Animals

Female C57BL/6J mice aged 8–10 weeks were obtained from the Experimental Animal Center of Anhui Medical University for this study. All experimental procedures were performed in compliance with the Institutional Animal Care Guidelines of Anhui Medical University and approved by the Animal Ethics Committee of Anhui Medical University (Approval No. LLSC20160052). The mice were housed in specific pathogen-free facilities under a 12-h light-dark cycle at a temperature of 24 ± 1 °C with controlled humidity, and had access to food and water.

Spinal cord Injury Model

The mice were anesthetized with intraperitoneal injection of sodium pentobarbital (50 mg/kg; Harbin Pharmaceutical Group Co., Ltd., Harbin, China) and placed in the prone position on the operating table. The dorsal surface around the T10 segment was shaved, and the skin was disinfected with iodophor. The T9-T11 dorsal skin was dissected, the back muscle was sequentially peeled off. Laminectomy was performed to expose the spinal cord at the T10 level, which served as the clamping site. Crush-induced spinal cord injuries were made using calibrated Dumont #5 forceps (11252-20, Fine Science Tools, Germany) without spacers, with a tip width of 0.5 mm, to completely compress the entire spinal cord laterally from both sides for 5 s [22]. Successful SCI induction was confirmed by observing transient spasms in the hind limbs and tail along with a continuous red clamp mark after saline irrigation. The wound was then sutured layer by layer. Postoperatively, the mice were subjected to auxiliary urination twice daily (morning and afternoon) until spontaneous urination was restored after SCI.

In situ injection of PDGFD or rGalectin-3

Mice in the experimental group received in situ injection of PDGFD or recombinant Galectin-3 (rGalectin-3) in the uninjured spinal cord. After the T10 spinal cord was exposed as described above, each mouse was placed in a stereotaxic device. The microinjection needle (7634-01 and 7803-05, Hamilton, Switzerland) was inserted 0.3 mm lateral to the midline and 0.8 mm deep into the mouse spinal cord [6]. Two microliters of 100 ng/µl recombinant human PDGFD (1159-SB/CF, R&D Systems, United States) dissolved in 4 mM HCl containing 0.1% BSA or 100 µg/ml rGalectin-3 (HY-P70309, MedChemExpress, United States) dissolved in 100 µl of ddH2O was injected into the uninjured spinal cord at a rate of 0.5 µl/min using a stereotaxic injector (KDS LEGATO 130, RWD, China). The control mice received only 2 µl of ddH2O. All in situ-injected mice were sacrificed at 7 days post-injection.

Intrathecal injection of TD139

Prior to injection into the lumbar 5–6 intervertebral space in mice, the dorsal surface was shaved, and the skin was disinfected with iodophor. The successful insertion of the needle into the intradural space was confirmed by sudden tail wagging [22]. Thirty milligrams of TD139 (GC19350, GLPBIO, United States) dissolved in 10 ml PBS (Servicebio, China) containing 3% DMSO (3304, R&D Systems, United States) was injected daily at 1 µl/4 s using a microinjection needle (1701, Hamilton, Switzerland). For the uninjured mice, TD139 was injected immediately after the injection of PDGFD or rGalectin-3 and then injected daily for 7 consecutive days. For the SCI mice, TD139 was injected post-operatively daily until 14 dpi. The control mice received the same volume of 10 µl of PBS containing 3% DMSO.

In situ injection of lentivirus

The lentiviral vectors utilized in this study were designed by GenePharma (Shanghai, China) and had a virus titer of Lv-shLgals3 at 1.0 × 10E9 TU/ml. The shRNA sequences targeting lgals3 (which encodes Galectin-3) were 5′-ACCCAAACCCTCAAGGATATC-3′ (Lv-shLgals3-144, referred to as Lv-shLgals3 below), 5′-GTAACACGAAGCAGGACAATA-3′ (Lv-shLgals3-657) and 5′-GCTCACCTACTGCAGTACAAC-3′ (Lv-shLgals3-785). A lentivirus expressing a nonspecific shRNA sequence (5′-TTCTCCGAACGTGTCACGT-3′) (Lv-shNC) was employed as a negative control.

The T10 spinal cord crush injury was established according to the experimental methods described above. Subsequently, 1 µl of lentivirus carrying Galectin-3-shRNA (Lv-shLgals3) or empty lentiviral vector (Lv-shNC) was injected in situ at the T10 injury site using a microinjection needle (7634-01 and 7803-05, Hamilton, Switzerland). All mice that received in situ injections of lentivirus were sacrificed at 28 dpi.

