Conjugation of Proangiogenic Peptide to Enhance a Soft Tissue Bioink

Abstract

Acellular methods are needed to vascularize 3D-printed tissue-engineered constructs for faster clinical translation. One method is to create bioactive materials that encourage migration and network formation of a patient’s own cells. This study utilizes QK peptide which replicates a binding sequence of vascular endothelial growth factor. This peptide has previously shown to maintain bioactivity when modified with an acrylate group and covalently bound to gelatin methacrylate (GelMA). However, this binding interferes with the crosslinking of the hydrogel matrix, requiring a characterization of the modified material. We investigated how binding QK peptide impacts GelMA crosslinking and material properties. A factorial design was employed to investigate the relationships between GelMA concentration, GelMA degree of substitution, and QK peptide concentration. These experiments were used to inform a rheology study, where the impact of QK peptide on crosslinking was investigated over a range of photocrosslinking times. We then investigated the printability of QK-GelMA bioink using rheology and a filament collapse test. The bioactivity induced by the peptide on the hydrogel surface was evaluated using endothelial cell network formation. QK peptide impacted key GelMA characteristics including swelling behavior, mesh size, and storage modulus through inhibition of temperature induced chain entanglement. We showed that while QK peptide affected the bioink on the nanoscale, QK-GelMA bioinks maintain high print fidelity and increase endothelial network formation on the surface of the hydrogel. Addition of QK peptide increases the vascularization potential of 3D-printed tissue-engineered constructs.

Introduction

Tissue engineering using 3D-printed biomaterials has the potential to create a wide variety of clinically relevant structures for enhancing, replacing, or repairing tissues. Additive manufacturing of biomaterials allows for complex architectures, patient individualization, automation, and repeatability. Advancements have been made in the last several years to create 3D bioprinted (3DBP) tissues from nearly every body system ^[1–3]. 3DBP tissues are often constructed from hydrogels due to their general biocompatibility and printability ^[4–6]. One obstacle to making these constructs a clinical reality is creating complex, functional vascular networks within at-scale tissues ^[7]. A 3DBP tissue requires sufficient vascularization for proper function and viability. A lack of sufficient vascularization throughout the architecture often leads to cell death and failure of the engineered tissue.

Previous vascularization approaches have included cells, bioactive factors, intentional scaffold design, or a combination of the above ^[8,9]. An array of cell types have been investigated for vascularizing 3DBP tissues, with focus largely on endothelial cells. However, limited expansion potential, patient isolation, donor variability, and tissue specificity makes selecting and obtaining cells challenging. Additionally, endothelial cells require a host of supportive cells to create functional vasculature, complicating design and delivery of cellularized 3DBP tissues. Scaffolds embedded with cells can provide a rudimentary vascular network ^[10]. However, concerns with culturing and maturation of encapsulated cells, patient immune response, and clinical translatability limit their near-future potential. While cellularized scaffolds are promising, acellular 3DBP scaffolds do not share the same concerns. Acellular constructs can utilize bioactive factors and engineered design to integrate with a patient’s own cells, facilitating faster translatability.

Acellular implants rely on vascularization by the surrounding tissue, a process promoted when signaling molecules are incorporated. This development of new blood vessels from preexisting vessels is called angiogenesis. Naturally derived 3DBP materials can inherently contain bioactive components that encourage angiogenesis through cell adhesion and migration. Modified materials can combine desired material properties to enhance proangiogenic behavior. Previous acellular bioinks have included bioactive molecules embedded within the material or bound to the material network. Types of modifications include ECM proteins^[11,12], ECM-derived peptides^[13,14], and signaling proteins^[10,15].

One such signaling protein, vascular endothelial growth factor (VEGF), is undeniably linked to migration and vascular restructuring of endothelial cells. QK peptide replicates the 17–25 α-helix region binding sequence of α1 VEGF-A.^[16] QK peptide is capable of inducing bioactivity levels similar to VEGF while being much smaller, easier to modify, uniquely thermally stable, and with higher selectivity ^[17–19]. QK peptide has been used in a variety of applications, both encapsulated ^[20] and covalently bound ^[21,22]. Prior studies have investigated QK peptide with a variety of materials, including but not limited to GelMA^[21], poly(ethylene) glycol^[22,23], decellularized extracellular matrix^[24], and elastin-like scaffolds^[25,26]. Acrylated QK peptide has previously been shown to induce encapsulated cell network formation when covalently bound to GelMA ^[21]. However, material properties and accessibility of the peptide could impact how endothelial cells behave on the hydrogel surface, potentially delaying endothelial cell response until hydrogel degradation releases the peptide. While the binding of acrylated QK peptide is a simple process with the addition of a photoinitiator, this modification directly impacts the hydrogel network crosslinking.

The amount of crosslinking within a GelMA hydrogel is directly tied to its physical properties such as swelling, mechanical strength, and pore size ^[27]. These characteristics in turn affect cell behavior including proliferation, migration, and adhesion ^[28]. Tissue formation and maturation relies on proper cell behavior and 3DBP materials should be well characterized to understand potential cellular implications for tissue-engineered constructs. To our knowledge, QK peptide covalently bound GelMA, or QK-GelMA as we will refer to it, has yet to be investigated as a bioink for developing 3DBP tissues.

