J Neurophysiol 105: 1306 –1317, 2011. First published January 12, 2011; doi:10.1152/jn.00692.2010.
Dynamics of extraclassical surround modulation in three types of V1 neurons Yong-Jun Liu, Maziar Hashemi-Nezhad, and David C. Lyon Department of Anatomy and Neurobiology, School of Medicine, University of California, Irvine, California Submitted 4 August 2010; accepted in final form 11 January 2011
Liu Y, Hashemi-Nezhad M, Lyon DC. Dynamics of extraclassical surround modulation in three types of V1 neurons. J Neurophysiol 105: 1306 –1317, 2011. First published January 12, 2011; doi:10.1152/jn.00692.2010.—Visual stimuli outside of the classical receptive field (CRF) can influence the response of neurons in primary visual cortex (V1). While recording single units in cat, we presented drifting sinusoidal gratings in circular apertures of different sizes to investigate this extraclassical surround modulation over time. For the full 2-s stimulus time course, three types of neurons were found: 1) 68% of the cells were “suppressive,” 2) 25% were “plateau” cells that showed response saturation with no suppression, and 3) the remaining 6% of cells were “facilitative.” Analysis of the response dynamics revealed that at response onset, activity of one-half of facilitative cells, 70% of plateau cells, and all suppressive cells is suppressed by the surround. However, over the next 20 –30 ms, surround modulation changes to stronger suppression for suppressive cells, substantial facilitation for facilitative cells, and weak facilitation for plateau cells. For all three cell types, these modulatory effects then stabilize between 100 and 200 ms from stimulus onset. Thus our findings illustrate two stages of surround modulation. Early modulation is mainly suppressive regardless of cell type and, because of rapid onset, may rely on feedforward mechanisms. Surround modulation that evolves later in time is not always suppressive, depending on cell type, and may be generated through different combinations of cortical circuits. Additional analysis of modulation throughout the cortical column suggests the possibility that the larger excitatory fields of facilitative cells, primarily found in infragranular layers, may contribute to the second stage of suppression through intracolumnar circuitry. area 17; center-surround; facilitation; feedback; feedforward; horizontal connections; infragranular; spatial summation; suppression
(V1) are typically composed of a classical receptive field (CRF) center and an extraclassical surround (Albright and Stoner 2002; Allman et al. 1985; Fitzpatrick 2000). Whereas cell responses are driven by center stimulation, the surround plays a more modulatory role that is thought to be important in figure ground segregation and in forming the building blocks of form perception (Seriès et al. 2003; Sillito et al. 1995; Zipser et al. 1996). Previous studies have suggested that the center responses of most V1 cells are suppressed by inclusion of surround stimuli (Jones et al. 2001; Sengpiel et al. 1997; Walker et al. 2000), with maximal suppressive modulation when the center and the surround have the same orientation and spatial frequency (Akasaki et al. 2002; DeAngelis et al. 1994; Knierim and Van Essen 1992; Levitt and Lund 1997; Walker et al. 1999). However, even under these optimal stimulus conditions, not all V1 cells are suppressed by surround stimuli. Instead, in about one-quarter of the cells, responses will saturate and exhibit a plateau-like tuning profile (DeAngelis et al. 1994;
NEURONS IN PRIMARY VISUAL CORTEX
Address for reprint requests and other correspondence: D. C. Lyon, Dept. of Anatomy and Neurobiology, School of Medicine, Univ. of California, Irvine, CA 92697-1275 (e-mail:
[email protected]). 1306
Sengpiel et al. 1997; Walker et al. 2000). In addition, there is a third, more rarely encountered type of V1 cell that shows neither suppression nor saturation of response to increasing stimulus size, but rather increases in firing rate (Cavanaugh et al. 2002a; Gilbert 1977; Li and Li 1994). Thus three types of V1 cells can be characterized, suppressive, plateau, and facilitative, on the basis of their responses to increasing stimulus size beyond their CRF. Results from previous studies examining the dynamics of surround suppression have helped unravel the underlying mechanisms of surround modulation for suppressive cells. For example, it has been shown that there is an early suppressive component (Alitto and Usrey 2008; Xing et al. 2005) that likely arises through relatively fast feedforward lateral geniculate nucleus (LGN) afferents that already exhibit extraclassical surround suppression (Alitto and Usrey 2008; Ozeki et al. 2004; Sceniak et al. 2006; Solomon et al. 2002; Webb et al. 2005a). In addition, there is a second component that develops later in time (Chen et al. 2005; Xing et al. 2005) and is likely to be mediated instead through cortical mechanisms, which propagate more slowly (Bringuier et al. 1999; Girard et al. 2001; Grinvald et al. 1994; Schwabe et al. 2006). Previous studies on surround dynamics, however, have not included analyses of nonsuppressive modulation. Therefore, in the present study we have investigated response dynamics of all three types of surround modulation found in V1 (suppressive, plateau, and facilitative) using drifting sinusoidal gratings of various radii. Our results reveal that there are two stages of surround modulation: 1) early surround modulation that is mainly suppressive regardless of cell type and may rely on feedforward mechanisms because of rapid onset, and 2) a secondary surround modulation that evolves later in time, can be suppressive or facilitative depending on the cell type, and may be generated through more slowly propagating cortical circuits. METHODS
Animal Preparation and Recording The experiments were performed in 13 adult cats weighing 2.4 –5.0 kg and of both sexes. All procedures were approved by the Institutional Animal Care and Use Committee of the University of California, Irvine. Animals were initially anesthetized with a mixture of ketamine (21 mg/kg im) and xylazine (3 mg/kg im). Anesthesia was maintained with isofluorane (0.2– 0.6%) in a 67:33 mixture of nitrous oxide and oxygen through artificial respiration. ECG, EEG, and expired CO2 were monitored throughout the entire experiment to ensure proper level of anesthesia. To prevent eye movements, neuromuscular blockade was induced with an initial bolus of vecuronium bromide (0.6 mg/ml iv) and maintained for the duration of the experiment with 0.15 mg/ml at a flow rate of 2.0 ml·kg⫺1·h⫺1 mixed with dexamethasone (0.5 mg·kg⫺1·h⫺1 iv) in a solution of 5% dex-
0022-3077/11 Copyright © 2011 the American Physiological Society
www.jn.org
EXTRACLASSICAL SURROUND DYNAMICS IN V1
