Visual
perception
gives
us
a
reliable
estimate
of
the
world
throughout
day,
thanks
to
computations
implemented
in
visual
circuits.
In
given
space
or
time,
relative
change
luminance,
contrast,
forms
basis
for
many
downstream
like
detection
edges,
shape,
orientation
and
motion.
Stable
contrast
estimation
is
challenged
by
immensely
changing
lightning
conditions
that
can
slowly
day
much
more
rapidly
when
viewing
natural
scenes.
To
ensure
behavioral
responses
same
(contrast)
stimulus,
systems
face
monumental
task
keeping
representations
stable.
Gain
mechanisms
operate
across
nervous
system
match
neural
relevant
stimulus
distribution.
systems,
photoreceptors
have
numerous
implement
sensory
gain
keep
stable
slow
changes
luminance
occurring
day.
However,
major
also
occur
at
fast
timescales
due
distributions
within
scenes
which
we
view
through
navigate.
The
extend
photoreceptor
often
insufficient
these
but
vision
still
luminance-invariantly
as
shown
human
perception,
neurons
vertebrate
cortex,
lateral
geniculate
nucleus
retina.
This
highlights
necessity
post-receptor
evidence
from
few
studies
suggests
this
happens
detailed
understanding
underlying
circuitry,
including
contribution
different
cell
types,
their
specialization
molecular
gain,
not
understood
any
system.
A
recent
study
Drosophila
melanogaster
showed
fly
behavior
stimuli
containing
decrements
(OFF
stimuli)
luminance-invariant
(Ketkar
et
al.
2020).
Two
second
order
located
lamina
post-synaptic
photoreceptors,
L2
L3,
contribute
OFF
encoding
respectively.
L3
underlie
providing
rapid
pathway
opened
up
possibility
investigating
using
advanced
genetic
arsenal.
first
part
thesis,
asked
if
ON
utilizes
second-order
support
this.
was
thought
receive
inputs
only
single
neuron
type,
L1
neurons,
instead
sensitive
parallel
pathways
formed
pathway.
Using
calcium
imaging
paradigms,
revealed
all
encode
distinct
features
scenery
distribute
both
showing
arise
neurons.
requires
provided
well
inputs.
For
provide
contributions
since
differ.
Thus,
combination
temporal
filtering
leading
contrast-luminance
specializations
vital
processing.
bipolar
cells,
retina,
form
channels
suggesting
serve
roles
study,
I
investigated
factors
shaping
feature
properties
cell-types.
Biophysical
are
largely
shaped
ion
channel
expression
significant
fraction
signal
processing
properties.
RNA-seq
data
endogenous
protein
tagging,
identified
specific
subtype
voltage
gated
potassium
channels,
Ka
Shaker
Shal,
being
highly
expressed
pharmacological
manipulations,
circuit
wide
role
enhancing
responses.
Additionally
cell-type
RNAi,
involved
sharpening
Similarities
other
species
suggest
conserved
cell-types
conferring
computational
circuits
third
gain.
characterized
celltypes
circuitry
dendrites
two
third-order
medulla
location
where
arises.
Tm1
Tm9
scale
ways.
Whereas
reach
estimations
boost
signals
low
luminances.
Spatial
pooling
underlies
glutamatergic
yet
use
Genetic
analysis
shows
relies
on
inhibition
mediated
GluClα
whereas
luminanceinvariance
remains
be
identified.
reveals
details
how
acting
specifically
dim
light
vertebrates
presence
convergent
strategies
vision.
Testing
exist
whether
similar
spatial
reveal
scene
species.
My
initial
added
examples
divergent
happening
One
outcome
such
complex
wiring
synaptic
connections
made
between
partners,
information
differently.
get
causal
connections,
manipulations
restricted
synapses
necessary.
However
current
maximally
achieve
specificity
affects
presynapses
one
thus
limit
level
detail
understand
function.
my
final
developing
tool,
Synapse
Targeted
Activity
Block
(STAB),
manipulate
partners.
STAB
based
components
each
partners
activates
coincide,
connections.
component
viral
protease
(TEV
-
TEVp)
cleaves
conditionally
integrated
cleavage
site
TEVcs)
an
constitutes
component.
We
generated
TEVp
fusion
proteins
localize
regions
function
extracellularly
vitro
vivo.
Concentrating
transsynaptic
postsynaptic
receptors
known
essential
function,
experiments
suggested
leads
degradation
receptor
TEVcs
levels
enough
recapitulate
LOF
phenotypes
optimized
version
higher
degradation.
with
enhanced
needs
validated
vivo
before
applying
it
example
synapse
opens
door
investigations
insights
circuits,
behavior.
Overall,
work
combined
computation
dynamically
environments.
development
tool
will
allow
resolution
than
possible
date,
applied
studying
computations,
type
analysis.
