A) FTIR spectra of water‐swollen EKGel‐H and EKGel. (B) State diagram of the aH‐CNC/gelatin aqueous mixture. Gel formation was examined in an inversion test performed 24 h after mixing aH‐CNCs and gelatin. (C) Variation in the shear storage modulus, G′, and loss modulus, G″ plotted as a function of time for the aH‐CNC‐8 1/gelatin aqueous mixture containing 0.028 mmol g−1 of aldehyde groups and 0.0047 mmol g−1 of amine groups. (D) Scanning electron microscopy image of water‐swollen EKGel‐H. Scale bar is 1 µm.
Importantly, gelation of the aH‐CNC/gelatin mixture occurred only in the particular range of concentrations of the free (non‐consumed) aldehyde groups of aH‐CNCs and amine groups of gelatin (Figure S6). Figure 2B shows a state diagram for the aH‐CNC/gelatin mixture, with a green area outlining the conditions of gel formation. Gelation occurred over the entire range of aldehyde group concentrations and only within a particular range of the content of amine groups provided by gelatin. At low concentrations of amine groups, their reaction with the aldehyde groups of aH‐CNCs was insufficient to cause gelation, while at high concentration of amine groups (at high gelatin content), the increased viscosity of the suspension, slowed down the aH‐CNC diffusion.
The role of histidine in gel formation was evaluated in control experiments performed for the mixture of a‐CNCs and gelatin, in which a greater amount of amine groups was required to achieve gel formation, compared to the aH‐CNC/gelatin mixture (Figure S7). This observation suggested that in the region of low amine group concentration in Figure 2B, gelation was caused both by the formation of covalent imine crosslinks and non‐covalent interactions between the histidine residues of aH‐CNCs and gelatin.
Figure 2C shows the variation in the shear storage modulus, G′, and loss modulus, G″, of the aH‐CNCs/gelatin mixture with incubation time, measured at 1% oscillatory strain and 1 Hz frequency. The value of G′ began to increase in 2 h after mixing aH‐CNCs and gelatin, and after 25 h, it reached the value of 28 Pa, while during this time, G″ increased only to 2.4 Pa. In contrast, a lower G' value of 10 Pa was reached after 25 h for EKGel (Figure S8). This was ascribed to the contribution of histidine moieties associating with gelatin in EKGel‐H.
Figure 2D shows a representative scanning electron microscopy (SEM) image of EKGel‐H. The gel had a characteristic fibrous structure with the average fiber diameter of 32 ± 1 nm. With histidine concentration in the hydrogel increasing from 0 to 0.114 mmol/g, the average fiber diameter increased from 21 ± 1 to 32 ± 1 nm, respectively (Figure S9), which was attributed to the histidine‐mediated association of aH‐CNCs and the enhanced incorporation of gelatin in the fibers due to gelatin‐histidine interactions.
Hydrogel Buffering Properties
2.3
To evaluate the buffering performance of EKGel‐H, a 0.01 m solution of HCl or NaOH was added to the hydrogel submerged in deionized water. Concurrently, the aliquots of these solutions were added to deionized water. The pH measurements of the medium surrounding the gel and the water were performed 10 min after adding an aliquot of the acid or base solution. The titration curve of water in Figure 3A shows that starting from the initial pH 7.4, a strong change in pH was observed upon acid or base addition, while the change in pH of the medium surrounding EKGel‐H was significantly shallower. Given that the pK_a_ of the imidazole group in histidine is 6.1,[48] the histidine residues contributed to the buffering effect primarily within the pH range of 5.1–7.1. Notably, the initial pH of the medium surrounding EKGel‐H was 5.8, as the sulfate groups of aH‐CNCs and the intrinsic acidity of gelatin led to a weakly acidic medium. The results for EKGel are shown in Figure S10, Supporting Information. Within the same pH range, the buffering efficiency of EKGel was markedly weaker than that of EKGel‐H.
(A) Titration curves for deionized water and the medium surrounding EKGel‐H. (B) Variation in the pH of the medium surrounding EKGel‐H after adding 0.5 µmol NaOH, shown at varying histidine concentrations in EKGel‐H. (C) Variation in the buffering capacity of EKGel‐H, plotted as a function of the gel concentration. The solid line is given for eye guidance. In A and C, the buffering tests were performed for 1 mL of EKGel‐H surrounded by 1.5 mL of water. The concentrations of aH‐CNCs and gelatin were 1.0 wt.% and 1.3 wt.%, respectively, corresponding to 0.028 mmol g−1 of aldehyde groups and 0.0047 mmol g−1 of amine groups.
The buffering capacity of EKGel‐H as a measure of its resistance to pH change was determined as
where B is the buffering capacity (mol L^−1^), C B is the concentration of the added acid or base (mol/L), V B is the volume of the acid or the base (L) added, ΔpH is the corresponding change in pH, and V g is the volume of the hydrogel (L) [54]. Figure 3B shows that the buffering capacity of the EKGel‐H in the pH range of 5.8–7.4 strongly increased at a higher concentration of histidine in the gel.
The buffering capacity of EKGel‐H in the pH range of 5.5–7.4 was also increasing with aH‐CNC/gelatin content in the gel (Figure 3C). With the EKGel‐H concentration changing from 0.16 to 0.64 g/mL, its buffering capacity increased from 0.2 × 10^−3^ to 10.2 × 10^−3^ mol/L, respectively. The results shown in Figure 3B,C highlight the ability to control EKGel‐H's buffering properties and the potential utility of this gel for controlling the pH of the PDTO microenvironments.