trose and lactated Ringer. Pupils were dilated with 1% atropine sulfate solution, and the nictitating membranes were retracted with 2.5% phenylephrine hydrochloride. Zero-power, air-permeable contact lenses were fitted to each cornea, and 3-mm artificial pupils were placed in front of the eyes. A craniotomy was made above the dorsal surface of area 17 (V1). Extracellular recordings of single cells were made with epoxy-insulated tungsten microelectrodes (5–7 M⍀; FHC, Bowdoin, ME). Electrode penetrations were made perpendicular to the cortical surface, and cells were recorded at all cortical depths. Visually evoked action potentials of single neurons were amplified and isolated by an Xcell-3 four-channel amplifier (FHC) and then fed into a computer running EXPO software (courtesy of Peter Lennie). Spikes were saved at 1-ms resolution for further analysis using custom MATLAB software. Visual Stimuli The visual stimuli were generated by a G5 Mac computer with an ATI Radeon 9200 graphics card running EXPO software. Stimuli were displayed on a gamma-calibrated Viewsonic Graphics Series G225f CRT monitor with a mean luminance of 50 cd/m2 and at a viewing distance of 37 cm. The refresh rate was set at 100 Hz on the noninterlaced monitor. The locations of the optic disc and the area centralis for each eye were plotted daily using a fiber-optic light source (Pettigrew et al. 1979). Through visual stimulation of the dominant eye, the receptive field center of each isolated cell was approximated under mouse control with an ⬃4° diameter aperture of drifting square wave gratings at the approximate preferred orientation. Maximal response was estimated audibly. The exact position of the receptive field center was then determined by reducing aperture size to as small as 1°. Spatial eccentricities of all receptive fields in our experiment were constrained within 2° above and 16° below the area centralis. Once the receptive field center was established, the CRF properties were tested in detail in the following sequence: First, we reassessed the preferred orientation of each neuron by varying the tilt of the drifting sinusoidal gratings in 22.5° increments. Gratings were shown at 100% contrast with spatial and temporal frequencies of 0.2 cycle/° and 4 Hz, respectively. Next, optimal spatial frequency was determined by presenting ⬃10° gratings at the preferred orientation while varying spatial period. It is worth noting that although this aperture size did not always match the optimal receptive field size of the cell (see below), it should nevertheless not affect spatial frequency tuning, since this aperture size was more than twice the diameter of the spatial frequency periods used (Mazer et al. 2002). Preferred temporal frequency was then assessed using gratings with the preferred orientation and spatial frequency. The starting phase of all drifting gratings was set to 0°. Finally, the optimal contrast for each cell was determined through contrast-response profiles (10 –100%, in 10% steps). Once the optimal parameters were determined, we then performed spatial summation measurements (aperture tuning test) using circular apertures of drifting sinusoidal gratings of various diameters (0.2–30°; see Fig. 1). The gratings were presented at optimal contrast at the preferred orientation, spatial and temporal frequencies, and in pseudorandom order. Each grating drifted for 2 s, followed by a 6-s interstimulus interval during which the animal viewed a blank screen presented at the mean luminance. This was repeated two to four times. For cells that showed a monotonic increase in response up to the largest aperture size (30°), a series of annuli with inner diameters ranging from 0.2–20° were presented to determine whether a CRF region was present within the large excitatory response field. Data Analysis Aperture tuning curve fitting. Aperture tuning curves were fitted by a difference of Gaussian (DoG) function J Neurophysiol • VOL
Rs ⫽ Ke
兰
s
e共⫺x ⁄ re兲2dx ⫺ Ki
⫺s
1307
兰
s
⫺s
e共⫺x ⁄ ri兲2dx ⫹ R0
(1)
in which Rs is the response evoked by different aperture sizes. The free parameters, Ke and re, describe the strength and the size of the excitatory space, respectively; Ki and ri represent the strength and the size of the inhibitory space, respectively; and R0 is the spontaneous activity of the cell. To evaluate how well our experimental data was explained by the model, we calculated the coefficient of determination (R2) using
冋
R2 ⫽ 1 ⫺
⁄
兺Nj⫽1 共Rn ⫺ Fn兲2 兺
N j⫽1
共 R n ⫺ R 0兲 2
册
(2)
where Rn is the responses to the jth stimulus, Fn is the predicted value, and R0 is the mean response of the actual data (Freeman et al. 2002). Across the population, the mean R2 for 2-s stimuli is 0.86 ⫾ 0.09. Definition of modulatory cell types and their CRF. In our study, cells with suppressive, plateau, and facilitative extraclassical surrounds were defined on the basis of their tuning profiles to a series of gratings drifting within circular apertures of varying diameter. Each aperture was presented for 2 s, and classification of the three modulatory cell types is based on this full time course. Suppressive cells were defined as those showing a response attenuation ⱖ5% to stimuli beyond their CRF (see examples in Figs. 1, A and B, and 2A). The CRF of suppressive cells was determined by parameter re from Eq. 1 (mean re ⫽ 9.3 ⫾ 4.6°), which was similar to the aperture size that evoked the maximum response (mean 9.6 ⫾ 3.9°; t-test, P ⫽ 0.14; Sceniak et al. 1999). Plateau cells were defined as those that either showed ⬍5% suppression or a saturation of response to aperture sizes larger than the CRF (see examples in Figs. 1C and 2B). As described above for suppressive cells, the CRF size of plateau cells was also determined by the parameter re from Eq. 1, consistent with earlier reports describing plateau cells (DeAngelis et al. 1994; Walker et al. 2000). Facilitative cells were defined as those that exhibited no saturation, but instead showed a monotonic response increase to increases in size of the stimulus aperture (see examples in Figs. 1D and 2C). For such cells, the size of the CRF (center) was determined by locating the annular minimum response field (5% above background; Cavanaugh et al. 2002a) to a series of annuli with variable inner diameters and a fixed 30° outer diameter (see gray trace in Fig. 2C). Thus, although we found that these cells always responded optimally to the largest aperture presented (30°), the CRF of these facilitative cells was much smaller, ranging from 4 to 10° in diameter (see arrow in Fig. 2C). Importantly, if a stimulus was not presented to this small CRF region, the facilitative cell was unresponsive to large stimuli. Overall, the average CRF (7.0 ⫾ 2.5°) was comparable to, although somewhat smaller than, the excitatory space constant derived from Eq. 1 (mean re ⫽ 10.0 ⫾ 3.4°). Simple/complex classification. The mean firing rate and first harmonic components of the accumulated response were computed for each stimulus. V1 neurons were classified as simple or complex by comparing the ratio of the first harmonic to the mean response to drifting grating stimuli (Skottun et al. 1991). Modulation index. To quantify the extraclassical surround modulation of all three modulatory cell types, we calculated a modulation index (MI) using MI ⫽ 共RECS ⫺ RCRF兲 ⁄ max共RECS, RCRF兲
(3)
where RECS is the response for maximum aperture size and RCRF is the response for the CRF. A negative MI indicates suppression, whereas a positive MI indicates facilitation. Temporal dynamics of the surround modulation. To determine the time course of extraclassical surround modulation, we fitted aperture tuning curves at successive time points using the DoG model in Eq. 1 and calculated MI values for each using Eq. 3. Cumulative time windows were created in steps of 10 ms for the first 200 ms from stimulus onset
105 • MARCH 2011 •
www.jn.org
1308
EXTRACLASSICAL SURROUND DYNAMICS IN V1
Fig. 1. Response time course of V1 cells to variations in stimulus diameter. A: suppressive simple cell. B: suppressive complex cell. C: plateau complex cell. D: facilitative simple cell. Plots at left show the fitted aperture tuning curve for each cell. Histograms of the cell’s response time to apertures corresponding to the black points in plots at left are shown at right. To explore the time course of response, firing rates were calculated in cumulative windows schematically illustrated by t1, t2, to tn with steps of 10 or 50 ms from stimulus onset (see METHODS for details). For some simple cells, such as that shown in D, response onset was delayed because stimulus phase was nonoptimal (phase was set to 0° for all cells). For these cells, response onset was shifted to be aligned with complex cells (see METHODS for details). Error bars are SE.
