Evidence for retinal pathology following interruption of neural regulation of choroidal blood flow: Müller cells express GFAP following lesions of the nucleus of Edinger-Westphal in pigeons.

Current Eye Research

Volume 9 number 6 1990

Evidence for retinal pathology following interruption of neural regulation of choroidal blood flow: Muller cells express GFAP following lesions of the nucleus of EdingerWestphal in pigeons Malinda E.C.Fitzgerald, Betty A.Vana and Anton Reiner

Curr Eye Res Downloaded from informahealthcare.com by University of Sydney on 01/01/15 For personal use only.

Department of Anatomy and Neurobiology , 875 Monroe Avenue, University of Tennessee, Memphis, TN 38163, USA

ABSTRACT

INTRODUCTION

Choroidal blood flow in pigeons is regulated by the medial part of the nucleus of EdingerWestphal (EW)via the ipsilateral ciliary ganglion. lnten-uption of this circuit by unilateral lesions of EW results in pathological modifications in the morphology of retinal photoreceptors in the ipsilateral eye in pigeons housed under 12hr light (400 lux)/l2hr dark conditions. In the present study, we examined the effects of unilateral EW lesions on glial fibrillary acidic protein (GFAP) expression by retinal Muller cells in pigeons housed under the same lighting conditions. Since Muller cells in the retina of land vertebrates express increased GFAP during conditions of retinal pathology or stress (e.g.. inflammation or hypoxia). this study would enable us to further evaluate the effects of disruption in the neural regulation of choroidal blood flow on the retina. We found that following EW lesions, retinal Muller cells expressed GFAP, with the precise intracellular location of the GFAP dependent on the amount of time elapsed following the lesion. One week after the EW lesions, GFAP labelling was restricted to the Muller cell endfeet in the nerve fiber layer and ganglion cell layer. By two-three weeks, the labelling had extended outward (or sclerad) into the portions of the Muller cells spanning the inner plexiform layer. Finally, by six weeks post-lesion, the entire extent of the Muller cell from the nerve fiber layer to the outer limiting membrane contained GFAP. No GFAP immunoreactivity in Miiller cells was observed in the eyes contralateral to the EW lesions or in eyes in which the pupil had been fixed and dilated by lesions of the pretectal region. Our results suggest that the retina is in a state of physiological stress following interruption of the neural regulation of choroidal blood flow by EW lesions. Although the precise mechanisms by which altered choroidal blood flow regulation affects Miiller cell GFAP production require elucidation. the results nonetheless highlight the importance of intact neural regulation of choroidal blood flow for retinal health.

Miiller cells are the glial cells that form the supporting framework of the retina and are involved in metabolic interactions with photoreceptors and other cells of the retina. The cell bodies of Miiller cells are located in the inner nuclear layer and these cell bodies give rise to two processes, one directed toward the inner retina and one directed toward the outer retina. Together these two radiallyoriented processes span almost the entire depth of the retina, extending to and forming the inner limiting membrane and the outer limiting membrane. In addition, microvillous extensions of the sclerally-directed processes extend slightly beyond the outer limiting membrane and surround and separate the inner segments of rods and cones. Miiller cells differ from other forms of glia. such as fibrous astrocytes. in that they normally do not contain glial fibrillary acidic protein (GFAP, 47-54 kD) throughout their total extent (1, 2). GFAP is the major constituent of glial intermediate filaments (10 nm) and is found throughout the total extent of glial cells such as astrocytes (1, 3-5). In the retina of land vertebrates, GFAP is only present in the endfeet of Mfiller cells, if at all (1, 4-9). However, in these animals. expression of GFAP by Mfiller cells increases in response to retinal injury, disease, or photoreceptor degeneration (1. 2. 8. 10-13). This increase in GFAP results in the presence of GFAP throughout the entire extent of Miiller cells.

Received on January 16, 1990; accepted on May 10, 1990

0 Oxford University Press

583


Current Eye Research In this study we investigated the immunoreactivity of GFAP in Miiller cells of pigeon retina following unilateral electrolytic lesions of the nucleus of Edinger-Westphal (EW), the parasympathetic component of the oculomotor nuclear complex. In previous anatomical studies, we have demonstrated that the medial subdivision of EW (mEW is part of a circuit that may be involved in the lightmediated reflexive regulation of choroidal blood flow (14-17). The serially connected components of this circuit are the retina - the suprachiasmatic nucleus (SCN) - mEW - the ciliary ganglion - the choroidal blood vessels (Fig. 1). Microstimulation studies have revealed that both SCN and EW mediate increases in choroidal blood flow (18. 19) and that retinal illumination in pigeons leads to increases in choroidal blood flow (Reiner and Fitzgerald unpub. obs.). These results suggest that the SCN-mEW circuit may be involved in increasing choroidal blood flow in response to increases in illumination levels. Since choroidal blood flow plays a vital role in the metabolic and thermoregulatory support of the photoreceptors in the retina (20. 21). adaptive neural control of choroidal blood flow may be essential to the ability of the choroidal vasculature to play its supportative role. This hypothesis is supported by our finding that unilateral EW lesions in pigeons lead to pathological modifications in the morphology of retinal photoreceptors in the ipsilateral eye (22). These modifications manifested themselves as reductions in the diameter of rod and cone outer segments and subtle changes in the inner segments and terminal regions of photoreceptors in the superior pole of the eye. Since Miiller cell expression of GFAP dramatically increases in conditions of retinal stress (e.g.. inflammation or hypoxia) or pathology (e.g.. degeneration or trauma). we have used GFAP expression by Miiller cells in the present study as a further means for evaluating the effects of EW lesions (and,

584

therefore, interruption of adaptive regulation of choroidal blood flow) on the health of the retina. We, in fact, found increased immunocytochemically detectible levels of GFAP in pigeon Miiller cells in the eyes ipsilateral to EW lesion. These changes in the Miiller cell GFAP staining were not observed in control eyes. Thus. interruption of the normal regulation of choroidal blood flow appears to place the retina in a state of physiological stress. A brief report of these findings was previously presented at the 1989 Society for Neuroscience meeting. MATERlALs AND METHODS

