The temporal expression and cellular localization of CS-A and CS-C proximal to the lesion following SCI.
In the present study, we first examined the expression of CS-A and CS-C moiety of CSPGs proximal to the lesion following SCI via immunofluorescent staining. Compared to very low levels of reactivity for CS-A in the sham-operated group, we found immediately increased staining for CS-A proximal to the lesion at 1-day post-SCI that reached a peak at 14-days post-SCI. Moreover, upregulated CS-A was detected far out as 60-days post-SCI. Next, we found that LDI treatment restored the expression of CS-A to a level near that of the sham-operated group (Fig. 1A). Immunohistochemical analysis showed that, at 1-day post-SCI, 6.0 ± 1.33 cells/cm2 were immunoreactive for CS-A in the sham-operated group, which increased to 16.3 ± 2.78 cells/cm2 in the SCI group and was attenuated to 10.02 ± 0.98 cells/cm2 after LDI treatment (F (2, 12) = 7.12, p = 0.009). At 14-days post-SCI, the CS-A expression level was 6.2 ± 1.58 cells/cm2 in the sham-operated group, which was increased to 32.7 ± 5.35 cells/cm2 in SCI group and was attenuated to 15.3 ± 2.19 cells/cm2 after LDI treatment (F (2, 12) = 7.39, p = 0.008).
The temporal expression of CS-C was different from that of CS-A. Our results demonstrated that no significant variation in CS-C expression was found among the sham-operated, SCI and LDI treatment groups at 1-day post-SCI. However, CS-C was increased by 14-days post-SCI and reached a peak by 60-days post-SCI. Strikingly, in contrast to the results of CS-A expression, LDI treatment showed no inhibitory effect on CS-C expression following SCI (Fig. 1B). Immunohistochemical analysis showed that 5.8 ± 2.23 cells/cm2 were immunoreactive for CS-C in the sham-operated group, which increased to 13.3 ± 3.27 cells/cm2 at 14-days post-SCI in the SCI group and was similarly 12.1 ± 1.88 cells/cm2 after LDI treatment (F (2, 12) = 7.59, p = 0.007). At a later time point, there were 6.1 ± 1.09 cells/cm2 in the sham-operated group, which was increased to 35.38 ± 2.29 cells/cm2 at 60-days post-SCI in SCI group and was similarly 33.51 ± 1.97 cells/cm2 after LDI treatment (F (2, 12) = 7.62, p = 0.007).
We next determined the spatial expression of CS-A and CS-C proximal to the lesion. Double-immunofluorescent staining was performed with the following specific antibodies at the time points when CS-A and CS-C expressions reached their corresponding peaks: GFAP for astrocytes; MAP-2 for neurons and axons; LY111 for CS-A; and MC21 for CS-C. At 14-days post-SCI, we found substantial staining for CS-A proximal to the lesion, and colocalization of CS-A and GFAP was also apparent. On the contrary, a thin expression of CS-A was observed in MAP-2 positive neurons/axons (Fig. 1C). At 14-days post-SCI, the percentage of CS-A-positive astrocytes was 8.7 ± 0.93% in the sham-operated group, which increased to 26.8 ± 2.09% after SCI (F (1,8) = 12.81, p = 0.000); the percentage of CS-A-positive neurons/axons was 5.5 ± 0.86% in the sham-operated group, which was similarly 5.8 ± 0.79% after SCI (F (1,8) = 1.25, p = 0.100). At 60-days post-SCI, CS-C was upregulated in both astrocytes and neurons/axons proximal to the lesion. The percentage of CS-C positive astrocytes was 5.5 ± 0.73% in the sham-operated group, which increased to 12.2 ± 1.29% after SCI (F (1,8) = 10.82, p = 0.000); the percentage of CS-C positive neurons/axons was 7.5 ± 1.29% in the sham-operated group, which was increased to 39.8 ± 1.28% after SCI (F (1,8) = 1.35, p = 0.118).
LDI decreases astrocyte-associated CS-A expression, restrains CS-A-enriched PNN accumulation.
