- Email: [email protected]
Photoreceptor topography of the retina in the New World monkey Cebus apella
Vision Research 40 (2000) 2395 – 2409 www.elsevier.com/locate/visres
Photoreceptor topography of the retina in the New World monkey Cebus apella Belmira Lara S. Andrade da Costa a, Jan Nora Hokoc¸ b,* a
Departamento de Fisiologia e Farmacologia, Centro de Cieˆncias Biolo´gicas, Uni6ersidade Federal de Pernambuco, Pernambuco, Brazil b Departamento de Neurobiologia, Instituto de Biofisica Carlos Chagas Filho, Centro de Cieˆncias da Sau´de, Bloco G, Uni6ersidade Federal do Rio de Janeiro, 21949 -900 Rio de Janeiro, Brazil Received 26 August 1996; received in revised form 25 February 2000
Abstract The number and topographical distribution of photoreceptors was studied in whole-mounted retinas of Cebus apella. It was estimated a total of 48 million rods and 3.8 million cones. The average peak foveal cone density and the Nyquist Limit at the foveola were estimated as 169, 127 cells/mm2 and 46.77 97.98 cyc/deg, respectively. A cone-enriched rim was found near the ora serrata, more noticeable in the nasal retina. Rod distribution was asymmetrical along horizontal and vertical meridians with a higher density in the dorsal retina. The rod/cone ratio was variable and asymmetrical along both meridians. © 2000 Elsevier Science Ltd. All rights reserved. Keywords: Primates; Visual system; Fovea; Photopic vision; Scotopic vision
1. Introduction In primates, the first study on photoreceptors mosaic was done by Osterberg (1935) in the human retina. Since then, many researchers, using various histological methods and a wide range of ages, have shown a great variability in the number of photoreceptors in various species of Old World primates (Rolls & Cowey, 1970; Young, 1971; Adams, Perez & Hawthorne, 1974; Wa¨ssle & Riemann, 1978; Miller, 1979; Borwein, Borwein, Medeiros & McGowan, 1980; Perry & Cowey, 1985; Hirsch & Miller, 1987; Krebs & Krebs, 1987, 1989; Schein, 1988; Wa¨ssle, Gru¨nert, Rohrenbeck & Boycott, 1989). Curcio, Packer and Kalina (1987) introduced a new method to visualise photoreceptors that minimises tissue shrinkage. As a consequence, rod and cone distribution of many species of Old World primates were re-examined (Packer, Hendrickson & Curcio, 1989; Curcio, Sloan, Kalina & Hendrickson, 1990; Wickler & Rakic, 1990; Wickler, Williams & Rakic, 1990). * Corresponding author. Tel.: +55-21-2906897; fax: + 55-212808193.
In New World primates, comparative studies have been performed on the nocturnal specie Aotus tri6irgatus, (Ogden, 1975; Wickler & Rakic, 1990) and on the diurnal species Saimiri sciureus and Callithrix jacchus (Rolls & Cowey, 1970; Troilo, Howland & Judge, 1993; Wilder, Gru¨nert, Lee & Martin, 1996). In the diurnal capuchin monkey Cebus apella, rod and cone density were also described, but only along the horizontal meridian (Silveira, Yamada, Perry & Picanc¸o-Diniz, 1994; Yamada, Silveira & Perry, 1996). In a more complete study, we aim in the present paper, to analyse spatial distribution, mosaic organisation and morphometric parameters of rods and cones along the entire retina of the Cebus apella. Cebus is a diurnal New World monkey that has been widely used as a suitable experimental model for comparative studies of the visual system at retinal (Silveira, Picanc¸o-Diniz, Sampaio & Oswaldo-Cruz, 1989; Lima, Silveira & Perry, 1993; Silveira et al., 1994; Lima, Silveira & Perry, 1996; Yamada et al., 1996; Andrade da Costa, Hokoc¸, Pinaud & Gattass, 1997; Guimara˜es & Hokoc¸, 1997; Pessoa, Tavares, Aguiar, Gomes & Tomaz, 1997; Silveira, Lee, Yamada, Kremers & Hunt, 1998; Jacobs, 1999; Martin & Gru¨nert, 1999; Silveira, Yamada, Kremers, Hunt, Martin & Gomes 1999); and cortical (Hess
0042-6989/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 2 - 6 9 8 9 ( 0 0 ) 0 0 1 0 4 - 8
2396
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
& Edwards, 1987; Gattass, Rosa, Sousa, Pinon, Fiorani & Neuenschwander, 1990 for revision; Amorim & Picanc¸o-Diniz, 1997; Pinon, Gattass & Sousa, 1998), levels. Using the unstained retinal wholemount preparation introduced by Curcio et al. (1987), we analysed qualitative and quantitative parameters of the photoreceptor organisation in Cebus apella retina in order to characterise aspects of the visual function of this monkey and compare them to those reported for other diurnal Old and New World primates.
2. Methods
2.1. Animals and tissue preparation Eyes from seven adults, four males and three females of the species Cebus apella were studied. Procedures for the use of animals were in accord with the ‘Principles of Laboratory animal care’ (NIH) and approved by the commission of animal care of the Instituto de Biofı´sica/ UFRJ. The animals weighed between 1.8 – 2.5 kg and were kept on a 12h light/dark cycle. They were also used for other experiments not correlated with the present study. The anaesthesia was induced first with Ketamine chloride, 50 mg/kg, i.m. and supplemented with sodium pentobarbital (20 mg/kg, i.v.). One eye from each animal was enucleated under deep anaesthesia, except for one animal from which both retinas were taken to check for inter-eye variability. Eyes were fixed by immersion in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) and a small amount of the fixative solution was injected into the vitreal chamber to keep the globe taut and prevent permanent folds in the foveal area (Curcio et al., 1987). At the end of the experiments, the animals were sacrificed by an overdose of the anaesthetic. After being fixated for one hour, the eyes were bisected at the equator and the cornea and lenses removed. The retinas were then dissected in phosphate buffer (0.1 M, pH 7.4) whole-mounted on a slide with the photoreceptor layer up and coverslipped with DMSO for 24 h. The next day, the coverslip was removed, DMSO was blotted and the retinas were permanently coverslipped with 100% glycerol and sealed with clear nail polish. The final areal expansion during processing was estimated by comparing drawings of the tissue outlined while in phosphate buffer and at several different stages during processing. One eye from an adult Cebus apella was used as a standard for the measurement of shrinkage and expansion during processing. The retina and pars plana of the ciliary body were dissected and flattened on a glass slide. An image of the retina contour was drawn by projecting through a photographic mag-
nifier and the distances from the optic disc to the fovea as well as to the temporal, nasal, dorsal and ventral edges of the retina were determined at each step of the procedure. There was a 2% linear variation due to the fixative. No correction was made for this slight shrinkage. Any retina with central or peripheral deformation was discarded.
2.2. Image acquisition Whole-mounted retinas were observed using Normarski Differential Interference Contrast microscopy (NDIC) with the aid of an Axioplan Zeiss optical microscope equipped with a high resolution video camera (GRUNDIG, Pal-G, 625/50 Hz). Images of the photoreceptor inner segments were acquired and intensified in brightness and contrast using a monitor ACEZeiss Microsystems. The video camera was attached to a microcomputer supplied with the software IBAS v. 2.0 (Interaktives Bild Analysen System, Zeiss), where selected images could be stored and analysed by more than one individual for cell counts and measurements of morphometric parameters.
2.3. Sampling procedure In and around the fovea, within the first 0.8 mm of eccentricity, photoreceptor arrays were sampled at 100 mm of distance. Starting from 1 mm of the foveola throughout the retina toward the periphery, sampling areas were photographed at 1–2 mm intervals. An 100× oil immersion objective and an intermediate lens of 2.0 × were used providing a final magnification of 3500× for images in the foveal region and of 2200× for images in the mid-periphery (with an intermediate lens, 1.25× ). The diameter of the inner segment, larger in cones than in rods, was used as a morphological criterion for differentiation. An important parameter for estimating the peak cone density is the size of the sampling window. It is known that a large sampling window for the foveal region may underestimate the peak cone density by 20% or more (Packer et al., 1989), especially because cone density declines sharply as you move away from the foveal centre. We tested several sizes of sampling window for the foveal region and adopted one measuring 40×40 mm, which revealed minimal variation in the morphometric parameters, cone density and inter-cone spacing. Sampling areas of 1600 mm2 were used to count cones and rods in the central retina. From 1 mm of the foveola and throughout the whole extension of the retina, rods were counted in sampling areas of 4049 mm2 (63.6× 63.6 mm). In mid-periphery, the cones were counted using a 40× objective in sampling areas of 26 500 mm2 (162.7×162.7 mm). Alternatively, several
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
adjacent windows of 4049 mm2 were used to obtain a minimum of 30 cones per sampling area. In the far periphery of the retina, adjacent to the ora serrata, cones were counted at 500 mm steps. The location of each sampling area was marked on a drawing of the retina and the X – Y coordinates of the microscope for each region recorded. Approximately 200 areas were sampled in each retina. Isodensity lines were drawn, linking the areas of similar density starting from the point of highest density (fovea). The centre of the fovea was used as a reference point for measurements of eccentricity. The total number of photoreceptors was calculated for each retina by multiplying the area between two adjacent lines by the average cell density for both lines. These values were added to give the total number of photoreceptors for each retina. The areas and diameters of cone inner segments at different eccentricities were measured directly from video displayed images, using a mouse to trace cell contours. These contours were printed, scanned and selected for mosaic analysis. Small routines were built for morphometric analysis and added to the IBAS system. For the conversion of eccentricity from mm to degrees we adopted the Retinal Magnification Factor (RMF) of 197.3 mm/deg estimated by Silveira et al. (1989), which was based on a posterior nodal distance (PND) of 11.3 mm for the Cebus eye. This value was useful within the central 3° of the retina. For eccentricities higher than 5°, the function obtained by Yamada et al. (1996), where RMF = 205−0.48x, (x, angular distance in degrees), was employed.
2.4. Mosaic analysis Starting from the foveola, 18 contiguous sampling areas (1600 mm2), at the photoreceptor inner segments’ level, were used to measure the regularity of the cone mosaic in the central retina. Outlines of the cones were stored and the centroids X and Y for each cell were obtained using NIH-Image version 5.5. Measures of mosaic regularity were performed using the software Program Spatial Point Patterns (PSPP), version 2.2 (a new version of Ferna´ndez, Cuenca & De Juan, 1993). This software compares the mosaic organisation of each sample to a random, to a regular or to a clustered distribution, and allows to obtain statistical indexes that correspond to each distribution. Using the PSPP we obtained the average nearest neighbour distance and standard deviation for each window. Center-to-centre distances between nearest
2397
neighbours were used to estimate the spatial resolution in each retinal sampling area at different eccentricities.
2.5. Estimate of spatial resolution in central retina According to the sampling theorem (Shannon & Weaver, 1949), density and packing mode of cones in the photoreceptor array determine the highest spatial frequency that can be detected unambiguously. As many authors have shown, in human and monkey retinas, the foveal photoreceptor lattice is packed according to a hexagonal array (Miller, 1979; Borwein et al., 1980; Hirsch & Hylton, 1984). In this array, the highest spatial frequency (referred as the Nyquist Limit) for each window sampling is 2/ 3 (1/2 dcc), where dcc corresponds to inter-cone distance in degrees of visual angle (Snyder & Miller, 1977; Williams, 1986). Assuming a perfect hexagonal packaging, the Nyquist Limit can also be estimated considering the RMF and cone density by using the following relation: F · Nyquist= (0.5) · (RMF) · [(2/ 3) · dmax]1/2, where dmax is the maximum density of cones. We used both formulas to estimate the Nyquist Limit in each sampling window in the fovea within the first three central degrees of eccentricity.
3. Results
3.1. Area of the retinas The mean area measured in fixed, non-dehydrated whole-mounted retinas obtained from seven adult Cebus monkey was 601.79 23.9 mm2. The average distance from the centre of the fovea to the centre of the optic disc was 3.279 0.23 mm and between the fovea and the retinal margins was 17.29 0.48 mm along the nasal axis and 13.39 0.38 mm along the temporal axis (Table 1).
