Molecular Vision 2004; 10:83-92 <http://www.molvis.org/molvis/v10/a12/> Received 10 April 2003 | Accepted 14 January 2004 | Published 10 February 2004

Download Reprint

Comparative study of the three neurofilament subunits within pig and human retinal ganglion cells

Javier Ruiz-Ederra,¹ Mónica García,¹ David Hicks,² Elena Vecino¹
 
 

¹Universidad del País Vasco, Facultad de Medicina, Departamento Biología Celular, Leioa, Vizcaya, Spain; ²Laboratoire de Physiopathologie Rétinienne, INSERM, CHUR, Strasbourg, France

Correspondence to: Corresponding author: Elena Vecino, Universidad del País Vasco, Facultad Medicina, Departamento Biología, 48940 Leioa, Vizcaya, Spain; Phone: 34-946012820; FAX: 34-944648966; email: gcpvecoe@lg.ehu.es

Abstract

Purpose: Neurofilaments (NF) are neuronal cytoskeletal components and immunostaining against them has been used to visualize retinal ganglion cells (RGC) and their axons. Since the RGC cytoskeleton exhibits differential damage in diseases such as glaucoma, we examined the distribution of light, medium, and heavy NF subunits (NF-L, NF-M, and NF-H respectively) within normal human and porcine retinas, as a function of RGC soma size and eccentricity.

Methods: NF subunits were visualized with immunofluorescence techniques using retinal sections and flatmounts from adult human and pig retinas that were incubated with specific antisera against the three NF subunits. Porcine RGCs were retrogradely labeled with fluorogold while human RGCs were identified based on their position within the inner retina and their relatively large somata.

Results: NF-H and NF-M were distributed widely within all RGC somata and dendrites, whereas NF-L was more restricted to the perinuclear area. In addition, phosphorylated NF-H distribution varied with retinal eccentricity so a subpopulation of large RGCs located in the peripheral retina was intensely labeled with the antiserum recognizing the phosphorylated NF-H.

Conclusions: We show that at least one of each of three NF subunits is present in all RGCs in porcine and presumably in human retina, and that NF distribution is very similar in RGCs of both species.

Introduction

Neurofilaments (NFs) are the predominant intermediate filaments in mature neurons. They are assemblies of three subunits, NF-L (68 kDa), NF-M (160 kDa), and NF-H (200 KDa) [1]. These three components form heteropolymeric 10 nm filaments that run parallel along the length of the axon and are also present within the soma, serving primarily a structural function. However, not all neurons have the same combination of NF subunits and even between species there are clear differences. Thus, rodent NFs are obligate heteropolymers requiring NF-L plus either NF-M or NF-H for filament formation. However, human NF-L is capable of self-assembly, an important distinction from rodent NF-L subunits. Moreover, in human neurons NF-M cannot form homopolymers and requires the presence of NF-L for incorporation into filaments [2]. The functional meaning of these differences in the association of the three NF subunits has not been elucidated, but differential NF subunit content typifies particular subpopulations of neurons [3]. In the retina, NF immunoreactivity has been previously described mainly in the axons and processes of retinal ganglion cells (RGCs) [3-7], but depending on the species studied was also found in other types of neurons. Thus, NFs can be expressed in a subset of horizontal cells of many mammalian species, including rat, mouse, hamster, rabbit, guinea pig, cow, cat, or pig [5,8-12], and in bipolar cells of animals like rabbit and guinea-pig [5]. NFs are normally absent from amacrine cells, although they have been described in a subpopulation of this cell type in cat [13] and rabbit [12]. No immunolabeling for the NFs has been detected in photoreceptors.

The three protein subunits of NFs are normally associated but there are exceptions: in the inner retina of mouse the presence of NF-L has been described throughout large RGCs, while NF-H is only found in the more distal portions of the optic axons [4]. Distinct patterns of RGC labeling have also been seen in hamster: all three subunits are present in intraretinal axons, while most somata expressed NF-L and to a lesser extent NF-H, whereas very few were revealed with NF-M [11].

Neurofilaments play a major role in brain development, maintenance, regeneration, and plasticity of the neural cytoskeleton. There is growing evidence that NFs can affect the dynamics and perhaps the function of other cytoskeletal elements, such as microtubules and actin filaments [14]. Disorganized NFs can induce selective neuronal degeneration and death, and interference of axonal transport has been proposed as one possible mechanism of NF-induced pathology [15]. Thus an abnormal distribution of highly phosphorylated NF protein in the soma and dendrites of some RGCs is observed after nerve crush [16]. Furthermore, the excitatory neurotransmitter glutamate, which exhibits elevated extracellular levels in pathologies like glaucoma, can enhance the phosphorylation of NFs [17] and induce the accumulation of NFs in the neuronal soma [18]. It has also been demonstrated that changes in intraocular pressure can modify cytoskeletal organization in astrocytes [19] and different types of ocular tissues [20]. In agreement with this, it has been reported that accumulation of NF-L in RGCs occurs during glaucoma, which is normally associated with increased intraocular pressure [21].

