Freeze-casting with 3D-printed Templates Creates Anisotropic Microchannels and Patterned Macrochannels within Biomimetic Nanofiber Aerogels for Rapid Cellular Infiltration

Abstract

A new approach is described for fabricating three-dimensional (3D) poly(ε-caprolactone) (PCL)/gelatin (1:1) nanofiber aerogels with patterned macrochannels and anisotropic microchannels by freeze-casting with 3D-printed sacrificial templates. Single layer or multiple layers of macrochannels are formed through an inverse replica of 3D-printed templates. Aligned microchannels formed by partially anisotropic freezing act as interconnected pores between templated macrochannels. The resulting macro/microchannels within nanofiber aerogels significantly increase pre-osteoblast infiltration in vitro. The conjugation of vascular endothelial growth factor (VEGF)-mimicking QK peptide to PCL/gelatin/gelatin methacryloyl (1:0.5:0.5) nanofiber aerogels with patterned macrochannels promote the formation of a microvascular network of seeded human microvascular endothelial cells. Moreover, nanofiber aerogels with patterned macrochannels and anisotropic microchannels show significantly enhanced cellular infiltration rates and host tissue integration compared to aerogels without macrochannels following subcutaneous implantation in rats. Taken together, this novel class of nanofiber aerogels holds great potential in biomedical applications including tissue repair and regeneration, wound healing, and 3D tissue/disease modeling.

Graphical Abstract

A facile approach is reported for engineering patterned macrochannels and aligned microchannels within aerogels by a 3D-printed sacrificial template-induced freeze-casting method. Such aerogels show faster cell infiltration and host tissue integration than the ones without patterned macrochannels after subcutaneous implantation in rats. This new class of aerogels hold great potential for many applications in biomedical and other fields.

1. Introduction

Aerogels, an emerging class of ultra-light, three-dimensional (3D), porous materials, are used in many fields including catalysis,^[1] chemical sensing,^[2] thermal insulation,^[3] energy storage,^[4] oil-water separation,^[5] drug delivery,^[6] wound healing,^[7] and tissue engineering.^[8] Traditionally, aerogels are prepared by wet chemical synthesis based on the sol-gel process.^[9] However, this approach has limitations in terms of solubility and covalent crosslinking at ambient temperature.^[10] For example, certain functional group (e.g., -OH, -COOH and -NH₂) containing precursors are necessary for gelation during the sol-gel process and covalent conjugation at ambient temperature.^[11] Moreover, many precursors need additional curing time and agents during the process in order to tune porosity and other material properties.^[12-15] Recently, freeze-casting suspensions of nano-/microscale materials has become a viable alternative approach to fabricate aerogels. This technique is dependent on many factors, including suspension concentration, type of suspending liquid, freezing temperature, and thermal conductivity of the freezing container.^[16] Compared to the sol-gel process, this method holds enormous potential for tuning aerogel porosity and anisotropy. During freezing, suspended nano-/microscale materials are forced into spaces between growing ice crystals. Once completely frozen, ice is removed by sublimation, leaving only aerogels.^[17,18] Other unique features can be incorporated by fabricating aerogels with carbon nanotubes,^[19] graphene oxide sheets,^[20] nanoparticles,^[21] nano-/microwires,^[22] nanofibers,^[23] or other specialized materials.^[24] Specifically, aerogels made short electrospun nanofibers have been subject to research in several innovative areas including pressure sensing,^[25] thermal insulation,^[26] protein separation,^[27] oil-water separation,^[28] bone regeneration,^[29] and tissue repair studies.^[30,31]

Due to their high surface area and porosities and overall biomimetic architecture, aerogels composed of short electrospun nanofibers have exhibited potential as scaffolds for wound healing and tissue regeneration.^[32] While the porosity of aerogels can be tuned within a limited range by changing the freezing temperatures during freeze-casting, the efficacy for tissue regeneration is often unsatisfactory, which may be attributed to the nonoptimal pore size and structure.^{[17, 18]} Nonoptimal pore size and structure can lead to poor cellular infiltration and heterogeneous cell distribution within the aerogels, thus inhibiting rapid cell movements and tissue regeneration.^[33,34] Scaffolds with channeled structures have enabled enhanced cell infiltration and uniform distribution of nutrients and oxygen in both computational simulations and experimental studies.^[35-38] We hypothesized that the creation of patterned macrochannels and aligned microchannels within nanofiber aerogels could promote passive transport of nutrients and metabolic wastes, promote rapid cellular infiltration, and induce angiogenesis. To test this hypothesis, we introduce a novel approach to generate patterned macrochannels and anisotropic microchannels within nanofiber aerogels using 3D-printed sacrificial templates during freeze-casting of nanofiber suspensions.

