Global phosphotyrosinylated protein profile of cell-matrix adhesion complexes of trabecular meshwork cells

Abstract

Dysregulation of the mechanical properties and cell adhesive interactions of trabecular meshwork (TM) are known to impair aqueous humor drainage and elevate intraocular pressure in glaucoma patients. The identity of regulatory mechanisms underlying TM mechanotransduction, however, remains elusive. Here we analyzed the phosphotyrosine proteome of human TM cell-extracellular matrix (ECM) adhesion complexes, which play a key role in sensing and transducing extracellular chemical and mechanical cues into intracellular activities, using a two-level affinity pull-down (phosphotyrosine antibody and titanium dioxide beads) method and mass spectrometry. This analysis identified ~1,000 tyrosine-phosphorylated proteins of TM cell-ECM adhesion complexes. Many consensus adhesome proteins were found to be tyrosine phosphorylated. Interestingly, several of the phosphotyrosinylated proteins found in TM cell-ECM adhesion complexes are known to be required for podocyte glomerular filtration, indicating the existence of molecular parallels that are likely relevant to the shared fluid barrier and filtration functions of the two mechanosensitive cell types.

INTRODUCTION

Aqueous humor, a clear fluid secreted by the ciliary epithelium, enters the anterior chamber of the eye, circulates, and nourishes avascular tissues, including the lens, cornea, and trabecular meshwork (TM). This fluid then passively exits the anterior chamber through the trabecular or conventional pathway in parallel with the uveoscleral or unconventional pathway. The trabecular pathway represents the major aqueous humor (AH) outflow pathway and consists of the TM, juxtacanalicular tissue (JCT), and Schlemm’s canal (SC). After traversing the TM and SC, the AH drains into the episcleral venous system (5). The balance between AH inflow via the ciliary epithelium and outflow through the trabecular and uveoscleral pathways determines intraocular pressure (IOP) (5). Physiologically, the TM, JCT, and SC serve to maintain the barrier or resistance to passive outflow of AH from the anterior chamber such that the IOP in the anterior chamber is optimal for attainment of the convex corneal curvature required for light focusing and vision. In glaucoma patients, however, increased resistance to AH drainage through the TM, JCT, and inner wall of SC leads to ocular hypertension (5). Elevated IOP caused by impaired AH drainage through the TM is recognized to hasten optic nerve atrophy and retinal ganglion cell death in glaucoma patients who can go irreversibly blind if not treated (26, 55). Although external cues (both chemical and physical) regulating cell adhesive interactions, actin filament organization, contractile properties and mechanosensing transcriptional activity, and extracellular matrix production and organization have all been reported to influence AH outflow through the trabecular pathway, the molecular mechanisms involved in AH filtration and resistance are poorly understood at large (15, 38, 49, 51). This knowledge is of paramount importance not only for understanding the etiology of ocular hypertension but also for developing efficacious and targeted therapy for glaucoma (38).

In experimental models of ocular hypertension, increased levels of TGF-β, endothelin-1, connective tissue growth factor, lysophosphatidic acid, and glucocorticoids are recognized to impair AH outflow in association with increased cell adhesive interactions, actin stress fiber formation, and extracellular matrix (ECM) production and stiffness (5, 37, 48). Pharmacological agents targeting regulators of actin cytoskeletal organization and cell adhesive interactions, including Rho GTPases, Rho kinase, integrins, protein kinase C, and myosin II, and perfusion studies with matrix metalloproteinases have been demonstrated to lower IOP, emphasizing the importance of cell-ECM and cell-cell adhesive interactions in modulating resistance to AH outflow (15, 38, 49, 51). Therefore, identifying the molecular mechanisms controlling cell-ECM adhesive interactions, and characterizing how external cues deriving from the ECM, mechanical stretch, and growth factors are transduced into intracellular responses through the adhesion complexes in TM cells is critical for our understanding of normal and dysregulated AH outflow.

Cell-ECM adhesive interactions mediated by integrins and other cell surface proteins initiate a wide array of signaling pathways that control vital cellular functions, including proliferation, differentiation, survival, and migration (22, 56). Cell adhesion complexes, which bridge the intracellular cytoskeleton and the ECM, act as key force-sensing and -transducing units in cells (21, 58). Importantly, dysregulated cell adhesive interactions and actin cytoskeletal organization are associated with disease etiologies, including fibrosis, hypertension, chronic kidney disease, and cancer (46, 56). Cell-ECM adhesive complex formation is orchestrated through recruitment of diverse proteins to the sites of adhesion, and protein phosphorylation is recognized to play an important role in cell-ECM adhesion complex function and stability (16, 42, 58). The mass-spectrometry-based proteomics analysis of cell-ECM adhesive complexes from different cell types has begun to identify a wide range of cytoskeletal, adhesion, scaffolding, kinases, phosphatases, ECM, integrins, and signaling proteins, and to unravel their role in both physiological and pathological processes (21, 56). We therefore applied a proteomics-based approach to characterizing the global phosphotyrosinylated protein profile of TM cell-ECM adhesive complexes. For this we isolated the cell-ECM adhesion complexes from primary cultures of human TM cells grown in plastic dishes, followed by enrichment and identification of tyrosine-phosphorylated proteins (phosphotyrosine adhesome) from the TM cell-ECM adhesive complexes using phosphotyrosine antibody and titanium dioxide affinity pulldown approaches and liquid chromatography tandem-mass spectrometry, respectively. The results of this study not only unraveled the identity of abundant and minor regulatory proteins of TM cell-ECM adhesion complexes but also enabled us to identify the presence of several proteins involved in podocyte barrier activity, indicating that the two cell types likely share certain parallel cell adhesive interaction- and actin cytoskeletal organization-based molecular mechanisms for fluid filtration and mechanosensing functions.

MATERIALS AND METHODS

Isolation of cell-ECM adhesion complexes from human trabecular meshwork cells.

The primary TM cell cultures described in these studies were used at passage 3 and were generated from human TM tissue isolated from the deceased donor corneal rings as we described earlier (34), with proteomic sample 1 and 2 being derived from 19-yr-old and 72-yr-old human donor eyes, respectively. Health Insurance Portability and Accountability Act (HIPAA) compliance guidelines were adhered for the use of human donor tissue. Donor specimens are exempt from the Department of Health and Human Services regulations at 45 CFR Part 46. The Institutional Review Board (IRB) protocol (00050810) has been approved by the Duke University Medical Center IRB review board members. TM cells were plated in 150 × 25-mm Falcon culture plastic dishes (Corning) and cultured at 37°C with 5% CO₂, in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS) and penicillin (100 U/mL)-streptomycin (100 µg/mL)-glutamine (4 mM). Cells were grown to confluence and then maintained for an additional 14 days with two culture media changes per week, before isolation of cell-ECM complexes.

Before decellularization of the monolayer, cells were incubated with 2 mM ATP (cat. no. A26209; Sigma-Aldrich, St. Louis, MO) for 15 min at 37°C, followed by aspiration of culture media and three rinses with 1× phosphate-buffered saline (PBS), pH 7.4. Cells were permeabilized with 15 mL of 0.2% Triton X-100/PBS and incubated at room temperature for 10 min. Permeabilized cells were carefully aspirated at low suction followed by the slow addition of 15 mL of 0.3% ammonium hydroxide to each cell culture plate and incubation for 5 min at room temperature. The plates were checked under a phase contrast microscope to ensure that there were no attached cells remaining before being rinsed three times with 1× PBS, pH 7.4. The adherent cell-ECM complexes were scraped into 500 μl of 50 mM ammonium bicarbonate buffer, pH 8.0, containing 8 M urea and protease/phosphatase inhibitors, as we described earlier (28). Cell-ECM extracts from 10 cell culture dishes were pooled for the affinity pulldown of tyrosine-phosphorylated proteins. Pooled samples were sonicated under ice-cold conditions before enrichment of tyrosine-phosphorylated proteins and proteomics analysis.

