Expression of Adenylate Cyclase Subtypes II and IV in the Human Outflow Pathway

Abstract

purpose. It has been demonstrated that low doses of pilocarpine and other muscarinics substantially increase outflow facility in the isolated human outflow system devoid of ciliary muscle. These cholinergic-induced facility responses were thought possibly to be due to elevation of cAMP as a result of the presence of adenylate cyclases II (AC-II) and IV (AC-IV). Therefore, whether these isoforms are present in outflow tissues was examined.

methods. Human anterior segments were perfused with carbachol (10⁻⁹–10⁻⁵ M), and outflow facility and cAMP levels in the perfusate were measured simultaneously. Isolated trabecular meshwork (TM) were incubated with carbachol (10⁻⁷ M), and the subsequent changes in cAMP were measured by radioimmunoassay. AC-II and AC-IV were characterized in ocular tissue with reverse transcription–polymerase chain reaction and in situ hybridization.

results. Outflow facility increased, in a dose–dependent manner, by 10%, 16%, and 27% in response to 10⁻⁹, 10⁻⁷, and 10⁻⁵ M carbachol, respectively. Similarly, cAMP increased by 9%, 70%, and 210% in response to 10⁻⁹, 10⁻⁷, and 10⁻⁵ M carbachol, respectively. In addition, cAMP levels significantly increased by 39% in isolated TM strips incubated with 10⁻⁷ M carbachol. AC-II was detected in most normal tissue examined, but not in any cultured cell lines or any glaucomatous tissue. AC-IV was also widely expressed in most normal tissues, faintly detected in some glaucoma tissue, but not detected in most cultured cells.

conclusions. The presence of AC-II and AC-IV in outflow tissues supports the hypothesis that cholinergics may indeed exert an effect on outflow facility, mediated by cAMP, which is independent of muscle contraction.

It has been known for some time that lipid soluble analogues of cAMP and agents that stimulate adenylate cyclase can increase outflow facility in human ^{1 2} and subhuman primate ^{3 4 5 6 7 8 9 10 11} eyes. Although the involvement of adenylate cyclase seems clear, the types of cyclases present are unknown, as are the downstream consequences of cyclase stimulation that ultimately result in the facility effect. 

Recent studies have shown that there are at least nine genes coding for distinctive adenylate cyclases. ^{12 13} It has been known for a number of years that the regulation of adenylate cyclase activity is via stimulatory or inhibitory alpha subunits of G proteins. Recently, it has been demonstrated that regulation by other factors, especially Ca²⁺, may be as important. ¹⁴ Accumulating data show that most cyclases, if not all, are multiply regulated, which may explain many previously conflicting results. Specifically, it is now known that besides the classically described regulation by G_α subunits, protein kinase C, Ca²⁺, and βγ subunits of G proteins all can regulate cyclase activity more effectively than G_α subunits. ^{15 16} Interestingly, stimulation of M1 and M3 receptors can lead to increased cAMP via activation of the βγ subunits of cyclases II and IV (which are stimulated by the βγ–G protein subunit and are calmodulin-insensitive). ¹⁷ The cyclases and their regulators can be classified into three groups as described by Cooper et al. ¹⁴ : group 1, stimulated by calcium (AC-I, -III, -VIII); group 2, nonstimulated by calcium (AC-II, -IV, -VII); and group 3, inhibited by calcium (AC-V, -VI). In situ hybridization has shown that the individual cyclases have discrete distributions in the central nervous system; certain isoforms are expressed uniquely in certain areas of the brain. ^{18 19 20 21 22 23} This differential distribution apparently serves to modulate differential signal transmissions. AC-II is detected by Northern blot analysis in rat brain, olfactory epithelium, olfactory bud, and lung but not in kidney, liver, intestine, or heart. ²³ AC-IV appears to be widely detected by polymerase chain reaction analysis in rat brain, heart, intestine, kidney, liver and lung, but not in testis. ²⁴ In addition, AC-II and AC-IV are both detected in increasing amounts by Northern blot analysis in human and rat myometrium during pregnancy. ²⁵ Knowledge of the types and distribution of adenylate cyclase isoforms present in the outflow apparatus will lend insight into the ultimate mechanisms responsible for outflow facility increases mediated by agents modulating cyclase activity. In the present study, we characterized the presence and distribution of AC-II and AC-IV in the human outflow pathway. 

