Chlorite oxidized oxyamylose differentially influences the microstructure of fibrin and self assembling peptide hydrogels as well as dental pulp stem cell behavior

Tailored hydrogels mimicking the native extracellular environment could help overcome the high variability in outcomes within regenerative endodontics. This study aimed to evaluate the effect of the chemokine-binding and antimicrobial polymer, chlorite-oxidized oxyamylose (COAM), on the microstructural properties of fibrin and self-assembling peptide (SAP) hydrogels. A further goal was to assess the influence of the microstructural differences between the hydrogels on the in vitro behavior of human dental pulp stem cells (hDPSCs). Structural and mechanical characterization of the hydrogels with and without COAM was performed by atomic force microscopy and scanning electron microscopy to characterize their microstructure (roughness and fiber length, diameter, straightness, and alignment) and by nanoindentation to measure their stiffness (elastic modulus). Then, hDPSCs were encapsulated in hydrogels with and without COAM. Cell viability and circularity were determined using confocal microscopy, and proliferation was determined using DNA quantification. Inclusion of COAM did not alter the microstructure of the fibrin hydrogels at the fiber level while affecting the SAP hydrogel microstructure (homogeneity), leading to fiber aggregation. The stiffness of the SAP hydrogels was sevenfold higher than the fibrin hydrogels. The viability and attachment of hDPSCs were significantly higher in fibrin hydrogels than in SAP hydrogels. The DNA content was significantly affected by the hydrogel type and the presence of COAM. The microstructural stability after COAM inclusion and the favorable hDPSCs' response observed in fibrin hydrogels suggest this system as a promising carrier for COAM and application in endodontic regeneration.


Results
COAM did not modify fibrin microstructure but affected SAP hydrogels leading to fiber aggregation. The inclusion of COAM did not alter the microstructure of the fibrin hydrogel at the fiber level as demonstrated by atomic force microscopy (AFM) images ( Fig. 1A-D) and quantitative analysis (Table 1). However, SAP hydrogel microstructure (homogeneity) at the fiber level was affected by the inclusion of COAM (Fig. 1E-), leading to fiber aggregation (visible in Fig. 1F,H). SEM images further confirmed the microstructural stability of the fibrin hydrogels (Fig. 1I,J). In contrast, the effect of COAM inclusion on the morphology of the SAP hydrogels was not detected by SEM (Fig. 1K,L). The fiber height distribution showed no significant impact of COAM inclusion on fibrin hydrogels when comparing those without COAM ( Fig. 2A,B) to those with COAM (Fig. 2C,D). SAP hydrogels without COAM (Fig. 2E,F) showed a distinct fiber height distribution with three peaks between 1 and 5 nm. After COAM inclusion (Fig. 2G,H), those three peaks disappeared, confirming aggregation at the fiber level.
Topographic and quantitative microstructural analysis for the AFM images showed that both fibrin and SAP hydrogels have a nanofibrous structure at different scales ( Table 1). The roughness average (Ra) was 8.1 (SD: 1.6) nm for fibrin hydrogels and 1.2 (SD: 0.1) nm for the SAP hydrogels. The inclusion of COAM in fibrin hydrogels increased the Ra to 16.9 (SD: 7.5) nm. However, this increase was not statistically significant (p > 0.05). www.nature.com/scientificreports/   (Fig. 1F,H); therefore, these measurements were not reported.
