Age-related aqueous humor (AH) and lens epithelial cell/capsule protein carbonylation and AH protein concentration in cataract patients who have pseudoexfoliative diseases

• Introduction
• Methods
• Results
• Discussion

Abstract

Purpose: The aim of this study is to investigate the age-correlation of oxidative stress (OS, assessed by the accumulative OS damage marker protein carbonyls) in aqueous humour (AH; together with protein concentration) and lens epithelial cells plus capsule (LECs/capsule) in patients with cataract (CAT), and also suffering from pseudoexfoliation syndrome (PEX), primary open-angle glaucoma (POAG) and pseudoexfoliation glaucoma (PXG).

Methods: AH samples from 78 male/female patients (21, 20, 19 and 18 with CAT, PEX, PXG, and POAG, respectively), and LECs/capsule samples from 104 male/female patients (34, 32, 18, and 20 with CAT, PEX, PXG and POAG, respectively) were collected during phacoemulsification CAT surgery. Average protein carbonyl concentrations were measured in patients grouped in 5-year age intervals (ranging from 56-60 to 86-90). The non-overlapping age ranges and numbers of the tested subjects did not allow comparative follow up studies for the tested diseases.

Results: There is an age-dependent increase of protein carbonyls in AH (nmol mg^-1 protein and ml^-1), and in the order CAT<PXG=~POAG<PEX and CAT<PEX=~POAG<PXG respectively. Moreover, protein concentration in AH accumulates in the order CAT<PEX<POAG<PXG but is not age-related. An age-dependent increase of protein carbonyls (nmol mg^-1 protein) is also observed in LECs/capsule, and in the order CAT<PEX=~POAG<PXG.

Conclusions: The present study shows for the first time an age-increased OS-induced protein damage (protein carbonyl formation) in the AH and LECs/capsule of CAT patients with PEX, POAG or PXG. The slow rate of change of protein carbonyls strongly suggests a long-term implication of OS in ocular disease pathogenesis. Additionally, protein concentration levels in the AH of CAT patients increase independently of age, and in same as with protein carbonyls increasing order levels for CAT<POAG<PXG in AH and LECs/capsule. This may suggest a protein cross-diffusion taking place between AH and LECs/capsule, most likely originating from PEX deposition and/or necrotic/apoptotic LECs/capsule. Moreover, the findings of the present study establish the use of protein carbonyls (together with a methodology for their more accurate quantification, which overcomes serious unreliability problems of past methods) as an age accumulative marker of OS damage, for future studies that investigate long-term OS involvement in pseudoexfoliative ocular disorders.

Although the pathogenesis of cataract, pseudoexfoliation (PEX), secondary open-angle glaucoma (PXG), and primary open-angle glaucoma (POAG) is not fully understood, there is growing evidence that oxidative stress plays an important role, given that it is implicated in the etiology of many ocular diseases [1-9], such as PEX [10-14], PXG [15], and POAG [16,17]. However, the central question is how reliably oxidative stress has been measured in these studies. The most reliable evaluation of oxidative stress is via the direct quantification of the reactive oxygen species (ROS) components. The most representative are the superoxide and hydroxyl free radicals. However, the in vivo quantification of these free radicals in patients is not applicable due to bioethical restrictions. Alternatively, indirect methods for assessing oxidative stress have been applied and focus mainly on the oxidative damage ROS cause in biomolecules, such are lipids, proteins, and DNA. However, more reliable are those oxidative modifications that are not repairable and thus, accumulated, with protein carbonylation the most representative.