Lesions Adult White Carneaux pigeons (supplied by Bowman Gray School of Medicine, WinstonSalem, N.C.) were used and all procedures were carried out in accordance with the Declaration of Helsinki and The Guiding Principals in the Care and Use of Animals (DHEW Publication. NIH 80-23) guidelines. The pigeons were anesthetized with ketamine (0.66 ml/kgl and xylazine (0.33ml/kg) prior to surgical procedures. Stereotaxic techniques were used to cause unilateral electrolytic lesions of either of two target structures: 1) EW in experimental pigeons (n=8). or 2) area pretectalis (AF')in control pigeons (n=2). Since mEW is involved in the regulation of choroidal blood flow, lesions of EW result in an impairment in the adaptive neural regulation of choroidal blood flow. Birds with such lesions, therefore, allowed us to examine the effects of experimentally-produced disruption in the neural regulation of choroidal blood flow on the health of the retina. These unilateral EW lesions, however, typically encroached on lateral EW PEW) (which is involved in the control of the pupil. see Fig. 1) and thereby produced dilation of the pupil and loss of the pupillary light reflex. To separate the effects of disrupted choroidal blood flow (by mEW damage) and the increased amount of light


Current Eye Research falling on the retina (by pupil dilation attending 1EW damage). some pigeons received unilateral lesions of AP, which projects to 1EW and controls the pupillary light reflex (15. 23) (Fig. 1). Such AP lesions lead to a fixed dilated pupil and allow increased amounts of light to fall on the retina, but have no effect on the neural control of choroidal blood flow. Birds with such lesions therefore served as controls for the possible effects of pupil dilation alone on the health of the retina in the EW-lesioned birds. The eyes contralateral to lesions in EWlesioned birds and ipsilateral to lesions in APlesioned birds were also used as controls. To assure the accuracy of electrode placement for both EW and AP lesions. microstimulation techniques were used. Electrode tip localization was considered accurate when stimulation with low threshold 0.5 msec, lOOHz current pulses (50 pA for EW, 100 pA for AP) yielded pupillary constriction (in the ipsilateral eye with EW stimulation. the contralateral eye with AP stimulation)(see Fig. 1). Once the electrode tip placement was confirmed. an electrolpc lesion was made with 1 mA constant current for 30 sec. Fixation The animals were returned to their home cages, allowed to recover and maintained on a 12-hr light/ 12-hr dark photoperiod. with fluorescent lighting for each cage providing 400 lux illumination in the center of each cage (a level that is less than the illumination level in a well-lit room, i.e. 800-900 lux). Pigeons with EW lesions were sacrificed 1-6 weeks (1-2 wks - n=4: 3 wks - n=2; 6 wks - n=2) after the lesion and birds with AP lesions were sacrificed 2-6 weeks post-lesion (n=2). To sacrifice the pigeons, the animals were deeply anesthetized at the appropriate time post-lesion and perfused transcardially, with all perfusions performed between 11:OO a.m. and 1:00 p.m so that the state of photoreceptor outer segment turnover was comparable for each animal (24). Prior to perfusion. 200 units of heparin were

injected into the heart to prevent clotting. The pigeons were then perfused with 50-100 ml of avian physiological saline (0.75% saline). followed by 200-300 ml of 4% paraformaldehyde in 0.1 M lysine-0.01 M sodium periodate-0.1 M phosphate buffer (pH 7.4) (termed PLP fixative). The brains were then

IRIS

1. Schematic diagram illustrating two pathways providing bisynaptic retinal input to the avian nucleus of Edinger-Westphal IEW). one to caudal-lateral EW via the area pretectalis (AP)and one to medial EW via the suprachiasmatic nucleus (SCN). These separate parts of the avian EM' in turn send separate projections to the ipsilateral eye via separate populations of neurons in the ciliary ganglion. The pathway of mEW to the eye via the ciliary ganglion terminates on choroidal vessels and is involved in the regulation of choroidal blood flow. Since the retinal projection to SCN is crossed and since the SCN projection to mEW is largely crossed (with a small ipsilateral component). this pathway provides a neural substrate by which the nervous system may be able to reflexively regulate choroidal blood flow (in response to light) in the eye of origin of the pathway. The pathway from AP to the caudallateral EW controls the pupillary light reflex in the eye of origin of the pathway. Abbreviations: TeO - optic tectum; V - ventricle.

585


Current Eye Research removed and placed in 20% sucrose-lo?? glycerol-0.02% sodium wide (NaAz) in 0.1 M phosphate buffer at 4OC until sectioned. Following perfusion. the eyes were enucleated and the cornea, lens and vitreous dissected away. The remaining eyecup was placed in PLF fixative for 1 hour a t 4OC. After a brief wash in 0.1 M phosphate buffer (pH7.4). the eyes were stored at 4OC in a 0.1 M PB-O.02Oh NaAz-25Oh sucrose solution until sectioned. The brains were sectioned on a sliding microtome at 40 pm and stained with cresyl violet to confirm histologically the accuracy and completeness of the EW and AP lesions. The eyes were sectioned on a cryostat at 20 pm a t -25OC and stored at -2OOC until processed for immunohistochemistry. Immunohistochemistry Slide-mounted sections were first thawed and washed in three - 5 min rinses in 0.1 M phosphate buffer (pH 7.4)(PB) and allowed to air dry. The sections were then incubated overnight a t 4OC in a rabbit-anti-bovine GFAP antiserum, diluted 1:150 - 1:300 with 0.1 M PB containing 0.3% Triton X-100 - 0.001% NaAz1% normal goat serum (NGS). The GFAP antiserum, obtained from Incstar (Stillwater, M N ) , is highly specific for GFAP and binds well to avian GFAP. After the primary antiserum incubation, the sections were rinsed with three 5 minute 0.1 M PB washes and allowed to air dry and then incubated for 1 hour at room temperature with fluorescein-isothiocyanate (FITC)-conjugated donkey anti-rabbit antiserum (Jackson Immunoresearch Lab., Inc., West Grove, PA). diluted to 1:50-1:200 with 0.1M PB containing 0.3Oh Triton X-100 - 0.001~hNaAz 5% normal horse serum (NHS). The sections were then rinsed in 0.1 M PB and allowed to air dry before being cover-slipped with 100 mg pphenylenediamine (PLD) in 10 ml PB with 90 ml of glycerol added (25). The slides were viewed using an Olympus fluorescence microscope and photographed using Kodak TMax 400 film. 586

