2.2. Preparation of CNTs-SS

Sericin was extracted from fibroin-deficient silkworm cocoons, Bombyx mori, 140 Nd-s, using LiBr isolation method as previously described [20]. These mutant silkworm cocoons are mainly composed of sericin. A gram of CNTs was added into 10 mL distilled water. After sonication for 30 min, the dispersed CNTs solution (10%, w/v; distilled water) was prepared. 200 μL genipin (1%, w/v; distilled water) was added into 1.2 mL sericin solution (2%, w/v; distilled water) to prepare the precursor solution. The dispersed CNTs solution and the precursor solution were thoroughly mixed at various volume ratios. After gelation and lyophilization, the CNTs-SS with different CNTs concentrations were prepared.

2.3. Characterizations of physical properties

The surfaces of the CNTs-SS were examined by a scanning electron microscope (SEM; SM-5610LV, Japan) with 25 kV working voltage. The pore sizes of different specimens were analyzed by ImageJ software. The mechanical properties of CNTs-SS were examined by a universal testing machine (Instron 5848 MicroTester, USA) equipped with a 100 N load cell. The dynamic swelling behaviors of the CNTs-SS were examined as previously described [20]. Briefly, CNTs-SS were immersed in phosphate-buffered saline (PBS) at 37 °C for 10 days. The dry weight (W_d) and swollen weight (W_s) of the scaffolds at a specified time point were recorded and the swelling ratios were calculated by following equation: Swelling (%) = (W_s – W_d)/(W_d) × 100%.

2.4. Cell culture

Bone marrow mesenchymal stem cells (BMSCs) were isolated from wild-type mice or GFP transgenic mice as previously reported [25]. Briefly, Iscove's Modified Dulbecco's Medium (IMDM, Hyclone, UT, USA) was infused into femurs and tibia lumens to flush marrow. The marrow was placed in a tissue culture flask. After a 3-day incubation, the unattached cells were removed. The adherent cells were cultured in IMDM supplemented with 10% fetal bovine serum (FBS, Gibco, CA, USA) and 1% penicillin-streptomycin (Hyclone, UT, USA). After 5 passages, the BMSCs were used for experiments. SHSY-5Y cells or GFP-expressing SHSY-5Y cells were cultured in complete Dulbecco Modified Eagle Medium (DMEM, Hyclone, UT, USA) supplemented with 10% FBS and 1% penicillin-streptomycin. All cells were incubated in a humidified incubator at 37 °C with 5% CO₂.

2.5. Seeding cells on prepared scaffolds

The CNTs-SS were sterilized by immersion in 75% ethanol for 1 h and washed with PBS to remove residual ethanol. Then 2 × 10⁵ cells (SHSY–5Y or BMSCs) were digested and suspended in 50 μL basal medium. The cell suspension was seeded on the top surface of the scaffold with 10 μL at 4 peripheral points and 10 μL at the central point. The cells in the scaffolds were cultured in corresponding medium at 37 °C with 5% CO₂.

2.6. Analysis of cell morphology and viability

To view cell adhesion, extension and growth, the GFP-expressing cells (SHSY–5Y or BMSCs) cultured in CNTs-SS were imaged using a confocal laser scanning microscope (Nikon A1Si, Japan) according to our previous study [21]. For CCK-8 assay (Dojindo, Japan), SHSY-5Y cells (5000 cells per well) were seeded in a 96-well plate coated with SS (control) or CNTs-SS. After a 3-day culture, a total of 100 μL DMEM mixed with 10 μL CCK-8 solution was added into each well and incubated for 45 min at 37 °C. The absorbance was measured at 450 nm using a microplate reader (TECAN, Mannedorf, Switzerland). The Live&Dead cell staining was performed as previously described [21]. Briefly, SHSY-5Y cells in CNTs-SS were digested and incubated with staining solution for 15 min at 37 °C. Then the cells were subject to analysis using a FACSCalibur flow cytometer (BD Canto II, NJ, USA).

