THE JOURNAL OF COMPARATIVE NEUROLOGY 502:339 –357 (2007)

Area Map of Mouse Visual Cortex

QUANXIN WANG AND ANDREAS BURKHALTER*
Department of Anatomy and Neurobiology, Washington University School of Medicine,
St. Louis, Missouri 63110

ABSTRACT
It is controversial whether mouse extrastriate cortex has a “simple” organization in
which lateral primary visual cortex (V1) is adjoined by a single area V2 or has a “complex”
organization, in which lateral V1 is adjoined by multiple distinct areas, all of which share the
vertical meridian with V1. Resolving this issue is important for understanding the evolution
and development of cortical arealization. We have used triple pathway tracing combined with
receptive field recordings to map azimuth and elevation in the same brain and have refer-
enced these maps against callosal landmarks. We found that V1 projects to 15 cortical fields.
At least nine of these contain maps with complete and orderly representations of the entire
visual hemifield and therefore represent distinct areas. One of these, PM, adjoins V1 at the
medial border. Five areas, P, LM, AL, RL, and A, adjoin V1 on the lateral border, but only LM
shares the vertical meridian representation with V1. This suggests that LM is homologous to
V2 and that the lateral extrastriate areas do not represent modules within a single area V2.
Thus, mouse visual cortex is “simple” in the sense that lateral V1 is adjoined by a single
V2-like area, LM, and “complex” in having a string of areas in lateral extrastriate cortex,
which receive direct V1 input. The results suggest that large numbers of areas with topolog-
ically equivalent maps of the visual field emerge early in evolution and that homologous areas
are inherited in different mammalian lineages. J. Comp. Neurol. 502:339 –357, 2007.
© 2007 Wiley-Liss, Inc.

Indexing terms: extrastriate visual cortex; visuotopic organization; cortical connections; visual
receptive fields

Mammalian visual cortex is broadly subdivided into cortex (Sur and Rubenstein, 2005; Mallamaci and
striate (primary visual area, V1) and extrastriate cortex, Stoykova, 2006; Rash and Grove, 2006). The investiga-
which surrounds V1. Extrastriate cortex is further parti- tions, however, have been limited to primary sensory ar-
tioned into multiple distinct areas, each of which contains eas, because there is no generally accepted area map of
a map of the visual field and has distinct cytoarchitecture, mouse neocortex. The urgent need for such a map is high-
connections, and functional properties. Striate and extra- lighted by the recent publication of the Allen Brain Atlas
striate areas are linked through feedforward and feedback (http://www.brain-map.org), which shows the expression
connections to a hierarchical network, which is the sub- patterns of 21,000 genes in a largely featureless layered
strate for perception and cognitive processing of visual sheet of mouse cerebral cortex (Brill, 2006).
information (Felleman and Van Essen, 1991; Coogan and The classic cytoarchitectonic map of mouse cortex shows
Burkhalter, 1993). that V1 is surrounded by five distinct areas (Rose, 1929).
Modeling studies support the notion suggested by All-
man and Kaas (1974) that topographic maps and areas
evolved under the pressure to minimize the length of
Grant sponsor: National Institutes of Health; Grant number: EY-05935;
connections between neurons with similar functional pref- Grant number: EY-016184; Grant number: HFSP123/2000-B; Grant spon-
erences (Chklovskii and Koulakov, 2004). Because the sor: McDonnell Center for Studies of Higher Brain Function.
formation of areas, creation of functional circuits, and *Correspondence to: Andreas Burkhalter, Department of Anatomy and
increased functional capacity go hand in hand (Changizi Neurobiology, Box 8101, Washington University School of Medicine, 660
South Euclid Avenue, St. Louis, MO 63110.
and Shimojo, 2005), much research has concentrated on E-mail: [email protected]
understanding the molecular mechanisms underlying the Received 26 July 2006; Revised 13 October 2006; Accepted 16 November
generation of cortical areas during development. Most 2006
studies have used molecular genetic approaches and have, DOI 10.1002/cne.21286
therefore, focused on the development of mouse cerebral Published online in Wiley InterScience (www.interscience.wiley.com).

© 2007 WILEY-LISS, INC.

 The Journal of Comparative Neurology. DOI 10.1002/cne

340 Q. WANG AND A. BURKHALTER

A more recent map shows that area 17 is adjoined by only implied that mouse extrastriate visual cortex resembles
two areas: 18a laterally and 18b medially (Caviness, the “complex” organization of rat; however, whether
1975). By using electrophysiological mapping of receptive mouse extrastriate cortex has a “simple” organization has
fields, Wagor et al. (1980) further subdivided 18a into area not been ruled out directly.
V2 and a more lateral area V3, whereas 18b was split into Here, we have traced the connections from three points
a medial rostral (Vm-r) and a medial caudal (Vm-c) area. of V1 with distinct receptive field locations in the same
Yet another scheme lumped Wagor’s V2 and V3 (Wagor et mouse and have mapped the topography of projections and
al., 1980) into a single area V2L, whereas Vm-r and Vm-c visual receptive fields in extrastriate visual cortex relative
was merged into a single area, V2ML (Paxinos and Frank- to fixed myeloarchitectonic and callosal landmarks. The
lin, 2001). As a result, V1 is almost completely surrounded results show that V1 is surrounded by at least nine dis-
by V2L and V2ML, except for the caudomedial corner, tinct areas, which contain complete, orderly maps of the
which is adjoined by V2MM (Paxinos and Franklin, 2001). visual field. All of these maps are registered across areal
Although V2L and V2ML are cytoarchitectonically uni- borders to minimize topological discontinuities.
form, single-tracer injections into mouse V1 label multiple
sites in lateral and medial extrastriate cortex (Olavarria
et al., 1982; Simmons et al., 1982; Olavarria and Montero, MATERIALS AND METHODS
1989). Similar projection patterns were found in lateral
Experiments were performed on 60 –90-day-old
extrastriate cortex of gray squirrels, which were inter-
C57BL/6J mice. All experimental procedures were ap-
preted as connections to repeating modules within a single
proved by the Washington University Animal Studies
area, V2 (Kaas et al., 1989). In contrast, Montero (1993)
Committee.
proposed, based on topographic mapping of connections in
rat, that each V1 projection target that adjoins the lateral Anatomical mapping of V1 connections
border of V1 represents a separate area and that all share
the vertical meridian with V1. Whether the organization Elevation and azimuth were mapped by simultaneously
of mouse extrastriate visual cortex is “simple” (Kaas et al., tracing axonal connections from three different points of
1989) or “complex” (Montero, 1993) is important to know, V1 with different fluorescent dextrans. In the same ani-
because a “complex” organization suggests that primate mal, retrograde tracing with bisbenzimide was used to
extrastriate visual areas are homologues of ancestral ar- label callosal connections, which served as fixed land-
eas in rodent visual cortex (Rosa and Krubitzer, 1999). marks for referencing the dextran-labeled connections. To
Although Montero’s (1993) study supports the presence of make dextran injections at known retinotopic locations,
multiple areas in rat extrastriate visual cortex, it does not animals were injected with atropine (2.5 mg/kg, SQ) and
directly demonstrate that the extrastriate projections anesthetized with urethane (1.2 g/kg, IP). The left V1 was
from different retinotopic coordinates in V1 vary relative exposed at different locations by making small (0.1 ⫻ 0.1
to fixed landmarks and that the location of the proposed mm) openings in the skull with sharp-tipped surgical
extrastriate maps are fixed relative to these landmarks. In blades (Feather No.11; Electron Microscopy Sciences, Fort
addition, how the anatomical maps relate to receptive field Washington, PA). To map receptive fields in V1, visually
maps has not been studied. Olavarria and Montero (1989) evoked action potentials of small clusters of neurons were
recorded, using conventional extracellular recording tech-
niques (see below). The receptive field map was used to
select three topographically distinct sites for simultaneous
injection with biotinylated dextran amine (BDA; 10,000
Abbreviations MW, Invitrogen, Eugene, OR; 10% in 0.01 M phosphate
buffer, pH 7.4), fluororuby (FR; dextran tetramethylrho-
A anterior area damine; 10,000 MW; Invitrogen; 5% in H2O), and fluoro-
AL anterolateral area
AM anteromedial area emerald (FE; dextran fluorescein; 10,000 MW; Invitrogen;
Au auditory cortex 5% in H2O). Injections were performed iontophoretically
Cg1 cingulate area 1 (5 ␮A, positive current, 7-second duty cycle, 10 minutes)
ENT entorhinal cortex 300 –350 ␮m below the pial surface, with glass pipettes
HMcl caudolateral branch of horizontal meridian
HMml mediolateral branch of horizontal meridian with tip inner diameters of 15–20 ␮m. For labeling cal-
HMrm rostromedial branch of horizontal meridian losal connections, a large opening was made on the right
LEA lateral entorhinal areas side between lambdoid suture, bregma, and the parietal
LI laterointermediate area ridge. Multiple pressure injections (20 nl each; Pico-
LL laterolateral area
LLA laterolateral anterior area spritzer III; Parker-Hannifin Corp., Fairfield, NJ) of bis-
LM lateromedial area benzimide (Sigma, St. Louis, MO; 5% in H2O) were dis-
MEA medial entorhinal area tributed in a 0.5 ⫻ 0.5 mm grid across the occipital,
MM mediomedial area parietal, and temporal cortices. Postsurgical survival was
P posterior area
PL posterolateral area 2–3 days.
PM posteromedial area
POR postrhinal area Electrophysiological mapping of receptive
RL rostrolateral area fields
RSA retrosplenial agranular cortex
S1 somatosensory area 1 Visual stimulation. Visual stimulation was per-
V1 visual area 1 formed through the right eye. A translucent spherical
V2 visual area 2
Vm-c medial visual area, caudal subdivision
dome (27.5 cm in diameter) was used as projection screen.
Vm-r medial visual area, rostral subdivision The dome was centered on the right eye. Eyelids were held
36p posterior area 36 open with loose sutures, and, after dilating the pupil by
The Journal of Comparative Neurology. DOI 10.1002/cne

