on Fos expression in the nucleus proprius, the region that contains primarily 
nonnociresponsive neurons, was flat; that is, no significant inhibition of Fos 
expression at any of the morphine doses tested was found. This is consistent 
with the fact that morphine predominantly inhibits nociresponsive neurons (Le 
Bars et al. 1976a, 1976b). By contrast, the inhibition of Fos immunoreactivity 
in the superficial dorsal horn, laminae V and VI, and the ventral horn of the cord 
was dose related. The dose-response curves in laminae V and VI and 
the ventral horn were steeper than in the superficial laminae. However, even 
at the highest dose tested, which produced 100 percent inhibition of behavioral 
analgesia, there was still significant residual "activity” in all areas, particularly 
the superficial cord. Since a major component of the residual staining was 
found in the marginal layer (lamina I) of the superficial dorsal horn, a region 
that contains neurons that respond almost exclusively to noxious stimulation, 
this finding is of particular interest. However, lamina I contains a mixture of 
projection and interneurons. In a double labeling study, Men^trey and 
colleagues (1989) found that approximately 45 percent of marginal layer 
neurons that express Fos following noxious stimulation project to regions 
of the brain that have been implicated in the rostral transmission of nociceptive 
information. Conceivably, opioids preferentially inhibit the population of lamina I 
projection neurons; that is, the residual labeling may be concentrated in a 
population of interneurons that do not contribute to the “pain”-related response. 
That retrograde labeling and Fos immunocytochemistry can be combined 
(figure 8) makes it possible to determine whether opioids produce a selective 
reduction in the number of retrogradely labeled marginal cells that express 
Fos immunoreactivity. This possibility is being tested. Again, it is important 
to emphasize that this sort of experiment could not be done without using Fos 
immunoreactivity as a marker for populations of neurons activated by noxious 
stimulation. 
THE EFFECTS OF SUPRASPINALLY ADMINISTERED OPIOIDS ON 
NOXIOUS STIMULUS-EVOKED FOS EXPRESSION 
Since systemically administered morphine produces analgesia via an action 
at both spinal and supraspinal sites (Yaksh and Noueihed 1985; Basbaum 
and Fields 1984), we could not distinguish the relative contribution of either 
site to the inhibitory effects found. One approach to addressing this issue was 
to separately test the effects of supraspinal and intrathecal (IT) administration 
of morphine on formalin-evoked Fos expression in the cord. It was found 
that administration of morphine into the third ventricle (ICV) or directly 
onto the spinal cord (IT) inhibits both formalin-evoked behaviors and Fos 
immunoreactivity (K.R. Gogas, G.l. Batchkine, H.J. Cho, and A.I. Basbaum, 
manuscript in preparation). Since morphine is relatively nonselective for \x-, 
6-, or K-receptors (Martin 1983), it is still not clear which opioid receptor subtype 
147 
