Hydrogel Microdroplet Based Glioblastoma Drug Screening Platform

TL;DR

This paper introduces a new drug screening platform using hydrogel microdroplets to study glioblastoma and test treatments like temozolomide in a more realistic and efficient way.

Contribution

A novel hydrogel microdroplet platform for high-throughput glioblastoma drug screening with controlled and physiologically relevant conditions.

Findings

GBM cells in microgels maintain metabolic activity and drug response similar to macro-scale models.

Microgel composition influences GBM cell morphology and drug screening outcomes.

The platform is compatible with liquid handlers for high-throughput screening of therapies.

Abstract

Glioblastoma is the most common primary malignant brain tumor with a 5-year survival rate < 5%. The standard of care involves surgical resection followed by treatment with the alkylating agent temozolomide (TMZ). GBM cells that evade surgery eventually become resistant to TMZ and lead to recurrence of tumors in patients. With only four drugs currently FDA-approved for GBM treatment, there is a need for a clinically relevant model capable of accelerating the identification of new therapies. Microgels are microscale (~10–1000 μm) hydrogel particles that can be used to encapsulate cells in a tailorable 3D matrix. Microdroplets offer short diffusion lengths relative to conventional hydrogel constructs (> 1 mm) to limit spatial distributions of hypoxia and potentially screen therapeutics in a controlled and physiologically relevant environment. Here, we establish a method to encapsulate GBM cells in gelatin and polyethylene glycol (PEG) microgels. We show that microgel composition can affect cell morphology and further, that collections of GBM-laden hydrogels can be used to quantify the effect of single versus metronomic doses of TMZ. GBM metabolic activity is maintained in microgel culture and GBM cells display drug response kinetics similar to previously established literature using macro-scale hydrogel constructs. Finally, we show microgels can be integrated with a liquid handler to enable high-throughput screening using cell-laden microgels.

Figures

A](#F1)). U87MG GBM cells were mixed with a 4% GelMAL precursor solution then run through a flow-focusing microfluidic device resulting in microgels with an elastic modulus of 3–7 kPa and an average diameter of 150 μm (Figure S2A,B). The cell concentration chosen, 2 × 10^6^ cells/mL, resulted in 1–3 cells per microgel (Figure S2C) and provided ample space for cell-matrix interaction and cell proliferation. Microgels were collected and washed then cultured in suspension for processing. Consistent with prior literature [[28](#R28)], we observed maintaining cell viability after microfluidic emuls

B](#F2)). Past day 3, we see a trend of increasing proliferating (EdU+) population and metabolic activity ([Figure 2C](#F2),[D](#F2), respectively). Actin staining of U87MG cells in GelMAL microgels cultured in free suspension ([Figure 2E](#F2)) reveals an increase in cell numbers and formation of cell aggregates, as well as a trend that suggests GBM cell growth may lead to significant GelMAL microgel remodeling and merging by day 5. After 5 days of culture, cells degrade a majority of the microgels and adhere to the bottom of the well to grow in a traditional 2D format. This suggests that ana

A](#F3)). Microgels were cultured for 3 days in free suspension to allow time for GBM cell proliferation to initiate before dosing microgels in culture with TMZ [0, 10, 100 μM]. Drug efficacy was assessed by changes in cell metabolism 2 days post-treatment. U87MG cells encapsulated in GelMAL microgels did not show a change in activity when exposed to a low (subphysiological) dose of TMZ [10 μM] ([Figure 3B](#F3)). However, cells exposed to a supraphysiological dose of TMZ [100 μM] showed a significant decrease in metabolic activity ([Figure 3B](#F3)). Further, repetitions of TMZ exposure to U8

A](#F4)), likely a consequence of the non-degradable nature of PEG-4MAL microgels that keeps cells constrained. We used the extended culture time of PEG-4MAL microgels to examine the effect of metronomic daily dosing of TMZ ([Figure 4B](#F4)). GBM-laden PEG microgels were cultured for 3 days after encapsulation, then were given either a single supraphysiological dose of 100 μM TMZ on day 3 or five daily physiologically relevant consecutive doses of 20 μM TMZ (days 3–7). We assessed changes in cell metabolic activity on day 5, 7, and 9 (2, 4, and 6 days after the initial TMZ dose was administer

A](#F5)). We first assessed the integrity and structure of microgels post-movement, finding microgel shape and cellular integrity were maintained after liquid handler and micropipette ([Figure 5B](#F5),[C](#F5)). Microgels were then maintained in 384 well plate culture for 2 days before we assessed cell metabolic activity, finding no changes in metabolic activity of microgels that were moved either by the liquid handler or those moved manually via micropipette ([Figure 5D](#F5)). These data confirm the ability to integrate GBM-laden microgels with high throughput screening (HTS) automation to

