N-cadherin–mimetic 3D hydrogels program pro-regenerative and immunomodulatory states in human adipose-derived mesenchymal stem cells

Abstract

Mesenchymal stromal cell (MSC) therapies critically depend on culture systems that preserve stem cell identity while appropriately configuring immunomodulatory functions. Here, we investigated how a biomimetic three-dimensional (3D) hydrogel that presents the N-cadherin–derived HAVDI motif shapes the phenotype and inflammatory programming of human adipose-derived MSCs (ADMSCs). Self-assembled HAVDI-functionalized peptide amphiphile hydrogels formed nanofibrous, mechanically stable networks that supported high cell viability and sustained encapsulation. Across all conditions—including tissue culture plastic, micromass aggregates, and 2D/3D peptide formulations—flow cytometry showed that ADMSCs remained > 90% positive for canonical MSC markers CD73, CD90, and CD105, indicating global preservation of MSC surface phenotype, with peptide-modified environments modestly stabilizing marker expression relative to uncoated plastic. In 3D HAVDI hydrogels, gene expression profiling revealed robust upregulation of p120-catenin and β-catenin, together with increased transcription of matrix-remodeling and angiogenesis-related genes (MMP2, PLAU, VEGFR2), consistent with a pro-regenerative program. Notably, 3D HAVDI cultures displayed markedly elevated basal expression of multiple immunoregulatory cytokine genes (IL-1α, IL-1β, IL-8, IFN-γ, TNF-α, GM-CSF) under LPS-negative conditions, followed by a pronounced downregulation upon LPS challenge, suggestive of a tolerance-like, controlled inflammatory response rather than unchecked activation. Collectively, these findings show that HAVDI-functionalized 3D hydrogels provide a bioinstructive niche that maintains MSC phenotype while priming ADMSCs into a regenerative and tightly regulated immunomodulatory state, highlighting their potential as an advanced platform for stem cell–based tissue repair and immune modulation.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-31276-8.

Introduction

Mesenchymal stromal cells (MSCs) are widely used in regenerative medicine owing to their multilineage differentiation capacity and potent immunomodulatory behavior. Among various MSC sources, adipose-derived MSCs (ADMSCs) are particularly advantageous because they are abundant, accessible through minimally invasive procedures, and exhibit robust proliferative and paracrine activity with low immunogenicity^1–4.

The behavior and therapeutic potency of MSCs are strongly regulated by the extracellular microenvironment⁵. Conventional two-dimensional (2D) culture on rigid plastic alters MSC morphology, promotes loss of stemness and secretory activity, and fails to mimic the native stem cell niche^6,7. In contrast, three-dimensional (3D) biomaterials recreate physiologically relevant features such as matrix stiffness, ligand presentation, and nanoscale topography^8–10. Biomimetic hydrogels provide tunable mechanics and biochemical functionality, enabling regulation of MSC adhesion, proliferation, and fate^11,12.

Peptide-functionalized hydrogels represent a next-generation strategy in 3D niche engineering. Short peptide motifs can be incorporated into hydrogels to emulate native adhesion mechanisms and activate specific signaling cascades. One such motif is HAVDI, derived from the extracellular domain of N-cadherin, a key cell–cell adhesion protein that regulates β-catenin signaling, cytoskeletal organization, and MSC mechanotransduction^13–16. Recent reports indicate that synthetic matrices displaying cadherin-derived peptides can preserve stemness, influence differentiation decisions, and modulate MSC–immune crosstalk^17,18,18.

Beyond biochemical cues, the mechanical and dimensional properties of the niche also influence MSC function. Unlike rigid 2D substrates that exhibit stiffness values in the gigapascal range, hydrogels approximate the physiological stiffness of the native niche (∼kPa) and support a rounded morphology associated with increased quiescence, stemness maintenance, and paracrine activity^19,21,21. Thus, integrating biophysical confinement with cadherin-mediated adhesion may provide synergistic control over MSC phenotype.

Importantly, MSC immunomodulation is not constitutive, but must be activated or “licensed” by microenvironmental cues. While inflammatory cytokines (e.g., IFN-γ or TNF-α) are recognized licensing agents^22,24,24, mechanotransductive inputs and cell–cell adhesion signals can also prime MSCs for immunoregulatory activity, even in the absence of inflammation²⁵. Despite the central role of N-cadherin in stem cell quiescence and fate specification, whether cadherin-mimetic interactions can function as a non-inflammatory licensing stimulus for ADMSC immunomodulation has not been explored.

This study, therefore, was designed to test the hypothesis that mimicking N-cadherin interactions within a physiologically relevant 3D hydrogel could program ADMSCs toward a dual phenotype: one that retains stemness while being functionally preconditioned for immunomodulatory and regenerative action. To explore, we cultured human adipose-derived MSCs (ADMSCs) within peptide amphiphile hydrogels presenting the HAVDI motif. This system was designed to mimic key aspects of the stem cell niche by providing both N-cadherin–like cell–cell interaction cues and a mechanically stiff 3D matrix. We systematically assessed ADMSC behavior in HAVDI-PA versus scrambled control hydrogels under 2D and 3D conditions, including proliferation, expression of stemness-associated surface markers, β-catenin pathway-related genes, pro-inflammatory cytokine transcripts and proteins, LPS-induced responses, and regenerative gene expression. This approach allowed us to characterize how HAVDI-PA–based hydrogels modulate the phenotypic and molecular profile of ADMSCs and provided insights into the design of synthetic niches that may support both regenerative function and immunological conditioning.

