Cyclooxygenase-2 expression in oligodendrocytes increases sensitivity to excitotoxic death

Abstract

Background: We previously found that cyclooxygenase 2 (COX-2) was expressed in dying oligodendrocytes at the onset of demyelination in the Theiler's murine encephalomyelitis virus-induced demyelinating disease (TMEV-IDD) model of multiple sclerosis (MS) (Carlson et al. J.Neuroimmunology 2006, 149:40). This suggests that COX-2 may contribute to death of oligodendrocytes.

Objective: The goal of this study was to examine whether COX-2 contributes to excitotoxic death of oligodendrocytes and potentially contributes to demyelination.

Methods: The potential link between COX-2 and oligodendrocyte death was approached using histopathology of MS lesions to examine whether COX-2 was expressed in dying oligodendrocytes. COX-2 inhibitors were examined for their ability to limit demyelination in the TMEV-IDD model of MS and to limit excitotoxic death of oligodendrocytes in vitro. Genetic manipulation of COX-2 expression was used to determine whether COX-2 contributes to excitotoxic death of oligodendrocytes. A transgenic mouse line was generated that overexpressed COX-2 in oligodendrocytes. Oligodendrocyte cultures derived from these transgenic mice were used to examine whether increased expression of COX-2 enhanced the vulnerability of oligodendrocytes to excitotoxic death. Oligodendrocytes derived from COX-2 knockout mice were evaluated to determine if decreased COX-2 expression promotes a greater resistance to excitotoxic death.

Results: COX-2 was expressed in dying oligodendrocytes in MS lesions. COX-2 inhibitors limited demyelination in the TMEV-IDD model of MS and protected oligodendrocytes against excitotoxic death in vitro. COX-2 expression was increased in wild-type oligodendrocytes following treatment with Kainic acid (KA). Overexpression of COX-2 in oligodendrocytes increased the sensitivity of oligodendrocytes to KA-induced excitotoxic death eight-fold compared to wild-type. Conversely, oligodendrocytes prepared from COX-2 knockout mice showed a significant decrease in sensitivity to KA induced death.

Conclusions: COX-2 expression was associated with dying oligodendrocytes in MS lesions and appeared to increase excitotoxic death of oligodendrocytes in culture. An understanding of how COX-2 expression influences oligodendrocyte death leading to demyelination may have important ramifications for future treatments for MS.

Figure 1

Expression of COX-2 in dying oligodendrocytes in an MS lesion. An MS spinal cord lesion was examined by confocal immunofluorescence after probing with antibodies to the oligodendrocyte marker CNPase (red), COX-2 (green) and activated caspase 3 (blue). (A) A low power image (10×) of the plaque shows the area of demyelination (bar = 80 μm). Two regions (see arrows 1 and 2) indicate the areas shown in higher magnification (40×) for panels B-F (bar in B = 20 μm). (B). The three color image for the region shown by arrow 1 in panel A. Note the co-expression of COX-2, activated caspase 3 and CNPase as indicated by white. (C-E) Individual channels are shown COX-2 (C), activated caspase 3 (D) and CNPase (E). The three color merged image is shown in panel F. Two examples of labeling of all three antigens (appearing white) are indicated with arrows.

Figure 2

Therapeutic effect of the COX-2 inhibitor CAY 10452 in TMEV-IDD. The COX-2 inhibitor CAY10452 decreases demyelination. Spinal cord sections from control and CAY10452 (35 days after infection with TMEV) were stained with H&E and LFB and scored for inflammation and demyelination as described in methods. Comparisons between control and CAY10452 groups were done using the Mann-Whitney nonparametric t-test and the p values indicated above the error bars (SEM).

Figure 3

Excitotoxicity in the spinal cord explant cultures. A) A spinal cord culture was stained for expression of the neuronal marker MAP-2 (Blue) and for the oligodendrocyte marker oligodendrocyte-specific protein (OSP) (Red). One hemisphere is boxed to show the regions from other slices which appear in panels B and C (bar = 400 μm). Hemispheres of a slice cultures stained for MAP-2,(Blue), OSP (Red) and activated caspase 3 (green) are shown for untreated (B) control and kainic acid (KA) (C). (Magnifications are 2× for A and 4× for B and C, bar = 80 μm for B, C and = 20 for D μm). (D). A higher magnification (60×) of the white matter region showing oligodendrocytes stained for CNPase (red) and activated caspase 3 (green) and co-labeling with both as yellow. Examples of oligodendrocytes containing activated caspase 3 are shown (see white arrows). An example of a cell labeled for activated caspase 3 which is not an oligodendrocyte (and likely a motor neuron) is shown with a gray arrow.

Figure 4

COX-2 inhibitor mediated decrease of KA-induced activated caspase 3 in white matter and gray matter. The appearance of cells stained with the marker for cell death (activated caspase-3) was assessed in four spinal cord sections for white matter (left) and gray matter (right) with treatments of vehicle (control), kainic acid (KA) or kainic acid with the COX-2 inhibitor NS398 (KA/NS398). Error bars are SEM and P values determined by ANOVA Tukey-Kramer. This is a representative experiment which has been repeated in two other experiments.

Figure 5

COX-2 is induced in oligodendrocytes by Kainic Acid (KA). Dispersed oligodendrocyte cultures were treated with either vehicle (Control) or KA and examined 24 hours later by confocal immunofluorescence. COX-2 expression (green) is seen in the cells labeled with the oligodendrocyte specific marker Olig-1 (red), co-labeling appears yellow (see arrow). Magnification bar = 20 μm.

Figure 6

Protection of oligodendrocytes with the COX-2 inhibitor (CAY10404). Dispersed oligodendrocyte cultures were treated with KA in the presence or absence of CAY 10404 (10 uM) and analyzed 24 hours later for cell death. Error bars are SEM.

Figure 7

Cultures of Oligodendrocytes derived from COX-2 transgenic mice over-express COX-2. Dispersed wild-type cultures were prepared from wild-type and transgenic mice the over-express COX-2 specifically in oligodendrocytes (generated with the CNPase promoter fused to the human COX-2 gene. The oligodendrocyte marker to Olig-1 appears red and COX-2 appears green. Co-expression of both appears yellow. Magnification bar = 40 μm.

Figure 8

Vulnerability of wild-type and Oligo-COX-2 transgenic oligodendrocytes to KA-induced excitotoxicity. Dispersed oligodendrocytes were treated to varying concentrations of KA and examined for cell death 24 hours later. Oligodendrocytes derived from the transgenic animals were 8-fold more sensitive to KA than wild-type. Note, the x-axis is not a linear scale.

Figure 9

Sensitivity of oligodendrocytes from COX-2 knockout mice to excitotoxic death and protection with COX-2 inhibitors. (A) Oligodendrocytes from COX-2 knockout mice are more resistant to KA-induced excitotoxicity. Dispersed oligodendrocyte cultures were prepared from wild-type (WT), heterozygous COX-2 knockout (COX-2 +/-) and homozygous COX-2 knockout (COX-2 -/-) and treated with KA. Surviving cells were scored 24 hours after KA treatment. This is the average of two independent experiments. (B) The COX-2 inhibitor CAY 10404 does not protect COX-2 knockout oligodendrocytes from KA-induced death. Dispersed oligodendrocyte cultures were prepared from COX-2 -/- mice and treated with KA in the presence or absence of CAY 10404 (10 uM). Viability was assessed 24 hours after treatment with KA. There was no significant increase in the number of surviving oligodendrocytes in the CAY 10404-treated group. Error bars are SEM.