SEARCH

SEARCH BY CITATION

Keywords:

  • papillofoveal bundle;
  • carbocyanine dye;
  • primate optic nerve;
  • age-related macular degeneration;
  • glaucoma

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Axons in the fovea are precisely organized to ensure accurate vision. We investigated the morphologic characteristics and localization of nerve bundles in the optic nerve in primates. Macaque eyes were studied for conventional and immunostaining, and also marmoset eyes for carbocyanine dye tracing. Locally confined lesions associated with similar findings to human age-related macular degeneration (ARMD) were also evaluated. Axons of retinal ganglion cells formed fasciculi near their origin, and these fasciculi formed bundles thereafter. In the retinal nerve fiber layer, ascending bundles assembled stratification adding proximal bundle underneath successively. Bundles in the arcuate zone displayed a characteristic fine, parallel arrangement, whereas those in the outside zone intermingled with undefined reticular bundles as they approached the optic nerve head. Macular bundles remained in groups and were distributed in the temporal wedge of the optic nerve head. Orthograde and retrograde tracing revealed that these bundles formed confined groups of various sizes and, ultimately, a specific group of small bundles located in the innermost row, near the central vessels. In addition, these bundles showed evidence of focal degenerative deterioration in eyes with ARMD. Papillomacular bundles have a characteristic alignment and configuration. Foveal bundles that compose the confined group closest to the optic trunk (which we term papillofoveal bundles) appear to have functional significance with respect to the isolated lesions that accompany central vision loss or preservation. Anat Rec, 2012. © 2011 Wiley Periodicals, Inc.

The processing of visual information is made possible by the convergence of a long pathway of receptor cells to the visual cortex. The organization of the macular fibers with respect to the positioning of axonal bundles in the optic nerve remains controversial, in particular, the reorganization that occurs at or near the optic nerve head in retinotopy. The most informative division of ascending retinal bundles between the optic nerve head is localized at the superior and inferior arched area of the temporal hemisphere. These bundles enter the optic nerve head between the 1 o'clock and 5 o'clock positions in the posterior retinal dome, the center of which is occupied by the macula. The nerve fiber distribution in this area is characteristically related to the “arcuate” visual field defect in glaucoma. The terminology used to describe bundles localized between the optic nerve head and fovea is confusing. They have been referred to as papillomacular bundles (PMBs) in the central papillomacular area (Pavlidis et al.,2006), arcuate bundles, and central papillomacular bundles (Ogden,1984). In the present study, nerve fiber bundles within the macula (an area approximately 800 μm in diameter in adult macaques) will be referred to as PMBs within the arcuate configuration. Among these PMBs, bundles of climbing axons within the fovea (approximately 100 μm in diameter) will be referred to as papillofoveal bundles (PFBs). We examined the morphologic characteristics and localization of PFBs in the nerve fiber layer (NFL) and optic nerve and used dye tracing to determine their axonal pathways. In addition, the functional relevance of PFBs was evaluated at lesions sites associated with spontaneous age-related macular degeneration (ARMD) in macaques.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

The study comprised nine pairs of eyes from adult Japanese monkeys (Macaca fuscata) weighing 5–9 kg and two pairs of eyes from common marmosets (Callithrix jacchus) of both sexes raised in our two institutes. The experiments were performed in accordance with the Institutional Guidelines for Animal Handling Protocol and Experimentation of the Tokyo Metropolitan Institute of Medical Science. After the animals were killed and exsanguinated under deep anesthesia, the eyes were fixed in a mixture of 4% formaldehyde and 0.2% picric acid in 0.1% phosphate buffer. The sections were stained with Gomori's aldehyde-fuchsin (Ald-fuch), silver-Luxol fast blue-hematoxylin (Ag-LFB-H), silver-Bielshowsky (Ag-Biel), and Bodian for 6-μm paraffin sections. For immunohistochemical analysis, monoclonal mouse antibody to phosphorylated (SMI-31) or dephosphorylated neurofilaments (SMI-32: Sternberger Monoclonal Inc. Lutherville, MD) was applied at a concentration 1:200 to the paraffin-embedded sections as a primary antibody. The sections were deparaffinized, dehydrated, and incubated in blockade for 1–2 hr to reduce nonspecific binding. Specimens were then incubated with one of the primary mouse antibodies at room temperature for 1–3 days. After three 5-min washes in 0.01 M phosphate-buffered solution, biotinylated horse anti-mouse secondary antibody at a concentration of 1:200 was applied for 2 hr, which was followed by application of avidin–biotin complex for another 1.5 hr. For substrate chromogen staining, 3,3′-diaminobenzidine (DAB: Sigma-Aldrich Inc, St Louis, MO)-nickel chloride (DABni) was used for color development.

