Light‐Responsive Surface Topographies Modulate Macrophage Immune Responses Through Dynamic Mechanical Cues
Oksana K. Savchak, Ruth M. C. Verbroekken, Burcu Gumuscu, Albert P. H. J. Schenning
TL;DR
This paper shows how light-controlled surfaces can change macrophage behavior, offering new ways to control immune responses for regenerative medicine.
Contribution
The study introduces light-responsive liquid crystal polymer films that dynamically modulate macrophage immune responses through reversible topographies.
Findings
Grooves on light-responsive surfaces trigger mixed inflammatory and anti-inflammatory macrophage responses.
Pillar-shaped topographies maintain an anti-inflammatory macrophage profile without broad activation.
Dynamic topographies induce distinct membrane morphologies, including migration-associated blebbing and lamellipodia formation.
Abstract
Understanding macrophage phenotype regulation by mechanical stimuli is a promising way to elucidate the body's inflammatory response and design new therapies. However, creating dynamic interfaces that allow precise, real‐time, and reversible control over mechanical cues remains a challenge. In this study, we report the immunomodulatory effects of dynamic liquid crystal (LC) polymer films on in vitro macrophage responses. By utilizing reversible light‐induced LC surface topographies, we generate dynamic mechanical stimuli on cells during topography formation and removal, enabling on‐demand and reversible reprogramming of cell behavior. Our findings reveal a strong topographical shape‐dependent cell response by examining the effects of flat, pillared, and grooved LC films on THP‐1‐derived macrophages. A strong increase in both pro‐ and anti‐inflammatory markers is observed on grooves,…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5- —Dutch Ministry of Education, Culture, and Science
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdvanced Materials and Mechanics · 3D Printing in Biomedical Research · Cellular Mechanics and Interactions
Introduction
1
The inflammatory response is inherently dynamic, involving continuous cell recruitment, phenotypical changes, and tissue restructuring [1, 2, 3, 4, 5]. These events are tightly interconnected, and regulation or dysregulation of any of them can alter the outcome of inflammation. With the rapid expansion of regenerative medicine, there is a growing interest in the on‐demand modulation of immune cell behavior to improve tissue repair and implant integration. For instance, understanding how macrophages, key players in the early immune response, are influenced by tissue remodeling and biophysical cues can be used to predict and prevent inflammatory responses [6, 7, 8, 9, 10]. Macrophages are highly mechanosensitive and integrate biophysical cues from their microenvironment to regulate and coordinate the cellular response at the injury site [11, 12, 13, 14]. To investigate and modulate macrophage mechanosensitivity, dynamic and reversible modulation of surface topographies provides a promising approach to alter activation states and polarization profiles. An upregulated anti‐inflammatory activity may drive chronic inflammation or excessive scar formation [9, 15, 16, 17]. However, the use of dynamic surface topographies to mechanically modulate macrophage immune responses has not yet been studied.
Cell environment modulation, such as physical modulation of the extracellular matrix [18, 19, 20] or surface topographies [21, 22, 23, 24, 25, 26, 27] have emerged as a powerful tool for guiding cell behavior, enabling researchers to mimic the mechanical and structural cues of native tissue environments. Amongst the materials that can offer dynamically changing surface patterns, light‐responsive polymers offer an attractive approach to studying such on‐demand cell modulation in conditions that better mimic the constantly changing extracellular matrix. Because light is a non‐invasive, biocompatible stimulus for surface modulation, photo‐responsive materials are suited for in vitro cellular behavior studies [28, 29, 30, 31, 32, 33, 34, 35, 36]. Recent studies have shown that light‐responsive materials, such as LC polymers [31, 36, 37] and hydrogels [34, 35, 38] enable reversible, photoinduced surface deformations. The deformation of light‐responsive LC polymers arises from the light‐driven modulation of molecular disorder. For example, incorporating an azobenzene photoswitch enables trans‐to‐cis photoisomerization [39], generating precise and controllable changes in the molecular disorder that lead to localized surface protrusions in the illuminated regions [28, 40, 41, 42]. Well‐defined, arbitrary topographies can be fabricated using light and various photomasks. Although the in vitro bioactivity of light‐responsive LC films has been demonstrated [36], their potential for controlling immune cell phenotypes on‐demand and reversible mechanical cues remains unexplored.
In this work, we report macrophage phenotype modulation using light‐responsive LC polymer films that generate on‐demand dynamic surface topographies of sub‐micrometer heights with features matching the size of cells [36]. To probe the adaptability and plasticity of macrophages, we employ a two‐step approach that combines cell‐compatible topographical generation and removal (Figure 1). The cells might experience this switching surface topography as a dynamic mechanical cue, enabling direct investigation of mechanical memory and phenotype reprogramming. The platform allows us to create on‐demand mechanical cues and to assess whether cells can adapt to repeated mechanical stimuli in their extracellular environment by means of morphology, inflammatory markers, and cytokine secretion. Our study demonstrates that light‐responsive LC surfaces represent a controllable and translational platform for adaptive immunomodulation, with potential for the rational design of immunoregulatory scaffolds in regenerative medicine.
Generation of dynamic surface topographies for cell modulation at physiological conditions. The initial film containing macrophages is bottom mask UV illuminated (365 nm), generating a topography of 50 µm diameter pillars spaced 50 µm apart, after which visible light (455 nm) illumination removes the topography. The first generated topography, T1, functions as a preconditioning step to prime macrophage mechanosensitivity, as such pre‐activation enhances the cellular capacity to respond to subsequent mechanical stimuli. A sequential second mask is applied for generating a new surface topography T2, pillars of 50 and 25 µm spacing on equal distance, and 50 µm grooves on 50 µm spacing, by bottom mask UV illumination. This design enables systematic investigation of macrophage adaptation to dynamic, sequential topographical cues.
