Processing of species-specific auditory patterns in the cricket brain by ascending, local, and descending neurons during standing and walking


Zorović, M.; Hedwig, B.

Journal of Neurophysiology 105(5): 2181-2194

2012


The recognition of the male calling song is essential for phonotaxis in female crickets. We investigated the responses toward different models of song patterns by ascending, local, and descending neurons in the brain of standing and walking crickets. We describe results for two ascending, three local, and two descending interneurons. Characteristic dendritic and axonal arborizations of the local and descending neurons indicate a flow of auditory information from the ascending interneurons toward the lateral accessory lobes and point toward the relevance of this brain region for cricket phonotaxis. Two aspects of auditory processing were studied: the tuning of interneuron activity to pulse repetition rate and the precision of pattern copying. Whereas ascending neurons exhibited weak, low-pass properties, local neurons showed both low- and band-pass properties, and descending neurons represented clear band-pass filters. Accurate copying of single pulses was found at all three levels of the auditory pathway. Animals were walking on a trackball, which allowed an assessment of the effect that walking has on auditory processing. During walking, all neurons were additionally activated, and in most neurons, the spike rate was correlated to walking velocity. The number of spikes elicited by a chirp increased with walking only in ascending neurons, whereas the peak instantaneous spike rate of the auditory responses increased on all levels of the processing pathway. Extra spiking activity resulted in a somewhat degraded copying of the pulse pattern in most neurons.

J
Neurophysiol
105:
2181-2194,
2011.
First
published
February
23,
2011;
doi:10.1152/jn.00416.2010.
Processing
of
species-specific
auditory
patterns
in
the
cricket
brain
by
ascending,
local,
and
descending
neurons
during
standing
and
walking
M.
Zorovie
and
B.
Hedwig
Department
of
Zoology,
University
of
Cambridge,
Cambridge,
United
Kingdom
Submitted
6
May
2010;
accepted
in
final
form
21
February
2011
Zorovie
M,
Hedwig
B.
Processing
of
species-specific
auditory
patterns
in
the
cricket
brain
by
ascending,
local,
and
descending
neurons
during
standing
and
walking.
J
Neurophysiol
105:
2181-2194,
2011.
First
published
Febru-
ary
23,
2011;
doi:10.1152/jn.00416.2010.—The
recognition
of
the
male
calling
song
is
essential
for
phonotaxis
in
female
crickets.
We
inves-
tigated
the
responses
toward
different
models
of
song
patterns
by
ascending,
local,
and
descending
neurons
in
the
brain
of
standing
and
walking
crickets.
We
describe
results
for
two
ascending, three
local,
and
two
descending
interneurons.
Characteristic
dendritic
and
axonal
arborizations
of
the
local
and
descending
neurons
indicate
a
flow
of
auditory
information
from
the
ascending
interneurons
toward
the
lateral
accessory
lobes
and
point
toward
the
relevance
of
this
brain
region
for
cricket
phonotaxis.
Two
aspects
of
auditory
processing
were
studied:
the
tuning
of
interneuron
activity
to
pulse
repetition
rate
and
the
precision
of
pattern
copying.
Whereas
ascending
neurons
exhibited
weak,
low-pass
properties,
local
neurons
showed
both
low-
and
band-pass
properties,
and
descending
neurons
rep-
resented
clear
band-pass
filters.
Accurate
copying
of
single
pulses
was
found
at
all
three
levels
of
the
auditory
pathway.
Animals
were
walking
on
a
trackball,
which
allowed
an
assessment
of
the
effect
that
walking
has
on
auditory
processing.
During
walking,
all
neurons
were
additionally
activated,
and
in
most
neurons,
the
spike
rate
was
correlated
to
walking
velocity.
The
number
of
spikes
elicited
by
a
chirp
increased
with
walking
only
in
ascending
neurons,
whereas
the
peak
instantaneous
spike
rate
of
the
auditory
responses
increased on
all
levels
of
the
processing
pathway.
Extra
spiking
activity
resulted
in
a
somewhat
degraded
copying
of
the
pulse
pattern
in
most
neurons.
interneurons;
pattern
copying
IN
ANIMAL
COMMUNICATION,
the
response
to
conspecifics
is
guided
by
the
recognition
of
their
signals,
which
may
come
in
shape
of
chemicals,
vibrations,
sound,
or
other
modalities.
In
some
frogs,
and
in
birds
and
mammals,
the
recognition
of
acoustic
communication
signals
emerges
from
integrated
pro-
cessing
of
several
song
parameters (frequency,
duration,
and
amplitude)
in
a
complex
network
(Fay
1992;
Viceic
et
al.
2006).
However,
most
acoustically
communicating
insects,
such
as
crickets,
bush-crickets,
grasshoppers,
and
cicadas,
communicate
with
stereotyped
patterns
of
sound
pulses
(Ger-
hardt
and
Huber
2002),
where
the
crucial
feature
for
species
recognition
is
the
temporal
pattern
of
their
calling
songs
(Pollack
and
Hoy
1979).
Because
of
the
simple
temporal
patterns
of
their
auditory
communication
signals,
crickets
have
long
been
used
as
models
for
investigating
the
neural
mecha-
nisms
underlying
recognition
and
orientation
(Hoy
1978;
Hu-
ber
et
al.
1989).
Address
for
reprint
requests
and
other
correspondence:
B.
Hedwig,
Depart-
ment
of
Zoology,
University
of
Cambridge,
Downing
St.,
Cambridge
CB2
3EJ,
UK
(e-mail:
bh202@cam.ac.uk).
Several
methods
have
been
developed
for
investigating
the
tuning
of
phonotaxis
in
walking
crickets,
all
involving
some
kind
of
a
walking
compensator
(Bohm
and
Schildberger
1992;
Doherty
and
Pires
1987;
Schildberger
and
Homer
1988;
Weber
et
al.
1981).
Using
a
trackball
system
with
improved
spatial
and
temporal
resolution,
Hedwig
and
Poulet
(2004)
showed
that
auditory
orientation
of
crickets
emerges
from
simple
reactive
steering
toward
individual
sound
pulses.
This
indi-
cated
that
the
pulse
structure
is
preserved
from
the
sensory
input
to
the
motor
output
during
cricket
phonotaxis.
The
reactive
steering
started
with
latency
as
low
as
55-60
ms,
while
the
recognition
of
the
song
pattern
appeared
to
control
the
gain
of
the
reactive
steering
(Poulet
and
Hedwig
2005).
Whichever
the
mechanism
of
pattern
recognition
underlying
phonotaxis
is,
the
pulse
structure
of
chirps
appears
to
be
forwarded
to
the
walking
motor
network
controlling
the
leg
movements.
This
must
be
reflected
in
the
neural
activity
of
the
crickets'
auditory-to-motor
interface
that
leads
from
the
pro-
cessing
of
the
sensory
inputs
to
the
generation
of
appropriate
motor
commands.
Besides
the
experiments
of
Bohm
and
Schildberger
(1992),
Schildberger
and
Homer
(1988)
,
Schildberger
et
al.
(1988),
and
Staudacher
(1998,
2001),
studies
of
neural
pathways
un-
derlying
cricket
phonotactic
behavior
were
usually
carried
out
in
restrained
animals.
However,
the
activity
of
the
neurons
involved
in
phonotaxis
appeared
to
be
strongly
affected
by
motor
activity-dependent
and
modulatory
influences
(Stau-
dacher
2001;
Staudacher
and
Schildberger
1998).
It
is
therefore
essential
to
perform
neural
recordings
in
behaving
animals
(Fee
2000).
To
understand
the
functioning
of
the
entire
neural
pathway
of
auditory
processing
from
the
sensory
input
to
motor
output,
it
is
also
vital
to
look
at
different
levels
of
processing.
We
developed
an
experimental
set-up
that
allowed
recording
the
activity
of
brain
neurons
in
walking
crickets.
Recording
the
activity
of
the
ascending,
local,
and
descending
neurons
from
the
brain
provided
a
unique
opportunity
to
analyze
the
sensory-
to-motor
pathway
at
three
different
levels
of
information
pro-
cessing.
Focusing
on
the
trends
in
song
pattern
selectivity
and
the
preservation
of
the
pulse
structure
of
chirps
through
these
levels
of
auditory
processing,
we
describe
new
types
of
in-
terneurons
and
point
toward
a
specific
premotor
region
of
the
brain,
which
may
be
involved
in
the
control
of
phonotaxis.
Additionally,
the
high-resolution
measurement
of
walking
be-
havior
enabled
us
to
investigate
the
correlations
between
walk-
ing
and
neuronal
activity
and
the
impact
of
walking
on
auditory
processing.
www.jn.org
0022-3077/11
Copyright
©
2011
the
American
Physiological
Society
2181
2182
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
MATERIALS AND
METHODS
Animals
and
dissection.
Experiments
were
performed
on
adult
female
crickets
(Gryllus
bimaculatus
de
Geer),
from
the
colony
at
the
Department
of
Zoology,
University
of
Cambridge
(Cambridge,
UK),
maintained
on
a
12-h:12-h
light:dark
photo
cycle.
Female
larvae
were
isolated
as
last
instars
and
raised
in
a
separate
room
to
ensure
physical
and
acoustic
isolation
from
males.
For
dissection,
the
crickets
were
positioned
upright
on
a
block
of
Plasticine
by
restraining
all
legs
with
metal
clamps.
The
Plasticine
block
was
fixed
to
a
holder
that
allowed
free
rotation
of
the
animal.
The
head
was
slightly
tilted
backwards,
and
using
a
mixture
of
beeswax
and
resin,
it
was
tightly
glued
into
a
tip
of
a
modified
Eppendorf
tube.
The
anterior
head
capsule
was
opened,
the
antennae
and
the
ocellar
nerves
were
cut,
and
the
ventral
side
of
the
brain
was
exposed.
The
clypeus
and
the
labrum
were
removed,
and
the
mandi-
bles
were
immobilized
with
the
beeswax/resin
mixture.
In
some
animals,
the
contracting
esophagus
was
cut
off
to
stabilize
the
brain.
The
tissue
was
covered
with
insect
saline
(ionic
composition
in
mmol
1
':
140.NaC1,
10.KC1,
4
CaCl
2
,
4
NaHCO
3
,
6
Nall
2
PO
4
).
Next,
the
leg
restraints
were
removed,
and
the
animal
was
posi-
tioned
on
the
trackball
(for
details,
see
Hedwig
and
Poulet
2005).
Care
was
taken
that
all
of
the
legs
were
intact
and
moved
freely.
A
small,
metal
platform
attached
to
a
micromanipulator
was
placed
underneath
the
brain
for
support,
and
a
metal
ring
was
lowered
onto
the
brain
to
stabilize
it.
The
platform
also
served
as
a
reference
electrode
for
intracellular
recordings.
Acoustic
stimulation.
Sound
stimuli
were
generated
with
the
soft-
ware
Cool
Edit
Pro
2000
(Syntrillium
Software,
Phoenix,
AZ)
running
on
a
personal
computer
(PC)
and
presented
by
two
piezoelectric
speakers
through
brass
tubes
with
frontal
openings
of
14
mm,
which
were
positioned
10
mm
away
from
the
crickets'
ears
and
at
45°
in
respect
to
the
animals'
longitudinal
axis.
The
tubes
were
used
to
reduce
possible
diffraction
of
sound
by
the
electrophysiological
setup.
Acoustic
stimuli
had
a
carrier
frequency
of
4.8
kHz
and
an
intensity
of
75-decibel
(dB)
sound
pressure
level
RMS
(root
mean
square)
relative
to
20
µPa.
Sound
amplitude
was
calibrated
with
an
accuracy
of
1
dB
at
the
position
of
the
ears
using
a
Briiel
&
Kjxr
(Nwrum,
Denmark)
measuring
amplifier
(type
2610)
and
a
free-field
micro-
phone
(type
4191).