2. Results and Discussion

Fabrication of the aforementioned aerogels involves three major components: poly(ε-caprolactone) (PCL)/gelatin (1:1) short nanofibers, 3D-printed sacrificial templates made of alginate, and predesigned molds. Figure 1 shows the procedures for fabricating nanofiber aerogels with templated macrochannels and anisotropic microstructures. Initially, we prepared PCL/gelatin (1:1) short nanofibers with lengths ranging from 20-50 μm using our previously reported protocols.^{[39, 40]} Next, short nanofibers were homogenized in water (25 mg/ml) with 5% (w/w) gelatin relative to short nanofiber content. The addition of gelatin as surfactant to the short nanofiber solution enhances the stability of the suspension for further process. A predetermined volume of the short nanofiber suspension was added to a copper (Cu) mold containing 3D-printed alginate meshes (sacrificial template). Subsequently, the nanofiber suspension in the mold was immediately frozen using one of three methods: isotropic, anisotropic, or partially anisotropic freezing configurations. Frozen short nanofiber suspensions were lyophilized to obtain sacrificially-templated aerogels. The freeze-dried samples were removed from the molds and crosslinked with glutaraldehyde (GA) vapor overnight to strengthen their mechanical properties using Schiff-base chemistry.^{[39, 40]} Subsequently, the sacrificial alginate templates were removed by immersing the aerogels in a 50-mM ethylenediaminetetraacetic acid (EDTA) solution for 4-6 h. The obtained aerogels were sterilized with 70% ethanol or ethylene oxide prior to in vitro and in vivo studies. Nanofiber aerogels without templated macrochannels were prepared based on the same procedure, except without using 3D-printed sacrificial templates during freezing casting.

Figure 1. Schematic illustrating the creation of patterned macrochannels within biomimetic nanofiber aerogels based on 3D-printed sacrificial templates:

i) Producing PCL/gelatin (1:1) nanofiber mats through electrospinning; ii) Cryocutting of PCL/gelatin nanofiber mats into small fragments; iii) Adding crosslinker (gelatin) and homogenizing nanofiber segments in water; iv) 3D printing the sacrificial templates; v) Molding with 3D-printed sacrificial templates; vi) Freeze-drying of short nanofiber suspensions with sacrificial templates; vii) Crosslinking under GA vapor, and viii) Removal of sacrificial templates.

Engineering the structure of biomaterial scaffolds with macrochannels and/or microchannels may be ideal for enhancing cellular infiltration in tissue repair and regeneration.^[35-38] One technique to produce such materials, directional ice-templating, is widely used for generating short nanofiber aerogels with controlled porous architectures.^[29-31] In this study, we attempted to create both macro- and microchannels in nanofiber aerogels by combining the 3D-printed sacrificial template and freeze-casting process. Figure S1A shows a schematic representation of isotropic freezing, illustrating that cold ethanol (−80 °C) provides a spatially-uniform freezing environment surrounding the side and bottom of the Cu mold during freeze-casting.^[30] After freeze-drying, aerogels were crosslinked by GA vapor to enhance their mechanical stability like our previously reported procedure.^{[39, 40]} Figure S1B (i)-(vi) and Figure S1C (i)-(vi) show scanning electron microscopy (SEM) images of the horizontal and vertical sections of crosslinked PCL/gelatin (1:1) nanofiber aerogels without and with 3D-printed sacrificial templates fabricated using isotropic freezing. Nanofiber aerogels produced without 3D-printed sacrificial templates had random porous structures from both a horizontal and vertical view (Figure S1B (ii) and (v)). Conversely, nanofiber aerogels fabricated with 3D-printed templates had patterned macrochannels and few aligned microchannels in some regions (Figure S1C (i), (ii), (v)).

