3D coaxial bioprinting of RADA16-I self-assembling peptide hydrogel

Abstract

Peptides that self-assemble into hydrogels provide a dynamic microenvironment for various cell types. Combining top-down extrusion 3D bioprinting with bottom-up self-assembly of peptide hydrogels offers an innovative approach to biofabrication. However, modest mechanical properties of peptide hydrogels pose challenges for extrusion 3D bioprinting. This study introduces RADA16-I peptide hydrogels for bioprinting by leveraging the potential of coaxial extrusion to print mechanically soft hydrogels. A coaxial 3D bioprinter was employed to co-extrude a RADA16-I peptide core supplemented with methylcellulose (MC) and sucrose, surrounded by an MC-alginate composite hydrogel shell. The phosphate-buffered MC-alginate shell provides stability and initiates the RADA16-I hydrogel self-assembly post-extrusion. Rheological characterization confirmed the increase in viscosity of the RADA16-I core solution without compromising self-assembly (G′ ≈ 100 Pa). Core extrusion ratio was set to 20% to balance filament stability and soft-core content. Printed scaffolds maintained excellent shape fidelity and structural integrity over a 21-day culture period, with gradual MC release (≈90%) creating an open-porous shell structure. Mesenchymal stem cells (MSCs) encapsulated in the RADA-MC core hydrogel tended to aggregate, forming a dense collagen network with calcium phosphate deposition. Bioprinted cell-laden scaffolds displayed a homogeneous distribution of viable cells (>90%). In conclusion, this approach successfully introduced self-assembling peptide hydrogels to bioprinting technology, offering a promising strategy for biofabrication.

Graphical abstract

Highlights

• •Methylcellulose improved printability without impairing peptide self-assembly.
• •MC-alginate shell buffered pH and stabilized printed RADA16-I hydrogel core.
• •Bioprinted scaffolds showed high MSC viability.
• •Gradual MC release formed porous scaffolds supporting long-term cell culture.
• •RADA-MC hydrogel supported osteogenic differentiation.

1. Introduction

Tissue engineering and regenerative medicine (TERM) aim to restore or replace damaged tissues by recreating the spatial and temporal organization of cells and extracellular matrix (ECM) [1]. Biofabrication technologies, particularly three-dimensional (3D) bioprinting, enable the controlled deposition of cells and biomaterials, facilitating the construction of spatially organized and structurally complex tissue analogues [2,3]. Among these techniques, extrusion-based bioprinting remains widely used due to its cost-effectiveness, compatibility with a broad range of materials, and ability to print multiple materials and cell types simultaneously [4].

Significant progress has been made in engineering biomaterials with reproducible chemical architectures and defined bioactivity, both from synthetic and biological origins [5,6]. However, materials optimized for 3D cell culture often need to be modified to meet the strict rheological requirements imposed by extrusion bioprinting [7,8]. Many biologically favorable hydrogels exhibit rheological properties outside the traditional biofabrication window, requiring additional strategies to support filament formation and shape retention during printing. To address this limitation, several approaches have emerged to push the rheological and mechanical boundaries of soft hydrogels [8]. These include in situ crosslinking during deposition [9,10], extrusion into support baths [11,12], and coaxial bioprinting [13,14], where a low-viscosity core is extruded within a load-bearing hydrogel shell. Such strategies enable the printing of otherwise unprintable, low-viscosity bioinks while maintaining high post-printing cell viability and enhanced biological performance.

Self-assembling peptide (SAP) hydrogels represent a particularly promising class of supramolecular biomaterials, offering access to chemical and sequence spaces that extend well beyond those routinely exploited by biology [15]. Through non-covalent interactions, synthetic biomimetic peptides spontaneously organize into nanofibrous networks with tunable biochemical functionality, enabling structural organization on the molecular scale [16]. These building blocks can be synthesized and chemically modified with relative ease, and they assemble with high efficiency into biocompatible materials with tunable biodegradability. Among them, RADA16-I is one of the most extensively studied SAPs and has been applied to fabricate 3D scaffolds for homeostasis, wound healing, and TERM, due to its shear-thinning properties and ability to avoid potential immune responses [[17], [18], [19]]. 3D cell cultures have been performed using hepatocytes [20], chondrocytes [21], neural [22] and mesenchymal stem cells (MSC) [[23], [24], [25]]. Integrating self-assembly with biofabrication offers exceptional potential for producing highly ordered structures, however, SAP hydrogels exhibit inherently low pre-gelation viscosity and modest mechanical strength [26]. These characteristics limit their processability in conventional extrusion systems, which require sufficient viscosity for precise filament deposition and an adequate yield stress to prevent cell sedimentation prior to printing.

Several strategies have been explored to improve the printability of SAP bioinks. For example, blending SAPs with shear-thinning polymers such as methylcellulose (MC) enhances printability while retaining self-assembly capacity [27]. Nevertheless, the peptide concentrations required were approximately 20-fold higher than the levels typically used for cell culture [25,[28], [29], [30]]. Such elevated concentrations can compromise cell viability due to the acidic pre-assembly pH and restrict cell movement and activity within the resulting densely packed fibrous network. Alternatively, low-viscosity SAP solutions have been printed using coaxial extrusion, in which the fast-gelling SAP self-assembles in situ at the nozzle orifice when in contact with physiological buffer solutions [[31], [32], [33]]. This approach has enabled deposition of nanofibrous networks into large, high-fidelity constructs for neural, intestinal, and chondrogenic applications. However, these methods still depend on high concentrations of rapidly gelling components, prioritizing printability over cellular functionality.