Effect of COAM inclusion on fibrin and SAP hydrogel stiffness. The elastic modulus for fibrin hydrogels at 3.5 mg/ml fibrinogen concentration was 752 ± 13 Pa before and 683 ± 6 Pa after the inclusion of COAM (Fig. 3). Furthermore, for the SAP hydrogels at 3.5 mg/ml RADA-16 concentration, the elastic modulus was 5425 ± 295 Pa before and 4821 ± 386 Pa after the inclusion of COAM (Fig. 3). The stiffness of the SAP hydrogels was sevenfold higher than the fibrin hydrogels (p < 0.05). The inclusion of COAM did not alter the stiffness of the fibrin and SAP hydrogels (p > 0.05).
hDPSCs show higher viability and better attachment in fibrin hydrogels. According (Fig. 4F). hDPSCs adopted a spread morphology, indicated by a lower circularity value, in the fibrin hydrogels both with and without COAM (Fig. 4A,B). In contrast, hDPSCs remained rounded in the SAP hydrogels, again for the hydrogels both with and without COAM (Fig. 4C,D). These results indicate superior cell attachment in the fibrin hydrogels ( Fig. 4A-D,F). Furthermore, the hDPSC morphology did not change significantly with time for any of the treatment conditions (Fig. 4F). The average circularity scores were 0.25 ± 0.05 in the fibrin hydrogels without COAM and 0.26 ± 0.02 in the fibrin hydrogels with COAM over the 7-day test period. The average circularity scores were 0.76 ± 0.01 in the SAP hydrogels without COAM and 0.74 ± 0.02 in the SAP hydrogels with COAM over the 7-day test period. Post hoc comparisons using the Tukey test showed a significantly higher DNA content in response to the fibrin hydrogels with and without COAM compared to the SAP hydrogels with and without COAM (p < 0.05). hDPSCs showed a proliferative pattern in the fibrin hydrogels with an average 1.3-fold increase in DNA content at day 7 compared to day 0 for fibrin without COAM (p > 0.05) and a significant 2.1-fold increase for fibrin with COAM (p < 0.05).
A low DNA content was observed in the SAP hydrogels without COAM at day 0 with tenfold and eightfold lower DNA content compared to the fibrin hydrogels without COAM (p < 0.05) and SAP hydrogels with COAM (p > 0.05), respectively. SAP hydrogels with COAM showed higher DNA content than SAP hydrogels without COAM ranging between fivefold higher at day 1 and threefold higher at day 7. However, these differences were not statistically significant (p > 0.05). Moreover, the DNA content was stable in both SAP hydrogels from day 1 up to day 7 (Fig. 5).

Discussion
In this study, we characterized the structural and mechanical characteristics of selected compositions of fibrin and SAP hydrogels. Nano-to micro-scale structural and mechanical cues are associated with biological responses in both native ECM and synthetic constructs 18 . Although the understanding of the association between the surface topography and the cellular response is still limited, it has been suggested that the nano-topography enhances cellular communication, such as in neural cell networks 36 , and protein adsorption, thus affecting the modulation of cellular interactions 37 . In this study, the nanoscale topographical features of fibrin and SAP hydrogels were different. The roughness average of the fibrin surface was sevenfold higher than that of the SAP hydrogel, along www.nature.com/scientificreports/ www.nature.com/scientificreports/ with a twofold increase in the average fiber diameter. Other features, such as fiber straightness and alignment, were comparable. Fibrin hydrogels showed structural stability after the inclusion of our experimental macromolecule, COAM, while SAP hydrogels were affected, leading to fiber aggregation. The effect of COAM inclusion on the morphology of the SAP hydrogels was not observed in the SEM images, which could be due to the sample preparation procedure that resulted in drying artifacts. The measured stiffness of fibrin hydrogels, with the composition tested in the current study, was in the range of the stiffness of the native pulp tissue, which has previously been reported to be 800 Pa 38 . In addition, the stiffness of SAP hydrogels was found to be sevenfold higher than the fibrin hydrogels. hDPSCs are mesenchymal stem/ stromal cells (MSCs) that pose the potential to differentiate into numerous cell types in vitro, including odontoblasts/osteoblasts, chondroblasts, adipocytes, and neuronal-like cells [13][14][15] . MSCs have been shown to specify lineage and commit to phenotypes with extreme sensitivity to tissue-level elasticity, as soft matrices induced neurogenic differentiation and stiffer matrices were osteogenic 39 . The current mechanical characterization results might aid in explaining previous observations of mineralized tissue formation within SAP (RADA-16) hydrogels encapsulating hDPSCs when they were implanted in an ectopic mouse model for 12 weeks 30 . Future studies should explore to which extent differences in matrix stiffness would affect hDPSC differentiation profiles in vivo.