Figure S2A is a schematic representation of anisotropic freezing and illustrates how the side and top of the Cu mold was covered with a Styrofoam insulator. In anisotropic freezing, the temperature gradient is oriented such that freezing occurs from bottom-up in Cu molds. After freeze-drying, nanofiber aerogels were crosslinked under GA vapor. Figure S2B (i)-(vi) and Figure S2C (i)-(vi) show the horizontal and vertical sections of PCL/gelatin (1:1) nanofiber aerogels produced without and with 3D-printed sacrificial templates using anisotropic freezing. In this case, nanofiber aerogels fabricated without templates showed an anisotropic porous honeycomb structure in the horizontal view (Figure S2B (ii)), consistent with previous reports.^[27-29] In the horizontal cross sections, patterned macrochannels were visible in the nanofiber aerogels fabricated with a sacrificial template (after removing the template with the EDTA solution) (Figure S2C (i) and (iv)). In addition, the regions between the macrochannels also exhibited an anisotropic porous honeycomb structure (Figure S2C (i) and (ii)). Vertical microchannels were observed from bottom to top in the longitudinal cross-sections of both nanofiber aerogels fabricated with and without the template through the anisotropic freezing.

Isotropic and anisotropic freezing of short nanofiber suspensions generated two distinct structures between patterned macrochannels within nanofiber aerogels. Isotropic freezing resulted in partially aligned microchannels, while anisotropic freezing resulted in interconnected honeycomb structures. To further establish interconnections between patterned macrochannels within nanofiber aerogels, partially anisotropic freezing was utilized because the metal plate on the bottom of the mold has higher thermal conductivity than the surrounding air, resulting in a faster directional freezing rate. Figure 2A shows schematic representation of partially anisotropic freezing. In this setup, we removed the insulative Styrofoam from the anisotropic freezing setup but kept the cold metal plate. Figure 2B (i)-(vi) and Figure 2C (i)-(ix) show SEM images of the horizontal and longitudinal cross-sections of the 3D PCL/gelatin (1:1) nanofiber aerogels produced without and with sacrificial templates using partially anisotropic freezing. Horizontal cross-sections of aerogels fabricated without 3D-printed sacrificial templates showed an unordered structure (Figures 2B (i)-(ii)). The vertical view showed no obvious anisotropic features (Figure 2B (iv) and (v)). Figure 2C (i)-(iii) shows SEM images of horizontal cross-sections of the PCL/gelatin (1:1) nanofiber aerogels produced with templates. The single layer of the patterned macrochannels within aerogels represented the inverse replica of the 3D-printed sacrificial templates. The presence of aligned microchannels (green arrows) indicated a high degree of interconnectivity between macrochannels, though the surface of these macrochannels appeared to be disorganized (Figure 2C (ii) and (iv)). Orientation analysis further confirmed the anisotropic topography between the macrochannels as shown in Figure S3A-C. Along the vertical direction, the seemingly anisotropic microchannels were probably due to the lamellar structure (Figure 2C (vii) and (viii)) from the partially anisotropic freezing.

Figure 2. SEM images of 3D nanofiber aerogels with and without patterned macrochannels.

(A) Schematic illustrating the fabrication processes of nanofiber aerogels with and without one layer of patterned macrochannels. (B i-vi) SEM images of horizontal and longitudinal cross-sections of 3D PCL/gelatin nanofiber aerogels without macrochannels. (C i-ix) SEM images of horizontal and longitudinal cross-sections of 3D PCL/gelatin nanofiber aerogels with one layer of patterned macrochannels. The green arrows indicate microchannels inside aerogels.

It is well-known that directional freezing can induce ice crystal formation along a temperature gradient and eventually lead to the formation of anisotropic structure during freeze-casting. ^{[18, 41, 42]} Based on previous studies and our own observations,^[18] we speculate the following mechanism for formation of nanofiber aerogels with patterned macrochannels and anisotropic microchannels (Figure S7A). Evidently, macrochannels are formed through inverse replication of 3D-printed sacrificial templates. During ice-templating, if short nanofibers are rejected by the solidification front, the interfacial energy satisfies the following equation:

$$\Delta {\gamma }_0={\gamma }_{fs}-({\gamma }_{fl}+{\gamma }_{sl}+{\gamma }_{ff})$$ (1)

Where γ_fs, γ_fl, γ_ff, and γ_sl are the interfacial free energies associated with the nanofiber-solid, nanofiber-liquid, nanofiber-nanofiber, and solid-liquid interfaces, respectively.^[18,43] Short nanofibers experience both repulsive (F_R) and attractive forces (F_A) from the advancing freezing front arising due to van der Waals interactions at the liquid-solid interface and viscous drag.^{[18, 44]}

$$F_R=2\pi r\Delta {\gamma }_o{\left(\frac{{\alpha }_0}{d}\right)}^{{}_{{}_n}}$$ (2)

$$F_A=\frac{6\pi \eta v{r}^2}{d}$$ (3)