The
standard
sound
pattern
to
mimic
the
natural
calling
song
consisted
of
six
sound
pulses
grouped
into
a
chirp. Each
pulse
was
21
ms
long
(with
2-ms
rise
and
fall
times).
The
intervals
between
individual
pulses
were
also
21
ms
long,
resulting
in
a
pulse
period
(PP)
of
42
ms.
The
chirp
duration
was
231
ms
and
the
chirp
period
500
ms.
This
corresponded
to
the
syllable
repetition
intervals
that
Thorson
et
al.
(1982)
used
in
phonotactic
experiments.
To
investigate
the
temporal
selectivity
of
neurons,
we
also
used
sound
patterns
with
10
ms
or
98
ms
PPs.
The
three
different
temporal
patterns
are
labeled
PP10, PP42,
and
PP98.
The
pulse
duty
cycle
within
a
chirp
was
50%
for
all
sound
patterns.
With
respect
to
the
direction
of
acoustic
stimulation,
the
terms
"ipsilateral"
and
"con-
tralateral"
refer
to
the
position
of
the
active
speaker
in
relation
to
a
neuron's
cell
body.
Intracellular
recording.
Recordings
were
made
using
microcapil-
laries
pulled
with
a
Flaming/Brown
micropipette
puller
(P-97;
Sutter
Instrument,
Novato,
CA)
from
thick-walled
Borosilicate
capillaries
(Harvard
Apparatus,
Edenbridge,
Kent,
UK;
1
mm
outer
diameter,
0.58
mm
inner
diameter).
The
tip
of
the
microcapillaries
was
filled
with
5%
Lucifer
Yellow
CH
(Sigma-Aldrich,
St.
Louis,
MO),
dis-
solved
in
aqueous
0.5
M
LiC1
and
the
shaft
backfilled
with
a
0.5-M
LiC1
solution
to
produce
electrodes
of
60-100
MS/
resistance.
For
intracellular
staining,
Lucifer
Yellow
was
injected
for
2-10
min
with
a
1-
to
8-nA
hyperpolarizing
current.
Signals
were
recorded
using
a
SEC-05LX
amplifier
(npi
electronic,
Tamm,
Germany)
in
bridge
mode.
Following
injection,
Lucifer
Yellow
was
allowed
to
spread
throughout
the
neuron
for
15-20
min.
Thereafter,
the
brain
was
dissected
and
fixed
in
4%
formaldehyde,
dehydrated
in
an
ethanol
series,
and
cleared
in
methylsalicylate.
Neurons
were
photographed
using
a
digital
single-lens
reflex
camera
(Canon
EOS
350D)
attached
to
a
Zeiss
Axiophot
fluorescence
microscope.
For
graphical
projec-
tions
of
neural
arborizations,
photo-stacks
were
traced
using
Photo-
shop
CS2
(Adobe,
San
Jose,
CA)
and
a
Graphire
4
pen
tablet
(Wacom
Europe
GmbH,
Krefeld,
Germany).
Data
sampling
and
analysis.
All
data
were
sampled
on-line
to
the
hard
disk
of
a
PC
with
Spike2
software
[Cambridge
Electronic
Design
(CED),
Cambridge,
UK]
using
a
CED
1401
plus
data
acquisition
interface
set
to
a
sampling
rate
of
10
kHz/channel.
Data
evaluation
was
carried
out
with
Spike2
and
NeuroLab
software
(Hedwig
and
Knepper
1992;
Knepper
and
Hedwig
1997).
To
present
the
response
characteristics
of
interneurons
to
different
auditory
stimuli,
we
evaluated
the
mean
spike
rates/s
and
the
peak
instantaneous
spike
rates
for
12-15
chirps
for
each
of
the
three
auditory
patterns
tested.
We
analyzed
the
responses
of
one,
two,
or
three
individual
neurons/neuron
type.
In
the
case
of
local
contralateral
brain
neuron
1
(soma
cluster
5)
[B-LC1(5)],
only
six
responses
were
analyzed
for
PP10
and
PP98,
due
to strong
adaptations
(see
RESULTS).
The
means
and
SD
were
calculated
for
each
neuron
separately;
where
more
than
one
neuron
of
the
neuron
type
was
analyzed,
we
present
the
grand
means.
The
instantaneous
spike
rate
was
defined
as
the
inverse
of
the
duration
of
each
interspike
interval.
The
peak
values
of
the
instantaneous
spike
rate
were
measured
for
individual
chirps,
and
their
means
and
the
SD
were
calculated.
The
physiological
characteristics
of
responses
to
PP42
chirps
for
one
neuron
of
each
type
are
presented
in
the
form
of
raster
plots,
peri-stimulus
time
histograms
(PSTHs),
and
changes
in
the
spike
rate
(note
that
here,
the
instantaneous
spike
rate
was
averaged
over
5-ms
intervals).
Where
numbers
of
spikes/pulse
are
given,
n
refers
to
the
number
of
pulses
evaluated.
The
degree
of
synchronization
of
action
potentials
(APs)
with
the
sound
pulses
of
different
auditory
patterns
was
analyzed
as
described
by
Schildberger
(1984).
After
latency
correction,
the
synchronization
coefficient
(SC)
was
determined:
all
spikes
occurring
during
a
pulse
(EPu)
and
those
occurring
during
a
pause
between
pulses
(EPa)
were
counted,
and
the
ratio
(EPu
EPa)/(EPu
+
EPa)
was
calculated.
In
the
case
of
perfect
copying,
SC
=
1,
and
SC
=
0,
if
there
is
no
copying
at
all.
SC
values
were
averaged
over
12-15
chirps
for
each
neuron
[six
in
the
case
of
PP10
and
PP98
patterns
and
neuron
B-LC1(5)].
Note
that
the
SC
value
is
independent
of
a
neuron's
overall
activity
level;
i.e.,
SC
may
be
high,
although
the
neuron
generated
only
few
spikes.
While
the
crickets
were
walking
on
top
of
the
trackball,
an
optical
sensor
recorded
their
left-right
and
forward-backward
movement
components.
From
the
forward-backward
component,
we
calculated
the
crickets'
forward
velocity
and
from
the
left-right
component,
their
lateral
velocity.
For
details
on
evaluation
of
trackball
data,
see
Hedwig
and
Poulet
(2005).
The
relation
between
walking
and
the
spike
rate
of
a
neuron
was
analyzed
by
calculating
Pearson's
corre-
lation
coefficient
between
the
forward-velocity
amplitude
and
the
spike
rate
for
consecutive,
300-ms
intervals
over
a
6-
to
10-s
period.
The
instantaneous
spike
rate
is
presented
in
the
figures.
In
text,
we
regarded
crickets
as
walking,
even
if
walking
occasionally
stopped
for
intervals
shorter
than
500
ms;
intervals
exceeding
500
ms
were
judged
as
episodes
of
standing.
Terminology.
Terminology
for
insect
neurons
is
not
uniform,
and
in
the
case
of
cricket
brain
neurons,
partially
redundant
acronyms
have
been
used.
We
therefore
follow
the
labeling
system
of
Hedwig
(1986).
For
instance,
"B"
stands
for
brain,
"TH1"
stands
for
1st
thoracic
ganglion,
"A"
for
ascending,
"L"
for
local,
and
"D"
for
a
descending
axon.
The
following
letters,
"I"
or
"C",
signify
an
ipsi-
or
contralateral
position
of
the
output
structures
with
regard
to
the
cell
body
in
the
ascending
and
local
neurons,
whereas
in
the
descending
neurons,
the
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
2183
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
Table
1.
Response
parameters
of
ascending,
local,
and
descending
auditory
intemeurons
of
Gryllus
bimaculatus
to
three
different
sound
patterns,
PP10,
PP42,
and
PP98
Mean
Spike
Rate
(Hz)
Peak
Spike
Rate
(Hz)
Synchronization
Coefficient
(SC)
PP42
PP42
PP42
PP10
PP98
PP10
PP98
PP10
w
PP98
TH1-AClt
51.5
70.1
86.1
79.2
134
201
228
220
0.29
0.87
0.79
0.95
TH1-AC2t
33.2
42.0
69.7
47.8
86
123
167
133
0.34
0.57
0.51 0.71
B-LI1*
21.3
35.5
27.7
32.7
56
76
99
71
0.23
0.54
0.38
0.86
B-LC1t
13.6
32.9
29.9
23.7
83
285
319
244
0.36
0.94 0.74
0.91
B-LC2*
40.0
113.4
81.0
31.8
237
257
221
240
0.07
0.34
0.37
0.89
B-DI1(5)
t
3.0
37.2
22.5
13.1
-
175
102
56
0.24
0.93
0.89
0.58
B-DC1(5)f
12.3
32.5
28.6
13.9
174
156
213
147
0.21
0.59
0.72
0.49
Values
represent
means
(B-LI1,
B-LC2)
and
grand
means
(other
neurons).
Number
of
neurons
analyzed/each
neuron
type:
*n
=
1,
tn
=
2,
fn
=
3.
For
each
individual
cell,
the
responses
elicited
by
6-15
chirps
were
analyzed
for
each
sound
pattern.
s,
standing;
w,
walking;
PP,
pulse
period;
TH1,
1st
thoracic
ganglion;
AC,
ascending
contralateral;
B-LI1,
local
ipsilateral
brain
neuron
1;
B-LC,
local
contralateral
brain
neuron
;
B-DI1(5),
descending
ipsilateral
brain
neuron
1
(soma
cluster
5);
B-DC1,
descending
contralateral
brain
neuron
1(5).
letters
I
or
C
signify
the
side
of
the
brain
(with
regard
to
the
cell
body)
through
which
the
axon
is
descending
toward
the
suboesophageal
ganglion.
Terms
ipsilateral
and
contralateral
in
figures
are
also
used
with
respect
to
the
cell
body.
Following
the
guidelines
on
neural
taxonomy
(Rowell
1989),
we
here
replace
the
established
acronyms
for
ascending
intemeurons
1
and
2
(AN1
and
AN2)
(Wohlers
and
Huber
1982)
by
TH1-AC1
and
TH1-AC2.
The
acronyms
PABN2
(Boyan
1980)
and
BNC1c
(Schil-
dberger
1984),
which
were
used
for
a
local
ipsilateral
brain
neuron,
were
replaced
by
B-LI1.
Following
the
scheme,
two
descending
intemeurons,
which
had
been
labeled
BDNi5
and
BDNc5
(Staudacher
1998),
as
belonging
to
the
cluster
5
of
descending
neurons,
were
assigned
B-DI1(5)
and
B-DC1(5).
In
this
case,
the
addition
of
a
parenthetical
number
indicates
the
soma
cluster
in
the
brain.
The
phrase
spontaneous
activity
is
used
to
describe
neural
activity
in
the
absence
of
sound
stimuli,
and
spontaneous
walking
refers
to
walking
in
the
absence
of
sound
stimuli.
RESULTS
When
placed
on
top
of
the
trackball,
most
crickets
started
walking
spontaneously.
The
animals
then
performed
episodes
of
walking
and
standing,
during
which
different
song
patterns
were
presented.
The
forward
velocity
ranged
in
amplitude
from
1
to
12
cm
s
-1
.
Lateral
velocity
ranged
from
1
to
5
cm/s
-1
.
We
present
physiological
and
morphological
data
from
seven
au-
ditory
intemeurons.
Intracellular
recordings
and
staining
were
obtained
in
the
proto-
and
deutocerebrum
and
lasted
up
to
30
min.
In
the
case
of
the
ascending
neurons,
only
a
single
TH1-AC1
and
a
single
TH1-AC2
are
present
on
each
side
of
a
cricket
(Hennig
1988;
Wohlers
and
Huber
1982).
The
data
from
Schildberger
(1984)
and
our
own
data
suggest
that
there
is
only
one
B-LI1
neuron
on
each
side
of
the
cricket
brain.
B-LC2,
however,
may
label
a
small
group
of
neurons
of
similar
morphology
and
physiology.