and in steps of 50 ms for the remainder of the 2-s stimulus time period. For a schematic of the cumulative time windows used for each of the aperture tuning curves, see Fig. 1. Figure 1, A and B, represent a simple suppressive cell and a complex suppressive cell, respectively; Fig. 1C represents a complex plateau cell; and Fig. 1D represents a simple facilitative cell. The reason for using cumulative time windows as opposed to sequential windows of equal size was to improve the signal-tonoise ratio and allow for more reliable aperture fits at each 10-ms bin. For some cells, the use of drifting sine wave gratings led to little or no response in several 10-ms bins (see Fig. 1, A and B). This was especially true for some simple cells that preferred lower temporal frequencies, where several consecutive 10-ms bins would be unresponsive (see example in Fig. 1D). Although the use of cumulative time windows greatly improves fits of the aperture tuning curves, it also sacrifices the precision at which changes in a cell’s response can be pinpointed in time (see DISCUSSION). J Neurophysiol • VOL
Response onset latency of simple cells depends on the initial phase of the stimulus (Movshon et al. 1978; Skottun et al. 1991; Spitzer and Hochstein 1985). The starting phase of our stimuli was always 0°, which was not optimal for a number of our simple cells. As a result, 63% (25/40) of them initiated firing substantially later, especially simple cells preferring low temporal frequencies (see example in Fig. 1D). To compare results of all simple and complex cells, we realigned the onset latency of these delayed simple cells (14/25 suppressive cells, 6/7 plateau cells, 5/8 facilitative cells) to the average onset latency of our complex cells. For this purpose, onset latency was defined as the earliest time that responses to the optimal aperture size (CRF) reached 25% of maximum; the mean latency of our complex cells was ⬃47 ms. Therefore, for each simple cell, the earliest time point that simple cell responses reached 25% of maximum was aligned to 47 ms (see schematic of the cumulative time windows shown in Fig. 1D). Orientation tuning properties. Responses to drifting gratings varying in orientation (presented at the optimal aperture size and the
105 • MARCH 2011 •
www.jn.org
EXTRACLASSICAL SURROUND DYNAMICS IN V1
preferred spatial and temporal frequencies) were plotted. The orientation tuning curve was fitted using 2 2 Ros ⫽ Ro ⫹ Rpe⫺共Os ⫺ Op兲2 ⁄ 共2 兲 ⫹ Rne⫺共Os ⫺ Op ⫹ 180兲2 ⁄ 共2 兲 (4)
in which Os is the stimulus orientation, Op is the preferred orientation, ROs is the response to different orientations, Rp and Rn correspond to the preferred and nonpreferred orientation response, respectively, Ro is the spontaneous response, and is the tuning width. The narrow-
1309
ness of the orientation tuning curve was measured as the half width at half height (HWHH), which equals 1.18. The orientation selectivity index (OSI), a measure of circular variance, was calculated using OSI ⫽
ⱍ兺 n
Rnexp共in兲
ⱍ 共兺 ⱍ ⁄
n
Rn
ⱍ兲
(5)
where n is the nth orientation of the stimulus and Rn is the corresponding response. Significance tests. Tests for significance were done using one-tailed t-tests or one-way ANOVA. RESULTS
We recorded from 158 neurons in V1. For each cell, aperture tuning curves were plotted for responses to aperture gratings with increasing diameters, shown drifting for a duration of 2 s (see examples in Figs. 1 and 2). Each curve was then fitted by a DoG function (Eq. 1). Three Types of Extraclassical Surround Modulation On the basis of fitted aperture tuning curves, cells could be classified into three types: suppressive, plateau, and facilitative (Fig. 2). The majority of our cells (108/158, 68%) were suppressive, for which responses decreased 5% or greater to aperture diameters beyond optimal. Figure 2A shows a representative suppressive cell. For this cell, the CRF (parameter re from Eq. 1) was 9° (95 spikes/s). The strongest suppression beyond the CRF occurred with the largest aperture diameter presented (30°), reducing spike rate 34% (63 spikes/s). Effects of the extraclassical surround were not always suppressive. Cells that exhibited no suppression to increasing aperture diameters beyond their CRF were classified into two groups, the more common plateau cells (40/158, 25%) and the rarely encountered facilitative cells (10/158, 6%). Figure 2B shows an example of a plateau cell. In this cell, responses saturated around 10°, and no suppression was observed when larger apertures were presented. This saturation may be caused by a balance between excitatory and inhibitory inputs (Sengpiel et al. 1997). For these plateau cells, their CRF can be represented by the excitatory space constant re in Eq. 1, as described in previous work (DeAngelis et al. 1994; Walker et al. 2000). Figure 2C shows an example facilitative cell, responses of which continuously increased with increasing aperture size. Importantly, for facilitative cells, we also tested responses to surround annuli and found that increasing the inner diameter led to a marked decrease in firing rate (see gray trace in Fig.
Fig. 2. Three types of extraclassical surround modulation exemplified by different V1 cells. Responses for each cell are plotted against the diameter of stimulus aperture. A: a suppressive cell shows peak response at 10°, which is close to the cell’s classical receptive field (CRF; size of the excitatory space re ⫽ 9°, indicated by arrow), and then rapidly suppresses in response to further increases in aperture size. B: an example plateau cell shows maximum response to apertures of 10° and larger. Thus the response saturates but is not suppressed by stimuli larger than 10°. The CRF (10°; indicated by arrow) for plateau cells is represented by parameter re extracted from Eq. 1. C: for the example facilitative cell, the response increases with increase in aperture size and does not suppress or saturate (black trace). To identify the CRF, the annular minimum response field was determined through a series of annuli of varying inner diameters (gray trace). The cell has a CRF of 10° (indicated by arrow). The modulation index (MI) is also plotted for each type of cell and equals ⫺0.34 for the suppressive cell in A, ⫹0.07 for the plateau cell in B, and ⫹0.62 for the facilitative cell in C. Error bars are SE. J Neurophysiol • VOL
105 • MARCH 2011 •
www.jn.org
1310
EXTRACLASSICAL SURROUND DYNAMICS IN V1
2C). This means that stimulation of the receptive field center is necessary for the surround to have a facilitative effect. In this way, facilitative cells, like suppressive cells, can also be considered to have a CRF. The inner diameter at which annulus responses dropped to within 5% of the mean background firing rate (the annulus minimum response field; Cavanaugh et al. 2002a) was used to define the CRF of these facilitative cells. For the example shown in Fig. 2C, this corresponded to 10° (indicated by arrow) where the cell response was 10 spikes/s. Center response was then facilitated 62% to 27 spikes/s by presenting the full 30° aperture. Simple and complex classification of modulatory cell types. In addition to categorizing each neuron based on the type of extraclassical surround modulation, we also classified them as simple or complex by plotting the first harmonic (F1) against the mean response (F0) to the drifting grating stimuli. The results are illustrated in Fig. 3, where cells below the diagonal (F1/F0 ⬎ 1) are classified as simple cells. With the use of this criteria, only 40 cells of our entire population were simple cells (25%), whereas the majority were complex cells (n ⫽ 118; 75%). Likewise, most of our suppressive (83/108, 77%) and plateau cells (33/40, 83%) were complex as well (open black and filled gray circles located above the diagonal in Fig. 3). Conversely, there was a dramatic difference in the classification of our facilitative cells, because 8 of 10 were simple cells (open black circles falling below the diagonal). Dynamics of V1 Size Tuning To determine the time course of extraclassical surround modulation in V1, we segregated responses to different-sized apertures into a series of cumulative time windows (see Fig. 1). For each cell, we plotted size tuning curves for each window and fitted them using Eq. 1.
