High-cytocompatible semi-IPN bio-ink with wide molecular weight distribution for extrusion 3D bioprinting

FTIR analysis of bio-inks after repeated heating/cooling circularly

Figure 2a, b shows the FTIR spectra in the percentage of transmittance for different concentrations of alginate-gelatin bio-ink over a spectral of 4000–200 cm⁻¹. Here, peak shifts, differences in peak shapes, and the appearance of new bands can be observed dependent on changes in the hydrogel composition³⁸. These main differences between SA/G hydrogels crosslinked with Ca²⁺ and SA/G hydrogels uncrosslinked reveal four peaks appearing at 1465 cm⁻¹, 1415 cm⁻¹, 1100 cm⁻¹ and 1050 cm⁻¹. The first two peaks at 1465 cm⁻¹ and 1415 cm⁻¹ are proposed to be the asymmetric and symmetric stretch vibration of –COO⁻ associated with carboxylic acid salts³⁹ and are specific to the ionic bonding. Moreover, the peaks around at 1100 cm⁻¹ relating to the C–C and C–O stretching, which considerably strengthens in the spectrum of SA/G bio-inks with Ca²⁺ ions, can be also attributed to the presence of crosslinking. The peak of C–C stretching (around 1050 cm⁻¹) shows a higher intensity, suggesting either a stronger O–H binding vibration or a stronger binding of the Ca²⁺ to the guluronic acids of alginate. There were no significant differences in other absorption bands between six different hydrogels. The characteristic absorption band over 3000 cm⁻¹ indicated the region of hydrogen-bonded –OH stretching or hydroxyl group⁴⁰. The absorption band at 1634 cm⁻¹ suggested that there was an interaction between negative charges of –COO^– groups in sodium alginate with the positive charges of –CONH₂ in gelatin⁴¹. The two main absorption bands lying between 1700–1600 cm⁻¹ and 1600–1500 cm⁻¹ were attributed to Amide I and Amide II, respectively. The Amide I band is attributed to C=O stretching vibrations of the peptide bonds, that were regulated by the secondary structure of protein such as the α-helix and β-sheet⁴². The Amide II band is associated with the C–N stretching vibrations and N–H bending of the amino acid. All samples modified by the repeated heating–cooling treatment showed no changes in wavenumbers for amide I and amide II peak positions indicating an intact primary protein structure of gelatin^40,43. Through heat-cool treatment, intramolecular hydrogen bonds are broken and replaced by hydrogen bonds to water molecules, thus preventing reconstruction of the inherent tertiary protein conformation^43,44.

Nevertheless, FTIR results showed that the primary structure of gelatin was intact and unaffected by dissolution temperature, allowing alginate cross-linking as previously described^43,45. Thus, the cyclic heating treatment did not destroy the main functional groups of gelatin and sodium alginate. In other words, the bio-inks were synthesized to contain the same proteins and polysaccharides to support the growth of the cells.

Thermal stability

TGA thermograms of SA/G-A, SA/G-B, and SA/G-C are shown in Fig. 2c, d.

The decomposition of SA/B-C is the fast in 0–200 °C (Fig. 2c) and then it started to degrade at a steady rate compared to the others over 200 °C (Fig. 2d). Nevertheless, their fast decomposition temperature of SA/G-A, SA/G-B, and SA/G-C all were 200 °C. This may be because the broken molecular chain decomposes most at 200 °C, but the long chain is rarely destroyed³¹. When the temperature rises again, the remaining long molecular chains are destroyed³¹. The number of short molecular chains in SA/G-C is the largest, so its initial degradation is very fast, and the final residual amount is almost the same^31,46,47. Thus, these results indicated that the thermal stability of gelatin/sodium alginate was obviously changed by cyclic heating–cooling treatment.

Molecular weight and its distribution

Figure 3a and Table 2 show the molecular weight and distribution of the three bio-inks. It can be seen from Fig. 3a that the molecular peak moved to the left and the distribution of copolymer molecular weight became broad with the increase of heating–cooling times. With the increase of heating and cooling times, the probability of molecular chain breakage of hydrogels increases⁴⁸. The molecular weight of the polymer has polydispersity, so the amorphous polymer has no clear viscous flow temperature, but a wide softening region⁴⁹. In this area, the polymer is easy to flow, which is conducive to processing and molding⁵⁰.

