Conceived and designed the experiments: ED HMC. Performed the experiments: ED ODB EW GR. Analyzed the data: ED ODB EW HMC. Contributed reagents/materials/analysis tools: EW LAW. Wrote the paper: ED ODB CC PD HMC.
The authors have declared that no competing interests exist.
Glaucoma is a widespread ocular disease and major cause of blindness characterized by progressive, irreversible damage of the optic nerve. Although the degenerative loss of retinal ganglion cells (RGC) and visual deficits associated with glaucoma have been extensively studied, we hypothesize that glaucoma will also lead to alteration of the circadian timing system. Circadian and non-visual responses to light are mediated by a specialized subset of melanopsin expressing RGCs that provide photic input to mammalian endogenous clock in the suprachiasmatic nucleus (SCN). In order to explore the molecular, anatomical and functional consequences of glaucoma we used a rodent model of chronic ocular hypertension, a primary causal factor of the pathology. Quantitative analysis of retinal projections using sensitive anterograde tracing demonstrates a significant reduction (∼50–70%) of RGC axon terminals in all visual and non-visual structures and notably in the SCN. The capacity of glaucomatous rats to entrain to light was challenged by exposure to successive shifts of the light dark (LD) cycle associated with step-wise decreases in light intensity. Although glaucomatous rats are able to entrain their locomotor activity to the LD cycle at all light levels, they require more time to re-adjust to a shifted LD cycle and show significantly greater variability in activity onsets in comparison with normal rats. Quantitative PCR reveals the novel finding that melanopsin as well as rod and cone opsin mRNAs are significantly reduced in glaucomatous retinas. Our findings demonstrate that glaucoma impacts on all these aspects of the circadian timing system. In light of these results, the classical view of glaucoma as pathology unique to the visual system should be extended to include anatomical and functional alterations of the circadian timing system.
The hallmark of glaucoma is the degenerative loss of retinal ganglion cells (RGCs) and their optic nerve fibers. In the absence of adequate treatment glaucoma inevitably leads to blindness and is expected to affect more than 60 million people worldwide by 2010
In the mammalian retina, the light sensitive RGCs that express the photopigment melanopsin
To explore the hypothesis that glaucoma leads to alterations in both the visual and non-visual systems we used a rat model of glaucoma with laser-induced chronically elevated IOP
Male Wistar rats (n = 30) were housed individually in propylene cages, under a 12:12 LD cycle, with food and water
Argon laser treatment (blue-green argon laser; Coherent, Palo Alto, CA) of the episcleral veins (responsible for aqueous outflow tissue) was used to induce chronic elevation of IOP according to previous methods
IOP in the operated eyes are compared with that in the unoperated eyes from the same individuals (n = 12). Arrows represent the times of the three laser surgeries. The serial sequence of laser treatments assures a long-term chronic elevation of IOP.
Rats with monocular laser surgery (n = 4 of the 12 rats described in laser treatment) were anesthetized with a mixture of ketamine (50 mg/kg) and xylazine (25 mg/kg). Eyes were additionally anesthetized locally using topical application of oxybuprocaïne chlorhydrate. Two rats received a 0.5 mg intraocular injection (6 µl) of Cholera toxin subunit b (CTb) Alexa Fluor 594 conjugate (red fluorescence, #C-22842, Molecular Probes, CA) in the operated right eye and a 0.5 mg intraocular injection (6 µl) of CTb Alexa Fluor 488 conjugate (green fluorescence, #C-22841, Molecular Probes, CA) in the unoperated (control) left eye. The 2 other rats received the same two intraocular injections of CTb fluorochromes but with the reverse Alexa Fluor 488 conjugate in the operated right eye and Alexa Fluor 594 conjugate in the control left eye.
Rats with monocular laser surgery (right eye, n = 8 of the 12 rats described in laser treatment) and controls with no laser surgery (n = 8) were anesthetized with a mixture of ketamine (50 mg/kg) and xylazine (25 mg/kg). Eyes were additionally anesthetized locally using topical application of oxybuprocaïne chlorhydrate. Rats with monocular induced glaucoma received a 0.5 mg injection (6 µl) of CTb (#130A, List Biological Laboratories Inc) in the operated eye. Control rats with no laser surgery received the same injection in their right eye.