Fig. 3. Simple and complex classification of the 3 modulatory cell types. Mean response (F0) to drifting gratings is plotted as a function of the first harmonic response (F1) for each cell. The majority of our neurons (118/158, 75%) fall above the diagonal (F1/F0 ⬍ 1) and are considered complex cells based on the criteria of Skottun et al. (1991). Conversely, the remaining 40 neurons fall below the diagonal (F1/F0 ⬎ 1) and are regarded as simple cells. The majority of suppressive (83/108; gray open circles) and plateau cells (33/40; gray filled circles) are complex cells, whereas 8 of 10 facilitative cells (black open circles) are simple cells. J Neurophysiol • VOL
Figure 4 shows several examples of size tuning curves over time for the different modulatory cell types. For suppressive (Fig. 4, A–C) and plateau cells (Fig. 4, D–F), three examples each are shown, whereas two facilitative cell examples (Fig. 4, G and H) are shown. In addition, Fig. 4I shows the dynamics of annulus response profiles for a facilitative cell. In Fig. 4, for each neuron, the blue curves represent earlier time courses, beginning as early as 30 ms from stimulus onset. Progressively warmer colored curves represent cumulative responses at later time windows up to 2 s (red traces). As a reminder, the classification of each cell as suppressive, plateau, or facilitative is based on the overall response curves to the full 2-s stimulus presentation. All but one of these examples (see description of Fig. 4H below) illustrate that extraclassical surround modulation of responses earlier in time was suppressive regardless of cell type and evolved later into stronger suppression for suppressive cells, no suppression for plateau cells, and facilitation for facilitative cells. A prime example of the evolution of increasing suppression strength for suppressive cells is illustrated in the traces of the complex cell shown in Fig. 4A. For this cell, the tuning curve at 40 ms (blue trace indicated by arrow) shows a CRF response at 14 spikes/s that was suppressed 64% (MI ⫽ ⫺0.64, as determined using Eq. 3) by the largest aperture (30°) to 5 spikes/s. However, at 120 ms (green trace indicated by arrow), the CRF response was 58 spikes/s and responses were suppressed nearly 80%, down to 13 spikes/s, by the largest aperture. Similar increases in suppression strength over time are shown for the two other suppressive cells in Fig. 4, B and C, with the latter of the two being a simple cell example. The evolution of modulation over time in our plateau cells from suppression to no suppression is best illustrated by the complex cell shown in Fig. 4D. For this cell, from 40 to 70 ms, suppression was quite strong. For example, the response at 40 ms (blue trace indicated by arrow) was 14 spikes/s for the CRF aperture and dropped to only 2 spikes/s for the largest aperture size, yielding 86% suppression (MI ⫽ ⫺0.86). However, by 90 ms and beyond, suppression was minimal such that the responses beyond optimal at 150 ms (green trace indicated by arrow) were suppressed only 3% (MI ⫽ ⫺0.03): ⬃31 spikes/s at CRF, down to 30 spikes/s at 30°. At even later cumulative time windows, no suppression was evident. Weakening of suppression strength over time is shown in the other two examples as well; one of these is another complex cell (Fig. 4E) and the other a simple cell (Fig. 4F). For facilitative cells, one-half (5/10) showed relatively strong suppression early in time followed later by facilitation. For the example complex cell shown in Fig. 4G, the response peaked for a relatively small-sized aperture at 50 ms (blue trace indicated by arrow) and was then suppressed 56% by apertures of 30°. By 120 ms (green trace indicated by arrow), peak response occurred for a larger stimulus and suppression was no longer evident. At most of the subsequent time points, the cell response increased as the size of the stimulus increases, showing neither suppression nor saturation. In some of the facilitative cells (5/10), on the other hand, suppression was not evident at early time points, but rather weak facilitation was observed, such as the blue 50-ms trace shown for the example facilitative cell in Fig. 4H. Nevertheless, as indicated by the green 90-ms trace in Fig. 4H, facilitation also became stronger later in time for these cells.
105 • MARCH 2011 •
www.jn.org
EXTRACLASSICAL SURROUND DYNAMICS IN V1
1311
Fig. 4. Temporal dynamics of size tuning curves for the 3 modulatory cell types. A–C: 3 suppressive cell examples. D–F: 3 plateau cell examples. G and H: 2 facilitative cell examples. Aperture tuning curves in each cumulative time window are represented as a collection of colored curves for each cell. Blue curves represent responses at earlier time points relative to stimulus onset. Warmer colored curves represent the cumulative responses at later time windows up to the full 2-s stimulus presentation, represented by red curves. Time window (ms) and R2 of the fits for each tuning curve are shown at right. Arrows in A–H indicate aperture tuning curves described in RESULTS. I: annulus response of a facilitative cell at different time windows. As indicated by the arrow, the annular minimum response field is consistent across time (8°). A, B, D, E, and G show complex cell examples; C, F, H, and I show simple cells.
Figure 4I shows the response profiles over time for annulus testing of an example facilitative cell. In this example, the minimum annulus response field used to define the CRF was very consistent, 8°, across all time points (see arrow in Fig. 4I). This invariance of CRF size over time was found for the population of facilitative cells as well (data not shown). Dynamics of the Extraclassical Surround Modulation We next examined detailed dynamics of extraclassical surround modulation for the three cell populations. To do this, we computed a MI for each cumulative time window of each cell (see examples in Dynamics of V1 Size Tuning). Actual time windows used for this analysis were in steps of 10 ms for the first 200 ms after stimulus onset and then in steps of 50 ms for the remaining time (see x-axis in Fig. 5). The resulting evolution of extraclassical surround modulation (MI) over time for the populations is plotted in Fig. 5. The black trace represents facilitative cells, the dark gray trace represents plateau cells, and the light gray trace represents suppressive cells. Consistent with the individual examples shown in Fig. 4, for the three populations, all cell types showed suppression early, even facilitative cells (mean MI ⬍ 0; at response onset, ⬃30 ms from stimulus onset in Fig. 5A). After this initial period, modulation evolved into facilitation (MI ⬎ 0) for facilitative cells 20 ms later (at ⬃50 ms from stimulus onset). Similarly, the MI of plateau J Neurophysiol • VOL
cells crossed from suppression into facilitation ⬃100 ms after stimulus onset. These evolving modulatory effects then stabilized by ⬃200 ms for facilitative and plateau cells. For suppressive cells, after the first 30 ms from response onset, surround suppression became even stronger as indicated by an increase in negative MI beginning at the transition from 60 to 70 ms from stimulus onset (see light gray trace in Fig. 5A). Suppression then hit maximum strength at ⬃100 ms from stimulus onset, which is mostly consistent with the maximum suppression latency shown for monkey V1 neurons (Bair et al. 2003). Thus, from our population results, we conclude that two types of extraclassical surround modulation are present over time: an early component that is suppressive even for facilitative and plateau cells, and a later component that is either suppressive or facilitative depending on the cell type. The early suppressive modulation may rely on feedforward mechanisms because of its rapid onset, and the surround modulation that evolves later in time may be generated through different combinations of more slowly propagating cortical circuits. Receptive Field Size of Different Cell Types Across Cortical Depth The dynamics of extraclassical surround modulation that we have observed thus far suggest the presence of suppression early in time, followed by either facilitation or increased
105 • MARCH 2011 •
www.jn.org
1312
EXTRACLASSICAL SURROUND DYNAMICS IN V1
For each cell, we plotted the aperture size of the maximum response within their respective columns relative to cortical depth (see Fig. 6A). Cells from the same electrode penetration are plotted along the same dashed vertical line. The size of each circle is shown proportionally, with the largest apertures representing 30° of visual angle. As is evident in Fig. 6A, all but one of the facilitative cells (black circles) were located in cortical depths corresponding to the infragranular layers of V1 (⬎1,200 m). On the other hand, suppressive (light gray filled circles) and plateau cells (gray open circles) were distributed evenly across all cortical layers. Furthermore, as expected, facilitative cells required larger aperture sizes for maximum response than the other two cell types (see Fig. 6B). Average optimal aperture sizes for suppressive (9.3 ⫾ 3.9°) and plateau cells (10.1 ⫾ 4.1°) were similar and not significantly different (P ⫽ 0.30), whereas aperture sizes for facilitative cells were
Fig. 5. Mean temporal evolution of the extraclassical surround modulation for the 3 modulatory cell types. A: the MI for suppressive (light gray trace), plateau (dark gray trace), and facilitative cells (black trace) are plotted against time. Each time point is cumulative. A negative MI represents suppressive modulation, and a positive MI represents facilitation. The value 0 on the x-axis represents stimulus onset time. B: average R2 of the fitted curves are plotted as a function of each cumulative time window. Error bars are SE.