Employing a load-bearing shell in 3D bioprinting reduces the mechanical and rheological demands on the core bioink, enabling the extrusion of low-viscosity, low-strength materials being less dependent on gelation kinetics. In coaxial extrusion, a mechanically supportive shell stabilizes the deposited filament, allowing immediate self-assembly post-extrusion and preserving filament fidelity. This approach combines top-down spatial control with the bottom-up molecular assembly of peptide hydrogels, overcoming the mechanical limitations that previously restricted their use in extrusion bioprinting.

Here, we present a biofabrication strategy that couples coaxial extrusion bioprinting with the bottom-up self-assembly of the soft RADA16-I peptide hydrogel. A printable RADA16-I solution supplemented with methylcellulose and sucrose was extruded as the core, while a phosphate-buffered methylcellulose–alginate blend formed a mechanically supportive outer shell that simultaneously triggered peptide self-assembly upon contact. This study represents, to our knowledge, the first coaxial extrusion strategy explicitly designed to use a mechanically supportive shell to enable bioprinting of inherently soft SAP hydrogels such as RADA16-I. Screening experiments identified core–shell extrusion ratios that maximized the proportion of soft, self-assembling core material while preserving strand stability. Using the optimized parameters, the printed constructs exhibited excellent shape fidelity, and post-printing ionic crosslinking of the shell ensured long-term structural stability. Encapsulated mesenchymal stem cells (MSCs) spread, interconnected, and remodeled the RADA–MC core, and under osteogenic conditions deposited a bone-like extracellular matrix. Overall, this approach overcomes the inherent mechanical limitations of self-assembling peptide hydrogels and expands their applicability in extrusion bioprinting for the fabrication of bioinspired tissue constructs.

2. Materials and methods

2.1. Materials

Methylcellulose (4000 cP), phosphate-buffered saline (PBS) tablets, sucrose, HEPES, calcium chloride, Minimum Essential Medium Eagle (αMEM), research-grade fetal bovine serum (FBS), penicillin-streptomycin cocktail (P/S), trypsin-EDTA, paraformaldehyde (PFA), Xylene, Eukitt mounting media, Triton X-100, alkaline phosphatase yellow liquid substrate system for ELISA, p-nitrophenol solutions, Alizarin Red S, Phalloidin-Atto 488, DAPI ready-made solution, and Thioflavin-T were all purchased from Sigma- Aldrich (Saint Louis, MO, USA). Sodium hydroxide was acquired from Merck KGaA (Darmstadt, Ger-many). DPBS solution, Propidium Iodide, calcein AM, GlutaMAX™, CellPath OCT Embedding Matrix, sterile syringe filters (polyethersulfone, 0.2 μm, 25 mm), and InvitrogenTM Rat OPN ELISA Kit were purchased from Thermo Fisher Scientific (Waltham, MA, USA). Ethanol absolute pure for molecular biology and sodium alginate were purchased from AppliChem (Darmstadt, Germany). RADA16-I peptide with >95% purity was purchased from Abyntek Biopharma (Bizkaia, Spain). Mykoval™ staining solution was purchased from Hund (Wetzlar, Germany), and Zinc chloride solution iodinated was purchased from VWR (Darmstadt, Germany). Picro-Sirius red staining solution was purchased from Ab-cam (Cambridge, United Kingdom).

2.2. Preparation of shell and core hydrogel

2.2.1. MC-alginate shell hydrogel

The shell bioink was prepared by dissolving 1% (w/v) sodium alginate powder (w/v) in a solution containing 3x PBS and 10% (w/v) sucrose. The ready sodium alginate solution and MC powder were steam-sterilized at 121 °C for 15 min. Aseptically, 7% (w/v) MC was dissolved in pre-heated sodium alginate solution (70 °C) using a SpeedMixer DAC 150.1 FVZ (FlackTek, Inc.) at 700 rpm for 5 min. The shell bioink formulation was stored at 4 °C overnight for complete hydration.

2.2.2. RADA16-I core hydrogel

The core bioink was prepared by dissolving RADA16-I peptide powder in Milli-Q water. Separately, steam-sterilized MC powder was dissolved in a filter-sterilized, pre-heated solution at 70 °C containing 30% (w/v) sucrose. Under aseptic conditions, the solution was mixed using the SpeedMixer at 700 rpm for 5 min and stored at 4 °C overnight. In the final step, 0.5% (w/v) MC-solution was mixed with 1% (w/v) RADA16-I peptide in solution in a ratio of 0.85:0.15 (MC:RADA), referred to as RADA-MC. The control was prepared following previous reports [29]. A 10% (w/v) sucrose solution was mixed with a 1% (w/v) RADA16- I peptide solution in a ratio of 0.85 to 0.15, referred to as RADA.

2.3. Rheological tests

Rheometric experiments were performed using a Discovery Hybrid Rheometer-2 (TA Instruments). All rheometric experiments were performed using a 20 mm stainless steel upper plate at 25 °C. For rheological measurements of bioink formulations, the measuring gap was set to 500 μm with a trim gap offset of 50 μm. Rotational tests were performed to assess the flow behavior of core and shell bioink formulations, increasing the shear rate from 0.01 to 100 s-1. To evaluate the crosslinking kinetics of MC-alginate hydrogel, oscillatory measurements were performed at a constant angular frequency of 10 rad/s and 1% strain. Samples were equilibrated for 60 s before adding a crosslinking solution containing 150 mM CaCl₂ and 50 mM HEPES. Similarly, the self-assembly of RADA16-I hydrogel was evaluated by oscillatory measurements at a constant angular frequency of 10 rad/s and 0.1% strain. Samples were equilibrated for 60 s before adding 3x PBS buffer solution. Viscoelastic properties of RADA16-I hydrogel with and without supplementation of MC were compared by performing oscillatory frequency sweeps ranging from 1 to 125 rad/s.

