Membrane Association of Proteins in the Aging Human Lens: Profound Changes Take Place in the Fifth Decade of Life

Abstract

purpose. To characterize age-related changes to proteins in the center of the human lens.

methods. Human lenses of different ages were dissected using trephines. Sucrose density gradient centrifugation was used to separate the proteins from two defined nuclear regions. Densitometry of Coomassie-stained protein bands was compared with lipid analysis with the use of mass spectrometry.

results. A profound change in the density gradient profiles of lenses occurred at approximately age 40. As soluble crystallins decreased, four higher density bands appeared that were absent in younger lenses. These four bands contained crystallins, as well as membrane lipids, and appear to have resulted from the interaction of denatured crystallins with fiber cell membranes.

conclusions. Changes in lens proteins and membranes can be detected in each decade of life; however, major changes to the lens crystallins of the nucleus take place between age 40 and 50, after the loss of free soluble alpha crystallin. These alterations are consistent with large-scale binding of crystallin aggregates to fiber cell membranes after middle age.

The human lens grows continuously throughout life. Defined stages of growth can be distinguished based on suture patterns. ¹ As depicted by the Kuszak ² and Garland ¹ groups, certain lens regions can be readily differentiated. For example, the fetal nucleus (4-mm diameter) is characterized by a Y suture pattern ^{1 2} and the infantile nucleus (5.5- to 6-mm diameter) by a suture pattern with six arms. ¹ The fetal nucleus is synthesized early in utero, whereas the infantile nucleus is formed in late gestation/early childhood. 

Because there is no turnover of protein, the lens polypeptides are present for a lifetime. ³ It is becoming clear that changes to these old proteins, and possibly to other lens macromolecules such as membrane lipids, may be responsible in large part, if not totally, for the profound changes in physical properties that take place as the human lens ages. As one example, a massive increase in lens stiffness with age underlies the development of presbyopia. ⁴ If we are to understand the molecular basis for such conditions, the changes that occur with age to the lens must be characterized. 

In this study we used a gentle procedure, involving aspiration through a syringe needle, to disrupt the cells in two defined regions of the center of human lenses. These two regions correspond to the fetal and infantile nucleus. We used sucrose density gradient centrifugation to separate the proteins and to characterize the profiles as a function of lens age. 

Methods

Dissection of Lenses

Normal human lenses of men and women were collected from eyes donated to the Lions NSW Eye Bank at the Sydney Eye Hospital. Enucleation occurred within 12 hours of death, with lenses immediately stored at −80°C until use. All lenses used in this study were stored for no longer than 12 months. The work was approved by the human research ethics committees at the University of Wollongong and the University of Sydney (Ethics #7292). Trephines with diameters of 6 and 4.5 mm were used to dissect lenses into inner (fetal nucleus) and core (infantile nucleus) regions, respectively. The trephines were precooled at −20°C. Decapsulated lenses were placed in a precooled, custom-made nonstick holder with a diameter of 9 mm. Cortical tissue was removed with a 6-mm trephine. A cold scalpel was used to remove 1 mm from each end of the cylindrical center (combined inner core), which was inserted into a 6-mm holder and refrozen. After refreezing, a 4.5-mm trephine was used to separate the core region from the inner region. 

Sucrose Gradient Centrifugation of Lens Tissues

Sucrose concentrations of 8%, 25%, 45%, 50%, 60%, 70%, and 80% were prepared in 10 mM Tris buffer containing 2 mM EDTA and 2 mM β-mercaptoethanol at pH 8.0. The sucrose density gradient was layered as follows: 1 mL, 80%; 2 mL, 70%; 1.5 mL, 60%; 1.5 mL, 50%; 1.5 mL, 45%; 1.5 mL, 25%. The sample (core or inner region) was aspirated 30 times through a 21-gauge syringe needle in 8% sucrose (100 μL). Coomassie stain (800 μL 8% sucrose buffer plus 100 μL Coomassie stain; #23236; Pierce, Rockford, IL) was added to the sample and gently vortexed just before loading onto the sucrose gradient. An additional 1 mL of 8% sucrose solution was layered on top of the sample. All sucrose gradients were kept chilled on ice before centrifugation. 