The specificity of the anti-GFAP labelling was confirmed in two ways: 1) omission of the primary antiserum and normal.incubatlonin secondary: 2 ) labelling of control and experimental tissue with other anti-GFAP antisera.

RESULTS Controls eves contralaterd to EW lesions and iDsilatera1 to A p lesions. No unequivocal labelling of Mtiller cells for GFAP was observed in these control eyes (Fig. 2A). There was, however, some diffuse labelling in the nerve fiber layer (NFL) that was apparently specific and appeared localized within the Mtiller cell endfeet. The low intensity of immunofluorescent staining in the NFL was variable between animals and among different retinal regions in individual animals. This diffuse labelling appeared to be specific, however. since it was largely eliminated by omission of the primary antiserum (and incubation with the secondary) (Fig. 2B),but was only somewhat attenuated by addition of 1% normal goat serum (NGSI to the primary antiserum and 5% normal horse serum (NHS) to the secondary (Fig. 2C). In contrast, these same manipulations showed the perikaryal labelling in the outer nuclear and inner nuclear layers (ONL.INL) to be nonspecific, since such labelling was not affected by omission of the primary (Fig. 2B). but was nearly eliminated by addition of NGS to the primary and NHS to the secondary (Fig. 2C). Thus, the low level of specific labelling in the NFL appears to represent a low level of GFAP in the endfeet of normal Miiller cells. Nonetheless, individual GFAP-labelled endfeet could not be visualized in the NFL (Fig. 2A). Specific GFAP immunoreactivity that appeared to be localized to astrocytes was observed in the NFL in the region of the optic streak (linear optic disc). These astrocytes may be associated with axonal bundles in the



2 . Sections through control eyes from birds with an EW lesion (A1 and a n AP lesion (C) and

a section through a control eye (ipsilateral to an lesion) in which the primary antibody was omitted during the staining (B). A photomicrograph of a control retina from the contralateral eye of a 1 week EW lesioned animal is shown in (A). Note that labelling of Muller cells for GFAP is not evident, with the possible exception that the light staining in the NFL may be present in Muller cell endfeet (NFL. A, C). Similar results were observed in control retina from eyes ipsilateral to AP lesions (not shown). Labelling was also observed in the nuclei of cell bodies in the

Ap

outer nuclear layer (ONL) and inner nuclear layer (INL). This labelling was shown to be nonspecific since it was also observed in sections that had been incubated in only secondary antibody, as shown in (B).Further, this labelling in the ONL and INL was eliminated if NGS was added to the primary antiserum and NHS to the secondary. The photomicrograph in C shows that GFAP labelling in the retina from the contralateral eye from a 6 week A p lesioned animal is also indistinguishable from that in the other control eye. Abbreviations: OPL - outer plexiform layer: IPL - inner plexiform layer: GCL - ganglion cell layer. Scale bar = 50 pm. 587


Current Eye Research NFL, as is the case in the optic disc region in rabbit retina (26). Control eves from animals with AP lesions Animals with this type of lesion, as confirmed histologically and by the dilated fixed pupil in the contralateral eye, possessed normal neural circuitry for control of choroidal blood flow. No increases in Muller cell GFAP staining were observed in the retina affected by the AP lesion (contralateral to the lesion) (Fig. 2C). Thus. both retinas from these animals were indistinguishable in their GFAP labelling pattern from normal retinas and from the control retinas from the EW lesion animals (i.e. contralateral to the EW lesion). As observed in normal retina (Fig. 2A). slight GFAP staining was observed in the NFL (Fig. 2C) and in the inferior retina near the optic streak.

Emenmental eves ipsilateral to EW lesions In the EW-lesioned birds whose results are reported here. all lesions except one

3 . Photomicrograph of a cresyl violet stained

section through the pigeon brain at the level of EW in an animal that was sacrificed two weeks after a unilateral EW lesion. The right side has a glial scar (arrows) at the site of the EW lesion. 588

resulted in the complete destruction of both medial and lateral subdivisions of EW. as confirmed histologically (Fig.3). The pupil ipsilateral to such complete EW lesions was totally dilated and unresponsive to light. In one bird (which was sacrificed three weeks after the lesion). the lesion destroyed all of medial EW and spared substantial portions of lateral EW. The pupil ipsilateral to this lesion was nearly indistinguishable from normal in its resting state of dilation and in its response to light. One week after complete EW lesions, GFAP immunoreactivity was unequivocally present in the endfeet of the Miiller cells in the nerve fiber layer (NFL) and extended into the portions of the Muller cell processes located in the ganglion cell layer (GCL) (Fig. 4A.B). The labelling was not diffuse as observed in control sections and, in contrast to the condition observed in the control eyes, many individual

Note that EW has been completely destroyed. In contrast, on the nonlesioned left side, EW is normal (EW outlined with arrowheads). The somatic divisions of the oculomotor nucleus (OMN) are ventral to EW . Scale bar = 100 pm.