Materials and methods

Materials

All amino acids, 2-(1 H-benzotriazol-1-yl)−1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU), rink amide MBHA resin (0.3–0.6 mmol/g, 100–200 mesh) and Lauric acid were purchased from Chem- Impex International (Wood Dale, IL, USA). Piperidine, trifluoroacetic acid (TFA), triisopropylsilane (TIS) were obtained from Acros (Beijing, China) and used as received. Water was double distilled using a Millipore simplicity 185 (Darmstadt, Germany) purification system (18.1 MΩ) which was fed with distilled water through an internal tank [44].

Peptide synthesis

Peptides were synthesized manually using the Fmoc-based solid-phase peptide synthesis (SPPS) method on Rink amide MBHA resin. This process was conducted at room temperature, under normal atmospheric conditions. Initially, the resin was allowed to swell in dimethylformamide (DMF) for at least 30 min. Amino acids were then introduced using HBTU as a potent coupling agent. For the coupling reaction, we used three equivalents of Fmoc-protected L-amino acids, nearly three equivalents of HBTU, and six equivalents of N, N-diisopropylethylamine (DIPEA) in DMF.

The Fmoc protecting groups were removed by treating the resin with a 20% piperidine solution in DMF, done twice for 10 min each for 0.5 g of resin. To prevent reactions at unreacted amines, we capped the N-terminus of the peptides with a DMF/Ac2O/Pyridine solution for 5 min. Following each step, we performed washes with DMF to clean the resin.

We monitored the synthesis progress using the Kaiser test, assessing peptide couplings and Fmoc removal. After peptide chains were fully assembled, the peptidyl resin was washed sequentially with DMF, dichloromethane (DCM), methanol (MeOH), and diethyl ether (Et2O), followed by air drying. For peptide cleavage, the resin was treated with a TFA/Triisopropylsilane (TIS)/H2O mixture for 2–3 h. The cleaved peptides were then separated from the resin by filtration, precipitated in cold diethyl ether, and centrifuged to isolate the solid peptides. This precipitation and washing step were repeated three times. After air drying, the peptides were purified using reversed-phase high-performance liquid chromatography (HPLC) and subsequently lyophilized to obtain them as pure powders. Peptides used in biological assays were derived from both in-house synthesis and commercial sources (Lifetein). Batch-to-batch consistency was confirmed by comparing retention times and molecular masses using HPLC and MS. No significant differences were observed, and representative batches were used across all experiments.

High-performance liquid chromatography (HPLC)

Analytical HPLC spectra of peptides were recorded on a Dionex UltiMate 3,000 HPLC system (Sunnyvale, CA, USA) equipped with a Gemini-NX C18 column dimensions 150 × 3 mm, particle size 3 μm, pore size 110 Å (Phenomenex, Torrance, CA, USA) and 20 µL injection loop. Peptides were eluted in acetonitrile containing 0.08% TFA (v/v) and water containing 0.1% TFA (v/v) gradient (5–100%, 1–35 min, flow 0.4 mL/min). Preparative HPLC of peptides was performed on a Dione UltiMate 3,000 HPLC system (Sunnyvale, CA, USA) equipped with a semi-preparative Gemini-NX C18 column-dimensions 150 × 10 mm, particle size 10 μm, pore size 110 Å (Phenomenex, Torrance, California, U.S.A) and 2 ml injection loop. For elution of peptides, a linear gradient of acetonitrile and water containing 0.08% TFA (v/v) and 0.1% (v/v) respectively (5–100%, 1–70 min) and a flow of 4.0 mL/min were used. All peptide amphiphile batches (HAVDI-PA, SCRAM-PA, and K-PA), including both in-house synthesized and commercially obtained materials, were analyzed by analytical HPLC to assess purity and confirm batch equivalence prior to biological testing. Representative chromatograms are presented in Figure S1.

Mass spectrometry

For the structural confirmation of the peptides, Agilent Technologies 6530 Accurate-Mass Q-TOF was used. Concentration of the samples for HRMS measurements was arranged approximately to 0.1 mg/ml. Mass spectra were recorded in positive-ion modes. Both laboratory-synthesized and commercially acquired batches were analyzed to confirm consistency in m/z values. Representative spectra are provided in Figure S1.

Circular dichroism (CD) spectroscopy

Circular dichroism spectra were recorded to assess the secondary structure of the synthesized peptides. Measurements were carried out using a Jasco J-810 Spectropolarimeter (Easton, MD, USA). Peptide samples were dissolved in phosphate-buffered saline (PBS) to a final concentration of 0.2 mg/mL. The spectra were obtained from 190 nm to 260 nm at room temperature, with a bandwidth of 1 nm, a response time of 4 s, and a scan speed of 50 nm/min. Each spectrum represented an average of three scans (Figure S2).

