Two Types of Receptive Field Dynamics in Area V4 at the Time of Eye Movements?

Abstract

How we perceive the world as stable despite the frequent disruptions of the retinal image caused by eye movements is one of the fundamental questions in sensory neuroscience. Seemingly convergent evidence points towards a mechanism which dynamically updates representations of visual space in anticipation of a movement (Wurtz, 2008). In particular, receptive fields (RFs) of neurons, predominantly within oculomotor and attention related brain structures (Duhamel et al., 1992; Walker et al., 1995; Umeno and Goldberg, 1997), are thought to "remap" to their future, post-movement location prior to an impending eye movement. New studies (Neupane et al., 2016a,b) report observations on RF dynamics at the time of eye movements of neurons in area V4. These dynamics are interpreted as being largely dominated by a remapping of RFs. Critically, these observations appear at odds with a previous study reporting a different type of RF dynamics within the same brain structure (Tolias et al., 2001), consisting of a shrinkage and shift of RFs towards the movement target. Importantly, RFs have been measured with different techniques in those studies. Here, we measured V4 RFs comparable to Neupane et al. (2016a,b) and observe a shrinkage and shift of RFs towards the movement target when analyzing the immediate stimulus response (Zirnsak et al., 2014). When analyzing the late stimulus response (Neupane et al., 2016a,b), we observe RF shifts resembling remapping. We discuss possible causes for these shifts and point out important issues which future studies on RF dynamics need to address.

Keywords: attention; extrastriate cortex; eye movements; receptive field; remapping.

Figure 1

V4 receptive field shifts at the time of saccades. (A) Centers of 71 V4 current receptive fields (cRFs) plotted together with the visual probe grid (white disks) drawn to scale. Responses of V4 neurons were probed during fixation at the fixation point (FP) or the saccade target (ST; red disks), long before and after an eye movement, or shortly before an eye movement from the FP to the ST. RFs were then estimated for each condition by fitting Gaussian functions at various times to the neuronal responses recorded in the three conditions. Blue diamonds indicate the centers of the estimated cRFs during fixation at the FP long before and after an eye movement (see “Materials and Methods” Section for details). The average cRF center was x = 0.15° and y = −3.07° (gold disk) relative to the ST (x = 0°, y = 0°). (B) Mean responses of the recorded V4 population to three probes flashed briefly before the onset of a saccade (probe 2). The blue line shows the mean response to a probe (blue disk) presented closest to the centers of the individual cRFs. The red line shows the mean response to a probe (red disk) presented 0.9° above the probe closest to the cRF, closer to the ST (STdir). The green line shows the mean response to a probe (green disk) presented closest to the centers of the estimated future fields (FFs). (C) RF shift estimates based on the immediate, early presaccadic probe responses (red shaded area in B). Each line indicates the difference between the center of the cRF (x = 0°, y = 0°), as measured long before an eye movement, and the center of the RF, as measured shortly before movement onset (alignment as in Tolias et al., 2001). Consistent with Tolias et al. (2001), V4 RF centers shifted towards the ST (see also Figure 2I). The amplitude of the RF shift towards the ST depended on the distance of the cRF center to the ST (upper inset). Solid line depicts best linear fit. Furthermore, consistent with Tolias et al. (2001) V4 RFs shrank by relative to their cRF size (lower inset). Solid line depicts line of unity. (D) Average population RF (pRF) based on the early visually evoked activity relative to saccade offset, as measured shortly before a saccade (probe 2) during fixation at the FP (see “Materials and Methods” Section for details). Full dynamic range of responses is shown. Consistent with individual RFs (C), the average RF shifts upwards, away from the current population RF (pRF) center as measured during fixation (probe 1; blue diamond; Figure 2A) and towards the STs (red diamonds; see also Figure 2I). (E) RF shift estimates based on the late, post-movement activity of presaccadic probe responses (green shaded area in B). RF centers shift in the direction of the FF (green diamond at x = 2.76°) consistent with Neupane et al. (2016a,b); see also Figure 2J. (F) Average pRF based on the late activity relative to saccade offset. The full dynamic range of responses is shown in the top panel. The bottom panel shows a strongly reduced dynamic range. Consistent with individual RFs (E), the average RF shifts rightwards, away from the pRF center as measured during fixation (probe 1; blue diamond) and towards the FF. (G) RF shift estimates based on the later, post-movement activity of presaccadic probe responses (purple shaded area in B). Each line indicates the difference between the center of the cRF (x = 0°, y = 0°), as probed (probe 1) long before an eye movement during fixation at the FP, and the center of the RF as probed (probe 2) shortly before movement onset. RF centers shift into the direction of the FF (x = 2.76°, y = 0°; see also Figure 2K). (H) Average pRF based on the later activity relative to saccade offset. Reduced dynamic range of responses is shown. The activity is still biased to the right, resembling the activity pattern shown in Figure 2B.