Full text

Introduction

1 |

Glioblastoma (GBM) is the most common and deadly primary malignant brain tumor with a five-year survival rate < 5% [1, 2]. The current standard of care involves maximal safe surgical resection followed by radiotherapy and adjuvant temozolomide (TMZ) [2–4]. Although treatment increases median survival from 6 to 14 months, most patients relapse due to the invasion of therapy-resistant GBM cells beyond the surgical margin [5–8]. A significant challenge for improved clinical outcomes is due to the slow progress of therapeutic advancement from lab to clinic [9–11]. Patient-derived xenografts (PDX), where patient tumor cells are implanted into immunocompromised mice, have shown the ability to preserve primary tumor heterogeneity and molecular composition [12–14], however, these studies can be costly and time-intensive, making them incompatible with many modern high-throughput drug screening (HTS) approaches. Whereas HTS offers the potential to rapidly screen a wide range of doses or drug combinations, the vast majority of these studies rely on two-dimensional (2D) cell culture that cannot recapitulate aspects of the native tumor microenvironment. The GBM tumor microenvironment includes a complex and evolving extracellular matrix (ECM) as well as a range of central nervous system (e.g., pericytes, microglia, astrocytes) and peripheral (e.g., macrophages) cells known to influence disease progression and drug resistance [15, 16]. While hydrogels have been implemented to mimic relevant cell–cell and cell-matrix interactions [17, 18] and are beginning to be used to explore drug resistance [19–22], many current efforts have used either short time frames or supraphysiological compound doses. Thus, there is a critical need for drug screening platforms amenable to HTS approaches that can also replicate aspects of the tumor microenvironment.

Our group has previously developed methacrylamide-functionalized gelatin (GelMA) hydrogels to investigate patterns of GBM invasion and drug response using both conventional cell lines and patient-derived xenograft cells [21, 23–25]. However, the majority of these studies have been conducted in large (~1 mm scale) hydrogel constructs whose fabrication is labor intensive and can require large amounts of both hydrogel material and cell populations. Their large size (1 mm thick; ~5 mm in diameter) can result in diffusional limitations creating a hypoxic core and hindering cell metabolism [26, 27]. Comparatively, microgels are hydrogels formed as microscale (~10–1000 μm) particles that can be used either as distinct units or jammed assemblies referred to as granular hydrogels [28]. Granular hydrogels possess inherent characteristics such as injectability and porosity as well as well-defined rheological properties [29] that have made them an appealing platform for biomedical applications [30]. To date they have been predominantly used as acellular hydrogel particles with cells cultured in voids between particles [28, 30–32]. The encapsulation of cells in microgels has emerged as a versatile tool for tissue engineering and therapy delivery [33–35]. Microgels can be rapidly formed using smaller quantities of starting materials and are not large enough to induce diffusional limitations that reduce the delivery of nutrients, oxygen, or drugs to encapsulated cells [36]. Furthermore, the matrix composition as well as the identity and concentration of cells encapsulated into the microgels can be altered to create discrete microgel populations that can be then mixed [29] to create a more complex model of the tumor microenvironment.

We have recently adapted our gelatin hydrogel system via the inclusion of maleimide crosslinking moieties along the gelatin backbone (GelMAL) to encapsulate primary cells for extended in vitro culture [35]. Here, we evaluate the degree to which hydrogel microdroplets can be integrated with HTS approaches to establish an in vitro GBM drug screening platform. GBM cells are encapsulated in microgels via microfluidic polymerization for precise control over particle size and cell density. We demonstrate microfluidic approaches to monitor cell activity post-encapsulation in vitro as well as successful manipulation (microgel pipettability) via a liquid handler to enable integration with conventional HTS workflows. And finally, we report GBM response to physiologically relevant doses of TMZ on encapsulated GBM cells.

Materials and Methods

2 |

Gelatin Maleimide (GelMAL) Synthesis and Characterization

2.1 |

GelMAL was synthesized as previously described [35, 37]. Briefly, porcine gelatin type A, 300 bloom (Sigma Aldrich, St. Louis, MO) was dissolved in a mixture of 4:5 (v/v) dimethyl sulfoxide (DMSO; Sigma-Aldrich): aqueous medium of pH 6.0 1.0 M 2-morpholinoethanesulfonic acid buffer (MES; Gold Biotechnology, St. Louis, MO) at 40°C. Then, 6× excess N-succinimidyl 3-maleimidopropionate (Tokyo Chemical Industry Co. Ltd) was dissolved in DMSO and added to the vial containing dissolved gelatin and capped. The pH of the reaction was then adjusted to 4.5 using 1.0 N HCl. Reaction proceeded for 24 h at 40°C and 720 rpm; the solution was dialyzed against distilled water acidified to pH 3.25 using 1.0 N HCl, at 40°C for 5 days. Product was frozen and lyophilized, then stored at −20°C until further use. ^1^H NMR was used to measure the degree of functionalization (DOF). GelMAL with DOF between 49% and 51% (data not shown) was used in this study. The compressive moduli of GelMAL macrogels were determined via unconfined compression using an Instron 5943 (Instron, Norwood, MA). Samples were compressed at a rate of 0.1 mm/min, with the Young’s modulus obtained from the linear region of the stress–strain curve (2.5%–17.5% strain) using a custom MATLAB (MathWorks, Natick, MA) code.

