Quantitative Phosphoproteomic Comparison of Lens Proteins in Highly Myopic Cataract and Age-Related Cataract

Purpose To investigate and compare the lens phosphoproteomes in patients with highly myopic cataract (HMC) or age-related cataract (ARC). Methods In this study, we undertook a comparative phosphoproteome analysis of the lenses from patients with HMC or ARC. Intact lenses from ARC and HMC patients were separated into the cortex and nucleus. After protein digestion, the phosphopeptides were quantitatively analyzed with TiO2 enrichment and liquid chromatography-mass spectrometry. The potential functions of different phosphopeptides were assessed by Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. Results In total, 522 phosphorylation sites in 164 phosphoproteins were identified. The number of phosphorylation sites was significantly higher in the cortex than in the nucleus, in both ARC and HMC lenses. The differentially phosphorylated peptides in the lens cortex and nucleus in HMC eyes were significantly involved in the glutathione metabolism pathway. The KEGG pathway enrichment analysis indicated that the differences in phosphosignaling mediators between the ARC and HMC lenses were associated with glycolysis and the level of phosphorylated phosphoglycerate kinase 1 was lower in HMC lenses than in ARC lenses. Conclusions We provide an overview of the differential phosphoproteomes of HMC and ARC lenses that can be used to clarify the molecular mechanisms underlying their different phenotypes.


Introduction
Cataract is an age-related degenerative disease and the principal cause of blindness worldwide [1]. It frequently accompanies other eye diseases, such as high myopia, glaucoma, uveitis, and trauma, all of which displaying clinical processes distinct from those of age-related cataract (ARC). Highly myopic cataract (HMC) is more prevalent in Asia than in other regions [2][3][4]. Compared with ARC, HMC is characterized by earlier onset and cataract with greater nuclear sclerosis and rapid progression [5,6], which implies that there are unique pathological processes involved in the development of each type of cataract. However, the underlying molecular differences between ARC and HMC remain unclear.
Lens proteins are some of the most long-lived proteins in the body and are the targets of numerous posttranslational modifications [7]. These modifications, especially phosphorylation, are implicated in the regulation of protein solubility and activities [8]. Previous studies demonstrated the differential expression of phosphorylated proteins in normal and cataractous lenses and suggested that protein phosphorylation affects the occurrence and development of cataract [9]. Given the vital role of phosphorylation in protein denaturation and the intriguing clinical differences between HMC and ARC, the differential phosphoproteomes in these two conditions warrant investigation. Differences in the phosphorylation of lens proteins could help distinguish the phenotypes of HMC and ARC lenses.
In this first comparative study of the differential phosphorylation status of HMC and ARC lenses, we quantified the phosphoproteomes of the lens cortex and nucleus separately. This was followed by Gene Ontology (GO) functional analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of the differentially phosphorylated proteins to determine the implications of phosphorylation to the unique phenotype of HMC.

Materials and Methods
In this study, we analyzed the phosphoproteomes of HMC and ARC lenses to detect differences between HMC and ARC. The workflow of this study is shown in Figure 1.

Tissue Collection.
In total, three HMC lenses and three age-matched ARC lenses were collected from patients, who provided informed consent before they underwent extracapsular cataract extraction surgery at the Eye and Ear, Nose, and Throat Hospital, Fudan University, Shanghai, China. The research strictly adhered to the tenets of the Declaration of Helsinki and was approved by the Ethics Committee of the Eye and ENT Hospital, Fudan University. Patients with an axial length of ≥26 mm were diagnosed with high myopia. In all lenses, the nucleus and cortex were separated by coring through the visual axis with a 4.5 mm diameter trephine. The lens tissues were divided into four groups: ARC lens cortex (ARC-C), ARC lens nucleus (ARC-N), HMC lens cortex (HMC-C), and HMC lens nucleus (HMC-N).

Sample Preparation.
Each tissue was lysed with 200 μL of lysis buffer (4% sodium dodecyl sulfate, 100 mM dithiothreitol, 150 mM Tris-HCl, pH 8.0), disrupted with agitation using a homogenizer, and boiled for 5 min. The samples were ultrasonicated and boiled again for another 5 min. Insoluble material was removed by centrifugation at 12,000 g for 15 min. The supernatant was then collected for protein digestion. The protein concentrations were quantified with a BCA Protein Kit (Bio-Rad, Shanghai, China).