Growth of PDTOs in EKGel‐H
2.4
The EKGel‐H properties were first explored in the cell culture media. The composition of the cell culture media for breast cancer PDTOs is detailed in the Supporting Information. Notably, the medium contains phenol red as a built‐in pH indicator for the pH range of 8.2 to 6.8 [55, 56]. The initial pH of the medium was 7.51, measured under 5% CO_2_ atmosphere.
The hydrogel containing 1.0 wt.% aH‐CNCs with the concentration of histidine groups of 11.4 mmol/g and 2.0 wt.% of gelatin formed after incubating the aH‐CNC‐8 1/gelatin mixture at 37 °C for 2 h. The concentration of aH‐CNC/gelatin imine crosslinking groups in this gel was 3.6 mmol/L. A representative SEM image in Figure 4A shows that the hydrogel had a fibrous structure with average fiber diameter of 28 ± 1 nm.
(A) Scanning electron microscopy image of cell culture media‐swollen EKGel‐H containing 1.0 wt.% aH‐CNCs with the concentration of histidine groups of 11.4 mmol/g and crosslinked with 2.0 wt.% of gelatin. Scale bar is 1 µm. (B) Variation in the shear storage modulus, G′, and loss modulus, G″, of the EKGel‐H swollen in culture medium over gelation time. (C) Live–dead assay of patient‐derived cells (DCXBTO.58 (ER‐/PR‐/HER2‐) encapsulated in EKGel‐H for 4 h (day 0) and 4 days, stained with calcein‐AM (live cells) and Propidium Iodide (dead cells). Scale bar is 100 µm.
The values of G′ and G′′ of the aH‐CNC/gelatin mixture in the cell culture medium reached 140 and 7 Pa, respectively, after 7 h incubation and changed insignificantly to 156 and 8 Pa, respectively, after 25 h (Figure 4B; Figure S11). These markedly higher G′ and G′′ values, compared to the water‐swollen EKGel‐H at the same time points (Figure 2B) were attributed to the ionically mediated physical crosslinking of aH‐CNCs in the cell culture medium,[57, 58] which occurred in addition to the formation of covalent imine crosslinks. In contrast, the values of G′ and G′′ of the EKGel in the cell culture medium after 7 h were 105 and 6 Pa, respectively. However, after 24 h of incubation at 37 °C, both G′ and G″ of the media‐swollen EKGel increased substantially, reaching 264 and 17 Pa, respectively (Figure S12).
The biocompatibility of EKGel‐H was examined for the metastasis‐derived ER‐/PR‐/HER2‐ (“triple negative”) patient‐derived breast cancer model DCXBTO.58 by staining live/dead cells with calcein‐AM (live cells) and Propidium Iodide (dead cells). Figure 4C shows high cell viability in EKGel‐H. The biocompatibility results of EKGel (used as a reference in PDTO growth) are shown in Figure S13.
The EKGel‐H swollen in the breast cancer organoid media was used as a scaffold for long‐term culture of PDTOs formed from patient line DCXBTO.58 or primary patient line BPTO.95 (see Table S2 for details). A recently reported microfluidic Recoverable‐Spheroid‐on‐a‐Chip with Unrestricted External Shape (ReSCUE) platform was used to generate uniformly sized rod‐shaped PDTOs in EKGel (Figure S14) [59]. After 4‐day culture, the PDTOs with dimensions of ∼1.8 mm × 0.4 mm × 200 µm were released from the microfluidic device, transferred into a chambered cell culture slide (one PDTO per chamber), and cultured for 21 days in unconstrained conditions in 100 µL either EKGel‐H or EKGel, with cell culture medium refreshed every 7 days. The details of these experiments are provided in Supporting Information. The PDTO areas and shapes were examined at 7‐day intervals.
Over time, initially rod‐shaped PDTOs acquired a sphere‐like shape in both EKGel‐H and EKGel (Figure 5A,B; Figure S15). This effect was characterized by the time‐dependent variation in the PDTO circularity, calculated as the ratio of the PDTO area to the squared PDTO perimeter (Supporting Information) [59, 60]. Figure 5C shows that after 21 days, the average normalized PDTO circularity underwent a 1.4‐ and 1.9‐fold increase for the PDTOs grown in EKGel and EKGel‐H, respectively. The results for the PDTOs derived from BPTO.95 cell line showed a similar trend (Figure S16). The increase in PDTO circularity was consistent with our earlier findings for PDTOs formed from DCXBTO.58 cell line [59] and ascribed to the preferential proliferation of cancer cells in high‐curvature PDTO regions.
(A,B) Brightfield images of the PDTOs formed from the DCXBTO.58 breast cancer cell line over 21‐day culture in (A) EKGel‐H and (B) EKGel. Scale bars are 200 µm. (C) Variation in the average DCXBTO.58 PDTOs circularity over 21‐day culture in EKGel‐H (green symbols) and EKGel (blue symbols). The data are shown as mean ± 95% confidence interval for 10 PDTOs. The average PDTO circularity was normalized by the average circularity on day 0. (D) Confocal microscopy images of DCXBTO.58 PDTO stained with F‐actin (red), Ki‐67 (green), and DAPI (blue) after 21‐day incubation in EKGel‐H (left image) and EKGel (right image). Insets in the images show the fraction of proliferating cells. The data are shown as mean ± standard error for 8 PDTOs. Scale bars are 200 µm. (E) 3D reconstruction image of DCXBTO.58 PDTOs after 21‐day culture in EKGel‐H (left) and EKGel (right), obtained using confocal microscopy. Scale bars are 500 µm.