Sample preparation for proteomics analysis.

The above described protein samples in 8 M urea were submitted to the Duke Proteomics Core Facility, with the rest of sample processing being performed by core facility personnel. Briefly, protein samples were subjected to three rounds of probe sonication for 10 s each with an energy setting of 30% and then centrifuged at 3,000 g at 4°C for 5 min. Protein concentration of supernatant samples was determined by the Bradford assay, with total protein content for sample 1 and 2 being 10.4 and 3.5 mg, respectively. Protein samples were reduced for 45 min at 32°C with 10 mM dithiothreitol, alkylated for 30 min at room temperature with 25 mM iodoacetamide, and then diluted to 1.6 M urea using 50 mM ammonium bicarbonate. Trypsin (cat. no. V5111, sequencing-grade enzyme from Promega, Madison, WI) was added at a 1:25 enzyme-to-total protein ratio, and digestion was allowed to proceed for 18 h at 32°C. Samples were acidified with trifluoroacetic acid (TFA), centrifuged at 3,000 g for 5 min at 4°C, subjected to C18 solid-phase extraction clean up (Sep-Pak, 500-mg bed; Waters Corp.), and lyophilized to dryness.

Antibody affinity-based phosphopeptide enrichment.

The above described lyophilized tryptic peptide samples were resuspended in 700 μL 1× IAP buffer (50 mM MOPS, pH 7.2, 10 mM sodium phosphate, and 50 mM NaCl; Cell Signaling Technology) using vortexing and brief bath sonication. Prealiquoted Phos-Tyrosine PTMScan beads (cat. no. 8803; Cell Signaling Technology, Danvers, MA) were diluted with 4 × 1 mL of 1× PBS buffer. Resuspended peptides in IAP buffer were then transferred directly onto beads and incubated for 2 h at 4°C using end-over-end mixing. After gentle spinning to settle the beads, supernatants were removed. The IAP resins containing the enriched phospho-peptidome were then washed with 1 mL of IAP buffer three times and one time with 0.1× IAP buffer. After the supernatant was removed, antibody-bound Tyr-phosphorylated peptides were eluted by incubation with a 50-μL aliquot of 0.15% TFA in water for ~10 min at room temperature, tapping gently on the bottom of the tube a few times during elution to ensure mixing. Beads were eluted a second time with 45 μL of 0.15% TFA in water. Combined eluents were lyophilized to dryness and further enriched using 10 μL GL Bioscience TiO₂ spin tips (cat. no. 5010–21310; GL Sciences) per the manufacturer’s protocol using 600-μg capacity tips, 80% MeCN (acetonitrile), 1% TFA binding and equilibrium buffer, 20% MeCN, and 5% aqueous ammonia elution buffer. Eluted phosphopeptides were lyophilized to dryness and resuspended in 12 μL of 10 mM citric acid in 1% TFA/2% acetonitrile containing 10 fmol/μL yeast alcohol dehydrogenase.

Liquid chromatography tandem mass spectrometry.

Liquid chromatography-MS/MS was performed on 4 μL of each sample, using a nanoAcquity ultra performance liquid chromatography (UPLC) system (Waters Corp., Milford, MA) coupled to a Thermo QExactive HF high-resolution accurate mass tandem mass spectrometer (Thermo Fisher Scientific, Waltham, MA) via a nanoelectrospray ionization source. Briefly, the sample was first trapped on a Symmetry C18 20 mm × 180 μm trapping column (5 μL/min at 99.9:0.1 vol/vol water-acetonitrile), after which the analytical separation was performed using a 1.7-μm Acquity BEH130 C18 75 μm × 250 mm column (Waters Corp.) with a 90-min linear gradient of 3 to 30% acetonitrile with 0.1% formic acid at a flow rate of 400 nL/min and a column temperature of 55°C. Data collection on the QExactive HF mass spectrometer was performed in a data-dependent acquisition mode with a r = 120,000 [at mass-to-charge ratio (m/z) 200] full MS scan from m/z 375–1,600 with a target AGC value of 3 × 10⁶ ions followed by 12 MS/MS scans at r = 15,000 (at m/z 200) at a target AGC value of 5 × 10⁴ ions. A 20-s dynamic exclusion was employed to increase depth of coverage. The total analysis cycle time for each sample injection was ~2 h.

Following UPLC-MS/MS analysis, the MS/MS data were searched against the SwissProt human database with additional proteins, including yeast alcohol dehydrogenase, bovine serum albumin, and an equal number of reversed-sequence “decoys” used to enable false discovery rate determination. Mascot Distiller and Mascot Server (version 2.5; Matrix Sciences) were used to produce fragment ion spectra and to perform the database searches. Database search parameters included fixed modification on cysteine (carbamidomethyl) and variable modifications on methionine (oxidation), serine/threonine/tyrosine (phosphorylation). Search tolerances were 5 ppm for precursor ions and 0.02 Da for product ions with full trypsin specificity rules. All searched spectra were imported into Scaffold (version 4.3, Proteome Software), and scoring thresholds were set to achieve a peptide false discovery rate of 1% using the PeptideProphet algorithm. Modification localization probabilities were calculated within ScaffoldP using the AScore algorithm (3).

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the data set identifier PXD015377 (http://www.ebi.ac.uk/pride/archive/projects/PXD015377).

Immunofluorescence.

Human TM cells were grown to confluency on 2% gelatin-coated glass coverslips (12-well plates) as described above. Cells were washed two times with PBS and then fixed with 4% paraformaldehyde for 15 min. After completing fixing, washing, permeabilization, and blocking as described earlier by us (34), cells were incubated with primary antibodies against ZO-1, junctional protein associated with coronary artery disease (JCAD), MAGI-2, phospho-Rho kinase 2 (ROCK2), NEPH1, lipoma preferred partner (LPP), P120 catenin, and FAT1 for 2 h at room temperature. After a thorough washing, specimens were incubated with the appropriate Alexa Fluor-conjugated secondary antibodies for 2 h at room temperature as described previously (34). The details of all antibodies (source and dilution) used in this study are given in Supplemental Table S1 (all Supplemental material is available at https://doi.org/10.6084/m9.figshare.12192600.v1). Coverslips were mounted onto glass slides using Shandon Immu-Mount (Thermo Fisher Scientific) and then viewed and imaged using a Nikon Eclipase 90i confocal laser-scanning microscope.

Immunoblotting.

The cell-ECM complex fractions (10 μg) derived from human TM cells of 52-yr-old and 66-yr-old donor eyes as described above were resolved on 4 to 20% gradient sodium dodecyl sulfate (SDS)-polyacrylamide gels (Bio-Rad) and transferred to nitrocellulose membranes, as we previously described (28). Membranes were probed with desired primary antibodies for 18 h at 4°C (details of antibodies are described in Supplemental Table S1). After being washed, nitrocellulose membranes were incubated with appropriate horseradish peroxidase-conjugated secondary antibodies (1:5,000 dilution; Jackson Immuno Research) for 2 h at room temperature, and immunopositive protein bands were detected with an enhanced chemiluminescent substrate as we described earlier using Bio-Rad ChemiDoc Touch Imaging System (28).

RESULTS

Identification and characterization of predominant tyrosine-phosphorylated proteins in cell-ECM adhesion complexes isolated from human TM cells.

