A method describing the microdissection of trabecular meshwork tissue from Brown Norway rat eyes

Abstract

Glaucoma is often associated with elevated intraocular pressure (IOP), generally due to obstruction of aqueous humor outflow within the trabecular meshwork (TM). Despite many decades of research, the molecular cause of this obstruction remains elusive. To study IOP regulation, several in vitro models, such as perfusion of anterior segments or mechanical stretching of TM cells, have identified several IOP-responsive genes and proteins. While these studies have proved informative, they do not fully recapitulate the in vivo environment where IOP is subject to additional factors, such as circadian rhythms. Thus, rodent animal models are now commonly used to study IOP-responsive genes in vivo. Several single-cell RNAseq studies have been performed where angle tissue, containing cornea, iris, ciliary body tissue in addition to TM, is dissected. However, it is advantageous to physically separate TM from other tissues because the ratio of TM cells is relatively low compared to the other cell types. In this report, we describe a new technique for rat TM microdissection. Evaluating tissue post-dissection by histology and immunostaining clearly shows successful removal of the TM. In addition, TaqMan PCR primers targeting biomarkers of trabecular meshwork (Myoc, Mgp, Chi3l1) or ciliary body (Myh11, Des) genes showed little contamination of TM tissue by the ciliary body. Finally, pitfalls encountered during TM microdissection are discussed to enable others to successfully perform this microsurgical technique in the rat eye.

1. Introduction

Glaucoma affects approximately 66 million people worldwide (Tham et al., 2014). Primary open-angle glaucoma is the most common form, where elevated intraocular pressure (IOP) is a primary risk factor (Kwon et al., 2009). Elevated IOP is most often due to dysfunction of aqueous humor outflow in the anterior eye, and specifically, the trabecular meshwork (TM). In the normal conventional outflow pathway, aqueous humor drains through the TM, exits via Schlemm’s canal to collector channels, and then to the episcleral venous circulation (Stamer and Acott, 2012). In glaucoma, aqueous outflow pathways become progressively occluded by excess extracellular matrix, which hinders outflow and increases IOP (Acott et al., 2021). Many laboratories have investigated gene expression changes associated with elevated IOP in ex vivo perfused anterior segments from human cadavers, pig, and bovine eyes (Erickson-Lamy et al., 1991; Ethier and Chan, 2001; Gonzalez et al., 2000; Johnson, 2005; Keller et al., 2011; Mao et al., 2011; Vittitow and Borras, 2004; Vranka et al., 2015). Other groups have studied gene expression in mechanically stretched TM cells in culture as a proxy for elevated IOP (Luna et al., 2009; Vittal et al., 2005). These important studies have revealed numerous pathways and specific genes associated with elevated IOP (Vittitow and Borras, 2004; Vranka and Acott, 2017; Vranka et al., 2020). However, the major disadvantage of these studies is that the perfusion cultures are ex vivo, and the anterior segments are not subject to the normal and complex aqueous dynamics of the living eye. To be able to study factors influencing IOP in vivo, animal models are needed to study TM gene expression.

Rodent models of ocular hypertension are valuable tools to study cellular mechanisms involved in IOP regulation (McDowell et al., 2022; Pang and Clark, 2020; Pang et al., 2015). The anatomy of the rodent aqueous outflow pathway is similar to humans (Morrison et al., 1995; Morrison et al., 1997; Overby et al., 2014). Evaluation of the rat eye by polychrome staining of Epoxy-embedded tissue shows the rat TM, angle, and aqueous humor drainage structures (Figure 1A). There are fenestrated beams of corneoscleral meshwork and a single lumen Schlemm’s canal (McMenamin and al-Shakarchi, 1989; Morrison et al., 1995; Tripathi, 1971; van der Zypen, 1977). A collector channel (CC) is also clearly visible in this section. High magnification images clearly show 2–3 layers of fenestrated TM beams and nuclei of Schlemm’s inner wall cells (Figure 1B) (Lei et al., 2011). Given these features, the rat is an attractive species to perform in vivo TM studies. Ocular hypertension can be induced using various methods to occlude outflow channels, or by injecting adenoviruses to over-express bioactive proteins (Giovingo et al., 2013; Mao et al., 2012; McDowell et al., 2015; McDowell et al., 2013; Shepard et al., 2010; Wang et al., 2008), by genetically modifying / manipulating mice (Pang and Clark, 2020; Zode et al., 2011), or by using various methods to occlude outflow channels (Levkovitch-Verbin et al., 2002; Morrison et al., 2015; Morrison et al., 1997; Sappington et al., 2010). Yet, many of these techniques compromise TM tissue in some manner rendering it unusable to study effects on TM. Recently, a Controlled Elevation of IOP (CEI) rat model was described where a thin cannula is inserted through the cornea (Morrison et al., 2016). The cannula is attached to a reservoir containing balanced salt solution (BSS+) the height of which is adjusted to allow BSS+ to perfuse at a defined pressure and duration. Most importantly, the cannula does not touch or disturb the TM, unlike the models described above. Therefore, while the CEI model was primarily developed to study mechanisms of pressure-induced optic nerve damage, it also has the potential to allow study of TM gene expression in response to IOP in an animal model.

