Cerebral Cortex, Vol. 10, No. 12, 1217-1232,
December 2000
© 2000 Oxford University Press
Modulatory Influence of Feedback Projections from Area 21a on Neuronal Activities in Striate Cortex of the Cat
1 Departments of Anatomy and Histology and , 2 Physiology, Institute for Biomedical Research, The University of Sydney, NSW 2006, Australia and , 3 Department of Neurophysiology, Nencki Insitute of Experimental Biology, 3 Pasteur Street, Warsaw 02-093, Poland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsEffects of Cooling of...DiscussionConclusionsNotesReferences |
|---|
We have examined the influence of ‘feedback’ projections from extrastriate visual cortical area 21a on the responses of neurons in area 17 of the cat, by cooling area 21a to 5–10°C while the temperature over the recording sites was kept at 36°C. Orientation, direction and length selectivities as well as contrast sensitivity were tested before and during cooling and after rewarming of area 21a. Overall, for the sample of cells recorded from the part of area 17 visuotopically corresponding to area 21a, the ‘spontaneous’ activity (at background illumination of 1 cd/m2) and the responsiveness to visual stimuli of standard contrast (15) were significantly reduced by inactivation of area 21a. In about half of the cells inactivation of area 21a affected substantially the sharpness of orientation-tuning. However, only in a minority of the cells were the direction or length selectivities significantly affected by inactivation of area 21a. Thus (i) the feedback projections from area 21a appear to exert mainly an excitatory influence on the background activity and responsiveness of area 17 cells and (ii) only in subgroups of area 17 cells does the feedback activity originating from area 21a appear to modulate specifically certain receptive field properties.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsEffects of Cooling of...DiscussionConclusionsNotesReferences |
|---|
Starting with the early ‘classical’ work of Hubel and Wiesel (Hubel and Wiesel, 1959
, 1962, 1965), a substantial amount of data about the functional organization of neurons in the mammalian visual cortices has been accumulated. The general consensus has emerged that the retino-geniculo-cortical pathway and the so-called ‘feedforward’ cortico-cortical association path- way provide the principal excitatory drive to cortical neurons in both the primary and the ‘higher-order’ mammalian visual cortices (Douglas and Martin, 1991; Domenici et al., 1995; Shao and Burkhalter, 1996) [for reviews see (Stone et al., 1979; Orban, 1984; Dreher, 1986; Maunsell and Newsome, 1987; Livingstone and Hubel, 1988; Felleman and Van Essen, 1991; Henry, 1991; Coogan and Burkhalter, 1993; Bullier et al., 1994; Salin and Bullier, 1995; Burke et al., 1998)].
On the other hand, much less is known about the functional role of the extensive cortico-cortical ‘feedback’ projections from the higher-order visual cortical areas to the lower-order cortical areas (Rockland and Pandya, 1979; Van Essen and Maunsell, 1983; Dreher, 1986; Zeki and Shipp, 1988; Shipp and Zeki, 1989a,b; Felleman and Van Essen, 1991; Van Essen et al., 1992; Rockland, 1994; Salin and Bullier, 1995). Most of the experimen- tal data so far available suggest that feedback projections from the higher-order cortical areas play only a modulatory role in pro- cessing of visual information in the lower-order cortical areas (including the primary visual cortices). Thus, inactivation of cytoarchitectonic area 18 in a New-World primate, the squirrel monkey, did not abolish the responses to visual stimuli of the great majority of neurons in the striate cortex, but did affect the magnitude of the responses of about one-third of area 17 cells and, in most cases, the magnitude of these responses was substantially reduced (Sandell and Schiller, 1982). Furthermore, Hupé and colleagues (Hupé et al., 1998) have demonstrated that many neurons in area V1 as well as those in two other relatively lower-order areas (V2 and V3) of the macaque visual cortex exhibit reduced ability to differentiate low-salience figures from the background when one of the motion areas, the middle tem- poral area (MT, or area V5), is reversibly inactivated by cooling. Thus, it appears that the feedback connections could affect the ‘functional performance’ of neurons located at a lower level of visual information-processing hierarchies.
In the cat, cytoarchitectonic area 17 (which, together with area 18, constitutes the primary visual cortex since both areas receive their principal thalamic input from the dorsal lateral geniculate nucleus) projects directly to and receives direct feedback input from a number of visuotopically organized cor- tical areas [Fig. 1A,B; for reviews see (Rosenquist, 1985; Dreher, 1986; Burke et al., 1998)]. We wished to examine a possible functional role of feedback projections to area 17 originating from area 21a, which is located at some distance from the primary visual cortices (Fig. 1A). Michalski and colleagues (Michalski et al., 1993, 1994) have demonstrated by reversible inactivation (cooling) of area 17 that the well documented and numerically large inputs from area 17 (Symonds and Rosenquist, 1984a,b; Dreher, 1986; Mizobe and Toyama, 1989; Dreher et al., 1996b; Morley et al., 1997) provide a major excitatory drive to area 21a neurons. The studies in our laboratory (Dreher, 1986; Dreher et al., 1993, 1996a; Wang and Dreher, 1996) as well as those in several other laboratories (Mizobe et al., 1988; Wimborne and Henry, 1992; Toyama et al., 1994; Tardiff et al., 1996; Morley and Vickery, 1997; Vickery and Morley, 1997, 1999) have shown that, despite their relatively large receptive fields and relatively low spatial frequency tuning, the neurons in area 21a are characterized by low temporal resolution, prefer- ence for slowly moving stimuli and sharp orientation-tuning. Furthermore, almost all area 21a neurons are binocular and in most of them responses are strongly modulated by binocular retinal disparities of a stimulus (Wang and Dreher, 1996; Vickery and Morley, 1997, 1999). Overall, the receptive field properties of area 21a neurons tend to be quite similar to those of neurons in area 17 of the cat [cf. for reviews (Orban, 1984; Dreher, 1986; Burke et al., 1992; Dreher et al., 1992)]. It appears that areas 17 and 21a are involved in the same functional information- processing stream, that is, form rather than motion analysis, and area 21a represents a gateway area to a distinct form-processing stream in the temporal cortex (Dreher, 1986; Wimborne and Henry, 1992; Dreher et al., 1993, 1996a, Dreher et al., b; Morley and Vickery, 1997; Burke et al., 1998).
