A Role for Clusterin in Exfoliation syndrome and Exfoliation glaucoma?

Abstract

The multifunctional protein clusterin (CLU) is a secreted glycoprotein ubiquitously expressed throughout the body, including in the eye. Its primary function is to act as an extracellular molecular chaperone, preventing the precipitation and aggregation of misfolded extracellular proteins. Clusterin is commonly identified at fluid-tissue interfaces, and has been identified in most body fluids. It is a component of exfoliation material, and CLU mRNA is reduced in eyes with exfoliation syndrome compared with controls. SNPs located in the CLU genomic region have been associated with Alzheimer disease (AD) at the genome-wide level and several CLU SNPs located in an apparent regulatory region have been nominally associated with XFS/XFG in Caucasians with European ancestry and in south Indians. Interestingly, clusterin associates with altered elastic fibers in human photo-aged skin and prevents UV-induced elastin aggregation in vitro. In light of the known geographic risk factors for XFS/XFG, which could include UV light, investigations of CLU-geographic interactions could be of interest. Future studies investigating rare CLU variation and other complex interactions including gene-gene interactions in XFS/XFG cases and controls may also be fruitful. While CLU has been considered as a therapeutic target in AD, cancer and dry eye, a role for clusterin in XFS/XFG needs to be better defined before therapeutic approaches involving CLU can be entertained.

Exfoliation syndrome (XFS) and the associated glaucoma (XFG) are genetically complex traits.¹ LOXL1 (lysyl oxidase-like 1) is a major risk factor for XFS/XFG. While LOXL1 genetic variants were initially identified in Iceland², the association with XFS/XFG has been replicated in populations throughout the world, including the United States.^3–7 Although LOXL1 is a major risk factor for XFS/XFG, the association is complex. While the LOXL1 risk alleles are present in the majority of cases worldwide, they are also frequently found in controls, arguing that other genetic and/or environmental factors are necessary for the disease to be fully manifested.⁸ A recent multi-ethnic genome wide association study (GWAS) identified common variants near CACNA1A (calcium channel, voltage-dependent, P/Q type, alpha 1A subunit) significantly associated with XFS.⁹ Here we review data suggesting that clusterin (CLU), an extracellular chaperone protein known to reduce protein aggregation, could contribute to XFS/XFG pathogenesis.

Clusterin is a secreted glycoprotein ubiquitously expressed throughout the body, including in the eye. Its primary function is to act as an extracellular molecular chaperone, preventing the precipitation and aggregation of misfolded extracellular proteins.¹⁰ Clusterin is commonly identified at fluid-tissue interfaces, and has been found in most body fluids.¹⁰ The protein is encoded by a single gene (CLU) that includes a 14-bp element in the promoter that is specifically recognized by the transcription factor HSF1 (heat shock factor 1) allowing for heat-shock-induced transcription.¹¹ CLU mRNA is also up-regulated in response to a broad range of cellular stress signals and conditions: oncogene activity, growth factors and cytokines (including TGFβ), and other stress- or apoptosis-inducing agents including UVA, UVB, proteotoxic stress, heavy metals, oxidants, hypoxia, ionizing radiation, and chemotherapeutic drugs.^12–14

The translated clusterin preprotein is cleaved prior to secretion from the cell to produce A and B subunits, and post-translational modifications include glycosylation at 6 different asparagine residues (N-glycosylation (Asn86, Asn103, Asn145, Asn291, Asn354, Asn374).¹⁵ Mature clusterin contains a disulfide-linked core region that involves at least 4 and possibly 5 disulfide bonds.¹⁶ Under normal conditions clusterin is constitutively secreted from cells and is found in extracellular fluids.¹⁷ Mature, post-translationally modified clusterin can be released from the ER/Golgi under conditions of ER stress, when it may also have a role in intracellular processes occurring outside the secretory system.¹⁸

After secretion, clusterin’s primary role is to bind to hydrophobic complexes and denatured proteins to promote their removal.¹⁹ Clusterin is a highly potent inhibitor of protein precipitation, with more activity than heat-shock proteins.²⁰ Interestingly, clusterin can form self-oligomers depending on the pH of the environment.^21–23 In the mildly acidic environment of the anterior chamber, clusterin would be expected to be disassociated (monomers) and have maximal chaperone activity.²¹ In some instances clusterin has been reported to form high molecular weight complexes with particular ligands such as fibrinogen.²²

Clusterin has been implicated in a wide variety of biological and pathological processes (Figure 1). The protein appears to be particularly important in pathological conditions related to cellular stress. CLU knockout mice are indistinguishable from their wild type littermates under normal conditions. With age or stress however, CLU knockout mice develop glomerular neuropathy, potentially due to the accumulation of pathological protein deposits in the absence of clusterin activity.²⁴

Figure 1.