Scanning electron microscopy (SEM)

Morphological characterization of the peptide hydrogels was performed using a FEI Quanta 250 FEG Scanning Electron Microscope (Hillsboro, OR, USA). Hydrogels were prepared by diluting the peptides in PBS to a concentration of 1% w/v and allowing gelation at room temperature for 24 h. Samples were then fixed with 2.5% glutaraldehyde, dehydrated in an ascending ethanol series, and dried under vacuum. The dried samples were sputter-coated with gold-palladium and imaged at an accelerating voltage of 5 kV.

Rheological measurements

The viscoelastic properties of the peptide amphiphile hydrogels were characterized using an AR-G2 rheometer (TA Instruments, New Castle, DE, USA) equipped with a 20 mm parallel plate geometry. Hydrogels were carefully loaded onto the rheometer stage, and excess material was trimmed. All measurements were conducted at 25 °C following a 5-minute equilibration period. Strain sweep tests were first performed over a range of 1% to 100% strain at a constant angular frequency of 6.28 rad/s (1 Hz) to determine the linear viscoelastic region (LVR). Subsequently, frequency sweep tests were conducted within the LVR (at 1% strain), using an angular frequency range of 0.1–100 rad/s (0.1–10 Hz) to evaluate storage modulus (G′) and loss modulus (G″) profiles.

Adipose derived mesenchymal stem cell culture

Human Adipose-Derived Mesenchymal Stem Cells (PCS-500-011 TM) were obtained from ATCC (Wesel, Germany). The cells were cultured in DMEM (Capricorn, Germany) supplemented with 20% Fetal Bovine Serum (FBS), 2 mM L-glutamine (Sigma-Aldrich, Milan, Italy), and 1% antibiotic-antimycotic solution (BIOIND, Israel). The cells were maintained in an incubator at 37 °C with 5% CO₂ and 100% humidity. The culture medium was replenished every two to three days until the cells reached a confluency of 70–80%. This optimized culture environment provided the necessary conditions for the growth and expansion of the adipose-derived mesenchymal stem cells, ensuring their viability and maintaining their characteristic properties for subsequent experiments and applications.

Designing of 3D and 2D culture environment and treatment

Equimolar solutions of HAVDI-PA, SCRAM-PA and K-PA were prepared, each at a concentration of 1 µM (refer to Supplementary Table S1). The role of the K-PA peptide amphiphile was to balance charges in the mixture with SCRAM-PA and HAVDI-PA. These solutions were then uniformly applied to coat the wells of a 24-well plate. Following the coating process, the plates were subjected to overnight drying under UV light. The next day, cells (3.5 × 10⁴ cells/well) were seeded onto the peptide-coated wells to examine the effects of the HAVDI bioactive sequence. Before treating the cells with LPS, they incubated for 24 h.

To investigate the HAVDI bioactive sequence in a three-dimensional context, a similar approach was employed using a 24-well plate. In this setup, 5 µM concentrations of HAVDI-PA, SCRAM-PA, and K-PA were prepared to enable gel formation, facilitating a 3D structure. Initially, a spheroid of HAVDI-PA solution was placed on the plate, into which cells were injected (1 × 10⁴ cells/gel). This was followed by the addition of K-PA on top of the cells to create a sandwich model, encapsulating the cells between two peptide layers (as illustrated in Figure S3). The plate was then incubated at 37 °C and 5% CO2 for 15 min, allowing the peptides to solidify into gels. Afterward, culture medium was added, and the plate was returned to the incubator. Parallel preparations of SCRAM and K-PA gels were conducted for control experiments. Before treating the cells with LPS, they incubated for 24 h.

For micro-mass culture generation, cells were deposited directly onto the 24-well plate as a droplet to form a spheroidal structure (as shown in Figure S3). To ensure cell aggregation and prevent dispersion, the droplet was incubated for 15 min prior to medium addition. All experimental groups were treated with 0.1 µg/mL LPS. For qRT-PCR analysis the LPS incubation period was 6 h.

MTT assay

The proliferation effects of HAVDI and SCRAM peptide sequences on ADMSCs in 2D cell culture environment were assessed using the 3-[4,5-dimethylthiazol-2-yl]−2,5-diphenyl tetrazolium bromide (MTT) assay in conjunction with a spectrophotometer. For each experimental group, 2,000 cells were seeded per well in 96-well plates provided by Nest Scientific USA Inc. The plates were then incubated at 37 °C with 5% CO₂. After 24 h of incubation, 10 µL of MTT solution with a final concentration of 0.5 mg/mL (Biotium, California, USA) was added to each well at pre-determined time points.