2.4. Core-shell 3D printing

A custom Python script was used to generate G-code for extruding a zig-zag pattern with varying core-to-shell extrusion rates [13]. The printer was equipped with two syringe extruders for extruding two bioinks simultaneously. The coaxial nozzle with an inner diameter of 250 μm, and an outer diameter of 840 μm was mounted to the primary syringe extruder which directly fed the shell bioink. The second syringe extruder fed the core bioink connected to the coaxial nozzle by a flexible silicone tube (inner diameter = 0.6 mm). Parameters for printing were set to 4 mm/s and 2 μl/s for printing speed and extrusion rate, respectively. The print bed temperature was set to 37 °C for all experiments.

2.4.1. Screening effect of core-shell extrusion ratio

The screening of different core-to-shell extrusion ratios was performed to identify printing conditions that balance mechanical support and soft-core content. Increasing the proportion of core material was expected to enhance the biological relevance of the construct by providing a larger volume of the peptide-based hydrogel, whereas higher shell fractions were anticipated to improve strand stability and structural integrity during layer stacking. G-code to extrude a zig-zag pattern with varying core-to-shell extrusion rates was generated (Table 1). The first line was printed by extruding solely through the shell nozzle and subsequent lines were printed by stepwise reducing shell extrusion and increasing core extrusion (Fig. 1, A). Immediately after printing, images of the extruded filaments were acquired using a stereomicroscope (Axio Observer Z1, Zeiss). For image analysis, first, a region of interest (ROI) was selected (Fig. 1, B), followed by the detection of filament structures using the Segment Anything Model (Fig. 1, C) [34,35]. Finally, the width of the detected filaments was quantified and compared between the different extrusion ratios.

Table 1.

Zig-zag pattern with varying core-to-shell extrusion rates.

filament number	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
% core	0	4	7	10	13	16	19	22	25	28	31	34	37	40	43
% shell	100	96	93	90	87	84	81	78	75	72	69	66	63	60	57

Fig. 1.

Core-shell extrusion ratio screening outline. (A) Image of printed screening design. For demonstration purposes, dH₂O supplemented with food dye was used as core material (B) Stereomicroscope image of the core-shell extruded strand for image segmentation and analysis. (C) Image showing the segmented area used for further analysis in red. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

2.4.2. Thioflaving T staining of self-assembling peptide hydrogel

Single filaments extruded in a core-shell ratio of 20/80 were crosslinked in 150 mM CaCl2 50 mM HEPES buffer solution for 5 min. Filaments were washed twice in dH20 followed by incubation in solution containing 100 μM Thioflavin T for 2 h. Samples were washed two more times with dH20 and imaged with a Leica TCS SP8 CLSM using excitation/emission wavelengths of 488/495–560, respectively.

2.4.3. Structural integrity

8-layer cubical structures (diameter = 13 mm) were printed with varying core-shell extrusion rates to assess structural integrity, as previously described [36]. Photographic images were taken at a fixed distance and angle to measure the height of the built structures using a custom-made camera tripod. The structural integrity of the 3D printed hydrogel formulations was quantified by comparing the relative height between the bioink formulations using ImageJ software [37]. Larger woodpile 12-layered structures were printed with a core-shell extrusion ratio of 20/80, and a line spacing set to 2 mm.

2.5. Assessment of MC release

The MC content within core-shell printed and Ca²⁺-crosslinked scaffolds was visualized using a chlorine-zinc-iodine solution. Images were taken after 1, 7, 14, and 21 days of incubation under cell culture conditions. Mykoval™, a staining solution that reacts with chitin and cellulose-emitting fluorescence signals, was used to detect MC release into the supernatant. Printed scaffolds were incubated in 1.5 ml of culture medium at 37 °C in a humidified incubator with 5% (v/v) CO2. On defined time points (0, 1, 2, 5, 7, 9, 12, 16, 19, 21 days), 750 μl of supernatant was taken and replaced by 750 μl of fresh medium. For measurements, 180 μl of supernatant was mixed with 20 μl Mykoval™ solution in a black F-bottom polystyrene 96 well plate (VWR, Darmstadt, Germany). The samples were incubated for 5 min, and the fluorescence signal was measured using a microplate reader (Infinite M Nano, TECAN, Switzerland). The excitation and emission wavelengths used were 400 and 450 nm, respectively. Regression analysis to obtain calibrated MC concentration via detected fluorescence signal was performed via exponential decay model using Python (Python Software Foundation, version 3.9) with NumPy [38] and Matplotlib [39] libraries.

2.6. Biological testing

2.6.1. Mesenchymal stem cell isolation and culture

MSCs were isolated from bone marrow flushed from the femurs and tibias of five-week-old male Sprague- Dawley rats as previously described [40]. In short, rats were euthanized by CO₂ inhalation for the collection of tibias and femoral bones from the hind legs, cells were harvested, passed through a 40 μm cell strainer, and centrifuged for 3 min at 200 rcf. Cells were cultured at 37 °C and 5% CO₂ in culture medium consisting of α-MEM, 10% (v/v) FBS, 1% (v/v) GlutaMAX, and 1% (v/v) penicilin/streptomycin (10 000 U ml⁻¹ and 10 mg ml⁻¹, respectively). The medium was changed every 2-3 days and subconfluent cells were detached from the culture flask by adding trypsin-EDTA solution. Only cells from passages 2 to 5 were used for experiments.