The three NFs polymerize in the neuronal soma to form the so-called NF-triplet. Once assembled, the NF triplet is transported down the axon via slow axonal transport [22,23]. This axonal transport is regulated by the phosphorylation of NF subunits [24]. It has been suggested that NF-phosphorylation and intermediate filament content are correlated with axonal caliber [25-27]. However, other studies have not observed this correlation [28-31]. In support of the latter, it has been demonstrated that NF content is not the only determinant of axon diameter in the optic nerve since decreases in axonal caliber occur much later than the drop in NF expression following optic nerve crush [32].

Since expression of NF subunits differs among RGCs and these cells degenerate in diseases like glaucoma [33], we were interested in characterizing and comparing the pattern of NFs distribution in the retina. In the present study we principally used human retina, but we examined in parallel the pig retina for the following reasons: (1) in this species the whole population of RGCs can be labeled by retrogradely tracing with fluorogold, which is not possible in humans [34]; (2) the structure of the eye and retina is very similar to humans [35,36]; (3) it is possible to reproduce ocular diseases similar to human [37].

Methods

Tissue collection

In vivo analysis of the distribution of NF subunits in pig retina was performed by using four adult pig retinas. All experiments were carried out following the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Pigs were deeply anaesthetized with an intramuscular injection of Ketolar plus Propofol (each 20 mg/kg). Once the animal was anesthetized, an intravenous canula was applied to the ear in order to provide the animal with additional anesthetic (Propofol at 1 ml every 15 min) maintaining deep anesthesia throughout the operation. A life-support machine was used to facilitate breathing and to monitor vital functions during the operation. To identify RGCs, we labelled them with the fluorescent tracer fluorogold (Fluorochrome, Englewook CO, USA) diluted in 0.9% NaCl containing 3% fluorogold and 0.1% dimethylsulfoxide. The optic nerve of the left eye was exposed and a massive injection (total volume 40 μl) of 3% fluorogold was applied to the optic nerve around 4 mm from the optic nerve head. The unoperated right eyes were used as controls. Pigs were kept alive for two days post-operation to allow fluorogold to fill the entire population of RGCs. Then animals were euthanized with an overdose of anesthesia, the eyes were enucleated, and the lens and vitreous extracted. The eyecups were lightly fixed with 4% paraformaldehyde in 0.1 M phosphate buffer saline (PBS, pH 7.4) for 2 h at 4 °C and then retinas were removed. Fixed retinas were washed in PBS and then transferred to 30% sucrose in PBS overnight at 4 °C. Tissue was frozen in Tissue-Tek (Sakura Finetek, Zoeterwoude, The Netherlands) and sectioned at 14 μm in a cryostat. Some retinas were flatmounted before fixation and then transferred to PBS for subsequent immunostaining.

Two adult human retinas were obtained from donors 10 and 12 h post-mortem. Cryostat sections and flat mounts of the two retinas were performed following the method described above for pig retina. All experiments were carried out following the Declaration of Helsinki [38].

Immunohistochemistry

Pig and human retinal sections were kept at -20 °C until immunohistochemical analysis was performed. To analyze the distribution of NF subunits in flatmounted retinas, free-floating retinas were incubated in phosphate buffer (TX-PB solution) with 0.25% Triton-X 100 for 24 h at 4 °C (Sigma-Aldrich, St. Louis, Mo.) to improve permeability. Blocking was performed by using 1% bovine serum albumin (BSA) in TX-PB solution for 1 h at 4 °C. Samples were incubated using all combinations resulting from mixing one monoclonal with one polyclonal antibody against each different NF subunit for 48 h at 4 °C. Primary antibodies used in the present study were: polyclonal antibody NF-Hp directed against the phosphorylated form of NF-H subunit (16 μg/ml); polyclonal NF-L directed against the NF-L subunit (1:500), monoclonal NF-H directed against the NF-H subunit (30 μg/ml, clone N52) and monoclonal antibody NF-M directed against the NF-M subunit (24 μg/ml, clone NN18). In relation to potential phosphorylation-dependency of the epitopes recognized by these antibodies, the data sheets and previous publications [39] report that NF-H antibody recognizes equally well both phosphorylated and non-phosphorylated forms of 160 kDa NF subunits, NF-Hp reacts predominantly with the phosphorylated form of 200 kDa NF subunit, and NF-L recognizes the non-phosphorylated form of 68 kDa NF subunit. With respect to NF-H (clone N52), although the data sheet indicates that this antibody recognizes both phosphorylated and non-phosphorylated forms of 200 kDa, N52 does not appear to be a phosphorylation-independent antibody [40], and recognizes only the non-phosphorylated form of the NF-H subunit. All primary antibodies were obtained from Sigma (St. Louis, Mo), except anti-NF-L that was obtained from Serotec (Oxford, U.K.).

