SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflicts of interest
  9. References

The activation of complement system can aggravate the secondary injury after spinal cord injury (SCI). Our previous study indicates that the interception of complement activation by C3 deficiency can reduce the secondary injury and improve the regeneration and functional recovery after SCI. However, recently, it was reported that C5a which was generated during the complement activation pathways also had a protective effect on neurons, but whether it has the similar effect after SCI is unknown. To investigate the possibility and mechanism of the protective effect of C5a on neurons, it is necessary to study the expression profiles of C5a and its receptor CD88 after SCI and the influence on their expression when C3 was knocked out. By immunohistochemistry and Western blot, we found that in wild-type (WT) mice, both the expression of C5a and its receptor CD88 increased significantly, and there were two peaks during their expression after SCI. However, in C3-deficient mice, the expression of C5a still increased after SCI, although it was lower than that in WT group at every time points after SCI, and the expression of CD88 remained stable. Our study suggests that the expressions of C5a and CD88 can be inhibited in different degrees after SCI when the activation of complement system is blocked through C3 deficiency, which can reduce the secondary injury caused by C5a after SCI on one hand but deprive neurons of the possible protective effect from C5a on the other hand.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflicts of interest
  9. References

C3 is the pivotal factor in the three complement activation pathways (classical, alternative and lectin) and is necessary for complement activation. Our previous study indicates that complement inhibition caused by C3 deficiency can reduce secondary injury and improve neural regeneration and functional recovery after spinal cord injury (SCI) [1]. However, it was reported recently that the activation of complement could possess a dual character – both secondary injury [2, 3] and neuroprotective effect [4-6], during which C5a was the most important factor. On one hand, among all kinds of complement factors, C5a is considered to be the most effective inflammatory factor which can induce the chemotactic response of neutrophilic granulocytes, the secretion of cytokines such as IL-2, IL-6, IL-8 and TNF-α [7] and then mediate the early phase of inflammation after SCI, which is harmful to cells at the injury site [8]. On the other hand, C5a can bind to its receptor CD88 and exert its direct neuroprotective effect by inhibiting caspase 3 and then the neuron apoptosis [4, 5, 9] or exert its indirect neuroprotective effect by promoting the phagocytosis of microglia and reducing the secondary injury [10, 11]. But there is no direct evidence for this dual character of C5a after SCI. Although the secondary injury caused by complement activation was taken into account when we promoted regeneration after SCI by C3 deficiency [1], the neuroprotective effect after complement activation was neglected. So the dual effect of complement activation, especially that of C5a after SCI, should be balanced to promote the neuronal regeneration efficiently.

Presently, the study on the relationship between C5a and SCI is mainly focused on the inflammatory effect and secondary injury caused by C5a. Whether C5a has a neuroprotective effect after SCI is unknown. Study on the expression profiles of C5a and its receptor CD88 after SCI and the influence on their expression when C3 was deficient will be helpful to our understanding about the possibility and mechanism of the protective effect of C5a on neurons after SCI.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflicts of interest
  9. References

Female wild-type (WT) and C3-deficient C57/BL6 mice (purchased from Jackson Laboratories, Bar Harbor, ME USA, weighing 20–25 g, 6–8 weeks old) were used in this study. Vaginal smear examination was performed, and animals in diestrus were used to reduce the influence of the oestrogen level to the minimum. Animals were randomized into sham (laminectomy, no SCI damage) or SCI groups, respectively. A weight-drop contusion injury (5 g, 6 cm) was performed using the procedure described previously [12] after a laminectomy at T12. Manual bladder expression was performed twice a day after surgery. All surgical operations were performed in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals.

Animals were perfused with 4% paraformaldehyde at 1 h, 6 h and 24 h after injury. Tissue segments containing the lesion area (3 mm on each side of the lesion) were sectioned in the longitudinal plane (20 μm) on a cryostat. Immunohistochemistry was performed according to the protocol previously described [13]. The following primary antibodies were used: rat-anti-C5a antibody (1:200, R&D) and rat-anti-CD88 antibody (1:1000, Abcam).

Proteins were resolved by SDS-PAGE on 12% gels, transferred to nitrocellulose membranes, and blocked overnight in 5% nonfat milk. Membranes were probed with rat-anti-C5a antibody (1:1000, R&D) or rat-anti-CD88 antibody (1:1000, Abcam), followed by biotinylated horse anti-rat antibody and then horseradish peroxidase streptavidin. Proteins were visualized by reaction with 3,3-diaminobenzidine-tetrahydrochloride. Blots were also probed for β-actin (1:1000, Abcam) as a control. The ratio of C5a or CD88 to β-actin product was obtained by analysing the integrated optical density (IOD) of the corresponding bands using Quantity One® (BIORAD).