Having recognized that the cell-ECM adhesion complexes (focal adhesions) and focal adhesion-associated components, including actomyosin cytoskeleton, integrins, ECM proteins, and cell-cell junctional proteins, influence AH outflow (15, 33, 38, 49, 51), we initiated studies to gain a broader understanding of the molecular mechanisms regulating cell-ECM adhesion complex function in TM cells. To this end, cell-ECM adhesion complexes were isolated from confluent human TM primary cell cultures grown in large plastic cell culture plates for 2 wk. Before monolayer decellularization, cultures were treated with 2 mM ATP for 15 min to ensure that ATP is not a limiting factor for extracellular and intracellular tyrosine kinases that mediate phosphorylation of ECM, cell surface, and cell adhesive complex proteins. The total protein content of decellularized material was 10.5 mg for sample 1 and 3.5 mg for sample 2, both of which were similarly and independently generated from TM cells derived from a 19- and a 72-yr-old human donor, respectively. The enriched phosphotyrosine peptides (isolated by affinity pulldown of tryptic digests using phosphotyrosine antibody and TiO2, as described in materials and methods) derived from samples 1 and 2 were subjected to LC/MS/MS analysis using a nanoAcquity UPLC system coupled to a Thermo QExactive HF high-resolution accurate mass tandem mass spectrometer.

LC/MS/MS analysis of phosphotyrosine-containing peptides isolated by immuonoaffinity pulldown of tryptic digests from sample 1 and 2 identified a total of 1,302 and 494 individual proteins (based on exclusive spectrum count), respectively, reflecting a concentration-dependent outcome with respect to the total number of proteins identified under the LC/MS/MS settings and identification criteria described in materials and methods. Of the individual proteins identified, 1,004 and 456 were confirmed to be phosphotyrosinylated proteins in sample 1 and 2, respectively. Refer to the data repository, which contains individual raw files and scaffold files for sample 1 and 2, with the option to assess total spectrum count, exclusive spectrum count, exclusive unique peptide count, exclusive unique spectrum count, percentage of the total spectrum, and percent coverage (data were deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the data set identifier PXD015377). Protein scoring thresholds were set to achieve a peptide false discovery rate of 1%. The results confirmed that ~77% and ~92% of the total identified proteins were tyrosine phosphorylated in samples 1 and 2, respectively. Supplemental Tables S2 and S3 (Excel format) list all of the individual phosphotyrosine proteins identified in samples 1 and 2, respectively, along with accession ID, protein name, Ascore, peptide sequence, peptide score, Mascot threshold, Mascot score, variable modifications, actual mass, and observed mass.

Although some proteins could not be confirmed to be phosphorylated in both samples 1 and 2, all proteins identified in both samples could be classified as participants in physiological processes and molecular functions involving cell adhesion, scaffolding, adherens junctions, cytoskeleton organization, and intracellular signaling. Moreover, this analysis also identified some peptides that were both tyrosine and serine/threonine phosphorylated (see Data Repository-Scaffold files). In both sample 1 and 2, phosphoserine and phosphothreonine peptides accounted for less than 6% and 2%, respectively, relative to phosphotyrosine peptides, confirming that the procedure used in this study enriches primarily for tyrosine-phosphorylated (p) proteins from cell-ECM complexes of TM cells. For sample1, there were a total of 141 p-serine, 42 p-threonine, and 2,179 p-tyrosine peptides identified, while sample 2 yielded a total of 30 p-serine, 11 p-threonine, and 643 p-tyrosine peptides. The pie charts in Fig. 1, A and B depict the relative percentages of the three phospho amino acid peptides detected in TM cell-ECM complexes isolated from sample 1 and 2, respectively.

Fig. 1.

A pie chart representation of the relative percent of phosphorylated (P) tyrosine, serine, and threonine peptides detected in sample 1 (A) and 2 (B) of human trabecular meshwork (TM) cell-extracellular matrix (ECM) complexes. In both samples, more than 90% of the total detected phosphopeptides in TM cell-ECM complexes were found to be tyrosine phosphorylated.

For authenticating the ability of our protocol to correctly identify tyrosine-phosphorylated proteins of the cell-ECM complexes of TM cell samples analyzed in this study, we examined the site-specific phosphorylation of well-characterized adhesome proteins, including p-Y397 focal adhesion kinase (FAK), p-Y722 ROCK2, p-Y188 paxillin, p-Y127 talin, p-Y783 β₁-integrin, p-Y822 vinculin and p-Y247, p-Y265 and p-Y313 connexin-43 (Cx43), among others. All proteins selected for this assessment were confirmed to be tyrosine phosphorylated at the expected residues in TM samples, based on mass spectrometry identification (Supplemental Tables S2 and S3).

Of the many phosphotyrosinylated peptides identified in our analyses of proteins found in TM cell-ECM complexes, those phosphorylated at previously unrecognized sites will be added to the phosphosite.org database. A few examples of new previously unidentified phosphotyrosine sites (based on the phosphosite.org database) from well-characterized integrin adhesome proteins include paxillin p-Y58, talin 1 p-Y199 and p-Y216, talin 2 p-Y201 and p-Y218, LPP p-Y58, filamin A p-Y1041 and p-Y2077, actinin alpha 4 p-Y338, IQGAP1 p-Y192, and ponsin 1 p-Y165 (Supplemental Tables S2 and S3).

Major subclasses of phosphotyrosine proteins identified in cell-ECM adhesion complexes of TM cells.

Because the primary goal of this study was to identify proteins with a potential role in regulating the formation and function of cell-ECM adhesion complexes in TM cells, we further characterized the tyrosine-phosphorylated protein profile of cell-ECM adhesion complexes isolated and identified from human TM cells. This was performed by initially using the gene ontology (GO) tool ShinyGO v0.60 for enrichment of GO terms. Evaluation of the hierarchical clustering tree and network output for biological processes, cellular components, and molecular function revealed enrichment of proteins belonging to functional categories, including cytoskeletal organization, actin filament-based processes, actin cytoskeletal organization, adherens junctions, anchoring junctions, focal adhesions, cell-substrate junctions, cell substrate-adherens junctions, cadherin binding, cell adhesion molecule binding, enzyme binding, and kinase binding. For these analyses, data were filtered using a P value cutoff of 0.05 and the 20 most significant terms. Figure 2 shows one example of protein enrichment carried out for molecular function. Consistent with results generated using the ShinyGO v0.60 tool, the GOrilla tool-based enrichment analysis of GO terms also revealed very similar results for molecular function and cellular components (data not shown). Similarly, functional pathway analysis using KEGG pathway mapper confirmed that most of the phosphotyrosine proteins identified in the cell-ECM adhesion complexes of human TM cells belong to a class or family of proteins involved in cell adhesion assembly, cell-cell junction assembly, and actin cytoskeletal organization or actomysin organization. Figure 3 shows the data output for focal adhesion proteins and the specific phosphotyrosine residues identified for proteins belonging to the focal adhesion pathway.

Fig. 2.

Enrichment analysis of human trabecular meshwork (TM) cell-extracellular matrix (ECM) complex phosphotyrosine proteins for molecular function using the ShinyGO v0.60 tool. Data are shown in enrichment false discovery rate (FDR, left), hierarchical clustering tree (middle), and network (right) format for the 20 most significant terms.

Fig. 3.

Enrichment of proteins involved in focal adhesion formation in the phosphotyrosine fraction derived from trabecular meshwork (TM) cell-extracellular matrix (ECM) complexes, evaluated using the KEGG pathway mapper tool. Enriched phosphotyrosine proteins from TM cell-ECM complexes are shown in red. Specific phosphotyrosine residues from identified focal adhesion pathway proteins present in the phosphotyrosine fraction of sample 1 are shown in blue.

Moreover, most of the TM cell-ECM complex proteins identified in our study were found to belong to the meta-adhesome and literature-curated adhesome (16, 21, 58), with nearly 65% of consensus adhesome proteins being tyrosine phosphorylated (21).