We had demonstrated that LDI inhibited CS-A expression, but not CS-C expression, proximal to the lesion following SCI. However, since CS-A is expressed in both astrocytes and neurons, we next explored whether LDI inhibits neuron-associated CS-A expression and/or astrocyte-associated CS-A expression at 14-days and 60-days post-SCI. There were rare CS-A positive neuron/axon in the sham-operated and SCI animals. LDI treatment did not alter CS-A expression in neuron/axon. At 14-days post-SCI, 3.2 ± 0.38% of neuron/axon were CS-A positive in the sham-operated group, which was similarly 3.7 ± 0.35% in the SCI group and 3.3 ± 0.02% after LDI treatment (F (2, 12) = 1.55, p = 0.252). At 60-days post-SCI, 3.5 ± 0.17% of neuron/axon were CS-A positive in the sham-operated group, which was similarly 3.8 ± 0.25% in the SCI group and 3.6 ± 0.55% after LDI treatment (F (2, 12) = 1.62, p = 0.238) (Fig. 2B). In contrast, LDI treatment significantly inhibited CS-A expression in astrocytes. At 14-days post-SCI ,1.2 ± 0.58% of astrocytes were CS-A positive in the sham-operated group, which increased to 35.3 ± 3.26% in the SCI group and was attenuated to 5.1 ± 0.91% after LDI treatment (F (2, 12) = 7.99, p = 0.006). At 60-days post-SCI, 1.5 ± 0.61% of astrocytes were CS-A positive in the sham-operated group, which increased to 39.8 ± 2.15% in the SCI group and was attenuated to 6.3 ± 1.03% after LDI treatment (F (2, 12) = 7.85, p = 0.007) (Fig. 2A). These results revealed that SCI increased astrocytic but not neuronal CS-A expression, and that LDI treatment induced significant inhibition of the aberrantly elevated astrocytic CS-A expression following SCI, but did not significantly affect the already stable neuronal CS-A expression in neurons following SCI.
We further used double-immunofluorescent staining and Western blotting to examine whether protein expression of C4ST-1 changed in the spinal cord tissue proximal to the lesion after LDI treatment. C4ST-1 expression was significantly increased in astrocytes following SCI in the spinal cord tissue proximal to the lesion compared to a time-matched sham-operated group (Fig. 2C) (see supplemental 2). Interestingly, after LDI treatment, C4ST-1 expression in astrocytes was significantly reduced to a level close to that of the sham-operated group. Additionally, double-immunofluorescent staining showed that 7.0 ± 0.78% of astrocytes were C4ST-1 positive in the sham-operated group, which increased to 20.5 ± 2.37% in the SCI group and was attenuated to 9.12 ± 0.56% after LDI treatment at 14-days post-SCI (F (2, 12) = 8.01, p = 0.006). Later, at 60-days post-SCI, 7.2 ± 0.83% of astrocytes were C4ST-1 positive in the sham-operated group, which was increased to 19.5 ± 1.17% in SCI group and was attenuated to 9.06 ± 0.88% after LDI treatment (F (2, 12) = 7.89, p = 0.006). Complementing these results, Western blotting showed that, in the SCI group, the protein expression of C4ST-1 increased to 213.5 ± 7.36% (14-days post-SCI) and 153.9 ± 5.61% (60-days post-SCI) compared to those of time-matched sham-operated animals, and LDI treatment restored the C4ST-1 protein expression to 136.3 ± 3.95% (14-days post-SCI) (F (2, 12) = 8.06, p = 0.006) and 120.9 ± 7.33% (60-days post-SCI) (F (2, 12) = 8.13, p = 0.005) of that of the sham-operated levels, respectively (Fig. 2D) (see supplemental 2).
We further investigated whether LDI treatment altered CS-A or CS-C-enriched PNNs following SCI. We used WFA to label CS-A-enriched PNNs, a CS56 antibody to label CS-C-enriched PNNs, and we quantitatively analyzed the CS-A/CS-C-enriched PNNs. We found that LDI decreased CS-A-enriched PNNs proximal to the lesion following SCI (Fig. 3A). At 14-days post-SCI, 6.3 ± 1.12% of PNNs were CS-A-enriched in the sham-operated group, which was increased to 29.3 ± 3.32% in SCI group and was attenuated to 8.85 ± 0.96% after LDI treatment (F (2, 12) = 7.93, p = 0.006). At 60-days post-SCI, 6.7 ± 1.07% of PNNs were CS-A-enriched in the sham-operated group, which was increased to 35.3 ± 1.89% in SCI group and was attenuated to 9.27 ± 1.03% after LDI treatment (F (2, 12) = 7.79, p = 0.007). However, LDI treatment had no significant inhibitory effect on CS-C-enriched PNNs expression. At 14-days post-SCI, 5.8 ± 0.89% of PNNs were CS-C-enriched in the sham-operated group, which increased to 22.5 ± 1.28% in the SCI group and was 19.25 ± 1.53% after LDI treatment (F (2, 12) = 1.89, p = 0.193). At 60-days post-SCI, 6.2 ± 1.01% of PNNs were CS-C-enriched in the sham-operated group, which increased to 31.5 ± 1.75% in the SCI group and was 28.2 ± 1.19% after LDI treatment (F (2, 12) = 1.92, p = 0.189) (Fig. 3B).