2.6.2. Cell encapsulation and 3D bioprinting

MSCs were detached and resuspended in 10% (w/v) sucrose solution, before mixing with RADA16-I solution at a concentration of 3 × 10⁶ cells per ml. To form microtissues, 40 μl RADA16-I solution was self-assembled in 24-well plate inserts pre-equilibrated in culture medium followed by 3 washes in culture medium.

2.6.3. Cell morphology and hydrogel contraction

Cell morphology was evaluated on days 0, and 3 by staining actin filaments of encapsulated MSCs. Samples were washed twice in DPBS, followed by fixation in 4% paraformaldehyde for 10 min, and three additional washes with DPBS buffer. Then, cells were permeabilized by submerging samples in 0.1% Triton-X in DPBS with soft shaking for 30 min, followed by three additional washes in DPBS. Next, samples were incubated in a solution of phalloidin-Atto 488 (0.1 μg ml⁻¹; in PBS) for 30 min, and 10 min in a solution of DAPI (20 μg ml⁻¹; in PBS). After an additional three washes with DPBS, cells were imaged with a Leica TCS SP8 CLSM, with excitation/emission wavelengths set to 405/440-480 and 488/495-545 for DAPI and phalloidin-Atto 488 detection. Cell aggregation was assessed by acquiring Z-stack images of MSCs encapsulated in hydrogels. The tendency to form aggregates was evaluated by comparing the mean volume of phalloidin-positive stained areas using the 3D Object Counter plugin in ImageJ software. Images of RADA and RADA-MC hydrogels were acquired on days 0, 1, and 3 of culture. Macroscopic contraction was quantified by measuring the hydrogel area using ImageJ software.

2.6.4. Osteogenic differentiation

The potential of RADA-MC to promote osteogenic differentiation of MSCs was investigated. The hydrogel culture was compared to conventional culture on 2D polystyrene surfaces selecting equal number of cells among experimental groups (Table 2). All samples were cultured for 21 days either in medium with or without osteogenic supplementation (10⁻⁸ M dexamethasone, 50 g ml⁻¹ ascorbic acid, 10 mM β-glycerol phosphate). On days 1, 7, 14, and 21 samples were washed twice in DPBS and frozen in DPBS containing 0.1% Triton X-100. Thawed samples were homogenized by vigorously pipetting up and down. Cell proliferation was assessed on days 1, 7, 14, and 21 using Quant-iT PicoGreen dsDNA Assay Kit. Quant-iT™ PicoGreen™ dsDNA Reagent was incubated with cell lysate for 5 min to quantify presence of dsDNA by reading fluorescence/emission spectrum at 480/520 nm respectively, using a microplate spectrophotometer system (Infinite M Nano, TECAN). Differentiation was studied by quantifying alkaline phosphatase (ALP) activity at day 1, 7, 14 and 21 of culture; and osteopontin (OPN) secretion at day 14 and 21. For ALP analysis, cell lysates were incubated with p-nitrophenol phosphate (pNPP) for 60 min at RT in the dark. A calibration curve was prepared using p-nitrophenol in 0.1% Triton X-100 in PBS. ALP activity was calculated as pg of reaction product (p-nitrophenol) per min (over 60 min in total). OPN quantification was performed from cell lysates using Invitrogen Rat OPN ELISA Kit. OPN and ALP activity was normalized against the total cell number.

Table 2.

Summary of experimental groups.

Experimental group (Cells per well/scaffold)	Description	MSC density
TCPS	Culture on 2D tissue culture polystyrene surfaces	0.2 × 106 cm−2
(1.2 × 105) RADA16-I bioink	Cells encapsulated in RADA16-I/MC hydrogel	3 × 106 ml−1
(1.2 × 105)

2.6.5. Histological analysis

Following 21 days of culture in medium with osteogenic supplementation, RADA–MC hydrogels containing MSCs were washed twice in DPBS and fixed in 4% paraformaldehyde for 15 min at room temperature. Samples were rinsed with DPBS, embedded in OCT compound, and frozen at −80 °C until sectioning. Cryosections of 30 μm thickness were obtained using a cryotome and mounted on glass slides for subsequent histological analysis. Mineralized matrix deposition was assessed using Alizarin Red S (ARS) staining. Cryosections were equilibrated to RT, rinsed once in distilled water, and incubated in 40 mM ARS solution (pH 4.2) for 30 min. Excess dye was removed by five washes in distilled water. Collagen deposition was evaluated using Sirius Red staining. Cryosections were rinsed in distilled water and incubated in picro-Sirius Red solution (0.5 g Direct Red 80 dissolved in saturated picric acid) for 1 h. After staining, sections were washed twice in acidified water (0.5% acetic acid), briefly blotted to remove excess fluid, and dehydrated through three changes of 100% ethanol. Slides were cleared in xylene and mounted using a resinous medium. Collagen fibril organization and tissue morphology were visualized using bright-field microscopy (INCLUDE MICROSCOPE MODEL).

