Cationic Ferritin Changes Outflow Facility in Human Eyes Whereas Anionic Ferritin Does Not

Abstract

purpose. To determine the effect of charged moieties within the outflow pathway on aqueous outflow facility in human eyes.

methods. After baseline facility measurement in human eye bank eyes (n= 10 pairs), one eye of each pair received anterior chamber exchange and continued perfusion with medium containing 10 mg/ml cationic ferritin. Contralateral eyes were treated in a similar manner with anionic ferritin (10.0 or 102 mg/ml). Eyes were fixed by anterior chamber exchange and perfusion with universal fixative at 8 mm Hg (corresponding to a physiologic pressure of 15 mm Hg in vivo) and examined by transmission electron microscopy. In a second series of human eyes (n = 8 pairs), facility was measured before and after anterior chamber exchange, with a solution containing 0.1 U/ml neuraminidase.

results. Perfusion of eyes with anionic ferritin at either 10.0 or 102 mg/ml caused a negligible 2% increase in facility, whereas cationic ferritin perfusion reduced facility by 66% (P < 0.00001). Perfusion with fixative reduced facility by approximately 60% in both cationic and anionic ferritin-perfused eyes, relative to facilities after perfusion with ferritin. Transmission electron microscopy showed that the distribution of ferritin was segmentally variable. Cationic ferritin consistently labeled the luminal surface of the inner wall of Schlemm’s canal, and variably labeled the juxtacanalicular connective tissue (JCT) and trabecular beam surfaces. Anionic ferritin was more prominent in the JCT and intertrabecular spaces and less so on the luminal surface of Schlemm’s canal. By scanning electron microscopy, cationic ferritin was seen to accumulate at intercellular margins of the inner wall. Neuraminidase perfusion had no significant effect on outflow facility.

conclusions. Cationic ferritin reduces outflow facility, presumably by binding to negatively charged sites in the outflow pathway. A possible mechanism is partial or complete blockage of intercellular clefts in the inner wall of Schlemm’s canal by the ferritin that accumulates on the luminal surface of the inner wall. Although they are possible targets for ferritin binding, sialyl residues themselves seem to have little direct effect on outflow facility. Our data indicate that positively charged molecules, especially if they can interact with inner wall pores, have the potential to markedly alter outflow facility.

The cause of the elevated aqueous humor outflow resistance characteristic of primary open-angle glaucoma (POAG) remains unknown. Management of POAG is based on reduction of intraocular pressure (IOP), and this will no doubt continue to play an important role for the foreseeable future. A better understanding of the factors controlling outflow facility in normal eyes would facilitate IOP control. 

Charged moieties within the juxtacanalicular connective tissue (JCT) and on Schlemm’s canal endothelium have been hypothesized to play a role in determining aqueous outflow facility. ^{1 2} In particular, Tripathi et al. ³ have suggested that differences in sialic acid content between normal and glaucomatous eyes play a role in ocular hypertension. Normal and glaucomatous human eyes showed different patterns of cationic ferritin (CF) labeling after postmortem perfusion, ⁴ possibly related to differences in the distribution of charged species within the outflow tract. 

In unrelated work, we discovered that CF perfusion markedly affects the outflow facility of human eyes. The goal of the present work was to investigate this effect in detail. First, we compared the effects of CF and anionic ferritin (AF) perfusion on outflow facility. Second, we examined the distribution of these tracers in the outflow tissues of perfused eyes. Third, we studied the effects of neuraminidase, an enzyme that hydrolyzes sialyl residue linkages, on outflow facility. Taken together, these studies were meant to improve our understanding of how charged moieties influence aqueous outflow facility. 