MOUSE VISUAL CORTEX AREA MAP 341

topical application of Topicamide (1%; Bausch and Lomb, were assessed by listening to an audiomonitor. X, Y, and Z
Tampa, FL) and phenylephrine hydrochloride (2.5%; coordinates of recording sites were displayed by the opti-
Bausch and Lomb) and protecting the cornea with oph- cal encoder of a micromanipulator (MP-225; Sutter Instru-
thalmic ointment (Puralube Vet; Pharmaderm, Melville, ments, Novato, CA) and were plotted onto enlarged im-
NY), the optic disk was viewed with a binocular indirect ages of the transcranial, neuronal labeling pattern. Selected
ophthalmoscope and a 78 diopter condensing lens. The recording sites were marked by coating the electrodes with
optic disk was back-projected onto the dome with a laser 1% 1,1⬘-dioctadecyl-3,3,3⬘,3-tetramethylindocarocyanine
pointer attached to the condensing lens and was located at perchlorate in ethanol (DiI; Invitrogen), 1% 3, 3⬘-
58° ⫾ 2° azimuth (relative to the 0° azimuth line, which dioctadecyloxacarbocyanine perchlorate in ethanol (DiO;
represents the vertical meridian) and at 31° ⫾ 2° above Invitrogen), or BDA (5% in 0.01 M PB), which in histolog-
the horizontal plane through the center of the eye. This ical sections was visualized with streptavidin-Alexa Fluor
plane, however, was not the 0° elevation line. Instead, the 647 (see below).
horizontal meridian was defined as the intersection of the
horizontal plane with the dome through the optic streak, Histology and tracer labeling
whose center is 15° above the optic disk (Dräger and For histological analysis of the visual cortex, mice were
Olson, 1980). anesthetized with ketamine and xylazine and perfused
Visual stimulation was done manually, by using an transcardially with PBS, followed by 4% paraformalde-
overhead projector for displaying edges and slits of vari- hyde in 0.1 M phosphate buffer, pH 7.4 (PB). The left
able widths and lengths (0.5– 40°) at different orientations cortex was separated from the rest of the brain, flattened
and speeds. The stimulus intensity was 55 cd/m2, and the between glass slides, and postfixed in the same fixative.
background illumination was 0.7 cd/m2. Receptive fields The tissue was cryoprotected in 30% sucrose. Cortex was
(position, size) were mapped onto the dome with colored sectioned on a cryostat at 50 ␮m in the tangential plane.
maker pens (see below). To reveal the myeloarchitectonic borders of V1, sections
Microelectrode recording. To permit visually guided were wet mounted onto glass slides, coverslipped with PB,
recordings in extrastriate cortex, we used a new approach viewed with dark field optics in a stereomicroscope, and
for in vivo transcranial imaging of FR- and/or FE-labeled digitally imaged with a CCD camera (Optronics Mag-
callosal and/or ipsilateral striate-extrastriate cortical con- naFire, Goleta, CA). The same sections were then imaged
nections (Wang et al., 2007). For tracer injections, mice in a compound fluorescence microscope (Nikon Eclipse
were anesthetized with a mixture of ketamine (86 mg/kg, E800) equipped with filter sets for bisbenzimide (330 –
IP) and xylazine (13 mg/kg, IP) and mounted in a head 380-nm excitation filter/420-nm barrier filter), FE (470 –
holder. To label callosal connections, multiple pressure 490-nm excitation filter/520 –560-nm barrier filter), and
injections of FR (5% in H2O) or FE (5% in H2O) were made FR (510 –560-nm excitation filter/610-nm barrier filter)
as described for bisbenzimide. For labeling of ipsilateral fluorescence. To reveal BDA, sections were removed from
striate-extrastriate connections, single injections of FR or glass slides, treated with 0.3% Triton X-100 in PB, and
FE were made into V1. Three days later, mice were used stained with streptavidin-Alexa Fluor 647 (1:200; Invitro-
for receptive field mapping. Mice were injected with atro- gen), which was visualized with an IR-filter set (590 –
pine (2.5 mg/kg, SQ) and anesthetized with urethane (Sig- 650-nm excitation filter/663–735-nm barrier filter). Sec-
ma; 20% in PBS; 1.2 g/kg, IP) and mounted in a head tions were mounted on glass slides, air dried, dehydrated
holder with the incisor bar 2.5 mm below the interaural in ethanol, cleared in xylenes, coverslipped in DPX (BDH
line, so that the roof of the mouth was horizontal. The Laboratory Supplies, Poole, United Kingdom), and studied
body temperature was maintained at 37°C with a in the microscope. Multiple digital images were taken at
feedback-controlled heating pad. Hydration was main- ⫻10 and ⫻20 magnification, and montages of BDA-, FR-,
tained by subcutaneous injection of 5% dextrose. The scalp and FE-labeled connection patterns were produced in Pho-
was opened along the midline, and a fluorescence ste- toshop 8.0 (Adobe Systems, San Jose, CA).
reomicroscope (Leica MZ16F) equipped with filter sets for
FR (560-nm excitation filter/610-nm barrier filter) and FE Assembly of connectional maps
(480-nm excitation filter/510-nm barrier filter) and a
cooled CCD camera (CoolSnap EZ; Roper Scientific, Tuc- Digital images of montages of bisbenzimide-labeled cal-
son, AZ) was used to capture images of FR- or FE-labeled losal connections, FR-, FE-, and BDA-labeled V1 connec-
connections through the intact skull (Wang et al., 2007). tions, and myeloarchitectonic borders were aligned by re-
The callosal labeling pattern was used as landmark to sizing and/or rotation, using blood vessels as reference
guide microelectrodes visually into specific extrastriate marks and Adobe Photoshop 8.0. Multicolor composite
visual areas. Ipsilateral V1 connections were used as a maps were generated to show the topographic organiza-
guide to map receptive fields at specific visuotopic loca- tion of connetions. Images were digitally enhanced by
tions within selected areas. Penetrations were made in making linear adjustments of brightness and contrast.
rows, whose direction was optimized to the particular part
of visual cortex to be studied. The distance between re- Assembly of receptive field maps
cording sites was 50 –150 ␮m. Recordings were performed Composite maps of recording sites were assembled by
with sharp-tipped, lacquered tungsten electrodes (1.2–1.4 digitally aligning (resizing and/or rotation) images of FR-
M⍀ at 1 kHz). The exposed brain was protected with labeled callosal patterns seen through the skull and in
warm saline. Raw neural signals from single units and flat-mounted cortex with the myeloarchitectonic V1 bor-
small clusters of neurons were amplified and bandpass ders, by using DiI-, DiO-, or BDA-labeled recording sites
filtered at 500 and 5,000 Hz with the Axoprobe-2A ampli- as fiduciary marks. The full set of recording sites was then
fier (Molecular Devices, Sunnyvale, CA). Signals were dis- transferred onto the composite map by using the X/Y
played on an oscilloscope, and minimal response fields readout of the micromanipulator. With optimal align-
 The Journal of Comparative Neurology. DOI 10.1002/cne

342 Q. WANG AND A. BURKHALTER

Fig. 1. Visuotopic map of mouse primary visual cortex (V1). connections (blue). Numerals represent recording/injection sites
A: Flattened left occipital cortex showing callosal connections labeled whose receptive fields are shown in D and whose axonal connections
by axonal tracing with fluororuby (red). Numerals indicate recording are illustrated in Figures 2, 3, 4B,D, 5, and 6. The colors indicate
sites in V1, whose receptive fields are shown in B. B: Plots of receptive different tracers: red (fluororuby), green (fluoroemerald), and yellow
fields recorded in left V1 after stimulation of right visual hemifield. (biotinylated dextran amine). D: Receptive fields recorded at the sites
Numerals correspond to recording sites shown in A. C: Flattened left shown in C. A, anterior; P, posterior; M, medial; L, lateral. See list for
occipital cortex showing retrogradely bisbenzimide-labeled callosal abbreviations. Scale bars ⫽ 1 mm.

ment, the remaining discrepancies between in vivo and 2003) was represented at the lateral V1 border, whereas
flat-mount images were ⬍100 ␮m. the temporal visual field (i.e., monocular segment ⬎30°
Receptive field diameter (D in degrees) was determined azimuth) was located in medial V1 (Fig. 1B). The vertical
by computing receptive field area (A) from rectangular meridian (i.e., 0° azimuth line) representation was con-
receptive field plots and substituting A in the equation fined to the segment of the V1 border, which adjoined the
D ⫽ 2公 (A/␲). Statistical significance was assessed by large extrastriate acallosal zone and represented the up-
one-way ANOVA, followed by post hoc Tukey test. per 70° of the visual field. More caudal parts of the lateral
V1 border, which represented the upper visual field pe-
riphery, lacked a nasal visual field representation (Fig.
RESULTS 1B). Similarly, the nasal visual field representation was
V1 contains topographic map of the absent rostral to the large acallosal zone (Fig. 1B). The
visual field representation of the lower field perimeter continued
We used transcranial imaging of FR-labeled callosal around the tip onto the medial side of V1 and terminated
connections to visually guide microelectrode recordings of at the horizontal meridian (Fig. 1B). The medial and pos-
receptive fields in V1 (Fig. 1A). The receptive field map terior V1 border behind the horizontal meridian repre-
showed that caudal V1 represented the upper visual field, sented the upper temporal visual field perimeter.
whereas the lower visual field was located in rostral V1
(Fig. 1B). The upper and lower visual fields were joined at
Extrastriate visual cortex contains nine
the horizontal meridian, which extended lateromedially complete maps of the visual field
from the rostral edge of the large extrastriate acallosal Topographic maps of V1 connections were obtained from
zone to the medial border of V1, midway from the poste- 20 dual-tracer injections and 15 triple-tracer injections.
rior pole of V1. The nasal visual field (i.e., binocular seg- The findings are illustrated by representative cases in
ment ⱕ30° azimuth; Dräger, 1975; Kalatsky and Stryker, which the extrastriate connections from three distinct to-
The Journal of Comparative Neurology. DOI 10.1002/cne