After washing three times in PBS for 10 min each, retinas were incubated for 1 h with goat anti-mouse IgG/Texas Red (10 μg/ml, for monoclonal primary antibody) or goat anti-rabbit IgG/Bodipy FL (10 μg/ml, for polyclonal primary antibody; Molecular Probes, Portland, OR). In human retinas, nuclei of cells were stained with 4,6-diaminodiphenyl-2-phenylindole (DAPI; 1 μg/ml; Sigma St. Louis, Mo), incubated together with fluorescent secondary antibody. Finally, retinas were washed three times in PBS and mounted in PBS-glycerol (1:1). Some human and porcine retinas were incubated with a cocktail of the four antibodies (each one used at the same concentration as for separate incubations) recognizing the three NF subunits, and detected with a mixture of goat anti-mouse and goat anti-rabbit IgG/Texas red antibodies. Immunocytochemical control experiments consisted of omission of the primary antibody, omission of the second antibody, and the use of a corresponding non-immune serum. All preparations were observed under an Axioskop 2 epifluorescent microscope (Zeiss, Jena, Germany) and photographs were taken by using a Coolsnap digital camera (RS Photometrics, Tucson, USA). Images from pig and human in vivo retinas were prepared using Adobe PhotoShop 5.0 (Adobe Systems, San Jose, CA).

Incidence of co-localization of fluorogold pre-labelled RGCs with each NF subunit

To quantify the number of RGCs that contained each NF subunit, we chose the dorsal region of the mid-peripheral retina (RGCs previously labeled with fluorogold) with an area of about 190 mm², that contained 152,000±15,500 RGCs [34]. This region corresponded to the central ring of the dorsal retina, located between the peripapillar and the peripheral regions. We selected this region of the retina because first, the three RGC size groups were abundant (composed of, 37%±1.2% large RGCs, 53%±0.9% medium RGCs and 10%±0.6% small RGCs, out of 152,000 RGCs analyzed), and second, axon bundles were not very densely packed in this region, allowing us to perform morphometric analyzes. We analyzed 6 fields of 650x480 μm² within the mid-peripheral retina from each of the 5 pig retinas incubated with either NF-H, NF-M or NF-L. This represented 1% of the mid-peripheral retinal area, and we analyzed a total of 1,116±162.9 RGCs, which represented 0.73% of the total number of RGCs in this area.

In order to determine whether a specific combination of NF subunits was expressed in any particular RGC type, in the present work we categorized the RGCs according to previous studies of pig RGCs performed in our laboratory, where three distinct size classes were correlated with differential survival (44). Thus we took into consideration the RGC soma size to analyze co-localization of fluorogold pre-labelled RGCs with each NF subunit: small (8 to 14 μm in diameter); medium (15 to 20 μm in diameter); and large RGCs (more than 21 μm in diameter).

Measurements were performed directly on the digital images by image analysis on computer (Scion Image; Scion, Frederick, MD). Statistical analysis was performed by computer (SPSS v10.0; SPSS Sciences, Chicago, IL).

Results

The distribution of NF subunits in human and pig retinas in vivo was analyzed using both retinal sections and flatmounts. The whole population of pig RGCs was retrogradely labeled with fluorogold. Moreover, the RGCs in human retinas were characterized by their location in the ganglion cell layer (GCL) and the nuclear DAPI staining. All NF subunits were localized within the inner retina, mainly in axon bundles and RGC somas and to a lesser extent in some fibers of the inner plexiform layer (IPL), both in pig and human retinas (Figure 1). In the outer plexiform layer (OPL) of the porcine retina, we observed immunostaining in fibers probably corresponding to horizontal cells. This staining was observed with antibodies against NF-H (Figure 1A) and NF-M, but not with NF-L. In human retina we did not observe any OPL labeling with any of the four antibodies used (Figure 1C). In pig (and in human) most RGCs were labeled with the three NF subunits, and the entire population of RGCs was labeled by at least one of the antibodies directed against one of the subunits (see Table 1 for data used in quantitative analysis performed in pig whose RGCs could be retrogradely labeled). To examine NF immunoreactivity in RGC soma in greater detail, double immunostaining was performed using both pig and human retinas, in sections and flat mounts, and with all possible combinations of antibodies. Thus, we compared NF-Hp versus NF-H (Figure 2); NF-Hp versus NF-M (Figure 3) and NF-L versus NF-H (Figure 4).

