Visual Neuroscience ~2005!, 22, 661–676. Printed in the USA.
Copyright © 2005 Cambridge University Press 0952-5238005 $16.00
DOI: 10.10170S0952523805225154
Structural and functional composition of the developing
retinogeniculate pathway in the mouse
LISA JAUBERT-MIAZZA, ERICK GREEN, FU-SUN LO, KIM BUI, JEREMY MILLS,
and WILLIAM GUIDO
Department of Cell Biology and Anatomy, Louisiana State Health Sciences Center, New Orleans
(Received March 15, 2005; Accepted July 27, 2005!
Abstract
The advent of transgenic mice has made the developing retinogeniculate pathway a model system for targeting
potential mechanisms that underlie the refinement of sensory connections. However, a detailed characterization of
the form and function of this pathway is lacking. Here we use a variety of anatomical and electrophysiological
techniques to delineate the structural and functional changes occurring in the lateral geniculate nucleus ~LGN! of
dorsal thalamus of the C570BL6 mouse. During the first two postnatal weeks there is an age-related recession in the
amount of terminal space occupied by retinal axons arising from the two eyes. During the first postnatal week,
crossed and uncrossed axons show substantial overlap throughout most of the LGN. Between the first and second
week retinal arbors show significant pruning, so that by the time of natural eye opening ~P12–14! segregation is
complete and retinal projections are organized into distinct eye-specific domains. During this time of rapid
anatomical rearrangement, LGN cells could be readily distinguished using immunocytochemical markers that stain
for NMDA receptors, GABA receptors, L-type Ca 2⫹ channels, and the neurofilament protein SMI-32. Moreover, the
membrane properties and synaptic responses of developing LGN cells are remarkably stable and resemble those of
mature neurons. However, there are some notable developmental changes in synaptic connectivity. At early ages,
LGN cells are binocularly responsive and receive input from as many as 11 different retinal ganglion cells. Optic
tract stimulation also evokes plateau-like depolarizations that are mediated by the activation of L-type Ca 2⫹
channels. As retinal inputs from the two eyes segregate into nonoverlapping territories, there is a loss of binocular
responsiveness, a decrease in retinal convergence, and a reduction in the incidence of plateau potentials. These data
serve as a working framework for the assessment of phenotypes of genetically altered strains as well as provide
some insight as to the molecular mechanisms underlying the refinement of retinogeniculate connections.
Keywords: Retina, Lateral geniculate nucleus, Retinal projections, Relay cells
operational, aggregates of neighboring retinal ganglion cells fire
spontaneously in rhythmic bursts of activity that traverse across
the retina in wave-like fashion ~Maffei & Galli-Resta, 1990;
Meister et al., 1991; Demas et al., 2003!. Alterations in the pattern
of spontaneous activity or the complete cessation of activity leads
to abnormally wide spread retinal arbors and a failure of retinal
projections to segregate properly in LGN ~Shatz & Stryker, 1988;
Sretavan et al., 1988; Feller et al., 1996; Penn et al., 1998;
Chapman, 2000; Huberman et al., 2002; Stellwagen & Shatz,
2002!.
In recent years, the molecular mechanisms underlying the
remodeling of retinogeniculate connections has been a topic of
intense inquiry. Transgenic mouse models have been used with
increasing regularity to either isolate or excise potential molecular
factors ~e.g. Mooney et al., 1996; Huh et al., 2000; Pham et al.,
2001; Rossi et al., 2001; Muir-Robinson et al., 2002; Grubb et al.,
2003; Torborg et al., 2005!. Since genetic manipulations are most
advanced in mice, they have become the model system of study for
visual system development. While some information about the
developing mouse retinogeniculate pathway exists ~Godement et al.,
Introduction
Much of our present understanding about the activity-dependent
refinement of sensory connections is based on work done in the
developing retinogeniculate pathway. Studies done primarily in the
cat and ferret indicate this pathway develops from a crudely
organized network of cells, into a highly ordered sensory system
comprised of precise retinotopic patterns of connectivity, separate
eye-specific laminae, distinct ON and OFF sublaminae, and at
least two morphologically and functionally distinct neuronal streams
~see Sherman & Spear, 1982; Shatz, 1990, 1996; Goodman &
Shatz, 1993; Cramer & Sur, 1995; Guido et al., 2003; Grubb &
Thompson, 2004!. The establishment of these orderly connections
has been attributed to the coordinated firing patterns of retinal
ganglion cells ~Wong, 1999!. Even before photoreceptors are
Address correspondence and reprint requests to: William Guido,
Department of Cell Biology and Anatomy, LSU Health Sciences
Center, 1901 Perdido Street, New Orleans, LA 70112, USA. E-mail:
[email protected]
661
662
1984; Mooney et al., 1993; Chen & Regehr, 2000!, a detailed
examination of ontogeny is lacking. To address this, we employed
a variety of anatomical and electrophysiological techniques to
ascertain the structural and functional state of the developing
retinogeniculate pathway in a commonly used pigmented strain,
the C570BL6 mouse. To examine the postnatal time course of
retinogeniculate axon segregation, we used anterograde labeling
techniques to visualize the pattern of retinal projections in LGN.
To assess the structural composition of developing LGN cells, we
applied a number of immunocytochemical markers to label neuronal elements suspected to play a key role in the refinement of
connections or which could be used to identify different cells types
in LGN. Finally, we examined the electrophysiological and synaptic properties of developing cells by utilizing an in vitro recording preparation in which the retinal projections and intrinsic circuitry
of LGN are preserved ~Lo et al., 2002; Ziburkus et al., 2003;
Ziburkus & Guido, 2005; see also Hu, 1993!.
Materials and methods
Subjects
Anatomical and electrophysiological experiments were conducted
on C570BL6 mice that ranged in age from postnatal ~P! day 2 to
adulthood ~35–50 days old!. Mouse pups were obtained from
time-pregnant females acquired from a commercial vendor ~Charles
River Laboratories, Wilmington, MA!. All surgical procedures
were performed in accordance with the IACUC protocol approved
by the Louisiana State University Animal Care and Use Committee.
Anterograde labeling of retinogeniculate projections
Mice were anesthetized by hypothermia ~⬍P7! or with isofluorane
vapors ~⬎P7!. Prior to natural eye opening, fused eyelids were
separated or cut to expose the temporal region of the eye. The
sclera was pierced with a sharp-tipped glass pipette and excess
vitreous was drained. Another pipette, filled with a 0.1–0.2%
solution of the B subunit of cholera toxin ~CTB, Molecular Probes,
Eugene, OR! conjugated to either Alexa Fluor 594 ~red! or Alexa
Fluor 488 ~green! dissolved in distilled water, was inserted into the
hole made by the first pipette. The pipette containing the CTB was
attached to a Picospritzer and a prescribed volume ~1–3 µl at
P3–10 and 3–5 µl for ages ⬎P10! of solution was injected into the
eye. Each eye was injected with a different fluorescent conjugate
so the terminal fields from both eyes could be visualized simultaneously in a single section of the LGN ~Figs. 1B & 2!. After a
2–3 day survival period, animals were deeply anesthetized ~halothane vapors or IP injection of nembutal! and transcardially perfused with a heparinized Tyrode solution ~25–50 ml! followed by
4% paraformaldehyde in 0.1 M phosphate buffer ~PB, pH 7.4,
100–250 ml!. The brains were postfixed overnight and sectioned at
60 mm in the coronal plane using a vibratome. Sections containing
the LGN were mounted in ProLong ~Molecular Probes! and imaged with epifluorescence microscopy.
