Cell-adhesive hydrogels composed of peptide nanofibers responsive to biological ions

Abstract

Novel calcium ion (Ca²⁺)-responsive hydrogels composed of designed β-sheet peptides were constructed. As the novel designed peptide, E1Y9, has a Glu residue to interact with Ca²⁺, the peptide in the sol-state self-assembled into hydrogels in the presence of Ca²⁺. The hydrogelation did not occur in the absence of Ca²⁺; therefore, Ca²⁺-dependent hydrogelation was achieved by the molecular design. The hydrogelation from the viscous sol-state solution can be induced by a slower self-assembly process of the β-sheet peptide involving a rapid process of Ca²⁺binding. When the sol-state peptide solution was injected with Ca²⁺, gel drops and strings with desired shapes could be constructed. Different cell lines can be cultured on the hydrogel, demonstrating its low toxicity, which is comparable to commercially available microtiter plate surfaces for cell culture. Furthermore, the hydrogels showed a high cell-adhesive ability that was similar in magnitude to fibronectin, which is a native cell-adhesive protein. The Ca²⁺-responsive peptide nanofiber-based hydrogelation system will facilitate novel studies exploiting self-assembling peptide nanomaterials that will lead to cell-based technology, such as three-dimensional cell culturing.

Introduction

The study of the self-assembled properties of biomolecules, such as peptides, proteins and nucleic acids, has developed in recent years into an attractive and diverse research field involving nanostructured materials.^{1, 2} In natural systems, peptides and proteins are spontaneously organized in supramolecular architectures through numerous weak interactions (noncovalent bonds), such as electrostatic interactions, hydrogen bonding, π-stacking, van der Waals interactions and hydrophobic effects. The ability of peptides and proteins to construct specific secondary structures provides a useful function for bioinspired material design that cannot be achieved with traditional organic molecules and polymers.^{3, 4} Peptides act as excellent building blocks for nanomaterials because of their ease of design and synthesis, small size and stability; therefore, peptide-based molecular designs have been used for the formation of novel smart materials, such as fibers or tubes,^{5, 6, 7, 8} fibrils,^{9, 10} capsules,^{11, 12} surfactants^{13, 14} and hydrogels.^{15, 16, 17} We have developed de novo designed short β-sheet peptides that self-assemble into nanofibers with a uniform width of ca 100 nm for potential applications ranging from scaffolds for protein immobilization to metal-biofusion materials.^{18, 19, 20, 21} More recently, we found that surface-functionalized self-assembling nanofibers, using a short peptide composed of nine amino-acid residues (Y9 peptide, sequence: Ac-YEYKYEYKY-NH₂), could be good scaffold candidates for cell adhesion.²² In nature, cells are surrounded by a complex three-dimensional (3D) extracellular matrix, which affords a structural function and physical strength to cells. Thus, artificial extracellular matrices, cell scaffolds, are fundamental to provide similar environments to the biological systems and also to form the required structures for tissue engineering. Self-assembling peptide hydrogels with properties responsive to local environmental stimuli lead to materials suitable for a wide range of biomedical applications.²³ As peptide interactions can be controlled by external stimuli, such as pH, ionic strength, temperature, light or the presence of small molecules, peptides could act as components for stimuli-responsive hydrogels.

In contrast, amino acids, such as His, Cys, Met, Glu, and/or Asp, interact with biological metal ions specifically to induce structural changes of proteins and to form supramolecular architectures. In particular, calcium is essential for living organisms, especially in cell biology, in which movement of the calcium ion (Ca²⁺) into and out of the endoplasmic reticulum functions as a signal for a number of cellular processes. Therefore, it might be possible to use hydrogels that respond to Ca²⁺ to construct exquisite scaffolds for cell engineering. In this study, we report the construction of a self-assembling peptide-based nanofiber and hydrogel system that is triggered by Ca²⁺ binding. Generally, a carboxylic group in Glu that is known to interact with Ca²⁺ was conjugated to the designed β-sheet Y9 peptides. The self-assembling ability of Y9 peptides is expected to be enhanced by the coordination of Ca²⁺ with intermolecular Glu residues. The novel Glu-conjugated designed peptides were self-assembled into highly networked peptide nanofibers, and the nanofiber formations were strongly dependent on the presence of Ca²⁺. Higher concentrations of the peptide solutions in the presence of Ca²⁺ resulted in the formation of self-supporting hydrogels. Furthermore, the injection of the peptide solutions into Ca²⁺ solutions resulted in the construction of gel strings that can be transformed into various desired shapes. Cell culture experiments on hydrogels composed of the designed peptide nanofibers conjugated with the cell-adhesive peptide (RGDS) suggested that the peptides could act as scaffolds for cell-adhesive nanomaterials. Furthermore, adhesion cells on the peptide gel strings were well spreadout and differentiated with a high rate of survival. The molecular design-based concept will lead to attractive technology that uses the novel strategy involving cell-adhesive peptide nanofiber-based stimuli-responsive materials.