LDI increases expression of 5-HT axons accompanied with reduction of CS-A-enriched PNNs
Using laser-scanning confocal microscopy and different fluorescent probes, we measured 5-HT axonal sprouting and CS-A-enriched/CS-C-enriched PNNs accumulation by three-dimensional image acquisition and reconstruction in sham-operated, SCI and LDI groups. Three-dimensional reconstructed images showed that prominent CS-A-enriched PNNs and CS-C-enriched PNNs surrounded axons with decreased 5-HT positive axons on the ipsilateral side following SCI when compared to that of the sham-operated group. LDI treatment reversed 5-HT-positive axonal spouting and decreased CS-A-enriched PNNs accumulation, but no inhibitory effect on CS-C-enriched PNNs.
We first measured the percentage of 5-HT axons/tissue and CS-A-enriched PNNs/tissue. At 14-days post-SCI, in the sham-operated group, the percentage of 5-HT axons/tissue was 29.6 ± 0.67%, the percentage of CS-A-enriched PNNs/tissue was 6.3 ± 1.12%; in the SCI group, 5-HT axons were repelled in a similar manner between CS-A-enriched PNNs proximal to the lesion, and the percentage of 5-HT axons/tissue was reduced to 5.1 ± 0.79%, the percentage of CS-A-enriched PNNs/tissue was 29.3 ± 3.32%; following LDI treatment, the percentage of 5-HT axons/tissue was increased to 25.2 ± 1.08%, the percentage of CS-A-enriched PNNs/tissue was 8.85 ± 0.96% (Fig. 3). At 60-days post-SCI, decreased 5-HT axon also accompanied with increased CS-A-enriched PNNs. In sham-operated group, the percentage of 5-HT axons/tissue was 28.3 ± 1.05%, the percentage of CS-A-enriched PNNs/tissue was 6.7 ± 1.07%; in SCI group, the percentage of 5-HT axons/tissue was 6.5 ± 0.82%, the percentage of CS-A-enriched PNNs/tissue was 35.3 ± 1.89%; in LDI group, the percentage of 5-HT axons/tissue was 25.5 ± 1.06%, the percentage of CS-A-enriched PNNs/tissue was 9.27 ± 1.03%.
The percentage of 5-HT axons/tissue and CS-C-enriched PNNs/tissue were also measured. At 14-days post-SCI, in the sham-operated group, the percentage of 5-HT axons/tissue was 27.3 ± 0.85%, the percentage of CS-C-enriched PNNs/tissue was 6.2 ± 0.97%; in the SCI group, the percentage of 5-HT axons/tissue was reduced to 7.5 ± 0.39%, the percentage of CS-C-enriched PNNs/tissue was 28.1 ± 1.15%; following LDI treatment, the percentage of 5-HT axons/tissue was increased to 10.3 ± 1.22%, the percentage of CS-C-enriched PNNs/tissue was 25.2 ± 0.96%. At 60-days post-SCI, in sham-operated group, the percentage of 5-HT axons/tissue was 23.8 ± 1.13%, the percentage of CS-C-enriched PNNs/tissue was 6.6 ± 1.10%; in SCI group, the percentage of 5-HT axons/tissue was 7.3 ± 1.01%, the percentage of CS-C-enriched PNNs/tissue was 36.3 ± 1.77%; in LDI group, the percentage of 5-HT axons/tissue was 11.2 ± 1.63%, the percentage of CS-C-enriched PNNs/tissue was 32.2 ± 0.89%.
