Impaired Neurite Outgrowth in the Retina of a Murine Model of Sandhoff Disease

Abstract

purpose. To investigate the effects of lysosomal storage on the morphologic appearance and the neurite outgrowth capability of the retina in a mouse model of G_M2 gangliosidosis (Sandhoff disease).

methods. Histopathologic appearances of retinas in Sandhoff (SD) mice at 3 and 4 months of age were examined by light and electron microscopy. Retinas of SD mice and wild-type (WT) mice at 1, 2, and 4 months of age were cultured in collagen gel in the presence or absence of brain-derived neurotrophic factor (BDNF), and neurite outgrowth was examined.

results. Morphologic studies revealed accumulation of G_M2 ganglioside in the retinal ganglion cells of SD mice in a time-dependent manner. The number of neurites from the retinal explants after 7 and 10 days in culture were significantly lower in 2- and 4-month-old SD mice than in the age-matched WT mice. The application of BDNF significantly improved neurite outgrowth from the retina in both SD and WT mice at 2 months of age. At 4 months of age, BDNF was much less effective at stimulating neurite outgrowth in the retina of SD mice than in retina of WT mice.

conclusions. These results indicate that lysosomal storage of G_M2 ganglioside impairs the capability of neurite outgrowth in retinal ganglion cells in culture and that BDNF is effective at diminishing this impairment during the early stage of the disease.

The G_M2 gangliosidoses are a group of inherited neurodegenerative disorders, characterized by a deficiency in the degradation of G_M2 ganglioside and related substrates. ¹ They are caused by mutations in any of three genes; the HEXA gene, causing Tay-Sachs disease; the HEXB gene, causing Sandhoff disease; and the GM2A gene, causing the G_M2 activator deficiency. ^{2 3} The HEXA and HEXB genes encode the α and β subunits of lysosomal β-hexosaminidase (Hex), which dimerize to produce three isozymes: HexA (αβ), HexB (ββ), and HexS (αα). The GM2A gene encodes the G_M2-activator protein. The HexA isozyme and G_M2-activator protein are necessary for the degradation of G_M2 ganglioside in lysosomes. ² Accordingly, the absence of either HexA (Tay-Sachs disease, Sandhoff disease) or G_M2-activator protein (G_M2 activator deficiency) leads to lysosomal storage of G_M2 ganglioside and other glycolipids throughout the nervous system ^{1 2 3} . These diseases, in both infantile and late-onset forms, have similar clinical courses and neuropathology. The infantile forms of G_M2 gangliosidoses are characterized by progressive muscular weakness, mental retardation, blindness, and death in early childhood. ¹ Ophthalmoscopic examination reveals a cherry-red spot surrounded by a white halo in the macula of the retina. ^{4 5} The perimacular halo is due to the thickening of the retinal ganglion cell layer caused by neuronal storage, and the cherry-red spot at the macula is the result of the relative absence of ganglion cells, which causes the underlying choroid to be clearly seen. ⁴ The histologic changes in the retina of G_M2 gangliosidoses, such as loss of ganglion cells, ballooning of the remaining neurons, and atrophy of the optic nerve fibers, are similar to those in the brain. ⁶  

We and others have created murine models of Tay-Sachs disease, Sandhoff disease, and G_M2-activator deficiency through the targeted disruptions of mouse Hexa, ^{7 8} Hexb, ^{9 10} and Gm2A genes, ¹¹ respectively. Unlike the human disorders, the three mouse models show markedly different phenotypes. The Tay-Sachs model displays late-onset neurologic manifestations (beginning at 11–12 months of age) with a mild clinical course and had a normal lifespan, ⁸ whereas the Sandhoff model shows progressive motor dysfunction and death before 5 months of age. ^{9 10} The model of G_M2-activator deficiency displays a normal lifespan, but exhibits impaired motor coordination. ¹¹ Among these models, the Sandhoff model best mimics clinical, pathologic, and biochemical features of human G_M2 gangliosidoses. ¹² This model has been used for developing effective therapies for the diseases, ^{13 14 15} and for elucidating the pathogenesis of neurodegeneration due to G_M2 ganglioside accumulation. ^{16 17 18 19} However, morphologic or functional abnormalities of retina in these mice have not yet been fully described. In the present study, we investigated microscopic changes of retinal ganglion cells in Sandhoff mice at 3 and 4 months of age. We also used a method for explant culture of retinal tissue, ^{20 21} to examine the effects of glycosphingolipid accumulation on the neurite outgrowth capability of retina. 