Methods

Ferritin Perfusion and Facility Measurements

All perfusions used Dulbecco’s phosphate-buffered saline with 5.5 mM glucose added (DBG) as mock aqueous humor. DBG fluid was prefiltered through a 0.22-μm filter (Millex-GS; Millipore, Bedford, MA) before use. CF (catalog number F7879) and AF (catalog number F4503) were obtained from Sigma (St. Louis, MO) and were diluted to the desired concentrations using DBG. The AF was type I from horse spleen; the CF was also from horse spleen and was rendered cationic by conjugation with N,N-dimethyl-1,3-propanediamine. The resultant solutions were sonicated for 6 to 10 minutes before use to break up clumps (model Bransonic 32 Sonicator Bath; Branson Instruments, Danbury, CT). The pH of the AF solution was 5.65, whereas for the cationic solution it was 5.87. 

Ostensibly normal human eyes were obtained within 24 hours after the donor’s death (n = 10 pairs). The mean donor age was 70.3 years (range, 40–85) and the mean time after death to the start of perfusion was 21.1 hours (range, 11–25 hours). Baseline facility was measured for 60 to 90 minutes by perfusion into the posterior chamber with DBG. All perfusions were performed at a constant pressure of 8 mm Hg (corresponding to 15 mm Hg in vivo), using previously described techniques. ⁵ At the completion of baseline facility measurement, one eye of each pair received anterior chamber exchange with a solution of 10.0 mg/ml CF. Contralateral eyes were treated in a similar manner with either 10.0 or 102 mg/ml AF. Care was taken during the anterior chamber exchange to maintain a constant IOP of 8 mm Hg. Eyes were then further perfused with their respective ferritin solutions at 8 mm Hg for 60 to 90 minutes, while facility was measured. 

We also obtained one pair of POAG-affected eyes (donor age, 88 years) perfused 24 hours after death. They were perfused with 10 mg/ml CF and 102 mg/ml AF, as described for the remaining eyes. 

We present the facility data in two ways. The first is by simple averaging of facility data between eyes, with the resultant values reported as mean ± SEM. The second is by normalizing the raw facility data, thereby allowing comparison of relative facility changes between pairs of eyes. Normalization involved dividing the measured facility at each time for a given eye by the average facility reading for that eye in the 30-minute period before anterior chamber exchange with ferritin. The percentage facility change due to ferritin (or fixative) perfusion was then computed as 100 · (C _norm,pre − C _norm,post), where C _norm,pre and C _norm,post are the normalized facilities before and after ferritin (or fixative) exchange, respectively. The statistical significance of facility changes due to ferritin (or fixative) was computed from a paired two-tailed Student’s t-test using the percentage normalized facility change as the statistic of interest. 

Neuraminidase Perfusion and Facility Measurements

Two types of neuraminidase were used: from Clostridium perfringens (Boehringer-Mannheim, Laval, Quebec, Canada; catalog number 1585 886) and from Arthrobacter ureafaciens (Boehringer-Mannheim; catalog number 269 611). Lyophilized powder was dissolved in DBG+0.3% weight bovine serum albumin (BSA; Sigma) to make a stock solution of 1 U/ml. Working solutions of 0.1 U/ml were then made up by dilution of stock solutions with DBG+0.3% BSA, and the pH was adjusted to 5.9 by titration with HCl. This pH is in the middle of the optimal range for enzyme activity. Supplementary data from the manufacturer on the neuraminidase preparation from C. perfringens indicated that the enzyme had some protease contamination (as much as 7.6 U protease per mg lyophilizate), whereas the preparation from A. ureafaciens was essentially protease free (<0.02 U protease/mg lyophilizate). 

The concentration of neuraminidase was based on measurements by Tripathi et al., ¹ who state that the human trabecular meshwork contains 3.6 micromoles sialic acid per gram wet weight. Considering the meshwork to have a cross-sectional (flow-wise) area of 0.05 to 0.13 cm^{2 6} and an average internal–external thickness of 200 μm, we estimated the wet weight of the meshwork to be approximately 1 to 3 mg, which corresponds to a total sialic acid content of approximately 4 to 11 nanomoles. One unit of the enzyme liberates 1 micromole of sialic acid per minute by hydrolyzing terminal bonds joining sialic acid to oligosaccharides, glycoproteins, and glycolipids. Considering the duration of exposure (typically several hours) and the continual flushing of the meshwork with new perfusate, we estimated that the described concentration of neuraminidase would be sufficient to largely hydrolyze bonds, attaching the sialyl residues within the meshwork. 

