Immobility and flight associated with antinociception produced by activation of the ventral and lateral/dorsal regions of the rat periaqueductal gray¹

Abstract

It has long been known that the periaqueductal gray (PAG) plays an important role in the modulation of nociception. Given that activation of the lateral PAG also produces wild running and tachycardia, it has been suggested that PAG mediated antinociception is part of an integrated defensive reaction. However, an alternative hypothesis is that these effects are merely a secondary response to aversive brain stimulation. If antinociception and flight reactions are caused by aversive brain stimulation, then these effects should always occur together. The objective of the present study was to determine whether antinociception and locomotion could be dissociated by microinjecting morphine and kainic acid into various subdivisions of the caudal PAG. Non-selective activation of lateral and dorsal regions of the PAG by microinjection of kainic acid produced wild running, while injections into the ventrolateral PAG produced immobility. Microinjection of morphine evoked similar locomotor effects, although the onset to effect was slower with morphine (approximately 5 min vs. 1 min for kainic acid), and the antinociceptive efficacy of microinjecting 0.2 μl of morphine was less than with kainic acid injections. In fact, microinjection of morphine evoked locomotor effects in the absence of antinociception on 39% of the tests. Increasing the injection volume to 0.4 μl (dose remained at 5 μg) greatly enhanced the likelihood that antinociception and locomotor effects (e.g. running, freezing, circling) occurred simultaneously (79%). These findings indicate that, although distinct locomotor effects are associated with antinociception from the ventral and more dorsal regions of the PAG, antinociceptive and locomotor effects can occur independently. This finding is consistent with the hypothesis that ventral and dorsal regions of the PAG integrate defensive freezing and flight reactions, respectively.

Introduction

It has long been known that electrical or chemical activation of the periaqueductal gray (PAG) produces antinociception [17][28][38]. In addition, activation of the PAG has been shown to produce running, jumping, an increase in blood pressure, tachycardia, and redistribution of blood flow to skeletal muscles [7][8][14][17][40]. These findings suggest that antinociception evoked from the lateral PAG is part of a coordinated defensive reaction [2][3][13]. However, stimulation of the lateral and dorsal PAG also has been shown to be aversive [15][33][34] raising the possibility that the resulting antinociception is secondary to aversive brain stimulation (i.e. a form of stress-induced analgesia) rather than an integrated component of the defense response.

In contrast to the overt defensive behaviors evoked in response to activation of lateral and dorsal regions of the PAG, activation of the ventrolateral PAG produces what has been described as `pure' antinociception [14]. However, this appears to be an overstatement given that microinjection of excitatory amino acids into the ventrolateral PAG produces immobility in a test of social interaction [11][12] and inactivation of the ventrolateral PAG disrupts fear-induced freezing [10][22][23]. Given that freezing is a defensive response to natural stimuli [13], antinociception resulting from activation of the ventrolateral PAG may be part of an integrated defensive freezing response or merely another form of stress-induced analgesia.

If antinociception resulting from activation of the PAG is a form of stress-induced analgesia, rather than an integrated defensive response, then antinociception and locomotor responses should always occur together. In an attempt to determine whether antinociceptive and locomotor effects can be dissociated, we took advantage of the observation that excitatory amino acids and morphine activate PAG neurons via different mechanisms: Excitatory amino acids activate PAG neurons directly, whereas morphine activates a subset of PAG neurons via disinhibition [35][42].

In addition, it may be possible to dissociate antinociceptive and locomotor effects by microinjecting drugs into different PAG regions. The effects of electrical stimulation of the PAG indicate that site specific differences in antinociception and locomotion exist. Electrical stimulation of the ventral, but not lateral and dorsal regions of the PAG produces antinociception that is attenuated by the opioid antagonist naloxone [6], shows tolerance [31], and modulates nociception via both ascending and descending pathways [32], whereas electrical stimulation of lateral and dorsal, but not ventrolateral regions of the PAG produces flight reactions [14]. Given that electrical stimulation activates both cell bodies and fibers of passage, it is difficult to localize effects using this technique. Thus, the present study is the first to systematically map PAG sites for both antinociceptive and locomotor effects using small injections of morphine and kainic acid.