To facilitate the MTT assay, the plates were incubated in the dark for 4 h at 37 °C within the incubator. Following this incubation period, the formazan crystals that had formed were dissolved in each well using 100 µL of Dimethyl sulfoxide (DMSO). The DMSO was pipetted and mixed thoroughly with the formazan crystals, and the plate was left for an additional 15 min to ensure complete dissolution. The absorbance of the dissolved formazan solution was then measured at a wavelength of 570 nm using a microplate reader (BioTek Synergy H1, BioTek Instruments, Winooski, VT, USA). This measurement allowed for the evaluation of the impact of the SCRAM and HAVDI peptide sequences on the proliferative ability of ADMSCs as well as the quantification of the cell proliferation.

Proliferation and cytocompatibility assay

To assess live cell proliferation, MSC cells were cultured in 24-well culture plates. Prior to seeding the cells, the wells were coated separately with HAVDI and SCRAM peptide sequences in both 2D and 3D culture environments. The cells were then washed carefully twice with PBS (phosphate-buffered saline) to remove any residual debris. Subsequently, the cells were incubated with a solution of 2 µM calcein-acetoxymethyl ester (obtained from Santa Cruz Biotechnology, Inc., Dallas, TX, USA) at 37 °C for 30 min in the dark within the incubator. To visualize the live cells, a Leica DM IL fluorescent inverted microscope (Leica, Wetzlar, Germany) was used. The live cells were observed under the microscope, and photographs were taken from four randomly selected areas. In the control group, the number of cells was counted using the NIH Image J program, and this count was considered as 100%. The same counting method was applied to the other treatment groups, and the cell numbers were customized accordingly. This allowed for a comparative analysis of cell proliferation among the different treatment groups.

To assess cytocompatibility, cells were seeded in 24-well plates at 1–2 × 10⁴ cells cm⁻² and cultured for 24 h prior to staining for TCP (2D). Peptide amphiphile (PA) precursor solutions (HAVDI-PA or SCRAM-PA) were prepared as described in the Hydrogel Preparation section and mixed with cell suspension to a final density of 1–2 × 10⁵ cells mL⁻¹; 50–100 µL gel drops were cast per well of 24-well plates and allowed to gel per protocol (typically ≤ 15 min) for 3D hydrogels,. All staining was performed 24–72 h post-seeding as indicated in figure legends.

After staining, samples were rinsed 1× with PBS (2–3 min) to reduce background and imaged immediately in PBS. Images were acquired on an epifluorescence microscope (e.g., Nikon/Zeiss) using a 10× or 20× objective. FITC channel for Calcein-AM (~ 488/520 nm) and TRITC/TexasRed for EthD-1 (~ 528/617 nm). Exposure/gain and camera settings were held constant across all conditions within an experiment. For 3D gels, focus was set to the mid-plane of the construct; where indicated, a short z-stack (± 20–30 μm, 3–5 planes) was captured and the central plane exported for analysis to avoid double-counting.

Phalloidin–DAPI staining: ADMSCs cultured on tissue culture plastic (TCP, 2D) or encapsulated in 3D HAVDI-PA or 3D SCRAM-PA hydrogels were fixed at room temperature for 15 min in 4% paraformaldehyde in PBS and rinsed three times with PBS. Samples were permeabilized with 0.1% Triton X-100 in PBS for 5 min and blocked with 1% BSA in PBS for 20–30 min. F-actin was labeled with fluorescent phalloidin (iFluor/Alexa Fluor 488 or 568; 1:100–1:200 in 1% BSA/PBS) for 20–30 min at room temperature protected from light, followed by nuclear counterstaining with DAPI (0.5–1.0 µg/mL in PBS) for 5 min and three PBS washes. No optical clearing was used. Imaging was performed on an epifluorescence or confocal microscope with identical exposure and detector settings across conditions.

Gene expression analysis

Total RNA extraction was conducted utilizing Trizol (SERVA, Germany) following the manufacturer’s specified protocol. For gene expression analysis, 60 ng of RNA was employed for reverse transcription PCR (RT-PCR), utilizing the Lightcycler^® 96 system (Roche Diagnostic Systems, Indianapolis, IN). Quantitative RT-PCR (qRT-PCR) analysis was performed using the A.B.T.™ 2X qPCR SYBR-Green Master Mix (ATLAS Biotechnology, Turkey). The primers targeting various genes, including Glyceraldehyde 3-phosphate dehydrogenase (GAPDH), Interleukin 1-alpha (IL-1-α), Interleukin 1-beta (IL-1β), Interleukin 8 (IL-8), Interferon-γ (IFN-γ), Granulocyte macrophage colony-stimulating factor (GM-CSF), Monocyte chemotactic activating factor (MCAF), tumor necrosis factor alpha (TNF-α), matrix metallopeptidase 2 (MMP-2), Plasminogen activator urokinase (PLAU), vascular endothelial growth factor receptor 2 (VEGFR2), Catenin delta-1 (p120) and Beta Catenin (β -Catenin) were utilized at a final concentration of 0.1 µM. The primer sequences used were as follows in Supplementary Table S2.