Biological and pathological processes involving clusterin.

Specific pathological or biological processes involving clusterin include induction by oxidative stress,²⁵ clearance of misfolded Alzheimer amyloid beta (Aβ) peptide across the blood brain barrier,²⁶ association with sperm cells to prevent their rejection by the female immune system,²⁷ modulation of macrophage activity,²⁸ induction of PI3K/AKT-signaling and other cytoprotective effects,²⁹ and alteration of autophagy and apoptosis.^30,31 Clusterin is known to interact with many proteins and ligands are part of these processes.

In addition to XFS/XFG, clusterin has been implicated in a number of human disorders including Alzheimer disease,³² various types of cancer and cancer therapies,³³ dry eye,^34,35 AMD,³⁶ and more recently diabetic retinopathy³⁷. The inhibition of formation of amyloid aggregates by clusterin is an important feature of Alzheimer disease. Binding to misfolded proteins is thought to be related to clusterin’s surface hydrophobicity which is enhanced by acidic pH.²¹

Clusterin in Exfoliation syndrome and Exfoliation glaucoma

Clusterin is known to be expressed in ocular tissues relevant to XFS/EG including the iris and ciliary body.^38,39 Clusterin is found in higher amounts in aqueous humor from XFS patients compared with POAG or cataract patients.⁴⁰ Three different studies identified clusterin as a component of exfoliation material.^41–43 The relatively acidic pH of the aqueous humor would be expected to promote clusterin chaperone activity.²¹

An important study provided evidence that CLU mRNA expression is reduced in the iris, lens, and ciliary processes of XFS patients⁴⁴ and a follow-up study showed that expression of both CLU mRNA and protein was reduced in eyes of XFS patients with and without glaucoma in comparison to normal and POAG eyes.⁴⁵ This study also showed that TGFβ reduced CLU expression in ciliary body epithelial cells in vitro, suggesting that TGFβ levels could regulate CLU expression in XFS eyes.

CLU genetic associations

Alzheimer’ Disease (AD)

Clusterin inhibits the aggregation of amyloid beta (Aβ) peptide and participates in the clearance of Aβ via the blood-brain barrier,^32,46 processes integrally related to AD risk. There are other interesting similarities when comparing clusterin activities in AD and XFS/XFG. Similar to clusterin’s presence in aqueous humor and in exfoliation material, in the brain, CLU is highly expressed in both the cerebrospinal fluid,⁴⁷ amyloid plaques⁴⁸ and neurofibrillary tangles.⁴⁹ Aβ and clusterin combine together in normal cerebrospinal fluid, suggesting that reduced CLU expression may contribute in Aβ aggregation.⁴⁹

Genome-wide association studies have identified SNPs located in the CLU genomic region significantly associated with AD at the genome-wide level.^50,51 These studies have been replicated in most populations, with exceptions of African-Americans and Hispanics.⁵² Mutations in the Cys-rich region of the clusterin protein result in reduced secretion of clusterin, providing additional support for a critical role in disease development.⁵³