Results and Discussion
2
LC Film Fabrication
2.1
Light‐induced surface topographies were created using azobenzene‐functionalized LC films. These LC films enable fabrication of precise and controllable surface topographies in vitro by applying bottom‐mask illumination (Figure 1) [36]. The LC polymer films were prepared with a mixture of LC mono‐ and diacrylate monomers, an acrylate photoisomerizable azobenzene, an acrylate chiral dopant, a chain transfer agent, and a photo‐initiator as earlier reported (SI). A ∼20 µm thick LC film is prepared by depositing the LC mix between two glass plates spaced on 20 µm distance, which is sequentially photopolymerized using visible light, followed by a 5 min post‐baking step (80°C). After removing the top glass plate, the flat LC film showed a mean surface roughness of ∼30 nm (SD = 5, Table S1). Sequential UV illumination (365 nm) without the application of a mask in air at 37°C without the presence of cells did not affect the roughness. However, visible‐light (455 nm) exposure significantly reduced the roughness to ∼10 nm (SD = 4), (Table S1), which might be explained by the *trans‐*azobenzene isomers requiring less free volume than the cis‐azo isomer [43]. UV light illumination triggers trans‐to‐cis photoisomerization of the azobenzene, while visible light illumination promotes *cis‐*to‐trans isomerization [36].
Macrophage LC‐Film Biocompatibility
2.2
For cell culturing, all the LC films were functionalized with a laminin coating prior to cell‐seeding. To study whether the LC films and UV illumination are compatible with the macrophage cell cultures, we first evaluated the macrophage response of nonexposed films (cells were maintained in the dark) and fully exposed films that were maskless illuminated with two cycles of UV light (365 nm) and one cycle of visible light (455 nm) exposure (vide infra).
After the cell differentiation period, the nonexposed LC films and fully exposed LC films were stained with a nuclear dye and a dead‐cell dye to evaluate cell viability. Fully differentiated macrophages on nonexposed LC films showed a 93% viability (SD = 2.36), and a viability of 82% (SD = 4.7) for the fully exposed film (Figure 2a). Besides cell viability, DNA damage was quantified by staining for phosphorylated H2AX histone, indicative of DNA double‐strand break (Figure 2a) [35]. These DNA damage measurements were performed as an initial cytocompatibility screening to exclude overt genotoxic effects and are not intended as statistically powered comparisons. Macrophages cultured on a nonexposed LC film showed 3.9% of DNA damage‐positive cells (SD = 1.28), and cells cultured on the fully exposed control showed 3.0% of DNA damage‐positive cells (SD = 2.10). In addition, cells differentiated and grown on a glass surface showed 3.3% of DNA damage‐positive cells (SD = 3.1). These results demonstrate that neither the LC film's physical and chemical properties nor the applied UV/visible‐light activation cycles cause DNA damage in cultured macrophages. Exposure to two cycles of UVA light and one cycle of visible light resulted in an 11% reduction in cell viability, without any detectable long‐term DNA damage. UVA and visible‐light wavelengths are generally considered less harmful than UVB or UVC [44, 45]. Both UVA and blue light, however, can generate reactive oxygen species, which may lead to cell death and DNA damage [46, 47, 48]. Considering that we expose the 20 µm film from the bottom side with low‐intensity light (10 mW/cm^2^) and given the concentration of azobenzene photoswitches in the film, the light exposure of the cells will be very limited [36]. Our results reveal that LC films and the light actuation protocol are compatible with macrophage culture and suitable for studying dynamic mechanomodulation.
LC film biocompatibility and THP‐1 macrophage interaction with LC films. (a) Cell viability quantification on nonexposed and fully exposed LC films. The total cell population is represented by the blue bar depicting the live cell population and the red bar showing the dead cell population. The fluorescence image is representative of a typical view of co‐staining of total cell DNA (blue staining) with a phosphorylated H2AX histone (γH2AX, red staining). Pink nuclei are a combination of both nuclear and γH2AX staining and show DNA damage in the cell. The graph represents the quantification of DNA damage in macrophages on nonexposed and fully exposed LC films. The total cell population is represented by the blue bar, depicting the entire cell population, and the purple bar, showing the DNA damage‐positive cell population. N = 1 (b) Morphological analysis of the cells on nonexposed and fully exposed LC films. Quantification derived from CellProfiler analysis of the actin staining of the macrophages on flat and topography LC films. All conditions were normalized to the nonexposed control and represented in a violin plot with the mean and median values, as well as data distribution, groups comparisons performed by an unpaired nonparametric Kolmogorov‐Smirnov test. N = 3, no significant differences detected.
Fluorescence microscopy revealed no detectable changes in cell surface features between the nonexposed and fully exposed films (Figure 2b; Figure S1). Morphological analysis of fluorescent staining images, furthermore, showed no significant differences in cell area (1.00‐fold, SD = 0.18; 1.78‐fold, SD = 1.82; respectively) but a larger cell size distribution in the fully exposed samples. Furthermore, eccentricity indicative of the cell body circularity (1.00‐fold, SD = 0.12; 0.98‐fold, SD = 0.13; respectively), and the cell body extent indicative of the number of membrane protrusions (1.00‐fold, SD = 0.07; 1.00‐fold, SD = 0.07; respectively) also showed no significant differences in between conditions.
