Selective Loss of Orientation Column Maps in Visual Cortex during Brief Elevation of Intraocular Pressure

Abstract

purpose. To compare the orientation column maps elicited by different spatial frequency gratings in cortical area 17 of cats before and during brief elevation of intraocular pressure (IOP).

methods. IOP was elevated by injecting saline into the anterior chamber of a cat’s eye through a syringe needle. The IOP was elevated enough to cause a retinal perfusion pressure (arterial pressure minus IOP) of approximately 30 mm Hg during a brief elevation of IOP. The visual stimulus gratings were varied in spatial frequency, whereas other parameters were kept constant. The orientation column maps of the cortical area 17 were monocularly elicited by drifting gratings of different spatial frequencies and revealed by a brain intrinsic signal optical imaging system. These maps were compared before and during short-term elevation of IOP.

results. The response amplitude of the orientation maps in area 17 decreased during a brief elevation of IOP. This decrease was dependent on the retinal perfusion pressure but not on the absolute IOP. The location of the most visible maps was spatial-frequency dependent. The blurring or loss of the pattern of the orientation maps was most severe when high-spatial-frequency gratings were used and appeared most significantly on the posterior part of the exposed cortex while IOP was elevated. However, the basic patterns of the maps remained unchanged. Changes in cortical signal were not due to changes in the optics of the eye with elevation of IOP.

conclusions. A stable normal IOP is essential for maintaining normal visual cortical functions. During a brief and high elevation of IOP, the cortical processing of high-spatial-frequency visual information was diminished because of a selectively functional decline of the retinogeniculocortical X pathway by a mechanism of retinal circulation origin.

It is well known that maintaining a normal level of intraocular pressure (IOP) is essential in animal and human vision. Elevation of IOP affects visual function by reducing or even completely abolishing the electroretinogram (ERG) and visual evoked potentials (VEPs). ^{1 2 3 4} It also causes retinal ischemia, interruption of axoplasmic transport, and axonal degeneration. ^{5 6 7 8} Abnormalities caused by extreme elevation of IOP may cause blindness. 

In the visual cortex, most cells respond selectively to bars or gratings specifically oriented in the visual field. Neurons with similar preferred orientations form subtle columnar structures extending from the pial surface to white matter in the visual cortex. ^{9 10} However, little is known about the effect of elevated IOP on response properties and functional organization of the visual cortical cells. We have reported that briefly increased IOP leads to preferential loss of X cell activities, both in the retina and thalamus. ^{11 12 13} Because X cells have a higher spatial-frequency preference than Y cells in the retina and the lateral geniculate nucleus, ^{14 15 16 17} we suggest that this underlies the loss of high-spatial-frequency vision in acute angle-closure glaucoma and predict that high-spatial-frequency response of cortical neurons, especially in area 17, should be compromised during elevation of IOP. We report herein the effect of selective retinal loss on the activity of the visual cortex. The orientation maps elicited by different spatial-frequency gratings in the cat’s visual cortex were examined and compared before and during brief elevation of IOP, using the in vivo optical imaging based on intrinsic signals. ^{18 19 20 21}  

Methods

Surgical Preparation

Eight adult cats were used in the experiments. They were first sedated with ketamine (25 mg/kg, intramuscularly), and surgical anesthesia continued with intravenous thiopental sodium (0.4 mg/mL). The tracheal and arterial cannulae were inserted, and all pressure points and incisions were infiltrated with a long-acting anesthetic (1% lidocaine HCl). A mixture of gallamine triethiodide (10 mg/kg per hour) and pentobarbital sodium (3 mg/kg per hour) was continuously infused intravenously to maintain anesthesia and paralysis. Electrocardiogram (ECG), electroencephalograph (EEG), and rectal temperature of the animal were monitored throughout the experiments. The end-tidal CO₂ was measured and maintained at approximately 4% by adjusting the respirator. The pupils were dilated with atropine (0.5%) and the nictitating membrane was retracted with phenylephrine (2%). The refractive errors of the eyes were tested carefully and corrected appropriately with contact lens and 3-mm artificial pupils. The eyes were routinely checked to ensure optimum optical conditions before and during elevation of IOP. In general, the difference was less than 0.50 D between the two conditions. In some cases, additional lenses were used for correcting the eye’s optics, if needed. 