Images of LGN were acquired with a Photometrix Coolsnap
camera attached to a Nikon Eclipse fluorescence microscope using
a 10⫻ objective. Fluorescent images of labeled sections were
acquired and digitized separately ~1300 ⫻ 1030 pixels0frame!
using the following filter settings: Alexa 488: Exciter 465– 495,
DM 505, BA 515–555, Alexa 594: Exciter 528–553, DM 565, BA
600– 660. Background fluorescence was subtracted and grayscale
L. Jaubert-Miazza et al.
images were normalized ~0–255! using Metamorph or Adobe
Photoshop software. The boundaries of LGN were delineated so as
not to include label from the optic tract, intrageniculate leaflet, and
ventral geniculate nucleus. To determine the spatial extent of
labeled retinal projections in the LGN, grayscale images were
converted into binary high contrast black and white images. A
threshold procedure was employed that allowed for the specification of a level that corresponded to a location in the gray scale
histogram where there was a clear distinction between signal and
residual background fluorescence ~Muir-Robinson et al., 2002;
Torborg & Feller, 2004!. Typically, threshold levels corresponded
to a value of 30–50 on a normalized range of 0–255. To help
illustrate the threshold procedure, we generated binary pseudocolored images of LGN ~Fig. 2, right panel!. In these, the red and
green fluorescent images were superimposed. Pixel intensity for
each channel was “normalized” so that every pixel greater than or
equal to to a defined threshold level was assigned a value of 255.
Pixels that contain both red and green fluorescence were considered as areas of overlap and represented as yellow.
For terminal fields belonging to crossed ~contralateral eye! or
uncrossed ~ipsilateral eye! axons, the total number pixels in the
defined area of LGN representing either green or red fluorescence
was measured. To determine the extent to which the projections
from each eye share common or overlapping space in LGN, the
pixels that contained both red and green signal ~represented as
yellow in Fig. 2! were counted. The spatial extent of uncrossed,
crossed, and overlapping projections were computed by summing
their area across five successive sections through the middle of
LGN and then expressing the labeled regions as a percentage of the
total area of LGN.
A threshold-independent method for analyzing eye-specific
segregation was also adopted ~Torborg & Feller, 2004!. For each
pixel in analyzed sections of LGN, the logarithm of the intensity
ratio ~R ⫽ log 10 FI 0FC !, where FI is the fluorescence intensity
associated with ipsilateral terminal fields and FC is the intensity of
the contralateral fields, was calculated and plotted as a frequency
histogram ~bin size 0.1 log units!. The variance of each R-distribution
was calculated and compared across mice of different ages.
Immunohistochemistry
Coronal sections containing the LGN were prepared using the
methods described above and were then processed using a standard
Nissl protocol or incubated with a number of different antibodies.
For immunocytochemistry, sections were preincubated for 30 min
in 10% normal goat serum ~NGS! or normal rabbit serum ~NRS!
in phosphate-buffered saline ~PBS; 0.01 MPO4 , pH 7.4, 0.9%
NaCl!. Treated sections were then transferred to their primary
antibody solutions ~diluted in 1% NGS or NRS in PBS! and
incubated overnight at 48C. The following antibodies were used: a
polyclonal rabbit antibody directed toward the pore forming a1C
subunit of L-type Ca 2⫹ channel ~1:1000, Alomone Labs, ACC-003,
Jerusalem, Israel!, the mouse monoclonal antibody SMI-32 which
stains for the nonphosphorylated form of a high-molecular-weight
protein ~1:5000, Sternberger Monoclonals, Bethesda, MD! that
labels thalamic relay cells ~Bickford et al., 1998; Carden et al.,
2000!, a mouse monoclonal antibody directed towards glutamic
acid decarboxylase ~GAD! the synthesizing enzyme for GABA
~1:2000, GAD, BD Pharmingen, 559931, San Diego, CA! that
labels thalamic interneurons ~Montero & Singer, 1985; Ziburkus
et al., 2000!, and a mouse monoclonal antibody directed toward the
pore forming R1 subunit of the NMDA receptor ~1:1000, BD
Developmental remodeling in the mouse LGN
Pharmingen, 54.1!. The next day the tissue was rinsed in 0.1 M PB
and incubated at room temperature for 1 h in 1% NGS0PBS
solution containing biotinylated secondary antibodies. Tissue treated
for L-type Ca 2⫹ channels were incubated in a biotinylated goat
anti-rabbit ~1:400, Sigma RS 5001, St. Louis, MO! while those
treated with SMI-32, NMDAR-1 or GAD antibodies were incubated in biotinylated goat anti-mouse ~1:400 Sigma, B8520!. Following a series of buffer rinses, tissue was incubated for 1 h in an
avidin and biotinylated horseradish peroxidase ~1:100, ABC Vector
Labs, Burlingame, CA! 1% NGS0PBS solution. Sections were
rinsed in buffer and reacted with a nickel intensified diaminobenzidine ~DAB! solution for 1–2 min. Following a sodium acetate
wash and a final buffer rinse, tissue was mounted on slides,
dehydrated, and coverslipped for light microscopy.
High-resolution digital images of labeled cells were acquired at
different levels of magnification with a Kodak CD 290 camera
mounted on the head of the Nikon Labophot light microscope.
Soma size and cell density measurements were obtained by using
Sigma Scan software. Labeled cells were included in soma size
and cell density analyses if their nucleus or a continuous cell
membrane was readily discerned in a given focal plane. Soma sizes
were obtained from a large field in the middle of LGN from at least
two representative sections of two P7 and two adult mice. At each
age, soma measurements were pooled and plotted as a frequency
distribution. Cell density measurements were computed by counting labeled cells in two representative sections through the middle
of LGN of 2–5 different mice at P7, P14, P21, and adult.
Electrophysiology
In vitro intracellular recordings of LGN cells were done in an
isolated brainstem preparation ~Lo et al., 2002; Ziburkus et al.,
2003; Ziburkus & Guido, 2005!. Mice were anesthetized with
halothane and killed by decapitation. The brain was excised and
placed in 48C solution of artificial cerebrospinal fluid ~ACSF,
in mM: 124 NaCl, 2.5 KCl, 1.25 NaH2PO4 , 1.0 MgSO4 , 26
NaHCO 3 , 10 Dextrose, 2 CaCl2 , saturated with 95% O2 05% CO2 ,
pH ⫽ 7.4!. While in solution, the excised brain was cut in half
along the midline axis, glued to a silver plate, and placed into a
well of a temperature-controlled recording chamber. The lateral
surface of the thalamus and midbrain were exposed by removing
the forebrain. In some cases, this surgical approach was modified
and large segments of each optic nerve were preserved ~Ziburkus
& Guido, 2005!. The brainstem preparation was submerged and
perfused continuously ~4–5 ml0min! with warmed ~31–338C! ACSF.
Recordings began 1–3 h after incubation. Electrode penetrations
were made along the dorsal lateral surface of the LGN and
restricted to depths ⱕ 150 mm. In the mature mouse, this region
receives retinal input exclusively from the contralateral eye ~see
Fig. 1, Godement et al., 1984; Muir-Robinson et al., 2002!.
Whole-cell recordings were obtained with pipettes made of
borosilicate glass filled with a solution containing ~in mM! 140 K
Gluconate, 10 HEPES, 1.1 EGTA-Na, 0.1 CaCl2 , 2 MgCl2 , 2
ATP-Mg, 0.2 GTP-Na, pH ⫽ 7.2! and pulled to a final tip resistance 5–7 MV. Sharp-tipped recordings were also employed with
electrodes filled with either 4 M potassium acetate ~KAC! or a 2%
solution of biocytin dissolved in 2 M KAC. Sharp-tipped electrodes were pulled to a final tip resistance of 70–90 MV. Intracellular recordings were done in current clamp mode with an
Axoclamp 2B amplifier using techniques described elsewhere
~Guido et al., 1997, 1998; Lo et al., 2002; Ziburkus et al., 2003!.
Neuronal activity was displayed on a storage oscilloscope, digi-
663
tized at 10 kHz, and stored directly on computer. During some
recordings ~n ⫽ 50!, the GABAA antagonist bicuculline ~10–
25 mM! or the GABAB antagonist 2-hydroxysaclofen ~100 mM!
were bath applied to block inhibitory activity. Nimodipine ~n ⫽ 3,
10–20 mM! was bath applied to block L-type Ca 2⫹ channel
activity.