The current study showed higher hDPSC survival in fibrin hydrogels compared to in SAP hydrogels. This agrees with the data reported previously 17 , where hDPSCs in fibrin hydrogels prepared at 10 mg/ml showed higher viability compared to hDPSCs in SAP (RADA-16) hydrogels when they were evaluated using an MTT assay. Moreover, other authors 40 reported hDPSCs' survival at day 4 to be just above 60% in SAP (RADA-16) hydrogels at 1.5 mg/ml, reflecting the cell viability data reported in the current study. The higher cell survival in fibrin hydrogels can be likely explained by the presence of natural cell adhesion motifs 41 facilitating cell attachment and elongated cellular morphology, which was demonstrated in the current study. In contrast, SAP hydrogels lack these cell adhesion motifs. Future studies could explore the possibility of improving cellular attachment to SAP hydrogels by conjugating bioactive short peptide motives such as the integrin-binding arginine-glycineaspartic acid (RGD) to the C-terminus of the RADA-16 peptide. Another possible explanation for the low cell survival in SAP hydrogels could be the initial acidic pH (3.0) that is only neutralized after adding medium to induce gelation. The DNA quantification results confirmed differences in cell viability, with a tenfold lower DNA content measured at day 0 for SAP hydrogels compared to fibrin hydrogels. The DNA content in SAP hydrogels with COAM was eightfold and fivefold higher compared to SAP hydrogels without COAM at day 0 and day 1, respectively, suggesting an initial protective influence for COAM that needs to be further investigated. The hDPSC viability and the DNA content for the SAP hydrogels with and without COAM were then relatively stable over the remaining period of the experiments, strengthening the hypothesis that this drop in viability is related to the low attachment and the pH conditions at the time of encapsulation.
One interesting outcome was the effect of COAM on the increase of the DNA content, suggesting that the hDPSCs were proliferating more. Such effect for the presence of COAM may be explained in terms of a biological influence since no effect was observed for COAM inclusion on the structural and mechanical properties www.nature.com/scientificreports/ of the fibrin hydrogels. COAM is a polyanionic polysaccharide derivative with antibacterial properties 32 and broad-spectrum antiviral activity that acts as an immunomodulator 33,34 . In previous studies, COAM has been shown to induce and bind chemokines such as granulocyte chemotactic protein-2 (GCP-2), leading to significant recruitment of myeloid cells in mice 42 . Furthermore, it has been demonstrated that COAM competes with GAGs for binding and recruitment of chemokines 35 . This COAM-chemokine binding complex influenced chemokine localization and the selectivity of leukocyte responses and migration 35 . hDPSCs and MSCs produce a plethora of soluble factors, cytokines, and chemokines influencing cellular growth, proliferation, migration, differentiation, and immune responses 43,44 . For example, insulin-like growth factor-1 (IGF-1), a cytokine produced by hDPSCs, was found to stimulate hDPSC proliferation in serum-free culture medium 45 . Moreover, hDPSCs overexpressing the chemokine stromal-derived factor-1 alpha (SDF-1a/CXCL12) showed higher cell proliferation compared to wild-type hDPSCs 46 . Therefore, one possible explanation for the higher DNA content in fibrin hydrogels with COAM could be the formation of a binding complex, increasing the availability of factors and chemokines involved in cellular proliferation inside the 3D hydrogel microenvironment. This is in line with preliminary experiments, in which we found that COAM binds SDF-1a/CXCL12, both on solid phase and in solution (unpublished data). Finally, the current study presents a comprehensive structural and mechanical characterization for two promising biomaterials for dental pulp tissue engineering in combination with an analysis of biological features such as cell viability, shape, and proliferation. Future research will explore the influence of different hydrogel properties such as matrix stiffness on hDPSC migration and differentiation. Furthermore, the molecular mechanisms underlying the effect of COAM on hDPSC proliferation need to be investigated in detail in order to obtain insight to optimize their use in tissue engineering.

