Secondary Logo

Journal Logo

LOXL1-associated Candidate Epithelial Pathomechanisms in Exfoliation Glaucoma

Laczko, Rozalia PhD; Szauter, Kornelia M. PhD; Csiszar, Katalin PhD

doi: 10.1097/IJG.0000000000000105
LOXL1 and Growth Factors

Results of the present study support ocular epithelia-specific LOXL1 functions in exfoliation glaucoma that may include both dysregulated extracellular matrix cross-linking activity and cellular mechanisms involving a role for LOXL1, in direct interaction with Snail1, in promoting epithelial to mesenchymal transition and a potential shift towards fibrogenic epithelial cell phenotypes.

John A. Burns School of Medicine, University of Hawaii, Honolulu, HI

Supported by The Glaucoma Foundation and NIH Grants U54RR022762 and U54RR026136 to K.C.

Disclosure: The authors declare no conflict of interest.

Reprints: Katalin Csiszar, PhD, John A. Burns School of Medicine, University of Hawaii, 1960 East-West Road, Biomed T415, Honolulu, HI, 96822 (e-mail:

Received July 31, 2014

Accepted August 1, 2014

Exfoliation syndrome (XFS) is a late-onset systemic disorder characterized by the pathologic deposition of fibrillar material in multiple tissues. XFS is the most common recognizable cause of exfoliation glaucoma (XFG) worldwide, comprising the majority of open angle glaucoma cases in some countries. XFG has an aggressive course, with a worse prognosis than primary open-angle glaucoma, and either predisposes to or is associated with a range of intraocular complications. In the XFG eye, fibrillogranular aggregates accumulate on the anterior lens capsule and in other tissues of the anterior segment. These aggregates are considered to result from aberrant production, conformation, and function of both structural (including microfibrils1), and enzymatic (such as LOXL1) components of the extracellular matrix (ECM). Exfoliation material (XFM) released from both the lens surface by iridolenticular friction and synthesized endogenously in the trabecular meshwork, combined with pigment released from ruptured iris pigmented epithelial cells, block the trabecular outflow pathway.

Genome-wide associations and genetic linkage studies have consistently demonstrated in populations worldwide, single-nucleotide polymorphisms at the chromosome 15q24 locus, including alleles of the LOXL1 gene, that predispose to a high risk for both XFS and XFG,2–5 yet the mechanisms underlying the involvement of LOXL1 remains to be elucidated.

LOXL1 is one of 5 members of the lysyl oxidase (LOX) family of copper-dependent ECM amine oxidases characterized by a conserved carboxyl terminal domain required for the catalysis of lysine-derived cross-linking of collagen and elastic fibers. Although the exact pathomechanisms of XFS and XFG remain unknown, characterization of the aggregated intraocular and extraocular material, and aberrant production of microfibrillar proteins, fibrillin-1, microfibril-associated glycoprotein-1 (MAGP-1), latent TGF-ß binding proteins (LTBP-1 and LTBP-2), fibulin-2, elastin, and LOXL1 have led to the concept that XFG may result from primary disturbances in LOXL1 regulation and elastic fiber6,7 or microfibrillar1 network homeostasis. More in-depth analyses; however, indicate disordered function of a wider range of ECM components, including those of the basement membrane, various glycoproteins, ECM remodeling enzymes, and cell-ECM signaling involving TGF-β. The TGF-β-regulated LOXL1 also has significant intracellular functions within regulatory networks that have not been analyzed in the context of XFS or XFG. XFM is known to be produced by mesenchymal, epithelial, and endothelial cell types, yet there is an overall lack of understanding of the contribution of cell type-specific mechanisms to disorderly ECM production and aberrant LOXL1 functions in XFS and XFG.