Cell Culture

Mouse mononuclear macrophage leukemia cells (RAW 264.7) were purchased from the Stem Cell Bank, Chinese Academy of Sciences (Shanghai, China) and cultured in complete medium (CM-0190, Procell Life Science & Technology, China). Mouse brain vascular pericytes (referred to as perivascular fibroblasts) were purchased from the National Collection of Authenticated Cell Cultures (Shanghai, China), cultured in complete medium (CM-M222, Procell Life Science & Technology, China), and induced with 250 ng/ml PDGF-BB (220-BB-050, R&D Systems) for 72 h [22]. The cells were cultured in a humidified atmosphere at 37 °C containing 5% CO2.

Cell treatment

RAW 264.7 macrophages in 24-well plates were transduced with Lv-shNC, Lv-shLgals3-144 (Lv-shLgals3), Lv-shLgals3-657, or Lv-shLgals3-785 at a multiplicity of infection of 100, according to the manufacturer’s protocol (GenePharma). The macrophages were transduced with these lentiviral vectors for 24 h.

For the Galectin-3 internalization experiments, the control group comprised fibroblasts cultured in a 1:1 mixture of fibroblast culture medium and macrophage complete culture medium, and the MP-CM group consisted of fibroblasts cultured with in a 1:1 mixture of fibroblast culture medium and macrophage supernatant conditioned culture medium (MP-CM) for 24 h. The rGalectin-3 group was composed of fibroblasts cultured in medium supplemented with rGalectin-3 (10 µg/ml) for 24 h, and the rGalectin-3 + TD139 group consisted of fibroblasts treated with rGalectin-3 (10 µg/ml) and TD139 (10 µM) for 24 h. The Lv-shNC and Lv-shLgals3 groups were composed of fibroblasts cultured for 24 h with conditioned medium from macrophages transduced with Lv-shNC (10 µl/ml) or Lv-shLgals3 (10 µl/ml), respectively. Immunofluorescence staining of PDGFRβ, Galectin-3, and phalloidin in fibroblasts was performed for the Galectin-3 internalization experiments.

Histology and immunofluorescence staining

At predetermined time points, each mouse was anesthetized, and cardiac perfusion was performed with cold 0.1 M PBS followed by 4% paraformaldehyde (PFA, Servicebio, China). The tissue near the injury core at the T10 level (3 mm above and below) was removed, fixed in PFA for 5 h, and dehydrated in 30% sucrose solution overnight at 4 °C, embedded in optimum cutting temperature (OCT, BL557A, Biosharp, China) compound, and cut into continuous 16-µm frozen sections with a cryostat (NX50, Thermo Fisher Scientific, United States). The frozen spinal cord sections were washed three times for 5 min each with 0.1 M PBS, blocked in 5% donkey serum in PBS containing 0.3% Triton X-100 (T8200, Solarbio, China) for 1 h at room temperature, and then incubated with primary antibodies in 1% donkey serum containing 0.3% Triton X-100 overnight at 4 °C. The sections were then incubated for 1 h at room temperature with secondary antibodies. The nuclei were stained and sealed with 4,6-diamidino-2-phenylindole (DAPI with an antifluorescence quencher, P0131-25 ml, Beyotime, China) for 5 min. The following primary antibodies were used at the indicated dilutions: goat anti-PDGFRβ (5 µg/ml, AF1042-SP, R&D Systems), rat anti-Galectin-3 (1:200, SC-23938, Santa Cruz Biotechnology), rabbit anti-fibronectin (1:100, 15613-1-AP, Proteintech), rabbit anti-laminin (1:100, 23498-1-AP, Proteintech), rabbit anti-F4/80 (1:100, 28463-1-AP, Proteintech), rabbit anti-neurofilament-heavy (NF-H) (1:400, ab207176, Abcam), rat anti-CD68 (1:300, MCA1957, AbD Serotec), rat anti-GFAP (1:300, 13–0300, Thermo Fisher Scientific), goat anti-5-HT (1:500, #20080, Immunostar), and rabbit anti-NeuN (1:400, ab177487, Abcam). The secondary antibodies used were donkey anti-rat Alexa Fluor 488, donkey anti-rabbit Alexa Fluor 488 (1:500, A-21206, A-21208, Invitrogen), donkey anti-rabbit Alexa Fluor 555, donkey anti-rat Alexa Fluor 555, and donkey anti-goat Alexa Fluor 555 (1:500, A-21432, A-31572, A-48270, Invitrogen). Immunofluorescence images were captured under an Axio Scope A1 microscope (Zeiss, Germany) with a confocal microscope (LSM 900, Zeiss, Germany). Staining colocalization images were created by Zen 3.3 software.

