proposed that the expression of lEGs, either by immunocytochemistry or in situ 
hybridization, provides a sensitive assay for assessing the activity of neurons in 
the CNS. 
Since the basal levels of Fos protein in the spinal cord are low, stimulus- 
induced increases in Fos expression are readily detected. In studies of c -fos 
expression after noxious peripheral stimulation, it was first demonstrated that 
the pattern of Fos immunoreactivity is consistent with the anatomical location 
of nociresponsive neurons in the dorsal and ventral horn. Thus, following 
noxious stimulation in anesthetized rats, Hunt and colleagues (1987) reported 
that Fos immunoreactivity was concentrated in regions that contain neurons 
that respond to noxious stimulation, including the superficial laminae (I and llo) 
and in lamina V of the dorsal horn. Two important questions arose from that 
study. First, would the pattern of Fos expression be similar in awake animals? 
Second, would the magnitude of the behavioral responses to noxious stimuli 
correlate with the numbers of Fos-immunoreactive neurons? To address these 
questions, the authors developed a model whereby they could administer a 
noxious stimulus to a rat, record its behavioral response to the stimulus, and, 
most important, correlate the behavior with the levels of Fos expression in the 
cord. 
FOS EXPRESSION AFTER SOMATIC NOXIOUS STIMULATION 
In the authors’ initial studies, Gogas and colleagues (1991) used the formalin 
pain model in which a dilute concentration of formalin is injected into the 
hindpaw of the rat. The formalin model has been well characterized both 
electrophysiologically and pharmacologically (Dickenson and Sullivan 1987a, 
1987b; Hunskaar et al. 1986; Drower et al. 1987). Formalin injection produces 
a characteristic behavioral response that lasts for as much as 4 hours and can 
be readily quantified in terms of intensity (Dubuisson and Dennis 1977). The 
basic protocol was to inject the formalin into the plantar surface of the hindpaw. 
The rats were then placed into an observation chamber so that their behavior 
could be observed and recorded over a 1-hour period (Gogas et al. 1991). 
The animals were then anesthetized and perfused, with the spinal cord tissue 
removed and postfixed as previously described (Presley et al. 1990; Gogas et 
al. 1991). The tissue was processed immunocytochemically with the avidin- 
biotin procedure of Hsu and colleagues (1981) using commercially available 
kits (Vectastain ABC; Vectorlabs, Burlingame, CA). We have used a rabbit 
polyclonal antiserum directed against an in vitro translated product of c -fos 
(provided by Dr. Dennis Slamon of the University of California, Los Angeles). 
An important advantage of Fos immunocytochemistry is that the results can be 
studied quantitatively. To this end, the outlines of three sections through the 
136 
