Different glutamate receptors convey feedforward and recurrent processing in macaque V1

Task and Behavior.

We used a version of the figure–ground segregation task (25) where the monkey was required to indicate the location of a texture-defined figure with a saccade (Fig. 1). On catch trials (25% of trials), the texture was presented with no figure present, and the monkey was required to maintain fixation. The figures were salient because previous reports have demonstrated that these produce strong FGM in V1 (25). The monkeys performed the task with high accuracy. If the figure was present, monkey S was correct on 95.8% and monkey E was correct on 98.0% of the trials. Accuracy on catch trials was somewhat lower (monkey S = 74.5% correct; monkey E = 91.7% correct).

Effect of NMDA-R Antagonists at Example Recording Sites.

We recorded the envelope of the multiunit activity (MUAe) simultaneously from the different laminae of V1 in three hemispheres of the two monkeys with a multicontact electrode (“U-probe”; Plexon). In addition to 24 recording sites, the electrode contained a fluid line, which allowed us to pressure-inject small quantities (25–100 nL) of drugs into V1 (Fig. S1). We placed the electrode at the desired depth in cortex using current-source density analysis (Fig. S2) and measured the location and size of the multiunit RFs for each recording site (Fig. S3). We then presented figure–ground textures and ensured that the RF fell either in the center of the figure or in the background region (Fig. 1B). We quantified the effect of the drugs using two measures: the peak response and FGM. The peak response was measured as the average activity across both conditions in a window from 50 to 100 ms after stimulus onset. FGM was measured as the difference between the figure and background response from 100 to 200 ms after figure onset expressed as a fraction of the peak response. All MUAe responses before and after drug application were normalized to the peak response of the predrug epoch for each recording site. An example recording in which we injected APV is illustrated in Fig. 2A. In the predrug control period, we observed the typical FGM profile. After ∼100 ms, the response evoked by the figure (thick, dark blue trace) became stronger than activity evoked by the homogeneous background (thin, light blue trace). After this baseline measurement, we injected 30 nL of 50 mM APV. The example in Fig. 2A shows the typical effect of APV. The drug caused a small reduction in peak response by 9% of the predrug level (P < 0.01; t test). In the later response phase, APV reduced the figure response, but it barely affected the background response. This led to a reduction in FGM, by 46.2% of the predrug level (ANOVA; F_1,463 = 5.6, P < 0.02). To investigate whether the decrease in FGM was caused by the change in response strength, we calculated a modulation index (MI), which is insensitive to the overall level of activity because it is defined as the difference between the figure and ground response in a window from 100 to 200 ms after stimulus onset divided by the summed activity in that window [MI = (figure − ground)/(figure + ground)]. APV reduced the MI from 0.13 to 0.07 (a 41.8% reduction), which indicates that the reduction in FGM was larger than the reduction in overall activity. In a few penetrations, the visual response increased after APV application, but FGM was reduced (Fig. S4). FGM recovered well within 60 min of drug application. We also studied injections of ifenprodil, which is an NMDA antagonist that predominantly inactivates NMDA-Rs containing the NR2B subunit. Fig. 2B shows a recording site where we injected ifenprodil, which to our surprise caused an increase in the peak response by 11.5% (P < 0.01; t test), an effect that was typical for ifenprodil injections. Despite this increase in activity, we found that ifenprodil reduced the level of FGM, in this example by 94% of the predrug level (ANOVA; F_1,450 = 4.73; P = 0.03). The combination of an increase in overall activity with a decrease in FGM reduced the MI from 0.14 to 0.005 (a 96% reduction). We did not observe much recovery from the ifenprodil injection, even though we waited for over 1.5 h after the injection, but this result is not unexpected because long recovery times have also been reported for ifenprodil in cortical slices (30) (Fig. S5).

Effects of NMDA-R Antagonists at the Population Level.