Back to Top | Article Outline


To analyze LOXL1-associated epithelial functions, panels of mesenchymal and epithelial cell lines, including kidney, and retinal epithelial cells (ARPE-19) suitable for LOXL1 analysis,8 were established and genotyped for the major risk-associated alleles R141G and G153D. In cultured cells, LOXL1 synthesis and processing were evaluated by Western blot analysis using LOXL1 polyclonal rabbit primary antibody (Zymed Laboratories, San Francisco), and polyclonal anti-BMP-1 antibody (Affinity Bioreagents). LOXL1 catalytic activity assays were performed in media fractions of cells cultured for 48 hours in phenol red and serum -free media using the Amplex Red assay (Life Technologies) as previously described.9 LOXL1 subcellular and ECM localizations were determined in fluorescent immunocytochemistry experiments. Cell membrane staining was performed with wheat germ agglutinin, and nuclear staining using 4',6-diamidino-2-phenylindole or propidium iodide. Stably transfected epithelial cell lines overexpressing the human full-length hLOXL1 (allele 153G/G), were generated and in these cells, LOXL1 overexpression, subcellular localization, processing, and activity were analyzed. LOXL1-induced cell phenotype changes were monitored under different culture conditions using cell migration assays, and tests for filopodia for cell mobility/spindling. LOXL1 activity-dependent and independent functions were evaluated using an irreversible inhibitor of LOXL1, BAPN, and catalase. Cultured cells were pretreated with BAPN or catalase for 48 hours before protein extraction or immunofluorescent microscopy, or for migration assays, 24 hours before assays and an additional 4 hours during the assays. Earlier established noncytotoxic concentrations of 100 μM for BAPN (Sigma), and 200 U/mL for catalase (Sigma) were used in these experiments.10 To monitor relative expression changes in the levels of E-cadherin, whole cell lysates were probed in Western blot protein assays using polyclonal rabbit anti–E-cadherin (Santa Cruz Biotechnology). Optical density on autoradiographs was normalized to GAPDH control. For LOXL1-Snail1 interaction tests, LOXL1 overexpressing adherent cells were washed with PBS and harvested in high salt RIPA lysis buffer, protein lysates were analyzed using SDS gel electrophoresis. LOXL1-Snail1 interactions were detected using coimmune precipitation of whole cell lysates and Snail1 (Cell Signaling Technology) and LOXL1-specific antibodies.

Back to Top | Article Outline


Although traditionally, mesenchymal cells are believed to be the primary source of LOXL1, in our experiments, the levels of LOXL1 in E-cadherin expressing cultured ARPE-19 retinal epithelial cells exceeded LOXL1 levels detected in mesenchymal fibroblasts (Fig. 1). In retinal epithelial cells, LOXL1 appeared in both cellular and secreted compartments in various molecular masses including the full length 65 and a 55 kD form likely resulting from furine-like processing.11 The secreted retinal epithelial cell type-specific forms (Fig. 1) (44, 40, and 30 kD), by analogy to known posttranslational intracellular and extracellular modifications of the LOXL1 protein in various cell types, were presumed to be the product of activating BMP-1 processing and suggested a high level of retinal epithelial LOXL1 activity. A further evaluation of BMP-1 processing of LOXL1 in retinal epithelial cells using Western blot analysis of the culture media, confirmed significant levels of BMP-1 (data not shown). In addition to cytoplasmic and secreted LOXL1, abundant amounts of the LOXL1 protein were detected within the nuclei of retinal epithelial cells using both Western and immunocytochemisty analyses (Fig. 2).





Overexpression of the full-length human LOXL1 (allele 153G/G) in stably transfected epithelial cells (HEK293), resulted in high levels of the LOXL1 protein. The hLOXL1-transfected (pIND-LOXL1) cells showed robust expression of both the 65 kD full-length protein and processed forms immunolocalized in the cytoplasm and nuclei and less prominently in the ECM. In activity assays of 3-day postconfluent pIND-LOXL1 stable overexpressing HEK293 cells, LOXL1 proved to have increased catalytic activity compared with native and empty vector (pIND)-transfected cells (Fig. 3).