MOUSE VISUAL CORTEX AREA MAP 343

pographic locations of V1 were labeled with different trac- nine distinct areas, we have mapped inputs from the re-
ers. To interpret the connectivity maps in terms of visual maining portions of the visual field. The findings are il-
field coordinates, recordings from small clusters of neu- lustrated in four representative cases, showing azimuthal
rons were made to map receptive fields at injection sites maps in the upper (Fig. 2) and lower (Fig. 5) visual fields
(Fig. 1C,D) and across selected projection fields (see Figs. and elevation maps in the temporal (Fig. 3) and nasal (Fig.
6 – 8). In all cases, the callosal connetions were labeled and 4) visual fields.
were used as fixed landmarks against which the variable Figure 2B,C shows that, in LM and AL, nasal fields
topographic (i.e., location-dependent) connections were (red) are represented near the V1 border and temporal
compared. The connetions were identified as red-, green-, fields (yellow) are located at the lateral edge of the acal-
and yellow-labeled clusters of axon branches and boutons. losal zone, indicating that, in LM and AL, azimuthal maps
In many cases, we observed multiple separate groups of of the upper visual field are rough mirror images of the
red, green, and yellow clusters in distinct extrastriate map in V1. The azimuthal map in LM was perpendicular
cortical fields at specific locations within the callosal pat- to the V1 border, whereas the map in AL had a more
tern. The relative position of the colored clusters within a oblique orientation. Compared with V1, the extrastriate
group reflected the visuotopic map within a portion of the cortical maps were condensed into smaller spaces. In ad-
cortical field. The complete map within each cortical field dition, neurons at well-separated V1 injection sites made
was reconstructed from different cases in which we sys- projections that were joined to seamless maps, suggesting
tematically varied the location of the injection sites. that feedforward connections to LM and AL diverge and
A striking example of multiple groups of red, green, and the recipient neurons represent a larger segment of the
yellow projections was found after injections at different visual field than the parent projection neurons in V1.
nasal and temporal locations of the upper visual field Similar to upper visual field injections into caudal V1,
(Figs. 1C,D, 2A–C). The projection clusters were widely lower field injections into rostral V1 labeled multiple ex-
distributed across lateral and medial extrastriate cortex trastriate cortical projections all around V1 and not just at
and were not preferentially targeted to locations that are nearby sites across the V1 border (Figs. 1C, 3B, 4B,D, 5).
closest to the injection sites, as was suggested in previous In addition, we found that large injections (e.g., Fig. 3B,
studies in rat visual cortex (Malach, 1989; Rumberger et red) did not label larger regions or more numerous patches
al., 2001). The projections in each group showed distinct than small injections (e.g., Fig. 3B, yellow). The overall
dimensions, densities, shapes, orientations, and locations orientation of the elevation map in LM was similar to that
relative to callosal and myeloarchitectonic landmarks. in V1 and showed that lower elevations of the visual field
Many of the locations mapped onto cortical fields that were represented in rostral parts of the acallosal zone
were similar to those described by Montero (1993) in rat (compare green and yellow in Figs. 1C,D, 3B,C). In con-
visual cortex. We, therefore, used Montero’s terminology trast, the elevation map in AL was inverted by 180° rela-
to designate cortical fields in mouse visual cortex and tive to the map in LM and represented higher elevations
outlined the areal borders to facilitate the interpretation (green) at the rostral edge of the acallosal zone, whereas
of the topographic labeling patterns illustrated in Figures lower elevations (yellow) were mapped at more caudal
2–5. It is important to note that these borders reflect best locations (Fig. 3B,C). The diametrically opposed axes of
estimates derived from multiple cases in which we in- the elevation maps in LM and AL are nicely illustrated by
jected the perimeter of V1 (see below). the fact that LM and AL projections from progressively
In the case shown in Figure 2B,C, the most prominent lower fields move closer and closer together (compare
groups of projections on the lateral side of V1 fell into LM green and yellow in Fig. 3B,C), until inputs from the
and AL in the large extrastriate acallosal zone. A third bottom of the visual field finally merge to a single cluster
group of projections was found in the posterior acallosal (Fig. 3B,C, red). Similar examples are shown in Figures 4
field, P, behind LM. A fourth group fell into the callosally and 5, in which red inputs from upper nasal fields (Figs.
connected postrhinal cortex (POR). Groups 5 and 6 were 1C,D, 4B,C) and green inputs from the temporal horizon-
found in the rostrolateral (RL) and anterior (A) fields near tal meridian (Figs. 1C,D, 5B,C) labeled two separate clus-
the rostral tip of V1 (Fig. 2B,D). Groups 7 and 8 repre- ters in LM and AL, whereas projections from lower tem-
sented projections to the posteromedial (PM) and antero- poral fields fused to single clusters (Figs. 1C,D, 4C,D,
medial (AM) fields in acallosal cortex, on the medial side of green; 1C,D, 5B,C, yellow). This elevation-dependent clos-
V1 (Fig. 2B,C). In this case, the only map that appeared to ing of the gap between LM and AL projections strongly
consist of only red and green projections was found in the suggests that inputs to the lower field perimeter of LM
lateral intermediate field, LI (Fig. 2B). Observations at and AL converge in a single “shared” cluster of two pro-
higher magnification, however, revealed a small yellow jections to nearby topographic locations on opposite sides
cluster that was joined to a larger yellow cluster in LM, of the LM/AL border. We have marked this border with a
indicating that V1 inputs to LI represent a ninth distinct line that bisects a single cluster with “shared” inputs to
group of projections, which adjoined LM on the lateral LM and AL (Figs. 3B,C, red; 4B,C, yellow, green; 5B,C,
side. A similar merger of LM and LI clusters was found in yellow).
the green projections from a nearly identical location of V1 To demonstrate the LM/AL border directly, we traced
(Figs. 1C,D, 3B,D). Together, these results suggest that lower field inputs from V1 with FR (red) and used in vivo
extrastriate visual cortex contains at least nine distinct transcranial imaging to visually guide receptive field re-
azimuthal maps. These maps are not identical and have cordings in labeled “shared” clusters. Figures 6A and 6B
clearly different dimensions, shapes, orientations, and lo- show a caudorostral recording sequence (sites 1–13),
cations relative to the callosal landmarks. which originated in the upper temporal periphery of LM.
To study whether the nine azimuthal maps shown in Progressively more rostral recordings showed a gradual
Figure 2B,C represent repeating modules within a single descent in receptive field location, until in the center of the
global map or correspond to upper field representations of red projection (Fig. 6A, site 8); responses were evoked
 The Journal of Comparative Neurology. DOI 10.1002/cne

344 Q. WANG AND A. BURKHALTER

Figure 2
The Journal of Comparative Neurology. DOI 10.1002/cne

MOUSE VISUAL CORTEX AREA MAP 345

from the bottom of the visual field (Fig. 6B). A small step representation of the vertical meridian (Figs. 6D, 9). Neu-
forward within the red projection (i.e., from sites 8 and 9) rons on the lateral side of this border failed to respond to
reversed the downward trend, and neurons at more ros- visual stimulation.
tral sites showed receptive fields in the upper visual field Thus, receptive field mapping strongly supports the in-
of AL (Fig. 6B). The map reversal strongly suggests that terpretation that the systematic and orderly topology of
the LM/AL border lies within the red projection, between V1 inputs to LM and AL reflects two complete maps of the
recording sites 8 and 9. The AL/RL border was more visual field and that each cortical field represents a dis-
rostral and coincided with a significant increase in recep- tinct area. Both areas are contained within the acallosal
tive field size (Fig. 6B, sites 12 and 9). zone, of which LM shares most of its border with V1. AL
Because two points are required to define an areal represents a wedge-shaped area in the anterior one-third
boundary, a second recording sequence (Fig. 6A, sites 14 – of the acallosal zone, which adjoins V1 at a single vertex at
23) was made adjacent to the V1 border, which traversed the rostromedial corner of the acallosal zone, representing
two FE-labeled projections from the upper field of V1. the lower nasal field perimeter near the horizontal merid-
Figure 6C shows that the LM and AL projections repre- ian. The remainder of AL’s vertical meridian extended
sent similar locations of the upper nasal visual field (Fig. perpendicularly to the V1 border and was represented
6C, sites 17 and 21). The mediolateral recording sequence within the callosal bridge between the large and the small
that passed between the green clusters showed a smooth acallosal zone, which formed the AL/RL border (Figs. 1D,
nasotemporal progression of receptive fields along the hor- 4D, yellow, red).
izontal meridian, without any indication of a map reversal RL overlapped with the callosal ring in front of the large
toward the vertical meridian at recording sites near the acallosal zone in lateral extrastriate cortex. The field ad-
posterior green cluster (Fig. 6A,D, sites 24 –28). This dem- joined AL in the back of the callosal ring and contained a
onstrates directly that the horizontal meridian in LM runs mirror image map of AL, in which increasing eccentricity
perpendicularly to an imaginary line that connects the was mapped along the rostromedial axis, lower fields were
green clusters next to the V1 border. The results further represented at the V1 border, and upper fields were rep-
suggest that pairs of projections labeled by a single V1 resented at the border with S1 (Figs. 1D, 2B,C, 3B,C). The
injection do not represent inputs to a spilt horizontal perimeter of the upper nasal field was represented along
meridian, which was proposed to represent the border the AL/RL border, which converged with the V1 and the
between areas V2 and V3 of Wagor et al. (1980). The LM/AL borders, slightly below the horizontal meridian
cortex between the green projections contained lower representation (Figs. 1D, 4B, yellow; 4D, red). The lower
fields and showed an elevation map reversal at the lower visual field perimeter continued along the V1/RL border
field perimeter, which was associated with a slight in- toward the tip of V1 and represented increasingly more
crease in receptive field size (Fig. 6A,C, sites 19 and 20). A temporal fields (Figs. 1A–D, 4B,C, yellow, green). Projec-
similar reversal and increase in receptive field size was tions from the tip of V1 labeled a single large cluster,
found at slightly more lateral recording sites (Fig. 6A,B, which included inputs to both sides of the RL/A border
sites 8 and 9). The most parsimonious explanation of these (Fig. 4B,C, green). Two distinct clusters, representing in-
elevation map reversals is that the two points correspond puts from the lower temporal field perimeter to RL and A,
to the LM/AL border. The mediolateral recording se- were labeled by injections at the rostromedial border of V1
quence across LM (Fig. 6A,D, sites 24 –31) further demon- (Figs. 1D, 5B–D, yellow).
strates that the LM/AL border contains a map of the lower Although V1 inputs to A were relatively sparse, the
visual field perimeter, which reverses from temporal to topographic organization was clearcut. The map was a
nasal within the FR-labeled projection (Fig. 6A,D, sites 28 highly compressed mirror image of RL. An example is
and 29) at the lateral edge of the acallosal zone. The shown in Figure 2B,D, which demonstrates that upper
reversal was accompanied by a sharp increase in receptive temporal fields are represented at the RL/A border, and
field size, suggesting that the recordings have entered LI nasal fields are located at the medial edge of A. Although
(see Fig. 9). The lateral border of LI was marked by the red, green, and yellow fiber densities peaked at specific
locations, Figure 2D shows that the inputs overlapped
substantially, suggesting that the azimuth map in A was
coarse. Elevation was mapped along the rostrocaudal axis,
with lower fields represented at the V1/A border and up-
Fig. 2. Representation of azimuth in extrastriate visual cortex per field inputs at the border with S1 (Figs. 3B,C, 4B,C).
shown in horizontal sections of left occipital cortex. The maps were
generated by making three simultaneous injections of fluororuby (FR;
LI is a narrow, highly elongated field at the lateral edge
red), fluoroemerald (FE; green), and biotinylated dextran amine of LM, in the transition zone between acallosal and cal-
(BDA; yellow) into V1 and triple anterograde tracing of intracortical losally connected cortex. The map in LI was found to be a
connections. A: Darkfield image showing heavy myelination in V1, S1, mirror image of LM in which the upper temporal field
and RSA. Arrowheads indicate myeloarchitectonic borders. Arrows perimeter was represented by a single joined cluster of LM
indicate injection sites in V1. B: Fluorescently labeled axonal connec- and LI inputs to both sides of the LM/LI border (Figs.
tions after injections of FR, FE, and BDA at different nasotemporal
locations (azimuth) of the upper visual field representation in V1. 2B,C, yellow; 3B,D, green). Progressively more nasal in-
Numerals correspond to injection/recording sites and receptive fields jections labeled green and red inputs closer to the lateral
shown in Figure 1C,D. Dashed lines indicate areal borders, which edge of LI (Fig. 2B,C). Elevation was mapped along the
were determined by mapping inputs from the perimeter of V1 (for rostrocaudal axis, with green upper field inputs terminat-
details see text). Solid lines indicate myeloarchitectonic borders. ing at posterior and red lower field inputs projecting to
C: Overlay of projections shown in B and bisbenzimide-labeled cal-
anterior locations (Fig. 3B,C). At the lower field perimeter,
losal connections (blue). D: Higher magnification image of axonal
labeling in area A (boxed area in B). A, anterior; P, posterior; M, LI adjoined LM and AL. This was indicated by the gradual
medial; L, lateral. See list for abbreviations. Scale bars ⫽ 1 mm in A; merger of red LI and red LM/AL clusters (Fig. 3B,C) and
1 mm in C (applies to B,C); 0.1 mm in inset; 0.3 mm in D. the ultimate fusion of green (Fig. 4B,C) and yellow (Fig.
 The Journal of Comparative Neurology. DOI 10.1002/cne