We then examined the phosphotyrosine proteomics data set of human TM cell-ECM complexes for major subclasses of proteins. This analysis confirmed that our data set contained proteins from several subclasses, including the kinases (tyrosine and serine/threonine), phosphatases, channels, integrins, adhesion, inositol, receptors, cadherins, catenins, caveolin, cytoskeleton, ephrin, discoidin, cyclin, E3 ligase, FERM domain, growth factors, heat shock, myosin, junctional, LIM domain, microtubule, mitogen, par, PDZ domain, Rho GTPase, Rap, Ras, SH3 domain, SH2 domain, calponin domain, pleckstrin homology (PH), spectrin, WD domain, ankyrin domain, tyrosine, tight junctions, tensin, RNA, tubulin, mechanosensitive, metalloproteinase, and autophagy (Supplemental Tables S2 and S3). Table 1 lists a few selected functional classes of human TM cell-ECM complex phosphotyrosinylated proteins. Additional proteins were also identified in human TM cell-ECM adhesion complexes (e.g., fibronectin and syndecan-4), uncovering the involvement of a variety of proteins in regulation of cell adhesive interactions. It is noteworthy to point out here that we not only found proteins belonging to integrin adhesome but also identified cadherin adhesion (cadhesome) and contractile proteome (contractome) proteins in TM cell-ECM complexes (57, 59). In addition to actin and actin-binding proteins, microtubule and various intermediate filament proteins were also present in human TM cell-ECM adhesion complexes (Supplemental Tables S2 and S3).

Table 1.

Phosphotyrosinylated proteins of trabecular meshwork cell-extracellular matrix complexes categorized based on function

