Vitamin A active metabolite, all-trans retinoic acid, induces spinal cord sensitization. I. Effects after oral administration

Abstract

Background and purpose: Retinoic acid is an active metabolite of vitamin A involved in the modulation of the inflammatory and nociceptive responses. The aim of the present study was to analyze the properties of spinal cord neuronal responses of male Wistar rats treated with all-trans retinoic acid (ATRA) p.o. in the normal situation and under carrageenan-induced inflammation. We also studied the expression and distribution of cyclooxygenases (COX) in the spinal cord.

Experimental approach: Properties of spinal cord neurons were studied by means of the single motor unit technique. The expression of COX enzymes in the spinal cord was assessed by Western blot analysis and immunohistochemistry.

Key results: Intensity thresholds for mechanical and electrical stimulation (C-fibers) were significantly lower in animals treated with ATRA than vehicle, either in normal rats or in rats with inflammation. The size of cutaneous receptive fields was also larger in animals treated with ATRA in the normal and inflammatory conditions. The expression of COX-2 enzyme, but not COX-1, was significantly higher in animals treated with ATRA. COX-2 labeling was observed in dorsal horn cells and in ventral horn motoneurons.

Conclusions and implications: In conclusion, the oral treatment with ATRA in rats induces a sensitization-like effect on spinal cord neuronal responses similar to that observed in animals with inflammation and might explain the enhancement of allodynia and hyperalgesia observed in previously published behavioral experiments. The mechanism of action involves an over-expression of COX-2, but not COX-1, in dorsal and ventral horn areas of the lumbar spinal cord.

Figure 1

Illustration of the methods followed in electrophysiological experiments. (a) Original recording of A- and C-volley responses recorded after an electrical stimulus. The threshold for activation of C-fiber volley was determined by the application of 2 ms width square electrical pulses. Data from electrical stimulation were analyzed by counting the responses evoked between 150 and 650 ms after each pulse. (b) Cutaneous receptive fields were mapped on dorsal and plantar sides of the paw using a 500 mN filament. (c) Illustration of an original recording of an unit activated by the application of a 500 mN filament on the most sensitive area of the receptive field during 5 s.

Figure 2

Threshold intensities for mechanical stimulation in animals treated with vehicle or ATRA in the normal and inflammatory conditions. As previously observed, the threshold intensity for the activation of the units was lower in animals with soft tissue inflammation. In normal animals treated with oral ATRA, the threshold was significantly lower than in vehicle-treated rats. Likewise, the threshold was lower in animals with inflammation treated with ATRA when compared to control rats. (^*P<0.05; ^**P<0.01; ^***P<0.001, statistical significance was calculated using the one-way analysis of variance, ANOVA with post hoc Dunnett's test).

Figure 3

Threshold intensities for electrical stimulation in animals treated with vehicle or ATRA in the normal and inflammatory conditions. The threshold intensity for the activation of C-fibre volley was lower in animals treated with ATRA in comparison to vehicle treated animal. As for mechanical stimulation, the difference was observed both in the normal and inflammatory situations. In addition, the threshold observed in normal animals treated with ATRA was very similar to that seen in the control group of rats with inflammation (statistical significance and layout as for Figure 2).

Figure 4

Cutaneous receptive fields of rats treated with vehicle and ATRA in the normal and inflammatory conditions. (a) Extension of cutaneous receptive fields represented for each unit on the dorsal and plantar faces of the paw of animals treated with vehicle or ATRA in the normal and inflammatory conditions. Note the similarities in the shape and size of fields in normal animals treated with ATRA and animals with inflammation treated with vehicle. (b) Quantification of the areas of the fields in the four experimental conditions. The size of fields from units studied in animals treated with ATRA were larger than those in animals treated with vehicle either in normal rats or in rats with inflammation (statistical comparison and layout as for Figure 2).

Figure 5

Expression of COX enzymes in the lumbar segment of the spinal cord of rats treated with ATRA or vehicle. The hemicords ipsilateral (Inflam) and contralateral (Normal) to the inflamed paw were studied separately. The results show a significant enhancement of COX-2, but not COX-1, expression both in the ipsilateral and contralateral hemicords of animals treated with ATRA when compared to that of animals treated with vehicle. Data are expressed as relative values of COX protein (Western blot analysis) normalized for β-tubulin protein levels (^*P<0.05, comparison vs vehicle, using the one-way analysis of variance, ANOVA with post hoc Dunnett's test). A typical Western blot is shown in the upper panel of the figure.

Figure 6

Effect of ATRA on spinal cord COX-2 expression of animals with carrageenan-induced inflammation. Immunohistochemistry for COX-2 in 4 μm sections of the rat lumbar segment of the spinal cord showed a rather homogeneous enhancement of immunoreactivity on the spinal cord of animals treated with ATRA when compared to vehicle treatment (a). A high number of positive cells were observed both on dorsal and ventral horn cells (b). Detailed image (c) shows labeling on ventral horn motoneurons (arrows). (a, b) scale bar: 250 μm; (c) scale bar: 50 μm.