Notably, distinct differences in cell morphology and membrane topography were observed when comparing macrophages on the nonexposed and fully exposed LC film surfaces to those on conventional glass coverslips (Figure S2). On glass, cells had a more extended morphology with a larger cell area, multiple extensions surrounding the cell anchoring it to the surface, and a “ruffled” membrane structure. These ruffles, lamellipodia, are sheet‐like integrin‐rich domains that play critical roles in substrate adhesion and endocytosis [49, 50]. Additionally, when the cell extends its membrane to bind to the surface, it forms what is known as filopodia—the thin crown extensions around the cell body. Filopodia's main function is to tether the cell to the surface [51, 52, 53]. Both lamellipodia and filopodia are responsible for the cell's monitoring of the extracellular environment mechanism and assist with cell motility and mechanosensing [54, 55, 56, 57]. Macrophages seeded on the LCs, however, displayed a “blebbed” membrane morphology regardless of the illumination state (Figure 3a,b; Figure S2), no lamellipodia formation, and a reduced amount of filopodia. Due to the identically applied cell treatment and differentiation protocol, this morphological difference stems from an intrinsic effect of the LC material on macrophage behavior. As previously reported, cells can adopt plasma membrane blebs to bypass unfavorable migration conditions and facilitate their long‐distance migration [58, 59, 60]. Membrane blebs are therefore a functional adaptation of the cell, as they serve as a reservoir for excess membrane [61]. The elevated presence of surface blebs is therefore presumably owed to the low attachment of macrophages to the LC film, as well as cytoskeletal tension caused by mechanical stimuli such as roughness changes or dynamic surface movement, generating higher cytoskeletal tension [49, 62, 63]. The macrophages on LC films show a differential behavior, where the cells have low attachment due to the dynamically changing surface roughness, suggesting a functional adaptation toward high motility and a modulatory effect on macrophage polarization.
Macrophage phenotype and morphology characterization on control LC film conditions. Showing from left to right: fluorescence microscopy (x2), and scanning electron microscopy on single macrophages on (a) nonexposed and (b) fully exposed LC films (cells on LC films subjected to the same UVA/visible illumination cycles used during actuation but without topography patterning). Nuclei are represented in blue, f‐actin in green, CD206 in orange, and iNOS in red. Scale bars 50, 20, and 5 µm from left to right. Scale bar of zoomed scanning electron microscopy image 1 µm. Colors are artificially increased for visualization purposes and do not correspond to real intensity ratios. (c) iNOS marker quantification of the macrophages on LC films. (d) CD206 anti‐inflammatory marker quantification in macrophages on LC films. (e) F‐actin intensity quantification in macrophages on LC films. (c‐e) normalized to the nonexposed control (f) Cytokine quantification in the extracellular media of macrophages on LC films. All values were normalized to the number of cells in each condition and represented as a concentration (pg/mL) per 1 × 105 cells. (c‐f) All statistical comparisons were performed by an unpaired nonparametric Kolmogorov‐Smirnov test. N ≥ 3, * p < 0.05, ** p < 0.01, *** p <0.001, **** p < 0.0001.
Next, we evaluated surface marker expression of macrophages cultured on LC films to probe the phenotypical response to the LC film polymers' surface properties (Figure 3c–e). Macrophages on the fully light‐exposed film exhibited a discernible phenotypic shift compared to the non‐exposed sample. Immunofluorescence staining revealed a slight reduction in the pro‐inflammatory marker iNOS (inducible nitric oxide synthase; 0.95‐fold of nonexposed control, SD 0.11, Figure 3c), and conversely, an increased expression of the anti‐inflammatory marker CD206 by 3.17‐fold (SD 1.75, Figure 3d), relative to the nonexposed control markers (1.00‐fold, SD 0.11) and (1.00‐fold, SD 0.87), respectively. This phenotypic shift was supported by cytoskeletal analysis, which showed increased actin staining intensity in macrophages on the fully exposed sample (Figure 3e), indicating enhanced actin polymerization or cytoskeletal remodeling, potentially triggered by mechanical cues associated with the LC film [64, 65, 66]. These results further substantiate that macrophages detect nm‐scale dynamic roughness changes and respond via cytoskeletal remodeling and phenotype modulation, supporting the idea that dynamic LC topographies can reprogram macrophages in situ [67, 68, 69, 70].
Cytokine profiling also revealed a differential macrophage response to the nonexposed and fully exposed LC film surfaces. Anti‐inflammatory cytokines (arginase, IL‐1RA, and IL‐10) were largely unchanged. In contrast, pro‐inflammatory cytokines exhibited selective downregulation: IFN‐γ secretion was significantly reduced in the fully exposed sample (Figure 3f), accompanied by a decrease in TNF‐α (Figure S3 and Table S2), but no significant changes in IL‐8 (Figure 3f), IL‐1β, IP‐10, or IL‐12p70 (Figure S3 and Table S2). Taken together, these findings suggest that light‐responsive LC film surfaces exert a baseline immunomodulatory effect through reversible nanoscale roughness changes, selectively reducing pro‐inflammatory cytokine secretion (e.g., IFN‐γ, TNF‐α) and promoting CD206 expression, even in the absence of patterned topographies. This effect reflects macrophage sensitivity to subtle mechanical cues rather than a complete polarization toward an anti‐inflammatory phenotype, highlighting the potential of LC films for adaptive immunomodulation without chemical or structural modifications.