This necessitates the investigation of past experimental approaches in their attempt to correlate oxidative stress and ocular diseases in humans in both the epithelial cells plus the capsule (LECs/capsule) and in the aqueous humor (AH). LECs/capsule exposure to various oxidative stress factors has been associated to several conditions, with CAT being more extensively studied [6,18-22]. Oxidative stress development in LECs/capsule has been assessed by certain biological indicators: activity of certain antioxidant enzymes (superoxide dismutase, SOD, catalase, CAT) [23,24]; genes encoding antioxidant defense enzymes (aldehyde dehydrogenase 1-ALDH1A1, OMIM 603687; microsomal glutathione S-transferase 1-MGST1, OMIM 138330; superoxide dismutase 2-SOD2, OMIM 147460) [25]; proteins involved in multiple protection paths (e.g., metallothioneins) [26,27]; levels of non-enzymatic antioxidants (reduced and oxidized glutathione, GSH and GSSG, respectively) [12,28]; DNA oxidative damage (8-OH-dG, 8-oxo-Gua, DNA strand breaks, pyrimidine dimers, and telomeres [29-34]); lipid peroxidation (malondialdehyde [MDA]) [12,28]; and protein oxidative modification (i.e., carbonyl groups) [21,28].

Similarly, oxidative stress changes in the AH may also be associated with ocular diseases [6,20,35-37]. Oxidative stress in human aqueous humor has been also assessed with certain biochemical indicators: DNA oxidative damage evaluation (8-OH-dG, 8-oxo-Gua) [17,38]; lipid peroxidation (MDA, 8-iso-PGF_2a, conjugated dienes, and lipofuscin-like fluorescent end products) [10,39–42]; antioxidant enzymes (SOD, CAT, glutathione peroxidase [GPx], and arylesterase, paraoxonase) [14,39,40,42-46]; prooxidant enzymes (xanthine oxidase) [47]; antioxidant enzyme gene expression (for SOD and glutathione transferase [GST]) [48]; non-enzymatic antioxidants (ascorbic acid, vitamin E, selenium, uric acid, GSH, and GSSG) [3,10,11,46,49-51]; superoxide radical (O₂^•−) scavenging activity [52]; total antioxidant capacity [14,17,39,41,43,50,53-55]; total oxidant status [14,54]; oxidant/antioxidant balance [44]; ROS levels (e.g., H₂O₂) [44,52]; total inorganic/lipid/protein/nucleic acid hydroperoxides [52]; and protein oxidative modification (carbonyl groups) [3,56,57].

Oxidative carbonylation of proteins is the most reliable indicator of oxidative stress for the following reasons: Proteins that have been carbonylated are accumulated because they are not readily recognized by proteasomes. Although moderately carbonylated proteins are degraded by the proteasomal system [58], the heavily carbonylated ones tend to form high-molecular-weight aggregates that escape proteolytic degradation by the proteasome [59,60]. This is possibly because carbonylated protein aggregates exhibit structural constraints that prevent recognition of the aggregates by the catalytic sites inside the cylinder of the proteasome complex. Such aggregates accumulate in cells as damaged or unfolded proteins [61]. Moreover, these carbonylated protein aggregates can inhibit proteasome activity [60], cause a progressive further increase in protein aggregation and cross-linking in non-dividing (post-mitotic) cells, and may eventually induce apoptosis [61,62]. Carbonylated aggregates can also become cytotoxic and have been associated with numerous diseases, including cataractogenesis [63,64]. Another reason for the limited carbonylated protein withdrawal by proteasomes is that they can become the target of carbonylation (e.g., subunit S6 ATPase) and other oxidative modifications (e.g., glycoxidation and modification with lipid peroxidation products) [65], ending up (e.g., 26S proteasome) in decreased activity [66]. Other advantages of using carbonylated proteins as indicators of oxidative stress is their chemical stability which allows easy detection even after long storage [67] and the existence of many assays for detection [68-75].

Protein oxidative damage is exerted on ROS-labile aromatic and sulfur-containing amino acids in proteins. This is especially relevant to lens proteins as they contain large amounts of tryptophan, tyrosine, phenylalanine, histidine, methionine, and cysteine amino acids, which can be modified by ROS, forming adducts and aggregates and affect enzyme function [76]. These events may be related to an increase in the protein concentration in AH, which has been observed in some studies on cataract [45], PEX [15,77-81], and PXG [78,82-84].