Surface marker analysis

To explore the cell surface markers of ADMSCs after their seeding on wells coated with either HAVDI-PA or SCRAM-PA peptide amphiphiles, an experiment was conducted wherein a total of 40,000 cells were introduced into each well of a 24-well cell culture plate. After their adherence to the plastic surface, the cells underwent a 48-hour incubation period. Following this, cell harvesting was accomplished by gently detaching them from the surface using a cell scraper and then transferring them to plastic tubes. The tubes were subsequently subjected to centrifugation at 1,000 x g for 5 min, with subsequent removal of the supernatant. The cells were subjected to staining using the Mesenchymal Stem Cell Marker Verification Kit (R&D Systems, Catalog #FMC020). In brief, the cell samples were subjected to a wash in a staining buffer. For each individual sample, 10 µL of positive MSC surface antibody markers, including anti-CD90, anti-CD105, and anti-CD73, were applied, alongside 10 µL of the Negative Marker Cocktail containing anti-CD45, anti-CD34, anti-CD11b, anti HLA-DR, and anti-CD79 antibodies, in addition to 10 µL of isotype control antibody. This mixture was allowed to incubate at room temperature in the dark for 30–45 min. Following incubation, cells were rinsed with the staining buffer and centrifuged to eliminate any remaining antibodies. For the subsequent flow cytometric analysis, the final cell pellet was resuspended in 100 µL of staining buffer, with the removal of cell debris being performed by the selection of intact cells based on forward scatter vs. side scatter dot plots. Fluorescence data were then represented as histograms utilizing Cell quest Pro software (BD, Biosciences, SAN JOSE, CA). The percentages of cells expressing CD90, CD73, and CD105 were determined for each experimental sample, drawing comparisons with isotype-matched control samples.

Statistical analysis

Statistical analyses were conducted using GraphPad Prism 9.0. One-way ANOVA followed by Dunnett post hoc test was used for multi-group comparisons. Data are reported as mean ± standard deviation (SD). A p-value < 0.05 was considered statistically significant. Experiments were performed in triplicate (n = 3) unless otherwise noted.

Results

Characterization of self-assembled peptide hydrogels

The molecular structures of K-PA, HAVDI-PA, and SCRAM-PA peptides are depicted, revealing distinct sequence variations. K-PA serves as the foundational sequence, while HAVDI/K-PA and SCRAM/K-PA represent modified variants with potential implications for their physical properties and biological function. The red highlighted segment in HAVDI-PA indicates the location of sequence variation, which may contribute to its unique structural attributes and gelation capabilities (Figure S2a). Similarly, the blue highlighted segments in SCRAM-PA suggest alterations that could affect its conformational behavior and self-assembly into hydrogel networks (Figure S2a).

SEM images of HAVDI/K-PA and SCRAM/K-PA hydrogels showcase the microstructural differences between the two gels (Figure S2a). The mechanical properties of hydrogels formed from the self-assembly of the designed peptide HAVDI/K-PA and its scrambled control sequence SCRAM/K-PA were investigated using oscillatory rheology. Rheological characterization highlighted the decisive role of peptide sequence in shaping the mechanical behavior of the resulting gels. First, the structural integrity and resistance to deformation of the hydrogels were evaluated by an amplitude sweep experiment (Figure S2d). In the linear viscoelastic region (LVR) at low strain, both samples exhibited solid-like behavior, with the storage modulus (G′) exceeding the loss modulus (G″), confirming gel network formation for both peptides. However, HAVDI/K-PA consistently showed a higher G′ than the SCRAM/K-PA control and maintained this elastic dominance over a broader stress window, whereas SCRAM/K-PA departed from the LVR earlier, with a more pronounced decrease in G′ and an earlier G′/G″ crossover. These observations demonstrate that the HAVDI/K-PA gel forms a mechanically stronger network that better resists deformation. To further probe the time-dependent viscoelastic behavior, a frequency sweep was performed within the LVR (Figure S2d). For both samples, G′ remained higher than G″ across the entire tested frequency range (1–100 rad/s), indicating stable gel-like structures. Consistent with the amplitude sweep, the G′ and G″ values of HAVDI/K-PA were systematically higher than those of SCRAM/K-PA at all frequencies, showing that HAVDI/K-PA forms a mechanically superior network (Figure S2d).

From this point onward in the manuscript, SCRAM/K-PA will be referred to as “SCRAM-PA” and HAVDI/K-PA as “HAVDI-PA”.

HAVDI-PA does not impair ADMSC viability and supports stable proliferation in 2D culture

The cytocompatibility of the HAVDI-PA nanofiber system was evaluated by culturing ADMSCs on HAVDI-PA– and SCRAM-PA–coated surfaces, with unmodified TCP as control. MTT analysis demonstrated that at early time points (24 h and 48 h), metabolic activity remained comparable across all groups, with no meaningful reduction associated with peptide-coated surfaces (Fig. 1a). By 72 h, a modest decline in MTT signal was observed in both HAVDI-PA and SCRAM-PA groups relative to their earlier time points. Importantly, the decrease was similar in magnitude for both peptide-coated substrates, suggesting that the effect is not sequence-specific, but rather reflects a shared peptide material-related phenomenon. A plausible explanation for this pattern is the progressive cellular adaptation to the peptide-coated surface, where initial attachment and spreading may alter mitochondrial activity per cell over time. Additionally, peptide-coated substrates can modulate cell–substrate interactions, influencing metabolic output independent of overall cell number. Thus, the reduction in 72 h MTT signal likely represents a physiological shift in metabolic status, rather than cytotoxicity or peptide-mediated inhibition (Fig. 1a).

