Hydrogel formulation: The precursor solution for microgel synthesis had a 3.0 wt% (w/v) polyethylene glycol (PEG) backbone and was prepared by mixing equal volumes of working solutions A and B. Working solution A consists of 5.64 mM 4-arm PEG-maleimide (10 kDa, Nippon Oil Foundry, Japan), 0.75 mM MethMal annealing macromer (synthesized as previously described66), and 2 mM RGD peptide (Ac-RGDSPGGC-NH2, Watson Bio). Working solution B consists of 10.78 mM enzymatically degradable crosslinker (Ac-GCGPQGIAGQDGCG- NH2, Watson Bio) and 10 µM biotin-PEG6-maleimide (TCI Chemicals). To synthesize microgels, working solution A was prepared using 10X PBS at pH 1.5, and working solution B was prepared using 1X PBS at pH 7.4. The two working solutions were then thoroughly mixed at a 1:1 ratio to create the precursor solution. To create the nanoporous implants, both working solutions were dissolved at twice their concentrations (solutions A2 and B2) in 1X PBS at pH 7.4 and aliquoted for use in surgery (40 µL for each aliquot). Aliquots (80 µL) of 1X PBS at pH 9.15 were also prepared. All solutions used for nanoporous implants were sterile filtered prior to aliquoting and frozen at -20°C afterward.
Mechanical matching
Bulk hydrogel constructs were used to mechanically match the hydrogel formulation to skeletal muscle (12 kPa68). Small hydrogel disks (100 µL) were formed between two glass slides treated with SigmaCote that were spaced 2 mm apart. After gelation was completed, the hydrogel disks were swollen to equilibrium in 1X PBS overnight at 37°C. Immediately before testing, the disks were removed from PBS and excess moisture was wicked away. The disks were then tested for compressive stiffness using an Instron Universal Testing System at a rate of 0.5 mm min− 1. BlueHill software and the MATLAB SLM package were used to create stress-strain curves and calculate the Young’s modulus (Pa).
Microgel size quantification: A dilute solution of microgels (1:100) was prepared in 1X PBS with 1 µM fluorescently labeled streptavidin. After at least 15 min incubation at 37°C, the solution was imaged in a 96-well plate (100 µL per well) using an ImageXpress Micro Confocal microscope (Molecular Devices) and particle diameters were quantified with an ImageXpress analysis module. Over 6000 particles were analyzed to determine average diameter and polydispersity index (PDI).
Microgel synthesis
Microgels were generated with a step-emulsification device (11.4 µm channel height) using a high-throughput method previously described67. A flow rate of 3.75 mL hr− 1 was used for the precursor solution, while a flow rate of 7.5 mL hr− 1 was used for the oil solution (2% Pico-Surf in NOVEC 7500). A solution of triethylamine (20 µL mL− 1 precursor solution) dissolved in NOVEC 7500 was added to the collection tube to accelerate gelation. After gelation was finished, surfactant was removed from the microgels with successive washes in NOVEC 7500, 1X PBS, and hexanes as previously described54. Unreacted maleimides were quenched by incubating the microgels in 100 mM N-acetyl-L-cysteine (Acros Organics) overnight at 37°C. Following three washes in 1X PBS, microgels were sterilized with three washes in 70% isopropyl alcohol (IPA) and stored in 70% IPA at 4°C until further use.
Preparation of MAP hydrogel for implantation: Microgels were transitioned from 70% IPA to PBS with 5 successive washes in sterile 1X PBS. A sterile solution of 1 mM lithium phenyl-2,4,6- trimethylbenzoylphosphinate (LAP) in 1X PBS was prepared, and microgels were incubated in the photoinitiator solution at a 1:1 ratio (v/v) for at least 15 minutes. The microgels were then centrifuged at 18,000 g for 5 min and the supernatant was removed. Centrifugation and removal of supernatant was repeated until no supernatant could be visualized. The microgels were loaded into sterile 1 mL syringes and protected from light with aluminum foil.