During intracellular recording, some LGN cells ~n ⫽ 10! were
filled with biocytin by passing alternating positive and negative
current pulses ~6 1 nA, 30 ms, 100–300 pulses! through the
recording electrode. Tissue containing biocytin-filled cells was
placed in a fixative solution of 4% paraformaldehyde in 0.1 M PB
for 72 h. The LGN was sectioned ~400 mm! in the coronal plane
and processed using the ABC method ~Horikawa & Armstrong,
1988; Guido et al., 1997; Li et al., 2003!. Labeled cells were
digitally photographed and drawn using a camera lucida attached
to a microscope.
To evoke synaptic activity in LGN, single square-wave pulses
~0.1–0.3 ms, 0.1–5.0 mA! were delivered at varying rates ~0.20–
100 Hz! through a pair of low impedance ~0.2–0.5 MV!, thin
gauged Ir wires whose exposed tips ~0.1 mm! were placed on the
surface of the optic tract or on each optic nerve. To examine retinal
convergence, we first established the minimum intensity needed to
evoke a postsynaptic response. Current intensity was then adjusted
in 1, 2, 5, or 10% increments above this initial value until an
excitatory postsynaptic potential ~EPSP! of maximal amplitude
was reached ~Lo et al., 2002; Ziburkus et al., 2003!. EPSP amplitude by stimulus intensity plots were constructed and used to
estimate the number of retinal inputs converging onto a single
LGN ~Bartlett & Smith, 1999; Chen & Regehr, 2000; Lo et al.,
2002; Lu & Constantine-Paton, 2004; Ziburkus & Guido, 2005!.
Since trial-to-trial variability is relatively small ~see above references!, an individual step size was defined by a change in amplitude of ⬎1 mV.
Results
Retinal innervation of the developing LGN
We began our study by comparing the cytoarchitecture of the LGN
of a P7 and an adult mouse ~Fig. 1A!. Even at early postnatal ages,
the LGN and adjacent thalamic nuclei are readily distinguished.
However, as is the case with other rodents, the LGN lacks an
obvious lamination pattern ~Reese, 1988!. In other mammalian
species ~e.g. cat, ferret, monkey!, the LGN is partitioned into
laminae that receive input from one eye or the other. In rodents,
eye-specific patterns are visualized in LGN by the anterograde
labeling of retinal ganglion cells ~Godement et al., 1984; Jeffery,
1984; Muir-Robinson et al., 2002; Torborg & Feller, 2004;
Ziburkus & Guido, 2005!. For these experiments, we made eye
injections of Alexa Fluor 488 ~green! and Alex Flour 594 ~red!
conjugated to cholera toxin B subunit ~CTB! to visualize retinal
projections from both eyes simultaneously in single sections of the
LGN. Fig. 1B shows the pattern of retinal projections in LGN of
the left and right hemisphere at P7 and adult. At both ages, the
regions of termination for crossed ~contralateral! and uncrossed
~ipsilateral! projections are well defined ~Muir-Robinson et al.,
2002!. Contralateral eye projections occupied most of the LGN
while ones from the ipsilateral eye resided in a small patch located
in the dorsal medial region of the nucleus. However, one notable
difference was in the spatial extent of the ipsilateral pathway. At
P7, it was diffusely organized and shared common terminal space
with contralateral projections.
664
L. Jaubert-Miazza et al.
Fig. 1. Anatomical organization of the lateral geniculate nucleus ~LGN! in the C570BL6 mouse. A: Coronal sections through the LGN
showing Nissl stain at postnatal day ~P! 7 ~left! and in the adult. At P7, the LGN is readily distinguished from the intrageniculate leaflet
~IGL! and the ventral geniculate nucleus ~VLG!. The cytoarchitecture of the LGN lacks an eye-specific lamination pattern. Scale bar ⫽
100 mm. B: Anterograde labeling of retinal projections with fluorescent conjugates of cholera toxin B ~CTB! reveals eye-specific
organization at P7 and in the adult. Shown are coronal sections through the LGN of the left and right hemisphere. Alexa Fluor 594 ~red!
conjugated to CTB was injected into the left eye. It labels the terminal fields of uncrossed ~ipsilateral! retinal axons in the left
hemisphere and crossed ~contralateral! axons in the right hemisphere. Alexa Fluor 488 ~green! conjugated to CTB was injected into
the right eye. It labels contralateral axons of the left hemisphere and ipsilateral axons in the right hemisphere. Scale bar ⫽ 100 mm.
To chart the time course of axon segregation in the developing
LGN, we examined the pattern of projections at several postnatal
~P3, P7, P12, P14, P17, P19, P21, and P28! ages. Representative
examples are shown in Fig. 2A. At early ages ~P3–7!, the inputs
from the two eyes shared substantial area in LGN. This is best
illustrated in the superimposed pseudocolored images of Fig. 2A
where the red- and green-labeled pixels represent contralateral and
ipsilateral inputs, respectively, and the yellow corresponds to areas
shared by both. After P7, retinal projections from the two eyes
begin to show signs of segregation, and certainly by the time of
natural eye opening ~P12–14!, they were well segregated and
resembled the pattern found at older ages. We compared the spatial
extent of contralateral, ipsilateral, and overlapping projections at
different postnatal ages and these estimates are summarized in
Fig. 2B. Contralateral projections show modest changes, and these
were largely restricted to the first postnatal week. For example,
between P3 and P7, contralateral projections occupied 92% and
84% of LGN, respectively. In contrast, ipsilateral projections
showed substantial rearrangement throughout the first two postnatal weeks. Initially, at P3 they occupied about 59% of the LGN
and overlapped with contralateral ones by as much as 57%. At P7,
the ipsilateral projections began to recede but were still quite
robust ~26%!, sharing about 18% of LGN with contralateral projections. By P12–14, ipsilateral inputs showed further attrition,
occupying about 12% of LGN and sharing little ~3%! if any
territory with contralateral projections. These values were indistinguishable from those of older animals ~P17–28!.
The difference between unsegregated and segregated retinal
projections can be further quantified by analyzing the fluorescence
intensity of individual pixels. Such measures have advantages over
those described above because they provide an unbiased or thresholdindependent index of segregation ~Muir-Robinson et al., 2002;
Torborg & Feller, 2004!. Examples of scatterplots of pixel intensity for a section of LGN at P3 and P28 are shown in Fig. 3A. For
Developmental remodeling in the mouse LGN
665
Fig. 2. A: Retinogeniculate axon segregation in the LGN of the developing mouse. Coronal sections in the left LGN of a P3, P7, P12, P14, P21, and P28
mouse. Alexa Fluor 594 ~red! conjugated to CTB was injected into the right eye and reveals the terminal fields of crossed ~contralateral eye! retinal inputs.
Alexa Fluor 488 ~green! conjugated to CTB was injected into the left eye and reveals uncrossed ~ipsilateral eye! inputs. Panels from left to right depict red
and green fluorescence labeling from the same section of LGN, the superimposed fluorescence pattern, and corresponding pseudocolored image. Regions
of overlap are colored yellow. Scale bar ⫽100 mm. B: Summary graphs plotting the percent area in LGN occupied by the crossed ~contralateral eye, red!,
uncrossed ~ipsilateral eye, green!, and overlapping ~yellow! terminal fields as a function of age. Shown are means and SEMs ~P3 n ⫽ 3, P7 n ⫽ 3, P12
n ⫽ 5, P14 n ⫽ 9, P17 n ⫽ 4, P19 n ⫽ 5, P21 n ⫽ 4, & P28 n ⫽ 8!. Calculations are based on five 60-µm-thick sections through the middle of LGN.
Contralateral projections show a modest recession during the first week. Ipsilateral projections show substantial retraction through postnatal day 12. The
amount of shared territory declines rapidly during the first week and by P12–14, inputs segregate into nonoverlapping territories.