It can be seen from Table 2 that the Mw/Mn ratio of SA/G-C is the largest. That is to say, it has the largest dispersion and the widest molecular weight distribution. Previous studies have shown that the wider the molecular weight distribution is, the more sensitive the material is to shear stress²⁶. At a low shear rate, long molecular chains with wide distribution are easy to entangle⁵⁰. At this time, the wider the molecular weight distribution, the higher the viscosity; At a high shear rate, the entangled long molecular chain is easy to be destroyed. At this time, the wider the molecular weight distribution is, the smaller the viscosity is^25,27. Thus, when bio-ink is extruded, the cells in SA/G-C may experience the least shear force.

Rheological properties

The phase transition temperatures of SA/G-A, SA/G-B, and SA/G-C are shown in Fig. 3b. The transition temperature is the temperature of the sol–gel transition, which is determined by the cross storage modulus (G′) and the loss modulus (G″). The phase transition temperatures of SA/G-A, SA/G-B, and SA/G-C are 31.5 °C, 30.48 °C and 29.2 °C, respectively. This may be related to the change of molecular structure caused by heating treatment^21,51. Therefore, data on phase transition temperatures also confirm heating–cooling treatment may make the molecular chain of the material shorter.

Shear-thinning is the principle of SA/G bio-ink extrusion molding^25,46. As shown in Fig. 3e, the microscopic explanation of shear-thinning is that when the shear effect is stronger than Brownian motion, the entangled molecular chains tend to stretch, deform, disperse, and be ordered. The shear-thinning behavior is generally reversible. When the shear force disappears, the molecular aggregates are reformed due to its Brownian motion, that is, the chain-like colloidal molecules return to their natural position of non-orientation and return to the entangled state^21,25. Figure 3c shows the shear-thinning characteristics of bio-inks. It can be seen from Fig. 3c that the slope of SA/G-C’s curve is the smallest, followed by SA/G-B, and SA/G-A is the smallest. In other words, the shear sensitivity of these three bio-inks is SA/G-A<SA/G-B<SA/G-C. The rheological data verified the previous guesses in sections “Thermal stability” and “Molecular weight and its distribution”. The recovery behavior of hydrogel inks was especially important for the post-printing behavior of each printed scaffold¹⁷. Physically, recovery allows the bio-ink to rapidly increase in viscosity after extrusion and to maintain a high shape fidelity^17,43. All samples show excellent recovery to their initial viscosity over the 180–480 s periods, demonstrating all changes in polymer structure or properties as a result of exposure to the high bioprinting shear conditions over the 180 s periods^17,25,35,46. SA/G-C showed the fastest recoveries after application of the high shear rate compared to the SA/G-A as well as SA/G-B at first 180 s. This could also be attributed to wider molecular weight distribution in comparison to SA/G-A as well as SA/G-B^17,25,35,46. Thus, it is suitable for semi-IPN with more molecular weight distribution to be printed.

When the extrusion molding effect is the same, the shear force of the bio-ink with high shear sensitivity is smaller and the damage degree of cells is the smallest^17,25,35,46. When the shear force disappears, the curing speed of high shear sensitivity biological ink is faster and the molding precision is high^21,25,46.

Figure 3 revealed the relationship between molecular structure and rheological properties and hinted at their effect on cells suffered extrusion. The rheological properties showed that the shear sensitivity of the materials is controlled by the number of heating–cooling cycles. The rheological behavior of bio-ink is generally related to molecular weight and distribution, molecular chain structure, and softness^21,25,26,27. The cells in bio-ink with high shear sensitivity suffered less damage during printing.

Scaffolds morphologies

SEM analysis revealed the influence of molecular structure on microstructure morphologies of freeze-dried hydrogels after printing. Figure 4a–c displayed printed macropore structure of scaffolds, which represented similar print accuracy. The printability of hydrogels was not weaken by the preparation method of cyclic heating preparation. Cross-section of Scaffolds (Fig. 4d–f) showed a continuous and porous structure by virtue of the freeze-drying step, resembling other macromolecular hydrogel system structures^46,47. And each bio-ink exhibited porous microstructures with different chamber diameters, densities, and distributions. And the porosity of the scaffold was quantified by SEM of the scaffolds with imageJ software, as shown in Fig. 4g. The average porosity of the three scaffolds were 46.48%, 59.59% and 72.29%, respectively. With the increase of molecular weight distribution, the hole wall becomes thinner (Fig. S1). As previously reported in the literature, the porous structure of these hydrogels suggests their potential as scaffolds for cell infiltration, growth, and migration²⁶.