Across the population of recording sites, we observed that APV (n = 388) and ifenprodil (n = 295) had opposite effects on the visually driven response. APV decreased visual responses by an average of 7.5% if we analyzed the entire response epoch (0–200 ms) (Fig. 3A). APV reduced both the peak response (50–100 ms) by 8.3% (Fig. 3B; t test; P < 0.05 in both monkeys) and the later sustained response (100–200 ms) by 8.1% (t test; both P < 0.001; Fig. 3C). In contrast, ifenprodil reliably increased neural responses (Fig. 3A). Across the population of recording sites, the average increase in peak response was 7.8%, and responsiveness also increased in the later analysis window (t tests; all P values, <0.001 for both monkeys) (green symbols in Fig. 3 B and C). Despite their opposite effects on visual responsiveness, both NMDA antagonists reduced FGM. APV reduced responses in both the figure and ground conditions, but the reduction was greater in the figure condition leading to the decrease in FGM (Fig. 3C). APV reduced FGM by 34.2% across all penetrations (Fig. 4A; t test; both monkeys, P < 0.001). Moreover, this decrease in FGM was stronger than predicted by the decrease in visual responsiveness, so that the MI was also reduced, from 0.14 to 0.10 (22.2% reduction; t test; both monkeys P < 0.05; Fig. 4B). Fig. 4C shows the effect of APV across individual penetrations. The majority of penetrations exhibited significant reductions in FGM. In some, FGM was not reduced, but we could not exclude that the injection failed in some of these experiments. Ifenprodil also reduced FGM. It increased responses evoked by the figure and by the background, but the increase was greater for the background responses, causing a decrease in FGM (Fig. 3C). Ifenprodil caused significant reductions in FGM in 10 out of 17 penetrations (Fig. 4D). Across all recording sites, the average reduction in FGM was 21.5% (t test; both monkeys, P < 0.001; Fig. 4A). Moreover, this reduction in FGM combined with an increase in visually driven activity reduced the MI from an average value of 0.19 to 0.13 (33.2% reduction; t test; both monkeys, P < 0.001; Fig. 4B). The time course of the effect of the two drugs differed. APV produced immediate effects, which recovered well within 30 min (yellow bars in Fig. 4 A and B), whereas the effects of ifenprodil took longer to reach their maximum and were relatively long-lasting, so that the changes in neuronal activity did not return to baseline, even if the recording session was extended to >2 h postinjection (30) (Fig. S5). The effects of the two drugs also differed across the different layers of V1 (Fig. S6). Responses during catch trials were very similar to responses during the ground condition (triangles in Fig. 3 B and C), although responses from monkey E were slightly, but significantly (t test; P < 0.001), higher during catch trials than ground trials. This difference indicates that the presence of a figure elsewhere caused a decrease in the response evoked by the background. The drugs did not influence this difference between catch and ground trials (t test, P > 0.05).

Effects of an AMPA-R Antagonist.

In separate experiments, we injected CNQX, which is an antagonist of the AMPA-R. Fig. 2C shows an example recording site where CNQX was injected. In this example, CNQX produced a large reduction in the peak response by 56.5% of the predrug level (P < 0.001; t test). Interestingly, the decrease in peak response was not accompanied by a significant change in FGM (ANOVA; F_1,785 = 1.19; P = 0.28). In this example, responses recovered within 90 min of the injection. The effects of CNQX on the overall response level were consistent across the population with 93.5% of recording sites (n = 252) showing a reduction in response (Fig. 3 A and B and Fig. S7A). The drug reduced the initial peak response on average by 37.5% of the predrug level (t test; both monkeys, P < 0.001; Fig. 3 A and B). At the population level, the decrease in activity during the later sustained response phase was similar to the reduction in peak response (reduction of 26.1%;t test; both monkeys, P < 0.01; Fig. 3C). The effects of CNQX on the figure and background responses were similar (Fig. 3C), and the drug, therefore, did not cause a significant reduction in either FGM (t test; P = 0.17) or MI (t test; P = 0.27) (Fig. 4 A and B and Fig. S7C). Recovery from CNQX injections was partial, and its effect on the visually driven response recovered by ∼85% (yellow bar in Fig. 3A). As a control for the effects of the pressure injection, we also carried out separate penetrations in which we injected water or artificial cerebrospinal fluid (aCSF) for a total of 224 recording sites. These control injections had no effect on the peak response (t test; P > 0.05) (black symbols in Figs. 3 and 4; also see Fig. S7 B and D), FGM (t test; P > 0.05), or MI (t test; P > 0.05). Control analyses demonstrated that the effects reported here could not be explained by small changes in fixational eye movements (SI Results).

Are Feedforward Connections AMPA-Ergic and Recurrent Connections NMDA-Ergic?