B-DI1(5)
and
B-DC1(5)
repre-
sent
clusters
of
several
neurons,
based
on
our
morphological
and
physiological
data
and
anatomical
data
presented
by
Stau-
dacher
(1998).
All
physiological
response
parameters
and
data
on
spiking
activity
during
spontaneous
walking
are
summa-
rized
in
Tables
1
and
2.
The
experimental
approach
allowed
stable
recording
of
brain
neurons
in
standing
and
walking
crickets.
Phonotactic
steering
toward
the
species-specific
song
occurred
infrequently.
Some
females
(ca.
15%)
readily
steered,
at
least
during
the
first
hour,
whereas
on
several
occasions,
the
animals
started
steering
toward
the
sound
source
several
hours
after
the
start
of
the
brain
recordings.
Preliminary
experiments
demonstrated
that
microinjections
of
neuroactive
substances
into
the
brain
may
be
a
suitable
method
to
elicit
phonotaxis
(M.
Zorovic'
and
B.
Hedwig,
unpublished
observations).
Arborization
patterns
of
neurons.
The
structures
of
ascend-
ing,
local,
and
descending
auditory
neurons
revealed
two
bilateral
brain
regions
that
were
involved
in
processing
the
auditory
stimuli
(Fig.
1).
One
is
the
ventral
protocerebrum,
lateral
to
the
a-lobe,
where
the
main
axonal
projection
fields
of
both
ascending
neurons,
TH1-AC1
and
TH1-AC2,
were
lo-
cated.
Dendrites
of
the
local
interneuron
B-LI1
overlapped
with
these
projection
areas
in
the
ventral
protocerebrum,
whereas
in
the
posterior
protocerebrum,
its
axonal
arboriza-
tions
overlapped
with
the
dendrites
of
two
bilateral
local
Table
2.
Spike
rates
of
ascending,
local,
and
descending
auditory
intemeurons
of
G.
bimaculatus
and
its
correlation
to
walking
velocity
during
spontaneous
walking
Spike
Rate
during
Spontaneous
Walking
(Hz)
Correlation
between
Spiking
Activity
and
Forward
Walking
Velocity
Correlation
between
Spiking
Activity
and
Lateral
Walking
Velocity
(C,
contralateral;
I,
ipsilateral)
TH1-AClt
20-100
Yes;
r=
0.80,
P
<
0.001
Yes
(C);
r
=
0.71,
P
<
0.001
TH1-AC2t
20-60
Yes;r=
0.6,
P<
0.01
Yes
(C);
r
=
0.47,
P
<
0.05
B-LI1*
10-120
Yes;
r=
0.69,
P
<
0.01
Yes
(I);
r
=
0.53,
P
<
0.05
B-LC1t
20-135
No;
r=
0.13,
P
>
0.05
Yes
(C);
r
=
0.64,
P
<
0.01
B-LC2*
20-85
No;
r=
0.33,
P
>
0.05
Yes
(C);
r
=
0.5,
P
<
0.05
B-DI1(5)t
5-40
Yes;
r=
0.55,
P
<
0.001
Yes
(I);
r
=
0.55,
P
<
0.01
B-DC1(5)f
30-200
No;
r
=
0.20,
P
>
0.05
No;
r=
0.22,
P
>
0.05
Number
of
cells
analyzed/cell
type:
*n
=
1,
tn
=
2,
fn
=
3.
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
2184
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
TH1-AC1
TH
1-AC2
B-LI1
Fig.
1.
Structures
of
two
ascending,
three
local,
and
two
descending
auditory
neurons
in
the
brain
of
Gryllus
bimaculatus.
First,
thoracic
ganglion-as-
cending
contralateral
1
(11-11-AC1)
and
TH1-AC2
project
from
the
first
thoracic
ganglion
to
the
brain;
only
their
axonal
arborizations
are
shown.
Local
ipsilateral
brain
neuron
1
(B-LI1)
projects
from
the
axonal
arborizations
of
the
ascending
intemeurons
toward
the
lateral
accessory
lobes
(LALs).
The
local
intemeurons,
local
contralateral
brain
neuron
1
(B-
LC1)
and
B-LC2,
connect
the
left
and
right
LALs.
The
dendrites
of
the
descending
intemeurons,
de-
scending
ipsilateral
brain
neuron
1
(soma
cluster
5)
[B-DI1(5)]
and
descending
contralateral
brain
neu-
ron
1(5)
[B-DC1(5)],
overlap
with
the
LAL.
The
LAL
regions
are
indicated
by
the
outlines
in
B-LC2.
)
-
7
B-LC1
B-LC2
)
B-DI1(5)
B-DC1
(5)
interneurons,
B-LC1
and
B-LC2.
Branches
of
these
two
bilat-
eral
neurons
occupied
the
area
in
the
posterior
median
proto-
cerebrum,
on
the
border
to
deutocerebrum,
which
was
previ-
ously
described
by
Schildberger
(1984)
as
the
branching
area
of
some
local
brain
neurons.
It
is
located
posterolaterally
to
the
central
body
complex
and
lies
on
the
anterior
border
of
the
antennal
lobes
(Fig.
1,
outlines
given
for
B-LC2)
in
the
area
of
the
lateral
accessory
lobes
(LALs)
(Sakura
et
al.
2007).
The
dendrites
of
both
descending
interneurons
were
also
located
in
the
LALs.
B-DI1(5)
had
an
ipsilaterally
descending
axon,
whereas
the
axon
of
B-DC1(5)
crossed
the
midline
before
entering
the
contralateral
connective.
The
ascending
neurons
TH1-AC1
and
TH1-AC2
are
two
first
order
interneurons
forwarding
auditory
information
from
the
first
thoracic
ganglion
to
the
brain
(Wohlers
and
Huber
1982).
In
TH1-AC2,
the
axon
split
in
the
posterior
protocer-
ebrum
and
the
main
branch
projected
anteriorly,
whereas
the
largest
side
branch
projected
along
the
dorsal
surface
of
the
lateral
neuropil
toward
the
optic
sta
k,
where
it
terminated
after
a
sharp
anteriomedial
loop
(Boyan
and
Williams
1982).
Data
were
derived
from
three
recordings
of
TH1-AC1
and
four
recordings
of
TH1-AC2.
We
recorded
and
stained
one
ipsilateral
and
two
bilateral
local
interneurons.
The
ipsilateral
neuron
B-LI1
corresponded
to
PABN2
described
by
Boyan
(1980)
and
BNC1c
described
by
Schildberger
(1984).
The
structure
of
B-LI1
with
anterior
dendrites
and
posterior
axonal
arborizations
indicated
that
it
could
forward
information
from
the
terminals
of
TH1-AC1
and
TH1-AC2
to
the
ipsilateral LAL.
Large
unitary
5
mV
excit-
atory
postsynaptic
potentials
(EPSPs)
and
spikes
were
recorded
from
the
dendrites
of
this
neuron.
The
bilateral
local
neurons
B-LC1
and
B-LC2
had
a
char-
acteristic
morphology
with
arborizations
restricted
to
the
LALs
on
either
side
of
the
brain,
connected
by
an
axon
that
crossed
the
midline
The
branches
on
the
ipsilateral
side
of
the
soma
had
a
smooth
appearance,
whereas
the
contralateral
ones
had
a
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
2185
varicose
structure.
Correspondingly,
both
EPSPs
and
spikes
were
recorded
from
the
ipsilateral
branches
and
only
spikes
from
contralateral
branches.
Although
B-LC1
and
B-LC2
shared
a
general
appearance,
they
differed
in
the
location
of
the
cell
body
and
the
shape
of
their
branching
patterns.
We
recorded
from
and
stained
two
cells
of
each
neuron
type.
The
cell
body
of
B-LC1
was
located
ventrally
in
the
median
deutocerebrum.
The
primary
neurite
looped
dorsolaterally
and
on
approaching
the
middle
of
the
ipsilateral
deutocerebrum,
turned
toward
the
LAL
region,
where
it
branched
extensively.
The
axon
crossed
the
midline
rising
toward
the
dorsal
surface
and
then
turned
ventrally
before
it
terminated
with
extensive
arborizations
in
the
contralateral
LAL.
The
cell
body
of
B-LC2
was
located
ventrally
in
the
middle
of
the
ipsilateral
posterior
protocerebrum.
The
axon
connecting
the
bilateral arborizations
was
located
dorsally
with
regard
to
the
cell
body,
whereas
the
axonal
and
dendritic
regions
lie
closer
to
the
ventral
surface
of
the
brain.
Two
types
of
descending
neurons
were
stained,
B-DI1(5)
and
B-DC1(5).
Recordings
were
obtained
from
three
B-DI1(5)
and
eight
B-DC1(5)
neurons.
B-DI1(5)
neurons
belong
to
a
ventral
cluster
i5
that
contains
between
4
and
19
perikarya
(Staudacher
1998).
The
primary
neurite
projected
mediodor-
sally
and
turned
posteriorly
toward
the
dendritic
arborizations,
which
were
located
dorsally,
posteriorly,
and
laterally
to
the
central
complex.
The
axon
descended
through
the
dorsal
deuto-
and
tritocerebrum
and
then
in
the
middle
of
the
connective
(Fig.
1).
Recordings
were
obtained
from
the
dendrites
and
revealed
spikes
and
synaptic
activity.
B-DC1(5)
neurons
belong
to
a
morphologically
homoge-
nous
group
of
four
to
seven
neurons
with
cell
bodies
in
the
posterior
part
of
the
ventromedial
protocerebrum
in
soma
cluster
c5
(Staudacher
1998).
Main
morphological
character-
istics
of
B-DC1(5)
neurons
were
bilateral
arborizations
over-
lapping
in
the
LALs
with
those
of
the
local
bilateral
interneu-
rons
B-LI1
and
B-LC1.
Dendrites
occurred
in
the
medial
and
dorsal
protocerebrum.
The
contralateral
axonal
arborizations
were
dorsal
in
the
posterior
proto-
and
anterior
deutocerebrum.
The
axon
descended
from
the
brain
contralaterally
through
the
median
part
of
the
connective.
We
encountered
some
variabil-
ity
in
physiological
properties.
All
eight
neurons
received
excitatory
inputs
upon
acoustic
stimulation
from
both
speakers.
Whereas
in
two
experiments,
the
neurons
responded
with
the
same
level
of
excitation
to
ipsi-
and
contralateral
acoustic
stimulation,
in
three
experiments,
the
neurons
received
stron-
ger
auditory
inputs
from
the
ipsilateral
side
and
in
three
other
experiments,
from
the
contralateral
side.
Physiological
data
shown
are
based
on
four
experiments,
including
neurons
from
all
three
groups.
Selectivity
for
temporal
patterns.
In
most
neurons,
the
se-
lectivity
for
song
patterns
PP10,
PP42,
and
PP98
was
tested
only
in
standing
crickets,
and
we
analyzed
the
responses
during
standing
to
compare
across
all
neuron
types
(Fig.
2).
In
the
case
of
TH1-AC1
and
TH1-AC2,
the
song
patterns
were
presented
from
the
contralateral
side
(note
that
in
the
ascending
auditory
neurons,
the
dendrites
are
opposite
the
cell
bodies),
and
in
the
case
of
the
local
and
the
descending
neurons,
the
song
patterns
were
played
from
the
ipsilateral
side.
The
only
neurons
to
show
any
spontaneous
activity
during
standing
were
B-LC2
neurons;
their
firing
rate
varied
between
5
and
70
Hz.
Chirps
of
the
song
pattern
PP42
elicited
responses
with
mean
spike
rates
between
30
and
115
Hz,
on
average,
across
all
neuron
types
(Fig.
2,
histograms
left),
with
the
highest
response
recorded
from
the
local
cells
B-LC2
(Table
1;
Fig.
2E).
With
the
exception
of
B-LC2,
overall,
the
mean
spike
rate
of
the
auditory
responses
decreased
for
attractive
as
well
as
unattract-
ive
patterns
along
the
processing
pathway
from
the
ascending
to
the
descending
neurons
(Table
1).