Dynamic Light‐Responsive Surface Topographies
2.3
The LC film allows for topography generation via bottom mask‐illumination employing UV light. For the dynamic mask exposed light‐responsive surface topographies, we focused on two distinct patterns, pillars and grooves (Figure 1, Figure 4iv, and Figure 5a–c), building on our previous studies [36]. As THP‐1‐derived macrophages have an average diameter of ∼20–25 µm, topography dimensions were chosen to match or exceed cell size, enabling effective mechanosensing. Such topographical dimensions are also widely explored in implant and device design, as micro‐ and submicron features comparable to ECM architectures that are known to regulate macrophage adhesion, elongation, and inflammatory signaling [23, 25, 26]. Pillars were therefore fabricated with diameters and pitch sizes matching the average macrophage size (25 µm, Figure 4b iv and Figure 5b) and double that size (50 µm, Figure 4a iv and Figure 5a). Grooves were 50 µm wide with a 50 µm pitch (Figure 4c iv and Figure 5c).
Overview on dynamic LC surface patterning at 37°C in air. i. The initial flat LC films showing nanometer surface roughness. ii. LC surface topographies after the first topography generation, employing a circular mask with 50 µm diameter features and 50 µm spacing. iii. Reset the LC surface by sequential visible light exposure, removing pillars to yield a flat surface. iiii. Second patterning step, employing 50 µm diameter pillars on 50 µm distance (a), 25 µm diameter pillars on 25 µm distance (b), and a 50 µm groove width on 50 µm distance.
Macrophage phenotype and morphology characterization on dynamic surface topographies for T2. Showing from left to right: fluorescence microscopy (x2), and scanning electron microscopy of single macrophages on (a) Pillars 50 × 50 µm, topography, (b) Pillars 25 × 25 µm, topography, and (c) Grooves 50 × 50 µm, topography LC films. Nuclei are represented in blue, f‐actin in green, CD206 in orange, and iNOS in red. Scale bar 50, 20, and 5 µm from left to right. Scale bar of zoomed SEM image 1 µm. Colors are artificially increased for visualization purposes and do not correspond to real intensity ratios. (d) Morphological analysis of the cells on nonexposed LC and T2 topographies. Quantification derived from CellProfiler analysis of the actin staining of the macrophages on flat and topography LC films. All conditions normalized to nonexposed control and represented in a violin plot with the mean and median values, as well as data distribution, groups comparisons performed by unpaired non‐parametric Kruskal‐Wallis test with Dunn's multiple comparisons test, N = 3, * p<0.05, ** p<0.01, *** p<0.001, **** p< 0.0001. (e) Inducible Nitric Oxide Synthase (iNOS) marker quantification of the macrophages on LC films. (f) CD206 anti‐inflammatory marker quantification in macrophages on LC films. (g) F‐actin intensity quantification in macrophages on LC films. (h) Cytokine quantification in the extracellular media of macrophages on LC films. All values were normalized to the number of cells in each condition and represented as a concentration (pg/mL) per 1 × 105 cells. (e‐h) All statistical comparisons were performed by an unpaired nonparametric Kruskal‐Wallis test with Dunn's multiple comparisons test. N ≥ 3, * p < 0.05, ** p < 0.01, *** p <0.001, **** p < 0.0001.
Topographical dimensions were first characterized in the absence of cells at 37°C. We employed the same two‐step approach as for the topographical generations and removals in the presence of the macrophages. In the first step, using a circular mask with circles of 50 µm in diameter with a 50 µm pitch, the LC films yielded topographies (T1) with averaged peak‐to‐valley heights of 710 nm (SD = 66 nm, N = 12, Figure 4 ii) [36]. Sequentially, the topographies were removed with a maskless visible light illumination [36]. After topography removal, bottom UV‐illumination with a new desired mask allowed for the fabrication of a desired new surface topographical cell‐environment (T2). In this second patterning step, three distinct masks were used to create three unique topographies: (i) pillars with 50 µm diameter and 50 µm spacing; (ii) pillars with 25 µm diameter and 25 µm spacing; and (iii) grooves with 50 µm width and 50 µm spacing (Figure 4 iv), showing peak‐to‐valley heights of 615, 501, and 508 nm (SD = 32, 53, 34 nm, N = 3), respectively. This dual‐patterning approach enables real‐time, reversible generation of dynamic topographies, allowing macrophages to experience on‐demand mechanical cues similar to dynamic extracellular matrices. The first topography acts as a preconditioning step, priming macrophage mechanosensitivity and enhancing their response to subsequent mechanical stimuli [71, 72, 73, 74].
