The polarization of microglia and infiltrated macrophages in the injured mice spinal cords: a dynamic analysis

Animals

A total of 150 specific-pathogen free adult female C57BL/6 mice (18–20 g) were obtained from Chang Zhou Cavens Laboratory Animal Ltd. (Chang Zhou, China; license No. SCXK (Su) 2016-0010). Animals were housed as previously described (Chen et al., 2021, 2020). All experimental designs and reports were referred to previous to the previous guidelines (Kilkenny et al., 2011). The surgery protocol was approved by the Animal Care Ethics Committee of Bengbu Medical College. The Animal Ethical Approval number was 2017-037. The mice were randomly divided into sham-operated (sham), 1-, 3-, 7-, 14-, 21- and 28-days post-injury (dpi) groups, using a computer based random order generator (Zhao et al., 2018). The comprehensive description of the total number of mice used is shown in Fig. 1.

Contusive SCI

The mice contusive SCI model was established as described previously (Chen et al., 2021, 2020). Specifically, the Infinite Horizon impactor was made by Precision Systems & Instrumentation (Lexington, KY, USA). Anesthetics ( 80 mg/kg ketamine and 10 mg/kg xylazine) for intraperitoneal injection were obtained from Sigma-Aldrich (St. Louis, MO, USA). The T9 spinal cord was impacted with 50 Kdynes force, and the diameter of the impact rod was 1.3 mm. After impact, the spinal cord was filled with blood and edema. Sham-operated mice only underwent laminectomy without contusion. After operation, the animal care and welfare were performed as previously described (Chen et al., 2021, 2020), which included bladder emptying three times per day, relieving pain with meloxicam (5 mg/kg; CSN pharm, Chicago, IL, USA), and preventing infection with chloramphenicol (50 mg/kg; Sangon Biotech, Shanghai, China) for 7 days after surgery. Inclusion criteria: the animals underwent successful contusive SCI, defined by the T9 site filled with blood and edema, and the spinal cord was intact and not ruptured. Exclusion criteria: the degree of injury was not up to standard, postoperative infection or sacrifice.

Flow cytometry

At the indicated time points post-injury, the samples were collected and single-cell suspensions were obtained using Percoll gradient centrifugation as previously described (Chen et al., 2021, 2020). Briefly, the chest cavity was opened with surgical scissors to expose the heart. The ventricle was clamped with a vascular clamp to fix the heart. The No. 7 needle was inserted into the left ventricle. At the same time, a small opening was cut on the right atrium so that the blood and lavage solution could be drained. Then, 10 ml of 0.01 M phosphate-buffered saline (PBS) buffer solution (pH = 7.4) was slowly injected at 250 ml/h with a microinjection pump. After perfusion, the 5 mm spinal cord segments which contained the injury center were taken, and the corresponding spinal cord segments were also obtained from sham group. The spinal cords were put into the 45-m nylon mesh and fully ground with the syringe plunger to obtain single cell suspensions. To obtain enough cells for analysis, three spinal cord segments were mixed for one test. The Percoll gradient centrifugation (Amersham Pharmacia Biotech, Piscataway, NJ, USA) was used to separate the single cells. Table 1 showed the fluorescent labeled antibodies used in this study to identify different immune cell subtypes. To eliminate the background staining caused by the non-specific binding of the antibody, the immunoglobulin with the same species, subtype, dose and fluorescein as the primary antibody was used as the isotype control. The cells were collected using a BD Accuri flow cytometer (Becton Dickinson, San Diego, CA, USA), and the data were analyzed using FlowJo7.6.1 software (FlowJo; LLC, Ashland, OR, USA).