For both the CTb fluorescence and diaminobenzidine (DAB) studies, 48 hr after the injection, all animals were deeply anesthetized with a lethal intraperitoneal injection of sodium pentobarbital (150 mg/kg) and perfused intracardially with warm (37°C) heparinized saline followed by 300 ml of Zamboni's fixative at 4°C. Brains were removed, and post-fixed overnight in a mixture containing the same fixative with 30% sucrose for cryoprotection at 4°C. Serial coronal sections were made at 50 µm on a freezing microtome and all brain sections were collected. Sections from the animals injected with the CTb anterograde tracer coupled to a fluorochrome were directly mounted on slides, dehydrated and coverslipped.
All sections from all animals injected with CTb were processed at the same time to obtain identical levels of tissue staining for data analysis. Endogenous peroxidase was first suppressed using a solution of 50% ethanol in saline with 0.03% H2O2. Free-floating sections were rinsed briefly in PBS (0.01 M, pH 7.2) containing 0.3% Triton and blocked with 1% bovine serum albumin. Sections were incubated in the anti-CTb primary antibody (dilution 1∶3,000) for 3 days at 4°C. Immunoreactivity was visualized using a Vectastain ABC Elite kit (PK-6100, Vector Laboratories, Burlingame, CA), followed by incubation in 0.2% 3,3′- DAB with 0.5% ammonium nickel sulfate and 0.003% H2O2 in Tris buffer (0.05 M, pH 7.6). Sections were then mounted on slides, dehydrated and coverslipped.
Retinal projections to the brain were quantified on sections processed for DAB immunocytochemistry, based on a published methodology
At the end of the behavioral study the two retinas were pooled from each of the animals (n = 5 binocular induced glaucoma rats and n = 5 control rats, age 14–16 months). Retinas were collected from the animals between ZT8-ZT9, after the animals had been re-entrained to a 100 lux LD cycle. Total RNA was extracted using GenEluteTM Mammalian Total RNA Miniprep Kit (Sigma) according to the manufacturer's instructions and subsequently subjected to DNase digestion. Total RNAs was reverse transcribed using random primers and MMLV Reverse Transcriptase (Invitrogen). Real-time PCR was then performed on a LightCycler™ system (Roche Diagnostics) using the light Cycler-DNA Master SYBR Green I mix. The efficiency and the specificity of the amplification were controlled by generating standard curves and carrying out melting curves and agarose gels of the amplicons respectively. Relative transcript levels of each gene were calculated using the second derivative maximum values from the linear regression of cycle number versus log concentration of the amplified gene. Amplification of the non cyclic control gene 36B4 was used for normalization. Each reaction was performed in duplicate. Primer sequences were the following:
SW opsin forward
SW opsin reverse
MW opsin forward
MW opsin reverse
Rhodopsin forward
Rhodopsin reverse
Melanopsin forward
Melanopsin reverse
Thy1 forward
Thy1 reverse
PACAP forward
PACAP reverse
36B4 forward
36B4 reverse
For monitoring locomotor activity, rats operated bilaterally to induce experimental glaucoma were used. A total of 10 rats (n = 5 with binocular raised IOP, n = 5 controls) were housed individually in cages equipped with passive infrared motion captors placed over the cages and a computerized data acquisition system (CAMS, Circadian Activity Monitoring System, INSERM, France). Rats were initially maintained for 26 days under a 12:12 LD cycle with broad-band white light (100 lux). Animals subsequently underwent a 6 hr phase delay of the LD cycle, associated with successive decreases of light intensity (from 100 to 10 lux and subsequently from 10 to 1 lux).