346 Q. WANG AND A. BURKHALTER

Figure 3
The Journal of Comparative Neurology. DOI 10.1002/cne

MOUSE VISUAL CORTEX AREA MAP 347

5B,C) projections to a single cluster with topographically Figure 2B,C shows that PM contained a map in which
matching inputs to LI, LM, and AL at the joined LI/LM/AL upper temporal fields (yellow projections) were repre-
border. Recordings within a similar projection cluster sented at the V1/PM border and upper nasal fields were
showed that the map reverses at the lower temporal field marked by red projections near the medial edge of the
periphery (Fig. 6A,D), which directly demonstrates the acallosal zone. The lower field perimeter was represented
presence of the LM/LI border at this location. at the tip of V1, where projections to PM merged with
P was a field in lateral extrastriate cortex associated projections to AM (Figs. 3B,C, red; 4B,C, green; 5B,C,
with a small acallosal zone behind LM, adjacent to V1 and yellow). Progressively more nasal points of the lower field
medial to postrhinal cortex. Triple labeling of projections perimeter were mapped to increasingly more medial and
from the upper field of V1, showed that P contained an caudal locations and demarcated the PM/AM border (Figs.
azimuthal map, in which temporal fields were represented 3B,C, red; 4B,C, yellow; 5B,C, red).
at the V1 border and nasal fields were located more lat- AM adjoined PM on the rostromedial side and occupied
erally and anteriorly (Figs. 2B,C, 4D). Thus, unlike the the anterior portion of the medial acallosal zone. The map
V1/LM border, which represents nasal fields, the V1/P in AM was a mirror image of PM in which upper nasal
border represents the perimeter of the upper temporal fields were represented at the medial edge of the AM/PM
visual field. Unfortunately, we were unable to produce border and the upper temporal perimeter was located at
high-resolution maps of the most extreme upper nasal the rostromedial margin of the acallosal zone (Fig. 2B,C).
field, because the posterior pole of V1 was buried below The lower temporal field perimeter was represented at the
the transverse sinus and was inaccessible to injections. AM/PM border near the juncture of V1, A, and PM (Figs.
Extrapolation of isoelevation and isoeccentricity lines, 3B,C, red; 4B,C, green, yellow; 5B,C, red). From this point,
however, predicts that the nasal perimeter is represented the PM/AM border turned caudomedially and increasingly
at the posterior margin of the V1/LM border and continues represented the upper nasal field perimeter.
laterally and caudally along the callosal band between LM
and P toward lower field representations near the rhinal
Areas with coarse or undetectable
sulcus (Fig. 4B,C, red, yellow; D, green). The perimeter of topographic maps
the lower nasal visual field marked the P/POR border at V1 inputs to area 36p of perirhinal cortex were found
the lateral edge of the acallosal zone (Fig. 4D, red). from all points of V1 (Figs. 2B,C, 4D). In area 36p, the red,
POR occupied callosally connected cortex behind LM green, and yellow projections overlapped completely and
and LI and lateral to P (Figs. 2B,C, 4D). The organization showed no discernible topographic map.
of red, green, and yellow inputs from V1 showed that nasal V1 projections to the lateral entorhinal area (LEA) were
fields were located at the medial POR/P border and tem- weak and nontopographic (not shown). Several cases
poral fields were represented at the lateral POR/36p showed V1 projections to MM, in callosally connected cor-
boundary (Figs. 2B,C, 4D, 5B,C). Thus, POR contained a tex located between PM, AM, and RSA (Figs. 2B,C, 3B,
mirror image map of P, in which the lower visual field 4B). In all cases, triple labeling was weak, which made it
perimeter was represented at the rhinal sulcus (Fig. 4D, difficult to determine with certainty whether MM was
green), and the upper field perimeter continued along the topographically organized.
posterior edge of the large acallosal zone (Fig. 4B,C, red; V1 projections to RSA consisted of thin, widely dis-
D). Extrapolation of isoeccentricity and isoelevation lines persed fibers, which were present in all cases but were
strongly suggests that POR adjoined LM and LI along the clearly visible only at high magnification (Fig. 2B, inset).
extreme upper nasal field perimeter (Fig. 4B–D). We were, Multilabeling experiments showed that inputs to different
however, unable to demonstrate directly the juncture of parts of RSA overlapped, suggesting that RSA is not to-
POR, LM, P, and LI because the extreme upper field pographically organized.
perimeter of V1 was inaccessible to injection. V1 connections to primary motor cortex terminated in
On the medial side, V1 was adjoined by PM, which layers 1–5 but were densest in layer 2 and were focused on
occupied the caudal half of a triangular acallosal zone. the caudal part of the medial granular Cg1 field (Brecht et
al., 2004; not shown). Triple-tracing experiments revealed
no discernible visuotopic map.
V1 projections to S1 were found in the forehead, earlobe,
and caudal whisker representations. In barrel cortex, V1
Fig. 3. Representation of elevation in temporal visual field shown axons preferentially terminated in septa (not shown) and
in horizontal sections of left visual cortex. The maps were generated
by making three simultaneous injections of fluororuby (FR; red),
showed a coarse map in which A-row barrels were prefer-
fluoroemerald (FE; green), and biotinylated dextran amine (BDA; entially innervated by the upper visual field and C-row
yellow) into V1 and triple anterograde tracing of intracortical connec- barrels favored inputs from the lower visual field. Al-
tions. A: Darkfield image showing heavy myelination in V1, S1, and though retrogradely BDA- and FR-labeled neurons were
RSA. Arrowheads indicate myeloarchitectonic borders. Arrows indi- rare throughout the cortex, they were consistently absent
cate FR, FE, and BDA injection sites in V1. B: Fluorescently labeled in S1, suggesting that S1 does not reciprocate input from
axonal connections after injections of FR, FE, and BDA into V1 at
different elevations of the visual field. Numerals correspond to V1.
injection/recording sites and receptive fields shown in Figure 1C,D. Connectional maps represent functional
Dashed lines indicate areal borders, which were determined by map-
ping inputs from the perimeter of V1 (for details, see text). Solid lines visuotopic maps
indicate myeloarchitectonic borders. C: Overlay of projections shown To study whether the connectional maps represent func-
in B and bisbenzimide-labeled callosal connections (blue). D: Higher
magnification image of axonal labeling in area LM and LI (boxed area
tional visuotopic maps, we recorded receptive fields in V1
in B). A, anterior; P, posterior; M, medial; L, lateral. See list for and the surrounding extrastriate cortex. Transcranial im-
abbreviations. Scale bars ⫽ 1 mm in A; 1 mm in C (applies to B,C); 0.3 aging of FR-labeled callosal connections was used to visu-
mm in D. ally guide microelectrode recordings (Wang et al., 2007).
 The Journal of Comparative Neurology. DOI 10.1002/cne

348 Q. WANG AND A. BURKHALTER

Figure 4
The Journal of Comparative Neurology. DOI 10.1002/cne

MOUSE VISUAL CORTEX AREA MAP 349

This approach was more flexible than using the tradi- the upper temporal periphery of the visual field (Fig. 8B).
tional equidistant recording grid and allowed high- This is in sharp contrast to receptive fields at slightly
resolution mapping and superior precision for referencing more rostral locations at the V1/LM border, which repre-
recording sites to myeloarchitectonic and callosal land- sented the upper nasal visual field (Figs. 7B, 8B) and
marks. suggests that the nasal visual field is not represented
The recordings shown in Figure 7A were concentrated above 70° in the upper visual field. Across the V1 border in
mainly on LM, AL, LI, and RL in lateral extrastriate P, receptive fields were significantly (P ⬍ 0.001) larger
cortex. The first sequence included sites 1–17 and tra- than in V1 and represented the upper temporal perimeter
versed the acallosal zone in rostrolateral direction (Fig. of the visual field (Figs. 8B, 9). As predicted from the
7A). At the V1/LM border, the receptive field map reversed connection map, the posterior lateral border of P repre-
from nasal to temporal, which was correlated with a sig- sented the lower nasal field (Fig. 8B). The recording se-
nificant (P ⬍ 0.001) increase in receptive field size in LM quence shown in Figure 8C,D confirmed that the acallosal
(Figs. 7B, 9). Progressively more lateral and rostral re- zone in front of P contained two mirror symmetric maps,
cordings in LM shifted receptive fields toward the lower LM and AL, whose shared border traversed the rostral
temporal perimeter of the visual field, which coincided part of the acallosal zone and represented the lower visual
with the LM/LI border at the lateral edge of the acallosal field perimeter (Figs. 6A–C, 7A,C,D). Figure 8D further
zone (Fig. 7B). Recordings at more lateral sites showed a shows that the V1/AL border is extremely short and is
map reversal back to the nasal visual field periphery, effectively reduced to a single point at the rostromedial
which was represented at the far lateral margin of LI. corner of the acallosal zone, which coincides with the
Similar azimuthal maps were found in the lower visual intersection of the horizontal and vertical meridian. As
field representations of LM and LI (Fig. 6A,D). The second shown in Figure 6A–D, most of AL’s vertical meridian was
recording sequence included sites 18 –28 and ran in a represented along the AL/RL boundary, which was per-
caudorostral direction, parallel to the V1 border (Fig. 7A). pendicular to the V1 border. In RL, receptive fields were
Stepping forward from POR (site 18) into the acallosal significantly (P ⬍ 0.001) larger than in AL (Fig. 9). Lower
zone, receptive fields shifted upward (site 19) and then fields were represented at the V1/RL border, upper fields
downward, indicating a map reversal at the POR/LM bor- at the RL/S1 border, and temporal fields in front of the
der (Fig. 7C). Advancing rostrally across LM, receptive small callosal ring, at the border with A (Fig. 8D). Cross-
fields continued to fall and ultimately reached the lower ing the RL/A border into A was associated with a signifi-
temporal perimeter of the visual field (site 21). At more cant (P ⬍ 0.001) increase in receptive field size (Fig. 9).
rostral sites, the map reversed toward the upper field, Area A contained a mirror-image map of RL, with nasal
indicating the crossing of the LM/AL border into AL (Fig. fields represented near the tip of V1 (Fig. 8D).
7C). As recordings approached the rostral edge of the Two recording sequences were made in medial extra-
acallosal zone, receptive fields continued to shift upward striate cortex (Fig. 8A). The anterior sequence started at
and finally ended up in the upper nasal visual field, close the V1/PM border at the representation of the lower tem-
to the vertical meridian. At the AL/RL border in the pos- poral field perimeter (Fig. 8E). Crossing into PM was
terior part of the callosal ring, the receptive fields signif- easily recognized by a significant (P ⬍ 0.001) increase in
icantly (P ⬍ 0.001) increased in size, and the map reversed receptive field size (Fig. 9). Moving away from the V1/PM
away from the vertical meridian toward lower temporal border toward the PM/A border was associated with a
representations at the V1/RL border. Similar map rever- shift to the lower nasal perimeter of the visual field (Fig.
sals at virtually identical locations are shown in Figure 8E). More medial recordings across the PM/AM border in
6A,B. AM showed a slight increase in receptive field size (Fig. 9)
The series of recordings shown in Figure 8A represents and a map reversal toward the lower temporal field (Fig.
a nearly complete circumnavigation of V1. Receptive fields 8E). Recordings in the posterior sequence revealed recep-
recorded at the V1/P border in posterior V1 represented tive fields at higher elevations (Fig. 8F). Advancing medi-
ally away from the V1/PM border shifted receptive fields
nasally and into the upper visual quadrant (Fig. 8F). At
the PM/AM border, the map reversed and, on the way to
the AM/MM border, represented progressively more tem-
Fig. 4. Representation of elevation in nasal visual field shown in poral fields (Fig. 8F). Exiting AM and entering MM was
horizontal sections of left visual cortex. The maps were generated by indicated by a dramatic increase in the size of receptive
making three simultaneous injections of fluororuby (FR; red), fluoro-
emerald (FE; green), and biotinylated dextran amine (BDA; yellow)
fields (Figs. 8F, 9).
into V1 and triple anterograde tracing of intracortical connections. Extrastriate areas have different receptive
A: Darkfield image showing heavy myelination in V1 and S1. Arrow-
heads indicate myeloarchitectonic borders. Arrows indicate FR, FE, field sizes
and BDA injection sites in V1. B: Fluorescently labeled axonal con- In all areas, receptive fields tended to be slightly larger
nections after injections of FR, FE, and BDA in lateral V1 at different at more eccentric locations. Over the range of receptive
elevations of the visual field. Numerals correspond to injection/
recording sites and receptive fields shown in Figure 1C,D. Dashed field locations that we have studied, however, the differ-
lines indicate areal borders, which were determined by mapping in- ence was not significant. We have, therefore, used mean
puts from the perimeter of V1 (for details see text). Solid lines indicate receptive fields size of pooled recordings for between-area
myeloarchitectonic borders. C: Overlay of projections shown in B and comparisons. The results show that, in every extrastriate
bisbenzimide-labeled callosal connections (blue). D: FR, FE, and BDA area, receptive fields were larger than in V1 (Fig. 9).
injection sites (indicated by numerals) and labeled V1 projections in
lateral and posterior extrastriate cortex. Injection/recording sites and
Within extrastriate cortex, receptive fields were smallest
receptive fields shown in Figure 1C,D. A, anterior; P, posterior; M, in LM and were significantly (P ⬍ 0.001) different from all
medial; L, lateral. See list for abbreviations. Scale bars ⫽ 1 mm in other areas, except AL and POR. The receptive fields in
A,D; 1 mm in C (applies to B,C). AL were slightly larger than those in LM and similar to
 The Journal of Comparative Neurology. DOI 10.1002/cne

350 Q. WANG AND A. BURKHALTER

Figure 5
The Journal of Comparative Neurology. DOI 10.1002/cne

MOUSE VISUAL CORTEX AREA MAP 351

those in P, LI, and PM (Fig. 9). The largest receptive fields opossum, hedgehog, rabbit, gray squirrel, tree shrew, and
were found in RL, A, and MM, which were significantly cat (Towns et al., 1977; Gould and Ebner, 1978; Symonds
(P ⬍ 0.001) larger than those in most other areas, except and Rosenquist, 1984; Kaas et al., 1989; Lyon et al., 1998;
in PM and AM, which were not significantly different from Kahn et al., 2000); and more than the 11 V1 targets
RL. observed in primates (Felleman and Van Essen, 1991).
Although the total numbers of V1 connections in mice,
cats, and primates are similar, it is important to note that
DISCUSSION mouse V1 provides direct inputs to somatosensory, motor,
The results of the experiments described here demon- and limbic structures, which in cats and primates either
strate that mouse extrastriate visual cortex contains at do not exist or originate from higher visual areas (Felle-
least nine complete, orderly maps of the visual field (Fig. man and Van Essen, 1991). The direct access of V1 to
10B). Each of these maps has a unique topographic orga- higher order areas suggests that mouse visual cortex has
nization, size, shape, and location relative to fixed land- few hierarchical levels and may be similar to rat extra-
marks such as myeloarchitectonic borders and callosal striate cortex (Coogan and Burkhalter, 1993). This inter-
connections (Fig. 10A). In some instances, the callosal pretation is consistent with the large receptive field sizes
connections represent the vertical meridian (Choudhury we have recorded in different areas and suggests that, in
et al., 1965) and demarcate areal borders (e.g., V1/LM, mouse cortex, the visual scene is represented over a 50 –
AL/RL). However, the callosal connections do not gener- 100 times narrower spatial range than in monkey (Rosa,
ally represent the vertical meridian (Lewis and Olavarria, 1997; this study).
1995), and the relationship between callosal connections The V1 connections that we have found are similar to
and areal boundaries is often quite complex, as shown those described in previous studies of mouse visual cortex,
previously in macaque monkey (Van Essen et al., 1982). but they are not identical (Olavarria et al., 1982; Olavar-
For example, in RL, the callosal ring demarcates the ria and Montero, 1989). The first difference is caused
boundaries with V1 and AL but does not line up with the merely by changing the names of the posterolateral field
RL/A and RL/S1 borders. This organization suggests that (PL; Olavarria and Montero, 1989) to POR and the latero-
interhemispheric connections not only link similar loca- lateral field (LL; Olavarria and Montero, 1989) to area
tions of the visual field but also mediate influences from 36p, which are more widely used terms than PL and LL
mirror-image locations in both visual hemifields. (Burwell, 2001). Second, we found no evidence for a lat-
Many of the connectionally defined maps correspond to erolateral anterior field (LLA), adjacent to LI, which was
functionally defined maps with diverse receptive field described in rat (Olavarria and Torrealba, 1978; Espinoza
sizes. Although the results show that V1 is adjoined by six and Thomas, 1983; Thomas and Espinoza, 1987; Montero,
distinct areas, we found that only LM shares the vertical 1993). Third, we found V1 inputs to RL and A, which have
meridian with V1 and therefore resembles V2 (Van Essen, previously been described only in rat and hamster (Olav-
1979). These findings argue against the view that mouse arria and Montero, 1990; Coogan and Burkhalter, 1993).
lateral extrastriate visual cortex is composed of repeating Fourth, we found inputs to a new region, MM, which
modules contained within a single large area V2 (Kaas et overlapped with V2MM of Paxinos and Franklin (2001).
al., 1989; Rosa and Krubitzer, 1999). Instead, the results Fifth, we found that V1 preferentially innervates the
support the notion that mouse V1 is adjoined laterally by septa of S1, which supports the notion that septa and
a string of distinct extrastriate visual areas. barrels belong to separate processing networks (Kim and
Ebner, 1999; Brecht et al., 2003; Shepherd and Svoboda,
V1 is connected to 15 visual, somatosensory, 2005). Sixth, we found input to Cg1 of primary motor
motor, and limbic areas cortex, which in rat receives V1 input and plays a role in
We found that V1 projects to P, POR, 36p, LEA, LM, LI, pupillary constriction and eye movements (Miller and
AL, RL, A, S1, PM, AM, MM, RSA, and Cg1. This is a Vogt, 1984; Coogan and Burkhalter, 1993; Guandalini,
larger number of targets than the nine sites described for 1998; Brecht et al., 2004). Finally, we found nontopo-
rat (Sefton et al., 2004); more than the 2 to 8 sites found in graphic input to LEA, which is consistent with the lack of
spatially selective neurons in LEA (Hargreaves et al.,
2005).
Vertical meridian representation in V1 is
Fig. 5. Representation of azimuth in lower visual field shown in confined to V1/LM border
horizontal sections of left visual cortex. The maps were generated by
making three simultaneous injections of fluororuby (FR; red), fluoro- The results from receptive field mapping in striate cor-
emerald (FE; green), and biotinylated dextran amine (BDA; yellow) tex show that the azimuthal and elevation maps resemble
into V1 and triple anterograde tracing of intracortical connections. published maps of rat and mouse V1 (Fig. 10B; Montero et
A: Darkfield image showing heavy myelination in V1 and S1. Arrow- al., 1973; Dräger, 1975; Wagor et al., 1980; Schuett et al.,
heads indicate myeloarchitectonic borders. Arrows indicate FR, FE, 2002; Kalatsky and Stryker, 2003; Gias et al., 2006). Re-
and BDA injection sites in V1. B: Fluorescently labeled axonal con-
nections after injections of FR, FE, and BDA into rostral V1 at differ- cordings from the perimeter of V1 further show that the
ent locations of the lower visual field. Numerals correspond to vertical meridian begins approximately 15° below the hor-
injection/recording sites and receptive fields shown in Figure 1C,D. izontal meridian and extends to approximately 70° in the
Dashed lines indicate areal borders, which were determined by map- upper visual field. This corresponds to predictions from
ping inputs from the perimeter of V1 (for details see text). Solid lines ophthalmoscopic examinations in rat (Hughes, 1979) and
indicate myeloarchitectonic borders. C: Overlay of projections shown
in B and bisbenzimide-labeled callosal connections (blue). D: Higher
indicates that in mice the visual field in the upper and
magnification image of projections to area RL (boxed area in B). A, lower nasal segments is blind (Fig. 10B).
anterior; P, posterior; M, medial; L, lateral. See list for abbreviations. The key observation, however, is that the vertical me-
Scale bars ⫽ 1 mm in A; 1 mm in C (applies to B,C); 0.05 mm in D. ridian representation is much shorter than the lateral
 The Journal of Comparative Neurology. DOI 10.1002/cne