The perfusion protocol was identical with that used for the ferritin experiments, except that the eyes received neuraminidase instead of ferritin. Specifically, in four pairs of eyes, one eye received neuraminidase from C. perfringens, and the contralateral eye received vehicle (DBG+0.3% weight BSA, pH adjusted to 5.9). In two pairs of eyes, one eye received neuraminidase from A. ureafaciens and the contralateral eye received vehicle. Finally, in two pairs of eyes, one eye received neuraminidase from C. perfringens, and the contralateral eye received neuraminidase from A. ureafaciens. The mean donor age for this group of eyes was 83.1 years (range, 72–92), and the mean time after death was 21.5 hours (range, 12–26). 

Neuraminidase activity was verified using a fluorometric in vitro assay based on the liberation of methylumbelliferone (MU) from 2′-(4-methylumbelliferyl)-α-d-N-acetylneuraminic acid (MU-NANA). ^{7 8} MU-NANA (0.4 ml of 0.25 mM; Toronto Research Chemicals, Toronto, Ontario, Canada) in 50 mM sodium phosphate buffer (pH 5.0) was incubated with 1 μl neuraminidase solution for 1 minute at 37°C, followed by addition of 0.6 ml 1 M sodium carbonate solution. After approximately 4 minutes, the stabilized fluorescence signal was read with a fluorescence spectrophotometer (QM-1; Photon Technology International, Lawrenceville, NJ), using excitation and emission wavelengths of 390 nm and 450 nm, respectively. Both enzyme types were assayed at concentrations of 0.0125, 0.05, and 0.1 U/ml, spanning the range of concentrations expected to occur in perfused eyes. All readings were corrected for background fluorescence, by using a control assay without enzyme. 

Morphology

At the conclusion of facility measurement after ferritin perfusion, seven pairs of eyes were fixed by anterior chamber exchange and perfusion with universal fixative (2.5% paraformaldehyde, 2.5% glutaraldehyde in Sörensen’s buffer) at 8 mm Hg for 60 to 90 minutes, followed by overnight immersion fixation. The remaining eyes (n = 3 pairs) were lightly fixed by anterior chamber exchange and perfusion with universal fixative at 2 to 3 mm Hg for only 15 minutes, followed by overnight immersion fixation. 

Tissue was processed for transmission electron microscopy (TEM). Briefly, radial segments of the limbal area were dissected, postfixed in 1% osmium tetroxide, dehydrated, infiltrated, and embedded in Epon-Araldite. Some ultrathin sections were stained with uranyl acetate and lead citrate, whereas other sections were not stained. Use of unstained samples allowed clear and definitive visualization ferritin particles within the extracellular matrix of the JCT. Adjacent tissue samples were microdissected and processed by conventional methods ⁹ to produce scanning electron microscopic montages of the inner wall of Schlemm’s canal. 

Results

Facility Effects of Ferritin

Perfusion with either 10.0 or 102 mg/ml AF had no discernible effect on outflow facility (Fig. 1 and Table 1 ). Before ferritin exchange, there was no significant difference in facility between the AF- and CF-perfused eyes. However, perfusion with 10.0 μg/ml CF caused a rapid decrease in facility, followed by a gradual time-dependent decrease (Fig. 1) . We could not measure facility during the anterior chamber exchange, and we were therefore unable to precisely measure the time course of the rapid facility decrease due to CF. The facility decrease due to CF was statistically significant when compared with the contralateral AF-perfused eyes (P < 0.00001 for the 30-minute period before fixation; Table 1 ). 