Figure 1. Histology of tissues comprising the rat conventional outflow pathway.

(A, B) Methylene blue and fuschin staining of a rat anterior segment embedded in epoxy. S=sclera, CC=collector channel, CB=ciliary body, C=cornea, AC=anterior chamber, SC= Schlemm’s canal, and TM=trabecular meshwork. Image modified from (Morrison et al., 2011).

Single cell RNAseq (scRNAseq) approaches have been successfully used to evaluate gene expression from TMs dissected from human anterior segments (Patel et al., 2020; van Zyl et al., 2020). Similar studies using mouse eyes were also performed, but due to their small size, all anterior segment tissues (cornea, iris, ciliary body, TM) were dissected. Because the rodent TM has few TM cells, the ratio of TM cells in the total population of cells isolated from anterior ocular tissues is low. This could lead to low capture efficiency and bias in transcript coverage, which are common limitations in scRNAseq studies (Chen et al., 2019). In order to enrich TM cell isolation for gene expression studies, a method to physically separate TM from other ocular tissues is needed. Laser capture microdissection of mouse TM tissue has been described, but the technique is cumbersome and requires expensive, highly specialized equipment (Sutherland et al., 2018). Here, we describe a technique for microdissecting TM tissue from rat eyes.

2. Materials and Supplies

2.1. Animals

Sixteen retired Brown Norway breeder rats (Charles River Laboratories, Wilmington, MA), both male and females aged 6–9 months old, were used for this study. All animal protocols were reviewed by the Institutional Animal Care and Use Committee at the Oregon Health & Science University, and were performed in accordance with the National Research Council’s Guide for the Care and Use of Laboratory Animals.

2.2. Dissection supplies

Jeweler’s micro-forceps Dumont #5SF (#11252-00) and Weck-Cel eye spears (#18105-01) (Fine Science Tools, Foster City, CA); Vannas Capsulotomy Curved Scissors (#E3387) (Storz Ophthalmic Instruments, Rochester, NY); 800 and 1500 grit sandpaper, and single-edged razor blades (#3962A8) (McMaster-Carr, Douglasville, GA); Minutien pins, stainless steel, 0.15mm diameter (Carolina Biological Supply Company, Burlington, NC); Hygenic Boxing Wax (#5475815) (Henry Schein Dental, Melville, NY).

2.3. Histopathology

Phosphate buffered saline (PBS) (ThermoScientific, Waltham, MA); paraformaldehyde (#15700), glutaraldehyde (#16100), and polychrome stain components (Methylene blue (#26075-20), Azur II (#11210), and Basic fuchsin (#26031-20) (Electron Microscopy Sciences, Hatfield, PA). Hematoxylin and eosin (H&E), Masson’s Trichrome, and Periodic Acid Schiff (PAS) histological stains were provided from the histopathology core facility at OHSU Knight Cancer Center.

