Structural Control for Tunable Hyperthermia-Induced Cellular Responses Using 3D-Printed Platforms

Abstract

Three-dimensional (3D) printing technologies have revolutionized bioengineering by enabling the fabrication of complex, customized structures with high morphological compatibility for specific functions. Most advances in the materials aspect of 3D printing have focused on developing inks that provide high stability and precise deposition for specific printing techniques. A new generation of printable materials not only ensures structural and mechanical integrity, but also incorporates additional functionalities directly into the material. The integration of rational structural design with functional materials offers powerful tools for biomedical applications. In this study, we developed a platform for investigating thermoresponsiveness in cell culture. By inducing controllable, localized heating, we examined the effects of hyperthermia on cancer cells, an emerging treatment modality gaining increasing attention as a promising anticancer strategy. We demonstrate that structurally controlled 3D-printed objects composed of polymer and iron oxide (IO) can generate defined thermal gradients upon exposure to infrared irradiation, thereby inducing differential cellular responses. Using precise spatial control with Digital Light Processing (DLP) printing, we created hyperthermia models. We demonstrated that the experimental conditions can detect changes in cell sensitivity, showing that pre-exposure of cancer cells to the cryoprotective compound trehalose alters their heat resistance. Moreover, repeated thermal cycles promoted the emergence of a cell subpopulation with enhanced heat resistance and increased aggressiveness, highlighting the platform’s ability to drive adaptive cell selection based on thermal tolerance. Our findings indicate that thermal conditioning via 3D-printed platforms can serve as a robust tool for studying cellular responses to hyperthermia and may contribute to optimizing hyperthermia-based cancer therapies.

Introduction

Three-dimensional printing has rapidly evolved as a transformative technology across biomedical sciences, engineering, and clinical medicine and was even highlighted in 2012 as the next industrial revolution. In biomedicine, the fabrication of complex, customized structures with high precision and versatility has the potential to revolutionize fields such as tissue engineering, drug delivery, regenerative medicine, and cancer research. − Among various 3D printing modalities, DLP stands out for its capacity to produce structures with high spatial resolution, often at the sub-100 μm scale. DLP utilizes a digital micromirror device to project patterned light onto a photosensitive resin, enabling layer-by-layer polymerization with high robustness and speed while also enabling smooth surface finishes. , Unlike extrusion methods, in DLP the printing involves a liquid open bath and not solid materials, enabling easier material manipulation. This ability of changing resin composition opened new possibilities for producing unique functional materials for various purposes, including in advanced cell studies. ,

Hyperthermia, the localized elevation of tissue temperature, has gained interest as an adjunctive therapy for cancer, leveraging the differential thermal sensitivity of malignant versus healthy cells. Printed scaffolds incorporating functional materials such as iron oxide (IO) nanoparticles were demonstrated to generate localized heat under external electromagnetic fields, offering the potential to selectively target tumor cells with a minimal effect on the surrounding healthy tissue. This approach has a noninvasive nature, is compatible with existing cancer therapies and enables the engineering of thermal gradients that can be tailored to specific tissue geometries and tumor microenvironments.

IO nanoparticles are widely used in biomedical 3D printing due to their excellent magnetic properties, biocompatibility, and ability to convert electromagnetic energy into heat. Embedding IO nanoparticles within printed scaffolds can generate spatially controlled thermal gradients, which are essential for mimicking the heterogeneous thermal profiles encountered in vivo and for studying the differential effects of hyperthermia on cell populations. Recent studies have demonstrated that the distribution of IOs, whether uniformly embedded or patterned at the surface, can be fine-tuned to achieve desired heating profiles, further enhancing the precision and safety of hyperthermic interventions. ,

Specifically in oncology, hyperthermia has emerged as a promising modality of treatment, particularly for solid tumors. By elevating the temperature of tumor tissue, typically up to 45 °C, hyperthermia can induce direct cell death, sensitize cancer cells to radiation and chemotherapy, and modulate the tumor microenvironment to enhance immune responses. Despite its clinical potential, the translation of hyperthermia into practice is challenging due to reproducibility issues in localizing heat delivery and the lack of systematic accuracy and standardization in assessing its biological effects. In order to provide a precise platform for studying thermal effects in cells, ex vivo, 3D-printed objects with integrated IO nanoparticles were suggested as a tool that addresses these challenges, enabling precise experimental method for hyperthermic protocols under controlled, physiologically relevant conditions. ,

Developing DLP-based heat responding objects with specific structures can be useful both for in vitro cell models and for in vivo uses, especially for hyperthermia research and cancer therapy.

In this study, we focused on developing an in vitro platform to study thermic effects in cancer cells. We introduce a combined approach that integrates functional materials with rationally designed architectures that together offer a systematic and robust platform for generating controllable heat for cell culture uses. We have successfully fabricated a range of 3D constructs with diverse geometries by combining a commercial 3D printing resin with IO nanoparticles. These hybrid constructs can generate controlled thermal gradients upon heating. This platform enables a systematic investigation of responses of various cell types to localized hyperthermia across a wide range of temperatures, including moderate (∼42–45 °C) and ablative levels (<70–100 °C), with a particular focus on cell viability.

Furthermore, we demonstrate that this approach allows for generation of thermoresistant cell subpopulation using direct evolution on a plate. By growing cells under repeated heat cycles, we demonstrated a selective enrichment of nonsmall cell lung carcinoma cells that exhibit increased resistance to heat, which was associated with more aggressive cellular phenotypes. For example, H460 cells that survived 6 heating cycles demonstrated an increase in postheat survival compared to nonpreheated cells (almost ×6 cell survival) along with enhanced migratory behavior and elevated Heat shock protein (HSP) 27 expression markers associated with tumor aggressiveness and stress adaptation. Furthermore, we showed that the experimental setup is sensitive to detecting effects of cryopreserving compounds, for example, trehalose, which led to a 2-fold increase in cell survival after 4 min infrared (IR) exposure.

Methods

Cell Culture

NIH/3T3 mouse fibroblast and HaCaT human keratinocytes cell lines were purchased from ATCC (VA, USA) and were maintained in Dulbecco’s Modified Eagle Medium (DMEM; Sigma Aldrich) supplemented with 1% penicillin–streptomycin and 10% fetal calf serum (FCS; Gipco, Brazil). (Each cell line had its medium.) Cell passaging was carried out via trypsinization using Trypsin-EDTA (Sigma Aldrich) after the culture medium had been removed and cells were washed with phosphate-buffered saline (PBS; Sigma Aldrich). Trypsin was neutralized after cell detachment using complete medium, and the cells were washed via centrifugation. Cells were passaged to maintain 20–80% confluency and discarded after passage 30. Cells were maintained under sterile conditions at 37 °C with 5% CO₂. All cells were tested for mycoplasma and were found to be mycoplasma-free. NCI-H460 large cell human lung cancer cells were purchased from ATCC (VA, USA) and were maintained in Roswell Park Memorial Institute (RPMI; Sigma Aldrich) supplemented with 1% penicillin–streptomycin and 10% fetal calf serum (FCS). The cells were subjected to identical treatment protocols as the other cell lines.