suppression depending on the cell type. The early onset of suppression is consistent with a feedforward mechanism from LGN and/or retina (Alitto and Usrey 2008; Nolt et al. 2004; Sceniak et al. 2006), whereas later modulation may be generated through different combinations of cortical circuits. In addition to the more commonly considered cortical sources for extraclassical surround modulation, feedback and longrange lateral connections (Angelucci et al. 2002; Gilbert 1992), another source is through deeper layer neurons within the same cortical column (Allison and Bonds 1994; Bolz and Gilbert 1986). To address this issue, we plotted the diameter of the aperture size eliciting a maximum response for each cell and sorted by cortical depth encountered via the electrode penetration (as shown in Fig. 6A). For facilitative cells, the optimal aperture was always the maximum stimulus diameter (30°). For suppressive cells, the optimal aperture was the aperture that corresponded to the peak response. For plateau cells, the optimal aperture was defined as the point of saturation (size corresponding to 95% of the maximum response). J Neurophysiol • VOL
Fig. 6. Receptive field size of maximum response of different modulatory cell types as a function of cortical depth. A: receptive fields of maximum response (illustrated by circle size) within each of 27 electrode penetrations relative to cortical depth (m). Circle sizes range from 2 to 30° and are shown to scale (see right). Cells located in the same penetration are plotted along the same dashed vertical line. Light gray filled circles are suppressive cells, gray open circles are plateau cells, and black open circles are facilitative cells. All but 1 of the facilitative cells are located ⬎1,200 m below the cortical surface in what is considered the infragranular layer, whereas suppressive and plateau cells are distributed broadly across all layers (⬍600 m approximates the location of superficial layers, 600 –1,200 m approximates the position of the granular layer, and ⬎1,200 m approximates the position of infragranular layers). B: distributions of receptive fields of maximum response for the 3 cell types.
105 • MARCH 2011 •
www.jn.org
EXTRACLASSICAL SURROUND DYNAMICS IN V1
1313
significantly larger than those for suppressive (P ⬍ 0.001) and plateau cells (P ⬍ 0.001). It has been proposed that the surround temporal dynamics and visual space covered by the full extraclassical receptive field can only be accounted for through feedback from higher order visual areas where individual neurons have large enough receptive fields with faster propagation speeds than multisynaptic long-range lateral connections (Bullier et al. 2001; Cantone et al. 2005; Cavanaugh et al. 2002a; Schwabe et al. 2006). However, the suppressive surround could arise from local circuits. The large excitatory receptive fields of the deep layer intrinsic facilitative cells may provide an alternative source for the suppressive surround of V1 cells, since the sizes of these receptive fields were comparable to receptive field sizes of neurons found in higher visual areas (see DISCUSSION). Spatial Frequency and Orientation Tuning of Suppressive, Plateau, and Facilitative Cells Spatial frequency and orientation tuning are related to extraclassical surround modulation in that the maximal suppression occurs when the center and surround are presented with the same orientation and spatial frequency (Akasaki et al. 2002; DeAngelis et al. 1994; Knierim and Van Essen 1992; Levitt and Lund 1997; Walker et al. 1999). We analyzed spatial frequency and orientation properties for all three cell types. Figure 7 shows examples for cells located in the same electrode track (track 3 from Fig. 6A), including four suppressive cells, two facilitative cells, and two plateau cells. As shown in Fig. 7A, the preferred spatial frequency was comparable regardless of modulatory cell type. Likewise, the preferred orientation varied by no more than 22.5° from cell to cell (see Fig. 7B). Furthermore, although orientation tuning width did vary, broad and narrow band widths were observed for all three cell types. Across the population, the spatial frequency distribution shown in Fig. 8A was similar for all three types of modulatory cells (ANOVA, P ⫽ 0.28). As illustrated in Fig. 8B, the orientation tuning width (HWHH) was linearly related to the OSI of the cells (R ⫽ 0.54, P ⬍ 0.001). Among the three types of modulatory cells, the HWHH was not significantly different (ANOVA, P ⫽ 0.14). On the other hand, the OSI of facilitative cells was significantly higher than that of suppressive and plateau cells (ANOVA, P ⫽ 0.02).This is perhaps due to most of the facilitative cells being simple cells, and the simple cells had higher average OSI values than our complex cells (Fig. 8D) (Heggelund and Albus, 1978; Leventhal and Hirsch, 1978; Ringach et al. 2002; Schummers et al. 2007). Likewise, the HWHH was generally smaller for simple cells, and the results are statistically significant for suppressive and plateau cells (P ⫽ 0.002 and P ⫽ 0.003, respectively; Fig. 8C). DISCUSSION
We found that three different kinds of extraclassical surround modulation exist depending on the type of V1 cell: suppressive, plateau, or facilitative. In addition, our response dynamic analyses showed that these types of surround modulations are not stationary over time. Shortly after stimulus onset, surround modulation is suppressive for all three types of cells. However, this modulation evolves into facilitation for facilitative and plateau cells. On the other hand, suppression does not turn into facilitation at later time points for suppresJ Neurophysiol • VOL
Fig. 7. Spatial frequency and orientation tuning curve examples for cells within the same penetration. Cells from this single penetration (track 3 in Fig. 6) include 4 suppressive, 2 plateau, and 2 facilitative cells. A: the preferred spatial frequencies are similar for each cell. B: orientation tuning curves vary but are comparable across the 3 modulatory cell types. In addition, preferred orientations are similar for all the cells.
105 • MARCH 2011 •
www.jn.org
1314
EXTRACLASSICAL SURROUND DYNAMICS IN V1
sive cells, but rather increases in suppressive strength. In this report we discuss our results in light of previous work and consider how they relate to possible underlying mechanisms involved in the evolution of different types of surround modulation over time. Three Types of Extraclassical Surround Modulation
Fig. 8. Spatial frequency preference and orientation selectivity distributions for the 3 modulatory cell types. A: the distributions of spatial frequency preference for suppressive, plateau, and facilitative cells are similar (ANOVA, P ⫽ 0.28). B: the orientation selectivity index (OSI) for all cells is plotted against the orientation tuning width [half width at half height; (HWHH), shown in radians], and they are linearly related (R ⫽ 0.54, P ⬍ 0.001). In addition, the OSI of facilitative cells (which are predominantly simple cells) is significantly larger than that of suppressive and plateau cells (ANOVA, P ⫽ 0.02). HWHH is similar for the 3 types of cells (ANOVA, P ⫽ 0.14). C and D: mean HWHH and OSI, respectively, for the 3 modulatory cell types. Across the population, orientation selectivity and tuning of simple cells (solid bars) is better than that of complex cells (hatched bars). For suppressive cells, these simple/complex differences are significant for both HWHH and OSI, whereas for plateau cells this difference is significant for only the HWHH. For facilitative cells, simple/complex cell differences in HWHH and OSI are not significant, probably due to only having 2 complex cells in the sample.