666
L. Jaubert-Miazza et al.
Fig. 3. Quantification of retinogeniculate axon segregation in the developing LGN. A: Scatterplots of pixel intensity for a single section of LGN at P3 and
P28. Each point represents a pixel in which the fluorescence intensity of the contralateral projection is plotted against the intensity of the ipsilateral
projection. At P3, the positive correlation between the ipsilateral and contralateral pixel intensities confirm the substantial overlap of projections from the
two eyes ~r ⫽ 0.53, total pixels in LGN ⫽ 105,511, P ⬍ 0.001!. At P28, the negative correlation between the ipsilateral and contralateral pixel intensities
confirm the complete segregation of projections from the two eyes ~r ⫽ ⫺0.73, total pixels in LGN ⫽ 227,544, P ⬍ 0.001!. B: R-distributions of pixel
intensity from a labeled section of LGN at P3 and P28. For each pixel ~see A!, the logarithm of the intensity ratio, R~log 10 FI 0FC !, is plotted as a frequency
histogram ~bin size ⫽ 0.1 log units!. Narrow R-distributions ~P3! reflect overlap and diffusely organized inputs, whereas wide distributions reflect highly
organized, segregated inputs. C: Graph showing the variance values obtained from R-distributions ~see B! at different postnatal ages. Each point depicts
the mean and SEM for ages and animals plotted in Fig. 2. Variance increases with age and corresponds to a progressive increase in eye-specific axon
segregation.
Developmental remodeling in the mouse LGN
667
Fig. 4. High-power images of LGN at P7 and adult showing the cellular labeling pattern for L-type ~alpha a1C ! Ca 2⫹ channels, SMI-32,
GAD, and NMDA. Scale bar ⫽ 10 mm.
these, each point represents a pixel in which the fluorescence
intensity of the contralateral projection is plotted against the intensity of the ipsilateral one. At P3, when contralateral and ipsilateral
inputs overlap, individual pixel intensity shows a positive correlation ~i.e. high contralateral and ipsilateral intensities in the same
pixels!. Conversely at P28, when inputs from the two eyes are
segregated, pixel intensity becomes inversely related ~i.e. high
contralateral and low ipsilateral intensities in the same pixels, or
visa versa!. Pixel intensity can also be expressed as the logarithm
of the ratio fluorescence intensities representing the ipsilateral and
contralateral projections ~R ⫽ log 10 FI 0FC , see Torborg & Feller,
2004!. R-values representing each pixel of LGN can then be
plotted as a frequency distribution. Examples of two R-distributions
for a section of LGN at P3 and P28 are shown in Fig. 3B. At P3,
the R-distribution was narrow indicating that the majority of pixels
had intensity values that were not dominated by one projection or
the other. In contrast, at P28, the distribution was wider and
bimodal, with a large peak between ⫺2 and ⫺1, and a smaller yet
distinct one between 0.5 and 1. The former represents the great
majority of pixels dominated by a high contralateral intensity
value, while the latter corresponds to a smaller group of pixels
dominated by a high ipsilateral value. We used the variance of the
R-distribution to compare the relative widths of distributions across
mice of different ages ~Torborg & Feller, 2004!. A summary plot
of the mean variance of R-distributions as a function of age is
shown in Fig. 3C. There was a steep increase in variance between
P3 ~ xS ⫽ 0.085! and P12 ~ xS ⫽ 0.47!. Values between P12–P28 show
little change, indicating that axon segregation is complete by P12–14.
668
L. Jaubert-Miazza et al.
Fig. 5. Distributions of LGN soma sizes stained for L-type Ca 2⫹ channels, SMI-32, GAD, and Nissl at P7 and adult.
Developmental remodeling in the mouse LGN
669
Structural properties of developing LGN cells
subunit of NMDA receptor. Prominent cellular staining was seen at
all ages, and examples of the labeling pattern found at P7 and in the
adult are shown in Fig. 4. Corresponding soma size distributions
for labeled cells are shown in Fig. 5. By comparing these with the
soma distributions of Nissl-stained cells, one can determine whether
labeling was confined to either relay cells ~medium and large cells!
or interneurons ~small cells! of LGN ~Bickford et al., 1998; Carden
et al., 2000; Ziburkus et al., 2000; Li et al., 2003!.
We assessed the structural composition of developing LGN using
immunocytochemical markers to label the following elements: the
pore forming a1C subunit of the L-type Ca 2⫹ channel, a nonphosphorylated form of a high-molecular-weight neurofilament protein
associated with large thalamic relay cells, the synthesizing enzyme
for GABA to label intrinsic interneurons, and the pore forming R1
Fig. 6. Density of labeled cells in the developing LGN. A: Representative coronal sections in LGN showing the cellular labeling pattern
obtained with an antibody directed toward the pore forming a1C subunit of the L-type Ca 2⫹ channel at P7, P14, and P21. Scale bar ⫽
100 mm. B: LGN cell-labeled density plots at different postnatal ages. Each point represents a cell count made from a section of LGN
of a different animal. Density of labeled cells for L-type Ca 2⫹ channels show close to a four-fold reduction with age. The density of
labeled SMI-32, GAD, and Nissl-stained cells remain relatively stable throughout postnatal development.
670
L. Jaubert-Miazza et al.
At P7, L-type Ca 2⫹ channel labeling showed dense soma and
cytosol staining. In the adult, staining was confined to soma membrane and proximal dendrites. Soma size distributions for L-type
Ca 2⫹ channels were relatively broad and resembled those of Nissl,
suggesting these channels are found on both interneurons and relay
cells. SMI-32 and NMDA labeled soma and proximal portions of
dendrites while GAD produced a dense neurophil and relatively
sparse but dense soma staining. Soma size distributions for SMI-32
indicated that staining was restricted to the largest cells of LGN and
thus labeling only relay cells ~Grossman et al., 1973, Bickford et al.,
1998; Carden et al., 2000!. In contrast, the distributions for GAD
were restricted to the smallest cells of LGN thereby labeling only
intrinsic interneurons ~Montero & Singer 1985; Li et al., 2003!.
Finally, NMDA staining was found on the soma and proximal dendrites of cells varying in size ~distribution not shown!, indicating
both relay cells and interneurons were labeled.
To determine whether the pattern of cellular staining in LGN
varied with age, we made cell density measurements at P7, P14,
P21, and adult. As shown in the plots of Fig. 6B, the density of
labeled cells for SMI-32, GAD, NMDA ~not shown!, and Nissl
remained relatively stable throughout postnatal development. However, L-type Ca 2⫹ channel staining declined steadily with age.
Examples of the labeling pattern for L-type Ca 2⫹ channels at P7,
P14, and P21 are shown in Fig. 6A. There was about a four-fold
decrease in the density of labeled cells between P7 ~ xS ⫽ 26.7
cells0100 mm 2 ! and adult ~ xS ⫽ 6.35 cells0100 mm 2 !.
Functional properties of developing LGN cells
We studied the electrophysiological properties and synaptic responses of 200 LGN cells in mice that ranged in age from P2–P21.
All cells exhibited a resting membrane level of ⫺55 to ⫺68 mV,
action potentials that exceeded 60 mV, and near-threshold synaptic
responses greater than 3 mV. During intracellular recording some
cells ~n ⫽ 10! were filled with biocytin and Fig. 7A provides
examples of camera lucida reconstructions for two cells at P8 and
P19. Overall, labeled cells had relatively large somata and multipolar dendritic arbors consistent with those of class A thalamocortical cells ~Grossman et al., 1973; Webster & Rowe, 1984!. However,
the dendritic architecture of early postnatal cells was simpler, with
arbors having fewer and shorter branches ~Parnavelas et al., 1977;
Ziburkus et al., 2003!.