Abstract
Originally
a
genetic
model
organism,
the
experimental
use
of
Drosophila
melanogaster
has
grown
to
include
quantitative
behavioral
analyses,
sophisticated
perturbations
neuronal
function,
and
detailed
sensory
physiology.
A
highlight
these
developments
can
be
seen
in
context
vision,
where
pioneering
studies
have
uncovered
fundamental
generalizable
principles
processing.
Here
we
begin
with
an
overview
vision-guided
behaviors
common
methods
for
probing
visual
circuits.
We
then
outline
anatomy
physiology
brain
regions
involved
processing,
beginning
at
periphery
ending
descending
motor
control.
Areas
focus
contrast
motion
detection
optic
lobe,
circuits
feature
selectivity,
computations
support
spatial
navigation,
contextual
associative
learning.
Finally,
look
future
fly
neuroscience
discuss
promising
topics
further
study.
bioRxiv (Cold Spring Harbor Laboratory),
Год журнала:
2023,
Номер
unknown
Опубликована: Март 13, 2023
Abstract
We
can
now
measure
the
connectivity
of
every
neuron
in
a
neural
circuit,
but
we
are
still
blind
to
other
biological
details,
including
dynamical
characteristics
each
neuron.
The
degree
which
measurements
alone
inform
understanding
computation
is
an
open
question.
Here
show
that
with
only
network,
predict
activity
underlying
computation.
constructed
model
network
experimentally
determined
for
64
cell
types
motion
pathways
fruit
fly
optic
lobe
unknown
parameters
single
and
synapse
properties.
then
optimized
values
these
using
techniques
from
deep
learning,
allow
detect
visual
motion.
Our
mechanistic
makes
detailed
testable
predictions
connectome.
found
agreed
experimental
across
24
studies.
work
demonstrates
strategy
generating
hypotheses
about
mechanisms
circuit
function
measurements.
this
more
likely
be
successful
when
neurons
sparsely
connected—a
universally
observed
feature
networks
species
brain
regions.
Nature Communications,
Год журнала:
2024,
Номер
15(1)
Опубликована: Фев. 21, 2024
Visual
systems
are
homogeneous
structures,
where
repeating
columnar
units
retinotopically
cover
the
visual
field.
Each
of
these
columns
contain
many
same
neuron
types
that
distinguished
by
anatomic,
genetic
and
-
generally
functional
properties.
However,
there
exceptions
to
this
rule.
In
800
Drosophila
eye,
is
an
anatomically
genetically
identifiable
cell
type
with
variable
properties,
Tm9.
Since
anatomical
connectivity
shapes
neuronal
we
identified
presynaptic
inputs
several
hundred
Tm9s
across
both
optic
lobes
using
full
adult
female
fly
brain
(FAFB)
electron
microscopic
dataset
FlyWire
connectome.
Our
work
shows
Tm9
has
three
major
sparsely
distributed
inputs.
This
differs
from
other
Tm
neurons,
which
have
only
one
major,
more
stereotypic
than
Genetic
synapse
labeling
showed
heterogeneous
wiring
exists
individuals.
Together,
our
data
argue
system
uses
heterogeneous,
circuit
properties
achieve
robust
processing.
bioRxiv (Cold Spring Harbor Laboratory),
Год журнала:
2023,
Номер
unknown
Опубликована: Окт. 13, 2023
A
catalog
of
neuronal
cell
types
has
often
been
called
a
"parts
list"
the
brain,
and
regarded
as
prerequisite
for
understanding
brain
function.
In
optic
lobe
bioRxiv (Cold Spring Harbor Laboratory),
Год журнала:
2024,
Номер
unknown
Опубликована: Апрель 8, 2024
Abstract
Normalization
is
a
fundamental
operation
in
image
processing.
Convolutional
nets
have
evolved
to
include
large
number
of
normalizations
(Ioffe
and
Szegedy
2015;
Ulyanov,
Vedaldi,
Lempitsky
2016;
Wu
He
2018),
this
architectural
shift
has
proved
essential
for
robust
computer
vision
(He
et
al.
Bjorck
2018;
Santurkar,
Tsipras,
Ilyas
2018).
Studies
biological
vision,
contrast,
invoked
just
one
or
few
model
psychophysical
(Mach
1868;
Furman
1965;
Sperling
1970)
physiological
(Carandini
Heeger
2011;
Shin
Adesnik
2024)
observations
that
accumulated
over
century.
Here
connectomic
information
(Matsliah
2023)
used
argue
interneurons
the
fly
visual
system
support
with
unprecedented
specificity.
Ten
interneuron
types
distal
medulla
(Dm)
optic
lobe,
example,
appear
chiefly
spatial
normalizations,
each
which
specific
single
cell
type
length
scale.
Another
Dm
supports
normalization
features
as
well
space.
Two
outlier
do
not
at
all.