The
peak
instantaneous
spike
rates,
however,
revealed
no
consistent
trend.
Ascending
neurons
exhibited
weak
low-pass
properties,
local
neurons
showed
either
low-
or
band-pass
properties,
and
descending
neurons
represented
clear
band-pass
filtering.
This
increase
in
selectivity
along
the
processing
pathway
was
reflected
in
the
mean
spike
rate
but
less
clear
for
the
peak
instantaneous
spike
rates.
When
the
latter
was
considered
(Fig.
2,
histograms
middle),
only
one
decending
neuron,
B-DI1(5),
exhibited
band-pass
selectivity
for
the
behaviorally
attractive
pattern
PP42
(Fig.
2F).
Note
that
no
value
is
given
at
PP10
for
neurons
B-DI1(5),
because
the
average
number
of
spikes/chirp
was
<1.
The
spike
rates
of
responses
to
pattern
PP42
across
all
neurons
ranged
from
75
to
285
Hz.
B-DC1(5)
exhibited
strong
response
decrement
toward
behaviorally
unattractive
chirps
of
song
patterns
PP10
and
PP98.
The
neuron
stopped
spiking
after
the
first
(for
PP10)
or
the
third
(for
PP98)
chirp
of
a
test
sequence.
The
response
traces,
the
number
of
spikes/chirp,
and
the
peak
instantaneous
spike
rates
(Fig.
2G)
therefore
reflect
only
the
responses
at
the
beginning
of
the
chirp
sequences.
B-DC1(5)
thus
demonstrated
band-pass
selectivity,
not
only
in
the
num-
ber
of
spikes/chirp
but
also
through
a
strong
response
decre-
ment
to
patterns
PP10
and
PP98.
Copying
of
the
pulse
pattern
in
the
auditory-to-motor
pathway.
The
level
of
pulse
pattern
copying
via
the
coupling
of
APs
to
individual
pulses
was
evaluated
for
different
temporal
patterns
by
calculating
the
SC
(Table
1).
Across
all
neuron
types,
the
SC
values
were
lower
for
pattern
PP10
than
for
PP42
or
PP98
(Fig.
2,
histograms
right).
The
SC
was
generally
high
within
the
sensory-to-motor
pathway,
reflecting
good
coupling
of
neuronal
activity
to
the
pulse
structure
of
the
chirp
for
many
neurons.
In
the
descending
neurons,
the
SC
values
were
higher
for
the
attractive
pattern
PP42
than
for
patterns
PP10
and
PP98.
The
highest
SC
values
overall,
0.94
and
0.93,
were
calcu-
lated
at
PP42
for
local
neurons
B-LC1
and
descending
neurons
B-DI1(5),
respectively.
The
only
neurons
with
a
SC
value
below
0.5
for
PP42
were
the
local
neurons
B-LC2
with
SC
0.34.
Figure
3,
left,
shows
the
coupling
of
spikes
to
individual
pulses
and
the
modulation
of
the
instantaneous
spike
rate
for
pattern
PP42
across
the
ascending,
local,
and
descending
brain
neurons
during
standing.
In
TH1-AC1,
over
the
course
of
one
chirp,
the
number
of
spikes/pulse
decreased
from
3.8
±
0.8
(n
=
20)
for
the
first
pulse
to
2.6
±
0.6
(n
=
20)
for
the
last
pulse.
The
firing
pattern
represented
each
PP42
pulse
as
a
distinctive
peak
in
the
spike
rate
of
180-220
Hz.
In
the
sound-pulse
intervals,
the
neuron
was
mostly
silent
(note
the
high
SC
value,
Table
1);
however,
the
spike
rate
still
remained
at
50
Hz,
as
the
spike
rate
value
was
determined
by
the
last
spike
of
one
burst
and
the
first
spike
of
the
next
burst.
The
PSTH
and
the
raster
plot
for
TH1-AC2
show
that
the
first
pulse
of
each
chirp
elicited
the
most
accurate
response
when
timing
of
spikes
is
considered
(Fig.
3B,
left).
Spiking
to
subsequent
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
Sync.
N=2
n=15
Coefficient
2186
A
TI-11-AC1
PROCESSING
OF
100
80
AUDITORY
Mean
Spike
N=
2
n=15
PATTERNS
Rate
IN
THE
(Hz)
250
200
Peak
N=
2
n=15
CRICKET
Spike
BRAIN
Rate
(Hz)
1,0
'
0,
8
60
Y
150
0,6
40
100
0,
4
AWL_
20
50
0,2
0
0
0
'
N=
2
n=12
200
N=
2
n=12
1,0
N=
2
n=12
B
75
TH1-AC2
60
160
0,
45
120
0,6
30
80
0,4
_110111-
15
40
0,
0
0
0
C
40
-
N=
1
N=
1
1,0-
N=
1
B-LI1
30-
n=15
100-
80-
n=15
I
0,8-
n=15
I
20-
I
60-
0,6
-
I
40-
0,4
-
10-
20-
0,2-
I
0
0
0
D
50
N=
2
N=2
N=2
B-LC1
40
n=12
300
n=12
1,0
n=12
30
240
0,8
180
0,6
20
120
0,4
10
60
0,2
0
N=
1
0
0
N=
1
1,0-
N=
E
B-LC2
125-
100-
n=15
300
-
240
-
n=15
0,8-
n=15
75-
180
-
0,6-
50-
120
-
0,4-
25-
I
60
-
0,2-
0
0
0
F
50
N=
2
N=2
N=2
B-DI1(5)
n=12
200
n=12
1,0
n=12
40
160
0,8
30
120
0,6
20
80
0,4
10
40
0,2
0
0
0
G
B-DC1(5)
40
N=
3
n=6-12
300
240
N=
3
n=6-12
1,0
0,8
N=3
n=12
30
180
0,6
20
120
0,4
1111111111111_
10
60
0,2
PP10
PP42
PP98
0
0 0
PP10
PP42
PP98
PP10
PP42
PP98
PP10
PP42
PP98
200
ms
Fig.
2.
Song
pattern
selectivity
and
copying
accuracy
in
ascending,
local,
and
descending
brain
neurons.
(A—G,
left)
Typical
responses
of
brain
neurons
to
chirps
of
different
temporal
patterns
during
standing;
the
song
patterns
are
shown
below
the
neuronal
activity.
Vertical
scale
bars:
10
mV
(A—D,
F,
G)
and
50
mV
(E).
Histograms
left:
mean
spike
rate
of
responses
elicited
by
chirps
of
different
temporal
patterns.
Histograms
middle:
peak
instantaneous
spike
rates
in
response
to
different
temporal
patterns.
Histograms
right:
synchronization
coefficient
for
responses
elicited
by
chirps
of
different
temporal
patterns.
(A—G)
The
gray-shaded
histogram
bars
indicate
the
values
during
standing,
and
the
white
bars
show
the
values
during
walking.
(A,
B,
D,
F,
G)
Symbols
represent
the
means
of
individual
cells
(±SD);
the
histogram
bars
represent
the
grand
means.
PP,
pulse
period;
N,
number
of
cells;
n,
number
of
chirps/cell.
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
In
0
100
200
‘.•
•••,
i
t
:•
.
1
TI
standing
-100
0
100
200
C
B-LI1
20-
Sweeps
10-
0
16
AP/bin
8-
0
-100
0
100
200
-
J
tfiih
,1
-
11
Far
A
TH
1-AC1
20
Sweeps
10:
2
0
-
0
AP/bin
10
0
-1
1
00
B
TH
1-AC2
20
Sweeps
10-
0
AP/bin
10
1
15
5
walking
.
.
X1••••fiV
.
if;
ir
.14.
WA'
.0•;:•'.
'ile
}
:
of•i
200
100
0
20
10
0
20
10
0
--
-100
0
300
100
200
i
200
AP/s
100
(Hz)
0
300
Time
(mS)
Time
(ms)
20
10
0
70
[5
3
15
10
0
0
5
300
Time
(mS)
-100
1
1
,
4
-
i[FRT
0
100
200
300
Time
(ms)
70
35
(H)
AP/s
D
B-LC1
20
Sweeps
10
38
-
AP/bin
15
0
-100
.
3
3:
;
100
200
300
20
-
10-
0
120
30
60
15-
0
0
-
c
Time
(ms)
-100
0
AP/bin
100
200
300
E
B-LC2
20
Sweeps
10
0
20
10
0
-100
20
10
150
20
75
10
O
0
Time
(ms)
-100
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
2187
20-
"•":
:'
•'••:•:•'
•••.;;'''••:,...:'
.
..y.:.
•.''..,'.....:.....::•:‘•::•'.
io
.
'::.
:.
:
.'1.
:
••••••).
'
1.•
:
1
..
....
:
.
...
..
16
8
40
AP/s
oVitilltit
far
di
-FIT
40
fa
Ii
-Er k
li"'Pritli
o
(Hz)
-100
0
100
200
300
Time
(ms)
80
40
0
300
Time
(ms)
.
:`‘.
[
120
AP/s
I
ft -,11k
kA7l
1
^
,
"^-
,
-
Viz
)
100
200
300
0
Time
(ms)
••••::.
Ayi
•:
'
r
••
150
AP/s
r4
75
(Hz)
100
200
300
Time
(ms)
F
B-DI1(5)
20
20
AP/bin
10
-
Sweeps
10
-
20
0
0
-100
•I
I
4
;•
100
200
300
Time
(ms)
-60
10
120
28
O
0
10
'100
0
100
200
300
Time
(ms)
.
••
. .
j120
60
(Hz)
0
AP/s
G
B-DC1(5)
20
20
•••
.•
'
10-
10
Sweeps
0
1
35
70
12
0
6-
12
70
[
AP/bin
6
35
AP/s
,
,r
0
(Hz)
I
s
m
FRE
0
0
0
-100
100
200
300
Time
(ms)
-100
100
200
300
Time
(ms)
Sound
Fig.
3.
(A-G)
Activity
patterns
of
ascending,
local,
and
descending
brain
neurons
upon
presentation
of
PP42,
presented
in
the
form
of
PSTHs,
raster
plots,
and
the
averaged
instantaneous
spike
rates
over
20
chirps
during
standing
(left)
and
walking
(right).
AP,
action
potential.
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
60
N
50
6
40
cn
30
a
°
20
co
(1)
10
0
cc
2188
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
80
(I)
70
I I
4-F
'
5
03
Fig.
4.
Response
latency
to
PP42
chirps
during
standing
in
ascending,
local,
and
descending
brain
neurons.
Responses
to
20
chirps
were
analyzed
for
each
individual
cell.
Number
of
cells
analyzed/cell
type:
*n
=
1,
tn
=
2,
fn
=
3.
Where
n
=
2
or
3,
the
histogram
bars
represent
the
grand
means.
pulses
appeared
more
scattered.
Since
the
neuron
responded
only
with
1.1
±
0.3
spikes/pulse
(n
=
120),
the
spike
intervals
were
rather
long,
and
the
sound
pulses
were
not
copied
by
the
instantaneous
spike
rate.
The
PSTH
of
B-LI1
(Fig.
3C,
left)
shows
weak
coupling
of
spikes
to
pulses;
the
high
scatter
in
the
raster
plot
reflects
the
relatively
low
SC
value
of
0.54.
The
spike
rate
reached
a
maximum
of
40
Hz
at
the
first
pulse;
however,
it
was
not
modulated
to
follow
the
pulse
pattern.
Neurons
B-LC1
fired
an
average
of
2.6
spikes
in
response
to
each
of
the
first
three
pulses,
whereas
the
last
pulse
elicited,
on
average,
only
1.1
spikes
(Fig.
3D,
left).
The
pulses
were
copied
by
the
timing
of
spikes
and
the
modulation
of
their
firing
rate.
In
response
to
pulses,
the
spike
rate
reached
up
to
220
Hz
and
decreased
to
ca.
