The diverse functional roles and regulation of neuronal gap junctions in the retina.

Abstract

Electrical synaptic transmission through gap junctions underlies direct and rapid neuronal communication in the CNS. The diversity of functional roles that electrical synapses have is perhaps best exemplified in the vertebrate retina, in which gap junctions are formed by each of the five major neuron types. These junctions are dynamically regulated by ambient illumination and by circadian rhythms acting through light-activated neuromodulators such as dopamine and nitric oxide, which in turn activate intracellular signalling pathways in the retina.The networks formed by electrically coupled neurons are plastic and reconfigurable, and those in the retina are positioned to play key and diverse parts in the transmission and processing of visual information at every retinal level.

Full text

Page 1

Rapid inter-neuronal communication is fundamental to
the integration and propagation of signals in the brain.
Although chemical synaptic transmission is the most
common mode of neuronal communication, demonstra-
tions of direct electrical signalling between neighbouring
neurons date back more than 50 years1,2. It was shown
subsequently that electrical synapses are specializations
of closely opposed membranes called gap junctions3,4,
but there remained a paucity of studies of electrical syn-
aptic transmission owing to the difficult and laborious
techniques that were needed to show the existence of
direct electrical interactions or to visualize gap junction
profiles. Over the past decade, however, there has been a
proliferation of investigations of electrical transmission
and gap junctions owing to new techniques, including
the use of gap junction-permeable tracers, patch record-
ings and molecular cloning. Possibly the most impor-
tant finding was the discovery of the connexin subunit
structure of gap junctions (FIG. 1). Connexins can be
identified through immunocytochemical techniques to
determine the distribution of gap junctions in the brain5.
Further, the use of mouse mutants in which selected gap
junctions are disrupted by targeting connexin genes has
become an important tool by which to characterize the
function of electrical synapses.
It is now clear that electrical synaptic transmission
through gap junctions is a common mode of inter cellular
communication in the CNS6. An elegant example of
such communication is provided by the vertebrate retina,
in which each of the five major neuron types is coupled
by gap junctions that express a number of different con-
nexin proteins5,7. In many cases the coupling strength is
regulated by ambient illumination or circadian rhythms
acting through neuromodulators8–12. Gap junctions’ broad
distribution, subunit makeup and regulatory pathways
suggest that they have a number of key and diverse func-
tional roles relating to neural processing of images. In this
regard the retina serves as arguably the best model system
in which to study the functional roles of neuronal gap
junctions and electrical transmission in the CNS. In this
Review we describe seven distinct functional roles that
have been elucidated for retinal gap junctions, related not
only to the propagation of signals, but also to the encod-
ing of specific visual information. These findings indicate
that electrical synapses play a complex part in neural sig-
nalling in the brain, with their regulation and plasticity
rivalling those described for chemical transmission.
Structure and regulation of gap junctions
Gap junctions, the morphological substrate of electrical
synapses, are composed of two hemichannels, or con-
nexons, that link across an extracellular space of 2–4 nm
(FIG. 1a). They form a channel that directly communicates
with the cytoplasm of neighbouring cells, providing an
intercellular pathway for the diffusion of molecules up
to 1,000 Da. Hemichannels are in turn composed of
six transmembrane protein subunits called connexins,
approximately 20 different isoforms of which have been
characterized in humans and mice7,13. A hemichannel
can be composed entirely of one connexin (homomeric)
Departments of Physiology &
Neuroscience and
Ophthalmology, New York
University School of Medicine,
550 First Avenue, New York,
New York 10016, USA.
Correspondence to S.A.B.
e‑mail: stewart.bloomfield@
nyumc.org
doi:10.1038/nrn2636
Published online 3 June 2009
The diverse functional roles and
regulation of neuronal gap junctions
in the retina
Stewart A. Bloomfield and Béla Völgyi
Abstract | Electrical synaptic transmission through gap junctions underlies direct and rapid
neuronal communication in the CNS. The diversity of functional roles that electrical synapses
have is perhaps best exemplified in the vertebrate retina, in which gap junctions are formed
by each of the five major neuron types. These junctions are dynamically regulated by ambient
illumination and by circadian rhythms acting through light-activated neuromodulators such
as dopamine and nitric oxide, which in turn activate intracellular signalling pathways in the
retina. The networks formed by electrically coupled neurons are plastic and reconfigurable,
and those in the retina are positioned to play key and diverse parts in the transmission and
processing of visual information at every retinal level.
REVIEWS
NATuRe RevIewS | NeuroscieNce
vOlume 10 | july 2009 | 495
© 2009 Macmillan Publishers Limited. All rights reserved

Page 2

20 nm
Nature Reviews | Neuroscience
Connexon type Homomeric
Channel type
HomotypicHomotypic
Homomeric
Heterotypic
Heteromeric
Heterotypic
Heteromeric
Connexon
a
b
2– 4 nm
Ions and
small molecules
Connexins
Intracellular
Extracellular
Membrane
M1
M2
M3
M4
E1
E2
C
N
Cytoplasm
of cell A
Extracellular
space
Cytoplasm
of cell B
Amacrine cell
An interneuron located in the
inner plexiform layer of
the retina, at the level where
bipolar cells and ganglion cells
synapse.
or can be composed of different connexin types (hetero-
meric), although it seems that only certain connexins can
combine to form functional hemichannels. In addition,
a gap junction can be composed of two hemichannels
with the same (homotypic) or different (heterotypic)
connexin makeup.
Similar to ion channels, the conductance of gap junc-
tions is regulated by a number of physiological factors
and agents. most gap junction connexins undergo post-
translational modification by phosphorylation — mainly
of serine amino acids found in their carboxyl termini
and intracellular loop regions14,15 (FIG. 1b). Indeed, con-
nexins are targeted by numerous protein kinases and
the resulting phosphorylation has been implicated in a
broad range of regulatory steps, including trafficking,
assembly and disassembly, and conductance gating of
gap junctions. These protein kinases are modulated by
a number of factors, including Ca2+–calmodulin, cyclic
AmP, cyclic GmP and neuromodulators.
Importantly, the light-activated neuromodulators
dopamine and nitric oxide (NO), which are released by
different amacrine cell subtypes16,17, activate a number
of intracellular pathways involving cAmP- and cGmP-
dependent protein kinases. This results in the phospho-
rylation or dephosphorylation of gap junction connexins,
altering the conductance of the gap junctions to ionic cur-
rents18–21 (FIG. 2). This modulation varies across the popu-
lation of retinal interneurons, such that the conductance
of some gap junctions is increased by light whereas that of
others is decreased10,11,22–26. In fact, light can increase or
decrease the conductance of gap junctions in the same
neuron, depending on the level of brightness8,10,25,26.
It has long been known that increases in cytosolic Ca2+
can close gap junctions27,28. Recent studies indicate that
Ca2+ action on gap junction conductance is dependent
on calmodulin, which interacts directly with connexin
subunits19,20,29,30. Consistent with this idea, inhibition of
calmodulin prevents the Ca2+-mediated reduction in cell
coupling through actions on a number of different con-
nexins, including CX43 (also known as GjA1) and CX32
(also known as GjB1)30–32.
Subtle changes in intracellular pH can also alter the
electrical coupling between neurons33,34. Acidification
usually results in decreased conductance of certain
gap junctions, whereas alkalinization produces a
conductance increase; however, in sharp contrast
to other connexins, the conductance of CX36 (also
known as GjD2) gap junctions decreases after alka-
losis rather than acidosis35. Because neural activity
produces changes in intracellular pH36, the proton
sensitivity of gap junctions provides a mechanism for
activity-dependent modulation of coupling between
neighbouring neurons.
Finally, most mammalian gap junctions are sen-
sitive to the voltage difference across the junctional
membrane37,38, the parameters of this sensitivity vary-
ing according to the connexin expressed. This sensitiv-
ity is a second mechanism by which gap junctions can
undergo activity-dependent modulation. Interestingly,
single gap junctions show both voltage-dependent and
-independent currents, which reflects multiple conduct-
ance sub-states of the junctional channel39–41. Thus, gap
junction opening and closing are graded rather than
all-or-none.
Figure 1 | structure and molecular organization of gap junctions. a | Gap junctions are formed between the
opposing membranes of neighbouring cells. Hemichannels on each side dock to one another to form conductive channels
between the two cells. An extended field of these channels forms a gap junctional plaque. Each hemichannel, or
connexon, is comprised of 6 connexin protein subunits that are oriented perpendicular to the cells’ membranes to form a
central pore. This central pore serves as a conduit for ions and low-molecular-mass molecules of up to 1,000 Da.
Connexons can contain only one type of connexin subunit (homomeric connexons) or a mixture of different connexins
(heteromeric connexons). Gap junctional channels can consist of two of the same connexon (homotypic channels) or of
connexons with different subunit compositions (heterotypic channels). b | Connexin subunits are proteins that have four
transmembrane domains, two extracellular loops (E1 and E2) and one intracellular loop, as well as carboxyl and amino
termini in the cytoplasm. Although the four transmembrane domains (M1–M4) share a conservative sequence that is
important for docking in the cellular membrane, their cytoplasmic domains vary in length and amino acid sequence.
Regulation of the three-dimensional connexin structure, which underlies the opening and closing of gap junction
channels, is mediated at the cytoplasmic regions.
REVIEWS
496 | july 2009 | vOlume 10
www.nature.com/reviews/neuro
© 2009 Macmillan Publishers Limited. All rights reserved