Conclusion
The microstructural stability after the inclusion of COAM as well as the preservation of cell viability, elongated morphology, and higher DNA content observed in the fibrin hydrogels suggest this system as a promising carrier for COAM and application in endodontic regeneration.

Materials and methods
All methods were performed in accordance with the relevant guidelines and regulations.

COAM synthesis, hydrogel composition, and preparation. Chlorite-oxidized oxyamylose (COAM)
was synthesized by two-step oxidation of amylose, purified, and fractionated according to molecular weight (MW) as described previously 33,47 . COAM was endotoxin-free and used as MW mixture (corresponding to protein molecular equivalent weights exceeding 100 kDa).
Fibrin hydrogels were prepared by mixing fibrinogen and thrombin components in equal volumes (pH 6.6), as described previously 22 . Plasminogen-depleted fibrinogen (Enzyme Research Laboratories, USA), derived from human plasma, was dissolved in 20 mM HEPES and 150 mM NaCl (fibrinogen buffer). Sterile stock solutions of thrombin (Sigma, USA), derived from human plasma, and factor XIII (Fibrogammin, CSL Behring, Germany) were prepared in 20 mM HEPES, 150 mM NaCl, 40 mM CaCl 2, and 0.1% BSA (thrombin buffer). Thrombin and factor XIII were mixed with the thrombin buffer and were kept in a water bath at 37 °C for 30 min to activate factor XIII to factor XIIIa. The control fibrin hydrogels were prepared at 3.5 mg/ml fibrinogen, 0.1 U/ml thrombin, and 0.1 U/ml factor XIII, whereas the test fibrin hydrogels were prepared at 3.5 mg/ml fibrinogen, 0.1 U/ml thrombin, 0.1 U/ml factor XIII, and 1 mg/ml COAM. SAP RADA-16 hydrogels were prepared according to the manufacturer's instructions by mixing the peptide solution (PuraMatrix Peptide Hydrogel, BD Biosciences, USA) with 20% sucrose solution followed by adding an equal amount of phosphate-buffered saline (PBS) for gelation. The control SAP hydrogels were prepared at 3.5 mg/ml RADA-16 peptide, whereas the test SAP hydrogels were prepared at 3.5 mg/ml RADA-16 peptide and 1 mg/ml COAM. www.nature.com/scientificreports/ Structural and mechanical characterization. Atomic force microscopy (AFM). AFM imaging was performed to characterize the microstructure of the different hydrogels at the fiber level. A 100 μl sample from each hydrogel composition (n = 3) was deposited on a silica sample holder and incubated at 37 °C for 30 min. After gelation, the hydrogel's top surface was carefully removed using gentle air blowing/drying to allow the imaging of the internal network. Agilent 5500 with MAC III controller and JPK Nanowizard 3 AFM systems were used for morphological imaging in intermittent contact mode in air. A sharp microlever probe MSNL-F (f = 120 kHz, k = 0.6 N/m, tip radius of curvature < 12 nm) was used. The AFM topography images were leveled, line-corrected, and measured (height and roughness profiles) using Gwyddion 48 . A fiber extraction algorithm, ct-FIRE 49 , was applied to the AFM images to characterize the fiber diameter, length, straightness, and alignment.
Scanning electron microscopy (SEM). A 100 μl sample from each hydrogel composition (n = 3) was prepared then fixed using 4% glutaraldehyde in PBS for 30 min. This was followed by drying in an ethanol series for the fibrin hydrogels and freeze-drying for the SAP hydrogels because the SAP hydrogels disintegrated in ethanol. Subsequently, the samples were attached to aluminum stubs and sputter-coated with a 5 nm thick platinum layer under vacuum. The microstructure was then observed using an XL30 FEG scanning electron microscope (Philips, Panama).