25
Hz
between
pulses.
An
interesting
response
characteristic
that
sets
B-LC2
apart
from
all
of
the
other
interneuron
types
described
here
is
a
considerably
weaker
response
to
the
first
pulse
of
a
chirp
than
to
any
of
the
subsequent
pulses
(Fig.
3E,
left).
Whereas
the
first
pulse
of
each
chirp
elicited
1.7
(n
=
20)
spikes
on
average,
pulses
two
to
five
elicited
3.9-4.6
(n
=
80)
spikes,
and
then
the
response
decreased
again
to
2.1
spikes
at
the
last
pulse.
Although
the
ras-
ter
plot
shows
considerable
scatter,
the
sound
pulses
were
copied
by
the
timing
of
spikes
and
the
modulation
of
the
spike
rate.
The
highest
spike
rate
of
148
Hz
occurred
in
response
to
the
second
sound
pulse,
whereas
the
average
spike
rate
of
the
response
to
the
first
pulse
was
only
40
Hz.
Between
the
responses
to
individual
pulses,
the
spike
rate
decreased
by
15-35%.
Note
that
the
song
patterns
were
all
presented
during
hyperpolarization
by
4.5
nA,
which
removed
any
background
spike
activity
(see
Fig.
3E).
Another
interesting
feature
of
this
neuron
was
the
long
response
latency
of
72.2
ms,
on
average,
as
measured
for
PP42
during
standing.
This
was
a
consequence
of
the
neuron
failing
to
respond
to
the
first
pulse
in
ca.
50%
of
the
chirps.
Overall,
the
latency
of
the
responses
increased
along
the
processing
pathway,
from
ca.
20-24
ms
in
the
ascending
to
45-50
ms
in
the
descending
neurons
(Fig.
4).
The
spikes
of
B-DI1(5)
were
coupled
to
pulses
(Fig.
3F,
left);
however,
after
the
forth
pulse,
the
response
became
weaker.
The
raster
plot
shows
very
little
scatter;
the
individual
pulses
were
copied
by
the
modulation
of
the
spike
rate
with
peaks
up
to
125
Hz.
The
second
pulse
elicited
the
strongest
response.
Between
individ-
ual
peaks,
the
spike
rate
decreased
to
ca.
30
Hz.
The
raster
plot
and
the
PSTH
of
B-DC1(5)
(Fig.
3G,
left)
revealed
that
the
spikes
and
the
instantaneous
spike
rate
were
only
weakly
coupled
to
pulses.
Neuron
activity
and
walking.
All
neurons
were
activated
during
walking,
even
when
no
sound
was
presented
(Table
2;
Fig.
5,
A—G);
spike
frequency
during
walking
ranged
from
5
to
200
Hz
across
all
neuron
types.
The
effects
of
walking
on
neurons'
responses
to
sound
stimuli
were
tested
for
PP42
chirps
(Table
1;
Fig.
2;
Fig.
3,
right).
The
number
of
spikes
elicited
by
a
chirp
during
walking
increased
compared
with
standing
only
in
the
ascending
neu-
rons,
whereas
there
was
a
10-65%
decrease
in
all
other
neurons.
The
peak
spike
rates,
on
the
other
hand,
increased
during
walking
in
all
neurons
except
B-LC2
and
B-DI1(5).
In
the
latter,
the
spike
rate
decreased
by
as
much
as
42%.
The
pulse
pattern
copying
properties
were
also
affected
by
walking:
in
most
cases,
the
SC
values
decreased
during
walking
by
4%
[B-DI1(5)]
to
30%
(B-LI1).
In
B-LC2,
there
was
a
small
increase
of
9%
in
SC,
whereas
in
B-DC1(5),
the
SC
value
increased
during
walking
by
22%.
During
spontaneous
walking,
both
ascending
neurons
showed
a
significant
correlation
between
spiking
activity
and
the
forward
walking
and
the
lateral
velocity
to
the
contralateral
side
(note,
again,
that
in
the
ascending
auditory
neurons,
the
dendrites
are
opposite
the
cell
bodies)
(Table
2).
In
TH1-AC1,
the
maximum
instantaneous
spike
rate
in
response
to
PP42
chirps
reached
200
Hz,
and
the
majority
of
spikes
fired
during
the
sound
stimulus
were
coupled
to
pulses
(Fig.
3A,
right;
Fig.
5A).
The
spiking
activity
of
TH1-AC2
copied
the
pulse
pattern
of
PP42
chirps
by
both
the
timing
of
spikes
(PSTH)
and
the
modulation
of
the
spike
rate
(Fig.
3B,
right).
The
spike
rate
reached
70
Hz
in
response
to
the
first
and
45
Hz
in
response
to
the
last
pulse.
Between
individual pulses,
the
spike
rate
de-
creased
to
ca.
30
Hz.
The
scatter
plot
shows
that
the
firing
pattern
was
more
dispersed
than
during
standing;
indeed,
during
walking,
the
SC
value
decreased.
During
recording
from
B-LI1,
the
cricket
was
very
active
during
periods
of
walking;
attempted
—1
jumps
and
running
with
a
forward
velocity
of
up
to
12
cm
s
resulted
in
an
unstable
recording
(Fig.
5C).
The
spiking
activity
was
correlated
to
the
forward
walking
velocity
and
the
lateral
velocity
to
the
ipsi-
lateral
side
(Table
2).
The
spiking
frequency
of
B-LI1
rose
to
ca.
50
Hz
in
response
to
PP42
chirps,
with
peaks
reaching
120
Hz.
Although
some
unitary
EPSPs
still
occurred,
the
neuron
now
mostly
generated
spikes.
The
spike
rate
of
B-LI1
in
response
to
PP42
chirps
was
modulated
and
weakly
copied
the
pulse
pattern;
the
firing
pattern
represented
each
pulse
as
a
peak
in
the
spike
rate
of
45-75
Hz
(Fig.
3C,
right).
During
the
interpulse
intervals,
the
spike
rate
decreased
to
about
35
Hz.
During
spontaneous
walking,
the
B-LC1
neurons
were
acti-
vated
with
a
spike
frequency
of
around
20
Hz,
with
occasional
Fig.
5.
(A—G,
upper)
Intracellular
recordings
of
ascending,
local,
and
descending
brain
neurons
during
standing
and
spontaneous
walking.
(A—G,
bottom)
Activity
of
ascending,
local,
and
descending
brain
neurons
upon
presentation
of
PP42
during
standing
and
walking.
(D,
bottom)
Arrows
indicate
a
rapid
turn-away
from
the
sound
source.
(F,
bottom)
Arrow
indicates
a
change
in
the
activity
pattern
when
the
animal
transiently
stopped
walking;
arrowheads
indicate
turning
movements.
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
120
mV
B
TH1-AC2
Spike
60
rate
30-
(Hz)
0
Lateral
1
steering
velocity
(cm
s
-1
)
1
Forward
1
velocity
0
(cm
s
-1
)
III
I40
mV
I
100
-
50-
5
0
5
12
0-
ipsilateral
contralateral
forward
forward
backward
ipstrateral
contralateral
0
0.5
1
1.5
Time
(s)
0
Sound
0.5
1
1.5
Time
(s)
B-LI1
Spike
120
rate
60
-
(Hz)
Lateral
4
steering
velocity
0
(
ere
e
-1)
-2
Forward
8
velocity
(cm
s
-1
)
0-
0
Sound
ipstrateral
contralateral
forward
110
mV
4
200
100
0
2-
0
2
Spike
rate
(Hz)
Lateral
steering
velocity
(cm
e)
Forward
8
velocity
(cm
s
-1
)
0
IpsIlateral
contralateral
forward
D
B-LC1
vt.
Spike
150
rate
75-
(Hz)
Lateral
1
steering
velociy
(cm
a
1
)
-1
Forward
1
velociy
(cm
s
-1
)
0
0
110
mV
ipsilateral
conbalateral
forward
3
4
Time
(s)
Spike
200
rata
100
(Hz)
0
Lateral
3
steering
n
velocity
(cm
a
l
)
3
Forward
3
velocity
0
(ana
l
)
0
ipsil
tm
ate
m
rL
i
forward
backward
Time
(s)
Sound
B-LC1
Spike
120
rate
60
(Hz)
Lateral
2
steering
0
ipsilateral
velociy
contralateral
(cm
s
-1
)
2
f
Forward
2
Fo
forward
velociy
0
ard
(cm
s'l)
.2
barkw
0
0.5
G
B-DC1(5)
10
mi/1
1.5
2
2.5
Time
(s)
110
mV
E
B-LC2
Spike
100-
rate
50
(Hz)
Lateral
2
steering
-
velocity
°
(cm
s
-1
)
2
Forward
4
velocity
0
(cm
0)
Sound
B-LC2
110
my
ipsilateral
contralateral
formal
backward
JUJU_
Time
(s)
1130
m
V
Spike
288
rate
125
-
(Hz)
_I
t
L I
0
2
0
2
0-
-1
0
Lateral
steering
velocity
(cm
s'i)
Forward
velocity
(cm
s
-1
)
rj
ipsilateral
contralateral
forward
backward
L
2
3
4
lime
(a)
TH1-AC2
Spike
100
rate
50
-
(Hz)
LLII
Illll
1
4
°
mV
ipsilateral
contralateral
lanyard
0
0.5
1
1.5
2
2.5
3
Time
(s)
Lateral
steering
velocity
(cm
s
-1
)
Forward
velocity
(cm
s
-1
)
0
B-L11
Spike
100
rate
50-
(Hz)
Lateral
steering
velocity
0
(cm
s
-1
)
-4
Forward
12
velocity
0
(cm
el)
-6
0
nN
L
ipsilateral
contralateral
forward
backyard
3
4
Time
(s)
04040
W
ooevi
lj
fiw440000,446,4404,0
1
4.
120
mV
i
B-Dll
(5)
Spike
40
rate
20
(Hz)
0
Lateral
4
steering
0
velocity
(cm
6
-1
)
4
Forward
71
velocity
(cm
s
-1
)
_FL
IpsIlateral
contralateral
forward
0
5
6
Time
(s)
F
Lateral
seri
ve
te
ioci
n
ty
g
(cm
s
-
i)
-3
Forward
velocity
06
forward
(cm
s
-1
)
.4
backward
0
2
4
6
8
10
12
Time
(
s
)
ipsilatwal
contralateral
Sound
ipsilat.
Sound
contralat.
B-D11(5)
Spike
121
rate
60
(Hz)
1111111111
1111111111
11111111
1
(041
20
mVI
66_
0
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
2189
A
TH1-AC1
Spike
rate
(Hz)
Lateral
steering
velocity
(cm
s
-1
)
Forward
velocity
(cm
s
-1
)
Sound
TH1-AC1
hi
Ji1u
11120
mV
0.5
1
1.5
2
2.5
Time
(s)
Sound
B-DC1(5)
110
mV
Spike
1
°°
rate
50
(Hi)
0
Lateral
2
steering
0
ipsilateral
velocity
contralateral
(
00,
s
-1
)
2
Forward
2
velocity
01
forward
(
cm
0)
1
Time
(s)
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
2190
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
peaks
up
to
135
Hz
(Fig.
5D).
The
spiking
activity
during
spontaneous
walking
showed
significant
correlation
only
with
the
lateral
walking
velocity
to
the
contralateral
(Table
2).
The
response
to
the
first
pulse
of
a
chirp
during
walking
was
most
pronounced
and
reached
3.5
±
0.9
spikes
(n
=
20),
whereas
only
1.2-1.6
spikes
were
elicited
by
any
of
the
following
pulses
(Fig.
3D,
right).
The
response
pattern
became
more
scattered
compared
with
the
response
in
a
standing
cricket.
As
a
consequence,
the
PSTH
and
the
peaks
in
the
spike
rate
became
slightly
broader
and
less
pronounced.
The
pulse
struc-
ture
was
again
copied
by
both
the
timing
of
spikes
and
modulation
of
the
spike
rate.