Immunofluorescence double-staining

At the indicated time points post-injury, mice were euthanized and perfused with PBS as described in “Flow cytometry”. Then, the mice were perfused with 10 mL of 4% paraformaldehyde (PFA) at a rate of 180 mL/h. After perfusion, the 5 mm spinal cord segments which contained the injury center were collected and fixed in 10 mL of 4% PFA solution at 4 °C overnight. The next day, the spinal cords were removed from 4% PFA solution and placed in 20% sucrose solution (prepared in PBS) at 4 °C overnight. The third day, the spinal cords were transferred to 30% sucrose at 4 °C, until the samples sinking to the bottom. This process usually needs 1 day. Next, the embedding agent (Tissue-Tek; Sakura Finetek USA Inc., Torrance, CA, USA) was used to embed the spinal cord segments at −20 °C. The 6 μm thick transverse sections were cut using a Leica CM1900 cryostat (Leica Microsystems, Bannockburn, IL, USA). The IHF assay was performed as previously described (Chen et al., 2021, 2020). Briefly, the slides were washed three times with 0.01 M PBS to completely clear the embedding agent. When the slides were left to dry, the blocking solution (0.01 M PBS containing 10% normal goat serum) were used for 2 h at room temperature to eliminate the background staining caused by the non-specific binding of antibodies. After cleaning the blocking solution, the primary antibodies with appropriate concentration were incubated overnight at 4 °C. The next day, the slides were washed three times with 0.01 M PBS to completely remove the unbound antibodies. Then, the secondary antibodies with appropriate concentration were incubated at 37 °C for 1 h. The primary and FITC and RHO-conjugated secondary antibodies were shown in Table 1. After the second antibody incubation, the 0.01 M PBS was used to wash the slides for three times, and the 1 μg/ml Hoechst 33342 (Cat# B2261; Sigma-Aldrich, St. Louis, MI, USA) containing medium was used to coverslip the slides. Finally, the slides were examined using a ZWISS Axio observation microscope (Carl Zeiss, Oberkochen, Germany). The cell quantification was performed as previously described (Chen et al., 2021, 2020). Specifically, for each spinal cord, the cells of 5 complete cross-sections containing the injury epicenter (0 mm), rostral (1 and 0.5 mm) and caudal (−1 and −0.5 mm) were counted.

Statistical analyses

The SPSS software v.14.0 (SPSS Inc., Chicago, IL, USA) was used to statistical analysis. The non-parametric Kruskal Wallis analysis of variance (ANOVA) following by the individual Mann-Whitney U test was used. The P < 0.05 was considered to be statistically significant.

Temporal pattern of MG and infiltrated Mø following SCI: the flow cytometry (FCM) analysis

To determine the temporal pattern of MG and infiltrated Mø, a panel of cell markers (CD11b, CD45 and CD68) was examined by FCM. Here, CD45^high population was peripheral infiltrated leukocytes, CD68⁺CD11b⁺ population was activated Mø and MG, CD45^highCD11b⁺ population was peripheral infiltrated Mø, CD45^−/lowCD11b⁺ population was MG, CD45^highCD68⁺CD11b⁺ population was activated peripheral infiltrated Mø, CD45^−/lowCD68⁺CD11b⁺ population was activated MG, and CD45^highCD68⁻CD11b⁻ population was peripheral infiltrated leukocytes excluding Mø (Fig. 2A).

Figure 2B showed that CD11b⁺ cells had no significant difference among sham, 1 and 3 dpi groups (P > 0.05, n = 6). However, at 7 dpi, the proportion increased significantly and reached to peak, although decreasing at the later time points (14, 21 and 28 dpi), they remained at high levels comparing with sham, 1 and 3 dpi groups (P < 0.01, n = 6).

Figure 2C showed that CD68⁺ cells were the lowest in the sham group comparing with the injured groups (P < 0.01, n = 6). The proportions increased significantly after injury, reached to peak at 7 dpi, and maintained at high levels at 14, 21 and 28 dpi. There were no significant differences among 7, 14, 21 and 28 dpi (P > 0.05, n = 6). However, CD68⁺ cells in these four groups were significant more comparing with sham, 1 and 3 dpi (P < 0.05 or 0.01, n = 6).