2.7. Viability of coaxial bioprinted RADA16-I cell-laden scaffolds

For bioprinting, G-code files of 4-layered square scaffolds, with 5 lines per layer, a line distance of 2 mm, and a layer height of 650 μm were generated using custom Python script. Scaffolds were printed using a core-shell ratio of 20/80. Printed scaffolds were crosslinked in a solution containing 150 mM CaCl₂ and 50 mM HEPES for 5 min, followed by 3 washes in culture medium. The viability of bioprinted MSCs encapsulated in RADA16-I hydrogels was evaluated after 1, 3, and 7 days of incubation at 37 °C in a humidified incubator with 5% CO₂. Live cells were stained using calcein AM cell permanent dye, and dead cells were counterstained using propidium iodide membrane impermeant dye. Hydrogel samples were incubated in a working solution containing 2 μM calcein AM and 750 nM propidium iodide (PI) at 37 °C and 5% CO₂. Then samples were washed twice in buffer containing 150 mM NaCl and 10 mM CaCl₂ and imaged with a Leica TCS SP8 CLSM using excitation/emission wavelengths of 494/510-540 and 535/600-640 nm, respectively.

2.8. Statistical analysis

Each experiment was performed in at least three replicates (n = 3). The distribution of values was considered normal for evaluating statistical differences among all experiments performed. Data are presented as the mean and standard deviation of the replicates. A one-way ANOVA was performed followed by Tukey HSD posthoc analysis to compare all means using Origin 2022 (OriginLab).

3. Results

The present study utilized a custom-modified coaxial 3D bioprinter to co-extrude RADA16-I self-assembling peptide solution surrounded by a load-bearing MC/alginate composite hydrogel (Fig. 2). Immediately after co-extrusion, the phosphate-buffered viscous hydrogel shell enabled the in-situ self-assembly of RADA16-I cell-laden hydrogel core. The alginate-containing shell was crosslinked in a HEPES-buffered CaCl₂ solution to ensure the stability of the fabricated hydrogel scaffolds under cell culture conditions. Over time, the non-crosslinked MC gradually released from the 3D printed scaffolds, resulting in an open-porous alginate shell.

Fig. 2.

Schematic of core-shell 3D bioprinting process.

3.1. Rheological characterisation of core and shell hydrogel

Methylcellulose (MC) has been effectively utilized as an internal stabilizing component for alginate to improve printability [41,42]. Adjusting the concentrations of MC, electrolytes, and carbohydrates allows for the fine-tuning of the rheological properties of MC-containing hydrogels [43]. In this study, two separate bioink formulations were prepared: MC was blended with RADA16-I for the core bioink, and MC was blended with alginate for the shell bioink to enhance printability while maintaining cytocompatibility. Rotational flow sweeps compared the RADA16-I solution with and without supplementation of MC (Fig. 3, A; i). The addition of MC increased the viscosity of the RADA16-I solution while maintaining good shear-thinning behavior across the tested shear-rate range. Increasing viscosity is expected to reduce cell sedimentation and support the formation of continuous filaments by limiting spillage of the low-viscosity RADA16-I core solution prior to self-assembly. The effect of MC on the self-assembly kinetics and viscoelastic properties of RADA16-I hydrogels was evaluated by monitoring their self-assembly under physiological pH for up to 10 min (Fig. 3, A; ii). Upon addition of PBS buffer solution, G′ rapidly increased, indicating the self-assembly of RADA16-I hydrogels. MC caused a modest delay in the onset of self-assembly, likely due to steric hindrance and increased solvent viscosity. Nevertheless, both formulations reached comparable final storage moduli (RADA: 105 ± 12 Pa, RADA–MC: 90 ± 2 Pa), indicating preserved peptide assembly. We hypothesize that the presence of MC introduces steric hindrance and increases the overall viscosity of the precursor solution, which transiently slows down peptide diffusion and assembly kinetics. However, since the final storage modulus (G′) of RADA16-I and RADA-MC hydrogels is comparable, this suggests that MC does not prevent RADA16-I self-assembly. This interpretation is consistent with previous findings by Cofiño et al., who reported that MC increases the viscosity of the formulation without disrupting the peptide's self-assembly mechanism [27]. Potential non-covalent interactions, such as hydrogen bonding between MC hydroxyl groups and peptide termini, may further modulate local organization, contributing to the delay observed during self-assembly. For the shell bioink, MC was blended with an alginate solution containing PBS and sucrose to improve flow proper-ties and provide pH-buffering capabilities for extrusion bioprinting. The MC-alginate blend displayed viscosity values two orders of magnitude higher at low shear rates compared to alginate alone (Fig. 3, B; i). The viscosity of alginate solutions decreased with increasing shear rate initially but plateaued at higher shear rates, indicating polymer chain alignment. In contrast, the MC-alginate composite displayed shear-thinning properties across the entire range of shear rates tested. The crosslinking kinetics of the MC-alginate hydrogels were evaluated for up to 15 min by incubating them in a HEPES-buffered CaCl₂ crosslinking solution (Fig. 3, B; ii). Before crosslinking, the MC-alginate hydrogel formulation displayed a tan(δ) value of approximately 0.26, which has been previously reported to enable extrusion printing of scaffolds with excellent shape fidelity [43]. While alginate alone rapidly crosslinked to form stiff hydrogels, achieving a tan(δ) of approximately 0.15 after 5 min, the MC-alginate hydrogels showed significantly lower elastic and a higher viscous portion, indicated by a tan(δ) of approximately 0.26. This result suggests that MC, which was not ionically crosslinked, reduced the rate of crosslinking of alginate through steric hindrance and effects on viscosity. The presence of MC polymer might physically obstruct the interaction between alginate chains and calcium ions, while increasing the viscosity of the solution hinders the mobility of both alginate chains and calcium ions, thereby slowing down the crosslinking process.

Fig. 3.