2.7. In vivo biocompatibility

All animal experimental protocols were approved by the Animal Care and Use Committee, Tongji Medical College, Huazhong University of Science and Technology (Wuhan, China). The C57BL/6J mice were anesthetized and a hole with a 4-mm diameter was drilled in the skull. Disk-shaped CNTs-SS with a diameter of 3 mm were implanted onto exposed brain surface of mice. After 14 days, the mice were anesthetized, perfused with 0.9% NaCl (w/v, distilled water) and then fixed with 4% paraformaldehyde (w/v, PBS). The fixed brains were embedded in paraffin and cut into 5-μm sections for immunofluorescence staining. The rabbit anti-mouse GFAP (1:100 diluted; BA0056, Boster, Wuhan, China) and rabbit anti-mouse IBA-1 (1:200 diluted; Ab178847, Abcam, MA, USA) were used to detect astrocytes and microglia, respectively.

2.8. Biodegradation

To test in vivo degradation of CNTs-SS, disk-shaped CNTs-SS with a 3-mm diameter were implanted onto the surface of exposed murine brain. At the indicated time point, the mice were sacrificed. Then, the scaffolds were taken out and dried for weighing (W_n). The initial weight (W₀) of CNTs-SS was recorded and the residual weight percentages were calculated by the following equation: Residual weight $(\%)=W_n/W_0\times 100\%$.

2.9. Analysis of cell migration from CNTs-SS in vivo

GFP-expressing BMSCs were seeded in a disk-shaped CNTs-SS with a 3-mm diameter and cultured for 3 days. Then the scaffolds carrying GFP-expressing BMSCs were implanted onto the surface of exposed murine brain. After 7 or 14 days, the mice were anesthetized, perfused with 0.9% NaCl (w/v, distilled water) and then fixed with 4% paraformaldehyde (w/v, PBS). The fixed brains were embedded in paraffin and cut into 5-μm sections. After being stained with 4’,6-diamidino-2-phenylindole (DAPI, Sigma, TX, USA) and Tuj1 antibody (1:100 diluted; Ab52623, Abcam, MA, USA), the brain sections were observed using an inverted fluorescence microscope (Olympus IX71, Japan).

2.10. Characterization of shape-memory property

The T-shaped, I-shaped and pentagram-shaped CNTs-SS were fabricated by mold casting. The scaffolds were immersed in PBS for 30 min and subjected to compression and rehydration in PBS. The compressed scaffolds and rehydrated scaffolds were imaged using a camera (Sony ILCE, Japan). For mathematical model establishment, a series of cubic CNTs-SS (10 mm × 10 mm × 10 mm) were fabricated. The scaffold was immersed in PBS for 30 min and then compressed into a square-shaped small graft. PBS was continuously dripped onto the scaffold and the video was captured. The images were extracted from the video and analyzed by ImageJ software. Multiple linear regression was performed to establish the mathematical model using MATLAB software (VisionR2012b, MathWorks, MA, USA). For in vivo experiments, a pentagram-shaped CNTs-SS was fabricated and compressed. After a C57BL/6J mouse was anesthetized, a 1-mm incision was opened on the abdomen and the compressed scaffold was implanted through the incision. The images were continuously captured by a versatile small animal imaging system (In-Vivo FX PRO, Bruker, Germany).

2.11. Characterization of injectability

The CNTs-SS were fabricated in three different shapes, including a 4 mm × 4 mm square shape, a 6 mm × 6 mm square shape and a 3-mm-diameter disk shape. After being immersed in PBS for 30 min, the scaffolds were suspended in 0.3 mL PBS and injected by a syringe through a 16G needle. During injection, the videos and the images were captured. For in vivo injection experiments, disk-shaped CNTs-SS were suspended in 0.3 mL PBS and then injected into different sites of mice, including skin, muscle, abdomen and brain, through a 16G needle. The mice were anesthetized and imaged using a versatile small animal imaging system. The mice receiving a subcutaneous injection of the scaffold were imaged by a camera. To test cell viability after injection, confocal laser imaging and the Live&Dead cell staining were performed. GFP-expressing SHSY-5Y cells or wild-type SHSY-5Y cells were seeded in the disk-shaped CNTs-SS and cultured for 24 h. After being injected through a 16G needle, the scaffolds loaded with cells were cultured for another 24 h. The non-injected scaffolds were set as a control. Then the scaffolds loaded with GFP-expressing SHSY-5Y cells were observed under a confocal laser scanning microscope. The wild-type SHSY-5Y cells in the scaffolds were digested and subjected to Live&Dead cell staining.