Figure 2D showed that CD11b⁺CD68⁺ cells were extremely rare in sham group, however, they increased significantly in the injured groups (P < 0.01, n = 6). The proportions had no significant differences between 1 and 3 dpi (P > 0.05, n = 6). However, it reached to peak at 7 dpi, and then decreased, but remained at high levels at 14, 21 and 28 dpi.

Figure 2E showed that CD45^high cells were extremely rare in sham group, and they gradually increased after injury, peaked at 7 and 14 dpi, and then decreased, but still maintained at high levels at 21 and 28 dpi.

Figure 2F showed that CD11b⁺CD45^high cells were also extremely rare in sham group, and they significantly increased after injury, peaked at 7 dpi, and then decreased, but still maintained at high levels at 14, 21 and 28 dpi. The proportion of CD11b⁺CD45^high cells in each group is significantly lower than that of their corresponding CD11b⁺ cells (Fig. 2B).

Figure 2G showed that CD11b⁺CD45^−/low cells had no significant difference among sham, 1 and 3 dpi groups (P > 0.05, n = 6). However, at 7 dpi, the proportion increased significantly and reached a peak, although decreased at the later time points (14, 21 and 28 dpi), they still maintained at high levels comparing with sham, 1 and 3 dpi groups (P < 0.01, n = 6). Except for the 1 and 3 dpi groups, CD11b⁺CD45^−/low cells constitute the majority of CD11b⁺ cells (Fig. 2B).

Figure 2H showed that CD68⁺CD45^high cells were also extremely rare in sham group, and they rapidly increased after injury. Up to 28 dpi, they still maintained at high levels. The proportion of CD68⁺CD45^high cells in each SCI group is significantly lower than that of their corresponding CD68⁺ cells (Fig. 2C).

Figure 2I showed that the percentages ofCD68⁺CD45^−/low cells in sham group was the lowest comparing with the injured groups (P < 0.05 or 0.01, n = 6). The proportions had no significant differences at 1 and 3 dpi (P > 0.05, n = 6). However, they reached to peak at 7 and 14 dpi, and remained at high levels at 21 and 28 dpi. The proportion of CD68⁺CD45^−/low cells in each SCI group is significantly lower than that of their corresponding CD68⁺ cells (Fig. 2C).

Figure 2K showed that CD68⁺CD11b⁺CD45^high cells were extremely rare in sham group, and they rapidly increased after injury, peaked at 1, 3 and 7 dpi. Compared with the sham group, these cells in the 1, 3 and 7 dpi groups were significantly more (P < 0.01, n = 6). Although, comparing with 1, 3 and 7 dpi groups, the proportions decreased to the lower levels at the later time points (14, 21 and 28 dpi) (P < 0.05, n = 6), they still maintained at higher levels comparing to the sham group (P < 0.05, n = 6). Comparing with Fig. 2H, the proportion of cells in Fig. 2K is lower. This indicated that the activated peripheral infiltrated Mø in the injured spinal cord are significantly inferior to activated MG.

Figure 2L showed that CD68⁺CD11b⁺CD^45−/low cells were also extremely rare in the sham group, and they gradually increased after injury, peaked at 7 dpi, and then decreased, but still maintained at high levels at 14, 21 and 28 dpi. Comparing to sham group, the percentages of CD68⁺CD11b⁺CD^45−/low cells in all SCI groups had significant differences (P < 0.01, n = 6).

Temporal pattern of MG and infiltrated Mø following SCI: the immunohistofluorescence (IHF) analysis

To verify the temporal pattern of MG and infiltrated Mø detected by FCM, the spinal cords from several representative time points after SCI (sham, 1, 7 and 28 dpi) were selected for IHF analysis. CD11b, CD68 and TMEM119 antibodies were used for immunofluorescence labeling (Fig. 3). Here, TMEM119⁺CD11b⁺ cells are total MG, TMEM119⁻CD11b⁺ cells are monocyte-derived Mø, TMEM119⁺CD68⁺ cells are activated MG, TMEM119⁻CD68⁺ cells are activated monocyte-derived Mø, respectively (Figs. 3A–3H).