Between
individual
peaks,
the
spike
rate
decreased
to
ca.
25
Hz.
During
the
recording
shown
in
Fig.
5D,
the
cricket
was
walking
relatively
slowly;
the
maximum
forward
velocity
did
not
exceed
2
cm
s
-1
.
The
strongest
activity
of
the
neuron
coincided
with
a
strong
turn
of
the
cricket
to
the
contralateral
side,
which
led
to
a
negative
value
of
the
forward
velocity
(Fig.
5D).
During
the
recordings
of
B-LC2,
the
crickets
were
walking
spontaneously,
albeit
with
the
forward
velocity,
never
exceed-
ing
4
cm
s
-1
(Fig.
5E).
During walking,
the
spike
rate
reached
up
to
85
Hz.
As
in
B-LC2,
the
spiking
activity
during
sponta-
neous
walking
showed
significant
correlation
only
with
the
lateral
walking
velocity
to
the
contralateral
but
not
with
the
forward
walking
velocity
(Table
2).The
scatter
in
the
raster
plot
for
responses
to
PP42
chirps
during
walking
indicated
a
high
variability
of
the
spike
pattern
(Fig.
3E,
right).
The
PSTH
showed
that
the
response
to
the
first
pulse
of
a
chirp
was
low
and
that
the
spikes
were
coupled
best
to
pulses
2-5.
Pulses
were
also
copied
by
the
modulation
of
the
spike
rate.
Between
subsequent
pulses,
the
spike
rate
dropped
by
30-50%.
The
high
variability
in
the
number
of
spikes/chirp
and
the
spike
frequency
modulation
in
response
to
PP42
during
standing
and
walking
are
illustrated
by
the
recording
in
Fig.
5E.
During
spontaneous
walking,
the
membrane
potential
of
B-DI1(5)
increased
by
4-5
mV.
The
spike
rate
was
5-15
Hz
with
peaks
up
to
40
Hz
(Fig.
5F).
There
was
significant
correlation
between
spiking
activity
and
the
forward
velocity
and
the
lateral
velocity
to
the
ipsilateral
side
(Table
2).
When
PP42
chirps
were
presented,
the
spikes
were
coupled
to
indi-
vidual
pulses
(Fig.
3F,
right).
The
first
three
pulses
of
a
chirp
were
copied
by
the
peaks
in
the
instantaneous
spike
rate,
whereas
the
last
three
pulses
were
not
mirrored
by
the
spike
rate.
The
scatter
of
the
spikes
still
remained
low,
and
accord-
ingly,
the
SC
value
was
still
very
high
at
0.89.
In
one
B-DI1(5)
neuron,
there
was
a
change
in
the
neuron's
response
to
the
species-specific
song
pattern,
which
was
related
to
the
animal's
standing
and
walking
behavior
(Fig.
5F).
While
a
sequence
of
PP42
was
presented,
the
neuron
hardly
responded
to
the
sound
pattern
in
the
walking
cricket.
The
animal
then
stopped
walk-
ing,
and
at
the
same
time,
the
neuron
changed
its
activity
pattern.
It
responded
to
the
next
chirp
with
a
series
of
six
large,
20
mV
EPSPs,
which
copied
the
pulse
pattern
and
elicited
two
to
three
spikes
(Fig.
5F).
Along
with
the
change
in
the
firing
pattern,
the
cricket
also
performed
several
strong
turns
in
the
direction
of
sound
(Fig.
5F).
The
negative
values
of
forward
velocity
in
this
case
were
the
consequence
of
a
strong
rotation
of
the
cricket
to
the
ipsilateral
side
without
any
forward
movement.
When
the
cricket
then
started
walking
again
(Fig.
5F,
from
7 s
to 11
s),
the
forward
velocity
increased,
and
the
response
to
the
sound
stimulation
from
the
contralateral
side
was
rather
weak
and
irregular.
After
the
animal
again
had
stopped
walking,
the
chirps
presented
from
the
ipsilateral
side
elicited
again
a
spiking
response
coupled
to
the
sound
pattern.
Thus,
the
neuron's
auditory
responses
were
clearly
reduced
during
walking.
B-DC1(5)
showed
low
activity
during
walking
with
no
correlation
between
the
spike
rate
and
forward
or
lateral
velocity
(Table
2;
Fig.
5G).
The
number
of
spikes
elicited
by
PP42
chirps
during
walking
was
slightly
lower
than
in
the
standing
animal.
During
sound
presentation,
there
was
a
main-
tained
depolarization
of
about
2.5
mV
(Fig.
5G).
The
raster
plot
shows
less
scatter
of
spikes
(Fig.
3G,
right).
During
walking,
the
spikes
were
better
coupled
to
pulses
than
during
standing,
which
was
reflected
in
an
increased
SC
value.
The
instantaneous
spike
rate
copied
the
first
two
pulses
of
a
chirp
only
and
reached
its
maximum
in
response
to
the
second
pulse.
DISCUSSION
The
experimental
setup
was
designed
to
explore
mecha-
nisms
of
auditory
pattern
recognition
and
steering
in
behavior-
ally
relevant
conditions.
It
allowed
intracellular
recordings
of
brain
neurons
in
tethered
crickets
during
standing
and
walking.
Recording
from
ascending,
local,
and
descending
neurons
enabled
us
to
describe
overall
trends
in
temporal
pattern
selectivity
and
copying
accuracy
along
the
processing
path-
way.
Arborization
patterns
of
neurons
and
processing
of
auditory
information
in
the
brain.
Our
staining
of
interneurons
at
different
levels
of
the
auditory
pathway
and
the
response
latencies
of
the
corresponding
interneurons highlight
the
flow
of
auditory
information
in
the
cricket
central
nervous
system.
From
the
auditory
neuropil
within
the
prothoracic
ganglion,
activity
reaches
the
brain
by
the
ANs
TH1-AC1
and
TH1-AC2
(Boyan
and
Williams
1982;
Wohlers
and
Huber
1982).
Both
interneurons
form
dense
axonal
arborizations
in
the
anterior
ventral
protocerebrum.
Auditory
activity
is
then
forwarded
toward
local
brain
neurons
such
as
B-LI1
(Boyan
1980;
Schil-
dberger
1984).
Its
structure
is
suited
to
receive
information
from
the
ascending
auditory
interneurons
and
to
activate
neu-
rons
in
the
posterior
median
protocerebrum
in
the
region
of
LALs.
The
dendrites
of
the
bilateral
local
neurons
B-LC1
and
B-LC2,
which
connect
the
left
and
right
LALs,
overlap
with
the
axonal
arborizations
of
B-LI1.
Interestingly,
also,
the
dendrites
of
the
descending
neurons
B-DI1(5)
and
B-DC1(5)
overlap
with
LALs.
These
structural
characteristics
of
the
ascending,
local,
and
descending
auditory
neurons
in
the
brain
point
toward
a
directed
flow
of
auditory
information
from
the
ascending
auditory
to
the
descending
premotor
interneurons
and
indicate
the
role
of
LAL
regions
(Fig.
6,
outlines)
as
an
integral
part
within
the
auditory-to-motor
pathway.
The
long
response
latency
of
B-LC2
suggests
that
this
neuron
is
not
suited
to
forward
commands
underlying
rapid
steering
re-
sponses;
it
may,
however,
be
part
of
a
pattern
recognition
mechanism
operating
on
a
larger
time
scale
(see
below).
The
structure
of
the
local
neurons
B-LC1
and
B-LC2
with
smooth
branches
on
the
ipsilateral
side
of
the
cell
body
and
varicose
arborizations
on
the
contralateral
side,
together
with
the
finding
that
graded
potentials
could
be
recorded
only
on
their
ipsilateral
side,
demonstrates
a
crossing
of
auditory
ac-
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
2191
TH1-AC
Thl
Fig.
6.
Suggested
flow
of
information
from
the
auditory
afferents
to
the
descending
premotor
intemeurons
in
the
cricket
nervous
system.
Circles
illustrate
dendritic
regions
of
auditory
intemeurons,
and
their
position
indicates
the
brain
regions,
which
may
be
a
part
of
the
auditory
pathway.
Thl,
prothoracic
ganglion.
The
outlines
indicate
the
LALs.
tivity
between
the
left
and
right
side
of
the
brain.
Even
if
sound
is
presented
only
to
one
side
of
the
animal,
a
resulting
motor
signal
may
descend
ipsi-
and
contralaterally
to
drive
phonot-
actic
walking
and
steering
toward
the
sound
source.
The
neurons
may
also
sharpen
any
bilateral
descending
motor
commands
by
inhibitory
interaction.
The
relevance
of
the
LALs
for
insect
motor
activity
has
been
described
in
connection
with
visual
control
of
flight
behavior
in
locusts
(Homberg
1994)
and
the
control
of
pheromone-guided
flight
in
moths
(Kanzaki
and
Shibuya
1986;
Kanzaki
et
al.
1991).
Additionally,
Bohm
and
Schildberger
(1992)
described
a
descending
brain
neuron
in
crickets
with
extensive
dendritic
arborizations
in
the
LAL
area,
which
was
directly
involved
in
control
of
walking.
Together
with
the
ventral
protocerebrum,
the
LAL
appears
to
form
a
premotor
center
in
insect
brains
(Bohm
and
Schildberger
1992).
Numerous
studies
also
imply
a
strong
connection
of
the
central
body
complex
to
insect
walk-
ing
and
turning
activity
(Homberg
1987;
Popov
et
al.
2004;
Ridgel
et
al.
2007;
Strauss
2002;
Strauss
and
Heisenberg
1993).
However,
so
far,
there
is
no
evidence
that
this
region
would
be
involved
in
the
control
of
cricket
phonotaxis.
Bilateral
brain
neurons
structurally
similar
to
B-LC1
and
B-LC2
were
described
as
high-frequency
neurons
and
might
be
involved
in
evasive
flight
behavior
in
the
cricket,
Teleogryllus
oceanicus
(Brodfuehrer
and
Hoy
1990).
Some
of
these
neurons
also
responded
weakly
to
5
kHz
pulses,
whereas
B-LC1
neu-
rons,
which
we
described
here,
responded
also
to
ultrasound
stimulation
(data
not
shown).
This
points
to
similarities
of
auditory
processing
by
bilateral
local
brain
neurons
in
different
cricket
species.
The
response
of
central
auditory
neurons
to
both
the
frequency
of
the
conspecific
song
and
the
frequency
of
bat
echolocation
raises
the
question
of
how
the
auditory-to-
motor
pathway
generates
opposing
motor
responses.
In
T.
oceanicus,
for
example,
an
identified
ascending
auditory
in-
terneuron
elicits
antipredator
behavior
only
in
flight
(Nolen
and
Hoy
1984)
but not
during
walking
(Hofstede
et
al.
2009).
Similarly,
the
findings
of
Schul
and
Shulze
(2001)
suggest
that
in
Tettigonia,
the
auditory
interneuron
TN-1
has
different
response
characteristics
during
walking
and
flying
because
of
differences
in
predation
pressure
(Schul
and
Schulze
2001).
Selectivity
for
temporal
patterns.
Various
models
have
been
proposed
for
auditory
pattern
recognition
in
insects.
Cross-
correlation
mechanism
involves
an
internal
template,
against
which
the
song
pattern
is
matched
(Hennig
2003;
Weber
and
Thorson
1989).
Another
mechanism
is
an
autocorrelation
filter
with
one
direct
and
one
delayed
pathway
that
converges
on
a
coincidence
detector
(Reiss
1964),
which
is
activated
when-
ever
the
time
interval
between
pulses
equals
the
delay.
Schil-
dberger
(1984)
proposed
a
neural
band-pass
system
for
pattern
recognition
in
G.
bimaculatus.
Bush
and
Schul
(2006)
provide
evidence
that
in
a
katydid,
pulse-rate
recognition
relies
on
the
resonant
properties
within
an
auditory
network.