Where r is the radius of the short nanofiber, v is the freezing velocity, α_o is the mean distance between the molecule in liquid phase, d is the thickness of the liquid layer between the solid–liquid interface and the short nanofiber (i.e., the distance between the ice front and the short nanofiber), η is the dynamic viscosity of the liquid and n is an empirical correction factor for the repulsive forces that generally ranges from 1 to 3.^[18]

Considering ice front velocity (V) and critical freezing front velocity (V_cr), when V<V_cr, short nanofibers are generally rejected and form lamellar walls within the freeze-casting scaffold, and for V≥Vcr (Figure S7A (ii)), a certain fraction of short nanofibers would generally be entrapped by ice crystals and create bridges between lamellar ice walls, yielding microscale porosity throughout bulk aerogel structure.^[18]

We speculated that inclusion of 3D-printed sacrificial templates may impart new thermal fluxes and thus alter uniform formation of lamellar ice walls during freeze-casting. Figure S7A (iii) shows that short ice plate crystals formed from the bottom of the mold to the templates and in between the templates. SEM images in Figure 2C (v) and Figure S4 clearly illustrate the presence of a lamellar structure during freeze-casting. The green arrows in the Figure 2C clearly illustrate the lamellar structure in both vertical and horizontal cross-sections. When V≥V_cr, nanofibers are entrapped within the ice crystals and form short nanofiber bridges between macrochannels, which is clearly seen in the templated aerogels. In addition, Figure S7B (i-ii) and Figure 2C (ii), (iii), (v), and (vi) show an ice plate crystal growth in between the 3D-printed sacrificial templates during freeze-casting. Formation of aligned microchannels between macrochannels occurred when V≥V_cr, because the nanofiber bridges between the lamellar structures were obvious, as shown in Figure 2C (ii) and (v). However, no such anisotropic microchannels were observed in aerogels fabricated with isotropic or completely anisotropic freezing with or without sacrificial templates. The results of SEM images are consistent with the above speculations, especially the short ice plate crystal formed between templates and template grids. We report for the first time the formation of anisotropic microchannels between patterned macrochannels within nanofiber aerogels by freeze-casting.

To create layers of macrochannels within aerogels, we stacked two or three 3D-printed sacrificial templates during freeze-casting. Figure S4A and S4B shows SEM images of the vertical cross-section of PCL/gelatin (1:1) nanofiber aerogels with two and three layers of templated macrochannels. Figure S4A (ii) and S4B (ii) shows the corresponding high-magnification images of Figure S4A (i) and S4B (i), indicating the formation of ordered microchannels between layers of templated macrochannels. A 3D-printed ring-shaped template was also used to demonstrate the versatility of macrochannel patterns (Figure S5). We also prepared functional aerogels with methacryloyl functionality. Figure S6 shows that PCL/gelatin/gelatin methacryloyl (GelMA) (1:0.5:0.5)-composed short nanofibers for tethering with functional peptides or molecules, using our previously reported procedure.^[39] One shared feature of each freeze-casting technique was the formation of randomly-oriented and dense structures near the bottom of the Cu molds, which could be due to rapid freezing rates, as these results are consistent with previous reports.^{[41, 42]} Among the three freeze-casting methods (e.g., isotropic, anisotropic and partially anisotropic) used, partially anisotropic freeze-casting was the only method that can form aligned lamellar microchannels between the templated macrochannels in the horizontal cross-sections of the resultant aerogels. In contrast, the microchannels in aerogels produced by the isotropic freezing were mainly disorganized and the regions between the macrochannels of aerogels produced by the anisotropic freezing process exhibited an anisotropic porous honeycomb structure. The aerogels with the interconnected microchannels between patterned macrochannels could promote the cellular infiltration compared with the ones having the honeycomb or disordered microchannels between patterned macrochannels. Therefore, we chose the aerogels fabricated by the partially anisotropic freezing for further in vitro and in vivo studies.