	Protein
Cell-matrix adhesions	LAP3, LASP1, FHL1, VIM, VCL, CTNNA1, LIMA1, GDI2, PPP1R12A, CNN2, RHOA, ACTN1, LIMS2, FERMT2, ACTB, TNS1, EPB41L2, ITGB5, FAT1, RAB10, CTTN, TRIP6, ADD1, PXN, SLC9A1, NRP1, TRIOBP, CD99L2, CAV1, PDLIM1, GIT1, HSPA8, PPFIBP1, CORO1C, TNS2, ARPC3, PDGFRB, FHL2, ARHGEF2, DOCK7, CNN3, LRP1, AHNAK, VASP, CDC42EP1, FLNC, ARHGAP22, PALLD, CNN1, NECTIN2, ACTN4, AKAP12, MPRIP, NUMB, ANXA1, TES, FLNB, LMO7, TNS3, TLN1, GIT2, IQGAP1, TGFB1I1, CLTC, EPHA2, RPS8, LPP, EGFR, ITGB1, TWF1, DST, GJA1, DAB2, CSRP1, ACTC1, ZYX, RPL8, ITGA5, JAK1, FBLIM1, NEXN, DDR2, CAPN2, ANXA5, HNRNPK, CTNNB1, ADAM9, LIMS1, PDCD6IP, TLN2, SYNPO2, CFL1, JUP, YES1, PLEC, FHL3, FLRT2, SPRY4, PPIA, AFAP1, FLNA, ANXA6, SVIL, PARVA, MPZL1, RPL10A, TGM2, LAYN, HSPA1A, PI4KA, SCARF2, CYFIP1, BCAR1, ASAP3, CAV2, PTK2B, PTK2, PEAK1, CLASP1, CPNE3, SORBS1, JAK2, LPXN, NUP214, LIMD1, PHLDB2, CLASP2, MAPRE2, SRP68, FES, PDLIM7, ARHGAP31, MYH9, NEDD9, EPB41L5, SDC4, PTPN12, AJUBA, DCTN4, SDCBP, ARHGAP24, ZNF185, SORBS2, GAK, NHS, LIMS3, MAPK1, MAPK3
Actin cytoskeleton	ACTB, BAIAP2L1, MYLK, WDR1, CAPZB, ABLIM1, PLS3, WASL, ARPC3, MACF1, SPECC1, DCTN4, ARHGAP32, TMOD3, PSTPIP1, CTTNBP2NL, TWF1, PSTPIP2, ASAP1, TTN, LMOD1, SHROOM1, WIPF2, SYNPO, SYNPO2, ABLIM3, PEAK1, DCTN2, BAIAP2, PAWR, FHL3, PARVA, FLNA, LIMA1, PPP1R12A, LIMCH1, CNN2, ACTN1, MARK2, CTTN, ADD1, PXN, MYO9B, TRIOBP, MYH9, SIPA1L3, ARHGAP21, SH3PXD2A, LPXN, CORO1C, TARS, CNN3, CALD1, AHNAK, MYO1B, PALLD, ACTN4, PDLIM4, MYH10, MPRIP, ANXA1, CDC42BPA, ALDOA, FER, FMN2, EEF1A1, ACTC1, ZYX, FBLIM1, DDR2, PDLIM5, CDC42EP3, PTK2, AFAP1, MYO18A, DAPK1, PDLIM7, SVIL, MYO5A, CD2AP, CDC42BPB, LASP1, FYN, VCL, CTNNA1, ARHGEF5, BCAR1, MYO9A, FERMT2, MYL6, ABL1, TJP1, ACTA2, PPP1R9B, DPYSL3, DBN1, SPTBN1, ARHGEF2, IVNS1ABP, DSTN, PTPN12, CGNL1, ACTR10, MTSS1L, FLNB, DBNL, IQGAP1, ZNF185, BAG3, UTRN, MYO1E, ARHGAP35, MICALL2, HNRNPK, CFL2, CRK, DAPK3, PDCD6IP, CFL1, JUP, YES1, ITSN1, SORBS1, TLN2, EPB41L2, VASP, ABL2, SORBS2, CAPN2, CLIC4, SPTAN1
Cell junction	LAP3, LASP1, BAIAP2L1, FHL1, VIM, VCL, CTNNA1, LIMA1, GDI2, PPP1R12A, CNN2, RHOA, ACTN1, LIMS2, FERMT2, ACTB, TNS1, EPB41L2, ITGB5, FAT1, RAB10, CTTN, TRIP6, ADD1, PXN, SLC9A1, NRP1, TRIOBP, CD99L2, CAV1, PDLIM1, GIT1, HSPA8, EIF4G2, PPFIBP1, CORO1C, KRT18, TNS2, ARPC3, PDGFRB, FHL2, ARHGEF2, DOCK7, CNN3, LRP1, AHNAK, VASP, CDC42EP1, FLNC, ARHGAP22, PALLD, CNN1, NECTIN2, ACTN4, AKAP12, MPRIP, NUMB, ANXA1, TES, FLNB, LMO7, TNS3, TLN1, TMOD3, GIT2, IQGAP1, TGFB1I1, CLTC, EPHA2, RPS8, LPP, EGFR, ITGB1, TWF1, DST, GJA1, DAB2, CSRP1, ACTC1, ZYX, RPL8, ITGA,5 JAK1, FBLIM1, NEXN, DDR2, CAPN2, PDLIM5, ANXA5, HNRNPK, CTNNB1, ADAM9, LIMS1, PDCD6IP, TLN2, SYNPO2, CFL1, JUP, BAIAP2, YES1, PLEC, ANXA2, FHL3, FLRT2, SPRY4, PPIA, AFAP1, FLNA, ANXA6, SVIL, PARVA, MPZL1, RPL10A, TGM2, LAYN, HSPA1A, PI4KA, SCARF2, CYFIP1, BCAR1, PKP2, DLG1, ASAP3, DSP, ARVCF, CAV2, PARD3B, PTK2B, AJUBA, DLG4, PKP4, LIMD1, PARD3, MAGI1, SHROOM1, AMOTL1, PTK2, CTNND2, PEAK1, NECTIN3, SCRIB, MAGI2, CTNND1, PKN2, WDR1, MPP5, CLASP1, CPNE3, SORBS1, JAK2, TJP1, PPP1R13L, SIPA1L3, DNMBP, ARHGAP21, LPXN, TJP2, NUP214, MACF1, PDGFRA, APC, TP53BP2, RAB13, PHLDB2, TNKS1BP1, MPP7, DLG5, FER, PTPRK, EPB41, TBC1D24, CLASP2, MIOS, CDK5, FRS2, CLMP, MAPRE2, SRP68, CLIC4, PPP1CA, PTPRM, HEG1, SYNM, FES, NHS, KAZN, PDLIM7, ARHGAP31, RAI14, GOPC, ARHGEF5, LZTS1, TNK2, ADGRL1, CBARP, MYH9, CDK16, CSK, MPDZ, SH3PXD2A, NEDD9, DBN1, FRMD4B, PIKFYVE, EPB41L5, TANC1, EPHA4, DENND1A, SDC4, PLCG1, PTPN12, CGNL1, PDLIM4, MAP1B, DCTN4, ARHGAP32, DBNL, ABHD17C, ABI1, SDCBP, UNC13C, ABI2, ARHGAP24, FRMD6, ARHGAP17, MINK1, APP, ANK2, USP53, PIK3R1, ZNF185, ADD3, FRMD4A, EPS8, UTRN, SORBS2, PSD3, MYO1E, NCK1, SHANK2, ATP1A1, PRKCD, MICALL2, SYNPO, GRB2, KCTD12, GAK, PRKD1, SHC4, BCR, SIPA1L1, CD2AP, UNC13B, CDC42BPB, DMD, PJA2, ITSN,1 LYN, LIMS3, CDC42BPA, MAPK1, MAPK3, ARHGEF18
Cell substrate adherens junction	LAP3, LASP1, FHL1, VIM, VCL, CTNNA1, LIMA1, GDI2, PPP1R12A, CNN2, RHOA, ACTN1, LIMS2, FERMT2, ACTB, TNS1, EPB41L2, ITGB5, FAT1, RAB10, CTTN, TRIP6, ADD1, PXN, SLC9A1, NRP1, TRIOBP, CD99L2, CAV1, PDLIM1, GIT1, HSPA8, PPFIBP1, CORO1C, TNS2, ARPC3, PDGFRB, FHL2, ARHGEF2, DOCK7, CNN3, LRP1, AHNAK, VASP, CDC42EP1, FLNC, ARHGAP22, PALLD, CNN1, NECTIN,2 ACTN4, AKAP12, MPRIP, NUMB, ANXA1, TES, FLNB, LMO7, TNS3, TLN1, GIT2, IQGAP1, TGFB1I1, CLTC, EPHA2, RPS8, LPP, EGFR, ITGB1, TWF1, DST, GJA1, DAB2, CSRP1, ACTC1, ZYX, RPL8, ITGA5, JAK1, FBLIM1, NEXN, DDR2, CAPN2, ANXA5, HNRNPK, CTNNB1, ADAM9, LIMS1, PDCD6IP, TLN2, SYNPO2, CFL1, JUP, YES1, PLEC, FHL3, FLRT2, SPRY4, PPIA, AFAP1, FLNA, ANXA6, SVI,L PARVA, MPZL1, RPL10A, TGM2, LAYN, HSPA1A, PI4KA, SCARF2, CYFIP1, BCAR1, ASAP3, CAV2, PTK2B, PTK2, PEAK1, CLASP1, CPNE3, SORBS1, JAK2, LPXN, NUP214, LIMD1, PHLDB2, CLASP2, MAPRE2, SRP68, FES, PDLIM7, ARHGAP31, MYH9, NEDD9, EPB41L5, SDC4, PTPN12, AJUBA, DCTN4, SDCBP, ARHGAP24, ZNF185, SORBS2, GAK, NHS, LIMS3, MAPK1, MAPK3
Anchoring junction	LAP3, LASP1, BAIAP2L1, FHL1, VIM, VCL, CTNNA1, LIMA1, GDI2, PPP1R12A, CNN2, RHOA, ACTN1, LIMS2, FERMT2, ACTB, TNS1, EPB41L2, ITGB5, FAT1, RAB10, CTTN, TRIP6, ADD1, PXN, SLC9A1, NRP1, TRIOBP, CD99L2, CAV1, PDLIM1, GIT1, HSPA8, EIF4G2, PPFIBP1, CORO1C, KRT18, TNS2, ARPC3, PDGFRB, FHL2, ARHGEF2, DOCK7, CNN3, LRP1, AHNAK, VASP, CDC42EP1, FLNC, ARHGAP22, PALLD, CNN1, NECTIN2, ACTN4, AKAP12, MPRIP, NUMB, ANXA1, TES, FLNB, LMO7, TNS3, TLN1, TMOD3, GIT2, IQGAP1, TGFB1I1, CLTC, EPHA2, RPS8, LPP, EGFR, ITGB1, TWF1, DST, GJA1, DAB2, CSRP1, ACTC1, ZYX, RPL8, ITGA5, JAK1, FBLIM1, NEXN, DDR2, CAPN2, PDLIM5, ANXA5, HNRNPK, CTNNB1, ADAM9, LIMS1, PDCD6IP, TLN2, SYNPO2, CFL1, JUP, BAIAP2, YES1, PLEC, ANXA2, FHL3, FLRT2, SPRY4, PPIA, AFAP1, FLNA, ANXA6, SVIL, PARVA, MPZL1, RPL10A, TGM2, LAYN, HSPA1A, PI4KA, SCARF2, CYFIP1, BCAR1, PKP2, ASAP3, ARVCF, CAV2, PARD3B, PTK2B, AJUBA, PKP4, LIMD1, PARD3, SHROOM1, PTK2, CTNND2, PEAK1, NECTIN3, SCRIB, CTNND1, MPP5, CLASP1, DLG1, CPNE3, SORBS1, JAK2, LPXN, NUP214, APC, PHLDB2, TNKS1BP1, MPP7, CLASP2, FRS2, MAPRE2, SRP68, PPP1CA, PTPRM, SYNM, FES, KAZN, PDLIM7, ARHGAP31, TNK2, DSP, MYH9, TJP1, NEDD9, FRMD4B, EPB41L5, TJP2, SDC4, PTPN12, DCTN4, SDCBP, ABI2, ARHGAP24, ZNF185, DLG5, MAGI1, FRMD4A, PTPRK, SORBS2, MYO1E, GAK, NHS, LYN, LIMS3, MAPK1, MAPK3
Rho GTPase binding	ARHGEF10, DNMBP, ARHGEF17, CDC42EP1, ARHGAP17, ARHGDIA, EPS8, CDC42EP3, PAK2, FARP2, ARHGEF2, DOCK7, DOCK4, ABI2, DAPK3, FLNA, CDKL5, TRIO, ARHGEF5, PKN2, SOS2, ARHGEF18, DVL1, CORO1C, ARHGEF6, MTSS1L, ARHGEF11, ROCK2, FGD4, IQGAP1, STRIP1, DOCK11, VAV2, ARHGEF40, BCR, ARHGEF12, ITSN2, CDC42BPB, ITSN1, ARHGEF28, CYFIP1, RALBP1, MYO9B, CAV1, OCRL, TRIOBP, DOCK1
Cell adhesion molecule binding	LASP1, BAIAP2L1, GPRC5A, VCL, CTNNA1, LIMA1, PKP2, CNN2, PKN2, PKM, DHX29, EPN2, MARK2, ENO1, DLG1, CAPZB, BZW1, COBLL1, ERC1, RAB10, CTTN, EPS15, CHMP5, ADD1, DOCK9, GOLGA3, HSP90AB1, SH3GLB1, CRKL, TBC1D10A, MICALL1, MYH9, EMD, EHD4, TJP1, EEF1D, PPP1R13L, PLIN3, ZC3HAV1, EIF4H, PDLIM1, HSPA8, EHD1, EIF4G2, PPFIBP1, KRT18, STK38, CLINT1, RARS, DBN1, SPTBN1, STAT1, HDLBP, PRDX1, CNN3, PRDX6, TJP2, TRIM25, CALD1, LRRFIP1, AHNAK, VASP, KTN1, EPS15L1, MACF1, PAICS, CDC42EP1, MYO1B, CC2D1A, MPRIP, NUMB, LDHA, ANXA1, TES, FLNB, DBNL, ABI1, TLN1, FNBP1L, TMOD3, UNC45A, IQGAP1, EPHA2, SERBP1, PHLDB2, EIF2A, EGFR, TNKS1BP1, ALDOA, MPP7, ITGB1, TWF1, BAG3, CAST, ASAP1, NCK1, TAGLN2, STX5, PDLIM5, HNRNPK, GAPVD1, CTNNB1, ATXN2L, PCBP1, FASN, JUP, RPL15, KLC2, BAIAP2, PLEC, PDXDC1, PAK2, SCRIB, ANXA2, PTPN1, FLNA, SND1, SPTAN1, PARVA, CD2AP, CTNND1, GIGYF2, HSPA1A, CLIC1, DDX3X, PI4KA, ITGB5, ARVCF, PKP4, TENM4, CTNND2, NECTIN3, ADAM9, ACTN1, FN1, PPP1CA, PTPRM, PXN, DSP, CTNNAL1, NECTIN2, SDCBP, ITGA5, PTPN11, DMD, LYN, PTPRF, DST, UTRN, ADGRL1, ACTN4
Integrin	ITGA5, ITGB1, ITGB5, ADAM9, ADAM12
Channels	PIEZO2, CLIC1, CLIC4, CBARP, SCN9A
Kinases	FYN, JAK2, ABL1, TYK2, JAK1, PRKCD, AXL, PTK2, YES1, PAK2, MTOR, LYN, MAP4K3, MAP4K5, TNK2, PKN2, MYLK, PHKA1, SPEG, MARK2, FGFR1, ARAF, TYRO3, MAPK1, PGK1, CDK16, MAPK3, RIPK2, MET, MAPK14, STK38 ICK, PRPF4B, PDGFRB, AAK1, EPHA4, PTK2B, CDK2, PRKAA1, PIK3C2B, EPHB2, CDK7, PDGFRA, TEC, RIPK1, DDR1, NEK1, MINK1, ERBB2, EPHA2, ABL2, PLK2, PPIP5K2, EGFR, LATS2, FER, CDK20, BUB1B, DDR2, TGFBR2, CDK5, MAPK7, DAPK3, PIK3CD, PEAK1, FES, TBK1, MAPK12, EPHB4, DAPK1, PAPSS2, MAP3K3, PI4KA, IKBKE, CDKL5, GSK3B, CPNE3, PRKD2, HIPK3, ROCK2, DYRK3, CDC42BPA, TTN, DYRK1A, CMPK1, PRKAG1, PRKD1, PRKG1, CDC42BPB, DYRK4, TRIO, PKM, NRP1, TAB1, CSK, EPHB6, PPP1R9B, PRKAR1A, PIKFYVE, NEK9, AKAP9, GIT2, NEK7, NPR2, HIPK1, SGMS2, GAK, HGS, BCR, MAGI2, CALM1, BRD4, SQSTM1, DLG1, TJP2, CRIM1
Matrix metalloproteinase	SRRM2, ADAM9, ADAM12, FN1, LMNA, TAGLN
PDZ and LIM domain	MAGI2, PDLIM5, PDLIM1, PDLIM4, PDLIM7, MAGI1, MPDZ, GOPC, PDLIM2, LMO7, FHL2, LIMA1, ABLIM1, LASP1, LIMS3, FHL1, LIMCH1, LIMD1, LIMS2, FHL3, FBLIM1, ABLIM3, AJUBA, LIMS1