|
Specifically, in the present study we have attempted to examine the influence of feedback projections from area 21a on ‘spontaneous’ (background) activity and responses of neurons in the part of area 17 visuotopically corresponding to the part of the visual field represented in area 21a. This was achieved by comparing the responses of area 17 neurons to visual stimuli presented before and during inactivation by cooling of area 21a as well as after rewarming of area 21a. Preliminary results were reported previously (Wang et al., 1999
). |
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsEffects of Cooling of...DiscussionConclusionsNotesReferences |
|---|
Animal Preparation
Experiments were carried out on eight adult cats of either gender weighing 2.5–4.0 kg. The animals were initially anaesthetized with 2.5–5.0% halothane in an N2O/O2 (67%/33%) gaseous mixture. A surgical level of anaesthesia was maintained with 1.5–2.5% halothane in the same gaseous mixture while the surgery, intravenous and tracheal cannulation and bilateral cervical sympathectomy were performed. The animals were then placed in a stereotaxic headholder and craniotomies were performed over area 21a (Horsley–Clarke or HC coordinates: lateral 8–16 mm; posterior 0–8 mm) and area 17 (HC coordinates: lateral 0–5 mm; posterior 0–8 mm). Area 21a was cooled by means of a specially shaped metal probe put over the surface of the area while a second probe was placed over area 17 to keep the local temperature constant at 36°C (Fig. 1A
).
During the recording sessions, the animals were paralyzed by intra- venous infusion of gallamine triethiodide at the rate of 7.5 mg/kg/h in a mixture of equal parts of 5% dextrose and sodium lactate (Hartmann's) solutions and artificially ventilated. Anaesthesia was maintained with gaseous mixture of halothane (0.5–0.7%) in N2O/O2 (67%/33%). Peak expired-CO2 was continuously monitored and maintained at 3.7–4.0% by adjusting the stroke volume or rate of the pulmonary pump. Body temperature was automatically maintained at ~37.5°C by an electric heating blanket. Heart rate and electroencephalogram (EEG) were monitored continuously. With the above regimen the EEG maintained a slow-wave record while the heart rate was kept below 180 beats/min. Antibiotic (amoxycillin trihydrate, 75 mg), dexamethasone phosphate (4 mg) and atropine sulphate (0.3 mg) were given intramuscularly on a daily basis. Experimental procedures followed the guidelines of the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes and were approved by the Animal Care Ethics Committee of the University of Sydney.
The corneas were protected with zero-power, air-permeable plastic contact lenses. Pupils were dilated and accommodation paralysed with 1% atropine sulphate solution. The nictitating membranes were retracted with 0.128% phenylephrine hydrochloride. Artificial pupils (3 mm in diameter) were placed in front of the contact lenses. Correcting lenses were used, if required, to focus the eyes on a tangent screen located 57 cm in front of the eyes. The locations of the optic discs and the areae centrales were plotted twice daily using a fibre-optic projector (Pettigrew et al., 1979).
Recording
A plastic cylinder was glued to the skull surrounding the opening over area 17 to form a well. A smaller opening in the dura was made above the recording site. A stainless steel microelectrode (11 M
; FHC, Brunswick, ME) was positioned medially to the warming probe and lowered until its tip was positioned at the cortical surface; the cylinder was then filled with 4% agar gel and sealed with warm wax to reduce brain pulsations. The microelectrode was advanced into the cortex by a step-motor-controlled hydraulic micromanipulator. The action potentials of single neurons were recorded extracellularly, amplified conventionally and monitored both visually and via the loudspeaker; triggered standard pulses (again mon- itored both visually and aurally) were then fed to a microcomputer.
Temperature Control
Temperature control of cortical areas, for example, cooling of cortical area 21a as well as the subsequent rewarming of the area, was achieved by using a gold-plated metal probe attached to a Peltier element (9 W; Marlow, Dallas, TX) which, in turn, was controlled by a laboratory-built device. The temperature on the cortex could be preset and automatically maintained by adjusting the polarity and the amount of current passing through the Peltier element. With our temperature control units the preset temperatures (in a range of 5–40°C) could be reached within 1 min and could be maintained constantly for as long as required. In the same regime, area 17 was kept warm at 36°C. The probe placed over area 21a was quadrant-shaped with a radius of 7 mm, while the probe for controlling temperature of area 17 was rectangular, 3 x 6 mm. The contact surfaces of the probes were shaped to match the curvatures of the cortical surfaces on which they were placed. Temperatures over areas 17 and 21a were continuously monitored by the thermocouples built into the foot of each probe. The warming probe for area 17 with temperature maintained at 36°C was placed lateral to the recording site, which, together with the recording electrode, was covered with 4% agar and wax in the well. When area 21a was not being cooled its temperature was also kept at 36°C. The responses of single neurons recorded in area 17 were first tested when the temperature of area 21a was maintained at 36°C, i.e. our control condition, and then the same test was performed again when area 21a was cooled to 5–10°C. To avoid possible irrevers- ible damage to the cortex, the cooling process was discontinued after 10–15 min and the cortex was slowly (over the period of 2–3 min) rewarmed back to 36°C. Data collections usually started 3 min after the preset temperature was reached. After rewarming of area 21a to 36°C several runs were made until either complete recovery or an incomplete but steady-state recovery had been achieved.