Protein Digestion.
We digested the proteins in 250 μg of each sample with the filter-aided sample preparation procedure described by Wisniewski et al. [10]. Briefly, 200 μL of uric acid (UA) buffer (8 M urea, 150 mM Tris-HCl, pH 8.0) was used to remove the detergent, dithiothreitol, and other low-molecular weight components with repeated ultrafiltration. To block the reduction of cysteine residues, 100 μL of 0.05 M iodoacetamide in UA buffer was added and the samples were incubated for 20 min in the dark. The filter was washed three times with 100 μL of UA buffer and then twice with 100 μL of 25 mM NH 4 HCO 3 . The protein suspension was then digested with 3 μg of trypsin in 40 μL of 25 mM NH 4 HCO 3 . The mixture was incubated at 37°C overnight, and the resulting peptides were collected as the filtrate.

Enrichment of Phosphorylated Peptides with TiO 2 Beads.
The phosphopeptides were captured according to the TiO 2 protocol [11], adapted for label-free quantitative proteo-mics. The peptides were concentrated with a vacuum concentrator and resuspended in 500 μL of loading buffer (2% glutamic acid, 65% acetonitrile (ACN), and 2% trifluoroacetic acid (TFA)). The TiO 2 beads were added and then agitated for 40 min and centrifuged for 1 min at 5,000 × g . The supernatant was mixed with another TiO 2 bead, resulting in the second beads which were collected as before. The beads were washed sequentially with 50 μL of washing buffer I (30% ACN, 3% TFA) and 50 μL of washing buffer II (80% ACN, 0.3% TFA) three times to remove the remaining unabsorbed material. The phosphopeptides were finally eluted with 50 μL of elution buffer (40% ACN, 15% NH 4 OH) [12]. The eluates were lyophilized for further analysis.

Liquid Chromatography-(LC-) Electrospray Ionization
Tandem Mass Spectrometry (MS/MS) Analysis with Q Exactive™. The peptides from each sample were desalted with a C18 Cartridge (Empore™ SPE Cartridges C18 (standard density), bed I.D. 7 mm, volume 3 mL; Sigma), concentrated with vacuum centrifugation, and equilibrated with 40 μL of 0.1% (v/v) TFA. MS experiments were performed on a Q Exactive mass spectrometer coupled to an Easy nLC™ liquid chromatography (Thermo Fisher Scientific). The phosphopeptide extract (5 μg) was injected onto a C18 reversed-phase column (Thermo Scientific EASY-Spray™ Column, 10 cm long, 75 μm I.D., 3 μm resin) in buffer A (2% ACN, 0.1% formic acid) and separated with a linear gradient of buffer B (80% ACN, 0.1% formic acid), at a flow rate of 250 nL/min over 60 min. The most abundant precursor ions from the survey scan (300-1800 m/z) for higher-energy C trap dissociation (HCD) fragmentation. The target value determination was based on predictive automatic gain control. The duration of dynamic exclusion was 25 s. Survey scans were acquired at a resolution of 70,000 at 200 m/z, and the resolution of the HCD spectra was set to 17,500 at 200 m/z. The normalized collision energy was 30 eV, and the underfill ratio, which specifies the minimum percentage of the target value likely to be reached at the maximum fill time, was defined as 0.1%. All MS experiments were performed in triplicate for each sample.
2.6. Sequence Database Searches and Data Analysis. All the raw data were identified with the MaxQuant software (version 1.3.0.5.) and screened against the UniProt human database, containing a total of 156,914 entries. The datasets were searched with a mass tolerance of 6 ppm. The search followed the enzymatic cleavage rule: trypsin/P, allowing two missed cleavages; tolerance on a mass measurement of 20 ppm; fixed modification; and carbamidomethylation of cysteines. Protein N-terminal acetylation and methionine oxidation were defined as variable modifications. The cutoff for the global false discovery rate for peptide and protein identification was set to 0.01. Label-free quantification was performed with MaxQuant, as previously described [13]. Protein abundance was calculated based on the normalized spectral protein intensity (label-free quantification (LFQ) intensity). In the

Statistical Analysis.
To identify quantitative differences in the phosphorylation states between each group, the degree of phosphorylation at the same site was estimated as the difference ratio and a ratio of >2 was considered to indicate overabundant phosphorylation. Conversely, a reduction in phosphorylation < 0:5-fold was considered to indicate less abundant phosphorylation [14,15]. Statistical significance was determined with a t-test. A p value < 0.05 was considered statistically significant.