Three-Dimensional Printing Using DLP

All 3D-printed molds and constructs were fabricated using the digital light-processing stereolithography printer Asiga Max (Sydney, Australia) with an LED light source of 385 nm UV. All devices were designed in Autodesk AutoCad (San Rafael, CA, USA). The final designs were exported as a stereolithography file and uploaded to the printer’s Asiga, a composer software for 3D printing (Sydney, Australia). The structures intended for creating heat gradients were all printed from FREEPRINT ortho (Detax GmbH), unless otherwise mentioned. The printed object was removed from the printing platform at the end of the printing process, and uncured resin was removed by washing the print with ethanol for 5 min in a bath sonicator. The print was then cured under UV light for 5 min (PCU Led, Dreve, Germany).

Printed IO Supplemented Objects: Wedge-Shaped and Wedge Plate

The commercial resin “FREEPRINT Ortho” (Detax GmbH) was used to create a new printable material. To incorporate a heating capability, IO nanoparticles (20–40 nm, US Research Nanomaterials, TX, USA) were added to the resin. Following preliminary tests, samples were prepared with concentrations of 0, 0.1, and 0.5% (w/v). Templates were prepared for use in a PCU LED (Dreve, Germany) device. Samples with concentrations of 0, 0.1, and 0.5% (w/v) were heated by using various methods, including a hot plate, oven, and IR lamp, to assess their thermal properties. A wedge-shaped sample was prepared by using resin concentrations of 0%, 0.1%, and 0.5% (w/v) IO to create a thermal gradient. This sample was heated by using an IR lamp to create a heat gradient effect. A wedge plate design was developed for the gradient plates at AutoCad (San Rafael, CA, USA): A fully printed plate with a built-in gradient, the design underwent a process to smooth its surface and activate it to allow cell adherence.

Thermal Imaging Analysis

All thermal monitoring was conducted using the IR Flash Pro software (Infrared Cameras Inc., TX, USA), which was connected to an ICI infrared thermal camera. For all experiments, the infrared lamp was positioned at approximately 4–5 cm from the printed object, at an inclined angle of ∼ 45°, ensuring homogeneous irradiation of the surface. The thermal camera was placed at 20–30 cm above the heating setup, thereby avoiding direct exposure of the sensor to the IR source while maintaining accurate focus on the sample. Calibration was performed in a way that Temperature values obtained from the camera and software were cross validated against an external digital thermometer and the deviation was recorded. Additional calibration was performed before each experimental session by heating a standard reference object (60 mm resin plate) and confirming stable temperature detection across repeated short pulses of irradiation. Before each trial, a brief preheating step was also used to verify that the lamp and camera alignment produced consistent and reproducible thermal profiles.

IO Discs Fabrication and Heating with an IR Lamp

Upper and down molds (70 mm × 50 mm × 10 mm) were fabricated as described above. The lower mold is structured with disks of varying heights (1, 2.5, 5, and 7.5 mm) and a 16 mm diameter matching the well diameter of a standard 24-well culture plate, aligned and assembled, forming a defined cavity between the molds. The assembled molds were further secured at both ends using clip-like holders to ensure tight sealing and prevent leakage during casting. PDMS (SYLGARD184 silicone elastomer, MI, USA) and cross-linker (SYLGARD184 silicone elastomer curing agent, MI, USA) in a 10:1 ratio mixture was cast into the cavity and placed in an oven set at 65 °C overnight. After curing, the PDMS cast was removed from the template and cured under UV light for 5 min (PCU Led, Dreve, Germany). Afterward, the PDMS discs were filled with IO (20–40 nm, US Research Nanomaterials, TX, USA) at a concentration of 0.5% (w/v) and 99.5% FREEPRINT Ortho (Detax GmbH) and cured under UV light for 5 min (PCU Led, Dreve, Germany). The IO discs were then heated with an IR lamp for 10 min and imaged using a thermal camera (ICI Infrared Camera, TX, USA), the images captured were analyzed using IR Flash Pro, a dedicated software for thermal field analysis to determine the maximum temperature achievable by each disc size upon heating.

Effect of the Height of IO Containing Discs and Duration of Heating on Cell Viability

NIH/3T3 fibroblast cells were seeded onto 24-well plates (50000 cells per well) and incubated for 24 h. After incubation, the plates were placed on the IO discs (for 4 min) that were positioned on a platform to allow equal contact between the different heights of discs and the wells (that were preheated with an IR lamp for different time intervals (1,5, 10, etc.)); all of the process was imaged using a thermal camera (ICI Infrared Camera, TX, USA) to control it. Subsequently, the cells were stained as follows: the medium was removed, and Hoechst (5 μg/ml) was added to cover the surface for 10 min. The solution was then removed, and cells were washed with 2 mL of fresh medium. After the medium was removed, Calcein (1 mM) was added to cover the surface for 10 min at room temperature. Cells were then imaged by using a microscope. Fluorescence microscopy was performed using a Nikon Eclipse Ti microscope (Tokyo, Japan) at 20× magnification, capturing images in the DAPI (excitation/emission: 350/461 nm) and GFP (excitation/emission: 485/530 nm) channels. The same heating procedure was done with another 24-well plate, but was followed by viability detection using MTT reagent (thiazolyl blue tetrazolium bromide purchased from Holland Moran). The MTT viability assay was conducted as follows: cells were seeded, incubated, and heated as described above. MTT stock was prepared at a concentration of 5 mg/mL in saline. MTT solution was added per well (10% v/v) and incubated for 2–4 h; after incubation, the medium was removed, and 200 μL of DMSO was added. The plate was placed gently on a vortex at 160 rpm, and then the absorbance was measured at 570 nm using a plate reader (Wallac 1420 VICTOR plate-reader, Perkin-Elmer Life Sciences, Shelton, CT, USA).