3.2. Estimate of rod and cone number We estimated a total of 489 1.62 million of rods and 3.89 0.13 million of cones in the Cebus retina (Table 2). In one animal, from which both retinas were taken to be measured, the difference in the number of cones and rods between the two eyes was less than 10%. Fig. 1A shows a representative map of isodensity lines for cones for one whole retina. The contours had a horizontally elongated appearance forming a streak of cones. The average total number of rods and cones was similar in males and females, indicating no significant
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
2398
difference between sexes. On the other hand, the regional analysis of cone number showed differences of approximately 20% in the fovea of individual monkeys. Fig. 1B is the magnification of foveal region of the retina of two different animals (B1=CB203 and B2= CB131) to illustrate the isodensity lines within this region. Retinal area within line a (20 000 cells/mm2) of both animals was the same (0.919 mm2 in B1 and 1.056 mm2 in B2), however, there was a difference of 18% in the total number of cones for these animals: 32 335 cones for B1 and 39 247 cones for B2. This finding is possibly due to the difference observed in the peak density of the foveolas (87 500 cones/mm2 in B1, and 281 887 cones/mm2 in B2). Photoreceptor mosaic at different eccentricities along the horizontal meridian is illustrated in Fig. 2. It is evident from this figure that there is an eccentricity-dependent variation of photoreceptor density and size. Rods first intrude into the cone mosaic at 150 mm from the centre of the foveola (Fig. 2A) and reach a peak density at eccentricities near the optical disc (Fig. 2D).
3.3. Topographic distribution of cones 3.3.1. Fo6ea Among the animals, there was a great variability in peak cone density (Table 2), with values ranging from 87 500 to 281 887 cells/mm2 at the foveola. The distribution of cone density within the central 0.7 mm (about 3.5°) of eccentricity along the horizontal meridian is seen in Fig. 3. The average peak cone density of four animals was estimated as 173 909974 139 cells/mm2. This peak value decreased approximately 60% within 40–60 mm from the centre of the fovea and was about 20 000 cones/mm2 at 0.7 mm of eccentricity. As shown in the inset of Fig. 3, the best logarithmic function for cone density along the temporal horizontal meridian is: Y= − 22.588 ln(x)+45.277
(R 2 = 0.9949)
where Y is the cone density and x is eccentricity in degree.
Table 1 Morphometric dimensions Animal
Animal weight (kg)
Retinal area (mm2)
OD-nasal distance (mm)
OD-fovea distance (mm)
OD-temp distance (mm)
CB229(LE)a CB229(RE)a CB255a CB131b CB129a CB232b CB203a CB579b
2.250
609.5 632.6 551.2 581.8 593.4 615.0 614.0 616.2
18.1 17.8 17.5 17.0 17.0 17.0 16.5 16.8
3.5 3.1 3.7 3.0 3.3 3.1 3.2
13.1 14.2 13.0 13.5 13.0 13.0 13.5 13.2
601.71 23.89
17.2 0.48
3.27 0.23
13.31 0.38
2.200 1.900 2.100 1.800 2.500 2.200
Mean SD a b
Male. Female.
Table 2 Estimate number of photoreceptors Animal
Cone number
Rod number
Peak cone density
Cone centre–peripheral gradient
CB229(LE)a CB229(RE)a CB255a CB131b CB129a CB232b CB203a CB579b
3 872 515 3 900 200 3 914 940 3 579 970 3 850 000 na na na
48 900 350 49 400 250 49 883 525 45 356 246 47 619 885 na na na
nac na na 281 887 150 000 199 375 87 500 126 875
na na na 62.63 33.33 44.3 19.4 28.19
Mean SD
3 828 925 128 199
48 232 051 1 623 802
169 127 66 998
37.57 14.89
a
Male. Female. c na, not analysed. b
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
2399
Fig. 1. (A) Map of cone density in a Cebus monkey retina (area =619 mm2). Optic disc and centre of the fovea are represented by a filled circle and an asterisk, respectively. Isodensity lines correspond to: A, 5000 cells/mm2; B, 6000 cells/ mm2; C, 8000 cells/mm2; D, 10 000 cells/mm2; E, 15 000 cells/mm2 and F, 25 000 cells/mm2. V, ventral; D, dorsal; T, temporal; N, nasal. Bar =4 mm. (B) Detail of foveal cone isodensity lines in two retinas that differ in peak density by a factor of 3.2. In B1 the peak density corresponds to 87 500 cells/mm2 (animal CB203, male, from Table 1) and in B2 it corresponds to 282 000 cells/mm2 (animal CB131, female, from Table 2). Characters: a, 20 000 cells/mm2; b, 30 000 cells/mm2; c, 40 000 cells/mm2; d, 50 000 cells/mm2; e, 60 000 cells/mm2; f, 70 000 cells/mm2; i, 100 000 cells/mm2. The total number of cones estimated in the fovea is 32 335 cones (B1) and 39 247 cones (B2). Bar = 500 mm.
Cone distribution across the horizontal meridian revealed a naso-temporal asymmetry, within 0.7 mm from the foveola, which was evident in some but not in all retinas. When apparent, this asymmetry was more conspicuous around 0.2 mm from the foveola, where the nasal density was about twice as the temporal density (Fig. 3). No association with sex or peak cone density was found.
3.3.2. Mid and far periphery In the mid-periphery cone density declined linearly with densities falling by approximately 500 cones/mm2 per mm along all axes (Fig. 4). Minimal densities
ranged among 4000–5000 cells/mm2 in five animals. The centre–peripheral gradient of cones ranged from 19.4X to 62X and was directed related to the peak density at the foveola (Table 2). Analysis of cone density distribution along the vertical (Fig. 4A) and horizontal (Fig. 4B) meridians shows a naso-temporal asymmetry along the horizontal meridian. This asymmetry starts at 5 mm (about 25°) of eccentricity and reaches a maximum between 8 and 10 mm (about 40–50°). At this point, cone density in the nasal retina was about 50% greater than in temporal retina. No asymmetry was found along the vertical meridian.
2400
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
In the far periphery of the retina, near the ora serrata, cone density increased in all quadrants. This peripheral increase in cone density was more conspicuous in the nasal quadrant of the retina and formed a cone-enriched rim, starting at 13 mm from the fovea and reaching a peak at 1 mm from the ora serrata. The average density along the nasal margin was 800091225 (from 6173 to 9630 cones/mm2) whereas at the margins of any other quadrant the average was 6373 91000 (from 5000 to 7655 cones/mm2). Fig. 4C illustrates the increase cone density at the nasal periphery averaged from four retinas. A photomicrograph of the region next to ora serrata, where cones reached the maximal density is illustrated in Fig. 5A. Their identification as cones was confirmed by NADPH-diaphorase histochemistry (Fig. 5B). Rods were not found within this region.
3.4. Cone inner segment diameter Cone inner segment diameters increased with increasing eccentricities along the horizontal (Fig. 6A) and
Fig. 3. Averaged cone density ( 9SD) from five animals (two females and three males), along the horizontal meridian. Measurements were made along a 700 mm wide strip (about 3.5°). In the foveola the peak cone density varied from 75 000 to 282 000 cells/ mm2. Inset: The best fit power function calculated for the average cone density along the horizontal meridian is Y = −22.588 ln(x) +45.277 (R 2 = 0.9949).
vertical (Fig. 6B) meridians. Diameters varied from a minimum of 4.0 mm at the foveola to a maximum of 10 mm in the temporal periphery (at 12 mm of eccentricity) and 9 mm in the nasal periphery (at 15 mm of eccentricity). Between 0 and 1 mm, cone inner segment diameters more than doubled in all quadrants. A naso-temporal asymmetry was observed from 4 mm of eccentricity, where the diameter of the cone inner segment reached higher values in the temporal than in the nasal retina. There were no obvious dorsal-ventral asymmetries. In the nasal far periphery at the cone enriched rim (at 17–18 mm of eccentricity) the estimated cone inner segment diameter was 10.019 0.79 mm (n=4 retinas, not shown). This average was slightly larger than the values obtained from eccentricities around 14–15 mm at the nasal periphery (9.09 0.5 mm).
3.5. Cone mosaic and anatomical resol6ing power in the central retina We measured the cone inner segments’ cross sectional area (CISA) in optical sections obtained from video screen images of samples along a 600 mm (2°) strip extending from the centre of the fovea to the temporal periphery. Cone area increased from 5 to 40 mm2 (Fig. 7B) with the best-fitting linear function being: CISA= 7.986+ 0.0412ecc Fig. 2. Photomicrographs of cones and rods inner segments, viewed in a video-enhanced differential interference contrast (DIC) system, at different eccentricities. A, 0.2 mm; B, 0.5 mm; C, 2.0 mm; D, 4.0 mm; E, 10 mm; F, 15 mm. From A to F note the variation in density and size of rods and cones. Magnification = 2200× .
(R =0.985)
where CISA is cone inner segment area and ecc is eccentricity in mm. Considering that the cone inner segment is the first component in capturing light (Snyder & Miller, 1977),
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
2401
we calculated the total area of cone inner segments by multiplying the average cone cross sectional area by the number of cones within each sampling window. The total cone area decreased about 15% within 1 mm of eccentricity (average= 97 132 9130.5) and fitted to the nearly linear function, as shown in Fig. 7A: Y = − 435.759x +1151.822
(R = −0.824)
Fig. 5. Photomicrographs showing the far periphery of the nasal retina (at 0.5 mm from the ora serrata). (A) View in video-enhanced DIC. Diameter of inner segments range from 9 to 11 mm, slightly larger than cones in adjacent regions. Rods are absent in this region. Magnification =2 200 ×. (B) Same area reacted for NADPH-diaphorase histochemistry, confirming that only cones are present. Bar = 40 mm.
Fig. 6. Cone inner segment diameter as a function of retinal eccentricity along the horizontal (A) vertical (B) meridians. Hatched box represents the optic disc region.
Fig. 4. Cone density variation along the mid and far periphery (from 1 to 18 mm). (A) Vertical meridian, no asymmetry is seen along this meridian and (B) Horizontal meridian, note the asymmetry along this meridian. The hatched box indicates the optic disc position. (C) Detail of cone density at the far periphery of nasal retina, showing an increase of density starting at 13 mm (0.5–1 mm from ora serrata) and reaching 8000 9 1225 cells/mm2.
where x is eccentricity and Y is total cone area/window. As defined by Packer et al. (1989), cone coverage corresponds to the cone inner segment area versus cone density. We calculated cone coverage for each sampling window and observed a decrease from 100 to 53% within the 1 mm of eccentricity (about 5°) while the cone inner segment area increased (Fig. 7B). Rods
2402
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
appeared to occupy the remaining area. Since cone density declines with increasing eccentricity while the area of individual cones increases, the overall decline in cone coverage occurs because increasing cone size only partially counterbalances decreasing cone density. Fig. 8 illustrates the variation of inter-cone distance with eccentricity for three retinas. Within the same sampling windows we observed that the mean centreto-centre cone spacing (dcc) increased by a factor of 1.5 – 2.6X as a function of retinal eccentricity within the central 2.0°. The best linear function for the increase of this parameter was: ID =2.575+6.584ecc
R =0.991
where ID is inter-cone distance and ecc is eccentricity. Assuming an hexagonal array for the mosaic of cone inner segments we calculated the Nyquist limits based on the intercone distances of four retinas (Table 3). At the centre of the fovea we found a highly variable Nyquist frequency, which decreased progressively as a function of retinal eccentricity up to central 2° (Fig. 8B). The decrease of the Nyquist frequency best fitted the quadratic polynomial:
Fig. 8. (A) Average intercone distances and standard deviation (mm and degrees) as a function of eccentricity (mm and degrees) for three retinas. The inset represents individual plots of each retina. (B) Nyquist Frequency (cyc/deg) as a function of eccentricity along the central 3° from four animals.
NF=76.945ecc2 − 80.603ecc+ 41464
Fig. 7. (A) Total area of cone inner segment per sampling window along the temporal horizontal meridian. The nearly linear decrease fits well with the equation: Y= − 435.759x+1151.822 R= − 0.824. (B) The decrease in cone coverage (triangles) is compared to the increase of the cone inner segment area ( ) within each sampling window, along the central 5° along the temporal horizontal meridian. Data were obtained using a sampling window size of 1600 mm2, at 40 mm steps.