Materials and methods

Peptide synthesis

Peptides were synthesized by a conventional solid-phase method by using a 9-fluorenylmethyloxycarbonyl (Fmoc) strategy.²⁴ The peptide chains were assembled on an NH-SAL-PEG resin (purchased from Watanabe Chemical Industries, Hiroshima, Japan) (N-termini were acetylated and C-termini were amidated) with Fmoc amino-acid derivatives (3 equiv), N,N-diisopropylethylamine (DIEA, 6 equiv) and 2-(1H-Benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (3 equiv) in N-methylpyrrolidone (NMP) for coupling and piperidine (20%)/NMP for Fmoc removal. In the case of coupling, 2-(aminoethoxy) ethoxy acetic acid (Aeea), 2-(1H-9-azabenzotriazole-1-yl)-1,1,3,3-tetramethyluroniumhexafluorophosphate (HATU, 1.5 equiv) and 3 equiv of DIEA were used. To cleave the peptide from the resin and remove the side-chain protecting groups, the peptide resin was treated with trifluoroacetic acid (TFA)/triisopropylsilane/water (95/2.5/2.5). All peptides were purified by reverse-phase HPLC on a Cosmosil 5C18-AR-II packedcolumn (10 × 250 mm²) with a linear gradient of acetonitrile/0.1% TFA at a flow rate of 3.0 ml min⁻¹. The peptides were identified satisfactorily by liquid chromatography-electron spray ionization mass spectrometry on a Shimadzu LCMS-2010 (Shimadzu, Tokyo, Japan). Synthesized peptides showed over 95% purities, and the observed mass values agreed with the calculated mass values within 1.0 mass unit (Y9: M_w calcd1389.5, m/z_calcd [M+2H]²⁺ 695.8, m/z_obsd [M+2H]²⁺ 695.2; E1Y9: M_w calcd 1518.6, m/z_calcd [M+2H]²⁺ 760.3, m/z_obsd [M+2H]²⁺ 760.3; E2Y9: M_w calcd 1647.7, m/z_calcd [M+2H]²⁺ 824.9, m/z_obsd [M+2H]²⁺ 825.3; E3Y9: M_w calcd 1776.8, m/z_calcd [M+2H]²⁺ 889.4, m/z_obsd [M+2H]²⁺ 889.4; E4Y9: M_w calcd 1906.0, m/z_calcd [M+2H]²⁺ 954.0, m/z_obsd [M+2H]²⁺ 954.3; E1Y9-RGDS: M_w calcd 2079.2, m/z_calcd [M+2H]²⁺ 1040.6, m/z_obsd [M+2H]²⁺ 1040.8, m/z_calcd [M+3H]³⁺ 694.1, m/z_obsd [M+3H]³⁺ 694.1).

Infrared (IR) analysis

Peptides were dissolved in ultrapure water with a concentration of 400 μM and incubated for 48 h for nanofiber formation. Attenuated total reflection-IR spectra of peptide nanofibers were obtained using refractive surfaces (IRPrestige-21 with DuraSampl IR-II, Shimadzu, Tokyo, Japan) in air at ambient temperature. Interferograms were co-added 100 times and Fourier transformed at a resolution of 4 cm⁻¹.

Thioflavin T (ThT) binding analysis

Peptides were dissolved in ultrapure water with a concentration of 400 μM containing 0, 0.4, 0.8, 1.6, 4.0 or 8.0 mM Ca²⁺. Then, 20 μl of the mixed solution was added to 80 μl of ThT solution with a concentration of 50 μM, and the mixtures were incubated for 1 h at ambient temperature. Subsequently, the amounts of peptide nanofibers were estimated by measuring the fluorescence intensities of binding ThT to the peptide nanofibers (λ_ex and λ_em were 440 and 490 nm, respectively).

Transmission electron microscopic (TEM) observation

A collodion-coated copper electron microscopic grid was placed coated-side down onto the peptide nanofiber solutions for 30 min, and then the grid was floated onto a droplet of ultrapure water for 1 min for washing. The sample was negatively stained with 2% (w/v) uranyl acetate for 1 min. The grid was blotted and then allowed to dry gradually at ambient temperature overnight. All images were taken using a Hitachi H-7500 electron microscope (Hitachi, Tokyo, Japan) operating at 80 kV.

Gel drop and string formation

Peptide solutions (2.0 wt%) were incubated for 48 h, and then the solutions were mixed with Coomassie brilliant blue or ThT solution (the final concentrations of the peptide, Coomassie brilliant blue and ThT are 1.8 wt%, 0.05 wt% and 150 μM, respectively). The mixed solutions were injected into Ca²⁺ solutions with a concentration of 100 mM. Then, the constructed hydrogels were observed by optical and fluorescence microscopies.

Cell culture on peptide hydrogels

3T3-L1 cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum and 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin. PC12 cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum, 10% Donor Horse Serum and 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin. The cells were seeded in a 10-cm diameter Petri dish and allowed to grow until ca 50% (3T3-L1) or 80% (PC12) confluence. All incubations were performed at 37 °C, 5% CO₂. RGDS conjugated E1Y9 (E1Y9-RGDS) and E1Y9 peptides were mixed with percentages of 0, 10 or 20 of E1Y9-RGDS peptides. The mixed peptide solutions were incubated in 24-well tissue culture-treated plates (TCTP, Techno Plastic Products AG, Trasadingen, Switzerland) or injected with 500 μl of 100 mM Ca²⁺ solutions in 24-well non-adhesive-treated plates (Asahi Glass, Tokyo, Japan) to construct flat gels or gel strings. Next, supernatants were removed and the gels were washed twice with serum-free medium. Cells in serum-free medium (5000 cells per 100 μl) were then applied to the well and incubated for several hours. Optical and fluorescence microscopic observations were used to evaluate cell adhesion and differentiation using Ti-U-PH-1 (Nikon, Tokyo, Japan). The percentages of strong adhered cells were estimated by comparing the changes in the fluorescence intensity before and after washing. For fluorescence observations, calcein-acetoxymethyl ester (AM) and propidium iodide were used for the quantification of cell numbers and live/dead assays.

Results and Discussion

Spectroscopic characterization of peptide nanofiber

One to four Glu residues were conjugated to the N-termini of Y9 peptide (one Glu residue-conjugated Y9 peptides were named E1Y9; the other designed peptides were named similarly). The peptides were dissolved in water, and the solutions were incubated for 48 h at ambient temperature. To investigate the effect of Glu conjugation, the secondary structures of the peptides were analyzed by attenuated total reflection-IR. IR spectra of the amide I region (1600–1700 cm⁻¹), which is primarily assigned to C=O stretching, was attributed to β-sheet (1620–1640 cm⁻¹), α-helix (1648–1655 cm⁻¹) and antiparallel β-sheet (1670–1690 cm⁻¹)^{25, 26, 27} structures. The IR spectra of Y9 (Ac-YEYKYEYKY-NH₂) and Glu-conjugated Y9 (EnY9, n=1, 2, 3, or 4) nanofibers are shown in Figure 1. The spectrum of the original Y9 nanofibers shows a sharp band at 1625 cm⁻¹ and a much smaller band at 1690 cm⁻¹ corresponding to antiparallel β-sheet structures. EnY9 peptides also show almost the same bands as Y9 peptides, as expected, demonstrating antiparallel β-sheet structures. Therefore, the conjugation of various Glu residues to Y9 peptides did not show significant effects on the secondary structural formation, which is important for construction of nanofibers, implying that the assembling abilities of EnY9 peptides are maintained. Circular dichroism measurements for analyzing the secondary structure of peptides could not be performed due to the strong exciton coupling of the Tyr residue in the nanofibers.

Figure 1

Secondary structural analyses of designed peptides. The infrared spectra of Y9 (Ac-YEYKYEYKY-NH₂) and EnY9 (Ac-E_nYEYKYEYKY-NH₂) peptides are shown. All samples were measured at 400 μM after 48 h incubation, and the scale of absorbance is shown in the figure.

ThT-binding analysis was performed to evaluate the fiber-forming abilities of Y9 and EnY9 peptides in the absence and presence of Ca²⁺. ThT is widely used to visualize and quantify the presence of self-assembled β-sheet peptides or proteins due to the fluorescence enhancement caused when it binds to β-structures. In this case, ThT can be used to quantify the nanofiber formation. In the case without Ca²⁺, the fluorescence intensities of the peptides showed almost the same values and are larger than that of ThT alone (Figure 2), suggesting nanofiber formation. Addition of different concentrations of Ca²⁺ to the original Y9 peptide solution resulted in subtle fluorescence enhancement, suggesting that Ca²⁺ might interact with Glu residues in the original Y9 peptides. In contrast, EnY9 peptide solutions with Ca²⁺ showed much higher fluorescence intensities than solutions without Ca²⁺ and the original Y9. Furthermore, different fluorescence profiles in each peptide clearly suggested that organized β-structures and nanofibers were distinct from one another. In all EnY9 peptides, the fluorescence intensities were enhanced, and the intensities were dependent on the concentration of Ca²⁺, demonstrating that the nanofiber formation derived from β-structure construction was facilitated by the interaction with Ca²⁺. In particular, the maximum fluorescence intensity value of E4Y9 with 8 mM of Ca²⁺was 18 times greater than that of the original Y9 with the same concentration of Ca²⁺, demonstrating the efficacy of Ca²⁺ for the EnY9 assembly. The fluorescence intensities of the peptide solutions under the same Ca²⁺ concentration were dependent on the conjugated Glu residue numbers (the order of the fluorescence intensity values was E4>E3>E2>E1>the original, in the presence of the same concentration of Ca²⁺), indicating that designed and conjugated Glu residues preferentially interact with Ca²⁺, which leads to intermolecular interactions of peptides via Ca²⁺. In other words, the Glu residues in the designed peptides were effective for β-structure and nanofiber formation due to the interactions with Ca²⁺.

Figure 2

Thioflavin T (ThT)-binding analyses of peptide anofibers. The designed peptide (100 μM) nanofibers and ThT (40 μM) were incubated in water in the presence and absence of Ca²⁺ with various concentrations. The fluorescence intensity values of ThT binding to nanofibers were measured (λ_ex and λ_em were 440 and 490 nm, respectively) to estimate the number of peptide nanofibers. Control experiments without peptides were also performed (ThT only).

Ca ²⁺ -dependent nanofiber formation and hydrogelation

In the previous section, the formation abilities of the peptides for β-structures and nanofibers were spectroscopically characterized. The fluorescence intensities, however, did not provide direct evidence for the nanofiber formation. To better understand the fiber formation abilities, direct observation should be performed by TEM analyses. Y9 and EnY9 peptide nanofibers were incubated with and without Ca²⁺ and applied to the TEM grid. Highly networked and disentangled peptide nanofibers constructed by Y9 or EnY9 peptides were observed (Figures 3a–e). All nanofibers showed a similar morphology with a width of approximately 10–20 nm and a length of several micrometers, suggesting that the conjugated Glu residues do not affect the self-assembling ability in the absence of Ca²⁺. When Ca²⁺ was added to the peptide solutions, the morphologies of EnY9 nanofibers were drastically changed (Figures 3g–j), even though that of the original Y9 nanofiber was not changed (Figure 3f). The EnY9 peptides in the presence of Ca²⁺were self-assembled into nanofibers with widths of 30–90 nm and lengths of several micrometers. The addition of Ca²⁺ resulted in increases of the width, supporting interactions of Ca²⁺ with the peptides. The Ca²⁺ might act as a cross-linking reagent for intermolecular interaction of the peptides, leading to the construction of thicker nanofibers. E1Y9 nanofibers are well disentangled, just as the original Y9 nanofibers are, although the number of entangled nanofibers increased with an increasing number of Glu residues. In particular, E4Y9 nanofibers are strongly entangled with each other (Figure 3j). These results demonstrated that the Glu residues introduced to the Y9 peptide interact with Ca²⁺, that is, the introduction of Glu residues afforded a responsive ability to Ca²⁺ for the nanofiber assembly.

Figure 3

TEM images of the peptide nanofibers with the networked structures. Peptide nanofibers were incubated with a concentration of 400 μM (a–e) without and (f–j) with Ca²⁺ (4.0 mM). The peptide names are shown in the each image. Glu-conjugated peptide (EnY9) nanofibers show different morphologies when Ca²⁺ was added. The widths of the nanofibers with Ca²⁺ were several times wider than those of the nanofibers without Ca²⁺. The entanglement degree increased with the conjugation number of the Glu residue. All scale bars represent 500 nm.

The peptide solutions at a higher concentration (0.75 wt%, approximately 5 mM) with and without Ca²⁺ (5 equiv) were prepared to evaluate the hydrogelation abilities. All peptide solutions in the absence of Ca²⁺and both original Y9 peptide solutions yielded subtly viscous solutions (Supplementary Figure S1). In contrast, EnY9 peptide solutions with Ca²⁺yielded highly viscous solutions due to the interactions between the peptides and Ca²⁺ (Supplementary Figure S2). In particular, the E1Y9 peptide solution behaved as a self-supporting hydrogel, as shown by the vial inversion test (Figure 4a), demonstrating that the E1Y9 peptide has a highly responsive ability to phase transition from a solution to a hydrogel via the addition of Ca²⁺. Considering the wider width and disentangled states of the E1Y9 nanofiber, the morphology is appropriate for construction of hydrogels. Owing to the E1Y9 peptide having only one additional Glu residue that can bind to Ca²⁺, Ca²⁺should act as a cross-linker when interacting with E1Y9 peptides, resulting in hydrogelation. Although E2-E4Y9 peptides showed higher abilities in fiber formation in the ThT assay than E1Y9, a longer Glu sequence might weaken the hydrogelation ability, possibly due to an increase in hydrophilicity. To apply the excellent Ca²⁺ responsive hydrogelation ability of E1Y9 peptides, the E1Y9 solution (2.0 wt%) in a higher concentration containing Coomassie brilliant blue or ThT was injected into a Ca²⁺ solution (100 mM). As a result, gel drops and strings composed of the E1Y9 peptides were generated (Figures 4b–e). Based on the optical and fluorescence staining, it was clearly demonstrated that the drops and strings were composed of peptides (Figures 4b–e). As the E1Y9 hydrogels were constructed by injection of the sol-state solution into a Ca²⁺ solution with a faster process after formation of a viscous sol solution with a slower self-assembly process, hydrogels with various shapes could be generated. That is, the hydrogels have high formability for tissue engineering, in which various shapes are desired for 3D cell culturing.

Figure 4

Photographs and microscopic images of the gel drops and strings. (a) E1Y9 peptide nanofibers with and without Ca²⁺ in vial tubes. The peptide solution (0.75 wt%) in the presence of Ca²⁺ self-assembled into self-supporting hydrogels. (b, c) Optical photographs of E1Y9 peptide gel drops (b) and strings (c) constructed by injection of the peptide solution into Ca²⁺ solutions. The scale bars represent 1 cm. (d, e) Bright-field (d) and fluorescence (e) microscopic images of the gel strings. The scale bars represent 500 μm. In the fluorescence image, the peptide nanofibers were stained by ThT.

Cell adhesion on hydrogels

To construct cell-adhesive hydrogels composed of the peptide nanofibers, the RGDS sequence, which is well known to introduce cell adhesion, was conjugated to C-termini of E1Y9 peptides via Aeea as a spacer, and the newly designed peptide was named E1Y9-RGDS peptide. IR, TEM and hydrogelation studies suggested that mixing of up to 20% of E1Y9-RGDS with E1Y9 did not affect to the nanofiber assembly and hydrogel formation (data not shown). Thus, E1Y9-RGDS peptide was used as a mixed substrate with E1Y9 peptide. For cell adhesive materials, a high cell-survival rate on the material is critically important. Flat hydrogels were constructed on the 24-well non-adhesive-treated plate surfaces. A live/dead assay for 3T3-L1 cells using calcein-AM and propidium iodide was performed on the hydrogel and TCTP surfaces, respectively. Fluorescence microscopic images clearly demonstrated the high survival rates on the peptide hydrogels compared with commercially available TCTP surfaces (Supplementary Figure S3). Furthermore, 3T3-L1 cells were cultured on the peptide hydrogel, fibronectin immobilized, TCTP or nonadhesive plate surfaces in the absence of serum-free Dulbecco’s modified Eagle’s medium. Cells were propagated by culture on all surfaces. Strongly adhered cells were estimated by comparing the fluorescence before and after washing (Figure 5). The peptide hydrogels showed three times higher cell-adhesive abilities compared with TCTP surfaces and comparable abilities with the fibronectin immobilized surface, which is a native adhesive protein surface. Interestingly, the adhered cell numbers on each peptide hydrogel were almost the same. Therefore, the content (percentage) of E1Y9-RGDS peptides does not have a major role in cell adhesive abilities, implying that some proportion of RGDS moieties might exist in the interior of the hydrogels. In other words, an E1Y9 peptide-alone hydrogel surface has an activity high enough for cell adhesion. The cell-adhesive ability might be derived from electrostatic interactions and/or hydrophobic effects between cell and peptide-gel surfaces. It was worth noted that hydrogels composed of charged peptide nanofibers have cell-adhesive abilities due to the electrostatic interactions with cell surfaces.^{28, 29}

Figure 5

Percentages of strongly attached 3T3-L1 cells. 3T3-L1 cells were incubated on the flat hydrogels composed of E1Y9 peptides containing 0, 10 or 20% E1Y9-RGDS (Ac-EYEYKYEYKY-Aeea-RGDS-NH₂) peptides, fibronectin (FN), tissue culture-treated plates (TCTP) or nonadhesive plate surfaces in Dulbecco’s modified Eagle’s medium containing 5 mM Ca²⁺. By comparing the amplified cell numbers and the cell numbers after washing, the percentages of strongly adhered cells were determined.

To use the high formability of the hydrogels for tissue engineering, cells were cultured on the constructed gel strings in non-adhesive microtiter plates. After 4 days of culturing the3T3-L1 cells, fluorescence microscopic analyses were performed (Figures 6a–c). Cells were homogeneously attached on all the peptide gel strings, which were several millimeters long. High-magnification images suggested that higher contents of E1Y9-RGDS resulted in increasingly high cell densities (Figures 6d–f). Similar differences in cell adhesion were also observed on the flat peptide hydrogels, as shown in Figure 5. Therefore, it was suggested that a portion of the RGDS moiety was effectively exposed on the hydrogel surfaces, leading to higher-density adhesions. That is, cell densities may be controlled by changing the mixing ratio of RGDS-peptide. When the cells were cultured in Ca²⁺-free phosphate-buffered saline (pH 7.4), sufficient cell adhesion did not occur due to lack of bivalent cations. Moreover, other cell lines (PC12), which can be differentiated to nerve-like cells, were also cultured on the hydrogels. After 24 h incubation, the cells were adhered on the gel strings (Figure 6g). After further culturing for 6 days with differentiation induction, neurite elongation of differentiated PC12 cells was observed (Figures 6h and ). These results strongly suggested that the novel peptide hydrogels composed of E1Y9 and E1Y9-RGDS peptides can be powerful scaffolds that can be formed into the shapes required for 3D culturing of various cell lines in controlled differentiation states.

Figure 6

Fluorescence microscopic images of cell-adhered peptide gel strings. Cells were cultured in Dulbecco’s modified Eagle’s medium containing 5 mM Ca²⁺. All scale bars represent 100 μm. (a–f) 3T3-L1 cells were cultured on the gel strings composed of E1Y9 peptides containing 0, 10 or 20% E1Y9-RGDS peptides. The contents of the gel strings are shown in the images. The cells were clearly adhered on the strings. Well-extended states were observed in the high-magnification images (d–f). (g–i) PC12 cells were cultured on the gel strings with differentiation induction for 24 h (g) and 6 d (days) (h, i). The cells were clearly increased with the incubation and were well differentiated.

Conclusions

We demonstrated Ca²⁺-induced formation of a peptide hydrogel by triggered cross-linking and subsequent self-assembly to nanofibers. The novel one Glu residue-conjugated β-sheet peptides, E1Y9, underwent hydrogelation on cue in response to Ca²⁺. The hydrogels were constructed by disentangled and wider nanofibers than the original Y9 nanofibers and E1Y9 nanofibers in the absence of Ca²⁺. The additional Glu residues in E1-E4Y9 peptides were effectively used for interaction with Ca²⁺, and the morphologies of the peptide nanofibers were changed. Introduced Glu numbers were important for the nanofiber morphologies following hydrogelation, and one Glu residue has the best ability to construct peptide hydrogels, as described above. Hydrogelation of the E1Y9 peptides was performed by the formation of a viscous sol state with a slower self-assembly process, followed by conjugation with Ca²⁺ solutions with a faster process. Therefore, the peptide hydrogels have the advantage of higher formability, which can be used to construct the desired shapes. In fact, gel strings can be easily created by the simple injection method. Cell culture experiments on the hydrogels were successfully performed with a low toxicity. Furthermore, the hydrogel was demonstrated to show adhesive abilities much higher than the cell culture plate and with a level similar to fibronectin surfaces. Different cell lines, 3T3-L1 and PC12, could be propagated and attached on the gel strings. Thus, the novel peptide-based materials will facilitate development studies for 3D cell culturing using various desired shapes, which will advance the field of novel tissue engineering.