Phosphoproteome Identification in HMC and ARC
Lenses. The clinical information for the lens samples is given in Table S1. In this study, we identified 451 unique phosphopeptides in 164 phosphoproteins from the HMC and ARC lens samples (Table S2). Among the 164 phosphoproteins, 84 contained a single phosphorylation site, 26 contained two phosphorylation sites, and 17 contained three phosphorylation sites (Figure 2   To investigate the quantitative differences in the phosphorylation status of the four groups (ARC-C, ARC-N, HMC-C, and HMC-N), a two-fold change was used as the cutoff to screen for differentially phosphorylated proteins. The level of phosphorylation was higher in the cortex than the nucleus, in both the HMC and ARC lenses. Twenty-six phosphopeptides were hyperphosphorylated in the HMC-C, whereas 10 were underphosphorylated compared with the HMC-N. In the ARC lenses, we detected 104 more abundant phosphopeptides and eight less abundant phosphopeptides in the cortex relative to the nucleus. A comparison of the HMC and ARC lenses revealed that the level of phosphorylation was higher in the HMC-N group than in the ARC-N group. We identified 58 phosphopeptides that were significantly altered in the HMC-C group: 23 were more abundant and 35 were less abundant than in the ARC-C group (Figures 3(a) and 3(c)-3(e)). Among the phosphosites identified, 303 (77.3%) were shared by all four groups (Figure 3(b)).
Seventeen phosphosites were exclusively detected in HMC-C, but not in HMC-N, that included cytoskeletal proteins, oxidoreductases, and binding proteins ( Table 1). The phosphorylation status of S-formylglutathione hydrolase and glyceraldehyde-3-phosphate dehydrogenase was greater in the HMC-C than in the HMC-N. Meanwhile, in comparisons of ARC-C and ARC-N, a total of 33 phosphosites were exclusively detected in ARC-C and the phosphorylation degree of β-crystallin A3 (phosphorylation at site Y36) was much higher in ARC-C, with a different ratio of 34.04. Other significantly overabundant phosphosites were also detected in phakinin, β-crystallin B, and filensin (p < 0:05) ( Table 2).
Twelve and 14 phosphosites were exclusively detected in the HMC-C and ARC-C, respectively, when comparing these groups. A protein with a high degree of phosphorylation was γ-crystallin D, which was phosphorylated at Y29 ( Table 3). Phosphorylation of the lens cytoskeletal proteins, phakinin and filensin, was significantly lower in the HMC-C than in the ARC-C. In the HMC-N and ARC-N, the predominant differentially expressed phosphosites were found in the αcrystallin B chain, at S21, T170, and S76. Table 4 shows that the proteins with the most abundant phosphorylation included crystallins and structural proteins, particularly βcrystallin B1 and filensin.