Animal care
This study was conducted in compliance with the University of Virginia Animal Care and Use Committee guidelines and all procedures were approved under protocol 4045. We used a well characterized rodent model of VML in the tibialis anterior (TA)71. Twelve week-old female Lewis rats were purchased from Charles River and used for all studies. Rats were pair-housed in a vivarium accredited by the American Association for the Accreditation of Laboratory Animal Care and provided with enrichment, food, and water ad libitum.
In vivo functional testing
We evaluated TA function before defect creation to establish a baseline and at 4, 8, and 12 weeks after surgery as previously described79. Briefly, rats were anesthetized (1.5–2.5% isoflurane) and placed in a supine position on a heated platform with their knee stabilized at a 90° angle and their foot secured to the footplate of the Aurora Scientific 305-LR-FP servomotor. Two percutaneous needle electrodes were superficially inserted into the skin of the anterior compartment of the lower leg along the peroneal nerve. Electrical stimulation was applied in a controlled manner using the Aurora Scientific Stimulator Model 701C and Dynamic Muscle Control software at a range of frequencies (1-200 Hz). Functional testing was performed on the left hindlimb only.
VML defect creation and treatment
Surgical creation was performed as previously reported. Briefly, rats were anesthetized (2-2.5% isoflurane), administered analgesia (Ethiqa XR, 0.65 mg kg− 1), and the left hindlimb was aseptically prepared. A longitudinal incision was made on the lateral aspect of the lower leg, and the skin was separated from the fascia. Another longitudinal incision was made along the lateral aspect the fascia over the TA. Blunt dissection was used to gently separate the fascia from the TA and expose the anterior crural muscles. The extensor digitorum longus (EDL) and extensor hallucis longus (EHL) were isolated and removed. The boundaries of the middle third of the TA (longitudinal axis, approximately 1 cm in length) were marked with a sterile surgical marker. Having previously calculated the theoretical weight of the TA (0.0017 × body weight), approximately 20% of the TA was removed, leaving 2–3 mm margins on the lateral and medial sides of the muscle and avoiding damage to the distal tendon. For animals treated with MAP scaffolds, 100 µL of microgels were injected into the defect and irradiated with 365 nm LED light (ThorLabs, 440 mA) for 5–10 s, until the scaffold was annealed. For animals treated with nanoporous implants, one aliquot each of working solution A2, working solution B2, and pH 9.15 1X PBS were thawed and their entire volumes were thoroughly mixed immediately before applying 100 µL of the mixture to the defect. Gelation of the nanoporous gel was verified before proceeding. The fascia was closed with interrupted sutures and 6 − 0 vicryl. The skin was closed with interrupted sutures and 5 − 0 prolene, and a thin layer of skin glue was applied over the skin sutures.
MRI quantification of implant
Animals were scanned at 4 weeks and 12 weeks post-surgery with a T2- mapping sequence using a small animal MRI machine (7 Tesla Bruker/Siemens ClinScan). Rats were anesthetized (1.5–2.5% isoflurane) and images were acquired from knee to ankle with slice thickness of 0.7 mm. No contrast agent was used. The images were analyzed on OSiriX to calculate the total volume of hydrogel implant remaining in the defect. One animal from the nanoporous group in the 12 week cohort was not imaged due to scheduling constraints at the core facility.
Tissue collection
At the endpoints, animals were anesthetized by CO2 inhalation. TA muscles were isolated, cut in half through the middle of defect (perpendicular to the longitudinal axis), and immediately snap-frozen in liquid nitrogen-cooled isopentane (-150°C). Samples were then embedded in OCT. Cryosections of 10 µm thickness were obtained, with consecutive sections on the same slide obtained from locations at least 200 µm apart on the block. Cryosections of 50 µm thickness were obtained for additional NMJ characterization.
Picrosirius red staining: Slides were rehydrated through a graded series of EtOH (100%, 95%, 70%; 3 min each) to deionized water and fixed in 4% PFA for 15 min before staining with picrosirius red according to the manufacturer’s instructions (Abcam, cat#: ab150681).
Immunostaining: Slides were allowed to warm to room temperature before staining. For Pax7 and macrophage stains (CD68 and CD163), we fixed the slides for 10 min in 4% PFA and then performed heat induced antigen retrieval. We submerged the slides in a citrate-based unmasking solution (Vector Laboratories, cat#: H-3300) and processed them in an Instant Pot pressure cooker for 20 minutes on the low pressure setting with natural release. The slides were then left to cool down to room temperature for 1 hr. Slides were rinsed with PBS and sections were circled with a hydrophobic pen (Vector Laboratories, H-4000). Slides were washed with 10 mM glycine in PBS for 10 min, then washed twice with 0.5% Tween 20 in PBS. They were incubated for 30 min in blocking buffer80 (2% goat serum, 50 mM glycine, 0.05% Tween 20, 0.1% Triton X-100, and 0.1% BSA in PBS) at room temperature, followed by overnight incubation with the primary antibody at 4°C. Primary antibodies were diluted in antigen signal enhancement solution80 (10 mM glycine, 0.05% Tween 20, 0.1% Triton X-100, 0.1% hydrogen peroxide). Slides were washed in 0.5% Tween 20 for 5 min, then washed twice in PBS for 5 min. Secondary antibodies were diluted 1:1000 in 0.1% Tween 20, and slides were incubated overnight at 4 ºC. Slides were washed four times in PBS for 5 min, incubated with DAPI (Thermo Scientific, cat#: 62248; 1:1000 in PBS) and wheat germ agglutinin (WGA) (Invitrogen, cat#: W32464; 5 µg mL− 1) for 30 min, and washed three times in PBS for 5 min. Slides were mounted with Prolong Gold (Invitrogen, P36930). For all other antibodies, slides were fixed in cold 100% MeOH for 10 min at 20 ºC, followed by three PBS rinses (5 min each). Sections were circled with a hydrophobic pen, then slides were permeabilized in PBS-T (0.1% Triton X-100 in PBS) for 10 min. Slides were incubated in Sea Block (Thermo Scientific, cat#: 37527) for 30 min, then incubated with primary antibodies overnight at 4 ºC. Primary antibodies were diluted in 5% Sea Block in PBS-T. Slides were washed in PBS-T for 5 min, then washed twice in PBS for 5 min. Secondary antibodies were diluted 1:1000 in PBS-T, and slides were incubated overnight at 4 ºC. Slides were washed four times in PBS for 5 min, incubated with DAPI (1:1000 in PBS) and WGA (5 µg mL− 1) for 30 min, and washed three times in PBS for 5 min. Slides were mounted with Prolong Gold. For slides stained with NF200, an endogenous biotin blocking kit (Invitrogen, cat#: R37628) was used according to manufacturer instructions prior to incubation with primary antibodies. Biotinylated α-Bungarotoxin (BTX) (Invitrogen, cat#: B1196) was included in the primary antibody diluent (1 µg mL− 1) to label neuromuscular junctions (NMJs), and fluorescently labeled streptavidin (Invitrogen, cat#: 84547) was included in the secondary antibody diluent (1 µg mL− 1). Slides were imaged with a 20x objective on an ImageXpress Micro Confocal microscope (Molecular Devices) or on an Axioscan 7 Slide Scanner (ZEISS). Slides with 50 µm cryosections were imaged with a Leica DMi8 inverted microscope using a 40x oil immersion objective.
Table 1
((Primary antibodies used for immunohistochemistry))
Primary antibody | Company | Catalog number | Dilution |
Pax7 | Santa Cruz Biotechnology | sc-81648 | 1:50 |
CD31 | Novus Biologicals | NB100-2284 | 1:200 |
CD68 | Bio-Rad | MCA341R | 1:100 |
CD163 | Abcam | ab182422 | 1:200 |
NF200 | Invitrogen | PA1-10016 | 1:200 |
MF20 | DSHB | MF 20 concentrate | 1:50 |
SV2 | DSHB | SV2 concentrate | 1:50 |
S100 | Invitrogen | MA1-26621 | 1:100 |
Table 2
((Secondary antibodies used for immunohistochemistry))
Secondary antibody | Company | Catalog number |
Goat anti-Mouse IgG, Alexa Fluor Plus 647 | Invitrogen | A32728 |
Goat anti-Rabbit IgG, Alexa Fluor Plus 555 | Invitrogen | A32732 |
Goat anti-Mouse IgG, Alexa Fluor Plus 488 | Invitrogen | A32723 |
Goat anti-Rabbit IgG, Alexa Fluor Plus 647 | Invitrogen | A32733 |
Goat anti-Rat IgG, Alexa Fluor Plus 488 | Invitrogen | A48262 |
Goat anti-Mouse IgG, Alexa Fluor Plus 594 | Invitrogen | A32742 |
Quantification of immune cells
Slides were imaged with a 20x objective on an ImageXpress Micro Confocal microscope (Molecular Devices), and stitched images for whole sections were exported as TIF files. The MATLAB Image Processing Toolbox and a custom code was used to isolate square 0.5 mm2 regions of interest from whole cross-section images. At least three sections were analyzed per animal, and three random regions of interest within the area of the defect were analyzed per section. Quantification of nuclei, CD68+ cells and CD68+/CD163+ cells was performed using CellProfiler.81
Quantification of fibrotic capsules
Slides stained with picrosirius red were imaged with on a Leica DMi8 microscope, and stitched images for whole sections were exported as TIF files. At least three sections were analyzed per animal. Masks of fibrotic capsules were created and quantification of the thickness of fibrotic capsules was performed using the MATLAB Image Processing Toolbox.
Quantification of nerves and NMJs
Slides stained with NF200 and BTX were imaged with a 20x objective on an AxioScan 7 Slide Scanner (ZEISS), and stitched images for whole sections were exported as CZI files. Image analysis was performed on QuPath82. Masks were created for the region of the implant/defect, and another mask was extended 500 µm into the muscle for the interface region. At least three tissue sections were analyzed per animal. The watershed cell detection module was used to identify nerves (NF200+ areas) and neuromuscular junctions (BTX+ areas) in the whole section, and the masks were used to quantify features by location. Data exported from QuPath was analyzed with RStudio. The density of NMJs was calculated for each region, then normalized to the density of the uninjured region of each tissue section.
Characterization of NMJs
Slides stained with BTX, NF200, S100, SV2, and DAPI were imaged at 40x with a Leica DMi8 microscope to generate z-stacks (1 µm slices) of regions of interest. Average intensity projections were created for visualization.
Quantification of blood vessels
Slides stained with CD31 were imaged with a 20x objective on an AxioScan 7 Slide Scanner (ZEISS), and stitched images for whole sections were exported as CZI files. Image analysis was performed on QuPath82. Masks were created for the region of the implant/defect, and another mask was extended 500 µm into the muscle for the interface region. At least three tissue sections were analyzed per animal. A machine learning approach was used to train a pixel classifier for the identification of blood vessels (CD31+ structures) in the whole section. A training image “collage” composed of least three regions of interest per animal was generated and annotated manually, followed by the creation of a Random Trees model. The image features used in the model were Gaussian filter, Laplacian of Gaussian, Weighted Deviation, and Gradient Magnitude. Only data from the channels corresponding to WGA and CD31 were included, using full resolution (0.35 µm/px) and 0.5 Gaussian scale. Data exported from QuPath was parsed and summarized with RStudio.
Quantification of satellite cells
Slides stained with Pax7 were imaged with a 20x objective on an AxioScan 7 Slide Scanner (ZEISS), and stitched images for whole sections were exported as CZI files. At least three sections were analyzed per animal, and 5–8 random regions of interest within the area of the defect were analyzed per section. QuPath82 was used to quantify Pax7+ nuclei using the cell detection module. Data exported from QuPath was parsed and summarized with RStudio.
Quantification of muscle fibers
Slides stained with MF20 were imaged with a 20x objective on an AxioScan 7 Slide Scanner (ZEISS), and stitched images for whole sections were exported as CZI files. Image analysis was performed on QuPath82. Masks were created for the region of the implant/defect, and another mask was extended 500 µm into the muscle for the interface region. At least three tissue sections were analyzed per animal. The threshold pixel classifier was used to identify muscle fibers (MF20+) in the whole section, and the masks were used to quantify the percentage of positively labeled pixels by location. Data exported from QuPath was parsed and summarized with RStudio.
Statistics
Statistical analyses were performed in GraphPad Prism 10. All values are expressed as the mean ± standard deviation. The ROUT method was used for outlier analysis.