Representative voltage responses to intracellular current injection and synaptic responses evoked by optic tract stimulation
recorded at P8 and P19 are shown in Figs. 7B–7C. These traces are
representative of our sample and underscore two points. First,
regardless of age LGN cells seem equipped with a full compliment
of active membrane properties ~Crunelli et al., 1987; Williams
et al., 1996; Ziburkus et al., 2003!. As shown in Fig. 7B, membrane
hyperpolarization evoked a mixed cation conductance ~H! that
results in a strong inward rectification. Termination of hyperpolarizing pulses led to passive membrane repolarization and the activation of a T-type Ca 2⫹ conductance that produced a large low
threshold Ca 2⫹ spike ~T! and burst firing ~B!. This event takes the
form of a triangular depolarization that is crowned with a highfrequency burst of action potentials. Cells also displayed a strong
outward rectification in response to membrane depolarization. This
response is attributed to the activation of an outward K⫹ conductance ~A! and leads to a delay in spike firing ~McCormick, 1991;
Li et al., 2003!. Strong and sustained levels of membrane depolarization also elicited a train of action potentials that showed spike
frequency accommodation. The latter is attributed to activation of
a K⫹ conductance that produces an after-hyperpolarizing response
between spikes ~AHP! ~Li et al., 2003!. Additionally, we found at
all ages, optic tract stimulation produced EPSPs that were followed
by inhibitory postsynaptic potential ~IPSP! activity ~Crunelli et al.,
1988; Ziburkus et al., 2003!. As shown in Fig. 7C, following the
EPSP was an early short hyperpolarizing response that reversed
near the equilibrium potential of Cl⫺ ~not shown! and corresponds
to the activation of GABAA receptors. The early IPSP was followed by a slower longer one that reversed near the equilibrium
potential of K⫹ ~not shown! and corresponds to the activation of
GABAB receptors. The passive repolarization of these long GABAB
mediated IPSPs often led to a rebound Ca 2⫹ spike and bursting.
While the functional properties of developing cells seemed
relatively mature, there were some aspects of synaptic transmission that varied considerably with age. First, was the prevalence of
synaptic responses that reflected the convergence of multiple
retinal inputs onto a single LGN cell. To obtain estimates of retinal
convergence, we activated the optic tract at various levels of
stimulus intensity and measured the amplitude of the evoked EPSP.
Representative examples at P8 and P19 are shown in Fig. 8. At
early ages, a progressive increase in stimulus intensity led to a
step-wise increase in EPSP amplitude. These graded changes
reflect the successive recruitment of different retinal inputs innervating a single LGN cell ~Chen & Regehr, 2000; Lo et al., 2002;
Ziburkus et al., 2003; see also Lu & Constantine-Paton, 2004!. At
P8, as many as nine different inputs were evident ~Fig. 8A!. In
contrast, at P19 the same progression in stimulus intensity led to
the activation of two inputs. We examined the pattern of innervation at different postnatal ages for 37 cells and the summary plot
in Fig. 8B illustrates there was a progressive decline in retinal
convergence with age ~r ⫽ 0.82, P ⬍ 0.001!. At early ages
Fig. 7. Morphological, electrophysiological, and synaptic properties of LGN cells at P8 ~left! and P19 ~right!. A: Camera lucida
reconstructions of biocytin-filled relay cells. Labeled cells have dendritic morphology consistent with class A thalamocortical cells.
Scale bar ⫽ 20 mm B: Examples of the voltage responses ~black! to current pulse injections ~red! reflect the activation of a number
of active conductances. Below each set of responses is the corresponding I–V plots measured at steady state ~250 ms!. Membrane
hyperpolarization evokes a mixed cation conductance ~H! that results in an inward rectification. Terminating the hyperpolarizing pulses
leads the activation of a T-type Ca 2⫹ conductance that produces a low threshold Ca 2⫹ spike ~T! and burst firing ~B!. Membrane
depolarization produces an outward rectification ~A! that delays spike firing. Larger depolarizing pulses can evoke spikes trains that
exhibit frequency accommodation due to the activation of an outward K⫹ conductance ~AHP!. C: Examples of postsynaptic activity
evoked by electrical stimulation of optic tract. The EPSP ~E! is followed by a pair of IPSPs ~I!, a short one that occurs just after the
EPSP, and a longer slower one that follows the early IPSP. The passive repolarization of the membrane associated with the termination
of the late IPSP leads to rebound LT spike and bursting. Resting membrane levels recorded between ⫺58 and ⫺70 mV.
Developmental remodeling in the mouse LGN
FIGURE 7
671
672
L. Jaubert-Miazza et al.
Fig. 8. Degree of retinal convergence in the developing LGN. A: Examples of synaptic responses at P8 and P19 evoked by stimulating
optic tract ~OT! fibers at progressively higher levels of stimulation. Below each set of responses are the corresponding amplitude by
stimulus intensity plots. Each step is numbered and corresponds to the recruitment of a separate input. At P8, an incremental change
in stimulus intensity leads to nine step-wise increases in EPSP amplitude. At P19, increases in stimulus intensity produce two steps.
B: Scatterplot depicting the number of retinal inputs LGN cells receive as a function of age. Each point depicts a single cell. For the
37 cells tested, there is a significant decrease ~r ⫽ 0.82, P ⬍ 0.001! in retinal convergence with age.
~P6–10!, cells typically received 8–11 inputs. However, between
P14–21 the degree of retinal convergence declined so that LGN
cells received input from just a few retinal ganglion cells. The
second transient feature was the presence of binocular responses.
At two representative ages ~P8 and P19!, we preserved large
segments of each optic nerve and shocked each one independently.
Examples of optic nerve evoked responses are shown in Fig. 9. At
P8, separate and distinct EPSPs were frequently ~6 of 12 cells!
encountered by stimulating the contralateral as well as the ipsilat-
eral optic nerve. However, when tested at P19, binocular responses
were not apparent. All cells tested ~n ⫽ 8! were monocular and
responded only to contralateral optic nerve stimulation. The final
transient feature was the prevalence of synaptically evoked plateau
potentials ~Lo et al., 2002!. At early ages strong activation of
optic tract fibers with either a single or repetitive ~25–100 Hz!
shock often evoked EPSPs that gave rise to a high amplitude
~25– 40 mV!, long-lasting ~300–1300 ms!, slow decaying depolarization. Fig. 10A provides examples of these plateau potentials
Developmental remodeling in the mouse LGN
673
Fig. 9. Binocular and monocular responses in the
developing LGN. Example of a binocular response
at P8 evoked by the separate stimulation of the
contralateral and ipsilateral optic nerve. Example
of a monocular response at P19 evoked by contralateral optic nerve stimulation. Responses recorded at ⫺60 mV.
~Lo et al., 2002! for a cell recorded at P8. Also shown are the
synaptic responses at P19, in which a similar form of stimulation
failed to evoke such a response. The summary plot in Fig. 10B
~n ⫽ 141 cells! indicates the incidence of plateau potentials decreased with age. Between P2– 4, the majority of encountered cells
~70% for single shock and 90% for repetitive shocks! exhibited
synaptically evoked plateau potentials. By P14–16, the incidence
showed a substantial decline ~0% for single shock and 50% for
repetitive shocks!. Bath application of nimodipine ~10 mM! com-
pletely abolished plateau potentials ~n ⫽ 3 not shown, see also Lo
et al., 2002! indicating these events are mediated by the activation
of L-type Ca 2⫹ channels.
Discussion
Our results indicate the retinogeniculate pathway of the mouse
undergoes a significant period of anatomical and functional remodeling during early postnatal life. At P3, retinal projections
Fig. 10. Synaptically evoked plateau potentials in the developing LGN. A: Examples of synaptic responses evoked by single or
repetitive stimulation of optic tract. At P8, single ~left! or repetitive ~right, 100 Hz! optic tract stimulation evokes a high-amplitude,
sustained-plateau potential. At P19, strong delivery of a single shock or repetitive stimulation ~20 Hz! fails to evoke plateau potential.
