Circadian IOP Rhythm in Rats Is Driven by Neural Signals From the Brain

Abstract

Purpose

The purpose of this study was to establish the origin of the circadian intraocular pressure (IOP) in the rat.

Methods

IOP was continuously monitored via wireless telemetry in adult Brown-Norway rats. After entrainment to a light/dark (LD) cycle, one cohort was transitioned to constant darkness (DD), and tetrodotoxin (TTX) was applied at different times and concentrations during subjective day (SD) and night (SN). Another cohort underwent superior cervical ganglionectomy (SCGx) and thereafter was maintained in LD. Effects of interventions were quantified via mean IOP and cosinor estimation of rhythm properties.

Results

All rats exhibited a pronounced circadian IOP rhythm that peaked during SN and persisted in DD. TTX had no effect during SD and reduced IOP during SN in a dose-dependent manner by up to 103% ± 15% of the nocturnal elevation. TTX had no effect on the IOP rhythm of the non-instilled contralateral eye. Rhythm baseline, peak-to-peak amplitude, period, and phase in DD averaged 9.5 ± 1.7 mm Hg, 4.9 ± 1.4 mm Hg, 24.0 ± 0.1 hour, and 2.3 ± 1.0 minute before TTX instillation and were unchanged afterward. IOP was not significantly different during SD and SN after SGCx, indicating complete knockout of the circadian IOP rhythm.

Conclusions

The IOP rhythm in rats is driven by sympathetic efferent signals from a central circadian clock. Ocular clocks and circulating humoral factors do not appear to contribute to rhythmogenesis. The neural pathway of circadian pressure control may be relevant to glaucoma pathophysiology.

Glaucoma is the leading cause of irreversible vision loss worldwide, and intraocular pressure (IOP) is a key parameter for detecting and managing disease onset and progression.^1–4 Changes in mean IOP garner most attention, but other aspects of IOP variation may contribute significantly to glaucoma pathophysiology as well. This is because IOP is highly dynamic, fluctuating over timescales of seconds to days as a result of various internal and external factors. Among the largest factors is the diurnal IOP rhythm that is expressed by many mammalian species,^5–10 including humans.^11–15 The rhythm causes IOP to peak during the night and trough during the day, even under constant illumination conditions.^7,16–19 This phase relation is conserved across species of differing locomotor posture^12,20,21 and daily activity pattern,^{5–7,16,22–33} indicating that the IOP rhythm reflects an endogenous circadian clock and not behavioral or environmental factors. In humans, the nocturnal IOP elevation is enhanced by recumbency during sleep^20,34,35 and rhythm disruption has been correlated with optic nerve damage, suggesting a possible role in glaucoma development.^11,36–41 It is thereby of paramount importance to understand the origin and function of circadian IOP rhythmicity.

A fundamental question is whether the IOP rhythm is driven by a central or peripheral clock. Circadian rhythms are synchronized throughout the body by a network of clock neurons in the hypothalamus of the brain that can entrain to ambient illumination cycles via visual input from the eye. The central clock sends circadian messages through neural and hormonal pathways that can impart rhythmicity to peripheral tissues or modify endogenous rhythms generated by local clocks in those tissues. Peripheral clocks have been implicated in liver metabolism,^42–45 adrenal glucocorticoid production,^46–48 and barrier functions of lung^49,50 and skin.^51,52 They have also been identified in several ocular tissues, including the retina,^53–57 cornea,^58–62 and iris-ciliary body.^31,63–66 The retinal clock is known to drive a melatonin rhythm that modulates retinal gene expression, circuit connectivity, and light sensitivity during the day and night.^67–71 None of the ocular clocks have been shown to drive the IOP rhythm as yet.^11,30,31

The mechanisms of IOP rhythmogenesis may differ across animals. In mice, peripheral clocks in the iris-ciliary body complex can influence the phase and amplitude of the IOP rhythm but cannot support the rhythm.^63,72 Non-local neural and hormonal pathways are required that involve circadian sympathetic nerve transmission and glucocorticoid/corticosterone circulation to mice eyes.^29–31,33 In rabbits, on the other hand, sympathetic nerve signals from the central clock increase IOP at night^8,73,74 and another unknown process lowers IOP during the day.^8,25 The objective of this work was to elucidate the origin of circadian IOP rhythmicity in rats. Rats are a popular animal model for glaucoma research, along with mice and rabbits, and they may provide valuable insights about circadian clock contributions to the disease in humans.