The optic nerve head, retinal vessels, and area centralis of the eye were mapped on the screen of the visual stimulator, by using the fundus reflective projection method. ²² The primary visual cortex at Horsley-Clarke coordinates A1 to P8.7 and L0 to L6 was exposed for optical imaging. A 16-mm diameter stainless chamber was cemented on the skull directly above the exposed cortex. After careful removal of the dura, the chamber was filled with warm silicone oil and sealed with a coverslip. Special care was taken to prevent the oil from leaking. All investigations involving animals adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the Policy of the Society for Neuroscience on the Use of Animals in Neuroscience Research. 

Elevation of IOP and Blood Pressure

It is known that the function of retinal ganglion cells depends on the ocular perfusion pressure (PP) rather than the absolute IOP. ^{6 12} To change the retinal perfusion pressure (PP is mean arterial blood pressure minus IOP), we first measured the femoral arterial blood pressure and IOP in the cats. The method used for elevating IOP has been reported in detail. ¹² In short, to elevate the IOP artificially, a needle was inserted into the anterior chamber of the eye and connected through a cannula to a three-way switch that was connected to a syringe filled with saline and a pressure transducer. IOP was elevated by injecting saline into the anterior chamber of the cat’s eye. Through a transducer and an analog-digital converter, the resultant IOP was displayed on the computer and adjusted manually to maintain a stable pressure within ±5 mm Hg. The mean arterial blood pressure was monitored continuously throughout the experiment by another blood pressure transducer through a needle that connected to the femoral artery. Thus, the ocular PP was readily determined. To minimize the detrimental effects of extended elevation of IOP, a PP of 30 mm Hg for a period of less than 4 minutes was used routinely. 

In some experiments, we elevated the blood pressure while maintaining a normal ocular PP. In two cats, to elevate the animal’s blood pressure, metaraminol bitartrate (Aramine, 0.25 mg/mL, 1–2 mL in 3 minutes; Harvest Pharmaceutical Co., Shanghai, China) was intravenously injected so that the mean blood pressure was raised to a level of 180 to 200 mm Hg, and a normal ocular PP was maintained during elevation of IOP. 

Visual Stimuli

A 30 × 40-cm² monitor (EIZO Nanao, Ishikawa, Japan) was set 57 cm away from the eyes of the animal. The ipsilateral or contralateral eye was stimulated monocularly. Drifting sinusoidal gratings of contrast 0.9 were used. The mean luminance of gratings was 19 cd/m². Various spatial frequencies of gratings from 0.5 to 2.0 cyc/deg were used in different experiments, whereas a 2-Hz temporal frequency was used in all experiments. Only the horizontal and vertical gratings were used to test the effect of high IOP on the functional orientation map. The two differently oriented grating stimuli were randomly presented. 

Optical Imaging

According to the method of Bonhoeffer and Grinvald, ²¹ a slow-scan charge-coupled device (CCD) camera (512 × 512 pixels, 24 × 24 μm/pixel; DALSA, Waterloo, Ontario, Canada) was used to record the optical images of intrinsic signals from the exposed cortex. ²³ Two tandem lenses were mounted on the camera to provide a narrow depth of field. ²⁴ The camera was mounted on a mechanical structure capable of moving three dimensionally with a fine adjustment of 2 μm/division along the z-axis. First, vessel maps were obtained from the cortex illuminated with green light (546 nm, half-peak width 20 nm), and then orientation maps were obtained with red light (640 nm, half-peak width 15 nm). Generally, the focus image plane was positioned at 450 μm below the surface of the cortex, and the focus depth of the plane was approximately 150 μm. For each trial, the visual stimulus was presented for 2 seconds followed by a 10-second interval, during which a uniform screen of mean luminance was presented. The cortical optical images, each of which had five frames in one trial, were captured from 1 second before the stimulation to 2 seconds after. 