Cellular overexpression of LOXL1, or exogenous LOXL1 in HEK293 or ARPE-19 cells treated for 48 hours with LOXL1 in conditioned culture media derived from either high-expressing ARPE-19 or overexpressing HEK293 cells cultures, did not result in visible cell phenotype changes. In cell migration assays, the number of migrating cells (the average number of migrated cells in 6 randomly chosen field in 3 parallel experiments per conditions) in pIND-LOXL1 stably transfected HEK293 cells, did not change significantly (P=0.42) compared with native (HEK) or empty vector-transfected (pIND HEK) cells (data not shown). To test for LOXL1 activity-dependent or activity-independent functions, cultured epithelial cells were treated with BAPN, an irreversible inhibitor of LOXL1 activity, or catalase that promotes the degradation of hydrogen peroxide, a byproduct of the LOXL1 cross-linking reaction and itself an important regulator of cellular functions. There were no significant differences in the number of filopodia formed under different culture conditions, suggesting that LOXL1 activity-related function did not have a significant effect on cell mobility/spindling, a phenotypic marker of epithelial to mesenchymal transition (EMT).

Among epithelial features affected by LOXL1, the epithelial marker and cell adhesion protein E-cadherin, highly expressed in ARPE-19 cells, was reduced to less than 50% by LOXL1 overexpression in stable transfected HEK293 cells (Figs. 4A, B). E-cadherin expression is controlled by Snail, a transcriptional regulator of EMT. Therefore, we considered the possibility that reduced E-cadherin was an indicator for the initiation of LOXL1-mediated EMT in these cells. The basal levels of Snail expression did not increase significantly in response to LOXL1 overexpression. However, in immunoprecipitation experiments, we detected direct interaction between Snail and the full-length LOXL1 in overexpressing cells, an interaction that was not detectable in native HEK293 cells or in vector-transfected cells (Fig. 4C). These results confirmed our earlier data reported for various epithelial cell types12–14 that LOXL1 in interactions with Snail may be directly involved in the induction of EMT and the generation of fibrogenic cell types, a mechanism that in ocular epithelial cells, may contribute to disorderly ECM and the pathomechanism of XFG.



Back to Top | Article Outline


To date, pathomechanistic explanations for XFG have largely focused on downstream scenarios, deduced from the components of the irreversible aberrant fibrillogranular aggregates in eye compartments. Attempts to identify the cell type-specific mechanisms in the initiating pathologic steps, meanwhile, have remained extremely difficult due to limited access to tissue and cell specimens suitable to experimentally test and challenge these hypotheses and lack of animal models that would recapture the XFG process. In this study, therefore, we have established a cell model to address firstly, the contribution of epithelial cells to LOXL1-associated pathologic mechanisms most relevant to its aberrant regulation and cross-linking function of ECM aggregates in XFG and, secondly, to explore cellular roles for LOXL1 within regulatory functions that may additionally contribute to the widespread ECM disturbance and aberrant production of fibrotic material that occurs in XFG.

Results from the current study demonstrated that, compared with mesenchymal cells considered as major sources of LOXL1, E-cadherin-expressing polarized epithelial cells produce higher levels of LOXL1, express abundant amounts of BMP-1 critical for LOXL1 activation, and process and activate LOXL1, supporting high levels of epithelia-derived LOXL1 cross-linking activity in the ECM. The broader epithelial cell type specificity and functional significance of some of the processed forms unique to retinal epithelial cells needs further study. Although there are variations in LOXL1 expression levels among the cell lines analyzed, this is unlikely due to allelic differences at R141L or G153D (Arg/Leu, Gly/Asp) associated with risk for XFS and/or XFG, similar to an earlier report on LOXL1 haplotypes (GG, GA, TG, TA) at these loci that did not affect LOXL1 catalytic activity in vitro.15

Overexpression or inhibition of cellular or exogenous LOXL1 did not induce cell phenotype changes in terms of migratory ability, filopodia, or cell mobility/spindling, visible signs of mesenchymal features. LOXL1, however, significantly downregulated the epithelial marker E-cadherin, suggesting the initiation of LOXL1-mediated EMT in these cells. Consistent with this novel cellular/nuclear LOXL1 function, LOXL1 localized not only to the ECM, but was also prominently present in nuclear compartments, in agreement with our earlier in vivo16 and in vitro12 observations.