2.12. In vitro implantation of CNTs-SS into PDMS cavity

Polydimethylsiloxane (PDMS) precursor (A) and crosslinker (B) were mixed at the ratio of 10:1. The mixture was poured into a 1.5 cm × 1.5 cm square-shaped mold. After being cured at 110 °C for 30 min, a piece of square-shaped PDMS was fabricated. Then a hole with a 3-mm diameter was punched at the centre of the PDMS. After being treated with oxygen PLASMA (PDC-32G-2, Harrick, USA), the punched PDMS was bonded to two pieces of un-punched PDMS to obtain a PDMS cavity. Three disk-shaped CNTs-SS were compressed into small grafts, loaded into the tip of a 16G flat-head needle, and then vertically, obliquely, or horizontally injected into the three PBS-filled PDMS cavities, respectively. All of these PDMS cavities were imaged before, during and after injection.

2.13. In vivo application of a customizable CNTs-SS for filling stroke cavities

The C57BL/6J mice were subjected to transient focal cerebral ischemia by middle cerebral artery blockade (MCAO) with a suture as previously described [19]. After 2 weeks, the MCAO mouse's brain was imaged using a magnetic resonance imaging (MRI) system. Using surface rendering and fluorescence thresholding in Imaris software, the MRI images of the stroke cavity were three-dimensionally reconstructed. The 3D images of stroke cavity were converted to STL-format files for designing 3D molds (RAYDATA, Wuhan, China). Then the corresponding 3D molds were printed by a 3D printer (Phoenix Touch Pro, S-MARKER, China). The CNTs-SS with customized shapes were fabricated through casting in these molds. To visualize the scaffolds in micro-CT scanning, the CNTs-SS were mixed with 0.1% BaSO₄ during fabrication. The MCAO mouse was anesthetized and the skull was drilled. The scaffold was compressed, loaded in the tip of a 16G flat-head needle, and transplanted into the stroke cavity by stereotactic injection. Then the mouse was scanned by a Skyscan 1176 micro-CT (Bruker, Germany).

2.14. Construction of TPIS (Tuj1 promoting imaging system)

The Tuj1 promoter sequence was cloned from the genomic DNA of the C57BL/6J mouse brain tissue. The promoter region spanning 2900 bp upstream of Tuj1 starting codon ATG was cloned using the following primers: forward, 5′-GATACTAGTCTAAGGGGGAGCAACT-3’; reverse, 5′-TCGAATTCGCTGACTTCACGCGGCT-3’. The restriction enzyme sites (SpeI and EcoRI, underlined) were designed into these primers. The Tuj1 promoter sequence was inserted into a pCDH vector (System Biosciences, CA, USA) to replace the MSCV promoter sequence on the backbone. The mCherry coding sequence (CDS) was cloned from a px330-mCherry vector (Addgene 98,750, USA) using the following primers [26]: forward, 5′-GCGAATTCATGGTGAGCAAGGGCGA-3’; reverse, 5′-GCGGATCC TTACTTGTACAGCTCGT-3’. The restriction enzyme sites (EcoRI and BamHI, underlined) were designed into these primers. The mCherry CDS sequence was inserted into the pCDH vector backbone downstream of the Tuj1 promoter sequence to yield the pTuj1-mCherry vector. GFP expression was driven by the EF1α promoter independently within the vector.