Theoretical studies have suggested that feedforward connections drive the neurons in a cortical column, whereas recurrent connections can only modulate the activity (3, 4). It is advantageous that recurrent connections do not drive their target neurons, because this could lead to positive-feedback loops causing uncontrolled activity (3). Models of cortical processing suggested that the properties of driving and modulatory connections could be related to differences in the proportion of AMPA-Rs and NMDA-Rs (38). Specifically, AMPA-Rs were suggested to propagate visual activity from lower to higher areas, whereas NMDA-Rs were proposed to cause modulatory effects mediated by recurrent connections (17, 18). On the one hand, AMPA-Rs directly depolarize the postsynaptic neuron, and this could explain the efficient activation of neurons by feedforward connections, which drive cells and determine their tuning (22, 32, 34). The presence of NMDA-Rs in recurrent connections, on the other hand, could explain why these connections are merely modulatory. If glutamate binds to NMDA-Rs, then the channels only pass current upon removal of the magnesium block by sufficient depolarization caused by concomitant activation of AMPA-Rs (19). NMDA-ergic recurrent connections could, thereby, amplify a selective subset of the neuronal responses evoked by AMPA-ergic feedforward connections, and this mechanism could explain why modulatory effects are most pronounced for neurons that are well driven by a visual stimulus (5, 20, 21). NMDA channels of neurons that are not driven by feedforward connections would remain blocked, so that recurrent connections would have little impact on these cells. The idea that feedforward connections use AMPA-Rs and recurrent connections NMDA-Rs is presumably only a first approximation, however, because cortical synapses in adult animals usually contain functional AMPA-Rs as well as NMDA-Rs (39). Nevertheless, there are substantial differences in the AMPA/NMDA ratio between cortical areas and also between layers within area V1 (34, 40–42), which suggest that recurrent connections may target synapses that are richer in NMDA-Rs than feedforward connections. A previous study in anesthetized cats (23) supported the different roles of these glutamate receptors by showing that activation of AMPA-Rs added a relatively constant number of spikes to the firing rate of a neuron, whereas activation of the NMDA-R caused an increase in response gain. Similarly, a study on direction selectivity (43) showed that AMPA-R activity produces directionally selective responses, whereas NMDA-R activity amplifies neuronal responses evoked by a stimulus moving in the preferred direction. In other words, NMDA-Rs increase the firing rate of active cells and have little effect on neurons not driven by the visual stimulus.

Other Mechanisms That Could Explain the Effect of NMDA Blockers on FGM.

We also considered other explanations for the effects of NMDA antagonists on FGM that do not depend on differences in glutamate receptor distributions. First, we considered the possibility that the differential effect of AMPA and NMDA blockers could be caused by NMDA channels opening more slowly than AMPA channels (19). However, we observed that the NMDA-R antagonists had significant, albeit opposite, effects on the initial transient response, which is incompatible with explanations solely based on differences in receptor dynamics. Second, an intriguing recent study (44) has opened up the possibility that NMDA-Rs play a crucial role in integrating synaptic inputs contacting the apical dendrites of pyramidal neurons, which are key targets for feedback projections (45). The activation of NMDA-Rs induces NMDA spikes, which are regenerative events in the dendritic tufts of cortical neurons that can give rise to a calcium spike. This calcium spike, in turn, propagates along the apical dendrite to the soma to influence the firing rate of the neuron. Blocking of NMDA-Rs reduces the ability of layer I synapses to generate a dendritic calcium spike (44). In our experiment the NMDA blockers may also have prevented the NMDA spikes induced by feedback connections that contribute to figure–ground segregation, leading to a reduction in FGM. We did not completely remove FGM with injections of APV or ifenprodil. It is, therefore, possible that other receptors also contribute to FGM. A recent study that used drugs that target acetylcholine receptors, for example, demonstrated that attentional feedback effects in area V1 depend on the activation of muscarinic receptors (46). However, it is also possible that the partial removal of FGM reflects the small amount of drug that we injected. Higher concentrations and larger quantities might lead to more complete blocks.

Is Feedback to V1 Necessary to Solve the Figure-Ground Task?

FGM correlates with the ability of monkeys to segment the image and detect the figure (25). Moreover, FGM does not occur in anesthetized animals (47). These findings suggest that recurrent interactions between area V1 and higher visual areas are necessary to solve the figure–ground segregation task. A strict test of this hypothesis would be to block or reduce FGM in area V1 while allowing the normal feedforward propagation of the visual response. In this study, we observed that NMDA-R antagonists come close to fulfilling these criteria. Both APV and ifenprodil reduce the level of FGM, and the effects of these drugs on the feedforward response are small compared with the effects of CNQX. Our behavioral paradigm was designed to produce strong FGM, and the figures were always well above the threshold for perception. Future studies could use more sensitive behavioral paradigms to test whether the reductions in FGM caused by NMDA blockers interfere with figure perception.