Distribution of NF-Hp in pig RGCs

NF-Hp distribution within RGCs was the only NF subunit studied in the present work that varied with retinal eccentricity: RGCs located in the central retina (Figure 2A) express NF-Hp mainly in their axons and dendrites, while RGC somas that were retrogradely labeled with fluorogold (Figure 2B) were not stained. However, we cannot discard the possibility that labeling occurred in the extreme cell body periphery, producing an unstained gap appearance in the GCL (Figure 2A and Figure 3A). Peripheral somatic immunolocalization of NF-Hp was more apparent when comparing NF-Hp distribution with other NF that was present throughout the cytoplasm, as was the case for NF-H (Figure 2C). In some areas of peripheral retina we observed a mosaic of small RGCs (major axis length: 10-13 μm) surrounding large RGCs (major axis length: greater than 30μm; Figure 2D). Large NF-Hp immunopositive RGCs (representing only 5x10^-3% of the total large RGCs located in the periphery) exhibited a polygonal shape, emitting 4 to 5 dendrites and a thick axon that could be easily followed from the peripheral retina to the optic nerve head. NF-Hp localization within these RGCs was distributed not only in axons and dendrites, but was also widespread in the cytoplasm (Figure 2D), contrasting with the distribution of the same NF subunit described above. NF-Hp immunoreactivity seemed to be present in the entire RGC population, but unambiguous counting of RGCs based on the number of NF-Hp labeled somas was difficult due to the reasons given.

Distribution of NF-Hp in human RGCs

Distribution of NF-Hp in human retina was very similar to that of the pig. In peripheral retina, staining was present in somas, axons, and dendrites, as seen in sections as well as whole mounts (Figure 2E,H). NF-Hp immunostaining in human peripheral retina was very similar to NF-H labeling, as seen in the pig (Figure 2F). Due to the limited tissue available, we could not confirm the presence of a mosaic-like pattern of stained cells as seen in the porcine retina. In central retina, NF-Hp was only observed in axons and dendrites (Figure 2G) with no visible labeling of cell bodies, as also seen in the porcine retina.

Distribution of NF-H, NF-M, and NF-L

The three NF subunits exhibited very similar distributions, being present in axons, dendrites, and throughout cell bodies of porcine and human RGCs. Below we describe the subtle differences observed in the distribution of each subunit.

Distribution of NF-H in pig RGCs

NF-H was expressed with similar intensity in all parts of the pig RGCs (axons, dendrites, and cell bodies). NF-H immunolabeling was present in 971 out of 1,116 RGCs analyzed (corresponding to 88%±2.74%). The number of RGCs labeled with NF-H, as a function of soma size, was 268 small RGCs (corresponding to 84%±3.67%), 422 medium RGCs (corresponding to 85%±2.55%), and 281 large RGCs (corresponding to 94%±1.99%). There were some RGCs that were stained for NF-H but not for NF-L (Figure 4D-F).

Distribution of NF-H in human RGCs

Distribution of NF-H in human RGCs was also similar to pig, being present in axons, dendrites, and filling the somas. The similar distribution of NF-H between both species was apparent in sections (Figure 4H) as well as flat mounts (Figure 4I). This subunit was present in most human RGC with few exceptions (Figure 1C and Figure 4H).

Distribution of NF-M in pig RGCs

Localization of the NF-M within the cell bodies and dendrites of the RGCs (Figure 3B,C,E) resembled that observed for NF-H (Figure 1A), being homogeneously distributed within the entire RGC. However, a distinct perinuclear ring of more intense NF-M labeling was visible, more clearly observed in flatmounted retinas (Figure 3E). Cytoplasmic NF-M immunolabeling of RGCs (Figure 3B,C) contrasted with the superficial labeling observed with NF-Hp (Figure 3A). The number of NF-M positive RGCs was lower (980 out of 1,422 RGCs analyzed, corresponding to 69%±4.76%), compared to the other two subunits. The number of RGCs labeled with NF-M as a function of soma size was 53 (corresponding to 36%±0.63%) small RGCs; 500 medium RGCs (corresponding to 75%±8.67%); and 427 large RGCs (corresponding to 78%±5%).

Distribution of NF-M in human RGCs

In human as in pig retinas, NF-M distribution was very similar to NF-H, being present in RGC axons, dendrites, and filling the somas (Figure 3G,H). NF-M immunostaining of RGC within flatmounted human retina (Figure 3H) was more distinct than for pig (Figure 3E), probably due to OPL staining occurring in the former slightly obscuring the signal.

Distribution of NF-L in pig RGCs

NF-L was distributed widely within RGC cell bodies (Figure 4A,B,D,E), but appeared slightly more internal than NF-H (Figure 4C). The number of NF-L-positive RGCs represented 988 out of 1,116 RGCs analyzed, with a similar percentage to that observed for NF-H (87%±0.55%). Taking into consideration the RGC sizes, NF-L was present in 204 small RGCs (64%±2.96%), 486 medium RGCs (97%±0.34%), and 295 large RGCs (99%±1.04%). There were some RGCs that were stained for NF-L but not for NF-H. (Figure 4D-F).