Page 3

g
g
NO
NO
g
D1 receptor
D2/4 receptor
DA
AMPcAMP
+
–
PKA
P
P
+
HC–HC, AII–AII, AC–AC, AC–GCR–C, GC–GC, GC–ACHC–HC, AII–CB
DA
ACy
AMPcAMP
PKA
ACy
GMP cGMP
PKG
GCy
Ph
_
+
P
P
P
P
Connexin
Dopaminergic
amacrine cell
+
Light
+
Dopamine
NADPH/NOS
amacrine cell
?
Electrical coupling between cones
electrical coupling between cone photoreceptors was
first demonstrated in the turtle retina, where a current
applied extrinsically to a cone produced activity in neigh-
bouring (within 40 μm) cones42. The electrical coupling
is derived from gap junctions formed between cones, a
feature that is conserved across many species, includ-
ing primates43–47 (FIG. 3). The gap junctions are found at
the endings of fine processes — called telodendria —
that emerge from the base of cone pedicles and contact
adjacent cones. CX36 gap junction plaques are present
here48–50, but whether other connexins are incorporated
in cone–cone gap junctions is unclear.
Recent studies have shown that the gap junctions
between neighbouring mammalian cone photorecep-
tors have an effective conductance of a few hundred
picosiemens, allowing for efficient transfer of visual
signals51,52. However, such lateral interaction between
adjacent cones would be expected to introduce blur-
ring of visual signals. Indeed, calculations indicate that
electrical coupling between cones in the human fovea,
where sampling is extremely dense, results in a blurring
of visual signals and a reduction of acuity51. This calcu-
lated ‘neural blur’ due to cone coupling was subsequently
confirmed by human psychophysical experiments51.
what, then, is the benefit of coupling between cone
photoreceptors? Phototransduction is an inherently
noisy process owing to random photon absorptions
and fluctuations in signalling molecules and ion chan-
nel conductances in the signal processing cascade. The
inherent noise in each cone is independent of that of its
neighbours. By contrast, the vision-evoked activity of an
individual cone is correlated with that of its neighbours
owing to the reception of shared stimuli in a viewed
image and the scattering of light as it passes through the
cornea and lens. Coupling between cones sums the cor-
related visual signals and attentuates the asynchronous
noise. It has been calculated that the coupling between
foveal cones improves the signal-to-noise ratio of each
cone by nearly 80%51. Overall, cone coupling improves
the sensitivity and fidelity of visual signals, but at the
cost of some neural blurring of the image. However,
this neural blurring seems to be far narrower than
the blurring produced by the optics of the eye51. Thus, the
improvement in the signal-to-noise ratio of the cone
signals outweighs the small degradation in acuity. An
additional benefit of direct coupling between cones is
that it removes noise before any distortions occur in the
downstream retinal circuitry. Finally, psychophysical
experiments do not detect any changes in neural blur-
ring over a several-fold change in ambient illumination,
suggesting that, in contrast to other retinal gap junctions
(see below), those that couple cones are not affected
by light51.
The retinas of humans and certain Old world mon-
keys contain three types of cone photoreceptors, which
preferentially absorb long (red), medium (green) and
short (blue) wavelengths of light53. This trio of cone
types underlies colour perception in the visual system.
Although blue cones do not seem to have gap junctions,
coupling between red and green cones is indiscriminate,
Figure 2 | Neuromodulators affect gap junction conductances through
intracellular pathways. A summary of the intracellular pathways by which dopamine
(DA) and nitric oxide (NO) are thought to affect the conductance of retinal gap junctions.
DA release from dopaminergic amacrine cells is increased by light. For some retinal
neurons DA binds to D1 receptors (left panel), activating adenylate cyclase (ACy) and
increasing the concentration of cyclic AMP. This in turn activates cAMP-dependent
protein kinase (PKA), the phosphorylation of connexins and a reduction in the
conductance (g) of gap junctions. (However, it has recently been suggested that in AII
amacrine cells PKA activates a phosphatase (Ph) that dephosphorylates connexins and
thereby causes reduced gap junction conductance21.) This D1 receptor mechanism
occurs at gap junctions between horizontal cells (HC–HC)18,19, between AII amacrine
cells (AII–AII)21,23,134, between other amacrine cell subtypes (AC–AC)154 and at the
amacrine cell hemichannel of amacrine cell–ganglion cell gap junctions (AC–GC)154. DA
also binds to D2/4 receptors (middle panel), which reduces the activity of adenylate
cyclase, resulting in a reduction of cAMP levels. This reduces the activity of PKA, resulting
in increased gap junction conductance. This mechanism occurs at gap junctions
between rods and cones (R–C)12, between ganglion cells (GC–GC)154 and at the ganglion
cell hemichannel of ganglion cell–amacrine cell gap junctions (GC–AC)154. Light also
increases the release of NO from NADPH/nitric oxide synthase (NOS)-positive amacrine
cells (right panel). NO diffuses into retinal neurons and activates guanylate cyclase (GCy),
resulting in an increase in cGMP levels, activation of a cGMP-dependent protein kinase
(PKG), phosphorylation of connexins and reduced gap junction conductance. This
mechanism occurs at gap junctions between horizontal cells (HC–HC)11,103,104 and
between AII amacrine cells and ON cone bipolar cells (AII–CB)23.
REVIEWS
NATuRe RevIewS | NeuroscieNce
vOlume 10 | july 2009 | 497
© 2009 Macmillan Publishers Limited. All rights reserved

Page 4

Nature Reviews | Neuroscience
OFF
GC
GC
ON
C
R
GCL
ONL
OPL
INL
IPL
AII
AII
GC
HCCB RB
AII
CB
ON
ON
OFF
AC
CX36 or CX45 or ?
CX36
CX36
CX36
CX36 or
CX45
f AII–CB
CX57 or CX50
??CX36CX36CX36
?
a C–C
b R–C
c R–R
d HC–HC
g GC–GC, GC–AC
e AII–AII
Gap junction plaque
A collection of up to thousands
of single gap junction channels.
in that green cones are coupled to green cones, red cones
are coupled to red cones, and red cones are coupled to
green cones50,54. This indiscriminate coupling blurs the
spectral discrimination between cones, as their signals
partially mix. However, this spectral blurring is modest,
with colour discrimination reduced by only 20%54.
Further, red and green cones form regular but separate
mosaics, resulting in a high probability of homologous
coupling (between like cones) and a lower probability of
heterologous coupling (between unlike cones) across the
retina. This characteristic patchiness of red and green
cone mosaics further minimizes the spectral blurring
associated with electrical coupling55.
Electrical coupling between rods and cones
In mammals, rod and cone photoreceptors make chemi-
cal synapses with different bipolar cells, segregating their
signals into parallel retinal streams56,57. Although more
than ten subtypes of cone bipolar cells have been identi-
fied, only a single subtype of rod bipolar cell exists58,59.
This would seem to suggest that the rod circuitry in the
retina is relatively simple, but it is now clear that multiple
pathways transmit rod signals across the retina60 (FIG. 4).
In the primary rod pathway, rod photoreceptors pass sig-
nals to rod bipolar cells, which carry the signals radially
to the inner retina and contact AII amacrine cells. The AII
cells, in turn, form sign-inverting chemical synapses with
OFF cone bipolar cell axons and sign-conserving electrical
synapses with ON cone bipolar cell axons. Thus, the pri-
mary rod pathway ‘piggybacks’ onto the cone circuitry
in the inner retina so that its signals can reach the output
ganglion cells.
In addition, there are gap junctions between the axon
terminals of rod and cone photoreceptors61 (FIG. 3). This
rod–cone coupling provides an alternative, ‘second-
ary’ rod pathway for the transmission of scotopic sig-
nals (FIG. 4). In this pathway, rod signals are transmitted
directly to cones, then to cone bipolar cells and then to
ganglion cells. experimental evidence for the functional-
ity of this pathway includes the detection of rod-generated
signals in cone photoreceptors62,63 and the survival of
rod signals in ganglion cells after blockade of the pri-
mary rod pathway64,65. Human psychophysical studies
support the existence of multiple rod pathways66–69.
Physiological studies indicate that the functions of
the two rod pathways relate to differences in the sen-
sitivities of the signals that they carry as well as in the
ganglion cells that they target. Studies using retinas
from CX36-knockout mice, in which the rod–cone gap
junctions are disrupted, indicate that the signals carried
by the primary rod pathway are the most sensitive to
stimulus brightness65,70. By contrast, rod signals car-
ried by the secondary pathway are approximately one
log unit less sensitive. The two rod pathways thus have
largely different operating ranges. Although the primary
rod pathway has high sensitivity, its nonlinear synaptic
transfer properties result in a narrow operating range71.
The secondary rod pathway thus provides a route for
scotopic signals to reach the ganglion cells when the
primary pathway is saturated72.
A second difference between the primary and sec-
ondary rod pathways relates to the ganglion cell sub-
types that they target. In rabbits the two pathways target
different ganglion cell subtypes64, whereas in mice they
converge onto most of the OFF ganglion cells73. A recent
study65 found both segregated and convergent signalling
Figure 3 | Gap junctions expressed by retinal neurons. This schematic shows seven
examples of electrical coupling, the different functions of which are detailed in the main
text. The coloured ovals represent gap junction hemichannels. The solid and dotted
arrows represent excitatory and inhibitory chemical synapses, respectively. a | Both
hemichannels of the gap junctions that couple neighbouring cones (C) express Cx36
(ReFs 48,49). b | In rod (R)–cone gap junctions, only the hemichannel on the cone side
contains Cx36; the connexin on the rod side remains unknown48,49,65,70. c | The type of
connexin in rod–rod gap junctions is also unknown. d | Horizontal cell (HC) dendrites are
extensively coupled. In mammals, axonless horizontal cells express Cx50 (ReF. 166)
whereas axon-bearing horizontal cells express Cx57 (ReFs 86,87). e,f | AII amacrine cells
(AII) form two types of gap junction. Gap junctions between AII cells seem to be
homotypic and comprised of homomeric hemichannels containing Cx36 (e)70,121,122. By
contrast, gap junctions between AII amacrine cells and ON cone bipolar cells (CB) can be
homotypic or heterotypic, with the AII cell hemichannels containing Cx36 and the cone
bipolar cell hemichannel containing either Cx36 or Cx45 (ReFs 70,121–125). g | Ganglion
cells (GC) are extensively coupled to each other and/or to neighbouring amacrine cells
(AC). To date, ganglion cell gap junctions have been reported to contain Cx36 or Cx45
(ReFs 167–169). GCL; ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform
layer; ONL, outer nuclear layer; OPL, outer plexiform layer; RB, rod bipolar cell.
REVIEWS
498 | july 2009 | vOlume 10
www.nature.com/reviews/neuro
© 2009 Macmillan Publishers Limited. All rights reserved