Gradients were centrifuged at 100,000g for 2 hours at 4°C using a rotor (Ti41; Beckman Coulter, Fullerton, CA) in a centrifuge (L-80; Beckman). The resultant interfaces were labeled consecutively from most to least dense: SG1 (70/80% interface), SG2 (60%/70% interface), SG3 (50%/60% interface), SG4 (45%/50% interface), SG5 (25%/45% interface), and SG6 (8%/25% interface). 

Densitometric Analysis of Sucrose Density Gradients

Sucrose density gradients were photographed with a digital camera, and the images were normalized with an optical density step tablet (#3; catalog no. 1527654; Eastman Kodak, Rochester, NY) taken alongside the sucrose gradient. Sucrose gradient images were imported as gray-scale images into the image analysis program ImageJ software (developed by Wayne Rasband, National Institutes of Health, Bethesda, MD; available at http://rsb.info.nih.gov/ij/index.html), version 1.34s. Densitometric plots were generated for each gradient. Relative quantification of each interface was obtained by area under curve. The amount of protein at each interface was estimated based on the proportion of staining intensity at each interface. This was converted to an amount of protein, based on the total amount of protein loaded onto the gradient. Previous thermogravimetric analysis experiments have shown protein content to be 35% in the human lens inner and core regions and to be independent of age. ⁵ For the purposes of analysis, it was assumed that there was no preferential staining of crystallins and that membranes did not affect binding. 

Phospholipid Extraction of Interfaces

Interfaces designated SG6 to SG1 were collected from the sucrose gradient using a Pasteur pipette. The sucrose remaining between interfaces was also collected for phospholipid and cholesterol analysis. To ensure total recovery of phospholipid and cholesterol after all fractions were removed, each centrifuge tube was washed with 10% methanol, and this fraction was also examined for the presence of phospholipid and cholesterol. Each sample was diluted to an approximate concentration of 8% sucrose to keep the extraction conditions the same. A methanolic solution containing deuterated cholesterol (D6_CHOL) and lauroyl dihydrosphingomyelin (m/z 649.6, d18:0/12:0) was added to each sample before lipid extraction using the method of Folch ⁶ with minor modifications. In brief, chloroform/methanol (2:1 vol/vol) containing 0.01% butylated hydroxytoluene was added to the diluted interfaces, vortexed, and centrifuged for 10 minutes at 2000g. The resultant organic phase was removed, and the extraction step was repeated three times. The organic phases were combined and dried under a stream of nitrogen at 37°C and reconstituted in 1 mL of 1:2 (vol/vol) chloroform/methanol. 

Mass Spectrometry

All lipid extracts were analyzed with a mass spectrometer (QuattroMicro; Waters, Manchester, UK) equipped with a z-spray electrospray ion source. Mass spectrometer conditions and details of quantification were as described previously. ⁷ Precursor ion scans for the phosphocholine cation (m/z 184.1) were performed with a collision energy of 35 eV and a mass range of m/z 640–850. Quantification of sphingomyelins (SMs) was determined by comparison of the internal standard peak area (SM d18:0/12:0) to that of the individual SMs and dihydrosphingomyelins (DHSMs) present. All phospholipid extracts were analyzed in triplicate, with a minimum of 100 scans acquired for each precursor ion scan. To quantify the amount of phospholipid, the six most abundant SM and DHSM species in the human lens were used. These were SM (d18:0/16:0) (DHSM 16:0) m/z 705, SM (d18:1/22:0) m/z 787, SM (d18:0/24:1) m/z 815, SM (d18:0/14:0) m/z 677, SM (d18:1/16:0) m/z 703, and SM (d18:1/24:1) m/z 813. ⁷ Other minor SM and DHSM species, accounting for approximately 5% to 7% of the total, were not included. 