Visual Stimulation and Qualitative Data Analysis
The receptive fields of recorded cells were first explored and plotted with a hand-held light slit projector on a tangent screen. The general features of the receptive field, such as size of the discharge field, receptive field organization, ocular dominance, orientation, direction and velocity pre- ferences, were first examined qualitatively. The excitatory receptive field (discharge field) was defined as the area of visual space within which visual stimuli elicited an increase in the cell's firing rate detectable by ear. Cells which exhibited spatially separate ON and OFF discharge regions to optimally oriented, stationary flashing slits and non-overlapping dis- charge regions to moving, optimally oriented elongated bars darker or brighter than the background were identified as S-type cells [see S-cells and A-cells (Henry, 1977); cf. simple cells (Hubel and Wiesel, 1962; Gilbert, 1977)]. Cells which gave transient responses at both onset and offset of stationary flashing slits of light positioned anywhere within their discharge fields (ON/OFF discharge fields), and/or their discharge fields plotted with moving stimuli darker than the background overlapped completely with those revealed with moving stimuli brighter than the background, were identified as C-type cells [see C-cells and B-cells (Henry, 1977); cf. complex cells (Hubel and Wiesel, 1962, 1965; Gilbert, 1977)]. End-zone inhibition or end-stopping was defined as the reduction of the response magnitude when the length of the optimally oriented bar exceeded the length of the discharge field (Hubel and Wiesel, 1962; Dreher, 1972; Gilbert, 1977). S-type cells with end-zone inhibition were identified as SH cells (Henry, 1977) [cf. hypercomplex type I cells (Dreher, 1972; Gilbert, 1977; Kato et al., 1978)] while the C-type cells with end-zone inhibition were identified as CH cells (Henry, 1977) [cf. hypercomplex cells (Hubel and Wiesel, 1965; Gilbert, 1977) and hypercomplex type II cells (Dreher, 1972; Kato et al., 1978)].
Quantitative Data Analysis
After qualitative analysis of the receptive field properties of a cell, the screen was then replaced with a computer-controlled CRT monitor (BARCO, Belgium) where the visual stimuli were presented at a distance of 57 cm from the cat's eyes. Light bars of different orientations, lengths and contrasts were employed to examine the possible contributions of feedback from area 21a on the receptive field properties of area 17 neurons. After determining the receptive field properties, such as velocity preferences, orientation selectivity and length selectivity, we examined the effect of cooling area 21a on a particular receptive field property. Finally, area 21a was rewarmed to 36°C and the receptive field properties of the neurons were re-examined.
The stimulus velocities used for examination of the effects of revers- ible inactivation of area 21a were usually optimal for the individual cell under test and ranged from 2 to 25°/s. When testing the orientation or length selectivity, a set of bars of different orientations or different lengths moving at optimal velocities were interleaved and presentations were repeated 5–10 times. Orientation selectivity was examined with light slits which were 5–10 times the length of the tested receptive field (along its optimal orientation axis) moving in the direction orthogonal to the axis of orientation. As a measure of the sharpness of orientation selectivity of an individual neuron we used the orientation-tuning width at half-height.
Since we used single bars on a constant background we used the following definition of contrast:
 |
where Lo and LB are respectively luminance of the object (in our case single bars) and luminance of the background. Thus defined, the contrast used by us for testing orientation, direction and length selectivity was 15 at the background luminance of 1 cd/m2. In testing contrast sensitivity a set of optimally oriented moving light bars of varying contrast was interleaved pseudo-randomly and presentations were repeated 5–10 times. Two ‘control runs’ were usually done before cooling was applied.
The peristimulus time histograms were constructed by summing the responses collected during the stimulus period. The peak discharge rate was determined as the maximum discharge rate after applying a smoothing algorithm, the Gaussian weighted average over five neighbouring bins.
The mean direction selectivity index (MDI), defined as the geometric mean of the direction selectivity indices (DI) at different orientation, contrast or length (Orban et al., 1981), was used to evaluate the direction selectivity of a given cell. The MDI was calculated using the following formula:
 |
where n is the number of orientations, contrasts or lengths tested and Ri is the response in the preferred direction for each orientation, contrast or length. DI was calculated by the following formula:
 |
where Rp and Rnp are the responses to the stimuli moving in the preferred and non-preferred direction respectively.
For determining the strength of response suppression in the end- stopped region of end-stopped cells we introduced an ‘end-stopped suppression index’ (ESI). It was calculated according to the following formula:
 |
where Rmax is the response produced by a stimulus of an optimal length for a given cell and Rmin is the minimum average response to a number of stimuli of greater than optimal length (where the slope of response declination was <10% of its maximum). Usually, the lengths used to calculate Rmin were >5 times the optimal length.