2.4. RNA isolation

RNAzap (#AM9780), TRIzol (#15596018), glycogen (#R0551), sodium acetate solution, pH5.2 (#R1181), nuclease-free water (#R0581), RNaseOUT^™ recombinant RNase inhibitor (#10777019), SuperScript^™ III reverse transcriptase (#18080044), RNase-free tris-EDTA, pH8.0 (#AM 9849), chloroform (#298-500) (ThermoFisher). Diethyl ether (#296082 (MilliporeSigma, St Louis, MO). DNase I (#E1011-A) (ZymoResearch, Irvine, CA). oligo(dT)₁₇ (Integrated DNA Technologies, Coralville, IA). dNTPs (#DA1011, DT1011, DC1011, DG1011) (Meridian Bioscience, Swedesboro, New Jersey).

2.5. TaqMan Gene Expression Reagents.

Pre-designed, rat-specific TaqMan gene expression assays were purchased from ThermoFisher (Table 1). Each primer-probe set consists of an unlabeled, gene-specific PCR primer pair and a probe, which is labeled with fluorescein amidite (FAM) at the 5’ end and a minor groove binder moiety and a non-fluorescent quencher at the 3’ end. TaqMan fast advanced master mix (#4444557), MicroAmp fast 96-well reaction plates (#4346907) and TaqMan preAmp 2X master mix (#4384266) (Applied Biosystems, Waltham, MA).

Table 1:

TaqMan primer probes used in the qPCR analyses.

Target Gene	Assay ID	Catalog #	Exon boundary	Amplicon length	TM or CB marker
Mgp	Rn00563463_m1	4331182	1–2	151 bp	TM
Myoc	Rn00578382_m2	4331182	1–2	70 bp	TM
Chi3l1	Rn01490608_m1	4331182	1–2	107 bp	TM
Myh11	Rn01530321_m1	4331182	27–28	62 bp	CB
Des	Rn00574732_m1	4331182	1–2	78 bp	CB
β-actin	Rn00574732_m1	4331182	4–5	91 bp	housekeeping

2.6. Equipment

A Leica MZ6 binocular dissecting microscope (Wetzlar, Germany); an Olympus BX51 microscope with a DP71 digital microscope camera (Olympus, Center Valley, PA); Olympus Fluoview FV1000 confocal microscope; NanoDrop One (ThermoScientific); QuantStudio 3 real-time thermocycler (Applied Biosystems); vacufuge 5301 (Eppendorf, Enfield, CT).

3. Detailed Methods

3.1. Trabecular Meshwork Microdissection

Following euthanasia by decapitation under deep isoflurane anesthesia, the rat eye globes were removed and transferred to chilled phosphate buffered saline (PBS). Using a binocular dissecting microscope, conjunctiva was removed from the limbus to expose the sclera to the equator. Beginning with a razor blade incision, approximately 1 mm behind and parallel to the limbus, the anterior and posterior sclera were separated using a Vannas scissors (Figure 2A). Following removal of the lens, the anterior corneoscleral tissue with attached ciliary body and iris was placed anterior side down on a sheet of boxing wax, and bisected with a single-edged razor blade (Figure 2B), stabilizing the tissue by pressing the sclera against the bone wax with the two blades of the jeweler’s forceps. Each half was then bisected again, creating four quadrants. These were performed by stabilizing corneal tissue with both points of the jeweler’s forceps against the wax, placing the razor blade between them, and then pulling the razor blade away from the forceps, slicing the tissue through the TM and ciliary body (Figure 2B). Quadrants were then placed together in a single drop of PBS to prevent drying out while awaiting TM dissection.

Figure 2. Microdissection of rat trabecular meshwork tissue.

(A) The anterior segment was removed and cut into quarters (B). (C) The iris and ciliary body were removed with fine jeweler’s forceps to expose the TM. Arrows indicate dissection pins used to anchor the cornea and sclera to the underlying wax. (D) Speckled TM (*) lies between pigmented Schwalbe’s line (white arrows) and ciliary body remnant (white arrowheads) attached to the sclera. (E) The TM (white arrow) was gently peeled off, revealing the white outer wall of Schlemm’s canal (*). (F) Region without TM (white bracket) contrasts with region still containing TM (black bracket).