Task and Training.

All procedures were approved by the Institutional Animal Care and Use Committee of the Royal Netherlands Academy of Arts and Sciences. We recorded from two adult macaque monkeys (Macaca Mulatta: monkeys S and E). The monkeys were implanted with a head post for head stabilization and a recording chamber over V1 using standard techniques (SI Materials and Methods). The animals were trained to fixate within a window of 1° diameter centered on a small red circular fixation point (0.3° in diameter). We recorded eye movements using a video eye tracker (Thomas Recording) sampled at 250 Hz. Every trial began with the fixation point presented on the gray background, and the monkeys triggered the beginning of the trial when their gaze entered a 1° fixation window centered on the fixation point. After 300 ms of fixation, a full-screen texture stimulus was presented. The texture consisted of black oriented lines (either 45° or 135°) on a white background (Fig. 1A). The texture was designed so that, on average, the image elements in the RF were constant across conditions (SI Materials and Methods). On figure and ground trials, the texture contained a 4° wide square figure. The location of the figure was varied so that the RFs of the neurons fell on the figure or on the background (Fig. 1B). After a further 300 ms, the fixation point was extinguished, and the monkey was required to make a saccade into a target window centered on the figure position. On 25% of trials, we presented a homogeneous texture without a figure (catch condition), and the animals were rewarded for carrying on fixating within the fixation window for a further 400 ms. The stimulus conditions were presented in a pseudorandom order. We made the recordings in epochs of 200 correct trials with each session consisting of at least one predrug epoch, one drug epoch, and one recovery epoch, although we typically recorded several recovery epochs.

Electrophysiological Methods.

We used laminar electrodes (“U-probes”; Plexon) for all recordings. We used a 24-contact version of the probe, with a diameter of 420 μm, a contact diameter of 25 μm, and an intercontact spacing of 100 μm. The impedance of the contact points was in the range of 300–700 kΩ at 1 kHz, allowing us to record multiunit data. The recordings were grounded and referenced to the guide tube used to stabilize the transition of the electrode across the dura. (The guide tube did not penetrate the dura.) The head stage amplified the signal by four times before the signal was passed to a preamplifier (Tucker-Davis Technologies). The signal was digitized at 24.4 kHz. As in previous studies (48, 49), we measured the MUAe (SI Materials and Methods). The signal was bandpass-filtered (500–5,000 Hz), full-wave-rectified, and then low-pass-filtered (200 Hz) and sampled at a rate of 763 Hz. The MUAe represents the pooled activity of a number of single units in the vicinity of the tip of the electrode, and the population response obtained with this method is, therefore, expected to be identical to the population response obtained by pooling across single units (49). Multiunit RFs were plotted using a moving bar stimulus as reported previously (49) (Fig. S3). Further details about the MUAe recording technique and comparisons with thresholded MUA are provided in SI Materials and Methods and Figs. S8–S10.

Data Analysis.

The MUAe traces for each trial were filtered with notch filters at 50 and 85 Hz to remove possible contamination from line noise and the monitor refresh rate. For each recording site, we calculated an average response after subtracting the pretrial baseline and then normalizing to the average response in the peak window (50–100 ms) from the predrug, catch-trial condition. The MUAe data we report are, therefore, expressed as fractions of the average peak response, e.g., a change of 0.1 units equals a change of 10% of the peak response. We removed recording sites with a signal-to-noise ratio (SNR) of less than 1 (12.8% of sites). SNR was calculated as the ratio between the peak response and the SD of the spontaneous activity level across trials. For statistical comparisons at the population level, we used two-sided t tests for every drug to test whether it changed the visual response or FGM. Because we performed four tests for each time window (three drugs plus control), we applied a Bonferroni correction for multiple comparisons. Results are only reported as significant if the corrected P value was less than 0.05 for both monkeys individually and when averaged across monkeys. The MI [(figure − ground)/(figure + ground)] is an unstable measure for recording sites in which the denominator is close to zero (i.e., sites where the overall response level during the sustained period drops back to baseline). To avoid the average MI across recording sites becoming dominated by extreme values, we removed recording sites in which the sustained period activity was lower than 5% of the peak response or the absolute z score of sustained period activity was greater than 3 (3.9% of sites).