Fluorescence microscopy analysis revealed distinct topography‐dependent morphology and marker expression changes (Figure 5a–d; Figure S1). When looking at the macrophage distribution over the patterned areas (Figure 5a–c), macrophages preferentially localized around patterned areas, clustering near topographical features. Moreover, when analyzing morphological parameters, the largest cell surface area was observed on 50 × 50 µm pillars (Figure 5d), with a 1.19‐fold increase (SD = 0.34) relative to control (1.00‐fold, SD = 0.18). This increase was statistically significant compared to both the nonexposed control (1.00‐fold, SD = 0.18) and grooves (0.99‐fold, SD = 0.62), but not significantly different from the 25 × 25 µm pillars condition (1.09‐fold, SD = 0.44). Overall, pillars, particularly 50 × 50 µm, promoted greater cell spreading than grooved and non‐patterned surfaces. For eccentricity, reflecting the cell elongation, macrophages on grooves (0.99‐fold, SD = 0.09) were comparable to the nonexposed control (0.98‐fold, SD = 0.07). In contrast, both pillar topographies reduced eccentricity with fold changes of 0.94 (SD = 0.10) for 25 × 25 µm and 0.94‐fold (SD = 0.11) for 50 × 50 µm pillars as compared to the nonexposed control. These values were significantly lower than those observed in both the control and grooved conditions, but not significantly different from each other. This indicates that pillars promote more elongated cell shapes, whereas grooves maintain baseline morphology. Extent (cell protrusions) was unaffected across conditions (Figure 5d). In all cases, surface topographies did not influence the protrusive behavior of macrophages. Following, we examined phenotype‐specific markers in the fixed cells, revealing a complex, topography‐dependent effect for T2 (Figure 5e–g). Expression of iNOS (Figure 5e) remained largely unchanged across all topographies, with a higher distribution of values, as well as a higher iNOS cytoplasmic content on the grooves topography (Figure S1e). On the other hand, CD206 expression (Figure 5f) was consistently upregulated, with grooves yielding the strongest CD206 upregulation (4.2‐fold, SD 2.23), and pillars showing modest increases (1.32–1.59‐fold). F‐actin intensity decreased on grooves (0.76‐fold, SD 0.23, Figure 5g; Figure S1), reflecting redistribution from condensed, migration‐focused cytoskeleton toward lamellipodia formation observed by scanning electron microscopy (Figure S2). Together, these results support a groove‐driven, pro‐remodeling/mixed phenotype, whereas pillars maintain a low‐activation, anti‐inflammatory profile.
SEM imaging provided further insight into topography‐directed phenotype modulation. On pillars (Figure 5a,b), macrophages retained the blebbed membranes with minimal filopodia formation, consistent with low attachment and high migration potential. Grooves (Figure 5c) uniquely induced a mixed membrane morphology with blebs and lamellipodia on the membrane, and increased filopodia formation, enhancing surface interaction. Moreover, formation of lamellipodia and more filopodia can also be seen in a single exposure step (T1) (Figure S4), when the macrophages were exposed to only grooves. This presents an intriguing interaction between the light‐induced mechanical cue and the specific response evoked by grooves. Lamellipodia have been linked to enhanced binding during foreign‐object detection and initiating phagocytosis. Furthermore, shape compatibility also plays a role: ellipsoid and rod‐like particles fit more easily between membrane ribbons, improving binding [75]. Grooves, with their continuous ribbon‐like profile, may encourage macrophages to adapt membrane structure for better surface binding, as well as provide them a continuous cue to anchor and form more filopodia. In contrast, discrete pillars are not recognized as continuous objects and therefore do not drive such membrane adaptations.
Dynamic plasticity of the macrophages was evident when comparing the membrane morphology under different topography conditions of a single and double exposure. Cells on pillars adopt the above‐discussed blebbed morphology with low surface tethering, while cells on grooves show a higher degree of attachment (Figure S2). Additionally, upon visual analysis, when the cells are first exposed to pillars (T1) and afterward to grooves (T2), the cell membrane morphology dynamically reacts to the new topography, increasing the number of filopodia and forming lamellipodia sheets on the surface (Figure 5c; Figure S2). Furthermore, when the grooves are used in a second exposure cycle, lamellipodia sheets are more distinctly formed and in a greater number, supporting the hypothesis that priming the cells with a first dynamic topography increases cells’ plasticity and ability to quickly adapt to the dynamic environment. This enhanced responsiveness suggests that macrophages retain a form of mechanical memory, allowing them to integrate prior topographical cues into subsequent morphological adaptations [72, 74].
Cytokine profiling (Figure 5h) revealed topography‐dependent functional modulation. Grooves consistently exhibited elevated expression of all measured cytokines, indicating a general increase in cellular activity rather than a shift toward a specific polarization state. Although grooves produced slightly higher anti‐inflammatory cytokine levels (arginase, IL‐1RA, IL‐10) compared to pillars, these increases were modest. These results indicate that while grooved topographies stimulate a higher overall cytokine output, the magnitude of increase in anti‐inflammatory cytokines remains relatively moderate when compared to the pillar topographies.
Pillars maintained minimal expression of pro‐inflammatory markers (IFN‐γ and IL‐8) with many samples measuring below the quantification threshold (Figure 5h). In contrast, grooves induced robust and consistent increases in both cytokines across replicates. Grooves also promoted higher and more consistent expression of additional cytokines, including both pro‐inflammatory (TNF‐α, IL‐1β) and regulatory (TGF‐β, IL‐2) markers (Figure S3 and Table S2). Notably, TNF‐α and IL‐2 levels were greater than those on pillars but still lower than in the nonexposed controls. In contrast, TGF‐β and IP‐10 reached their maximum levels across all LC conditions on grooves (Figure S3 and Table S2), highlighting the influence of groove geometry on a regulatory/mixed macrophage phenotype. Interestingly, IL‐4 (an anti‐inflammatory cytokine) was consistently detectable in the grooves, but it remained unchanged versus controls (Figure S3 and Table S2). These findings suggest that grooves may simultaneously activate both pro‐ and anti‐inflammatory pathways, driving a mixed macrophage phenotype. It should be noted that cytokine production after a single versus double exposure (Figure S4 and Figure 5h, respectively) revealed that the response is not strictly time‐dependent: one exposure lasted 3 h, and two exposures 6 h. Thus, IL‐10 levels increased more than 2‐fold as compared to the single topography (T1) exposure, while arginase levels increased up to 10‐fold after the second exposure (Figure 5h; Figure S4). Conversely, IL‐1RA decreased, potentially reflecting upregulation of IL‐1 signaling and transient suppression of its receptor antagonist. These amplified cytokine responses following repeated topographical stimulation align with the observed increase in membrane morphology differences after dual exposure, suggesting that macrophages retain mechanical memory of prior surface interactions. This memory may enhance their capacity to integrate sequential mechanical cues into a more robust and adaptive immunological response.