The use of protein carbonylation to assess oxidative stress in ocular diseases is limited, and the methodologies employed quite unreliable. Specifically, protein carbonyl levels were assessed in the lens epithelial cell fraction of smokers or diabetic patients with cataract [21], in non-cataractous human LEC cultures [28], and in the AH of patients with PEX [57] with an unreliable version of the 2,4-dinitrophenylhydrazine (DNPH)-based assay [85,86]. In another study, carbonyl groups were measured semiquantitatively (with enzyme-linked immunosorbent assay [ELISA] using an anti-DNP antibody) in the AH of patients with cataract [56]. These studies did not investigate oxidative stress as a function of a patient’s sex and age.

The present study proposes the very reliable oxidative stress marker protein carbonyls, and the methodology for its quantitative assessment, in order to evaluate more accurately the role of oxidative stress in pseudoexfoliative ocular diseases. The study specifically evaluates the protein carbonyl levels, as a function of age and sex, in the AH and LECs/capsule of CAT patients suffering also from PEX, POAG or PXG, and also correlates the protein carbonyls in AH with protein concentration. Since some of the carbonyls measured in the LEC samples may have come from oxidized proteins in the capsule basement membrane, this is why these samples is more appropriate to be referred as LECs/capsule.

LECs/capsule and AH average carbonyl value for each 5-year interval age group in patients with CAT, PEX, PXG, and POAG was plotted as a curve (fitted straight line) vs the studied 5-year-age intervals. Age-correlated protein carbonyl content was expressed as the resulting straight line curve slope (representing the carbonyl content change versus increasing 5-year age interval), and defined as nmoles carbonyl groups mg^-1 aqueous protein vs age for AH and LECs/capsule, while for AH as nmoles carbonyls ml^-1 versus age. In addition, the study investigated the age-correlated association of protein concentration in AH.

The present study investigated for the first time the age (and sex)-related association of ROS in the AH and LECs/capsule of CAT patients that also suffer from PEX, POAG or PXG. The study used the specific oxidative marker of protein carbonylation, which is a very reliable representative of high oxidative stress mainly due to its intracellular accumulation as not being repairable. It should be noted that it became possible to quantify this marker with high reliability and sensitivity only until recently [89].

Until now, the association of ocular diseases with oxidative stress via protein carbonyls in AH was limited to three studies [3,56,57], which used methodologies that are known to have serious reliability problems [89]. This may be one reason that the findings of these studies are partially correlated to those of the present study. Other reasons are that (a) protein carbonyl levels were measured only in PEX (also using cataract as control), (b) these levels were not correlated with age (or sex), and (c) measurements were performed with either ELISA [56] or a version of the standard DNPH photometric assay [57], both known to be unreliable [89].

Specifically, Yagci et al. showed a 66% increase in the protein carbonyl concentration (per volume) over control cataract in AH of patients with PEX with an age span of 66.9 ± 8.30 years [57]. This result is in agreement with the approximate 60% increase observed in the present study (after the carbonyl values were averaged for the similar patient age span 56 to 70 years). However, the protein carbonyl level increase (0.87 nM) over that of the cataract control derived from the data presented in the Yagci et al. study is about 2,300-fold lower (in absolute number) than that for PEX in the present study (Figure 1A). Nonetheless, this carbonyl value (0.87 nM) is questionable because it is about 1,000-fold lower than the sensitivity limit 0.91 µM of the standard DNPH assay (determined at minimum absorbance 0.02 and for DNPH molar absorptivity 22.000 M⁻¹ cm⁻¹ at 360 nm [85,86]). Similarly, Alamdari et al. showed a 47% increase in the protein carbonyl concentration per milligram of AH protein (over control patients with cataract) in patients with PEX of unspecified, however, age [56]. This is also in agreement with the approximate 48% increase using the average value of the patients tested in the present study (covering an age span of 71 to 90 years; Figure 1B). However, the protein carbonyl increase over control derived from the data from the Alamdari et al. study in absolute number (0.32 nmol mg⁻¹) is about 40-fold lower (Figure 1B). The present study, however, for the first time presented systematic data for the carbonyl rate change values per AH milligram of protein and volume in the increasing disease order cataract<PXG=~POAG<PEX and cataract<PEX=~POAG<PXG, respectively (Table 1).