Carbocyanine dye (DiIC: Molecular Probes, Eugene, OR) was used for tracer staining along the axonal lipophilic membranes and was applied to the entire posterior dome of fixed, postmortem marmoset eyes, which were selected because the distance between the optic nerve and macula is shorter in the marmoset than in the macaque. For orthograde tracing, several crystals moisturized by 95% ethanol were used to impale the macula via a small hole that had been opened in the sclera with a blunt-tip micropipette, so as to avoid rupturing the internal limiting membrane. For retrograde labeling, the dye was injected into the temporal optic nerve adjacent to the retinal central artery, about 3 mm behind the bottom of the optic disc excavation. The eyes were kept in a dark box within an immersing solution of 4% formaldehyde and 0.1% phosphate buffer. The solution was changed weekly, and after 8 weeks it was altered to 30% sucrose with azide-sodium for a week. During the ninth week, 14-μm-thick tissue sections of the eyes were prepared by cryocut (Finetech, Tokyo, Japan) section and the path of the bundle was traced. Images were examined and photographed by using an optic, phase-differential, fluorescent light, and confocal microscope (Nikon TE2000 C1, Tokyo, Japan).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Ascending Macular Nerve Fibers (Fig. 1)

The papillofoveal distance (i.e., distance from the fovea to the central retinal artery) was approximately 3.6 mm in adult macaque (Fig. 1A) and 3.0 mm in adult marmoset. Ascending bundles from the macula remained parallel along their course toward the optic nerve head (Fig. 1B,C). Fiber strata in the retinal NFL were organized into lamella that overlay the proximal bundle along the centripetal pathway; that is, there was no zone of reorganization after the central aggregation (Fig. 1B,C). Axons of retinal ganglion cells formed a fasciculus at their points of origin (Fig. 1D(a)), and these fasciculi aggregated into bundles (Fig. 1D(b)).

thumbnail image

Figure 1. A: The distance between the fovea (M) and central retinal artery (D) is 36 mm in this macaque (frontal section, Ald-Fuch staining). B: Ascending papillo-macular bundles (PMBs) formed layers of overlying proximal bundles (black arrows) in the centripetal pathway to the optic disc (sagittal section, Ag-Biel). C: Bundles containing phosphorylated neurofilaments took parallel alignment and had a lamellar organization (sagittal section, SMI-31, DABni). D(a): Retinal ganglion cell axons formed fasciculi near their proximal origin. (b) Fasciculi grouped into bundles, and several bundles formed groups that traversed the NFL (SMI-31, DABni).

Download figure to PowerPoint

Configuration of Papillomacular Bundles on Optic Nerve Head (Fig. 2)

In the optic nerve head, the size and arrangement of bundles differed by retinal hemisphere (Fig. 2A). PMBs had a fine structure and ran parallel within the intercalating spaces (Fig. 2B(a)), whereas bundles in the nasal half formed large and complex networks, without demarcating lines (Fig. 2B(c)). Outside the arcuate zone, bundles with larger diameters and a reticular configuration shifted initially parallel and intermingled thereafter (Fig. 2B(b)). This configuration was maintained as the bundle passed throughout the lamina cribrosa pores, with little lateral dispersion.

thumbnail image

Figure 2. Difference in bundle alignment in the optic nerve head (frontal section, SMI-31, DABni). B(a): PMBs including papillo-foveal bundles (PFBs) run parallel and radially, with clearly demarcated interspaces between bundles. B(b): The border between arcuate and nasal bundles was characterized by a gradual transition from parallel to bridged bundles. B(c): Nasal bundles formed a complex network. A: Figures B(a–c) are indicated by boxes.