The representative images showed that TMEM119⁺CD11b⁺ cells could be detected in all groups (Figs. 3A–3D). The statistical results (Fig. 3E) showed that the number (cells/mm²) of TMEM119⁺CD11b⁺ cells had no significant difference between sham (117.50 ± 19.30) and 1 dpi (200.33 ± 16.59) groups (P > 0.05, n = 6). Comparing with the other three groups, there were more TMEM119⁺CD11b⁺ cells in the 7 dpi (537.33 ± 99.80) (P < 0.01, n = 6). Although, the cells were decreased at 28 dpi (308.33 ± 50.27), the number still significantly more comparing with sham and 1 dpi groups (P < 0.01, n = 6). In the sham group, TMEM119⁻CD11b⁺ cells were extremely rare (Fig. 3A), and they significantly increased in the three SCI groups (Fig. 3B-D). The statistical results (Fig. 3F) showed that the numbers of TMEM119⁻CD11b⁺ cells in all three SCI groups were significant more than that of sham (70.17 ± 7.65) group (P < 0.01, n = 6). Comparing with the other three groups, there were also more TMEM119⁻CD11b⁺ cells in 7 dpi (201.80 ± 42.53) group (P < 0.05, n = 6).

In Fig. 4A, both TMEM119⁺CD68⁺ and TMEM119⁻CD68⁺ cells were extremely rare in sham group. However, both of them could be detected in all SCI groups (Figs. 4B–4D). The statistical results (Figs. 4E and 4F) showed that the numbers of these two types of cells had significant differences among sham and SCI groups (P < 0.01, n = 6). Compared with the three other groups, there were most TMEM119⁺CD68⁺ cells in 7 dpi (473.50 ± 64.48) group (P < 0.01, n = 6). Although, the number of these cells decreased at 28 dpi (269.67 ± 42.49), it still significantly more comparing with sham (6.67 ± 6.31) and 1 dpi (156 ± 43.75) groups (P < 0.01, n = 6). Comparing with the other two groups, TMEM119⁻CD68⁺ cells were most in 1 (155.00 ± 18.51) and 7 dpi (124.75 ± 34.88) groups (P < 0.01, n = 6).

Temporal pattern of SCI-induced M1 and M2 differentiation of Mø and MG: the FCM analysis

To further explore the temporal pattern of SCI-induced M1 and M2 differentiation of Mø and MG, FCM was used by combining CD68, CD45, CD11b and CCR7 antibodies.

As shown in Fig. 5A, the same size “region” of total CD11b⁺ cells (R1) were set for each sample in the pseudocolor plots of CD45/CD11b, and then the percentage of each cell population was analyzed in the pseudocolor plots of CD68/CCR7 by setting the boundary between negative and positive with isotype-matched antibodies. The statistical results (Fig. 5B) showed that the percentage of CD11b⁺CD68⁺CCR7⁺ M1 cells in the sham group was the lowest. The proportions significantly increased after injury, reached to peak from 7 dpi, and maintained at high levels at 14, 21 and 28 dpi. There were no significant differences among 7, 14, 21 and 28 dpi (P > 0.05, n = 6). However, the percentages in these four groups were significant higher comparing with sham and 1 dpi groups (P < 0.01, n = 6).

Figure 5C showed that CD11b⁺CD68⁺CCR7⁻ M2 cells had no significant differences among all groups (P > 0.05, n = 6). However, when converted to ratio (Fig. 5D), the total M1/M2 ratios in all SCI groups were significant increased comparing with the sham group (P < 0.01, n = 6).