Methods

All experiments were conducted on healthy adult Brown-Norway rats (male, retired breeder, 300–400g, >12 months) in compliance with the National Institutes of Health guide for the care and use of laboratory animals, ARVO Statement for the use of animals in ophthalmic and vision research, and protocols approved by the Institutional Animal Care and Use Committee at the University of South Florida. Animals were housed before experimentation under a 12-hour light (6 AM–6 PM)/12-hour dark (6 PM–6 AM) cycle in a humidity- and temperature-controlled room with food and water freely available. Animals were then outfitted with an in-house wireless telemetry system that continuously recorded IOP round-the-clock at 0.25 Hz. Details of the IOP telemetry system have been published.^28,75 Briefly, a drug-loaded silicone microcannula was surgically implanted in the anterior chamber of one or both eyes, secured to the sclera with sutures, and routed subdermally to a skull-mounted coupler that was connected fluidically to a pressure sensor worn on the animal's back. After cannula surgery, animals were returned to housing for at least 3 days to record baseline IOP rhythmicity under the ambient light/dark (LD) cycle and then subjected to testing.

Neurotoxin Instillation

A cohort of unilaterally-canulated animals (n = 8) and a cohort of bilaterally-cannulated animals (n = 4) animals were transferred to an environmental control unit (ECU, BIO-C36; Tecniplast, Buguggiate, Italy) that allowed for manipulation of ambient lighting. After three days of LD entrainment, animals were placed in constant darkness (DD) to eliminate any direct effects of light on IOP. A 10 µL drop of saline solution or tetrodotoxin (TTX), a voltage-gated sodium channel blocker, was then instilled on the cannulated eye at different times of day and concentrations (0.001, 0.1, 0.5, 1 mg/mL). The drop was applied to only one eye of bilaterally-cannulated animals. Instillations were performed under dim red light and brief (<3 minutes) 2% isoflurane at least 24 hours apart. Each was administered at a specific phase of the free-running IOP rhythm in DD, with trough and peak phases corresponding in circadian time (CT) to subjective day (SD) and subjective night (SN) when light would be, respectively, on and off. SD thereby spanned from CT0–12 and SN from CT12–24.

Superior Cervical Ganglionectomy

A separate cohort of unilaterally-cannulated animals (n = 3) underwent a bilateral superior cervical ganglionectomy (SCGx). The animals were administered extended-release analgesics and anesthetized via intraperitoneal injection of ketamine (1 mg/mL) and xylazine (0.1 mg/mL). The nape of the neck was shaved and disinfected, and a vertical incision (∼25 mm) made to expose the mandibular glands. The glands were slid aside to access the carotid triangle, and a 4–0 suture was fed under the carotid arteries to gently retract the vessels without completely obstructing blood flow. The underlying superior cervical ganglion (SCG) was removed by transecting the sympathetic trunk and carotid nerves alongside the ganglion. The procedure was repeated on the contralateral SCG, the neck incision was closed with resorbable sutures, and topical antibiotics were applied for two to three days. SCGx success was confirmed qualitatively by monitoring eyelid ptosis,⁷⁶ and animals that did not exhibit signs of SCGx success were omitted from the study. Following surgery, animals were returned to housing and IOP was monitored under the ambient LD cycle.