Activity records were analyzed with the Clocklab software package (Actimetrics, Evanston, IL). The time of locomotor activity onsets was determined using the Clocklab onset fit algorithm. Animals were considered to be entrained when the fit of the least squares regression line of the activity onsets was stable in relation to lights-off for at least 7 days of in LD. The phase angle, defined as the time difference between the onset of the activity rhythm and time of lights-off was determined for each animal (see
Data were analyzed using the SigmaStat software (Systat Software Inc., Point Richmond, CA). The
All operated eyes of animals treated with laser surgery displayed a consistent chronic increase in IOP (
The unoperated control eye was injected with a red fluorescent CTb anterograde tracer (A) and the operated glaucomatous eye with a green fluorescent CTb tracer (B). The images in A and B are taken from the same sections but using different excitation filters. For each of the structures illustrated, the brain hemisphere ipsilateral to the injected eye is to the left and contralateral to the injection is to the right. Thus, the ipsilateral dLGN and SC (red: control) correspond to the same hemisphere as the contralateral dLGN and SC (green: operated). Note that projections to the SCN and SC (montage of 3 photomicrographs) are markedly reduced and that the topographical distribution of the projection in the dLGN is also considerably altered. Part of the material shown is modified from a previous review
While the double fluorescent tracing method allows a direct qualitative assessment of the topographical distributions of RGC fibers from glaucomatous and control eyes, the use of different fluorescent filters precludes a precise quantitative comparison. For this reason we used optical density analysis of digitalized images from brain sections that were processed for DAB immunocytochemistry
(A) Retinal projections to the brain from a control eye illustrating the normal pattern of dense fiber innervation in the different visual structures. Several cases of experimental glaucoma are illustrated to show the variation of RGC fiber loss in the brain that are (B) reduced but conserve an even distribution of retinal fibers in each visual structure, (C) reduced and show an irregular patchy distribution or (D) are almost completely abolished except for a few small pathes of label. Only contralateral sides of the dLGN and SC are shown. SCN, suprachiasmatic nucleus; dLGN, dorsal lateral geniculate nucleus; vLGN) ventral lateral geniculate nucleus; IGL, intergeniculate leaflet; SC, superior colliculus.
Chronically elevated IOP alters these topographical distributions and causes a reduction of the RGC projections in brain structures, but these features vary depending on the individual case. Some laser operated animals showed a uniform reduction of retinal projections in all structures, although the fiber distribution appeared almost normal within the nuclei (
The quantitative analysis of the mean densities of retinal projections to different brain structures for operated and control groups and the relative percent reduction (compared to the average values for the controls) are illustrated in
Structures analyzed in control animals (n = 8, black) and operated animals (n = 6, white) include the suprachiasmatic nucleus (SCN), dorsal lateral geniculate nucleus (dLGN), ventral lateral geniculate nucleus (vLGN), intergeniculate leaflet (IGL), pretectum (PRT) and superior colliculus (SC). For each structure, the ipsi and contralateral densities are summed and represent the integral optical density (IOD) for all the sections obtained from the entire structure. The scale on the right corresponds to values for the total summed density of RGC projections for all structures). The difference in IOD was significant for all structures and for the total amount of projections. (Mann Whitney; * p<0.05). B shows the percent reduction of the IOD value of retinal fiber projections in visual structures of laser-operated eyes compared to the mean IOD value of the projections from unoperated control eyes. Percent reductions range from ∼50–70% depending on the structure. Errors bars in A and B are S.E.M.
