First published online July 26, 2004
Journal of Experimental Biology 207, 3043-3053 (2004)
Published by The Company of Biologists 2004
doi: 10.1242/jeb.01136
The biomechanical and neural control of hydrostatic limb movements in Manduca sexta
Sheri Mezoff,
Nicole Papastathis,
Anne Takesian and
Barry A. Trimmer*
Department of Biology, Dana Laboratory, Tufts University, Medford, MA
02155, USA
View larger version (48K):
[in a new window]
|
Fig. 1. Anatomy of a Manduca proleg. (A) A late 5th instar larva is
illustrated from the right-hand side. Prolegs are found on abdominal segments
3-6 (A3-A6). The specialized legs on the most posterior segments are called
claspers. (B) A single right-hand side proleg is illustrated in a
three-quarter frontal view and is representative of each proleg on body
segments A3-A6. The pseudo segments, subcoxa and coxa are named by analogy
with segments in the articulated thoracic legs. The tip of the proleg is
called the planta and it carries the curved cuticular hooks (crochets) used
for gripping. The principal and accessory planta retractor muscles (PPRM and
APRM, respectively) have their origin on the lateral body wall near the
spiracle (not illustrated) and insert at the planta and coxa-subcoxa boundary,
respectively. The medial hairs are located along the inner surface of the
proleg.
|
|
View larger version (37K):
[in a new window]
|
Fig. 2. Kinematics of adduction. (A) Markers were placed to span the cuticle of the
subcoxa (circles), coxa (diamonds), planta (squares) and subcoxa-coxa membrane
(triangles). These were tracked in three dimensions during adduction and the
distance between them calculated to estimate the main site of extension. Most
of the total extension results from an increase in the spacing across the
subcoxa-coxa membrane with a small contribution from cuticle stretching in the
subcoxa. (B) In a related experiment, markers were placed at lateral
(squares), central (triangles) and medial (circles) positions spanning the
subcoxa-coxal membrane and their spacing tracked during adduction. Both the
rate and extent of membrane stretching are greatest at the lateral margin and
decrease towards the medial surface causing adduction. In all cases, extension
and adduction are coincident. In each figure, the results are the mean length
changes (± S.E.M.) from three discrete
adductions in one larva. Data from each adduction were aligned relative to the
transition point from decreasing to increasing length at the subcoxa-coxal
membrane but were otherwise not normalized. Different larvae were used for A
and B and each panel is representative of nine adductions in three larvae.
|
|
View larger version (28K):
[in a new window]
|
Fig. 3. Identification of medial hair (MH)-activated motor activity. (A) Whole
nerve cords from the thoracic to the terminal ganglion were removed along with
the attached A4 proleg. Recordings were made from nerves contralateral to the
removed proleg. The dorsal nerve (DN) was dissected such that the anterior
(DNA), lateral (DNL) and posterior (DNP)
branches were distinct, and extracellular recordings were made from each
branch. The ventral nerve (VN) was not dissected into separate branches.
Representative traces are shown of the spike activity before, during and after
MH stimulation, which is indicated by the marker trace below each record. (B)
Histograms showing mean responses of each nerve branch to an MH stimulus. The
number of distinct events was counted for 1 s before the stimulus (open bars)
and for 1 s beginning 1 s after the start of an MH stimulus (shaded bars).
Each bar is the mean (± S.E.M.) of 3-6
preparations. Asterisks indicate significant differences (paired
t-tests, P<0.05).
|
|
View larger version (18K):
[in a new window]
|
Fig. 4. Medial hair (MH) stimulation excites VIL bilaterally and multisegmentally.
(A) The preparation consisted of a whole animal with the gut removed and
pinned to display the ventral muscles and nerve cord. A suction electrode was
placed on muscle fibers of VIL. For illustrative purposes, the size of the
nerve cord has been exaggerated relative to the segmental muscles. (B)
Representative traces are shown from extracellular recordings taken from VIL
in segments A3 (N=3), A4 (N=10), A5 and A6 (N=3)
(left side) and A5 (right side, N=2). Stimulus bars below the
recordings show the duration of mechanical MH stimulation on the left side A4
proleg.
|
|
View larger version (27K):
[in a new window]
|
Fig. 5. Proleg movements are highly correlated with the electromyogram (EMG)
activity of the principal planta retractor muscle (PPRM). (A) Spontaneous
proleg movements were monitored in a restrained larva by measuring the
separation of the right and left plantas in one segment. Upward deflections
are retractions with the peaks representing complete withdrawal and the
troughs full adduction. The activity of PPRM on one side of the same segment
was monitored with a bipolar electrode inserted into the origin on the body
wall. Each retraction is concurrent with a burst of activity in PPR. One
exception to this finding (arrow) is shown at the beginning of the trace and
this corresponds to the unilateral withdrawal of the opposite leg. (B) An
expanded part of the record shown in A to illustrate the start of adduction
corresponds to the end of a burst of activity in PPRM. These responses are
typical of several hundred cycles of both spontaneous and evoked proleg
movements in six preparations.
|
|
View larger version (28K):
[in a new window]
|
Fig. 6. The activity of the principal planta retractor muscle (PPRM) is highly
correlated to that of the ventral muscles. (A) Electromyogram (EMG) recordings
from the origin of PPRM and the insertion point of VIL in abdominal segment A5
were made in a restrained larva. Cycles of retraction (gray regions) and
adduction were evoked by mechanically stimulating the planta hairs and the
medial hairs, respectively. The activity index for each muscle was calculated
by demeaning and rectifying the EMG (no smoothing), then integrating the
voltage using a 200 ms bin. In most cases, activity in the two muscles was
coincident but not identical. (B) A cross-correlation analysis (MatLab,
function xcorr) of the activity index for each muscle shows that they are
highly correlated with little or no response lag.
|
|
View larger version (36K):
[in a new window]
|
Fig. 7. Body pressure changes in response to adduction and retraction. (A) Pressure
changes were measured at the base of the subcoxa in segment A4 (upper trace)
while monitoring the separation of the prolegs in all body segments (lower
traces). An upward deflection indicates an increase in pressure and retraction
of the prolegs. Proleg movements tend to occur concurrently in different body
segments. The pressure scale bar corresponds to 980 Pa (10 cmH2O).
(B) The first 9 s of data for the pressure and prolegs in A5 are shown on an
expanded scale. The magnitude of the pressure trace has been rescaled to
demonstrate the close relationship between pressure changes and proleg
movements. (C) A cross-correlation plot of movements and pressure change in
A5. A peak at the dotted line would indicate exact coincidence of movement and
pressure. Here, the peak lags behind, showing that the pressure pulse precedes
the proleg movement by 200 ms. This correlation is very similar for the other
prolegs at the start of the recording but the relationship breaks down
completely for the rest of the recording (see text).
|
|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
Twitter What's this?
© The Company of Biologists Ltd 2004