The overall objective of this study is to characterize QK-GelMA bioink for future use as an acellular avenue for vascularizing 3DBP tissue-engineered constructs. The first aim is to delineate how the addition of QK peptide affects GelMA properties after crosslinking. The second aim is to identify printability of QK-GelMA bioink. The third aim is to investigate endothelial cell network formation on the surface of QK-GelMA. The factorial design in this study gives a basis of bioink properties allowing for specific applications with various tissue types and circumstances. The use of QK-GelMA within 3D-printable tissues provides greater potential for vascularization and successful integration.

Methods

Hydrogel Preparation

Sterile, lyophilized gelatin methacryloyl (GelMA) was purchased at two different levels of substitution (PhotoGel 50% DS, PhotoGel 95% DS, CELLINK). The degree of substitution for the lots used was verified using 1H-Nuclear Magnetic Resonance (NMR) spectroscopy. Briefly, samples were prepared at 25 mg/ml in deuterium oxide with 1 mg/ml of 3-(Trimethylsilyl)propionic acid-d4 sodium salt (TMSP, Fisher Scientific) as an internal reference. TopSpin was used for phase corrections, baseline correction, and chemical shift scale adjustment to the TMSP signal before analysis. The amount of methacrylation (Equation 1) and amount of total free amino groups (Equation 2) were used to calculate the degree of substitution (Equation 3):

$$\mathit{\text{Amount of methacrylation}}\left(\mathit{\text{mmole}}{g}^{-1}\right)=\frac{\int \mathit{\text{methacryloyl}}(\mathit{\text{peaks}}5.6-5.8\mathit{\text{ppm}})}{\int \mathit{\text{TMSP}}\left(0\mathit{\text{ppm}}\right)}\times \frac{9H}{1H}\times \frac{5.8048\mathit{\text{mmole TMSP}}}{25g\mathit{\text{GelMA}}}$$ (1)

$$\mathit{\text{Amount of total free animo groups}}\left(\mathit{\text{mmole}}{g}^{-1}\right)=\frac{\int \mathit{\text{methylene}}(\mathit{\text{peaks}}3.1-2.9\mathit{\text{ppm}})}{\int \mathit{\text{TMSP}}\left(0\mathit{\text{ppm}}\right)}\times \frac{9H}{2H}\times \frac{5.8048\mathit{\text{mmole}}\mathit{\text{TMSP}}}{25g\mathit{\text{GelMA}}}$$ (2)

$$\mathit{\text{Degree of substitution}}\left(\%\right)=\frac{\mathit{\text{Amount of methacrylation}}}{\mathit{\text{Amoung of the total free amino groups}}}\times 100$$ (3)

Hydrogel solutions were created by dissolving the lyophilized GelMA at the desired concentration (5, 10, or 15% w/v) in phosphate buffered saline. Photoinitiator lithium phenyl 2,4,6trimethylbenzoylphosphinate (LAP, Tocris) was added at 0.1% w/v. Acrylated QK peptide, Acetyl-{Lys(Acryl)}LTWQELYQL{Lys-Ac}Y{Lys-Ac}GI-Amide, (Genscript) was added at a concentration of 1, 10, or 100 μM and the solution was mixed until homogenous. Peptides had >85% purity. The hydrogel solution was then pipetted into molds with a height of 10 mm and a diameter of 5 mm. Each sample was UV crosslinked for 2 minutes in a clear mold.

Factorial Design

A 3×2×3 factorial design was used to assess the effect of GelMA concentration, GelMA degree of substitution, and QK peptide concentration (Table 1). GelMA concentration was selected over a range of common concentrations generally considered printable for extrusion 3D printing. GelMA is commercially available for purchase from CELLINK at a “medium” and “high” level of degree of substitution. QK peptide has been shown to elicit a response from endothelial cells at a variety of concentrations.

Table 1:

Factors and concentrations of the 3×2×3 factorial design. A total of 18 formulations were tested for swelling and mesh size, compression modulus, and enzymatic degradation rate via collagenase.

GelMA Concentration (weight/volume) (%)	GelMA Degree of Substitution	QK Peptide Concentration (μM)
5	N/A	1
10	Medium (Degree of methacrylation = 45–55%)	10
15	High (Degree of methacrylation >75%)	100

Swelling and Mesh Size Calculation

Mesh size is a measure of the linear distance between adjacent crosslinks. To determine the mesh size scaffolds were weighed post-cast, weighed after incubation for 48 hours at 37°C in PBS to reach equilibrium, and then lyophilized for 24 hours and weighed again to determine the dry mass.^[29] These measurements were then used to calculate the mesh size using the Flory-Rehner theory for polymers crosslinked in the presence of solvent. First, the equilibrium swelling ratio and relaxed swelling ratio were calculated using Equations 4 and 5:

$$Q_m=\frac{m_{\mathit{\text{post}}\mathit{\text{swell}}}-m_{dry}}{m_{dry}}$$ (4)

$$Q_{mr}=\frac{m_{\mathit{\text{post}}\mathit{\text{cast}}}-m_{dry}}{m_{dry}}$$ (5)