In the seven pairs of eyes in which facility was measured after fixation, the fixative caused a further facility decrease of 58% (AF-perfused eyes) or 62% (CF-perfused eyes). These decreases were not statistically different from one another. The facility response of the single pair of glaucomatous eyes was similar to that of the normal eyes. Specifically, there was a small (several percentage points) decrease in facility due to AF perfusion and a large, rapid decrease due to CF perfusion. The net facility decrease due to CF perfusion (relative to the AF-perfused eye) was 56%, which is comparable to the 66% decrease seen in normal eyes. 

Facility Effects of Neuraminidase

Both neuraminidase preparations showed an approximately linear dose–response curve over the concentration range from 0 to 0.1 U/ml in the in vitro assay and the expected amount of total activity. The activity of the enzyme from C. perfringens was approximately 50% higher than that from A. ureafaciens. We conclude that the enzyme preparations we used had biological activity. 

In eyes perfused with neuraminidase from C. perfringens, there was a time-dependent increase in facility for 3 to 4 hours after anterior chamber exchange. After 180 minutes of neuraminidase perfusion, the net facility increase was 71% ± 18% (mean ± SEM), which was different from zero at P = 0.03. However, similar perfusions with neuraminidase from A. ureafaciens (negligible protease contamination) showed no significant increase in facility (net increase of only 4%). This was consistent with experiments in which the two types of neuraminidase were perfused into paired eyes: the eyes perfused with neuraminidase from C. perfringens showed a mean facility increase of 71% when compared with eyes perfused with neuraminidase from A. ureafaciens. Although a relatively small number of eyes were perfused in this part of the study, the results are therefore consistent with a facility effect from the protease-contaminated neuraminidase but no effect from the protease-free enzyme preparation. 

Morphologic Findings in Ferritin-Perfused Eyes

By TEM, cationic and AF showed markedly different distribution patterns within the meshwork, similar to those reported by deKater et al. ⁴ Generally speaking, both AF and CF distributions were highly variable, with some regions showing extremely dense labeling and others showing little or no labeling (Figs. 2 and 3) . The CF was usually present as clumps, whereas the AF was more punctate and uniform. 

In CF-perfused eyes, we consistently observed a coating on the luminal surface of the inner wall of Schlemm’s canal. The CF on the inner wall was almost invariably clumped and of variable thickness. The abluminal surface of the inner wall (including the linings of giant vacuoles) was very sparingly coated with CF, whereas the lumens of the giant vacuoles themselves were never filled with CF (Fig. 2A) . The CF distribution within the JCT was variable, with most regions being devoid of CF and others showing a modest number of discrete CF particles or clumps. Very rarely, isolated regions of the JCT were heavily filled with CF (Figs. 2B 2D) . CF was seen to fill intercellular clefts between inner wall cells, with a marked tendency to accumulate near the luminal side of the cleft (Fig. 2C) . There was also patchy accumulation of CF in extracellular material immediately under the inner wall. The beams were usually coated with a thin layer of CF, but this coating did not typically extend to any appreciable extent into the intertrabecular spaces. One exception to this was in the uveal meshwork, where intertrabecular spaces were occasionally filled with CF. However, this seemed to occur only over the nonfiltering parts of the meshwork. CF was also seen coating the endothelial lining of collector channels. 

In AF-perfused eyes, there was much less ferritin coating on the inner wall of Schlemm’s canal on the luminal surface (Fig. 3) . Giant vacuoles were sometimes entirely filled with AF, and the JCT frequently demonstrated extremely heavy labeling, particularly at the higher concentration (102 mg/ml) of AF. However, even at this high concentration, there were always some giant vacuoles and regions of the JCT that were entirely devoid of label (Fig. 3A , right side). AF was only very rarely seen in intercellular clefts between inner wall cells (Fig. 3C) . AF was observed in the extracellular matrix immediately under the inner wall, where the accumulation seemed similar to that observed for CF (Fig. 3C) . Particularly striking was the distribution of AF in the corneoscleral meshwork, where the majority of the intertrabecular spaces were completely and densely filled with AF (Figs. 3A 3B) . 