Rheological characterization of core and shell solution. (A; i) Flow behavior of RADA16-I core solution. (A; ii) Self-assembly of RADA16-I solution after adding 3x PBS. (B; i) Flow behavior of alginate-containing shell material. Shaded areas indicate the standard deviation. (B; ii) Crosslinking of alginate/MC blend in comparison to alginate after adding CaCl₂ solution.

3.2. Core-shell 3D printing

3.2.1. Morphology of core-shell extruded filaments

The viscous alginate/MC bioink shell was designed to buffer the peptide-based core fluid to physiological pH immediately after nozzle extrusion, thereby triggering the self-assembly of peptide hydrogels and enabling the extrusion of stable filaments. Screening experiments were conducted to identify the optimal core-shell extrusion ratios for the production of stable filaments, with a focus on maximizing the proportion of soft core hydrogel (Fig. 4, A). Screening experiments demonstrated the extrusion of stable filaments when the percentage of core extrusion was below 30% (Fig. 4, A; i). At higher core extrusion ratios, printed strands displayed a marked increase in width and reduced geometric uniformity, reflected by higher standard deviations in filament width (Fig. 4,A; ii). At the highest core ratios, intermittent core breakthrough was observed, consistent with insufficient shell confinement under these conditions. Because this effect was not reproducible across replicates, it is reported qualitatively. The self-assembly of RADA16-I peptide into fibrillar structures to form stable hydrogels within co-extruded strands was confirmed by ThT staining (Fig. 4, B).

Fig. 4.

Screening the effect of relative core-shell extrusion rate on 3D printed strand morphology. (A) Stereomicroscopic images of hydrogel strands extruded using different core-shell extrusion ratios. Scale bar = 1 mm. (A; i) Quantitative evaluation of strand width of extruded hydrogel filaments using different core-shell extrusion ratios. Error bars denote standard deviation. (B) Hydrogel strand with a core-shell extrusion ratio of 20/80 showing the ThT-stained self-assembled RADA16-I core in green and bright field image of MC/alg shell. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.2.2. Shape fidelity of printed scaffolds

After establishing a core extrusion limit (≤30%) that allows for the extrusion of stable filaments, the capability of stacking multiple filaments in the Z-direction to create 3D objects was investigated. Therefore, 8-layered cubical structures were printed, and their relative heights were compared (Fig. 5). Scaffolds printed at core extrusion percentages of 20% and 25% exhibited good shape fidelity, with the structural height being at least 90% of the control fabricated from shell material only (Fig. 5, A; i, ii). It is important to note that slight adjustments to the relative extrusion rate were made to account for increasing filament collapse at higher core extrusion percentages. In contrast, scaffolds printed at a core extrusion percentage of 30% displayed impaired shape fidelity despite increasing the relative extrusion rate. The cubical structures collapsed inwards, and filaments started to fuse between layers. Consequently, a core extrusion percentage of 20% was selected for further experiments. Moreover, the capability to print larger woodpile structures by adjusting the layer height accordingly was demonstrated (Fig. 5, B). The printed structures exhibited excellent shape fidelity and the presence of open pores.

Fig. 5.

Printability and structural integrity. (A; i) Images of 3D printed 8-layered, square structures extruded using different core-shell extrusion ratios. Scale bar = 5 mm. (A; ii) Quantification of structural integrity. Error bars denote standard deviation. Statistical significant differences were marked with ∗p < 0.05, or ∗∗p < 0.01. (B) 3D printed 12-layered scaffold comparing 0%, to 20% core extrusion rate. (B; i) Image from top-view. Scale bar = 5 mm. (B; ii) Stereomicroscope image from top-view. Scale bar = 1 mm. (B, iii) Image taken from side-view. Scale bar = 5 mm.

3.3. Scaffold morphology and MC release

The morphology of printed hydrogel scaffolds cultivated under cell culture conditions was evaluated over a period of 21 days (Fig. 6). First, the alginate/MC bioink shell of the 3D printed structures was crosslinked in CaCl2 solution (Fig. 6, A; i, ii). The release of the MC polymer from the scaffold, which is not ionically crosslinked by Ca²⁺, was visualized using chlorine-zinc-iodine staining (Fig. 6, A; iii). Hydrogels stained on the first day of incubation appeared dark brown, indicating the presence of MC. After 21 days of incubation, the intensity of chlorine-zinc-iodine staining decreased, indicating the release of MC. Quantitative assessment using Mykoval™ staining revealed that both groups released approximately 20% of the total MC content after crosslinking and subsequent washings (Fig. 6, B). After 7 days, scaffolds printed with the shell material only and 20% core extrusion released 52% and 63% of the total MC content, respectively. By 21 days, around 90% of the initial MC content had been released in both conditions. The gradual release of MC is expected to generate an open, interconnected pore network within the hydrogel, thereby improving diffusion and nutrient exchange. This mechanism has been well documented in previous studies, where leaching of MC from alginate/MC blends led to the formation of a porous architecture [44,45]. Thus, the MC release profile observed here is indicative of progressive pore formation during culture.

Fig. 6.

Scaffold morphology and MC release. (A) Images of 3D printed 4-layered scaffold comparing 0%, to 20% core extrusion rate (A; i) before crosslinking, (A; ii) and after crosslinking for 5 min in 150 mM CaCl2 solution. (A, iii) Scaffolds in culture stained after 1, and 21 days with zinc-iodine-solution. Scale bar = 5 mm. (B) Quantitative measurement of cumulative percentage of MC release over time. Error bars denote standard deviation.