Flow-Focusing Microfluidic Device Construction

2.2 |

Polydimethylsiloxane (PDMS) flow-focusing microfluidic devices with a 200 μm nozzle were cast from either silicon or SU8 masters using the Sylgard 184 silicone elastomer kit (Dow Corning, Midland, Michigan) [34, 35]. Ports were incised using a 1 mm biopsy before being bonded directly to glass slides after plasma treatment.

U87MG Cell Culture

2.3 |

U87MG cells (American Type Culture Collection [ATCC], HTB-14) were maintained in DMEM, high glucose (Thermo Fisher Scientific, Waltham, MA) supplemented with 10% (v/v) fetal bovine serum (FBS; R&D Systems, Minneapolis, MN) and 1% (v/v) antibiotic-antimycotic (Thermo Fisher Scientific) in a humidified 5% CO_2_ incubator at 37°C.

Cell-Laden Microgel Fabrication

2.4 |

GelMAL microgels were formed using a flow-focusing microfluidic device with a 200 μm nozzle [34, 35]. 3 vol% SPAN80 (Sigma-Aldrich) in light mineral oil (Sigma-Aldrich) comprised the oil phase. 1,4-dithiothreitol (DTT) (Sigma-Aldrich) was freshly prepared at 20 mg/mL aqueous solution emulsified with the oil mixture at a ratio of 1:15 (v/v). 4.8 wt% GelMAL was dissolved in PBS with 12% (v/v) OptiPrep (Sigma Aldrich) and 0.06% (w/v) Pluronic F-108 (Sigma Aldrich). Addition of OptiPrep and Pluronic F-108 was incorporated to increase the cell medium density and aid cell distribution. U87MG cells were passaged using TrypLE Express (Thermo Fisher Scientific) before treating with ethylenediaminetetraacetic acid (0.52 μM) (EDTA; Thermo Fisher Cat# 15575020) for 15 min at 37°C. Cells were washed with PBS three times and kept on ice until mixed with the GelMAL precursor. The GelMAL precursor and cells were mixed at a 1:5 ratio (final cell concentration: 2 × 10^6^ cells/mL), respectively. The oil mixture, DTT crosslinker, and gel precursor were loaded into syringes then independently pumped into the microfluidic device using syringe pumps (Pump 11 Elite, Harvard Apparatus, Holliston, MA) at rates of 0.5 μL/min (oil mixture), 30 μL/min (crosslinker emulsion), and 5 μL/min (gel precursor + cells), respectively. When encapsulating cells, the volume of precursor pumped through the device was limited to 60 μL to limit the total time cells were exposed to shear stress from the pumps. Microgels were collected in culture medium on an ice bath then shaken at 60 rpm at 4°C for 20 min to allow completion of crosslinking. Microgels were then washed once with PBS, 0.1% (v/v) Tween-20 (Fisher Scientific) in PBS, and culture medium. Washes were performed at 300 rcf × 3 min of centrifugation. PEG microgels were formed as stated previously utilizing PEG-4MAL (JenKem Technology USA Inc., Plano, TX). PEG hydrogels were created with the inclusion of 0.3 mM RGD peptide (MedChem Express, GRGDSPC) containing a free cysteine to facilitate adhesion to the PEG-4MAL monomer (added before mixing with cells).

Viability Measurement of Encapsulated Cells

2.5 |

Non-encapsulated cells and cell-laden microgels were stained with live/dead indicators calcein AM and BOBO-3 iodide (Thermo Fisher Scientific, R37601) and Hoechst 33342 (1:1000) for 15 min before washing and imaging. Microgels were imaged using a DMi8 Yokogawa spinning disk confocal microscope equipped with a Hamamatsu EM-CCD digital camera (Leica Microsystems Inc., Deerfield, IL).

Measurement of DNA Content

2.6 |

Double-stranded DNA (dsDNA) was extracted from non-encapsulated cells and cell-laden microgels using the QIAamp DNA Micro Kit (QIAGEN, Hilden, Germany) following the manufacturer’s protocol, with the following modifications: resulting lysate was passed through a QIAshredder column (QIAGEN). The lysate was then transferred to the QIAamp column, and DNA was extracted according to the kit instructions and eluted in molecular-grade water. Quantification of dsDNA was performed using the Quant-iT PicoGreen dsDNA Assay Kit (Thermo Fisher Scientific), and fluorescence was measured using a Synergy LX Multi-Mode Microplate Reader (BioTek, Winooski, VT) at 485 nm excitation wavelength and 528 nm emission wavelength. dsDNA extracted from cell pellets of 1, 5, or 10 × 10^4^ cells was used to correlate dsDNA levels to cell number (Figure S1A).