These values were then used to calculate the equilibrium volumetric swelling and relaxed volumetric swelling using Equations 6 and 7. The gelatin polymer density (${\rho }_p=1.35\text{g}/{\text{cm}}^3$)^[30,31] and the solvent density of PBS (${\rho }_s=1.01\text{g}/{\text{cm}}^3$)^[32] were used.

$$Q_v=1+\frac{{\rho }_p}{{\rho }_s}(Q_m-1)$$ (6)

$$Q_{vr}=1+\frac{{\rho }_p}{{\rho }_s}(Q_{mr}-1)$$ (7)

The equilibrium polymer fraction and relaxed polymer fraction were calculated with Equations 8 and 9:

$$v_{2s}=\frac{1}{Q_v}$$ (8)

$$v_{2r}=\frac{1}{Q_{vr}}$$ (9)

The Flory Rehner equation (Equation 10) was then used to calculate the molecular weight between crosslinks in the hydrogel^[33]. M_n is the number average molecular weight before crosslinking (63,565 g/mol)^[34], $\stackrel{-}{v}$ is the specific volume of the polymer (0.7407 mL/g)^[31], V₁ is the molar volume of water (18.01 mL/mol) and X₁ is the Flory’s Chi parameter (0.497)^[35]

$$\frac{1}{M_c}=\frac{2}{M_n}-\frac{\frac{\stackrel{-}{v}}{V_1}[\text{ln}\left(1-v_{2s}\right)+v_{2s}+X_1{v_{2s}}^2]}{v_{2r}[\frac{{v_{2s}}^{\frac{1}{3}}}{v_{2r}}-\frac{1}{2}\left(\frac{v_{2s}}{v_{2r}}\right)]}$$ (10)

The mesh size was then calculated using Equation 11. Here, l is the amino acid bond length (4.28 Å)^[30], M_r is the molecular weight of the repeat unit (91.19 g/mol)^[31] and C_n is the Flory Characteristic ratio (8.8785)^[31].

$$\xi ={v_{2s}}^{-\frac{1}{3}}*l*{\left(\left(2\frac{M_c}{M_r}\right)C_n\right)}^{\frac{1}{2}}$$ (11)

Compression Modulus

The compression modulus was used to evaluate how the addition of QK peptide affects the bulk mechanical properties of the hydrogel. Scaffolds reached swelling equilibrium in PBS for 48 hours. Each scaffold underwent compression testing using an Instron with a 1 kN load cell and a compression rate of 1.3 +/− 0.3 mm/min. The compression modulus of each was determined by calculating the slope of the linear region (4–6% strain) of the stress-strain curve.

Enzymatic Degradation via Collagenase

Each scaffold was placed in PBS containing 2 U/ml of collagenase type II (Worthington Biochemical) at 37°C. Collagenase solutions were replaced every 2 days to maintain activity of the enzyme. The percent degradation was calculated (Equation 12) using the initial weight and the final weight.

$$\%D=\frac{W_0-W_f}{W_0}\times 100\%$$ (12)

Rheological Analysis of Photocrosslinked Hydrogel

Both 50% DS and 95% DS GelMA were reconstituted at 5% (w/v) with 0.1% LAP and 100 μM of QK peptide. Control samples contained no QK peptide. GelMA was cast into cylindrical molds with 8 mm diameter and 2 mm height and UV photocrosslinked for either 15 seconds, 30 seconds, 1 minute, or 2 minutes. The ARES-G2 Rheometer outfitted with an 8 mm diameter parallel plate was used for an oscillation frequency test at 20°C with 1% strain from 0.1 to 10 Hz. The average storage modulus was calculated for each.

Rheological Analysis of Bioink

Based on the results from the factorial design experiment, only the 5% GelMA groups underwent rheological testing. GelMA at 5% (w/v) was reconstituted as above, excluding LAP, with 100 μM of QK peptide. Control samples contained no QK peptide. Rheology testing was performed with an ARES-G2 Rheometer (TA Instruments) and 40 mm diameter parallel plate. Each sample underwent three rheological tests. First, an oscillation frequency test was performed at 20°C with 1% strain from 0.1 to 10 Hz. Next followed a flow sweep at 20°C with a shear rate of 0.1 to 100 s⁻¹. Last was an oscillation temperature sweep from 40°C to 20°C with a 5°C step and a 30 second soak time at 5% strain and 1 Hz frequency.

Filament Collapse Printability Assessment

5% (w/v) GelMA with 100 μM of QK peptide was reconstituted, added to a print cartridge, and flash-frozen on ice for 5 minutes. Control samples contained no QK peptide. Samples were then placed in a temperature-controlled print head tool for 30 minutes in the BioAssembly Bot (Advanced Solutions). A platform with six pillars (length = 2.0 mm, width = 2.0 mm, height = 4.0 mm) was printed using a Bambu Lab X1 Carbon. Pillars were distanced at 1.0, 2.0, 4.0, 8.0, and 16.0 mm. Bioink filament was extruded bridging over the pillars using a 22-gauge conical nozzle, a speed of 6 mm/s, and an acceleration of 500 mm/s². Control and QK-containing inks of the same DS were printed at the same temperature, while pressure was varied to obtain an extrudable filament. 95% DS GelMA was printed at 24°C and 10.5 psi and QK-95% DS was printed at 9 psi. 50% DS GelMA was printed at 26°C and 10 psi while QK-50% DS was printed at 6 psi. Each filament was imaged, and the angle of deflection was measured between the pillar and the filament using ImageJ.