Exfoliation syndrome and exfoliation glaucoma

While CLU appears to be a promising candidate for an XFS/XFG risk factor, genetic studies have not provided consistent association results. A single CLU SNP (rs3087554) located in the 3’ untranslated region (Figure 2) was nominally associated with XFS in the Blue Mountain Eye Study (86 cases and 2,422 controls) at the genotypic level (P = 0.044), but not at the allelic level or when the age of controls was restricted to those over age 73 (P >0.07).⁵⁴ Significant association between the rs3087554 variant and XFS was not observed in a German (661 cases and 342 controls; P > 0.08) and Italian (209 cases and 190 controls; P > 0.70) case-control set, although a positive association for another CLU SNP, rs2279590 (also associated with AD) was reported in the German dataset only (P = 0.035).⁵⁵ A study of 136 cases and 89 controls from India did not find an association of rs3087554 with XFS (P >0.06), but did find a significant association with rs2279590 (P = 0.039).⁵⁶ However, a recent study of 299 cases and 224 controls from South India did not find significant association of rs3087554 or rs2279590 with XFS (P >0.43).⁵⁷ Additionally, a recent GWAS in a Japanese dataset of 1,484 cases and 1,188 controls showed nominal association of rs3087554 with XFS (P = 0.029), although the direction of effect for the minor allele ‘G’ is in the opposite direction compared with the effect in Caucasians with European ancestry.⁹ rs2279590 was not included in the Japanese study.

Figure 2. Features of the Clusterin gene (CLU) genomic region.

Clusterin gene (CLU) genomic region. The genomic location of the CLU exons (larger blue rectangles) and introns (blue lines) as depicted in the UCSC genome browser (http://ucsc.genome.edu) on chromosome 8 are shown. Beneath the exon/intron are the locations of the SNPs associated with Alzheimer disease (AD) (green type) and XFS/XFG (black type). rs2279590 is associated with both AD and XFS/XFG. The layered H3K27Ac histone marks (frequently found in regulatory regions) for 8 cell types studied by ENCODE are shown beneath the SNPs. rs3087554 and rs2279590 fall in a H3K27Ac peak in HUVEC cells (human umbilical vein). Rs3087554 and rs2279590 are also in the vicinity of a CLU miRNA binding site (miR-370) indicated in green below the H3K27Ac peaks.

Our group performed a meta-analysis of data from two US and Israeli datasets (314 cases and 446 controls) with published data from other Caucasian populations.⁵⁸ We found a statistically significant association of rs2279590 with XFS/XFG overall (summary OR = 1.18, 95% CI: 1.03–1.33, P = 0.01). No significant association was found for rs3087554 in Caucasian populations (summary OR = 0.90, 95% CI: 0.77–1.05, P = 0.17). Other CLU SNPs were not investigated.

Despite these important observations, genome-wide association for common CLU variants and XFS/XFG has not yet been observed. Significant genetic heterogeneity for CLU association has been observed among populations,⁵⁸ which could confound results in multi-ethnic GWAS. Alternatively, common CLU genetic variation may not underlie any clusterin function that could be involved in XFS/XFG. Rare CLU variation and complex genetic processes such as gene-gene and gene-environment interactions have not yet been fully evaluated in XFS/XFG cases and controls.

Rare CLU genetic variants have been evaluated in AD. In a sequencing study of AD cases an increased frequency of rare non-synonymous mutations in the CLU β-chain domain were identified in AD patients.⁵⁹ A subsequent report showed that three patient-specific CLU mutations in the B-chain caused abnormal subcellular CLU localization and diminished CLU transport through the secretory pathway.⁵³

Complex genetics and interactions

Complex genetic interactions such as gene-gene and gene-environment interactions are expected to contribute to diseases with complex inheritance such as XFS/XFG. CLU is an excellent candidate genetic risk factor for XFS/XFG, and while individual CLU SNPs have demonstrated only nominal association with disease,⁵⁸ these results do not exclude the possibility of involvement of CLU in complex interactions.

Interestingly, clusterin associates with altered elastic fibers in human photo-aged skin and prevents UV-induced elastin aggregation in vitro.⁶⁰ Recent studies suggest that CLU expression and clusterin glycosylation may be markers of cellular redox and oxidative stress which can be induced by UV light, suggesting that clusterin could link environmental factors related to UV light exposure with disease risk. Considering the significant effects of latitude on XFS risk,^61–64 which may involve ocular UV light exposures, gene-environment studies investigating interactions for UV light exposure and CLU SNPs could be of interest in a well-powered sample.

In AD, a significant gene-gene interaction has been identified for CLU and MS4A4E, coding for a membrane protein of unknown function⁶⁵. The odds ratio (OR) for the interaction on disease risk was substantially higher than that for CLU or MS4A4E alone, providing good support that the interaction could contribute to disease pathogenesis.