Page 5

AII
AII
C
R
ON
ON
ON
ON
OFF
R
R
OFF
OFF
OFF
ONON
ON
ON
ON
OFF
OFF
C
GC
GC
GC
GC
GC
GC
CB
CB
CB
CB
CB
CB
RB
RB
RB
AII
AII
AII
BC
C
AII
AII
AII
AII
AII
AII
AII
AII
b
c
a
Bipolar cell
A cell that receives information
formed by the interactions of
horizontal cells with cone or
rod photoreceptors and
conveys it to the inner retina.
ON (cone or rod) bipolar cells
respond to increases in
intensity, whereas OFF cone
bipolar cells respond to
decreases in intensity.
AII amacrine cell
A subtype of retinal amacrine
cell with a small dendritic field
that conveys the rod signal to
cone bipolar cells.
Sign-inverting synapse
A synapse that inverts the
polarity of the signal passed
from the pre- to the
postsynaptic neuron.
Sign-conserving synapse
A synapse that preserves the
polarity of the signal passed
from the pre- to the
postsynaptic neuron.
Ganglion cells
The output neurons of the
retina, the axons of which form
the optic nerve. ON ganglion
cells respond to increases in
light intensity, whereas OFF
ganglion cells respond to
decreases in light intensity.
Scotopic
Relating to dim ambient light
conditions under which only
rod photoreceptors are active.
of the rod pathways to single ganglion cells. This indi-
cates not only the complexity of the pathways by which
rod signals are propagated in the retina, but also the
complex organization of these signals’ transmission to
central visual areas.
The conductance of the gap junctions between rod
and cone photoreceptors is regulated by a circadian
clock in the retina12. In this scheme, the circadian clock
increases dopamine release during the day, which acti-
vates D2/D4 dopamine receptors on rods and cones.
This in turn lowers intracellular cAmP and protein
kinase A (PKA) activity, which reduces the conductance
of rod–cone gap junctions (FIG. 2). By contrast, reduced
dopamine release at night allows for robust electrical
coupling between rods and cones. This circadian control
ensures that the secondary rod pathway is operational
in night time conditions, to facilitate the detection of
dim objects. Further, the reduction of rod–cone coupling
during the day ensures that cone signals are not passed
to a saturated — and thus non-operational — network of
rods, which would attenuate the signals created in bright
conditions.
Electrical coupling between rods
The contribution of rod–cone gap junctions to a second-
ary rod pathway was challenged by a study of mice that
were genetically engineered to lack cone photorecep-
tors73. In these animals, rod responses in OFF ganglion
cells occurred even when the primary rod pathway was
blocked. As the mice had no rod–cone gap junctions, it
was suggested that the alternative rod pathway is sub-
served by direct chemical synapses between rods and
OFF bipolar cells rather than by rod–cone gap junc-
tions. Although such contacts were thought not to exist
in mammals45,56,74, subsequent studies described them in
a number of species47,75–77.
Indeed, there is now functional evidence that direct
chemical synapses between rods and OFF bipolar cells
form a third rod pathway in the mammalian retina65
(FIGs 3,4). Interestingly, only one in five rods in the
mouse retina forms a chemical synapse with an OFF
bipolar cell, suggesting that this pathway may play a
relatively limited part in scotopic signal transmission47.
However, gap junctions exist between the axon terminals
of mammalian rod photoreceptors, including those of
primates47,52. It has been proposed that rod–rod coupling
pools the scotopic signals at the photoreceptor level for
conveyance to the ganglion cells through the third path-
way. This third rod pathway might thus be useful at dusk
and dawn, when more photons are available than during
starlight, and the pooled signal might thereby efficiently
encode faintly backlit objects47.
Physiological evidence for this third pathway
includes the finding of rod signals in ganglion cells
after disruption of the primary and secondary rod path-
ways65. Further, these signals showed scotopic sensitivi-
ties that were lower than those conveyed by the other
two rod pathways, supporting the hypothesis that the third
rod pathway subserves rod vision during dusk and dawn.
The suggestion that the third rod pathway has a unique
function is further supported by the finding that most of
Figure 4 | The three rod pathways in the mammalian
retina. a | The primary rod pathway involves electrical
synapses between AII amacrine cells (AII), and between AII
cells and cone bipolar cells (CB). In this pathway, signals are
transmitted from rods (R) to rod bipolar cells (RB) and
subsequently to AII cells. AII cells make sign-conserving
electrical synapses with ON cone bipolar cells and
sign-inverting inhibitory chemical synapses with OFF cone
bipolar cells. In turn, the ON and OFF cone bipolar cells
make excitatory chemical synapses with ON and OFF
ganglion cells (GC), respectively. b | The secondary rod
pathway involves electrical synapses between rod and
cone photoreceptors (C). In this pathway, rod signals are
transmitted directly from rods to cone photoreceptors
through interconnecting gap junctions. The rod signals are
then relayed to ON and OFF cone bipolar cells, which carry
the signals to ganglion cells in the inner retina. c | The
tertiary rod pathway involves electrical synapses between
rods only. In this pathway, rods make direct chemical
synapses with a subset of OFF bipolar cells (BC), which
transmit the signals to some OFF ganglion cells. This
pathway does not seem to have a counterpart in the ON
circuitry. The solid and dotted arrows represent excitatory
and inhibitory chemical synapses, respectively. Figure is
modified, with permission, from ReF. 65  (2006) Society
for Neuroscience.
REVIEWS
NATuRe RevIewS | NeuroscieNce
vOlume 10 | july 2009 | 499
© 2009 Macmillan Publishers Limited. All rights reserved

Page 6

Horizontal cells
Retinal neurons that form a
network just beneath the
photoreceptors that is
responsible for averaging visual
activity over space and time,
which is important for contrast
signalling.
Receptive field
A dynamic area of the retina in
which stimulus presentation
leads to the response of a
particular cell.
Ephaptic
Relating to the direct electrical
interaction between
neighbouring neurons,
mediated by current flow
through the extracellular space
that separates them.
the ganglion cells targeted by this pathway do not receive
input from the other two rod pathways65. Interestingly,
the signals that are transmitted by the third rod pathway
survive in the CX36-knockout mouse retina, in which
both AII–bipolar cell and rod–cone gap junctions are
disrupted, suggesting that pooling of signals by rod–rod
coupling is intact in these mice65 and that CX36 is not
expressed by rod photoreceptors.
Electrical coupling of horizontal cells
Horizontal cells are second-order retinal neurons that
extend processes laterally in the outer plexiform layer,
where they contact both rod and cone photorecep-
tors78. In fact, yamada and Ishikawa79 first described
“fused membrane structures” between neighbouring
horizontal cells ~5 years before Goodenough and Revel3
coined the term gap junction. Subsequently, there have
been numerous demonstrations of coupling between
horizontal cells using dyes and tracers16,80–83 (FIG. 3).
The extensive gap junctional coupling between
horizontal cells results in their characteristically large
receptive fields, which can be 25 times the size of an
individual cell’s dendritic arbor81,84,85. This indicates an
extremely efficient lateral spread of visual signals through
electrical synapses81. Indeed, the deletion of horizontal
cell gap junctions in the retina of CX57 (also known
as GjA10)-knockout mice86 resulted in a significant
reduction in the cells’ receptive field sizes87.
Bipolar cells, the radially projecting second-order
neurons in the retina, have antagonistic centre–surround
receptive fields. This results in light-evoked responses
that signal the difference in illuminance of an attended
object (the centre) and that of the background envi-
ronment (the surround)88,89. Bipolar cells are thus con-
trast detectors. There is considerable evidence that the
large receptive fields of horizontal cells provide the sur-
round receptive field of bipolar cells and the output gan-
glion cells90–93. It remains unclear whether bipolar cell
surrounds are created by direct inhibitory horizontal
cell synapses onto bipolar cells or by feedback inhibi-
tory synapses of horizontal cells onto cone photorecep-
tors. To some controversy, it has been suggested that the
feedback inhibition is not chemical, but ephaptic, and
that electrical charge thus moves across horizontal cell
hemichannels into the extracellular space to modify
the activity of cone photoreceptors94. Regardless of the
above, horizontal cell coupling serves to form a homo-
geneous electrical syncytium by which a signal of the
average ambient background illuminance is created. In
essence, gap junctional coupling between horizontal
cells is thought to form the initial mechanism for con-
trast detection in the visual system, although this idea
has recently been challenged95.
There are now converging sources of evidence that
horizontal cell coupling is under dynamic regulation.
For example, dopamine decreases the gap junctional
conductance of horizontal cells and, acting through the
intracellular messenger cAmP, produces a concomitant
reduction in receptive field size22,96,97 (FIG. 2). Dopamine
modifies gap junctional conductance by affecting
junctional densities and/or mean open conductance
times98–101. Nitric oxide also alters the electrical coupling
of horizontal cells, acting through the intracellular mes-
senger cGmP19,102–104 (FIG. 2). In rabbits, exposure to either
dopamine105 or the nitric oxide substrate l-arginine106
reduces electrical coupling between horizontal cells.
Both the extracellular levels of dopamine and the
production and release of nitric oxide are modulated
by changes in illuminance, suggesting that horizontal
cell coupling can be modulated by light17,107–110. Indeed,
both prolonged darkness and light adaptation have been
shown to uncouple horizontal cells in a number of ver-
tebrate retinas10,24,25,111–113. There is a complex triphasic
relationship between the conductance of horizontal
cell gap junctions and illuminance, whereby coupling
is poor under both dark- and light-adapted condi-
tions and strong only under intermediate illuminance
levels10,25.
As expected, light-induced changes in horizontal
cell coupling affect the centre–surround receptive field
organization of retinal neurons and contrast signalling.
Prolonged darkness attenuates the surround receptive
fields of ganglion cells, which is thought to increase
sensitivity at dim light at the expense of contrast detec-
tion114–116. likewise, a reduction of horizontal cell cou-
pling under bright light results in smaller surround
receptive fields and thus more local contrast detection,
which is consistent with higher acuity. Overall, the
light-induced modulation of horizontal cell coupling
optimizes the extraction of important cues in natural
images, including contrasts and edges117,118.
AII amacrine cell gap junctions
Rod bipolar cells do not make contacts directly with
ganglion cells, but instead form synapses with the inter-
mediary AII amacrine cells (FIGs 3,4). These AII cells
express two types of gap junctions: neighbouring AII
cells form gap junctions with one another and with the
axon terminals of cone bipolar cells45,119,120. As described
earlier, these latter contacts form a conduit for rod sig-
nals to use the cone pathways before reaching the output
ganglion cells.
Plaques of CX36 are found at dendritic crossings
of AII cells, suggesting that AII–AII gap junctions
are homotypic70,121. However, both CX36 and CX45 are
expressed at cone bipolar cell hemichannels, indicating
that at least some of the AII cell–cone bipolar cell gap
junctions might be heterotypic70,121–125, although the exist-
ence of CX45–CX36 heterotypic junctions has been ques-
tioned126. Such variations in composition could explain the
different conductances and pharmacologies of AII–AII
and AII cell–cone bipolar cell gap junctions127,128.
AII cell–cone bipolar cell gap junctions. Deletion of CX36
disrupts both the AII–bipolar cell and the rod–cone gap
junctions, resulting in the loss of signalling in the pri-
mary and secondary rod pathways, respectively65,70,129. As
a result, all rod-driven input to ON ganglion cells is lost.
These results not only show that electrical synapses play
an essential part in the rod pathways, but also provide
the first demonstration that gap junctions are obligatory
elements in a defined circuit in the CNS.
REVIEWS
500 | july 2009 | vOlume 10
www.nature.com/reviews/neuro
© 2009 Macmillan Publishers Limited. All rights reserved