Interneuron
likely
be
normalizers
are
identified
only
but
also
all
other
families
lobe.
For
diversity
appears
an
inevitable
consequence
specificity
normalizations.
The
accurate
processing
of
contrast
is
the
basis
for
all
visually
guided
behaviors.
Visual
scenes
with
rapidly
changing
illumination
challenge
computation
because
photoreceptor
adaptation
not
fast
enough
to
compensate
such
changes.
Yet,
human
perception
stable
even
when
visual
environment
quickly
changing,
suggesting
rapid
post
receptor
luminance
gain
control.
Similarly,
in
fruit
fly
Drosophila,
control
leads
invariant
behavior
moving
OFF
stimuli.
Here,
we
show
that
behavioral
responses
ON
stimuli
also
utilize
a
gain,
and
ON-motion
depends
on
inputs
from
three
first-order
interneurons
L1,
L2,
L3.
Each
these
neurons
encodes
differently
distributes
information
asymmetrically
across
both
contrast-selective
pathways.
Behavioral
rely
luminance-based
correction
provided
by
L1
L3,
wherein
supports
linearly,
L3
non-linearly
amplifies
dim
Therefore,
are
specific
pathways
but
lamina
serves
as
separate
layer
distinct
support
varying
conditions.
Current Biology,
Год журнала:
2023,
Номер
33(13), С. 2632 - 2645.e6
Опубликована: Июнь 6, 2023
Animals
navigating
in
natural
environments
must
handle
vast
changes
their
sensory
input.
Visual
systems,
for
example,
luminance
at
many
timescales,
from
slow
across
the
day
to
rapid
during
active
behavior.
To
maintain
luminance-invariant
perception,
visual
systems
adapt
sensitivity
changing
different
timescales.
We
demonstrate
that
gain
control
photoreceptors
alone
is
insufficient
explain
invariance
both
fast
and
timescales
reveal
algorithms
adjust
past
fly
eye.
combined
imaging
behavioral
experiments
with
computational
modeling
show
downstream
of
photoreceptors,
circuitry
taking
input
single
luminance-sensitive
neuron
type
L3
implements
This
computation
bidirectional
it
prevents
underestimation
contrasts
low
overestimation
high
luminance.
An
algorithmic
model
disentangles
these
multifaceted
contributions
shows
occurs
The
a
nonlinear
interaction
contrast
achieve
correction
dark-sensitive
channel
improve
detection
dim
stimuli
Together,
our
work
demonstrates
how
neuronal
performs
diverse
computations
implement
multiple
are
together
important
navigation
environments.
Abstract
The
compound
eyes
of
insects
exhibit
stunning
variation
in
size,
structure,
and
function,
which
has
allowed
these
animals
to
use
their
vision
adapt
a
huge
range
different
environments
lifestyles,
evolve
complex
behaviors.
Much
our
knowledge
eye
development
been
learned
from
Drosophila
,
while
visual
adaptations
behaviors
are
often
more
striking
better
understood
studies
other
insects.
However,
recent
insects,
including
bees,
beetles,
butterflies,
have
begun
address
this
gap
by
revealing
the
genetic
developmental
bases
differences
morphology
key
new
aspects
structure
function.
Furthermore,
technical
advances
facilitated
generation
high‐resolution
connectomic
data
insect
species
that
enhances
understanding
information
processing,
impact
changes
processes
on
evolution
behavior.
Here,
we
review
breakthroughs
propose
future
integrated
research
function
systems
within
among
represents
great
opportunity
understand
remarkable
diversification
vision.
Scientific Reports,
Год журнала:
2025,
Номер
15(1)
Опубликована: Март 14, 2025
Abstract
Sex-specific
behaviours
are
common
across
animals
and
often
associated
with
sexual
dimorphism
in
the
nervous
system.
Using
micro-CT
scanning
we
standardized
sex-specific
brain
atlases
tested
for
of
orchid
bee
Euglossa
dilemma
,
a
species
marked
sex
differences
social
behaviour,
mating
strategies
foraging.
Males
show
greater
investment
all
primary
visual
processing
neuropils
uniquely
integrated
central
complex,
evidenced
by
strong
positive
covariation.
This
suggests
that
males
invest
more
on
locomotor
control,
flight
stability
sky-compass
navigation
which
may
have
evolved
response
to
behaviours,
like
courtship
display.
In
contrast,
females
larger
mushroom
bodies
strongly
positively
covary
optic
lobes
increased
volume
Kenyon
cell
cluster,
implying
capabilities
associative
memory.
We
speculate
this
is
an
adaptation
nest-building
reliance
learning
landmarks
required
place
Our
study
provides
first
record
sexually
dimorphic
morphological
integration
insect,
approach
revealed
traits
lack
apparent
signal.
These
subtle
provide
further
evidence
causal
link
between
architecture
behaviour.