Current Eye Research be observed to pass between and around the cells of the GCL (Figs. 4-6). By two-three weeks post-lesion, a further

4. Photomicrographs of a retina ipsilateral to an EW lesion in an animal sacrificed 1 week post-lesion (A, B). Muller cell processes that are positive for GFAF' span the nerve fiber layer (NFL) and ganglion cell layer (GCL). GFAPimmunoreactive Miiller cell endfeet (curved arrows) can be seen in the NFL and GFAP-

positive processes can be seen, some of which wrap around cells of the GCL, sclerad to the NFL. Muller cell processes were GFAP-positive throughout the retina in all EW lesioned animals. Abbreviations: INL - inner nuclear layer: IPL - inner plexiform layer. Scale bar = 50 pm


Muller cell processes and endfeet could be observed (Fig 4k curved arrow). Miiller cell processes, sclerad to the NFL. could

589


Current Eye Research increase in GFAP labelling was evidenced by the presence of GFAP in the portions of the Muller cells extending through the inner plexiform layer (IPL) (Fig. 5A. B). This labelling pattern was observed both in animals with complete unilateral EW lesions and in the only bird with the lesion largely restricted to mEW. By 6 weeks after complete destruction of EW, GFAP immunoreactivity appeared to span the entire length of the Muller cells (Fig. 6).At all times after the EW lesion and in all EW-lesioned animals, the GFAP immunoreactivity was present throughout all quadrants of the retina. Regardless of the extent of the labelling within individual Miiller cells, the most intense GFAP staining within Muller cells was of their endfeet and their processes in the NFL (Fig. 4-6). lmmunohistochemical controls Omission of the primary antiserum resulted in a total loss of Muller cell labelling in experimental tissue that yielded good Muller cell labelling with anti-GFAP (not shown). The use of other anti-GFAP antisera (e.g. mouseanti-porcine GFAP from Chem Credential, or mouse-anti-astrocytoma-derived GFAP from Lipshaw) yielded labelling patterns in our experimental tissue identical to that seen with the Incstar anti-GFAP. DISCUSSION These results show that, in birds housed under moderate illumination levels, interruption of the normal adaptive regulation of choroidal blood flow by lesions of the EW results in dramatic increases in retinal Miiller cell expression of GFAP. These observed increases in the level of Muller cell GFAP immunoreactivity were not observed In either the control eyes contralateral to the EW lesions, or in either eye from pigeons that received an AP lesion. The increased GFAP levels were also observed in one animal that had a lesion largely restricted to mEW. Thus, the increased levels of GFAP in Muller cells are

590

attributable to a disturbance in the neural regulation of choroidal blood flow and apparently not to a disturbance in the regulation of pupil diameter. The accumulation of GFAP progresses from the vitread portions of the Muller cells toward the scleral portions, with GFAP immunoreactivity clearly evident in the Muller cell endfeet in the NFL by one week and throughout the Muller cell by 6 weeks post-lesion (Fig. 7). Since GFAP labelling throughout the Muller cell, in land vertebrates, has been found in other studies to occur only in conditions of retinal pathology or stress (1. 6. 8. 9. 10-13.27-31),we take our results to support the view that interruptions of the neural regulation of choroidal blood flow can have deleterious consequences on the health of the retina. C0rnDat-l'Sons to Drevious studies In previous studies showing the increased presence of GFAP in Miiller cells during conditions of retinal pathology or trauma, investigators have examined conditions such as hereditary and light induced photoreceptor degeneration, optic nerve section, retinal transplantation, and retinal stab wounds (1. 6. 8, 9. 10-13,27-31).These increases in GFAP levels in the Miiller cells under these conditions appear to be due to induction of GFAP synthesis (9). Since GFAP is thought to exist as a heteropolymer with vimentin within the cytoskeleton of glial cells (32-351, it seems likely that the alteration in GFAP expression in Miiller cells is also accompanied by either a change in vimentin synthesis or a change in the ratio between vimentin and GFAP (36).This possibility, however, has not been extensively explored. In many of the pathological conditions previously reported to lead to increased GFAP expression in Muller cells, considerable trauma and attending cell death occurs in the retina. However, increases in GFAP expression have also been observed to occur in the absence of



5 . Photomicrographs of a retina ipsilateral to an EW lesion in a n animal sacrificed 3 weeks post-lesion (A, B). An increase in GFAP labelling is observed in the Muller cell processes. The labelled portions of the Muller cell processes span the NFL, ganglion cell layer (GCL) and extend through the inner plexiform

layer (IPL). GFAP labelling terminates just vitread to the inner nuclear layer (INL) (arrows). The labelled processes in the nerve fiber layer (NFL) and GCL were often observed to deflect around cells of the GCL (curved arrow, B). Scale bar= 50 pm. .

59 1



6. Photomicrograph of a retinal section ipsilateral to an EW lesion in an animal sacrificed 6 weeks post-lesion. The GFAP labelling now spans the total length of the Miiller cell from the nerve fiber layer (NFL) to the outer nuclear layer (ONL) (arrows). The

GFAP immunoreactivity is still most apparent and intense, however, in the NFL and ganglion cell layer (GCL).Abbreviations: INL - inner nuclear layer: IPL - inner plexiform layer. Scale bar=50 pm.

any detectable cell death (37). For example, Penn and co-workers (37)exposed newborn rats to high oxygen levels (60%)from birth to 14 days of age. The animals were then returned to room air. Physiological and anatomical studies on these animals a t various time intervals (2-8 weeks) after being returned to room air revealed: 1) a temporary paucity of retinal blood vessels: 2) a sustained increase in the expression of GFAP by Miiller cells: and 3)a sustained reduction in the b-wave of the electroretinogram (which originates from the Miiller cell) (38). The poor retinal vascularization of the peripheral retina appears attributable to retardation of the normal development of retinal vessels by the hyperoxic conditions. Although the basis of the increased expression of GFAP in their study was uncertain, it is possible that retinal hypoxia

stemming from the inadequate development of retinal vessels contributed to the increased GFAP expression. The reduction in the b-wave of the electroretinogram further confirms that Miiller cells were functioning abnormally. It is significant to also note that there was no observable retinal cell death or pathology in the Penn et al. study. Previous authors have reported that the normalcy of retinal morphology may not reveal the actual state of retinal pathophysiology in early disease states of the eye. For example. in rats that have been exposed to light levels that subsequently lead to the appearance of retinal histopathology. there is a reduction of up to 5O0h in the amplitude of the a-wave of the electroretinogram prior to any histological evidence of retinal pathology (39).Thus, even seemingly mild alterations in retinal