Page 7

Nature Reviews | Neuroscience
Starlight
Twilight
Daylight
AII amacrine cells
Mesopic
Relating to the ambient light
condition under which both rod
and cone photoreceptors are
active.
Paired recordings indicate that the AII–cone bipolar
cell gap junctions can pass current bidirectionally128:
rod signals are passed from AII amacrine cells to cone
bipolar cells under dark-adapted conditions, but there
is a reversal in the direction of signal flow under light-
adapted conditions and, as a result, cone signals move
into the network of coupled AII cells130. The conductances
of the AII–cone bipolar cell gap junctions are modulated
by nitric oxide acting through a cGmP cascade8,23 (FIG. 2).
However, no light-induced changes in the conductances
of these gap junctions have yet been found8.
AII–AII amacrine cell gap junctions. Based on com-
putational models, Smith and vardi131 speculated that
AII–AII cell coupling serves to sum synchronous signals
and subtract asynchronous noise, thereby preserving the
high sensitivity of signals carried by the primary rod path-
way. Consistent with this idea, the intensity response pro-
files of the most sensitive ganglion cells in the retina show
a rightward shift when electrical synapses between AII
amacrine cells are deleted in CX36-knockout mice65.
AII–AII cell coupling thus underlies a unique func-
tion of the primary rod pathway: maintaining the high
sensitivity of rod signals arriving in the inner retina.
The dendrites of the dopaminergic subtype of
amacrine cell form a dense plexus that surrounds
the AII amacrine cells132, and dopamine modulates
the conductance of AII–AII cell gap junctions through
a cAmP-mediated PKA cascade21,127,133,134 (FIG. 2). As
dopamine release is modulated by light16, it is not surpris-
ing that changes in light adaptation are found to affect the
coupling between AII amacrine cells in a triphasic man-
ner, as described earlier for horizontal cells8 (FIG. 5). Thus,
dark-adapted AII cells are coupled in relatively small
groups and show relatively small receptive fields, but expo-
sure to dim background light increases both parameters
approximately sevenfold130. Further light adaptation
brings about a decrease in coupling to levels similar to
those seen in dark-adapted retinas. These robust con-
comitant changes in tracer coupling and receptive field
size of AII cells indicate a clear modulation of AII–AII
cell coupling under different adaptational states.
The light-induced changes in AII–AII cell coupling
enable these cells, as vital elements in the rod pathway,
to remain responsive throughout the entire scotopic and
mesopic range26. In this scheme (FIG. 5), dark adaptation
is analogous to starlight conditions, under which rods
only sporadically absorb photons of light. Accordingly,
the AII cells are relatively uncoupled in the sense
that the few correlated signals are carried by only closely
neighbouring AII cells; extensive coupling would
dissipate signals into a largely inactive network, thereby
attenuating them. Presentation of dim background light,
analogous to twilight conditions, brings about a greater
than tenfold increase in AII cell coupling. This increased
coupling allows summation of synchronous activity over
a wider network of active AII cells, thereby preserving sig-
nal fidelity. This transition in coupling suggests that there
are two basic operating states for AII cells under scotopic
or mesopic light conditions: first, the ability to respond
to single photon events; and second, summing signals
over a relatively large area to augment synchronized
events above the background noise.
Electrical coupling between ganglion cells
Possibly the most complex electrical synaptic networks
occur in the inner retina, where most ganglion cell sub-
types show gap junction-mediated tracer coupling with
neighbouring ganglion and/or amacrine cells80,135–137
(FIGs 3,6). Interestingly, coupling between different sub-
types of ganglion cells has never been reported, suggest-
ing that ganglion cell gap junctions subserve separate
electrical networks in the inner retina.
At first glance, the extensive coupling displayed by
ganglion cells seems counterintuitive, in that it suggests
that there is lateral intercellular propagation of signals
across the inner plexiform layer. This would result in
a reduction of the visual acuity of neuronal signals just
as they exit the retina. However, the receptive fields of
ganglion cells approximate the extent of their dendritic
arbors, irrespective of the extent of tracer coupling138.
This is because the conductance of ganglion cell gap
junctions is low, which restricts the movement of both
electrical current and tracers. Thus, ganglion cell gap
junctions probably underlie local operations rather than
the global processing that is exemplified by the exten-
sive electrical syncytia formed by horizontal cells in the
outer retina.
Figure 5 | electrical coupling between Aii amacrine cells is regulated by
background light conditions. The extent of coupling between AII amacrine cells under
three different background light conditions, mimicking starlight, twilight and daylight34.
Each group of red symbols provides the average extent of Neurobiotin tracer coupling of
rabbit AII amacrine cells. Under dim starlight conditions, the conductance of the gap
junctions connecting neighbouring AII amacrine cells is relatively low, and so tracer
movement is limited to only a few cells. As the ambient background light increases to
twilight conditions, the conductance of the gap junctions increases and so the electrical
syncytium of the coupled cells enlarges dramatically. Under bright daylight conditions,
the conductance of the gap junctions is once again reduced and electrical
communication is limited to a small group of cells. This triphasic modulation of AII cell
coupling to light is similar to that seen for horizontal cells and ensures that the fidelity of
the signals carried by the primary rod pathway is maintained under different scotopic
conditions152. Under bright background conditions, the limited coupling between AII
amacrine cells limits lateral interactions that would blur the image, thereby maintaining
the high acuity that is essential for daylight vision144. Figure is modified, with permission,
from ReF. 26  (2004) Elsevier.
REVIEWS
NATuRe RevIewS | NeuroscieNce
vOlume 10 | july 2009 | 501
© 2009 Macmillan Publishers Limited. All rights reserved

Page 8

a
b
Time
0
Spike frequency
Spike Frequency
Time
0
AC
GCGC
GCGC
Accessory optic system
A visuosensory pathway with a
direct retinal input to the
midbrain.
Optokinetic response
A compensatory eye
movement that stabilizes an
image on the retina during slow
head rotation.
Beginning with the seminal work of mastronarde139–141,
a number of studies have shown that ganglion cell cou-
pling underlies the coherent firing of neighbouring cells,
ranging from broad correlations spanning several tens of
milliseconds to finely tuned spike synchrony with 1–3 ms
latencies142–146. This concerted firing accounts for up to
one-half of retinal spike activity, suggesting that electri-
cal coupling plays an important part in encoding visual
information147,148.
The homologous ganglion cell–ganglion cell and the
heterologous ganglion cell–amacrine cell coupling that
are found in the inner retina are thought to produce dif-
ferent patterns of concerted activity in neighbouring gan-
glion cells143–145 (FIG. 6). Direct ganglion cell–ganglion cell
coupling is thought to mediate a fast (<2 ms) and recipro-
cal excitation that is reflected by prominent dual peaks
in cross-correlograms of the spike activity of ganglion
cell neighbours. By contrast, ganglion cells that are cou-
pled indirectly through gap junctions with a common
cohort of amacrine cells produce cross-correlograms
with a narrow, unimodal contour. This latter correlation
profile, which accounts for most of the concerted spike
activity in the retina148, probably reflects electrical synap-
tic inputs from common amacrine cells that give rise to
synchronous spikes. Consistent with this idea, ganglion
cell–amacrine cell coupling is the most common pattern
of gap junctional coupling in the inner retina136,137.
Correlated firing is thought to compress informa-
tion for efficient transmission and thereby increase the
bandwidth of the optic nerve149. In this scheme, synchro-
nous activity forms a separate stream of information to
the brain, in addition to the asynchronous signals from
individual ganglion cells. Concerted spike activity is
also thought to enhance the saliency of visual signals
by increasing the temporal summation at central tar-
gets150–152. In this way, concerted ganglion cell activity
might provide the temporal precision by which retinal
signals are reliably transmitted to central targets153.
The regulation of ganglion cell gap junctions is pres-
ently unclear. However, alpha ganglion cell coupling is
dramatically increased under light-adapted conditions,
resulting in an increase in concerted spike activity145.
Interestingly, this increase in coupling seems to be
generated by a light-induced elevation of extracellular
dopamine that binds to D2/D4 receptors (FIG. 2). It has
been proposed that activated D2/D4 receptors initiate an
intracellular cascade involving the attenuation of adeny-
lyl cyclase activity, thereby inhibiting PKA and reducing
phosphorylation of alpha ganglion cell connexins154. The
light-induced regulation of ganglion cell gap junction
conductance is thus opposite to that described above for
horizontal cell and for AII amacrine cell coupling10,130.
Intercellular communication through gap junc-
tions also plays an important part in the development
of neuronal circuits, including cell differentiation and
pathfinding155,156. Gap junctions are thought to have a
crucial role in regulating concerted ganglion cell activity
in the developing retina, which is seen as spontaneous
waves of depolarization157,158. These spontaneous waves
in turn are crucial to the refinement of retinal-thalamic
and intraretinal connections157,158.
ON direction-selective ganglion cells
As we described in the previous section, the general
functional role of ganglion cell coupling is to synchro-
nize the activities of neighbouring cells. A recent study
of ON direction-selective (DS) ganglion cells revealed a
specific role for synchronization of neighbouring cells
in encoding the direction of stimulus motion159. The DS
ganglion cells are a unique subtype of ganglion cell that
respond vigorously to stimulus movement in a preferred
direction but weakly to movement in the opposite, or
null, direction160. The ON subtype of DS cells project to
the accessory optic system161–163, where their directional
signals underlie the optokinetic response164.
Neighbouring ON DS ganglion cells are coupled
indirectly through gap junctions with a subtype of
polyaxonal amacrine cell159 (FIG. 7a) and show both
Figure 6 | Ganglion cell gap junctions underlie two patterns of concerted spike
activity. a | Direct electrical coupling between neighbouring ganglion cells (GC) (left
panel, red and brown) results in concerted activity. Paired recordings from coupled
ganglion cell neighbours generate a cross-correlogram function with a distinct pattern
consisting of two peaks with short latencies of approximately 1–3 ms (right panel). These
two peaks reflect reciprocal interactions in which a spike in one cell gives rise to a spike
in a coupled neighbour. b | Indirect electrical coupling between ganglion cells (red and
brown) through gap junctions with a mutual intermediary amacrine cell (AC) (blue) (left
panel) results in synchronous activity. Paired recordings from neighbouring ganglion cells
give rise to a cross-correlogram with a peak at time 0, indicative of extensive
synchronous spike activity (right panel). This pattern of synchrony reflects spiking in an
amacrine cell producing spike activity in neighbouring ganglion cells with identical
temporal properties.
REVIEWS
502 | july 2009 | vOlume 10
www.nature.com/reviews/neuro
© 2009 Macmillan Publishers Limited. All rights reserved