To expose the TM, the first triangular-shaped quadrant was oriented with the corneoscleral juncture facing the surgeon and the cornea away from the dominant hand (pointing left for a right-handed surgeon). Two insect dissecting pins (arrows; Figure 2C), one in the cornea and one in the sclera, on either side of the proximal limbus, were used to fix the tissue to the wax. As seen in Figure 2C, the pars plicata of the ciliary body and attached iris root overlie the TM, and needed to be removed first. This was accomplished using jeweler’s micro-forceps, custom-sharpened with 800 and 1500 grit sandpaper so that the tips of the forceps precisely meet. A video showing how the forceps are sharpened is included (Figure 3).

Figure 3. Movie showing how the jeweler’s forceps are sharpened.

Holding the forceps at about 45 degrees to the dissecting surface, the posterior aspect of the ciliary body/iris complex was grasped proximally, between the dissecting pins, and gently pulled away from the cornea at an angle of about 45 degrees from the limbus, using a Weck-Cel eye spear in the non-dominant hand to help stabilize the tissue. Grasping the posterior edge of the ciliary processes where it attaches to the sclera generally allowed simultaneous removal of iris and ciliary body, leaving the underlying TM in place. For RNA analyses, the removed ciliary body and iris tissue from one quadrant was placed in a RNase-free Eppendorf collection tube containing 300 μl TRIzol and stored at −80 °C.

Once exposed, the TM in these pigmented animals could be identified as an irregularly, lightly pigmented band of tissue (asterisk; Figure 2D). This lay between the heavily pigmented Schwalbe’s line anteriorly (white arrows) and the remnants of the ciliary body insertion posteriorly (white arrowheads). Once identified, using forceps reserved for TM dissection, proximal TM was gently grasped, lifting it away from the posterior wall of Schlemm’s canal and pulling it along the length of the meshwork (arrow; Figure 2E), again using a Weck-Cel eye spear to help stabilize the tissue. This maneuver was generally easier if the tissue was turned approximately 30° so that the forceps blades could be oriented perpendicular to the TM, allowing a better grasp of the friable, 1 – 2 cell layer thick tissue. In addition, the TM periodically had to be re-grasped over a single quadrant. Once dissected, the white outer wall of Schlemm’s canal was exposed (asterisk; Figure 2E). To highlight these differences, Figure 2F shows where the TM has been removed to reveal the glistening, white outer wall of Schlemm’s canal (white bracket), and where the TM is still to be dissected (black bracket). A video showing each step of the microdissection is included (Figure 4). Once surgeons are experienced, the entire dissection process takes approximately 30 minutes for one rat eye.

Figure 4. Movie showing the microdissection of rat trabecular meshwork tissue.

Following removal of TM from the first quadrant, tissue (still adherent to the forceps tips) was transferred to an RNase-free Eppendorf collection tube containing 300 μL TRIzol. This was performed by gently swishing the forceps tip in the solution until the TM came free from the forceps, determined by inspection through the dissecting microscope. TM from the remaining three quadrants was then dissected and transferred in similar fashion to the same collection tube. The tube was then closed and stored at −80 °C for further analysis. To limit cross-contamination between eyes, all instruments were sprayed with RNAzap, wiped with alcohol wipes and rinsed in RNase-free water.

3.2. Histopathology

To assess TM angle tissue pre-and post-dissection, quadrants of anterior segments with ciliary body, iris, and lens removed (as described above), were cut in half and fixed in 4% paraformaldehyde for 24 hours (8 eyes from 4 rats). In some cases, TM had also been removed. Fixed tissues were embedded in paraffin and 5 μm radial cross-sections of the anterior chamber angle were cut. Sections were stained with hematoxylin-eosin (H&E) and imaged using an Olympus BX51 microscope with a DP71 digital microscope camera. These tissue sections showed that, where the ciliary body and iris had been carefully removed, the TM remained (Figure 5A) and that TM dissection resulted in removal of the majority of the TM (Figure 5B). In addition, we immunostained sections with an α-smooth muscle actin (αSMA) monoclonal antibody, which strongly immunolabels mouse TM and ciliary muscle (Overby et al., 2014). Images were acquired using an Olympus Fluoview FV1000 confocal microscope. Similar to the mouse, rat TM had strong immunoreactivity against αSMA before TM dissection (Figure 5C). However, following TM dissection (Figure 5D), the majority of immunostaining was gone, with minimal immunoreactivity remaining primarily near where the TM attaches to the sclera and cornea. The H&E and αSMA immunostained sections clearly support the near-total removal of the TM during microdissection.