Download figure to PowerPoint

The Grouped Bundles at Innermost Row (Fig. 3)

Adjacent to PMBs, we observed a characteristic group of bundles in the temporal lamina cribrosa (Fig. 3A). These bundles were small in diameter and encircled by wide columns containing glial cells (Fig. 3B), capillaries, and extracellular matrices (ECMs) (Fig. 3C). A group of these bundles was aligned in the innermost row, adjacent to the central vessels (Fig. 3A(a)). These fibers had a more reactive to phosphorylated neurofilaments (Fig. 3D(a)) than to dephosphorylated ones (Fig. 3D(b)). We refer the bundles with these characteristics to PFBs.

thumbnail image

Figure 3. PFBs in the prelaminar region (frontal section). A(a): Fibers from bundle traversed through the septa. A(b): High magnification view of the septa showing astrocyte with process and oligodendroglia (Ag-LFB-H). C: The septum was rich in collagenous and elastic ECMs (EvG). D: The bundles were more reactive to phosphorylated neurofilaments (a: SMI-31, DABni) than to dephosphorylated neurofilaments (b: SMI-32, DABni).

Download figure to PowerPoint

Tracing of the Papillofoveal Bundles (Fig. 4)

A tracer applied to the post-vital posterior segment revealed the orthograde and retrograde pathways of PFBs. Several ascending bundles formed a group unit (Fig. 4A(a)), and these units converged at the optic nerve head, without dispersion (Fig. 4A(b)). The distribution of the ascending grouped bundles was limited in the wedge of the temporal hemisphere; there were fewer than ten in the two examined eyes. Retrograde uptake was dense in nasal region of the foveal cave and a decrease in light was noted in solid bundles in the circumferential row of the fovea (Fig. 4C(a)). Bidirectional tracing revealed that PMBs formed groups of bundles and situated near the central vessel in the optic nerve.

thumbnail image

Figure 4. Track of PFBs on orthograde (A) and retrograde (C) diffusible tracing in a postmortem posterior section of marmoset. B: Stereoscopic view of injection site (black arrow) and disc margin showed limited lateral dispersion. A(b): Photomontage of six frames in a figure that is recomposed with the use of a phase-differential background (frontal section). Ascending PFBs were grouped and traced together (white arrows in (a) [sagittal] and (b) [frontal] views). The advanced unit extended to a region near the central vessel (*). C(a): Retrograde tracer uptake (fontal section) in nasal (marked by right angles) and temporal surrounding bundles around the fovea (*). Upper scale bar: 1 mm. (b) The diffusive pathway in the optic nerve (white dotted arrow) and the disc margin (right angles, sagittal section) is indicated.

Download figure to PowerPoint

The Papillofoveal Bundles in Age-Related Macular Degeneration (Fig. 5)

Four eyes from very old animals showed the characteristic pathologic findings of human ARMD, and the features of PFBs were evaluated in the stages applied to dry, exudative, and cicatricose. In dry stage, focal deterioration of nerve fibers was found in a group of PFBs, and a surrounding column had glial cells with large nuclei and collagenous ECM proliferation (Fig. 5A(a,b)). The exudative stage was characterized by focal loss of PFBs. ECM proliferation was not severe, probably due to the short duration of disease (Fig. 5B(a, b)). Distinctive degeneration of PFBs was found in cicatricose stage. Ascending fibers from the cavity of the fovea accompanied glial cells with large nuclei (Fig. 5C(a)). The neurofilaments of grouped bundles were disrupted in the NFL near the optic nerve head (Fig. 5C(b)). In the prelaminar region, focal defects were observed in two groups of PFBs with glial proliferation (Fig. 5C(c)). These three confined lesions, which were noted in cases with central vision loss, closely corresponded to the morphologic localization of the PFB.

thumbnail image

Figure 5. Focal lesion of the innermost row caused by spontaneous ARMD in the dry (A), exudative (B), and cicatricose stages (C). A(a): Focal degeneration of inner temporal bundles in the lamina cribrosa (sagittal section, Ald-Fuch). (b) High magnification of box in (a) showing a decrease in the number of bundles as well as wide columns with proliferation of large glial cells and collagenous extracellular matrices. B(a): Localized loss in several of the innermost group of bundles (frontal section, Bodian). (b) High magnification view of focal deterioration of bundles (box in (a).) showing a large defect in bundles (white arrows) and residual dots in the space demarcated by the column. C(a): Climbing fibers from the foveal cavity (*) were accompanied by large glial nuclei (sagittal section, Bodian). (b): Focally disrupted group of bundles (white arrows) in NFL (frontal sections, SMI-31, DABni. D, toward disc; M, toward macula). (c) Focal loss of bundles in two groups of the innermost bundles (white arrows) and proliferation of large glial cells were characteristic of more severe deterioration (frontal section, Bodian).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