Page 9

–1
0
1
Time (s)
0
Spikes per s
–101
Time (s)
0
15
Spikes per s
Preferred
Null
50
Preferred
Cell 1
Cell 2
Cell 1
Cell 2
Null
GC
AC
AC axon
AC
dendrite
ab
correlated and synchronous activity in response to
moving and stationary light stimuli. The correlated
activity, which is reflected as a broad profile in the cross-
correlogram, is unaffected by disruption of gap junctions,
suggesting that it results from common inputs derived
from the conventional excitatory chemical synapses
between the ON DS ganglion cells and bipolar cells. By
contrast, gap junction blockade eliminates synchronous
spiking between ON DS cell neighbours, indicating that
the synchronous spiking is due to electrical coupling.
As described above, this type of ganglion cell spike syn-
chrony is thought to result from indirect coupling to
common amacrine cells. most interesting is the finding
that the degree of spike synchrony between neighbour-
ing ON DS cells was dramatically affected by the direc-
tion of stimulus movement159 (FIG. 7b). Synchronized
spiking was evoked by stimulus movement in all but
the null direction, which produced a dramatic desyn-
chronization of activity. Although previous studies have
shown that changes in the state of dark–light adapta-
tion can globally modulate the coupling between retinal
neurons8,10,111, this finding indicates that specific light
stimulation can also effectively modulate the coupling
between neurons to alter their response activity.
As described in the previous section, a number of
global functions have been attributed to spike synchrony,
including the enhanced saliency of visual signals or the
increased temporal precision by which retinal signals are
reliably transmitted to thalamic and cortical areas147,150,153.
By contrast, the synchronous spiking of neighbour-
ing ON DS ganglion cells seems to have a more clearly
defined and focused role, namely to encode the direction
of stimulus motion. Interestingly, it is not the synchrony
but the desynchronization to stimulus movement in the
null direction that seems to be a key component of how
ON DS cells signal direction of movement. Overall, these
results show that a dynamic modulation of the intercel-
lular current transmitted through gap junctions forms
a mechanism by which cell groups can encode specific
information.
Conclusions and future directions
like at other CNS loci, it is now clear that electrical
coupling through gap junctions is ubiquitous in the
vertebrate retina. Gap junctions and their subunit con-
nexin proteins are not only widely distributed in both
synaptic layers, but converging evidence suggests that
they are also expressed by most of the ~60 subtypes of
retinal neurons. The finding that gap junctional con-
ductances are affected by neuromodulators that mediate
the changes in light adaptation and circadian rhythms
indicates that electrical synaptic transmission forms a
complex and dynamic mode of cellular communication.
Despite the fact that we are just beginning to elucidate
Figure 7 | synchronous activity of coupled oN direction-selective ganglion cells encodes the direction of
moving light stimuli. a | An ON direction-selective (DS) ganglion cell (GC) in the rabbit retina labelled with Neurobiotin.
This ganglion cell is coupled to an array of polyaxonal amacrine cells (AC), the somata of which lie in the inner nuclear
layer and are thus out of focus. However, the dendritic and axonal processes of the coupled polyaxonal amacrine cells are
well labelled. b | An illustration of simultaneous extracellular recordings from two neighbouring ON DS ganglion cells,
showing the responses to a rectangular slit of light moving in the preferred (top panels) and opposite (null) (bottom panels)
directions. When the stimulus moves in the preferred direction, most of the light-evoked spikes are synchronized (red). By
contrast, the movement in the opposite, null direction results in a complete loss of spike synchrony. The modulation of
synchronous activity can be visualized in the cross-correlogram functions of the simultaneous paired recordings. When
the slit is moved in the preferred direction the correlogram shows a large peak at time 0, corresponding to synchronized
spiking. However, the peak is lost when the stimulus is moved in the null direction. The change in spike synchrony is due to
a modulation of the intercellular current that flows through gap junctions between ON DS ganglion cells and polyaxonal
amacrine cells. The change in response synchrony modifies the summation of the signal at central targets in the accessory
optic system, thereby signalling the direction of stimulus motion to the brain. Figure is modified, with permission, from
ReF. 159  (2006) Society for Neuroscience.
REVIEWS
NATuRe RevIewS | NeuroscieNce
vOlume 10 | july 2009 | 503
© 2009 Macmillan Publishers Limited. All rights reserved