Figure 5. Histology and immunohistochemistry of rat anterior segments.

(A, C) Anterior segment tissue pre-TM dissection. (B, D) Anterior segment tissue after TM microdissection. Representative images from (A, B) H&E stained paraffin sections and (C, D) α-smooth muscle actin immunostained (green) sections. DAPI stained the nucleus (blue). Note the minimal staining in D where the TM attaches to the sclera (arrow), indicating that the majority of the TM tissue was removed. TM= trabecular meshwork; SC=Schlemm’s canal; SL=Schwalbe’s line. Scale bar = 20 μm.

3.3. RNA isolation and cDNA synthesis

TM samples, which also contain JCT and inner wall of Schlemm’s canal, and ciliary body-iris (CBI) samples (n=6 from 5 rats), stored at −80 °C in 300μl TRIzol, were thawed on ice. Samples were transferred to a Dounce homogenizer, homogenized until tissues were no longer distinguishable, and transferred into a fresh RNAse-free 1.5ml microcentrifuge tube on ice. The tube and homogenizer were washed with an additional 200μL of TRIzol. Two hundred μL each of phenol and chloroform were added to each sample, mixed, followed by centrifugation at 14,000 rpm for 10 minutes at room temperature. Two phases formed: an upper aqueous phase containing the RNA, and a lower organic phase. The top aqueous phase (~200μL) was transferred to a new tube and kept on ice. To ensure maximum RNA yield, 300μL RNase-free Tris-EDTA buffer was added to the lower organic phase, mixed, and centrifuged as above. The top layer (~300μL) was combined with the previously transferred aqueous phase (~200μl) in a 1.5ml tube. An equal volume of chloroform was added, the sample was mixed and centrifuged. The aqueous layer was transferred to a new tube and the chloroform extraction was repeated one more time. The aqueous phase (~500μL) was again transferred to a new tube and an equal volume of diethyl ether was added, which aided removal of residual phenol or chloroform (Moore, 2001). After mixing and centrifuging, the top layer was discarded carefully without disturbing the interface. The tubes containing the bottom aqueous layer (~400μl) were placed in a chemical fume hood with their lids open for 20 minutes to ensure evaporation of residual diethyl ether.

To reduce loss of RNA during the precipitation step, glycogen was added as an inert carrier to each sample. For this, RNA sample volume was estimated and one tenth volume of 3M sodium acetate, 0.15 mg/mL glycogen, and 2.5X volume 100% ethanol was added. After mixing well, samples were placed at −80 °C overnight and then centrifuged at 14,000 rpm at 4°C for 30 minutes. The supernatant was discarded, 700μL of 70% ethanol was added, and the sample was mixed and centrifuged at 14,000 rpm at 4°C for 10 minutes. The ethanol wash step was repeated. With the tube lids open, the samples were rotated in a vacufuge at 45°C for 10 minutes. Nuclease-free water (16μL) was added and samples were placed in a 50°C water bath for 10 minutes until the RNA pellet was completely dissolved. RNA concentration was measured using a NanoDrop One. To eliminate genomic DNA, samples were digested with DNase I by adding 1/10^th volume of 10x DNase I reaction buffer and 1 unit of DNase I per 1–2μg of RNA. Samples were incubated at 37°C for 15–30 minutes. The DNase I was inactivated by adding 1μL of 0.05M EDTA and samples were heated at 65–75°C for 5–10 minutes. In all, 6 TM and 6 CBI samples were measured, which originated from 5 biological replicates. The mean total RNA concentration from TM from one whole eye was 39.8 ng/μl (range: 17.5–88.1 ng/μl; yield: 280–1410 ng), while the mean total RNA concentration from CBI from one quadrant was 73.5 ng/μl (range: 46.2–115.2 ng/μl; yield: 1478–3686 ng).