Cell Culture and Endothelial Cell Network Formation

Pooled human umbilical vein endothelial cells (HUVEC, Lonza) were cultured in tissue culture polystyrene flasks in Endothelial Growth Medium-2 (EGM-2, Lonza). Cells were grown to approximately 80% confluency and passaged using 0.25% trypsin/ethylenediaminetetraacetic acid. HUVECs were used at passage 4–5.

GelMA was reconstituted as previously described at 5% w/v at both 50% and 95% degree of substitution. QK peptide was added to test hydrogels at 100 μM at >95% purity with trifluoroacetic acid (TFA) exchanged for acetate (guaranteed <1% TFA level, Genscript). GelMA was added to 96-well tissue culture treated plates at 50 μL per well and UV crosslinked for 2 minutes. HUVECs were added on top of the gels at 6×10⁴ cells/cm². Plates were then incubated for 24 hours at 37°C. Cells were then fixed with 4% paraformaldehyde for 10 minutes, permeabilized with 0.25% Triton X-100 for 5 minutes, blocked with 2% bovine serum albumin (BSA) for 30 minutes, stained with Texas Red-X Phalloidin (1:400) in 1% BSA for 30 minutes, then stained with 4’,6-diamidino-2-phenylindole (DAPI, 1:1000) for 5 minutes. Nine images were taken of each well 20x and then stitched together. Images were then cropped to a centered circle to include only the cells in focus. Using Fiji, phalloidin stained images were made binary and then skeletonized. Skeletonized images were analyzed for number of branches, number of junctions, average branch length, and longest shortest path. DAPI stained images were also processed to be made binary, with overlapping nuclei resolved using watershed segmentation. Cells were automatically counted using the “analyze particles” Fiji function.

Statistics

All material experiments were conducted with n = 5 and cell experiments with n=3. GraphPad Prism was used for statistical analysis. The datasets were first assessed for normality using the Shapiro-Wilk test. Normal datasets were analyzed using ANOVA with Tukey’s multiple comparisons test. Non-normal datasets were analyzed using Kruskal-Wallis test with Dunn’s multiple comparisons test. Cell experiments were analyzed using Welch’s t test to compare the control and QK-containing group of each degree of substitution. A p-value < 0.05 was considered significant with *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

The results from the 3×3×2 factorial design resulted in seven factor effects: three main factor effects (GelMA concentration, GelMA degree of substitution, and QK peptide concentration), three 2-factor interaction effects, and one 3-factor interaction effect. To analyze the significance of the main and interaction effects the F-value, F-critical value, and p-value were calculated for each effect. An effect with a p-value < 0.05 was considered significant.

Results

Acrylated QK peptide can easily be bound to the GelMA network as it uses the same mechanism to bind as photocrosslinking of the hydrogel, depicted in Figure 1A. This covalent binding competes with the crosslinking of the GelMA network, requiring study to determine if this molecular-level interference would significantly impact material properties. A factorial design was employed for a comprehensive review of interactions between GelMA concentration, GelMA degree of substitution, and QK peptide concentration. A rheology study followed, to investigate material properties relevant during 3D extrusion printing. A filament collapse test assessed the printability of QK-GelMA bioinks. A network formation assay investigated the vascular promoting properties of QK-GelMA on the surface of the construct.

Figure 1:

Schematic of acrylated QK peptide conjugation to gelatin methacryloyl (GelMA) and GelMA 1H-NMR spectra. (A) Acrylated QK peptide will bind to acrylate groups on either methacrylate groups (ester containing) or methacrylamide groups (amide containing). Binding of QK peptide inhibits crosslinking within the GelMA matrix. Image created using ChemDraw. (B) Depiction of QK peptide distribution and hypothesized presentation on the hydrogel surface. In highly substituted GelMA, QK peptide may be more available on the surface and more evenly distributed than in lower substituted GelMA. QK peptide activates endothelial cells by interacting with vascular endothelial growth factor receptors (VEGFR). (C) NMR spectra of commercially produced GelMA with approximately 50% degree of substitution (DS) and 95% DS. Boxes show peaks of interest with A being the methacryloyl peak and B being the lysine group or free amino peak. Peak integral values and calculations for the DS can be found in Table 2. Image created using TopSpin.

1H-NMR was used to verify the degree of substitution of the purchased GelMA. Calculations from NMR showed the degree of substitution to be 53.7% for PhotoGel 50% DS and 138.6% for PhotoGel 95% DS. Calculations were performed as a ratio of free amino groups (Table 2). As methacrylation can happen at both amino and hydroxyl groups on a gelatin molecule, the degree of substitution can be over 100% when calculated as a ratio of the free amino groups that are discernable on the NMR spectra. The NMR spectra (Figure 1B) for the highly substituted GelMA had a lack of peak at 3 ppm and higher peaks at 5.6–5.8 ppm, showing a reduction of free amino groups and an increase in methylene groups when compared to the 50% DS GelMA. Degree of substitution was calculated to be as expected.