To assess the proliferative activity of the PDTOs cultured in EKGel and EKGel‐H, immunostaining using Ki‐67 was performed after 21‐day culture. Quantitative analysis revealed a higher fraction of proliferating cells in the PDTOs cultured in EKGel‐H (45 ± 3%), compared to those in EKGel (25 ± 3%) (Figure 5D). The higher fraction of Ki‐67‐positive proliferative cells in the PDTOs grown in EKGel‐H relative to those cultured in EKGel was evident through the entire PDTO body (Figure S17). Collectively, the results shown in Figure 5 indicate that EKGel‐H provided a more favorable environment for the spatial organization and proliferation of tumor cells within the PDTOs.
Notably, while the increase in the average area of the PDTOs grown in EKGel and EKGel‐H was similar over 21 days (Figure S18), the average height of PDTOs cultured in EKGel‐H for 21 days was 379 ± 19 µm, compared to 260 ± 13 µm in EKGel (Figure 5E) which corresponded to the estimated difference between the PDTO volumes of approximately 0.10 ± 0.02 cm^3^ (Figure S19).
During the course of 21‐day PDTO growth (Figure 6A) and up to further 36‐day culture (Figure S20), the EKGel‐H maintained its shape and uniformity, with no visible sign of degradation or detachment from the cell culture chamber walls. In contrast, starting from day 7 the volume of EKGel in the chamber underwent a significant decrease. Analysis of the culture medium surrounding each hydrogel using sulfuric acid fuchsin (Schiff's reagent for aldehydes) revealed a‐CNC leaching from the EKGel, while no aH‐CNC release was detected for the media surrounding EKGel‐H (Figure 6B).
(A) Photographs of EKGel and EKGel‐H containing PDTOs, taken after removing the cell culture medium. Scale bar is 1 mm. (B) Absorbance spectra of the cell culture media surrounding DCXBTO.58 PDTO in EKGel and EKGel‐H, acquired on day 21. (C, D) Representative pH maps of the DCXBTO.58 PDTOs encapsulated in EKGel‐H (C) and EKGel (D). The images were acquired following media exchange from day 14 of culture on days 14–21. The images were obtained by ratiometric analysis of the transmitted light at 440 and 560 nm (corresponding to the absorbance maxima of the acidic and basic forms of Phenol Red, respectively) and converted to pH values using a calibration curve. Scale bar is 400 µm.
Since the most significant pH changes were expected in the PDTO vicinity, we utilized the spectroscopy‐based mapping to examine the local acidification of the PDTO environment in EKGel and EKGel‐H. Prior to the imaging, the PDTOs were cultured for 14 days, with fresh medium replacement every 7 days. Figure 6C,D shows representative ratiometric transmission images of the PDTOs over 7 days. The color bars in these images reflect the pixel‐by‐pixel ratio of the transmitted light intensities at 440 and 550 nm, corresponding to the absorbance maxima of phenol red (the pH indicator in the media) in its acidic and basic forms, respectively [56, 61]. The ratio of the intensities of transmitted light was converted to pH values using a calibration graph (Figure S21), which enabled the visualization of degree of acidification in the vicinity of the PDTO–hydrogel interface. Following the introduction of fresh media on day 14, the pH value measured after 6 h at the distance of 50 µm from the PDTO surface was 7.61 ± 0.04 and 6.42 ± 0.02 in EKGel‐H and EKGel (Figure 6C,D and Video S1, S2, respectively). On day 21, the corresponding values of pH were 7.02 ± 0.04 and 5.99 ± 0.03.
A strong difference in the local pH (more than 1.0) unit between highlighted the effective buffering capacity of the histidine‐modified hydrogel. Notably, we also observed strong and similar acidification at the inner periphery of the PDTOs in both hydrogels, due to the higher density of proliferating cells in this PDTO region.
To examine the impact of medium acidification on the hydrogel integrity, we studied the relationship between the DCXBTO.58 PDTO growth over 21‐day culture, the corresponding variation in the average pH of the surrounding medium, and the changes in the structure and mechanical properties of EKGel‐H and EKGel (Figure 7A–F).
(A) Variation in the DCXBTO.58 PDTO area over 21‐day culture in EKGel‐H and EKGel. The data are shown as mean ± standard error for 4 PDTOs. The PDTO area was normalized by the area on day 0. (B) Variation in pH of the cell culture media during 21‐day DCXBTO.58 PDTO growth in EKGel‐H and EKGel. The data are shown as mean ± standard error for 4 PDTOs. (C) Variation in the concentration of imine crosslinking groups in EKGel‐H and EKGel over 21‐day culture of DCXBTO.58 PDTOs. All concentrations were normalized by the imine group concentration on day 0. The data are shown as mean ± standard error for 3 PDTOs. (D) Variation in the normalized shear storage modulus, G′, of EKGel‐H and EKGel over 21‐day culture of DCXBTO.58 PDTOs. The value of G′ was normalized by G′ of the corresponding gel on day 1. The data are shown as mean ± standard error for 3 PDTOs. (E, F) SEM images of EKGel‐H (E) and EKGel (F) on day 0 and after 21‐day growth of DCXBTO.58 PDTOs. Scale bar is 1 µm. The insets show the average fiber diameter, d, of the corresponding gels as mean for analyzed 150 fibers.