Methods

Animals

A mouse model of Sandhoff disease (Hexb ^{− / −}) was developed by disruption of the Hexb gene, ⁹ and the Hexb genotype was determined by polymerase chain reaction (PCR), as described previously. ²² Hexb ^−/− and Hexb ^+/+ mice at 3 and 4 months of age were used for morphologic study, and the mice at 1, 2, and 4 months of age were used for in vitro study. All experiments were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the Guide of the Animal Ethics Committee, Yokohama City University School of Medicine. 

Optic fundi in these mice were observed and photographed under a stereoscopic zoom microscope (SMZ-1; Nikon, Tokyo, Japan) equipped with a 35-mm camera (Nikon). 

Light Microscopy

The mice examined at 3 months and 4 months of age were killed by chloroform exposure. Retinas were dissected from the bilateral eyeballs of the mice, embedded in optimal cutting temperature (OCT) compound (Tissue-Tek; Sakura Finetechnical Co. Ltd, Tokyo, Japan), and processed for frozen sections. The sections were stained with periodic acid-Schiff (PAS). The PAS reaction is routinely used for the detection of carbohydrate-rich macromolecules, ²³ and the storage materials in the brain of Hexb ^−/− mice were intensely stained with PAS on frozen sections. ⁹  

Electron Microscopy

The mice were anesthetized by intraperitoneal injection of pentobarbital sodium (50 mg/kg) and then briefly perfused through the left cardiac ventricle with 100 mM phosphate-buffered saline (PBS; pH 7.4) followed by 2% paraformaldehyde and 2% glutaraldehyde. Retinas were dissected from the mice, and osmicated, dehydrated through graded ethanol, and embedded in epoxy resin. One-micrometer-thick sections were stained with toluidine blue and examined with a light microscope. Areas appropriate for further ultrathin sectioning were selected. Ultrathin sections were double stained with uranyl acetate and lead citrate and examined with an electron microscope (Hitachi, Tokyo, Japan). 

Neurite Extension from Retinal Explants in Culture

The explant culture of adult mouse retina was performed as previously described. ^{20 21} Briefly, retinas were dissected from the bilateral eyeballs of each mouse at 1, 2, or 4 months of age and isolated from the choroid, the pigment epithelium, and adherent vitreous. The retina was then cut into 16 small pieces (approximately 500 μm² each) with a sharp razor blade and embedded in collagen gel ²⁴ on poly-l-lysine (10 μg/mL)-precoated dishes. These retinal pieces were cultured at 37°C in a 95% air-5% CO₂ atmosphere in serum-free Minimum Essential Medium (MEM, Invitrogen) supplemented with 2.7 mg/mL glucose, 5 μg/mL insulin, 16.1 μg/mL putrescine, 792 μg/mL bovine serum albumin, 5.2 ng/mL sodium selenite, 3.7 mg/mL sodium bicarbonate, and 3.6 mg/mL HEPES. Recombinant human brain derived-neurotrophic factor (BDNF; Pepro Tech, NJ) was added to half of the retinal explants from 2- and 4-month-old mice to be a final concentration at 100 ng/mL in the culture medium. The number of outgrowing neurites longer than 50 μm that were elongated outside the retinal explants was counted under a phase-contrast light microscope after 3, 7, and 10 days in culture. The explant culture of retina and subsequent measurement of the regenerating neurites were performed in each mouse. All measurements are given as mean ± SEM calculated from 64 explants from two animals, and the Mann–Whitney nonparametric test was performed for statistical analysis. 

Results

Ophthalmic Findings

Under a substantial microscope, optic nerve fiber bundles spreading radially from the optic papilla were clearly seen in the Hexb ^+/+ mice (Fig. 1A) . In contrast, optic fundi were spotted with white dots, and nerve fibers were hardly visible in the Hexb ^−/− mice (Fig. 1B) . Unlike the eyes of patients with G_M2 gangliosidoses, cherry-red spots were not observed in the eyes of Hexb ^−/− mice. 

Light Microscopic Findings

The Hexb ^−/− mice at 3 and 4 months of age exhibited marked storage of PAS-positive materials in the ganglion cell layer on frozen sections (Fig. 2B) . The inner plexiform layer, inner nuclear layer, and outer plexiform layer of these mice were also intensely stained with PAS. In sharp contrast, the retina of Hexb ^+/+ mice at any age were faintly stained with PAS (Fig. 2A) . 