Figure 2

V4 receptive field shifts during fixation and at the time of saccades. (A) Average population RF (pRF) based on early visually evoked activity (53–106 ms) relative to probe onset (probe 1), as measured long before a saccade during fixation at the FP (see “Material and Methods” Section for details). Blue diamond indicates the center of all RFs (pRF^probe1). Red diamonds indicate the STs relative to the individual current RF (cRF) centers (Figure 1A). Full dynamic range of responses is shown. (B) Average RF based on later visually evoked activity relative to probe onset (249–347 ms) during fixation of the FP. The full dynamic range of responses is shown in the top panel. The bottom panel shows a strongly reduced dynamic range. The activity is shifted to the right resembling a “negative RF”. (C) RF shift estimates based on the later responses to fixation probes. Each line indicates the difference between the center estimated with the Gaussian fit of the cRF (x = 0°, y = 0°), as measured long before an eye movement during fixation at the FP and based on the immediate probe response, and the center of the RF based on the later probe responses. RF centers shift to the right into the direction of the FF (x = 2.76°, y = 0°) and upward with an average amplitude of 2.72° (p < 10⁻⁷, Wilcoxon rank test; see also L). (D) Average RF based on early visually evoked activity (53–106 ms) relative to probe onset (probe 3), as measured long after a saccade during fixation at the ST. Green diamond indicates the center of all RFs (pRF^probe3). Full dynamic range of responses is shown. (E) Population RF based on the later visually evoked activity (249–347 ms) relative to probe onset during fixation at the ST. The full dynamic range of responses is shown in the top panel. The bottom panel shows a strongly reduced dynamic range. This time the activity is shifted to the left resembling again a “negative RF”. (F) RF shift estimates based on the later responses to fixation probes. Each line indicates the difference between the center of the cRF (x = 0°, y = 0°), as measured long after an eye movement during fixation and based on the early probe response, and the center of the RF based on the later probe responses. RF centers shift to the left into the direction of the pre eye movement cRF as measured during fixation at the FP with an average amplitude of 3.66° (p < 10⁻¹⁰, Wilcoxon rank test; see also M). (G) Nominal saccade vector (2.76°) from the FP to the ST (x = 0°, y = 0°). (H) Average displacement (2.70°) of current RF (cRF) centers as measured long before and after a saccade during fixation at the FP (probe 1; x = 0.15°, y = −3.07°) and at the ST (probe 3; x = 2.85°, y = −3.1°). cRF estimates are based on the immediate, early probe responses (A,D). Shaded regions depict one standard deviation around the respective means. (I) Average shift of RF centers (0.41°) towards (x = 0.09°, y = −2.67°) the ST based on the immediate, early responses to presaccadic probes (probe 2; Figure 1B). Shaded regions depict one standard deviation around the respective means. (J) Average shift of RF centers (1.75°) into the FF (cRF^probe3) direction to the right (x = 1.9°, y = −2.9°). RF estimates are based on the late presaccadic probe (probe 2) responses (Figure 1B). Shaded regions depict one standard deviation around the respective means. (K) Average shift of RF centers (1.77°) into the direction of the FF (x = 1.9°, y = −2.7°). RF estimates are based on the later responses to presaccadic probes (Figure 1B). Shaded regions depict one standard deviation around the respective means. (L) Average shift of RF centers (2.72°) into the FF direction and upwards away from the ST (x = 2.3°, y = −1.4°). RF estimates are based on the later probe (probe 1) responses during fixation at the FP (B). Shaded regions depict one standard deviation around the respective means. (M) Average shift of RF centers (3.66°) into the direction of the cRF^probe1 (x = −0.77°, y = −2.53°). RF estimates are based on the later probe (probe 3) responses during fixation at the ST (E). Shaded regions depict one standard deviation around the respective means.