Cell Viability Using Microscopic Imaging with “Wedge in Plate”

NIH/3T3 cells were seeded onto a 60 mm plate (cells were seeded at a density of 4 × 10³ cells/cm²) and incubated for 24 h. After incubation, the plate was placed on top of the “wedge in plate” 3D-printed construct and heated with an IR lamp for 3 min and imaged using a thermal camera (ICI Infrared Camera, TX, USA). The images captured were analyzed using IR Flash Pro software. Subsequently, the cells were stained as follows: the medium was removed, and Hoechst (5 μg/ml) was added to cover the surface for 10 min. The solution was then removed, and cells were washed with 2 mL of fresh medium. After removing the medium, Calcein (1 mM) was added to cover the surface for 10 min at room temperature. Cells were then imaged using a microscope. Fluorescence microscopy was performed using a Nikon Eclipse Ti microscope (Tokyo, Japan) at 20× magnification, capturing images in the DAPI (excitation/emission: 350/461 nm) and GFP (excitation/emission: 485/530 nm) channels. Analysis was conducted using the NIS Elements software.

Cell Viability Using Imaging Using “Multiwedge”

All procedures performed with the “wedge-in-plate” setup were replicated with the Multiwedge construct. We designed the “Multiwedge” platform using computer-aided design (CAD) software, incorporating a radial structure embedded with four stair-like compartments of varying heights (3, 5, 7, and 9 mm). The platform was fabricated using DLP-based stereolithographic 3D printing and filled with a composite mixture consisting of 0.5% w/v IO nanoparticles and 99.5% Freeprint resin (Figure S4).

To assess heat-induced cytotoxicity, NIH/3T3 fibroblasts, NCI-H460 lung carcinoma cells, and HaCaT keratinocytes were seeded onto 60 mm culture plates (each cell line at the density that suits them) and incubated for 24 h. Each plate was then placed onto the Multiwedge platform and exposed to IR irradiation for 4 min. Following thermal treatment, live/dead fluorescence staining was performed using Hoechst (labels the nuclei of all cells) and Calcein AM (labels the cytoplasm of viable cells), enabling spatial visualization of cell viability across the generated thermal gradient and analyzed using the same fluorescence microscopy method.

Effect of Trehalose on Heat Resistance in Cells

Cells supplemented with trehalose were subjected to the same procedures as described for the “Multiwedge” setup, with the only modification being the addition of 100 mM trehalose to the culture medium (the trehalose-supplemented medium must be freshly prepared before each experiment). 3T3, HaCaT, and NCI-H460 cells were seeded onto 60 mm plates, either treated with trehalose-containing medium or as a control not treated with trehalose-containing medium, and then incubated for 24 h, placed on the “Multiwedge” platform for infrared heating under identical conditions, and analyzed using the same fluorescence microscopy method.

Repeated Heating Cycles

NIH/3T3 and NCI-H460 fibroblast cells were seeded onto 24-well plates (50000 cells per well and 100000, respectively) and incubated for 24 h. After incubation, the plates were placed on the IO discs (which were positioned on a platform to allow equal contact between the different heights of discs and the wells), heated with an IR lamp, each plate for 4 min (after the IO discs were preheated with an IR lamp for 10 min), and then incubated for 24 h. After incubation, cells were trypsinized using Trypsin-EDTA (Sigma Aldrich) after the culture medium had been removed and cells were washed with phosphate-buffered saline (PBS; Sigma Aldrich). Trypsin was neutralized after cell detachment using complete medium, and cells were washed via centrifugation. The supernatant was removed; then, we added 1 mL of medium, took 100 μL from the cell suspension, and mixed it with 100 μL of Trypan Blue at a 1:1 ratio. Afterward, Cell counting was performed using a DeNovix CellDrop automated cell counter (DeNovix Inc., Wilmington, DE, USA). Next, we reseeded the cells in new culture plates according to the measured cell number. Cells were then allowed to grow for 3 days to recover and proliferate. After this incubation period, they were reseeded in 24-well plates at the same density described above and subjected to another round of heating under identical conditions. This process of heating, collecting surviving cells, counting, and reculturing was repeated for 6–7 consecutive cycles.

Heat Shock Protein 27 (HSP27) Human ELISA

Quantification of HSP27 protein expression in NCI-H460 cells (heated vs nonheated) was performed using the Human Heat Shock Protein 27 ELISA Kit (ab108862, Abcam, Cambridge, UK), according to the manufacturer’s instructions. Briefly, cell lysates were prepared using lysis buffer (PBS supplemented with 1% Triton X-100 and protease inhibitor cocktail) and incubated on ice for 1 h. Lysates were then centrifuged at 13000 rpm for 30 min at 4 °C, and the supernatants were collected and stored at −80 °C until analysis. A total of 50 μL of the samples and standards were added in duplicates to wells precoated with anti-HSP27 antibody, followed by a 2 h incubation with biotinylated HSP27 detection antibody and 30 min with streptavidin-peroxidase (SP) conjugate. TMB substrate solution was used to develop the color. After 15 min, stop solution was added, and absorbance was measured at 450 nm using a microplate reader (Wallac 1420 VICTOR, Perkin-Elmer Life Sciences, Shelton, CT, USA). Between each step, the wells were washed three times with washing buffer. HSP27 concentrations were calculated based on a standard calibration curve (0–80 ng/mL) generated in parallel.

Proliferation Assay

H460 cells (nontreated and preheated) were seeded in 4-microwell agarose molds using the method described in the work by Steinberg et al. Cells were seeded at a density of 5000 cells/microwell in molds placed in a 96-well plate. Twenty-four h after seeding, spheroids were extracted from molds (via ups and downs using a pipet) and transferred to a 96-well plate (single spheroid per well). Thirty μL of Matrigel was added, and the plate was placed in a 37 °C and 5% CO₂ incubator for 10 min. After 10 min, 100 μL of medium was added per well and placed back in the incubator. Spheroids were imaged after 1 and 5 days via a Nikon Eclipse Ti microscope (Tokyo, Japan). Growth analysis was conducted by using the NIS Elements software.

Scratch Wound Migration Assay

The assay was done using the same method described in the work by Steinberg et al. Cells (regular and heated) were seeded at a density of 60000 cells/well in an IncuCyte ImageLock 96-well plate (Sartorius, USA) and incubated at 37 °C and 5% CO₂ for 24 h. A scratch wound was made in each well using the IncuCyte 96-well WoundMaker Tool (ESSEN, BioScience). The plate was placed in an incubator containing an IncuCyte S3 Live-Cell Analysis System (ESSEN, BioScience) for 48 h. Real-time automated images were taken every 2 h, and cell migration analysis was performed using the IncuCyte Scratch Wound Cell Migration Software Module. Final images were processed by using ImageJ.

