Serum sCD95L concentration in patients with spinal cord injury

Abstract

Objective

To determine serum concentrations of soluble CD95 ligand (sCD95L) in patients with traumatic spinal cord injury.

Methods

Patients with traumatic spinal cord injury were recruited. Blood was collected on admission to hospital and at 4 h, 9 h, 12 h, 24 h, 3 days, 7 days, and 2, 4, 8 and 12 weeks postadmission. Serum concentrations of sCD95L were determined via immunoassay.

Result

The study included 23 patients. Mean sCD95L concentrations were significantly lower at 4 h, 9 h, 12 h and 24 h than at admission, and were significantly higher at 8 and 12 weeks, compared with admission.

Conclusion

The serum sCD95L concentration fell significantly during the first 24 h after traumatic spinal cord injury. Concentrations then rose, becoming significantly higher than admission levels at 8 weeks. sCD95L may represent a possible therapeutic target for traumatic spinal cord injury.

Introduction

Spinal cord injury has an enormous impact on health-related quality-of-life due to its substantial medical, psychological and economic consequences for patients and their families. ¹ Despite advances in its treatment, there is no therapy capable of effectively reversing neurological damage. ²

The acute phase of spinal cord injury lasts up to 2 days and is characterized by inflammation, neuronal and axonal changes, haemorrhage, demyelination and apoptosis.^3,4 Apoptosis plays an important role in the progressive destruction of grey and white matter^5–8 characterized by ischaemia, inflammation, glutamate excitotoxicity, oxidative stress and the production of free radicals. ³ This cascade leads to the destruction of neurons by both necrotic and apoptotic cell death, increasing the damage that was initially produced by traumatic injury. ⁹

The type II transmembrane protein CD95L can be proteolytically cleaved into a membrane-bound (mCD95L) and a soluble (sCD95L) form. ¹⁰ sCD95L is activated and secreted by cleavage of mCD95L by the serine matrix metalloproteinases (MMP)-7 and MMP-3. ¹¹ Binding of CD95L to CD95 (Fas) leads to trimerization of the intracellular death domain and binding of FADD (Fas-associated protein with death domain). This leads to caspase-8 activation and formation of DISC (death-inducing signalling complex). Caspase-8 autoproteolysis results in activation of effector caspases (caspase-3 and -7), inducing a mitochondrial apoptosis signal and death signal amplification.^12–14 Binding of CD95L to CD95 also leads to production of inflammatory cytokines. ¹⁵ CD95 also has effects on cellular activation, differentiation and proliferation, ¹⁴ and is involved in the branching of developing neurons in the CNS, axonal sprouting of dorsal root ganglion neurons, migration of malignant glioma cells, and differentiation of neuronal stem cells. ¹⁶

Blocking the CD95 pathway (via subarachnoid infusion of sFasR, ¹⁷ systemic antibody administration ¹⁸ or genome deletion ¹⁹ ) has been found to improve functional outcome and reduce apoptotic cell death following spinal cord injury in rodents. ¹⁷ The findings of these studies may not be applicable to humans, although a similar pattern of inflammatory cytokine production was seen in rodent spinal cords and human cerebrospinal fluid (CSF) following spinal cord injury. ²⁰ Our pilot study of patients with spinal cord injury revealed an initial increase followed by a decrease in blood sCD95L concentrations, but the sample size was insufficient to show statistical significance. ²¹ The aim of the current study, therefore, was to evaluate sCD95L levels in a larger cohort of patients with spinal cord injury, to determine whether sCD95L could be used as a therapeutic target.

Patients and methods

The study included all patients with acute spinal cord injury who were treated at the BG Trauma Centre Ludwigshafen, Department of Paraplegiology, Ludwigshafen, Germany, between November 2010 and January 2013. Exclusion criteria were: nontraumatic spinal injury; brain injury; infectious disease; rheumatological conditions; endocrinological disease; cancer; autoimmune conditions; coma. Injuries were classified using the AO classification system. ²²

The study was approved by the Ethics Commission of the Landesärztekammer Rheinland-Pfalz, Germany (No. 837.188.12 (8289-F)), and all participants provided written informed consent prior to inclusion in the study.

Blood samples

Venous blood (7.5 ml) was taken using standard methods on admission to hospital, and at 4 h, 9 h, 12 h, 24 h, 3 days, 7 days, and 2, 4, 8, and 12 weeks after injury. Blood was allowed to coagulate for 20 min, then centrifuged at 2500  g for 10 min at room temperature, and the resulting serum was stored at −80℃ until analysis.

sCD95L quantification

Serum sCD95L was quantified using the Quantikine® Human Fas Ligands/TNFSF6 Immunoassay (R&D Systems, Minneapolis, MN, USA), according to the manufacturer’s instructions. Analysis was performed in duplicate in a double-blind manner. Normal reference values for sCD95L are 39.8–145 pg/ml.

Neurological examination

Neurological deficit was classified according to the International Standards for Neurological Classification of SCI. ²³ Neurological examinations were conducted in awake and responsive patients at admission and discharge (after 12 weeks).

Statistical analyses

Continuous data were expressed as mean ± SD or SEM; categorical data were expressed as absolute and relative rates. Independent variables were analysed using Mann–Whitney U-test; dependent variables were compared with Wilcoxon signed-rank test. A test for normal distribution between groups was not used because of the small patient pool. Between-group comparisons of categorical data were made with χ²-test. Statistical analyses were performed using SPSS® version 19.0 (SPSS Inc., Chicago, IL, USA) for Windows®; graphs were created using SigmaPlot® Software (Systat Software, San Jose, CA, USA). P-values <0.05 were considered statistically significant.

