actions involving substance P (SP), a peptide present in both unmyelinated
primary afferents and interneurons within the dorsal horn. We examined pote
ntial functional sites for interactions between SP and MOR by using dual el
ectron microscopic immunocytochemical localization of antisera against SP a
nd a sequence-specific antipeptide antibody against MOR in rat cervical spi
nal dorsal horn. The distribution was compared with that of the functionall
y analogous dorsal horn of the trigeminal nucleus caudalis. Many of the SP-
immunoreactive terminals in the dorsal horn contacted dendrites that contai
n MOR (53% in trigeminal; 70% in cervical spinal cord). Conversely, within
the cervical spinal dorsal horn 79% of the MOR-labeled dendrites that recei
ved any afferent input were contacted by at least one SP-containing axon or
terminal. Although SP-immunoreactive dendrites were rare, many of these (4
8%) contained MOR, suggesting that the activity of SP-containing spinal int
erneurons may be regulated by MOR ligands. A few SP-labeled terminals also
contained MOR (12% in trigeminal; 6% in cervical spinal cord). These data s
upport the idea that MOR ligands produce antinociception primarily through
modulation of postsynaptic second-order nociceptive neurons in the dorsal h
orns of spinal cord and spinal trigeminal nuclei, some of which contain SP,
They also suggest, however, that in each region, MOR agonists can act pres
ynaptically to control the release of SP and/or glutamate from afferent ter
minals. The post- and presynaptic MOR sites are likely to account for the p
otency of MOR agonists as analgesics. (C) 2000 Wiley-Liss, Inc.