2.15. Analysis of neuro-differentiation-promoting activity of CNTs-SS

BMSCs were seeded in CNTs-SS as the methods described above, and cultured in IMDM containing 10% FBS for 7 or 14 days. The mRNA was extracted from BMSCs by Trizol (Invitrogen, CA, USA) for reverse transcription. Quantitative Real-Time PCR (qRT-PCR) was performed using ThermoScript RT-PCR System (Invitrogen, CA, USA). To examine cell morphology, the BMSCs cultured in the scaffolds for 14 days were stained with Calcein-AM (Biovision, CA, USA) according to the manufacturer's instructions. The BMSCs in the scaffolds were observed using a confocal laser microscope (Nikon A1Si, Japan). A portion of BMSCs in the scaffolds were fixed with 2.5% glutaraldehyde and imaged by scanning electron microscopy (SEM).

The Western blot analyses were performed as previously described [21]. Briefly, the samples were subjected to SDS-PAGE electrophoresis and then transferred to nitrocellulose membranes. After the membranes were incubated with the antibodies: anti-Tuj1 antibody (Ab52623, Abcam, MA, USA) and anti-GAPDH antibody (AC002, Abclonal, Wuhan, China), the blots were examined using a Luminescent Image Analyzer (Fujifilm LAS-4000, Tokyo, Japan). The immunofluorescence staining assay was performed as previously described [21]. The BMSCs cultured in the scaffolds for 14 days were fixed with 4% paraformaldehyde (w/v, PBS). The cells were incubated with the antibodies: anti-Tuj1 antibody (Ab52623, Abcam, MA, USA) and a fluorescein (FITC)-conjugated secondary antibody (111-095-003, Jackson ImmunoResearch, PA, USA). The cells were imaged using a confocal laser microscope (Nikon A1Si, Japan).

The BMSCs were transduced with the pTuj1-mCherry vector using Lipofectamine 2000 (Thermo Fisher Scientific, MA, USA). Then the cells were seeded in CNTs-SS. After being cultured for 14 days, the BMSCs were digested from the scaffolds and sorted by a flow cytometer. The GFP-Cherry double positive cells were collected and frozen in liquid nitrogen for transportation. The samples’ quality control, RNA sequencing and gene expression analysis were carried out in Beijing Genomics Institute (BGI, China). Then, the hierarchical clustering analysis of 40 neuron-associated genes was performed as previously reported [27].

3.2. In vitro cyto-compatibility and in vivo biocompatibility of CNTs-SS

To determine whether CNTs-SS could act as a cell carrier to effectively support cells’ adhesion and proliferation, we generated a GFP-expressing SHSY-5Y neuronal cell line for growing on the scaffolds. 3D reconstruction of confocal images showed that over 85% of neuronal cells directly contacting the scaffolds spread well adopting fusiform morphology (Fig. 2a and b) with cell viability over 90% (Fig. 2c and d). Moreover, CNTs-SS also exhibited good in vivo bio-safety and biocompatibility. One and two weeks after a disk-shaped CNTs-SS was implanted onto the brain surface after partial skull removal (Figure S6a), astrocytes and microglia, two main types of inflammation-responding cells in brain, were not functionally activated (no characteristic morphological alterations) (Figure S6b) and not quantitatively increased (Figure S6c). Given that biodegradability of a scaffold is critical for brain regeneration as it would provide additional space for formation of new tissues, we analyzed in vivo degradation dynamics of CNTs-SS (Figure S7a-b) and found that the CNTs-SS became structurally loose by Day 28, underwent significant structural dissociation by Day 42 (Week 6), and almost disappeared by Day 56 (Week 8). With the limited regenerative capability, the injured brain tissues would take several months for the differentiation of stem cells and the formation of neuronal networks after the stem cell therapy [[31], [32], [33]]. The 56-day degradation period of CNTs-SS could provide the long-term structural support for the regeneration of de novo neural tissues. Moreover, the degradation of CNTs-SS leaves the addition space for the formation of new tissues, facilitating the brain tissue repair.