3.4. Cell morphology and hydrogel contraction

Supplementation of MC to RADA16-I solution altered the distribution of encapsulated cells. To track cell distribution and cluster formation, actin filaments of MSCs encapsulated in RADA and RADA-MC were stained (Fig. 7). Images showed cells evenly distributed in RADA, while they formed aggregates in RADA-MC (Fig. 7A–i–ii). Image analysis revealed a significant increase in both the number and size of cell aggregates in RADA-MC compared to RADA (Fig. 7B). Representative confocal images taken after three days of culture further illustrate these differences (Fig. 7C–i–ii). Cells encapsulated in RADA alone exhibited greater spreading and visibly contracted the hydrogel, whereas in RADA-MC, cell spreading was reduced but still apparent between neighboring aggregates. Quantitative analysis of gel contraction showed that, within 24 h, encapsulated cells strained and contracted RADA-MC gels to ≈86 % of their original area, further decreasing to ≈69 % after 72 h (Fig. 7D). In the absence of MC, actin filaments of encapsulated MSCs formed a continuous interconnected network, and hydrogel samples significantly contracted, forming high cell-density structures that decreased to ≈25 % of their original area after 72 h. Notably, aggregates were visible immediately after printing, indicating that clustering likely resulted from physicochemical interactions between the RADA–MC–sucrose matrix and cell membranes during mixing, rather than from cell-mediated self-assembly processes. Recent work has demonstrated that methylcellulose and other crowding polymers can induce rapid, passive cell aggregation through depletion interactions and reduced electrostatic repulsion [46]. This is consistent with our observation that aggregates are present immediately after mixing and prior to culture, supporting the interpretation that polymer-induced interactions govern early cluster formation. Interestingly, characterization of viscoelastic properties showed similar results for RADA and RADA-MC across the tested frequency range (Fig. S1). Both hydrogels displayed a gel-like behavior with a loss tangent (tan δ = 0.05 at 1 Hz) within a range comparable to that of biological tissues and reconstituted ECM (tan δ = 0.1 at 1 Hz) [47]. Since hydrogel viscoelasticity is known to influence cell–matrix interactions and processes such as spreading, proliferation, and differentiation, similar rheological behavior suggests that RADA and RADA-MC would elicit comparable mechanobiological responses. However, the distinct aggregation patterns suggest that biochemical or colloidal interactions rather than bulk viscoelasticity alone, dominate early cell organization in this system. Dedicated comparative studies will be required to disentangle the relative contributions of these factors.

Fig. 7.

Cell morphology and hydrogel contraction. (A) Day 0 of MSCs encapsulated in (A; i) RADA-MC, and (A; ii) RADA hydrogel stained with phalloidin-DAPI. Scale bar = 500 μm (B) Quantification of cell aggregate volume stained positive for phalloidin. Samples were analyzed from 3 independent experiments. (C) Day 3 of MSCs encapsulated in (C; i) RADA-MC, and (C; ii) RADA hydrogel stained with phalloidin-DAPI. Scale bar represents for 500 μm in the bottom right (10x magnification), and 200 μm in the bottom left (40x magnification). (D) Gel contraction of RADA, and RADA-MC measured for up to 3 days. Statistical significant differences were marked with ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001.

3.5. Osteogenic differentiation

RADA16-I provides a dynamic hydrogel nanoarchitecture that supports the survival and differentiation of encapsulated cells. The dynamic nature of the hydrogel allows cells to aggregate and form cell-cell contacts, driving tissue maturation. MSCs were encapsulated in RADA-MC hydrogel, cultured in proliferation or osteogenic medium, and compared to cells cultivated in 2D (Fig. 8). The activity of ALP and OPN was measured to monitor bone tissue maturation in the tested samples (Fig. 8A and B). ALP activity showed an initial increase at 24 h in 3D cultures, which may reflect a transient stress response to encapsulation and brief exposure to the acidic RADA16-I solution, as reported in other stress-associated ALP upregulation studies [48,49]. In this study, the early ALP peak appeared in both osteogenic and non-osteogenic media and was markedly higher in 3D than in 2D conditions, suggesting that encapsulation rather than osteogenic stimulation was responsible.

Fig. 8.

Osteogenic differentiation of MSCs. (A) ALP and (B) OPN produced by MSCs in 2D culture (TCPS) and RADA-MC hydrogel. Symbols indicate statistically significant differences between groups (p < 0.05). (C) Histological sections of MSCs in RADA-MC hydrogel after 21 days, stained with (i, ii) Sirius Red and (iii) Alizarin Red. Scale bars represent 500 μm for (i) and (iii), and 200 μm for (ii). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Over time, ALP activity either decreased or plateaued, showing relatively similar values across the different culture conditions. In encapsulated cells cultured in osteogenic medium, ALP activity gradually declined, consistent with the transition from early osteogenic activity toward later stages of differentiation in which matrix mineralization becomes predominant. This trend was supported by the increased secretion of OPN, a marker associated with bone matrix maturation, in 3D constructs maintained in osteoinductive medium. Histological analysis further confirmed osteogenic progression, demonstrating the deposition of collagen and phosphate-containing mineral within the hydrogel matrix (Fig. 8, C). A dense collagen network was observed in MSCs encapsulated in RADA–MC hydrogels and cultured in both proliferation and osteogenic media (Fig. 8C–i, ii). In contrast, pronounced calcium phosphate deposition occurred only in constructs maintained in osteogenic differentiation medium (Fig. 8C–iii). These findings indicate that RADA–MC hydrogels supported collagen synthesis by encapsulated MSCs under basal conditions and promote the formation of bone-like extracellular matrix when osteogenic cues were provided.