Cholesterol was quantified as previously described. ⁷ In brief, a mass spectrometer (QP5050; Shimadzu, Kyoto, Japan) with an electron ionization source and a direct insertion probe was used. Approximately 1 μL chloroform/methanol extract containing cholesterol was applied to the sealed end of a glass capillary tube and allowed to dry at room temperature. A temperature program heated the probe to 200°C to provide thermal desorption of the analytes. Selected ion monitoring was used to quantify cholesterol by reference to a standard curve. Cholesterol was determined by the peak area ratio of cholesterol, and the D6_CHOL internal standard was added before extraction. 

High-Performance Liquid Chromatography

SG6 and SG6a were fractionated with an HPLC system (Shimadzu) by size exclusion chromatography (SEC) on HPLC SEC columns (BioSep S-3000 and S-400; Phenomenex, Torrance, CA) linked in tandem. The buffer was 10 mM Tris buffer, pH 7.4, containing 2 mM EDTA and 50 mM NaCl at a flow rate of 0.5 mL/min. Protein samples (approximately 1 mg) were loaded, and absorbance was monitored at 280 nm. 

Results

One objective of the present study was to characterize changes as a function of age within discrete regions of the human lens. Reproducible regions were isolated by dissection with a trephine, as described previously. ⁴ Two central lens regions were examined. The first was the center, or core, of the lens, a region 4.5 mm in equatorial diameter delineated by Y sutures that has also been referred to as the fetal nucleus. ¹ The second was a region 6 mm in equatorial diameter, distinguished by the presence of a six-armed suture pattern that surrounds this core and will be referred to as the inner region. The latter region, which has also been denoted as the infantile nucleus, ¹ was examined to determine whether the same changes occurred to this region as in the core and, if so, whether the changes took place at a similar age. 

The primary technique used was a modification of that used by Chandrasekar and Cenedella, ⁸ who used sucrose density gradient centrifugation to separate the proteins. Homogenates, obtained from lens regions of different ages by aspiration through a syringe needle, were loaded onto the gradients; then the tubes were centrifuged at 100,000g. To visualize the bands, a small amount of Coomassie dye was added to the homogenate. Results for the core and inner regions are shown in Figure 1 . As can readily be seen, an obvious change occurred to the lenses between the ages of 40 and 50. For this reason it is convenient to discuss the changes in three groups: those that occur before age 40, those that occur between 40 and 50, and those that continue past age 50. 

Changes That Take Place before Age 40

Before age 40, only two significant bands, labeled SG6 and SG5 (Fig. 1) , were observed. SG6 corresponds to the soluble lens crystallins, and a size-exclusion HPLC trace of SG6 was indistinguishable from that of soluble protein obtained by conventional homogenization methods (for example, see Heys et al. ⁴ ; Fig. 2a ). SG5 was found to contain cell membranes. This was demonstrated in two ways. First, the 8 M urea-insoluble fraction obtained by conventional homogenization of young and old lenses, which has been used previously to isolate lens cell membranes, ⁹ was found to sediment at SG5. Interestingly, the membranes stripped of extrinsic proteins by urea did not differ in sedimentation behavior compared with young lens tissue run on the gradient, suggesting that the extrinsic proteins do not contribute greatly to the density. Second, lipid analysis of all the density gradient bands from young lenses revealed that most sphingomyelin and cholesterol migrated with SG5 (see 3 4 Fig. 5 ). 

Two major changes in the appearance of the density gradients were noted before age 40. SG6 became progressively more diffuse, and a higher density band appeared (SG6a; Fig. 1 ). When SG6a was examined by size-exclusion HPLC, it was apparent that it was composed of high molecular weight (HMWt) protein complexes and some lower MWt crystallins (Fig. 2b) . These might have dissociated from HMWt protein because of exposure to the higher salt HPLC buffer. The time scale for the appearance and disappearance of SG6a (Fig. 3)mirrors that found previously for HMWt protein. ⁴ The other change with age was that the density of staining of the membrane band at SG5 became more intense. Densitometry revealed a progressive increase of Coomassie staining that would be consistent with incrementally more protein binding to the cell membranes, and this process continued until the mid-40s, when density began to decline (Fig. 4b) . 