Methods

All chemicals, tissue culture media, and supplements were obtained from Sigma Chemical (St. Louis, MO) unless otherwise noted. All human donor eyes were obtained from the National Disease Research Interchange (Philadelphia, PA). 

Human Tissue and Cell Cultures

Three pairs of normal human eyes (average age ± SEM, 76.33 ± 3.18 years), and four eyes from individuals with a documented history of glaucoma (average age ± SEM, 74.25 ± 2.69 years) were embedded in paraffin and sectioned (6-μm-thick) according to standard procedures. ²⁶ Four pairs of normal human eyes (average age ± SEM, 67.75 ± 6.73 years) were embedded in OCT compound (TissueTek, Torrance, CA), frozen, and sectioned (6-μm-thick) onto glass slides (Superfrost plus, Fisher Scientific, Pittsburgh, PA). Tissue pooled from six pairs of normal human eyes (average age ± SEM, 74.50 ± 2.22 years), and four pairs of glaucomatous eyes (average age ± SEM, 68.75 ± 1.55 years) were used for total RNA extraction. 

Human cell lines derived from ciliary muscle (H7CM; 1-day-old infant), trabecular meshwork (H4TM; 17-year-old), corneal endothelium (H4 cornea; 17-year-old), and anterior sclera (H312 sclera; 2-year-old) were used. H7CM cells (passage 9) were grown according to methods previously reported. ²⁷ H4 cornea cells (passage 4) and H312 scleral cells (passage 4) were grown in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum and 2 mM glutamine until confluent. 

Stimulation with Carbachol

Normal human donor eyes (average age ± SEM, 81.80 ± 3.51 years) were used in these studies. A detailed description of the tissue dissection, the perfusion apparatus, and the perfusion method can be found in previous articles. ^{28 29} In the present experiment, DMEM (control) and carbachol (10⁻⁹–10⁻⁵ M), a cholinergic agonist, were perfused into the anterior segment. The perfusate that exited the perfused eye was collected every 15 minutes, and cAMP was measured as described previously. ³⁰  

Stimulation of cAMP in isolated trabecular meshwork (TM) was performed according to a method previously reported. ³⁰ Normal human donor eyes (average age ± SEM, 76.33 ± 7.86 years) were dissected, and 1- to 4-mm strips of TM were separated and divided equally into three vials. The vials were then incubated with 300 μl DMEM (control media), 300 μl of 100 μM of forskolin in DMEM (passivity control), or 300 μl of 10⁻⁷ M carbachol in DMEM at 37°C, 5% CO_2, for 30 minutes. After incubation, 30 μl of 750 mM sodium acetate was added to each tube, and the samples were homogenized, boiled for 4 minutes, and centrifuged at 1200 rpm and the supernatant was then frozen. cAMP in the supernatant was measured by radioimmunoassay (RIA kit; Biomedical Technologies, Stoughton, MA). Each pellet was solubilized in 1N NaOH, and protein was determined by the BCA Protein Assay Kit (Pierce, Rockford, IL). Data were analyzed for statistical significance by a Student’s t-test. 