Results

Preparation and Characterization of Functional Resin

To create a heat gradient in 3D-printed structures, we used IO nanoparticles Fe₂O₃, 20–40 nm. To examine which concentrations of IO mixed with the resin provide optimal heat distribution and printing compatibility, we have tested 1–200 mg/mL (Fe₂O₃/mL resin). We found a wide range of concentrations that could be successfully mixed in the resin; however, 10% (w/v) was the upper limit with precipitation. A concentration of 5% yielded a homogeneous mix with the resin, but did not cure well under UV light exposure. The 0.5% and 0.1% concentrations created homogeneous mixtures and cured well; consequently, these two concentrations were selected for further investigation (Figure A).

1.

Thermal evaluation of IO-containing transparent platforms and development of an integrated wedge structure. (A) Photocured samples (UV-cured) of resin (“Freeprint”) containing 0, 0.1, and 0.5% (w/v) IO (from left to right). (B) Heating samples of (0, 0.1, 0.5% w/v from left to right) IO inks by IR radiation for 3 min, the thermal image demonstrates localized heat generation, with temperature values color-coded from blue (30 °C) to white (>50 °C), indicating the range of temperatures reached during IR exposure. (C) Wedge-shaped constructs fabricated using 3D printing and filled with inks containing different concentrations of iron oxide nanoparticles (0,0.1, 0.5% w/v). (D) Thermal image of the wedge-shaped constructs following 3 min IR exposure, color-coded temperature values ranging from blue (∼30 °C) to white (>70 °C), demonstrating heat distribution. (E) CAD design structure of a 60 mm plate that incorporates a built-in wedge of IO. (F) Printed structure of a 60 mm plate that incorporates a built-in wedge of IO.

The composite materials (resin/IO nanoparticles) were tested in 20 × 20 × 5 mm object prepared in a PDMS mold. To study heat induction, samples were subjected to various heating methods. While heating in an oven or on a hot plate (Figure S1A) yielded no significant differences in heating properties between the different concentrations, irradiation with an IR lamp allowed for more delicate distinctions in heating behavior (Figure B). Therefore, in our further experiments, we have used IR radiation as the heat induction method.

The concept of structure control to induce diverse and controlled heat was studied in a special design that enables gradient in a single printed object by using a wedge-shaped object. A temperature gradient using 3 min IR radiation confirmed a range of 30–70 °C on the wedge-shaped sample. This gradient was successfully created across all tested concentrations. As the concentration of IO increased, the resulting heat gradient became more pronounced, reaching higher temperatures > 70 °C under infrared exposure. (Figure C,D).

To enable the generation of a defined thermal gradient, we fabricated a custom 60 mm plate. This design integrated the previously developed wedge geometry, which allows for spatial variation in the heat distribution upon infrared exposure. The initial design (Figure E,F) successfully created a temperature gradient but was too thick and lacked transparency, making it unsuitable for microscopic imaging in case of direct cell seeding. Several calibrations were performed to determine the maximum thickness of the resin that would allow visibility. However, it was found that only a very thin layer (∼0.5 mm) of polymer permitted adequate visibility. To address these limitations, a new design was created consisting of two components: one printed to form a heating gradient (Figure A­(i–iii)). The design involved the 3D printing of a 60 mm diameter plate that includes an internal wedge-shaped compartment intended for localized heat generation. This compartment was filled with a mixture composed of 99.5% Freeprint ink and 0.5% IO. The other component was a commercial cell culture plate that enables cell visibility under a microscope. This revised design enables the use of a standard plate heated atop a heating gradient plate (Figure A,iv). This configuration successfully creates a temperature gradient, while allowing for direct microscopic imaging.

2.

Design and integration of a 3D-printed “Wedge-in-plate” setup, enabling heat distribution, with corresponding thermal mapping and cellular response. 3D printed construct filled using a mixture consisting of 99.5% “FREEPRINT Ortho” resin and 0.5% (w/v) IO. Following 3 min IR lamp exposure, cells were stained with Hoechst and Calcein to evaluate viability. (A) (i) CAD model of “wedge-in-plate” design (60 mm diameter). (ii) image of the 3D-printed “wedge-in-plate”. (iii) Wedge filled with IO (0.5% w/v) and “FREEPRINT Ortho” (99.5%). (iv) Final system consisting of a 60 mm plate placed on top of the “wedge-in-plate”. (B) Left: thermal image captured after 3 min of IR lamp exposure. The color scale represents temperature values ranging from blue (∼40 °C) to white (>70 °C). Right: fluorescence images of Live/dead staining achieved by Hoechst (stains the nucleus of all cells-blue) and Calcein (stains the cytoplasm of live cells-green), corresponding to the temperature zones marked in the thermal image. Scale bar = 100 μm.

The effect of heating on cellular viability is demonstrated in Figure B. Following three min of infrared exposure, a distinct thermal gradient was observed along the wedge-filled region of the 3D-printed plate. Thermal imaging revealed that temperatures ranged from ∼40 °C at the cooler end to ∼70 °C at the hotter end, confirming the structure’s capacity to generate controlled spatial heat distribution. Corresponding fluorescence microscopy illustrates the contrast between the two regions of the plate. In the staining images, Hoechst stains the nuclei of all cells (blue), while Calcein-AM selectively stains the cytoplasm of live, active cells (green). Higher Calcein intensity is observed in the cooler zones, indicating greater cell viability, whereas greater blue fluorescence in the hotter regions reflects thermal-induced cell damage or death.

In order to provide standardized analytical tools that are compatible with standard cell culture plates that generate higher throughput and distinct thermal environments within individual wells of a standard multiwell plate, we designed and fabricated IO nanoparticles containing discs with varying heights (1.0, 2.5, 5.0, and 7.5 mm) while maintaining a constant diameter across all samples.

To fabricate the discs, we used 4 upper and lower molds designed in Autodesk AutoCAD (Figure S2A), each measuring 70 mm in length, 50 mm in width, and 10 mm in height. The only difference between mold types was the height of the cylindrical structures incorporated into the lower mold, which defined the final height of each IO disc. Afterward, the molds were aligned and assembled, forming a defined cavity between the molds. PDMS, cross-linker in a 10:1 ratio mixture, was cast into the cavity and placed in an oven set at 65 °C overnight. After curing, the PDMS cast was removed from the template and cured under UV for 5 min (Figure S2B). Subsequently, the PDMS discs were filled with IO at a concentration of 0.5% (w/v) and 99.5% “FREEPRINT Ortho” and cured under UV for 5 min (Figure S2C).

