5HT and OCT (Chang et al. 1999b). Studies presently
underway in our laboratory are exploring possible
linkages between the dierent hormonal systems, but the
results are too preliminary to allow their inclusion in this
article. The focus here, therefore, will be entirely on
5HT. Despite the narrow focus, we believe that useful
generalizations and speculations can be made, that
should be testable, and therefore worthy of perusal.
The enigma of amines lies in their ubiquity. They are
important in behavior, everyone agrees, but they are
selective in the behaviors they in¯uence, and in the sign
and magnitude of their in¯uence. 5HT, for example, has
been suggested to serve important roles in memory
formation (cf. Dale et al. 1987; Kandel et al. 1987), in
pain perception (cf. Leung and Mason 1999), in feeding
behavior (cf. Rosen et al. 1983; Schachtner and BraÈunig
1993; Yeoman et al. 1994), in aective conditions like
depression (cf. Stockmeier et al. 1998), in aggression
(cf. Coccaro 1989; Raleigh et al. 1991; Miczek et al.
1994; Olivier et al. 1995; Edwards and Kravitz 1997),
and in any number of other important aspects of animal
and human behavior. Despite its obvious importance, in
all species, the numbers of 5HT neurons are small
compared to the total number of neurons in the nervous
system (only a fraction of a percent), while their sphere
of in¯uence is large, with virtually all areas of the central
nervous system receiving neuronal processes from these
few cells. Often, in target area, dendrites and dendritic
spines closely apposed to 5HT-containing endings show
none of the specialization typical of normal synaptic
contacts (for reviews see Beaudet and Descarries 1978;
Descarries et al. 1990). Thus, the concept of a local
hormone action of amines like 5HT has emerged (see
Beaudet and Descarries 1978; Descarries et al. 1990), in
which speci®city is imparted by the selective distribution
of amine receptors on subsets of cells in the vicinity of
the amine endings (Aghajanian et al. 1990). In some
species, like crustaceans (see below), amines also are
released from neurosecretory neurons into the general
circulation where they may function as hormones, akin
to the amines, steroids and peptide hormones released
from specialized vertebrate endocrine tissues or from
brain neuroendocrine regions. Possibly the greatest
mystery of all, is how amines and amine neurons in¯u-
ence the organization and/or activation of the complex
patterns of behavior with which they are involved.
This communication will focus on serotonergic neu-
rons and agonistic (®ghting) behavior in lobsters, and
will attempt to formulate a speculative synthesis of how
these might be related.
Why a lobster model of aggression?
Amine neuron systems have been mapped and well
studied in a wide variety of invertebrate species
(cf. SchuÈrmann and Klemm 1984; for reviews see NaÈssel
1988; Callaway and Stuart 1999; HoÈrner 1999). Many of
those species show aggression, and in some, for example
Drosophila (Homann 1987; Dow and von Schilcher
1975), powerful molecular and genetic methods in prin-
ciple allow the analysis to be brought to the level of the
genes important for the behavior (Neckameyer and
White 1992; Monasterioti et al. 1996). Our focus remains
on crustacean systems, however, because these animals
are particularly good for studies at many dierent levels
of analysis. First, lobsters and cray®sh are highly ag-
gressive, form long-term stable dominance relationships,
and the behavior has proven amenable to quantitative
analysis (Huber and Kravitz 1995; see also below). Sec-
ond, detailed physiological studies have been carried out
in the isolated central and peripheral nervous systems of
these animals. Many of the neurons found in the nervous
system are large and uniquely identi®able from prepa-
ration to preparation (Otsuka et al. 1967). This allows
extensive examination of the roles served by particular
neurons throughout the lifetime of these animals, and in
animals of diering behavioral status. Moreover, the
functions of such cells have been examined at both cel-
lular and systems levels. Finally, in recent years, genes
relevant to the physiological functioning of lobster neu-
rons have been cloned (cf. Baro et al. 1996a; McClintock
et al. 1997; Xu et al. 1997) and methods have been
elaborated that allow their quantitative measurement in
single identi®ed cells (Baro et al. 1994, 1996b; Schneider
et al. 1999). While methods of mutational analysis are
not practical in these animals because of their long
generation time, the use of molecular methods to search
for changes in the levels of expression of genes relating to
social status and experience now are possible. A future
goal is to develop methods of manipulating the levels of
expression of particular genes in selected neurons at
precise times during the lives of these animals.
Localization and identi®cation of serotonergic neurons
Amines in the lobster nervous system
Serotonin is the major amine derived from tryptophan in
lobsters. OCT is the main amine formed from tyrosine,
but small amounts of dopamine also are found in the
lobster nervous system. Thus, enzyme systems capable
of hydroxylating both the phenol ring of tyrosine (to
dihydroxyphenylalanine and then by decarboxylation to
dopamine) and the side chain of phenylethylamines
(tyramine to OCT) exist in lobster nervous systems, but
not in the same neurons, as no norepinephrine has been
found. Speci®c sets of neurons containing OCT (Schn-
eider et al. 1993) and dopamine (Cournil et al. 1994)
have been mapped in the lobster nervous system, but less
is known about their function. 5HT and OCT appear to
have opposite actions in postural regulation (Living-
stone et al. 1980; Harris Warrick and Kravitz 1984), and
thereby may serve opposite functional roles in aggres-
sion, but studies with OCT have lagged until very re-
cently (R. Heinrich et al., unpublished observations),
and for the most part, will not be considered here.
223