Evaluation of hydrogel stiffness. The stiffness of hydrogels of each composition (n = 3) was determined using a Chiaro Nanoindenter (Optics11, the Netherlands) by applying serial indentations with a spherical glass probe (r = 24.5 µm) attached to a flexible cantilever (k = 0.063 N/m). Loading and unloading velocities of the probe were set to 2 µm/s, with 2 s of holding period in between. For each sample, matrix scans (5 × 5 points) from three random locations were obtained. Load vs. displacement curves were extracted individually for each indentation point, and the Elastic Modulus (E) was calculated by using a Hertzian Contact Model (Poisson's ratio = 0.5) with Piuma Dataviewer Software (Optics11, the Netherlands), using Eq. (1): where F is the applied force, E is the elastic modulus, R is the radius of the probe, h is the indentation depth, and ϑ is Poisson's ratio.
Biological characterization. Primary cell cultures. Dental pulp tissues were acquired with informed consent from patients (15-20 years of age, male and female) undergoing extraction of third molars for therapeutic or orthodontic reasons as described previously 50 . Written informed consent was obtained from the patients or their parents, as approved by the medical ethical committee of Hasselt University, Belgium (protocol 13/0104U). The dental pulp tissue was harvested with forceps after mechanically fracturing the disinfected tooth with surgical chisels. Pulp tissues were then rinsed and transported at 37 °C in Eagle's Minimal Essential Medium, alpha modification (αMEM, Sigma-Aldrich, USA) supplemented with 2 mM l-glutamine (Sigma-Aldrich), 100 U/ml penicillin (Sigma-Aldrich), 100 μg/ml streptomycin (Sigma-Aldrich), and 10% fetal bovine serum (FBS, Gibco, Thermo Fisher Scientific, USA). hDPSCs were isolated according to the explant method and expanded in culture as described previously 50 . Cells were cultured in α-MEM, enriched with 10% heat-inactivated FBS (Biowest, Nuaillé, France), 2 mM l-glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin (Sigma-Aldrich). Only mycoplasma negative cells, screened with the PlasmoTest kit (InvivoGen), were used. All hDPSC cultures were tested for the expression of the following (stem) cell markers at the protein level by means of flow cytometry as described previously 50 : positive for CD29, CD73, CD90, and CD105 and negative for CD31, CD34, and CD45.
Evaluation of hDPSC viability. To obtain enhanced fluorescent protein (eGFP) labeled cells, pooled hDPSCs from three donors were transduced with a lentiviral vector encoding eGFP and a blasticidin resistance cassette.
The selection was performed with blasticidin (10 µg/mL, InvivoGen, Toulouse, France). Stem cells were used until passage 15. These labeled hDPSCs at 1 × 10 6 cells/ml seeding density were encapsulated in 100 μl hydrogels (n = 9) with and without COAM and deposited in a glass-bottom 96 well plate (CELLview Slide, Greiner, Austria). After gelation, an equal amount of maintenance culture medium was added (αMEM supplemented with 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 1% FBS). Tranexamic acid at 0.5 mg/ ml (Exacyl, Eumedica, Belgium) was added to the medium of fibrin hydrogels to prevent fibrin degradation. After 1, 4, and 7 days in culture, the nucleus of the cells was labeled with Hoechst 33342 (Invitrogen, USA), and the dead hDPSCs were labeled using the nucleic acid dye propidium iodide (PI) (Invitrogen) according to the manufacturer's instructions using an incubation time of 15 min at 37 °C. The images were collected using laser scanning confocal microscopy (LSM 880, ZEISS, Germany) using a 20 × objective (EC Plan-Neofluar 20 ×/0.50 M27). The fluorescence excitation/emission was measured at 490/552, 597/695, and 410/490 nm for GFP, PI, and Hoechst 33342, respectively. The number of live cells and dead cells were analyzed from 5 different regions per well (425 μm × 425 μm × 10 μm) in Fiji (Image J, National Institutes of Health, USA) 51 . Viability was calculated as a percent of live cells among the total number of live and dead cells.