B: Summary graph plotting the incidence of plateau potentials at different ages. Each point depicts the percentage of cells exhibiting
a plateau response evoked by single ~open circles! or repetitive ~25–100 Hz! stimulation. All responses recorded between ⫺60 and
⫺67 mV.
674
from the two eyes share common terminal space in LGN. However, by natural eye opening ~P12–14! these initially diffuse projections retract and segregate to form distinct eye specific domains.
At early ages, LGN cells are binocularly responsive and receive
input from many retinal ganglion cells. Retinal afferent stimulation
also evokes plateau-like depolarizations that are mediated by the
activation of L-type Ca 2⫹ channels. As retinal projections from the
two eyes retract and segregate into nonoverlapping territories,
there is a loss of binocular responsiveness, a decrease in retinal
convergence, and a reduction in the incidence of plateau potentials.
Interestingly, many of the active membrane properties, repetitive
firing characteristics, and basic EPSP0IPSP responses of developing LGN cells are remarkably mature and remain stable throughout
this remodeling period.
A previous study using single eye injections of HRP to label
retinogeniculate projections concluded that eye-specific segregation occurs within the first week of life ~Godement et al., 1984!.
However, our results along with those of Muir-Robinson et al.
~2002! indicate that segregation occurs more gradually, extending
well into the second week and reaching an adult-like pattern by the
time of natural eye opening ~P12–14!. This discrepancy may be
attributed to two factors. First, the use of a binocular fluorescence
labeling technique that makes possible the separate visualization of
projections from the two eyes in a single section of LGN. Secondly, the adoption of quantitative methods to analyze the spatial
extent and pixel intensity of labeled retinal projections in LGN.
Indeed, the application of threshold-independent methods to analyze pixel intensity has proven to be a valuable tool for the
quantitative assessment of segregation in normal developing mice
as well as genetically altered ones ~Torborg & Feller, 2004!.
Our results also confirm a report by Muir-Robinson et al.
~2002! who suggested that segregation occurs in two stages.
During the first week, a more macroscopic level of organization is
achieved where the final positioning of contralateral and ipsilateral
projections are established and the initial pruning of arbors begin.
Between the first and second week, a more focalized form of
retraction occurs, as ipsilateral projections undergo extensive attrition and discrete nonoverlapping fields are formed. These events
appear to be regulated by two types of spontaneous retinal activity;
an early phase ~P0–8! of cholinergic transmission that contributes
to a large-scale regionalization of eye specific territories, and a late
one ~P10–P14! involving glutamate signaling that drives local
patterns of segregation ~Feller, 2002; Muir-Robinson et al., 2002!.
It is important to note that the anatomical rearrangements in
LGN translate directly into functional changes in connectivity.
Initially, overlapping projections from the two eyes provide LGN
cells with as many as 11 separate inputs ~see also Lo et al., 2002;
Ziburkus & Guido, 2005 for similar estimates in the rat!. Indeed,
our estimates in rodents seem conservative compared to a report in
mouse showing that relay cells receive in excess of 20 retinal
inputs even during the second postnatal week ~Chen & Regehr,
2000!. Despite these differences, it is clear that as retinal projections from the two eyes recede and overlapping territories dissipate, synapses are eliminated and cells receive far fewer inputs
~Chen & Regehr, 2000; Lo et al., 2002! from just one eye ~Shatz
& Kirkwood, 1984; Ziburkus et al., 2003; Ziburkus & Guido,
2005!. This form of synaptic refinement correlates well with the
maturation of receptive-field properties reported in a number of
mammalian species. Immature receptive fields are large, irregularly shaped and lack distinct on- and off-subregions. Mature fields
are much smaller, have well-defined concentric center-surround
organization, and are dominated by input from 1–3 retinal ganglion
L. Jaubert-Miazza et al.
cells ~Tootle & Friedlander, 1986; Mastronarde, 1987; Usrey et al.,
1999; Tavazoie & Reid, 2000; Grubb & Thompson, 2003!.
An intriguing aspect of the present results is that functional
changes in connectivity seem to persist even after the anatomical
segregation of retinal inputs is nearly complete ~see also Chen &
Regehr, 2000!. Similar observations have been made in other
species. In the cat, a high incidence of binocular responses is
evident at birth ~Shatz & Kirkwood, 1984!, a time when retinal
axons have sorted into eye specific layers ~Shatz, 1983!. In the rat,
binocular responses ~Ziburkus et al., 2003, Ziburkus & Guido,
2005! and a high degree of retinal convergence ~Lo et al., 2002;
Ziburkus & Guido, 2005! are still present at the time of natural eye
opening when eye-specific zones are newly established. Taken
together, these results suggest the bulk labeling of retinal terminal
fields cannot by itself be used to fully resolve the synaptic rearrangements of individual relay cells. Clarification of this issue
would seem to require single axon labeling to be done in conjunction with an ultrastructural analysis of developing retinogeniculate
synapses.
Finally, it is important to consider that the developmental
remodeling of retinogeniculate pathway coincides with the expression and activity of L-type Ca 2⫹ channels occurring in LGN. The
density of cells expressing L-type Ca 2⫹ channels and the incidence
of L-type mediated activity in the form of synaptically evoked
plateau potentials are at their highest at early postnatal ages. Both
events then show age-related decreases that parallel axon segregation and synaptic refinement. A similar trend exists in the rat
where it was shown that plateau potentials are evoked by strong
and0or repetitive stimulation of retinal afferents, have a voltage
dependency consistent with the high threshold voltage-gated L-type
Ca 2⫹ channel, and are completely abolished by L-type antagonists
~Lo et al., 2002!. Other events that seem to contribute to the
developmental regulation of L-type activity is the strong postsynaptic depolarization provided by the summation of convergent,
multiple retinal inputs and high-frequency discharge patterns of
spontaneously active retinal ganglion cells ~Mooney et al., 1996;
Lo et al., 2002!.
Taken together these results indicate that developing LGN cells
can readily acquire significant amounts of Ca 2⫹ through L-type
channels. Indeed, NMDA receptors need not be the sole source of
Ca 2⫹ acquired during synaptic transmission, but a much larger and
longer influx can actually occur via the activation of L-type Ca 2⫹
channels ~Budde et al., 1998!. The L-type activity recorded in the
rodent LGN is identical to the synaptically evoked plateau potentials recorded in neurons of the developing rodent superior colliculus ~Lo & Mize, 2000! and brainstem trigeminal nuclei ~Lo &
Erzurumlu, 2002! and shares some similarities to plateau-related
activity reported in the rodent brainstem ~Rekling & Feldman,
1997!, spinal cord ~Kiehn & Eken, 1998!, and invertebrate motor
neurons ~Dicaprio, 1997!. Thus, this event may reflect a highly
conserved mechanism by which cells can acquire large amounts of
Ca 2⫹ in an activity dependent manner. A role for L-type Ca 2⫹
channels in synaptic plasticity has also been well documented.
Activity through these channels can induce long-term depression
and potentiation in the hippocampus ~Magee & Johnston, 1997!,
superior colliculus ~Lo & Mize, 2000!, and the principal nucleus of
brainstem ~Guido et al., 2001!. Retinal axons also fail to segregate
properly in the developing LGN and superior colliculus of transgenic mice that show reduced levels of L-type channel activity
~Cork et al., 2001!.
The Ca 2⫹ influx associated with the synaptically evoked plateau potential could also contribute to signaling events and gene
Developmental remodeling in the mouse LGN
expression involved in the stabilization of developing connections
~Ghosh & Greenberg, 1995; Greenberg & Ziff, 2001!. For example, Ca 2⫹ entry via L-type channels favors signaling cascades
brought on by heightened periods of neural activity ~Mermelstein
et al., 2000!. One in particular involves the cAMP response
element ~CRE0CREB! transcription pathway. The CRE binding
protein ~CREB! is a calcium and cAMP-regulated transcriptional
activating protein shown to be important for thalamic circuit
development and retinogeniculate axon segregation ~Pham et al.,
2001!. Future studies that target the role of L-type Ca 2⫹ channels
should shed light on the molecular mechanisms involved in mediating the activity-dependent remodeling of retinogeniculate
connections.