The successfully fabricated IO discs produced at defined heights (1, 2.5, 5, and 7.5 mm) are displayed in Figure A. To account for the varying heights of the IO discs and ensure uniform positioning during infrared exposure, a custom 3D-printed platform was designed in Autodesk AutoCAD and fabricated to equalize the height of the discs (Figure B). The full illustration and experimental image of the setup are presented in Figure C and Figure S2D, showing a 24-well plate positioned above preheated IO discs of varying heights, placed on the previously described platform. The setup was subjected to IR lamp exposure, with the IO discs heated before plate placement and thermal monitoring performed using a thermal camera.

3.

Design, setup, and thermal performance of IO discs under IR irradiation. (A) IO discs with varying heights (7.5, 5, 2.5, and 1 mm), diameter of 16 mm, arranged from left to right. (B) Side view of the experimental setup showing a standard 24 well plate on top of IO discs that are placed on multistep platform that enable alignment. In bottom, 3D CAD design and printed object of a custom multistep platform with three discrete height levels. (C) full experimental setup image including 24-well plate positioned above preheated IO discs placed on the platform, subjected to 10 min IR lamp exposure (the IO discs were subjected to exposure before plate placement), and monitored by thermal camera. (D) graphical representation of the maximum temperatures achieved by IO discs of varying sizes following heat with an infrared lamp over time, 1–10 min, mean ± SD, n = 6.

To characterize the thermal behavior of the IO discs, we performed a 10 min IR heating protocol while continuously monitoring surface temperatures using a thermal camera. As shown in Figure D, each disc exhibited a distinct, height-dependent heating profile. The tallest disc (7.5 mm) reached a maximum temperature of 100 ± 5 °C after 10 min exposure, while the shortest disc (1 mm) reached 62 ± 3 °C. Intermediate discs of 2.5 and 5.0 mm achieved 70 ± 3 °C and 90 ± 4 °C, respectively. The graph presents the maximum temperature reached by each disc at one min intervals over the 10 min exposure period. Values represent mean ± standard deviation from six independent replicates (n = 6). These results demonstrate a consistent and geometry-driven thermal gradient with temperature increasing progressively over time and correlating positively with disc height. Figure S3 displays a chronological sequence of thermal images, acquired via IR Flash software, illustrating the progressive and reproducible accumulation of heat across the IO discs over 10 min.

To examine how the geometry of IO discs and the duration of infrared heating affect cell viability, we exposed cells to graded thermal stress generated by discs of increasing height for 1, 5, or 10 min. NIH/3T3 cells were seeded onto three 24-well plates (50,000 cells per well) and incubated for 24 h. Following incubation, the plates were placed onto preheated IO discs positioned on a custom 3D-printed platform. The IO discs had been preheated using an IR lamp for three different durations: 1, 5, or 10 min. Each plate was placed onto discs corresponding to one heating duration. The entire heating process was monitored in real time by using a thermal camera to control the process.

Cell viability was assessed using the MTT assay and expressed as relative viability (%) normalized to unheated control cells. As shown in Figure A, viability progressively declined with increasing disc height and heating duration. After 10 min of IR exposure, cells in contact with 7.5 mm discs exhibited a viability of approximately 28%, while those exposed to 1 mm discs retained 63% viability. Notably, even a brief 1 min exposure to 7.5 mm discs resulted in a reduction in viability to 73%. A similar viability decline pattern was observed for the intermediate-height discs (2.5 and 5 mm). Specifically, exposure to 5 mm discs reduced viability to ∼79% after 1 min and further down to ∼37% after 10 min. A statistically significant difference in relative cell viability was observed between all heated IO disc conditions at each exposure duration (1, 5, and 10 min), with the most pronounced difference occurring between the 7.5 and 1 mm disc heights.

4.

Effect of IO disc height and heating duration on cell viability. MTT assay was conducted following IR lamp exposure; in addition, cells were stained with Hoechst and Calcein to evaluate viability. (A) Quantitative analysis of relative cell viability assessed by MTT assay after exposure to IO discs of varying heights (1 mm, 2.5, 5, and 7.5) following IR lamp heating for 1, 5, or 10 min. Statistical significance between groups was calculated using unpaired Student’s t tests (mean ± SD, n = 4; ***p < 0.001, **p < 0.01). (B) Representative images of Live/Dead fluorescence staining of cells following treatment with IO discs exposed to IR heating for 1, 5, and 10 min. Staining achieved with Hoechst (blue-stains the nucleus of all cells) and Calcein (green-stains the cytoplasm of live cells). Images are arranged to correspond with increasing time from left to right and decreasing disc height from top to bottom, matching the conditions shown in panel A. Scale bar = 100 μm in all images.

To validate these results visually, live and dead staining using Hoechst and Calcein-AM was performed. As shown in Figure , Calcein-positive (live) cells decreased with higher heat conditions, while Hoechst-only nuclei became more prominent.

To explore how cell-cell variations respond to localized thermal stress, we developed a printed “Multiwedge” geometry designed to generate a controlled heat gradient. A comparison of cell viability across distinct cell types: fibroblasts, epithelial cells, and cancer cells was done based on their known differences in heat sensitivity. As shown in Figure A,B, thermal imaging revealed a well-defined vertical temperature gradient across the Multiwedge following 4 min of IR lamp exposure. Specifically, the maximum temperatures reached at each stair region were ∼46.6 °C (3 mm), ∼53.8 °C (5 mm), ∼73.9 °C (7 mm), and ∼82.4 °C (9 mm). Following thermal treatment, cells were stained, as described previously. Figure and Figure S5 show increased thermal intensity across the Multiwedge correlated with reduced viability. Comparative microscopical analysis of cell viability at the 5 mm stair region, where the maximal temperature reached approximately 54 °C, revealed differences in thermal sensitivity among the tested cell types, as illustrated in the representative images (Figure ). After 4 min of IR exposure, HaCaT keratinocytes exhibited the most pronounced reduction in viability, declining to 60% ± 3.2%. In contrast, NIH/3T3 fibroblasts maintained the highest viability under the same conditions, with viability levels of 83% ± 2.5%. NCI-H460 lung cancer cells exhibited an intermediate profile, with a viability of 74% ± 3.0% and moderate Calcein signal loss (Figure B). Comparison between the three cell types revealed statistically significant differences in viability: HaCaT vs 3T3 (p < 0.001), HaCaT vs H460 (p < 0.01), H460 vs 3T3 (p < 0.05), n = 3.

5.