EMT has been reported to occur, as part of various eye pathologies, in eye epithelial compartments,17 in the trabecular meshwork,18 under conditions such as ocular tissue fibrosis and injury,19 and with the involvement of Snail1.20 EMT-like phenomenon has been also proposed to directly result in aberrant conditions in the aqueous outflow pathway in glaucomatous eyes.18

We have previously reported the involvement of LOX family members, LOXL2 and LOXL3, in EMT including interactions with Snail1 to downregulate E-cadherin expression. Furthermore, we have also noted that Snail1 lysine residues 98 and 137 are required for stability, functional cooperation with LOXs, and induction of EMT.12,13 We have also described an in vivo role for LOXL1 in cooperation with Snail1 during the transition of liver epithelial cells towards a fibrogenic phenotype.14 Results of the current study further support interactions between LOXL1 and Snail1 in epithelial cells, and LOXL1-induced repression of E-cadherin, a hallmark of EMT, and a mechanism that in eye epithelial cells, may contribute to disorderly ECM production during the pathogenesis of XFG.

Transdifferentiation reported in retinal epithelial cells induced by TGF-β1 and EGF, included gain of α-smooth muscle actin, an indicator of a transition towards a myofibroblast phenotype, and enhanced migration.21 Although the progression of LOXL1-induced EMT including expression patterns of early (vimentin) and late (α-smooth muscle actin) markers need to be defined, enhanced epithelial cell migration does not appear to be a LOXL1-associated feature.

Not only epitihelial, but mesenchymal cells also become activated in XFG and under reactive conditions including injury, repair, and inflammation, and differentiate toward profibrotic myofibroblast phenotypes resulting in persistent and excessive production and disorderly deposition of fibrotic material. Under these conditions, there is an additional, TGF-β-driven, expansion of connective tissue cell population.22 In XFG, the profibrotic TGF-β has been noted to induce LOXL1 in Tenon’s capsule cells.23 In cells of the trabecular meshwork, TGF-βI induced LOXL1 expression involving both Smad and non-Smad signaling pathways, and the resulting increase in LOXL1 cross-linking activity was proposed to be partly responsible for TGF-β-mediated intraocular pressure elevation.24 Of interest is the fact that in the characteristic XFG fibrotic profile, although numerous fibrotic elements are present, unlike in tissue fibrosis, there is no increase in fibrillar and type IV collagens. In epithelial cells; however, TGF-β did not appear to affect LOXL1 expression (A546 cells, NCBI GEO array data). Thus alterations in multiple epithelial and mesenchymal cell-specific regulatory pathways may converge and contribute to the XFG phenotype.

Data collectively support that in XFG, LOXL1-associated pathology involves dysregulated ECM activity of the secreted LOXL1 produced by both mesenchymal and epithelial cells that may directly contribute to irreversible cross-linking of fibrillar ECM aggregates, and additionally, in eye epithelia, a cellular mechanism with a role for LOXL1 in interactions with Snail1, in the regulation of EMT, a process that activates multiple regulatory networks to promote fibrogenic cell phenotypes.