Distribution of NF-L in human RGCs

Distribution of NF-L in human RGCs resembled that for pig, with NF-L staining less intense toward the outer soma with respect to NF-H (Figure 4G,H). NF-L was expressed by most RGCs (Figure 4G,I).

Discussion

RGCs are especially affected during pathologies like glaucoma, in which increased intraocular pressure frequently occurs. These pressure changes lead to alterations in the RGC cytoskeleton, and appear to preferentially affect large RGCs [21]. We have preliminary data, obtained from retinas of pigs under experimental glaucoma based on sustained increase of intraocular pressure, suggesting down-regulation of NF subunit expression among some RGC somatas (unpublished data).

It was hence plausible that NF expression would exhibit specific differences between RGC subpopulations, and we examined whether NF expression varied in RGCs in relation to soma size or location in both normal human and pig retina. The latter was included since it may represent a useful animal model for human retinal pathophysiology, and permitted unambiguous identification of RGCs by retrograde labeling with fluorogold. The anatomy of the pig eye is very similar to that of the human, but relatively few studies have analyzed the differences between retinas from both species [35,36,41,42]. RGC identification in human retina was based upon their large nuclear size in relation to soma [43]. The results obtained in the present work show a high similarity between the distribution of NFs within RGCs of both species, supporting the use of pig retina in studies related to human vision.

Immunodetection of NF protein

Different fixation protocols can affect NF immunogenicity [44], so the immunostaining patterns compared in the present study were performed in retinas that were fixed under identical conditions to minimize such differences. Triton X-100 treatment reduces the levels of NF-H and NF-Hp detected in Purkinje cell dendrites and somata, while enhancing them in axonal [45], and hence the NF-Hp immunoreactivity observed in superficial regions of RGC somata in the central retina could be due in part to Triton X-100 treatment. However, the very intense labeling in the cytoplasm of the peripheral mosaic-like RGCs indicate that Triton X-100 did not affect staining. Moreover, the specificity of the antibodies has been previously monitored [46]. Examination of flatmounted retinas revealed somatic labeling and overall neurite morphology, and permitted the visualization of mosaic arrangements of NF-Hp immunolabeled RGCs in the retinal periphery. In some parts of the mid-peripheral retina where low density of axon bundles allow observation of the majority of the RGC somas, we were able to measure the degree of co-localization of NF-immunopositive cells with fluorogold pre-labeled RGCs. Moreover we studied the percentage of cells immunolabeled with each NF subunit in relation to RGC soma size. The high incidence of co-localization of fluorogold pre-labeled RGCs obtained in the present analysis together with results obtained in the retinas incubated with the cocktail of all three NF subunits, demonstrate that the entire RGC population expresses at least one NF subunit.

Cells other than RGCs stained with NF antibodies

In the retina, the only neuronal population that is consistently labeled with antibodies directed against NF is RGC, which is, together with some horizontal cells, the only axon-bearing retinal neuron. We can speculate that only axon-bearing neurons in the retina contain NF, in addition possibly to type A horizontal cells. Although the presence of axon-bearing horizontal cells has been reported in the human OPL [47], we did not find labeling for any NF subunit in this layer. This difference in OPL staining was the greatest difference between pig and human retina NF distribution. Detailed studies on the horizontal cells of mammalian retina have been done and species and cell type differences have been found [5,10,11,48].

Neurofilament organization in RGCs

Silver staining methods to visualize NF within RGCs in adult mammals like pigs showed staining to be mainly present in large alpha RGCs, attributable to the large amounts of NFs contained in them [49]. Moreover, previous studies on NF distribution within the retina using specific antibodies also showed specific labeling within large RGCs [6]. On the other hand, while antibodies against NFs have been commonly used as specific RGC markers in cultures [50], this is not the case for in vivo studies where it is commonly assumed that NFs are located mainly in RGC axons [3-7,51]. In the present work we have demonstrated that the entire RGC population expresses at least one NF subunit, by incubating retinas in which RGCs have been retrogradely labeled with fluorogold with a cocktail of NFs. It is possible that other authors employing antibodies directed against only one of the subunits failed to label the entire population of RGCs. In this regard, some RGCs showed labeling to NF-Hp but not NF-H (Figure 2E,F); to NF-L but not NF-H (Figure 4D,F); to NF-H but not NF-L (Figure 4G,H); to neither NF-H (Figure 4C,F) nor NF-L (Figure 4A,C) concomitantly; or that were not labeled with NF-M (Figure 3E).

It has been shown that there are great differences in NF polymerization between species. Thus studies on NF assembly have shown that rodent NF-L cannot form homopolymers while human NF-L can. A single amino acid change can drastically alter the dimerization potential [2]. In vitro studies of purified porcine NF-L have probed the ability of individual NF proteins to form homopolymers [52], supporting the similar NF structure between pig and human NFs. However, the biological significance of the differential expression of NFs between species or their selective subunit composition remains unknown.