Supporting Information

• 1.Felleman DJ, Van Essen DC. Distributed hierarchical processing in the primate cerebral cortex. Cereb Cortex. 1991;1:1–47. doi: 10.1093/cercor/1.1.1-a. [DOI] [PubMed] [Google Scholar]
• 2.Hubel DH, Wiesel TN. Receptive fields and functional architecture of monkey striate cortex. J Physiol. 1968;195:215–243. doi: 10.1113/jphysiol.1968.sp008455. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 3.Crick F, Koch C. Constraints on cortical and thalamic projections: The no-strong-loops hypothesis. Nature. 1998;391:245–250. doi: 10.1038/34584. [DOI] [PubMed] [Google Scholar]
• 4.Sherman SM, Guillery RW. On the actions that one nerve cell can have on another: Distinguishing “drivers” from “modulators”. Proc Natl Acad Sci USA. 1998;95:7121–7126. doi: 10.1073/pnas.95.12.7121. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 5.Moore T, Armstrong KM. Selective gating of visual signals by microstimulation of frontal cortex. Nature. 2003;421:370–373. doi: 10.1038/nature01341. [DOI] [PubMed] [Google Scholar]
• 6.Roelfsema PR. Cortical algorithms for perceptual grouping. Annu Rev Neurosci. 2006;29:203–227. doi: 10.1146/annurev.neuro.29.051605.112939. [DOI] [PubMed] [Google Scholar]
• 7.Desimone R, Duncan J. Neural mechanisms of selective visual attention. Annu Rev Neurosci. 1995;18:193–222. doi: 10.1146/annurev.ne.18.030195.001205. [DOI] [PubMed] [Google Scholar]
• 8.Zipser K, Lamme VA, Schiller PH. Contextual modulation in primary visual cortex. J Neurosci. 1996;16:7376–7389. doi: 10.1523/JNEUROSCI.16-22-07376.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 9.Hupé JM, et al. Cortical feedback improves discrimination between figure and background by V1, V2 and V3 neurons. Nature. 1998;394:784–787. doi: 10.1038/29537. [DOI] [PubMed] [Google Scholar]
• 10.Angelucci A, et al. Circuits for local and global signal integration in primary visual cortex. J Neurosci. 2002;22:8633–8646. doi: 10.1523/JNEUROSCI.22-19-08633.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 11.Lamme VA. The neurophysiology of figure-ground segregation in primary visual cortex. J Neurosci. 1995;15:1605–1615. doi: 10.1523/JNEUROSCI.15-02-01605.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 12.Lamme VA, Roelfsema PR. The distinct modes of vision offered by feedforward and recurrent processing. Trends Neurosci. 2000;23:571–579. doi: 10.1016/s0166-2236(00)01657-x. [DOI] [PubMed] [Google Scholar]
• 13.Lamme VA. Towards a true neural stance on consciousness. Trends Cogn Sci. 2006;10:494–501. doi: 10.1016/j.tics.2006.09.001. [DOI] [PubMed] [Google Scholar]
• 14.Hupé JM, James AC, Girard P, Bullier J. Response modulations by static texture surround in area V1 of the macaque monkey do not depend on feedback connections from V2. J Neurophysiol. 2001;85:146–163. doi: 10.1152/jn.2001.85.1.146. [DOI] [PubMed] [Google Scholar]
• 15.Sandell JH, Schiller PH. Effect of cooling area 18 on striate cortex cells in the squirrel monkey. J Neurophysiol. 1982;48:38–48. doi: 10.1152/jn.1982.48.1.38. [DOI] [PubMed] [Google Scholar]
• 16.Johnson RR, Burkhalter A. Evidence for excitatory amino acid neurotransmitters in forward and feedback corticocortical pathways within rat visual cortex. Eur J Neurosci. 1994;6:272–286. doi: 10.1111/j.1460-9568.1994.tb00270.x. [DOI] [PubMed] [Google Scholar]
• 17.Dehaene S, Sergent C, Changeux JP. A neuronal network model linking subjective reports and objective physiological data during conscious perception. Proc Natl Acad Sci USA. 2003;100:8520–8525. doi: 10.1073/pnas.1332574100. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 18.Lumer ED, Edelman GM, Tononi G. Neural dynamics in a model of the thalamocortical system. I. Layers, loops and the emergence of fast synchronous rhythms. Cereb Cortex. 1997;7:207–227. doi: 10.1093/cercor/7.3.207. [DOI] [PubMed] [Google Scholar]