Back to Top | Article Outline


1. Kuchtey J, Kuchtey RW. The microfibril hypothesis of glaucoma: implications for treatment of elevated intraocular pressure. J Ocul Pharmacol Ther. 2014; 30:170–180.
2. Challa P. Glaucoma genetics. Int Ophthalmol Clin. 2008; 48:73–94.
3. Liu Y, Allingham RR. Molecular genetics in glaucoma. Exp Eye Res. 2011; 93:331–339.
4. Schlötzer-Schrehardt U. Genetics and genomics of pseudoexfoliation syndrome/glaucoma. Middle East Afr J Ophthalmol. 2011; 18:30–36.
5. Nakano M, Ikeda Y, Tokuda Y, et al.. Novel common variants and susceptible haplotype for exfoliation glaucoma specific to Asian population. Sci Rep. 2014; 4:5340.
6. Schlötzer-Schrehardt U. Molecular pathology of pseudoexfoliation syndrome/glaucoma—new insights from LOXL1 gene associations. Exp Eye Res. 2009; 88:776–785.
7. Schlötzer-Schrehardt U, Hammer CM, Krysta AW, et al.. LOXL1 deficiency in the lamina cribrosa as candidate susceptibility factor for a pseudoexfoliation-specific risk of glaucoma. Ophthalmology. 2012; 119:1832–1843.
8. Coral K, Madhavan J, Pukhraj R, et al.. High glucose induced differential expression of lysyl oxidase and its isoform in ARPE-19 cells. Curr Eye Res. 2013; 38:194–203.
9. Szauter KM, Ordas A, Laxer RM, et al.. A novel fibrotic disorder associated with increased dermal fibroblast proliferation and downregulation of genes of the microfibrillar network. Br J Dermatol. 2010; 163:1102–1115.
10. Kirschmann DA, Seftor EA, Fong SF, et al.. A molecular role for lysyl oxidase in breast cancer invasion. Cancer Res. 2002; 62:4478–4483.
11. Borel A, Eichenberger D, Farjanel J, et al.. Lysyl oxidase-like protein from bovine aorta. Isolation and maturation to an active form by bone morphogenetic protein-1. J Biol Chem. 2001; 276:48944–48949.
12. Peinado H, Del Carmen Iglesias-de la Cruz M, Olmeda D, et al.. A molecular role for lysyl oxidase-like 2 enzyme in snail regulation and tumor progression. Embo J. 2005; 24:3446–3458.
13. Moreno-Bueno G, Salvador F, Martin A, et al.. Lysyl oxidase-like 2 (LOXL2), a new regulator of cell polarity required for metastatic dissemination of basal-like breast carcinomas. EMBO Mol Med. 2011; 3:528–544.
14. Diaz R, Kim JW, Hui JJ, et al.. Evidence for the epithelial to mesenchymal transition in biliary atresia fibrosis. Hum Pathol. 2008; 39:102–115.
15. Kim S, Kim Y. Variations in LOXL1 associated with exfoliation glaucoma do not affect amine oxidase activity. Mol Vis. 2012; 18:265–270.
16. Hayashi K, Fong KS, Mercier F, et al.. Comparative immunocytochemical localization of lysyl oxidase (LOX) and the lysyl oxidase-like (LOXL) proteins: changes in the expression of LOXL during development and growth of mouse tissues. J Mol Histol. 2004; 35:845–855.
17. Li P, Jing J, Hu J, et al.. RNA interference targeting snail inhibits the transforming growth factor beta 2-induced epithelial-mesenchymal transition in human lens epithelial cells. J Ophthalmol. 2013; 2013:869101.
18. Takahashi E, Inoue T, Fujimoto T, et al.. Epithelial mesenchymal transition-like phenomenon in trabecular meshwork cells. Exp Eye Res. 2014; 118:72–79.
19. Saika S, Yamanaka O, Flanders KC, et al.. Epithelial-mesenchymal transition as a therapeutic target for prevention of ocular tissue fibrosis. Endocr Metab Immune Disord Drug Targets. 2008; 8:69–76.
20. Li H, Li M, Xu D, et al.. Overexpression of Snail in retinal pigment epithelial triggered epithelial-mesenchymal transition. Biochem Biophys Res Commun. 2014; 446:347–351.
21. Lee H, O’Meara SJ, O’Brien C, et al.. The role of gremlin, a BMP antagonist, and epithelial-to-mesenchymal transition in proliferative vitreoretinopathy. Invest Ophthalmol Vis Sci. 2007; 48:4291–4299.
22. Hinz B, Gabbiani G. Fibrosis: recent advances in myofibroblast biology and new therapeutic perspectives. F1000 Biol Rep. 2010; 2:78.
23. Zenkel M, Krysta A, Pasutto F, et al.. Regulation of lysyl oxidase-like 1 (LOXL1) and elastin-related genes by pathogenic factors associated with pseudoexfoliation syndrome. Invest Ophthalmol Vis Sci. 2011; 52:8488–8495.
24. Sethi A, Mao W, Wordinger RJ, et al.. Transforming growth factor-beta induces extracellular matrix protein cross-linking lysyl oxidase (LOX) genes in human trabecular meshwork cells. Invest Ophthalmol Vis Sci. 2011; 52:5240–5250.

exfoliation syndrome; Lysyl oxidase-like 1; exfoliation glaucoma; epithelial to mesenchymal transition

© 2014 by Lippincott Williams & Wilkins.