Localization of Cooling and Recording Sites
Area 21a was localized on the basis of its relationship to the lateral suprasylvian and marginal (lateral) sulci (Fig. 1A). The reconstruction of electrode penetrations in area 17 was based on small electrolytic lesions (5–7 µA for 20 s) at the end of the first and last electrode penetrations. At the end of the recording session (lasting usually 3–5 days) the animal was deeply anaesthetized with an intravenous injection of 120 mg of sodium pentobarbitone and perfused transcardially with warm (37°C) Hartmann's solution followed by a 4% solution of paraformaldehyde in 0.2 M phosphate buffer (pH 7.4). Localization of the position of the cooling probe was aided by the post-mortem location of the imprint left by the probe on the brain. Finally, the brains were stereotactically blocked and sectioned coronally at 50 µm on a freezing microtome, mounted on slides and counterstained with cresyl violet.
Statistics
Three non-parametric statistical tests were used: the Wilcoxon matched- pairs signed-ranks test, the Mann–Whitney U-test and the 
2 test (Siegel, 1956). The tests were applied to single neurons and to the populations of neurons. Statistical significance of the differences between the two sets of data was accepted if the probability (P) using the two-tailed criterion was <0.05. Correlations were examined with the Pearson statistic (r). The statistics are given as mean values ± SEM. For individual cells a difference between the state when areas 17 and 21a were at 36°C and the state when area 17 was at 36°C while area 21a was at 5–10°C was deemed significant only if : (i) a significant difference was found between the initial control state (areas 17 and 21a at 36°C) and the cooled state (area 21a at 5–10°C); and (ii) a significant difference was found between the cooled state and the subsequent rewarmed state (areas 17 and 21a again at 36°C).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsEffects of Cooling of...DiscussionConclusionsNotesReferences |
|---|
Cooling of Area 21a vs neighbouring areas and area 17
Multiunit activities (both ‘spontaneous’ and visually evoked) in area 21a were effectively abolished when the temperature of the area dropped to ~10°C. Despite the fact that our cooling probe was restricted to area 21a, due to the temperature gradient, cooling of area 21a resulted also in lowering the temperature in the number of cortical areas in proximity to area 21a. Indeed, even in area 17, which is relatively remote from area 21a (Fig. 1A), when the temperature of area 21a was at ~10°C and the warming probe was not placed over area 17, the temperature measured at 1 mm below the surface of the cortex dropped to 30–33°C. The dual temperature control approach used in the present study allowed us to maintain the physiological temper- ature at the recording sites in area 17 irrespective of the tem- perature of area 21a. This was true not only when the recording sites were near the marginal (lateral) gyrus, that is, in close proximity of the warming probe, but also when the recording sites in area 17 were 3–4 mm below the cortical surface along the medial aspect of the hemisphere. Since in our experimental set-up the warming probe covered not only area 17 but also most, if not all, of the part of area 18 visuotopically corres- ponding to area 21a, this area also was probably well protected against the cold spreading across the marginal sulcus from area 21a. However, since our warming probe (i) did not cover the areas in the immediate proximity of area 21a, such as area 19, the postero-medial lateral suprasylvian area (PMLS) and area 7, and (ii) the temperature gradient measured by us was ~5°C/mm, the temperature in parts of those areas was almost certainly sub- stantially reduced during cooling of area 21a (see Discussion).
Receptive Field Location of Tested Area 17 Cells
The effects of reversible inactivation (by cooling and rewarm- ing) of area 21a on the responses to visual stimuli were tested in 62 single neurons recorded from area 17. Fifty-one of the cells had their receptive fields located within 12° from the area centralis in the part of area 17 visuotopically corresponding to the part of the visual field represented in area 21a (Tusa et al., 1978, 1981; Tusa and Palmer, 1980; Dreher et al., 1993). For brevity, this region of area 17 will henceforth be referred to as the VC21a region. The remaining 11 cells tested by us had receptive fields located in the lower visual field at eccentricities exceeding 10° (range 14–23°), that is, outside the VC21a region (Tusa and Palmer, 1980; Tusa et al., 1981; Wimborne and Henry, 1992; Dreher et al., 1993).
Typical Result
Figure 2A shows peristimulus time histograms of responses of an S-type cell recorded from the VC21a part of area 17 to elongated light bars of different orientations moving across the cell's discharge field. Fairly typically, with background luminance of 1 cd/m2, this cell had relatively low ‘spontaneous’ activity and responded selectively to particular stimulus orientations. Cool- ing area 21a to 10°C resulted in a 36% decrease in the magnitude of responses to an optimally oriented (almost horizontal) bar moving across the cell's receptive field (second row from the top in Fig. 2A). A very similar relative decrease in magnitude of responses was apparent when non-optimally oriented stimuli moved across the cell's discharge field. However, the cell's response profile (Fig. 2A) and its orientation-tuning curve were largely unchanged by inactivation of area 21a (Fig. 2B). Seven minutes after area 21a was rewarmed to 36°C, the responsive- ness of the cell to stimuli of any orientation returned almost completely to the control (precooling) levels (Fig. 2A,B). Overall, the decrease in the magnitude of the responses following in- activation of area 21a in relation to the precooling data and the data collected after rewarming of area 21a was statistically significant (P = 0.018 in both cases, Wilcoxon test).
|
Effects of Cooling of Area 21a on Background Activities of Area 17 Neurons
Figure 3A shows the overall relationship between the mean background (‘spontanous’) activity of area 17 cells when the temperature of both area 21a and area 17 was 36°C and that during the time when area 21a was inactivated by cooling. It is apparent from the figure that, although with few exceptions background activity of area 17 was lower when area 21a was inactivated, all but eight cells exhibited some background activity during the cooling of area 21a. Furthermore, three of these cells did not exhibit any spontaneous activity even when area 21a was not inactivated by cooling.
|
For the entire sample of area 17 cells recorded from the VC21a region (51 cells) cooling area 21a resulted in a significant decrease in background activity (from 1.3 ± 0.3 spikes/s to 0.7 ± 0.2 spikes/s, P = 0.0003, Wilcoxon test). By contrast, the spontaneous activity of cells with receptive fields located ouside the VC21a region (11 cells) was not significantly affected by inactivation of area 21a (prior to cooling: mean background activity 1.2 ± 0.45 spikes/s; during cooling: mean background activity 1.5 ± 0.55 spikes/s, P > 0.20, Wilcoxon test).