(R= 0.962)
where NF is Nyquist frequency and ecc is eccentricity in mm. The nearest neighbour distance between cones revealed a regular mosaic pattern in all the sampling areas. As adopted by Hirsch and Miller (1987) the ratio standard deviation/inter-cone distance was used here as an index of fractional spacing disorder for each sampling area. This index of lattice disorder was used to normalise the position variability against the effects of increasing centre-to-centre spacing with retinal eccentricity. As shown in Fig. 9A, the index of fractional spacing disorder ranged from about 0.08 to 0.11 (which represents 8–11% of variation of the distance between cones) within the central 3°, in Cebus retina. Fig. 9B compares the fractional spacing disorder index obtained in Cebus retina with that which was described by Hirsch and Miller (1987) for Macaca fascicularis and human retinas, along the central horizontal meridian. Thus, in view of differences between the ideal and the actual lattice, and the possible effect of the sampling
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
2403
Table 3 Cone lattice Eccentricity (°)
Average intercone distance (mm9SD)
Average Nyquist frequency (cyc/deg9SD)
Fractional spacing disorder (range inter-animal)
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
2.40 9 0.24 2.85 9 0.29 2.98 9 0.30 3.279 0.37 3.80 9 0.42 3.86 9 0.36 4.00 90.40 4.73 90.45 4.849 0.42 5.049 0.46 5.13 90.24 5.45 90.54 5.869 0.57 5.809 0.55 6.61 9 0.60 6.629 0.59
46.77 97.98 35.42 93.27 33.65 9 3.38 31.30 92.65 28.84 9 1.48 27.72 91.46 26.83 9 2.24 25.55 9 2.34 24.38 9 1.99 23.75 91.34 23.19 9 2.17 21.69 9 0.96 20.66 9 0.94 20.00 90.31 18.64 9 1.06 18.00 9 1.08
0.100–0.105 0.067–0.101 0.062–0.114 0.720–0.109 0.062–0.094 0.070–0.095 0.075–0.095 0.068–0.096 0.070–0.093 0.052–0.075 0.061–0.099 0.086–0.098 0.074–0.095 0.066–0.091 0.089–0.105 0.094–0.108
disorder (estimated as at least 8%) it should be considered that the Nyquist frequency calculated above yields an estimate of the maximum or best possible anatomical resolving power for each window.
62 505 cells/mm2 was four times higher than temporal rod density estimated as 15 062 cells/ mm2.
3.6. Topographic distribution of rod 3.6.1. Rod peak density A map of rod density from a typical Cebus retina is illustrated in Fig. 10. We found a high density of rods (117 172910 840 rods/mm2) in an annulus approximately 4 mm away from the rod-free foveola. However, the density of rods inside this annulus was not uniform. Several spots of high density were found along the horizontal and the vertical meridians. The highest density of rods (142 483 9 1532 rods/mm2) was seen in the dorsal retina (line H in Fig. 10), a discovery similar to that described as the dorsal rod peak (DRP) by Wickler et al. (1990) for the Macaca retina. The DRP was about 18% higher than the density of rods in any other region. Rod density decreased linearly from the peak of the annulus towards the fovea and towards the far periphery, although the foveal density declined more conspicuously. Lowest density of rods was observed at the retinal borders and this density was highly variable among the quadrants (13 000 – 50 000 rods/ mm2). 3.6.2. Asymmetric distribution Analysis of rod density along the horizontal meridian revealed a naso-temporal asymmetry (Fig. 11A) while a higher rod density in nasal retina at eccentricities greater than 5 mm (about 25°) was observed. This disparity increased towards the periphery and at around 12 mm (60°) nasal rod density estimated as
Fig. 9. (A) Average fractional disorder of intercone spacing (S/M, standard deviation/mean) as a function of eccentricity within the central 3°. Data pooled from four retinas. (B) Comparison of the fractional disorder (S/M) within the central 3° in Cebus (C) Macaca fascicularis (M) (Hirsch & Miller, 1987) and human (H) (Hirsch & Curcio, 1989).
2404
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
Fig. 10. Rod isodensity map of one Cebus retina (area = 619 mm2). A, 45 000 cells/mm2; B, 80 000 cells/mm2; C, 100 000 cells/mm2; D, 110 000 cells/mm2; E, F and G, 120 000 cells/mm2; H, 140 000 cells/mm2. Counts were made at 0.1 mm steps in the foveal region and at 1 – 2 mm steps in other eccentricities. Filled circle indicates the optic disc and the asterisk represents the foveola. D, dorsal; V, ventral; N, nasal; T, temporal. Bar = 4 mm.
Fig. 11. Rod density as a function of eccentricity from the fovea to the periphery along the horizontal (A) and vertical (B) meridians. Asymmetries are evident in both meridians, from 2.5 to 12 mm along the horizontal meridian and from 2.5 to 9.5 mm along the vertical meridian. Rod density is higher in the dorsal retina. The hatched box indicates the optic disc position.
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
Along the vertical meridian (Fig. 11B), rod density was about 17–30% higher in dorsal than in ventral retina at eccentricities ranging from 2.5 to 9.5 mm (about 13–49°). The topographic distribution of rods (Fig. 10) revealed a pattern of circular isodensity contours with a slight displacement towards the nasal and dorsal hemiretinas. Compared to the topographic distribution of cones, there was no tendency to form a visual streak in the distribution of rods.
3.7. Ratio of rods to cones The proportion of rod/cone density in the Cebus retina varied as a function of eccentricity along the horizontal and vertical meridians (Fig. 12). Along both meridians we observed the highest ratio of rods to cones in a region located 6 – 8 mm (about 30 – 40° of eccentricity) from the foveola in all quadrants. However the peak value for each quadrant was different. Along the horizontal meridian (Fig. 12A) the highest ratio in the temporal retina (15:1) was in average 20% higher than in the nasal retina (12.5:1) at equivalent eccentricities. This ratio decreased to 2:1 towards the temporal far periphery, but remained stable across the nasal mid periphery, and declined three times between the peak and the far periphery. The ratio of rods to cones along the vertical meridian (Fig. 12B) revealed a large asymmetry between ventral and dorsal retina. A great variability was observed between individuals (n =5) in the ventral retina, with an average highest value of 13.4291.82. On the other hand, in the dorsal retina the ratio of rods to cones were similar between individuals and reached a highest
2405
value compared to other quadrants (20.69 1.39). Thus, highest values in the dorsal meridian were about 53.5% (20.6/13.42) higher than in the ventral meridian.
4. Discussion
4.1. Estimated anatomical resolution power We found a significant degree of variability in the estimated peak cone density for five retinas of Cebus monkey, ranging from 87 500 to 281 887 cells/mm2 (169 1279 66 998 cells/mm2). These values were comparable to those obtained in other studies (Packer et al., 1989; Wickler et al., 1990; Curcio et al., 1990). Differences in peak cone density may reflect individual variability or may be due to age differences, taking into consideration the fact that foveal cone migration can last several years during primate development (Curcio & Hendrickson, 1991). The average intercone distance in central fovea varied from 2.2 to 4.0 mm among individuals, a value that is consistent with the large range for cone peak densities estimated for Cebus retina. If one assumes a triangular packing and a conversion factor of 197.3 mm/deg (Silveira et al., 1989), the calculated angular spacing in the region of peak density ranges from 0.63 to 1.2 min arc. This value is similar to that observed for the retinas of Macaca (0.54–1.4 min arc) (Perry & Cowey, 1985; Packer et al., 1989; Wickler et al., 1990), however slightly smaller than that described for retinas of New World primate Callithrix jacchus (1–2 min arc) (Troilo et al., 1993).
Fig. 12. Ratio of rod to cone densities along horizontal (A) and vertical (B) meridians from four retinas average (9 SD). The highest ratio (20:1) corresponds to the dorsal rod peak (DRP). Outside this region the ratio ranges from 12.5 to 15:1, being smaller along the nasal horizontal meridian.
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
2406
Table 4 Comparative optic and morphometric parameters Species
Area of the retina (mm2)
Axial length of the eye (mm)
PND (mm)
RMF (mm/deg)
Total number of cones (×106)
Macaca mulatta Macaca fascicularis
7509 110a
20.8b 18.8c
12.8b 11.6b
223.0b 200.7b
3.10 9 0.13a
17.929 0.37
243.0c 213 97.15d 280e 291f 197.3g 128 9 0.004
Humans
1108.69 103.7i
24.0
13.9c 12.7d 16.7e
Cebus apella Callithrix jacchus h
602.79 25.5 209.59 13.8
18.49 0.63g 10.9
11.3g 7.63
4.60 9 0.45i 3.83 9 0.3 2.76 9 0.33
a
Wickler et al. (1990). Perry and Cowey (1985). c Miller (1979). d Lapuerta and Schein (1995). e Drasdo and Fowler (1974). f Westheimer (1972). g Silveira et al. (1989). h Troilo et al. (1993). i Curcio et al. (1990). b
We estimated the fractional spacing disorder, which is a good index to indicate the lattice disorder and to normalise the position uncertainty (jitter positional) against the effects of increasing centre-to-centre spacing with increasing retinal eccentricities. In Cebus retina, the estimated values of spatial disorder remained constant within the central three degrees of eccentricity. Such values are comparable to what was described for humans (Hirsch & Curcio, 1989) and Macaca fascicularis (Hirsch & Miller, 1987) however, only within the first 1.5°, differing substantially at higher eccentricities (see Fig. 9). In Macaca fascicularis, Hirsch and Miller (1987) corrected the sampling limits for the fractional spacing disorder and found that the anatomical resolving power, estimated from inter-cone distance, became coincident with the pooled human acuity within 1.5 and 4.0°. According to these authors, the positional jitter observed in these eccentricities could have some influence on the reduced acuity measured by conventional detection procedures. If we assume that spatial disorder can influence power sampling in Cebus retina, the regularity in cone mosaic observed within the central 2° suggests that the anatomically estimated resolution might be close to values obtained by psychophysical methods. However, we do not know yet if the relative regularity observed at higher eccentricities has any positive influence on the final acuity of this animal in the perifoveal region of this animal.
and compared to data described by Samy and Hirsch (1989) for Macaca fascicularis (RMF= 243 mm/deg, Table 4) and humans (RMF=291 mm/deg, Table 4). Within the central 3.5° of eccentricity the angular diameter of cones in Cebus apella was higher than in Macaca fascicularis, and that found in human retinas. Recently, Lapuerta and Schein (1995) re-estimated the RMF for Macaca fascicularis and found a value of 213.09 7.15 mm/deg. Using this new value of RMF, the average cone angular diameter in Macaca reaches values closer to those found for Cebus and farther from those estimated for humans. These data indicate that the eyes of Cebus and Macaca fascicularis are not scaled to human eye with respect to focal length. This
4.2. Functional implications of the photoreceptor mosaic in the Cebus eye The angular size of cones was plotted (RMF of 197.3 mm/deg, Table 4) as a function of eccentricity (Fig. 13)
Fig. 13. Comparisons of cone angular diameters (°) as a function of eccentricity Cebus (this study); Macaca fascicularis and human (Samy & Hirsch, 1989).
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
Fig. 14. Comparisons of cone (A) and rod (B) densities as a function of eccentricity. Cebus (data from this study and RMF 197.3 mm/deg, Silveira et al., 1989) Macaca fascicularis and human (Samy & Hirsch, 1989). For details see text.
Fig. 15. Comparison of peripheral cone density along the horizontal meridian. Cebus apella (this study), Callithrix jacchus (Troilo et al., 1993), Macaca mullata (Perry & Cowey, 1985) and human (Curcio et al., 1990) as a function of eccentricity in degrees. Note that Cebus cone density is higher than Macaca and human, but lower than Callithrix.
2407
lack of equivalence to the human eye is also observed for Callithrix jacchus, whose eye size seems to be a limiting in its resolution on the fovea (Troilo et al., 1993). Posterior nodal distance/axial length ratio of the eye is directly proportional to spatial resolution and inversely proportional to sensitivity level. Comparing this ratio among the different species we found 0.61 for Cebus; 0.68 for Macaca fascicularis (Lapuerta & Schein, 1995); 0.69 for humans (Southall, 1943) and 0.7 for Callithrix jacchus (Troilo et al., 1993), which suggests a slight gain in sensitivity in the Cebus eye when compared with the other species. A comparison of cone densities along the temporal retina (central 3°) of Cebus, Macaca and humans revealed a similar distribution for the three species (Fig. 14A). The best power function was also similar for Macaca and Cebus (Y= 39 539x − 0.56, R= 0.90, for Macaca and Y= 39 285x − 0.3478, R= 0.91, for Cebus). Troilo et al. (1993) also reported a similarity in cone density of Callithrix jacchus, human and Macaca mulatta, within the five central degrees of eccentricity. Surprisingly, rod density values within the central three degrees in Cebus retina were found to be similar to those obtained in humans, but very different from values estimated for Macaca fascicularis (Fig. 14B). It is possible that this feature in Cebus retina is also different from Callithrix jacchus, where rod density was shown to be smaller than in the central retina of Old World primates (see discussion in Wilder et al., 1996).