Tracing of Morphological Changes in GelMAL Microgels

2.7 |

Microgels were injected into a microfluidic chip (idenTx 3, AIM Biotech, Singapore) to maintain in place and trace morphological changes. Culture medium was changed daily following the manufacturer’s protocol. Microfluidic chips were placed in a humidity chamber in an incubator at 37°C. Microgels were imaged using a DMi8 Yokogawa spinning disk confocal microscope equipped with a Hamamatsu EM-CCD digital camera (Leica Microsystems).

Tracking of Metabolic Activity of Cell-Laden GelMAL Microgels

2.8 |

Cell-laden GelMAL microgels were maintained under the previously described culture conditions. Metabolic activity was assessed using alamarBlue Cell Viability Reagent (Thermo Fisher Scientific). The reagent was added to each well to a final concentration of 10% (v/v) and incubated for 30 min in the dark at 37°C with 5% CO_2_. Fluorescence was measured using a Synergy LX Multi-Mode Microplate Reader (BioTek) at an excitation wavelength of 530 nm and an emission wavelength of 590 nm. Blank controls, including wells containing acellular microgels and wells with only media, were included for each treatment condition.

Proliferation Analysis of Encapsulated Cells

2.9 |

The fraction of proliferating cells was assessed using the Click-iT EdU Cell Proliferation Kit for Imaging (Thermo Fisher Scientific) to measure the fraction of EdU+ cells in microgel culture on day 3, 4, and 5 post-encapsulation. On day 2, 3, and 4, half the volume of media was replenished with fresh culture medium. The other half of medium contained 20 μM EdU (final concentration: 10 μM EdU). Microgels were incubated with EdU for 24 h, after the media was removed and microgels were fixed in formalin for 15 min. The click reaction to attach Alexa Fluor 647 to the alkyne-containing EdU was performed per manufacturer’s instructions. Microgels were stained with Hoechst 33342 (1:1000) for 15 min before washing and imaging.

TMZ Drug Response

2.10 |

Cell-laden microgels were cultured in suspension for 3 days in complete media. After 3 days, microgels were manually pipetted into a 96 well plate with each well containing on average 3.3 × 10^4^ cells (Figure S1B). Temozolomide (TMZ, Selleck Chemicals, Houston, TX) dissolved in 130 mM DMSO (Sigma Aldrich) was diluted in culture medium and added to wells. 0.0077% v/v DMSO was used as a vehicle control to match the highest volume of DMSO present in the treatment group. Changes in cell activity were assessed using alamarBlue Cell Viability Reagent (Thermo Fisher Scientific). For GelMAL microgels, a single dose of TMZ was given on day three and changes in cell behavior were assessed 2 days post-treatment. PEG-4MAL microgels were placed into an HTS Transwell-96 permeable support plate (Corning, Tewksbury, MA) to enable treatment supplements and ease medium changes. PEG-4MAL microgels were treated for 5 days consecutively beginning on day three. Changes in cellular activity were assessed on day five, seven, and nine post-encapsulation.

Morphological Changes in Microgel Culture

2.11 |

Morphological changes in the microgel cultures were assessed by staining actin filaments using phalloidin on day 1, 3, 5, 7, and 9 post-encapsulation. Cell-laden microgels were cultured in free suspension then fixed in formalin for 15 min at room temperature on the respective day. Following fixation, samples were permeabilized with 0.5% Triton X-100 (Sigma Aldrich) in PBS for 15 min. Phalloidin conjugated to Alexa Fluor 488 (1× final concentration from a 400× stock in DMSO; Thermo Fisher Scientific) was diluted in PBS and incubated with the microgels for 30 min at room temperature in the dark. Microgels were stained with Hoechst 33342 (1:1000) for 15 min before washing and imaging.

Liquid Handler Movement of Microgels

2.12 |

Cell-laden PEG microgels were cultured for 3 days in suspension. On day three, microgels were pipetted into a 384 well plate (#781090, Greiner Bio-One, Germany) using the Biomek i5 automated liquid handler (Beckman Coulter Life Sciences, Indianapolis, IN) within our campus High Throughput Screening Facility (Cancer Center at Illinois). Microgels diluted in media were mixed 5 times inside a 15 mL conical tube (Thermo Fisher) by aspirating 2 mm from the bottom at 50 μL/s and dispensing 4 mm from the bottom at 50 μL/s. Microgels were then transferred into a 384 well plate by aspirating 25 μL of diluted microgels 2 mm from the bottom at 20 μL/s and dispensing 1 mm from the bottom at 20 μL/s. Each well is designed to contain ~3.3 × 10^4^ cells, equivalent to the seeding density used for drug-screening experiments. Cell-laden PEG-4MAL microgels were manually pipetted into a 384 well plate as a control. Cell activity was measured at 2 days post-transfer using alamarBlue Cell Viability Reagent (Thermo Fisher Scientific). Microgel integrity was measured by imaging microgels automatically moved by a liquid handler or manually with a micropipette using the EVOS M5000 Imaging System (ThermoFisher Scientific).