We further analyzed the correlation between of 5-HT axonal immunoreactivity and CS-A/CS-C-enriched PNNs immunoreactivity through Pearson correlation following LDI treatment. There was a significant correlation between a reduction of CS-A-enriched PNNs and an increase of 5-HT axonal immunoreactivity (r2 = 0.810 at 14-days post-SCI, r2 = 0.706 at 60-days post-SCI, P < 0.05), but no significant correlation between change of CS-C-enriched PNNs and an increase of 5-HT axonal immunoreactivity (r2 = 0.050 at 14-days post-SCI, r2 = 0.249 at 60-days post-SCI, P > 0.05).
LDI increases 5-HT receptor expression but does not affect Nogo receptors, RPTPσ or LAR expression following SCI.
In order to identify the effect of LDI treatment on functional receptors of CSPGs and 5-HT receptors, we compared the expression levels of 5-HT receptors, Nogo receptors, RPTPσ and LAR in neurons through double-immunofluorescent staining in the sham, SCI and LDI groups.
As shown in Fig. 4, the sham-operated group showed a very weak expression of 5-HT2c/a and 5-HT7, and these expression levels were similar in the SCI group. In contrast, the LDI group showed a significantly increased expression of 5-HT2A, 5-HT2C and 5-HT7 in neurons (Fig. 4). The percentage of 5-HT2c/a-positive neurons was 12.5 ± 1.08% in the sham-operated group, 18.2 ± 2.16% in the SCI group, and was increased to 26.7 ± 1.67% in the LDI group (F (2, 12) = 8.65, p = 0.005). Similarly, the percentage of 5-HT7-positive neurons was 11.7 ± 0.88% in the sham-operated group, 18.6 ± 1.03% in the SCI group, and was increased to 27.3 ± 1.25% in the LDI group (F (2, 12) = 8.82, p = 0.004).
We next investigated to what extent LDI could be involved in the regulation of the Nogo receptors, NgR1 and NgR3. The sham-operated group only showed minimal expression of NgR1 and NgR3 in neurons. In comparison to the sham-operated group, the SCI group showed an increased expression of NgR1 and NgR3 in neurons. There was no significant difference in the expression of NgR1 or NgR3 in neurons between the SCI group and LDI group (Fig. 5A,C). The percentage of NgR1-positive neurons was 8.3 ± 0.58% in the sham-operated group, 9.3 ± 0.86% in the SCI group, and 9.8 ± 0.67% in the LDI group (F (2, 12) = 1.58, p = 0.250). The percentage of NgR3-positive neurons was 8.5 ± 0.63% in the sham-operated group, 9.6 ± 0.59% in the SCI group, and 10.1 ± 0.82% in the LDI group (F (2, 12) = 1.32, p = 0.303).
We also measured the expression of RPTP-σ and LAR in neurons. Similar to the expression of NgR1 and NgR3, there was no significant difference in the expression of RPTP-σ or LAR in neurons between the SCI group and LDI group (Fig. 5B, D). The percentage of RPTP-σ-positive neurons was 6.8 ± 0.56% in the sham-operated group and increased to 15.8 ± 1.01% in the SCI group and 16.2 ± 1.67% in the LDI group (F (2, 12) = 1.01, p = 0.393). The percentage of LAR-positive neurons was 5.9 ± 0.63% in the sham-operated group and increased to 22.9 ± 1.58% in the SCI group and 23.7 ± 1.33% in the LDI group (F (2, 12) = 0.83, p = 0.459).
LDI promotes motor function recovery following SCI.
The modified-Tarlov grading system was used to test hindlimb motor-function. The motor function of the hindlimbs was intact before SCI. After SCI, the injured side lower limbs were paralysis (Tarlov Grade 1) and incontinent, which suggested a severe and consistent lesion. After injury, serious hindlimb paralysis with no residual motor function was observed in all dogs, and hindlimb scores neared zero. At 7 days post-injury, some animals regained a slight improvement for weight support due to spontaneous recovery. Compared to SCI-treated animals, enhanced hindlimb motor-function recovery occurred in the LDI-treated animals. At 14 days post-SCI, the modified Tarlov scale in the sham-operated group was 5(0); this level decreased to 3(0) in the SCI group and was ameliorated to 3(1) with LDI treatment (F (2, 12) = 10.67, p = 0.002); At 60 days post-SCI, the modified Tarlov scale in the sham-operated group was 5(0); this level decreased to 3(1) in the SCI group and was ameliorated to 4(0) with LDI treatment (F (2, 12) = 10.25, p = 0.002) (Fig. 6). These data suggest that LDI treatment improved hindlimb motor-function recovery after SCI.