3.6. Core-shell bioprinting RADA16-I hydrogel

Upon co-extrusion, RADA16-I hydrogel core self-assembles provide encapsulated cells with a well-ordered and dynamic microenvironment. The viability of MSCs printed with a RADA16-I peptide hydrogel core was assessed for up to 7 days (Fig. 9). Modulating the rheological properties of the RADA16-I core solution with MC and sucrose slowed down cell sedimentation prior to printing. In addition, sucrose likely contributed to protecting encapsulated cells from the acidic nature of RADA16-I [50]. This modification allowed for the extrusion of continuous and stable filaments containing viable cells homogeneously distributed throughout the 3D-printed hydrogel scaffold structure. Live/dead analysis revealed that MSC viability remained consistently high during culture, with >85% viable cells detected at days 1, 3, and 7. These results indicate that both the printing procedure and the RADA-MC hydrogel core provide a cytocompatible environment that supports sustained cell survival. Throughout the culture period, cells were observed exclusively within the peptide-based core, with no evidence of migration into the surrounding alginate/MC shell. This behavior is consistent with the bioinert nature of alginate, which lacks adhesive ligands and does not support cell spreading or active migration [51]. Similar to RADA-MC hydrogel self-assembled in filter inserts, encapsulated MSCs tend to aggregate in co-extruded hydrogel filaments. Despite the increase in viscosity due to the additives, the core solution maintained favorable rheological behavior for bioprinting, including shear-thinning properties with a shear-thinning index of n = 0.257 and a consistency coefficient of K = 1.53. Computational fluid dynamics (CFD) analysis indicated that the maximum shear stress experienced by encapsulated cells during nozzle extrusion did not exceed 20 Pa. This value is substantially lower than thresholds reported in previous bioprinting studies, where cell viability decreases exponentially with rising shear stress. Ouyang et al. showed that embryonic stem cells maintained >90% viability when shear stress remained below 100 Pa, whereas higher stresses progressively reduced survival [52]. Similarly, Emmermacher et al. reported that MSC viability decreased to ∼60% at shear stresses of 3050 Pa and identified 1300 Pa as a critical upper limit for preserving high viability [53]. The shear stress estimated in our system is therefore at least an order of magnitude lower than levels known to compromise cell survival, which is consistent with the high viability observed in our live/dead staining assays.

Fig. 9.

Cell viability of core-shell bioprinted RADA16-I hydrogel cell-laden scaffolds. (A) Overlay of bright field image of entire scaffolds and fluorescence image of live/dead stained MSCs encapsulated in RADA-MC core hydrogel taken on (A; i) day 1, and (A; ii) day 7. (B) Magnification of fluorescence image of live/dead stained MSCs encapsulated in RADA-MC core hydrogel taken on (B; i) day 1, and (B; ii) day 7. (C) Image of live/dead stained MSCs showing a single strand of printed hydrogel scaffold on (C; i) day 1, and (C; ii) day 7. Scale bar = 1 mm for all images.

4. Conclusion

In this study, we introduced a novel biofabrication approach that combined top-down coaxial 3D extrusion bioprinting with the bottom-up self-assembly of a soft RADA16-I peptide hydrogel. This methodology aimed to address the challenges associated with using self-assembling peptide hydrogels for extrusion bioprinting, particularly their modest mechanical properties, by leveraging the strengths of MC and alginate as a load-bearing composite shell material. Our results demonstrated that supplementing the RADA16-I core with MC and sucrose significantly improved the viscosity and printability of the bioink without compromising the peptide's self-assembly capabilities.

Rheological characterization revealed that the MC-enhanced RADA16-I solution maintained shear-thinning properties while ensuring the formation of stable hydrogels upon self-assembly. The MC/alginate composite shell effectively buffered the core fluid to physiological pH immediately after extrusion, promoting rapid self-assembly of the peptide hydrogels. Optimal core-shell extrusion ratios were determined through screening experiments, establishing that a core extrusion percentage of 20% provided the best balance between maintaining filament stability and maximizing the soft core content. Morphological analyses confirmed the successful co-extrusion of stable filaments, and subsequent 3D printing of scaffold structures exhibited excellent shape fidelity and structural integrity. Over a 21-day culture period, the gradual release of MC from the alginate shell created an open-porous scaffold, facilitating nutrient diffusion and supporting cell growth. Encapsulated MSCs in the RADA-MC hydrogel demonstrated high viability, homogeneous distribution, and the ability to form aggregates. Importantly, these cells retained their capacity for osteogenic differentiation, evidenced by increased OPN expression, and the formation of a dense collagen network with calcium phosphate deposition when cultured in osteoinductive media. In conclusion, the integration of coaxial 3D extrusion bioprinting with self-assembling RADA16-I peptide hydrogels, supplemented with MC, presents a promising strategy for fabricating stable, cell-laden scaffolds. This strategy improves the printability of inherently soft peptide hydrogels without sacrificing function, providing a practical route toward fabricating biologically relevant, peptide-based tissue constructs.

CRediT authorship contribution statement

M. Jergitsch: Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. S. Perez-Amodio: Formal analysis, Investigation, Methodology, Writing – review & editing. L.M. Delgado: Formal analysis, Investigation, Methodology, Writing – review & editing. R.A. Perez: Funding acquisition, Resources, Writing – review & editing. M.A. Mateos-Timoneda: Conceptualization, Funding acquisition, Resources, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Miguel Angel Mateos-Timoneda reports financial support was provided by Spain Ministry of Science and Innovation. Roman A. Perez reports financial support was provided by Generalitat de Catalunya Ministry of Research and Universities. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Data availability

Data will be made available on request.