Electron Microscopic Findings

Ultrastructurally, we observed numerous intracytoplasmic inclusion bodies in all retinal ganglion cells in the Hexb ^−/− mice (Fig. 3B) , but not in the Hexb ^+/+ mice (Fig. 3A)at 3 months of age. Apparent cell death was hardly detected in the ganglion cell layer in either group of mice at this stage. In the Hexb ^−/− mice at 4 months of age, however, some ganglion cells showed apoptotic signs (Fig. 3C , arrow); the cytoplasm was occupied with the inclusion bodies, and the nucleus was darkened, shrunk, and partly disrupted. For quantitative analysis, we counted the number of apoptotic and nonapoptotic RGCs under an electron microscope, and calculated the ratio of apoptotic RGCs in each group of mice. The average ratios (mean ± SEM) were 0.2% ± 0.2% in 3-month-old Hexb ^+/+ mice, 1.2% ± 0.5% in 3-month-old Hexb ^−/− mice and 10.9% ± 1.8% in 4-month-old Hexb ^−/− mice, respectively (Fig. 4) . The invasion and phagocytotic action of macrophages/microglia were not observed in any group of mice. 

Impaired Neurite Outgrowth from Retinal Explants in Culture

Neurites growing out from each retinal explant in collagen gel could be easily identified under a phase-contrast microscope by moving the focus point (Fig. 5) . The neurites could be distinguished from the extending processes of fibroblasts and other non-neuronal cells by the unique structure of the spread growth cones at the tips of neurites. ^{21 25} The number of non-neuronal processes were extremely small in the explant culture in the absence of serum, and most processes of the non-neuronal cells were much shorter than the neurites. These findings were confirmed by immunohistochemical staining. The neurites were immunoreactive to the anti-neurofilament antibody, ²¹ whereas the fibroblasts and glial cells were immunoreactive to the anti-fibronectin and the anti-GFAP antibodies, respectively (Horie and Takano, unpublished data, 2002). There was, however, a possibility that the counted number of neurites in the present experiments might have erroneously include the processes of non-neuronal cells, but their numbers were negligibly small. 

Before the comparison between the Hexb ^−/− and Hexb ^+/+ mice, we observed that the number of neurites at any days in culture declined as the age of mice advanced (from 1 to 2 months and from 2 to 4 months; Table 1 ). At 1 month of age, the average number of neurites after 3 and 7 days in culture was significantly lower in the Hexb ^−/− mice than in the Hexb ^+/+ mice (5.6 ± 1.2 vs. 9.6 ± 0.8 after 3 days and 44.5 ± 2.9 vs. 76.8 ± 5.8 after 7 days). After 10 days in culture, however, the number in the Hexb ^−/− mice reached 103 ± 7.5, almost equal to that in the Hexb ^+/+ mice, 116 ± 14. The neurite outgrowth capability of retinal explants in the 2- and 4-month-old Hexb ^−/− mice was more severely affected than that in the 1-month-old mice (Fig. 5) . The numbers of neurites after 10 days in culture was significantly lower in these mice than in the age-matched Hexb ^+/+ mice (20.7 ± 2.6 vs. 69.2 ± 3.7 at 2 months and 11.7 ± 1.7 vs. 55.1 at 4 months, respectively). 

Treatment with BDNF markedly enhanced neurite outgrowth from both Hexb ^−/− and Hexb ^+/+ retinas at 2 months of age (Fig. 6) . The number of neurites increased from 14.6 ± 1.8 to 40.9 ± 5.2 and from 47.2 ± 2.5 to 82.0 ± 3.4 after 7 days and from 20.7 ± 2.6 to 49.7 ± 5.8 and from 69.2 ± 3.7 to 114.0 ± 4.1 after 10 days in culture, respectively. Such enhancement by BDNF was also observed in the 4-month-old Hexb ^+/+ retinas. The number increased from 26.7 ± 2.5 to 64.6 ± 5.9 after 7 days and from 55.1 ± 4.7 to 129.0 ± 12.2 after 10 days in culture. In contrast, BDNF exhibited much less activity for the neurite regeneration in the 4-month-old Hexb ^−/− retinas. The number changed from 4.4 ± 0.8 to 11.6 ± 1.4 after 7 days and from 11.7 ± 1.7 to 22.0 ± 1.8 after 10 days in culture. 