J Neurophysiol • VOL
Stimuli extending beyond the CRF of a V1 neuron will often lead to suppression of the cell’s excitatory center (Allman et al. 1985). It has been well documented that the strength of suppression induced by extraclassical stimuli depends on the relationship between the stimulus represented in the center and in the surround. For example, stronger suppression can be obtained when stimulus properties such as orientation, direction, and spatial frequency of the center and the surround match, even when these parameters are not optimal (Akasaki et al. 2002; Cavanaugh et al. 2002b; DeAngelis et al. 1994; Knierim and Van Essen 1992; Levitt and Lund 1997; Walker et al. 1999). The stimuli we employed (full contrast gratings that varied in diameter but always drifted in the preferred orientation and temporal and spatial frequencies) were most likely to lead to maximum suppression. Accordingly, we found that 68% of our V1 cells were suppressed by an extraclassical surround, which is consistent with previous work in both cat and monkey (Jones et al. 2001; Sengpiel et al. 1997; Walker et al. 2000). Although the majority of our cells were suppressive, 32% (50/158) did not show suppression, since they either reached a plateau in their excitatory response (plateau cells, n ⫽ 40) or continued to increase their firing rate with increasing stimulus size (facilitative cells, n ⫽ 10). Plateau and facilitative cells have been reported several times before (Cavanaugh et al. 2002a; DeAngelis et al. 1994; Kapadia et al. 1999; Knierim and Van Essen 1992; Li and Li 1994; Polat et al. 1998; Sengpiel et al. 1997; Walker et al. 2000). As such, it has been postulated that responses of plateau cells may be generated by a balance between excitatory and inhibitory inputs (Sengpiel et al. 1997). Responses of our facilitative cells increased with aperture size and showed no saturation. In our entire experiment, facilitative cells were rarely encountered (10/158, 6%), consistent with the idea that accurate estimation of the receptive field center primarily yields a high incidence of nonfacilitative cells (Cavanaugh et al. 2002a; Fitzpatrick 2000; Walker et al. 2000). Interestingly, most of our facilitative cells were located in infragranular layers of V1 (⬎1,200 m), and a larger proportion were simple cells (8 of 10). This is consistent with earlier work showing that a majority of neurons in layer 6 of V1 are simple and that some have very large excitatory receptive fields (Bolz and Gilbert 1989; Gilbert 1977; Leventhal and Hirsch 1978). The novel observation in our present work is that we also found deep layer facilitative cells to have a relatively small CRF, ranging from 4 –10°, and a facilitative extraclassical surround of at least 30°. Similar results using narrow bars with drifting gratings have been reported by Li and Li (1994) for cat V1, although they were not attributed specifically to deep layer simple cells. More recently, Haider et al. (2010) found that fast-spiking interneurons and regular/thin-spiking pyramidal neurons in cat V1 increased responses to larger natural stimuli that exceeded the CRF. This subpopulation of GABAergic and
105 • MARCH 2011 •
www.jn.org
EXTRACLASSICAL SURROUND DYNAMICS IN V1
pyramidal cells is consistent with the small population of facilitative cells found in our experiment and indicates that our facilitative cells could be either inhibitory or excitatory. Furthermore, 50% of the facilitative cells in our sample showed suppression early in time, followed later by facilitation, whereas the other 50% of cells showed no evidence of suppression early in time. It could be possible that the inhibitory and excitatory cells with large receptive fields described by Haider et al. (2010) correspond to the two different suppression dynamics we observed. Mechanisms Underlying Extraclassical Surround Modulation Across the great majority of the neurons we recorded, extraclassical surround modulation was suppressive at early time points (at response onset) but later diverged into either facilitation or even stronger suppression (Fig. 5). A feedforward mechanism from the LGN may account for this early universal suppression, since LGN neurons already exhibit a suppressive extraclassical surround (Alitto and Usrey 2008; Solomon et al. 2002; Webb et al. 2005b). Furthermore, a feedforward model without lateral inhibition, used to explain suppression for simple cells, could also contribute to the early suppression (Finn et al. 2007). A third possibility is fast suppression arising from nearby cells within V1, especially for cells where suppression comes even before the excitatory input to the CRF reaches spike threshold, as shown in monkeys (Bair et al. 2003). In monkey, faster feedforward propagation of the magnocellular compared with the parvocellular pathway (Alitto and Usrey 2008; Lyon et al. 2010; Schiller and Malpeli 1978; Vidyasagar et al. 2002) may account for such an effect. Likewise, in cat, parallel pathways of feedforward geniculocortical inputs from the slower X and faster Y channels (Cleland et al. 1971; Lennie 1980; Sur et al. 1987) can provide the biological basis for such a scenario. Although on average early modulation of all the V1 cell types was suppressive, later in time, surround modulations evolved into two types. That is, modulation became positive for facilitative and plateau cells, and for suppressive cells modulation became even more strongly suppressive. These later modulations may be generated through different combinations of cortical circuits. One candidate is long-range lateral connections in V1, which tend to link cells with similar functional properties, such as orientation preference (Buzás et al. 2006; Callaway and Katz 1990; Gilbert and Wiesel 1989; Kisvarday et al. 1997), direction preference (Roerig and Kao 1999), and ocular dominance (Yoshioka et al. 1996). Longrange lateral connections have been demonstrated to account for extraclassical surround modulation, including facilitation and suppression (Mizobe et al. 2001; Polat et al. 1998; Sceniak et al. 2001). Importantly, the speed of horizontal connections covering the extraclassical surround is slow (⬃0.1– 0.2 mm/ ms; Bringuier et al. 1999; Girard et al. 2001; Grinvald et al. 1994), which may contribute to the later modulation of our results. For instance, as measured through cumulative time windows, the change to stronger suppression for suppressive cells and the transition from suppression to facilitation for facilitative and plateau cells occurs ⬃20 –30 ms after response onset (Fig. 5), which would correspond to long-range connections arising from cells located ⬃2–3 mm laterally. In addition, J Neurophysiol • VOL
1315
suppression strength that peaks ⬃70 ms after response onset for suppressive cells or facilitation that rises for another ⬃100 ms is indicative of even longer range lateral connections or possibly multisynaptic lateral circuits. An intriguing possibility for facilitative cells may be the involvement of Meynert cells, which like most of our facilitative cells are found in infragranular layers of cat and monkey V1 (i.e., Gabbott et al. 1987; Rockland and Knutson, 2001). Meynert cells are large pyramidal neurons that have been shown to have extremely longrange projecting axons, up to 8 mm (Rockland and Knutson, 2001), and could thus provide excitatory monosynaptic inputs to facilitative cells that arrive as late as 80 ms after response onset. As explained in METHODS, the reason for using cumulative time windows was to improve the signal-to-noise ratio and allow for more reliable fits at each 10-ms bin. However, it is important to note that the cumulative time point at which suppression strength increases for suppressive cells or when suppression switches to facilitation for the population of our facilitative cells (20 –30 ms as shown in Fig. 5) is not the exact point at which this signal begins to affect the cell’s response; rather, it is when this influence has outweighed the influence of an early suppressed response. Therefore, the true latencies for these components are likely to be modestly shorter, by some amount that is difficult to estimate. Another likely contributor to extraclassical surround modulation is feedback from extrastriate visual cortex, which can cover large regions of visual field corresponding to the extraclassical surround of V1 neurons (Angelucci et al. 2002; Bair et al. 2003; Bullier et al. 2001; Cantone et al. 2005; Cavanaugh et al. 2002a; Chen et al. 2005; Levitt and Lund 2002; Xing et al. 2005) and usually targets excitatory neurons (Johnson and Burkhalter 1996; Salin and Bullier 1995; Shao and Burkhalter 1996). In this way, feedback connections arising from a large region of the visual field may contribute to the larger facilitative surround modulation that occurs later in time for facilitative and plateau cells. Feedback may also contribute to the later increase in strength of suppression for suppressive cells via intrinsic V1 relays onto local inhibitory cells (Schwabe et al. 2006). Because feedback axons are myelinated, these inputs are likely to arrive similarly in time, if not sooner, to the typical ⬃2-mm long-range lateral connections described above (Angelucci et al. 2002; Bair et al. 2003; Schwabe et al. 2006). Therefore, the effects of feedback projections should occur within the first 20 ms beyond response onset (Xing et al. 2005). In our experiments this time frame corresponds to the change from suppression to facilitation for our facilitative cell population and the point where suppression strength begins to increase for our suppressive cells (Fig. 5). In addition to long-range lateral and feedback connections, another possibility for later extraclassical surround modulation is through connections within columns. Most of our facilitative cells were located in deeper layers of V1 and were simple cells, consistent with previous results showing that such cells are found in layer 6 (Gilbert 1977). End inhibition of layer 4 cells will disappear after inactivating layer 6 with GABA, indicating that layer 6 cells account, at least partly, for surround suppression of cells in layer 4 (Bolz and Gilbert 1986; Bolz et al. 1989). Moreover, the inhibitory effects mediated by infragranular layers could be propagated throughout the column (see Allison and Bonds 1994). In this regard, our dynamic