Gene Ontology and KEGG Pathway Enrichment Analyses.
The potential functions of the phosphoproteins differentially        The GO analysis showed that the proteins differentially phosphorylated in the HMC-C and HMC-N were enriched for proteins involved in the cellular compartment, especially in the extracellular exosome, cytoplasm, and plasma. The molecular functions of these proteins were mainly related to the structure of the lens (Figure 4(a)). The GO analysis showed that the differentially phosphorylated phosphopeptides in the ARC lenses were enriched for visual perception, nerve impulse transmission, and protein homodimerization. The differentially phosphorylated phosphoproteins also included cell components, mostly related to the extracellular exosome and cytoplasm, that are involved in the structure of the lens and ATP binding. When the HMC and ARC groups were compared, the differentially phosphorylated proteins were functionally related to the structure of the lens or were crucial for visual perception in the cortical and nuclear regions. Concerning cellular component, the differentially expressed phosphoproteins in the HMC-C and ARC-C were mainly cytoplasmic proteins. However, the differentially phosphorylated proteins in the nuclear regions of the HMC and ARC groups (HMC-N and ARC-N, respectively) were predominantly extracellular exosome proteins (Figures 4(c) and 4(d)).
We also performed KEGG pathway enrichment analysis to identify the biological pathways associated with the differentially phosphorylated proteins. The 20 most abundant enrichment terms with p < 0:05 are shown in Figure 5. When the cortex and nucleus of HMC were compared, the most significantly enriched pathway was glutathione metabolism. However, when the HMC and ARC lenses were compared, glycolytic enzymes were most frequently differentially expressed.
As shown in Figure 6, when we compared the HMC and ARC lenses, the differentially phosphorylated proteins were enriched in the glycolysis and glutathione metabolism pathways. The key glycolytic enzyme, phosphoglycerate kinase 1 (PGK1), was the least phosphorylated protein in the HMC lenses. However, glutathione synthetase (GSS) and glutathione-disulfide reductase (GSR), the key enzymes in glutathione synthesis, were hyperphosphorylated in HMC.

Discussion
To improve the efficiency of identifying the phosphorylation sites in this study, we used TiO 2 enrichment combined with LC-MS/MS. For the first time, two parts of the lens with known histological differences, the lens cortex and nucleus, were compared separately. Proteomic differences were detected in the different regions of the lens, and by quantifying the differences in the phosphorylated proteins between HMC and ARC, we clarified the different pathogeneses in these two phenotypes.
We identified 522 phosphorylation sites in 164 phosphoproteins in this study. Previous studies have reported 73 phosphorylation sites and 32 phosphoproteins in normal and cataractous lenses, using immobilized metal affinity chromatography and nano-LC-coupled MS/MS [15]. α-Crystallin A and α-crystallin B are the most abundantly phosphorylated proteins in the porcine lens [16]. Our data show that besides these two crystallin proteins, beta A3, beta B1, beta B2, beta S, and gamma D crystallins were also phosphorylated at many peptide sites.
In this study, the number of phosphorylation sites was significantly greater in the lens proteins of the cortex than in those of the nuclear region, in both HMC and ARC. One possible explanation to this finding is that the lens epithelium cells immediately adjacent to areas of the cortex are metabolically relatively active and metabolites decrease from the lens cortex toward the lens nucleus. Protein enzymatically phosphorylated in the outer cortex could gradually dephosphorylate nonenzymatically in the metabolically inactive nucleus. Among the phosphoproteins in these groups, a high percentage of the differentially phosphorylated proteins were crystallins and lens structural proteins, including β-crystallin, α-crystallin, phakinin, and filensin.
α-Crystallin is a small heat shock protein that maintains the transparency of the lens. Phosphorylation is considered to change its chaperone activity by inducing a change in the protein's structure and altering the subunit exchange dynamics [17]. The phosphorylation of α-crystallin B has been shown to regulate the protein's activity in both lenticular and extralenticular tissues [5,[18][19][20][21][22][23]. The commonest functional modification sites in α-crystallin B are S19, S45, and S59. We detected these phosphopeptides in the lenses of patients with both HMC and ARC. When we compared the S19 site between these two groups, the level of phosphorylation was higher in the HMC-C than in the ARC-C. However, the pathological significance of these proteomic changes requires further analysis. There was a slight difference in the phosphorylation at S59 between the ARC-C and ARC-N and between the HMC-N and ARC-N. The phosphorylation of α-crystallin B at S59 is thought to be associated with actin nucleation and the migration of lens epithelial cells [18]. We detected no difference in the phosphorylation at S45 in any paired comparison of the four groups. Filensin and phakinin are two unique protein components of the lens fibers that assemble to form an intermediate filament, known as the beaded filament [24][25][26][27]. As previously reported, filensin and phakinin in the lens fiber cells are essential for maintaining the transparency of the lens [17]. As the lens fiber differentiates and with aging, these proteins become the targets of phosphorylation as a posttranslational modification [18]. It has also been reported that the phosphorylation of intermediate filament proteins plays an essential role in regulating the kinetics of these proteins, including their solubility, conversion, and the fiber structure [19][20][21].
In the HMC group, the phosphosites that differed strongly in their phosphorylation between the cortex and nucleus predominantly comprised enzymes involved in glutathione synthesis, including GSS and S-formylglutathione hydrolase. Our KEGG pathway analysis also showed that the largest proportion of phosphoproteins was associated with glutathione metabolism.
GSS and S-formylglutathione hydrolase catalyze key steps in glutathione synthesis. Glutathione is an essential