• 19.Daw NW, Stein PS, Fox K. The role of NMDA receptors in information processing. Annu Rev Neurosci. 1993;16:207–222. doi: 10.1146/annurev.ne.16.030193.001231. [DOI] [PubMed] [Google Scholar]
• 20.Treue S, Martínez Trujillo JC. Feature-based attention influences motion processing gain in macaque visual cortex. Nature. 1999;399:575–579. doi: 10.1038/21176. [DOI] [PubMed] [Google Scholar]
• 21.Ekstrom LB, Roelfsema PR, Arsenault JT, Bonmassar G, Vanduffel W. Bottom-up dependent gating of frontal signals in early visual cortex. Science. 2008;321:414–417. doi: 10.1126/science.1153276. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 22.Hagihara K, Tsumoto T, Sato H, Hata Y. Actions of excitatory amino acid antagonists on geniculo-cortical transmission in the cat’s visual cortex. Exp Brain Res. 1988;69:407–416. doi: 10.1007/BF00247586. [DOI] [PubMed] [Google Scholar]
• 23.Fox K, Sato H, Daw N. The effect of varying stimulus intensity on NMDA-receptor activity in cat visual cortex. J Neurophysiol. 1990;64:1413–1428. doi: 10.1152/jn.1990.64.5.1413. [DOI] [PubMed] [Google Scholar]
• 24.Roelfsema PR, Tolboom M, Khayat PS. Different processing phases for features, figures, and selective attention in the primary visual cortex. Neuron. 2007;56:785–792. doi: 10.1016/j.neuron.2007.10.006. [DOI] [PubMed] [Google Scholar]
• 25.Supèr H, Spekreijse H, Lamme VA. Two distinct modes of sensory processing observed in monkey primary visual cortex (V1) Nat Neurosci. 2001;4:304–310. doi: 10.1038/85170. [DOI] [PubMed] [Google Scholar]
• 26.Lamme VA, Supèr H, Spekreijse H. Feedforward, horizontal, and feedback processing in the visual cortex. Curr Opin Neurobiol. 1998;8:529–535. doi: 10.1016/s0959-4388(98)80042-1. [DOI] [PubMed] [Google Scholar]
• 27.Angelucci A, Bullier J. Reaching beyond the classical receptive field of V1 neurons: Horizontal or feedback axons? J Physiol Paris. 2003;97:141–154. doi: 10.1016/j.jphysparis.2003.09.001. [DOI] [PubMed] [Google Scholar]
• 28.Köhr G. NMDA receptor function: Subunit composition versus spatial distribution. Cell Tissue Res. 2006;326:439–446. doi: 10.1007/s00441-006-0273-6. [DOI] [PubMed] [Google Scholar]
• 29.Williams K. Ifenprodil discriminates subtypes of the N-methyl-D-aspartate receptor: Selectivity and mechanisms at recombinant heteromeric receptors. Mol Pharmacol. 1993;44:851–859. [PubMed] [Google Scholar]
• 30.Kumar SS, Huguenard JR. Pathway-specific differences in subunit composition of synaptic NMDA receptors on pyramidal neurons in neocortex. J Neurosci. 2003;23:10074–10083. doi: 10.1523/JNEUROSCI.23-31-10074.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 31.Sato H, Hata Y, Tsumoto T. Effects of blocking non-N-methyl-D-aspartate receptors on visual responses of neurons in the cat visual cortex. Neuroscience. 1999;94:697–703. doi: 10.1016/s0306-4522(99)00334-6. [DOI] [PubMed] [Google Scholar]
• 32.Armstrong-James M, Welker E, Callahan CA. The contribution of NMDA and non-NMDA receptors to fast and slow transmission of sensory information in the rat SI barrel cortex. J Neurosci. 1993;13:2149–2160. doi: 10.1523/JNEUROSCI.13-05-02149.1993. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 33.Miller KD, Chapman B, Stryker MP. Visual responses in adult cat visual cortex depend on N-methyl-D-aspartate receptors. Proc Natl Acad Sci USA. 1989;86:5183–5187. doi: 10.1073/pnas.86.13.5183. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 34.Fox K, Sato H, Daw N. The location and function of NMDA receptors in cat and kitten visual cortex. J Neurosci. 1989;9:2443–2454. doi: 10.1523/JNEUROSCI.09-07-02443.1989. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 35.Bair W, Cavanaugh JR, Movshon JA. Time course and time-distance relationships for surround suppression in macaque V1 neurons. J Neurosci. 2003;23:7690–7701. doi: 10.1523/JNEUROSCI.23-20-07690.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 36.Li Z. Visual segmentation by contextual influences via intra-cortical interactions in the primary visual cortex. Network. 1999;10:187–212. [PubMed] [Google Scholar]