105 • MARCH 2011 •
www.jn.org
1316
EXTRACLASSICAL SURROUND DYNAMICS IN V1
surround modulation results (Fig. 5) show that the evolutions of stronger suppression and stronger facilitation occur over a similar time course. Probably the most noteworthy feature of these rarely encountered facilitative cells is that they have very large excitatory receptive fields, far bigger than the excitatory fields of other cells in V1, even within the same column (Fig. 6). Most critically, however, these large facilitative receptive fields match the diameter of the extraclassical suppressive surrounds for the more commonly encountered suppressive cells. As such, they are capable of contributing to the more distal surround effects. Because facilitation develops later in time and coincides with the onset and continued development of the stronger suppression observed for suppressive cells, we suggest that facilitative cells may play a key role, perhaps directly as GABAergic interneurons or through excitatory synapses onto inhibitory neurons within the same cortical column. Overall, our present results suggest that mechanisms underlying surround modulation combine a number of different sources, including feedforward, intracolumnar, long-range horizontal, and feedback connections, and that these sources differ depending not only on the cumulative time relative to stimulus onset but also on the type of V1 cell. ACKNOWLEDGMENTS We thank Emily Grossman for helpful comments on this manuscript. GRANTS This work was partially supported by the Whitehall Foundation. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s). REFERENCES Akasaki T, Sato H, Yoshimura Y, Ozeki H, Shimegi S. Suppressive effects of receptive field surround on neuronal activity in the cat primary visual cortex. Neurosci Res 43: 207–220, 2002. Albright TD, Stoner GR. Contextual influences on visual processing. Annu Rev Neurosci 25: 339 –379, 2002. Alitto HJ, Usrey WM. Origin and dynamics of extraclassical suppression in the lateral geniculate nucleus of the macaque monkey. Neuron 57: 135–146, 2008. Allison JD, Bonds AB. Inactivation of the infragranular striate cortex broadens orientation tuning of supragranular visual neurons in the cat. Exp Brain Res 101: 415– 426, 1994. Allman J, Miezin F, McGuinness E. Stimulus specific responses from beyond the classical receptive field: neurophysiological mechanisms for local-global comparisons in visual neurons. Annu Rev Neurosci 8: 407– 430, 1985. Angelucci A, Levitt JB, Walton EJS, Hupe JM, Bullier J, Lund JS. Circuits for local and global signal integration in primary visual cortex. J Neurosci 22: 8633– 8646, 2002. Bair W, Cavanaugh JR, Movshon JA. Time course and time-distance relationships for surround suppression in macaque V1 neurons. J Neurosci 23: 7690 –7701, 2003. Bolz J, Gilbert CD. Generation of end-inhibition in the visual cortex via interlaminar connections. Nature 320: 362–365, 1986. Bolz J, Gilbert CD. The role of horizontal connections in generating long receptive fields in the cat visual cortex. Eur J Neurosci 1: 263–268, 1989. Bolz J, Gilbert CD, Wiesel TN. Pharmacological analysis of cortical circuitry. Trends Neurosci 12: 292–296, 1989. Bringuier V, Chavane F, Glaeser L, Frégnac Y. Horizontal propagation of visual activity in the synaptic integration field of area 17 neurons. Science 283: 695– 699, 1999. J Neurophysiol • VOL
Bullier J, Hupé JM, James AC, Girard P, Casanova C, Ptito M. The role of feedback connections in shaping the responses of visual cortical neurons. In: Progress in Brain Research. Vision: From Neurons to Cognition, edited by Casanova C, Ptito M. Amsterdam, The Netherlands: Elsevier, 2001, vol. 134, chapt. 13, p. 193–695–204. Buzás P, Kovács K, Ferecskó AS, Budd JML, Eysel UT, Kisvárday ZF. Model-based analysis of excitatory lateral connections in the visual cortex. J Comp Neurol 499: 861– 881, 2006. Callaway E, Katz L. Emergence and refinement of clustered horizontal connections in cat striate cortex. J Neurosci 10: 1134 –1153, 1990. Cantone G, Xiao J, McFarlane N, Levitt JB. Feedback connections to ferret striate cortex: direct evidence for visuotopic convergence of feedback inputs. J Comp Neurol 487: 312–331, 2005. Cavanaugh JR, Bair W, Movshon JA. Nature and interaction of signals from the receptive field center and surround in macaque V1 neurons. J Neurophysiol 88: 2530 –2546, 2002a. Cavanaugh JR, Bair W, Movshon JA. Selectivity and spatial distribution of signals from the receptive field surround in macaque V1 neurons. J Neurophysiol 88: 2547–2556, 2002b. Chen G, Dan Y, Li CY. Stimulation of non-classical receptive field enhances orientation selectivity in the cat. J Physiol 564: 233–243, 2005. Cleland BG, Dubin MW, Levick WR. Sustained and transient neurones in the cat’s retina and lateral geniculate nucleus. J Physiol 217: 473– 496, 1971. DeAngelis GC, Freeman RD, Ohzawa I. Length and width tuning of neurons in the cat’s primary visual cortex. J Neurophysiol 71: 347–374, 1994. Finn IM, Priebe NJ, Ferster D. The emergence of contrast-invariant orientation tuning in simple cells of cat visual cortex. Neuron 54: 137–152, 2007. Fitzpatrick D. Seeing beyond the receptive field in primary visual cortex. Curr Opin Neurobiol 10: 438 – 443, 2000. Freeman TC, Durand S, Kiper DC, Carandini M. Suppression without Inhibition in visual cortex. Neuron 35: 759 –771, 2002. Gabbott PLA, Martin KAC, Whitteridge D. Connections between pyramidal neurons in layer 5 of cat visual cortex (area 17). J Comp Neurol 259: 364 –381, 1987. Gilbert C, Wiesel T. Columnar specificity of intrinsic horizontal and corticocortical connections in cat visual cortex. J Neurosci 9: 2432–2442, 1989. Gilbert CD. Horizontal integration and cortical dynamics. Neuron 9: 1–13, 1992. Gilbert CD. Laminar differences in receptive field properties of cells in cat primary visual cortex. J Physiol 268: 391– 421, 1977. Girard P, Hupe JM, Bullier J. Feedforward and feedback connections between areas V1 and V2 of the monkey have similar rapid conduction velocities. J Neurophysiol 85: 1328 –1331, 2001. Grinvald A, Lieke E, Frostig R, Hildesheim R. Cortical point-spread function and long-range lateral interactions revealed by real-time optical imaging of macaque monkey primary visual cortex. J Neurosci 14: 2545– 2568, 1994. Haider B, Krause MR, Duque A, Yu Y, Touryan J, Mazer JA, McCormick DA. Synaptic and network mechanisms of sparse and reliable visual cortical activity during nonclassical receptive field stimulation. Neuron 65: 107–121, 2010. Heggelund P, Albus K. Orientation selectivity of single cells in striate cortex of cat: the shape of orientation tuning curves. Vision Res 18: 1067–1071, 1978. Johnson RR, Burkhalter A. Microcircuitry of forward and feedback connections within rat visual cortex. J Comp Neurol 368: 383–398, 1996. Jones HE, Grieve KL, Wang W, Sillito AM. Surround suppression in primate V1. J Neurophysiol 86: 2011–2028, 2001. Kapadia MK, Westheimer G, Gilbert CD. Dynamics of spatial summation in primary visual cortex of alert monkeys. Proc Natl Acad Sci USA 96: 12073–12078, 1999. Kisvarday Z, Toth E, Rausch M, Eysel U. Orientation-specific relationship between populations of excitatory and inhibitory lateral connections in the visual cortex of the cat. Cereb Cortex 7: 605– 618, 1997. Knierim JJ, Van Essen DC. Neuronal responses to static texture patterns in area V1 of the alert macaque monkey. J Neurophysiol 67: 961–980, 1992. Lennie P. Parallel visual pathways: a review. Vision Res 20: 561–594, 1980. Leventhal AG, Hirsch HV. Receptive-field properties of neurons in different laminae of visual cortex of the cat. J Neurophysiol 41: 948 –962, 1978. Levitt JB, Lund JS. Contrast dependence of contextual effects in primate visual cortex. Nature 387: 73–76, 1997. Levitt JB, Lund JS. The spatial extent over which neurons in macaque striate cortex pool visual signals. Vis Neurosci 19: 439 – 452, 2002.