In
G.
bimaculatus,
sound
patterns
elicited
the
best
phonot-
actic
steering
response
when
the
PP
was
in
the
range
of
26-48
ms,
which
included
the
PP
of
the
natural
calling
song
(Doherty
1985;
Thorson
et
al.
1982).
Both
ascending
auditory
neurons,
TH1-AC1
and
TH1-AC2,
respond
to
song
patterns
across
a
broad
range
of
PP
values,
well
below
and
above
the
phonot-
actically
effective
range;
however,
they
exhibit
weak
low-pass
properties
for
both
the
mean
spike
rate
and
the
peak
instanta-
neous
spike
rate.
Schildberger
(1984)
proposed
that
local
brain
neurons,
act-
ing
as
low-pass
and
high-pass
filters,
are
combined
to
act
as
a
band-pass
filter
system
for
pattern
recognition
in
G.
bimacu-
latus.
In
this
system,
the
information
on
the
occurrence
of
a
conspecific
calling
song
is
converted
from
an
accurately
copied
representation
to
a
temporally
filtered
signal
capable
of
trig-
gering
phonotaxis.
The
selectivity
of
these
local
neurons
for
specific
temporal
patterns
was
reflected
in
the
intensity
of
the
response,
i.e.,
the
number
of
spikes/chirp.
Based
on
these
values,
two
of
the
local
neurons
described
here,
B-LC1
and
B-LC2,
demonstrated
strong
band-pass
selectivity
for
the
spe-
cies-specific
pattern
(see
Table
1;
Fig.
2,
D
and
E).
High-pass
properties
were
not
observed
in
any
of
the
neurons
described.
The
activity
of
the
band-pass
responding
neuron
B-LC2
came
with
extremely
long
response
latency
(Figs.
3E
and
4)
due
to
recurrent
failure
to
respond
to
the
first
pulse
of
the
chirp.
Whereas
in
most
other
auditory
neurons,
the
first
pulse
of
a
chirp
elicited
the
strongest
response,
a
response
starting
with
the
second
sound
pulse
will
require
specific
temporal
process-
ing.
The
response
pattern
of
B-LC2
could
be
explained
by
an
initial
mixed
excitatory—inhibitory
input,
or
it
may
point
to-
ward
an
autocorrelation
mechanism
involving
a
time-delay
pathway
that
matches
the
pulse
intervals
(Weber
and
Thorson
1989).
In
the
barn
owl,
adaptations
in
the
microsecond
range
of
delay
lines
underlying
azimuthal
sound
localization
are
achieved
by
extending
the
axon length
of
magnocellular
neu-
rons
projecting
into
the
laminar
nucleus
(Carr
and
Boudreau
1993).
In
bats,
the
delays
are
created
by
different
lengths
of
Auditory
afferents
B
LI
Brain
B-D
BL
B
LC
B-D
TH1-AC
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
2192
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
delay
lines
and
additional
interneurons
(Kamata
et
al.
2004).
A
40-ms
delay
line
in
the
cricket
brain
may,
however,
be
ques-
tionable.
A
central
resonance
mechanism,
as
proposed
for
bush-crickets
(Bush
and
Schul
2006)
and
supported
by
the
oscillatory
responses
recorded
from
an
auditory
neuron
in
the
brain
of
T.
oceanicus
(Hennig
et
al.
2004),
may
also
be
involved
in
pattern
recognition.
Descending
neurons
described
here,
B-DI1(5)
and
B-DC1(5),
both
exhibit
band-pass
pattern
selectivity
in
the
mean
spike
rate
and
the
first
type
also
in
the
peak
instantaneous
spike
rate
values.
The
strong
response
decrement
of
B-DC1(5)
to
behav-
iorally
unattractive
chirps
might
mirror
the
selectivity
of
the
preceding
networks
for
species-specific
patterns.
In
the
local
and
descending
neurons,
the
selectivity
for
the
behaviorally
attractive
pulse
pattern
is
more
obvious
for
the
mean
than
the
peak
instantaneous
spike
rates.
This
suggests
that
although
the
instantaneous
spike
rate
might
be
the
appro-
priate
coding
principle
for
the
pulse
pattern
at
the
level
of
the
thoracic
omega
neuron
(ON1)
(Nabatiyan
et
al.
2003),
time-
averaged
neural
activity
appears
to
be
more
important
at
higher
levels
of
the
processing
pathway.
The
overall
decrease
in
the
number
of
APs/chirp
along
the
processing
levels
suggests
a
transformation
of
a
temporal
code
into
a
labeled
line
code,
although
the
coupling
of
neuronal
activity
to
sound
pulses
was
maintained.
Copying
of
the
pulse
pattern
in
the
auditory-to-motor
pathway.
During
cricket
phonotaxis,
the
temporal
structure
of
the
song
is
preserved
from
the
auditory
input
to
the
motor
output
(Hedwig
and
Poulet
2004).
Whichever
the
mechanism
of
pattern
recognition
underlying
phonotaxis
in
the
auditory-
to-motor
pathway,
a
representation
of
sound
pulses
needs
to
be
forwarded
to
the
thoracic
motor
network
controlling
leg
walk-
ing
and
steering
movements.
Our
experiments
aimed
to
ana-
lyze
if
the
temporal
structure
of
the
song
pattern
is
forwarded
toward
the
thoracic
ganglia
after
auditory
processing
in
the
brain.
Whereas
the
pool
of
auditory
afferents
faithfully
copied
the
pulses
of
a
chirp,
even
at
PP10
(Nabatiyan
et
al.
2003),
Schildberger
(1984)
and
Wohlers
and
Huber
(1982)
showed
that
in
the
spike
rate
of
the
ascending
neurons,
the
high
fidelity
of
pulse
copying
is
lost
below
PP20.
Our
data
on
responses
to
song
pattern
PP10
confirm
this
for
all
levels
of
the
auditory
pathway.
For
local
band-pass
brain
neurons,
Schildberger
(1984)
re-
ported
SC
values
below
0.3;
here,
however,
we
show
that
overall,
the
SC
values
for
the
attractive
PP
of
42
ms
did
not
drop
below
0.3
along
the
processing
pathway.
At
the
level
of
local
brain
neurons,
the
SC
values
of
B-LC2
were
slightly
above
the
pulse
copying
accuracy
reported
in
previous
studies
(Bohm
and
Schildberger
1992;
Schildberger
1984),
whereas
for
B-LI1
and
B-LC1,
the
copying
fidelity
far
exceeded
that
of
all
of
the
previously
described
local
auditory
neurons
in
the
cricket
brain.
Moreover,
even
in
the
descending
neuron
B-DI1(5),
the
SC
value
exceeded
that
of
the
ascending
neurons,
and
during
walking
in
B-DC1(5),
the
SC
reached
the
SC
level
of
TH1-
AC1.
Thus,
there
is
good
evidence
that
the
pulse
pattern
is
copied
and
forwarded
to
descending
premotor
interneurons.
This
is
a
major functional
prerequisite
so
that
pulse-related
activity
patterns
can
be
forwarded
from
the
brain
toward
the
thoracic
ganglia.
However,
the
data
so
far
do
not
provide
evidence
that
any
of
the
local
or
descending
neurons
described
here
would
be
directly
involved
in
generating
or
mediating
the
commands
for
auditory
steering
responses.
Neuron
activity
and
walking.
During
walking
in
the
absence
of
sound
stimuli,
all
of
the
auditory
neurons
were
activated.
Some
degree
of
correlation
between
walking
and
a
neuron's
spike
rate
was
present
at
all
levels
of
the
pathway.
TH1-AC1
showed
a
strong
and
statistically
significant
correlation
be-
tween
the
spike
rate
and
both
the
forward
velocity
and
lateral
velocity
to
the
contralateral
side.
TH1-AC2
showed
lower
levels
of
correlation
to
both
walking
parameters.
The
only
brain
neurons
that
exhibited
positive
correlation
between
the
spike
rate
and
walking
to
the
side
contralateral
to
their
input
region
were
the
local
neurons
with
an
axon
crossing
the
midline,
B-LC1
and
B-LC2.
The
activity
of
these
neurons
had
no
significant
correlation
to
the
forward
velocity.
Only
one
type
of
neuron,
the
B-DC1(5),
showed
no
statistically
signifi-
cant
correlation
to
walking
at
all,
although
its
spike
rate
increased
during
walking.
The
neuronal
activity
that
was
correlated
to
walking
may
arise
from
a
motor
activity-induced
oscillation
of
the
tympa-
num
and
hence,
an
excitation
of
auditory
afferents.
This
was
proposed
as
a
possible
cause
for
modulation
of
auditory
re-
sponses
during
locomotion
in
TH1-AC1
and
the
ON1
in
G.
bimaculatus
by
Schildberger
et
al.
(1988).
An
unspecific
acti-
vation
of
auditory
afferents
would
have
contributed
to
the
correlation
coefficients
between
walking
and
activity
of
audi-
tory
neurons.
The
effect
of
walking
on
processing
of
sound
signals
and
signals
of
other
sensory
modalities
was
described
in
several
studies
(Bohm
and
Schildberger
1992;
Staudacher
2001;
Stau-
dacher
and
Schildberger
1998).
Most
descending
neurons
were
gated
by
walking
and
started
to
respond
or
respond
more
strongly
to
acoustic
stimuli
when
the
crickets
were
walking.
Here,
we
found
that
walking
resulted
in
a
general
decrease
in
response
intensity
of
the
band-pass
neurons
(Table
1;
Fig.
3).
This
may
be
explained
by
supression
of
auditory
responses
at
particular
phases
of
the
leg
movement
(Schildberger
et
al.
1988).
During
walking,
the
overall
mean
spike
rates
in
response
to
chirps
decreased,
whereas
the
peak
instantaneous
spike
rates
mostly
increased
along
the
processing
pathway.
Most
neurons
fired
some
additional
APs
that
were
not
related
to
the
acoustic
stimuli
but
could
have
been
evoked
by
unspecific
activity
of
auditory
afferents.
In
the
ascending
and
local
auditory
neurons,
the
pattern
copying
was
slightly
weaker
during
walking
than
during
standing.
In
contrast,
the
SC
of
the
descending
neuron
B-DC1(5)
improved
with
walking
by
22%.
Furthermore,
among
the
remaining
neurons,
the
decrease
in
SC
was
lowest
for
the
other
descending
neuron,
B-DI1(5)
(Table
1).
As
crickets
steer
to
sound
pulses
(Hedwig
and
Poulet
2004),
one
would
expect
high
SC
values
during
phonotactic
steering.
A
general
decrease
in
the
pulse
copying
accuracy
during
walking,
shown
here
in
crickets,
which
were
mostly
not
phonotactically
active,
might
indicate
that
"nonphonotactic"
and
phonotactic
walking
may
need
to
be
regarded
as
two
distinct
behaviors
and
that
with
the
onset
of
phonotaxis,
certain
pattern
recognition
pathways,
inactive
in
solely
walking
crickets,
are
switched
on.
In
one
B-DI1(5)
recording,
the
neuron
responded
to
the
sound
pattern
during
standing,
whereas
the
auditory
response
failed
during
walking,
which
is
just
the
opposite
of
the
gated
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
2193
responses
described
by
Staudacher
(2001).
However,
neurons
involved
in
steering
need
to
be
activated
during
phonotaxis.
It
is
worth
noting
that
although
in
the
aforementioned
B-DI1(5),
the
responses
to
conspecific
chirps
occurred
after
the
cessation
of
walking,
they
were
accompanied
by
several
strong
turns
toward
the
sound,
which
may
be
interpreted
as
transient
pho-
notactic
turns.
ACKNOWLEDGMENTS
We
thank
our
colleagues
K.
Kostarakos,
S.
Schoeneich,
and
F.
Dupuy
from
the
University
of
Cambridge
and
A.
Cokl
and
N.
Stritih
from
the
National
Institute
of
Biology
in
Ljubljana,
Slovenia
(current
address
of
M.