Similarly, there are no studies available on the association of PEX, POAG and PXG neither with oxidative stress (via protein carbonyls) nor with age (and sex) in patients’ LECs/capsule. Thus, there are no data from other studies to corroborate the findings of the present study that there is an age-related increase of protein carbonyl levels per mg protein and in the increasing order CAT<PEX=~POAG<PXG (Figure 2). Previous studies limited protein carbonyl assessment in smokers or diabetic patients with cataract (an age group of 58 to 75 years [21]) and in non-cataractous human LEC cultures [28], in both cases determined with an unreliable version of the DNPH-based assay [89].

The present study also investigated the association of the protein concentration in AH and found that it is not age-related and increased among the studied ocular diseases in the following order cataract=~PEX<POAG<PXG (Figure 3). Interestingly, this increasing order in AH is in agreement with that of the protein carbonyl level change per volume for POAG and PXG. Most of the studies on AH-associated ocular diseases have determined the protein concentration in one or more of the same diseases [15,77-80,82-84,92-102]. The only study of patients who had the same diseases (male and female, non-age correlated, and using an age mean value of 72.1 ± 7.2 years) [78] reported a different order (cataract<POAG<PEX<PXG) than that of the present study. However, when grouping in the same disease type and averaging the protein concentration values reported by the previous studies, their resulting mean values showed an increasing protein concentration order (cataract<PEX<POAG<PXG with corresponding values 0.28<0.326<0.431<0.473 mg ml⁻¹), that coincides (in terms of increasing order and absolute values) with those of the present study.

LECs/capsule and AH are of great interest for oxidative stress related studies in ocular diseases since LECs/capsule are strongly metabolizing and in contact with AH, which is also involved in metabolic exchange (i.e., supplies nutrients and oxygen through diffusion, removes metabolic wastes etc.) [103,104]. Moreover, pseudoexfoliation disorders are characterized by abnormal fibrillary material production and its deposition in the anterior segment, such as the iris, lens surface, ciliary body, zonula, trabecular meshwork, and corneal endothelium, which can lead to the development of PXG and POAG.

The fact that the age-dependent increasing rate of change in the protein carbonyl levels (in AH and LECs/capsule) and the age-independent increase of the protein levels (in AH) coincide in increasing order with the increasing levels of protein carbonyls at a disease order CAT<POAG<PXG (with the levels in PEX being variable; Table 1), the following can be suggested: The causative factors of oxidative stress in CAT lens may produce proteins (as part of the deposited PEX material or possibly originating from necrotic/apoptotic LECs/capsule), with some of them being carbonylated and possibly become dissolved in the AH and also cross-diffused between AH and LECs/capsule.

Concluding, the present study correlates, for the first time, increasing age-associated oxidative stress in the AH and LECs/capsule of patients with CAT, PEX, POAG and PXG. The relatively slow age-related rate of change observed for the levels of protein carbonyls in the AH and LECs/capsule of patients with the tested ocular diseases, together with the accumulative nature of protein carbonylation, strongly suggest a long-term implication of OS in their pathogenesis. Moreover, the findings of the present study establish the use of protein carbonyls (together with a new methodology for their more accurate quantification, which overcomes serious unreliability problems of past methods) as an age accumulative marker of OS damage. Nonetheless, protein carbonyls assess only indirectly the long-term association of OS with the pathogenesis of each of the aforementioned pseudoexfoliative ocular disorders. Moreover, AH and LECs/capsule extremely small sample size (30-60µl and 5mm diameter single-layer cells, respectively) limits the simultaneous complementary testing, ideally in each sample, of other indirect OS markers. Therefore, this association needs to be verified also by direct markers and initiators of OS, such as the superoxide and hydroxyl free radicals. We plan to conduct such studies upon finalizing the development of specific methodologies for the in vivo quantification of these free radicals.