To generate cDNA, we followed the SuperScript^™ III reverse transcriptase manufacturers protocol (ThermoFisher) with some minor adjustments. Each reaction contained 500ng total RNA, 1μL oligo(dT)₁₇, 1μL 10mM dNTP mix, and sterile water to 15μL. The samples were heated at 65°C for 5 minutes and 4°C for 1 minute. The following were added to each tube: 5μL 5x first-strand buffer, 2.5μL 0.1M dithiothreitol (DDT), 1μL RNaseOUT^™ recombinant RNase inhibitor, 1μL SuperScript^™ III RT, and 0.5μL nuclease-free water for a total volume of 25μL. Samples were incubated at 25°C for 10 minutes, then at 50°C for 1 hour and 70°C for 15 min.

3.4. TaqMan assays

To evaluate the quality of TM microdissection, we performed quantitative RT-PCR on RNA isolated from TM and CBI tissues. We selected genes reported to be enriched in each of these tissues in human and/or rodent eyes, which were identified by single-cell RNAseq and other gene expression studies (Borras, 2003; Liton et al., 2005; Patel et al., 2020; Stone et al., 1997; van Zyl et al., 2020). We first selected genes that show high expression in TM (Tomarev et al., 2003), but low in CBI. These included matrix gla protein (Mgp), myocilin (Myoc), and chitinase-3-like protein 1 (Chi3L1). We also selected genes that are highly expressed in CBI, but low in TM. These included desmin (Des) and myosin-11 (Myh11).

To ensure enough template was available for all qPCR reactions, we preamplified the cDNA. The preamplication reaction, containing 25μL 2X TaqMan preAmp master mix, 12.5μL 0.2X pooled assay mix, 200 ng cDNA, and nuclease-free water up to 50μL, was amplified as follows: 95°C for 10 minutes, 10 cycles of 95°C for 15 seconds and 60°C for 4 minutes, and 99°C for 10 minutes, with a 4°C hold. Preamplification products were diluted 1:5 for the qPCR reaction. Each well of the 96-well plate contained 0.5μL 20X TaqMan primer-probe assays, 4.5μL preamplified cDNA, and 5μL 2X Fast Advanced master mix. The amplification protocol was as follows: 50°C for 2 minutes, 95°C for 20 seconds, and 60 cycles of 95°C for 20 seconds and 60°C for 20 seconds. Each PCR reaction was run in triplicate, and the raw Ct values were averaged.

To calculate relative gene expression in the two tissues (TM versus CBI), TM biomarkers (Mgp, Myoc, Chi3l1) were normalized to one CBI reference sample, while CBI biomarkers (Des, Myh11) were normalized to one TM reference sample. The following calculations were performed for all genes studied, but we will use Mgp as an example. First, ΔC_t values were calculated by deducting β-actin Ct values from Mgp Ct values. Mgp ΔC_t of one CBI sample was used as a reference. Then, Mgp ΔΔC_t values for all TM samples, as well as all the remaining CBI samples, were calculated against this one reference CBI sample. Fold change for each individual sample was then calculated using 2^^(−ΔΔCt) and the mean fold change for the reference tissue group (CBI for TM samples) was determined. All 2^^(−ΔΔCt) values were divided by the mean of the reference group. Finally, the mean (± standard error) relative fold expression of Mgp in TM and CBI was plotted. These calculations were then repeated for the remaining TM biomarker genes (Myoc and Chi3l1). To determine the relative expression of CBI biomarkers (Des and Myh11), the same steps were followed except the ΔC_t of one TM sample was used as a reference sample to which all the other samples were compared, and the mean fold change of the TM samples was used as the reference group.

For statistical analyses, fold change was transformed to log2(2^^(−ΔΔCt)), which provides a more robust, symmetrical, and accurate method for calculating P-values in linear mixed-effects models (Livak and Schmittgen, 2001). The Kenward-Roger method was used to estimate degrees of freedom (Kenward and Roger, 1997). P<0.05 was considered significant. All computations were done using R statistical language (www.R-project.org).

As shown in Figure 6, Mgp, Myoc, and Chi3l1 genes were highly expressed in TM tissue, as expected, while they showed low expression levels in CBI tissue. Conversely, Des and Myh11, genes that were expected to be biomarkers of CBI tissue, showed high expression in CBI and very low expression in TM tissue. These results indicate that our dissection technique yielded samples that were enriched for TM while avoiding significant cross-contamination with CBI, and vice versa.