Table 2:

Degree of substitution (DS) calculation from NMR spectra. Calculations were made using the integral of the peaks of each range. 3-(Trimethylsilyl)propionic acid-d4 sodium salt (TMSP) was used as an internal standard.

GelMA DS Type	Methacryloyl (5.6–5.8)	TMSP (−0.05–0.5)	Methylene (3.1–2.9)	Methacrylation (mmole/g)	Free Amino (mmole/g)	DS (%)
50	0.0425	1.069	0.1583	0.0831	0.1547	53.70
95	0.1656	0.9985	0.2390	0.3466	0.2501	138.6

A swelling study was used to calculate the theoretical mesh size to quantify the effect of QK peptide on the nanoscale. The mesh size was calculated using the Flory Rehner theory for polymers crosslinked in the presence of a solvent ^[30]. As an overall trend, the mesh size decreased as GelMA concentration and degree of substitution increased (Figure 2A). Significant differences between QK concentration groups were seen in the 5% GelMA groups, with 100 μM of peptide increasing the mesh size of both 50% substituted (**p<0.01) and 95% substituted (**p<0.01) GelMA. A 10 μM concentration of QK increased the mesh of 5% 95% DS GelMA as well (*p<0.05). All main effect factors and interaction effects were found to be statistically significant. The main effect factors GelMA concentration (****p<0.0001), QK peptide concentration (**p<0.01), and degree of substitution (****p<0.0001) were all statistically significant. Interaction effects were significant as well (GelMA concentration and QK peptide concentration (****p<0.0001), GelMA concentration and DS (****p<0.0001), DS and QK peptide concentration (*p<0.05)). QK peptide caused significant differences in mesh size for 5% GelMA, but no major differences were seen at 10% or 15% GelMA.

Figure 2:

Material characterization of QK-GelMA using a factorial design. (A) Mesh size of GelMA cylinders calculated using the Flory Rehner theory. QK peptide caused significant differences within the 5% GelMA groups. Statistical analysis compared the impact of QK peptide concentration within each group. Normality of each group was tested by Shapiro-Wilk. Normally distributed groups were analyzed with one-way ANOVA with Tukey’s multiple comparison test while non-normal distributed groups were analyzed by Kruskal-Wallis with Dunn’s multiple comparison test. (B) Compression modulus of GelMA cylinders containing QK peptide. An increase in GelMA concentration and DS correlated with an increase in compression modulus. No consistent trends were seen with the addition of QK peptide. Statistical analysis was done within each group. Normality of each group was tested by Shapiro-Wilk. Normally distributed groups were analyzed with one-way ANOVA with Tukey’s multiple comparison test while non-normal distributed groups were analyzed by Kruskal-Wallis with Dunn’s multiple comparison test. (C-E) Enzymatic degradation of GelMA cylinders containing QK peptide. 5% GelMA (C), 10% GelMA (D), and 15% GelMA (E) all saw an increase in degradation time when DS was increased, while QK peptide had no discernible effect. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. n=5 for each group. DS, degree of substitution.

Compression testing was employed to test the impact of QK peptide on the overall GelMA scaffold modulus. Compression modulus had an overall trend of increasing with GelMA concentration and degree of substitution, as shown in Figure 2B. GelMA concentration (****p<0.0001) and degree of substitution (****p<0.0001) were found to be statistically significant along with one 2-way interaction effect (GelMA concentration and degree of substitution, ****p<0.0001) and the 3-way interaction effect (*p<0.05). QK peptide alone was only found to cause significant differences in compression modulus for 15% GelMA, 50% DS albeit with no consistent trend or behavior.

We quantified the enzymatic degradation time of QK-GelMA scaffolds. Figure C shows the degradation of the GelMA cylinders, ranging from day 1 to day 9. QK peptide had no significant effect on degradation time. All constructs with 50% DS had faster degradation time than their 95% DS counterparts. As expected, the constructs with the longest degradation time were 15% GelMA 95% DS. Increasing the GelMA concentration and degree of substitution resulted in a longer degradation time.

The storage modulus of 5% GelMA with two different degrees of substitution was quantified using rheology across four different photocrosslinking time points. Figure 3 depicts the trend across both 50% DS and 95% DS GelMA. QK peptide significantly decreased the storage modulus of 50% substituted GelMA at all time points tested (****p<0.0001). QK peptide had no significant impact on 95% DS GelMA. No statistical significance was seen between QK groups of different DS, except for at 1 minute (**p<0.01).

Figure 3:

Storage modulus of crosslinked QK-GelMA across a range of photocrosslinking times. A frequency sweep was used to calculate the average storage modulus. Groups were normally distributed and analyzed with two-way ANOVA with Tukey’s multiple comparison test. Comparisons were made within each DS group between the control and QK containing groups and between the matching groups of 50% and 95% degree of substitution (DS). **p<0.01, ***p<0.001, ****p<0.0001. n=5 for each group. DS, degree of substitution.