105 • MARCH 2011 •
www.jn.org
EXTRACLASSICAL SURROUND DYNAMICS IN V1 Li CY, Li W. Extensive integration field beyond the classical receptive field of cat’s striate cortical neurons-classification and tuning properties. Vis Neurosci 34: 2337–2355, 1994. Lyon DC, Nassi JJ, Callaway EM. A disynaptic relay from superior colliculus to dorsal stream visual cortex in macaque monkey. Neuron 65: 270 –279, 2010. Mazer JA, Vinje WE, McDermott J, Schiller PH, Gallant JL. Spatial frequency and orientation tuning dynamics in area V1. Proc Natl Acad Sci USA 99: 1645–1650, 2002. Mizobe K, Polat U, Pettet MW, Kasamatsu T. Facilitation and suppression of single striate-cell activity by spatially discrete pattern stimuli presented beyond the receptive field. Vis Neurosci 18: 377–391, 2001. Movshon JA, Thompson ID, Tolhurst DJ. Spatial summation in the receptive fields of simple cells in the cat’s striate cortex. J Physiol 283: 53–77, 1978. Nolt MJ, Kumbhani RD, Palmer LA. Contrast-dependent spatial summation in the lateral geniculate nucleus and retina of the cat. J Neurophysiol 92: 1708 –1717, 2004. Ozeki H, Sadakane O, Akasaki T, Naito T, Shimegi S, Sato H. Relationship between excitation and inhibition underlying size tuning and contextual response modulation in the cat primary visual cortex. J Neurosci 24: 1428 –1438, 2004. Pettigrew J, Cooper M, Blasdel G. Improved use of tapetal reflection for eye-position monitoring. Invest Ophthalmol Vis Sci 18: 490 – 495, 1979. Polat U, Mizobe K, Pettet MW, Kasamatsu T, Norcia AM. Collinear stimuli regulate visual responses depending on cell’s contrast threshold. Nature 391: 580 –584, 1998. Ringach DL, Shapley RM, Hawken MJ. Orientation selectivity in macaque V1: diversity and laminar dependence. J Neurosci 22: 5639 –5651, 2002. Rockland KS, Knutson T. Axon collaterals of meynert cells diverge over large portions of area V1 in the macaque monkey. J Comp Neurol 441: 134 –147, 2001. Roerig B, Kao JPY. Organization of intracortical circuits in relation to direction preference maps in ferret visual cortex. J Neurosci 19: 44RC-, 1999. Salin PA, Bullier J. Corticocortical connections in the visual system: structure and function. Physiol Rev 75: 107–154, 1995. Sceniak MP, Chatterjee S, Callaway EM. Visual spatial summation in macaque geniculocortical afferents. J Neurophysiol 96: 3474 –3484, 2006. Sceniak MP, Hawken MJ, Shapley R. Visual spatial characterization of macaque V1 neurons. J Neurophysiol 85: 1873–1887, 2001. Sceniak MP, Ringach DL, Hawken MJ, Shapley R. Contrast’s effect on spatial summation by macaque V1 neurons. Nat Neurosci 2: 733–739, 1999. Schiller PH, Malpeli JG. Functional specificity of lateral geniculate nucleus laminae of the rhesus monkey. J Neurophysiol 41: 788 –797, 1978. Schwabe L, Obermayer K, Angelucci A, Bressloff PC. The role of feedback in shaping the extra-classical receptive field of cortical neurons: a recurrent network model. J Neurosci 26: 9117–9129, 2006.
J Neurophysiol • VOL
1317
Schummers J, Cronin B, Wimmer K, Stimberg M, Martin R, Obermayer KH, Koerding K, Sur M. Dynamics of orientation tuning in cat V1 neurons depend on location within layers and orientation maps. Front Neurosci 1: 14, 2007. Sengpiel F, Sen A, Blakemore C. Characteristics of surround inhibition in cat area 17. Exp Brain Res 116: 216 –228, 1997. Seriès P, Lorenceau J, Frégnac Y. The “silent” surround of V1 receptive fields: theory and experiments. J Physiol (Paris) 97: 453– 474, 2003. Shao Z, Burkhalter A. Different balance of excitation and inhibition in forward and feedback circuits of rat visual cortex. J Neurosci 16: 7353– 7365, 1996. Sillito AM, Grieve KL, Jones HE, Cudeiro J, Davls J. Visual cortical mechanisms detecting focal orientation discontinuities. Nature 378: 492– 496, 1995. Skottun BC, De Valois RL, Grosof DH, Movshon JA, Albrecht DG, Bonds AB. Classifying simple and complex cells on the basis of response modulation. Vision Res 31: 1078 –1086, 1991. Solomon SG, White AJR, Martin PR. Extraclassical receptive field properties of parvocellular, magnocellular, and koniocellular cells in the primate lateral geniculate nucleus. J Neurosci 22: 338 –349, 2002. Spitzer H, Hochstein S. Simple- and complex-cell response dependences on stimulation parameters. J Neurophysiol 53: 1244 –1265, 1985. Sur M, Esguerra M, Garraghty PE, Kritzer MF, Sherman SM. Morphology of physiologically identified retinogeniculate X- and Y-axons in the cat. J Neurophysiol 58: 1–32, 1987. Vidyasagar TR, Kulikowski JJ, Lipnicki DM, Dreher B. Convergence of parvocellular and magnocellular information channels in the primary visual cortex of the macaque. Eur J Neurosci 16: 945–956, 2002. Walker GA, Ohzawa I, Freeman RD. Asymmetric suppression outside the classical receptive field of the visual cortex. J Neurosci 19: 10536 –10553, 1999. Walker GA, Ohzawa I, Freeman RD. Suppression outside the classical cortical receptive field. Vis Neurosci 17: 369 –379, 2000. Webb BS, Dhruv NT, Solomon SG, Tailby C, Lennie P. Early and late mechanisms of surround suppression in striate cortex of macaque. J Neurosci 25: 11666 –11675, 2005a. Webb BS, Tinsley CJ, Vincent CJ, Derrington AM. Spatial distribution of suppressive signals outside the classical receptive field in lateral geniculate nucleus. J Neurophysiol 94: 1789 –1797, 2005b. Xing D, Shapley RM, Hawken MJ, Ringach DL. Effect of stimulus size on the dynamics of orientation selectivity in macaque V1. J Neurophysiol 94: 799 – 812, 2005. Yoshioka T, Blasdel GG, Levitt JB, Lund JS. Relation between patterns of intrinsic lateral connectivity, ocular dominance, and cytochrome oxidasereactive regions in macaque monkey striate cortex. Cereb Cortex 6: 297– 310, 1996. Zipser K, Lamme VAF, Schiller PH. Contextual modulation in primary visual cortex. J Neurosci 16: 7376 –7389, 1996.
105 • MARCH 2011 •
www.jn.org