• 37.Johnson RR, Burkhalter A. Microcircuitry of forward and feedback connections within rat visual cortex. J Comp Neurol. 1996;368:383–398. doi: 10.1002/(SICI)1096-9861(19960506)368:3<383::AID-CNE5>3.0.CO;2-1. [DOI] [PubMed] [Google Scholar]
• 38.Wang XJ. Synaptic reverberation underlying mnemonic persistent activity. Trends Neurosci. 2001;24:455–463. doi: 10.1016/s0166-2236(00)01868-3. [DOI] [PubMed] [Google Scholar]
• 39.Kharazia VN, Phend KD, Rustioni A, Weinberg RJ. EM colocalization of AMPA and NMDA receptor subunits at synapses in rat cerebral cortex. Neurosci Lett. 1996;210:37–40. doi: 10.1016/0304-3940(96)12658-6. [DOI] [PubMed] [Google Scholar]
• 40.Petralia RS, Yokotani N, Wenthold RJ. Light and electron microscope distribution of the NMDA receptor subunit NMDAR1 in the rat nervous system using a selective anti-peptide antibody. J Neurosci. 1994;14:667–696. doi: 10.1523/JNEUROSCI.14-02-00667.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 41.Rosier AM, Arckens L, Orban GA, Vandesande F. Laminar distribution of NMDA receptors in cat and monkey visual cortex visualized by [3H]-MK-801 binding. J Comp Neurol. 1993;335:369–380. doi: 10.1002/cne.903350307. [DOI] [PubMed] [Google Scholar]
• 42.Eickhoff SB, Rottschy C, Zilles K. Laminar distribution and co-distribution of neurotransmitter receptors in early human visual cortex. Brain Struct Funct. 2007;212:255–267. doi: 10.1007/s00429-007-0156-y. [DOI] [PubMed] [Google Scholar]
• 43.Rivadulla C, Sharma J, Sur M. Specific roles of NMDA and AMPA receptors in direction-selective and spatial phase-selective responses in visual cortex. J Neurosci. 2001;21:1710–1719. doi: 10.1523/JNEUROSCI.21-05-01710.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 44.Larkum ME, Nevian T, Sandler M, Polsky A, Schiller J. Synaptic integration in tuft dendrites of layer 5 pyramidal neurons: A new unifying principle. Science. 2009;325:756–760. doi: 10.1126/science.1171958. [DOI] [PubMed] [Google Scholar]
• 45.Rockland KS, Virga A. Terminal arbors of individual “feedback” axons projecting from area V2 to V1 in the macaque monkey: A study using immunohistochemistry of anterogradely transported Phaseolus vulgaris-leucoagglutinin. J Comp Neurol. 1989;285:54–72. doi: 10.1002/cne.902850106. [DOI] [PubMed] [Google Scholar]
• 46.Herrero JL, et al. Acetylcholine contributes through muscarinic receptors to attentional modulation in V1. Nature. 2008;454:1110–1114. doi: 10.1038/nature07141. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 47.Lamme VA, Zipser K, Spekreijse H. Figure-ground activity in primary visual cortex is suppressed by anesthesia. Proc Natl Acad Sci USA. 1998;95:3263–3268. doi: 10.1073/pnas.95.6.3263. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 48.Xing D, Yeh CI, Shapley RM. Spatial spread of the local field potential and its laminar variation in visual cortex. J Neurosci. 2009;29:11540–11549. doi: 10.1523/JNEUROSCI.2573-09.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 49.Supèr H, Roelfsema PR. Chronic multiunit recordings in behaving animals: Advantages and limitations. Prog Brain Res. 2005;147:263–282. doi: 10.1016/S0079-6123(04)47020-4. [DOI] [PubMed] [Google Scholar]

Supplementary Materials