Fig. 1.

Evaluation of ADMSC viability and proliferation on peptide-coated 2D surfaces. MTT assay showing cell metabolic activity after 24, 48, and 72 h of culture on uncoated tissue culture plastic (TCP), HAVDI-PA–coated, and SCRAM-PA–coated surfaces. Values are expressed as percentages normalized to the 24-hour TCP group, which was defined as 100%. (b) Quantitative analysis of relative cell proliferation across groups over time, expressed as a percentage normalized to the 24 h TCP control. (c) Representative fluorescence microscopy images of Calcein-AM–stained live ADMSCs at 24 h, showing cell distribution and morphology under each condition. Scale bar = 100 μm. Data are presented as mean ± standard deviation (SD) from three independent experiments (n = 3). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Proliferation assessed by Calcein AM fluorescence intensity showed relatively similar values across all groups at 24, 48, and 72 h, without statistically meaningful differences (Fig. 1). The comparable levels of fluorescence indicate that HAVDI-PA does not impede ADMSC growth and supports a proliferation rate consistent with TCP and SCRAM controls (Fig. 1b). Live-cell Calcein AM imaging further showed uniformly intense intracellular fluorescence and characteristic spindle-shaped ADMSC morphology in all groups, with no indications of membrane compromise or nuclear condensation (Fig. 1). Complementary Live/Dead staining using Calcein AM and EthD-1 detected only rare EthD-1–positive cells across all conditions, and Phalloidin/DAPI imaging demonstrated preserved nuclear morphology together with an organized F-actin cytoskeleton, without evidence of cytoskeletal collapse, apoptotic blebbing, or nuclear fragmentation (Figure S4).

HAVDI-PA preserves stemness and activates β-catenin signaling in 3D-cultured ADMSCs

To determine whether HAVDI-PA influences the maintenance of mesenchymal stem cell identity in different culture formats, we quantified the expression of canonical MSC surface markers (CD90, CD73, CD105) by flow cytometry (Fig. 2a). Across all conditions—TCP, micromass, 2D HAVDI, 3D HAVDI, 2D SCRAM, and 3D SCRAM—ADMSCs remained highly positive (> 90%) for all three markers, indicating that the MSC phenotype was globally preserved. Only small but statistically significant differences were detected. For CD73 and CD105, peptide-based conditions (both HAVDI- and SCRAM-containing, in 2D and 3D) showed slightly higher percentages of positive cells than TCP, whereas micromass values were generally comparable to the other groups. For CD90, the differences among conditions were minimal and all values clustered within the high-positivity range. Overall, these data indicate that neither 3D encapsulation nor peptide presentation leads to a loss of MSC surface phenotype; if anything, peptide-modified environments modestly stabilize the expression of certain markers relative to uncoated plastic (Fig. 2a).

Fig. 2.

Flow cytometric and gene expression analysis of ADMSCs cultured under different 2D and 3D conditions. (a) Quantification of mesenchymal stem cell surface markers CD90, CD73, and CD105 by flow cytometry in ADMSCs cultured on uncoated TCP, micromass aggregates, 2D peptide-coated surfaces (HAVDI-PA or SCRAM-PA), and 3D peptide hydrogels (HAVDI-PA or SCRAM-PA). (b) Relative mRNA expression of N-cadherin–associated signaling molecules β-catenin and p120-catenin, assessed by qRT-PCR in ADMSCs cultured in 3D HAVDI-PA, 3D SCRAM-PA, and micromass conditions. Gene expression levels were normalized to GAPDH. (c) Schematic illustration of HAVDI-PA hydrogel interaction with the N-cadherin receptor and associated intracellular signaling components (p120 and β-catenin). Data are presented as mean ± standard deviation (SD) from three independent experiments (n = 3). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Importantly, gene expression analysis demonstrated a substantial upregulation of p120-catenin (6.5-fold) and β-catenin (9.8-fold) in HAVDI-PA–treated 3D cultures compared to control micromass conditions, while SCRAM-PA had no significant effect (Fig. 2b–c). These findings confirm that HAVDI-PA not only preserves MSC identity but also activates intracellular signaling cascades associated with N-cadherin engagement, including β-catenin–mediated transcription, which is known to support self-renewal and stemness maintenance.