Data Collection and Analysis of Function Maps

Data analysis was mostly performed with one data- and image-analysis program (MatLab; The MathWorks, Natick, MA). The first step was the so-called first-frame analysis, which was to subtract all five frames from the first frame to reduce slow-wave noise. The strongest optical signals in the visual cortex were always produced 3 to 4 seconds after the onset of the stimulus, ^{20 21} and therefore only the fourth frame was taken for analysis. We then averaged all the fourth frames obtained at the repeatedly identical stimuli to produce the resultant orientation map. 

The orientation maps were recorded into two blocks of eight averaged trials (eight trials at 0° stimulation and eight trials at 90° stimulation, for a total of approximately 4 minutes) under normal IOP and during elevation of IOP, alternatively. To reduce the random noise, orientation maps were repeatedly recorded and averaged. Generally, averaging of a total of 64 trials (approximately 16 minutes) recorded was enough to produce a clear orientation map. For recording during elevation of IOP, a 15-minute interval under normal IOP between two 16-trial blocks was routinely used to avoid retinal damage. 

The subtracted 90° to 0° functional map was used for data analysis based on the fact that any two orthogonal orientation maps are spatially complementary to each other. ^{20 21 23} To quantify the degree of orientation selectivity, the response amplitude of an orientation map was defined as the averaged contrast of a raw orientation map. The contrast was defined as (L _white −L _black)/(L _white + L _black), where L is the luminance of the white patch or the black. The contrast was calculated for each of six to nine paired black-and-white circular areas located in the middle of each patch (a pair of areas denoted by two arrows in Fig. 2A ). The location of the paired regions were randomly chosen from a map obtained under normal IOP, and their width was approximately half of a white or black patch (approximately 250 μm in diameter). These identical six to nine paired circular areas were fixed for a cortical area and used for calculating the mean contrast—that is, the response amplitude of each map. Only original data without equalization were used for the quantitative analysis of contrast and cross correlation. 

For sharpness of orientation maps, the high-pass and low-pass filtering was performed for the 90° to 0° map. Then, the histogram equalization technique of the data-analysis program (Matlab; The MathWorks) was used to enhance the contrast of the map in a nonlinear way. In detail, all the pixels in each orientation map were equally divided into 256 groups according to their brightness values. Each pixel in the brightest group was given a relative value of 256 and that in the lowest one a relative value of 1. The relative values of all pixels in the orientation map were used to produce a clearer map. In this way, the map contrast would be sharpened while keeping the visible patterns unchanged. We used this method only for map display, not for quantitative analysis. 

To examine the similarity of two orientation maps, a two-dimensional cross-correlation analysis was used. ²⁵ We first chose a fixed square area (3 × 3 mm), as shown in Fig. 3A , without obvious vessel patterns on the visual cortex in one orientation map (control), and then varied the corresponding position of this same square area in another comparison orientation map (such as the elevated-IOP one). The varied maximum distances in both x- and y-axes were from −30 to 30 pixels (equivalent to from −720 to 720 μm) away from the original position, which means that the bin shifted across 30 pixels in either direction in 1-pixel steps, and the size of the cross-correlation maps was 60 × 60 pixels. For each varied position, the cross-correlation coefficient (CCC) of the square areas for the tested two maps was calculated as follows:  

\[CCC{=}\ \frac{{{\sum}_{i{=}1}^{n}}\ {{\sum}_{j{=}1}^{m}}P1(i,j){\times}P2(i,j)}{({[}\ {{\sum}_{i{=}1}^{n}}\ {{\sum}_{j{=}1}^{m}}P1(i,j){\times}P1(i,j){]}{\times}{[}\ {{\sum}_{i{=}1}^{n}}\ {{\sum}_{j{=}1}^{m}}P2(i,j){\times}P2(i,j){]})^{1/2}}\]

where n and m are the length and width pixels, respectively, of the outlined square matrix; P1 is the fixed square luminance matrix outlined from the control map; and P2 is the moving square luminance matrix in a certain position from the tested map, such as an elevated-IOP map. 

The CCC ranges from −1.0 to 1.0; the higher the CCC, the more similar the two maps. If the two maps do not have any correlation, their CCC is zero. Two spatially complementary maps have a CCC of −1.0, and the identical maps a CCC of 1.0. Finally, we obtained a 60 × 60-pixel matrix of CCCs. The maximum CCC among them is shown in the results. 