Data Analysis

IOP data were processed in MATLAB (The Mathworks, Natick, MA, USA) using median and mean filters of 28-second width to remove outliers and smooth records. The IOP rhythm in DD can drift out of synch with the external LD cycle so the amplitude, phase, and period of the free-running rhythm and TTX-induced changes in these parameters were estimated by cosinor analysis.¹⁶ Specifically, IOP records of at least 36 hours’ duration before and after TTX instillation were, respectively, fit to a single cosinor function:

$$\begin{array}{c}\hfill P\left(t\right)=\overline{P}+0.5\tilde{P}(1+cos\left(2\pi \left(t+\theta \right)/T\right)),\end{array}$$

where $\overline{P}$ is baseline IOP during SD, $\tilde{P}$ is peak-to-trough rhythm amplitude, T is rhythm period, and θ is rhythm phase with respect to midnight. The magnitude and duration of TTX effects were, respectively, quantified by comparing mean IOP for the two-hour interval immediately before and after instillation and the time from instillation until IOP recovery to pre-instillation level. TTX concentration effects were quantified by fitting IOP data over a 24-hour period with a difference-of-pulses function given by:

$$\begin{array}{c}\hfill P\left(t\right)=\overline{P}+P_1·rect\left(t;T_1,{\theta }_1\right)-P_2·rect\left(t;T_2,{\theta }_2\right),\end{array}$$

where $\overline{P}$ is baseline IOP during SD, P₁ and P₂ are the pulse amplitudes, T₁ and T₂ are the pulse durations, and θ₁ and θ₂ are pulse phases with respect to instillation time. Fits were obtained by nonlinear regression using criteria of T₁ > T₂ and θ₂ > 0 > θ₁. SCGx effects were quantified by comparing mean IOP over a four-hour interval during SD (CT4-CT8) and SN (CT16-CT20) across three days before and after surgery. Statistical analyses were performed with SigmaPlot software (Systat, San Jose, CA, USA) using paired Student's t-test or two-way ANOVA, with Holm-Sidak method for multiple comparisons on normally-distributed data and Mann-Whitney rank-sum test on non-normal data. Results are reported as mean ± standard deviation for normal data and median [lower quartile, upper quartile] for non-normal data, with significance defined as P < 0.05.

Results

Data were collected from a total of 14 rats. All animals exhibited a pronounced IOP rhythm that was synchronized to the ambient LD cycle. Figure 1A shows that the rhythm persisted in DD as well. To determine whether the IOP rhythm depends on action potentials, TTX was instilled at various times of day and night to block voltage-gated sodium channels. These channels are only expressed in nerve fibers and not any other anterior segment tissues to our knowledge. Figure 1B shows two-day segments of an IOP recording from an animal in DD during which TTX (1 mg/mL) was instilled on different days at CT3, CT11, CT15, and CT19, which corresponded to approximately 9 AM, 5 PM, 9 PM, and 1 AM on a 24-hour clock. TTX had no effect during SD, but the pre-dusk (CT11) instillation delayed the onset of the nocturnal IOP elevation by a few hours compared to the previous night. In contrast, TTX had a marked effect during SN, reducing nocturnal IOP to daytime levels for several hours when instilled at CT15 and for the remainder of the night when instilled at CT19. Figure 1C shows that the nocturnal IOP reduction was specific to TTX, because saline solution instillation during SD had no effect on nocturnal IOP (pre-saline: 21.7 ± 7.0 mm Hg, post-saline: 21.5 ± 5.7 mm Hg, n = 4, P = 0.82). Figure 1D summarizes the effect of TTX on IOP across instillation times and animals (n = 4 per condition). TTX did not noticeably alter IOP at CT3 (ΔIOP = 1.2 ± 1.4 mm Hg, P = 0.45) or CT7 (ΔIOP = 0.5 ± 1.2 mm Hg, P = 0.76), but it did significantly reduce IOP at CT11 (ΔIOP = −3.5 ± 2.4 mm Hg, P = 0.03), CT15 (ΔIOP = −8.9 ± 2.4 mm Hg, P < 0.001), CT19 (−9.0 ± 4.4 mm Hg, P < 0.001), and CT23 (−7.5 ± 3.8 mm Hg, P < 0.001). The time dependence suggests that IOP regulation during SD involve mechanisms not susceptible to sodium-channel blockade and that neural messages from the circadian clock elevate IOP during SN above a baseline daytime level.