To examine the effect of patterned macrochannels and anisotropic microchannels on cell proliferation and infiltration, MC3T3-E1 pre-osteoblasts were seeded on nanofiber aerogels with and without patterned macrochannels, cultured for 1, 3, and 7 days, and subsequently stained for LIVE/DEAD assay analysis (Figure S8). Cell proliferation occurred only on the surfaces of aerogels without templated macrochannels. In contrast, cells proliferated throughout aerogels with templated macrochannels. Cell proliferation was quantified using a CCK-8 assay (Figure S9). Significantly more cells (higher optical density (OD) value at 450 nm) were observed in aerogels with patterned macrochannels compared to aerogels without macrochannels on day 7. These findings indicate that patterned macrochannels might provide a route for uniform nutrient supply throughout the aerogels, which is consistent with previous studies.^[35-38] Additionally, the cytoskeletons and nuclei of MC3T3-E1 cells were stained with Alexa Flour 647 phalloidin in red and 4’,6-diamidino-2-phenylindole (DAPI) in blue (Figure S10A). Cell imaging indicated rapid migration and proliferation of MC3T3-E1 cells in templated nanofiber aerogels at different time intervals under static culturing conditions. MC3T3-E1 cells seeded on templated nanofiber aerogels showed rapid cell migration through the macrochannels and occupied the inner space of the aerogels, while those seeded on non-macrochanneled aerogels showed proliferation only on the aerogel surfaces. Figure S10C and D, and Figure S11 images shows that templated macrochannels provide migratory pathways for cells, and 3D reconstruction of confocal images show that cell nuclei occupied both macrochannels and inner space of the aerogels. To quantify cell migration, we measured migration based on 3D confocal images (Figure 3). MC3T3-E1 cells migrated ~200 μm inside templated aerogels on day 1, ~230 μm on day 3, and ~350 μm on day 7 (which is almost 50-60% of the aerogel thickness). In contrast, cells migrated a significantly shorter distance (~60 μm on day 1 and day 3 and ~100 μm on day 7) on aerogels without templated macrochannels (merely 10% of the aerogel thickness). These results clearly reveal that macrochannels and structural anisotropy provided sufficient room and orientation for cell migration and proliferation.

Figure 3. MC3T3-E1 cell migration on PCL/gelatin short nanofiber composed 3D aerogels without and with patterned macrochannels.

(A) Depth index analysis of the migration of MC3T3-E1 cells through the aerogels with and without patterned macrochannels by measuring the depth of the cells on the aerogels in μm from the 3D confocal images. (B) The migration of the cells through the aerogels with and without patterned macrochannels quantified by measuring the depth index of the cells on the aerogels in μm. Aerogel without macrochannel: 3D PCL/gelatin nanofiber aerogels without patterned macrochannels. Aerogel with macrochannel: 3D PCL/gelatin nanofiber aerogels with patterned macrochannels. Data represents mean ± standard deviation (n = 3). The significance was calculated with an ordinary two-way ANOVA followed by a post-hoc Tukey’s comparison test within the group, ****p < 0.0001.

In tissue repair and regeneration, the vascularization has a significant impact on the longevity of transplanted cells and rapid tissue integration.^[45] Vascular endothelial growth factors (VEGF) play a critical role in angiogenesis and microvasculature formation. QK peptide is a VEGF-mimicking peptide consisting of 15 amino acids which can enhance the angiogenesis by activating the VEGF receptors.^[39,46] To demonstrate the feasibility of incorporating signaling molecules, we prepared PCL/gelatin/GelMA (1:0.5:0.5) nanofiber aerogels with macrochannels (Figure S6) that exhibited similar structures as the PCL/gelatin (1:1) nanofiber aerogels. Then, OCTAL-tethered QK (QK-OCTAL) peptides were conjugated with the PCL/gelatin/GelMA nanofiber aerogels with templated macrochannels following our previously established protocols.^[39] To examine induced microvascularization, we seeded 1×10⁴ human microvascular endothelial cells (HMECs) onto QK peptide-conjugated PCL/gelatin/GelMA nanofiber aerogels with and without patterned macrochannels. Figure S12A shows endothelial cell marker CD31 (in pink) and nuclei (in blue) staining of HMECs on aerogels with and without QK-OCTAL conjugation after 3 and 7 days of culture. No significant microvascular tube formation was observed on aerogels with and without patterned macrochannels on day 3. In contrast, endothelial tubule roots were observed on QK-OCTAL-conjugated aerogels with patterned macrochannels on day 3 and dense microvascular networks were formed after 7 days. A similar degree of cell infiltration was observed for HMECs seeded on QK peptide-conjugated PCL/gelatin/GelMA templated nanofiber aerogels (Figure S12B). Cross-sectional confocal microscopy images (Figure S12B) showed cells also migrated through the templated macrochannels.

Generally, the rapid cellular infiltration is critical for tissue repair and regeneration. However, the cellular infiltration efficacy of the scaffolds like hydrogels and 2D nanofiber mats is very poor due to their limited porosity and small pores. Towards this end, the significance of microchannel in the cell-laden hydrogel was investigated.^[37] It was found that the hydrogel constructs containing channels were provided a way for the active transport of both nutrient and oxygen to improve cell viability when compared with the hydrogel construct without channels. Recently, an effective approach for the fractal design and optimization of bifurcated channels was reported in a cell-seeded anisotropic scaffold for oriented tissue engineering.^[38] Therefore, channels in scaffolds played a significant role in promoting cellular infiltration, cell growth, and vascularization.