Shear thinning behavior is characterized by a decrease in viscosity with increasing shear rate and is required for an extrudable bioink. The shear-thinning behavior of QK-GelMA was characterized with a shear rate sweep (Figure 4A). All tested inks exhibited shear thinning behavior. QK peptide had no significant effect on shear thinning behavior of 50% DS GelMA. However, QK peptide decreased the viscosity of 95% DS GelMA at low shear rates, specifically at 0.25 s⁻¹ (***p<0.001) and below (****p<0.0001).

Figure 4:

Rheological evaluation of 5% GelMA bioinks with or without suspended QK peptide. (A) Shear rate sweep verified the shear-thinning properties of the bioink. 50% DS groups were assessed using Kruskal-Wallis test with Dunn’s multiple corrections. 95% groups were assessed using one-way ANOVA with Tukey’s multiple corrections. Significant differences in viscosity at low shear rates (<0.25 s⁻¹) were seen for 95% DS GelMA. (B) Storage and loss modulus was assessed using a frequency sweep. Storage moduli were normally distributed and 50% DS storage modulus and statistically assessed using one-way ANOVA with Tukey’s multiple comparison test. Loss modulus of 95% DS was non-normally distributed and assessed with Kruskal-Wallis test with Dunn’s multiple corrections. The storage modulus for 95% DS bioinks was significantly different with QK peptide (0.1–0.4 Hz **p<0.01, 0.63–10 Hz ***p<0.001). (C) Storage and loss modulus were assessed again using a temperature sweep. (D) Thermal gelation temperature calculated from the crossover modulus from the temperature sweep. Statistical analysis was performed using a Kruskal-Wallis test followed by a Dunn’s multiple comparisons test. *p<0.05, n=5 for each group. Normality of all groups were assessed using Shapiro-Wilk test. DS, degree of substitution.

Frequency sweeps can be used to measure the viscosity of hydrogels and bioinks across a range of shear rates. Storage (elasticity) modulus (G’) and loss (viscous) modulus (G”) for 5% GelMA with or without QK peptide is quantified in Figure 4B. The addition of QK peptide significantly decreased the storage modulus of 95% DS GelMA from 0.1–10 Hz (0.1–0.4 Hz **p<0.01, 0.63–10 Hz ***p<0.001).

We used a temperature sweep from 40°C to 20°C to identify the cross point between the storage and loss modulus, interpreted as the thermal gelation point of the GelMA bioinks. Results from the temperature sweep (Figure 4C) were used to calculate the thermal gelation point of the bioinks (Figure 4D). QK peptide did not have a significant impact on GelMA’s thermal gelation temperature. However, increasing the degree of substitution decreased the gelation temperature (*p<0.05), an important factor to note for 3DBP.

A filament collapse test was used to assess the printability of 5% (w/v) GelMA containing 100 μM QK peptide (Figure 5A). No statistical significance was found between the control (Figure 5B) and QK-containing groups (Figure 5C) of angles at each pillar position.

Figure 5:

Printability assessment via filament collapse test of 5% GelMA with and without 100 uM QK peptide. (A) Representative image of pillared platform with GelMA filament. Numbers in red represent the length of the gap (mm). L designates the left angle while R designates the right angle. (B) Angle of deflection measurements for 5% (w/v) 50% DS GelMA. (C) Angle of deflection measurements for 5% (w/v) 95% DS GelMA. Statistical analysis was performed with Kruskal-Wallis test with Dunn’s multiple comparisons test between the control and QK group for each pillar position. No statistical significance was seen between any control and the corresponding QK containing group. For all groups n=5. All replicates were performed on the same day with consistent environmental conditions. DS, degree of substitution.

Network formation of HUVECs in 2D culture on QK peptide-containing GelMA was assessed, shown in Figure 6. The number of branches, number of junctions, and cell count were not significantly different between control and QK-containing GelMA. For 50% DS GelMA average branch length and longest shortest path were not significantly different. However, for 95% DS GelMA average branch length (**p<0.01), and longest shortest path (*p<0.05) were significantly different (Figure 6C, F).

Figure 6:

Immunofluorescent staining for F-actin and nuclei to assess endothelial cellular network formation. (A) Cells were stained with Texas Red X-Phalloidin and DAPI. Images were binarized, skeletonized, and analyzed using Fiji. (B) The number of cells in a binarized DAPI image were automatically counted using Fiji. No significant differences in cell count were seen between the control and corresponding QK containing hydrogel for each type of GelMA. Fiji skeleton analysis quantified the extent of cellular network formation. (C) Highly substituted, QK peptide containing GelMA had significantly higher longest shortest path. The number of branches (D) and the number of junctions (E) were not significantly different between corresponding groups. (F) Average branch length was also significantly higher on highly substituted, QK-GelMA. The addition of QK peptide had positive impact on HUVEC 2D network formation on highly substituted GelMA. Statistical analysis was performed using Welch’s t test between the control and corresponding QK containing group within each degree of substitution. For all groups n=3. Scale bar is 50 μm. *p<0.05, **p<0.01. DAPI, 4’6-diamidino-2-phenylindole; DS, degree of substitution.