Page 10

Pannexins
Proteins expressed in both
vertebrates and invertebrates
that can form intercellular gap
junction channels.They are
genetically related to the
invertebrate innexin family but
are not related to connexins.
the types of connexins (and pannexins) that are expressed
in the retina, it is already clear that gap junctions have a
wide variety of functions in propagating and integrating
signals. unlike studies carried out in other parts of the
CNS that often assign generic functions to electrical cou-
pling, such as increased spike synchrony, studies in the
retina have been able to detail the specific roles of indi-
vidual gap junctions in visual processing. This makes the
retina arguably the best model system in which to study
the roles of electrical synaptic transmission in the CNS.
One way to determine the function of a gap junc-
tion is to remove it, and so pharmacological blockers
and connexin-knockout mice have become important
resources for studying the roles of electrical synaptic
transmission in the brain. Future studies using more
selective genetic formulations such as cell-specific
and inducible connexin-knockout mouse models will
accelerate our understanding of the contribution of
particular retinal gap junctions to visual signalling. As
we learn more about the properties that distinguish the
different connexins, it will be important to determine
the link between these properties, such as voltage and
neuromodulator sensitivity, and the function of the par-
ticular gap junctions in which the different connexins
are expressed.
Defects in gap junctions have been linked to a
number of neurological pathologies in both the CNS and
the PNS165. Thus, developing a further understanding
of the regulation of gap junctions, as well as the dynamic
relationship between electrical and chemical transmis-
sion, is an important challenge for the future. There is
no doubt that future studies will reveal new and crucial
roles for gap junctions in neural processing and that the
retina will remain a vital resource in this endeavour.
1.
Furshpan, E. J. & Potter, D. D. Mechanism of nerve-
impulse transmission at a crayfish synapse. Nature
180, 342–343 (1957).
Watanabe, A. The interaction of electrical activity
among neurons of lobster cardiac ganglion. Jpn
J. Physiol. 433, 283–305 (1958).
Goodenough, D. A. & Revel, J. P. A fine structural
analysis of intercellular junctions in the mouse liver.
J. Cell Biol. 45, 272–290 (1970).
Bennett, M. V. L. in Cellular Biology of Neurons,
Handbook of Physiology, The Nervous System (ed.
Kandel, E. R.) 357–416 (Williams & Wilkins,
Baltimore, 1977).
Söhl, G., Maxeiner, S. & Willecke, K. Expression and
functions of neuronal gap junctions. Nature Rev.
Neurosci. 6, 191–200 (2005).
Meier, C. & Dermietzel, R. Electrical synapses-gap
junctions in the brain. Results Probl. Cell Differ. 43,
99–128 (2006).
Söhl, G. & Willecke, K. An update on connexin genes
and their nomenclature in mouse and man. Cell
Commun. Adhes. 10, 173–180 (2003).
Bloomfield, S. A., Xin, D. & Osborne, T. Light-induced
modulation of coupling between AII amacrine cells in
the rabbit retina. Vis. Neurosci. 14, 565–576 (1997).
Weiler, R., Pottek, M., He, S. & Vaney, D. I. Modulation
of coupling between retinal horizontal cells by retinoic
acid and endogenous dopamine. Brain Res. Brain Res.
Rev. 32, 121–129 (2000).
10. Xin, D. & Bloomfield, S. A. Dark- and light-induced
changes in coupling between horizontal cells in
mammalian retina. J. Comp. Neurol. 405, 75–87
(1999).
11. Xin, D. & Bloomfield, S. A. Effects of nitric oxide on
horizontal cells in the rabbit retina. Vis. Neurosci. 17,
799–811 (2000).
12. Ribelayga, C., Cao, Y. & Mangel, S. C. The circadian
clock in the retina controls rod-cone coupling. Neuron
59, 790–801 (2008).
This study showed that a circadian clock in the
retina controls the extracellular dopamine
concentration and thereby the conductance of
rod–cone gap junctions.
13. Willecke, K. et al. Structural and functional diversity of
connexin genes in the mouse and human genome.
Biol. Chem. 383, 725–737 (2002).
14. Lampe, P. D. & Lau, A. F. Regulation of gap junctions
by phosphorylation of connexins. Arch. Biochem.
Biophys. 384, 205–215 (2000).
15. Lampe, P. D. & Lau, A. F. The effects of connexin
phosphorylation on gap junctional communication.
Int. J. Biochem. Cell Biol. 36, 1171–1186 (2004).
16. Witkovsky, P. & Dearry, A. Functional roles of
dopamine in the vertebrate retina. Prog. Retin. Res.
11, 247–292 (1991).
17. Koistinaho, J., Swanson, R. A., de Vente, J. &
Sagar, S. M. NADPH-diaphorase (nitric oxide
synthase)-reactive amacrine cells of rabbit retina:
putative target cells and stimulation by light.
Neuroscience 57, 587–597 (1993).
18. Lasater, E. M. Retinal horizontal cell gap junctional
conductance is modulated by dopamine through a
2.
3.
4.
5.
6.
7.
8.
9.
cyclic AMP-dependent protein kinase. Proc. Natl Acad.
Sci. USA 84, 7319–7323 (1987).
19. DeVries, S. H. & Schwartz, E. A. Modulation of an
electrical synapse between solitary pairs of catfish
horizontal cells by dopamine and second messengers.
J. Physiol. 414, 351–375 (1989).
20. Patel, L. S., Mitchell, C. K., Dubinsky, W. P. & O’Brien,
J. O. Regulation of gap junction coupling through the
neuronal connexin Cx35 by nitric oxide and cGMP.
Cell Commun. Adhes. 13, 41–54 (2006).
21. Kothmann, W. W., Massey, S. C. & O’Brien, J.
Dopamine D1-receptor-mediated modulation of
connexin36 phosphorylation in AII amacrine cells.
Invest. Ophthal. Vis. Sci. 49, 1515 (2008).
22. Lasater, E. M. & Dowling, J. E. Dopamine decreases
conductance of the electrical junctions between
cultured retinal horizontal cells. Proc. Natl Acad. Sci.
USA 82, 3025–3029 (1985).
23. Mills, S. L. & Massey, S. C. Differential properties of
two gap junctional pathways made by AII amacrine
cells. Nature 377, 734–737 (1995).
This report showed that the gap junctions between
AII cells are regulated by dopamine, whereas the
gap junctions between AII amacrine cells and ON
cone bipolar cells are regulated by nitric oxide.
24. Umino, O., Lee, Y. & Dowling, J. E. Effects of light
stimuli on the release of dopamine from interplexiform
cells in the white perch retina. Vis. Neurosci. 7,
451–458 (1991).
25. Baldridge, W. H., Weiler, R. & Dowling, J. E. Dark-
suppression and light-sensitization of horizontal cell
responses in the hybrid bass retina. Vis. Neurosci. 12,
611–620 (1995).
26. Bloomfield, S. A. & Völgyi, B. Function and plasticity of
homologous coupling between AII amacrine cells.
Vision Res. 44, 3297–3306 (2004).
27. Rose, B., Simpson, I. & Loewenstein, W. R. Calcium ion
produces graded changes in permeability of membrane
channels in cell junction. Nature 267, 625–627 (1977).
28. Peracchia, C. Calcium effects on gap junction structure
and cell coupling. Nature 271, 669–671 (1978).
29. Peracchia, C., Wang, X., Li, L. & Peracchia, L. L.
Inhibition of calmodulin expression prevents
low-pH-induced gap junction uncoupling in Xenopus
oocytes. Pflugers Arch. 431, 379–387 (1996).
30. Peracchia, C., Sotkis, A., Wang, X. G., Peracchia, L. L.
& Persechini, A. Calmodulin directly gates chemical
channels. J. Biol. Chem. 275, 26220–26224 (2000).
31. Török, K., Stauffer, K. & Evans, W. H. Connexin 32 of
gap junctions contains two cytoplasmic calmodulin-
binding domains. Biochem. J. 326, 479–483 (1997).
32. Lurtz, M. M. & Louis, C. F. Intracellular calcium
regulation of connexin43. Am. J. Physiol. Cell Physiol.
293, 1806–1813 (2007).
33. Spray, D. C., Harris, A. L. & Bennett, M. V. Gap
junctional conductance is a simple and sensitive
function of intracellular pH. Science 211, 712–715
(1981).
34. Church, J. & Baimbridge, K. G. Exposure to high-pH
medium increases the conductance and extent of dye
coupling between rat hippocampal CA1 pyramidal
neurons in vitro. J. Neurosci. 11, 3289–3295 (1991).
35. González, D. et al. Regulation of neuronal connexin-36
channels by pH. Proc. Natl Acad. Sci. USA 105,
17169–17174 (2008).
36. Chesler, M. Regulation and modulation of pH in the
brain. Physiol. Rev. 83, 1183–1221 (2003).
37. Spray, D. C., Harris, A. L. & Bennett, M. V. Voltage
dependence of junctional conductance in early
amphibian embryos. Science 204, 432–434
(1979).
38. Srinivas, M. et al. Voltage dependence of macroscopic
and unitary currents of gap junction channels formed
by mouse connexin50 expressed in rat neuroblastoma
cells. J. Physiol. 517, 673–689 (1999).
39. Moreno, A. P., de Carvalho, A. C., Verselis, V., Eghbali, B.
& Spray, D. C. Voltage-dependent gap junction
channels are formed by connexin32, the major gap
junction protein of rat liver. Biophys. J. 59, 920–925
(1991).
40. Moreno, A. P., Rook, M. B., Fishman, G. I. & Spray,
D. C. Gap junction channels: distinct voltage-sensitive
and -insensitive conductance states. Biophys. J. 67,
113–119 (1994).
41. Spray, D. C., Chanson, M., Moreno, A. P.,
Dermietzel, R. & Meda, P. Distinctive gap junction
channel types connect WB cells, a clonal cell line
derived from rat liver. Am. J. Physiol. 260, 513–527
(1991).
42. Baylor, D. A., Fuortes, M. G. & O’Bryan, P. M.
Receptive fields of cones in the retina of the turtle.
J. Physiol. 214, 265–294 (1971).
43. Cohen, A. I. Some electron microscopic observations
on inter-receptor contacts in the human and macaque
retinae. J. Anat. 99, 595–610 (1965).
44. Raviola, E. & Gilula, N. B. Gap junctions between
photoreceptor cells in the vertebrate retina. Proc. Natl
Acad. Sci. USA 70, 1677–1681 (1973).
45. Kolb, H. The organization of the outer plexiform layer
in the retina of the cat: electron microscopic
observations. J. Neurocytol. 6, 131–153 (1977).
46. Tsukamoto, Y., Masarachia, P., Schein, S. J. & Sterling, P.
Gap junctions between the pedicles of macaque foveal
cones. Vision Res. 32, 1809–1815 (1992).
47. Tsukamoto, Y., Morigiwa, K., Ueda, M. & Sterling, P.
Microcircuits for the night vision in mouse retina.
J. Neurosci. 21, 8616–8623 (2001).
48. Lee, E. J. et al. The immunocytochemical localization
of connexin 36 at rod and cone gap junctions in the
guinea pig retina. Eur. J. Neurosci. 18, 2925–2934
(2003).
49. Feigenspan, A. et al. Expression of connexin36 in cone
pedicles and OFF-cone bipolar cells of the mouse
retina. J. Neurosci. 24, 3325–3334 (2004).
50. Li, W. & DeVries, S. H. Separate blue and green cone
networks in the mammalian retina. Nature Neurosci.
7, 751–756 (2004).
51. DeVries, S. H., Qi, X., Smith, R., Makous, W. &
Sterling, P. Electrical coupling between mammalian
cones. Curr. Biol. 12, 1900–1907 (2002).
This study showed that coupling between cone
photoreceptors results in minor blurring of the
image, which is overshadowed by an increased
signal‑to‑noise ratio of cone responses.
REVIEWS
504 | july 2009 | vOlume 10
www.nature.com/reviews/neuro
© 2009 Macmillan Publishers Limited. All rights reserved