16
BioMed Research International antioxidant that protects the lens from oxidative damage [22]. The level of glutathione synthesis is lower in the cataractous lenses than in the lenses of emmetropic eyes but is lowest in myopic lenses [23]. It has also been demonstrated that eyes with high myopia are susceptible to oxidative damage and are associated with an increased incidence of nuclear cataract (with an adjusted odds ratio of 3.01) [5]. The lens typically exists in a low-oxygen environment [28,29], and increased exposure to oxygen appears to cause cataract. Previous studies have shown that the degree of vitreous liquefaction is positively correlated with the level of nuclear opacity in the lens after adjustment for age [30]. As a possible mechanism, vitreous liquefaction increases the flow of fluid in the vitreous cavity and allows oxygen to flow from the retina to the lens. In patients with high myopia, vitreous liquefaction often occurs in the early stage of myopia and the severity of this complication increases as myopia worsens [31]. The glutathione content varies between different types of cataract. Subcapsular cataract, with an additional secondary nuclear cataract, shows a particularly rapid reduction in glutathione [32]. As a result, highly myopic eyes are more susceptible to oxidative damage than less myopic eyes, which leads to the formation of nuclear cataract. Consistent with this, our experimental data show that the number of phosphorylated glutathione synthase molecules was significantly higher in the cortex of the HMC lens than in the nucleus. By comparing the catalogues of differentially phosphorylated protein in the cortical regions of HMC and ARC, results showed that the degree of phosphorylation of GSS and GSR, the key enzymes of glutathione synthesis, was higher in HMC than in ARC. However, the precise roles of GSR and GSS phosphorylation remain unclear.
In our comparison of the two lens tissues in HMC and ARC, the number of phosphopeptides was much higher in the HMC-N than in the ARC-N, which may be associated with the severity of nuclear cataract in patients with high myopia. Truscott [33] proposed that there is a barrier to the transport of metabolites within the lens. This barrier may increase the half-lives of reactive molecules, thus promoting the posttranslational modification of proteins in the nucleus, and may also prevent an adequate flux of antioxidants reaching the lens interior, thus allowing the oxidation of the nuclear components. Other authors have suggested that a common underlying mechanism in the pathology of cortical and nuclear cataract is the failure of the microcirculatory system to regulate the cell volume in the lens cortex or to deliver antioxidants to the lens nucleus [34]. Therefore, we suggest that the nuclear region of the cataractous lens may be a meaningful target region for the posttranslational modification of proteins. The HMC lens may be the best model to study the transfer of antioxidants to the nucleus through the barrier.
When we examined the glycolysis and metabolic pathways, which were enriched in differentially phosphorylated proteins in HMC and ARC, we found that the degree of phosphorylation of PGK1 was lower in HMC-C and HMC-N than in the ARC-C and ARC-N, respectively. A previous study showed that phosphorylation of PGK1 reduces its activity, thus reducing the glycolytic activity [35]. Glycolysis is the main source of energy generation in the unique low-oxygen environment of the eye. A reduction in energy metabolism impairs the activity of NA + -K + -ATPase in the lens, and the cascade reaction leads to an imbalance in lens homeostasis.

Conclusion
In summary, we analyzed the phosphoproteomes of the cortex and nucleus of HMC and ARC lenses, while considering the clinical features of the lenses. We found significant differences in the extent of protein phosphorylation and the types of proteins phosphorylated between different regions of the lens. Our results will be valuable for the future investigation of the molecular characteristics and pathological pathways underlying HMC and ARC.

Data Availability
The data supporting the findings of this study are available within the article and its supplementary materials.

Conflicts of Interest
The authors have no relevant interests to declare.