Molecular architecture and modifications of full-length myocilin

Antibody-mediated clearance of an ER-resident aggregate that causes glaucoma

Recombinant antibodies are a promising class of therapeutics to treat protein misfolding associated with neurodegenerative diseases, and several antibodies that inhibit aggregation are approved or in clinical trials to treat Alzheimer's disease. Here, we developed antibodies targeting the aggregation-prone β-propeller olfactomedin (OLF) domain of myocilin, variants of which comprise the strongest genetic link to glaucoma and cause early onset vision loss for several million individuals worldwide. Mutant myocilin aggregates intracellularly in the endoplasmic reticulum (ER). Subsequent ER stress causes cytotoxicity that hastens dysregulation of intraocular pressure, the primary risk factor for most forms of glaucoma. Our antibody discovery campaign yielded two recombinant antibodies: anti-OLF1 recognizes a linear epitope, while anti-OLF2 is selective for natively folded OLF and inhibits aggregation in vitro. By binding OLF, these antibodies engage autophagy/lysosomal degradation to promote degradation of two pathogenic mutant myocilins. This work demonstrates the potential for therapeutic antibodies to disrupt ER-localized protein aggregates by altering the fate of folding intermediates. This approach could be translated as a precision medicine to treat myocilin-associated glaucoma with in situ antibody expression. More generally, the study supports the approach of enhancing lysosomal degradation to treat proteostasis decline in glaucoma and other diseases.

Myocilin misfolding and glaucoma: A 20-year update

Mutations in the gene MYOC account for approximately 5% of cases of primary open angle glaucoma (POAG). MYOC encodes for the protein myocilin, a multimeric secreted glycoprotein composed of N-terminal coiled-coil (CC) and leucine zipper (LZ) domains that are connected via a disordered linker to a 30 kDa olfactomedin (OLF) domain. More than 90% of glaucoma-causing mutations are localized to the OLF domain. While myocilin is expressed in numerous tissues, mutant myocilin is only associated with disease in the anterior segment of the eye, in the trabecular meshwork. The prevailing pathogenic mechanism involves a gain of toxic function whereby mutant myocilin aggregates intracellularly instead of being secreted, which causes cell stress and an early timeline for TM cell death, elevated intraocular pressure, and subsequent glaucoma-associated retinal degeneration. In this review, we focus on the work our lab has conducted over the past ∼15 years to enhance our molecular understanding of myocilin-associated glaucoma, which includes details of the molecular structure and the nature of the aggregates formed by mutant myocilin. We conclude by discussing open questions, such as predicting phenotype from genotype alone, the elusive native function of myocilin, and translational directions enabled by our work.

Quantitative differentiation of benign and misfolded glaucoma-causing myocilin variants on the basis of protein thermal stability

Accurate predictions of the pathogenicity of mutations associated with genetic diseases are key to the success of precision medicine. Inherited missense mutations in the myocilin (MYOC) gene, within its olfactomedin (OLF) domain, constitute the strongest genetic link to primary open-angle glaucoma via a toxic gain of function, and thus MYOC is an attractive precision-medicine target. However, not all mutations in MYOC cause glaucoma, and common variants are expected to be neutral polymorphisms. The Genome Aggregation Database (gnomAD) lists ∼100 missense variants documented within OLF, all of which are relatively rare (allele frequency <0.001%) and nearly all are of unknown pathogenicity. To distinguish disease-causing OLF variants from benign OLF variants, we first characterized the most prevalent population-based variants using a suite of cellular and biophysical assays, and identified two variants with features of aggregation-prone familial disease variants. Next, we considered all available biochemical and clinical data to demonstrate that pathogenic and benign variants can be differentiated statistically based on a single metric: the thermal stability of OLF. Our results motivate genotyping MYOC in patients for clinical monitoring of this widespread, painless and irreversible ocular disease.

Structure‒function‒pathogenicity analysis of C-terminal myocilin missense variants based on experiments and 3D models

MYOC is a common pathogenic gene for primary open-angle glaucoma and encodes the protein named myocilin. Multiple MYOC variations have been found, with different clinical significance. However, the pathogenesis of glaucoma induced by MYOC mutations has not been fully clarified. Here, we analyze the molecular and cellular biological differences caused by multiple variant myocilins, including protein secretion characteristics, structural changes, subcellular localization, cellular autophagic activity and oxidative stress. Denaturing and nondenaturing electrophoresis showed myocilin to be a secreted protein with the tendency to self-oligomerize. The full-length myocilin and its C-terminal cleavage fragment are secreted. Secretion analysis of 23 variant myocilins indicated that secretion defects are closely related to the pathogenicity of MYOC variants. Structural analysis showed that the alteration of steric clash is associated with the secretion characteristics and pathogenicity of myocilin variants. Immunocytochemistry results demonstrated that mutated myocilins are retained in the endoplasmic reticulum and disrupt autophagy. MTT assay, MitoTracker staining, and DCFH-DA staining showed increased oxidative injury in cells expressing MYOC mutants. Taken together, MYOC mutations are able to induce cell dysfunction via secretion defects and intracellular accumulation resulting from steric clash alterations.

Recombinant antibodies recognize conformation-dependent epitopes of the leucine zipper of misfolding-prone myocilin

Recombinant antibodies with well-characterized epitopes and known conformational specificities are critical reagents to support robust interpretation and reproducibility of immunoassays across biomedical research. For myocilin, a protein prone to misfolding that is associated with glaucoma and an emerging player in other human diseases, currently available antibodies are unable to differentiate among the numerous disease-associated protein states. This fundamentally constrains efforts to understand the connection between myocilin structure, function, and disease. To address this concern, we used protein engineering methods to develop new recombinant antibodies that detect the N-terminal leucine zipper structural domain of myocilin and that are cross-reactive for human and mouse myocilin. After harvesting spleens from immunized mice and in vitro library panning, we identified two antibodies, 2A4 and 1G12. 2A4 specifically recognizes a folded epitope while 1G12 recognizes a range of conformations. We matured antibody 2A4 for improved biophysical properties, resulting in variant 2H2. In a human IgG1 format, 2A4, 1G12, and 2H2 immunoprecipitate full-length folded myocilin present in the spent media of human trabecular meshwork (TM) cells, and 2H2 can visualize myocilin in fixed human TM cells using fluorescence microscopy. These new antibodies should find broad application in glaucoma and other research across multiple species platforms.