Results

The study included 23 patients (16[69.6%] male/ seven [30.4%] female; mean age 42.9±23.11 years; age range 18–85 years). Seven patients showed improvements in neurological deficit between arrival and discharge, 16 had no neurological improvement and no patient had a worsening in condition. Full clinical and demographic details of the patient cohort are given in Table 1.

Table 1.

Demographic and clinical characteristics of patients with traumatic spinal cord injury included in a study investigating postinjury serum concentrations of soluble CD95 ligand (n = 23).

Characteristic	n
Male/female	16/7 (69.6/30.4)
Age, years	42.9±23.1
Cause of injury
Fall	13 (56.5)
Traffic accident	9 (39.1)
Other	1 (4.4)
AO injury classification 22
A	15 (65.2)
B	6 (26.1)
C	2 (8.7)
Paraplegia	15 (65.2)
Partial paralysis	8 (34.7)
Complete paralysis	7 (30.4)
Tetraplegia	8 (34.7)
Partial paralysis	7 (30.4)
Complete paralysis	1 (4.3)
Injury location
Cervical spine	8 (34.8)
Thoracic spine	9 (39.1)
Lumbar spine	6 (26.1)

Data presented as n (%) or mean±SD.

Data regarding mean serum sCD95L concentrations at all timepoints are given in Table 2. Mean sCD95L concentrations were significantly lower at 4 h, 9 h, 12 h, and 24 h than at admission (P<0.05 for each comparison; Table 2). sCD95L concentrations were significantly higher at 8 and 12 weeks compared with admission (P < 0.05 for each comparison; Table 2).

Table 2.

Serum concentrations of soluble CD95 ligand (sCD95L) in patients with traumatic spinal cord injury (n = 23).

Timepoint	sCD95L concentration pg/ml
Admission	70.66 ± 27.33
4 h	63.73 ± 21.82*
9 h	64.08 ± 24.15*
12 h	59.83 ± 24.53*
24 h	57.16 ± 25.03*
72 h	60.52 ± 24.70
1 week	64.45 ± 40.46
2 weeks	70.56 ± 35.95
4 weeks	69.49 ± 30.26
8 weeks	77.89 ± 31.06*
12 weeks	76.92 ± 24.98*

Data presented as mean ± SD.

*

P<0.05 vs admission; Wilcoxon signed-rank test.

There were no significant differences in sCD95L concentration at any timepoint between patients with grade A injuries (n = 13) and those with grade B/C (n = 10) (data not shown).

Discussion

The present study found that serum concentrations of sCD95L significantly decreased in the first 24h after hospital admission, compared with initial levels. sCD95L concentrations then gradually increased, becoming significantly higher than admission at 8 weeks. These findings are in accordance with those of our pilot study ²¹ and the work of others. ²⁴

It is well documented that blocking the sCD95L signalling pathway leads to an improvement in neurological deficit after spinal cord injury.^{17–19,25,26} In rats, CD95 concentrations in white brain matter increase in the first 8 h following injury. ²⁷ There was a significant fall in sCD95L concentration immediately after injury in humans in the present study, followed by an increase after 24 h. It is possible that blocking the CD95 signalling pathway in its earliest phase might be therapeutically effective. More precise temporal information concerning sCD95L concentration is needed in order to determine the timepoint when a goal-oriented therapy would be most effective in paitents with acute spinal cord injury. Studies concerning cytokine expressionhave shown that rodents and humans release the same inflammatory cytokines after traumatic injury to the spinal cord, though they were detectable later in human CSF than in mice. ²⁰ This may explain the differences observed between human and animal studies relating to the timing of changes in CD95L concentration.

A postmortem human study found that the CD95/CD95L system was involved in both apoptosis and inflammatory response in the acute and subacute phases after spinal cord injury, ²⁸ and that both proapoptotic and proinflammatory activity were functionally important. The role of autoimmunity is unclear, with some studies indicating a neurotoxic effect^29–31 and others a neuroprotective one. ³² It remains to be elucidated whether there are different roles for the soluble and membrane-bound forms of CD95L. T-cells from mice lacking CD95 were unable to kill targeted cells, suggesting that mCD95L is responsible for cytotoxic activity and initiating apoptosis, rather than for immunological activity. ³³ Mice injected with sCD95L developed an autoimmune response, ³⁴ but others have shown that mCD95L confers autoimmune activity. ¹⁰ Some studies have shown that recombinant sCD95L, in contrast to human sCD95L, has no effect on apoptosis.^10,35 In this context, the relationship between neurological outcome and sCD95L concentration in a larger cohort population would be of interest. Our study showed no significant difference in serum sCD95L concentration between patients with, or without, an improved neurological outcome. This is most likely due to the small number of patients after stratification according to American Spinal Injury Association Impairment Scale grades. ³⁶ The current study has several limitations, most notably the absence of a control group comprising patients with vertebral bone fracture and without neurological deficit. This is due to the considerably shorter hospitalization of such control patients. A further limitation is the short duration of follow-up; a longer follow-up would be interesting but is difficult to manage. Finally, it would be of interest to compare sCD95L and mCD95L concentrations. Furthermore, properly controlled, larger-scale studies are required to support our findings.

In conclusion, the serum sCD95L concentration falls significantly during the first 24 h after traumatic spinal cord injury. sCD95L concentrations then rise, becoming significantly higher than admission levels at 8 weeks. sCD95L may represent a possible therapeutic target for traumatic spinal cord injury.