We then assayed the ability of rats with binocular induced glaucoma to entrain their daily locomotor activity to successive 6 hr delays of LD cycles each coupled with 1 log unit decreases in light levels (100, 10 and 1 lux). Measures of IOP in both glaucomatous eyes showed increases comparable to those recorded for the monocularly operated animals, ranging from 16.1±1.14 mm Hg before surgery to 31.0±1.90 mm Hg following the surgical procedures. Locomotor activity rhythms (double plotted actograms) are shown for a control and operated rat with binocular experimental glaucoma (
Animals are first entrained under a 12:12 light:dark (LD) cycle at 100 lux light (actograms shown in A). After 26 days, the LD cycle was shifted 6 hrs (delay) and the light level was decreased to 10 lux. After 35 days, the light LD cycle was again shifted by 6 hrs and the light level decreased to 1 lux (45 days). Animals were then released into constant darkness to assess whether the animals were entrained to the previous LD cycle. The three successive 12L:12D light cycles (from 100-10-1 lux) are shown above the actograms and the days corresponding to the lux levels of the light phase are indicated on the right. The black bars indicate constant darkness. Although both groups of animals entrained to each of the shifted LD cycles, glaucomatous rats show a greater variability in locomotor activity onsets with respect to the beginning of the dark phase. Some glaucomatous rats also show variability in activity offsets at lights on and components of activity drift during the dark phase. This is illustrated in B for the phase angle plots of the activity onsets of individual control rats (n = 5, left) and rats with binocular glaucoma (n = 5, right). (C) Quantification of the number of days necessary to entrain to a new light-dark cycle (left) and quantification of the activity onset variability with respect to the beginning of the dark phase (right) for both groups. Onset variability was calculated from the last 15 days of each LD cycle, when all animals displayed stable entrainment. (t-test, * p<0.05; ** p< = 0.005).
This is illustrated more clearly in the group analysis in
At the end of the experiment, rats were released in constant darkness to assess whether the animals were entrained to the previous LD cycle at 1 lux light level or if there was a masking component (
Finally, real time quantitative PCR was used to evaluate the expression of different retinal opsins as well as the expression of a specific ganglion cell marker (Thy1) and pituitary adenylate cyclase peptide (PACAP) a neurotransmitter co-expressed in melanopsin RGCs (
Short wavelength cone (SW), mid-wavelength cone (MW) opsins, rhodopsin and melanopsin mRNA expression are all significantly reduced in experimental glaucoma. The specific RGC marker Thy1 was also significantly reduced whereas PACAP expression is unchanged (t-test, * p<0.05; ** p<0.005.) Both of the retinas from each of the binocularly operated (n = 5) and control rats (n = 5) were used.
Previous studies of animal models of experimental glaucoma secondary to chronically elevated IOP have demonstrated significant RGC loss ranging from 30% to 90% depending on the method employed, time course and experimental model
Anterograde tracing in the hypertensive rat model indicates that experimental glaucoma induces an overall reduction of the retinal projections to the brain (∼60%) while the same analysis following severe degeneration of outer retinal photoreceptors reveals no effects
A consistent feature of the reduction of RGC projections, despite the reliable increase of IOP induced by the laser surgery, was the variability observed between individual animals in both the extent and the topographical pattern of loss of RGC fibers. The variation of RGC, optic nerve fiber and visual field loss following increased IOP is a characteristic feature in both experimental and human glaucoma
The reduction in mRNA of melanopsin and of the ganglion cell marker Thy1 also confirms the overall decreases of the RGC population and of melanopsin RGCs. In contrast, PACAP, which is co-expressed in melanopsin RGCs
An unexpected but significant finding is that mRNA opsin expression from outer retinal photoreceptors (MW cones, SW cones, rods) were all found to be under expressed. Although the question of whether other retinal cell types are altered in glaucoma is still a matter of debate, Jakobs et al.
The reduction of melanopsin RGCs and their innervation of the SCN impacts on the ability of glaucomatous rats to entrain to light. We used an entrainment paradigm since this assay is reported to be more sensitive for detecting entrainment deficits compared to single light-pulse phase shifts in both animals
In humans, it has been reported that patients with different degrees of blindness resulting from various ocular pathologies show sleep disturbances and abnormal circadian rhythms
Here, we provide evidence that chronic increase IOP in a rodent model of glaucoma leads to a decrease in the melanopsin RGC axonal innervation of the SCN and an alteration in the photic response to light. The alteration of entrainment of locomotor activity in this model suggests that patients with severe bilateral glaucoma may show an increased propensity for chronobiological disturbances. Our data and that of previous studies support the idea that RGC loss due to glaucoma affects both visual and non-visual functions. Concerns for health and quality of life for patients with glaucoma should thus not be limited to the detection and prevention of visual impairments but should also consider the potential impacts on altered circadian entrainment and sleep disturbances.
We thank C. Gronfier for critical reading of the manuscript and W. De Vanssay for technical assistance.