Distribution of phosphorylated NF

The phosphorylation state of NF proteins is thought to be important in the formation of NF bundles. Phosphorylated NF-H is generally regarded as the component that cross-links and/or spaces NFs to form bundles and thus it has been concluded that NF-H should be predominantly phosphorylated where NFs are abundant [27,44,53]. This could be the case for the large and small RGCs forming a mosaic in the peripheral retina as seen by NF-Hp immunoreactivity. The NF-H antibody, which recognizes mainly the non-phosphorylated NF high molecular weight subunit (200 kDa), is distributed in the same fashion within RGC somata irrespective of RGC location within the retina. On the contrary, anti-NF-Hp, which reacts predominantly with the phosphorylated form, binds RGC axons and soma edges in central retina, whereas its distribution is widespread in large and small RGCs located in the peripheral retina. We do not have a physiological explanation for this differential distribution of phosphorylated NFs between RGCs from central versus peripheral retina, but it may be implicated in regional-selective susceptibility to death of certain types of RGCs during pathologies like glaucoma, that have long been claimed to predominate in the retinal periphery [54]. With respect to NF-M (clone NN18), it has been previously reported to recognize both phosphorylated and non-phosphorylated NF-M [39], though it has never been formally shown to be phosphate-dependent in its binding properties. Moreover, the clone NN18 has been previously reported to recognize not all NF-M structures, but only some NF-M fibres [55].

Concluding remarks

The present study was unable to demonstrate subpopulation-specific differences in NF expression within human and pig RGC that might underlie regional variations in RGC degeneration. Alternative reasons for such preferential loss should therefore be investigated, such as selective glutamate receptor expression or mechanical differences. With regard to the former possibility, previous studies have failed to reveal glutamate receptor subunit variation in primate retinas [56], although developmental variations have been reported in rat retinas [57].

This study also reinforces the value of the pig as an animal model for examining human retinal pathophysiology, and it should prove useful in the present context of exploring cellular and molecular mechanisms underlying RGC degeneration.

Acknowledgements

Grant support was provided by European Community PRO AGE RET (QLK6-2001-00385) the University of the Basque Country (E-14887/2002 and 15350/2003); and MCYT (BFI 2003-07177) to EV. MG holds a European Community postdoctoral contract. JRE is supported by the University of the Basque Country.

Presented in part at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Laurderdale, Florida, May 2003.

References

1. Lee MK, Cleveland DW. Neuronal intermediate filaments. Annu Rev Neurosci 1996; 19:187-217.

2. Carter J, Gragerov A, Konvicka K, Elder G, Weinstein H, Lazzarini RA. Neurofilament (NF) assembly; divergent characteristics of human and rodent NF-L subunits. J Biol Chem 1998; 273:5101-8.

3. Vickers JC, Costa M. The neurofilament triplet is present in distinct subpopulations of neurons in the central nervous system of the guinea-pig. Neuroscience 1992; 49:73-100.

4. Drager UC, Hofbauer A. Antibodies to heavy neurofilament subunit detect a subpopulation of damaged ganglion cells in retina. Nature 1984; 309:624-6.

5. Shaw G, Weber K. The intermediate filament complement of the retina: a comparison between different mammalian species. Eur J Cell Biol 1984; 33:95-104.

6. Straznicky C, Vickers JC, Gabriel R, Costa M. A neurofilament protein antibody selectively labels a large ganglion cell type in the human retina. Brain Res 1992; 582:123-8.

7. Wang L, Dong J, Cull G, Fortune B, Cioffi GA. Varicosities of intraretinal ganglion cell axons in human and nonhuman primates. Invest Ophthalmol Vis Sci 2003; 44:2-9.

8. Drager UC, Edwards DL, Barnstable CJ. Antibodies against filamentous components in discrete cell types of the mouse retina. J Neurosci 1984; 4:2025-42.

9. Kivela T, Tarkkanen A, Virtanen I. Intermediate filaments in the human retina and retinoblastoma. An immunohistochemical study of vimentin, glial fibrillary acidic protein, and neurofilaments. Invest Ophthalmol Vis Sci 1986; 27:1075-84.

10. Lohrke S, Brandstatter JH, Boycott BB, Peichl L. Expression of neurofilament proteins by horizontal cells in the rabbit retina varies with retinal location. J Neurocytol 1995; 24:283-300.

11. Kong WC, Cho EY. Antibodies against neurofilament subunits label retinal ganglion cells but not displaced amacrine cells of hamsters. Life Sci 1999; 64:1773-8.

12. Volgyi B, Bloomfield SA. Axonal neurofilament-H immunolabeling in the rabbit retina. J Comp Neurol 2002; 453:269-79.

13. Gabriel R, Straznicky C. Immunocytochemical localization of parvalbumin- and neurofilament triplet protein immunoreactivity in the cat retina: colocalization in a subpopulation of AII amacrine cells. Brain Res 1992; 595:133-6.