The efficiency of the visual function of the subcortex is likely a result of the uniform step-by-step concentric organization of ascending axons. Previous research has shown that approximately 130 million RGCs connect to about 1.2 million nerve fibers (Jonas et al.,1990) and that a million nerve fibers make up approximately one thousand fasciculi in the NFL (Ogden,1984). Fasciculi aggregate into bundles that are surrounded by columns, and almost a thousand of these bundles traverse through the pores of the lamina cribrosa. With the addition of underlying proximal bundles, PMBs in the NFL increase in thickness by a factor of 10 as they approach the optic nerve head (Fig. 1B,C). Furthermore, bundles form groups with contiguous bundles, and these macular groups maintain their organization as they traverse through the radial wedge of the optic nerve head within the arcuate hemisphere (Fig. 4B). The parallel alignment of bundles in the temporal arcuate zone is clearly visible in vivo by ophthalmoscopy, and its sectional shadow has been described as an “NFL defect,” which is responsible for superior and/or inferior arcuate field loss in glaucoma. There have been a number of topographical studies relevant to the arcuate fibers (Radius and Anderson,1979; Minkler,1980; Ogden,1983,1984; Pavlidis et al.,2006); however, the importance of foveal fibers has been largely unexamined (Jonas et al.,1990) because central vision loss is present only in advanced glaucoma.

Most RGCs in the papillomacular area are midget cells, that is, cells with small cell bodies and relatively thin proximal axon segments (Pavlidis et al.,2006). Immunofluorescent staining of phosphorylated neurofilaments revealed that axons formed a fasciculus originating from each primary dendrite (Fig. 1D(a)). These fasciculi aggregated into a bundle in the NFL (Fig. 1D(b)). In this study, to identify specific relations to foveal function, the term PFB is used to describe a bundle that has the aforementioned distinctive morphologic features and is within the papillomacular area. Approximately, 90% of foveal RGCs were midget cells, that is, small-diameter axons, (Pavlidis et al.,2006) and the fiber size of PFBs was smallest within the PMB (Ogden,1984). Although the size of individual fibers within bundles could not be discerned, bundles in the temporocentral region were smaller than those outside of that area (Fig. 3). Neurofilaments in the PMB were fine and the arrangement was well defined (Fig. 2B(a)), whereas those in nasal bundles were intermingled with bridges between bundles (Fig. 2B(c)). In addition, bundles being characterized as PFBs were small and had large intercalating spaces (Fig. 3A–C).

In orthograde tracing, the area of traced fasciculi that was occupied by PFBs was found to be approximately 40% of that of PMBs (FitzGibbon et al.,1996). However, the bundle containing fasciculi was not diffusely distributed. In our tracing study using the same dye, the bundles formed groups (Fig. 4B). Orthograde tracing showed several dispersions in the optic nerve head (Fig. 4B), whereas retrograde tracing from the innermost region of the optic nerve allowed us to distinguish PFBs from PMBs by examining focal uptake in climbing fibers throughout the fovea (Fig. 4C(a)). Thus, the localization of PFBs in the optic nerve was clearly revealed.

The age-dependent decline in the number of fibers was found to be approximately five thousand per year, which could affect the differential light threshold in perimetry (Jonas et al.,1992). Because light perception is independent from foveal function, the finding of central vision loss in ARMD is characteristic of its distinct foveal pathology. The tendency of fibers to remain parallel as they pass through the optic nerve head indicates that focal degeneration in ARMD represents PFB localization. In addition, selective deterioration of PFBs depends on the severity of foveal malfunction resulting from time-dependent pathologic gliosis accompanied by proliferation of surrounding ECMs (Fig. 5). In a previous study, the role of anatomic reserve capacity was considered with regard to preservation of central visual in advanced glaucoma (Jonas et al.,1990). PFBs have larger intercalating spaces, more glial cells, and a richer vascular supply, but the reasons for these findings were not apparent in the present study (Fig. 3, B(b),C(b)). The fact of the preservation of central vision in advanced glaucoma might be depending on the morphological characteristics in PFBs.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Authors thank Teiko Kuroda, Michiko Imanishi, and Hiroko Ueda for their excellent technical assistance in preparing the histologic specimens, and Dr. Hironobu Tokuno and Ikuko Tanaka of the Laboratory of Brain Structure for donating the marmoset eyes. Authors also thank Dr. Haruo Okado, director of the division, for his help in conducting the experiment.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED