Welcome
to the Edwards Lab!
Neuroscience
Institute
Georgia State University
850 Petit
Science Center
Atlanta, GA 30303
Tel: (404) 413-5394 (office)
Tel: (404) 413-5357 (Lab)
FAX: (404)
413-5446
email: dedwards@gsu.edu
Current Research
Neuro-Robotics
Reverse-Engineering an Animal
Dominance Hierarchy Formation
Synapses, Serotonin and Social Status
News Accounts
"Science Bits"
References
Coincidence Detection
Growth
and Neuronal Integration
Publications
List of papers
Abstracts
of recent papers
Programs
and Research Centers
Neuroscience
at GSU
Teaching
Neur 8010
Bio 3840
Donald
H. Edwards (Regents'
Professor)
David W. Cofer (Postdoc)
Giselle Linan-Velez (Ph.D. Student)
Eric Randall (Ph.D. Student)
Bryce Chung (Ph.D. Student)
Maryanne Dos Santos (M.S. Student)
Uzma Tahir
(Undergraduate)
Lab Alumni
Undergraduates
Marjorie Sen (Emory
University)
Fadi Issa (Georgia State University)
Daniel Adamson (Paidea
School, Atlanta)
Eric James (St. Johns University, NY)
Katie McAlister (Swarthmore College)
Aaron Miller (Oberlin College)
Catherine McCurdy (Georgia State University)
M.S.
Rodney
Hillis, M.S. 1988
Jiangyan
Shi, M.S., 1999
Lisa
Blumke, M.S., 2008
Currently: Instructor Biology, Georgia Highlands College
Ph.D.
Dr. Ted W. Simon, Ph.D. 1988 Currently: US EPA
Dr.
Shih-Rung Yeh, Ph.D., 1996; Postdoc:
1996-1998 Currently Assoc. Prof. Biology
National Tsing-Hua University, Taiwan
Dr. Steven Versteeg Ph.D. student, Univ. Melbourne, 2004;
Dissertation under my direction, Currently: Research Staff, CA Labs, Melbourne,
Australia
Dr. Cha-Kyong Song Ph.D.,
2006) Currently postdoc,
Ewha Women’s University, Seoul, Korea
Dr. Nadja Spitzer, Ph.D. 2006. Postdoc, 2007-2007; Currently, Postdoc,
Dept. Biology, Marshall Univ., Huntington, W.Va.
Dr.
Barbara E. Musolf, Ph.D. 2007 Currently: Assoc. Professor of Biology, Clayton
State University, Georgia.
Dr.
Fadi A. Issa, Ph.D.
2008. Currently postdoc,
Department of Physiology, UCLA School of Medicine.
Dr. David Wayne Cofer, Ph.D.
2009 Currently postdoc
in this lab.
Postdoctoral Associates:
Dr.
Esther M. Leise, Postdoc:
1989-1990 Currently: Prof. of Biology,
UNC Greensoboro
Dr. Peri Nagappan,
Postdoc: 1990-1994
Currently with the Morehouse School of Medicine
Dr.
Marc Weissburg, Postdoc:
1994 Currently: Prof. of Biology,
Georgia Institute of Technology
Dr. Clifford Opdyke, Postdoc: 1994-1995
Currently: Georgia Environmental Protection Division
Dr. Linda Anderson, Postdoc:
1995-1996
Dr.
Shih-Rung Yeh, 1996-1998 Currently Assoc. Prof. Biology National Tsing-Hua University, Taiwan
Dr.
Joanne Drummond, Postdoc: 1996-1998 Currently with GlaxoSmithKline Australia
Dr.
Jens Herberholz, Postdoc:
1999-2005 Currently: Assist. Prof.
Psychology, Univ. Maryland,
College Park
Dr.
Brian Antonsen, Postdoc:
1999-2007 Currently: Assist. Prof.
Biology, Marshall Univ., Huntington, W.Va.
Dr.
Jeffrey Triblehorn Postdoc:
2003-2006 currently: Assist Prof. College of Charleston, Charleston, SC
Dr. Nadja Spitzer, Postdoc:
2007-2007; Currently: Postdoc, Dept. Biology,
Marshall Univ., Huntington, W.Va.
Dr.
Fadi A. Issa, Postdoc, 2008 Currently postdoc,
Department of Physiology, UCLA School of Medicine.
Dr. David W. Cofer, Postdoc: 2009-2011.
Technicians
Mr. Christopher Mims (Tech. 2001-04) Currently: Editor, Scientific American.com
Ms. Elizabeth DeGoursac
(Tech. 2002-03): Currently: Emory Univ. Neuroscience Grad. Prog.
Ms. Laurel Johnstone (Tech.
2000-06)
Photo Gallery
































Current Research: Neuro-Robotics
We use a "reversed-engineering" approach to
discover how animals work. We first
study the behavior, anatomy, and physiology of an animal to determine how its
parts fit together and interact. We then build computer models of these parts,
including some or all of the body and the nervous system, that capture their
key properties that allow them to function.
We then put the models together in an overall model of the animal, place
it in a virtual Newtonian world, and then test it to see whether the model's
behavior resembles that of the animal, and how the its different parts
contribute to the behavior.
|
Arm Flexion Sensory feedback from muscle spindles and
other receptors mediates reflexes that stabilize the body against outside
perturbations. During voluntary limb
movement, the reflex circuitry has to be prevented from interfering with the
movement while also assisting the limb to reach the new position, stabilized
against additional perturbation. We
studied this using an AnimatLab model of human arm flexion (Fig. 1, Movie 1), and described it in Cofer et al., 2010a.
Locust Jump The
neural circuitry and biomechanics of kicking in locusts have been studied to
understand their roles in the control of both kicking and
jumping. It had
been hypothesized that the same neural circuit and biomechanics governed both
behaviors, but this hypothesis was not testable with current technology. We built a neuromechanical model according to
current knowledge of the leg biomechanics and the anatomy and physiology of the
neural circuit that produced kicking (Fig. 2). When excited appropriately, the
circuit produced model jumping and kicking behaviors (see Fig. 3, Movie) that we compared to published results of real kicks
and jumps and to the kicking and jumping behavior of 5 live locusts. We found that the model replicated the live
data for both kicks and jumps in all particulars. This confirmed that the kick
neural circuitry can produce the jump behavior.
This is described in Cofer et
al., 2010b and Cofer et al., 2010c.

Crayfish We are interested in how the nervous system
controls the behavior of animals. To
this end, we have focused on the crayfish, a small invertebrate that has been
extensively studied for the past 100 years (see www2.biology.ualberta.ca/palmer/thh/crayfish.htm for Thomas Huxley's classic description), and has a
large behavioral repertoire and a relatively small nervous system. crayfish live primarily underwater, in
streams, rivers, ponds and swamps, but they also venture out on land when it's
wet, sometimes migrating considerable distances overland between
watersheds. They dig multichambered
burrows in mud banks for homes, and live and have their babies in them. They often live in communities of other
crayfish, and compete with them for food, shelter, and mates. They have hierarchical social relationships
based on their social interactions with each other and on their size and
personality. They forage for food on land and underwater, eating plants and
preying on other animals. Our research in the past has concerned the escape
behavior of crayfish, the circuits that produce escape, and how they and the
behavior is modulated by serotonin and by changes in the animal's social
status. Descriptions of this research
are found below.
Crayfish Posture and Walking We have been
studying the control of posture and locomotion in crayfish, with particular
attention to the role of sensory feedback from proprioceptors
like the coxa-basipodite chordotonal
organ (CBCO), which measures the angle and rate of change of the angle of the
depression-levation joint of the legs. When crayfish walk, each leg elevates and
promotes during the swing phase, and then depresses and remotes during the
stance phase. Depressor motorneuron bursts alternate with anterior levator motorneuron bursts; the posterior levator overlaps both,
beginning during the depression phase (Fig. 4). A movie of the animal walking,
the motor nerves firing, and plots of the joint angles of the 5th left leg can
be seen here: Crayfish Walking.
To start, have constructed a 4-legged
AnimatLab model of the crayfish based on published descriptions of the animal's
anatomy, the origins, insertions, and properties of the walking leg muscles,
and the physiology of its sensory and motor neurons (Fig. 5). (A 4-legged model is significantly less
complex than an 8-legged model). The rhythmic movements of the legs during
walking are governed by a set of central pattern generators (CPGs). Because the
precise circuitry that produces the relative phasing of each of the joints in a
leg has not been described, we have made up circuitry that does that. We are using the model to study the role of
the CBCO in mediating resistance and assistance reflexes during standing and
locomotion. As can be seen here (Crayfish
Walking Model), the model captures
many of the features of crayfish walking.

Our primary focus has been on one of the
best-understood circuits in any nervous system, the tail flip escape circuit in
crayfish. We have used this circuit and behavior to study a variety
of issues, including changes in the brain during social hierarchy formation (Yeh, et al., 1996; Yeh, et al., 1997) (Issa et al., 1999), coincidence detection (Edwards, et al., 1998), and the neural bases for behavioral choice (Edwards,
1991;
see the web-based simulation of our model by Steven Versteeg
at the University of Melbourne (http://www.cs.mu.oz.au/~scv/sim/simcray.html). Many of
these results are summarized in a review in Trends
in Neurosciences. We are also
interested in understanding the neural bases for behavioral choice (Edwards, 1991; see the web-based simulation of our model by
Steven Versteeg at the University of Melbourne (http://www.cs.mu.oz.au/~scv/sim/simcray.html). This work was supported by the NSF and
by the NIH.
Dominance Hierarchy Formation Social animals form hierarchies so as to divide resources without unnecessary conflict. Conflict may occur at the outset, however, when two unfamiliar animals meet, and they are of about equal size and strength. Then agonistic interactions may escalate to full fighting. Usually these fights are brief, and end when one animal suddenly withdraws. The new dominant may pursue briefly, so as to 'pound in' the lesson of who won and who lost. The sudden change in behavior, from fighting to retreat, is indicative of a corresponding change in the brain about which we know very little. We have found that in crayfish, this change in behavior is reflected in the frequency with which several agonistic behaviors (approach, attack, and offensive tailflip) and several defensive behaviors (retreat, three forms of escape tailflip) are displayed (Herberholz
et al., 2001).
Before the decision to withdraw by one animal, both animals displayed similar
patterns of attack and offensive tailflip,
accompanied by very few defensive behaviors. Suddenly one animal would
break contact and initiate a rapid series of tailflips
that would carry it away from the dominant.
The tailflips were activated by two different neural circuits, the 'medial
giant' and 'non-giant' circuits, that trigger rearward escapes in response to
frontal attacks. The sudden release of these tailflips indicates that the
stimulus threshold for activating the circuits must have changed from very high
before the decision to withdraw to very low thereafter. We don't yet know
the mechanism for this transition, except to say that both circuits are
under strong inhibitory control; it is likely that one consequence of the
decision to withdraw is removal of that inhibition. Interestingly, the
lateral giant circuit (see below), which triggers an upward tailflip escape in
response to an attack from the rear, was activated only once during fights
between 8 pairs of animals, suggesting that it is inhibited throughout such
contests.
Synapses, Serotonin and Social Status. It
has been known for some time that serotonin, a psychoactive neurochemical
implicated in the control of mood, aggression, blood flow, and digestion, also
affects the response of the lateral giant (LG) neuron (picture at right) in
crayfish to its normal sensory inputs (Glanzman and Krasne, 1983). The
LG neuron is excited by a tap on the tail that might occur when the animal is
attacked, and it triggers a tail flip escape response that moves the animal
quickly away from the stimulus (see picture above). Recently we found
that the effect of serotonin on LG's response depends on the social status of
the crayfish (Yeh, et al., 1996; Yeh, et al., 1997).
Serotonin increases the responsiveness of the LG neuron in crayfish that have
been isolated for a month or more ("isolates"), and this increase
persists for several hours after the serotonin is removed. In social
subordinates, however, serotonin inhibits LG's response, whereas in social
dominants, LG's response is increased again. In these last two, the
effects of serotonin persist only as long as the drug is present.
Social hierarchies form
when unfamiliar animals get together (Issa et al., 1999). The larger animal
will usually become the dominant member of the pair, with first choice of all
the available resources (e.g., food, shelter), whereas the smaller will become
subordinate. Although the social status of the two animals is decided
within the first hour of pairing, usually after a few brief bouts of
fighting, the differences in the effect of serotonin on LG's response
take nearly two weeks to develop fully. Other aspects of the
animals' behavior changes gradually during this period as they settle into
their new social roles.
Should the subordinate be
reisolated, or should it be re-paired with another subordinate and become
dominant to that new animal, the inhibitory effect of serotonin will change to
the facilitatory effect characteristic of isolate or dominant
animals. Should a dominant animal be paired with and become
subordinate to another dominant animal, the effect of serotonin does not change
to that typical of the other subordinates. Rather it retains its
facilitatory character, even after more than a month of subordinate
status. These results suggest that if dominant status is the preferred
social state, then the facilitatory effect of serotonin is a preferred
physiological state.
Many different drugs activate
different classes of receptor molecules for serotonin in vertebrates.
When two of these were substituted for serotonin in the crayfish experiments,
one had inhibitory effects on LG's response in both dominant and subordinate
animals, and the other had facilitatory effects in both animals. These
and other data suggest that the changes in the effect of serotonin following a
change in social status resulted from changes in the receptors for
serotonin in the LG neuron.
This work is significant because it
is the first report that a change in an animal's social status can change the
population of molecular receptors for, and the effect of, a neurochemical
like serotonin in the nervous system. In view of this result, it
appears likely that the obvious difference in the behavior of dominant and
subordinate animals occurs in part because their nervous systems are different,
so that serotonin now has different effects in the two animals. This
suggestion is supported by recent behavioral experiments which showed that
LG-evoked tail flips become much more difficult to evoke in subordinate
crayfish than in dominant crayfish during a fight, when serotonin may be
naturally released in the nervous system (Krasne et al., 1997).
News Accounts of this Work. Several press accounts of this work
are available. These include:
· Glanzman,
D.L. and Krasne, F.B. Serotonin and octopamine have opposite modulatory effects
on the crayfish's lateral giant escape reaction. J Neurosci 3:2263-2269, 1983.
· Krasne,
F.B., Shamasian, A., and Kulkarni, R. Altered excitability of the crayfish
lateral giant escape reflex during agonistic encounters. J.Neurosci.
17(2):692-699, 1997.
(Return to Contents)
Coincidence
Detection. Coincidence detection is
important for functions as diverse as Hebbian learning, binaural localization,
and visual attention. We have found that extremely precise coincidence
detection is a natural consequence of the normal function of rectifying
electrical synapses (Edwards et al., 1998). Such synapses open to bidirectional
current flow when presynaptic cells depolarize relative to their postsynaptic
targets, and remain open until well after completion of presynaptic
spikes. When multiple input neurons fire simultaneously, the synaptic
currents sum effectively and produce a large EPSP. However, when some
inputs are delayed relative to the rest, their contributions are reduced by the
early EPSP and by postsynaptic current shunts through junctions already opened
by the earlier inputs. These mechanisms account for the ability of the lateral
giant neurons of crayfish to sum synchronous inputs, but not inputs separated
by only 100 msec. This coincidence detection
enables crayfish to produce reflex escape responses only to very abrupt
mechanical stimuli. In light of recent evidence that electrical synapses
are common in the mammalian CNS, the mechanisms of coincidence detection
described here may be widely used in many systems. (Return to Contents)
Growth
and Neuronal Integration.
Neurons must work both when they are small in young animals and when they are
larger in adult animals. The increase in size that neurons experience,
however, changes the way in which electrical current flows through them, and so
changes they way in which they respond to synaptic inputs. We have
found that two sets of giant neurons, one in cricket and one in crayfish,
exemplify two different patterns of neuronal growth that have different effects
on neuronal integration. The medial giant interneurons (MGI) in cricket
maintains its response properties as it grows. It grows approximately
uniformly, but such that the diameters of its processes increase as the square
of their increase in length (Hill et al., 1994). The lateral giant (LG) neuron in
crayfish becomes more of a low-pass filter as it grows, which causes it to
switch input circuits and become susceptible to response habituation. It
grows approximately isometrically (Edwards et al., 1994a, b) . A cable analysis of these
different patterns of uniform growth indicates that, regardless of the initial
shape or distribution of active and passive membrane properties, a cell that
grows uniformly and with neurite diameters increasing as the square of their
increase in length will not change the way voltage is distributed through the
cell (Olsen et al., 1996). This "isoelectrotonic"
pattern of growth will enable a synaptic potential created at corresponding
positions in the small and large cells to evoke identical responses at all
other corresponding points in the two cells. Conversely, uniform isometric growth
causes the cell to become electrically larger, and to become a low-pass
filter. These analytical results account for the different effects of
growth on the two cells, in which MGI grows isoelectrotonically, and LG grows
isometrically. (Return to Contents)