We first assessed the change in the lateral dimensions of PDTOs over time. As shown in Figure 7A, the PDTOs cultured in EKGel and EKGel‐H displayed a similar trend in the expansion of their area, with accelerated growth occurring after day 14. The progression of PDTO area correlated with the change in the medium pH (ΔpH) (Figure 7B), which was calculated as the difference between the average pH values at the start and the end of each 7‐day medium exchange interval. While on day 7, the ΔpH values were 0.14 and 0.22 for EKGel‐H and EKGel, respectively; by day 21, these values increased to 0.52 and 0.90, indicating a more pronounced acidification in EKGel. Thus, despite the larger density of proliferating cells (Figure 5D) and stronger volumetric growth of PDTOs in EKGel‐H (Figure 5E), which would be expected to intensify medium acidification, the average pH of the medium in EKGel‐H remained approximately 0.4 units higher than in EKGel (corresponding to a 2.3‐fold lower H⁺ concentration). Notably, the extent of medium acidification differed between the two PDTO models used in the present work. By day 21, the medium around the BPTO.95 PDTOs exhibited substantially lower ΔpH values (0.26 in EKGel and 0.21 in EKGel‐H) (Figure S22), compared to the PDTOs grown from the metastatic DCXBTO.58 line, reflecting the slower growth rate and lower metabolic acid production by the primary‐tumor–derived BPTO.95 cells. These observations underscore the importance of the buffering capabilities of EKGel‐H for rapidly proliferating, metabolically active PDTOs.
The structural stability of the EKGel‐H network, compared to EKGel, was assessed by monitoring the time‐dependent concentration of imine crosslinks in both gels (Figure 7C). Figure 7C shows the variation in the concentration of imine crosslinking groups normalized to their initial concentration in each gel on day 0 (prior to the PDTO culture). A close‐to‐constant concentration of imine groups was maintained in EKGel‐H over 21‐day culture, while EKGel exhibited a 73% decrease in the content of imine crosslinks by day 21, consistent with their acid‐catalyzed hydrolysis (Figure S23).
Since intrafibrillar imine crosslinking impacts hydrogel stiffness,[62] the pH‐dependent difference in the content of intrafibrillar imine groups in EKGel and EKGel‐H affected the difference in their mechanical properties, as shown in Figure 7D. While EKGel‐H exhibited invariant normalized G′ value over the entire 21‐day duration of PDTO culture, the EKGel underwent a 32% decrease in G′ by day 21.
The difference in the mechanical properties of the two hydrogels correlated with the change in their fibrous structure. In Figure 7E,F, the SEM images of the gels before and after PDTO culture for 21 days revealed that the fiber diameter in EKGel‐H underwent an insignificant change, while in EKGel it decreased by 30%.
In summary, a biomimetic fibrous EKGel‐H with built‐in buffering capability enabled control of the pH of the environment surrounding PDTOs in the range of 6.5–6.9, characteristic of the extracellular environment for human breast tumors in vivo [63]. We admit that while breast cancer tumoroid medium contains 4‐(2‐hydroxyethyl)‐1‐piperazineethanesulfonic acid (HEPES) that acts as a buffering agent, its effective range limited to pH ≥ 6.8 [20] was insufficient to counteract the excessive acidification of the medium. In contrast, EKGel‐H maintained a 2.8 lower concentration of protons than in EKGel (corresponding to the average pH of the culture medium approximately 0.4 units higher (Figure 7B), while the pH difference in the immediate neighborhood of the PDTOs for these two gels reached approximately 1.0 unit on day 21 (Figure 6D).
The proliferative activity of cells in the PDTOs was substantially enhanced when they were cultured in EKGel‐H, in comparison with EKGel. After 21 days, the PDTOs grown in EKGel‐H contained a nearly twofold higher fraction of Ki‐67‐positive proliferating cells, which suggested that without histidine a stronger acidic environment and/or degradation of the supportive matrix caused reduced cell proliferation or quiescence. We admit, however, that the incorporation of histidine in EKGel resulted in its 45% reduced stiffness. To assess the effect of this change on PDTO growth, we cultured PDTOs in two EKGel‐H scaffolds with a storage modulus G′ of 130 and 220 Pa and observed no statistically significant difference in PDTO (Figure S24). Yet, the change in the chemical environment with introduction of histidine could contribute to cell proliferation, in addition to the buffering performance of EKGel‐H.
In addition, after 21‐day culture, the PDTOs grown in the buffering gel had a larger volume. Together, a greater fraction of proliferative cells and a stronger PDTO expansion highlight the capacity of EKGel‐H to provide a more favorable and homeostatic niche.
The enhanced proliferative activity of the PDTOs grown in EKGel‐H likely contributed to the stronger increase in their circularity, which could occur via two synergistic mechanisms. First, a higher fraction of Ki‐67‐positive cells in the PDTOs grown in EKGel‐H could accelerate cell expansion in higher curvature regions. Second, a higher density of proliferating cells in these PDTOs could amplify internal compressive stresses and favor the reduction in interfacial tension [59, 64]. The resulting 1.9‐fold increase in the circularity of PDTOs cultured in EKGel‐H, compared to a 1.4‐fold increase in EKGel reflected both biological and mechanical advantages conferred by the buffering hydrogel.