As shown in Fig. 5E, in the pseudocolor plots of CD11b/CD45, the same size region of CD11b⁺CD45^−/low cells (R2) were set for each sample, and then the percentages of cell populations were analyzed same as Fig. 5A. The statistical results (Fig. 5F) showed that the percentage of CD11b⁺CD45^−/lowCD68⁺CCR7⁺ MG-derived M1 cells in sham group was the lowest comparing with the injured groups, and the differences were statistically significant (P < 0.01, n = 6). The proportions significantly increased after injury, and reached to peak at 14 and 21 dpi, and continued at high level at 28 dpi. The percentages in the groups of 3, 7, 14, 21 and 28 dpi were significantly higher than that of 1 dpi (P < 0.05 or 0.01, n = 6).

Figure 5G showed that CD11b⁺CD45^−/lowCD68⁺CCR7⁻ MG-derived M2 cells had no significant differences among all groups (P > 0.05, n = 6). However, when converted to ratio (Fig. 5H), the M1/M2 ratio of MG was very low in sham group, and they increased after SCI, peaked at 7 dpi, and then decreased at 21 and 28 dpi. In the group of 7 dpi, the ratio was significantly higher than those of the other groups (P < 0.05, n = 6).

In Fig. 5I, in the pseudocolor plots of CD11b/CD45, the same size region of CD45^highCD11b⁺ cells (R3) were set for each sample, and then the percentages of cell populations were analyzed same as Fig. 5A. Figure 5J showed that the percentages of CD11b⁺CD45^highCD68⁺CCR7⁺ peripheral infiltrated M1 cells showed an increasing trend after SCI, and reached the highest levels at 28 dpi. The 7, 14, 21 and 28 dpi groups had statistically significant comparing with sham and 1dpi (P < 0.05 or 0.01, n = 6). Figure 5K showed that the percentages of CD11b⁺CD45^highCD68⁺CCR7⁻ peripheral infiltrated M2 cells were highest at 3 and 7 dpi, which were statistically significant comparing with the other groups (P < 0.01, n = 6). When converted to ratio (Fig. 5l), the infiltrated M1/M2 ratio was very low in sham group, and there was a transient rise at 1 dpi. Then, the ratios decreased to sham level at 3 and 7 dpi, and then showed an increasing trend from 14 to 28 dpi, it reached the highest levels at 28 dpi. Among sham, 3 and 7 dpi groups, the ratios had no significant differences (P > 0.05, n = 6). However, comparing with the other groups, the ratios were significant lower (P < 0.05 or 0.01, n = 6).

Temporal pattern of SCI-induced M1 and M2 differentiation of Mø and MG: the IHF analysis

In IHF analysis, CD68⁺CCR7⁺ and CD68⁺Arg1⁺ were used to label total M1 and M2 cells, respectively (Fig. 6); TMEM119⁺ CCR7⁺ and TMEM119⁺Arg1⁺ cells were M1 and M2 MG, respectively (Figs. 7A–7J). Therefore, (CD68⁺CCR7⁺ minus TMEM119⁺CCR7⁺) and (CD68⁺Arg1⁺ minus TMEM119⁺Arg1⁺) were M1 and M2 monocyte-derived Mø, respectively (Figs. 7K and 7L).

The representative images showed that CD68⁺CCR7⁺, CD68⁺Arg1⁺, TMEM119⁺CCR7⁺ and TMEM119⁺Arg1⁺ cells were both extremely rare in sham-operated spinal cords (Figs. 6A and 6E, 7A and 7E). However, these cells could be detected in all SCI groups (Figs. 6B–6D, 6F–6H, and 7B–7D, 7F–7H).

The statistical results (Fig. 6I) showed that in the groups of sham, 1, 7 and 28 dpi, the numbers of CD68⁺CCR7⁺ cells were 2.67 ± 2.50, 235.33 ± 5.13, 577.17 ± 40.18 and 543.17 ± 31.35, respectively. All SCI groups were significant more than that of sham (P < 0.01, n = 6). Up to 7 dpi, the cell number reached to peak, and continued at high levels at 28 dpi.