Statistics

2.13 |

Statistics were performed using GraphPad Prism (GraphPad Software Inc., Boston, MA). %LIVE population, double stranded DNA levels, metabolic activity, and EdU+ population are reported as the mean ± standard deviation. Normality of data was determined using the Shapiro–Wilk test, and equality of variance was determined using the Brown-Forsythe Test. Data satisfying normality were analyzed using the unpaired t-test, one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test, or two-way ANOVA followed by Dunnett’s post hoc test. When data were not normal, comparisons between groups were performed using a Mann–Whitney test or Kruskal-Wallis test with Dunn’s post hoc test. Significance is denoted as *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. A minimum of three biological samples and three technical replicates were used for all experiments.

Results

3 |

A Method to Reliably Encapsulate GBM Cells in GelMAL Microgels and Retain Viability Post-Encapsulation

3.1 |

We established a consistent and robust method to rapidly encapsulate GBM cells into hydrogel microdroplets of 15 nanoliter volume (Figure 1A). U87MG GBM cells were mixed with a 4% GelMAL precursor solution then run through a flow-focusing microfluidic device resulting in microgels with an elastic modulus of 3–7 kPa and an average diameter of 150 μm (Figure S2A,B). The cell concentration chosen, 2 × 10^6^ cells/mL, resulted in 1–3 cells per microgel (Figure S2C) and provided ample space for cell-matrix interaction and cell proliferation. Microgels were collected and washed then cultured in suspension for processing. Consistent with prior literature [28], we observed maintaining cell viability after microfluidic emulsion required limiting the length of each fabrication run to reduce the level of shear stress experienced by traveling to and through the microfluidic device. For runs of up to 12 min (60 μL of cell and precursor suspension pumped at a rate of 5 μL/min), U87MG cell viability displayed an immediate ~20% decrease (Figure 1B). Although the encapsulation process induced an immediate loss of viability, the live fraction remains stable at 1 and 3-days post-encapsulation (Figure 1C), suggesting the initial cell loss does not have a long-term effect on the culture of our microgels.

Characterization of Cellular Behavior Post-Encapsulation

3.2 |

Next, we set out to define how cellular behavior is impacted by the encapsulation process. We assessed morphological changes after encapsulation, using cell-laden droplets packed into an idenTx 3 microfluidic chip to make repeated measurements of GBM cells (Figure 2A). On the initial day of encapsulation (Day 0), encapsulated cells were largely spherical in shape, consistent with the morphology of cells passage from a flask. On day 1, cells began to display protrusions which became more pronounced by day 3. This elongated morphology is comparable to the morphology observed in both 2D cell culture as well as 3D gelatin hydrogel culture for the U87MG cell line.

We also investigated intracellular activities to further understand the cellular behavior inside microgel culture. Double stranded DNA (dsDNA) content, proliferation levels, and metabolic activity of cell-laden microgels were assessed from samples cultured in free suspension for up to 5 days. dsDNA levels suggest cell proliferation begins on day 3 where we see a significant increase of dsDNA compared with the initial population on day 0 (Figure 2B). Past day 3, we see a trend of increasing proliferating (EdU+) population and metabolic activity (Figure 2C,D, respectively). Actin staining of U87MG cells in GelMAL microgels cultured in free suspension (Figure 2E) reveals an increase in cell numbers and formation of cell aggregates, as well as a trend that suggests GBM cell growth may lead to significant GelMAL microgel remodeling and merging by day 5. After 5 days of culture, cells degrade a majority of the microgels and adhere to the bottom of the well to grow in a traditional 2D format. This suggests that analysis of GBM cell culture in GelMAL microgels may have a maximum timepoint of 5 days.

Assessing Single-Dose Drug Response in GelMAL Microgel Culture

3.3 |

Based on our work that showed GelMAL microgels supported 5-day cell culture, we then performed single TMZ dose response assays to establish a microgel drug screening platform (Figure 3A). Microgels were cultured for 3 days in free suspension to allow time for GBM cell proliferation to initiate before dosing microgels in culture with TMZ [0, 10, 100 μM]. Drug efficacy was assessed by changes in cell metabolism 2 days post-treatment. U87MG cells encapsulated in GelMAL microgels did not show a change in activity when exposed to a low (subphysiological) dose of TMZ [10 μM] (Figure 3B). However, cells exposed to a supraphysiological dose of TMZ [100 μM] showed a significant decrease in metabolic activity (Figure 3B). Further, repetitions of TMZ exposure to U87MG cells in GelMAL culture over different weeks and months resulted in a similar decrease in cellular activity with 100 μM TMZ doses (Figure S3). The recorded changes in cellular activity of GBM cells in response to TMZ demonstrate our ability to reliably study drug response in GelMAL microgel culture.