Reverse Transcription–Polymerase Chain Reaction and Southern Hybridization

Total RNAs from human cultured cells and tissue were prepared according to methods previously reported. ^{27 31} Primers were designed by Primer Premier 4.04 program (Biosoft International, Palo Alto, CA) according to the sequence in GenBank (No. M80550 for AC-II ²³ and M80633 for AC-IV ²⁴ ). Details of the primers are specified in Table 1 . The primers of glyceraldehyde-3-phosphate dehydrogenase (G3PDH), a housekeeping gene, were a kind gift from Shuhua Nong (Boston Medical Center, Boston University School of Medicine, Boston, MA). Reverse transcription–polymerase chain reaction (RT–PCR) was performed according to the instructions of RT and PCR Systems (Promega, Madison, WI) with the exception of annealing at 50°C for AC-II primers and 57°C for AC-IV primers. Rat brain, which is known to express AC-II and AC-IV, ^{23 24} was used as a positive control. The PCR mixture without template was used as a negative control. Amplification of the first strain of all samples with G3PDH primers was performed as an internal control. All RT–PCR products were electrophoresed onto a 2.5% Amplisize agarose gel (Bio-Rad, Hercules, CA), photographed, and then transferred to nylon membranes (CUNO, Meriden, CT). A Southern blot assay was then performed to confirm the expression of transcripts for AC-II and AC-IV. ^{32 33} The cDNA probes encoding AC-II and AC-IV subtypes were kind gifts from Alfred G. Gilman (University of Texas, Dallas, TX). The type II cyclase cDNA (4123 bp) was in pBS KS(+)-AC2 and the type IV cyclase cDNA (3665 bp) was in pBS SK(+)-AC4. ^{23 24} To prepare the probes for Southern hybridization, cDNAs of AC-II were excised from the plasmid, pBS KS(+)-AC2, by digestion with HindIII and BamHI, and cDNAs of AC-IV were excised from the plasmid, pBS SK(+)-AC4, by digestion with BamHI and KpnI. The cDNAs were then labeled with [α–³²P] dCTP by a random priming labeling system according to the instruction manual (New England Nuclear/DuPont, Boston, MA). The specific activity of ³²P–DNA probes was 4 to 6 × 10⁸ counts per minute/μg DNA. The PCR products on the nylon membranes were hybridized with[ ³²P]-labeled cDNA probes for AC-II and AC-IV according to standard instructions (Life Science, Meriden, CT) as previously reported. ^{32 33} Identification of the RT–PCR products was verified by their size and Southern hybridization assay. 

In Situ Hybridization

The antisense probes of AC-II and AC-IV were transcribed with T7 RNA polymerase on a HindII linearized pBS KS(+)-AC2 or a BamHI linearized pBS SK(+)-AC4 DNA template, respectively. The sense probes of AC-II and AC-IV were transcribed with T3 RNA polymerase on a BamHI linearized pBS KS(+)-AC2 or a KpnI linearized pBS SK(+)-AC4 DNA template, respectively. The RNA transcripts were labeled with DIG-11-uridine triphosphate, sized to approximately 200 bp, and purified according to the DIG RNA Labeling Kit instructions (Boehringer Mannheim, Indianapolis, IN). In situ hybridization of human eye sections with the probes was performed according to the DIG Nucleic Acid Detection Kit instructions (Boehringer Mannheim) and the methods described previously. ^{27 34} The slides were scanned by a Spot Digital Camera (Diagnostic Instruments, Sterling Heights, MI). All the experiments described above were repeated at least twice. 

Results

Effect of Carbachol on Outflow Facility and cAMP

Perfusion of human anterior segments with carbachol resulted in an increase in outflow facility in a dose–dependent manner (Table 2) . cAMP accumulation in the perfusate also increased in a dose–dependent manner after incubation with carbachol; however, the percent increase in cAMP levels was more variable than the outflow effect (Fig. 1) . The only exception to this was eye 365, which received 10⁻⁵ M carbachol and had no change in outflow facility (whereas cAMP decreased). Incubation of TM strips with carbachol resulted in a similar, and significant, rise in cAMP (Table 3) as determined by a Student’s t-test. 

RT–PCR and Southern Blot Analysis

In our initial experiments, we amplified AC-II and AC-IV mRNA with gene–specific primers. RT–PCR results showed mRNA expression for AC-II (380 bp) in all normal human tissue tested (TM, ciliary muscle[ CM], ciliary processes [CP], corneal endothelium [CE], corneal stroma [CS], anterior sclera [AS], nonpigmented epithelium [NPE], and pigmented epithelium [PE]) but not in any of the cultured cells (TM, CM, CE, or AS cells; Fig. 2A , row a; Table 4 ). AC-II was not detected in any glaucomatous tissue (Fig. 2B , row a; Table 4 ). mRNA expression for AC-IV (285 bp) was found, as expected, in the normal human TM, CM, CP, AS, NPE, and PE tissues and cultured AS cells but not in CE, CS tissue, or cultured TM, CM, or CE cells (Fig. 2A , row c; Table 4 ). However, AC-IV could be detected faintly in the glaucomatous TM, CP, and PE tissue (Fig. 2B , row c; Table 4 ). No signals were found in the negative control, and no genomic DNA contamination was found in the RNA preparation (data not shown). AC-II and AC-IV were detected markedly in rat brain, a positive control (Figs. 2A and 2B , rows a and c in both; Table 4 ). As expected, PCR products in the ethidium bromide–stained agarose gel showed bands at 983 bp for G3PDH in all samples (Figs. 2A and 2B , row e in both). 