When the data were examined for effects on single neurons it was found that cooling of area 21a resulted in significant changes in background activity (P < 0.05, Wilcoxon test) in a substantial proportion of cells recorded from the VC21a region (21/51; 41.2% of the sample). Of these, the great majority (17 cells) showed significant decreases in background activity, with only four cells showing significant increases. For cells recorded from the ouside of the VC21a region (11 cells) one cell showed a significant decrease in the spontaneous activity while three cells showed significant increases in background activity.
Effects of Cooling of Area 21a on Responsiveness of Area 17 Neurons to Visual Stimuli
Peak discharge rates were determined by measuring the re- sponse to an approximately optimal stimulus, i.e. at optimal orientation, length and velocity. The peak discharge rates of the cells recorded from the VC21a region of area 17, like their back- ground activities (see above), were also significantly affected by inactivating area 21a (Fig. 3B). When area 21a was cooled, the mean peak discharge rate of cells with receptive fields located within the VC21a region at 45.3 ± 4.55 spikes/s was substantially lower than that (58.0 ± 4.7 spikes/s) when area 21a was at control (36°C) temperature. The difference between the two sets of data was highly significant (P = 0.003, Wilcoxon test). Furthermore, the changes in the peak discharge rates were significantly correlated with changes in background activity (r = 0.30, n = 51, P < 0.05).
However, for the whole sample of cells with receptive fields located outside the VC21a region the peak discharge rates prior to cooling of area 21a (like their background activity, see above) were not significantly different from those during cooling of area 21a (mean peak discharge rate prior to cooling: 73.9 ± 13.9 spikes/s; during cooling: 70.6 ± 13.25 spikes/s, P = 0.42, Wilcoxon test).
When individual neurons were examined, in almost one-third of cells located within the VC21a region of area 17 (15/51; 29.5% of the sample) the magnitude of responses to visual stimuli was significantly affected (P < 0.05, Wilcoxon test) by inactivation of area 21a. In the great majority of these cells (14/15; seven S-type or SH-type cells, five C-type cells and two unclassified cells) the magnitude of responses to visual stimuli was significantly reduced (the decreases ranged from 8 to 86%, with the median and mean decreases being 39 and 42% respectively) while in one cell (an S-type cell) there was a substantial (78%) and significant increase (P < 0.05, Wilcoxon test) in the magnitude of responses.
Of the 11 cells outside the VC21a region only one cell (an S-type cell) showed a significant change in the magnitude of responses (an increase).
Eye Dominance
Unlike in areas 17 and 18, which provide the principal cortical inputs to area 21a (Dreher, 1986; Dreher et al., 1996b; Morley et al., 1997), the majority of cells in area 21a are either binocular cells dominated by the ipsilateral eye (class 4 cells) or monocular (class 5) cells activated exclusively via the ipsilateral eye (Dreher et al., 1993; Wang and Dreher, 1996). One might have expected on this basis that inactivation of area 21a could have had a greater effect on responsiveness of class 4 and class 5 cells than on the responsiveness of class 1 and class 2 cells. However, the overall reduction in the magnitude of responses of class 4 and class 5 cells (mean peak discharge prior to cooling: 63.7 ± 8.6 spikes/s; during cooling: 50.55 ± 9.9 spikes/s) resulting from inactiva- tion of area 21a was not significantly different (P = 0.84, Mann–Whitney U-test) from the reduction in the magnitude of responses of class 1 and class 2 cells (mean peak discharge prior to cooling: 56.55 ± 7.4 spikes/s; during cooling: 42.35 ± 5.2 spikes/s). Similarly, the proportion of class 4 and class 5 cells showing significant decrease in the magnitude of response was not significantly different from the proportion of such cells in classes 1 and 2 (35 versus 29%; P > 0.9, 2 test).
Recovery of Responsiveness of Area 17 Cells following Rewarming of Area 21a
For most area 17 cells (53/62; 85.5% of the sample) in which we examined the effect of cooling of area 21a we were also able to study the extent and time course of the recovery of their respons- iveness following rewarming of area 21a to 36°C.
In the substantial majority of cells (39/53; 73.6% of the sample) the magnitude of the response recovered gradually to at least 80% of the precooling value. In a few cases the magni- tude of responses following rewarming of area 21a was 10–20% greater than the magnitude of the control precooling responses. The mean level of recovery for the entire sample was 85 ± 21% of the precooling values. While the changes in background activity and responsiveness of area 17 neurons followed almost immedi- ately on cooling of area 21a to 10°C, recovery of responsiveness of area 17 cells after rewarming of area 21a to 36°C was much more gradual. The time for the complete or an incomplete but steady state recovery varied from 5 to 50 min and the mean recovery time for the entire sample was 14.7 ± 10.9 min. We note that the recovery of responsiveness of area 21a after rewarming area 17 to physiological temperature had a very similar time- course (Michalski et al., 1993).