352 Q. WANG AND A. BURKHALTER

Fig. 6. Delineation of LM/AL and LM/LI borders in acallosal zone are shown in B (sites 1–13), C (sites 13–23), and D (sites 24 –31).
of lateral extrastriate cortex by recording map reversals at lower Recordings within the red cluster show two map reversals at the lower
visual field perimeter. A: Flattened left occipital cortex, showing V1 visual field perimeter: 1) lower to upper (sites 8 and 9), corresponding
projections in lateral extrastriate cortex from upper (green fluoroem- to LM/AL border, and 2) temporal to nasal (sites 28 and 29), corre-
erald injection site 6A) and lower (red fluororuby injection site 6A) sponding to the LM/LI border. Recordings in the green clusters show
representations of the visual field. The receptive fields recorded at the that each cluster contains upper field representations, indicating that
injection sites are shown in Figure 1C,D. The callosal connections are the two nearby clusters next to the lateral V1 border do not represent
retrogradely labeled with bisbenzimide (blue). The large acallosal a split horizontal meridian. Dashed lines indicate areal borders,
zone contains two separate green clusters of upper field inputs to LM which were determined by mapping inputs from the perimeter of V1
and AL and a single red cluster of lower field inputs, which terminate (for details, see text). Solid lines indicate myeloarchitectonic borders.
at adjacent locations in LM, AL, and LI. The numerals represent A, anterior; P, posterior; M, medial; L, lateral. See list for abbrevia-
recording sites, some of which are marked by yellow deposits of BDA. tions. Scale bar ⫽ 1 mm.
The corresponding receptive fields, including receptive field centers

border of V1. This feature was reported previously by confined to the portion of the V1 border that adjoins the
Wagor et al. (1980) and is shown in the map of Kalatsky large lateral acallosal zone, which, except for a short strip
and Stryker (2003), but, without reference to callosal land- at the juncture of LM, AL, and RL, corresponds to the
marks, these findings were difficult to interpret. Our re- V1/LM border (Fig. 10A,B). Because, in some rodents and
sults clearly demonstrate that the vertical meridian is in all nonrodent mammals, V2 is the only area that ad-
The Journal of Comparative Neurology. DOI 10.1002/cne

MOUSE VISUAL CORTEX AREA MAP 353

lateral extrastriate cortex represents repeating modules

within a single area, V2 (Kaas et al., 1989; Malach, 1989;
Rosa and Krubitzer, 1999). Similar to observations in rat
(Montero, 1993), we found that in mice lateral V1 is ad-
joined by multiple areas. However, because in mice only
LM resembles V2, the string of areas cannot represent
modules within a single area, V2.
The organization of mouse extrastriate cortex differs
from that in primates, in which V1 is encircled completely
by V2 (Rosa and Tweedale, 2005). Mouse extrastriate cor-
tex also differs from that in many nonprimate species (i.e.,
quoll, hedgehog, guinea pig, rabbit, flying fox, opossum,
squirrel, tree shrew, cat) in which V1 is adjoined on the
lateral side by a single area, V2 (Kaas et al., 1970, 1989;
Hall et al., 1971; Towns et al., 1977; Payne, 1993; Rosa,
1999; Rosa et al., 1999; Kahn et al., 2000), and therefore
has a “simple” organization (Rosa and Krubitzer, 1999).
However, mouse extrastriate cortex is similar to that in
species with a “simple” organization, in that V1 is adjoined
on the medial side by parastriate or splenial visual areas,
which in hedgehog, rat, hamster, degu, guinea pig, and cat
receive direct V1 input and contain visually responsive
neurons (Kalia and Whitteridge, 1973; Choudhury, 1978;
Gould and Ebner, 1978; Olavarria and Mendez, 1979;
Bravo et al., 1990; Olavarria and Montero, 1990; Spatz et
al., 1991; Montero, 1993). Thus, mouse extrastriate cortex
is “simple” in the sense that it has a single V2-like area,
LM, on the lateral side of V1, but is “complex” in that
lateral V1 is adjoined by multiple visual areas.
Extrastriate cortex contains multiple
visuotopic areas
The principal result of triple-anterograde tracing of in-
tracortical connections of V1 is that the retinotopic map of
V1 is remapped multiple times in nine distinct extrastri-
ate cortical fields. Previous single and multitracer exper-
iments in mouse, rat, hamster, degu, guinea pig, and
squirrel have labeled similar targets, which were consid-
ered either distinct areas (Olavarria and Montero, 1989,
1990; Bravo et al., 1990; Spatz et al., 1991; Montero, 1993)
or repeating modules within a single area, V2 (Kaas et al.,
1989; Malach, 1989; Rumberger et al., 2001). By mapping
the connections from a wide range of topographic loca-
tions, we found that, unlike stripes in monkey V2, which
create map discontinuities within a single global map (Roe
Fig. 7. Receptive field maps of different areas in lateral extrastri-
ate visual cortex. A: Flat map of left occipital cortex showing callosal and Ts’o, 1995), each of the labeled targets contains a
connections labeled by axonal tracing with fluororuby (red). Numerals complete map of the entire visual field, and together the
represent recording sites of receptive fields shown in B and C. Differ- maps form a seamless mosaic of extrastriate areas (Fig.
ent colors indicate recordings in different areas. A, anterior; P, pos- 10B). The most direct demonstration of this organization
terior; M, medial; L, lateral. Scale bar ⫽ 1 mm. derives from experiments in lateral extrastriate cortex in
which we successfully determined the LM/AL, AL/RL, and
LM/LI borders by mapping receptive fields in and around
joins V1 at the vertical meridian (Kaas et al., 1970, 1972, labeled projection clusters. The connectivity studies show
1989; Hall et al., 1971; Towns et al., 1977; Choudhury, that lower field injections label single patches that are
1978; Payne, 1993; Rosa, 1999; Rosa et al., 1999; Kahn et “shared” by LM and AL (Figs. 3B, 4B). As expected, re-
al., 2000; Rosa and Tweedale, 2005), LM can be considered cordings from lower field connections show that a single
homologous to V2. The homology of LM and V2 was sug- patch contains two mirror-image maps of the lower visual
gested previously, based on location (Montero, 1993) and field, suggesting that the elevation map reversal at the
position in the hierarchy of rat extrastriate visual cortex lower field perimeter represents the LM/AL border (Fig.
(Coogan and Burkhalter, 1993). The idea, however, was 6A–C). More rostral parts of the acallosal zone contain the
controversial, because rat V1 is adjoined by a string of upper field of AL, whose representation reverses from
areas that were all thought to share the vertical meridian nasal to temporal at the AL/RL border (Figs. 6A, 10B).
with V1 and therefore resemble V2 (Espinoza and Recordings from lower temporal field patches further
Thomas, 1983; Thomas and Espinoza, 1987; Montero, show that, in addition to the LM/AL border, a single patch
1993). This led to the proposal that the string of areas in contains a temporal-to-nasal map reversal, which repre-
 The Journal of Comparative Neurology. DOI 10.1002/cne

354 Q. WANG AND A. BURKHALTER

Fig. 8. Receptive field maps of different areas in lateral and me- shown in B–F. Different colors indicate recordings in different areas.
dial extrastriate visual cortex. A: Flat map of left occipital cortex Note that, for clarity of presentation, for some of the recording sites
showing callosal connections labeled by axonal tracing with flu- shown in A, only the centers of receptive fields are indicated in C and
ororuby (red). Numerals represent recording sites of receptive fields D. A, anterior; P, posterior; M, medial; L, lateral. Scale bar ⫽ 1 mm.

