N E W S & V I E W S

animal models of relapse and neuroimagingstudies of craving addicts9,10.

When Choi et al. measured glutamate-mediated excitatory postsynaptic currents(EPSCs) in the nucleus accumbens ofENT1–/– mice, they found an increased fre-quency and magnitude of spontaneousEPSCs compared to ENT1+/+ mice.Moreover, although blocking A1 receptorsdid not enhance evoked EPSCs, stimulatingthe receptors inhibited the evoked responses.This demonstrated that the increase in gluta-mate transmission seen in the ENT1–/– miceresulted from a lack of endogenous adeno-sine tone, which was not due to a loss of A1receptors or decreases in A1 receptor affinity.Although the levels of endogenous extracel-lular adenosine were not determined, thesedata clearly point to a reduced A1 recep-tor–mediated inhibition of excitatory trans-mission in animals lacking ENT1.

A consequence of elevated glutamate anddopamine transmission in the accumbensobserved with many other drugs of abuse isincreased CREB phosphorylation (pCREB)11.Importantly, Choi et al. also found increasedpCREB in association with the increase inglutamate transmission observed in theaccumbens of ENT1–/– mice. Because theincrease in pCREB is linked with intracellular

changes thought to be critical for transition-ing between social or recreational drug useand addictive patterns of drug use11,12, thesedata suggest an involvement of ENT1 in thedevelopment of alcoholism. Figure 1 illus-trates the sequence of events whereby ENT1deletion, and perhaps chronic alcohol use,alters adenosine regulation of glutamaterelease in cortical inputs to the accumbens,and thereby induces pCREB.

Although the experiments by Choi et al.2

implicated ENT1 and decreased A1 adeno-sine tone in promoting alcohol consump-tion, one more critical experiment helpedstrengthen the authors’ proposition. Torestore A1 adenosine tone, the researchersadministered a compound to stimulate A1receptors in the ENT1–/– mice, which signifi-cantly diminished ethanol consumption andpreference in the mice. In contrast, the com-pound had no effect in wild-type ENT1+/+

mice. These data directly demonstrate thatincreased voluntary alcohol consumption inENT1–/– mice is related to a decrease in A1adenosine tone.

Demonstrating the importance of A1adenosine tone on glutamate receptor signal-ing in the nucleus accumbens points investiga-tors toward glutamate input from the cortex tothe accumbens as a potential site of primary

pathology in alcohol addiction. Moreover,Choi et al. have shown that this projection canalso be a site of genetic disposition for alco-holism. In spite of an intense worldwideresearch effort, alcoholism and other addic-tion disorders have proven extremely difficultto ameliorate. These new findings indicatethat the development of compounds capableof regulating neurotransmission and cell sig-naling at A1 or glutamate receptors could pro-vide valuable new pharmacotherapeutic toolsfor treating alcoholism.

1. McBride, W.J. & Li, T.K. Crit. Rev. Neurobiol. 12,339–369 (1998).

2. Choi, D.S. et al. Nat. Neurosci. 7, 855–861 (2004).3. Diamond, I. & Gordon, A. Physiol. Rev. 77, 1–19

(1997).4. Nagy, L.E. et al. Mol. Pharmacol. 36, 744–748

(1989).5. Schuckit, M.A. Am. J. Psychol. 15, 991–996

(1994).6. Jennings, L.L. et al. Neuropharmacology 40,

722–731 (2001).7. Weiss, F. & Porrino, L.J. J. Neurosci. 22, 3332–3337

(2002).8. Harvey, J. & Lacey, M.G. J. Neurosci. 17,

5271–5280 (1997).9. Kalivas, P.W. & McFarland, K. Psychopharmacology

168, 44–56 (2003).10. Goldstein, R.A. & Volkow, N.D. Am. J. Psychiatry

159, 1642–1652 (2002).11. Nestler, E.J. Nat. Rev. Neurosci. 2, 119–128

(2001).12. Carlezon, W.A. Jr. et al. Science 282, 2272–2275

(1998).