Characterization of temperature gradient and assessment of cellular viability following spatially localized thermal exposure using a multiwedge-printed structure. (A)­(i) CAD design of “Multiwedge” (60 mm diameter). (ii) Image demonstrating the heating gradient following 4 min of heating. (B) Table summarizing the minimum, average, and maximum temperatures each stair zone reaches following 4 min exposure to IR. (C) Finite element thermal image demonstrating the vertical temperature gradient generated by IR lamp heating of the custom-fabricated multiwedge structure, temperature distribution is color-coded from blue (∼40 °C) to white (>80 °C), the structure is designed to induce localized heat exposure zones. Left - representative fluorescence microscopy images of 3T3 cells, live/dead Staining achieved by Hoechst (blue, stains the nucleus of all cells) and Calcein (green, stains the cytoplasm of live cells) corresponding to the temperature zones marked in the thermal image. Right- images of HaCaT, 3T3, and NCI-H460 cells stained with Hoechst (blue, nuclei) and calcein­(green, cytoplasm), captured at the same thermal zone aligned along the Multiwedge after being heated for the same period. Scale bar = 100 μm in all images. (D) Quantitative comparison of relative cell viability (%) among the three cell types following 4 min of heating, reaching temperatures up to ∼54 °C at the thermal zone aligned in the Multiwedge. Viability was highest in 3T3/NIH fibroblasts and lowest in HaCaT keratinocyte cells, mean ± SD (t test comparing cell lines viability%, HaCaT vs 3T3 ***p < 0.001, n = 3).

The fluorescence microscopy images (Figure S5) revealed a distinct thermotolerance of each tested cell type, positioning NIH/3T3 fibroblasts as the most heat-resistant and HaCaT keratinocytes as the most heat-sensitive under these localized hyperthermic conditions.

To assess whether our platform is sufficiently sensitive to detect subtle effects of thermal sensitivity mediators, we investigated the protective effect of trehalose under localized thermal stress (Figure A). We seeded 3T3, HaCaT, and NCI-H460 onto 60 mm plates and incubated them for 24 h. Plates were then placed directly over the Multiwedge structure and exposed to IR irradiation, inducing temperature gradient between ∼40 °C and ∼80 °C. To detect a safe concentration of Trehalose, several concentrations were tested (25 mM to 100 mM). An optimal concentration selected 100 mM, was identified as a dose that has cytoprotective effects without compromising cell viability due to hyperosmolarity. To further detect these effects visually, cells were cultured for 24 h with 100 mM trehalose prior to IR exposure. After heating, live/dead staining was performed, demonstrating a spatial correlation between temperature intensity and cell viability remained consistent across all tested cell types (Figure S6). As shown in Figure B,C, representative fluorescence microscopy images of NCI-H460 cells revealed that trehalose-treated cultures exhibited a markedly enhanced Calcein (green) signal across all thermal zones, including areas subjected to moderate and high levels of heat. This enhancement of cell viability was most prominent in the hottest region (corresponding to the highest step of the Multiwedge) but was also obtained in the cooler zones.

6.

Trehalose enhances cellular viability under spatially controlled thermal stress applied via a Multiwedge structure. (A) Chemical structure of trehalose, a disaccharide composed of two glucose molecules. (B) Left - thermal image showing the spatial temperature gradient generated following 4 min of IR lamp heating of a Multiwedge structure, temperature distribution is color-coded from blue (∼40 °C) to white (>80 °C); Right - representative fluorescence microscopy images of NCI-H460 cells cultured with trehalose, live/dead staining achieved by Hoechst (stains the nucleus of all cells) and Calcein (stains the cytoplasm of live cells) corresponding to the temperature zones marked in the thermal image. Scale bar = 100 μm in all images. (C) Representative fluorescent microscopy images of heat-exposed NCI-H460 cells, stained with Hoechst (blue, nuclei) and calcein (green, cytoplasm) representative images show the difference in live cells between trehalose-free and trehalose-supplemented media culturing (left to right) following thermal exposure. Scale bar = 100 μm in both images. (D) Quantification of relative cell viability (mean ± SD, n = 4) in HaCaT, 3T3, and NCI-H460 cells following 4 min of infrared thermal exposure, under two conditions: without trehalose (control) and with trehalose supplementation. Statistical significance: *p < 0.05, **p < 0.01 (t test, comparing trehalose-supplemented vs trehalose-free for each cell line).

Viability analysis of the maximum temperature zone, as marked in the thermal image, was performed using NIS Elements software and further confirmed the protective effect of trehalose on cell survival (Figure D). In HaCaT cells, viability significantly increased from 12 ± 2.9% in control conditions to 27 ± 3.0% with trehalose (p < 0.01), whereas, in NIH/3T3 fibroblasts, from 30 ± 2.3% to 37 ± 2.0% (p < 0.05), and in NCI-H460 cells, from 15 ± 2.5% to 30 ± 2.2% (p < 0.01).

To investigate whether repeated heat cycles can generate selectivity toward heat resistant cells, we subjected NIH/3T3 fibroblasts to multiple rounds of IR-induced heating using either 5 or 7.5 mm IO discs. Each disc was preheated for 10 min using an IR lamp, followed by placement of a 24-well plate containing cells directly above the heated discs for an additional 4 min, as described previously. After each heating cycle, viable cells were collected, counted using the DeNovix CellDrop automated cell counter, reseeded, and allowed to recover before the next round of thermal exposure. As shown in Figure A, 3T3 cells exhibited a consistent increase in viable cell numbers across successive heating cycles. Notably, a sharp rise in cell count was observed between the fifth and sixth cycles. Specifically, in NIH/3T3 fibroblasts subjected to repeated heating using 7.5 mm IO discs, a 5-fold rise in survival under the same thermal conditions was obtained. Based on these results, we applied the same heating protocol to NCI-H460 lung carcinoma cells. As shown in Figure B, NCI-H460 cells displayed similar enrichment trend with a marked change in cell number between cycle 5 and cycle 6. A statistically significant increase in cell count was observed between cycle 1 and cycle 6 under both thermal conditions (5 mm and 7.5 mm discs, p < 0.01). Specifically, NCI-H460 lung carcinoma cells heated with the 7.5 mm IO discs, showed an increase of cell count from 40000 to over 200000 after the sixth cycle at the same heating conditions, representing a 5-fold increase.

7.