In addition to enhanced PDTOs culture, the EKGel showed the mechanical, chemical, and structural stability during the course of 21‐day PDTO culture, thus, addressing the challenge in changing hydrogel properties due to the acidification of the medium. The design of this hydrogel involved a balance between the attachment of the high amount of histidine molecules to the a‐CNCs (to confer a strong buffering capacity of the gel) without compromising the ability to crosslink a‐CNCs with gelatin. Importantly, the stoichiometric consumption of the aldehyde groups of a‐CNCs led to the hydrogel's biocompatibility, as non‐consumed aldehyde groups could impose cytotoxicity by reacting with nucleophilic groups of proteins and nucleic acids and disrupting essential cellular functions [51].
By recapitulating the level of acidification of the microenvironments of solid tumors, the buffering EKGel‐H can provide enhanced modeling of tumor growth or their response to therapies, particularly, over long‐term PDTO culture. Enhanced control of extracellular conditions may be particularly important for the evaluation of the rapidly expanding class of antibody‐drug conjugates, which is becoming a major form of anticancer therapy. In such therapies, a potent cytotoxic payload is linked to an antibody with linkers that can be influenced by acidic conditions.
Conclusion
3
The reported biomimetic buffering hydrogel enabled control of the local acidification of the PDTO environment to the physiologically relevant level. Such control led to enhanced cell proliferation and stronger PDTO growth over 21 days. The structural and mechanical resilience of the buffering hydrogel against the acidification‐induced gel degradation led to a chemically and mechanically stable microenvironment for PDTO growth and enabled their long‐term PDTO applications, such as studies of tumor evolution, drug resistance mechanisms, and long‐term drug screening applications. The principles of the hydrogel design, that is, the integration of the buffering capacity without compromising the hydrogel's structural integrity, offer a valuable strategy for the future development of advanced hydrogel scaffolds incorporating dynamic pH‐sensitive covalent bonds. By reproducing the pH‐buffering function of the native ECM, this hydrogel represents a step toward functional synthetic tumor microenvironments.
Materials and Methods
4
Materials
4.1
Type A gelatin (300 g bloom), L‐histidine powder (≥99% purity, later termed as “histidine”), ethylene glycol (≥99% purity), fuchsine sulfurous acid, and sodium periodate NaIO4 (≥99% purity) were purchased from Sigma–Aldrich, Canada. 1X and 10X formulations of Hank's Balanced Salt Solution (HBSS) were purchased from Fisher Scientific, Canada. An aqueous 10.6 wt.% CNC suspension was purchased from the University of Maine Process Development. All chemicals were used as received without purification.
Modification of CNCs with Aldehyde Groups
4.2
Aldehyde‐functionalized CNCs (a‐CNCs) were prepared by surface oxidation of CNCs with NaIO4 [65]. Briefly, NaIO_4_ was added to 1.0 wt.% CNCs suspension at a mass ratio of 0.6:1 or 4:1 to obtain a‐CNCs_0.6_ and a‐CNCs_4_, respectively. The flask was covered with aluminum foil to prevent photodecomposition of NaIO_4_, and the mixture was stirred at room temperature for 2 h (to obtain a‐CNCs‐2) or 8 h (to obtain a‐CNCs‐8) (see Table S1). The reaction was quenched with 600 µL of ethylene glycol. The resulting suspension of a‐CNCs was dialyzed (cellulose membrane, 12 kDa cutoff) for two weeks against Milli‐Q grade deionized water (DI, 18.2MΩ•cm resistivity) with the water being changed three times on day 1 and twice daily thereafter. Subsequently, the a‐CNC suspension was concentrated to >3.0 wt.% using rotary evaporation and the solid content was then determined by gravimetric analysis. The suspensions were diluted with deionized water to the desired concentration.
The amount of aldehyde groups on the a‐CNC surface was determined by titrating them with hydrochloric acid that was released during the reaction of a‐CNCs with hydroxylamine using protocol described elsewhere [66].
Modification of a‐CNCs with Histidine
4.3
Aldehyde‐modified CNCs modified with histidine molecules (aH‐CNCs) were prepared via the reaction between a‐CNCs and histidine. The histidine was added as a powder to the 3.0 wt.% of a‐CNCs_4_‐2 or a‐CNCs_4_‐8 suspension at a mass ratio of 1:1, 1:2 or 1:4 to obtain aH‐CNCs‐2 1, aH‐CNCs‐8 1; aH‐CNCs‐2 2, aH‐CNCs‐8 2, and aH‐CNCs‐2 4, aH‐CNCs‐8 4, respectively (see Table S1). The mixture was vortex‐mixed for 3 min and incubated at 37 °C for 40 min. The suspension of aH‐CNCs was dialyzed for two weeks against deionized water (18.2MΩ•cm resistivity) with the water being changed twice daily using cellulose membrane, 12 kDa cutoff). The concentration of aH‐CNCs was determined by gravimetric analysis.
The amount of residual aldehyde groups in aH‐CNCs was determined by analyzing the change in absorbance intensity of 1.0 wt.% suspension of aH‐CNCs with sulfuric acid fuchsin (Schiff's reagent for aldehydes) [67, 68]. The intensity of absorbance at λ = 565 nm was measured using a Varian Cary 5000 UV–vis Spectrometer. The concentration of histidine in aH‐CNCs was determined as the change in the concentration of aldehyde groups before and after a‐CNC modification with histidine.