Clusterin as a therapeutic target

The important molecular chaperone activity of CLU, and its increased expression in response to cellular stress suggests that modulating CLU expression could have therapeutic potential, especially in diseases such as AD^32,48 and cancer⁶⁸ where CLU expression is increased. CLU is also a therapeutic target in dry eye.^34,35

CLU expression is increased in several cancers and is also increased after administration of chemotherapy, especially cisplatin, doxorubicin, Herceptin, Hsp90 inhibitors, and HDAC inhibitors.⁶⁶ In prostrate cancer decreasing CLU expression using an antisense oligonucleotide molecule reduced TGFβ signaling and blocked metastases in mice.^69,70 Subsequent studies have shown that reducing CLU expression improved the cytotoxicity of hormone-, radiation-, and chemo-therapies.^68,71 A CLU antisense molecule (OGX-011) showed promising safety and efficacy in phase I and II clinical trials for prostate and lung cancers in humans^72–74 but failed to demonstrate improved survival in a large phase III study for prostate cancer.⁷⁵

CLU’s role in protecting cells at fluid-tissue interfaces from the toxicity of protein oligomers has identified CLU as a possible therapeutic target in AD. However, while CLU is clearly involved in mediating the safe clearance and disposal of excess Aβ,⁴⁷ CLU activity in animal models is complex, making it difficult to design a therapeutic molecule or strategy. For example, in a mouse model of AD, reduction in CLU expression also reduced Aβ deposition and neurotoxicity; however if APOE was also reduced (also expected in AD patients) then reducing CLU resulted in increased Aβ amyloid deposition.^78,79 In in vitro studies, CLU can either potently inhibit Aβ amyloid formation or, at low ratios of CLU:Aβ, (e.g. 1:500), actually increase amyloid formation.⁴⁹ Therefore, currently its not clear if increasing or decreasing CLU expression would be beneficial in AD, a complex situation that could be similar to XFS/XFG (see discussion below).

In the eye, topical CLU has been shown to protect the corneal surface from desiccating stress in a pre-clinical mouse model of dry eye. The protection offered by CLU is likely due to its interaction with LGALS3 (Galectin-3),⁸⁰ a lectin known to be present on corneal epithelial cells, and MMP9, a metalloproteinase known to be involved in dry eye development.⁸¹ CLU is present in tears and perhaps protects the eye from desiccating stress by ‘sealing’ the surface through its interaction with these proteins.^34,35

Conclusions

Clusterin is an important extracellular molecular chaperone that protects cells at fluid-tissue interfaces from physical stress and removes potentially toxic protein oligomers. These features suggest that CLU could have a significant role in the development of XFS/XFG, a disease characterized by the accumulation of protein aggregates in fluid-filled spaces. The observations that clusterin is a component of XFS material and that CLU expression is altered in XFS/XFG also support a role for clusterin in disease pathogenesis. Genetically, CLU common variants are associated at the genome-wide level with AD, a disease also involving accumulation of aggregated protein. In XFS/XFG common CLU variants are only associated with disease risk at the nominal level. Future studies investigating rare CLU variation and complex interactions including gene-gene and gene-environment interactions in XFS/XFG cases and controls could be of interest.

Other than genetic studies there are important areas of research that could provide insight into the role of CLU in EXF/XFG. First, it is interesting that CLU expression is decreased in XFS/XFG when in other pathogenic conditions, such as AD and cancer, the expression is increased. Are there secondary factors that decrease CLU expression in XFS/XFG? Does the reduction in CLU expression contribute to increased aggregation of proteins? In this regard, the demonstration that increased TGFβ can decrease CLU expression may be relevant. It could also be of interest to investigate the ocular phenotype of CLU knock-out mice at older ages and in stressed environments.

Similar to AD, is it also possible that CLU contributes to the formation of the XFS aggregated protein? CLU is known to be a component of XFS aggregate material and is also known to have enhanced binding to ligands in mildly acidic environments such as would be expected in aqueous humor.

Finally, is clusterin a therapeutic target for XFS/XFG? It is tempting to speculate that increasing clusterin protein activity could result in reduction and/or removal of aggregated protein complexes in XFS/XFG eyes. However, the experience in AD suggests that the CLU-aggregated protein interaction is complex. Further investigations into the effects of overexpression of CLU, as well as reduction of CLU expression, ideally in relevant animal models, are needed before approaching CLU as a therapeutic target for XFS/XFG.