Updated Review-13

Several motivating stimuli for defense appear to be innate and relatively independent of learning effects: defense in response to pain; and defense in response to certain defense pheromones and odors. The midbrain central gray receives input from spinal pain systems (e.g. Keay et al., 2001); this can provoke defense in decerebrate animals as previously reviewed (Adams, 1979a). The sensory analyzers of defense pheromones, such as the cat odor which provokes defense in rats, may also be innate. This has received extensive analysis using c-Fos immunoreactivity as well as chemical and electrolytic lesions in recent years (Canteras et al., 1997; Dielenberg et al., 2001; Blanchard et al., 2003) and appears to involve the anterior olfactory nucleus, posteroventral medial amygdaloid nucleus, and areas of the medial hypothalamus including the dorsal premammillary nucleus. From the medial hypothalamus and dorsal premammillary nucleus there are direct projections to the dorsal midbrain central gray (Canteras, 2002) where the c-Fos studies also show intense neural activity. This may help explain the functional significance of escape behavior elicited by chemical stimulation of the medial hypothalamus (Di Scala et al., 1984). Some motivating stimuli of defense, including sudden noises and visual movement, are influenced by learning but may not require it; this is suggested by findings that they can produce defense in rats or cats after removal of the forebrain, as previously reviewed (Adams, 1979a). The studies of Brandao and colleagues have implicated the inferior colliculus and superior colliculus in the mediation of auditory and visual elicitation of defense, respectively (Brandao et al., 1999). There is evidence for the involvement of the inferior colliculus in unlearned auditory-elicited defense (Brandal et al., 1993); although the inferior colliculus is not necessary for a learned escape response to noise (Lyon, 1964). The role of the superior colliculus in visually-elicited defense is supported by the finding that lesions of this structure reduce defense by wild rats to approach of the experimenter (Blanchard et al., 1981).

There are learning effects on most of the motivating stimuli of defense, as shown in the figure at points 1-8. Research suggests that, to some extent, the learning requires participation of the amygdala, which is a particularly complex structure, situated at a neural crossroads with strong inputs to the rest of the forebrain (cf. Petrovich et al., 2001).

Fear conditioning appears to take place in the basolateral amygdala which may also be called the frontotemporal region of the amygdala, according to recent research (Davis et al., 2003; Fanselow and Gale, 2003; Fanselow and LeDoux, 1999). This is illustrated in the figure as learning point 1. In apparent contradiction to this, an earlier study found that avoidance of a previously attacking animal depends not on the basolateral amygdala but on the corticomedial amygdala (Luiten et al., 1985).

The mechanism of neophobia which is a motivating stimulus for defense may involve the caudate-putamen and lateral amygdala according to lesion studies by Cigrang et al (1986) and Misslin and Ropartz (1981), respectively. This is shown in the figure as learning point 4.

Developmental learning effects on defensiveness include increased reactivity to noises and movements following exposure to an environment where escape is possible (Clark and Galef, 1977) and taming of animals by handling during infancy (Galef, 1970). These effects are shown in the figure as learning points 5 and 6, respectively. It would be interesting to know if these types of learning also involve structures of the amygdala.

Recognition of a consociate, as mentioned above, is essential to the operation of the consociate modulator which has been hypothesized to shift an animal from anti-predator to consociate defense. This proposal, made in the previous review (Adams, 1979a) was used to explain increased levels of anti-predator defense obtained by lesions of the septum and ventromedial hypothalamus in many studies (see review by Albert and Walsh , 1984), as well as the interaction between these structures and those of the amygdala where lesions have the opposite effect. Hence, taming effects of amygdala lesions can be reversed by lesions of the septum in the rat (King and Meyer, 1958) and ventromedial nucleus in the cat (Kling and Hutt, 1958). On the other hand, hyperdefensiveness caused by septal lesions in the rat can be reversed by combining them with amygdala lesions (King and Meyer, 1958; Blanchard et al., 1979).

(continued on next page)