Effect of repeated heating cycles on cell proliferation, HSP27 expression and aggressiveness following localized thermal stress. (A) Quantification of cell proliferation following multiple heating cycles using either 7.5 mm or 5 mm IO discs heated by an IR lamp. 3T3 Cells were exposed to 1–7 cycles of localized heating followed by recovery and growth. (B) Quantification of cell proliferation following multiple heating cycles using either 7.5 mm or 5 mm IO discs heated by an IR lamp. NCI-H460 cells were exposed to 1–6 cycles of localized heating followed by recovery and growth, mean ± SD (**p < 0.01, n = 4). (C) ELISA quantification of HSP27 protein levels in cells subjected to 6 thermal cycles (U1) versus nonheated control cells (U2). A significant upregulation of HSP27 expression was detected in heated cells, mean ± SD (***p < 0.001, n = 6). (D) Percent growth of the spheroid in heated vs nonheated NCI-H460 cells after 5 days relative to day 1 area ± SD (*p < 0.05, n = 5). (E)­(i) Representative image of spheroids formed from nonheated (top) and heated (bottom) NCI-H460 cells 1 day after seeding. (ii) Representative image of spheroids formed from nonheated (top) and heated (bottom) NCI-H460 cells 5 days after seeding. Scale bar = 100 μm in all images. (F)­(i) Representative microscope image of heated NCI-H460 cells at 48 h postscratch. (ii) Representative microscope image of Unheated NCI-H460 cells at 48 h postscratch. Scale bar = 400 μm. The images are shown in contrast colors to enhance visibility. (G) Comparison of wound confluence (%) between heated and unheated cells, mean ± SD (***p < 0.001, n = 6).

To provide a mechanical explanation for the heat residence, we analyzed the level of a central cell heat shock protein, HSP27. ELISA-based quantification revealed a significant elevation (p < 0.001) in protein levels in cells grown after six heating cycles (U1, 8.2 ± 0.3 ng/mL) compared to nonheated controls (U2, 7.2 ± 0.2 ng/mL) (Figure C). Importantly, we found that the upregulation persisted over time (Figure S8), where three weeks after the final heating cycle, HSP27 levels remained elevated in the thermally adapted population.

To further explore whether heat resistance is associated with changes in cell functionality, we have measured various key activities that are also critical in cancer biology: cell proliferation in 3D and cell mobility. Cell proliferative capacity and tumorigenic potential. A 3D spheroid proliferation assay was conducted using cells derived from the heated and control populations. On day 1, spheroid sizes appeared comparable across groups; however, by day 5, spheroids generated from repeatedly heated cells exhibited greater relative area expansion compared to their nonheated counterparts (Figure S7B). Specifically, spheroids derived from heated cells increased in size by 174.3%, while those from unheated cells increased by 112.9% between days 1 and 5 (Figure D). This difference in percentage growth was statistically significant (p < 0.05, n = 5). Representative images (Figure E) visually illustrate this observation.

An additional important cell function is motility, which is important in cancer progression and in fibrosis in tumor sites. In a wound closure assay, we found that the heat-resistant cells (post six heating cycles) exhibited significantly faster wound closure (p < 0.001, n = 6) compared to nonheated controls. Heated cells closed approximately 80% of the wound after 24 h post scratch, and after 48 h approximately 95% of the wound was confluent with cells, in contrast, nonheated cells closed only ∼43% of the wound after 24 h post scratch, and after 48 h, approximately 60% of the wound was closed with cells. Representative images visually illustrate this observation (Figure S9 and Figure F).

Discussion

This study focuses on studying cellular heat responses through the integration of functional material with rationally designed architectures. Structure function understanding of hyperthermia was enabled by spatially controlled hyperthermia using DLP printed objects embedded with IO. By applying IR irradiation to these geometrically tailored objects, we created reproducible thermal gradients ranging from ∼40 °C to > 70 °C, capturing both sublethal and ablative temperature zones within the same system. In order to provide a systematic well calibrated platform that can be compatible with standard cell culture plates, we evaluate cellular responses across multiple cell types and various formats.

Iron oxide nanoparticles, Fe₂O₃, were chosen since they exhibit excellent thermal conductivity properties, effective absorption and distribution of heat, when exposed to an external energy source such as infrared light. Previous studies that used IO combined with polymers for 3D printing typically used extrusion or fused deposition modeling (FDM). , Unlike in filaments used in FDM, with UV polymerization in DLP, the IO nanoparticles can be distributed with higher spatial precision, enabling micron-scale patterning and more accurate engineering of heat diffusion profiles, as highlighted in a recent review.

One of the key advantages of our platform is the generation of a continuous thermal gradient within a single construct. This addresses a key limitation in hyperthermia research, which commonly relies on water bath heating or metallic wires, that lack a specific spatial focus of heat and require multiple samples to study temperature-dependent responses. More recently, integration of heating via microfluidic platforms or magnetic nanoparticle-loaded hydrogels has shown improved local temperature control, but often at the expense of complex fabrication steps or dependence on external field manipulation. Our approach balances between precision, simplicity, and standardization and provides a high-throughput, self-contained platform that can be easily tailored for various biological systems and experimental setups.

Biologically, our results confirm the well-established differences in thermotolerance among cell types. Cell viability is highly sensitive to temperature, with well-defined thresholds for thermal damage. Hyperthermia during late S–M phases disrupts the spindle, stalls replication forks, and degrades BRCA2/RAD51, leading to chromosomal mis-segregation and unrepaired DNA breaks. While some variability in thermotolerance has been reported, it remains insufficiently characterized under carefully controlled and physiologically relevant heating conditions. The Multiwedge geometry was created to address this challenge and enable side-by-side comparison under identical thermal exposure and experimental conditions. Moreover, the Multiwedge configuration not only allows controlled comparison of thermotolerance but also exhibits how distinct thermal zones can be engineered in geometries more advanced than a single wedge, thus enhancing the flexibility of the platform.

In our study, we have found that keratinocyte HaCaT cells were the most sensitive to heat, H460 showed an intermediate sensitivity, and 3T3 cells were the most resistant cell type tested. Experimental studies show that in HaCaT keratinocytes, physiologically relevant heat shock provokes an early necrotic burst followed by apoptosis, and that cell fate is critically dependent on a modest but inducible rise in protective HSP70 and related heat-shock proteins. In contrast, 3T3 fibroblasts were found to be the most thermotolerant cell type, an observation that is in line with a study that showed that NIH-3T3 cells exposed to heat shock increased ceramide levels and induced the expression of αB-Crystallin, which is a small heat shock protein with cytoprotective properties. Fibroblasts in general are known for their resilience to environmental and metabolic stresses, with slower proliferation rates and potentially more efficient stress response pathways, suggesting a more pronounced molecular defense against heat. While direct evidence for H460 cells is lacking in the literature, cancer cell lines, in general, may exhibit increased HSP expression and altered stress response pathways, potentially manifest a wider range of sensitivity level.