796 VOLUME 7 | NUMBER 8 | AUGUST 2004 NATURE NEUROSCIENCE

A shrewd insight for visionFarran Briggs & W Martin Usrey

The circuitry responsible for generating orientation-specific responses in primary visual cortex remains controversial. A new studyidentifies an anatomical substrate for orientation selectivity and suggests the mechanism may be conserved across species.

For over 40 years, scientists have hotlydebated how orientation tuning emerges inthe visual system. One reason for suchintense scrutiny is that this work providesinsight into the general question of howstructure underlies function in the brain. Aswith other emergent properties of corticalneurons, identifying the circuitry responsi-ble for orientation tuning has provenextremely difficult. This difficulty is partlydue to the complexity of intracortical con-nections, along with the strong possibilitythat more than one mechanism may be atplay. Hubel and Wiesel suggested that orien-

tation tuning results from an anatomicallyprecise organization of feedforward projec-tions from the lateral geniculate nucleus(LGN) of the thalamus to layer 4 of visualcortex1. In contrast, alternative models assertthat feedforward projections lack sufficientstrength and precision to support orienta-tion tuning; these models emphasize the roleof intracortical mechanisms in amplifyingand sharpening orientation tuning2–5. Inthis issue, Mooser, Bosking and Fitzpatrick6

identify an elegant anatomical substrate thatmay support orientation tuning in the treeshrew. This substrate shares features withboth types of models—feedforward andintracortical—and could serve to establish afundamental principle governing connec-tions within the cerebral cortex.

‘Orientation tuning’ refers to the observa-tion that cortical neurons respond strongly to

stimuli presented at a particular orientationand less to stimuli presented at other orienta-tions. In the visual system of the cat, wherethis property has been studied most exten-sively, orientation-tuned neurons are presentat the first stage of cortical processing—layer4. Individual layer 4 neurons in cat havereceptive fields that are elongated along theaxis of preferred orientation1, whereas LGNinputs have circular receptive fields and lackorientation tuning7. Based on these differ-ences, Hubel and Wiesel proposed a straight-forward model for the generation oforientation tuning1. In this model, orienta-tion-tuned neurons in layer 4 receive conver-gent input from several LGN neurons whoseindividual receptive fields are displaced alonga line of visual space. The combined inputfrom the ensemble of LGN neurons estab-lishes an elongated receptive field that,

The authors are in the Center for Neuroscience,

University of California, Davis, California 95616,

USA.

e-mail: wmusrey@ucdavis.edu

©20

04 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce

N E W S & V I E W S

together with mechanisms of summationand threshold, confers orientation tuning onthe cortical neuron.

Although the cat has served as a modelanimal for studies of the visual system, thereare important species-specific differences inthe functional organization of the visualpathway. In particular, although orientationtuning is a prominent feature of layer 4 neu-rons in cat visual cortex, neurons in layer 4 ofmany primates and insectivores lack orienta-tion tuning, which instead emerges one ortwo synapses after the geniculocorticalsynapse. Thus, input from the LGN to layer 4cannot be directly responsible for the estab-lishment of orientation tuning, as it is in thecat8–13. Instead, orientation tuning must relyon intracortical mechanisms. Can the Hubeland Wiesel model work given the organiza-tion of the primate and insectivore visualsystems? The new paper from Mooser et al.6

demonstrates that the intracortical, feedfor-ward projections from layer 4 to 2/3 in thetree shrew are organized in a fashion similarin spirit to the geniculocortical connectionsin the Hubel and Wiesel model. Thus, thereappears to be a generalized and conservedmechanism contributing to the generationof orientation tuning across species.

The tree shrew is a small, diurnal insectivorefrom Southeast Asia. Once considered a pri-mate, these animals have a highly developed

visual system packaged in a relatively smallbrain. These features, along with a lack of ocu-lar dominance columns and an extremely pre-cise retinotopic map across the visual cortexwith very little distortion, make the tree shrewan ideal animal for studying structure–func-tion relationships in the cortex14.