Results

IOP and Response Amplitude of the Orientation Maps

The response of the ERG in the cat decreases with elevation of IOP and reduction of ocular PP and ceases at a critical PP of approximately 20 mm Hg. ⁴ In our optical imaging experiments, the response amplitude of orientation maps decreased with elevation of IOP monotonically in optical imaging. In the maps shown in Fig. 1 , the orientation map became blurred, and the response amplitude decreased significantly to 73% (Figs. 1D 1H) of normal (Fig. 1A) when the IOP was 60 mm Hg (PP = 60 mm Hg; t-test, P < 0.05). The response amplitude (55%, Figs. 1E 1H ) decreased more significantly as IOP increased to 90 mm Hg (PP = 30 mm Hg) compared with normal (Fig. 1B ; t-test, P < 0.05). The map disappeared completely when IOP reached 110 mm Hg (PP = 10 mm Hg; Figs. 1F 1H ), showing a significant difference from normal (Fig. 1C ; t-test, P < 0.01). The higher the IOP, the lower the response amplitude. 

Spatial-Frequency-Sensitive Visual Cortex

The response amplitude of different areas of an orientation map varied in accordance with the spatial frequency of the grating stimuli used. The clearest part of orientation maps shifted from the anterior side to the posterior of the visual cortex with increasing grating spatial frequency, as shown in Figs. 2A 2B 2C 2D 2E . A complete map could be seen clearly in the entire cortex exposed (Fig. 2A) as elicited by the lowest spatial frequency (0.5 cyc/deg), whereas the interdigitating orientation pattern appeared in only approximately half the area (Fig. 2C) with a spatial frequency of 1.2 cyc/deg. The map elicited by 2.0 cyc/deg could be detected only in a small area near the posterior corner of area 17 (Fig. 2E) . A previous electrophysiological study showed that the preferred spatial frequencies of almost all cells in area 17 were higher than 0.3 cyc/deg, and those in area 18 were less than 0.3 cyc/deg. ²⁶ This observation was also confirmed by optical imaging studies. ^{27 28 29} Therefore, in the experiments, spatial frequencies of 0.5 cyc/deg and higher were always chosen to assure that we were studying cortical area 17. 

To quantify this observation, the exposed cortex was divided by two dotted lines (Fig. 2A) into three parts from the anterior to the posterior for measuring the spatial frequency versus response amplitude curves. As shown in Figure 2G , the anterior side of the cortex is more sensitive to the low-spatial-frequency gratings and less sensitive to the high ones. The posterior side of the cortex is more sensitive to the high-spatial-frequency gratings and less sensitive to the lower ones, and the middle part more sensitive to the medial frequencies in between. This spatial frequency sensitivity of visible orientation maps was location dependent and was observed in all cats studied. 

Selective Loss of Maps Elicited by Different Gratings during Elevation of IOP

Although the basic pattern of visible maps in the visual cortex did not change during a brief period of elevated IOP, some selective losses of orientation maps were repeatedly found when animals were stimulated with different spatial-frequency gratings. A typical case is shown in Fig. 3 . The response amplitudes of all orientation maps elicited by gratings at spatial frequencies of 0.5, 0.8, and 1.2 cyc/deg were lower (Figs. 3B 3D 3F) compared with normal (Figs. 3A 3C 3E) . The response amplitude clearly depended on spatial frequency. The orientation map of 1.2 cyc/deg (Fig. 3F) completely disappeared when IOP was elevated, showing a relative response of approximately 30%, which was significantly different from normal (t-test, P < 0.001). The response amplitudes of maps obtained with 0.8 cyc/deg (Fig. 3D) and 0.5 cyc/deg (Fig. 3B) declined during elevation of IOP, and both showed relative responses of approximately 65%, which differed significantly from normal (t-test, both P < 0.0001) and from that of 1.2 cyc/deg during elevation of IOP as well (Fig. 3F , t-test, both P < 0.001). Fig. 3G shows that elevation of IOP always caused the heaviest loss of orientation maps of the highest spatial frequency at the most posterior position. Similar phenomena were observed in all the cats measured. 