Immunocytochemistry

The cells were fixed with 4% PFA at room temperature for 10 min, blocked in 5% donkey serum in PBS containing 0.3% Triton X-100 for 1 h at room temperature, and then incubated overnight at 4 °C with primary antibodies in 1% donkey serum containing 0.3% Triton X-100. The cells were then incubated for 1 h at room temperature with secondary antibodies and washed with PBS. The nuclei were stained and sealed with DAPI for 5 min. The following primary antibodies were used at the indicated dilutions: goat anti-PDGFRβ (5 µg/ml, AF1042-SP, R&D Systems) and rat anti-Galectin-3 (1:200, SC-23938, Santa Cruz Biotechnology). The secondary antibodies used included donkey anti-rat Alexa Fluor 555 (1:500, A-48270, Invitrogen), donkey anti-goat Alexa Fluor 647 (1:500, A-21447, Invitrogen), and Actin-Tracker Green-488 (Phalloidin) (1:100, C2201S, Beyotime). Immunofluorescence images were captured under an Axio Scope A1 microscope and a confocal microscope Zen 3.3 software was used to create images of the stained cells.

Quantitative analysis

Every 5th 16-µm frozen sagittal section that included the entire injured spinal cord was quantified. To evaluate spatiotemporal changes in PDGFRβ+ fibroblasts and Galectin-3+ macrophages, the maximum range of the positive cells spanning the craniocaudal range at the center of the injury site was quantified for fibroblasts and macrophages at 3–56 dpi [23]. To determine the source of Galectin-3 after SCI, the percentage of Galectin-3+ macrophages relative to the total number of F4/80+ macrophages at the injured spinal cord was quantified. To quantify PDGFRβ+ cell density at injury site after TD139 and Lv-shLgals3 injection, only PDGFRβ+ cells that were also DAPI+ were counted in the GFAP− region. The number of PDGFRβ+ cells in each section was normalized to the area of the GFAP− region. For each mouse, sections including the injury site and two adjacent sagittal sections spaced 160 μm apart were quantified to the area of the GFAP− region, and the counts from at least three sections were averaged [8]. To quantify the area of the fibrotic scar, the immunoreactivities of PDGFRβ, fibronectin, and laminin were normalized through threshold processing, and the areas of the spinal cord segment spanning the lesion site covered by threshold regions were calculated in a 4 × magnification [8]. The immunoreactivity of Galectin-3 or the percentage of the GFAP− area was also normalized through threshold processing to the area of the spinal cord segment spanning the lesion site in a 4 × magnification. The PDGFRβ+, fibronectin+, laminin+, Galectin-3+, and GFAP− area were expressed as a percentage of the area of the 2 mm spinal cord segment evaluated. To assess the density of fibers growing at the injury core, threshold regions of the NF-H were measured through threshold processing and according to our previous research [22]. To evaluate axonal regeneration passing through the central core of the injury, the distance from the tip of 5-HT+ axons to the GFAP+ edge at the rostral end was measured. To assess neuronal preservation, the number of NeuN+ cells in the Z1 (0–250 μm), Z2 (250–500 μm), and Z3 (1000–1250 μm) zones adjacent to the lesion core was quantified [24]. The results from each section were averaged with 3–5 samples per group. Image processing was performed using ImageJ/Fiji version 2.3 (NIH, United States). For the in vitro assay, the overall fluorescence intensity of Galectin-3 and PDGFRβ was quantified using ImageJ software after intensity thresholding. The mean fluorescence intensity was calculated for analysis [25].

Behavioral tests

The Basso Mouse Scale (BMS) was used to evaluate hindlimb motor function after SCI in mice [6]. The walking and limb activity scores of the hindlimbs were observed and recorded according to the protocol developed by Basso and colleagues [26]. Two experienced examiners scored each mouse at 0, 1, 3, 7, 14, 21, and 28 dpi. Gait footprint analysis and motor coordination were assessed at 28 dpi [22]. The forelimb and hind paw were coated with dye of different colors, and the mice were placed on a 4 cm × 80 cm runway covered with white paper. The animals were encouraged to walk straight to the finish line so that representative images of their gaits could be obtained, and stride length, stride width, and paw rotation were assessed to evaluate motor function.

Statistical analysis

All the data were expressed as the mean ± standard error of the mean (s.e.m). GraphPad Prism 7.0 (GraphPad Software Inc, United States) was used for statistical analysis. One-way analysis of variance (ANOVA), and two-way ANOVA followed by Tukey’s post hoc test were performed to detect differences among multiple groups, or a two-tailed Student’s t test was used to compare two groups. A p value < 0.05 was considered to indicate statistical significance.