4.3. Naso-temporal asymmetry and centre-periphery gradient The naso-temporal cone asymmetry along the horizontal meridian of Cebus retina was smaller than that described for Macaca (Packer et al., 1989) but comparable to what has been observed in humans, where the nasal retina has 40–45% more cones than temporal retina at equivalent eccentricities (Curcio et al., 1990). A naso-temporal asymmetry has also been observed in the primary visual cortex of Cebus (Rosa, Gattass & Fiorani, 1988). The centre-periphery gradient of cones along the horizontal meridian was higher in Cebus (37.6X; see Table 2) than in Callithrix jacchus (27X and 15X along temporal and nasal meridian, respectively, Troilo et al., 1993), but smaller than in Macaca (50X along nasal meridian and 80–140X along temporal meridian; Perry & Cowey, 1985; Packer et al., 1989; Wickler & Rakic, 1990). When plotting cone density against retinal eccentricity (in degrees), beyond the central 5° along the temporal horizontal meridian, we observed that Cebus cone density, at each eccentricity, has values which remain intermediate between those described for human, Macaca and Callithrix jacchus (Fig. 15).
2408
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
4.4. Rod/cone ratio
References
The maximum rod/cone ratio in Cebus retina was approximately 15:1 along the horizontal meridian and 20:1 along the vertical meridian. These ratios were lower than those found for human retina (25:1, Curcio et al., 1990) and in the same range as those found for Macaca (20:1, Wickler et al., 1990). A low ratio of rod/cone along the horizontal meridian reflects a high cone density along this meridian. Conversely, the high ratio detected in the dorsal retina is probably due to the dorso-ventral asymmetry in rod distribution, which was not accompanied by a cone asymmetry. The functional meaning of rod to cone ratio is not yet understood. However, based on a study of interactions between rods and cones, Wickler et al. (1990) proposed that the proportion of photoreceptors in the monkey retina may indicate a pattern of synaptic circuitry that mediates functional interactions between scotopic and photopic systems. A higher ratio of rod to cone at the dorsal portion of the retina may retain a gain in light sampling at scotopic conditions, which suggests a functional advantage in visual sensitivity. The rod/cone ratio in Cebus along the vertical meridian declined twice throughout the periphery as in Macaque (Wicker et al., 1989). However, along the horizontal meridian this ratio decreased from three to sevenfold in Cebus, being always lower along the nasal retina. The meridian asymmetry, the variability in rod/cone ratio as well as the differences in cone coverage in Cebus retina suggest regional differences in rod and cone interactions. Such features in Cebus retina seem to be distinct from those found in diurnal Old World primates studied so far, in which rod/cone ratio remains relatively constant along the retina.
Adams, C. K., Perez, J. M., & Hawthorne, M. N. (1974). Rod and cone densities in the Rhesus. In6estigati6e Ophthalmology, 13, 885 – 888. Amorim, A. K., & Picanc¸o-Diniz, C. W. (1997). Horizontal projections of area 17 in Cebus monkeys: metric features, and modular and laminar distribution. Brazilian Journal of Medical and Biological Research, 30, 1489 – 1501. Andrade da Costa, B. L. S., Hokoc¸, J. N., Pinaud, R. R., & Gattass, R. (1997). GABAergic retinocollicular projection in the New world monkey Cebus apella. NeuroReport, 8, 1797 – 1802. Borwein, B., Borwein, D., Medeiros, J., & McGowan, J. W. (1980). The ultrastructure of monkey foveal photoreceptors, with special reference to the structure, shape, size and spacing of the foveal cones. American Journal of Anatomy, 159, 125 – 146. Curcio, C. A., & Hendrickson, A. E. (1991). Development of the primate photoreceptor mosaic. Progress in Retinal Research, 10, 89 – 120. Curcio, C. A., Packer, O., & Kalina, R. E. (1987). A wholemount method for segmental analysis of photoreceptor and ganglion cell topography in a single retina. Vision Research, 27, 9 – 15. Curcio, C. A., Sloan, K. R., Kalina, R. E., & Hendrickson, A. E. (1990). Human photoreceptor topography. Journal of Comparati6e Neurology, 292, 497 – 523. Drasdo, N., & Fowler, C. W. (1974). Non-linear projection of the retinal image in a wide-angle schematic eye. British Journal of Ophthalmology, 58, 709 – 714. Ferna´ndez, E., Cuenca, N., & De Juan, J. (1993). A compiled BASIC program for analysis of spatial point patterns: application to retinal studies. Journal of Neuroscience Methods, 50, 1 –15. Gattass, R., Rosa, M. G. P., Sousa, A. P. B., Pinon, M. C. G., Fiorani Jr, M., & Neuenschwander, S. (1990). Cortical streams of visual information processing in primates. Brazilian Journal of Medical Biological Research, 23, 375 – 393. Guimara˜es, M. Z. P., & Hokoc¸, J. N. (1997). Tyrosine hydroxilase expression in the Cebus monkey retina. Visual Neuroscience, 14, 1 – 11. Hess, D. T., & Edwards, M. A. (1987). Anatomical demonstration of ocular segregation in the retinogeniculocortical pathway of the New World Capuchin monkey (Cebus apella). Journal of Comparati6e Neurology, 264, 409 – 420. Hirsch, J., & Curcio, C. A. (1989). The spatial resolution capacity of the human fovea. Vision Research, 29, 1095 – 1101. Hirsch, J., & Hylton, R. (1984). Quality of the primate photoreceptor lattice and limits of spatial vision. Vision Research, 24, 347–356. Hirsch, J., & Miller, W. H. (1987). Does cone positional disorder limit resolution? Journal of the Optical Society of America, 4, 1481 – 1492. Jacobs, G. H. (1999). Prospects for trichromatic color vision in male Cebus monkeys. Beha6ior Brain Research, 101, 109 – 112. Krebs, W., & Krebs, I. P. (1987). Quantitative morphology of primate retina (Macaca irus). In6estigati6e Ophthalmology of Visual Science (Suppl.), 28, 261. Krebs, W., & Krebs, I. P. (1989). Quantitative morphology of the central fovea in the primate retina. American Journal of Anatomy, 184, 225 – 236. Lapuerta, P., & Schein, S. J. (1995). A four-surface schematic eye of macaque monkey obtained by an optical method. Vision Research, 35, 2245 – 2254. Lima, S. M. A., Silveira, L. C. L., & Perry, V. H. (1993). The M-ganglion cell density gradient in New World monkeys. Brazilian Journal of Medical and Biological Research, 26, 961 –964. Lima, S. M. A., Silveira, L. C. L., & Perry, V. H. (1996). The distribution of M retinal ganglion cells in diurnal and nocturnal New World monkeys. Journal of Comparati6e Neurology, 368, 538 – 552.
Acknowledgements The authors are grateful to P. Ahnelt, R. Gattass, A.P.B. Sousa and M.L. Simas for their helpful discussion on this manuscript. M. Farina, U.C. Lins and M. Costa taught us and gave technical support on IBAS-system and NIH image software. M.M.M. Oliveira helped with the photography. E. Ferna´ndez provided the Spatial Point Patterns (SPP) software. A.M.M. Moraes reviewed the countings in some retinas and E.N. Yamasaki gave good suggestions for graphic construction. Part of this report was previously presented at the Annual Meeting of the Society for Neuroscience (1995). This work was supported by grants awarded to JNH by MCT-PRONEX, FINEP, CNPq and FAPERJ.
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409 Martin, P. R., & Gru¨nert, U. (1999). Analysis of the short wavelengthsensitive (‘blue’) cone mosaic in primate retina: comparison of New World and Old World monkeys. Journal of Comparati6e Neurology, 29 (406), 1 – 14. Miller, W. H. (1979). Ocular optical filtering. In H. Autrum, Handbook of sensory physiology, vol. 7 (pp. 70–143). Berlin: Springer-Verlag. Ogden, T. E. (1975). The receptor mosaic of Aotes tri6irgatus: distribution of rods and cones. Journal of Comparati6e Neurology, 163, 193 – 202. Osterberg, G. (1935). Topography of the layer of rods and cones in the human retina. Acta OphthalmologicaI (Suppl.), 6, 1–6. Packer, O., Hendrickson, A. E., & Curcio, C. A. (1989). Photoreceptor topography of the adult pigtail macaque (Macaca nemestrina) retina. Journal of Comparati6e Neurology, 288, 165–183. Perry, V. H., & Cowey, A. (1985). The ganglion cell and cone distributions in the monkey’s retina; Implications for central magnification factors. Vision Research, 12, 1795–1810. Pessoa, V. F., Tavares, M. C., Aguiar, L., Gomes, U. R., & Tomaz, C. (1997). Color vision discrimination in the capuchin monkey Cebus apella: evidence for trichromaticity. Beha6ior Brain Research, 89, 285 – 288. Pinon, M. C., Gattass, R., & Sousa, A. P. (1998). Area V4 in Cebus monkey: extent and visuotopic organisation. Cerebral Cortex, 8, 685 – 701. Rolls, E. T., & Cowey, A. (1970). Topography of the retina and striate cortex and its relationship to visual acuity in Rhesus monkeys and squirrel monkeys. Experimental Brain Research, 10, 298–310. Rosa, M. G. P., Gattass, R., & Fiorani Jr, M. (1988). Complete pattern of ocular dominance stripes in V1 of a New World monkey, Cebus apella. Experimental Brain Research, 72, 645–648. Samy, C. N., & Hirsch, J. (1989). Comparison of human and monkey retinal photoreceptor sampling mosaics. Visual Neuroscience, 3, 281 – 285. Schein, S. J. (1988). Anatomy of macaque fovea and spatial densities of neurons in foveal representation. Journal of Comparati6e Neurology, 269, 479 – 505. Shannon, C. C., & Weaver, W. (1949). The mathematical theory of communication. Urabana, IL: University of Illinois Press. Silveira, L. C. L., Picanc¸o-Diniz, C. W., Sampaio, L. F. S., & Oswaldo-Cruz, E. (1989). Retinal ganglion cell distribution in the Cebus monkey: A comparison with the cortical magnification factors. Vision Research, 29, 1471–1483.
.
2409
Silveira, L. C. L., Yamada, E. S., Perry, V. H., & Picanc¸o-Diniz, C. W. (1994). M and P retinal ganglion cells of diurnal and nocturnal New-World monkeys. NeuroReport, 5, 2077 – 2081. Silveira, L. C. L., Lee, B. B., Yamada, E. S., Kremers, J., & Hunt, D. M. (1998). Post-receptoral mechanisms of colour vision in new world primates. Vision Research, 38, 3329 – 3337. Silveira, L. C. L., Yamada, E. S., Kremers, J., Hunt, D. M, Martin, P. R., & Gomes, F. L. (1999). Ganglion cells of a short-wavelengthsensitive cone pathway in New world monkeys: morphology and physiology. Visual Neuroscience, 16, 333 – 343. Snyder, A. W., & Miller, W. H. (1977). Photoreceptor diameter and spacing for highest resolving power. Journal of the Optical Society of America, 67, 696 – 698. Southall, J. P. C. (1943). Mirrors, prisms and lenses: a text-book of geometrical optics. New York: Macmillan. Troilo, D., Howland, H. C., & Judge, S. J. (1993). Visual optics and retinal cone topography in the common marmoset (Callithrix jacchus). Vision Research, 33, 1301 – 1310. Wa¨ssle, H., Gru¨nert, U., Rohrenbeck, J., & Boycott, B. B. (1989). Cortical magnification factor, spatial resolution, and retinal ganglion cell density in the primate. Nature, 341, 643 – 646. Wa¨ssle, H., & Riemann, H. J. (1978). The mosaic of nerve cells in the mammalian retina. Proceedings Royal Society of London Series B, 200, 441 – 461. Wickler, K. C., & Rakic, P. (1990). Distribution of photoreceptor subtypes in the retina of diurnal and nocturnal primates. Journal of Neuroscience, 10, 3390 – 3401. Wickler, K. C., Williams, R. W., & Rakic, P. (1990). Photoreceptor mosaic: number and distribution of rods and cones in the Rhesus monkey. Journal of Comparati6e Neurology, 297, 499 – 508. Wilder, H. D., Gru¨nert, U., Lee, B. B., & Martin, P. R. (1996). Topography of ganglion cells and photoreceptors in the retina of a New World monkey: the marmoset Callithrix jacchus. Visual Neuroscience, 13, 335 – 352. Williams, D. R. (1986). Seeing through the photoreceptor mosaic. Trends in Neuroscience, 9, 193 – 198. Yamada, E. S., Silveira, L. C. L., & Perry, V. H. (1996). Morphology, dendritic field size, somal size, density and coverage of M and P retinal ganglion cells of dichromatic Cebus monkey. Visual Neuroscience, 13, 1011 – 1029. Young, R. W. (1971). The renewal of rod and cone outer segments in the rhesus monkey. Journal of Cell Biology, 49, 303 – 318.
Photoreceptor topography of the retina in the New World monkey Cebus apella Belmira Lara S. Andrade da Costa a, Jan Nora Hokoc¸ b,* a
Departamento de Fisiologia e Farmacologia, Centro de Cieˆncias Biolo´gicas, Uni6ersidade Federal de Pernambuco, Pernambuco, Brazil b Departamento de Neurobiologia, Instituto de Biofisica Carlos Chagas Filho, Centro de Cieˆncias da Sau´de, Bloco G, Uni6ersidade Federal do Rio de Janeiro, 21949 -900 Rio de Janeiro, Brazil Received 26 August 1996; received in revised form 25 February 2000
Abstract The number and topographical distribution of photoreceptors was studied in whole-mounted retinas of Cebus apella. It was estimated a total of 48 million rods and 3.8 million cones. The average peak foveal cone density and the Nyquist Limit at the foveola were estimated as 169, 127 cells/mm2 and 46.77 97.98 cyc/deg, respectively. A cone-enriched rim was found near the ora serrata, more noticeable in the nasal retina. Rod distribution was asymmetrical along horizontal and vertical meridians with a higher density in the dorsal retina. The rod/cone ratio was variable and asymmetrical along both meridians. © 2000 Elsevier Science Ltd. All rights reserved. Keywords: Primates; Visual system; Fovea; Photopic vision; Scotopic vision
1. Introduction In primates, the first study on photoreceptors mosaic was done by Osterberg (1935) in the human retina. Since then, many researchers, using various histological methods and a wide range of ages, have shown a great variability in the number of photoreceptors in various species of Old World primates (Rolls & Cowey, 1970; Young, 1971; Adams, Perez & Hawthorne, 1974; Wa¨ssle & Riemann, 1978; Miller, 1979; Borwein, Borwein, Medeiros & McGowan, 1980; Perry & Cowey, 1985; Hirsch & Miller, 1987; Krebs & Krebs, 1987, 1989; Schein, 1988; Wa¨ssle, Gru¨nert, Rohrenbeck & Boycott, 1989). Curcio, Packer and Kalina (1987) introduced a new method to visualise photoreceptors that minimises tissue shrinkage. As a consequence, rod and cone distribution of many species of Old World primates were re-examined (Packer, Hendrickson & Curcio, 1989; Curcio, Sloan, Kalina & Hendrickson, 1990; Wickler & Rakic, 1990; Wickler, Williams & Rakic, 1990). * Corresponding author. Tel.: +55-21-2906897; fax: + 55-212808193.
In New World primates, comparative studies have been performed on the nocturnal specie Aotus tri6irgatus, (Ogden, 1975; Wickler & Rakic, 1990) and on the diurnal species Saimiri sciureus and Callithrix jacchus (Rolls & Cowey, 1970; Troilo, Howland & Judge, 1993; Wilder, Gru¨nert, Lee & Martin, 1996). In the diurnal capuchin monkey Cebus apella, rod and cone density were also described, but only along the horizontal meridian (Silveira, Yamada, Perry & Picanc¸o-Diniz, 1994; Yamada, Silveira & Perry, 1996). In a more complete study, we aim in the present paper, to analyse spatial distribution, mosaic organisation and morphometric parameters of rods and cones along the entire retina of the Cebus apella. Cebus is a diurnal New World monkey that has been widely used as a suitable experimental model for comparative studies of the visual system at retinal (Silveira, Picanc¸o-Diniz, Sampaio & Oswaldo-Cruz, 1989; Lima, Silveira & Perry, 1993; Silveira et al., 1994; Lima, Silveira & Perry, 1996; Yamada et al., 1996; Andrade da Costa, Hokoc¸, Pinaud & Gattass, 1997; Guimara˜es & Hokoc¸, 1997; Pessoa, Tavares, Aguiar, Gomes & Tomaz, 1997; Silveira, Lee, Yamada, Kremers & Hunt, 1998; Jacobs, 1999; Martin & Gru¨nert, 1999; Silveira, Yamada, Kremers, Hunt, Martin & Gomes 1999); and cortical (Hess
0042-6989/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 2 - 6 9 8 9 ( 0 0 ) 0 0 1 0 4 - 8
2396
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
& Edwards, 1987; Gattass, Rosa, Sousa, Pinon, Fiorani & Neuenschwander, 1990 for revision; Amorim & Picanc¸o-Diniz, 1997; Pinon, Gattass & Sousa, 1998), levels. Using the unstained retinal wholemount preparation introduced by Curcio et al. (1987), we analysed qualitative and quantitative parameters of the photoreceptor organisation in Cebus apella retina in order to characterise aspects of the visual function of this monkey and compare them to those reported for other diurnal Old and New World primates.
2. Methods
2.1. Animals and tissue preparation Eyes from seven adults, four males and three females of the species Cebus apella were studied. Procedures for the use of animals were in accord with the ‘Principles of Laboratory animal care’ (NIH) and approved by the commission of animal care of the Instituto de Biofı´sica/ UFRJ. The animals weighed between 1.8 – 2.5 kg and were kept on a 12h light/dark cycle. They were also used for other experiments not correlated with the present study. The anaesthesia was induced first with Ketamine chloride, 50 mg/kg, i.m. and supplemented with sodium pentobarbital (20 mg/kg, i.v.). One eye from each animal was enucleated under deep anaesthesia, except for one animal from which both retinas were taken to check for inter-eye variability. Eyes were fixed by immersion in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) and a small amount of the fixative solution was injected into the vitreal chamber to keep the globe taut and prevent permanent folds in the foveal area (Curcio et al., 1987). At the end of the experiments, the animals were sacrificed by an overdose of the anaesthetic. After being fixated for one hour, the eyes were bisected at the equator and the cornea and lenses removed. The retinas were then dissected in phosphate buffer (0.1 M, pH 7.4) whole-mounted on a slide with the photoreceptor layer up and coverslipped with DMSO for 24 h. The next day, the coverslip was removed, DMSO was blotted and the retinas were permanently coverslipped with 100% glycerol and sealed with clear nail polish. The final areal expansion during processing was estimated by comparing drawings of the tissue outlined while in phosphate buffer and at several different stages during processing. One eye from an adult Cebus apella was used as a standard for the measurement of shrinkage and expansion during processing. The retina and pars plana of the ciliary body were dissected and flattened on a glass slide. An image of the retina contour was drawn by projecting through a photographic mag-
nifier and the distances from the optic disc to the fovea as well as to the temporal, nasal, dorsal and ventral edges of the retina were determined at each step of the procedure. There was a 2% linear variation due to the fixative. No correction was made for this slight shrinkage. Any retina with central or peripheral deformation was discarded.
2.2. Image acquisition Whole-mounted retinas were observed using Normarski Differential Interference Contrast microscopy (NDIC) with the aid of an Axioplan Zeiss optical microscope equipped with a high resolution video camera (GRUNDIG, Pal-G, 625/50 Hz). Images of the photoreceptor inner segments were acquired and intensified in brightness and contrast using a monitor ACEZeiss Microsystems. The video camera was attached to a microcomputer supplied with the software IBAS v. 2.0 (Interaktives Bild Analysen System, Zeiss), where selected images could be stored and analysed by more than one individual for cell counts and measurements of morphometric parameters.
2.3. Sampling procedure In and around the fovea, within the first 0.8 mm of eccentricity, photoreceptor arrays were sampled at 100 mm of distance. Starting from 1 mm of the foveola throughout the retina toward the periphery, sampling areas were photographed at 1–2 mm intervals. An 100× oil immersion objective and an intermediate lens of 2.0 × were used providing a final magnification of 3500× for images in the foveal region and of 2200× for images in the mid-periphery (with an intermediate lens, 1.25× ). The diameter of the inner segment, larger in cones than in rods, was used as a morphological criterion for differentiation. An important parameter for estimating the peak cone density is the size of the sampling window. It is known that a large sampling window for the foveal region may underestimate the peak cone density by 20% or more (Packer et al., 1989), especially because cone density declines sharply as you move away from the foveal centre. We tested several sizes of sampling window for the foveal region and adopted one measuring 40×40 mm, which revealed minimal variation in the morphometric parameters, cone density and inter-cone spacing. Sampling areas of 1600 mm2 were used to count cones and rods in the central retina. From 1 mm of the foveola and throughout the whole extension of the retina, rods were counted in sampling areas of 4049 mm2 (63.6× 63.6 mm). In mid-periphery, the cones were counted using a 40× objective in sampling areas of 26 500 mm2 (162.7×162.7 mm). Alternatively, several
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
adjacent windows of 4049 mm2 were used to obtain a minimum of 30 cones per sampling area. In the far periphery of the retina, adjacent to the ora serrata, cones were counted at 500 mm steps. The location of each sampling area was marked on a drawing of the retina and the X – Y coordinates of the microscope for each region recorded. Approximately 200 areas were sampled in each retina. Isodensity lines were drawn, linking the areas of similar density starting from the point of highest density (fovea). The centre of the fovea was used as a reference point for measurements of eccentricity. The total number of photoreceptors was calculated for each retina by multiplying the area between two adjacent lines by the average cell density for both lines. These values were added to give the total number of photoreceptors for each retina. The areas and diameters of cone inner segments at different eccentricities were measured directly from video displayed images, using a mouse to trace cell contours. These contours were printed, scanned and selected for mosaic analysis. Small routines were built for morphometric analysis and added to the IBAS system. For the conversion of eccentricity from mm to degrees we adopted the Retinal Magnification Factor (RMF) of 197.3 mm/deg estimated by Silveira et al. (1989), which was based on a posterior nodal distance (PND) of 11.3 mm for the Cebus eye. This value was useful within the central 3° of the retina. For eccentricities higher than 5°, the function obtained by Yamada et al. (1996), where RMF = 205−0.48x, (x, angular distance in degrees), was employed.
2.4. Mosaic analysis Starting from the foveola, 18 contiguous sampling areas (1600 mm2), at the photoreceptor inner segments’ level, were used to measure the regularity of the cone mosaic in the central retina. Outlines of the cones were stored and the centroids X and Y for each cell were obtained using NIH-Image version 5.5. Measures of mosaic regularity were performed using the software Program Spatial Point Patterns (PSPP), version 2.2 (a new version of Ferna´ndez, Cuenca & De Juan, 1993). This software compares the mosaic organisation of each sample to a random, to a regular or to a clustered distribution, and allows to obtain statistical indexes that correspond to each distribution. Using the PSPP we obtained the average nearest neighbour distance and standard deviation for each window. Center-to-centre distances between nearest
2397
neighbours were used to estimate the spatial resolution in each retinal sampling area at different eccentricities.
2.5. Estimate of spatial resolution in central retina According to the sampling theorem (Shannon & Weaver, 1949), density and packing mode of cones in the photoreceptor array determine the highest spatial frequency that can be detected unambiguously. As many authors have shown, in human and monkey retinas, the foveal photoreceptor lattice is packed according to a hexagonal array (Miller, 1979; Borwein et al., 1980; Hirsch & Hylton, 1984). In this array, the highest spatial frequency (referred as the Nyquist Limit) for each window sampling is 2/ 3 (1/2 dcc), where dcc corresponds to inter-cone distance in degrees of visual angle (Snyder & Miller, 1977; Williams, 1986). Assuming a perfect hexagonal packaging, the Nyquist Limit can also be estimated considering the RMF and cone density by using the following relation: F · Nyquist= (0.5) · (RMF) · [(2/ 3) · dmax]1/2, where dmax is the maximum density of cones. We used both formulas to estimate the Nyquist Limit in each sampling window in the fovea within the first three central degrees of eccentricity.
3. Results
3.1. Area of the retinas The mean area measured in fixed, non-dehydrated whole-mounted retinas obtained from seven adult Cebus monkey was 601.79 23.9 mm2. The average distance from the centre of the fovea to the centre of the optic disc was 3.279 0.23 mm and between the fovea and the retinal margins was 17.29 0.48 mm along the nasal axis and 13.39 0.38 mm along the temporal axis (Table 1).
3.2. Estimate of rod and cone number We estimated a total of 489 1.62 million of rods and 3.89 0.13 million of cones in the Cebus retina (Table 2). In one animal, from which both retinas were taken to be measured, the difference in the number of cones and rods between the two eyes was less than 10%. Fig. 1A shows a representative map of isodensity lines for cones for one whole retina. The contours had a horizontally elongated appearance forming a streak of cones. The average total number of rods and cones was similar in males and females, indicating no significant
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
2398
difference between sexes. On the other hand, the regional analysis of cone number showed differences of approximately 20% in the fovea of individual monkeys. Fig. 1B is the magnification of foveal region of the retina of two different animals (B1=CB203 and B2= CB131) to illustrate the isodensity lines within this region. Retinal area within line a (20 000 cells/mm2) of both animals was the same (0.919 mm2 in B1 and 1.056 mm2 in B2), however, there was a difference of 18% in the total number of cones for these animals: 32 335 cones for B1 and 39 247 cones for B2. This finding is possibly due to the difference observed in the peak density of the foveolas (87 500 cones/mm2 in B1, and 281 887 cones/mm2 in B2). Photoreceptor mosaic at different eccentricities along the horizontal meridian is illustrated in Fig. 2. It is evident from this figure that there is an eccentricity-dependent variation of photoreceptor density and size. Rods first intrude into the cone mosaic at 150 mm from the centre of the foveola (Fig. 2A) and reach a peak density at eccentricities near the optical disc (Fig. 2D).
3.3. Topographic distribution of cones 3.3.1. Fo6ea Among the animals, there was a great variability in peak cone density (Table 2), with values ranging from 87 500 to 281 887 cells/mm2 at the foveola. The distribution of cone density within the central 0.7 mm (about 3.5°) of eccentricity along the horizontal meridian is seen in Fig. 3. The average peak cone density of four animals was estimated as 173 909974 139 cells/mm2. This peak value decreased approximately 60% within 40–60 mm from the centre of the fovea and was about 20 000 cones/mm2 at 0.7 mm of eccentricity. As shown in the inset of Fig. 3, the best logarithmic function for cone density along the temporal horizontal meridian is: Y= − 22.588 ln(x)+45.277
(R 2 = 0.9949)
where Y is the cone density and x is eccentricity in degree.
Table 1 Morphometric dimensions Animal
Animal weight (kg)
Retinal area (mm2)
OD-nasal distance (mm)
OD-fovea distance (mm)
OD-temp distance (mm)
CB229(LE)a CB229(RE)a CB255a CB131b CB129a CB232b CB203a CB579b
2.250
609.5 632.6 551.2 581.8 593.4 615.0 614.0 616.2
18.1 17.8 17.5 17.0 17.0 17.0 16.5 16.8
3.5 3.1 3.7 3.0 3.3 3.1 3.2
13.1 14.2 13.0 13.5 13.0 13.0 13.5 13.2
601.71 23.89
17.2 0.48
3.27 0.23
13.31 0.38
2.200 1.900 2.100 1.800 2.500 2.200
Mean SD a b
Male. Female.
Table 2 Estimate number of photoreceptors Animal
Cone number
Rod number
Peak cone density
Cone centre–peripheral gradient
CB229(LE)a CB229(RE)a CB255a CB131b CB129a CB232b CB203a CB579b
3 872 515 3 900 200 3 914 940 3 579 970 3 850 000 na na na
48 900 350 49 400 250 49 883 525 45 356 246 47 619 885 na na na
nac na na 281 887 150 000 199 375 87 500 126 875
na na na 62.63 33.33 44.3 19.4 28.19
Mean SD
3 828 925 128 199
48 232 051 1 623 802
169 127 66 998
37.57 14.89
a
Male. Female. c na, not analysed. b
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
2399
Fig. 1. (A) Map of cone density in a Cebus monkey retina (area =619 mm2). Optic disc and centre of the fovea are represented by a filled circle and an asterisk, respectively. Isodensity lines correspond to: A, 5000 cells/mm2; B, 6000 cells/ mm2; C, 8000 cells/mm2; D, 10 000 cells/mm2; E, 15 000 cells/mm2 and F, 25 000 cells/mm2. V, ventral; D, dorsal; T, temporal; N, nasal. Bar =4 mm. (B) Detail of foveal cone isodensity lines in two retinas that differ in peak density by a factor of 3.2. In B1 the peak density corresponds to 87 500 cells/mm2 (animal CB203, male, from Table 1) and in B2 it corresponds to 282 000 cells/mm2 (animal CB131, female, from Table 2). Characters: a, 20 000 cells/mm2; b, 30 000 cells/mm2; c, 40 000 cells/mm2; d, 50 000 cells/mm2; e, 60 000 cells/mm2; f, 70 000 cells/mm2; i, 100 000 cells/mm2. The total number of cones estimated in the fovea is 32 335 cones (B1) and 39 247 cones (B2). Bar = 500 mm.
Cone distribution across the horizontal meridian revealed a naso-temporal asymmetry, within 0.7 mm from the foveola, which was evident in some but not in all retinas. When apparent, this asymmetry was more conspicuous around 0.2 mm from the foveola, where the nasal density was about twice as the temporal density (Fig. 3). No association with sex or peak cone density was found.
3.3.2. Mid and far periphery In the mid-periphery cone density declined linearly with densities falling by approximately 500 cones/mm2 per mm along all axes (Fig. 4). Minimal densities
ranged among 4000–5000 cells/mm2 in five animals. The centre–peripheral gradient of cones ranged from 19.4X to 62X and was directed related to the peak density at the foveola (Table 2). Analysis of cone density distribution along the vertical (Fig. 4A) and horizontal (Fig. 4B) meridians shows a naso-temporal asymmetry along the horizontal meridian. This asymmetry starts at 5 mm (about 25°) of eccentricity and reaches a maximum between 8 and 10 mm (about 40–50°). At this point, cone density in the nasal retina was about 50% greater than in temporal retina. No asymmetry was found along the vertical meridian.
2400
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
In the far periphery of the retina, near the ora serrata, cone density increased in all quadrants. This peripheral increase in cone density was more conspicuous in the nasal quadrant of the retina and formed a cone-enriched rim, starting at 13 mm from the fovea and reaching a peak at 1 mm from the ora serrata. The average density along the nasal margin was 800091225 (from 6173 to 9630 cones/mm2) whereas at the margins of any other quadrant the average was 6373 91000 (from 5000 to 7655 cones/mm2). Fig. 4C illustrates the increase cone density at the nasal periphery averaged from four retinas. A photomicrograph of the region next to ora serrata, where cones reached the maximal density is illustrated in Fig. 5A. Their identification as cones was confirmed by NADPH-diaphorase histochemistry (Fig. 5B). Rods were not found within this region.
3.4. Cone inner segment diameter Cone inner segment diameters increased with increasing eccentricities along the horizontal (Fig. 6A) and
Fig. 3. Averaged cone density ( 9SD) from five animals (two females and three males), along the horizontal meridian. Measurements were made along a 700 mm wide strip (about 3.5°). In the foveola the peak cone density varied from 75 000 to 282 000 cells/ mm2. Inset: The best fit power function calculated for the average cone density along the horizontal meridian is Y = −22.588 ln(x) +45.277 (R 2 = 0.9949).
vertical (Fig. 6B) meridians. Diameters varied from a minimum of 4.0 mm at the foveola to a maximum of 10 mm in the temporal periphery (at 12 mm of eccentricity) and 9 mm in the nasal periphery (at 15 mm of eccentricity). Between 0 and 1 mm, cone inner segment diameters more than doubled in all quadrants. A naso-temporal asymmetry was observed from 4 mm of eccentricity, where the diameter of the cone inner segment reached higher values in the temporal than in the nasal retina. There were no obvious dorsal-ventral asymmetries. In the nasal far periphery at the cone enriched rim (at 17–18 mm of eccentricity) the estimated cone inner segment diameter was 10.019 0.79 mm (n=4 retinas, not shown). This average was slightly larger than the values obtained from eccentricities around 14–15 mm at the nasal periphery (9.09 0.5 mm).
3.5. Cone mosaic and anatomical resol6ing power in the central retina We measured the cone inner segments’ cross sectional area (CISA) in optical sections obtained from video screen images of samples along a 600 mm (2°) strip extending from the centre of the fovea to the temporal periphery. Cone area increased from 5 to 40 mm2 (Fig. 7B) with the best-fitting linear function being: CISA= 7.986+ 0.0412ecc Fig. 2. Photomicrographs of cones and rods inner segments, viewed in a video-enhanced differential interference contrast (DIC) system, at different eccentricities. A, 0.2 mm; B, 0.5 mm; C, 2.0 mm; D, 4.0 mm; E, 10 mm; F, 15 mm. From A to F note the variation in density and size of rods and cones. Magnification = 2200× .
(R =0.985)
where CISA is cone inner segment area and ecc is eccentricity in mm. Considering that the cone inner segment is the first component in capturing light (Snyder & Miller, 1977),
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
2401
we calculated the total area of cone inner segments by multiplying the average cone cross sectional area by the number of cones within each sampling window. The total cone area decreased about 15% within 1 mm of eccentricity (average= 97 132 9130.5) and fitted to the nearly linear function, as shown in Fig. 7A: Y = − 435.759x +1151.822
(R = −0.824)
Fig. 5. Photomicrographs showing the far periphery of the nasal retina (at 0.5 mm from the ora serrata). (A) View in video-enhanced DIC. Diameter of inner segments range from 9 to 11 mm, slightly larger than cones in adjacent regions. Rods are absent in this region. Magnification =2 200 ×. (B) Same area reacted for NADPH-diaphorase histochemistry, confirming that only cones are present. Bar = 40 mm.
Fig. 6. Cone inner segment diameter as a function of retinal eccentricity along the horizontal (A) vertical (B) meridians. Hatched box represents the optic disc region.
Fig. 4. Cone density variation along the mid and far periphery (from 1 to 18 mm). (A) Vertical meridian, no asymmetry is seen along this meridian and (B) Horizontal meridian, note the asymmetry along this meridian. The hatched box indicates the optic disc position. (C) Detail of cone density at the far periphery of nasal retina, showing an increase of density starting at 13 mm (0.5–1 mm from ora serrata) and reaching 8000 9 1225 cells/mm2.
where x is eccentricity and Y is total cone area/window. As defined by Packer et al. (1989), cone coverage corresponds to the cone inner segment area versus cone density. We calculated cone coverage for each sampling window and observed a decrease from 100 to 53% within the 1 mm of eccentricity (about 5°) while the cone inner segment area increased (Fig. 7B). Rods
2402
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
appeared to occupy the remaining area. Since cone density declines with increasing eccentricity while the area of individual cones increases, the overall decline in cone coverage occurs because increasing cone size only partially counterbalances decreasing cone density. Fig. 8 illustrates the variation of inter-cone distance with eccentricity for three retinas. Within the same sampling windows we observed that the mean centreto-centre cone spacing (dcc) increased by a factor of 1.5 – 2.6X as a function of retinal eccentricity within the central 2.0°. The best linear function for the increase of this parameter was: ID =2.575+6.584ecc
R =0.991
where ID is inter-cone distance and ecc is eccentricity. Assuming an hexagonal array for the mosaic of cone inner segments we calculated the Nyquist limits based on the intercone distances of four retinas (Table 3). At the centre of the fovea we found a highly variable Nyquist frequency, which decreased progressively as a function of retinal eccentricity up to central 2° (Fig. 8B). The decrease of the Nyquist frequency best fitted the quadratic polynomial:
Fig. 8. (A) Average intercone distances and standard deviation (mm and degrees) as a function of eccentricity (mm and degrees) for three retinas. The inset represents individual plots of each retina. (B) Nyquist Frequency (cyc/deg) as a function of eccentricity along the central 3° from four animals.
NF=76.945ecc2 − 80.603ecc+ 41464
Fig. 7. (A) Total area of cone inner segment per sampling window along the temporal horizontal meridian. The nearly linear decrease fits well with the equation: Y= − 435.759x+1151.822 R= − 0.824. (B) The decrease in cone coverage (triangles) is compared to the increase of the cone inner segment area ( ) within each sampling window, along the central 5° along the temporal horizontal meridian. Data were obtained using a sampling window size of 1600 mm2, at 40 mm steps.
(R= 0.962)
where NF is Nyquist frequency and ecc is eccentricity in mm. The nearest neighbour distance between cones revealed a regular mosaic pattern in all the sampling areas. As adopted by Hirsch and Miller (1987) the ratio standard deviation/inter-cone distance was used here as an index of fractional spacing disorder for each sampling area. This index of lattice disorder was used to normalise the position variability against the effects of increasing centre-to-centre spacing with retinal eccentricity. As shown in Fig. 9A, the index of fractional spacing disorder ranged from about 0.08 to 0.11 (which represents 8–11% of variation of the distance between cones) within the central 3°, in Cebus retina. Fig. 9B compares the fractional spacing disorder index obtained in Cebus retina with that which was described by Hirsch and Miller (1987) for Macaca fascicularis and human retinas, along the central horizontal meridian. Thus, in view of differences between the ideal and the actual lattice, and the possible effect of the sampling
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
2403
Table 3 Cone lattice Eccentricity (°)
Average intercone distance (mm9SD)
Average Nyquist frequency (cyc/deg9SD)
Fractional spacing disorder (range inter-animal)
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
2.40 9 0.24 2.85 9 0.29 2.98 9 0.30 3.279 0.37 3.80 9 0.42 3.86 9 0.36 4.00 90.40 4.73 90.45 4.849 0.42 5.049 0.46 5.13 90.24 5.45 90.54 5.869 0.57 5.809 0.55 6.61 9 0.60 6.629 0.59
46.77 97.98 35.42 93.27 33.65 9 3.38 31.30 92.65 28.84 9 1.48 27.72 91.46 26.83 9 2.24 25.55 9 2.34 24.38 9 1.99 23.75 91.34 23.19 9 2.17 21.69 9 0.96 20.66 9 0.94 20.00 90.31 18.64 9 1.06 18.00 9 1.08
0.100–0.105 0.067–0.101 0.062–0.114 0.720–0.109 0.062–0.094 0.070–0.095 0.075–0.095 0.068–0.096 0.070–0.093 0.052–0.075 0.061–0.099 0.086–0.098 0.074–0.095 0.066–0.091 0.089–0.105 0.094–0.108
disorder (estimated as at least 8%) it should be considered that the Nyquist frequency calculated above yields an estimate of the maximum or best possible anatomical resolving power for each window.
62 505 cells/mm2 was four times higher than temporal rod density estimated as 15 062 cells/ mm2.
3.6. Topographic distribution of rod 3.6.1. Rod peak density A map of rod density from a typical Cebus retina is illustrated in Fig. 10. We found a high density of rods (117 172910 840 rods/mm2) in an annulus approximately 4 mm away from the rod-free foveola. However, the density of rods inside this annulus was not uniform. Several spots of high density were found along the horizontal and the vertical meridians. The highest density of rods (142 483 9 1532 rods/mm2) was seen in the dorsal retina (line H in Fig. 10), a discovery similar to that described as the dorsal rod peak (DRP) by Wickler et al. (1990) for the Macaca retina. The DRP was about 18% higher than the density of rods in any other region. Rod density decreased linearly from the peak of the annulus towards the fovea and towards the far periphery, although the foveal density declined more conspicuously. Lowest density of rods was observed at the retinal borders and this density was highly variable among the quadrants (13 000 – 50 000 rods/ mm2). 3.6.2. Asymmetric distribution Analysis of rod density along the horizontal meridian revealed a naso-temporal asymmetry (Fig. 11A) while a higher rod density in nasal retina at eccentricities greater than 5 mm (about 25°) was observed. This disparity increased towards the periphery and at around 12 mm (60°) nasal rod density estimated as
Fig. 9. (A) Average fractional disorder of intercone spacing (S/M, standard deviation/mean) as a function of eccentricity within the central 3°. Data pooled from four retinas. (B) Comparison of the fractional disorder (S/M) within the central 3° in Cebus (C) Macaca fascicularis (M) (Hirsch & Miller, 1987) and human (H) (Hirsch & Curcio, 1989).
2404
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
Fig. 10. Rod isodensity map of one Cebus retina (area = 619 mm2). A, 45 000 cells/mm2; B, 80 000 cells/mm2; C, 100 000 cells/mm2; D, 110 000 cells/mm2; E, F and G, 120 000 cells/mm2; H, 140 000 cells/mm2. Counts were made at 0.1 mm steps in the foveal region and at 1 – 2 mm steps in other eccentricities. Filled circle indicates the optic disc and the asterisk represents the foveola. D, dorsal; V, ventral; N, nasal; T, temporal. Bar = 4 mm.
Fig. 11. Rod density as a function of eccentricity from the fovea to the periphery along the horizontal (A) and vertical (B) meridians. Asymmetries are evident in both meridians, from 2.5 to 12 mm along the horizontal meridian and from 2.5 to 9.5 mm along the vertical meridian. Rod density is higher in the dorsal retina. The hatched box indicates the optic disc position.
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
Along the vertical meridian (Fig. 11B), rod density was about 17–30% higher in dorsal than in ventral retina at eccentricities ranging from 2.5 to 9.5 mm (about 13–49°). The topographic distribution of rods (Fig. 10) revealed a pattern of circular isodensity contours with a slight displacement towards the nasal and dorsal hemiretinas. Compared to the topographic distribution of cones, there was no tendency to form a visual streak in the distribution of rods.
3.7. Ratio of rods to cones The proportion of rod/cone density in the Cebus retina varied as a function of eccentricity along the horizontal and vertical meridians (Fig. 12). Along both meridians we observed the highest ratio of rods to cones in a region located 6 – 8 mm (about 30 – 40° of eccentricity) from the foveola in all quadrants. However the peak value for each quadrant was different. Along the horizontal meridian (Fig. 12A) the highest ratio in the temporal retina (15:1) was in average 20% higher than in the nasal retina (12.5:1) at equivalent eccentricities. This ratio decreased to 2:1 towards the temporal far periphery, but remained stable across the nasal mid periphery, and declined three times between the peak and the far periphery. The ratio of rods to cones along the vertical meridian (Fig. 12B) revealed a large asymmetry between ventral and dorsal retina. A great variability was observed between individuals (n =5) in the ventral retina, with an average highest value of 13.4291.82. On the other hand, in the dorsal retina the ratio of rods to cones were similar between individuals and reached a highest
2405
value compared to other quadrants (20.69 1.39). Thus, highest values in the dorsal meridian were about 53.5% (20.6/13.42) higher than in the ventral meridian.
4. Discussion
4.1. Estimated anatomical resolution power We found a significant degree of variability in the estimated peak cone density for five retinas of Cebus monkey, ranging from 87 500 to 281 887 cells/mm2 (169 1279 66 998 cells/mm2). These values were comparable to those obtained in other studies (Packer et al., 1989; Wickler et al., 1990; Curcio et al., 1990). Differences in peak cone density may reflect individual variability or may be due to age differences, taking into consideration the fact that foveal cone migration can last several years during primate development (Curcio & Hendrickson, 1991). The average intercone distance in central fovea varied from 2.2 to 4.0 mm among individuals, a value that is consistent with the large range for cone peak densities estimated for Cebus retina. If one assumes a triangular packing and a conversion factor of 197.3 mm/deg (Silveira et al., 1989), the calculated angular spacing in the region of peak density ranges from 0.63 to 1.2 min arc. This value is similar to that observed for the retinas of Macaca (0.54–1.4 min arc) (Perry & Cowey, 1985; Packer et al., 1989; Wickler et al., 1990), however slightly smaller than that described for retinas of New World primate Callithrix jacchus (1–2 min arc) (Troilo et al., 1993).
Fig. 12. Ratio of rod to cone densities along horizontal (A) and vertical (B) meridians from four retinas average (9 SD). The highest ratio (20:1) corresponds to the dorsal rod peak (DRP). Outside this region the ratio ranges from 12.5 to 15:1, being smaller along the nasal horizontal meridian.
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
2406
Table 4 Comparative optic and morphometric parameters Species
Area of the retina (mm2)
Axial length of the eye (mm)
PND (mm)
RMF (mm/deg)
Total number of cones (×106)
Macaca mulatta Macaca fascicularis
7509 110a
20.8b 18.8c
12.8b 11.6b
223.0b 200.7b
3.10 9 0.13a
17.929 0.37
243.0c 213 97.15d 280e 291f 197.3g 128 9 0.004
Humans
1108.69 103.7i
24.0
13.9c 12.7d 16.7e
Cebus apella Callithrix jacchus h
602.79 25.5 209.59 13.8
18.49 0.63g 10.9
11.3g 7.63
4.60 9 0.45i 3.83 9 0.3 2.76 9 0.33
a
Wickler et al. (1990). Perry and Cowey (1985). c Miller (1979). d Lapuerta and Schein (1995). e Drasdo and Fowler (1974). f Westheimer (1972). g Silveira et al. (1989). h Troilo et al. (1993). i Curcio et al. (1990). b
We estimated the fractional spacing disorder, which is a good index to indicate the lattice disorder and to normalise the position uncertainty (jitter positional) against the effects of increasing centre-to-centre spacing with increasing retinal eccentricities. In Cebus retina, the estimated values of spatial disorder remained constant within the central three degrees of eccentricity. Such values are comparable to what was described for humans (Hirsch & Curcio, 1989) and Macaca fascicularis (Hirsch & Miller, 1987) however, only within the first 1.5°, differing substantially at higher eccentricities (see Fig. 9). In Macaca fascicularis, Hirsch and Miller (1987) corrected the sampling limits for the fractional spacing disorder and found that the anatomical resolving power, estimated from inter-cone distance, became coincident with the pooled human acuity within 1.5 and 4.0°. According to these authors, the positional jitter observed in these eccentricities could have some influence on the reduced acuity measured by conventional detection procedures. If we assume that spatial disorder can influence power sampling in Cebus retina, the regularity in cone mosaic observed within the central 2° suggests that the anatomically estimated resolution might be close to values obtained by psychophysical methods. However, we do not know yet if the relative regularity observed at higher eccentricities has any positive influence on the final acuity of this animal in the perifoveal region of this animal.
and compared to data described by Samy and Hirsch (1989) for Macaca fascicularis (RMF= 243 mm/deg, Table 4) and humans (RMF=291 mm/deg, Table 4). Within the central 3.5° of eccentricity the angular diameter of cones in Cebus apella was higher than in Macaca fascicularis, and that found in human retinas. Recently, Lapuerta and Schein (1995) re-estimated the RMF for Macaca fascicularis and found a value of 213.09 7.15 mm/deg. Using this new value of RMF, the average cone angular diameter in Macaca reaches values closer to those found for Cebus and farther from those estimated for humans. These data indicate that the eyes of Cebus and Macaca fascicularis are not scaled to human eye with respect to focal length. This
4.2. Functional implications of the photoreceptor mosaic in the Cebus eye The angular size of cones was plotted (RMF of 197.3 mm/deg, Table 4) as a function of eccentricity (Fig. 13)
Fig. 13. Comparisons of cone angular diameters (°) as a function of eccentricity Cebus (this study); Macaca fascicularis and human (Samy & Hirsch, 1989).
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
Fig. 14. Comparisons of cone (A) and rod (B) densities as a function of eccentricity. Cebus (data from this study and RMF 197.3 mm/deg, Silveira et al., 1989) Macaca fascicularis and human (Samy & Hirsch, 1989). For details see text.
Fig. 15. Comparison of peripheral cone density along the horizontal meridian. Cebus apella (this study), Callithrix jacchus (Troilo et al., 1993), Macaca mullata (Perry & Cowey, 1985) and human (Curcio et al., 1990) as a function of eccentricity in degrees. Note that Cebus cone density is higher than Macaca and human, but lower than Callithrix.
2407
lack of equivalence to the human eye is also observed for Callithrix jacchus, whose eye size seems to be a limiting in its resolution on the fovea (Troilo et al., 1993). Posterior nodal distance/axial length ratio of the eye is directly proportional to spatial resolution and inversely proportional to sensitivity level. Comparing this ratio among the different species we found 0.61 for Cebus; 0.68 for Macaca fascicularis (Lapuerta & Schein, 1995); 0.69 for humans (Southall, 1943) and 0.7 for Callithrix jacchus (Troilo et al., 1993), which suggests a slight gain in sensitivity in the Cebus eye when compared with the other species. A comparison of cone densities along the temporal retina (central 3°) of Cebus, Macaca and humans revealed a similar distribution for the three species (Fig. 14A). The best power function was also similar for Macaca and Cebus (Y= 39 539x − 0.56, R= 0.90, for Macaca and Y= 39 285x − 0.3478, R= 0.91, for Cebus). Troilo et al. (1993) also reported a similarity in cone density of Callithrix jacchus, human and Macaca mulatta, within the five central degrees of eccentricity. Surprisingly, rod density values within the central three degrees in Cebus retina were found to be similar to those obtained in humans, but very different from values estimated for Macaca fascicularis (Fig. 14B). It is possible that this feature in Cebus retina is also different from Callithrix jacchus, where rod density was shown to be smaller than in the central retina of Old World primates (see discussion in Wilder et al., 1996).