Confirmation of the expression transcripts for AC-II and AC-IV was obtained by Southern blot analysis of the PCR products hybridized with a [³²P]-labeled cDNA probe for AC-II and AC-IV (Figs. 2A and 2B , rows b and d in both; Table 4 ). All Southern hybridization results correspond to the RT–PCR results in both patterns and intensities (Fig. 2 ; Table 4 ). 

In Situ Hybridization

Normal tissue sections were hybridized with a DIG-labeled RNA probe to determine the distribution of AC-II and AC-IV mRNA. In situ hybridization detected both AC-II and AC-IV in normal human TM, CM, CP, AS, PE, and the outer wall of Schlemm’s canal (SC-O) as indicated by the dark blue staining around the nuclear area (Figs. 3 4 ; Table 5 ). Signals for AC-II were also detected in CE, CS, and NPE (Fig. 3 ; Table 5 ). Both signals were only detected in the circular portion of CM, not in the longitudinal portion (Figs. 3A 4A) . AC-II and AC-IV mRNA were not detected in the inner wall of Schlemm’s canal (SC-I), but staining was found in SC-O (Figs. 3C 4C ; Table 5 ). The distribution of AC-II and AC-IV appeared differently among the various tissues. Both signals seemed to be stronger in the circular portion of CM and CP than in the TM, CS, AS, and SC-O. The signals were the weakest in the CE, NPE, and the PE of CP. 

Discussion

Our results, which are similar to those of a previous study, ³⁵ show that carbachol increases outflow facility in perfused human ocular anterior segments. The facility increase is accompanied by a rise in the perfusate of cAMP levels. Elevation of cAMP also occurs in freshly excised TM tissue after incubation with carbachol. Finally, we demonstrated the presence, in the outflow pathway, of AC-II and IV, which are regulated by cholinergic agents. Collectively these results suggest the possibility that the mechanism of the action of muscarinics on outflow facility may be mediated directly on the TM with an increased expression of cAMP. 

cAMP has long been recognized as an important cellular messenger capable of regulating such diverse functions as sugar metabolism in the liver, steroidogenesis in the ovary, and cardiac contractility. Recently, it has been recognized that in addition to these crucial physiological functions where the cAMP pathway functions as the primary signal transmittal pathway, there are a number of other key functions, such as differentiation, proliferation, and synaptic plasticity. ¹³ Ectopic expression of AC-II attenuates PDGF-induced signaling, suggesting that AC-II may function as a conditional modulator of regulation. The presence of AC-II can serve to integrate signals between different signaling pathways and as such, modulate proliferation responses (which can cause pathophysiological disorders such as cancer). ¹³  

There are substantial differences between the tissue distributions of the adenylate cyclase subtypes. ¹⁶ Gao and Gilman found that the most obvious functional differences of the ACs are their sensitivity to calmodulin and their response to the G protein βγ subunit complex. ²⁴ AC-II and AC-IV show potentiative interaction between G_sα and forskolin and are the most similar in terms of sequence and structure, both lacking the C_2b domain at the carboxyl terminus. Based on these general criteria, they may have similar regulatory properties. ²⁴ Interestingly, we found that both AC-II and AC-IV were detected only in the circular portion of the normal CM, which subserves changes in accommodation, and not in the longitudinal portion of the CM, which subserves changes in outflow facility. Notably, RT–PCR, the most sensitive method to detect low level expression of target mRNA, could not detect either AC-II or AC-IV mRNA in cultured TM, CM, or CE cells. These results are supported by our previous work; cultured TM and CM cells were incubated with the muscarinic agonists carbachol, aceclidine, and pilocarpine (10⁻³–10⁻⁹ M) without any change in cAMP levels. ³⁶ Surprisingly, AC-IV subtype was not detected by in situ hybridization in normal NPE, whereas PCR detected it. Although in situ hybridization is an effective method to make semiquantitative estimates of relative concentrations of mRNA in cells and tissues, ²⁶ it may not be sensitive enough to detect the low level of expression of AC-IV in our samples. Indeed, RT–PCR is frequently used because it is the most sensitive assay for lower levels of expression. ³⁷  

We found an intervening sequence of ≈255 bp in the AC-IV gene in normal human CP and PE tissue (Fig. 2A , lanes c and d), which is flanked by the primers used for PCR. This finding is similar to a previous report by Gao and Gilman, ²⁴ in which an intervening sequence (≈200 bp) compared with the expected 283-bp sequence. 