Acknowledgments
This study was supported by grants from the National Eye Institute ~RO1
12716! and the Whitehall Foundation. We thank R. Couvillion and P.
Kinney for their technical assistance. We also thank M. Feller for her
assistance and advice in quantifying CTB labeling patterns in LGN.
Finally, a special thanks to M. McCall and P. Nashina for their efforts in
organizing the workshop entitled The Laboratory Mouse in Vision Research sponsored by The Jackson Laboratory and National Eye Institute.
References
Bartlett, E. & Smith, P.H. ~1999!. Anatomic, intrinsic, and synaptic
properties of dorsal and ventral division in rat medial geniculate body.
Journal of Neurophysiology 81, 1999–2016.
Bickford, M.E., Guido, W. & Godwin, D.W. ~1998!. Neurofilament
proteins in Y-cells of the cat lateral geniculate nucleus: Normal expression and alteration with visual deprivation. Journal of Neuroscience
18~16!, 6549– 6557.
Budde, T., Munsch, T. & Pape, H.C. ~1998!. Distribution of L-type
calcium channels in rat thalamic neurones. European Journal of Neuroscience 10~2!, 586–597.
Carden, W.B., Guido, W., Ziburkus, J., Datskovskaia, A., Godwin,
D.W. & Bickford, M.E. ~2000!. A novel means of Y cell identification
in the developing lateral geniculate nucleus of the cat. Neuroscience
Letters 295, 5–8.
Chapman, B. ~2000!. Necessity for afferent activity to maintain eyespecific segregation in ferret lateral geniculate nucleus. Science 287,
2479–2482.
Chen, C. & Regehr, W.G. ~2000!. Developmental remodeling of the
retinogeniculate synapse. Neuron 28~3!, 955–966.
Cork, R.J., Namkung, Y., Shin, H.S. & Mize, R.R. ~2001!. Development
of the visual pathway is disrupted in mice with a targeted disruption of
the calcium channel beta~3!-subunit gene. Journal of Comparative
Neurology 440, 177–191.
Cramer, K.S. & Sur, M. ~1995!. Activity dependent remodeling of
connections in the mammalian visual system. Current Opinion in
Neurobiology 5~1!, 106–111.
Crunelli, V., Haby, M., Jassik-Gerschenfeld, D., Leresche, N. &
Pirchio, M. ~1988!. Cl- and K⫹ dependent inhibitory postsynaptic
potentials evoked by interneurones of the rat lateral geniculate nucleus.
Journal of Physiology 399, 153–176.
Crunelli, V., Kelly, J.S., Leresche, N. & Pirchio, M. ~1987!. The
ventral and lateral geniculate nucleus of the rat: Intracellular recordings. Journal of Physiology 384, 587– 601.
Demas, J., Eglen, S.J. & Wong, R.O. ~2003!. Developmental loss of
synchronous spontaneous activity in the mouse retina is independent of
visual experience. Journal of Neuroscience , 2851–2860.
Dicaprio, R. ~1997!. Plateau potentials in motor neurons in the ventilatory
system of the crab. Journal of Experimental Biology 200, 1725–1736.
Feller, M.B. ~2002!. The role of nAChR-mediated spontaneous retinal
activity in visual system development. Journal of Neurobiology 53,
556–567.
Feller, M.B., Wellis, D.P., Stellwagen, D., Werblin, F.S. & Shatz,
C.J. ~1996!. Requirement for cholinergic synaptic transmission in the
propagation of spontaneous retinal waves. Science 272, 1182–1187.
675
Ghosh, A. & Greenberg, M.E. ~1995!. Calcium signaling in neurons:
Molecular mechanisms and cellular consequences. Science 268,
239–247.
Godement, P., Salaun, J. & Imbert, M. ~1984!. Prenatal and postnatal
development of retinogeniculate and retinocollicular projections in the
mouse. Journal of Comparative Neurology 230~4!, 552–575.
Goodman, C.S. & Shatz, C.J. ~1993!. Developmental mechanisms that
generate precise patterns of neuronal connectivity. Cell 72 ~Suppl.!
77–98.
Greenberg, M.E. & Ziff, E.B. ~2001!. Activity dependent regulation of
gene expression. In Synapses, eds. Cowan, W.M., Sudhof, T.C. &
Stevens, C.F., pp. 357–392. Baltimore, Maryland: Johns Hopkins
University Press.
Grossman, A., Lieberman, A.R. & Webster, K.E. ~1973!. A Golgi study
of the rat lateral geniculate nucleus. Journal of Comparative Neurology
150, 441– 446.
Grubb, M.S., Rossi, F.M., Changeux, J.-P. & Thompson, I.D. ~2003!.
Abnormal functional organization in the dorsal lateral geniculate nucleus of mice lacking the beta 2 subunit of the nicotinic acetylcholine
receptor. Neuron 40, 1161–1172.
Grubb, M.S. & Thompson, I.D. ~2003!. Quantitative characterization of
visual response properties in the mouse dorsal lateral geniculate nucleus. Journal of Neurophysiology 90, 3594–3607.
Grubb, M.S. & Thompson, I.D. ~2004!. The influence of early experience
on the development of sensory systems. Current Opinion in Neurobiology 14, 503–512.
Guido, W., Günhan-Agar, E. & Erzurumlu, R.S. ~1998!. Developmental changes in the electrophysiological properties of brainstem
trigeminal neurons during pattern ~barrelette! formation. Journal of
Neurophysiology 79, 1295–1306.
Guido, W., Lo, F.S. & Erzurumlu, R.S. ~1997!. An in vitro model of
kitten retinogeniculate pathway. Journal of Neurophysiology 77,
511–516.
Guido, W., Lo, F.S. & Erzurumlu, R.S. ~2001!. Synaptic plasticity in the
trigeminal nucleus during the period of barrelette formation and consolidation. Developmental Brain Research 132, 97–102.
Guido, W., Ziburkus, J. & Lo, F.S. ~2003!. Synaptic plasticity in the
developing visual thalamus. In The Brain and Sensory Plasticity:
Language Acquisition and Hearing, ed. Berlin, C.I. & Weyand, T.G.,
pp. 75–100. New York: Thompson Delmar Learning Press.
Horikawa, K. & Armstrong, W.E. ~1988!. A versatile means of intracellular labeling: Injection of biocytin and its detection with avidin
conjugates. Journal of Neuroscience Methods 25, 1–11.
Hu, B. ~1993!. Membrane potential oscillations and corticothalamic connectivity in rat associational thalamic neurons in vitro. Acta Physiologica Scandanavia 148, 109–113.
Huberman, A.D., Stellwagen, D. & Chapman, B. ~2002!. Decoupling
eye specific segregation from lamination in the lateral geniculate
nucleus. Journal of Neuroscience 22~21!, 9419–9429.
Huh, G.S., Boulanger, L.M., Du, H., Riquelme, P.A., Brotz, T.M. &
Shatz, C.J. ~2000!. Functional requirement for class I MHC in CNS
development and plasticity. Science 290~5499!, 2155–2159.
Jeffery, G. ~1984!. Retinal ganglion cell death and terminal field retraction in the developing rodent visual system. Developmental Brain
Research 13, 81–96.
Kiehn, O. & Eken, T. ~1998!. Functional role of plateau potentials in
vertebrate motor neurons. Currrent Opinion in Neuroscience 8, 746–752.
Li, J., Bickford, M.E. & Guido, W. ~2003!. Distinct firing properties of
higher order thalamic relay neurons. Journal of Neurophysiology 90,
291–199.
Lo, F.S. & Erzurumlu, R.S. ~2002!. L-type calcium channel-mediated
plateau potentials in barrelette cells during structural plasticity. Journal
of Neurophysiology 88, 794–801.