Zorovic'),
for
critically
reading
and
commenting
on
the
manuscript.
GRANTS
This
work
was
supported
by
a
Biotechnology
and
Biological
Sciences
Research
Council
grant.
DISCLOSURES
No
conflicts
of
interest,
financial
or
otherwise,
are
declared
by
the
author(s).
REFERENCES
Boyan
GS.
Auditory
neurons
in
the
brain
of
the
cricket
Gryllus
bimaculatus
(De
Geer).
J
Comp
Physiol
A
140:
81-93,
1980.
Boyan
GS,
Williams
JLD.
Auditory
neurons
in
the
brain
of
the
cricket
Gryllus
bimaculatus
(De
Geer):
Ascending
intemeurons.
J
Insect
Physiol
6:
493-501,
1982.
Bohm
H,
Schildberger
K.
Brain
neurons
involved
in
the
control
of
walking
in
the
cricket
Gryllus
bimaculatus.
J
Exp
Biol
166:
113-130,
1992.
Brodfuehrer
PD,
Hoy
RR.
Ultrasound
sensitive
neurons
in
the
cricket
brain.
J
Comp
Physiol
A
166:
651-662,
1990.
Bush
SL,
Schul
J.
Pulse-rate
recognition
in
an
insect:
evidence
of
a
role
for
oscillatory
neurons.
J
Comp
Physiol
A
192:
113-121,
2006.
Carr
CE,
Boudreau
RE.
Organization
of
the
nucleus
magnocellularis
and
the
nucleus
laminaris
in
the
barn
owl:
Encoding
and
measuring
interaural
time
differences.
J
Comp
Neurol
334:
337-355,
1993.
Doherty
JA.
Temperature
coupling
and
"trade-off'
phenomena
in
the
acoustic
communication
system
of
the
cricket,
Gryllus
bimaculatus
deGeer
(Grylli-
dae).
J
Exp
Biol
114:
17-35,
1985.
Doherty
JA,
Pires
A.
A
new
microcomputer
based
method
for
measuring
walking
phonotaxis
in
the
field
crickets
(Gryllidae).
J
Exp
Biol
130:
425-432,
1987.
Fay
RR.
Structure
and
function
in
sound
discrimination
among
vertebrates.
In:
The
Evolutionary
Biology
of
Hearing,
edited
by
Webster
D,
Fay
RR,
Popper
AN.
New
York:
Springer-Verlag,
1992,
p.
229-263.
Fee
MS.
Active
stabilization
of
electrodes
for
intracellular
recording
in
awake
behaving
animals.
Neuron
27:
461-468,
2000.
Gerhardt
CH,
Huber
F.
Acoustic
Communication
in
Insects
and
Anurans.
Chicago:
The
University
of
Chicago
Press,
2002.
Hedwig
B.
On
the
role
in
stridulation
of
plurisegmental
intemeurons
of
the
acridid
grasshopper
Omocestus
viridulus
L.
I.
Anatomy
and
physiology
of
descending
cephalothoracic
intemeurons.
J
Comp
Physiol
A
158:
413-427,
1986.
Hedwig
B,
Knepper
M.
NEUROLAB,
a
comprehensive
program
for
the
analysis
of
neurophysiological
and
behavioral
data.
J
Neurosci
Methods
45:
135-148,
1992.
Hedwig
B,
Poulet
J.
Complex
auditory
behaviour
emerges
from
simple
reactive
steering.
Nature
430:
781-785,
2004.
Hedwig
B,
Poulet
J.
Mechanisms
underlying
phonotactic
steering
in
the
cricket
Gryllus
bimaculatus
revealed
with
a
fast
trackball
system.
J
Exp
Biol
208:
915-927,
2005.
Hennig
RM. Acoustic
feature
extraction
by
cross-correlation
in
crickets?
J
Comp
Physiol
A
189:
589-598,
2003.
Hennig
RM.
Ascending
auditory
intemeurons
in
the
cricket
Teleogryllus
commodus
(Walker):
comparative
physiology
and
direct
connections
with
afferents.
J
Comp
Physiol
A
163:
135-143,
1988.
Hennig
RM,
Franz
A,
Stumpner
A.
Processing
of
auditory
information
in
insects.
Microsc
Res
Tech
63:
351-374,
2004.
Hofstede
HM,
ICillow
J,
Fullard
JH.
Gleaning
bat
echolocation
calls
do
not
elicit
antipredator
behaviour
in
the
Pacific
field
cricket,
Telleogryllus
oce-
anicus
(Orthoptera:
Gryllidae).
J
Comp
Physiol
A
195:
769-776,
2009.
Homberg
U.
Structure
and
functions
of
the
central
complex
in
insects.
In:
Arthropod
brain:
its
evolution,
development,
structure
and
functions,
edited
by
Gupta
AP.
New
York:
Wiley,
1987,
p.
347-367.
Homberg
U.
Flight-correlated
activity
changes
in
neurons
of
the
lateral
accessory
lobes
in
the
brain
of
the
locust
Schistocerca
gregaria.
J
Comp
Physiol
A
175:
597-610,
1994.
Hoy
RR.
Acoustic
communication
in
crickets:
a
model
system
for
the
study
of
feature
detection.
Fed
Proc
37:
2316-2323,
1978.
Huber
F,
Moore
TE,
Loher
W.
Cricket
Behavior
and
Neurobiology.
Ithaca,
NY:
Cornell
University
Press,
1989.
Kamata
E,
Inoue
S,
Zheng
M,
Kashimori
Y,
Kambara
T.
A
neural
mechanism
for
detecting
the
distance
of
a
selected
target
by
modulating
the
FM
sweep
rate
of
biosonar
in
echolocation
of
bat.
Biosystems
76:
55-64,
2004.
Kanzald
R,
Arbas
EA,
Hildebrand
JG.
Physiology
and
morphology
of
descending
neurons
in
pheromone-processing
olfactory
pathways
in
the
male
moth
Manduca
sexta.
J
Comp
Physiol
A
169:
1-14,
1991.
Kanzald
R,
Shibuya
T.
Descending
protocerebral
neurons
related
to
the
mating
dance
of
the
male
silkworm
moth.
Brain
Res
377:
378
—382,
1986.
Knepper
M,
Hedwig
B.
NEUROLAB,
a
PC-program
for
the
processing
of
neurobiological
data.
Comput
Methods
Programs
Biomed,
52:
75-77,
1997.
Nabatiyan
A,
Poulet
JFA,
de
Polavieja
GG,
Hedwig
B.
Temporal
pattern
recognition
based
on
instantaneous
spike
rate
coding
in
a
simple
auditory
system.
J
Neurophysiol
90:
2484
—2493,
2003.
Nolen
TG,
Hoy
RR.
Initiation
of
behavior
by
single
neurons:
the
role
of
behavioral
context.
Science
226:
992-994,
1984.
Pollack
GS,
Hoy
RR.
Temporal
pattern
as
a
cue
for
species-specific
calling
song
recognition
in
crickets.
Science
204:
429-432,
1979.
Popov
AV,
Peresleni
AI,
Savvateeva-Popova
EV,
Wolf
R,
Heisenberg
M.
The
role
of
the
mushroom
bodies
and
of
the
central
complex
of
Drosophila
melanogaster
brain
in
the
organization
of
courtship
behavior
and
communicative
sound
production.
J
Evol
Biochem
and
Physiol
40:
641-652,
2004.
Poulet
J,
Hedwig
B.
Auditory
orientation
in
crickets:
Pattern
recognition
controls
reactive
steering.
Proc
Natl
Acad
Sci
USA
102:
15665-15669,
2005.
Reiss
RF.
A
theory
of
resonance.
In:
Neural
Theory
and
Modeling,
edited
by
Reiss
RF.
Stanford,
CA:
Stanford
University
Press,
1964,
p.
105-15665-
137.
Ridgel
AL,
Alexander
BE,
Ritzmann
RE.
Descending
control
of
turning
behavior
in
the
cockroach,
Blaberus
discoidalis.
J
Comp
Physiol
A
193:
385-402,
2007.
Rowell
CHF.
The
taxonomy
of
invertebrate
neurons:
a
plea
for
a
new
field.
Trends
Neurosci
12:
169-174,
1989.
Sakura
M,
Lambrinos
D,
Labhart
T.
Polarized
skylight
navigation
in
insects:
Model
and
electrophysiology
of
e-vector
coding
by
neurons
in
the
central
complex.
J
Neurol
99:
667-682,
2007.
Schildberger
K.
Temporal
selectivity
of
identified
auditory
neurons
in
the
cricket
brain.
J
Comp
Physiol
A
155:
171-185,
1984.
Schildberger
K,
Horner
M.
The
function
of
auditory
intemeurons
in
cricket
phonotaxis.
I.
Influence
of
hyperpolarisation
of
identified
neurons
on
sound
localization.
J
Comp
Physiol
A
163:
621-631,
1988.
Schildberger
K,
Milde
JJ,
Horner
M.
The
function
of
auditory
intemeurons
in
cricket
phonotaxis.
II.
Modulation
of
auditory
responses
during
locomo-
tion.
J
Comp
Physiol
A
163:
633-640,
1988.
Schul
J,
Schulze
W.
Phonotaxis
during
walking
and
flight:
are
differences
in
selectivity
due
to
predation
pressure?
Naturwissenshaften
88:
438-442,
2001.
Staudacher
E. Distribution
and
morphology
of
descending
brain
neurons
in
the
cricket
Gryllus
bimaculatus.
Cell
Tissue
Res
294:187-202,
1998.
Staudacher
E.
Sensory
responses
of
descending
brain
neurons
in
the
walking
cricket,
Gryllus
bimaculatus.
J
Comp
Physiol
A
187:
1-17,
2001.
Staudacher
E,
Schildberger
K.
Gating
of
sensory responses
of
descending
brain
neurons
during
walking
in
crickets.
J
Exp
Biol
201:
559-572,
1998.
Strauss
R.
The
central
complex
and
the
genetic
dissection
of
locomotor
behaviour.
Curr
Opin
Neurobiol
12:
633-638,
2002.
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org
2194
PROCESSING
OF
AUDITORY
PATTERNS
IN
THE
CRICKET
BRAIN
Strauss
R,
Heisenberg
M.
A
higher
control
center
of
locomotor
behavior
in
the
Drosophila
brain.
J
Neurosci
13:
1852-1861,
1993.
Thorson
J,
Weber
T,
Huber
F.
Auditory
behaviour
of
the
cricket.
II.
Simplicity
of
the
calling-song
recognition
in
Gryllus,
and
anomalous
pho-
notaxis
at
abnormal
carrier
frequencies.
J
Comp
Physiol
A
146:
361-378,
1982.
Viceic
D,
Fornari
E,
Thiran
JP,
Maeder
PP,
Meuli
R,
Adriani
M,
Clarke
S.
Human
auditory
belt
areas
specialized
in
sound
recognition:
a
functional
magnetic
resonance
imaging
study.
Neuroreport
17:
1659-1662,
2006.
Weber
T,
Thorson
J.
Phonotactic
behaviour
of
walking
crickets.
In:
Cricket
Behavior
and
Neurobiology,
edited
by
Huber
F,
Moore
TE,
Loher
W.
Ithaca,
NY:
Cornell
University
Press,
1989,
p.
310-1659-339.
Weber
T,
Thorson
J,
Huber
F.
Auditory
behaviour
of
the
cricket.
I.
Dynamics
of
compensated
walking
and
discrimination
paradigms
on
the
Kramer
treadmill.
J
Comp
Physiol
A
141:
215-232,
1981.
Wohlers
DW,
Huber
F.
Processing
of
sound
signals
by
six
types
of
neurons
in
the
prothoracic
ganglion
of
the
cricket
Gryllus
campestris
L.
J
Comp
Physiol
A
146:
161-173,
1982.
-
1
(
oy
q
Al
t.
%
J
Neurophysiol
VOL
105
MAY
2011
www.jn.org