Page 11

52. Hornstein, E. P., Verweij, J., Li, P. H. & Schnapf, J. L.
Gap-junctional coupling and absolute sensitivity of
photoreceptors in macaque retina. J. Neurosci. 25,
11201–11209 (2005).
53. Dartnell, H. J. A. in The Eye Vol. 2 (ed. Davson, H.)
323–533 (Academic Press, New York, 1962).
54. Hornstein, E. P., Verweij, J. & Schnapf, J. L. Electrical
coupling between red and green cones in primate
retina. Nature Neurosci. 7, 745–750 (2004).
55. Hsu, A., Smith, R. G., Buchsbaum, G. & Sterling, P.
Cost of cone coupling to trichomacy in primate fovea.
J. Opt. Soc. Am. A. Opt. Image Sci. Vis. 17, 635–640
(2000).
56. Boycott, B. B. & Kolb, H. The connections between
bipolar cells and photoreceptors in the retina of the
domestic cat. J. Comp. Neurol. 148, 91–114 (1973).
57. Ghosh, K. K., Bujan, S., Haverkamp, S., Feigenspan, A.
& Wässle, H. Types of bipolar cells in the mouse retina.
J. Comp. Neurol. 469, 70–82 (2004).
58. Boycott, B. B. & Wässle, H. Morphological
classification of bipolar cells of the primate retina. Eur.
J. Neurosci. 3, 1069–1088 (1991).
59. Euler, T. & Wässle, H. Immunocytochemical
identification of cone bipolar cells in the rat retina.
J. Comp. Neurol. 361, 461–478 (1995).
60. Bloomfield, S. A. & Dacheux, R. F. Rod vision:
pathways and processing in the mammalian retina.
Prog. Retin. Eye Res. 20, 351–384 (2001).
61. Raviola, E. & Gilula, N. B. Intramembrane organization
of specialized contacts in the outer plexiform layer of
the retina. A freeze-fracture study in monkeys and
rabbits. J. Cell Biol. 65, 192–222 (1975).
62. Nelson, R. Cat cones have rod input: a comparison of
the response properties of cones and horizontal cell
bodies in the retina of the cat. J. Comp. Neurol. 172,
109–135 (1977).
This study was the first to show that rod signals can
be detected in cones, presumably as a result of
their interconnecting gap junctions.
63. Schneeweis, D. M. & Schnapf, J. L. Photovoltage of
rods and cones in the macaque retina. Science 268,
1053–1056 (1995).
64. DeVries, S. H. & Baylor, D. A. An alternative pathway
for signal flow from rod photoreceptors to ganglion
cells in mammalian retina. Proc. Natl Acad. Sci.
USA 92, 10658–10662 (1995).
65. Völgyi, B., Deans, M. R., Paul, D. L. & Bloomfield, S. A.
Convergence and segregation of the multiple rod
pathways in mammalian retina. J. Neurosci. 24,
11182–11192 (2004).
66. Blakemore, C. B. & Rushton, W. A. The rod increment
threshold during dark adaptation in normal and rod
monochromat. J. Physiol. 181, 629–640 (1965).
67. Conner, J. D. The temporal properties of rod vision.
J. Physiol. 332, 139–155 (1982).
68. Hess, R. F. & Nordby, K. Spatial and temporal
properties of human rod vision in the achromat.
J. Physiol. 371, 387–406 (1986).
69. Sharpe, L. T. & Stockman, A. Rod pathways: the
importance of seeing nothing. Trends Neurosci. 22,
497–504 (1999).
70. Deans, M. R., Volgyi, B., Goodenough, D. A.,
Bloomfield, S. A. & Paul, D. L. Connexin36 is
essential for transmission of rod-mediated visual
signals in the mammalian retina. Neuron 36,
703–712 (2002).
This article showed that Cx36‑containing gap
junctions form obligatory elements in the
transmission of rod signals in the retina.
71. Dunn, F. A., Doan, T., Sampath, A. P. & Rieke, F.
Controlling the gain of rod-mediated signals in the
mammalian retina. J. Neurosci. 26, 3959–3970
(2006).
72. Smith, R. G., Freed, M. A. & Sterling, P. Microcircuitry
of the dark-adapted cat retina: functional architecture
of the rod-cone network. J. Neurosci. 6, 3505–3517
(1986).
73. Soucy, E., Wang, Y., Nirenberg, S., Nathans, J. &
Meister, M. A novel signaling pathway from rod
photoreceptors to ganglion cells in mammalian retina.
Neuron 21, 481–493 (1998).
74. Boycott, B. B. & Dowling, J. E. Organization of the
primate retina: light microscopy. Philos. Trans. R. Soc.
Lond. B Biol. Sci. 255, 109–184 (1969).
75. Hack, I., Peichl, L. & Brandstätter, J. H. An alternative
pathway for rod signals in the rodent retina: rod
photoreceptors, cone bipolar cells, and the localization
of glutamate receptors. Proc. Natl Acad. Sci. USA 96,
14130–14135 (1999).
76. Fyk-Kolodziej, B., Qin, P. & Pourcho, R. G.
Identification of a cone bipolar cell in cat retina which
has input from both rod and cone photoreceptors.
J. Comp. Neurol. 464, 104–113 (2003).
77. Li, W., Keung, J. W. & Massey, S. C. Direct synaptic
connections between rods and OFF cone bipolar cells
in the rabbit retina. J. Comp. Neurol. 474, 1–12
(2004).
78. Kolb, H. The connections between horizontal cells and
photoreceptors in the retina of the cat: electron
microscopy of Golgi preparations. J. Comp. Neurol.
155, 1–14 (1974).
79. Yamada, E. & Ishikawa, T. The fine structure of the
horizontal cells in some vertebrate retinae. Cold Spring
Harb. Symp. Quant. Biol. 30, 383–392 (1965).
80. Vaney, D. Many diverse types of retinal neurons show
tracer coupling when injected with biocytin and
neurobiotin. Neurosci. Lett. 125, 187–190 (1991).
This study was the first to demonstrate that
biotinylated tracers permeate gap junctions and
can be used to visualize electrical synapses.
81. Bloomfield, S. A., Xin, D. & Persky, S. E. A comparison
of receptive field and tracer coupling size of horizontal
cells in the rabbit retina. Vis. Neurosci. 12, 985–999
(1995).
82. Dacheux, R. F. & Raviola, E. Horizontal cells in the
retina of the rabbit. J. Neurosci. 2, 1486–1493
(1982).
83. Dacey, D. M. Primate retina: cell types, circuits and
color opponency. Prog. Retin. Eye Res. 18, 737–763
(1999).
84. Naka, K. I. & Rushton, W. A. The generation and
spread of S-potentials in fish (Cyprinidae). J. Physiol.
192, 437–461 (1967).
85. Bloomfield, S. A. & Miller, R. F. A physiological and
morphological study of the horizontal cell types in the
rabbit retina. J. Comp. Neurol. 208, 288–303
(1982).
86. Hombach, S. et al. Functional expression of
connexin57 in horizontal cells of the mouse retina.
Eur. J. Neurosci. 19, 2633–2640 (2004).
87. Shelley, J. et al. Horizontal cell receptive fields are
reduced in connexin57-deficient mice. Eur.
J. Neurosci. 23, 3176–3186 (2006).
This report showed that disruption of horizontal
cell gap junctions in the connexin57‑deficient
mouse retina results in a dramatic reduction in
their receptive field sizes.
88. Werblin, F. S. & Dowling, J. E. Organization of the
retina of the mudpuppy, Necturus maculosus. II.
Intracellular recording. J. Neurophysiol. 32, 339–355
(1969).
89. Kaneko, A. Physiological and morphological
identification of horizontal, bipolar and amacrine cells
in goldfish retina. J. Physiol. 207, 623–633 (1970).
90. Naka, K. I. & Nye, P. W. Role of horizontal cells in
organization of the catfish retinal receptive field.
J. Neurophysiol. 34, 785–801 (1971).
91. Naka, K. I. & Witkovsky, P. Dogfish ganglion cell
discharge resulting from extrinsic polarization of the
horizontal cells. J. Physiol. 223, 449–460 (1972).
92. Marchiafava, P. L. Horizontal cells influence membrane
potential of bipolar cells in the retina of the turtle.
Nature 275, 141–142 (1978).
93. Mangel, S. C. & Miller, R. F. Horizontal cells contribute
to the receptive field surround of ganglion cells in the
rabbit retina. Brain Res. 414, 182–186 (1987).
94. Kamermans, M. et al. Hemichannel-mediated
inhibition in the outer retina. Science 292,
1178–1180 (2001).
95. Dedek, K. et al. Ganglion cell adaptability: does
coupling of horizontal cells play a role? PLoS ONE 5,
e1714 (2008).
96. Hedden, W. L. & Dowling, J. E. The interplexiform cell
system. II. Effects of dopamine on goldfish retinal
neurones. Proc. R. Soc. Lond. B Biol. Sci. 201, 27–55
(1978).
97. Piccolino, M., Neyton, J. & Gerschenfeld, H. M.
Decrease of gap junction permeability induced by
dopamine and cyclic adenosine 3′:5′-monophosphate
in horizontal cells of turtle retina. J. Neurosci. 4,
2477–2488 (1984).
98. Baldridge, W. H., Ball, A. K. & Miller, R. G.
Dopaminergic regulation of horizontal cell gap junction
particle density in goldfish retina. J. Comp. Neurol.
265, 428–436 (1987).
99. McMahon, D. G., Knapp, A. G. & Dowling, J. E.
Horizontal cell gap junctions: single-channel
conductance and modulation by dopamine. Proc. Natl
Acad. Sci. USA 86, 7639–7643 (1989).
This report showed that dopamine can directly
modulate the conductance of gap junctions
connecting neighbouring horizontal cells.
100. Kurz-Isler, G., Voigt, T. & Wolburg, H. Modulation of
connexon densities in gap junctions of horizontal cell
perikarya and axon terminals in fish retina: effects of
light/dark cycles, interruption of the optic nerve and
application of dopamine. Cell Tissue Res. 268,
267–275 (1992).
101. McMahon, D. G. & Brown, D. R. Modulation of gap-
junction channel gating at zebrafish retinal electrical
synapses. J. Neurophysiol. 72, 2257–2268 (1994).
102. DeVries, S. H. & Schwartz, E. A. Hemi-gap-junction
channels in solitary horizontal cells of the catfish
retina. J. Physiol. 445, 201–230 (1992).
103. Lu, C. & McMahon, D. G. Modulation of hybrid bass
retinal gap junctional channel gating by nitric oxide.
J. Physiol. 499, 689–699 (1997).
104. Pottek, M., Schultz, K. & Weiler, R. Effects of nitric
oxide on the horizontal cell network and dopamine
release in the carp retina. Vision Res. 37, 1091–1102
(1997).
105. Hampson, E. C., Weiler, R. & Vaney, D. I. pH-gated
dopaminergic modulation of horizontal cell gap
junctions in mammalian retina. Proc. Biol. Sci. 255,
67–72 (1994).
106. Mills, S. L. & Massey, S. E. l-Arginine uncouples A-type
horizontal cells in rabbit retina. Invest. Ophthalmol.
Vis. Sci. (Suppl.) 34, 1382 (1993).
107. Godley, B. F. & Wurtman, R. J. Release of endogenous
dopamine from the superfused rabbit retina in vitro:
effect of light stimulation. Brain Res. 452, 393–395
(1988).
108. Kirsch, M. & Wagner, H. J. Release pattern of
endogenous dopamine in teleost retinae during light
adaptation and pharmacological stimulation. Vision
Res. 29, 147–154 (1989).
109. Zemel, E., Eyal, O., Lei, B. & Perlman, I. NADPH
diaphorase activity in mammalian retinas is
modulated by the state of visual adaptation. Vis.
Neurosci. 13, 865–871 (1996).
110. Neal, M., Cunningham, J. & Matthews, K. Selective
release of nitric oxide from retinal amacrine and
bipolar cells. Invest. Ophthalmol. Vis. Sci. 39,
850–853 (1998).
111. Mangel, S. C. & Dowling, J. E. Responsiveness and
receptive field size of carp horizontal cells are reduced
by prolonged darkness and dopamine. Science 229,
1107–1109 (1985).
112. Mangel, S. C. & Dowling, J. E. The interplexiform-
horizontal cell system of the fish retina: effects of
dopamine, light stimulation and time in the dark.
Proc. R. Soc. Lond. B Biol. Sci. 231, 91–121
(1987).
113. Tornqvist, K., Yang, X. L. & Dowling, J. E. Modulation
of cone horizontal cell activity in the teleost fish retina.
III. Effects of prolonged darkness and dopamine on
electrical coupling between horizontal cells.
J. Neurosci. 8, 2279–2288 (1988).
114. Rodieck, R. W. & Stone, J. Analysis of receptive fields
of cat retinal ganglion cells. J. Neurophysiol. 28,
832–849 (1965).
115. Peichl, L. & Wässle, H. The structural correlate of the
receptive field centre of α cells in the cat retina.
J. Physiol. 341, 309–324 (1983).
116. Muller, J. F. & Dacheux, R. F. Alpha ganglion cells of
the rabbit retina lose antagonistic surround responses
under dark adaptation. Vis. Neurosci. 14, 395–401
(1997).
117. Balboa, R. M. & Grzywacz, N. M. The role of early
retinal inhibition: more than maximizing luminance
information. Vis. Neurosci. 17, 77–89 (2000).
118. Balboa, R. M. & Grzywacz, N. M. The minimal local-
asperity hypothesis of early retinal lateral inhibition.
Neural Comput. 12, 1485–1517 (2000).
119. Famiglietti, E. V. Jr & Kolb, H. A bistratified amacrine
cell and synaptic circuitry in the inner plexiform layer
of the retina. Brain Res. 84, 293–300 (1975).
120. Strettoi, E., Dacheux, R. F. & Raviola, E. Synaptic
connections of rod bipolar cells in the inner plexiform
layer of the rabbit retina. J. Comp. Neurol. 295,
449–466 (1990).
121. Feigenspan, A., Teubner, B., Willecke, K. & Weiler, R.
Expression of neuronal connexin36 in AII amacrine
cells of the mammalian retina. J. Neurosci. 21,
230–239 (2001).
122. Mills, S. L., O’Brien, J. J., Li, W., O’Brien, J. & Massey,
S. C. Rod pathways in the mammalian retina use
connexin36. J. Comp. Neurol. 436, 336–350
(2001).
123. Lin, B., Jakobs, T. C. & Masland, R. H. Different
functional types of bipolar cells use different gap-
junctional proteins. J. Neurosci. 25, 6696–6701
(2005).
REVIEWS
NATuRe RevIewS | NeuroscieNce
vOlume 10 | july 2009 | 505
© 2009 Macmillan Publishers Limited. All rights reserved