These findings highlight the critical role of surface topography in shaping macrophage behavior. Pillars maintain a stable anti‐inflammatory profile, likely driven by baseline LC film roughness rather than the discrete geometry itself, whereas grooves trigger strong, dynamic responses. Grooves increase marker expression, alter cytoskeletal organization, and elevate both anti‐inflammatory and pro‐inflammatory cytokine production, demonstrating the macrophages’ intrinsic plasticity and capacity to adopt dynamic mixed phenotypes.
We hypothesize that grooves provide a continuous mechanical stimulus that persists throughout the exposure time due to the gradual decay in topography height. The uninterrupted spatial cues along grooves promote ongoing adaptation to these features. Overall, grooves exert a stronger topographical influence while retaining these dynamic and roughness‐related features. This combination creates a more intricate microenvironment, prompting macrophages to adopt a mixed phenotype with more complex and variable functional characteristics. Pillars, by contrast, consist of discrete, unconnected features. Macrophages may treat each pillar as an isolated cue, responding locally without integrating input from surrounding structures. This discontinuity leads to more localized and less sustained cellular adaptations compared to the grooves. Therefore, cellular response to pillars, where the influence of surface geometry appears reduced, may be governed mainly by baseline LC film roughness changes. Similar macrophage behavior in relation to topographies has been reported previously [70, 76]. The emergence of a mixed macrophage phenotype on grooves highlights the potential of engineered topographies to fine‐tune immune responses in regenerative biomaterial design.
Conclusion
3
In this work, we modulated the macrophage responses using light‐responsive dynamic LC surface topographies. Two distinct topographies were studied: pillars, providing discrete/discontinuous mechanical cues, and grooves, offering continuous cues capable of sustaining dynamic, real‐time stimulation. The dynamic topographies promote an immunomodulatory response in macrophages. On all LC films except for grooves, macrophages presented a blebbed surface morphology with a large amount of reserve membrane not attached to the surface. The grooves, on the other hand, showed increased formation of surface lamellipodia and more filopodia interacting with the surface. These results demonstrate that dynamic topography and nanoscale roughness actuation can initiate and actively modulate macrophage behavior, revealing their inherent potential for morphological and phenotypic plasticity. Moreover, macrophages exhibit remarkable plasticity: even when initially exposed to pillars, a subsequent exposure to grooves elicits a response like that observed in cells that were exposed to grooves as their initial condition. Dual exposure induced even greater morphological adaptation than initial groove exposure alone, with increased lamellipodia and filopodia formation, indicating macrophage plasticity and a form of “mechanical memory” where prior cues influence subsequent responses.
Although the dynamic topographies reported in this study do not allow the induction of any specific phenotypical response, they show topography‐dependent profiles. Pillars do not induce a strong response but help maintain a consistent anti‐inflammatory profile in macrophages. In contrast, grooves triggered a distinct cellular response characterized by significant morphological changes and elevated cytokine release. Interestingly, the phenotype induced by grooves did not align strictly with classical macrophage phenotypes but instead suggested a mixed or regulatory phenotype, marked by simultaneous upregulation of both anti‐ and pro‐inflammatory markers, increased surface attachment, and cell‐surface interaction.
Overall, understanding macrophage plasticity in response to dynamic environmental cues highlights the potential of LC surfaces in the design of regenerative and adaptive immunomodulatory biomaterials. The ability to reversibly switch LC surfaces in in vitro cell culture studies provides an appealing cell‐material microenvironment that mimics the naturally dynamic 3D surroundings found in vivo. Dynamic topographies present a platform for next‐generation implantable technologies that can be reversibly actuated to guide immune responses, allowing precise control over tissue regeneration and personalized immunomodulatory therapies. Future studies should be expanded to utilize targeted topographies that induce desired responses at the injury site.
Experimental Section
4
Preparation of the Liquid Crystal Polymer Film
4.1
See SI for detailed information on the preparation of the cholesteric LC mix. [36] The prepared LC mix was deposited on acrylate and fluorinated functionalized glass slides, bottom and top, respectively. Film thickness was controlled using glue containing 20 µm spacer beads, placed at the corners of the glass slide. In‐plane shearing of the LC film yields the desired aligned planar cholesteric LC film, which is next photopolymerized for 15 min at 30°C in a dark box under air, with 455 nm wavelength light at an intensity of 27 mW/cm^2^. Subsequently, the LC confined between two glass plates was heated to 80°C for curing. After polymerization, the fluorinated glass slide can be removed, yielding the desired LC film.
Functionalization of the Liquid Crystal Polymer Film
4.2
LC films were polymerized on glass substrates with a PDMS well attached to facilitate cell culturing. Sterilization was performed by incubating the films in 70 vol% ethanol: water for 15 min, followed by 15 min evaporation, and four successive MilliQ water rinses. The sterilized LC film was coated with a laminin solution of 5 µg/mL in MilliQ, for 1 h., RT in the dark. Coated LC films were gently rinsed once with sterile PBS and instantly used for cell seeding. Control group coverslips were treated using the same protocol.