Discussion

Extensive Glycolipid Storage and Neural Degeneration in Retina of the Hexb ^−/− Mouse

In the present study, we investigated whether a mouse model of Sandhoff disease exhibits degenerative changes in the neural retina. The ophthalmic signs of lysosomal storage diseases include the presence of a macular cherry-red spot, corneal clouding, retinal pigment epithelial degeneration, optic atrophy, vascular tortuosities, and choroidal white dots. ²⁶ We observed opaque fundi spotted with white dots in the Hexb ^−/− mice (Fig. 1B) , which can be a sign of lysosomal storage. The absence of a cherry-red spot in the Hexb ^−/− mice seemed to be due to anatomic differences in the retina between human and mouse. Macular structure with a fovea centralis is observed in human retinas, whereas rodents and many other mammals possess an area centralis. ²⁷  

Consistent with the fundus appearances described, light microscopy with PAS stain showed extensive storage throughout the neural retina (especially in the retinal ganglion cell layer) of the Hexb ^−/− mice. These findings were further supported by the results of electron microscopy: the intracytoplasmic inclusion bodies were observed in the ganglion cell layer, inner plexiform layer, inner nuclear layer, and outer plexiform layer in the Hexb ^−/− mice. The involvement by inclusions, particularly with membranous and lamellar shapes, became more conspicuous as the age of mice increased. It is obvious that continuous accumulation of G_M2 ganglioside and related glycolipids from the beginning of life causes neurodegeneration in both human Sandhoff disease and its mouse model, but the precise mechanisms remain unclear. Wada et al. ¹⁷ observed apoptotic cell death of neurons and microglial invasion in caudal regions of the central nervous system (CNS; i.e., thalamus, brain stem, and spinal cord) in symptomatic Hexb ^−/− mice (3 and 4 months of age). Because no apoptotic neurons were detected in presymptomatic mice (1 and 2 months of age) in that study, it seems likely that it takes approximately 3 months for lysosomal storage to cause neuronal cell death and microglial activation in those regions of the mouse. It is of interest to note that apoptotic cell death and reactive gliosis was not detected in other CNS regions (i.e., cerebral cortex and cerebellum) in 4-month-old Hexb ^−/− mice, even though the storage was abundant in those regions. ¹⁷ In the present study, we observed no apoptotic cells in the retina of 3-month-old Hexb ^−/− mice. At 4 months of age, a part of RGCs appeared to undergo apoptotic changes by electron microscopy, but there was no indication of massive cell death in neural retina. Taking these findings into consideration, ganglioside storage in neural retina, as well as in cerebral cortex and cerebellum, throughout the lifespan of Hexb ^−/− mice (4–5 months) is unlikely to reach a degree that causes significant cell loss. In contrast to the mouse model, ubiquitous neuronal cell death and reactive gliosis was detected throughout the CNS of patients with Tay-Sachs and Sandhoff diseases. ^{16 17 28} The human samples, which were from autopsy, represented the end stage of the diseases (∼4 years of age or longer), when the process of neurodegeneration due to ganglioside storage may be advanced in all regions of nervous tissues, including the neural retina. 

Impaired Neurite Outgrowth from Retinal Explants of Hexb ^−/− Mice In Vitro

We previously demonstrated impaired survival but normal neurite outgrowth in dorsal root ganglion (DRG) neurons cultured from presymptomatic Hexb ^−/− mice (4–5 weeks of age). ²² In contrast, Pelled et al. ²⁹ reported normal viability but reduced rates of axonal and dendritic growth in neurons cultured from the hippocampus of embryonic Hexb ^−/− mice, suggesting that even small levels of ganglioside accumulation can impair the neurite outgrowth capability of hippocampal neurons. In those studies, however, neurite outgrowth was examined at only early (embryonic and presymptomatic) stages of the disease. Since the Hexb ^−/− mice showed progressive neurologic manifestations beginning approximately 3 months after birth, the neurite outgrowth capability of neurons in vitro may be more severely affected in symptomatic mice (older than 3 months after birth) than in presymptomatic mice. It had been considered to be difficult for adult mammalian CNS neurons, including RGCs, to survive and extend neurites in vitro. However, we have established a three-dimensional collagen gel culture system of mammalian retinas and have observed neurites growing from retinal pieces of adult mice. ²⁰ Considering that the neurites outside the retinal explants were mostly from RGCs, ²¹ this explant culture system appears to be useful for the evaluation of the neurite outgrowth capability of RGCs in Hexb ^−/− mice at presymptomatic (1 and 2 months of age) and symptomatic stages (4 months of age). 