Open in a separate window

Fig. 2

The neuronal cyto-compatibility of CNTs-SS. (a) The confocal images of GFP-expressing SHSY-5Y cells (green) cultured in SS (red) or CNTs-SS (red) for 3 days. The regions boxed with yellow dashed lines were enlarged in the lower panels. (b) Quantification of the percentages of spreading cells and round cells in (a). (c) The CCK-8 assay for determining cell viability of SHSY-5Y cells cultured in SS or CNTs-SS for 3 days. *, P < 0.05; ANOVA. (d) Flow cytometry analyses after Live&Dead cell staining were performed to assess cell viability of SHSY-5Y cells cultured in SS or CNTs-SS for 3 days. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

3.3. CNTs-SS as an in vivo carrier of stem cells

To investigate if CNTs-SS were capable of delivering cells to target brain regions in vivo. GFP-expressing bone marrow mesenchymal stem cells (BMSCs) from GFP transgenic mice were isolated and seeded onto 0.5CNTs-SS. At Day 7 after implantation of the cell-carrying 0.5CNTs-SS onto the brain surface, the GFP-expressing cells delivered by CNTs-SS already migrated out of the scaffold and infiltrated into the neighboring brain tissues (Fig. 3a) by the average depth of 147 μm (Fig. 3b). This migration progressed to the distance of 196 μm at Day 14 (Fig. 3a–b), a 33.3% increase. The average area occupied by infiltrating cells at Day 14 (3.6 × 10⁴ μm²) was 1.6-time larger than that on Day 7 (1.4 × 10⁴ μm²) (Fig. 3c). Notably, the total numbers of infiltrating cells were similar between Day 7 (3.9 cells/10⁴ μm²) and Day 14 (3.4 cells/10⁴ μm²) (Fig. 3d), indicating no obvious reduction in the number of living cells, in contrast to rapid cell loss often observed for hydrogels-delivered stem cells [34,35].

Open in a separate window

Fig. 3

The CNTs-SS as a cell carrier effectively delivered stem cells to target brain regions. (a) Immunofluorescence staining of the brain sections 7 days or 14 days after brain tissue (Tuj1 positive, red) was implanted with a disk-shaped 0.5CNTs-SS (white asterisks; diameter, 3 mm) carrying GFP-expressing BMSCs (green). Nuclei were stained blue (DAPI). These BMSCs migrated out of the scaffold, crossed the interface (yellow dashed lines) between the scaffold and the brain, and infiltrated into surrounding brain tissue (outlined by white dashed lines). The white arrows indicate BMSC's infiltrating direction. (b–d) Quantification of infiltrating distance (b), infiltrating area (c), and infiltrating cell numbers of GFP-expressing BMSCs (d) after the scaffolds were implanted for 7 days or 14 days. **, P < 0.01; N·S·., not significant; Student's t-tests. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Apart from the migration of these exogenous BMSCs, the infiltration of host cells also facilitates the tissue healing [36]. The physically cross-linked hydrogel reportedly supports the infiltration of cells and promotes in situ tissue regeneration [37]. The interconnected porous microstructure of CNTs-SS could provide favorable space to support the infiltration of host cells, therefore benefiting the tissue repair. Moreover, give the abundantly active groups of sericin, bioactive factors or inductive factors, such as RGD, N-cadherin peptide, BDNF and TGF-β, could be included in this CNTs-SS through encapsulation or chemical conjugation, promoting cells adhesion (RGD), survival and differentiation [20,38,39], further enhancing the therapeutic outcomes of CNTs-SS.

3.4. The programmable shape-memory property of CNTs-SS

Unexpectedly, we found that CNTs-SS possessed a shape-memory property. 0.5CNTs-SS with four different shapes (“T”, “I”, pentagram, and cube shapes) could be volumetrically compressed by roughly 90% into small masses that then restored their original shapes within 10 s upon rehydration (Fig. 4a and b; Table S2; Movie S1). We established a mathematical model to describe the quantitative relations across CNTs' concentrations, recovery time, and volumetric changes via polynomial binary cubic forms (Fig. 4c). In the model, at a given CNTs' concentration, compressed CNTs-SS within a fixed time (10s) (Fig. 4c) appeared to have a fast recovery phase followed by a relatively slow recovery phase. As CNTs' concentrations were elevated, (1) the fast recovery phase was progressively extended, while the slow phase was shortened; (2) the recovery rates (slopes of the curves) in the fast phase were gradually reduced, but were increased in the slow phase. Using this model (Fig. 4c), the recovery behaviors of CNTs-SS could be conveniently programmed by varying its CNTs' concentrations. Since the overly rapid shape restoration of a scaffold would cause rather strong extrusion pressure to damage surrounding normal brain tissues, it is ideal to have restoration dynamics precisely controlled. Aided by this model, we could design a scaffold with on-demand restoration dynamics towards optimizing extrusion pressure during shape recovery. Next, we examined in vivo realization of shape-memory property by time-lapse imaging the fluorescence emitted from a pentagram-shaped 0.5CNTs-SS. This scaffold was compressed and implanted into the abdominal cavity of a mouse, and successfully recovered to its original shape (Fig. 4d). The sericin's α-helices, that are arranged in coiled-coil architectures, undergo continuous structural transition into metastable β-sheets when load is applied [40], and this reversible transition could be restored in response to hydration, probably endowing the shape-memory properties to CNTs-SS.

Open in a separate window

Fig. 4

The shape-memory property of CNTs-SS. (a) The compressed 0.5CNTs-SS in three different shapes recovered to their original shapes after rehydration. (b) Dynamic shape-recovery process of a compressed 0.5CNTs-SS upon rehydration. Frames were extracted from Movie S1. (c) The mathematical model (left) with the corresponding formulation (right) described the programmable recovery behaviors of CNTs-SS. p00, p10, p01, p20, p11, p30 and p21 were coefficients (with 95% confidence bounds); x, y and F (x,y) were three variables as indicated. (d) The compressed 0.5CNTs-SS restored its original pentagram shape after in vivo intraperitoneal implantation. Images of implanted scaffolds were enlarged and boxed with red dashed lines. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Supplementary data related to this article can be found at https://doi.org/10.1016/j.bioactmat.2020.12.017.

The following is the supplementary data related to this article:

3.5. The injectability of CNTs-SS

Similar to water-responsive shape-memory materials (such as methacrylated alginate scaffolds and glycidyl methacrylate functionalized chitosan cryogels) that reportedly have injectability [15,41,42], CNTs-SS were also injectable. The disk-shaped or square-shaped 0.5CNTs-SS in 300 μL phosphate buffer saline (PBS) was readily injected through a conventional 16G needle (Fig. 5a; Movie S2) with no obvious structural destruction detected (Figure S9a-c). Further, several sterile disk-shaped 0.5CNTs-SS suspended in 300 μL PBS were injected into different sites of mice, including skin, muscle, abdomen and brain, and effectively recovered to their original disk shapes (Fig. 5b and c), indicating CNTs-SS′ in vitro and in vivo injectability. Moreover, CNTs-SS could well protect cells from compressive damage of injection process, which might lead to certain-level microscopic collapse of scaffolds’ polymeric networks. Both confocal microscopy and flow cytometry analyses revealed that 92% of GFP-expressing SHSY-5Y cells carried by 0.1CNTs-SS and 0.5CNTs-SS scaffolds remained viable after injection (Fig. 5d–f). This cell protective effect was presumably attributable to extensive connection of pores and physical flexibility of polymeric networks.

Open in a separate window

Fig. 5

The injectability of CNTs-SS. (a) Photographs showed that a disk-shaped 0.5CNTs-SS (white arrowheads) in a 1-mL syringe (before injection) was injected through a 16G needle (after injection) and restored its original disk shape (shape recovery). (b) The shape recovery of a disk-shaped 0.5CNTs-SS (white arrowhead; red dashed lines) after subcutaneous injection. The region boxed by white dashed lines was enlarged on the lower panel. (c) Fluorescent images of a disk-shaped 0.5CNTs-SS (white arrowheads) after being injected underneath skin and into muscle, abdomen or brain (excitation 700 nm; emission 750 nm). The regions boxed with white dashed lines were enlarged at the lower right corner. (d) The confocal images of GFP-expressing SHSY-5Y cells (green) carried by CNTs-SS (red) that were injected (After injection) or not injected (No injection). (e) Quantification of the percentages of spreading cells and round cells from the indicated groups in (d). (f) Flow cytometry analysis on cell viability of Live&Dead stained SHSY-5Y cells carried by 0.1CNTs-SS or 0.5CNTs-SS that was injected (After injection) or not injected (No injection). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Supplementary data related to this article can be found at https://doi.org/10.1016/j.bioactmat.2020.12.017.

