Original Paper Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
Received: August 8, 2013 Returned for revision: September 10, 2013 Accepted after third revision: June 16, 2014 Published online: October 23, 2014
Comparative Study of Photoreceptor and Retinal Ganglion Cell Topography and Spatial Resolving Power in Dipsadidae Snakes Einat Hauzman a, b Daniela M.O. Bonci a, b Sonia R. Grotzner c Maritana Mela c André M.P. Liber a, b Sonia L. Martins a, b Dora F. Ventura a, b a
Departamento de Psicologia Experimental, Instituto de Psicologia, and b Núcleo de Neurociências e Comportamento, Universidade de São Paulo, São Paulo, and c Departamento de Biologia Celular, Universidade Federal do Paraná, Curitiba, Brazil
Abstract The diurnal Dipsadidae snakes Philodryas olfersii and P. patagoniensis are closely related in their phylogeny but inhabit different ecological niches. P. olfersii is arboreal, whereas P. patagoniensis is preferentially terrestrial. The goal of the present study was to compare the density and topography of neurons, photoreceptors, and cells in the ganglion cell layer in the retinas of these two species using immunohistochemistry and Nissl staining procedures and estimate the spatial resolving power of their eyes based on the ganglion cell peak density. Four morphologically distinct types of cones were observed by scanning electron microscopy, 3 of which were labeled with anti-opsin antibodies: large single cones and double cones labeled by the antibody JH492 and small single cones labeled by the antibody JH455. The average densities of photoreceptors and neurons in the ganglion cell layer were similar in both species (∼10,000 and 7,000 cells · mm–2, respectively). The estimated spatial resolving
© 2014 S. Karger AG, Basel 0006–8977/14/0843–0197$39.50/0 E-Mail [email protected] www.karger.com/bbe
power was also similar, ranging from 2.4 to 2.7 cycles · degree–1. However, the distribution of neurons had different specializations. In the arboreal P. olfersii, the isodensity maps had a horizontal visual streak, with a peak density in the central region and a lower density in the dorsal retina. This organization might be relevant for locomotion and hunting behavior in the arboreal layer. In the terrestrial P. patagoniensis, a concentric pattern of decreasing cell density emanated from an area centralis located in the naso-ventral retina. Lower densities were observed in the dorsal region. The ventrally high density improves the resolution in the superior visual field and may be an important adaptation for terrestrial snakes to perceive the approach of predators from above. © 2014 S. Karger AG, Basel
Introduction
Snakes comprise a highly diversified reptilian group (suborder Serpentes) that has successfully radiated into a wide range of ecological niches. Representatives of the ∼3,400 extant species are found in virtually every portion Einat Hauzman Instituto de Psicologia, Universidade de São Paulo Avenida Professor Mello Moraes, 1721, Bloco A, Sala D-9 São Paulo, SP 05508-030 (Brazil) E-Mail hauzman.einat @ gmail.com
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
Key Words Visual ecology · Snakes · Retina · Photoreceptors · Ganglion cells · Topography · Visual acuity
198
Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
(‘basal’) snakes and belong to the Rh1, SWS1, and LWS opsin groups, which generate rods and SWS and LWS cones, respectively [Davies et al., 2009]. In addition to the different types of photoreceptors and visual pigments expressed in vertebrate retinas, comparative studies have shown that the distribution and density of retinal neurons are variable across species and that retinal topography appears to be correlated with the species’ habitat or lifestyle [Hughes, 1977; Collin, 1999, 2008]. Numerous studies have used ganglion cell spacing in the areas of highest cell density in combination with the eye’s focal length to estimate the maximum spatial resolution in a number of vertebrate species [e.g. Pettigrew et al., 1988; Collin and Pettigrew, 1989; Pettigrew and Manger, 2008; Hart et al., 2012; Coimbra et al., 2013]. To date, the distribution of neurons in snake retinas has been evaluated only in a few of the great diversity of snakes species: the terrestrial T. sirtalis [Wong, 1989] and 3 species of sea snakes, i.e. Lapemis curtus, Aipysurus laevis, and Disteira major [Hart et al., 2012]. The present study investigated the retinas of 2 species of Dipsadidae snakes, i.e. Philodryas olfersii and P. patagoniensis (fig. 1), which are closely related, sympatric, and widely distributed in South America [Thomas, 1976]. They are both active only during the day, and both are fast moving snakes that constrict or envenom their prey [Sazima and Haddad, 1992; Hartmann and Marques, 2005]. They are dietary generalists, feeding on a large variety of small vertebrates, such as amphibians, lizards, other snakes, birds, and mammals [Amaral, 1978; Lopez, 2003; Hartmann and Marques, 2005]. Despite dwelling in both open and forested areas, P. olfersii is found mostly in forested areas and forest edges and is regarded as a semiarboreal snake, whereas P. patagoniensis is more frequently observed in open habitats and is viewed as essentially terrestrial [Sazima and Haddad, 1992; Fowler and Salomão, 1994; Marques et al., 2001; Hartmann and Marques, 2005]. This difference in preferred habitat is consistent with some morphological characteristics, such as body mass, body length, and color. The arboreal P. olfersii has a more slender body and a longer tail length than the terrestrial P. patagoniensis, characteristics that benefit locomotion in the arboreal layer by allowing greater equilibrium and better distribution of the body mass on branches [Lillywhite and Henderson, 1993; Martins et al., 2001]. The color differences of the two species are also consistent with differences in the frequency of habitat and microhabitat use. The green P. olfersii is well camouflaged against the background of green foliage in the forest, Hauzman /Bonci /Grotzner /Mela /Liber / Martins /Ventura
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