Inspired from these studies, to further examine the cellular infiltration in vivo, we subcutaneously implanted PCL/gelatin nanofiber aerogels with and without patterned macrochannels (1-mm-thick, 8-mm-diameter) in rats for 2 and 4 weeks (Figure S13 and Figure 4A). After 2 weeks of implantation, hematoxylin and eosin (H & E) staining results showed that host cells infiltrated very limited depth into the aerogels without templated macrochannels, and a large cell-void area where the aerogel was implanted was observed (Figure 4A (i) and (v)). In contrast, host cells penetrated much deeper into the nanofiber aerogels with templated macrochannels, though several small cell-void areas were noticed (Figure 4A (iii) and (vii)). After 4 weeks of implantation, superficial penetration of host cells was seen on aerogels without patterned macrochannels and a large cell-void area remained (Figure 4A (ii) and (vi)). Conversely, host cells completely penetrated throughout the templated aerogels (Figure 4A (iv) and (viii)). Deformation of nanofiber aerogels without macrochannels occurred 4 weeks after implantation, likely due to mechanical compression and the large cell-void area (Figure 4A (ii)). Aerogels with patterned macrochannels were able to maintain their shapes after 4 weeks after implantation, which might be attributed to the rapid cellular infiltration and extracellular matrix (ECM) deposition, leading to increased stability from new tissue formation (Figure 4A (iv)). The area of cellular infiltration or migration in aerogels was quantified based on H & E staining images (n=4) at each time point (Figure 4B). Nanofiber aerogels with patterned macrochannels showed 70% cellular infiltration at week 2 after implantation, and reached 100 % at week 4. In comparison, the aerogels without templated macrochannels had only 10% cellular infiltration after 2 weeks of implantation, but reached roughly 30% after 4 weeks. Templated macrochannels enhanced cellular infiltration and formation of new tissues entirely from host cells. Additionally, Masson’s trichrome revealed collagen deposition and angiogenesis within the aerogels (Figure 4C). Collagen deposition formed around non-templated aerogels (indicated by green arrows) (Figure 4C (i)-(vi)), indicating the occurrence of fibrosis a foreign body response.^[47] However, collagen deposition and blood vessel formation were observed inside templated aerogels 2 and 4 weeks after implantation (Figure 4C (iii)-(viii)), suggesting the patterned macrochannels can minimize the foreign body response and enhance the biocompatibility of implanted nanofiber aerogels, which is critical for tissue regeneration. The number of multinucleated giant cells per aerogel was shown in Figure S14A. The number of multinucleated giant cells was higher in the aerogels with microchannels as compared with aerogels without macrochannels. The aerogels without macrochannels were surrounded by giant cells, while giant cells were distributed throughout the aerogels with macrochannels. This could be due to the fast cell infiltration via patterned macrochannels while not possible in aerogels without macrochannels. The number of blood vessels per aerogel was quantified at each time point (Figure S14B). More blood vessels formed at week 4 in the aerogels with macrochannels than aerogels without microchannels. A similar strategy has been proposed by Ratner to modulate the porous structure of implants for reducing and eliminating foreign body reaction.^[48] Figure 4D shows a schematic illustrating the difference of cellular response between nanofiber aerogels with and without patterned macrochannels in vitro and in vivo. Cells mainly grew on the surface layer in vitro and only limited depth of cell penetration was observed after subcutaneous implantation for 4 weeks for nanofiber aerogels without patterned macrochannels, while the cells could migrate and proliferate on both the surface and inside in vitro and the host cells penetrated throughout the whole aerogel after implantation for 4 weeks for nanofiber aerogels with patterned macrochannels. According to the H&E or trichrome staining images the size decreased to 4 - 5.5 mm which could be due to the shrinking/deformation of aerogels caused by the degradation of gelatin portion under the skin and the deviation of the direction from the exact horizontal direction during the cryocutting of embedded tissues. These results indicate that templated nanofiber aerogels hold enormous potential in tissue repair and regeneration because they facilitate rapid formation of new tissues by enabling cell infiltration, ECM deposition, and angiogenesis after implantation. Additionally, these aerogels can be functionalized with signaling molecules to further regulate the response of surrounding cells for better host cell recruitment, ECM production, and angiogenesis.

Figure 4. Subcutaneous implantation of 3D PCL/gelatin nanofiber aerogels with and without patterned macrochannels in rats for 2 and 4 weeks.

(A) H & E staining of 3D aerogels with and without patterned macrochannels and their surrounding tissues. (B) Percentages of cell-infiltrated areas within aerogels with and without patterned macrochannels after subcutaneous implantation. (C) Masson’s trichrome staining indicating collagen deposition and neovascularization within subcutaneously implanted 3D aerogels. (D) Schematics illustrating cell distributions on aerogels without and with patterned macrochannels after cell seeding in subcutaneous implantation in rats, suggesting significantly improved cellular infiltration and neovascularization within aerogels having patterned macrochannels. Aerogel without channel: 3D PCL/gelatin nanofiber aerogels without patterned macrochannels. Aerogel with channel: 3D PCL/gelatin nanofiber aerogels with patterned macrochannels. White arrows indicate the area of aerogels with cellular infiltration. Orange arrows indicate the area of aerogels without cellular infiltration. Green arrows indicate the deposited collagen. Red arrows with yellow stroke indicate blood vessels. Yellow dashed lines represent the boundary between the tissue and the area without cellular infiltration. Data represents mean ± standard deviation (n = 4). The significance was calculated with an ordinary two-way ANOVA followed by a post-hoc Tukey’s comparison test within the group **p < 0.01

3. Conclusions

In summary, we demonstrated a simple approach for engineering biomimetic nanofiber aerogels with patterned macrochannels and anisotropic microchannels. Patterned macrochannels created within the nanofiber aerogels served as a migratory pathway for cells, nutrients, and oxygen. GelMA-containing nanofiber aerogels allowed for conjugation with QK peptides that can regulate microvascular network formation of seeded endothelial cells. Most importantly, the nanofiber aerogels with patterned macrochannels could reduce foreign body reaction, form new tissues, deposit collagen, and grow new blood vessels 4 weeks after subcutaneous implantation. The in vitro and in vivo studies suggest that nanofiber aerogels with patterned macrochannels provided an ideal matrix for cell infiltration and host tissue integration. Such nanofiber aerogels can be used in combination with signaling molecules and cells for applications in wound healing, tissue repair, and tissue modeling.

4. Experimental Section

Materials:

PCL (MW = 80,000 g/mol), Type A Gelatin from the Porcine skin, GelMA, Irgacure 2959, Pluronic-F-127, gelatin, and Triton X-100 were all purchased from Sigma-Aldrich (St. Louis, MO, USA). Glutaraldehyde (alcoholic solution) was purchased from Thermo Fisher Scientific Inc (Waltham, MA, USA). Dichloromethane (DCM) and N, N-dimethylformamide (DMF) were purchased from BDH Chemicals (Dawsonville, GA, USA) while minimum essential medium (MEM), fetal bovine serum (FBS), and penicillin-streptomycin were purchased from Gibco, Thermo Fisher Scientific Inc. (Waltham, MA, USA). Hexafluoroisopropanol (HFIP) was purchased from Oakwood chemicals (Estill, SC, USA).

Preparation of Various Nanofiber Mats by Electrospinning:

In this study, various nanofiber mats including (i) PCL/gelatin (1:1) and (ii) PCL/gelatin/GelMA (1:0.5:0.5) were fabricated by electrospinning following our previously established protocols.^[39,40] A spinning solution 8% (w/v) of PCL/gelatin (1:1), and 8% (w/v) of PCL/gelatin/GelMA (1:0.5:0.5) were prepared by fully dissolving the predetermined mass of polymer in HFIP to achieve the desired polymer concentration. The resulting polymer solution was electrospun with the following set parameters: DC voltage = 15 kV, Flow rate = 0.4-0.6 mL/h, and distance between the spinneret to collector = 10-15 cm. During spinning, fibers were continuously collected on a rotating drum collector spinning at high and low speeds. When spinning was complete, the nanofiber mat was removed crosslinked in a glutaraldehyde (GA) + 25% EtOH vapor chamber overnight.

Fabrication 3D-printed Sacrificial Template:

Direct 3D printing of alginate hydrogel scaffolds largely followed a method previously reported with minor modifications.^[49,50] The STL files of the digital models were fed into a 3D printer (Allevi 1) for creating the desired patterns. Following printing, the alginate scaffolds were immediately physically crosslinked with a cold CaCl₂ bath (2.5%) for 10 min, and then stored in water at 4 °C until use.