592



NORMAL

1WK EW LESION

3 WK EW LESION

6 W K EW LESION

VITREOUS

7. Schematic drawing of a series of Muller cells demonstrating the gradually increasing extent of GFAP labelling in individual Muller cells at increasing time points after an ipsilateral EW lesion. The shading indicates the localization of the GFAP labelling. The GFAP localization is restricted to the portions of the Muller cells in the NFL and GCL at one week post-lesion, is additionally found in the portions of Muller cells in the IPL and vitread to the INL at twothree weeks and extends throughout the Muller cell and its processes by six weeks. Abbreviations: OLM - outer limiting membrane: ONL - outer nuclear layer: OPL - outer plexiform layer: INL - inner nuclear layer: IPL inner plexiform layer: GCL - ganglion cell layer: NFL - nerve fiber layer.

morphology may be associated with alterations in retinal physiology and/or health that are severe enough to lead to Muller cell GFAP expression. It is possible that the Muller cells are among the first to respond to retinal insults: and therefore, Muller cell GFAP immunoreactivity may be a more sensitive indicator at the LM level of a n ongoing challenge to the health of the retina than the histological appearance of the retina with standard LM staining techniques (e.g. toluidine blue). The results of our present study and that of our previous study (22) also show that increased GFAP immunoreactivity in Muller cells can occur without any readily detectable

cell death or trauma in the retina. In our previous study, we have observed that electrolytw lesions of EW result in alterations in the morphology of the outer segments of the cones and rods, and of inner segments of some photoreceptors (22). b u t do not result in any evident death of photoreceptors. The changes in photoreceptors were seen as reductions in the cross-sectional diameter of many outer segments. These alterations, which were only readily evident at the E M level and only present in the superior retina, were observed beginning at one week post-lesion and lasted for three months. In contrast, the increased GFAP expression following EW lesions was very apparent at the LM level even at one week postlesion and was observable throughout the retina. Thus, even seemingly subtle alterations in retinal morphology may be associated with alterations in retinal physiology and/or health that are severe enough to lead to Miiller cell GFAP expression. Although the exact cause of the increased GFAP response in Miiller cells in our study is currently unknown, it seems likely that a n adverse biochemical or physiological alteration may have occurred in the retina as a result of the alteration in the control of choroidal blood flow after EW lesions. A further interesting aspect of our findings is that although the photoreceptor changes we have observed following EW lesions are mainly evident in the superior retina (22). Miiller cell GFAP immunoreactivity is increased throughout the entire retina following such lesions. Localized stab wounds of the eye, however, are also reported to lead to GFAP expression in Muller cells throughout the retina (1, 2. 30. 40. 41). The basis of the pan-retinal response to a localized wound is unclear. Compartmentalization of GFAP exmession Some investigators have reported GFAP expression to be present, in retinal Miiller cells of land vertebrates, b u t confined to the Miiller cell endfoot region (1. 4-9). In this study, we observed low levels of GFAP immunoreactivity

593


Current Eye Research in the nerve fiber layer that might be attributable to the presence of low levels of expression of GFAP in Muller cell endfeet. Following EW lesions, Muller cells in the ipsilateral eye were observed to increase their GFAP immunoreactivity, resulting in the progressive accumulation over time of GFAP throughout the entire Miiller cell. This accumulation process was initiated in the endfeet in the NFL and progressed sclerad. However, even with the longer post-lesion survival times that resulted in the accumulation of GFAP throughout the entire cells. GFAP labelling still appeared most intense in the endfeet in the NFL and in Muller cell processes in the GCL. A similar progression in the increase in GFAP levels has been observed by other investigators during other retinal pathological conditions (9. 10, 12, 13. 37). The seemingly consistently higher levels of GFAP in the endfeet suggests that these portions of the cell may be specialized in function. For example, potassium channels are more abundant at the endfeet (42) and the intermediate filaments with which GFAP is associated may in some currently unknown way foster such regional specialization (4). Thus, the increased immunoreactivity of GFAP by the Muller cells may be indicative of a change in the regional functional specialization of Muller cells. The Decten and the GFAP response While the choroid is the major blood supply to the pigeon retina, it is not the only vascular supply. Although pigeons lack blood vessels within their retina, the central retinal artery in pigeons provides a rich vascular supply to a f a n shaped intraocular structure termed the pecten (43,44). The pecten protrudes from the retina into the vitreous a t the head of the optic nerve, which is in the inferior retina. All birds lack retinal blood vessels and in all birds a highly vascularized pecten is present and is thought to serve as an inner retinal blood supply (43.44). It is interesting that the Muller cells located in

594

the inferior retina near the origin of the pecten were indistinguishable in their GFAP reponse from those located in parts of the retina distant from the pecten. suggesting that the vascular role played by the pecten in relationship to the inner retina is inadequate to prevent the GFAP response following interrupted regulation of choroidal blood flow. Mechanisms of retinal stress following interruDtio n of neural control of choroidal blood flow Our results can be taken to indicate that retinal Muller cells in pigeons express GFAP In response to the physiological stress placed on the retina following lesions that disrupt normal adaptive regulation of choroidal blood flow. I t should be noted that the EW lesions in this study did not eliminate choroidal blood flow. Rather the EW lesions disrupted the ability of the choroidal blood flow to respond adaptively to the needs of the retina. thereby presumably placing the retina in a state of frequent insufficiency in relation to its needs. This interruption of adaptive regulation in choroidal blood flow may result in an insufficient supply of oxygen and nutrients for the retina, or a n insufficient ability to remove waste products. Additionally, loss of normal regulation of choroidal blood flow may have led to a deficiency in the ability to use the choroidal vascular bed a s a heat sink to thermoregulate the retina. Consistent with this possibility, increases in ocular tissue temperature have been reported following increased exposure to light, and increased blood flow in the choroidal vasculature has been shown to reduce the heat load on the retina (45-48). In our experiments, we can not totally rule out the possible contributing role of increased pupil diameter following complete EW lesions to the GFAP response. Increased pupil diameter would lead to increased retinal illumination in pigeons with total EW lesions and sustained light levels that are greater than normal are known to be damaging to the retina.