Scanning electron microscopic examination of the inner wall of Schlemm’s canal showed significant differences between AF- and CF-perfused eyes. In CF-perfused eyes, most regions of the inner wall showed some ferritin decoration. Most frequently, this took the form of discrete ferritin clumps (Fig. 4) . The intercellular margins were often prominently elevated above the surrounding endothelial surface in CF-perfused eyes (Fig. 4) . This seemed to be due to accumulation of ferritin at the margin of the inner wall cells and lifting of the overlying cell where inner wall cells overlapped one another at their margins. In AF-perfused eyes, we observed only very infrequent and sparse ferritin accumulation on the inner wall, and the intercellular borders looked similar to those previously reported in control eyes (Fig. 5) . 

We spent considerable time searching for inner wall pores in CF-perfused eyes. However, due to the heavy ferritin labeling on the inner wall of Schlemm’s canal, it was very difficult to find and unambiguously identify pores in these eyes. In the left panel of Figure 6 an unambiguous border pore appears; however, it is important to emphasize that such unlabeled pores were extremely rare in CF-perfused eyes. Much more representative was the finding shown in the right panel of Figure 6 , where two structures that appear to be pores are surrounded by many ferritin clumps. We did not attempt to quantify inner wall pores in CF-perfused eyes, because the CF clumps probably impaired the view of pores. 

Because we were concerned that ferritin could have been washed out of the meshwork during fixation, we also examined tissue from lightly fixed eyes (15-minute fixative perfusion at low pressure). The AF and CF distributions in these lightly fixed eyes were the same as those in conventionally fixed eyes. This suggests that absence of ferritin in a given preparation is due to absence of binding, rather than washout by fluid. 

Discussion

The major conclusion of the present work is that CF perfusion profoundly affects outflow facility in enucleated human eyes, whereas the same (or 10 times higher) concentration of AF has no effect on facility. We would not have predicted this from a priori examination of transmission electron micrographs of the outflow tissues in perfused eyes, because the amount of AF labeling in the trabecular meshwork was generally much greater than the amount of CF labeling, especially at the higher AF concentration. This implies that CF was much more efficient at obstructing outflow than AF (per unit concentration), that the CF accumulated in a location that was particularly effective in blocking aqueous outflow, or both. 

Of these two possibilities, we think that the first is unlikely. The cationization procedure acts by modifying the carboxyl terminals of the native ferritin subunits ¹⁰ and as such is not expected to markedly change protein configuration. This is supported by measurements of the diffusion coefficient and Stokes radius of both types of ferritin, which show that the hydrodynamic characteristics of the two molecules are not significantly different from one another. ¹¹ Further, the clumping of CF that we observed would be expected to produce a lower flow resistance (per unit mass of ferritin) than the more homogeneous distribution seen with AF, because clumping invariably produces ferritin-poor channels through which flow can pass. In short, AF should be at least as hydrodynamically resistant as CF, per unit mass of ferritin. 

It therefore seems that the different effects that CF and AF have on facility are best explained by preferential accumulation of CF in a location that is effective at blocking aqueous outflow. This is consistent with the very rapid time course of CF-induced facility alterations, pointing to a hydrodynamic blockage effect by accumulating CF molecules. In this regard it is noteworthy that the one location where CF was more consistently seen than AF was on the luminal surface of the inner wall of Schlemm’s canal. This luminal accumulation of CF was most striking along the intercellular junctions, and seems to be consistent with observations of CF clumps near the mouths of inner wall openings (pores), although identification of such pores was usually not unambiguous (e.g., Fig. 6 , right panel). 