Effects of Cooling of Area 21a on Orientation Selectivity
The effect of cooling area 21a on orientation selectivity was examined in forty- two area 17 cells recorded from the VC21a region and in nine cells recorded from outside the VC21a region. Most of them (38/51; 74.5% of the sample) were considered to be orientation-selective since their orientation-tuning width at half-height of response ranged from 14.5 to 81.3°. In the remain- ing cells orientation-tuning widths at half-height were either within the range 90–150° (8/51 ‘orientation-biased’ cells; 15.5% of the sample) or exceeded 150° (5/51 non-orientation selective cells; 10% of the sample).
Figure 4A–D shows examples of orientation-tuning curves before and during cooling of area 21a and (with exception of Fig. 4C) after its rewarming. For the cell whose orientation- tuning is illustrated in Figure 4A inactivation of area 21a resulted in substantial reversible increases in the magnitude of responses to stimuli of most of the orientations presented and overall broadening of the orientation-tuning curve (measured at half- height). By contrast, for the cell whose orientation-tuning curves are illustrated in Figure 4B, cooling of area 21a resulted in sub- stantial reversible reductions in the magnitude of responses to stimuli of most of the orientations presented and some narrowing of the orientation-tuning curve. For the cell whose orientation-tuning curves are illustrated in Figure 4C inactivation of area 21a also resulted in substantial reductions of the magnitude of responses to stimuli of different orientations. However, at the same time there was a broadening, rather than a narrowing, of the orientation-tuning curve. Finally, in the case of the broadly tuned cell whose orientation-tuning curves are illustrated in Figure 4D cooling of area 21a resulted in substantial reductions in the magnitude of responses and a clear reduction in orientation-tuning width. This reduction in the width of the orientation-tuning curve was accompanied by some increase in direction selectivity. Irrespective of the direction of changes in the orientation-tuning width (broadening or narrowing), the pre- ferred orientations of the cells were unaffected by inactivation of area 21a (Figs 2A,B and 4A–D).
|
Figure 4E
graphs the relative changes (%) in the orientation- tuning widths of 41 area 17 cells during inactivation of area 21a against the orientation-tuning widths before cooling. Of the re- maining 10 cells (not included in Fig. 4E) in which the orientation selectivity was tested quantitatively, five were non-orientation selective and inactivation of area 21a did not have any effect on their orientation-tuning. Furthermore, inactivation of area 21a resulted in complete loss of orientation selectivity in two orienta- tion-tuned cells while in the other three orientation-tuned cells, inactivation of area 21a resulted in double-peak orientation- tuning curves.
It is apparent from Figure 4E that: (i) in a substantial propor- tion of cells (20/41; 48.5% of the sample) inactivation of area 21a resulted in clear changes (over 20%) in orientation-tuning width; (ii) all but one cell (eight cells) which, following inactivation of area 21a, exhibited a substantial (>20%) broadening of their orientation-tuning curves did not exhibit significant changes in the magnitude of their responses; and (iii) orientation-tuning of cells which showed significant decreases in the magnitude of responses to visual stimuli when area 21a was inactivated tended to become narrower. Indeed, as a group these cells (n = 10) exhibited greater changes (mean narrowing 19 ± 5%) of their orientation-tuning curves than the group of cells (n = 29) which did not show significant changes in the magnitude of their responses (mean broadening 7 ± 4%). The difference between the two groups was highly significant (P = 0.004, Mann–Whitney U-test). Finally, in a subpopulation of area 17 cells (seven cells) which were characterized by broad orientation-tuning (width > 90°) cooling of area 21a resulted in significant narrowing of orientation-tuning curves (P < 0.03, Mann–Whitney U-test).
However, the mean orientation-tuning width for the entire sample of cells tested for orientation when area 21a was at 36°C (52.4 ± 5.05°) was very similar to the mean orientation-tuning width when area 21a was inactivated (49.3 ± 3.75°) and the dif- ference between the two sets of data was not significant (P = 0.44, Wilcoxon test). Furthermore, (i) the preferred orientations of the cells were unaffected by inactivation of area 21a (Figs 2A,B and 4A–D) and (ii) cooling of area 21a in most cases did not change the general shape of individual orientation-tuning curves.
Effects of Cooling of Area 21a on Contrast Sensitivity
Contrast sensitivity functions were examined before and during cooling of area 21a in 25 area 17 cells (21 cells located inside and four cells located outside the VC21a region). For about half of the cells (13/25; 52% of the sample) neither the magnitudes of responses at different contrasts nor contrast sensitivity functions were affected by reversible inactivation of area 21a (P = 0.60, Wilcoxon test, Fig. 5A).
|
However, in almost half of the cells (12/25; 48% of the sample) inactivation of area 21a resulted in significant changes (P < 0.05, Wilcoxon test) in magnitude of responses to visual stimuli of different contrasts. In nine of these cells (including two cells located outside the VC21a region) inactivation of area 21a resulted in a substantial reduction (range 21–63%) in response magnitude at different contrasts, and in three cells (all three located in the VC21a region) there were substantial increases (ranging from 27 to 135%) in response magnitude to different contrasts.