sents the LM/LI border. This suggests that, at the lateral visual field. Mediolateral recording sequences across LM
edge of the acallosal zone, LM and LI share a common and AL clusters at the horizontal meridian demonstrate
border, which represents the lower temporal perimeter of clearly that this is not the case (Fig. 6A,D), suggesting
the visual field (Figs. 6A,D, 10B). Although these experi- that lateral V1 is not surrounded by a map with a second-
ments strongly suggest that LM and AL contain complete order transformation of the visual field (Allman and Kaas,
maps of the entire visual field, it was important to rule out 1974). In contrast, the results strongly suggest that lat-
that projection clusters in LM and AL from locations near eral V1 is adjoined by several distinct areas that contain
the horizontal meridian (e.g., Fig. 3B yellow) represent the first-order representations of the visual field (Fig. 10B).
split horizontal meridian of a single area, V2, and corre- This conclusion is consistent with findings that rodent
spond to the V2/V3 border (Wagor et al., 1980). The pre- visual cortex lacks a modular organization (Van Hooser et
diction from such an organization is that the horizontal al., 2006) and with observations that the different maps in
meridian coincides with a map reversal back to the nasal lateral extrastriate cortex have distinct connections, re-
The Journal of Comparative Neurology. DOI 10.1002/cne

MOUSE VISUAL CORTEX AREA MAP 355

Fig. 9. Mean ⫾ SD diameter of receptive fields in different visual

cortical areas. Numerals indicate N.

ceptive field sizes (Fig. 9), and stimulus selectivities

(Sanderson et al., 1991; Coogan and Burkhalter, 1993;
Wang and Burkhalter, 2004, 2005; Gao et al., 2006).
Area map differs from published maps
The most widely used map of mouse visual cortex shows
that V1 is adjoined laterally by two concentric areas, V2
and V3, and medially by two smaller areas, Vm-r and
Vm-c (Fig. 10C; Wagor et al., 1980). In contrast, we found
that V1 abuts six visuotopic areas. Despite the differences
in the partitioning scheme, the overall visuotopic organi-
zation of the present map shows similarities to Wagor’s
receptive field map (Wagor et al., 1980) as well as to the
intrinsic optical signal maps by Schuett et al. (2002) and
Kalatsky and Stryker (2003). The similarities are most Fig. 10. Diagrammatic representation of areas in flat maps of left
obvious in medial extrastriate cortex, in which Wagor’s hemisphere of mouse visual cortex. A: Spatial relationship between
areas Vm-c (AM of Schuett et al., 2002; V5 of Kalatsky and callosal connections (blue) and visuotopically organized visual areas V1,
Stryker, 2003) and Vm-r match the topology, shape, and P, LM, AL, RL, A, AM, PM, LI, and POR. B: Visuotopic maps of striate
size of PM and AM, respectively (Wagor et al., 1980). and extrastriate areas in mouse visual cortex. The maps were con-
structed by tracing the intracortical connections of known visuotopic
However, the partitioning scheme of lateral extrastriate locations of V1. The different quadrants of the visual hemifield are color
cortex is at variance. A likely explanation for these dis- coded (inset). Arrows indicate the orientation of the maps as shown in
crepancies is that the present maps were generated in the inset. Nasal visual field: ⱕ30° eccentricity. Temporal visual field:
individual animals by a combination of closely spaced ⬎30° eccentricity. Each area contains a complete and orderly visuotopic
recordings, tracing of connections, and referencing record- map, which is topologically equivalent to the contralateral visual hemi-
ing and projection sites to callosal and myeloarchitectonic field. Note that the visuotopic maps are registered across areal borders to
minimize map discontinuities. C: Visuotopic map of striate and extra-
landmarks, which eliminated the difficult task faced by striate visual cortex adapted from the receptive field mapping study
Wagor et al. (1980) of registering recording sites across performed by Wagor et al. (1980). Colors indicate the different quadrants
animals. of the visual field. A, anterior; P, posterior; M, medial; L, lateral.
Wagor’s interpretation of the map of lateral extrastriate
cortex was derived from Allman and Kaas’ (1974) descrip-
tion of the topographic organization of monkey visual cor-
tex (Wagor et al., 1980). An adaptation of this map (Fig. resent the V2/V3 border and that the mediolateral
10C) shows that the caudolateral (HMcl) and rostromedial (HMml) branch separates the upper from the lower visual
(HMrm) branches of the horizontal meridian jointly rep- field in V3. A similar map of the horizontal meridian was
 The Journal of Comparative Neurology. DOI 10.1002/cne

356 Q. WANG AND A. BURKHALTER

described by Kalatsky and Stryker (2003) and was ob- that mouse LM is too small, and there was little pressure
served in the present study (Fig. 10B). However, unlike for reducing the length of connections with homotopic
Wagor, we found a topographic reversal between HMcl locations of V1 (Allman and Kaas, 1974). Optimization of
and HMml and an additional polarity change between connection length, however, may explain why the maps
HMml and HMr (Fig. 10B,C). Similar reversals are shown are in register across common borders (Chklovskii and
in the animation of the intrinsic optical signal map pro- Koulakov, 2004). Indeed, it has been suggested that wire-
duced by stimulating the mouse visual system with length optimization plays a role in the development of the
upward- and downward-drifting bars (V.A. Kalatsky; http: areal mosaic in cerebral cortex (Kaas and Catania, 2005;
val.ee.uh.edu/about-maps.html). Such reversals are typi- Rosa and Tweedale, 2005).
cally found at areal boundaries (but see Roe and Ts’o,
1995). Thus, the different branches of Wagor’s horizontal
meridian representation are better fit by a scheme in ACKNOWLEDGMENTS
which HMcl represents the horizontal meridian of LM, We thank David Van Essen for many insightful discus-
HMml marks the horizontal meridian of AL, and HMrm sions and comments on earlier versions of the manuscript,
indicates the horizontal meridian of RL and A (Fig. Greg DeAngelis for his help with single-unit recording,
10B,C). In this scheme, both LM and AL contain a com- and Raj Apte for showing us how to perform ophthalmos-
plete representation of the visual field and therefore con- copy in mice. We also thank Katia Valkova for excellent
stitute distinct areas. AL differs from RL in visuotopic technical assistance.
organization, receptive field size, and callosal connections.
RL differs from A by its mirror-image map, callosal con-
nections, and receptive field size. P differs from LM by the LITERATURE CITED
absence of a vertical meridian representation at the V1/P Allman JM, Kaas JH. 1971. Representation of the visual field in striate and
border and the size of receptive fields. An impressive dem- adjoining cortex of the owl monkey (Aotus trivirgatus). Brain Res
onstration of P is shown in the animated azimuthal map of 35:89 –106.
the intrinsic optical signal (V.A. Kalatsky; http: Allman JM, Kaas JH. 1974. The organization of the second area (VII) in the
val.ee.uh.edu/about-maps.html). LI was first described by owl monkey: a second order transformation of the visual hemifield.
Brain Res 76:247–265.
Wagor et al. (1980) as a reversal of the azimuthal map at
Allman JH, Kaas JH, Lane RK. 1973. The middle temporal visual area
the lateral border of V2. Kalatsky and Stryker (2003) (MT) in the bushbaby, Galago senegalensis. Brain Res 57:197–202.
found a map at a similar location, which they designated Bravo H, Olavarria J, Torrealba F. 1990. Comparative study of visual inter
V3 and which is most clearly seen in the animated azi- and intrahemispheric cortico-cortical connections in five native Chilean
muthal map (V.A. Kalatsky; http:val.ee.uh.edu/about- rodents. Anat Embryol 181:67–73.
maps.html). In our map, LI is identified as a mirror-image Brecht M, Roth A, Sakmann B. 2003. Dynamic receptive fields of recon-
map of LM with significantly larger receptive fields (Figs. structed pyramidal cells in layer 3 and 2 of rat somatosensory barrel
cortex. J Physiol 533:243–265.
9, 10B). POR resembles V4 of Kalatsky and Stryker (2003)
Brecht M, Krauss A, Muhammad S, Sinai-Esfahani L, Bellanca S, Mar-
and differs from P by its mirror-image map and the pres- gerie TW. 2004. Organization of rat vibrissa motor cortex and adjacent
ence of callosal connections. Thus, the partitioning scheme areas according to cytoarchitectonics, microstimulation, and intracel-
of mouse lateral extrastriate cortex resembles the “com- lular stimulation of identified cells. J Comp Neurol 479:360 –373.
plex” organization of rat in which V1 is adjoined in the Brill L. 2006. Mouse brain map is complete. Nature 443:380 –381.
lateral side by a string of distinct areas (Espinoza and Burwell RD. 2001. Borders and cytoarchitecture of the perirhinal and
Thomas, 1983; Thomas and Espinoza, 1987; Montero, postrhinal cortices in the rat. J Comp Neurol 437:17– 41.
1993). Although mouse and rat extrastriate cortex look Caviness VS. 1975. Architectonic map of neocortex of the mouse. J Comp
Neurol 164:247–264.
alike, they differ in that, in mouse, only LM resembles V2,
Changizi MA, Shimojo S. 2005. Parcellation and area–area connectivity as
suggesting that mouse extrastriate cortex is neither “sim- a function of neocortex size. Brain Behav Evol 66:88 –98.
ple” nor “complex” as defined by Rosa and Krubitzer Chklovskii DB, Koulakov AA. 2004. Maps in the brain: what can we learn
(1999). However, except for the small size of LM (i.e., V2) from them? Annu Rev Neurosci 27:369 –392.
and the multiple small areas on the lateral side of V1, the Choudhury BP. 1978. Retinotopic organization of the guinea pig’s visual
organization of mouse extrastriate cortex is similar to that cortex. Brain Res 144:19 –29.
of many nonrodent mammals. Choudhury BP, Whitteridge D, Wilson ME. 1965. The function of the
Unlike the areal organization proposed by Wagor et al. callosal connections of the visual cortex. Q J Exp Physiol 50:214 –219.
(1980; Fig. 10C), each map in mouse visual cortex was Coogan TA, Burkhalter A. 1993. Hierarchical organization of areas in rat
visual cortex. J Neurosci 13:3749 –3772.
found to be topologically equivalent to the contralateral
Dräger UC. 1975. Receptive fields of single cells and topography in mouse
visual field, and, in this respect, the maps resemble V1 visual cortex. J Comp Neurol 160:269 –290.
and MT in primates (Allman and Kaas, 1971, 1974; All- Dräger UC, Olson JF. 1980. Origins of crossed and uncrossed retinal
man et al, 1973; Van Essen et al., 1981). This includes LM, projections in pigmented and albino mice. J Comp Neurol 191:383– 412.
which, like V2, shares the vertical meridian with V1 but Espinoza SG, Thomas HC. 1983. Retinotopic organization of striate and
whose retinotopic map is not a second-order transforma- extrastriate visual cortex in the hooded rat. Brain Res 272:137–144.
tion of the visual field as suggested for, gray squirrel, Felleman DJ, Van Essen DC. 1991. Distributed hierarchical processing in
hamster, and guinea pig V2 (Hall et al., 1971; Allman and the primate cerebral cortex. Cereb Cortex 1:1– 47.
Kaas, 1974; Tiao and Blakemore, 1976; Choudhury, 1978). Gao E, DeAngelis GC, Burkhalter A. 2006. Specialized areas for shape and
motion analysis in mouse visual cortex. Soc Neurosci Abstr 641.6.
Combined high-resolution connectional and receptive field
Gias C, Hewson-Stoate N, Jones M, Johnston D, Mayhew JE, Coffey PJ.
mapping studies are necessary to settle whether V2 in 2006. Retinotopy within rat primary visual cortex using optical imag-
these rodents differs from mouse LM. It seems, however, ing. Neuroimage 24:200 –206.
reasonable to assume that LM is homologous to primate Gould HJ, Ebner FF. 1978. Connections of the visual cortex in the hedge-
V2, which raises the question of why mouse LM lacks a hog (Parachinus hypomelas). J Comp Neurol 177:473–502.
topologically nonequivalent, V2-like map. It is possible Guandalini P. 1998. The corticocortical projection of the physiologically
The Journal of Comparative Neurology. DOI 10.1002/cne