In Fig. 6J, the numbers of CD68⁺Arg1⁺ cells in the groups of sham, 1, 7 and 28 dpi were 3.83 ± 3.43, 105.33 ± 9.56, 72.83 ± 10.55 and 150.17 ± 22.21, respectively. Although, CD68⁺Arg1⁺ cells in all SCI groups were significant more than that of sham group (P < 0.01, n = 6), the 7 dpi had the least number of cells among the three SCI groups, and there were most cells in 28 dpi group.

In Fig. 7I, TMEM119⁺CCR7⁺ cells in the groups of sham, 1, 7 and 28 dpi were 3.17 ± 1.83, 205.17 ± 9.97, 412.33 ± 18.04 and 410.17 ± 50.92, respectively. The overall change trend was similar to that of CD68⁺CCR7⁺ cells. Figure 7J showed that TMEM119⁺Arg1⁺ cells in the groups of sham, 1, 7 and 28 dpi were 4.33 ± 2.42, 85.83 ± 10.68, 50.00 ± 4.29 and 125.00 ± 13.33, respectively. The overall trend was also similar to that of CD68⁺Arg1⁺ cells.

In Fig. 7K, the infiltrated CD68⁺CCR7⁺ cells in the groups of sham, 1, 7 and 28 dpi were 0.00 ± 0.00, 30.17 ± 11.72, 164.83 ± 45.64 and 133.00 ± 56.67, respectively. All SCI groups were significant more than that of sham group (P < 0.01, n = 6). Up to 7 dpi, the cell number reached to peak. Although the number of cells had a decreasing trend, it remained at a higher level at 28 dpi. Figure 7L showed that although the infiltrated CD68⁺Arg1⁺ cells could be detected in SCI groups, they were very rare. The numbers in the groups of sham, 1, 7 and 28 dpi were 0.00 ± 0.00, 19.50 ± 12.65, 22.83 ± 12.04 and 25.17 ± 15.74, respectively. All three SCI groups were significant more than that of sham group (P < 0.01, n = 6).

Summary of dynamic changes of MG, infiltrated Mø and their subsets

As shown in the first row of Fig. 8, the proportions of total Mø/MG in the groups of SCI were gradually increased and peaked at 7 dpi. Although, decreasing at 14, 21 and 28 dpi, they were still maintained at high levels compared with sham, 1 and 3 dpi groups.

The second row of Fig. 8 showed that most of the Mø/MG were activated following SCI. The third row of Fig. 8 showed that MG were absolutely dominant and Mø were few in the sham-operated spinal cords. In the acute phase of SCI (1 and 3 dpi), the proportions of activated Mø increased significantly. However, with the progression of the SCI, the proportions of activated MG increased nearly to 90% in subacute phase (7 and 14 dpi) and chronic phase (21 and 28 dpi).

Finally, we further analyzed the proportion of M1 and M2 subsets in the activated MG and infiltrated Mø. The fourth row 4 of Fig. 8 showed that in sham, 1, 3, 7, 14, 21 and 28 dpi groups, the proportions of M1 Mø were 7%, 43%, 51%, 7%, 9%, 17% and 12%, respectively, while M2 Mø were 6%, 6%, 23%, 3%, 1%, 1% and 0%, respectively. The proportions of M1 MG were 31%, 43%, 20%, 69%, 79%, 66% and 68%, respectively, and M2 MG were 56%, 8%, 6%, 21%, 11%, 16% and 20%, respectively. These results showed that there are very few peripheral Mø in the sham-operated spinal cords, while the proportion of MG is absolutely dominant, and the MG are mainly M2 subtype. In the acute phase of SCI, the proportion of peripheral infiltrated Mø increased transiently, and M1 Mø are absolutely dominant, but in the subacute phase, M1 MG were absolutely dominant and continued to the chronic phase (28 dpi, the longest time point observed in this study).