Page 12

124. Han, Y. & Massey, S. C. Electrical synapses in retinal
ON cone bipolar cells: subtype-specific expression of
connexins. Proc. Natl Acad. Sci. USA 102,
13313–13318 (2005).
125. Dedek, K. et al. Localization of the heterotypic gap
junctions composed of connexin45 and connexin36 in
the rod pathway of the mouse retina. Eur. J. Neurosci.
24, 1675–1686 (2006).
126. Li, X. et al. Connexin45-containing neuronal gap
junctions in rodent retina also contain connexin36 in
both apposing hemiplaques, forming bihomotypic gap
junctions, with scaffolding contributed by zona
occludens-1. J. Neurosci. 28, 9769–9789 (2008).
127. Mills, S. L. & Massey, S. C. A series of biotinylated
tracers distinguishes three types of gap junction in
retina. J. Neurosci. 20, 8629–8636 (2000).
128. Veruki, M. L. & Hartveit, E. Electrical synapses
mediate signal transmission in the rod pathway of the
mammalian retina. J. Neurosci. 22, 10558–10566
(2002).
129. Güldenagel, M. et al. Visual transmission deficits in
mice with targeted disruption of the gap junction gene
connexin36. J. Neurosci. 21, 6036–6044 (2001).
130. Xin, D. & Bloomfield, S. A. Comparison of the
responses of AII amacrine cells in the dark- and light-
adapted rabbit retina. Vis. Neurosci. 16, 653–665
(1999).
131. Smith, R. G. & Vardi, N. Simulation of the AII amacrine
cell of mammalian retina: functional consequences of
electrical coupling and regenerative membrane
properties. Vis. Neurosci. 12, 851–860 (1995).
132. Voigt, T. & Wässle, H. Dopaminergic innervation of AII
amacrine cells in mammalian retina. J. Neurosci. 7,
4115–4128 (1987).
133. Hampson, E. C., Vaney, D. I. & Weiler, R. Dopaminergic
modulation of gap junction permeability between
amacrine cells in mammalian retina. J. Neurosci. 12,
4911–4922 (1992).
134. Urschel, S. et al. Protein kinase A-mediated
phosphorylation of connexin36 in mouse retina results
in decreased gap junctional communication between
AII amacrine cells. J. Biol. Chem. 281, 33163–33171
(2006).
135. Vaney, D. I. Territorial organization of direction-
selective ganglion cells in rabbit retina. J. Neurosci.
14, 6301–6316 (1994).
136. Xin, D. & Bloomfield, S. A. Tracer coupling pattern of
amacrine cells in the rabbit retina. J. Comp. Neurol.
383, 512–528 (1997).
137. Völgyi, B., Chheda, S. & Bloomfield, S. A. Tracer
coupling patterns of the ganglion cell subtypes in the
mouse retina. J. Comp. Neurol. 512, 664–687
(2009).
138. Bloomfield, S. A. & Xin, D. A comparison of receptive-
field and tracer-coupling size of amacrine and ganglion
cells in the rabbit retina. Vis. Neurosci. 14, 1153–1165
(1997).
139. Mastronarde, D. N. Correlated firing of cat retinal
ganglion cells. I. Spontaneously active inputs to X- and
Y-cells. J. Neurophysiol. 49, 303–324 (1983).
140. Mastronarde, D. N. Correlated firing of cat retinal
ganglion cells. II. Responses of X- and Y-cells to single
quantal events. J. Neurophysiol. 49, 325–349
(1983).
141. Mastronarde, D. N. Interactions between ganglion cells
in cat retina. J. Neurophysiol. 49, 350–365 (1983).
This study was the first to report concerted activity
between neighbouring ganglion cells, suggesting
direct electrical coupling years before gap junctions
between ganglion cells were demonstrated.
142. Arnett, D. & Spraker, T. E. Cross-correlation analysis of
the maintained discharge of rabbit retinal ganglion
cells. J. Physiol. 317, 29–47 (1981).
143. Brivanlou, I. H., Warland, D. K. & Meister, M.
Mechanisms of concerted firing among retinal
ganglion cells. Neuron 20, 527–539 (1998).
This study showed that different types of ganglion
cell and amacrine cell electrical coupling result in
different patterns of correlated spike activity.
144. DeVries, S. H. Correlated firing in rabbit retinal
ganglion cells. J. Neurophysiol. 81, 908–920 (1999).
145. Hu, E. H. & Bloomfield, S. A. Gap junctional coupling
underlies the short-latency spike synchrony of retinal α
ganglion cells. J. Neurosci. 23, 6768–6777 (2003).
146. Shlens, J., Rieke, F. & Chichilnisky, E. Synchronized
firing in the retina. Curr. Opin. Neurobiol. 18,
396–402 (2008).
147. Castelo-Branco, M., Neuenschwander, S. & Singer, W.
Synchronization of visual responses between the
cortex, lateral geniculate nucleus and retina in the
anasthetized cat. J. Neurosci. 18, 6395–63410
(1998).
148. Schnitzer, M. J. & Meister, M. Multineuronal firing
patterns in the signal from eye to brain. Neuron 37,
499–511 (2003).
149. Meister, M. & Berry, M. The neural code of the retina.
Neuron 22, 435–450 (1999).
150. Alonso, J. M., Usrey, W. M. & Reid, R. C. Precisely
correlated firing in cells of the lateral geniculate
nucleus. Nature 383, 815–819 (1996).
151. Stevens, C. F. & Zador, A. M. Input synchrony and
irregular firing of cortical neurons. Nature Neurosci. 1,
210–217 (1998).
152. Usrey, W. M. & Reid, R. C. Synchronous activity in the
visual system. Annu. Rev. Physiol. 61, 435–456 (1999).
153. Singer, W. Neuronal synchrony: a versatile code for the
definition of relations? Neuron 24, 49–65 (1999).
154. Mills, S. L. et al. Dopaminergic modulation of tracer
coupling in a ganglion-amacrine cell network. Vis.
Neurosci. 24, 593–608 (2007).
155. Naus, C. C. & Bani-Yaghoub, M. Gap junctional
communication in the developing central nervous
system. Cell Biol. Int. 22, 751–763 (1998).
156. Grubb, M. S. & Thompson, I. D. The influence of early
experience on the development of sensory systems.
Curr. Opin. Neurobiol. 14, 503–512 (2004).
157. Syed, M. M., Lee, S., He, S. & Zhou, Z. J. Spontaneous
waves in the ventricular zone of developing
mammalian retina. J. Neurophysiol. 91, 1999–2009
(2004).
158. Sernagor, E., Eglen, S. J. & Wong, R. O. Development
of retinal ganglion cell structure and function. Prog.
Retin. Eye Res. 20, 139–174 (2001).
159. Ackert, J. M. et al. Light induced changes in spike
synchronization between coupled ON direction
selective ganglion cells in the mammalian retina.
J. Neurosci. 26, 4206–4215 (2006).
This study showed that changes in the synchronous
activity of coupled ON direction‑selective ganglion
cells provide a mechanism to encode the direction
of stimulus motion.
160. Oyster, C. W. The analysis of image motion by the
retina. J. Physiol. 199, 613–635 (1968).
161. Oyster, C. W., Simpson, J. I., Takahashi, E. S. &
Soodak, R. E. Retinal ganglion cells projecting to the
rabbit accessory optic system. J. Comp. Neurol. 190,
49–61 (1980).
162. Buhl, E. H. & Peichl, L. Morphology of rabbit retinal
ganglion cells projecting to the medial terminal
nucleus of the accessory optic system. J. Comp.
Neurol. 253, 163–174 (1986).
163. Pu, M. L. & Amthor, F. R. Dendritic morphologies of
retinal ganglion cells projecting to the nucleus of the
optic tract in the rabbit. J. Comp. Neurol. 302,
657–674 (1990).
164. Simpson, J. I. The accessory optic sytem. Annu. Rev.
Neurosci. 7, 13–41 (1984).
165. Nakase, T. & Nasus, C. C. G. Gap junctions and
neurological disorders of the central nervous system.
Biochim. Biophys. Acta 1662, 149–158 (2004).
166. O’Brien, J. J. et al. Coupling between A-type
horizontal cells is mediated by connexin 50 gap
junctions in the rabbit retina. J. Neurosci. 26,
11624–11636 (2006).
167. Schubert, T. et al. Connexin36 mediates gap
junctional coupling of alpha-ganglion cells in mouse
retina. J. Comp. Neurol. 485, 191–201 (2005).
168. Schubert, T., Maxeiner, S., Krüger, O., Willecke, K. &
Weiler, R. Connexin45 mediates gap junctional
coupling of bistratified ganglion cells in the mouse
retina. J. Comp. Neurol. 490, 29–39 (2005).
169. Völgyi, B., Abrams, J., Paul, D. L. & Bloomfield, S. A.
Morphology and tracer coupling of alpha ganglion
cells in the mouse retina. J. Comp. Neurol. 492,
66–77 (2005).
Acknowledgements
The authors wish to acknowledge the National Eye Institute
of the US National Institutes of Health for support of their
research programmes (grants EY007360 (S.A.B.) and
EY017832 (B.V.)).
DATABASES
Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.
fcgi?db=gene
Cx32 | Cx36 | Cx43 | Cx57
All liNks Are AcTive iN The oNliNe pdf
REVIEWS
506 | july 2009 | vOlume 10
www.nature.com/reviews/neuro
© 2009 Macmillan Publishers Limited. All rights reserved