Figure 1.

Time dependence of IOP responses to TTX. (A) Representative IOP data from a rat entrained to LD and transitioned to DD. Black and white bars indicate light and dark phases of ECU illumination, and gray shading indicates SN of the animal. (B) IOP responses to TTX instillation at four different times of the circadian cycle (CT3, CT11, CT15, and CT19). Symbols indicate instillation time. (C) IOP response to saline solution instillation at CT15. Symbol indicates instillation time. (D) Summary of IOP data pre- and post- TTX (white and gray columns) for different instillation times. Connected symbols indicate measurements from individual animals (n = 4 per condition). Error bars give standard deviation. Asterisks indicate significance level: *P < 0.05; **P < 0.01.

The dose-response relationship was examined to determine whether TTX could reduce nighttime IOP partway to daytime levels. Figure 2A shows IOP responses to different concentrations instilled during SN. TTX effects were graded over ∼2 log units in concentration and quantified by fitting responses with a sum of two pulses to model the nocturnal elevation and TTX-induced reduction of IOP. Figure 2B summarizes the dose-response relationship across animals (n = 3). TTX drops of 0.01, 0.1, and 1 mg/mL lowered IOP by 38% ± 17%, 77% ± 13%, and 103% ± 15% of the nocturnal elevation, respectively. The graded changes were significant (0.01:0.1 mg/mL, P = 0.05; 0.01:1 mg/mL, P = 0.03; 0.1:1 mg/mL, P = 0.02), implying that progressively higher dosages blocked an incremental percentage of circadian clock input to the anterior segment of the eye.

Figure 2.

Dose-dependence of IOP response to TTX. (A) IOP records from a rat in SD during which TTX was applied at 1, 0.1, and 0.01 mg/mL (gray traces). Symbols indicate instillation time. Black trace gives the fit of a difference-of-pulses function to the raw data. (B) Summary of TTX dosage effect across animals. IOP reduction is expressed as a percent of the peak-to-peak rhythm amplitude. Connected symbols indicate measurements from individual animals (n = 3). Asterisks indicate significance level: *P < 0.05.

The time course of TTX effects was assessed in greater detail. Figure 3A shows an IOP record of another animal during which TTX (1 mg/mL) was instilled midway through SN. The times to half-maximum IOP reduction and half-return to nighttime IOP were measured and respectively averaged 12.7 ± 7.3 and 253.1 ± 96.2 minutes across animals (n = 6). It was noted that the IOP response to TTX is phenomenologically similar to its response to subjective dawn, so the response dynamics were compared. Figure 3B shows that the time to half-max-effect of TTX was indistinguishable from the time to half-return of the IOP rhythm to daytime levels (12.7 ± 7.3 min, P = 0.90). The equivalent dynamics suggests that circadian signals are rapidly extinguished by TTX and rhythm offset and that the time course mainly reflects slow readjustment of aqueous inflow and outflow mechanisms to the cessation of clock input.

Figure 3.

Dynamics of IOP response to TTX. (A) A 24-hour IOP record from a rat in DD during which TTX was applied midway through SN (black trace). Symbol indicates instillation time. Red trace and dots give a linear fit of TTX onset and offset responses and the times to max-effect (left) and half-recovery (right), respectively. (B) Summary of onset dynamics of IOP lowering by TTX and offset dynamics of the IOP rhythm reduction at subjective dawn (DL). Connected symbols indicate measurements from individual animals (n = 6). n.s., not significant.