Our results, which demonstrated a differential distribution of both of these subtypes between normal and glaucomatous eyes, suggest a common functional role in aqueous outflow physiology, which may be compromised in glaucoma. Of course, further studies are needed to determine whether the absence of these subtypes in ocular tissue from glaucomatous eyes is of primary importance or is simply artifactual (i.e., the result of chronic drug treatment or an end-stage effect). Given this limitation in interpretation, it is notable that the expression of AC-II and AC-IV is absent in many glaucomatous tissues, including the TM. To the best of our knowledge, this is the first report documenting the expression of adenylate cyclase subtypes in fresh human ocular tissue from both normal and glaucomatous eyes.  

In conclusion, the presence of AC-II and AC-IV in normal human ocular outflow tissue supports the hypothesis that cholinergics may indeed exert an effect on outflow facility mediated by cAMP, which is independent of muscle contraction and Ca²⁺ movement. This is further supported by our data, which show that outflow facility increases due to carbachol are accompanied by a rise in cAMP. It is of interest that AC-II and AC-IV are absent from most glaucomatous outflow tissue. Further studies should be continued to determine the nature of the role of these cyclases in both normal and glaucoma outflow physiology. 

 Supported by National Institutes of Health Grant EYO7321 (Bethesda, Maryland); Research to Prevent Blindness; and the Massachusetts Lions Eye Research Fund (Northborough, Massachusetts).

 Submitted for publication April 26, 1999; revised October 26, 1999; accepted November 16, 1999.

 Commercial relationships policy: N.

 

Table 1.

 

Primers Used for PCR Amplification of the AC-II and AC-IV Genes

Primer Name	Primer Sequence (5′–3′)	Gene Sequence Position*
AC-II-3F	GCG TCT ACC TGT	*** >587-610, **
	CTG CAA CAC CAG
AC-II-3R	GTG AAG CCA ACA	942-965, **
	ATG TCA GCG TAT
AC-IV-3F	TTG GGC TGA GGA	1456-1479, ***
	GGA AGA CGA GAA
AC-IV-3R	GTT CGA TGA CCT	1716-1739, ***
	GGA AGA ACT TGG

 F, Forward primer; R, reverse primer.

*  Position in Genbank:

**  M80550,

***  MM80633.

Table 2.

 

Outflow Facility and cAMP in Anterior Segments Perfused with Carbachol

Dose (M)	Co	Cd/Co	cAMPo	cAMPd/cAMPo
10−9	0.314 ± 0.03	1.10 ± 0.05	0.47 ± 0.08	1.09 ± 0.06
10−7	0.314 ± 0.03	1.16 ± 0.03*	0.47 ± 0.08	1.70 ± 0.62
10−5	0.314 ± 0.03	1.27 ± 0.10, **	0.47 ± 0.08	3.10 ± 1.72

 Outflow facility was calculated in perfused human ocular anterior segments and perfusate cAMP was measured, as described in the Methods section. After a baseline facility was determined (C_o), the chamber contents were exchanged with carbachol and post-drug facilities (C_d) were determined. Outflow facility data C_o and C_d are expressed as microliters per minute per millimeters of mercury. Baseline cAMP (cAMP_o) and post-drug (cAMP_d) measurements are expressed as pmoles per sample. Data are mean ± SEM; n = 5.

*   P < 0.01,

**   P < 0.005.

Figure 1.

 

The effect of carbachol on outflow facility and cAMP accumulation in human eyes. Outflow facility and perfusate cAMP accumulation were determined in the perfused ocular anterior segment from a 91-year-old donor, as described in the Methods section. Sequential dosages of carbachol (10⁻⁹, 10⁻⁷, and 10⁻⁵ M) were added to the perfusion medium at the indicated times. The data shown are from a single representative experiment.

Table 3.

 

cAMP Accumulation in Excised TM after Exposure to 10⁻⁷ M Carbachol

	f mol/mg	Carbachol/Control	Forskolin/Control
Control	134.49 ± 39.62	—	—
Carbachol	178.56 ± 41.79	1.39 ± 0.12*	—
Forskolin	350.41 ± 75.60	—	2.77 ± 0.23, **

 One to 4-mm pieces of TM were excised from a pair of human donor eyes and incubated in DMEM (control), 10⁻⁷ M carbachol, or 100 μM forskolin for 30 minutes. cAMP (fmol/mg protein) was measured by radioimmunoassay. Data are mean ± SEM from triplicate determination.