Actuation of the Liquid Crystalline Polymer Film
4.3
Reconfigurable surface topographies were fabricated by bottom UV mask illumination of the LC film. 4 min UV light exposure (365 nm, 10 mW/cm^2^) yielded the optimal topography generations, whereas surfaces are sequentially smoothened employing 10 min of visible light (455 nm, 10 mW/cm^2^). Chrome masks were obtained from JD photo data. The LED lamps obtained from Thorlabs, M365L2 and M455L3, were both mounted with Thorlabs COP1‐A collimators. Surface topographies and roughness were measured using the Sensofar S Neox white light interferometer. Data was processed using SensoView 2.3.1 software.
THP‐1 Cell Culture and Seeding
4.4
The THP‐1 monocyte cell line was purchased from ATCC. Cells were cultured in RPMI 1640 media (D5796, Sigma–Aldrich) supplemented with 10% Fetal Bovine Serum (Serana, origin Brazil) and 1% Penicillin/Streptomycin (Gibco) and passaged until maximum passage number 30. Cells were seeded on the LC film surface at a density of 1 × 10^5^ cells/cm^2^ and supplemented with 50 ng/mL phorbol 12‐myristate 13‐acetate (PMA, Sigma–Aldrich) for 24 h to induce differentiation and cell attachment to the LC film surface. After 24 h, PMA was removed, and macrophages were left to further differentiate for 48 h in fresh media.
Cell Stimulation with LC Film Actuations
4.5
After 72 h of differentiation, the cell culture medium was replaced with fresh medium immediately prior to exposure, ensuring that cytokine production measured reflects only the exposure period. LC films were exposed as per the corresponding condition and kept protected from light in a cell incubator at 37°C, 5% CO_2_ until the next cycle of actuation or cell fixation.
Cell Staining for Phenotype Assessment
4.6
At the end of each assay, THP‐1 cells were fixed with 3.7% paraformaldehyde in PBS for 30 min at room temperature, washed four times with PBS, and permeabilized with 0.1% Triton‐X in PBS for 10 min. Blocking solution (3% Bovine Serum Albumin, 5% Goat Serum, 0.3 m Glycine) was applied for 60 min at RT. The samples were washed with PBS 3 times for 5 min with mild agitation in between each step. THP‐1 cells were incubated with the mix of anti‐iNOS (1:200, ab3523, Abcam), and anti‐CD206 (1:200, MCA2155T, Bio‐Rad) primary antibodies overnight and secondary antibody mix for 1 h the following day. Afterward, cells were stained with phalloidin dye for actin staining for 15 min and DAPI dye for nuclear staining for 5 min. The samples were finally washed four times with PBS, bonded with Mowiol solution to a coverslip and left to dry over 48 h. Fluorescent imaging was performed by High Content Screening Microscope Nikon (Eclipse Ti2‐D‐PD) and Zeiss Axio Observer 7 with Apotome.
Cell Staining for DNA Damage
4.7
At the end of each assay, THP‐1 cells were fixed, permeabilized, and blocked as described above. Samples were incubated overnight with Anti‐gamma H2A.X antibody (1:200, ab22551, Abcam), followed by secondary antibody incubation for 1 h at RT the following day. Afterward, cells were stained with DAPI (5 min), washed four times with PBS, mounted with Mowiol to a coverslip, and left to dry for over 48 h. Imaging was performed by Zeiss Axio Observer 7 with Apotome with an average for 20 images per condition.
Live‐Dead Staining
4.8
Separate samples were used to assess viability. At the end of the assay, as described above, cells were incubated with 20 µg/mL Hoechst 33342 nuclear dye and 10 µm propidium iodide in PBS and incubated for 10 min. After that, the cells were immediately imaged. Per condition, 20 images were acquired, with 200–400 cells per field. Viability was calculated as the percentage of cells stained only with Hoechst versus cells co‐stained by Hoechst and propidium iodide.
Cytokine Secretion Measurement and Quantification
4.9
Cytokine release was quantified using LEGENDplex Human Essential Immune Response Panel (13‐plex) and LEGENDplex Human Macrophage/Microglia Panel (13‐plex) (BioLegend) following the manufacturer's instructions. Sample preparation included 25 µL of sample or standard, 25 µL of assay buffer, and 25 µL of beads per well, in a V‐bottom polypropylene 96‐well plate, and incubated at RT (shaking at 800 rpm) for 2 h. Plates were centrifuged at 1100 rpm (∼250 × g) for 5 min using a swinging bucket rotor with a microplate adaptor. Supernatant was removed by rapid inversion and blotting the plate. Plates were washed with 200 µL of 1× wash Buffer per well, incubated for 1 min, and centrifuged again. After supernatant removal, 25 µL of Detection Antibodies was added to each well. Plates were resealed, protected from light, and incubated at 800 rpm for 1 h. Without washing, 25 µL of SA‐PE was added directly to each well, followed by a 30 min incubation under the same conditions. A final wash step was performed as described above. Beads were resuspended in 200 µL of 1× wash buffer and transferred into flow cytometry tubes. The samples were read using a BD FACSCanto II flow cytometer. The acquired data were analyzed using the Biolegend LEGENDplex Data Analysis Software Suite. Cytokine concentration calibration curves are presented in Tables S3 and S4.