Enabling Metronomic Dosing in PEG Microgel Culture

3.4 |

To overcome challenges of significant cell-mediated degradation of GelMAL microdroplets and subsequent shorter culture period, we also encapsulated U87MG GBM cells in PEG-4MAL microdroplets. U87MG cells were encapsulated in RGD-functionalized PEG-4MAL, using the same process as GelMAL hydrogels. This expanded the culture timeframe of cell-laden microgels from a maximum of 5 days to 9 days post-encapsulation. The PEG-4MAL microgels fabricated had a lower elastic modulus of 1–3 kPa, higher average diameter of 190 μm, and similar cell distribution of 1–4 cells per microgel (Figure S4A–C), comparative to GelMAL microgels. Actin staining of GBM cells in PEG microgels revealed GBM cells form spherical aggregates within microgels that increased in size with time (Figure 4A), likely a consequence of the non-degradable nature of PEG-4MAL microgels that keeps cells constrained. We used the extended culture time of PEG-4MAL microgels to examine the effect of metronomic daily dosing of TMZ (Figure 4B). GBM-laden PEG microgels were cultured for 3 days after encapsulation, then were given either a single supraphysiological dose of 100 μM TMZ on day 3 or five daily physiologically relevant consecutive doses of 20 μM TMZ (days 3–7). We assessed changes in cell metabolic activity on day 5, 7, and 9 (2, 4, and 6 days after the initial TMZ dose was administered). We observed similar levels of high cell metabolic activity as a result of both dosing regimens at days 5 and 7; however, there was a significant reduction in GBM cell metabolic activity on day 9, 2 days after the final metronomic dose was administered, compared with a single supraphysiological dose of TMZ (Figure 4C–E). These results display the versatility a cell-laden microgel platform can provide to investigate drug response and administration regimens.

Evaluating the Handleability of Microgels With a Liquid Handler to Explore High-Throughput Screening

3.5 |

Successful extended culture and the ability to measure the therapeutic efficacy of alkylating agents (e.g., TMZ) at physiological doses suggest the microgel system may have significant potential to automate testing of large compound libraries against cells within highly controlled and physiologically relevant tissue microenvironments. To assess the integration of our microgel platform with automatic handling, U87MG cells were encapsulated in PEG-4MAL microgels, allowed to culture for 3 days, then 5 μL of PEG-4MAL microgels were moved automatically by a liquid handler or manually with a micropipette into a 384 well plate containing 25 μL of media per well (Figure 5A). We first assessed the integrity and structure of microgels post-movement, finding microgel shape and cellular integrity were maintained after liquid handler and micropipette (Figure 5B,C). Microgels were then maintained in 384 well plate culture for 2 days before we assessed cell metabolic activity, finding no changes in metabolic activity of microgels that were moved either by the liquid handler or those moved manually via micropipette (Figure 5D). These data confirm the ability to integrate GBM-laden microgels with high throughput screening (HTS) automation to expand the scope of drug testing in physiologically relevant matrix environments.

Discussion

4 |

A significant challenge to improving survival rates for glioblastoma remains the identification of new therapeutic approaches. While TMZ improves median survival to ~14 months [6], GBM remains the most deadly primary malignant brain tumor. Reliable and predictive preclinical models that recapitulate aspects of the local microenvironment offer the potential to accelerate the discovery of GBM treatments. Hydrogels are highly tunable and can be tailored to study cell–cell and cell-matrix interactions in a highly controlled environment. Previous studies have encapsulated GBM cells in bulk hydrogels to measure the impact of biomaterial stiffness [38–40], structure [41, 42], and composition [43–45] on morphology, migration, glioma stem cell maintenance, and drug response [46–49]. However, due to their size and the relatively complex process of making and moving large hydrogel volumes between culture conditions, macro-scale hydrogels can exhibit scalability limitations [50] and may not accommodate small cell populations often obtained from patient specimens. The encapsulation of cells into microgels offers an appealing alternative to overcome potential limitations of large hydrogel networks [51] and provides a pathway towards biomaterials-based high throughput drug screening.