The following is the supplementary data related to this article:

3.6. CNTs-SS as customizable scaffolds precisely fill stroke cavitary lesion in vivo

To test whether injection direction relative to the position of a stroke cavity affected the final orientation of the shape-restored scaffold within the cavity and its shape-fitting performance to the cavity, we conducted in vitro polydimethylsiloxane (PDMS) “stroke” cavity implantation experiments (Figure S10a). Independent of the various orientations of injection needles relative to the PDMS cavity, the compressed scaffold was able to effectively restore its original shape and well fit the cavity (Figure S10b-d). Next, we determined whether CNTs-SS could be precisely and geometrically fill stroke cavity in vivo as pre-designed in a mouse stroke model (MCAO) where a middle cerebral artery was occluded (Fig. 6a; Movie S3) [19]. The cavitary lesion in the MCAO mouse was scanned by magnetic resonance imaging (MRI) (Fig. 6b). The stacks of MRI image slices were 3D reconstructed using threshold segmentation algorithm to obtain the cavitary contour (Fig. 6c; Movie S4). Using a digital light processing (DLP) 3D printer, the customized molds based on this contour were printed for casting a 0.5CNTs-SS precisely matching the cavitary geometry (Fig. 6d). The cast scaffold was then intracranially injected into the cavity of the MCAO mouse that was MRI imaged. The scaffold faithfully restored its pre-cast shape (Fig. 6e; Movie S5), which was verified by micro-CT images. These results collectively exemplify the potential utilization of CNTs-SS by showing how CNTs-SS can be morphologically fabricated on demand, minimal-invasively implanted in vivo, and geometrically fit an irregular-shaped cavitary lesion, thus demonstrating the feasibility of using CNTs-SS for in vivo stroke repair.

Open in a separate window

Fig. 6

The in vivo application of a shape-customized CNTs-SS for filling a random stroke cavity. (a) Schematics showing the whole process of this application: extraction of image information of a stroke cavity in a MCAO mouse for 3D image reconstruction, fabrication of a customized CNTs-SS using a 3D-printed mold, and assessment of its shape recovery after being injected into the cavity. (b) The continuous MRI image series (left) covering the MCAO mouse's brain. Six representative slices from these MRI images showed the site and shape of the stroke cavity (white arrowheads). (c) The 3D reconstruction (left) of MRI images of the MCAO mouse's full brain (green) with the volumetrically-rendered stroke cavity (yellow). The images were extracted from Movie S4. (d) The 3D-printed mold (left) for casting the CNTs-SS with the customized shape geometrically matching the cavity. The newly cast CNTs-SS (right) was viewed from the different angles. (e) The micro-CT images of the injected CNTs-SS (orange; white arrowheads) within the stroke cavity in the brain (skull, light blue) viewed from different angles. The images were extracted from Movie S5. The implanted 0.5CNTs-SS was mixed with 0.1% BaSO₄ during fabrication for micro-CT imaging. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Supplementary data related to this article can be found at https://doi.org/10.1016/j.bioactmat.2020.12.017.

The following are the supplementary data related to this article:

J.W. and X.L. contributed equally to this work. This work was supported by the National Key Basic Research Program of China [2015CB5540007], the National Natural Science Foundation of China [81671904, 81572866, 81773104, and 81873931], the Science and Technology Program of Chinese Ministry of Education [113044A], the Major Scientific and Technological Innovation Projects in Hubei Province [2018ACA136], the Integrated Innovative Team for Major Human Diseases Program of Tongji Medical College of HUST, and the Academic Doctor Supporting Program of Tongji Medical College of HUST.