In the present study, a reduced number of neurites was observed in Hexb ^−/− mice after 3 and 7 days in culture and even at an early presymptomatic stage (1 month after birth), which was consistent with the findings of Pelled et al., ²⁹ who used embryonic hippocampal neurons. After 10 days in culture, however, the number of neurites in the 1-month-old Hexb ^−/− mice reached a level close to that in the Hexb ^+/+ mice. Therefore, it is plausible that lysosomal storage by 1 month after birth could retard the rate of neurite outgrowth from RGCs in vitro, but the degree of storage may not be sufficient to have a detrimental effect on the neurite outgrowth capability of RGCs. In contrast to the 1-month-old mice, the number of neurites after 10 days in culture was significantly reduced in the 2- and 4-month-old Hexb ^−/− mice. These results clearly show that the neurite outgrowth capability of RGCs in vitro deteriorates as the neuronal glycolipid storage in vivo escalates. It remains unclear how ganglioside accumulation affects the neurite outgrowth from RGCs. One of the most likely explanations for this is that lysosomal storage could affect the survival of RGCs when they are cultured as explants in vitro. We ²⁵ and others ³⁰ reported that optic nerve transection in normal adult rats resulted in the cell death of nearly 50% of the axotomized RGCs by 1 week after the injury. Because the environmental change from in vivo to in vitro may result in RGCs being in a condition that makes it more difficult for them to survive than the optic nerve transection in vivo, the number of surviving RGCs in the retinal explants would rapidly decline during the culture period. Degeneration of the neural retina induced by lysosomal storage could make RGCs more vulnerable to that condition, and thus their cell death in culture might be accelerated in Hexb ^−/− mice. This possibility is supported by the results of electron microscopy, which indicated apoptotic changes of some RGCs at the terminal stage of the disease. In addition to the impaired viability of RGCs, metabolic changes due to deficient ganglioside degradation could affect neurite regeneration. It has been suggested that an increased hexosamine content as well as a decreased free amino-acid pool in the brain of patients with Tay-Sachs disease may limit protein synthesis. ⁴ In a recent study by Buccoliero et al., ³¹ phospholipid synthesis was found to be decreased in neural tissue of the Hexb ^−/− mice. Since these materials are essential for the initiation and elongation of neurites, ³² reduction in their synthesis in RGCs may be one cause of impaired neurite outgrowth in vitro. 

BDNF is a member of the neurotrophin family ³³ and supports the survival of axotomized RGCs both in vivo ³⁰ and in vitro. ³⁴ In addition, neurite outgrowth from axotomized rat RGCs is enhanced by the administration of BDNF in vivo ³⁵ and in vitro. ³⁶ These reports led us to believe that BDNF might exert restoring effects on impaired neurite outgrowth caused by various degenerative disorders. In fact, exogenous application of BDNF can stimulate neurite outgrowth from retinal explants from aged humans ²¹ and streptozotocin-diabetic mice (Takano et al., unpublished data, 1999). In a manner similar that used in those studies, we showed that BDNF significantly enhanced neurite outgrowth from the retina in presymptomatic Hexb ^−/− mice (2 months after birth). Although further studies are needed to clarify the precise mechanism of this enhancement, BDNF is likely to restore both viability and neurite outgrowth capability of RGCs in the mice in vitro. In contrast, the application of BDNF was less effective for the neurite outgrowth in severely affected Hexb ^−/− mice (4 months after birth) than in presymptomatic mice. These findings are in accord with the histopathology of the neural retina by electron microscopy and support the possibility of time- and storage-dependent degeneration and decrease in the neurite outgrowth capability of RGCs. 

²  Contributed equally to the work and therefore should be considered equivalent authors.

 Supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture, Japan, and by funds from Mitsui Life Social Welfare Foundation and Suzuken Memorial Foundation.

 Submitted for publication January 12, 2005; revised April 8, 2005; accepted April 22, 2005.

 Disclosure: K. Sango, None; M. Takano, None; K. Ajiki, None; A. Tokashiki, None; N. Arai, None; H. Kawano, None; H. Horie, None; S. Yamanaka, None

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

 

Figure 1.

 

Fundus appearances of 3-month-old Hexb ^+/+ (A) and Hexb ^−/− mice (B). In the Hexb ^+/+ mouse, the optic nerve fiber bundles (spreading radially from the optic papilla in A) were clearly seen. In contrast, the optic fundus of the Hexb ^−/− mouse (B) was spotted with many white dots, and nerve fibers were hardly visible.