Current Eye Research Nonetheless, our control AP lesions also resulted in a pupil that was fixed and dilated, but these lesions did not result in a GFAP response. Thus, the increased level of light falling on the retina in our EW lesioned birds did not by itself lead to the GFAP response. The GFAP response in these birds appears directly attributable to the loss in the ability to adaptively regulate choroidal blood flow by this circuit. Nevertheless, in conjunction with the loss of compensatory regulation of choroidal blood flow, the increased levels of light falling on the retina in our EW lesioned birds may have contributed to the retinal stress that reveals itself as the increased expression of GFAP. To explore this possibility it will be necessary to carry out studies in which we make lesions that interrupt blood flow regulation, but not pupil diameter, or carry out studies in which we make total EW lesions and house the birds in total darkness. If increased light levels falling on the retina do augment the GFAP response, we would expect less of a GFAP response in the two lines of study suggested. Mechanism of GFAP resDonse to disruDtion of neural regulation of choroidal b lood flow Our results suggest that after EW lesions that disrupt normal adaptive regulation of choroidal blood flow, the physiological state of the retina is altered, thereby resulting in GFAP expression by the Muller cells. It is currently unclear whether the Muller cell GFAP expression is a direct response to the choroidal blood flow insufficiency or whether it is an indirect effect. Since the Muller cells function as support cells of the retina, it seems likely that the GFAP expression is in large part an indirect response to the effects of the blood flow insufficiency on other retinal cells. For example, the Miiller cells may be responding to a change in their ionic environment. Increases in subretinal potassium levels have been reported following hypoxia (49). In our studies, transient episodes of retinal hypoxia may have resulted following EW lesions due to transient

failures of blood flow increases in conditions of heightened need. As we suggested earlier in the discussion, the increased levels of GFAP in Miiller cells in the Penn et al. (37)study might also be attributable to hypoxia. Increases in extracellular potassium levels, which (in uiuo) lead to changes in neuronal activity. have been observed to increase the expression of GFAP in cultured glial cells (50). If increased extracellular potassium concentrations are present in the retina of pigeons with EW lesions, it should be possible to verify this either by recording with potassium electrodes or by measuring the standing potential of the retina (49). Alternatively. the Miiller cells may be expressing GFAP due to growth and/or migration in which they are extending their processes or cell bodies into space left by degenerating or shrunken photoreceptors or other retinal cell types (10. 13). Such growth or migration, however, seems less likely since we have not observed significant cellular loss in the retinas of EW lesioned birds (22). Summarv and implications Our studies indicate that the homeostatic state of the retina is deleteriously affected by lesions that disrupt the neural regulation of choroidal blood flow. This alteration in the homeostatic state of the retina was revealed by the increased expression of GFAP in retinal Muller cells following lesions of the nucleus of Edinger -Westphal. These results indicate the importance of the intact neural regulation of choroidal blood flow for the health of the pigeon eye and suggest that alterations in such control could be a contributing or causal factor in some disease states of the eye.

ACKNOWLEDGEMENTS We gratefully thank Betty Cook, Ellen Karle. Donna Purifoy, and Wes Sweeney and for their excellent technical assistance. We would also like to thank Drs. Ruth B. Caldwell, Nigel G.F. Cooper, and Roue1 S. Roque for their

595

Current Eye Research suggestions and discussions during the course of this study.


From: the Department of Anatomy and Neurobiology, University of TN. Memphis T N . Supported by NIH Research Grants: No. EY05298 (AR), NS-19620 (AR). CORRESPONDNG AUTHOR Anton Reiner. Ph.D., Department of Anatomy and Neurobiology, 875 Monroe, University of TN. Memphis, TN 38163. REFERENCES 1. Bignami. A. and Dahl, D. (1979) The radial glia of Mfiller in the rat retina and their response to injury. An immunofluorescence study with antibodies to the glial fibrillary acidic protein. Exp. Eye Res. 28. 63-69. 2. Shaw, G. and Weber, K. (1983) The structure and development of the rat retina: an immunofluorescence microscopical study using antibodies specific for intermediate filament proteins. Eur. J. Cell Biol. 3.Q. 219-232. 3. Bignami, A.. Eng. L.F.. Dahl. D. and Uyeda, C.T. (1972) Localization of the glial fibrillary acidic protein in astrocytes by immuno-fluorescence. Brain Res. 43,429435. 4. Lewis. G.P., Erickson. P.A.. Kaska, D.D. and Fisher. S.K. (1988) An immunocytochemical comparison of Muller cells and astrocytes in the cat retina. Exp. Eye Res. 47, 839-853. 5. Lewis, G.P., Erickson, P.A., Guerin. C.J.. Anderson D.H. and Fisher, S.K. (1989) Changes in the expression of specific Mfiller cell proteins during long-term retinal detachment. Exp. Eye Res. 93-111. 6. Bjorklund, H.. Bignami, A. and Dahl, D. (1985) lmmunohistochemical demonstration of glial fibrillary acidic protein in normal rat Mtiller glia and retinal astro363-368. cytes. Neuroscience Lett. 7. Karschin A.. Wassle. H. and Schnitzer, J. (1986) Shape and distribution of astrocytes in the cat retina. Invest. Ophthalmol. Vis. Sci. 27. 828-83 1 . 8. Erickson. P.A.. Fisher, S.K., Guerin. C.J., Anderson, D.H. and Kasha, D.D. (1987) Glial fibrillary acidic protein increases in Miiller cells after retinal detachment. Exp. EyeRes. 94, 37-48.