of the biosphere, with the exception of polar regions, some islands, and the deep ocean [Cadle, 1987; Ford and Burghardt, 1993; Lillywhite and Henderson, 1993]. The great diversity of the Ophidian group can be explained by a number of adaptive radiations, and their evolutionary success is related to the development of some specialized sensory organs [Ford and Burghardt, 1993]. Vision may represent an important sensory modality that is employed by some snake species, and the anatomical variety of snake eyes indicates how environmental conditions have influenced the evolution of ocular structure and function [Walls, 1942]. The morphological diversity of photoreceptors in the retinas of different snake species has been described in several studies [Walls, 1942; Underwood, 1967, 1970; Wong, 1989; Hart et al., 2012]. Both rods and cones are found in numerous nocturnal species from different families [Walls, 1942; Underwood, 1967, 1970; Sillman et al., 1999, 2001; Davies et al., 2009]. However, diurnal Caenophidian (‘advanced’) snakes tend to have only cones [Walls, 1942; Underwood, 1967, 1970; Wong, 1989; Sillman et al., 1997; Hart et al., 2012]. For example, Wong [1989] found no evidence of rod photoreceptors in the retina of the garter snake Thamnophis sirtalis, a diurnal Colubridae species, and described 3 classes of cones according to their morphologies: large single cones, small single cones, and double cones that consist of principal and accessory members. Later studies on the same species [Sillman et al., 1997] described a fourth cone-like photoreceptor, a very small single cone. In sea snakes from the Hidrophiidae family, Hart et al. [2012] also described the same morphologically distinct types of cones and the absence of rod-like photoreceptors. The spectral absorbance of photoreceptors has been measured in a number of species of Boidae, Colubridae, Viperidae, and Hydrophiidae snakes [Govardovskii and Chkheidze, 1989; Jacobs et al., 1992; Sillman et al., 1997, 1999, 2001; Davies et al., 2009; Hart et al., 2012]. In the Caenophidian snake T. sirtalis, 3 absorption peaks were distinguished [Sillman et al., 1997]: an LWS visual pigment (554 nm) in large single cones and the principal and accessory members of double cones, an SWS pigment (360 nm) in small single cones, and an MWS pigment (482 nm) in very small single cones. In retinas of Hydrophiidae sea snakes, double cones and large single cones also contain LWS visual pigments (555–559 nm). Small single cones have SWS pigments (428–430 nm), and a second class of small single cones contains an MWS pigment [496 nm; Hart et al., 2012]. The visual pigments were genetically classified into 2 species of Henophidian
Color version available online
a
We tested 4 hypotheses. (1) Because both Caenophidian snakes are diurnal, we expected to find 4 morphologically distinct types of cones and the absence of rods in their retinas. (2) Considering their phylogenetic proximity and similar ecological characteristics, such as hunting strategy and type of prey, we expected to find similar values of cell density and spatial resolving power. (3) Given the different preferred habitats and microhabitats and considering the terrain theory proposed by Hughes [1977], we predicted the presence of an area centralis in the retinas of the arboreal P. olfersii and a visual streak in the retina of the terrestrial P. patagoniensis, which inhabits predominantly open areas. (4) Because snakes are usually predated by aerial predators, we predicted a higher cell density in the ventral retina of both species.
Materials and Methods
Fig. 1. Photographs of the arboreal P. olfersii (a) and the terrestrial P. patagoniensis (b). Courtesy of and reproduced with permis-
sion from Otavio Augusto Vuolo Marques, Brazil.
whereas the brown P. patagoniensis is camouflaged against the terrestrial background in grasslands. Based on ecological diversity and phylogenetic relatedness, Philodryas snakes represent an interesting model to test hypotheses regarding the correlation between retinal topographic specialization and the behavioral ecology and evolutionary history of a given species. In the present study, we used immunohistochemistry and Nissl staining in retinal wholemounts and stereological methods to quantitatively assess the total number and topographic distribution of photoreceptors and retinal ganglion cells and estimate the spatial resolving power of Philodryas eyes. Light microscopy and scanning electron microscopy were used to verify the morphologically distinct photoreceptor types.
Light Microscopy and Scanning Electron Microscopy One eyecup of each species was fixed by immersion in ALFAC solution for 16 h (85 of 80% ethanol, 10 of 40% formaldehyde, and 5% of glacial acetic acid) and then rinsed in 70% ethanol. The tissues were embedded in paraffin, and 5-μm radial sections were cut in a microtome and placed on glass slides for hematoxylin and eosin staining. For scanning electron microscopy, one retina of each species was fixed overnight in 3% glutaraldehyde in 0.1 M phosphate buffer, pH 7.2, at 4 ° C. After fixation, the tissues were washed with buffer, dehydrated with increasing concentrations of ethyl alcohol, and critical-point-dried using liquid CO2. The tissues were then fragmented with the fine tip of either a pair of forceps or a pulled glass pipette, mounted in a specimen stub, and sputter coated with
Retinal Topography and Visual Acuity in Dipsadidae Snakes
Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
199
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
b
Specimens Twelve adult snakes of each species were provided by the Butantan Institute, São Paulo, Brazil. The animal procedures were in accordance with the ethical principles of animal management and experimentation established by the Brazilian Animal Experiment College (COBAE). After 2 h of dark adaptation to promote pigment epithelium retraction, the snakes were euthanized with a lethal injection of 30 mg/kg sodium thiopental (Thionembutal). Following euthanasia, gender was determined, and the following anatomical measures were obtained: body mass, body length, head width, and head length. After enucleation, the axial length of the eyes was measured, and the cornea and ciliary body were removed. A small radial incision was made in the dorsal region for later orientation of the retina. The lenses were removed, and their diameters were measured. The retinas were gently dissected from the choroid, and the pigment epithelium was easily separated from the retina at the moment of dissection. The dense vitreous humor was cut from the retina, and the remnant could be removed during the wholemounting process. The retinas were fixed in different solutions according to the following procedures.