14. Perrone Capano C, Pernas-Alonso R, di Porzio U. Neurofilament homeostasis and motoneurone degeneration. Bioessays 2001; 23:24-33.

15. Julien JP, Mushynski WE. Neurofilaments in health and disease. Prog Nucleic Acid Res Mol Biol 1998; 61:1-23.

16. Dieterich DC, Trivedi N, Engelmann R, Gundelfinger ED, Gordon-Weeks PR, Kreutz MR. Partial regeneration and long-term survival of rat retinal ganglion cells after optic nerve crush is accompanied by altered expression, phosphorylation and distribution of cytoskeletal proteins. Eur J Neurosci 2002; 15:1433-43.

17. Asahara H, Taniwaki T, Ohyagi Y, Yamada T, Kira J. Glutamate enhances phosphorylation of neurofilaments in cerebellar granule cell culture. J Neurol Sci 1999; 171:84-7.

18. Ackerley S, Grierson AJ, Brownlees J, Thornhill P, Anderton BH, Leigh PN, Shaw CE, Miller CC. Glutamate slows axonal transport of neurofilaments in transfected neurons. J Cell Biol 2000; 150:165-75.

19. Salvador-Silva M, Ricard CS, Agapova OA, Yang P, Hernandez MR. Expression of small heat shock proteins and intermediate filaments in the human optic nerve head astrocytes exposed to elevated hydrostatic pressure in vitro. J Neurosci Res 2001; 66:59-73.

20. Wax MB, Tezel G, Kobayashi S, Hernandez MR. Responses of different cell lines from ocular tissues to elevated hydrostatic pressure. Br J Ophthalmol 2000; 84:423-8.

21. Vickers JC, Schumer RA, Podos SM, Wang RF, Riederer BM, Morrison JH. Differential vulnerability of neurochemically identified subpopulations of retinal neurons in a monkey model of glaucoma. Brain Res 1995; 680:23-35.

22. Hoffman PN, Lasek RJ. The slow component of axonal transport. Identification of major structural polypeptides of the axon and their generality among mammalian neurons. J Cell Biol 1975; 66:351-66.

23. Black MM, Lasek RJ. Slow components of axonal transport: two cytoskeletal networks. J Cell Biol 1980; 86:616-23.

24. Jung C, Yabe TJ, Shea TB. C-terminal phosphorylation of the high molecular weight neurofilament subunit correlates with decreased neurofilament axonal transport velocity. Brain Res 2000; 856:12-9.

25. Friede RL, Samorajski T. Axon caliber related to neurofilaments and microtubules in sciatic nerve fibres of rats and mice. Anat Rec 1970; 167:379-88.

26. Weiss PA, Mayr R. Organelles in neuroplasmic ("axonal") flow: neurofilaments. Proc Natl Acad Sci U S A 1971; 68:846-50.

27. Nixon RA, Paskevich PA, Sihag RK, Thayer CY. Phosphorylation on carboxyl terminus domains of neurofilament proteins in retinal ganglion cell neurons in vivo: influences on regional neurofilament accumulation, interneurofilament spacing, and axon caliber. J Cell Biol 1994; 126:1031-46.

28. Parhad IM, Clark AW, Griffin JW. Effect of changes in neurofilament content on caliber of small axons: the beta,beta'-iminodipropionitrile model. J Neurosci 1987; 7:2256-63.

29. Monaco S, Autilio-Gambetti L, Lasek RJ, Katz MJ, Gambetti P. Experimental increase of neurofilament transport rate: decreases in neurofilament number and axon diameter. J Neuropath Exp Neurol 1989; 48:23-32.

30. Watson DF, Hoffman PN, Fittro KP, Griffin JW. Neurofilament and tubulin transport slows along the course of mature motor axons. Brain Res 1989; 477:225-32.

31. Monteiro MJ, Hoffman PN, Gearhart JD, Cleveland DW. Expression of NF-L in both neuronal and nonneuronal cells of transgenic mice: increased neurofilament density in axons without affecting caliber. J Cell Biol 1990; 111:1543-57.

32. Minzenberg M, Berkelaar M, Bray G, McKerracher L. Changes in retinal ganglion cell axons after optic nerve crush: neurofilament expression is not the sole determinant of calibre. Biochem Cell Biol 1995; 73:599-604.

33. Glovinsky Y, Quigley HA, Dunkelberge GR. Retinal ganglion cell loss is size dependent in experimental glaucoma. Invest Ophthalmol Vis Sci 1991; 32:484-91.

34. Ruiz-Ederra J, Garcia M, Urcola JA, Araiz J, Duran JA, Vecino E. Topography of retinal ganglion cells from the pig retina. Ophthalmic Res 2002; 34:S1-263.

