Previous anatomical studies have been unsuccessful in demonstrating signifi
cant cortical inputs to cholinergic and somatostatinergic striatal interneu
rons in rats. On the other hand, electrophysiological studies have shown th
at cortical stimulation induces monosynaptic EPSPs in cholinergic interneur
ons. It has been proposed that the negative anatomical findings might have
been the result of incomplete labeling of distal dendrites. In the present
study, we reinvestigated this issue using m2 muscarinic receptor antibodies
as a selective marker for cholinergic and somatostatinergic interneurons i
n the striatum. This was combined with injections of either the anterograde
tracer biotinylated dextran amine (BDA) in the monkey prefrontal cortex or
aspiration lesion of the sensorimotor cortex in rats. The results showed t
hat, in both species, a small percentage (1-2%) of cortical terminals make
asymmetric synaptic contacts with m2-immunoreactive interneurons in the str
iatum. Interestingly, the majority of these synapses are onto small dendrit
ic spines or spine-like appendages, as opposed to dendritic shafts and/or c
ell bodies. Thus, m2-containing striatal interneurons do receive direct cor
tical inputs and can, therefore, integrate and modulate cortical informatio
n flow through the striatum. Although the density of cortical terminals in
contact with individual striatal interneurons is likely to be relatively lo
w compared to the massive cortical input to projection neurons, both cholin
ergic and somatostatinergic interneurons display intrinsic properties that
allow even small and distal inputs to influence their overall state of neur
onal activity. (C) 2000 Wiley-Liss, Inc.