Fabrication of Aerogels with and without Patterned Macrochannels:

All the aerogels with and without patterned macrochannels in this study were fabricated by a freeze-casting technique with minor alterations. Initially, the nanofiber mats of both PCL/gelatin and PCL/gelatin/GelMA were cryocut, collected, and freeze-dried and kept at 4 °C for further use. For the fabrication of aerogels, the segmented nanofibers were typically dispersed in water at a concentration of 25 mg/ml using a homogenizer with the amplitude of 20% with 10/20 seconds of ON/OFF cycles for 40 min in an ice bath. A certain amount of gelatin (5%, concerning the fiber content) was added during the homogenization after 40 min and will continue the homogenization for another 20 min in the above-mentioned condition. Then, we prepared aerogels using three different setups. (i) Isotropic freezing configuration: The homogenous mixture was pipetted into a Cu ring mold glued to an aluminum plate. Then, the sacrificial template was immersed into the mold, and the mixture was submerged in cryopreserved ethanol at −80 °C in directions from side and bottom for 3-5 min, transferred to a −80 °C refrigerator, and stored for 1 h. (ii) Anisotropic freezing configuration: The homogenous mixture was pipetted into a Cu ring mold glued to an aluminum plate and the sacrificial template was immersed into the mold. Then, the exterior of mold was covered by Styrofoam except the bottom of the mold and which was frozen on a −80 °C metal plate in a deep freezer. The mold was quickly transferred to a −80 °C refrigerator for 1 h. (iii) Partially anisotropic freezing configuration: The homogenous mixture was pipetted into a Cu ring mold glued to an aluminum plate. Then, the sacrificial template was immersed into the mold and rapidly frozen on a −80 °C metal plate in a deep freezer and immediately moved to a −80 °C refrigerator for 1 h. Then the frozen mold was moved to a freeze dryer and lyophilized for 24 h. Subsequently, the obtained aerogel was crosslinked by GA vapor in a closed chamber containing 1 ml of 50% GA in ethanol for 12 h to strengthen the mechanical property. We further immersed the aerogel containing the 3D-printed template in a 50 mM EDTA solution for 4-6 h to remove the sacrificial template, and then sterilized them with 70% ethanol prior to in vitro and in vivo studies.

Characterization of Morphology and Size by Scanning Electron Microscopy:

The morphologies of aerogels with and without patterned macrochannels were characterized by SEM (FEI Quanta 200, Hillsboro, OR, USA). After removing the template, the aerogels were frozen with water and section in two ways like both horizontally and longitudinally by using cryostat and followed by freeze drying to dry aerogels. The freeze-dried aerogels were fixed onto a metallic stub the support of double-sided conductive carbon tape, sputter-coated with an Au-Pd target at a peak current of 15 μA for 5 min. The aerogels were subsequently imaged using an accelerating voltage of 15-25 kV. The ImageJ plugin OrientationJ was employed to determine the frequency of angle distribution based on structure tensors using at 1 pixel resolution through a Gaussian window.^[51]

Subcutaneous Implantation:

This animal study was conducted following approval by the University of Nebraska Medical Center’s IACUC in accordance with animal protocol # 17-103-11-FC. Briefly, rats were anesthetized under 4% isoflurane in oxygen for roughly 2 min and placed on a heating pad to maintain their body temperature. Rats were continuously anesthetized by 2% isoflurane during surgery. An area of 4 × 4 cm² on the back of each animal was shaved and the povidone-iodine solution was applied three times on the shaved skin. The aerogels with and without patterned macrochannels were implanted subcutaneously. Two rats were used at each time point. Each rat received four implanted aerogels. The total number of samples was 8 at each time point. Rats were euthanized after implantation for 2 and 4 weeks. Each explant with surrounding tissue was gently dissected out of its subcutaneous pocket and then immersed in formalin for at least 3 days before histology analyses.

Statistical Analysis:

The experiments were performed in multiple replicates (n = 3 or n = 4) and all data is presented as mean ± standard deviation. All physical measurements were taken on a calibrated digital caliper and all photographic measurements were made using ImageJ’s measure function following scale calibration. Each data set was analyzed and graphed using Graph pad Prism Version 9.0.0. Ordinary one-way ANOVAs and two-way ANOVAs (displaying significance within-group) followed by Post-hoc testing (Tukey’s) were used for statistical analysis. p-value ≥ 0.05 was denoted as not statistically significant (ns) difference and p < 0.05 was denoted as statistically significant difference. All graphics were original and were prepared using Bio-Render, Microsoft PowerPoint and photoshop.