ng
ng gc
Labeling Photoreceptors Using Immunohistochemistry One eyecup of each species was used to obtain radial sections. The eyecups were fixed by immersion in 4% PFA diluted in 0.1 M phosphate-buffered saline (PBS), pH 7.4, for 3 h and then rinsed in 0.1 M PBS. After cryoprotection with 30% sucrose in PBS for 72 h, the eyecups were embedded in Tissue-Tek OCT, frozen, and cryosectioned at 12-μm thicknesses (JUNG GM 3000 Leica, Nussloch, Germany). The sections were collected onto gelatinized glass slides and stored at –20 ° C until use. Radial sections were incubated with different antibodies to test their immunoreactivity and specificity. For retinal wholemount preparations, the retinas were carefully dissected free of the eyecup as described above. The isolated retinas were fixed by immersion in 4% PFA in 0.1 M PBS, pH 7.4, for 2 h, rinsed in 0.1 M PBS, and stored at 4 ° C until use. For opsin labeling, whole retinas were processed free-floating and preincubated for 1 h in 10% normal goat serum (Sigma-Aldrich, St. Louis, Mo., USA) diluted in 0.1 M PBS with 0.3% Triton X-100 to block nonspecific binding sites. The retinas were then incubated in the primary antiserum (JH492 or JH455) diluted 1: 3,000 in 0.1 M PBS with 0.3% Triton X-100 for 3 days at 4 ° C and then rinsed 3 times in 0.1 M PBS with 0.3% Triton X-100 for 10 min each. Radial sections were incubated at room temperature in the primary antiserum for 1 day. The tissues were incubated in the secondary antibody goat anti-rabbit immunoglobulin G (whole molecule; 1: 200; Jackson Immunoresearch Laboratories, West Grove, Pa., USA) coupled to the fluorescent molecule rhodamine (TRICT) diluted in 0.1 M PBS with 0.3% Triton for 2 h. Subsequently, they were rinsed 3 times in 0.1 M PBS for 10 min each and mounted with the photoreceptor layer facing up on a glass slide. Secondary antibody specificity was assessed by incubating retinal sections in buffer solution with omission of the primary antibody. No labeling by the secondary antibody was detected. Strategic cuts were made in the whole retinas to allow them to be flattened onto the slide. The slides were mounted with paraphenylenediamine (0.001%; Sigma), diluted in glycerol and PBS (1: 1), coverslipped, and observed under a fluorescent microscope (Leica DMRXE) with a ×100 oil immersion objective [numerical aperture (NA) = 1.25] or a ×40 oil immersion objective (NA = 0.7). The immunolabeled wholemounts were observed with a set of filters specific for TRITC (excitation green, emission red). The photoreceptor population was viewed by adjusting the focus on the photoreceptor inner segment level. The different photoreceptor types, labeled by the antibodies, were observed by adjusting the focus on the photoreceptor outer segment layer (fig. 3). These same procedures have been successfully used in our laboratory to identify turtle photoreceptors [Grotzner et al., 2004, 2005].
Fig. 2. Photomicrograph of the Nissl-stained retinal GCL of P. pa-
tagoniensis showing ganglion cells bodies (gc) and examples of nonganglion cells (ng), which may be displaced amacrine cells or glial cells. This digital image was processed using Adobe Photoshop CS3 for scaling, resolution, and adjustment of the levels of brightness and contrast. Scale bar = 20 μm. No pixel adjustments or manipulation of the captured images were made, with the exception of adjustment of the contrast and brightness using Adobe Photoshop CS3.
gold. The samples were examined using a JEOL JSM-6360LV scanning electron microscope. For each individual retina, images of 10 fields were captured and analyzed. Tissue Processing and Preparation of Nissl-Stained Retinal Wholemounts Neurons in the ganglion cell layer (GCL) were Nissl stained, and their density and distribution were determined in retinal wholemounts. The dissected retinas, previously fixed in 4% paraformaldehyde (PFA), were wholemounted on gelatinized slides with the GCL facing up, and radial cuts were made to allow them to be flattened and adhere to the slide. The retinas were immersed in ethanol (90%) and formaldehyde (10%) solution for 24 h, immersed in 4% PFA for 1 h, and rehydrated by passing through a decreasing series of ethanol. The tissues were Nissl stained in an aqueous solution of 0.05% cresyl violet for approximately 10 min, dehydrated in ethanol, cleared in xylene, and coverslipped with DPX (Aldrich). Because the retinal wholemounts were well attached to the slide during all of the staining steps, shrinkage was confined to the borders of the ora serrata and the edges of radial cuts and disregarded [Wässle et al., 1981; Peichl, 1992; Coimbra et al., 2012]. In the Nissl-stained wholemounts, the same cytological criteria proposed by Wong [1989] for the snake retina were used to distinguish neuronal and nonneuronal cells in the GCL (fig. 2). Glial cells, identified by their small soma size and round and darkly stained profile [Ehrlich, 1981; Hart et al., 2012], were not included in the counts. However, ganglion cells were not differentiated from ‘displaced’ amacrine cells [Ehrlich, 1981; Hayes, 1984;
200
Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
Antibody Characterization and Specificity Primary incubation with anti-opsin antibodies was performed with 2 rabbit polyclonal antibodies, i.e. JH492 and JH455, both developed and kindly provided by Jeremy Nathans and collaborators [Nathans et al., 1986] from The Johns Hopkins University School of Medicine, Baltimore, Md., USA.
Hauzman /Bonci /Grotzner /Mela /Liber / Martins /Ventura
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
gc
Hart, 2002] in any of the wholemounts because we could not reliably distinguish between the two cell types using cytological criteria.
a
b
sc Fig. 3. Sampled field images of wholemounted retinas of P. patagoniensis immunolabeled with the antibodies JH492 (a, b) and JH455 (c). a Focus on the photoreceptor inner segment layer. All
Antisera JH492 and JH455 have been characterized by Wang et al. [1992]. DNA segments that encode the last 38 amino acids of the human red pigment (RQFRNCILQLFGKKVDDGSELSSASKTEVSSVSSVSPA; GenBank accession No. NM020061) and the last 42 amino acids of the human blue pigment (NKQFQACIMKMVCGKAMTDESDTCSSQKTEVSTVSSTQVGPN; GenBank accession No. NM001708) were separately inserted into the polylinker of the T7 gene 10 expression vector pGEMEX (Promega). Each cone pigment-derived peptide was produced as a carboxyterminal extension of the T7 gene 10 protein [Studier et al., 1990]. The 4 fusion proteins were purified by preparative sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and used to immunize rabbits. A full description of their characteristics and staining properties was reported by Schiviz et al. [2008].