Proteomic profile of human TM cell-ECM adhesion complexes identifies proteins known to regulate podocyte morphology, slit diaphragm integrity, and barrier function.

Analysis of the proteomic profile of human TM cell-ECM adhesion complexes revealed the presence of several proteins that had not previously been described in the TM, including NEPH1 (Kirrel1), CD2AP, myosin 1e, BCAR1, NEDD9, microtubule-associated protein 1 (MAP1B), α-actinin-4 (ACTN4), partitioning defective 3 homolog (PARD3), LPP, JCAD, AHNAK, PEAK1, MAGI2, GSK3B, ABL2, ROBO1, ephrin-β receptor 4, NEK1, Tyro3, and many others. Interestingly, several of these proteins (NEPH1, CD2AP, myosin 1e, ZO-1, MAGI2, ACTN4, KANK2, FAT1, synaptopodin, ARHGDIA, ARHGAP24, NCK1 and 2, p130Cas/BCAR1, glycogen synthase kinase 3B, EPB41L5) and other cell adhesive proteins have been recognized to play vital role(s) in podocyte morphology, adhesive interactions, slit diaphragm barrier integrity, and glomerular fluid dynamics and filtration (35, 43, 45, 46). Given this intriguing finding, we carefully evaluated the protein profile of TM cell-matrix complexes to identify proteins demonstrated to be required for podocyte function. Table 2 lists some well-characterized proteins with roles in podocyte function and glomerular filtration that are also present in human TM cell-matrix adhesion complexes.

Table 2.

Detection of phosphotyrosine proteins involved in podocyte function in human trabecular meshwork cell-extracellular matrix complexes

NEPH1 (Kirrel)	Tyro3
CD2AP	CFL1
MYOSIN1E	FYN
ZO-1	p130Cas/BCAR1
MAGI2	mTOR
ACTN4	GSK3B
KANK2	Myh9
FAT1	EPB41L5
Synaptopodin	Connexin 43
ARHGDIA	IQGAP
ARHGAP24	Robo
NCK1	cAbl
SRGP2	Palladin
Plectin	PTEN
Rho	YAP
Rho kinase	TAZ
shroom	ITGβ1
JAM-A	FilaminA
Afadin	Talin
Fermitin2 (Kindlin 2)	FAK
Synaptojanin-1	ITA3

Although there were prior reports documenting phosphotyrosinylated sites in certain podocyte-enriched proteins, including NEPH1, CD2AP, synaptopodin, and FAT1, these observations were derived from analyses of rodent or non-primate samples (4, 40, 41). In human TM samples, we identified NEPH1 as being heavily tyrosine phosphorylated, including phosphorylation of tyrosine residues 572, 596, 605, 606, 622, 625, 647, 653, 661, 687, 701, 704 721, 724, and 745. Residues Y361 and Y548 of CD2AP; Y273, Y584, Y728, and Y738 of synaptopodin; and Y4242 of FAT1 were also found to be phosphorylated in human TM cell-ECM complexes. Details of several other podocyte-enriched proteins that are also tyrosine phosphorylated at specific sites in TM cells can be viewed in the scaffold files provided (PRIDE Data Repository) and in Supplemental Tables S2 and S3.

To obtain secondary confirmation of the presence of select proteins involved in podocyte function, including NEPH1, FAT1, MAGI2, and CD2AP in human TM cells, we performed immunoblotting analysis of cell-ECM complex fractions derived from human TM cells (derived from 52- and 66-yr-old donor eyes) and immunofluorescence analyses of human TM cells, with representative results shown in Fig. 4, A and B.

Fig. 4.

A: immunofluorescence-based confirmation of the presence and distribution of selected cell-extracellular matrix (ECM) complex proteins in human trabecular meshwork (TM) cells. Arrows indicate localization to the cell-cell junctions. Scale bar indicates image magnification. B: immunoblotting analysis-based detection of selected cell-ECM complex proteins in human TM cell-derived cell-ECM fractions from the replicates of 52- and 66-yr-old human donor eyes (lanes 1 and 2). LC, loading control [equal amounts of TM cell-ECM fractions (10 µg protein) derived from different samples were subjected to SDS-polyacrylamide gel electrophoresis, with the GelCode blue-stained protein bands shown].

Cell-cell junction proteins present in TM cell phosphotyrosine adhesome.

The presence of cell-cell junction and associated proteins, including ZO-1, ZO-2, JCAD, EPS8, FAT1, NEPH1, CD2AP, NOTCH2, nectin, P120 catenin, α-catenin, protocadherins 7, 10, and 18, CLIC1 and 4, ephrin receptors, MAGI-2, MAGI-1, JAM3, MAGUK proteins (MPP5 and MPP7), NOTCH2, kazrin, connexin 43, and junction plakoglobin in trabecular meshwork cells (Supplemental Tables S2 and S3), is of particular interest because it is assumed that cell-cell junctions in the SC rather than TM are relevant to barrier activity in the context of AH outflow and IOP (5, 48).