9.

10.

11.

12.

13.

14.

15.

16.

s,

m,

596

17.

18.

19.

Sarthy. P.V. and Fu,M. (1988) Photoreceptor degeneration leads to induction of the glial intermediate filament protein gene in the mouse retina. In "Molecular Biology of the Eye: Genes, Vision, and Ocular Disease". (eds. Piatigorsky. J.. Shinohara. T.. and Zelenka. P.S.). Pp. 305315. Liss, A.R.. Inc.. New York. Eisenfeld, A.J., Bunt-Milam, A.H. and Sarthy. P.V. (1984) Muller cell expression of glial fibrillary acidic protein after genetic and experimental photoreceptor degeneration in the rat retina. Invest. Ophthalmol. Vis. Sci. 25. 1321-1328. Hiscott, P.S.. Grierson, I., Trombetta, D.J., Rahi, A.H.S., Marshall, J. and McLeod. D. (1984) Retinal and epiretinal glia- a n immuno-histochemical study. Br. J. Ophthalmol. 68. 698-707. Sheedlo. H.J. and Turner. J.E. (1988) Immunocytochemical investigation of GFAP and (Na.K)-ATPasein the retina of control and RCS dystrophic (rdy/rdy) rats during development. SOC. Neuro. Sci. Abstr. l 4 ( 1 ) . 359. Roque. R.S. and Caldwell, R.B. (1989) Glial cell migration accompanies RPE cell alterations and vascular proliferation in the RCS retina. Invest. Ophthalmol. Vis Sci. 30 ISUDD 1.1. 464. Gamlin. P.D.R., Reiner. A. and Karten. H.J. (1982) Substance P-containing neurons of the avian suprachiasmatic nucleus project directly to the nucleus of EdingerWestphal. Proc. Natl. Acad. Sci. U.S.A. 79. 389 1-3895. Reiner. A..Karten. H.J.. Gamlin. P.D.R. and Erichsen. J.T. (1983) Functional subdivisions and circuitry of the avian nucleus of Edinger-Westphal. TINS, 6, 140-145. Reiner. A. (1987) The presence of substance P/CGRP-containing fibers, VIPcontaining fibers and numerous cholinergic fibers on blood vessel of the avian choroid. Invest. Ophthalmol. Vis. Sci. 2 8 ( s U D D l . l , 81. Meriney, S.D. and Pilar. G. (1987) Cholinergic innervation of the smooth muscle cells in the choroid coat of the chick eye and its development. J . Neurosci. Z, 3827-3839. Reiner, A. and Fitzgerald, M.E.C. (1989) Control of choroidal blood flow by the nucleus of Edinger-Wesphal: A laserDoppler study. Invest. Ophthalmol. Vis. Sci. 3 0 I S u plJ, ~ 464. Reiner, A, Fitzgerald. M.E.C. and Gamlin, P.D.R. (1990) Central neural circuits controlling choroidal blood flow: A laserDoppler Study. Invest. Ophthalmol. Vis. Sci. ~ S U DInDpress. ~ ~


Current Eye Research 20. Bill, A. (1984) The circulation in the eye. In "The Microcirculation. Part 2 The Handbook of Physiology, Section 2."(eds. Renkin E.M. and Michel. C.C.). Pp10011035. The American Physiological Society. 21. Bill, A. (1985) Some aspects of the ocular circulation. Invest. Ophthalmol. Vis. Sci. 26. 410-424. 22. Fitzgerald, M.E.C. and Reiner, A. (1989) Lesions of the nucleus of Edinger-Wesphal deleteriously affect photoreceptors in avian retina. Invest. Opthalmol. Vis. Sci. ~ O ~ S U 464. D D ~ ~ 23. Gamlin, P.D.R., Reiner, A., Erichsen. J.T.. Karten. H.J. and Cohen, D.H. (1984) The neural substrate for the pupillary light reflex in the pigeon (Columba liuia). J. Comp. Neurol. 226. 523-543. 24. Young, R.W. (1978) The daily rhythm of sheding and degradation of rods and cone outer segment membranes in the chick retina. Invest. Ophthalmol. Vis. Sci. 17, 105-1 16. 25. Platt, J.L. and Michael, A.F. (1983) Retardation of fading and enhancement of intensity of immunofluorescence by pphenylenedine. J. Histochem. Cytochem. 31, 840-842. 26. Stone, J. and Dreher, 2. (1987) Relationship between astrocytes. ganglion cells and vasculature of the retina. J. Comp. Neurol. 35-49. 27. Miller, N.M. and Oberdorfer M. (1981) Neuronal and neuroglial responses following retinal lesions in the neonatal rats. J. Comp. Neurol. 202. 493-504. 28. Rodrigues, M.M.. Wiggert, B., Tso, M.O.M. and Chader. G.J. (1986) Retinitis pigmentosa immunohistochemical and biochemical studies of the retina. Can. J . Ophthalmol. 21. 79-83. 29. Nork. T.M., Wallow, H.L.. Sramek. S.J. and Anderson, G. (1987) Miiller's cell involvement in proliferative diabetic retinopathy. Arch. Ophthalmol. 105, 1424-1429. 30. Seiler. M. and Turner, J.E. (1988) The activities of host and graft glial cells following retinal transplantation into the lesioned adult rat eye: developmental expression of glial markers. Dev. Brain Res. 43,111-122. 31. Schnitzer, J. (1988) Immunocytochemical studies on the development of astrocytes. Muller (glial) cells, and oligodendrocytes in the rabbit retina. Dev. BrainRes. 44. 59-72. 32. N6ell. WK.. Walker, V.S.. Kang. B.S. and Berman. S . (1966) Retinal damage by light in rats. Invest. Ophthalniol. 5,450473.

