The trabecular meshwork (TM) is the main site of drainage of the aqueous humor, and its dysfunction leads to intraocular pressure elevation, which is one of the main risk factors of glaucoma. We aimed to compare the effects on cytoskeleton organization and extracellular matrix (ECM) of latanoprost (LT) and a Rho-kinase inhibitor (ROCKi) on a transforming growth factor beta2 (TGF-β2)-induced glaucoma-like model developed from primary culture of human TM cells (pHTMC). The TGF-β2 stimulated pHTMC were grown and incubated with LT or a ROCKi (Y-27632) for 24 h. The expression of alpha-smooth muscle actin (αSMA) and fibronectin (FN), and phosphorylation of the myosin light chain (MLC-P) and Cofilin (Cofilin-P) were evaluated using immunofluorescence and Western blot. The architectural modifications were studied in a Matrigel^TM 3D culture. TGF-β2 increased the expression of αSMA and FN in pHTMC and modified the cytoskeleton with cross-linked actin network formation. LT did not alter the expression of αSMA but decreased FN deposition. The ROCKi decreased TGF-β2-induced αSMA and FN expression, as well as MLC-P and Cofilin-P, and stimulated the cells to recover a basal cytoskeletal arrangement. In the preliminary 3D study, pHTMC organized in a mesh conformation showed the widening of the TM under the effect of Y-27632. By simultaneously modifying the organization of the cytoskeleton and the ECM, with fibronectin deposition and overexpression, TGF-β2 reproduced the trabecular degeneration described in glaucoma. The ROCKi was able to reverse the TGF-β2-induced cytoskeletal and ECM rearrangements. LT loosened the extracellular matrix but had no action on the stress fibers.

This study aimed to investigate the human proteome profile of samples collected from whole (W) Schirmer strips (ScS) and their two parts—the bulb (B) and the rest of the strip (R)—with a comprehensive proteomic approach using a trapped ion mobility mass spectrometer, the timsTOF Pro. Eight ScS were collected from two healthy subjects at four different visits to be separated into three batches, i.e., 4W, 4B, and 4R. In total, 1582 proteins were identified in the W, B, and R batches. Among all identified proteins, binding proteins (43.4%) and those with catalytic activity (42.2%) constituted more than 80% of the molecular functions. The most represented biological processes were cellular processes (31.2%), metabolic processes (20.8%), and biological regulation (13.1%). Enzymes were the most represented protein class (41%), consisting mainly of hydrolases (47.5%), oxidoreductases (22.1%), and transferases (16.7%). The bulb (B), which is in contact with the conjunctiva, might collect both tear and cell proteins and therefore promote the identification of more proteins. Processing B and R separately before mass spectrometry (MS) analysis, combined with the high data acquisition speed and the addition of ion-mobility-based separation in the timsTOF Pro, can bring a new dimension to biomarker investigations of a limited sample such as tear fluid.

Annotation of lipids in untargeted lipidomic analysis remains challenging and a systematic approach needs to be developed to organize important datasets with the help of bioinformatic tools. For this purpose, we combined tandem mass spectrometry-based molecular networking with retention time (t_R) prediction to annotate phospholipid and sphingolipid species. Sixty-five standard compounds were used to establish the fragmentation rules of each lipid class studied and to define the parameters governing their chromatographic behavior. Molecular networks (MNs) were generated through the GNPS platform using a lipid standards mixture and applied to lipidomic study of an in vitro model of dry eye disease, i.e., human corneal epithelial (HCE) cells exposed to hyperosmolarity (HO). These MNs led to the annotation of more than 150 unique phospholipid and sphingolipid species in the HCE cells. This annotation was reinforced by comparing theoretical to experimental t_R values. This lipidomic study highlighted changes in 54 lipids following HO exposure of corneal cells, some of them being involved in inflammatory responses. The MN approach coupled to t_R prediction thus appears as a suitable and robust tool for the discovery of lipids involved in relevant biological processes.

The ocular surface (OS) enzymes are of great interest due to their potential for novel ocular drug development. We aimed first to profile and classify the enzymes of the OS to describe major biological processes and pathways that are involved in the maintenance of homeostasis. Second, we aimed to compare the enzymatic profiles between the two most common tear collection methods, capillary tubes (CT) and Schirmer strips (ScS). A comprehensive tear proteomic dataset was generated by pooling all enzymes identified from nine tear proteomic analyses of healthy subjects using mass spectrometry. In these studies, tear fluid was collected using CT (n = 4), ScS (n = 4) or both collection methods (n = 1). Classification and functional analysis of the enzymes was performed using a combination of bioinformatic tools. The dataset generated identified 1010 enzymes. The most representative classes were hydrolases (EC 3) and transferases (EC 2). Phosphotransferases, esterases and peptidases were the most represented subclasses. A large portion of the identified enzymes was common to both collection methods (n = 499). More enzymes were specifically detected in the ScS-extracted proteome. The major pathways in which the identified enzymes participate are related to the immune system and protein, carbohydrate and lipid metabolism. Metabolic processes for nucleosides, cellular amides, sugars and sulfur compounds constituted the most enriched biological processes. Knowledge of these molecules highly susceptible to pharmacological manipulation might help to predict the metabolism of ophthalmic medications and develop novel prodrug strategies as well as new drug delivery systems. Combining such extensive knowledge of the OS enzymes with new analytical approaches and techniques might create new prospects for understanding, predicting and manipulating the metabolism of ocular pharmaceuticals. Our study reports new, essential data on OS enzymes while also comparing the enzyme profiles obtained via the two most popular methods of tear collection, capillary tubes and Schirmer strips.

The goal of this study was to explore the specific signaling pathways related to inflammation in two experimental mouse dry eye (EDE) models. Female C57BL/6 mice housed for 10 days in a controlled desiccative environment were either treated with scopolamine (EDE-1; n = 18) or subjected to extraorbital lacrimal gland excision bilaterally (EDE-2; n = 10). Non-induced mice (n = 20) served as healthy controls. A corneal fluorescein staining (CFS) scoring was used at baseline through to day (D) 10 to evaluate epitheliopathy. At D10, corneas and conjunctivas were collected for multiplexed transcriptomic analysis with the NanoString^® mouse inflammatory CodeSet. Both EDE-1 and EDE-2 mice presented a change in corneal integrity, with a significant increase in CFS scores at D10. More gene transcripts were identified in EDE-2 compared with EDE-1 (116 vs. 96, respectively), and only a few were common to both models, 13 for the cornea and 6 for the conjunctiva. The gene functional annotation analysis revealed that the same inflammatory pathways were involved in both models. Comparative profiling of gene expression in the two EDE models leads to the identification of various targets and signaling pathways, which can be extrapolated to and confirmed in human disease.