sc
vc vc X c
Density and Topographic Distribution of Neurons across the Retinas The distribution of the photoreceptor population and Nisslstained cells in the GCL of each wholemount was assessed using systematic random sampling and the fractionator principle
[Gundersen, 1977; Coimbra et al., 2009, 2012; Ullmann et al., 2012; Lisney et al., 2013]. The coordinates of the outer edges of the retinas were obtained and plotted in an Excel file. Counts were made at regular intervals that were defined by a sampling grid (0.5 × 0.5 mm) by viewing the tissue directly under a Leica DMRXE compound microscope with a ×100 oil immersion objective (NA = 1.25) or a ×40 oil immersion objective (NA = 0.7) equipped with a Nikon Digital Sight DS-U3 DSRi1 camera and NIS-Elements AR Microscope Imaging software (Nikon Instruments, Melville, N.Y., USA). An unbiased counting frame (74 × 74 μm or 184 × 184 μm) was imposed at each sampling frame (table 1), and cells were counted if they lay entirely within the counting frame or if they touched the acceptance lines without touching the rejection lines [Gundersen, 1977]. The cell counts at each sample point were entered into a Microsoft Excel spreadsheet and converted to density counts (in cells · mm–2) [Lisney et al., 2013]. The total number of cells was determined for each retina by multiplying the total number of sampled cells by the inverse of the area sampling fraction (asf), which is the area of the counting frame divided by the area of the
Retinal Topography and Visual Acuity in Dipsadidae Snakes
Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
201
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
of the cells in the sampled fields were counted to estimate the photoreceptor density. b Same sampled field shown in a, with a focus on the immunolabeled outer segments. Large single and double cones could not be distinguished from each other. Examples of 2 small single cones unlabeled with the antibody JH492 are identified in a and b (*). c Small single cone outer segments (sc) immunolabeled with JH455. Some very small unlabeled single cones can be identified (vc) by their smaller inner segments. The digital images were processed using Adobe Photoshop CS3 for scaling, resolution, and adjustment of the levels of brightness and contrast. Scale bars = 20 μm.
Table 1. Stereological parameters used to estimate the number and distribution of retinal GCL cells, photoreceptors, large single and
double cones, and small single cones in P. olfersii and P. patagoniensis retinas using the optical fractionator method Cell type
Counting frame, μm Grid, μm
asf
Objective/NA
GCL cells Photoreceptors Large single and double cones Small single cones
74 × 74 74 × 74 74 × 74 184 × 184
0.0219 0.0219 0.0219 0.1354
100×/1.25 oil 100×/1.25 oil 100×/1.25 oil 40×/0.7 oil
500 × 500 500 × 500 500 × 500 500 × 500
Table 2. Stereological analysis of photoreceptors in the retinas of P. olfersii and P. patagoniensis Species
Retinal area, mm2
Sites Estimated CE counted, photoreceptors, n n
Average photoreceptor density, cells∙mm–2
Estimated CE SSC, n
Average SSC, % SSC density, cells∙mm–2
Estimated LSC and DC, n
CE
Average LSC and DC density, cells∙mm–2
LSC and DC, %
P. olfersii Female No. 1 Male No. 2 Female No. 3 Female No. 4 Male No. 5 Mean ± SD
40.5 43.4 46.2 49.6 50.6 46.1 ± 4.2
153 159 155 189 179
388,628 444,334 342,565 417,278 468,959 412,353 ± 49,205
0.010 0.009 0.010 0.009 0.009
10,160 ± 3,000 11,178 ± 2,508 8,840 ± 1,775 8,831 ± 1,748 10,480 ± 1,907 9,898 ± 1,037
15,817 29,168 – – –
0.023 0.016 – – –
388 ± 199 731 ± 229 – – – 598 ± 187
4.1 6.6 – – – 4.8 ± 2.5
– – – 276,548 348,118 391,080 338,582 ± 57,858
– – – 0.013 0.012 0.012
– – – 7,183 ± 1,819 7,597 ± 1,873 8,883 ± 1,952 7,888 ± 886
– – – 80.7 83.4 83.4 82.5 ± 1.5
P. patagoniensis Female No. 1 Female No. 2 Female No. 3 Female No. 4 Male No. 5 Mean ± SD
45.56 41.67 39,12 51.18 43.51 44.21 ± 4.6
159 142 140 188 161
382,133 351,690 381,416 538,293 372,410 405,188±75,415
0.010 0.010 0.010 0.008 0.010
9,613 ± 2,649 9,907 ± 2,427 10,898 ± 2,899 11,453 ± 2,951 9,252 ± 2,595 10,225 ± 920
18,845 16,535 22,142 – –
0.020 0.020 0.050 – –
508 ± 163 505 ± 146 566 ± 243 – – 526 ± 35
5.3 5.1 5.8 – – 5.2±0.1
– – – 438,765 304,369 371,567 ± 95,032
– – – 0.010 0.010
– – – 9,436 ± 2,513 6,995 ± 2,034 8,216 ± 1,726
– – – 81.5 81.7 81.6 ± 0.2
CE = Scheaffer’s coefficient of error; SSC = small single cones; LSC = large single cones; DC = double cones.
202
Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
Statistical Analysis Statistical analyses were performed using SPSS Statistics v.20 (IBM Corporation, Armonk, N.Y., USA). All of the data were log10 transformed prior to analysis. The nonparametric Mann-Whitney test was performed to compare the total number of cells and the average and peak cell densities of each group of cells, photoreceptors, large single and double cones, small single cones, and GCL cells between the two species. To compare wholemount retinal areas, which had a higher sample size and a normal distribution, a parametric t test was performed. Anatomical Assessment of the Focal Length and Estimation of the Spatial Resolving Power Estimates of the theoretical peak spatial resolving power were calculated using the maximum density of presumed ganglion cells and the estimated focal length. Focal length was assessed by freezing and sectioning one eye of each species on the axial and equatorial planes [Lisney and Collin, 2008]. The eyes were rapidly frozen in a mixture of dry ice and ethanol [Sivak, 1978] and stored at –80 ° C. After noting the orientation, each eye was embedded in Tissue-Tek OCT compound (Sakura Finetechnical Co., Ltd., To
Hauzman /Bonci /Grotzner /Mela /Liber / Martins /Ventura
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
sampling grid, according to the following formula: N total = RQ × 1/asf, where RQ is the sum of the total neurons counted [West et al., 1991; Coimbra et al., 2009, 2013]. The counting frame, the asf, and the objective magnification used to estimate the total number and topographic distribution of the different cell types are listed in table 1. Coefficients of error were calculated using the method of Scheaffer et al. [1996] and were
Received: August 8, 2013 Returned for revision: September 10, 2013 Accepted after third revision: June 16, 2014 Published online: October 23, 2014
Comparative Study of Photoreceptor and Retinal Ganglion Cell Topography and Spatial Resolving Power in Dipsadidae Snakes Einat Hauzman a, b Daniela M.O. Bonci a, b Sonia R. Grotzner c Maritana Mela c André M.P. Liber a, b Sonia L. Martins a, b Dora F. Ventura a, b a
Departamento de Psicologia Experimental, Instituto de Psicologia, and b Núcleo de Neurociências e Comportamento, Universidade de São Paulo, São Paulo, and c Departamento de Biologia Celular, Universidade Federal do Paraná, Curitiba, Brazil
Abstract The diurnal Dipsadidae snakes Philodryas olfersii and P. patagoniensis are closely related in their phylogeny but inhabit different ecological niches. P. olfersii is arboreal, whereas P. patagoniensis is preferentially terrestrial. The goal of the present study was to compare the density and topography of neurons, photoreceptors, and cells in the ganglion cell layer in the retinas of these two species using immunohistochemistry and Nissl staining procedures and estimate the spatial resolving power of their eyes based on the ganglion cell peak density. Four morphologically distinct types of cones were observed by scanning electron microscopy, 3 of which were labeled with anti-opsin antibodies: large single cones and double cones labeled by the antibody JH492 and small single cones labeled by the antibody JH455. The average densities of photoreceptors and neurons in the ganglion cell layer were similar in both species (∼10,000 and 7,000 cells · mm–2, respectively). The estimated spatial resolving
© 2014 S. Karger AG, Basel 0006–8977/14/0843–0197$39.50/0 E-Mail [email protected] www.karger.com/bbe
power was also similar, ranging from 2.4 to 2.7 cycles · degree–1. However, the distribution of neurons had different specializations. In the arboreal P. olfersii, the isodensity maps had a horizontal visual streak, with a peak density in the central region and a lower density in the dorsal retina. This organization might be relevant for locomotion and hunting behavior in the arboreal layer. In the terrestrial P. patagoniensis, a concentric pattern of decreasing cell density emanated from an area centralis located in the naso-ventral retina. Lower densities were observed in the dorsal region. The ventrally high density improves the resolution in the superior visual field and may be an important adaptation for terrestrial snakes to perceive the approach of predators from above. © 2014 S. Karger AG, Basel
Introduction
Snakes comprise a highly diversified reptilian group (suborder Serpentes) that has successfully radiated into a wide range of ecological niches. Representatives of the ∼3,400 extant species are found in virtually every portion Einat Hauzman Instituto de Psicologia, Universidade de São Paulo Avenida Professor Mello Moraes, 1721, Bloco A, Sala D-9 São Paulo, SP 05508-030 (Brazil) E-Mail hauzman.einat @ gmail.com
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
Key Words Visual ecology · Snakes · Retina · Photoreceptors · Ganglion cells · Topography · Visual acuity
198
Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
(‘basal’) snakes and belong to the Rh1, SWS1, and LWS opsin groups, which generate rods and SWS and LWS cones, respectively [Davies et al., 2009]. In addition to the different types of photoreceptors and visual pigments expressed in vertebrate retinas, comparative studies have shown that the distribution and density of retinal neurons are variable across species and that retinal topography appears to be correlated with the species’ habitat or lifestyle [Hughes, 1977; Collin, 1999, 2008]. Numerous studies have used ganglion cell spacing in the areas of highest cell density in combination with the eye’s focal length to estimate the maximum spatial resolution in a number of vertebrate species [e.g. Pettigrew et al., 1988; Collin and Pettigrew, 1989; Pettigrew and Manger, 2008; Hart et al., 2012; Coimbra et al., 2013]. To date, the distribution of neurons in snake retinas has been evaluated only in a few of the great diversity of snakes species: the terrestrial T. sirtalis [Wong, 1989] and 3 species of sea snakes, i.e. Lapemis curtus, Aipysurus laevis, and Disteira major [Hart et al., 2012]. The present study investigated the retinas of 2 species of Dipsadidae snakes, i.e. Philodryas olfersii and P. patagoniensis (fig. 1), which are closely related, sympatric, and widely distributed in South America [Thomas, 1976]. They are both active only during the day, and both are fast moving snakes that constrict or envenom their prey [Sazima and Haddad, 1992; Hartmann and Marques, 2005]. They are dietary generalists, feeding on a large variety of small vertebrates, such as amphibians, lizards, other snakes, birds, and mammals [Amaral, 1978; Lopez, 2003; Hartmann and Marques, 2005]. Despite dwelling in both open and forested areas, P. olfersii is found mostly in forested areas and forest edges and is regarded as a semiarboreal snake, whereas P. patagoniensis is more frequently observed in open habitats and is viewed as essentially terrestrial [Sazima and Haddad, 1992; Fowler and Salomão, 1994; Marques et al., 2001; Hartmann and Marques, 2005]. This difference in preferred habitat is consistent with some morphological characteristics, such as body mass, body length, and color. The arboreal P. olfersii has a more slender body and a longer tail length than the terrestrial P. patagoniensis, characteristics that benefit locomotion in the arboreal layer by allowing greater equilibrium and better distribution of the body mass on branches [Lillywhite and Henderson, 1993; Martins et al., 2001]. The color differences of the two species are also consistent with differences in the frequency of habitat and microhabitat use. The green P. olfersii is well camouflaged against the background of green foliage in the forest, Hauzman /Bonci /Grotzner /Mela /Liber / Martins /Ventura
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
of the biosphere, with the exception of polar regions, some islands, and the deep ocean [Cadle, 1987; Ford and Burghardt, 1993; Lillywhite and Henderson, 1993]. The great diversity of the Ophidian group can be explained by a number of adaptive radiations, and their evolutionary success is related to the development of some specialized sensory organs [Ford and Burghardt, 1993]. Vision may represent an important sensory modality that is employed by some snake species, and the anatomical variety of snake eyes indicates how environmental conditions have influenced the evolution of ocular structure and function [Walls, 1942]. The morphological diversity of photoreceptors in the retinas of different snake species has been described in several studies [Walls, 1942; Underwood, 1967, 1970; Wong, 1989; Hart et al., 2012]. Both rods and cones are found in numerous nocturnal species from different families [Walls, 1942; Underwood, 1967, 1970; Sillman et al., 1999, 2001; Davies et al., 2009]. However, diurnal Caenophidian (‘advanced’) snakes tend to have only cones [Walls, 1942; Underwood, 1967, 1970; Wong, 1989; Sillman et al., 1997; Hart et al., 2012]. For example, Wong [1989] found no evidence of rod photoreceptors in the retina of the garter snake Thamnophis sirtalis, a diurnal Colubridae species, and described 3 classes of cones according to their morphologies: large single cones, small single cones, and double cones that consist of principal and accessory members. Later studies on the same species [Sillman et al., 1997] described a fourth cone-like photoreceptor, a very small single cone. In sea snakes from the Hidrophiidae family, Hart et al. [2012] also described the same morphologically distinct types of cones and the absence of rod-like photoreceptors. The spectral absorbance of photoreceptors has been measured in a number of species of Boidae, Colubridae, Viperidae, and Hydrophiidae snakes [Govardovskii and Chkheidze, 1989; Jacobs et al., 1992; Sillman et al., 1997, 1999, 2001; Davies et al., 2009; Hart et al., 2012]. In the Caenophidian snake T. sirtalis, 3 absorption peaks were distinguished [Sillman et al., 1997]: an LWS visual pigment (554 nm) in large single cones and the principal and accessory members of double cones, an SWS pigment (360 nm) in small single cones, and an MWS pigment (482 nm) in very small single cones. In retinas of Hydrophiidae sea snakes, double cones and large single cones also contain LWS visual pigments (555–559 nm). Small single cones have SWS pigments (428–430 nm), and a second class of small single cones contains an MWS pigment [496 nm; Hart et al., 2012]. The visual pigments were genetically classified into 2 species of Henophidian
Color version available online
a
We tested 4 hypotheses. (1) Because both Caenophidian snakes are diurnal, we expected to find 4 morphologically distinct types of cones and the absence of rods in their retinas. (2) Considering their phylogenetic proximity and similar ecological characteristics, such as hunting strategy and type of prey, we expected to find similar values of cell density and spatial resolving power. (3) Given the different preferred habitats and microhabitats and considering the terrain theory proposed by Hughes [1977], we predicted the presence of an area centralis in the retinas of the arboreal P. olfersii and a visual streak in the retina of the terrestrial P. patagoniensis, which inhabits predominantly open areas. (4) Because snakes are usually predated by aerial predators, we predicted a higher cell density in the ventral retina of both species.
Materials and Methods
Fig. 1. Photographs of the arboreal P. olfersii (a) and the terrestrial P. patagoniensis (b). Courtesy of and reproduced with permis-
sion from Otavio Augusto Vuolo Marques, Brazil.
whereas the brown P. patagoniensis is camouflaged against the terrestrial background in grasslands. Based on ecological diversity and phylogenetic relatedness, Philodryas snakes represent an interesting model to test hypotheses regarding the correlation between retinal topographic specialization and the behavioral ecology and evolutionary history of a given species. In the present study, we used immunohistochemistry and Nissl staining in retinal wholemounts and stereological methods to quantitatively assess the total number and topographic distribution of photoreceptors and retinal ganglion cells and estimate the spatial resolving power of Philodryas eyes. Light microscopy and scanning electron microscopy were used to verify the morphologically distinct photoreceptor types.
Light Microscopy and Scanning Electron Microscopy One eyecup of each species was fixed by immersion in ALFAC solution for 16 h (85 of 80% ethanol, 10 of 40% formaldehyde, and 5% of glacial acetic acid) and then rinsed in 70% ethanol. The tissues were embedded in paraffin, and 5-μm radial sections were cut in a microtome and placed on glass slides for hematoxylin and eosin staining. For scanning electron microscopy, one retina of each species was fixed overnight in 3% glutaraldehyde in 0.1 M phosphate buffer, pH 7.2, at 4 ° C. After fixation, the tissues were washed with buffer, dehydrated with increasing concentrations of ethyl alcohol, and critical-point-dried using liquid CO2. The tissues were then fragmented with the fine tip of either a pair of forceps or a pulled glass pipette, mounted in a specimen stub, and sputter coated with
Retinal Topography and Visual Acuity in Dipsadidae Snakes
Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
199
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
b
Specimens Twelve adult snakes of each species were provided by the Butantan Institute, São Paulo, Brazil. The animal procedures were in accordance with the ethical principles of animal management and experimentation established by the Brazilian Animal Experiment College (COBAE). After 2 h of dark adaptation to promote pigment epithelium retraction, the snakes were euthanized with a lethal injection of 30 mg/kg sodium thiopental (Thionembutal). Following euthanasia, gender was determined, and the following anatomical measures were obtained: body mass, body length, head width, and head length. After enucleation, the axial length of the eyes was measured, and the cornea and ciliary body were removed. A small radial incision was made in the dorsal region for later orientation of the retina. The lenses were removed, and their diameters were measured. The retinas were gently dissected from the choroid, and the pigment epithelium was easily separated from the retina at the moment of dissection. The dense vitreous humor was cut from the retina, and the remnant could be removed during the wholemounting process. The retinas were fixed in different solutions according to the following procedures.
ng
ng gc
Labeling Photoreceptors Using Immunohistochemistry One eyecup of each species was used to obtain radial sections. The eyecups were fixed by immersion in 4% PFA diluted in 0.1 M phosphate-buffered saline (PBS), pH 7.4, for 3 h and then rinsed in 0.1 M PBS. After cryoprotection with 30% sucrose in PBS for 72 h, the eyecups were embedded in Tissue-Tek OCT, frozen, and cryosectioned at 12-μm thicknesses (JUNG GM 3000 Leica, Nussloch, Germany). The sections were collected onto gelatinized glass slides and stored at –20 ° C until use. Radial sections were incubated with different antibodies to test their immunoreactivity and specificity. For retinal wholemount preparations, the retinas were carefully dissected free of the eyecup as described above. The isolated retinas were fixed by immersion in 4% PFA in 0.1 M PBS, pH 7.4, for 2 h, rinsed in 0.1 M PBS, and stored at 4 ° C until use. For opsin labeling, whole retinas were processed free-floating and preincubated for 1 h in 10% normal goat serum (Sigma-Aldrich, St. Louis, Mo., USA) diluted in 0.1 M PBS with 0.3% Triton X-100 to block nonspecific binding sites. The retinas were then incubated in the primary antiserum (JH492 or JH455) diluted 1: 3,000 in 0.1 M PBS with 0.3% Triton X-100 for 3 days at 4 ° C and then rinsed 3 times in 0.1 M PBS with 0.3% Triton X-100 for 10 min each. Radial sections were incubated at room temperature in the primary antiserum for 1 day. The tissues were incubated in the secondary antibody goat anti-rabbit immunoglobulin G (whole molecule; 1: 200; Jackson Immunoresearch Laboratories, West Grove, Pa., USA) coupled to the fluorescent molecule rhodamine (TRICT) diluted in 0.1 M PBS with 0.3% Triton for 2 h. Subsequently, they were rinsed 3 times in 0.1 M PBS for 10 min each and mounted with the photoreceptor layer facing up on a glass slide. Secondary antibody specificity was assessed by incubating retinal sections in buffer solution with omission of the primary antibody. No labeling by the secondary antibody was detected. Strategic cuts were made in the whole retinas to allow them to be flattened onto the slide. The slides were mounted with paraphenylenediamine (0.001%; Sigma), diluted in glycerol and PBS (1: 1), coverslipped, and observed under a fluorescent microscope (Leica DMRXE) with a ×100 oil immersion objective [numerical aperture (NA) = 1.25] or a ×40 oil immersion objective (NA = 0.7). The immunolabeled wholemounts were observed with a set of filters specific for TRITC (excitation green, emission red). The photoreceptor population was viewed by adjusting the focus on the photoreceptor inner segment level. The different photoreceptor types, labeled by the antibodies, were observed by adjusting the focus on the photoreceptor outer segment layer (fig. 3). These same procedures have been successfully used in our laboratory to identify turtle photoreceptors [Grotzner et al., 2004, 2005].
Fig. 2. Photomicrograph of the Nissl-stained retinal GCL of P. pa-
tagoniensis showing ganglion cells bodies (gc) and examples of nonganglion cells (ng), which may be displaced amacrine cells or glial cells. This digital image was processed using Adobe Photoshop CS3 for scaling, resolution, and adjustment of the levels of brightness and contrast. Scale bar = 20 μm. No pixel adjustments or manipulation of the captured images were made, with the exception of adjustment of the contrast and brightness using Adobe Photoshop CS3.
gold. The samples were examined using a JEOL JSM-6360LV scanning electron microscope. For each individual retina, images of 10 fields were captured and analyzed. Tissue Processing and Preparation of Nissl-Stained Retinal Wholemounts Neurons in the ganglion cell layer (GCL) were Nissl stained, and their density and distribution were determined in retinal wholemounts. The dissected retinas, previously fixed in 4% paraformaldehyde (PFA), were wholemounted on gelatinized slides with the GCL facing up, and radial cuts were made to allow them to be flattened and adhere to the slide. The retinas were immersed in ethanol (90%) and formaldehyde (10%) solution for 24 h, immersed in 4% PFA for 1 h, and rehydrated by passing through a decreasing series of ethanol. The tissues were Nissl stained in an aqueous solution of 0.05% cresyl violet for approximately 10 min, dehydrated in ethanol, cleared in xylene, and coverslipped with DPX (Aldrich). Because the retinal wholemounts were well attached to the slide during all of the staining steps, shrinkage was confined to the borders of the ora serrata and the edges of radial cuts and disregarded [Wässle et al., 1981; Peichl, 1992; Coimbra et al., 2012]. In the Nissl-stained wholemounts, the same cytological criteria proposed by Wong [1989] for the snake retina were used to distinguish neuronal and nonneuronal cells in the GCL (fig. 2). Glial cells, identified by their small soma size and round and darkly stained profile [Ehrlich, 1981; Hart et al., 2012], were not included in the counts. However, ganglion cells were not differentiated from ‘displaced’ amacrine cells [Ehrlich, 1981; Hayes, 1984;
200
Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
Antibody Characterization and Specificity Primary incubation with anti-opsin antibodies was performed with 2 rabbit polyclonal antibodies, i.e. JH492 and JH455, both developed and kindly provided by Jeremy Nathans and collaborators [Nathans et al., 1986] from The Johns Hopkins University School of Medicine, Baltimore, Md., USA.
Hauzman /Bonci /Grotzner /Mela /Liber / Martins /Ventura
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
gc
Hart, 2002] in any of the wholemounts because we could not reliably distinguish between the two cell types using cytological criteria.
a
b
sc Fig. 3. Sampled field images of wholemounted retinas of P. patagoniensis immunolabeled with the antibodies JH492 (a, b) and JH455 (c). a Focus on the photoreceptor inner segment layer. All
Antisera JH492 and JH455 have been characterized by Wang et al. [1992]. DNA segments that encode the last 38 amino acids of the human red pigment (RQFRNCILQLFGKKVDDGSELSSASKTEVSSVSSVSPA; GenBank accession No. NM020061) and the last 42 amino acids of the human blue pigment (NKQFQACIMKMVCGKAMTDESDTCSSQKTEVSTVSSTQVGPN; GenBank accession No. NM001708) were separately inserted into the polylinker of the T7 gene 10 expression vector pGEMEX (Promega). Each cone pigment-derived peptide was produced as a carboxyterminal extension of the T7 gene 10 protein [Studier et al., 1990]. The 4 fusion proteins were purified by preparative sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and used to immunize rabbits. A full description of their characteristics and staining properties was reported by Schiviz et al. [2008].
sc
vc vc X c
Density and Topographic Distribution of Neurons across the Retinas The distribution of the photoreceptor population and Nisslstained cells in the GCL of each wholemount was assessed using systematic random sampling and the fractionator principle
[Gundersen, 1977; Coimbra et al., 2009, 2012; Ullmann et al., 2012; Lisney et al., 2013]. The coordinates of the outer edges of the retinas were obtained and plotted in an Excel file. Counts were made at regular intervals that were defined by a sampling grid (0.5 × 0.5 mm) by viewing the tissue directly under a Leica DMRXE compound microscope with a ×100 oil immersion objective (NA = 1.25) or a ×40 oil immersion objective (NA = 0.7) equipped with a Nikon Digital Sight DS-U3 DSRi1 camera and NIS-Elements AR Microscope Imaging software (Nikon Instruments, Melville, N.Y., USA). An unbiased counting frame (74 × 74 μm or 184 × 184 μm) was imposed at each sampling frame (table 1), and cells were counted if they lay entirely within the counting frame or if they touched the acceptance lines without touching the rejection lines [Gundersen, 1977]. The cell counts at each sample point were entered into a Microsoft Excel spreadsheet and converted to density counts (in cells · mm–2) [Lisney et al., 2013]. The total number of cells was determined for each retina by multiplying the total number of sampled cells by the inverse of the area sampling fraction (asf), which is the area of the counting frame divided by the area of the
Retinal Topography and Visual Acuity in Dipsadidae Snakes
Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
201
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
of the cells in the sampled fields were counted to estimate the photoreceptor density. b Same sampled field shown in a, with a focus on the immunolabeled outer segments. Large single and double cones could not be distinguished from each other. Examples of 2 small single cones unlabeled with the antibody JH492 are identified in a and b (*). c Small single cone outer segments (sc) immunolabeled with JH455. Some very small unlabeled single cones can be identified (vc) by their smaller inner segments. The digital images were processed using Adobe Photoshop CS3 for scaling, resolution, and adjustment of the levels of brightness and contrast. Scale bars = 20 μm.
Table 1. Stereological parameters used to estimate the number and distribution of retinal GCL cells, photoreceptors, large single and
double cones, and small single cones in P. olfersii and P. patagoniensis retinas using the optical fractionator method Cell type
Counting frame, μm Grid, μm
asf
Objective/NA
GCL cells Photoreceptors Large single and double cones Small single cones
74 × 74 74 × 74 74 × 74 184 × 184
0.0219 0.0219 0.0219 0.1354
100×/1.25 oil 100×/1.25 oil 100×/1.25 oil 40×/0.7 oil
500 × 500 500 × 500 500 × 500 500 × 500
Table 2. Stereological analysis of photoreceptors in the retinas of P. olfersii and P. patagoniensis Species
Retinal area, mm2
Sites Estimated CE counted, photoreceptors, n n
Average photoreceptor density, cells∙mm–2
Estimated CE SSC, n
Average SSC, % SSC density, cells∙mm–2
Estimated LSC and DC, n
CE
Average LSC and DC density, cells∙mm–2
LSC and DC, %
P. olfersii Female No. 1 Male No. 2 Female No. 3 Female No. 4 Male No. 5 Mean ± SD
40.5 43.4 46.2 49.6 50.6 46.1 ± 4.2
153 159 155 189 179
388,628 444,334 342,565 417,278 468,959 412,353 ± 49,205
0.010 0.009 0.010 0.009 0.009
10,160 ± 3,000 11,178 ± 2,508 8,840 ± 1,775 8,831 ± 1,748 10,480 ± 1,907 9,898 ± 1,037
15,817 29,168 – – –
0.023 0.016 – – –
388 ± 199 731 ± 229 – – – 598 ± 187
4.1 6.6 – – – 4.8 ± 2.5
– – – 276,548 348,118 391,080 338,582 ± 57,858
– – – 0.013 0.012 0.012
– – – 7,183 ± 1,819 7,597 ± 1,873 8,883 ± 1,952 7,888 ± 886
– – – 80.7 83.4 83.4 82.5 ± 1.5
P. patagoniensis Female No. 1 Female No. 2 Female No. 3 Female No. 4 Male No. 5 Mean ± SD
45.56 41.67 39,12 51.18 43.51 44.21 ± 4.6
159 142 140 188 161
382,133 351,690 381,416 538,293 372,410 405,188±75,415
0.010 0.010 0.010 0.008 0.010
9,613 ± 2,649 9,907 ± 2,427 10,898 ± 2,899 11,453 ± 2,951 9,252 ± 2,595 10,225 ± 920
18,845 16,535 22,142 – –
0.020 0.020 0.050 – –
508 ± 163 505 ± 146 566 ± 243 – – 526 ± 35
5.3 5.1 5.8 – – 5.2±0.1
– – – 438,765 304,369 371,567 ± 95,032
– – – 0.010 0.010
– – – 9,436 ± 2,513 6,995 ± 2,034 8,216 ± 1,726
– – – 81.5 81.7 81.6 ± 0.2
CE = Scheaffer’s coefficient of error; SSC = small single cones; LSC = large single cones; DC = double cones.
202
Brain Behav Evol 2014;84:197–213 DOI: 10.1159/000365275
Statistical Analysis Statistical analyses were performed using SPSS Statistics v.20 (IBM Corporation, Armonk, N.Y., USA). All of the data were log10 transformed prior to analysis. The nonparametric Mann-Whitney test was performed to compare the total number of cells and the average and peak cell densities of each group of cells, photoreceptors, large single and double cones, small single cones, and GCL cells between the two species. To compare wholemount retinal areas, which had a higher sample size and a normal distribution, a parametric t test was performed. Anatomical Assessment of the Focal Length and Estimation of the Spatial Resolving Power Estimates of the theoretical peak spatial resolving power were calculated using the maximum density of presumed ganglion cells and the estimated focal length. Focal length was assessed by freezing and sectioning one eye of each species on the axial and equatorial planes [Lisney and Collin, 2008]. The eyes were rapidly frozen in a mixture of dry ice and ethanol [Sivak, 1978] and stored at –80 ° C. After noting the orientation, each eye was embedded in Tissue-Tek OCT compound (Sakura Finetechnical Co., Ltd., To
Hauzman /Bonci /Grotzner /Mela /Liber / Martins /Ventura
Downloaded by: Universitätsbibliothek Giessen 134.176.129.147 - 5/17/2015 3:58:01 PM
sampling grid, according to the following formula: N total = RQ × 1/asf, where RQ is the sum of the total neurons counted [West et al., 1991; Coimbra et al., 2009, 2013]. The counting frame, the asf, and the objective magnification used to estimate the total number and topographic distribution of the different cell types are listed in table 1. Coefficients of error were calculated using the method of Scheaffer et al. [1996] and were