Lo, F.S. & Mize, R.R. ~2000!. Synaptic regulation of L-type Ca 2⫹ channel
activity and long-term depression during refinement of the retinocollicular pathway in developing rodent superior colliculus. Journal of
Neuroscience 20, 1– 6.
Lo, F.S., Ziburkus, J. & Guido, W. ~2002!. Synaptic mechanisms regulating the activation of a Ca 2⫹ mediated plateau potential in developing
relay cells of the lateral geniculate nucleus. Journal of Neurophysiology
87, 1175–1185.
Lu, W. & Constantine-Paton, M. ~2004!. Eye opening rapidly induces
synaptic potentiation and refinement. Neuron 43, 237–249.
Maffei, L. & Galli-Resta, L. ~1990!. Correlation in the discharges of
neighboring rat retinal ganglion cells during prenatal life. Proceedings
of the National Academy of Sciences of the U.S.A. 87~7!, 2861–2864.
676
Magee, J.C. & Johnston, D. ~1997!. A synaptically controlled, associative
signal for Hebbian plasticity in hippocampal neurons. Science 275~5297!,
209–213.
Mastronarde, D.N. ~1987!. Two classes of single input X cells in cat
lateral geniculate nucleus. II Retinal inputs and the generation of
receptive field properties. Journal of Neurophysiology 57, 381– 413.
McCormick, D.A. ~1991!. Functional properties of a slowly inactivating
potassium current in guinea pig dorsal lateral geniculate relay neurons.
Journal of Neurophysiology 66, 1176–1189.
Meister, M., Wong, R.O., Baylor, D.A. & Shatz, C.J. ~1991!. Synchronous bursts of action potentials in ganglion cells of the developing
mammalian retina. Science 252, 939–943.
Mermelstein, P.G., Bito, H., Deisseroth, K. & Tsien, R.W. ~2000!.
Critical dependence of cAMP response element-binding protein phosphorylation on L-type calcium channels supports a selective response
to EPSPs in preference to action potentials. Journal of Neuroscience
20~1!, 266–273.
Montero, V.M. & Singer, W. ~1985!. Ultrastructural identification of
somata and neural processes immunoreactive to antibodies against
glutamic acid decarboxylase ~GAD! in the dorsal lateral geniculate of
the cat. Experimental Brain Research 59, 151–165.
Mooney, R., Madison, D.V. & Shatz, C.J. ~1993!. Enhancement of
transmission at the developing retinogeniculate synapse. Neuron 10~5!,
815–825.
Mooney, R., Penn, A.A., Gallego, R. & Shatz, C.J. ~1996!. Thalamic
relay of spontaneous retinal activity prior to vision. Neuron 17~5!,
863–874.
Muir-Robinson, G., Hwang, B.J. & Feller, M.B. ~2002!. Retinogeniculate axons undergo eye specific segregation in the absence of eye
specific layers. Journal of Neuroscience 22~13!, 5259–5264.
Parnavelas, J.G., Mounty, E.J., Bradford, R. & Lieberman, A.R.
~1977!. The postnatal development of neurons in the dorsal lateral
geniculate nucleus of the rat: A Golgi study. Journal of Comparative
Neurology 171, 481–500.
Penn, A.A., Riquelme, P.A., Feller, M.B. & Shatz, C.J. ~1998!. Competition in retinogeniculate patterning driven by spontaneous activity.
Science 279~5359!, 2108–2112.
Pham, T.A., Rubenstein, J.L., Silva, A.J., Storm, D.R. & Stryker, M.P.
~2001!. The CRE0CREB pathway is transiently expressed in thalamic
circuit development and contributes to refinement of retinogeniculate
axons. Neuron 31~3!, 409– 420.
Rekling, J.C. & Feldman, J.L. ~1997!. Calcium-dependent plateau potentials in rostral ambiguous neurons in the newborn mouse brain stem
in vitro. Journal of Neurophysiology 78, 2483–2492.
Reese, B.E. ~1988!. ‘Hidden lamination’ in the dorsal lateral geniculate
nucleus: The functional organization of this thalamic region in the rat.
Brain Research 472~2!, 119–137.
Rossi, F.M., Pizzorusso, T., Porciatti, V., Marubio, L.M., Maffei, L. &
Changeux, J.P. ~2001!. Requirement of the nicotinic acetylcholine
receptor beta 2 subunit for the anatomical and functional development
of the visual system. Proceedings of the National Academy of Sciences
of the U.S.A. 98~11!, 6453– 6458.
L. Jaubert-Miazza et al.
Shatz, C.J. ~1983!. The prenatal development of the cat’s retinogeniculate
pathway. Journal of Neuroscience 3, 482– 489.
Shatz, C.J. ~1990!. Impulse activity and the patterning of connections
during CNS development. Neuron 5~6!, 745–756.
Shatz, C.J. ~1996!. Emergence of order in visual system development.
Proceedings of the National Academy of Sciences of the U.S.A. 93~2!,
602– 608.
Shatz, C.J. & Kirkwood, P.A. ~1984!. Prenatal development of functional
connections in the cat’s retinogeniculate pathway. Journal of Neuroscience 3, 482– 489.
Shatz, C.J. & Stryker, M.P. ~1988!. Prenatal tetrodotoxin infusion blocks
segregation of retinogeniculate afferents. Science 242~4875!, 87–89.
Sherman, S.M. & Spear, P.D. ~1982!. Organization of visual pathways in
normal and visually deprived cats. Physiological Reviews 62~2!, 738–855.
Sretavan, D.W., Shatz, C.J. & Stryker, M.P. ~1988!. Modification of
retinal ganglion cell axon morphology by prenatal infusion of tetrodotoxin. Nature 336~6198!, 468– 471.
Stellwagen, D. & Shatz, C.J. ~2002!. An instructive role for retinal
waves in the development of retinogeniculate connectivity. Neuron 33,
357–367.
Tavazoie, S.F. & Reid, R.C. ~2000!. Diverse receptive fields in the lateral
geniculate nucleus during thalamocortical development. Nature Neuroscience 3~6!, 608– 616.
Tootle, J.S. & Friedlander, M.J. ~1986!. Postnatal development of
receptive field surround inhibition in kitten dorsal lateral geniculate
nucleus. Journal of Neuroscience 56, 523–541.
Torborg, C.L. & Feller, M.B. ~2004!. Unbiased analysis of bulk axonal
segregation patterns. Journal of Neuroscience Methods 135, 17–26.
Torborg, C.L., Hansen, K.A. & Feller, M.B. ~2005!. High frequency,
synchronized bursting drives eye-specific segregation of retinogeniculate projections. Nature Neuroscience 8~1!, 72–78.
Usrey, W.M., Reppas, J.B. & Reid, C.R. ~1999!. Specificity and strength
of retinogeniculate connections. Journal of Neurophysiology 82,
3527–3540.
Webster, M.J. & Rowe, M H. ~1984!. Morphology of identified relay cells
and interneurons in the dorsal lateral geniculate nucleus. Experimental
Brain Research 56, 468– 474.
Williams, S.R., Turner, J.P., Anderson, C.M. & Crunelli, V. ~1996!.
Electrophysiological and morphological properties of interneurons in
the rat dorsal lateral geniculate nucleus. Journal of Physiology 490,
129–147.
Wong, R.O. ~1999!. Retinal waves and visual system development. Annual
Review of Neuroscience 22, 29– 47.
Ziburkus, J., Bickford, M.E. & Guido, W. ~2000!. NMDAR-1 staining in
the lateral geniculate nucleus of normal and visually deprived cats.
Visual Neuroscience 17, 187–196.
Ziburkus, J. & Guido, W. ~2005!. Loss of binocular responses and
reduced retinal convergence during the period of retinogeniculate axon
segregation. Journal of Neurophysiology (accepted with revision).
Ziburkus, J., Lo, F.-S. & Guido, W. ~2003!. Nature of inhibitory postsynaptic activity in developing relay cells of the lateral geniculate nucleus.
Journal of Neurophysiology 90, 1063–1070.