If it is accepted that aqueous humor passes through pores of the inner wall, often associated with intercellular junctions, ¹² then accumulation of ferritin at or over these locations would be expected to have a significant effect on outflow resistance, because the pores typically account for only approximately 0.1% of the inner wall surface ⁹ and represent “funnels” ¹³ or choke points where outflow can be relatively easily obstructed. The implication is that any substance that interacts with these pores to increase their resistance can very effectively reduce aqueous outflow facility. Negatively charged glycoproteins are thought to be present in the intercellular cleft of microvascular endothelia, particularly near the luminal surface, and are thought to play a central role in controlling the permeability of such endothelia. ¹⁴ If similar glycoproteins were present in the intercellular clefts between inner wall endothelium, then it seems plausible that cationic molecules with large Stokes radii (hydrodynamic resistance) could bind to negative sites on the glycoproteins and efficiently block intercellular pores. 

Our facility results are qualitatively and quantitatively consistent with experimental data from a study in which Turner et al. ¹⁵ perfused frog mesenteric capillaries with CF and AF. In this system, perfusion with CF with concentrations in the range of 1 to 25 mg/ml caused the hydraulic permeability of the vessel wall to be reduced by 70% (compared with 66% in our system), whereas perfusion with AF had minor effects on hydraulic permeability. Turner et al. also observed CF and AF labeling patterns on the capillary endothelium very similar to those reported in the current study. Our observed patterns are also consistent with studies by Rounds and Vaccaro, ¹⁶ who showed an intrinsic preference for CF to bind to the luminal (vs. the abluminal) surface of rat capillary endothelium. 

Perfusions with protease-free neuraminidase did not show any significant effect on outflow facility. This does not prove that sialic acid residues do not influence outflow facility—for example, binding of charged moieties to sialyl residues could indirectly alter outflow facility. However, it suggests that acute alteration of sialic acid residue levels does not alter facility per se. 

It is noteworthy that the percentage of facility reduction due to fixative infusion was almost identical in the AF- and CF-perfused eyes. Although it is not understood how fixation changes outflow facility, this result gives us some insight into meshwork function. More specifically, it implies that fixation effects in these eyes depended on the outflow resistance before fixation. This rules out a scenario in which fixation increases the resistance of one part of the meshwork, while CF increases the resistance of a second part that does not hydrodynamically interact with the first part. In such a noninteracting model, the magnitude of the resistance increase due to fixation would be independent of resistance increase due to ferritin, which was not observed. We conclude that fixation and CF must be interacting somehow. Perhaps fixation affects the flow resistance of the JCT and ferritin affects the pores of the inner wall, with interaction occurring through a funneling effect. ¹³  

It is remarkable how segmentally variable the CF distribution was and how CF was never seen filling the lumens of giant vacuoles, even though AF usually filled these lumens. We do not have a good explanation for these effects, which were first reported by deKater et al. ⁴ Because the hydrodynamic properties of CF and AF are similar, it seems probable that CF and AF were carried into the meshwork in a similar way during perfusion. Differences between CF and AF distributions (and possibly segmental variations) might therefore reflect the ability of the tissue to bind and/or retain the tracer. 

In summary, CF markedly reduced outflow facility in human eyes, possibly due to preferential accumulation of CF molecules at one or more critical locations in the outflow system. Our data suggest that intercellular clefts (pores) of the inner wall of Schlemm’s canal are a likely site of CF accumulation and flow resistance increase. Note that this does not imply that intercellular clefts have significant flow resistance under normal circumstances (i.e., in the absence of CF), but it suggests that cationic entities can lower outflow facility by blocking such sites. 

 Presented at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 1999.

 Supported by Grant MA-10051 from the Medical Research Council of Canada and the Glaucoma Research Society of Canada.

 Submitted for publication January 16, 2001; accepted February 16, 2001.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked“ advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Plot of normalized facility versus time, averaged over all pairs of eyes (n = 10 pairs before fixative exchange; n= 7 pairs after fixative exchange). Time zero is taken as the beginning of anterior chamber exchange for ferritin infusion. Lines are the mean; error bars represent SEM at each time point. Data from AF-perfused eyes (10 and 102 mg/ml) were pooled, because the facility response was statistically indistinguishable. Transient dips in facility after fixative exchange were due to re-establishment of steady state perfusion conditions after anterior chamber exchange and should be ignored in favor of stable values after approximately 120 minutes.