3D HAVDI-PA induces high basal cytokine transcription and a refractory response to LPS

To evaluate whether HAVDI-PA hydrogels modulate the immunological behavior of ADMSCs, we performed qPCR analysis, which revealed distinct inflammatory transcriptional signatures across the different culture conditions. In the TCP group, LPS stimulation induced clear upregulation of IL-1α, IL-1β, IL-8, TNF-α, and GM-CSF compared with LPS-negative controls, indicating that cells on standard 2D plastic mount a conventional early inflammatory response (Fig. 3). In the peptide-coated 2D conditions (2D HAVDI and 2D SCRAM), basal cytokine expression remained uniformly low, like TCP, and LPS elicited only modest increases that did not exceed the response observed on uncoated plastic. Micromass cultures showed a distinct pattern reflective of densely aggregated cell networks. LPS triggered pronounced induction of IFN-γ and IL-8, confirming that these compact 3D aggregates are responsive to endotoxin, although the magnitude of regulation for other cytokines remained variable relative to TCP (Fig. 3). The most striking phenotype emerged in the 3D HAVDI group. Under LPS-negative conditions, 3D HAVDI cultures exhibited markedly elevated basal expression of all measured cytokines—including IL-1α, IL-1β, IL-8, IFN-γ, TNF-α, and GM-CSF—surpassing all other conditions. This indicates that HAVDI-mediated cell–cell interactions induce a strongly primed transcriptional state even in the absence of stimulation. Upon LPS exposure, however, these transcripts did not increase further; instead, each cytokine displayed a substantial decrease relative to its elevated basal level. This sharp downward shift reflects a tolerance-like or refractory transcriptional response, in which pre-activated cells actively reduce inflammatory gene expression following a secondary stimulus. In contrast, the 3D SCRAM condition exhibited intermediate basal transcription and a moderate LPS response, confirming that the unique priming-plus-suppression phenotype observed in 3D HAVDI is driven by the HAVDI motif rather than 3D architecture alone. Collectively, these data indicate that while TCP and Micromass maintain typical LPS responsiveness, and 2D peptide coatings remain transcriptionally quiescent, 3D HAVDI uniquely induces high basal activation followed by a paradoxical suppression upon LPS challenge, consistent with an N-cadherin–dependent priming-and-tolerance mechanism.

Fig. 3.

Expression of inflammatory cytokine genes in ADMSCs cultured under 2D and 3D conditions with or without LPS stimulation. Relative mRNA levels of IL-1α, IL-1β, IFN-γ, IL-8, TNF-α, and GM-CSF were quantified by qRT-PCR in ADMSCs grown on uncoated tissue culture plastic (TCP), 2D peptide-coated surfaces (HAVDI-PA or SCRAM-PA), micromass aggregates, and 3D peptide hydrogels (HAVDI-PA or SCRAM-PA), under unstimulated (− LPS) and LPS-stimulated (+ LPS, 0.1 µg/mL for 18 h) conditions. Statistical comparisons for each condition were performed against the corresponding TCP group within the same LPS condition (− LPS or + LPS). Data are presented as mean ± standard deviation (SD) from three independent experiments (n = 3). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

HAVDI-PA upregulates regenerative gene expression in 3D-cultured ADMSCs

To evaluate whether HAVDI-PA hydrogels enhance the regenerative potential of ADMSCs, we assessed the expression of key matrix-remodeling and angiogenic genes—PLAU, MMP2, and VEGFR2—under both 2D and 3D culture conditions. While HAVDI-PA coating in 2D induced only minimal changes, 3D encapsulation within HAVDI-PA hydrogels resulted in a marked upregulation of all three genes compared to both micromass and SCRAM-PA controls (Fig. 4a–b). Notably, expression of MMP2 and PLAU, which are essential for extracellular matrix remodeling and cell migration, was significantly elevated. In parallel, VEGFR2 expression, a key marker of pro-angiogenic potential, was also increased.

Fig. 4.

Regenerative Gene Expression and Matrix Cues in 3D HAVDI-PA Hydrogels. (a) Relative mRNA expression of MMP2, VEGFR2, and PLAU measured by qRT-PCR in ADMSCs cultured on 2D tissue culture plastic (TCP), 2D peptide-coated substrates (HAVDI-PA or SCRAM-PA), and within 3D hydrogel scaffolds composed of the same peptide formulations (3D HAVDI-PA or 3D SCRAM-PA). Gene expression was normalized to GAPDH. Data are shown as mean ± standard deviation (SD) from three independent experiments (n = 3). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (b) Schematic representation illustrating how matrix stiffness modulates cellular behavior: softer hydrogels facilitate more dynamic and spread morphologies, while stiffer matrices restrict cell spreading and induce compact organization, influencing regenerative signaling outcomes.

These findings suggest that HAVDI-PA scaffolds not only preserve the stemness and immunomodulatory functions of ADMSCs but also activate transcriptional programs associated with tissue regeneration. This effect appears to be enhanced in the 3D environment, likely due to the combined influence of N-cadherin mimetic signaling and the physical properties of the hydrogel matrix (e.g., stiffness, cellular confinement) (Fig. 4c).