Similarity and Stability of Orientation Maps during Elevation of IOP

The CCC was used to analyze the comparability and similarity of the orientation maps of the same cortical area under normal IOP (Figs. 3A 3C 3E) and while IOP was elevated (Figs. 3B 3D 3F) . The same fixed square areas denoted by dotted lines on the maps under normal IOP (Fig. 3A) were compared with the moving ones in the corresponding maps when IOP was elevated. The maximum CCCs of Figures 3A and 3B and 3C and 3D , in the identical corresponding areas were 0.593 and 0.663, respectively, indicating no significant difference in basic pattern between maps of the normal and elevated IOP. In contrast, the maximum CCC for Figures 3E and 3F was 0.179, indicating that the two maps were very different because of loss of the activated areas. 

Effect of Increasing Blood Pressure on Maps during Elevation of IOP

In two cats, we compared the orientation map obtained under the elevated IOP and high blood pressure condition with that under elevated IOP alone. Figure 4 shows a typical example. The high IOP (IOP = 110 mm Hg) reduced retinal PP (10 mm Hg) made the orientation map disappear (Fig. 4B) and, significantly, the response amplitude decreased to zero (Fig. 4E ; t-test, P < 10⁻⁶). However, when the blood pressure was elevated to 200 mm Hg by an intravenous injection of metaraminol bitartrate, which left the retinal PP near normal level (PP = 90 mm Hg), the orientation map (Fig. 4D) remained despite the elevated IOP (110 mm Hg). The relative response was 71% of normal. In comparison with the virtual abolishment under elevated IOP with normal blood pressure (Fig. 4B) , the response shown in map D was significantly greater than that in map B (t-test, P < 10⁻⁵) but was still less than the control map C (t-test, P < 0.05). Again, it is the retinal perfusion pressure, but not the absolute IOP that causes the effect that we observed of elevated IOP on cortical function. 

Discussion

Using the intrinsic signal method, we have demonstrated for the first time that a selective loss of orientation maps elicited by relatively high-spatial-frequency gratings occurred in the visual cortex of the cat during short-term elevation of IOP and that all the orientation maps decreased in response amplitude. In a previous study we showed that the X-type ganglion cells in the retina and relay cells in the dorsal geniculate nucleus of the cat are more sensitive to brief elevation of IOP than the Y cells. ^{11 12 13} All these studies demonstrate that both the subcortical structures and the cortical area responsible for high-spatial-frequency preferentially lose their fine visual function during brief elevation of IOP. The present results thus suggest how differential cortical function is impaired by brief elevation of IOP that affects different classes of retinal ganglion cells. 

In the retina, the functional X-type ganglion cells correspond to the morphologic β-type in cats. The density of X ganglion cells is highest in the retinal area centralis and decreases drastically with the increase in retinal eccentricity, whereas that of Y cells, corresponding to α cells in cats, changes relatively little with eccentricity. ^{14 15 16 17 30} Furthermore, X cells possess the smallest receptive field center, the smallest dendritic field, and medium-caliber axons as well, whereas Y cells have the largest center and dendritic field and the thickest axons. ^{14 15 16 17 31} Because of these and other behavioral studies, it has been hypothesized (e.g., see Sherman and Spear ³² ) that X cells concentrated in the center of the retina are responsible for the fine vision in the cat and Y cells for gross vision and motion discrimination. The part of area 17 representing the central retinal projection in the optical image prefers a grating of higher spatial frequency, whereas the periphery representation prefers a lower one. This is consistent with the finding that the clearest orientation map of high spatial frequency appeared near the central retinal projection in the cortex (Fig. 2) and weakened severely during elevation of IOP, whereas the map of low frequency remained less affected (Fig. 3) . Therefore, we hypothesize that the preferred decline in X cell function in the retina leads to the selective loss of orientation maps during elevation of IOP. 