*   P < 0.025,

**   P < 0.005.

Figure 2.

 

(A) Row a, RT–PCR amplification of AC-II transcripts in normal ocular tissue and cultured cell lines with AC-II primers as described in Table 1 . Row b, Southern blot analysis of RT–PCR products from row a probed with[ ³²P]-labeled cDNA of AC-II. Row c, RT–PCR amplification of AC-IV transcripts in normal ocular tissue and cultured cells with AC-IV primers as described in Table 1 . All samples were the same as in row a. Row d, Southern blot analysis of RT–PCR products from row c probed with [³²P]-labeled cDNA of AC-IV. Row e, RT–PCR amplification of G3PDH transcripts with G3PDH primers. All samples were the same as in row a. Rat brain is used as a positive control. (B) Row a, RT–PCR amplification of AC-II transcripts in glaucoma ocular tissues with AC-II primers as described in Table 1 . Row b, Southern blot analysis of RT–PCR products from row a probed with [³²P]-labeled cDNA of AC-II. Row c, RT–PCR amplification of AC-IV transcripts in glaucomatous ocular tissue with AC-IV primers as described in Table 1 . All samples were the same as in row a. Row d, Southern blot analysis of PCR products from row c probed with [³²P]-labeled cDNA of AC-IV. Row e, RT–PCR amplification of G3PDH transcripts in glaucoma ocular tissue, with G3PDH primers. All samples were the same as in row a.

Table 4.

 

AC-II and AC-IV Expression in Human Ocular Tissue by RT–PCR and Southern Blot Analysis

	Rat Brain	TM	CM	CP	CE	CS	AS	NPE	PE
AC-II
Normal	+	+	+	+	+	+	+	+	+
Glaucoma	+	−	−	−	*	−	−	−	−
AC-IV
Normal	+	+	+	+	−	−	+	+	+
Glaucoma	+	+	−	+	*	−	−	−	+

 TM, trabecular meshwork; CM, ciliary muscle; CP, ciliary processes; CE, cornea endothelium; CS, cornea stroma; AS, anterior sclera; NPE, non-pigmented epithelium; PE, pigmented epithelium.

*  Tissue not available.

Figure 3.

 

Expression of AC-II mRNA in normal human ocular tissue by in situ hybridization with DIG-labeled antisense probe (A, C, E, G, and I) and sense probe (B, D, F, H, and J) as described in the Methods section. All tissues are from a 70-year-old donor. Note the significantly positive staining of AC-II mRNA on TM, CM (circular portion), CP, CE, CS, AS, NPE, and PE when hybridized with the antisense probe. No substantial staining is seen when hybridized with the sense probe. Original magnification, (A, B) ×82; (C through J) ×820. L, longitudinal ciliary muscle; C, circular ciliary muscle.

Figure 4.

 

Expression of AC-IV mRNA in normal human ocular tissue by in situ hybridization with DIG-labeled antisense probe (A, C, E, and G) and sense probe (B, D, F, and H) as described in the Methods section. The tissue in (A) and (B) is from an 80-year-old donor. All other tissues are from a 70-year-old donor. Note the significantly positive staining of AC-IV mRNA on TM, CM (circular portion), CP, AS, and PE when hybridized with the antisense probe. No substantial staining was observed with the sense probe. Magnification, (A, B) ×82; (C through H) ×820. L, longitudinal ciliary muscle; C, circular ciliary muscle.

Table 5.

 

AC-II and AC-IV Expression in Human Ocular Tissue Detected by In Situ Hybridization

	TM	CM*	CP	CE	CS	AS	NPE	PE	SC-O	SC-I
AC-II	++	+++	+++	+	++	++	+	+	++	−
AC-IV	++	+++	+++	−	−	++	−	+	++	−

 TM, trabecular meshwork; CM, ciliary muscle; CP, ciliary processes; CE, cornea endothelium; CS, cornea stroma; AS, anterior sclera; NPE, non-pigmented epithelium; PE, pigmented epithelium; SC-O, outer wall of Schlemm’s canal; SC-I, inner wall of Schlemm’s canal.

*  Circular portion only.

The authors thank Huanming Yang for providing careful review of the manuscript and invaluable comments.