Discussion

This study demonstrates that the N-cadherin–mimetic HAVDI/K-PA hydrogel provides a three-dimensional microenvironment that preserves, rather than exaggerates, the mesenchymal phenotype of ADMSCs while selectively reprogramming their transcriptional and immunoregulatory profiles. Across all culture conditions, cells remained positive for canonical MSC markers (CD73, CD90, CD105), indicating overall maintenance of MSC identity. Nevertheless, 3D HAVDI/K-PA hydrogels uniquely promoted a coordinated upregulation of p120-catenin, β-catenin and regenerative genes such as MMP2, PLAU and VEGFR2, pointing to a shift toward a pro-regenerative transcriptional state that is not observed on TCP, micromass, or scrambled peptide controls.

The concomitant increase in β-catenin and p120-catenin in HAVDI/K-PA–treated cells is consistent with active engagement of N-cadherin and assembly of the canonical cadherin–catenin adhesion complex. p120-catenin binds the juxtamembrane region of N-cadherin and prevents its endocytosis, thereby stabilizing junctions at the cell surface, whereas β-catenin links the cadherin tail to the actin cytoskeleton and is retained at the membrane rather than accumulating in the nucleus^{26,28–30,30}. In the context of a mechanically supportive HAVDI/K-PA matrix, this organization likely promotes stable adherens junctions that physically anchor the cells and simultaneously limit Wnt-driven transcriptional programs that drive premature differentiation. Thus, N-cadherin–mimetic signaling in 3D appears to preserve stemness via a coupled biochemical–mechanical mechanism.

HAVDI/K-PA also imposed a distinctive immunological phenotype that was not reproduced by any other condition. ADMSCs encapsulated in 3D HAVDI/K-PA showed elevated basal transcription of IL-1α, IL-1β, IFN-γ, IL-8, TNF-α and GM-CSF, consistent with N-cadherin–mediated priming in the absence of exogenous stimulation. Remarkably, upon LPS challenge these transcripts declined rather than exhibiting the classical induction observed in TCP and micromass cultures, indicating a tolerance-like or refractory response. Neither 2D peptide coatings nor 3D SCRAM-PA hydrogels elicited this “priming-and-down-tuning” pattern, supporting the conclusion that it is specific to the HAVDI motif presented within a 3D architecture. Collectively, these findings position HAVDI/K-PA as an immuno-instructive niche that prepares MSCs to encounter inflammatory cues while preventing excessive activation, in agreement with reports that 3D microenvironments can pre-activate and tune MSC immunomodulation^31,32.

Comparison of 2D and 3D formats further underscores that ligand presentation alone is insufficient to unlock the full potential of N-cadherin mimetic signaling. Whereas 3D HAVDI/K-PA robustly enhanced immunomodulatory priming and regenerative gene expression, the 2D HAVDI coating had neutral or modest effects, suggesting that a flat, rigid substrate cannot reproduce the spatial clustering and force-dependent engagement of N-cadherin receptors that occur in compliant matrices^35,37,38,38. Even SCRAM-PA hydrogels induced moderate transcriptional changes relative to 2D cultures, indicating that dimensionality and matrix mechanics contribute to regenerative programming independently of sequence-specific ligation. Within this framework, the superior performance of HAVDI/K-PA likely reflects the synergy between N-cadherin–mimetic biochemical cues and a 3D mechanical context that together approximate key features of the native stem cell niche.

Conclusion

In summary, this study shows that an N-cadherin–mimetic peptide amphiphile hydrogel (HAVDI/K-PA) establishes a three-dimensional microenvironment that stabilizes the mesenchymal phenotype of ADMSCs while actively reprogramming their mechanical, transcriptional and immunological states. HAVDI/K-PA forms a mechanically robust, elastic gel in which N-cadherin engagement is associated with assembly of the cadherin–catenin adhesion complex, maintenance of stemness markers and upregulation of β-catenin/p120-catenin together with key regenerative genes involved in matrix remodeling and neovascularization. At the same time, 3D HAVDI/K-PA elicits a distinctive “primed yet tolerant” inflammatory profile, characterized by elevated basal cytokine expression and LPS-induced down-tuning, positioning this material as an immuno-instructive niche rather than a passive scaffold. The muted responses observed on 2D HAVDI coatings and scrambled 3D controls underscore that N-cadherin–mimetic signaling requires an appropriate three-dimensional mechanical context to fully manifest. Collectively, these findings identify HAVDI/K-PA hydrogels as a rational platform for preconditioning MSCs toward pro-regenerative and controlled immunomodulatory phenotypes, with potential applications in tissue repair, implant integration and inflammatory disease, and motivate future in vivo studies to define dose, timing and target indications.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Seher Yaylacı conceived and designed the study, supervised all experiments, and oversaw data interpretation and manuscript preparation. Yelda Yüregir, Demet Kaçaroğlu and Alper M. Ulaşlı performed the experiments, collected and analyzed the data, and contributed to writing and editing the manuscript. Melek Parlak Khalily and Batuhan Baytekin synthesized and provided the peptide amphiphile materials used in this study.

Data availability

The datasets used during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics committee approval

An ethical approval for the study is not required. In this study, commercial cell was used.