Figure 6. TaqMan PCR to quantitate TM and CBI biomarkers in each tissue.

(A) Mgp, Myoc, and Chi3l1 are TM biomarkers, while (B) Des and Myh11 are CBI biomarkers. Results are from n=6 eyes from n=5 biological replicates. Mgp, p=0.033; Myoc, p=0.0001; Chi3l1 and Des, not significant; Myh11, p=0.0025.

4. Potential Pitfalls and Troubleshooting

4.1. Animals

We highly recommend using pigmented animals, such as Brown Norway rats, for this procedure. Pigmentation of the iris and ciliary body allows visualization of Schwalbe’s line and the attachment of the ciliary body to the sclera. Together, these define the location of the TM. Also, the light pigment speckling of the TM is highly beneficial; this allows one to identify the presence of the TM, where to begin dissecting, and to detect breaks in the TM that might occur mid-dissection. In addition, the appearance of the white outer wall of Schlemm’s canal helps confirm that the TM has been removed. We have not attempted this in Albino rats. However, for the above reasons, we anticipate that accurate removal of the TM will be more difficult in Albino rats. It may be possible to use younger adult rats. Rats should be old enough that their TMs are fully developed and not soft, which makes it harder to grasp the TM with the forceps. We recommend using adult rats to perfect the microdissection procedure before moving to younger animals. This microdissection technique has not been adapted to the mouse, which would be difficult due to their much smaller size.

4.2. Microdissection

The most significant problem arises if the forceps are either too blunt or do not meet precisely at their tips. Either of these problems makes it impossible to grasp the TM, without engaging surrounding tissues or the posterior wall of Schlemm’s canal. Both can be overcome by careful sharpening the forceps under a dissecting microscope with sandpaper (see Figure 3).

Throughout dissection, the forceps tips need to be pointed precisely at/over the TM, to avoid engaging and removing CBI remnants at the edge of Schlemm’s canal (see Figure 2D). In addition, the initial tendency is to grasp too much tissue too deeply, resulting in grabbing the outer wall of Schlemm’s canal, which resists peeling. If this occurs, use the forceps tips to engage only the superficial, lightly pigmented TM. This sometimes requires higher magnification of the dissecting microscope. While it is ideal to remove the TM in a single piece, this was often difficult given its fragile nature, consisting of only 2–3 lamellae. Generally, TM removal is also aided by orienting the forceps perpendicular to the meshwork, to allow the posterior forceps blade to provide support for the TM as it is being stripped away. Even with this, it is often necessary to periodically re-grasp a new TM edge of TM to complete the dissection. While occasionally it is necessary to pause in order to transfer material from the forceps tip to collection tubes and re-orient the tissue, after practice, transfer of material can generally be done once for each quadrant of TM.

It is also essential to keep the tissue properly hydrated. Heat from the microscope lights can quickly dry out tissue pieces. If too dry, the TM becomes brittle and breaks frequently. For this reason, we recommend adding a couple of drops of PBS to cover quadrants that are awaiting dissection. However, the TM becomes more difficult to identify and grasp definitively if the tissue is too wet, resulting in difficulty peeling it away from Schlemm’s canal. During dissection, either slightly drying or wetting the tissues with the aid of a Weck-Cel eye spear can help to stabilize the tissue at the corneal side of the TM.

TM tissue naturally adheres to the forceps and it can be difficult to transfer TM to the TRIzol. Occasionally, a 25-gauge needle can be used to gently scrape unusually adherent material off of the forceps and into the TRIzol.

4.3. RNA isolation

There remains the possibility of getting cross-contamination from adjacent tissues, or inadvertently introducing RNAases. To minimize cross-contamination while dissecting an eye, two pairs of forceps are recommended. One is used to peel off the ciliary body and iris, while a second pair is used exclusively for TM microdissection. To minimize cross-contamination in between eyes from different animals, all instruments are sprayed with RNAse Zap, wiped with 70% alcohol swabs, and then rinsed with sterile water. It is also important to use RNAase-free plasticware for tissue collection.