Discussion

This work details a comprehensive characterization of QK peptide covalently bound to GelMA for use as a bioink in 3DBP tissues. QK peptide has been shown to induce embedded endothelial cell network formation, a first step towards vascularization. As acrylated QK peptide binds to methacrylate groups along the GelMA molecule, it interferes with crosslinking within the hydrogel, depicted in Figure 1. Along with this change in crosslinking, steric effects from the addition of peptide may contribute to changes at the molecular level, such as an increase in mesh size or a decrease in storage modulus. This change in material properties requires characterization to better understand the bioink potential for creating 3DBP tissues. This characterization was done through addressing three aims: the first exploring how the addition of QK peptide impacts crosslinking of the hydrogel, the second investigating how QK peptide may affect hydrogel material properties pertinent to 3D extrusion bioprinting, and the third analyzing how bound QK peptide impacts cellular network formation on the surface of GelMA.

QK peptide impacted mesh size and storage modulus of crosslinked GelMA. Specifically, QK peptide had significant impact on mesh size of crosslinked hydrogel for 5% GelMA. Because of the QK binding mechanism we hypothesized that addition of peptide would increase the mesh size of GelMA. Smaller mesh size, and the resulting smaller pores, can inhibit cell proliferation and migration, negatively impacting complete integration of the hydrogel with a biological system ^[36]. QK peptide had a significant impact on the 5% GelMA groups, where an increase in QK peptide concentration correlated with an increase in mesh size. Average mesh size increased by approximately 4–5 nm from control to the highest peptide concentration for both types of 5% GelMA. This can be attributed to the reduced availability of methacrylate groups as crosslinking points on the 5% GelMA versus the 10% or 15%. We predict that if QK peptide concentration was increased above the experimental high of 100 μM, similar effects would be seen in the 10% and 15% GelMA groups. We also observed correlating mesh sizes between GelMA concentrations, demonstrating the impact of degree of substitution. Addition of acrylated QK peptide to the GelMA network affected the material at the nano-scale. While this not be relevant to the size and scale of a cell, this change could impact diffusion of nutrients through a scaffold. QK peptide also had a significant impact on the storage modulus of crosslinked 50% DS hydrogel across a range of UV cure times. Photocrosslinking can occur both during the 3DBP process and after the construct has been completed. Binding peptide significantly impacted 50% DS GelMA but not 95% DS GelMA. As this experiment was performed below the gelation point of GelMA (20°C), we hypothesize the higher storage modulus of 50% DS is due to the QK peptide inhibiting molecular chain entanglement and thermal-induced gelation. While controls between the DS groups all significantly differed, only the 1-minute time point saw a difference in storage modulus with peptide.

Addition of QK peptide had no impact on compression modulus. There is no clear trend in how QK peptide affects compression modulus within each group. However, overall compression modulus increased with GelMA concentration and DS, along with a notable increase in variability. We suggest this expansion of variability is due to the increased brittleness of the hydrogels due to heterogeneous crosslinking and entanglements, creating a denser but less uniform scaffold ^[37]. QK peptide appears to not affect the bulk compressive properties of GelMA. Bulk compressive testing may not be sensitive enough to capture the subtle differences in GelMA mesh size at the scale tested. QK peptide had no impact on enzymatic degradation time. GelMA is largely composed of gelatin, or denatured collagen, allowing enzymatic degradation via collagenase. Although in vitro and in vivo degradation can be quite different ^[38], this experiment allowed us to analyze the impact of QK peptide on macroscale material characteristics. QK peptide does not have a significant effect on enzymatic degradation of GelMA via collagenase. As expected, 50% substituted GelMA degraded before their 95% substituted counterpart. Highly substituted GelMA had a much wider range in complete degradation time, from 2 to 9 days.

Overall, binding of QK peptide had no significant impact on the printability of GelMA. We hypothesize that the length and type of modification, such as adding a poly(ethylene) glycol spacer, would result in more dramatic changes at the molecular level and influence printability but maintain bioactivity. Multiple rheology experiments were performed to assess material properties pertaining to extrudability and general printability. While addition of peptide resulted in a decrease in storage modulus and viscosity at low shear rates of 95% DS GelMA bioink, no other viscoelastic properties were affected. Shear-thinning describes the non-Newtonian behavior of fluids, where viscosity decreases under shear strain. This relates to the ease of extrusion and can impact cell viability. All inks exhibited shear-thinning behavior with small changes at low shear rates for highly substituted GelMA with the addition of peptide. A value of 10² to 10³ s⁻¹ is generally considered to simulate the shear rate inside the printing nozzle.^[39] As QK impacted viscosity at such low shear rates, minor changes could be seen in the initial pressure required to decrease the viscosity of the peptide-GelMA. Temperature sweeps are useful for characterizing gelatin based bioinks due to their thermal gelation properties. QK peptide had no statistically significant effect on the thermal gelation temperature of GelMA. However, increasing the degree of substitution caused an approximately 4°C decrease in gelation temperature. Replacing natural gelatin side groups with methacrylate causes a loss in gelation properties and reduces the degree of molecular entanglements. Due to the higher amount of methacrylation, the molecule loses gelatin-like properties thereby decreasing thermal gelation properties. Printability assessment via filament collapse test also showed no significant difference in printability with the addition of peptide. While QK containing inks required less pressure to create extrudable filaments, inks were still printable. This is likely a result of peptide inhibiting molecular entanglements, making it easier to extrude.