Junctional protein associated with coronary artery disease (JCAD), a newly characterized cell-cell junctional protein whose mutation is linked to coronary artery disease (2, 13), is tyrosine phosphorylated at several residues, with Y397 and Y1146 found to be phosphorylated in both samples 1 and 2 and in a replicate of sample 1 (data not shown). Figure 5 shows the sequence and mass spectra for the p-Y397- and p-Y1146-containing JCAD peptides identified in TM cell-ECM complexes. Sample 1 also contained several other tyrosine-phosphorylated JCAD peptides (phosphorylated on residues Y212, Y322, Y340, Y357, Y397, Y423, Y545, and Y686; Supplemental Table S2). Phosphorylation of JCAD at Y357 and Y686 has not been previously documented based on information contained in the phosphosite.org database. Using immunoblotting analysis, we have confirmed JCAD expression in human TM cells and determined that this protein localizes to the cell-cell junctions of human TM cells by immunofluorescence analysis (Fig. 4).

Fig. 5.

The sequence and mass spectra of junctional protein associated with coronary artery disease (JCAD) phosphopeptides containing residues Y397 and Y1146, as identified through analysis of trabecular meshwork (TM) cell-extracellular matrix (ECM) complexes. m/z, Mass-to-charge ratio.

Although it remains to be definitively demonstrated, the intercellular junctions of TM cells might also play a key role in TM tissue mesh architecture/fenestrations, since TM cells in the JCT region are known to develop cell-cell junctions and maintain cell adhesive interactions with SC cells (32, 51). However, the TM is also known to maintain resistance to AH outflow (24), and dexamethasone, which is known to increase resistance to AH outflow, has been reported to increase ZO-1-based cell-cell junctions in TM cells (50). Twenty five tyrosine residues of ZO-1 were found to be phosphorylated in TM cells, with the phosphorylation of Y198, Y1045, and Y1499 not having been previously reported. Similarly, ZO-2 is tyrosine phosphorylated in TM cells (Supplemental Tables S2 and S3). Moreover, we also detected MAGI proteins (MAGI-1 and MAGI-2), which are well-characterized tight junction- and adherens junction-associated proteins (23, 30), in the TM cell-ECM complexes (Supplemental Table S2). Immunofluorescence-based distribution analysis confirmed localization of MAGI-2 to cell-cell junctions in human TM cells (Fig. 4). MAGI-2 Y277, Y544, Y550, and Y890 residues were phosphorylated in TM cells, and these sites have not been documented previously (Supplemental Table S2). Collectively, it is reasonable to believe that cell-cell junctions in TM cells are physiologically relevant.

Tyrosine phosphorylation of diverse mechanosensing proteins derived from human TM cell-ECM complexes.

Mechanical stretch of TM tissue resulting from elevated IOP, shear stress associated with AH outflow, and ECM rigidity is presumed to play a critical role in homeostasis of IOP and AH outflow (51). Therefore, mechanotransduction, the process by which cells sense physical forces (both extracellular and intracellular) and translate them into biochemical and biological responses, is highly relevant to TM cell physiology and IOP. Integrin-regulated focal adhesions, cadherin-regulated cell-cell adhesion, and actomyosin-regulated contractile activity collectively and interdependently play a vital role in mechanotransduction (7, 19, 20, 44). Moreover, cell and tissue stiffness control transcriptional activation (11, 12, 36). In support, the phosphotyrosine proteome analysis of TM cell-ECM complexes reported in this study identified extensive tyrosine phosphorylation of diverse and several well-characterized mechanosensing proteins involved in mechanotransduction including kindlin, PINCH (LIMS1), VASP, talin, vinculin, p130Cas, MAGI, α₅β₁, zyxin, FAK, paxillin, α-actinin, myosin II, Rho, Rho kinase, and several regulatory proteins of Rho GTPases (Supplemental Tables S2 and S3). Additionally, phosphorylated Yap1 Y391 and Y407, TAZ Y305, LATS2 Y286, PTPN14, JCAD, AJUBA, LIMD1, angiomotin-like 1, and Salvador adaptor protein of mechanotranscription were detected in the TM cell phosphoadhesome (Supplemental Tables S2 and S3).

Interestingly, we also consistently identified phosphorylated Y445 piezo-2 in TM cell-ECM complexes (Supplemental Tables S2 and S3). Piezo channels are well-characterized pressure-sensitive channels involved in various physiological activities (17, 53). Figure 6A shows the mass spectrum of piezo-2 phosphopeptide containing the p-Y445 residue, which was detected in both samples 1 and 2 and in a replicate of sample 1 as well (data not shown). The sequence of this piezo-2 peptide (Fig. 6A) is highly conserved (100%) among various mammalian species, indicating its possible significance in regulation of piezo-2 channel function. We used Protter (31), a web-based tool that supports integrated visual analysis of membrane proteins, to visualize the location of piezo-2 Y445. As shown in Fig. 6B, Y445 appears to be localized in the nontransmembrane extracellular helical domain of piezo-2 (31, 52). Piezo-2 Y450 has also been shown to be phosphorylated based on the phosphosite.org database, but, in TM cell-ECM complexes, we did not detect this phosphopeptide. Immunoblotting analysis confirmed the presence of piezo-2 and piezo-1 channel proteins in human TM cell-ECM complex extracts (Fig. 4B).

Fig. 6.

Detection of tyrosine-phosphorylated piezo-2 pressure-sensitive channel protein in the phosphoproteome of trabecular meshwork (TM) cell-extracellular matrix (ECM) complexes. A: sequence and mass spectrum of piezo-2 Y445 phosphopeptide. m/z, Mass-to-charge ratio. B and C: Protter tool-based visualization of piezo-2 channel protein transmembrane topology and location of phosphorylated Y445 (indicated with arrow) in the nontransmembrane extracellular helical domain of piezo-2. The magnified region of piezo-2 sequence shown in C is marked in blue in B.

DISCUSSION

Although it is well recognized that cell adhesive interactions, actin cytoskeletal organization, integrins, ECM, contractile activity, and Rho GTPases play crucial role(s) in modulating AH drainage through the trabecular meshwork, the mechanisms mediating these interdependent activities and thereby influencing fluid dynamics are not well understood (15, 38, 51). Because tyrosine phosphorylation is known to modify the activity of several ECM and cell adhesion proteins (28, 42), we used a global approach to identify and characterize the tyrosine-phosphorylated proteins of the cell-ECM adhesion complexes of human TM cells by proteomics analysis. Importantly this study not only established by far the most comprehensive phosphotyrosine proteome of cell-ECM adhesion complexes but also identified various subclasses of regulatory proteins involved in cell-ECM adhesion and cell-cell junctions relevant to fluid barrier function, and mechanotransduction. Notably, this study also uncovered the presence of several proteins that have been demonstrated to be crucial for podocyte diaphragm barrier integrity and cell adhesive interactions in human TM cells, suggesting the existence of similar cell adhesive molecular mechanisms in these two cell types that are very likely linked to their common role in fluid dynamics and filtration and providing significant and novel insight into the role of TM in maintaining resistance to AH outflow and IOP.