We established a robust and consistent method to encapsulate GBM cells into nanoliter-volume GelMAL and PEG-4MAL microdroplets utilizing the U87MG cell line. This cell line has been widely studied in the field to investigate GBM invasion and drug resistance in 2D, bulk culture, and in vivo studies [39, 40, 46, 52, 53]. The extensive information available on U87MG behavior provided multiple sources for the optimization of this platform. Microgel fabrication utilizing microfluidic polymerization allows for precise control over particle size and shape compared with other methods [33, 34, 54]. The stiffness of produced hydrogels (2–7 kPa) was similar to those of GBM patient samples [55], allowing us to recapitulate the native tumor microenvironment. However, achieving high cell viability within microfluidic emulsion is limited by the level of shear stress cells are subjected to while traveling to and through the microfluidic device [28]. By limiting fabrication runs to < 15 min we achieve an encapsulation method that retains ~80% viability of the encapsulated population that exhibits subsequent growth (dsDNA, metabolic activity) in culture. Interestingly, with cellular health intact and a growing demand for culture area, GBM cells showed significant remodeling-associated changes to GelMAL microgels, leading to the formation of interconnected cell-microgel constructs by day five of culture. These results are consistent with macroscale studies of GBM-hydrogel interactions, where GBM cells showed significant matrix remodeling capacity [25, 27, 56]. While potentially interesting from the perspective of using initial microgel cultures as a granular niche to support de novo generation of interconnected tissue constructs at a much larger scale, we chose to limit GelMAL microgel culture for this work to the period where cells remained within the biomaterial environment. While RGD-functionalized PEG-4MAL microgels enabled extended culture of GBM cells, GBM cells in PEG-4MAL microgels formed spherical aggregates. Consistent with a PEG network that does not contain enzymatically degradable motifs but does contain RGD cell adhesion sites, we did not observe evidence of microgel degradation but rather cell aggregation within the hydrogel matrix. And while not shown, we observed in cases of extended culture (> 12 days), GBM cell-laden PEG-4MAL microgels exhibited rupture, suggesting extended cell proliferation without remodeling fractured the microgel network. Future extensions of this work could explore the use of PEG or hyaluronic acid functionalized PEG microgels containing enzymatically degradable motifs in order to explore cell mediated remodeling processes. While such materials may be of value for studies of GBM invasion, the viability and proliferative activity of GBM cells in PEG-4MAL and GelMAL microgels made them an ideal substrate to examine responses to physiologically relevant therapeutic strategies.

Temozolomide (TMZ) is an alkylating agent used to treat GBM [57, 58]. While delivered as a prodrug, it undergoes hydrolysis that induces DNA methylation and subsequent cellular apoptosis. The methylation of DNA is thought to be the principal mechanism responsible for TMZ cytotoxicity [57–60]. TMZ acts on proliferative cells, which has led many in vitro studies to rely on supraphysiological TMZ doses to induce rapid changes in cell viability amenable to conventional in vitro analyses. Recently, we used extended (> 7 day) in vitro culture to measure response to physiologically relevant (< 30 μM) and metronomic TMZ doses in GelMA macrogels [61]. Adapting this protocol, here we encapsulated U87MG GBM cells in GelMAL or PEG-4MAL microgels then cultured the cell-laden microgels for 3 days before TMZ treatments. In GelMAL microgels, U87MGs were unaffected after 48 h at a low, but physiologically relevant dose of 10 μM TMZ, but showed significant reduction in metabolic activity 48 h after a larger dose of 100 μM TMZ. These results are consistent with prior findings for U87MG in 2D culture that showed the half maximal inhibitory concentration (IC50) of TMZ ranges from 92.0 to 590.1 μM [52]. Hence, we demonstrate U87MG cells in microgel culture reveal similar trends in response to low IC50 doses of TMZ. Additionally, we are able to replicate these findings multiple times in the span of weeks and months reliably, showcasing the robustness of our drug screening platform. Future studies can encapsulate patient-derived cell lines with higher drug sensitivity profiles, enhancing patient sample relevance.

While measuring dose–response to supraphysiological TMZ dosages is an important first finding, TMZ concentrations in patient samples often register below 10 μM [62–64]. To achieve the antitumor effect of TMZ, patients instead receive metronomic administrations (e.g., TMZ orally 5 days per week for 4 weeks) [65, 66]. We previously showed macroscale gelatin hydrogels enabled the study of metronomic TMZ dosages [61], but the GelMAL microgel degradation limited the length of our studies. We used RGD-functionalized PEG microgels to extend our experimental timeline to accommodate microgel pre-culture, 5 days of metronomic dosing, then 2 additional days of culture to register responses. Cells were cultured for 3 days before being treated by either a single 100 μM dose or five consecutive daily doses of 20 μM for a cumulative dose of 100 μM TMZ. Interestingly, metronomic dosing exhibited a cumulative negative effect on GBM metabolic activity while a single 100 μM dose did not. While different from the significant effect of 100 μM TMZ on U87MGs in GelMAL microgels, the result is likely induced by the change in cellular morphology caused by our microgel composition. The reported IC50 doses of U87MG spheroids are reported to be higher than those of 2D culture (669.2 to 1500 μM [67, 68]), therefore, the efficacy of a single TMZ dose on aggregates formed inside PEG-4MAL microgels was highly reduced. Here, the ability to pursue an extended study of the effect of metronomic dosing is made possible by cell-laden PEG microgels, offering a platform for future studies using physiologically relevant drug dosages. Subsequent studies can take advantage of this highly tunable platform via the use of PEG-4MAL/GelMAL mixtures or the inclusion of additional components of the tumor extracellular microenvironment such as hyaluronic acid [20, 44, 69] to create a more complex environment tailored for cell remodeling, proliferation, and drug testing.