4.3. Naso-temporal asymmetry and centre-periphery gradient The naso-temporal cone asymmetry along the horizontal meridian of Cebus retina was smaller than that described for Macaca (Packer et al., 1989) but comparable to what has been observed in humans, where the nasal retina has 40–45% more cones than temporal retina at equivalent eccentricities (Curcio et al., 1990). A naso-temporal asymmetry has also been observed in the primary visual cortex of Cebus (Rosa, Gattass & Fiorani, 1988). The centre-periphery gradient of cones along the horizontal meridian was higher in Cebus (37.6X; see Table 2) than in Callithrix jacchus (27X and 15X along temporal and nasal meridian, respectively, Troilo et al., 1993), but smaller than in Macaca (50X along nasal meridian and 80–140X along temporal meridian; Perry & Cowey, 1985; Packer et al., 1989; Wickler & Rakic, 1990). When plotting cone density against retinal eccentricity (in degrees), beyond the central 5° along the temporal horizontal meridian, we observed that Cebus cone density, at each eccentricity, has values which remain intermediate between those described for human, Macaca and Callithrix jacchus (Fig. 15).
2408
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409
4.4. Rod/cone ratio
References
The maximum rod/cone ratio in Cebus retina was approximately 15:1 along the horizontal meridian and 20:1 along the vertical meridian. These ratios were lower than those found for human retina (25:1, Curcio et al., 1990) and in the same range as those found for Macaca (20:1, Wickler et al., 1990). A low ratio of rod/cone along the horizontal meridian reflects a high cone density along this meridian. Conversely, the high ratio detected in the dorsal retina is probably due to the dorso-ventral asymmetry in rod distribution, which was not accompanied by a cone asymmetry. The functional meaning of rod to cone ratio is not yet understood. However, based on a study of interactions between rods and cones, Wickler et al. (1990) proposed that the proportion of photoreceptors in the monkey retina may indicate a pattern of synaptic circuitry that mediates functional interactions between scotopic and photopic systems. A higher ratio of rod to cone at the dorsal portion of the retina may retain a gain in light sampling at scotopic conditions, which suggests a functional advantage in visual sensitivity. The rod/cone ratio in Cebus along the vertical meridian declined twice throughout the periphery as in Macaque (Wicker et al., 1989). However, along the horizontal meridian this ratio decreased from three to sevenfold in Cebus, being always lower along the nasal retina. The meridian asymmetry, the variability in rod/cone ratio as well as the differences in cone coverage in Cebus retina suggest regional differences in rod and cone interactions. Such features in Cebus retina seem to be distinct from those found in diurnal Old World primates studied so far, in which rod/cone ratio remains relatively constant along the retina.
Adams, C. K., Perez, J. M., & Hawthorne, M. N. (1974). Rod and cone densities in the Rhesus. In6estigati6e Ophthalmology, 13, 885 – 888. Amorim, A. K., & Picanc¸o-Diniz, C. W. (1997). Horizontal projections of area 17 in Cebus monkeys: metric features, and modular and laminar distribution. Brazilian Journal of Medical and Biological Research, 30, 1489 – 1501. Andrade da Costa, B. L. S., Hokoc¸, J. N., Pinaud, R. R., & Gattass, R. (1997). GABAergic retinocollicular projection in the New world monkey Cebus apella. NeuroReport, 8, 1797 – 1802. Borwein, B., Borwein, D., Medeiros, J., & McGowan, J. W. (1980). The ultrastructure of monkey foveal photoreceptors, with special reference to the structure, shape, size and spacing of the foveal cones. American Journal of Anatomy, 159, 125 – 146. Curcio, C. A., & Hendrickson, A. E. (1991). Development of the primate photoreceptor mosaic. Progress in Retinal Research, 10, 89 – 120. Curcio, C. A., Packer, O., & Kalina, R. E. (1987). A wholemount method for segmental analysis of photoreceptor and ganglion cell topography in a single retina. Vision Research, 27, 9 – 15. Curcio, C. A., Sloan, K. R., Kalina, R. E., & Hendrickson, A. E. (1990). Human photoreceptor topography. Journal of Comparati6e Neurology, 292, 497 – 523. Drasdo, N., & Fowler, C. W. (1974). Non-linear projection of the retinal image in a wide-angle schematic eye. British Journal of Ophthalmology, 58, 709 – 714. Ferna´ndez, E., Cuenca, N., & De Juan, J. (1993). A compiled BASIC program for analysis of spatial point patterns: application to retinal studies. Journal of Neuroscience Methods, 50, 1 –15. Gattass, R., Rosa, M. G. P., Sousa, A. P. B., Pinon, M. C. G., Fiorani Jr, M., & Neuenschwander, S. (1990). Cortical streams of visual information processing in primates. Brazilian Journal of Medical Biological Research, 23, 375 – 393. Guimara˜es, M. Z. P., & Hokoc¸, J. N. (1997). Tyrosine hydroxilase expression in the Cebus monkey retina. Visual Neuroscience, 14, 1 – 11. Hess, D. T., & Edwards, M. A. (1987). Anatomical demonstration of ocular segregation in the retinogeniculocortical pathway of the New World Capuchin monkey (Cebus apella). Journal of Comparati6e Neurology, 264, 409 – 420. Hirsch, J., & Curcio, C. A. (1989). The spatial resolution capacity of the human fovea. Vision Research, 29, 1095 – 1101. Hirsch, J., & Hylton, R. (1984). Quality of the primate photoreceptor lattice and limits of spatial vision. Vision Research, 24, 347–356. Hirsch, J., & Miller, W. H. (1987). Does cone positional disorder limit resolution? Journal of the Optical Society of America, 4, 1481 – 1492. Jacobs, G. H. (1999). Prospects for trichromatic color vision in male Cebus monkeys. Beha6ior Brain Research, 101, 109 – 112. Krebs, W., & Krebs, I. P. (1987). Quantitative morphology of primate retina (Macaca irus). In6estigati6e Ophthalmology of Visual Science (Suppl.), 28, 261. Krebs, W., & Krebs, I. P. (1989). Quantitative morphology of the central fovea in the primate retina. American Journal of Anatomy, 184, 225 – 236. Lapuerta, P., & Schein, S. J. (1995). A four-surface schematic eye of macaque monkey obtained by an optical method. Vision Research, 35, 2245 – 2254. Lima, S. M. A., Silveira, L. C. L., & Perry, V. H. (1993). The M-ganglion cell density gradient in New World monkeys. Brazilian Journal of Medical and Biological Research, 26, 961 –964. Lima, S. M. A., Silveira, L. C. L., & Perry, V. H. (1996). The distribution of M retinal ganglion cells in diurnal and nocturnal New World monkeys. Journal of Comparati6e Neurology, 368, 538 – 552.
Acknowledgements The authors are grateful to P. Ahnelt, R. Gattass, A.P.B. Sousa and M.L. Simas for their helpful discussion on this manuscript. M. Farina, U.C. Lins and M. Costa taught us and gave technical support on IBAS-system and NIH image software. M.M.M. Oliveira helped with the photography. E. Ferna´ndez provided the Spatial Point Patterns (SPP) software. A.M.M. Moraes reviewed the countings in some retinas and E.N. Yamasaki gave good suggestions for graphic construction. Part of this report was previously presented at the Annual Meeting of the Society for Neuroscience (1995). This work was supported by grants awarded to JNH by MCT-PRONEX, FINEP, CNPq and FAPERJ.
B.L.S. Andrade da Costa, J.N. Hokoc¸ / Vision Research 40 (2000) 2395–2409 Martin, P. R., & Gru¨nert, U. (1999). Analysis of the short wavelengthsensitive (‘blue’) cone mosaic in primate retina: comparison of New World and Old World monkeys. Journal of Comparati6e Neurology, 29 (406), 1 – 14. Miller, W. H. (1979). Ocular optical filtering. In H. Autrum, Handbook of sensory physiology, vol. 7 (pp. 70–143). Berlin: Springer-Verlag. Ogden, T. E. (1975). The receptor mosaic of Aotes tri6irgatus: distribution of rods and cones. Journal of Comparati6e Neurology, 163, 193 – 202. Osterberg, G. (1935). Topography of the layer of rods and cones in the human retina. Acta OphthalmologicaI (Suppl.), 6, 1–6. Packer, O., Hendrickson, A. E., & Curcio, C. A. (1989). Photoreceptor topography of the adult pigtail macaque (Macaca nemestrina) retina. Journal of Comparati6e Neurology, 288, 165–183. Perry, V. H., & Cowey, A. (1985). The ganglion cell and cone distributions in the monkey’s retina; Implications for central magnification factors. Vision Research, 12, 1795–1810. Pessoa, V. F., Tavares, M. C., Aguiar, L., Gomes, U. R., & Tomaz, C. (1997). Color vision discrimination in the capuchin monkey Cebus apella: evidence for trichromaticity. Beha6ior Brain Research, 89, 285 – 288. Pinon, M. C., Gattass, R., & Sousa, A. P. (1998). Area V4 in Cebus monkey: extent and visuotopic organisation. Cerebral Cortex, 8, 685 – 701. Rolls, E. T., & Cowey, A. (1970). Topography of the retina and striate cortex and its relationship to visual acuity in Rhesus monkeys and squirrel monkeys. Experimental Brain Research, 10, 298–310. Rosa, M. G. P., Gattass, R., & Fiorani Jr, M. (1988). Complete pattern of ocular dominance stripes in V1 of a New World monkey, Cebus apella. Experimental Brain Research, 72, 645–648. Samy, C. N., & Hirsch, J. (1989). Comparison of human and monkey retinal photoreceptor sampling mosaics. Visual Neuroscience, 3, 281 – 285. Schein, S. J. (1988). Anatomy of macaque fovea and spatial densities of neurons in foveal representation. Journal of Comparati6e Neurology, 269, 479 – 505. Shannon, C. C., & Weaver, W. (1949). The mathematical theory of communication. Urabana, IL: University of Illinois Press. Silveira, L. C. L., Picanc¸o-Diniz, C. W., Sampaio, L. F. S., & Oswaldo-Cruz, E. (1989). Retinal ganglion cell distribution in the Cebus monkey: A comparison with the cortical magnification factors. Vision Research, 29, 1471–1483.
.
2409
Silveira, L. C. L., Yamada, E. S., Perry, V. H., & Picanc¸o-Diniz, C. W. (1994). M and P retinal ganglion cells of diurnal and nocturnal New-World monkeys. NeuroReport, 5, 2077 – 2081. Silveira, L. C. L., Lee, B. B., Yamada, E. S., Kremers, J., & Hunt, D. M. (1998). Post-receptoral mechanisms of colour vision in new world primates. Vision Research, 38, 3329 – 3337. Silveira, L. C. L., Yamada, E. S., Kremers, J., Hunt, D. M, Martin, P. R., & Gomes, F. L. (1999). Ganglion cells of a short-wavelengthsensitive cone pathway in New world monkeys: morphology and physiology. Visual Neuroscience, 16, 333 – 343. Snyder, A. W., & Miller, W. H. (1977). Photoreceptor diameter and spacing for highest resolving power. Journal of the Optical Society of America, 67, 696 – 698. Southall, J. P. C. (1943). Mirrors, prisms and lenses: a text-book of geometrical optics. New York: Macmillan. Troilo, D., Howland, H. C., & Judge, S. J. (1993). Visual optics and retinal cone topography in the common marmoset (Callithrix jacchus). Vision Research, 33, 1301 – 1310. Wa¨ssle, H., Gru¨nert, U., Rohrenbeck, J., & Boycott, B. B. (1989). Cortical magnification factor, spatial resolution, and retinal ganglion cell density in the primate. Nature, 341, 643 – 646. Wa¨ssle, H., & Riemann, H. J. (1978). The mosaic of nerve cells in the mammalian retina. Proceedings Royal Society of London Series B, 200, 441 – 461. Wickler, K. C., & Rakic, P. (1990). Distribution of photoreceptor subtypes in the retina of diurnal and nocturnal primates. Journal of Neuroscience, 10, 3390 – 3401. Wickler, K. C., Williams, R. W., & Rakic, P. (1990). Photoreceptor mosaic: number and distribution of rods and cones in the Rhesus monkey. Journal of Comparati6e Neurology, 297, 499 – 508. Wilder, H. D., Gru¨nert, U., Lee, B. B., & Martin, P. R. (1996). Topography of ganglion cells and photoreceptors in the retina of a New World monkey: the marmoset Callithrix jacchus. Visual Neuroscience, 13, 335 – 352. Williams, D. R. (1986). Seeing through the photoreceptor mosaic. Trends in Neuroscience, 9, 193 – 198. Yamada, E. S., Silveira, L. C. L., & Perry, V. H. (1996). Morphology, dendritic field size, somal size, density and coverage of M and P retinal ganglion cells of dichromatic Cebus monkey. Visual Neuroscience, 13, 1011 – 1029. Young, R. W. (1971). The renewal of rod and cone outer segments in the rhesus monkey. Journal of Cell Biology, 49, 303 – 318.