Scanning Electron Microscopy
4.10
Macrophages were cultured and stimulated on LC films as previously described. Cells were fixed with 3.7% paraformaldehyde in PBS for 30 min at RT, followed by four washes with PBS. To remove residual salts, samples were washed three times in deionized water for 10 min each. Subsequently, samples were dehydrated in a graded ethanol series: 10%, 30%, 50%, 75%, 90%, and 100% ethanol in water, each for 10 min. An additional incubation in 100% ethanol was performed to ensure complete dehydration. Samples were then incubated in 50% hexamethyldisilane (HMDS) in ethanol for 10 min, followed by 100% HMDS for another 10 min. Residual HMDS was allowed to evaporate in a fume hood for 2 h. Dried samples were sputter‐coated with a 10 nm layer of gold and imaged using a scanning electron microscope (FEI Quanta 600F). Imaging was performed at an accelerating voltage of 5 kV, spot size 3.0, and a working distance of approximately 11 mm.
Analytical and Statistical Analysis
4.11
LC surface roughness measurements were taken on three different regions of 3 independent samples (N = 3) of 1.66×1.39 mm measurement areas each, applying ISO 25178/height using Sensoview software (SensoVIEW 2.3.1). Topography peak‐to‐valley heights were calculated with N profiles taken, from which four peak‐to‐valley distances were measured as calculated by the peak height average (peak_max_ + peak_min_ / 2) deducted by the valley, for all profiles. Statistical analysis on all LC surface measurements was done using Origin (Version 2022). All cell experiments were performed in at least 3 independent replicates (N = 3) unless otherwise stated. For imaging, 20–25 images were analyzed, totaling 1000–2000 cells per condition. Cytokines were measured in duplicates per sample. Cell morphology and antibody staining intensity were quantified by CellProfiler software (CellProfiler 4.2.1, Copyright 2003–2021 Broad Institute, Inc.). Statistical analysis was done using GraphPad Prism (GraphPad Prism version 10.4.0, GraphPad Software, Boston, Massachusetts, USA). Outliers were identified using ROUT, Q = 1%, and normality was assessed with Shapiro‐Wilks test. For two‐group comparisons, the non‐parametric Kolmogorov‐Smirnov test was used to assess the differences between conditions with a significance level of 0.05. For ≥3 groups, the unpaired non‐parametric Kruskal‐Wallis test with Dunn's multiple comparisons test was used to assess the differences between conditions with a significance level of 0.05.
Author Contributions
R.M.C.V. and A.P.H.J.S. designed the LC experiments. O.K.S. and B.G. designed the cell study experiments. R.M.C.V. conducted the liquid crystal fabrication, characterization, and exposures. O.K.S. conducted the cell study experiments and cell characterization. R.M.C.V. and O.K.S. analyzed the results. A.P.H.J.S and B.G. supervised the study. All authors reviewed and have given approval to the final version of the manuscript.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Supporting File: mabi70167‐sup‐0001‐SuppMat.docx.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1J. M. Cicchese , S. Evans , C. Hult , et al., “Dynamic Balance of Pro‐ and Anti‐Inflammatory Signals Controls Disease and Limits Pathology,” Immunological Reviews 285 (2018): 147–167.30129209 10.1111/imr.12671 PMC 6292442 · doi ↗ · pubmed ↗
- 2A. L. Favier and K. Nikovics , “Molecular and Cellular Mechanisms of Inflammation and Tissue Regeneration,” Biomedicines 11 (2023): 1416.37239087 10.3390/biomedicines 11051416 PMC 10216784 · doi ↗ · pubmed ↗
- 3X. Di , X. Gao , L. Peng , et al., “Cellular Mechanotransduction in Health and Diseases: From Molecular Mechanism to Therapeutic Targets,” Signal Transduction and Targeted Therapy 8 (2023): 1501.10.1038/s 41392-023-01501-9PMC 1038748637518181 · doi ↗ · pubmed ↗
- 4I. Cambré , D. Gaublomme , A. Burssens , et al., “Mechanical Strain Determines the Site‐Specific Localization of Inflammation and Tissue Damage in Arthritis,” Nature Communications 9 (2018): 4613.10.1038/s 41467-018-06933-4PMC 621847530397205 · doi ↗ · pubmed ↗
- 5R. Mata , Y. Yao , W. Cao , et al., “The Dynamic Inflammatory Tissue Microenvironment: Signality and Disease Therapy by Biomaterials,” Research 2021 (2021): 4189516.33623917 10.34133/2021/4189516 PMC 7879376 · doi ↗ · pubmed ↗
- 6M. Torres , J. Wang , P. J. Yannie , S. Ghosh , R. A. Segal , and A. M. Reynolds , “Identifying Important Parameters in the Inflammatory Process With a Mathematical Model of Immune Cell Influx and Macrophage Polarization,” PLOS Computational Biology 15 (2019): 1007172.10.1371/journal.pcbi.1007172 PMC 669055531365522 · doi ↗ · pubmed ↗
- 7S. R. Petrousek , G. S. Kronemberger , G. O'Brien , et al., “Mechano‐Immunomodulation of Macrophages Influences the Regenerative Environment of Fracture Healing Through the Regulation of Angiogenesis and Osteogenesis,” Acta Biomaterialia 200 (2025): 187–201.40409508 10.1016/j.actbio.2025.05.045 · doi ↗ · pubmed ↗
- 8B. Saha , D. Tornai , K. Kodys , et al., “Biomarkers of Macrophage Activation and Immune Danger Signals Predict Clinical Outcomes in Alcoholic Hepatitis,” Hepatology 70 (2019): 1134–1149.30891779 10.1002/hep.30617 PMC 6752989 · doi ↗ · pubmed ↗