To further validate our system and confirm its sensitivity to detect changes in heat resilience of cells, we have used biochemical approach in addition to baseline biological studies. Trehalose markedly enhanced survival across all cell types under heating. This nonreducing disaccharide composed of two glucose units is known for its function as a chemical chaperone and an osmoprotectant, stabilizing proteins and cellular membranes under environmental stresses, including heat, desiccation, and oxidant challenge. , Evidence for trehalose-mediated protection against acute heat-induced cell death was found in porcine intestinal epithelial cells, where trehalose supplementation significantly increased viability and reduced apoptosis following a 43 °C heat shock. Our results align with these findings and suggest that trehalose also prevents heat-induced cell death in the context of acute IR exposure.

One of the most intriguing questions is whether heat resilience is a feature that can be enriched in a given cell population and a parameter that can lead to direct evolution due to thermal adaptability or due to cell population heterogeneity in terms of heat-resistance. To study this question, we have induced a repeated heating protocol, simulating chronic sublethal stress. Over six heating cycles, we observed increased survival in NCI-H460 cells, indicating the selection for a thermotolerant subpopulation. Between the fifth and sixth cycles, we observed the most remarkable increase in viable cell population when subjected to the same heating conditions as in the first cycles. This may suggest that certain adaptive mechanisms promote survival and proliferation under repeated thermal stress.

Several studies support our findings. For example, Anderson et al. demonstrated that repeated cycles of heating in vitro selected for thermotolerant subpopulations in murine fibrosarcoma (RIF-1) cells, which showed a marked increase in heat resistance compared to the parental line. Similarly, Tao et al reported the progressive selection of heat-resistant clones in B16 melanoma cells following serial heat exposures. The survival of cell populations through repeated heat cycles may have significant clinical consequences, potentially enabling the regrowth of aggressive, heat-resistant cells after hyperthermia therapy.

In addition to the phenotype, we observed a consistent and sustained rise in HSP27 expression levels via ELISA quantification, which can partially explain the biological mechanism of the resistant cells. HSP27, a small heat shock protein, known to play a central role in adaptation during stress conditions, including oxidative damage and thermal injury, is implicated in cytoprotection, antiapoptotic signaling, and metastasis in several cancers.

HSP27 directly protects against heat-induced damage, and its suppression sensitizes cells to thermal stress. , Mechanistically, HSP27 contributes to cell protection following heat stress by stabilizing the cytoskeleton, preventing protein aggregation, and inhibiting apoptotic pathways. Our findings that HSP27 expression increases during repeated thermal exposure provide functional validation of its role in adaptation. This elevated level remained even 3 weeks after cells were in culture post the last cycle, This long-lasting shift in HSP27 levels suggest a form of “thermal memory”, meaning that preheat leaves a stable imprint on the cellular stress response that sustain beyond the acute stimulus.

Beyond the direct impact of heat resistance in hyperthermia as a therapeutic approach in cancer, the indirect effects on other biological functions that can affect cancer progression are critical to measure. We have found that heat-conditioned NCI-H460 cells that had an elevated survival also displayed enhanced aggressiveness in other key functional assays. Specifically, in the scratch assay for cell mobility, these resistant cells had higher migratory activity and capacity to close wounds significantly faster than their nonheated counterparts. This observation is consistent with other studies, which demonstrated that exposure of cancer cells to heat stress leads to increased aggressiveness, as reflected by enhanced migration, invasion, and epithelial-to-mesenchymal transition (EMT) markers. , This effect was found to be strongly linked to the induction of HSPs. For example, following sublethal heat shock, elevated HSP70 levels have been shown to drive aggressive phenotypes via stabilization of key metastasis-promoting factors and modulation of pathways such as HIF-1α Sumoylation.

While most mechanistic studies use a single acute heat shock, there is also evidence from repeated or stepwise heat exposure protocols that cancer cells acquire persistent increases in migration and invasion potential, together with upregulation of HSPs, including HSP70 and HSP27. These findings support the view that repeated heat stress not only selects for more aggressive cancer cell populations but also activates a sustained HSP-dependent response, which underlies the enhanced malignant behaviors observed after multiple rounds of thermal conditioning. Furthermore, in 3D culture conditions that promote spheroid formation, heat-conditioned cells exhibited significantly increased proliferation, as evidenced by real-time imaging and area-based analysis, indicative of increased aggressiveness and tumorigenic potential.

Despite its advantages, the platform also has several limitations. IR-based heating, while simple and noncontact, is sensitive to lamp geometry and exposure time; therefore, integrating a fixed-stage source or thermal feedback system could enhance reproducibility. The printed constructs’ opacity limits high-resolution imaging, partially mitigated by culturing cells on standard plates above the heaters. Thermal mapping was restricted to representative images, wide-field imaging would better validate gradients and enable dynamic analysis. Regarding biological readouts, cell viability was assessed using Hoechst/Calcein staining, which may overlook early apoptotic events. Future use of PI or EthD-1 could improve the mechanistic resolution of cell death pathways. Furthermore, Trehalose was used as a cryoprotectant, but assessing additional compounds (e.g., dextran, PVA) could broaden our understanding of heat-induced cytoprotection and expand drug-discovery potential. Finally, while validated in both 2D and 3D models, the system remains limited to in vitro use; in vivo adaptation will require material and heating refinements.

Conclusion

In summary, this study establishes DLP-printed IO constructs as a versatile platform for structure-controlled spatially defined thermal modulation, providing a powerful tool for biological research. Unlike conventional bulk or uniform heating approaches, this system enables precise, reproducible, and tunable thermal control within a single construct, minimizing intersample variability and allowing detailed mechanistic investigations under standardized conditions. The multiassay format further permits parallel evaluation of multiple cell types or experimental conditions exposed to identical heating profiles. This in vitro tool holds potential for drug discovery, enabling preclinical screening of thermoprotective compounds, and for personalized medicine, supporting the optimization of hyperthermia regimens using patient-derived models. Using this system, we revealed that repeated heat exposure induces heat resistance via HSP upregulation, promoting adaptive phenotypes that may mimic tumor evolution under thermal stress. These findings underscore the clinical challenge of tumor recurrence following hyperthermia therapy and highlight the platform’s potential to advance both mechanistic understanding and translational applications in thermal oncology.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsbiomaterials.5c01435.

• Supporting Figures S1–S9 (PDF)

The authors declare no competing financial interest.