In the experiments, the posterior surface of the occipital lobe was always exposed to an obliquely oriented CCD camera (approximately 30° to the vertical axis) for optical imaging. The observed orientation maps mostly reflect the input of the superior retina that receives the information from the inferior visual field. The more posterior on the orientation map, the closer to the central projection of the retina. This explains why the selective loss of the map observed during elevation of IOP always appeared on the more posterior part of the cortex that preferred the higher spatial frequency. The central part of retinotopic topography we observed was usually displaced approximately 3 to 4 mm posterior from that reported by Tusa et al. ³³ in most of the cats (17/20) in their study. However, they found a similar displacement in a minority of the cats (3/20). This systematic displacement we observed may be due to different species of cats used. 

The cat’s area 17 receives X and Y inputs from the lateral geniculate nucleus, whereas area 18 receives predominantly Y input. This should strongly support the differential effects of elevated IOP on the X and Y pathways shown in our observations as well when the imaging area also includes a part of area 18. An optical imaging study by Issa et al. ²⁷ showed that in area 17, the median preferred spatial frequency was approximately 0.5 cyc/deg, which was more than double that of area 18 (the median preferred spatial frequency, 0.18 cyc/deg). Other groups have recently reported similar findings. ^{28 29} Furthermore, cells in area 17 respond preferentially to spatial frequencies greater than 0.3 cyc/deg, whereas those in area 18 prefer less than 0.3 cyc/deg. ²⁶ In our experiments, all the spatial frequencies used were greater than or equal to 0.5 cyc/deg. Thus, the orientation maps observed herein should be primarily in area 17 and possibly somewhat in area 18. The selective loss during elevation of IOP seems mainly due to a different projection of the inputs between the X and Y pathways to area 17 and partially due to the differentiation in spatial frequency between areas 17 and 18. 

In two cats, we observed that the loss of orientation map caused by elevation of IOP was prevented by increasing the animals’ blood pressure. Therefore, the effect of elevation of IOP on orientation maps also depends on the retinal perfusion pressure, but not absolute IOP, as did the retinal ganglion cells. ^{6 12} Therefore, the study provides functional evidence at the cortical level to support the vasogenic hypothesis that during acute elevation of IOP, retinal ischemia may be the most critical factor, rather than the direct mechanical effect on the ganglion cell, per se. 

There still is the possibility that optical blurring induced by elevated IOP could be the cause of the spatial-frequency-selective loss of cortical function observed. However, it seems unlikely. First, we checked the eye’s optics repeatedly to ensure the experiment was performed under good optical conditions. Second, as shown in our previous study, the on-center ganglion cells are more sensitive to elevation of IOP than the off-center cells in the cat’s retina as well. ¹² The on- and off-center cells are evenly distributed in the retina. ^{34 35} Third, increasing blood pressure prevents the orientation map from degrading, which suggests a retinal vasogenic mechanism rather than blurring of optics in the eye. Overall, eye blurring is unlikely to be the mechanism of the effect we observed. 

It interested us that the identical basic pattern of the visible orientation maps elicited by the same grating could be maintained as long as 6 days, regardless of whether IOP was elevated. This clearly indicates that the orientation column has a rather stable functional organization, as shown by Chapman et al. ³⁶ in their long-term optical imaging study of ferrets during development. The orientation selectivity of visual cortical neurons have been reported to have origins in the retina and the lateral geniculate nucleus of the cat. ^{37 38 39 40 41} However, early visual deprivation decreases the orientation sensitivity of visual cortical cells in the cat significantly, but does not affect that of relay cells in the lateral geniculate nucleus. ⁴² Furthermore, silencing on-center retinal ganglion cells during development also affects the form of orientation maps. ⁴³ Therefore, one possible implication is that high IOP has extremely detrimental and long-lasting effects on human vision, especially during the critical period of cortical development. Medical treatment early in development may be needed to protect children from loss of high-spatial-frequency vision. 

The brief elevation of IOP used in the current experiments was rather high (approximately 100 mm Hg; i.e., 30 mm Hg lower than mean arterial pressure). The animal’s eye condition was similar to that of acute angle-closure glaucoma at the breaking-out stage, which results in a sharp decrease of visual acuity and even rapidly causes blindness. Thus, this study provides a two-dimensional orientation map of the primary visual cortex in the cat during brief, sharp elevation of IOP, and creates a cortical model for studying acute angle-closure glaucoma. 

 Supported by grants from the National Natural Science Foundation of China, Chinese Academy of Sciences, Ministry of Education and Shanghai-Unilever Research and Fund.

 Submitted for publication February 25, 2002; revised May 30, 2002; accepted July 8, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Functional orientation maps of the visual cortex under normal and elevated IOP. (A–C) Control maps with normal IOP (20 mm Hg) and perfusion pressure (PP = 100 mm Hg). (D–F) The orientation maps with elevated IOPs of 60 mm Hg (PP = 60 mm Hg), 90 mm Hg (PP = 30 mm Hg), and 110 mm Hg (PP = 10 mm Hg), respectively. (G) Blood vessel map on the surface of a part of the visual cortex studied. (H) Relative response values (mean ± SD) of maps in (D), (E), and (F), left to right, compared with the control. The response decreased monotonically with increasing IOP. The animal’s blood pressure was 120 mm Hg during the experiments. The center of all maps was located at Horsley-Clarke coordinates P3.5 and L2.0. The grating spatial frequency used was 0.5 cyc/deg. A, anterior; L, lateral; Scale bar, 2 mm.

Figure 2.

 

Functional orientation maps in the visual cortex elicited by various gratings of different spatial frequencies. The spatial frequency was 0.5, 0.8, 1.2, 1.6, and 2.0 cyc/deg in (A), (B), (C), (D), and (E), respectively. (F) Blood vessel map of the same cortex studied. (A, arrows) A pair of circular areas located in the centers of black-and-white patches was chosen to calculate a sample response amplitude (contrast). The position of the clearest orientation map shown was spatial-frequency dependent. The higher the spatial frequency, the more to the posterior the well-activated area was located; the lower the spatial frequency, the more anterior it was. (G) Three spatial-frequency-tuning curves of the different parts of the cortex, which were divided by two dotted lines from the anterior to posterior as shown in (A). The y-axis represents the response amplitude (i.e., mean contrast ± SD) of each part of the cortex. The center of all maps was located at Horsley-Clarke coordinates P3.0 and L1.8. A, anterior; L, lateral; Scale bar, 2 mm.

Figure 3.

 

Selective loss of orientation map elicited by high-spatial-frequency grating during elevation of IOP. (A, C, E) Orientation maps obtained under normal IOP. (B, D, F) Maps obtained with IOP elevated to 80 mm Hg (PP = 30 mm Hg). Three different stimulating gratings of spatial frequencies 0.5, 0.8, and 1.2, cyc/deg were used for maps (A, B), (C, D), and (E, F), respectively. (A) Dotted square: the 3 × 3-mm area for calculation of the CCC. (G) Relative responses (mean ± SD) of the whole, the anterior part, and the posterior parts of maps (B), (D), and (F), elicited by different spatial frequencies during elevation of IOP. Brief elevation of IOP resulted in the most severe decrease in signal on the high-spatial-frequency map, although the relative response of all the maps decreased. The center of all maps was located at Horsley-Clarke coordinates P3.2 and L2.3. A, anterior, L, lateral; Scale bar, 2 mm.

Figure 4.

 

Survival of the orientation map by increasing blood pressure while IOP was elevated. (A, C) Orientation maps of a cat with normal IOP. (B) Map in cat with an IOP of 110 mm Hg (PP = 10 mm Hg). (D) Map obtained when the blood pressure was pharmacologically increased to 200 mm Hg with an IOP of 110 mm Hg (PP = 90 mm Hg). (E) Response amplitudes (mean ± SD) of orientation maps (A–D) in various conditions. In condition D, the loss of signal during elevation of IOP was clearly prevented by the increase of blood pressure that results in keeping a normal ocular PP. The response amplitude of map D was approximately 71% of normal (C), as shown in (E). The center of all maps was located at Horsley-Clarke coordinates P3.2 and L1.8. A, anterior, L, lateral; Scale bar, 2 mm.

The authors thank Deming Su and Jason Clower for reviewing and editing the manuscript.