MOUSE VISUAL CORTEX AREA MAP 357

defined eye fields in the rat medial frontal cortex. Brain Res Bull In: Rockland KS, Kaas JH, Peters A, editors. Cerebral cortex: extra-
47:377–385. striate cortex in primates, vol 12. New York: Plenum Press. p 127–203.
Hall WC, Kaas JH, Killackey H, Diamond IT. 1971. Cortical visual areas in Rosa MGP. 1999. Topographic organization of extrastriate areas in the
grey squirrel (Sciurus carolinensis): a correlation between cortical flying fox: implications for the evolution of mammalian visual cortex.
evoked potential maps and architectonic subdivisions. J Neurophysiol J Comp Neurol 411:503–523.
34:437– 452. Rosa MGP, Krubitzer LA. 1999. The evolution of visual cortex: where is
Hargreaves EL, Rai G, Lee I, Knierim JJ. 2005. Major dissociation between V2? Trends Neurosci 22:242–248.
medial and lateral entorhinal inputs to dorsal hippocampus. Science Rosa MGP, Tweedale R. 2005. Brain maps, great and small: lessons from
308:1792–1794. comparative studies of primate visual cortical organization. Philos
Hughes A. 1979. A schematic eye for the rat. Vis Res 19:569 –588. Trans R Soc B Biol Sci 360:665– 691.
Kaas JH, Catania KC. 2005. How do features of sensory representations Rosa MGP, Krubitzer LA, Molnar Z, Nelson JE. 1999. Organization of
develop? Bioessays 24:334 –343. visual cortex in the Northern quoll, Dasyurus hallucatus: evidence for
Kaas JH, Hall WC, Diamond IT. 1970. Cortical visual area I and II in the a homologue of the second visual are in marsupials. Eur J Neurosci
hedgehog: relations between evoked potential maps and architectonic 11:907–915.
subdivisions and connections from the visual thalamus. J Neurophysiol Rose M. 1929. Cytoarchitektonischer Atlas der Grosshirnrinde der Maus.
33:595– 615. J Psychol Neurol 40:1–51.
Kaas JH, Hall WC, Killackey HP, Diamond IT. 1972. Visual cortex of the Rumberger A, Tyler CJ, Lund JS. 2001. Intra- and inter-areal connections
tree shrew (Tupaia glis): architectonic subdivisions and representa- between the primary visual cortex V1 and the area immediately sur-
tions of the visual field. Brain Res 42:491– 496. rounding V1 in the rat. Neuroscience 102:35–52.
Kaas JH, Krubitzer LA, Johanson KL. 1989. Cortical connections of areas Sanderson KJ, Dreher B, Gayer N. 1991. Prosencephalic connections of
17 (V-I) and 18 (V-II) of squirrels. J Comp Neurol 281:426 – 446. striate and extrastriate areas of rat visual cortex. Exp Brain Res
Kahn DM, Huffman KJ, Krubitzer L. 2000. Organization and connections 85:324 –334.
of V1 Monodelphis domestica. J Comp Neurol 428:337–354. Schuett S, Bonhoeffer T, Hübener M. 2002. Mapping retinotopic structure
Kalatsky VA, Stryker MP. 2003. New paradigm for optical imaging: tem- in mouse visual cortex with optical imaging. J Neurosci 22:6549 – 6559.
porally encoded maps of intrinsic signal. Neuron 38:529 –545. Sefton AJ, Dreher B, Harvey A. 2004. The visual system. In: Paxinos H,
Kalia M, Whitteridge G. 1973. The visual areas in the splenial sulcus of the editor. The rat nervous system. Amsterdam: Elsevier. p 1083–1165.
cat. J Physiol 232:275–283. Shepherd GMG, Svoboda K. 2005. Laminar and columnar organization of
Kim U, Ebner FF. 1999. Barrels and septa: separate circuits in rat barrel ascending excitatory projections to layer 2/3 pyramidal neurons in rat
field cortex. J Comp Neurol 408:489 –505. barrel cortex. J Neurosci 25:5670 –5679.
Lewis JW, Olavarria JF. 1995. Two rules for callosal connectivity in striate Simmons PA, Lemmon V, Pearlman, AL. 1982. Afferent and efferent con-
cortex of the rat. J Comp Neurol 361:110 –137. nections of the striate and extrastriate visual cortex of the normal and
reeler mouse. J Comp Neurol 211:295–308.
Lyon DC, Jain N, Kaas JH. 1998. Cortical connections of striate and
extrastriate visual areas in tree shrews. J Comp Neurol 401:109 –128. Sur M, Rubenstein JLR. 2005. Patterning and plasticity of the cerebral
cortex. Science 310:805– 810.
Malach R. 1989. Patterns of connections in rat visual cortex. J Neurosci
9:3741–3752. Spatz WB, Vogt DM, Illing RB. 1991. Delineation of the striate cortex, and
the striate peristriate projections in the guinea pig. Exp Brain Res
Mallamaci A, Stoykova A. 2006. Gene networks controlling early cerebral 84:495–504.
cortex arealization. Eur J Neurosci 23:847– 856.
Symonds LL, Rosenquist AC. 1984. Corticocortical connections among
Miller MW, Vogt BA. 1984. Direct connections of rat visual cortex with visual areas in cat. J Comp Neurol 229:1–38.
sensory, motor, and association cortices. J Comp Neurol 226:184 –202.
Thomas HC, Espinoza SG. 1987. Relationship between interhemispheric
Montero VM. 1993. Retinotopy of cortical connections between the striate cortical connections and visual areas in hooded rats. Brain Res 417:
cortex and extrastriate visual areas in the rat. Exp Brain Res 94:1–15. 214 –224.
Montero VM, Rojas A, Torrealba F. 1973. Retinotopic organization of Tiao Y-C, Blakemore C. 1976. Functional organization in the visual cortex
striate and peristriate visual cortex in the albino rat. Brain Res 53: of the golden hamster. J Comp Neurol 168:459 – 482.
197–201.
Towns LC, Giolli RA, Haste DA. 1977. Corticocortical fiber connections of
Olavarria J, Mendez B. 1979. The representations of the visual field on the the rabbit visual cortex: a fiber degeneration study. J Comp Neurol
posterior cortex in Octodon degus. Brain Res 161:539 –543. 173:537–560.
Olavarria J, Montero VM. 1989. Organization of visual cortex in the mouse Van Essen DC. 1979. Visual areas of the mammalian cerebral cortex. Annu
revealed by correlating callosal and striate-extrastriate connections. Rev Neurosci 2:227–263.
Vis Neurosci 3:59 – 69.
Van Essen DC, Maunsell JHR, Bixby JL. 1981. The middle temporal visual
Olavarria J, Montero VM. 1990. Elaborate organization of visual cortex in area in the macaque: myeloarchitecture, connections, functional prop-
the hamster. Neurosci Res 8:40 – 47. erties and topographic organization. J Comp Neurol 199:293–326.
Olavarria J, Torrealba F. 1978. The effects of acute lesions of the striate Van Essen DC, Newsome WT, Bixby JL. 1982. The pattern of interhemi-
cortex on the retinotopic organization of the lateral prestriate cortex in spheric connections and its relationship to extrastriate visual areas in
the rat. Brain Res 151:386 –391. the macaque monkey. J Neurosci 2:265–283.
Olavarria J, Mignano LR, Van Sluyters RC. 1982. Pattern of extrastriate Van Hooser SD, Heimel LA, Chung S, Nelson SB. 2006. Lack of patchy
visual areas connecting reciprocally with striate cortex in the mouse. horizontal connectivity in primary visual cortex of a mammal without
Exp Neurol 78:775–779. orientation maps. J Neurosci 26:7680 –7692.
Payne BR. 1993. Evidence for visual cortical area homologs in cat and Wagor E, Mangini NJ, Pearlman AL. 1980. Retinotopic organization of
macaque monkey. Cereb Cortex 3:1–25. striate and extrastriate visual cortex in the mouse. J Comp Neurol
Paxinos G, Franklin KBJ. 2001. The mouse brain in stereotaxic coordi- 193:187–202.
nates. New York: Academic. Wang Q, Burkhalter A. 2004. Dorsal and ventral streams for processing of
Rash BG, Grove EA. 2006. Area and layer patterning in the developing visual information in mouse cerebral cortex. Soc Neurosci Abstr 300.5.
cerebral cortex. Curr Opin Neurobiol 16:25–34. Wang Q, Burkhalter A. 2005. Separate output streams from V1 to higher
Roe AW, Ts’o DY. 1995. Visual topography in primate V2: multiple repre- areas of mouse visual cortex. Soc Neurosci Abstr 854.1.
sentations across functional stripes. J Neurosci 15:3689 –3715. Wang Q, Gao E, Burkhalter A. 2007. In vivo transcranial imaging of
Rosa MGP. 1997. Visuotopic organization of primate extrastriate cortex. connections in mouse visual cortex. J Neurosci Methods 159:268 –276.