Statistical analysis was performed by analysis of variance with repeated measures using Scheff's test for Post hoc comparisons. A P-value <0.05 denoted a statistically significant difference. As to immunohistochemistry, five sections of each mouse from each group were analysed by Image Pro. 5.0, and the average optical densities (AOD) of positive staining neurons were measured.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflicts of interest
  9. References

C5a

Immunohistochemistry

The immunohistochemical-positive products were mainly distributed in cytoplasm of neurons (Fig. 1 A–J, A'–J', K). In WT mice, the C5a immunoreactivity was weak in sham group (Fig. 1 A, A'), but it increased significantly (P < 0.01 versus WT sham) 1 h after injury (Fig. 1 A, A') until it reached a plateau at 6 h fter injury (Fig. 1 C, C'). However, it increased significantly again at 12 h after injury (P < 0.05 versus WT 6 h) (Fig. 1 D, D') until 24 h (Fig. 1 E, E') when it reached a second plateau. In other words, there were two expression peaks of C5a after injury (1 h and 12 h) in WT group. In C3-deficient mice, the C5a immunoreactivity in sham group (Fig. 1 F, F') was as weak as that in WT sham group, and it did not change significantly at 1 h after injury (Fig. 1 G, G'). However, it increased significantly at 6 h (P < 0.05 versus C3 1 h) (Fig. 1 H, H') and reached a plateau after then until 24 h after injury (Fig. 1 I–J, I'–J'). The C5a expression in C3-deficient mice was significantly lower than that in WT mice at each time point after injury, and the only peak time was advanced.

image

Figure 1. C5a immunohistochemistry in WT and C3-deficient mice after SCI. (A–J, A'–J', K) Longitudinal sections showing the C5a immunohistochemical reaction in ventral horn of sham group (A, A', F, F') and 1 h (B, B', G, G'), 6 h (C, C', H, H'), 12 h (D, D', I, I'), 24 h (E, E', J, J') after SCI in WT(A–E, A'–E') and C3-deficient mice (F–J, F'–J'). (K) Average optical density of C5a in WT and C3-deficient mice after injury. Scale bar: (A–E,F–J) 100 μm, (A'–E', F'–J') 20 μm. *: P < 0.01 versus WT sham; **: P < 0.05 versus WT 6 h; △: P < 0.05 versus C3 1 h. n = 4 per group.

Download figure to PowerPoint

CD88

Immunohistochemistry

The immunohistochemical-positive products were mainly localized in cytoplasm of neurons (Fig. 2. A–J, A'–J', K). In WT mice, the CD88 immunoreactivity was weak in sham group (Fig. 2 A, A'), but it increased significantly (P < 0.01 versus WT sham) 1 h after injury (Fig. 2 B, B'), then it decreased to sham level (Fig. 2 C–D, C'–D'). At 24 h after injury it increased significantly again (P < 0.01 versus WT 12 h) (Fig. 2 E, E'). In C3-deficient mice, the CD88 immunoreactivity in sham group (Fig. 2 F, F') was as weak as that in WT sham group, and it did not change obviously at each time point after injury (Fig. 2 G–J, G'–J'). In addition, the expression of CD88 in C3-deficient mice was significantly lower than in WT mice at 1 h and 24 h after injury (P < 0.01).

image

Figure 2. CD88 immunohistochemistry in WT and C3-deficient mice after SCI. (A–J, A'–J', K) Longitudinal sections showing the CD88 immunohistochemical reaction in ventral horn of sham group (A, A', F, F') and 1 h (B, B', G, G'), 6 h (C, C', H, H'), 12 h (D, D', I, I'), 24 h (E, E', J, J') after SCI in WT(A–E, A'–E') and C3-deficient mice (F–J, F'–J'). (K) Average optical density of CD88 in WT and C3-deficient mice after injury. Scale bar: (A–E, F–J) 100 μm, (A'–E', F'–J') 20 μm. *: P < 0.01 versus WT sham; **: P < 0.01 versus WT 12 h; △: P < 0.01 versus the same time point in WT group. n = 4 per group.

Download figure to PowerPoint

Western blot

Results from Western blot were similar to those from immunohistochemistry. In addition, the CD88 expression in C3-deficient mice was significantly lower than in WT mice at 1 h, 12 h and 24 h after injury (P < 0.05) (Fig. 3 A–C).

image

Figure 3. Expression of CD88 in WT and C3-deficient mice. (A–C) CD88 expression was measured by Western blot in WT(A) and C3-deficient mice (B) aftrer SCI. Results were expressed as the relative content of CD88 at different time points after SCI(C). *: P < 0.01 versus WT sham; **: P < 0.01 versus WT 12 h; △: P < 0.05 versus the same time point in WT group. n = 4 per group.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflicts of interest
  9. References