Quantification of the Content of Amine Groups in Gelatin
4.4
The concentration of amine groups in gelatin was determined by using the 2,4,6‐trinitrobenzylsulfonic acid (TNBS) assay [69, 70]. Gelatin was dissolved in 0.1 m sodium carbonate buffer (pH = 9.0) to a final concentration of 0.5 wt.%. Subsequently, TNBS was added to this solution to a final concentration of 0.1 w/v%, and the mixed solution was incubated for 4 h at 37 °C. The absorbance of this solution was measured at λ = 500 nm using a CLARIOstar plate reader. To determine the concentration of amine groups in gelatin, a calibration curve was constructed using glycine standard solutions.
EKGel‐H Preparation
4.5
Water‐Swollen EKGel‐H
4.5.1
A 1.1 wt.% aH‐CNC suspension and a 10 wt.% aqueous gelatin solution were mixed to obtain suspensions containing 1.0 wt.% of aH‐CNCs and 0.5‐3 wt.% of gelatin. The resulting suspensions were incubated for 2.5 h at 37 °C, which was followed by the inversion test. The EKGel‐H with the composition 1.0 wt.% aH‐CNCs‐8 1 and 1.3 wt.% gelatin was used for subsequent FTIR, rheological, and SEM characterization, unless otherwise specified.
Media‐Swollen EKGel‐H
4.5.2
The EKGel‐H swollen in the breast cancer organoid medium was prepared by vortex‐mixing a 2.5–3.0 wt.% aH‐CNCs‐8 1 suspension in 1X HBSS, 10 wt.% gelatin solution in breast organoid medium and breast organoid medium to form a mixture containing 1.0 wt.% of aH‐CNCs and 2.0 wt.% of gelatin. The mixed suspension was incubated for 2.5 h at 37 °C for gelation.
The same procedures were performed to prepare EKGel using a‐CNC_0.6_‐2 suspension. This 1.0 wt.% CNCs/2.0 wt.% gelatin formulation was previously used for PDTO growth [23]. The aH‐CNCs‐8 1 were selected because of the high histidine content, which provided a high buffering capacity and a moderate residual aldehyde concentration (2.8 mmol g^−1^), comparable to that of a‐CNC_0.6_‐2 (0.9 mmol g^−1^), thereby enabling their efficient and cytocompatible crosslinking with gelatin in EKGel‐H.
Transmission Electron Microscopy
4.6
A JEOL TEM‐210 high‐resolution transmission electron microscope (TEM) at 80 kV was used to image CNCs, a‐CNCs, and aH‐CNCs. A 2 µL droplet of 1.0 wt.% CNC, a‐CNC or aH‐CNC suspension was deposited on a plasma‐treated 200 mesh carbon‐coated copper grid (supplied by Electron Microscopy Sciences) and dried under ambient conditions. The average diameter and length of CNCs, a‐CNCs, and aH‐CNCs were determined by image analysis using ImageJ software (NIH).
Fourier‐transform Infrared Spectroscopy
4.7
The chemical structures of a‐CNCs, aH‐CNCs, and EKGel and EKGel‐H were characterized using attenuated total reflectance Fourier transform infrared spectroscopy (ATR‐FTIR) using a Bruker Vertex 70 spectrometer with a 1.85 mm diameter diamond crystal. The respective suspensions and gels were freeze‐dried, and the spectra were acquired for the powder. The FTIR spectra were recorded in the transmittance mode.
Scanning Electron Microscopy
4.8
The hydrogel structure was examined by scanning electron microscopy (SEM). Supercritical point drying was used to prepare the hydrogel samples. The solvent (water) was replaced with ethanol by sequentially immersing the hydrogels for 30 min in 30, 50, 70, and 90% (v/v) ethanol/water mixtures and subsequently, in anhydrous ethanol. Ethanol was removed from the samples via supercritical point drying in an Autosamdri‐810 Tousimis critical point dryer. The samples were fractured, gold‐coated by using an SC7640 high‐resolution sputter coater (Quorum Technologies) and imaged on a Quanta FEI scanning electron microscope (5 kV).
Buffering Performance of EKGel‐H
4.9
The titration curves of water‐swollen EKGel‐H and EKGel were obtained by using incremental addition of a 0.01 m solution of HCl (acid) or NaOH (base) in 0.1 µmol portions to 1.0 mL of the hydrogel submerged in 1.5 mL of the deionized water (pH = 7.0) under constant stirring. The pH of the medium surrounding the gel was measured 10 min after each acid or base addition using a Starter 300 pH Portable pH meter with an ST210 electrode (Ohaus, USA). The same procedure was used to titrate 2.5 mL of the deionized water.
To determine the buffering capacity of water‐swollen EKGel‐H, the gel samples with a different mass varying from 0.4 to 1.6 g were submerged into 2.5 mL of the deionized water to achieve the gel concentration of 0.16, 0.32, 0.48, 0.56, or 0.64 g/mL of gel. To test the buffering capacity, 0.5 µmol of either 0.1 m NaOH solution (to assess the basic buffering capacity) or 0.1 m HCl solution (for the acidic buffering capacity) was added to the submerged gels. The pH value of the medium surrounding the gel was measured after 1.5 h stirring. To assess the dependence of the buffering capacity on histidine concentration in the EKGel‐H, the same procedure was applied to 0.32 g/mL samples of EKGel‐H in water prepared from aH‐CNCs‐8 1, aH‐CNCs‐8 2, and aH‐CNCs‐8 4, as well as in the EKGel.