In a few cells the contrast sensitivity was selectively altered during cooling of area 21a. Thus, the magnitude of responses could be fairly selectively reduced at high contrasts (Fig. 5B) or at low contrasts (Fig. 5C). In three cells the reduction in the responsiveness at low contrasts was significantly greater (P < 0.05, Mann–Whitney U-test) than that at high contrasts while in two cells reduction in the magnitude of responses was signif- icantly greater at high contrasts than that at low contrasts (P < 0.05, Mann–Whitney U-test).
Effects of Cooling of Area 21a on Direction Selectivity
The direction selectivities of area 17 cells (n = 62) were evaluated by calculating the geometric mean of the direction selectivity indices (MDI; see Materials and Methods) at different stimulus orientations (n = 51) and in some cells (n = 11) at different stimu- lus lengths or different contrasts.
In a substantial minority of cells recorded from the VC21a region of area 17 (11/51; 21.5% of the sample) inactivation of area 21a resulted in clear reductions in the magnitude of re- sponses to stimuli moving in the preferred direction while a reduction in the magnitude of the responses to stimuli moving in the opposite direction was not as pronounced (Fig. 6A,C). However, although reductions in MDIs ranged from 0.21 to 0.59, in only one of those cells (the graph in Fig. 6A and the peri- stimulus time histograms in Fig. 6E) was the reduction in MDI significant (P < 0.05, Wilcoxon test). In a small proportion of cells (3/51 cells; 6% of the sample; Fig. 6B,D) inactivation of area 21a resulted not only in reductions in the magnitude of respon- ses to stimuli moving in either direction but also in increases in their MDI (range 0.37–0.49). Only in one of those cells was the increase significant (P < 0.05, Wilcoxon test).
|
Overall, inactivation of area 21a did not result in a statistically significant change (P = 0.38, Wilcoxon test) in the mean MDI of the entire sample of area 17 cells recorded from the VC21a region (0.51 ± 0.04 versus 0.54 ± 0.04 when area 21a was kept at the control temperature).
Inactivation of area 21a resulted in a substantial change in direction selectivity (decrease of MDI from 0.60 to 0.27) in only one of the 11 area 17 cells with receptive fields located outside the VC21a region; this change, however, was not statistically significant.
|
Effects of Cooling of Area 21a on Length Selectivity |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsEffects of Cooling of...DiscussionConclusionsNotesReferences |
|---|
A small number of cells (10/51 cells; 19.6% of the sample) recorded in the VC21a part of area 17 were classified as end- stopped cells [hypercomplex cells (Hubel and Wiesel, 1965
)] (Dreher, 1972; Gilbert, 1977) since they exhibited a reduction in the magnitude of their responses to optimally oriented moving bars when the bars extended outside the discharge fields. To assess quantitatively the strength of such a reduction for each cell we calculated the ESI, defined as the percentage reduction of responses to bars of optimal length, produced by long bars extending outside the cell's discharge field (see Materials and Methods). In eight of the end-stopped cells inactivation of area 21a resulted in reduction in their responsiveness to optimally oriented bars of any length.
Figure 7 shows the effects of cooling area 21a on the length selectivity of the end-stopped cells recorded from the VC21a part of area 17. Before cooling area 21a and after rewarming of area 21a to 36°C, the cell whose responses are illustrated in Figure 7A,B responded maximally to a 5° long bar and exhibited strong end-zone suppression (ESI = 0.75). Cooling area 21a resulted in some change of the optimal length: the cell gave the strongest response when a shorter bar of 4° was moved across its receptive field. Furthermore, the maximal response was ~90% of the maximal response before cooling of area 21a. On the other hand, inactivation of area 21a resulted in a much smaller reduction in the magnitude of the responses when the bar extended into the suppressive, end-stopped region (ESI = 0.26). The end-stopped suppression recovered gradually after the temperature in area 21a was restored to 36°C. Changes in length selectivity resulting from inactivation of area 21a were significant (P < 0.05, Wilcoxon test). A similar effect was observed in another end-stopped cell.
|
A more typical effect of inactivation of area 21a on the length selectivity index of end-stopped cells is illustrated in Figure 7C
. Although inactivation of area 21a resulted in a substantial reduc- tion in the magnitude of the responses to bars of any length, there was hardly any change in the length selectivity index (ESI = 0.79 versus 0.78 before cooling area 21a).
Overall, for the whole sample of end-stopped cells the charac- teristics of length tuning were largely unaffected by inactivation of area 21a (Fig. 7D). Indeed, the mean ESI for the whole sample of end-stopped cells before cooling of area 21a (0.64 ± 0.07) was not significantly different (P = 0.51, Wilcoxon test) from that (0.58 ± 0.10) when area 21a had been reversibly inactivated.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsEffects of Cooling of...DiscussionConclusionsNotesReferences |
|---|
The Effectiveness and Selectivity of Inactivation of Area 21a
We cooled area 21a to 5–10°C since multiunit activities in area 21a were only completely abolished when the temperature of the area dropped to ~10°C. Similarly, Michalski and colleagues (Michalski et al., 1993), who monitored the temperature of the cortex using microthermocouples very similar to those used in the present study (25 µm thick wires), observed that virtually all cells in area 17 of the cat ceased both spontaneous and visually evoked activity when the local temperature was in the 10–20°C range. Despite the fact that our cooling probe was restricted to area 21a, due to the temperature gradient of ~5°C/mm, cooling of area 21a resulted in lowering the temperature also in a num- ber of visual or polysensory cortical areas in close proximity to area 21a (see Fig. 1). Even in area 17, which is relatively remote from area 21a (see Fig. 1), when the temperature of area 21a was at ~10°C and the warming probe was not placed over area 17, the temperature measured at 1 mm below the surface of cortex dropped to 30–33°C. Indeed, a very similar temperature drop was observed in area 21a when area 17 was cooled to 10°C (Michalski et al., 1993, 1994). Even such a small temperature drop appears to have some direct effects on the responsiveness of cells in the cooled area. Thus, on the one hand, when the temperature of area 17 was directly (without cooling of area 21a) reduced to 30–33°C, the magnitude of responses of area 17 neurons was usually moderately (10–20%) reduced (Michalski et al., 1993).