a,

33. Sharp G.A., Osborn. M. and Weber, K. (1982) Occurence of two different intermediate filament proteins in the same filament in situ within a human cell line. Exp. Cell Res 141, 385-395. 34. Quinlan, RA. and Francke, W.W. (1983) Molecular interactions in intermediatesized filaments revealed by chemical cross-linking. Heteropolymers of vimentin and glial filament protein in cultured human glioma cells. Eur. J . Biochem. 132, 477-484. 35. Tolle. H-G.. Weber, K. and Osborn, M. (1986) Microinjection of monoclonal antibodies to vimentin. desmin and GFA in cells which contain more than one IF type. Exp. Cell Res. 162. 462-474. 36. Bignami. A. (1984) Glial fibrillaxy acidic IGFA) protein in Muller glia. Immunofluorescence study of the goldfish retina. Brain Res. m,175-178. 37. Penn. J.S., Thum. L.A.. Rhem M.N., and Dell, S.J. (1988) Effects of oxygen rearing on the electroretinogram and GFAprotein in the rat. Invest. Ophthalmol. Vis. Sci. 29. 1623-1630. 38. Miller, R.F. and Dowling. J.E. (1970) Intracellular responses of the Muller (glial) cells of mudpuppy retina: Their relation to b-wave of the electroretinogram. J. Neurophysiol. 33. 323-341 39. Bromberg, J.S. and Schachner. M. (1978) Localization of nervous system antigens in retina by immunohistology. Invest. Ophthalmol. Vis. Sci. 17,1322-1328. 40. Ekstrom, P.. Sanyal. S . . Narfstrom. K.. Chader. G.J. and van Veen, T. (1988) Accumulation of glial fibrillary acidic protein in Miiller radial glial during degeneration in the rat retina. Invest. Ophthalmol. Vis. Sci. a . 1 3 6 3 - 1 3 7 1 . 41. Seiler. M. and Turner, J.E. (1988) The activities of host and graft glial cells following retinal transplantation into the lesioned adult rat eye: developmental expression of glial markers. Dev. Brain Res. Q, 111-122. 42. Brew, H.. Gray, P.T.A.. Mobbs, P. and Attwell, D. (1986) Endfeet of retinal glial cells have higher densities of ion channels that mediate K+ buffering. Nature (London). 324. 466-468. 43. Meyer, D. (1977) The avian eye and its adaptations. In "The visual system in vertebrates. Handbook of Sensory Physiology Vol VII/5". (ed. Crescitelli, F.). Pp 550-1035 Springer Verlag, Berlin. 44. Pettigrew. J.D., Wallman, J. and Wildsoet. C.F. (1990) Saccadic oscillations facilitate ocular perfusion from the avian pecten. Nature 343.362-363.

597

Current Eye Research 45. Parver, L., Auker, C. and Carpenter, D. (1980) Choroidal blood flow as a heat

dissipating mechanism in the macula. Am.

J. Ophthalmol. 46. Parver,

89.641-646.

L.. Auker. C.. Carpenter, D. and

Doyle, T. (1982) Choroidal blood flow. 11. Reflexive control in the monkey. Arch. Ophthalmol. 1327-1330. Parver, L., Auker. C. and Carpenter, D. (1983) Choroidal blood flow. 111. Reflexive control in human eyes. Arch. Ophthalmol. 101.1604- 1606. Bill, A., Sperber. G. and Ujiie. K. (1983) Physiology of the choroidal vascular bed. Int. Ophthalmol. fi, 101-107. Steinberg, R.H. (1987) Monitoring communications between photoreceptors and pigment epithelial cells: effects of 'mild" systemic hypoxia. Invest. Ophthalmol. Vis. Sci. 1888-1904. Canady. KS.,Ali-Osman, F. and Rubel. E.W. (1988) Extracellular potassium and osmolarity influence DNA systhesis and glial fibrillary acidic protein expression in cultured glial cells. SOC.Neuro. Sci. Abstr.

m.

47.


48. 49.

a,

50.

14(1), 432.

598

Visual performance of pigeons following hippocampal lesions.

Control of retinal and choroidal blood flow.

Nucleus striae terminalis lesions affect agonistic behavior of pigeons.

The effects of halothane on retinal and choroidal blood flow in cats.

New techniques for interruption of gastroesophageal venous blood flow.

The Edinger-Westphal nucleus: sources of input influencing accommodation, pupilloconstriction, and choroidal blood flow.

Cholecystokinin-like immunoreactive retinal ganglion cells project to the ventral lateral geniculate nucleus in pigeons.

Evidence for activation of the complement cascade in the hypoglossal nucleus following peripheral nerve injury.

Effect of cyclooxygenase inhibition on retinal and choroidal blood flow during hypercarbia in newborn piglets.

Choroidal mast cells in retinal pathology: a potential target for intervention.

Temperature modulating action of choroidal blood flow.

Evidence for involvement of nitric oxide in the regulation of hypothalamic portal blood flow.

No Evidence for Retinal Damage Evolving from Reduced Retinal Blood Flow in Carotid Artery Disease.

Muscle blood flow following thermal injury.

Long-term changes in phosphoinositide hydrolysis following colchicine lesions of the nucleus basalis magnocellularis.

Neural crest origin of retinal and choroidal pericytes.

Ethanol drinking following 6-OHDA lesions of nucleus accumbens and tuberculum olfactorium of the rat.

Craniofacial alterations following electrolytic lesions of the trigeminal motor nucleus in actively growing rats.

Symptom correlates of cerebral blood flow following acute concussion.

Treatment of serous retinal detachment associated with choroidal ischemia with intravitreal bevacizumab following brain surgery.

Visual discrimination performance after lesions of the ventral lateral geniculate nucleus in pigeons (Columba livia).

Acute myocardial ischemia secondary to interruption of coronary blood flow.

Regulation of cortical blood flow by the dorsal raphe nucleus: topographic organization of cerebrovascular regulatory regions.

Entrainment of the rat's activity rhythm by cyclic light following lateral geniculate nucleus lesions.