The addition of QK peptide impacted the 2D network formation of endothelial cells only on highly substituted GelMA. Previous work has verified that QK peptide covalently bound to GelMA encourages network formation of encapsulated HUVECs.^[21] However, activation of endothelial cells at the surface of a 3D printed construct may enhance rapid vascularization, rather than waiting until construct degradation and peptide release. This experiment aims to quantify if QK peptide enhances surface network formation. QK peptide positively impacted network formation in highly substituted GelMA and had no impact on 50% DS GelMA. The addition of QK peptide had no effect on the number of branches or junctions. While the complexity of the networks may not have increased, network connectivity was positively impacted on highly substituted GelMA. The longest shortest path and average branch length were significantly higher with the addition of peptide in highly substituted GelMA. Overall peptide availability could be causing this distinction between substitution levels of GelMA. As substitution increases, the potential binding sites for QK peptide also increases, as acrylated peptide binding occurs on the methacryloyl groups. This could cause a more uniform distribution of the peptide and subsequent higher surface presentation. The increase in cellular response could also be attributed to steric effects. The high crosslinking density of highly substituted GelMA may create steric effects that orient or extend tethered QK peptide toward the hydrogel surface, making it more accessible to cells. Additional exposure and release of the peptide is expected as the hydrogel degrades, causing further cellular activation. Further studies should investigate how degradation kinetics of GelMA increases peptide release and availability, increasing the network formation of cells.

In addition to peptide availability, material properties of the highly substituted GelMA may be responsible for affecting endothelial cell network formation. The higher crosslink density and resulting higher stiffness of 95% GelMA can affect cell adhesion and downstream signaling response for network formation and actin deposition ^[40]. The increase in methacryloyl groups could cause a reduction in competitive binding between natural gelatin domains and QK peptide. As more methacryloyl groups replace the amino or hydroxyl groups along the gelatin molecule, this could sterically “block” the naturally occurring cell binding motifs, leaving the QK peptide exposed.

QK-GelMA is better suited than plain GelMA or other extra cellular matrix (ECM)-mimicking hydrogels for supporting complex tissues over time. GelMA contains cell adhesion motifs from gelatin but lacks specific bioactive cues for angiogenesis and vascularization. QK-GelMA directly activates endothelial cell, with sustained local signaling due to covalent tethering of the peptide. The addition of QK peptide encourages vascularization, improving oxygen and nutrient delivery and negating necrosis, fibrotic encapsulation, and overall failure of a tissue engineered scaffold. As a GelMA scaffold is degraded, the rate cellular infiltration and ECM deposition must match to counteract the loss of structure and shape. Increasing nutrient and oxygen flow, resulting from QK peptide induced vascularization, can promote ECM deposition. The value of QK peptide in vivo has been explored.^[41] In a mouse cornea micropocket angiogenesis assay, bound QK increased the number of branch points and vessel density.^[22] QK has also been conjugated to thermoresponsive poly(polyethylene glycol citrate-co-N-isopropylacrylamide) and shown to improve mouse hindlimb ischemic limb tissue regeneration and function.^[42] Future in vivo work, including ongoing preliminary studies, will investigate the use of QK-GelMA for subcutaneous implants and long-term integration.

Further work is needed before clinical translation of QK peptide in 3D printed tissue engineered constructs. Sterilization, scalability, and cost of QK peptide must be considered. However, due to peptide size and relative complexity, scalability and cost will mostly likely be reduced when compared to VEGF or other large, complex growth factors. Future studies should also investigate the relationship between UV wavelength, intensity, and exposure time with QK peptide bioactivity. QK peptide contains a tryptophan and two tyrosine residues, making the peptide susceptible to oxidation or photoinduced modifications from UV light exposure during crosslinking. Previous studies have reported an increase in QK-scaffold bioactivity after crosslinking functionalized fibrous scaffolds with 10 mW cm⁻² for 10 minutes^[43], suggesting that bioactivity is maintained under standard photoinitiation conditions. Degree of substitution had a significant impact throughout experimentation. Lot and batch variability, from purchased or lab synthesized GelMA, should be considered and quantified for large scale studies.

There is a wide basis of knowledge for using methacrylated gelatin as a bioink for 3DBP. As QK peptide incorporation had minimal impact on printability, subsequent studies can utilize established printing parameters while increasing the bioactivity of the bioink.

Conclusions

The effects of QK peptide on GelMA as both a crosslinked hydrogel and extrudable bioink were characterized in this work. Acrylated QK peptide interferes with GelMA crosslinking, requiring analysis of this augmented material. The results showed that QK peptide increased the mesh size of hydrogels and inhibited thermal-induced gelatin chain entanglement. GelMA containing QK peptide proved to be a printable bioink. QK peptide enhanced network formation on the surface of highly substituted GelMA, potentially leading to the rapid invasion of implanted, 3D printed hydrogel constructs. As QK peptide has previously been shown to induce vascularization both in vitro and in vivo, addition of this peptide increases the vascularization potential of GelMA in tissue-engineered constructs. The range of GelMA concentrations and degrees of substitution combinations used in this study gives a solid basis for future work desiring specific material properties.