On the other hand, it has been reported that some neurons in the mammalian central nervous system at temperatures slightly below physiological temperature could become hyper-respons- ive (Klee et al., 1974) [for review see (Brooks, 1983)]. Similarly, Michalski and colleagues (Michalski et al., 1993) have reported that some neurons in area 21a (but not those in area 17) of the cat increase their responsiveness when the temperature of the area drops to 30–33°C. Furthermore, most recently it has been reported that cooling slices of rat visual cortex to room temperature results in hyperexcitability of cortical neurons (Volgushev et al., 2000). These documented direct effects of a small temperature drop on the responsiveness of cells in area 17 prompted us to use in the present study the dual temperature control approach which allowed us to maintain the physiological temperature at the recording sites in area 17 irrespective of the temperature of area 21a [cf. Sherk for a similar approach to test- ing effects of selective reversible inactivation of area 17 on area 18 neurons (Sherk, 1978)].
It should be mentioned that although area 21a sends a direct projection to area 17 it can also influence area 17 indirectly via projections to areas 18, 19, PMLS, 20a and 21b, via several thalamic nuclei (including several components of the LGNd complex, the lateral posterior-pulvinar complex and rostral intralaminar nuclei) and via the dorsolateral (visual) claustrum [for reviews see (Symonds and Rosenquist, 1984a,b; Rosenquist, 1985; Bullier, 1986; Dreher, 1986; Dreher et al., 1996b)]. In the experiments described here we cannot distinguish between the effects exerted via direct and indirect pathways, although it is probable that the direct pathway is the more important if only because it involves fewer synapses.
While the dual temperature control approach allowed us to maintain the physiological temperature at the recording site in area 17 and the VC21a part of area 18 covered by our warming probe, our warming probe did not cover the areas in the immedi- ate proximity of area 21a. Thus, lowering the temperature of area 21a would also result in lowering the temperature of, but not silencing, area 7, the PMLS area and area 19 (Fig. 1). Poly- sensory area 7 can influence area 17 only indirectly (Symonds and Rosenquist, 1984a). Part of the PMLS area corresponding visuotopically to area 21a is located in the deep part of the posterior suprasylvian sulcus, at least 3 mm from our cooling probe (Turlejski and Michalski, 1975; Palmer et al., 1978; Tusa et al., 1981; Djavadian and Harutiunian-Kozak, 1983; Grant and Shipp, 1991; Sherk and Mulligan, 1993). Since the temperature depth gradient measured along the centre of the cooling probe was 5–6°C/mm (Girard and Bullier, 1989; Michalski et al., 1993; Zhang et al., 1996) one can assume that, when area 21a was cooled to 10°C, the PMLS neurons located in the visuotopically corresponding part, although exposed to a temperature lower than physiological (perhaps as low as 30°C), will not be silenced. Visuotopically organized area 19 (Hubel and Wiesel, 1965; Tusa et al., 1979, 1981; Mulligan and Sherk, 1993) is the area most likely to be affected by the cooling probe placed on area 21a. This area is strongly connected both directly and indirectly to area 17 (see Fig. 1), and an appreciable influence of altered activity of area 19 neurons on the area 17 neurons cannot be ruled out.
Overall, although cooling of area 21a might result also in a lowering of temperature, and therefore altered neuronal activity, of several surrounding areas, the effects we observe must be mainly due to silencing area 21a. Thus, (i) only area 21a was cooled to temperatures as low as 5–10°C, that is, to the tem- perature at which neuronal activity ceases; and (ii) neither the ‘spontaneous’ nor the evoked activity of cells recorded from the part of area 17 outside VC21a (but not outside the parts receiving direct inputs from areas 19 or PMLS) were significantly affected by cooling.
Changes in Background Activity and Peak Firing Rate of Area 17 Neurons following Inactivation of Area 21a
For the entire sample of cells located in the VC21a part of area 17, inactivation of area 21a resulted in: (i) a significant reduction in the ‘spontaneous’ (background) activity at the standard back- ground illumination of 1cd/m2; and (ii) a significant reduction in the magnitude of responses to visual stimuli of standard contrast 15. These population effects are supported by significant de- creases in peak discharge rates in a substantial number of cells.
Consistent with the present findings, substantial morpho- logical and functional evidence indicates that at least in primates, carnivores and rodents both intrinsic and extrinsic long- range associational cortico-cortical connections (including the feedback projections to the primary visual cortices) tend to be excitatory [for review see (Salin and Bullier, 1995); see also (Johnson and Burkhalter, 1996, 1997)]. The evidence for excitatory function of long-range associational connections includes: (i) orthodromic, monosynaptic activation or mono- synaptic depolarization of area 17 neurons in cats and rats by electrical stimulation of higher-order visual areas (Bullier et al., 1988; Shao and Burkhalter, 1996); (ii) reduction of the magni- tude of responses to visual stimuli of substantial proportions of cells in area 17 of squirrel monkeys, cats and macaque monkeys following reve