To examine the organization of layer 4connections to layers 2/3, where orientationtuning emerges in the tree shrew, Mooser andcolleagues made small injections of dye intothe middle of layer 4 and compared the dis-tribution of labeled boutons in layers 2/3 tothe map of orientation preference, as deter-mined with optical imaging techniques.Although all local orientation columnsreceived input from the labeled layer 4 pro-jections, Mooser et al.6 noticed an organiza-tion to the projections whereby boutonsclustered in orientation columns whose ori-entation preference matched the retinotopicdisplacement of the boutons from theirinjection site. From the perspective of a sin-gle, unoriented layer 4 neuron, synapses aremade with a variety of layer 2/3 neurons withoverlapping receptive fields and differing ori-entation preferences. From the perspective ofa single layer 2/3 neuron, inputs come frommultiple layer 4 neurons with receptive fieldsdisplaced along a line of visual space (Fig. 1).

Although the results of Mooser et al.6

indicate that intracortical, feedforward pro-

jections from layer 4 neurons establish asubstrate for orientation tuning, additionalcortical mechanisms, both excitatory andinhibitory, are almost certainly involved inthe refinement and maintenance of orienta-tion tuning4,5,13. In particular, previouswork from the authors’ laboratory hasshown that layer 2/3 neurons in the treeshrew give rise to long-distance, patchy,horizontal projections15. These horizontalprojections not only link neurons with sim-ilar orientation preferences, but also extendanisotropically across the cortex, linkingneurons along a retinotopic axis thatmatches the orientation preference of theprojecting neuron15. Thus, although feed-forward connections from layer 4 to 2/3 areinvolved in the emergence of orientation, itis important to keep in mind that this pathway is not operating in isolation,but rather represents one component of alarger machine.

By combining clever thinking with state-of-the-art techniques and an excellent ani-mal model, Mooser et al.6 have significantlyadvanced our progress in the ongoing questto understand structure–function relation-ships in the cortex. The finding that a common rule applies to feedforward con-nections and the emergence of orientationtuning in both the shrew and the catstrongly suggest that a similar plan mayexist in other species.

1. Hubel, D.H. & Wiesel, T.N. J. Physiol. (Lond.) 160,106–154 (1962).

2. Sillito, A.M., Salt, T.E. & Kemp, J.A. Vision Res. 25,375–381 (1985).

3. Somers, D.C., Nelson, S.B. & Sur, M. J. Neurosci 15,5448–5465 (1995).

4. Sompolinsky, H. & Shapley, R. Curr. Opin. Neurobiol.7 (1997).

5. Martin, K.A.C. Curr. Opin. Neurobiol. 12, 418–425(2002).

6. Mooser, F., Bosking, W.H. & Fitzpatrick, D. Nat.Neurosci. 7, 872–879 (2004).

7. Hubel, D.H. & Wiesel, T.N. J. Physiol. (Lond.) 155,385–398 (1961).

8. Chapman, B., Zahs, K.R. & Stryker, M.P. J. Neurosci.11, 1347–1358 (1991).

9. Reid, R.C. & Alonso, J.-M. Nature 378, 281–284(1995).

10. Ferster, D., Chung, S. & Wheat, H. Nature 380,249–252 (1996).

11. Reid, R.C. & Alonso, J.-M. Curr. Opin. Neurobiol. 6,475–480 (1996).

12. Ferster, D. & Miller, K.D. Annu. Rev. Neurosci. 23,441–471 (2000).

13. Miller, K.D. Cereb. Cortex 13, 73–82 (2003).14. Fitzpatrick, D. Cereb. Cortex 6, 329–341 (1996).15. Bosking, W.H., Zhang, Y., Schofield, B. &

Fitzpatrick, D. J. Neurosci 17, 2112–2127 (1997).

NATURE NEUROSCIENCE VOLUME 7 | NUMBER 8 | AUGUST 2004 797

Figure 1 Schematic diagram illustrating cortical circuits reported by Mooser et al.6. The orientationmap in cortical layer 2/3 is constructed by combining optical imaging responses to stimuli presented atvarious orientations, coded in different colors (upper right). Anatomical tracing experiments suggestthat orientation tuning in layer 2/3 could arise from input from a group of layer 4 neurons whosereceptive fields and cell bodies are arranged in a line (adapted from ref. 15).

©20

04 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

uren

euro

scie

nce