Measured response durations allow for possible spread of TTX beyond the instillation site via ocular diffusion and systemic circulation. To further assess whether the clock controlling IOP rhythmicity resides inside or outside the eye, rhythm properties were examined on subsequent days to determine if TTX not only arrested clock output but also interrupted clock timekeeping. Figure 4A shows a four-day IOP record from an animal in DD that was instilled with TTX during the second night. A cosine waveform was fit to data collected pre- and post-instillation, and Figure 4B summarizes TTX effects on IOP rhythm properties across animals (n = 8). No change was detected in rhythm amplitude (Δ$\tilde{P}$ = −0.3 ± 0.9 mm Hg, P = 0.39), period (ΔT = −0.1 ± 0.5 hour, P = 0.38), phase (Δθ = 0.0 ± 1.4 minutes, P = 0.98), or baseline daytime level (Δ$\overline{P}$ = 0.3 ± 1.0 mm Hg, P = 0.39), indicating that TTX did not alter the free-running IOP rhythm. This suggests the clock resides in the brain and modulates IOP via efferent nerve impulses transmitted to the eye. The rhythm could still originate within the eye, though, if the clock does not use voltage-gated sodium channels.

Figure 4.

Long-term effects of TTX on IOP rhythmicity. (A) Four-day IOP record from a rat in DD during which TTX was applied during one SD. Symbol indicates instillation time. Horizontal lines demark data segments used for cosinor analysis to quantify rhythm properties before and after instillation. (B) Summary of rhythm amplitude, period, phase, and baseline level before and after TTX instillation (white and gray columns). Connected symbols indicate measurements from individual animals (n = 8). n.s., not significant.

TTX is a potent neurotoxin that can have near-immediate secondary effects ranging from paresthesia to respiratory failure and cardiac dysrhythmia if absorbed systemically.^77–80 In some experiments, IOP was concurrently recorded from both eyes to assess whether the observed IOP response was the result of systemic TTX effect. Figure 5A shows a four-day segment of paired IOP recordings from an animal in DD. It can be seen that the free-running rhythm was identical in both eyes, as were many slower IOP fluctuations. TTX was instilled the following night in one eye, and Figure 5B shows that IOP in that eye decreased to the daytime level for a few hours whereas IOP in the non-instilled contralateral eye stayed at its nocturnal level. Figure 5C summarizes the effect of unilateral TTX instillation across experiments (n = 6). TTX consistently reduced nocturnal IOP of the instilled eye (ΔIOP = −4.2 ± 2.0 mm Hg, P < 0.01) but had no significant impact on the non-instilled eye (ΔIOP = −2.1 ± 2.6 mm Hg, P = 0.11), confirming that its effect on IOP rhythmicity reflects local non-systemic blockade of circadian nerve signals to the eye.

Figure 5.

Effect of TTX on IOP of bilaterally cannulated animals. (A) IOP traces collected concurrently for several DD days from the OD (black) and OS (red) eye of the same rat. (B) The 24-hour IOP records recorded concurrently from both while TTX was applied only to the OD eye (black). Symbol indicates instillation time. (C) Summary of IOP data before and after (white and gray bars) TTX instillation on the treated (OD + TTX) eye and contralateral control (OS) eye. Connected symbols indicate measurements from individual animals. Asterisks indicate significance level: **P < 0.01; n.s., not significant.

In other experiments, the eye was neurally disconnected from the master circadian clock by severing sympathetic input via bilateral removal of the superior cervical ganglion. The SCG is the bulbous structure positioned behind the carotid bifurcation in Figure 6A. IOP was recorded after SCGx for 6.4 ± 3.8 days (n = 3), and Figure 6B shows a six-day record under LD from one such animal. The rhythmic elevation of nighttime IOP was abolished after SCGx, and IOP remained thereafter at daytime levels. Figure 6C summarizes the effect on IOP rhythmicity across SGCx experiments. No difference in mean IOP was detected during SD and SN after SCGx (ΔIOP = 0.5 ± 0.3 mm Hg, P = 0.28), further demonstrating that the IOP rhythm does not originate in the eye but from sympathetic signals generated by the master clock in the brain.