Preparation of Breast Cancer Organoid Medium
4.10
Organoid media was stored at 4 °C. Breast organoid medium was prepared as described elsewhere [71]. A detailed list of the components is provided in Table S3.
Image Acquisition and Analysis
4.11
During the 21‐day culture period, brightfield images of PDTOs were captured every 7 days using a Nikon Ti Eclipse microscope. The PDTOs areas and perimeters were analyzed using a Python script (https://github.com/Fatemeh‐EngBio/PDTO‐image‐processing). The variation in the normalized PDTO area calculated as the ratio of PDTO area at each time point to the initial PDTO area, was plotted as a function of culture time. The circularity as a measure of the PDTO roundness was calculated as the ratio of the average PDTO area to the average PDTO perimeter squared (see Supporting Information).
Immunofluorescence Staining and Confocal Imaging
4.12
The PDTOs were washed with 1X HBSS three times to remove the cell culture medium and subsequently fixed with 400 µL of 5.0 wt.% formalin for 30 min. The PDBTOs were then submerged for 30 min into 0.1 m solution of glycine solution in HBSS (400 µL), which was subsequently replaced for 20 min with 400 µL of 0.5 vol% solution of Triton‐X‐100 in 1X HBSS to permeabilize the cells. The PDTOs were then washed three times with 400 µL of immunofluorescence washing solution (0.05 wt.% sodium azide, 0.1 wt.% bovine serum albumin, 0.2 vol% Triton‐X‐100, and 0.05 vol% Tween‐20 in HBSS). A primary blocking solution (10 wt.% goat serum in immunofluorescence solution) was then added for 90 min to the PDTOs and subsequently, replaced with 250 µL of fluorophore‐conjugated antibody solution. For F‐actin and Ki‐67 staining, AlexaFluor 488 Ki‐67 rabbit monoclonal antibody and AlexaFluor 568 Phalloidin (Life Technologies) were diluted at the volume ratio of 1:50 with the primary blocking solution. The stained PDTOs were incubated overnight at 4 °C and then submerged into 400 µL of the immunofluorescence washing solution for 40 min. For nuclei staining, Hoechst or DRAQ5 was added in 1X HBSS (1:1000) for 2 h. Finally, the stained PDTOs were then stored at 4 °C in 1X HBSS. The Leica SP8 STELLARIS confocal microscope (Leica Microsystems, Wetzlar, Germany) was used to acquire high‐resolution fluorescence images of the stained PDTOs.
Ki‐67 expression was quantified for both EKGel‐H and EKGel scaffolds using an automated image analysis pipeline in Python (https://github.com/Fatemeh‐EngBio/PDTO‐image‐processing).
Monitoring Hydrogel Stability
4.13
To assess time‐dependent changes in gel morphology, EKGel and EKGel‐H encapsulating PDTOs were incubated in breast cancer organoid medium for up to 36 days. On day 0, 7, 14, 21, 28, and 36, the culture medium was carefully removed from each chamber using a pipette, and a photograph of the gel was taken immediately before the addition of fresh medium.
Ratiometric pH Imaging
4.14
To assess the dynamics of local media acidification during PDTO culture, transmitted light images of the EKGel‐H and EKGel‐encapsulating PDTOs were acquired using a Zeiss Axio Observer inverted microscope (Zeiss, Germany) over a 7‐day period, starting on day 14 of the PDTO culture and immediately after medium replacement. Images were taken every 6 h at 37 °C and 5% CO_2_ using sequential illumination at 440 nm (channel 1) and 560 nm (channel 2). Pixel‐by‐pixel ratios of transmitted light intensities (I_440_/I_560_) were calculated from paired images and converted to pH values using a calibration graph. To construct this graph, the EKGel sample was incubated in the PDTO culture medium with pH adjusted to the pH values in the range of 3.5–8.0 by adding 1 m HCl or NaOH solutions. The pH value of each medium sample was measured using a calibrated pH meter (accuracy ± 0.01 pH units). The calibration graph was constructed by plotting the resulting I_440_/I_560_ ratios of transmitted light intensities as a function of pH (Figure S21).
With regards to the imaging experiments, for each PDTO sample, 28 sequential images were collected and compiled into a time‐lapse video showing the progression of PDTO growth and the corresponding acidification of the local microenvironment (Videos S1 and S2).
Determination of the pH of the Medium During PDTO Cultures
4.15
The change in acidity of the medium during PDTO culture was monitored by measuring its absorbance spectra, as it contains a phenol red pH indicator [61]. Briefly, prior to medium replacement every 7 days, 200 µL of the medium was collected and transferred to a 24‐well plate. The absorbance spectra were recorded in the spectral range of 300–800 nm using a CLARIOstar plate reader. The pH value was determined by relating the absorbance intensity ratio at 435 and 559 nm, I_435_/I_559_, to the calibration graph (Figure S25).
Determining the Concentration of Imine Crosslinks in the Hydrogels
4.16
To assess the change in the concentration of imine crosslinks in EKGel‐H and EKGel during PDTO culture, on day 7, 14, or 21, the hydrogels were transferred from the culture chambers into a 24‐well plate (50 µL per well). In parallel, a 2.0 wt.% gelatin solution and 1.0 wt.% aH‐CNC‐8 1 suspension were prepared. Next, 600 µL of 0.2 w/v% TNBS solution in a 0.1 m sodium carbonate buffer (pH 9) was added to each well. The samples were stained at room temperature for 20 min, and their absorbance at λ = 500 nm was recorded using a CLARIOstar plate reader. To minimize light scattering by the gels, during TNBS staining they were placed in tissue culture plate inserts with a polycarbonate membrane with pore size of 0.4 µm. To determine the concentration of amine groups in the EKGel‐H and EKGel samples, a calibration curve constructed from standard solutions of glycine was used. The concentration of imine crosslinks was calculated by subtracting the measured amine group content in the hydrogel from the initial content of amine groups in the precursor components.
Rheology Experiments
4.17
The variation in the shear moduli of the EKGel‐H and EKGel samples encapsulating PDTOs was assessed on day 7, 14, or 21. The cell culture medium and the PDTO were removed under a microscope from the PDTO culture chamber using a pipette. The remaining hydrogel was transferred onto the rheometer plate. Rheological measurements were performed using a plate rheometer (AR‐1000, TA Instruments) equipped with an integrated Peltier plate for temperature control and a solvent trap to minimize evaporation. Experiments were conducted at strain of 1.0%, frequency of 1 Hz, and temperature of 37 °C.
Dynamic Light Scattering
4.18
Dynamic light scattering (DLS) was used to determine the effective hydrodynamic diameter (D_h_) of CNCs, a‐CNCs, and aH‐CNCs. The measurements were performed using a Zetasizer Nano ZS instrument (Malvern Instruments, UK) at a backscattering angle of 173°. The samples were prepared as 1.0 wt.% suspensions in Milli‐Q grade deionized water.
Statistical Analysis
4.19
All data in the Results section are presented as mean ± 95% confidence interval, unless otherwise specified. Statistical significance was determined using one‐way ANOVA test. The differences with p > 0.05 were considered not significant (ns); 0.01 < p ≤ 0.05, significant (^^); 0.001 < p ≤ 0.01, highly significant (^^); 0.0001 < p ≤ 0.001, very highly significant (^^); and p ≤ 0.0001, extremely significant (^****^). All ANOVA tests were performed using Origin software. For the circularity analysis of the PDTOs the sample size was n = 10. For the fraction of proliferating cells, growth, and pH analyses, n = 8 was used. For the ratiometric imaging, G′, and imine group concentration analyses, n = 3 was used.
Conflicts of Interest
D.W.C. reports consultancy and advisory relationships with AstraZeneca, Daiichi Sankyo, GenomeRx, Gilead, GlaxoSmithKline, Inivata/NeoGenomics, Lilly, Merck, Novartis, Pfizer, Roche, and SAGA and research funding to their institution from AstraZeneca, GenomeRx, Guardant Health, Grail, Gilead, GlaxoSmithKline, Inivata/NeoGenomics, Knight, Merck, Pfizer, ProteinQure, and Roche. The other authors declare that they have no competing interests.
Supporting information
Supporting file 1: adhm70610‐sup‐0001‐SuppMat.docx
Supporting file 2: adhm70610‐sup‐0002‐VideoS1.mp4
Supporting file 3: adhm70610‐sup‐0003‐VideoS2.mp4
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1H. Clevers and D. A. Tuveson , “Organoid Models for Cancer Research,” Annual Review of Cancer Biology 3 (2019): 223–234, 10.1146/annurev-cancerbio-030518-055702. · doi ↗
- 2P. W. Nagle , J. T. M. Plukker , C. T. Muijs , P. van Luijk , and R. P. Coppes , “Patient‐derived Tumor Organoids for Prediction of Cancer Treatment Response,” Seminars in Cancer Biology 53 (2018): 258–264, 10.1016/j.semcancer.2018.06.005.29966678 · doi ↗ · pubmed ↗
- 3S. F. Boj , C.‐I. Hwang , and L. A. Baker , “Organoid Models of human and Mouse Ductal Pancreatic Cancer,” Cell 160 (2015): 324–338, 10.1016/j.cell.2014.12.021.25557080 PMC 4334572 · doi ↗ · pubmed ↗
- 4E. Driehuis , K. Kretzschmar , and H. Clevers , “Establishment of Patient‐derived Cancer Organoids for Drug‐screening Applications,” Nature Protocols 15 (2020): 3380–3409, 10.1038/s 41596-020-0379-4.32929210 · doi ↗ · pubmed ↗
- 5X. Ren , W. Chen , Q. Yang , X. Li , and L. Xu , “Patient‐derived cancer organoids for drug screening: Basic technology and clinical application,” Journal of Gastroenterology and Hepatology 37 (2022): 1446–1454, 10.1111/jgh.15930.35771719 · doi ↗ · pubmed ↗
- 6R. Zent and A. Pozzi , Cell–Extracellular Matrix Interactions in Cancer (Springer, 2010), 10.1007/978-1-4419-0814-8. · doi ↗
- 7L. Gu and D. J. Mooney , “Biomaterials and Emerging Anticancer Therapeutics: Engineering the Microenvironment,” Nature Reviews Cancer 16 (2016): 56–66, 10.1038/nrc.2015.3.26694936 PMC 4790726 · doi ↗ · pubmed ↗
- 8Y. Li and E. Kumacheva , “Hydrogel Microenvironments for Cancer Spheroid Growth and Drug Screening,” Science Advances 4 (2018): aas 8998, 10.1126/sciadv.aas 8998.PMC 592279929719868 · doi ↗ · pubmed ↗