Table 1.

 

Summary of Facility Data in Ferritin Perfusions

Perfusion Protocol	Facility*
		Baseline	After Ferritin	After Fixation
Anionic ferritin (10.0 μg/ml and 102 μg/ml; data pooled)	0.289 ± 0.045	0.304 ± 0.050	0.105 ± 0.038
Cationic ferritin (10.0 μg/ml)	0.291 ± 0.050	0.092 ± 0.021	0.035 ± 0.008

 Data are expressed in microliters/minute × mm Hg.

*  Baseline facility obtained by averaging over a 30-minute period before ferritin exchange. After ferritin facility obtained by averaging over a 30-minute period before fixative exchange but after ferritin exchange. After fixation facility obtained by averaging over a stable 30-minute period after fixative exchange.

Figure 2.

 

Transmission electron micrographs of outflow pathway tissue from 10 mg/ml CF-perfused eye. (A, B) Overview of JCT and Schlemm’s canal (unstained specimens), showing CF labeling on the inner and outer walls. (A) Relatively unlabeled JCT. (B) An essentially ferritin-free region immediately adjacent to a zone of densely packed ferritin. (C) High-magnification view of the inner wall of Schlemm’s canal, showing ferritin partially filling an interendothelial cleft and accumulating at the luminal surface of the cleft (stained specimen). (D) Higher magnification of boxed region in (B). This eye from an 84-year-old donor was perfused 24 hours after death. Baseline facility was 0.20 μl/min · mmHg and facility after CF was 0.11 μl/min · mmHg. Arrowheads: CF clumps. SC, Schlemm’s canal; GV, lumen of giant vacuole.

Figure 3.

 

Transmission electron micrographs of outflow tissue from AF-perfused eye (102 mg/ml). (A) Overview of Schlemm’s canal, JCT, and corneoscleral meshwork (unstained specimen). Note the extensive filling of intertrabecular spaces in the corneoscleral meshwork. (B) JCT and the inner wall of Schlemm’s canal (unstained specimen), showing extensive AF labeling. (C) High magnification view of inner wall of Schlemm’s canal, showing spare labeling of the inner wall and relative absence of AF in intercellular cleft. This is the contralateral eye to the one in Figure 2 . Baseline facility was 0.10μ l/min · mmHg, and facility after AF was 0.10 μl/min · mmHg. Arrowheads: AF particles. SC, Schlemm’s canal; SCE, Schlemm’s canal endothelium.

Figure 4.

 

Scanning electron microscopic overview of inner wall of Schlemm’s canal from a CF-perfused eye. Inset: higher magnification of the boxed area of the main image. Note the ferritin clumps on the inner wall, as well as the prominent intercellular margins (arrowheads). This image is of the eye depicted in Figure 2 .

Figure 5.

 

Scanning electron microscopic overview of the inner wall of Schlemm’s canal from an AF-perfused eye. Inset: higher magnification of the boxed area at the upper left of the main image. Little or no ferritin labeling was observed. This image is of the eye depicted in Figure 3 . p, inner-wall pore; J, junction between two inner-wall cells.

Figure 6.

 

Scanning electron microscopic images of pores from the inner wall of Schlemm’s canal from a CF-perfused eye. Left: border pore from a relatively unlabeled portion of the inner wall. Such unlabeled pores were very unusual. Right: two structures that appear to be pores (arrows) with large accumulations of ferritin around their margins. An artifactual pore ( ) is clearly identified by its rough, uneven margins. This image is of the eye depicted in Figure 2 .

The authors thank Doug Johnson for helpful suggestions during the course of this work and the donors’ families and the staff at the Eye Bank of Canada (Ontario Division) and National Disease Research Interchange (Philadelphia, PA) for providing tissue used in this work.