35. Prince JH, Diesem CD, Eglitis I, Ruskell GL. The pig. In: Prince JH. Anatomy and histology of the eye and orbit in domestic animals. Springfield (IL): Thomas; 1960. p. 210-30.

36. Hendrickson A, Hicks D. Distribution and density of medium-and short-wavelength selective cones in the domestic pig retina. Exp Eye Res 2002; 74:435-44.

37. Li ZY, Wong F, Chang JH, Possin DE, Hao Y, Petters RM, Milam AH. Rhodopsin transgenic pigs as a model for human retinitis pigmentosa. Invest Ophthalmol Vis Sci 1998; 39:808-19.

38. World Medical Association declaration of Helsinki. Recommendations guiding physicians in biomedical research involving human subjects. JAMA 1997; 277:925-6.

39. Veerana, Amin ND, Ahn NG, Jaffe H, Winters CA, Grant P, Pant HC. Mitogen-activated protein kinases (Erk1,2) phosphorylate Lys-Ser-Pro (KSP) repeats in neurofilament proteins NF-H and NF-M. J Neurosci 1998; 18:4008-21.

40. Guidato S, Bajaj NP, Miller CC. Cellular phosphorylation of neurofilament heavy-chain by cyclin-dependent kinase-5 masks the epitope for monoclonal antibody N52. Neurosci Lett 1996; 217:157-60.

41. Beauchemin ML. The fine structure of the pig's retina. Albrecht Von Graefes Arch Klin Exp Ophthalmol 1974; 190:27-45.

42. De Schaepdrijver L, Lauwers H, Simoens P, de Geest JP. Development of the retina in the porcine fetus. A light microscopic study. Anat Histol Embryol 1990; 19:222-35.

43. Curcio CA, Allen KA. Topography of ganglion cells in human retina. J Comp Neurol 1990; 300:5-25.

44. Trojanowski JQ, Obrocka MA, Lee VM. Distribution of neurofilament subunits in neurons and neuronal processes: immunohistochemical studies of bovine cerebellum with subunit-specific monoclonal antibodies. J Histochem Cytochem 1985; 33:557-63.

45. Gotow T, Tanaka J. Phosphorylation of neurofilament H subunit is related to arrangement of neurofilaments. J Neurosci Res 1994; 37:691-713.

46. Garcia M, Forster V, Hicks D, Vecino E. Effects of muller glia on cell survival and neuritogenesis in adult porcine retina in vitro. Invest Ophthalmol Vis Sci 2002; 43:3735-43.

47. Kolb H, Fernandez E, Schouten J, Ahnelt P, Linberg KA, Fisher SK. Are there three types of horizontal cell in the human retina? J Comp Neurol 1994; 343:370-86.

48. Peichl L, Gonzalez-Soriano J. Unexpected presence of neurofilaments in axon-bearing horizontal cells of the mammalian retina. J Neurosci 1993; 13:4091-100.

49. Peichl L, Ott H, Boycott BB. Alpha ganglion cells in mammalian retinae. Proc R Soc Lond B Biol Sci 1987; 231:169-97.

50. Torelli S, Sogos V, Marzilli MA, D'Atri M, Gremo F. Developmental expression of intermediate filament proteins in the chick embryo retina: in vivo and in vitro comparison. Exp Biol 1989; 48:187-96.

51. Villegas-Perez MP, Lawrence JM, Vidal-Sanz M, Lavail MM, Lund RD. Ganglion cell loss in RCS rat retina: a result of compression of axons by contracting intraretinal vessels linked to the pigment epithelium. J Comp Neurol 1998; 392:58-77.

52. Geisler N, Weber K. Self-assembly in Vitro of the 68,000 molecular weight component of the mammalian neurofilament triplet proteins into intermediate-sized filaments. J Mol Biol 1981; 151:565-71.

53. Matus A. Neurofilament protein phosphorylation--where, when and why. Trends Neurosci 1988; 11:291-2.

54. Laquis S, Chaudhary P, Sharma SC. The patterns of retinal ganglion cell death in hypertensive eyes. Brain Res 1998; 784:100-4.

55. Weber K, Shaw G, Osborn M, Debus E, Geisler N. Neurofilaments, a subclass of intermediate filaments: structure and expression. Cold Spring Harb Symp Quant Biol 1983; 48:717-29.

56. Hof PR, Lee PY, Yeung G, Wang RF, Podos SM, Morrison JH. Glutamate receptor subunit GluR2 and NMDAR1 immunoreactivity in the retina of macaque monkeys with experimental glaucoma does not identify vulnerable neurons. Exp Neurol 1998; 153:234-41.

57. Tehrani A, Wheeler-Schilling TH, Guenther E. Coexpression patterns of mGLuR mRNAs in rat retinal ganglion cells: a single-cell RT-PCR study. Invest Ophthalmol Vis Sci 2000; 41:314-9.

Typographical corrections