Mechanical properties

The mechanical properties of different 3D printing gelatin/alginate hydrogels with different heating–cooling times were also studied. Strain–stress curves of 3D printed scaffolds are shown in Fig. 4h. All scaffolds demonstrated good elasticity, with strains all greater than 80%. The fracture strain of scaffolds tended to increase with decreased heating–cooling times. As depicted in Fig. 4i, with increasing heating–cooling times, Young’s modulus was getting weakened. The fracture stress of SA/G-A, SA/G-B, and SA/G-C scaffolds were 0.08962, 0.08392, and 0.0524 MPa, respectively. A decrease in mechanical properties results from higher porosity and lower crosslink density, which were caused by molecular structure⁴⁶. The difference in molecular structure is the size and distribution of molecular weight, which is based on GPC data and FTIR spectra. Thus, molecular weight is the fundamental reason that affects the mechanical properties of scaffolds.

Studies have shown clear correlations between elastic moduli of hydrogel matrices and proliferation as well as differentiation of encapsulated cells⁴³. Compared to the results of SA/G-A and SA/G-B, significantly lower stiffness values were achieved in SA/B-C despite the same composition (sodium alginate and gelatin hydrogels), the same crosslinking agent as well as the same crosslinking time. Besides, the material preparation method leads to lower molecular weight and increasing molecular weight distribution according to the analysis in section “Materials and methods”⁴³. In conclusion, the method of heating–cooling decreases the mechanical properties, while leading to super shape fidelity and bioactivity.

In vitro cytocompatibility after bioprinting

Since cell activity directly affects the proliferation, differentiation, and protein expression ability of scaffold cells, it is important to study the effect of cell survival during and after printing²⁵. To determine the effects of bio-inks on cell behavior, the cell distribution, cell survival, and migration for 1 week were analyzed. Figure 5 shows the cell live/dead staining of SA/G bio-inks before printing and after printing. Before and after printing, the distribution trend of cells in semi-IPN bio-ink is the same. The distribution of cells in SA/G-A was the most uneven, which is far worse than that in SA/G-B and SA/G-C. This phenomenon is related to the viscosity of printing materials^34,35,43. Nevertheless, the distribution of cells after printing is relatively more uneven. High viscosity materials are easy to block the nozzle, resulting in nonuniform extrusion of materials²⁵. As the initial distribution of cells in construct made of SA/G-A is not uniform, the cells also showed uneven proliferation and diffusion.

Cytocompatibility of bioprinted semi-IPN bio-ink-containing cells was shown in Fig. 6. Cell survival rate at day1 increased significantly with increasing viscosity-shear sensitivity of bio-ink. The chondrocyte survival rates in the scaffolds made of SA/G-A, SA/G-B, and SA/G-C bio-inks were around 80%, 88%, and 95%, respectively (Fig. 6g). Indeed, fewer dead cells were observed in SA/G-C (Fig. 5f, i) compared with others. It confirmed our hypothesis that semi-IPN bio-ink with increasing molecular weight distribution can protect cells from stress and mechanical damage during the extrusion process^20,21,25,27, resulting from which semi-IPN bio-inks with increasing molecular weight distribution is sensitive to shear stress^17,25. SA/G-C can not only protect cells from mechanical pressure but also improve the printing accuracy.

In addition, A 3D view of the LIVE/DEAD staining on day4 and day7 after printing in Fig. 6a–f. A three-dimensional diagram showed that structures were “hollow”⁵² at day 7, especially Fig. 6f. The matrix strength of SA/G-C may be contributed to cell migration. The scaffold has appropriate matrix stiffness and high porosity, which is conducive to cell migration. The wide molecular weight distribution and low molecular weight resulted in a highly connected pore structure and appropriate matrix strength of SA/G-C³¹. This may be related to the loosening of the hydrogel network caused by the breakage of molecular chains. Recent studies have shown that the transport network is conducive to cell migration⁴⁷.

F-actin (red) and DAPI (blue) staining were used to evaluate the cell spreading and proliferation within the bioprinted constructs (Fig. 7). The cells diffused in situ and proliferated into spherical aggregates in the 3D bioprinted constructs at day 7 post-printing. As noted in Fig. 7, one cell in SA/G-C proliferated to four to seven cells after 7 days of culture, and there was no obvious proliferation for cells in SA/G-A. According to the staining data and quantitative analysis by image J, the cell proliferation rates in the three materials were 125%, 225%, and 475%, respectively (Fig. 7g). Cells in SA/G-C bio-ink showed fast proliferation. Broken molecules that are not involved in the cross-linking process dissolve, making room for cell proliferation^34,35,43.

Overall, SA/G-C bio-ink with a wide molecular weight distribution by cyclic heating–cooling treatment was prepared. And the relative molecular mass of this bio-ink became smaller, promoting a uniform cell distribution of cells; This semi-IPN bio-ink is more shear sensitive and can reduce the damage caused by the extrusion process and its interconnected porous structure provides space for cell proliferation^34,35,43.