Our previous findings revealed that C3 deficiency could inhibit complement activation and inflammation, reduce the secondary injury and improve the nervous regeneration and functional recovery after SCI [1]. There are many by-products produced during complement activation, including C5a. Known as an anaphylatoxin and chemotactic factor, C5a plays a substantial role in the initiation and regulation of inflammation. Overexpression of C5a can promote inflammatory response and aggravate the secondary injury after SCI [14, 15]. However, it was recently reported that C5a which was produced during the complement activation pathways also had a protective effect on neurons [4-6, 9-11], and whether it has the similar effect after SCI is unknown. If it does, there will be a defect in this potential therapy for SCI through complement inhibition by C3 deficiency [1]. When the expression of C5a is inhibited through C3 deficiency, both the secondary injury and the possible neuroprotective effect of C5a after SCI are deprived. Presently, studies on the relationship between C5a and SCI are mainly focused on the inflammation and secondary injury caused by C5a. Whether C5a has a protective effect on neurons after SCI is unknown. To investigate the possibility and mechanism of this question, it is necessary to study the expression profiles of C5a and its receptor CD88 after SCI and the influence on their expression when C3 was knocked out.

Our results showed that the expression of C5a in WT mice increased significantly and peaked at 1 h and 12 h, respectively, after SCI. While its expression was inhibited obviously in C3-deficient mice after SCI due to the interception of complement activation pathways caused by C3 deficiency. However, there was still an expression peak at 6 h after SCI in C3-deficient mice. These may be attributed to the three conventional complement activation pathways (classical, alternative and lectin) and the fourth coagulation pathway [16]. Each of the three conventional complement activation pathways can lead to the production of C5a. This may account for the expression peak at 1 h in WT mice. In addition, it was recently reported that the coagulation pathway could generate C5a in the absence of C3 [16]. This fourth substituted pathway might account for the expression peak of C5a at 6 h in C3-deficient mice and the second expression peak of C5a at 12 h in WT mice. However, the peak time points probably caused by this fourth pathway were different between WT and C3-deficient mice. This difference may result from two possibilities. Firstly, the fourth complement activation pathway may be inhibited and delayed normally by the conventional three activation pathways because of an as-yet unknown mechanism. So the C5a expression peak time point caused by the fourth activation pathway in WT mice was later than that in C3-deficient mice. Secondly, due to the sensitivity of immunohistochemistry, the possible C5a expression peak at 6 h in WT mice cannot be detected, even if the fourth activation pathway was not inhibited under normal condition. However, it is worthwhile to note that the coagulation pathway can generate C5adesArg as well as C5a. Thrombin (the coagulation pathway) can activate thrombin activatable fibrinolysis inhibitor (TAFI), a carboxypetidase, which can cleave C5a to generate C5adesArg. Although it has ~10–100-fold reduced affinity for CD88 than intact C5a, C5adesArg still reserve signalling activity for CD88 and most other pro-inflammatory activities [17]. As the antibody used here cannot distinguish the two forms of the C5a, the results demonstrated by immunohistochemistry that C5a is elevated, does not necessarily mean that the intact C5a is increased. However, even a proportion of the increased C5a might be composed of C5adesArg, as an important pro-inflammatory factor, it still plays a vital role during the secondary injury after SCI.

C5a exerts its role through binding to its receptor after the injury to the central nervous system (CNS), and CD88 is considered to be the main functional receptor of C5a [18, 19]. The relative levels of CD88 expression in the CNS may be influenced by the inflammatory state [20]. Many products generated during the complement activation are mediators of inflammation [21]. On one hand, excessive complement activation can lead to harmful inflammation, tissue damage and elevation of many cytokines such as TNF-α. Blockade of CD88 using PMX-53 can partially prevents the cytokine production [21, 22]. On the other hand, TNF-α can upregulate CD88 expression in the brain [23]. Therefore, a positive feedback between CD88 and inflammatory factors such as TNF-α is formed and the CD88 expression increases in diseased states generally [20]. Our previous work indicated that the expression of TNF-α increased significantly after SCI in WT mice [1], which may induce the elevation of CD88. Due to the positive feedback between CD88 and TNF-α mentioned above, the expression of CD88 may increased continuously and form the second expression peak. As for C3-deficient mice, the expression of TNF-α was inhibited and maintained at a low level [1], so the expression of CD88 did not change obviously at each time point after injury.

In conclusion, C5a can result in the secondary injury after SCI by binding to its receptor CD88. When C3 is knocked out, expressions of both C5a and CD88 are inhibited in different degrees. So C3 deficiency can reduce secondary injury and improve neural regeneration [1]. However, as C5a can inhibit caspase 3 and then the neuron apoptosis by binding to CD88 [4, 5, 9], the above-mentioned C3 deficiency treatment also has its negative effect on regeneration and can only partially improve the regeneration. Further study is needed to balance the favourable and unfavourable aspects of C5a after SCI.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflicts of interest
  9. References

This work was supported by the National Natural Science Foundation of China (No. 30900572).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflicts of interest
  9. References