Maintenance of normal IOP is required for the shape and architecture of the eye anterior segment, since the convex shape of the cornea is necessary for focusing of incident light onto the retina. Although IOP is determined by the balance between AH inflow from the ciliary epithelium and its drainage through the trabecular and nonconventional pathways, it is well recognized that elevated IOP, a major risk factor for glaucoma, is the result of impaired drainage of AH through the TM due to increased resistance to fluid flow, rather than impaired inflow from the ciliary epithelium (5, 48). Because cell adhesive interactions and cell-cell junctions influence the permeability characteristics of fluid flow through cells (6, 9), in this study we focused on identifying the various regulatory proteins involved in cell adhesive interactions of TM cells. For this we isolated the cell-ECM adhesion complexes of human TM cells and characterized their tyrosine-phosphorylated proteome by affinity pulldown and mass spectrometry analysis, identifying approximately ~1,000 individual phosphorylated proteins. Consistent with the integrin adhesome profile previously identified from various cell types (21, 56), human TM cell-matrix adhesion complexes contained most well-characterized adhesome proteins, including kinases, phosphatases, integrins, Rho GTPases and their regulatory proteins, scaffolding proteins, actin cytoskeleton-associated proteins, and cell-cell junctional and channel proteins. We identified that the integrins β₁, β₅, and α₅, which are crucial players in connecting the ECM to cytoskeletal proteins, were tyrosine phosphorylated in the cell-ECM complexes of human TM cells. While ECM proteins, including fibronectin, perlecan, versican, collagen type 1, thrombospondin-1, laminin-β and -γ, fibulin, and fibrillin, were also identified in the phosphoproteome of TM cell-ECM complexes, we could not definitively confirm that these proteins were tyrosine phosphorylated, even though they were isolated using phosphotyrosine antibody affinity pulldown described in this study. This discrepancy could be related to their limited percent coverage by mass spectrometry analysis.

To the best of our knowledge, this study provides the most comprehensive proteomics data for tyrosine-phosphorylated proteins of TM cell-ECM adhesion complexes (42). For TM cells, this is the first systematic study focused on characterizing the cell-ECM adhesion complex protein profile. Although there is a previous study that identified certain cell adhesive proteins of human TM cells using proteomics analysis, the study was based on analysis of soluble and membrane-enriched fractions rather than from the isolated cell-ECM adhesive complexes (8). Therefore, the profile that we report here is not only focused on cell-ECM adhesion complexes of TM cells but has also identified many more proteins than currently recognized as components of TM cell adhesive complexes (8). Some of the abundant proteins in TM cell-ECM adhesion complexes included tensin, MAP1B, BCAR1, NEDD9, AHNAK, FAK1, CTNND1, LPP, ZO-1, PEAK1, CTTN, talin, paxillin, ACTN4, caveolin-1, MET, Kirrel (NEPH1), LMO7, PEAR1, PARD3B, DOC1, PKP4, JCAD, and various Rho GTPase regulatory proteins. ARHGEF12 and optineurin, whose mutations are linked to open-angle glaucoma, were also found to be present in human TM cell-ECM complexes (39, 47).

The TM cells used in this study were cultured in flat plastic dishes for a period of 2 wk before the extraction of cell-ECM complexes. It may therefore be noteworthy to recognize that the higher stiffness of the plastic dish surface relative to the in vivo natural tissue environment of TM cells likely activates the formation of focal adhesions in these cells that could in turn augment mechanotransduction-dependent modification of the TM cell-ECM complex phosphoproteome (19).

There are two aspects in this study that are somewhat intriguing but are novel and significant for the role of TM in AH drainage. We were surprised at the diversity and vast number of cell-cell junction proteins and their regulatory proteins found in TM cells because thus far, little importance has been given to cell-cell junction proteins and their role in either in TM tissue fenestrations or barrier function (25, 33, 54). Some of the well-characterized cell-cell junction regulatory proteins included ZO-1, ZO-2, cingulin-like protein, nectins, afadin, ponsin, ephrin receptors, par proteins, desmoplakin, P120 catenin, JAM3, FAT1, both clustered and nonclustered protocadherins, shroom, IQGAP1, Cx43, MET, MAGI-1, MAGI-2, Crk and Crkl, JCAD, GSK3B, various catenins, Scrib, DLGs, EPLIN/Lima1, Nck, Slit/Robo, Abl kinases and their interacting proteins (Abi1 and Abi2), PAK2 and various Rho GTPases (Rho, Rac, and Cdc42), and Rap1 and their regulators. Despite the well-recognized fact that the TM tissue possesses a sponge-like architecture with varying sizes of fenestrations typically with wider gaps in the uveal TM compared with the corneoscleral region, and bears resistance to AH outflow (27), the molecular basis for the TM tissue fenestrations is not understood or explored. Similarly, little is known about the cell-cell junctions formed between the cells of the JCT and whether these cell-cell interactions are involved in barrier function or resistance to AH outflow (1, 5, 54). Because we did not see any VE-cadherin-positive cells in the TM cell cultures used in this study, the presence of the different cell-cell junctional and their regulatory proteins in the TM cell-matrix complexes suggests that these proteins likely participate in barrier function of TM and JCT cells. Moreover, because most of these proteins are tyrosine phosphorylated, it is possible that phosphorylation of cell-ECM adhesion complex proteins regulated by different tyrosine kinases (both intracellular and extracellular) is relevant for barrier function of TM and JCT cells in both normal and glaucomatous eyes. The ability of dexamethasone and other agents to stimulate an increase in ZO-1-based cell-cell junctions in TM cells has previously been reported (50). Also, Rho kinase inhibitors, which are used to lower IOP in glaucoma patients, have been shown to increase AH drainage partly through widening of the TM and JCT area resulting from cellular relaxation and decreased cell adhesion (38).

The second novel aspect of this study is evidence for the expression and distribution in TM cells of several cell adhesion and cell-cell junction proteins demonstrated to be required for podocyte slit diaphragm barrier integrity and cell adhesion. TM cell-ECM adhesion complexes contained high levels of NEPH1 (Kirrel 1), CD2AP, FAT1, Synaptopodin, ZO-1, KANK2, Myo1E, MAGI-2, ARHGDIA, ARHGAP24, ACTN4, EPB41L5, Robo, Nck, Par, and other proteins whose mutation or absence has been demonstrated to impair glomerular filtration and cause podocyte slit diaphragm defects, leading to kidney failure (12, 35, 43, 45, 46). Moreover, it is intriguing that cell morphological, cell adhesive, actin cytoskeletal organization, and cell biomechanical characteristics are reported to impact both AH drainage and podocyte barrier activity and glomerular filtration, indicating the existence of molecular parallels between TM cells and podocytes relative to their common role in barrier activity and fluid dynamics and the cell-ECM adhesive properties of these two cell types (12, 38, 43, 45, 48, 49, 51). It is equally important to recognize the role played by integrins, ECM, actin cytoskeleton, scaffolding proteins, myosin, and Rho/Rho kinase and other Rho GTPases in both TM and podocytes, in the context of the physiological role played by these cells in fluid dynamics and barrier activity (14, 18, 29, 38, 45). In future studies, it would be interesting to determine whether JCAD, which is expressed and heavily tyrosine phosphorylated in TM-ECM complexes and is known to play a role in the shear stress-dependent mechanotransduction pathway of endothelial cells, is also expressed in podocytes (10) and SC cells.

Having identified a vast number of tyrosine-phosphorylated proteins recognized to belong to different pathways, support diverse activities, and localize to the cell surface, extracellular or intracellular compartments, we believe that it is necessary, in our future studies, to identify and characterize the extracellular and intracellular kinases regulating tyrosine phosphorylation of TM cell-ECM complex proteins. For example, the extracellular domain of piezo-2 (Y445) and the CLIC1 (Y117, Y 214, and Y233) and CLIC 4 (Y128, Y225, and Y244) channel proteins are tyrosine phosphorylated in TM cell-ECM complexes, indicating the involvement of extracellular kinases. Therefore, once we gain more insight into the identity of some of these kinases, they could be explored as therapeutic targets for lowering IOP in glaucoma patients.

GRANTS

This study was funded by National Eye Institute Grants R01EY018590 and R01EY028823.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

R.M. and P.V.R. conceived and designed research; R.M. performed experiments; R.M. and P.V.R. analyzed data; R.M. and P.V.R. interpreted results of experiments; R.M. and P.V.R. prepared figures; R.M. and P.V.R. drafted manuscript; R.M. and P.V.R. edited and revised manuscript; R.M. and P.V.R. approved final version of manuscript.