Evaluation of hDPSC circularity (shape analysis). Live cells from 5 different regions per well (425 μm × 425 μm × 10 μm) were segmented using a combination of watershed segmentation, thresholding, and manual contour correction for cell boundaries. Shape (circularity) of segmented cells per region, excluding cells on the image edges, was analyzed using the particle analysis plug-in in Fiji (Image J, National Institutes of Health, www.nature.com/scientificreports/ USA) 51 . The circularity score was averaged for each well, yielding a final circularity score ranging between 0 and 1, where the closer the score to 1, the closer the shape to a circle, which would indicate low cellular attachment.
Evaluation of hDPSC proliferation (PicoGreen/Quant-iT DNA Quantification). hDPSCs at 5 × 10 5 cells/ml seeding density were encapsulated in 100 μl hydrogels (n = 3 per gel and per time point) with and without COAM and deposited in a 96 well plate coated with 50 μl of the same hydrogel devoid of cells or COAM (TPP Tissue Culture Plates, Sigma-Aldrich). Hydrogels were prepared devoid of cells as blank replicates. In addition, 2D controls of 5 × 10 3 cells were seeded in a 96 well plate. After gelation an equal amount of serum-free mesenchymal stem cell (MSC) medium (MesenCult-ACF Plus Medium, Stem Cell Technologies, Canada) supplemented with 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin was added to each well. At baseline (day 0) and after 1, 2, 3, and 7 days in culture, a PicoGreen/Quant-iT kit (Invitrogen) was used to investigate the effect of different hydrogels on cellular proliferation. The DNA content of three hydrogels per condition and per time point was calculated for three independent experiments. Fibrin hydrogels were first digested in a buffer composed of 50 FU/ml nattokinase in 5 mM EDTA in PBS for 2 h at 37 °C. SAP hydrogels were digested in a buffer composed of 1 mg/ml Pronase (Thermo Fisher Scientific) in PBS for 2 h at 37 °C. The cells seeded in the control wells were released using Trypsin-EDTA. The contents of the wells were collected, and cell pellets were retrieved by centrifugation. Retrieved cell pellets were then lysed to extract DNA using 100 μl cell lysis buffer composed of 0.029% Sodium EDTA, 0.112% Sodium pyrophosphate decahydrate, 0.88% Sodium chloride, 0.315% Tris HCl, 1% Triton-X-100, 0.038% EGTA, 0.0001% Leupeptin, 0.019% Sodium orthovanadate, 0.0216% β-glycerophosphate and 1 mM PMSF (ab152163, Abcam) and centrifuged at 14,000 rpm at 4 °C to collect the supernatant. A 200 μl working solution representing each well (hydrogel) was prepared and aliquoted directly into black 96-well plates (Chimney well, Fluotrac, Greiner, Austria), according to the manufacturer's instructions, and incubated for 5 min protected from light at room temperature. The fluorescence excitation/emission was measured at 481/520 nm using a microplate reader (Varioskan, Thermo Fisher Scientific). A standard curve was performed with λDNA, provided with the kit, and treated equally to the sample plates. The standards ranged from 10 ng/ml to 1 μg/ml λDNA and were used to calculate the final DNA content per ml of the sample.
Statistical analysis. Statistical analysis was performed using the statistical software package GraphPad Prism 8.00 (GraphPad Software, La Jolla California USA). Comparison of the fiber measurements from AFM images was performed using a one-way analysis of variance (ANOVA). Comparison of the stiffness of the hydrogels was performed using a two-way ANOVA. The influence of the different experimental conditions and the time factor on cell viability, shape, and DNA quantity was modeled using a three-way ANOVA. All ANOVA tests were followed by Tukey's correction for multiple comparisons. Statistical significance was determined at p < 0.05. Descriptive statistics are represented as mean and standard deviation (SD), or standard error of the mean (±), where appropriate.