From a translational aspect, microgels require fewer cells compared with traditional bulk hydrogel culture, an advantage for patient samples that have small cell populations. The encapsulation of samples without prior expansion using traditional flask culture may provide a novel way to process GBM samples and exploit microgel drug screening. The success of PEG-4MAL microgels for examining low physiologically relevant TMZ dosing inspired efforts to integrate microgels with infrastructure essential for high-throughput compound screening. Here, GBM cells encapsulated in microgels could be manipulated via a liquid handler into well-plate formats. GBM-laden PEG-4MAL microgels were cultured 3 days prior to liquid handler manipulation to assess changes in cellular activity versus (low-throughput) manual micropipetting typically used for microgel cultures. Microgel integrity was retained after liquid handler manipulation, with U87MG microgels demonstrating similar levels of cellular activity after liquid handler (vs. manual micropipetting) manipulation. However, there is a large variability in recorded responses from wells seeded with a liquid handler, indicated by the larger error bars. This is likely due to uneven microgel plating by the automatic process. Further work will need to be done to achieve an even microgel number in wells; this can be countered by assessing drug response with an assay that can account for cell numbers in each well. Overall, these results provide confidence for future use of PEG-4MAL microgels for high capacity and efficient studies of drug discovery using cells maintained in well-defined 3D matrix environments.

Conclusions

5 |

Tissue engineering models of the tumor microenvironment have provided a valuable tool to investigate processes of GBM cell matrix remodeling, invasion, and drug response. However, the large size of conventional hydrogel materials precludes their use for high-throughput compound screening workflows. Here, we describe an approach to encapsulate GBM cells in GelMAL and PEG-4MAL microgels. Microgel culture reveals an increase in cell activity and matrix remodeling with time. GelMAL microgels can be used for single dose response studies using TMZ (< 5 days). PEG-4MAL microgels allowed for extended culture and study of clinically relevant metronomic dosing strategies for cells in 3D matrix environments. Lastly, these microgels can be integrated into a conventional workflow (automated liquid handler) necessary to use microgels in high-throughput drug screening facilities. Together, this suggests cell-laden microgels offer a valuable system for high-dimensional studies of drug efficacy with cells maintained in well-defined matrix microenvironments [70, 71].

Supplementary Material

Supplementary Information

Additional supporting information can be found online in the Supporting Information section. Figure S1: Quantification of cells in GelMAL microgels. (A) dsDNA measurements from cell pellets composed of 1, 5, or 10 × 10^4^ U87MG cells was used to create a linear regression model to calculate cell number in microgels. (B) Number of U87MG cells in each well of a 96 well plate based on day post-encapsulation in GelMAL microgels. Graph and bar plot report mean ± standard deviation. n = 3 independent experiments. Figure S2:. (A) Compressive modulus of bulk GelMAL hydrogels composed with 10% (v/v) OptiPrep and 0.05% Pluronic F-108. (B) Diameter of cell-laden GelMAL microgels on initial day of formation. (C) Relative frequency of microgels containing × (0, 1, 2…) cells. Box plot reports median ± minimum and maximum. n = 8 hydrogels. Violin plot reports frequency distribution of microgel diameter. Bar plot reports mean ± standard deviation. A minimum of 100 microgels were measured for each n = 3 independent experiments. Figure S3: Metabolic activity of U87MG cells in GelMAL microgels 2 days post-TMZ treatment [0, 10, 100 μM] at different weeks and months of the year 2025. A minimum of n = 2 independent experiments was recorded for each timepoint. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. Figure S4: (A) Compressive modulus of bulk RGD-functionalized PEG4MAL hydrogels composed with 10% (v/v) OptiPrep and 0.05% Pluronic F-108. (B) Diameter of cell-laden PEG4MAL microgels on initial day of formation. (C) Relative frequency of microgels containing × (0, 1, 2…) cells. Box plot reports median ± minimum and maximum. n = 6 hydrogels. Violin plot reports frequency distribution of microgel diameter. Bar plot reports mean ± standard deviation. A minimum of 100 microgels were measured for each n = 3 independent experiments.