Figure 6.

Effect of superior cervical ganglionectomy on IOP rhythmicity. (A) Image of the SCG after retracting carotid vessels. Dashed lines mark incision sites in the sympathetic trunk (STx) and internal- (ICNx) and external- (ECNx) carotid nerves. (B) IOP record from a rat in LD during which SCGx was performed to sever sympathetic input to the eye. Symbol indicates SCGx time. (C) Comparison of IOPs during SD and SN (white and gray columns) after SCGx. Connected symbols indicate measurements from individual animals (n = 5). n.s., not significant.

Discussion

In this study, the circadian IOP rhythm of rats was disrupted pharmacologically and surgically and the effects on rhythmicity were captured with unprecedented temporal detail in ambulatory animals. As prior tonometry- and telemetry-based studies in rats have shown,^7,16–19,75 IOP exhibits a pronounced diurnal rhythm that peaks during the nighttime phase and troughs during the daytime phase of the circadian cycle. The rhythm continues unabated in DD, indicating that it arises from an endogenous circadian clock. Our results show that this clock acts solely at night to raise IOP, because blocking voltage-gated channels with TTX during SN transiently lowered nocturnal IOP to daytime levels yet had no impact on IOP during SD. The clock appears to reside within the brain of rats because IOP rhythmicity was abolished after silencing sympathetic input to the eye via SCGx. Rhythmicity should have persisted if it were intrinsic to the eye or other locus. In addition, TTX did not change the amplitude, frequency, or phase of the IOP rhythm. If the clock were located in the eye, rhythm properties should have been altered on subsequent days by the transient disruption of circadian circuit operation. A shift in rhythm phase might specifically be expected based on the similar IOP-lowering effect of light pulses during SN.¹⁶ This further implies that TTX did not get transported by systemic circulation to the master clock in the brain, which was confirmed by the maintenance of IOP rhythmicity in non-instilled eyes. Instead, TTX must have blocked clock output along the way to the instilled eye. The presumptive blockage site is the ocular surface, where sympathetic nerve fibers project into the anterior segment, because the magnitude and speed of pharmacologic effects favor local diffusion over systemic transport and mirror the dynamics of IOP rhythm offset at subjective dawn.

IOP Rhythmicity in Other Animals

Our results regarding mechanisms of circadian IOP rhythmicity in rats echo research findings in other animals, with some species-specific nuances. In rabbits, the master clock in the suprachiasmatic nucleus of the brain also drives nocturnal IOP elevations.⁸¹ It communicates circadian messages to the eye via sympathetic nerve fibers as SCGx reduces IOP elevations during SN,^8,73 electrical stimulation of SCG increases IOP and aqueous norepinephrine level during SD,⁸² and adrenergic receptor blockers alter aqueous production and outflow resistance.⁸³ Unlike rats, SCGx did not fully eliminate the IOP rhythm. IOP was reduced during the light phase relative to the mean daily IOP as well, so it has been suggested that a separate non-sympathetic mechanism depresses IOP during SD in rabbits.⁸ Such a mechanism does not apparently exist in rats because TTX lowered nocturnal IOP to the daytime level at most dosages. Hence, their baseline IOP innately rests at the daytime level and not the mean daily level like rabbits. The regulation of IOP rhythmicity is more complex in mice. In addition to adrenergic signals from a central clock, there is a significant endocrine component from adrenal glucocorticoids.^30,33 The two pathways appear to operate in concert to regulate IOP because adrenalectomy and SCGx are both needed to eliminate rhythmicity. SCGx alone had minor effect on the IOP rhythm of mice,³⁰ in sharp contrast to rabbits and rats. Our results offer little support for a glucocorticoid rhythm in rats because SCGx completely abolished IOP rhythmicity and TTX transiently and completely and unilaterally suppressed IOP. Several clock genes have also been found in anterior segment tissues of mice, with rhythmic expression levels that correlate to diurnal IOP variations.⁶³ These ocular clocks cannot drive IOP rhythmicity because they do not entrain to LD cycles³¹ or alter IOP when genetically knocked out,³⁰ but could perhaps modify how aqueous inflow and outflow tissues respond to central clock cues. Because mice, rabbits, and rats all exhibit an IOP rhythm that is controlled—in whole or part—by neural feedback to the eye from a circadian clock in the brain, it is likely that circadian neuroregulation plays a role in the human IOP rhythm as well. Further research in these animals could thus yield important insights about clock control of aqueous humor dynamics and its potential contributions to glaucoma.

Alternative Possible Interpretations

Several alternative explanations for our results were considered. One is the potential impact of pupil changes given that the iris is under sympathetic nerve control. A recent study reported that pupil dilation crowds the iridocorneal angle, increasing outflow resistance and elevating IOP.⁸⁴ Because the pupil would be dilated in darkness, it might be argued that TTX lowered IOP by relaxing the dilator muscle. However, experiments were performed in DD, so this would not explain the lack of TTX effect on IOP during SD. In addition, the IOP elevation induced by pupil dilation is likely insufficient to explain rhythm amplitude. No study has yet quantified the dependence of rat IOP on pupil size, but pharmacologic dilation of the rabbit pupil elevated IOP up to 5 mm Hg⁸⁵ whereas rhythm amplitude ranges up to 17 mm Hg.²⁷ Another is the potential impact of bilateral SCGx on the pineal gland. Loss of sympathetic innervation to pinealocytes is known to arrest the systemic melatonin rhythm that synchronizes circadian clocks throughout the body,^86–89 so the loss might disrupt IOP rhythmogenesis if melatonin is involved. However, melatonin rises at night and has been reported in rabbits to lower IOP.^90,91 Loss of melatonin rhythmicity after SCGx should thereby have not just abolished the IOP rhythm but also chronically raised IOP above the daytime level, which was not observed in rats. TTX also temporarily mimicked SCGx effects on IOP, presumably without interfering in pineal melatonin production given that IOP rhythmicity was unaltered in non-instilled eyes. Hence, our results do not offer much support for a role of melatonin in IOP rhythmogenesis. Finally, the potential impact of TTX on corneal sensory nerves and retinal ganglion cells was considered. Corneal nerve endings function as mechanonociceptors, sensing tissue stretch from external and internal forces such as IOP,^92–94 so loss of this mechanosensory information could upset central mechanisms of IOP regulation.^95,96 These mechanisms would presumably increase IOP to correct the loss and maintain homeostasis, but TTX caused IOP to decrease. TTX could have diffused to the retina and upset visual information transmission as well. In this case, the loss would strip the central clock of light input. Because experiments were performed in DD, TTX should not have any effect on IOP rhythmicity. These alternative explanations of our results were therefore rejected.

Study Limitations

Several limitations of this study should be acknowledged. First, although IOP rhythmogenesis and central control are unlikely to depend on phenotypic traits, rhythm amplitude, phase, and other properties could vary with ocular health, sex, age, or species because all experiments were performed on aged male rats. Second, ocular clock gene expression was not assessed under control or treatment conditions. As such, the persistence of IOP rhythmicity and the absence of phase or period shifts after TTX instillation may argue against a peripheral oscillator, but the possibility that these interventions altered clock activity in the eye cannot be excluded. Third, TTX and SCGx experiments implicate sympathetic nerve signaling in the rhythmic elevation of nocturnal IOP, but they do not resolve which adrenergic receptors or mechanisms of aqueous humor dynamics are involved. Pharmacologic studies targeting alpha- and beta-receptor subtypes while measuring aqueous production, trabecular, and uveoscleral outflows and episcleral venous pressure will be needed to understand how circadian signals regulate rat IOP. Last, the precise targets of sympathetic fibers within the anterior segment of rats were not identified. Future work incorporating anatomical maps and molecular assays will be essential to link neural input to specific aqueous humor dynamics pathways.