Figure 2.

 

PAS-stained frozen sections of retinas from 4-month-old Hexb ^+/+ (A) and Hexb ^−/− (B) mice. PAS-positive material was abundant in the ganglion cell layer of the Hexb ^−/− mouse (B, arrows). The inner plexiform layer, inner nuclear layer, and outer plexiform layer were also intensely stained. In contrast, little or no storage was visible in a similar section from the Hexb ^+/+ mouse (A). Magnification, ×400.

Figure 3.

 

Electron micrographs of retinal ganglion cells (RGCs) from 3-month-old Hexb ^+/+ (A) and Hexb ^−/− (B) mice and a 4-month-old Hexb ^−/− mouse (C). Numerous inclusion bodies were observed in the cytoplasm of RGCs in the Hexb ^−/− mouse (B, arrow), but not in the Hexb ^+/+ mouse. No apparent cell death was detected in either Hexb ^+/+ or Hexb ^−/− mice at 3 months of age. In 4-month-old Hexb ^−/− mice, however, apoptotic RGCs with darkened and partly disrupted nuclei (C, arrow) were occasionally seen.

Figure 4.

 

Ratios (%) of apoptotic RGCs: Data are the mean ± SEM of calculations from three to four different retinal sections from two animals. More than 300 cells in each group of mice were examined. *P < 0.01 (by Bonferroni-Dunn post hoc analysis) compared with 3-month-old Hexb ^+/+ and Hexb ^−/− mice.

Figure 5.

 

Neurite regeneration from Hexb ^+/+ (A–C) and Hexb ^−/− (D–F) mouse retinal explants after 10 days in culture. Retinas were obtained from 1-month-old (A, D), 2-month-old (B, E) and 4-month-old (C, F) mice, and cultured in three-dimensional collagen gels. Note that the number of neurites in (E) 2- and (F) 4-month-old Hexb ^−/− mouse retinas are significantly lower than those in (B) 2- and (C) 4-month-old Hexb ^+/+ mouse retinas, respectively.

Table 1.

 

Number of Neurites from Retinal Explants of Hexb ^+/+ and Hexb ^−/− Mice after 3, 7, and 10 Days in Culture

Age of Mice	Number of Neurites
		3 Days	7 Days	10 Days
1 Month
Hexb +/+	9.6 ± 0.8	76.8 ± 5.8	116.0 ± 14.0
Hexb −/−	5.6 ± 1.2	44.5 ± 2.9	103.0 ± 7.5
2 Months
Hexb +/+	2.5 ± 0.4	47.2 ± 2.5	69.2 ± 3.7
Hexb −/−	2.2 ± 0.4	14.6 ± 1.8	20.7 ± 2.6
4 Months
Hexb +/+	2.4 ± 0.4	26.7 ± 2.5	55.1 ± 4.7
Hexb −/−	0.2 ± 0.1	4.4 ± 0.8	11.7 ± 1.7

 Data are the mean ± SEM calculated from 64 explants from two animals. Significant differences (P < 0.05, by Mann-Whiney test) are noted between age-matched Hexb ^+/+ and Hexb ^−/− mice as follows: 1 month of age, at 3 and 7 days in culture; 2 months of age, at 7 and 10 days in culture; and 4 months of age, at all culture time points.

Figure 6.

 

Number of neurites from retinal explants of Hexb ^+/+ and Hexb ^−/− mice after 3, 7, and 10 days in culture in the presence or absence of BDNF (100 ng/mL). Data are the mean ± SEM of results in 32 explants from two animals at 2 (top) and 4 (bottom) months of age. Treatment with BDNF significantly enhanced neurite regeneration from both Hexb ^+/+ and Hexb ^−/− retinas at 2 months of age. In contrast, the significant enhancement of neurite regeneration by BDNF is observed in Hexb ^+/+, but not in Hexb ^−/− retinas at 4 months of age. *P < 0.01 compared with untreated Hexb ^+/+. **P < 0.01 compared with untreated Hexb ^−/−.

The authors thank Richard L. Proia for reading the manuscript; Seiichi Ishiguro, Mitsuru Nakazawa, and Takayuki Harada for helpful suggestions on light and electron micrographs; Miwa Sango-Hirade for help in preparing figures; and Tadashi Tai and Kazuhiko Watabe for providing the antibodies.