Changes That Take Place between Ages 40 and 50

The most noticeable change in Figure 1seen is the appearance of two high-density bands designated SG1 and SG2. These are essentially absent from lenses younger than 43 years. Also present in older lens samples were two less obvious bands of lower density labeled SG3 and SG4, which were almost undetectable before age 43. The appearance of these four higher density bands coincides with a marked decrease in the soluble protein (SG6; Fig. 4a ). 

Changes That Continue Past Age 50

Soluble protein content in the core did not alter appreciably beyond age 50 (Fig. 4a) ; however, other bands in the centrifugation profile did change. The age-dependent variation in the levels of SG5, SG4, SG3 SG2, and SG1 are shown in Figure 4 . 

SG3 and SG4

SG3 and SG4 behaved similarly. SG4 in both the core and inner regions was essentially absent before age 30, peaked by age 50, and then declined (Fig. 4c) . There was greater scatter for SG3. In the core it peaked between approximately age 50 and 55, but in the inner region maximum levels were found approximately 10 years later (Fig. 4d) . The appearance of the graphs of SG3 and SG4 suggests that these may be intermediate stages en route to bands at even higher densities (e.g., SG2 and SG1). 

SG2 and SG1

The most pronounced change in Figure 1was the appearance of bands at SG2 and SG1 in the older lenses. The shapes of the graphs of the densitometric analyses were similar for the core and inner regions (Figs. 4e 4f) . SG2 and SG1 were typically absent before the age of 45. Both SG1 and SG2 increased steeply after this age, and the increase in SG2 was observed in the core approximately 10 years before it was observed in the inner region. A similar time difference was noted for SG1, with this high-density band appearing at approximately age 45 in the core but not until age 55 to 60 in the inner region. 

Origin of the Higher Density Bands

Since the higher density bands in Figure 1are clearly major components of the older lens and their appearance coincided with a decrease in soluble protein (SG6) and in membranes (SG5), we hypothesized that SG1–SG4 may arise from the binding of formerly soluble proteins to fiber cell membranes. To investigate this, the bands in the sucrose gradients were examined for the presence of characteristic membrane components. 

Lipid Analyses

We attempted to correlate change in protein density with that of membranes. Shotgun lipidomics ¹⁰ was used to identify and obtain relative quantification of phospholipids present in the “native” membrane isolated at SG5 from the nucleus. Previously, it had been established that phospholipid concentrations of human lenses can be determined by the use of electrospray-ionization tandem mass spectrometry of crude human lens lipid extracts. Sphingomyelin and dihydrosphingomyelin were found to be the major phospholipids in the human lens, contributing 66% of total phospholipids in a 60-year-old human lens ⁷ in agreement with previous data. ¹¹ These abundant molecules were used as markers for changes in total phospholipid content in each interface and by inference membranes. 

Analyses of the centrifugation profiles for SM and cholesterol showed clear changes in distribution as a function of lens age (Fig. 5) . In younger lenses, these membrane components were localized at SG5; some cholesterol was found to be associated with SG6. In lenses older than 45, the membrane markers were consistently detected in SG4 and SG3 as well as in SG1 and SG2. The appearance of the mass spectra of the phospholipids from the “native membranes” (SG5; Fig. 5b ) and those from the higher density bands were essentially identical (data not shown). 

Sucrose gradients across the age range were examined for the content of SMs, and the results are depicted in Figure 6 . SG4 and SG3 displayed similar profiles for SM with peaks at age 45 (SG4) and 50 (SG3), after which the relative amounts decreased almost to zero in the oldest lenses examined. This pattern is similar to that documented earlier for protein content in these two bands (see Fig. 4 ). It was of interest that SM content at SG4 and SG3 peaked earlier than at either SG1 or SG2. 

The appearance of the profiles for SG1 or SG2 were also similar and showed marked increases with age (Figs. 6c 6d) . In particular, SG1 rose steeply after age 40 and reached a plateau after age 60. 

By comparison, preliminary experiments using cortical tissue from the same lenses showed distinctly different protein sedimentation patterns from the inner and core regions. Although some protein was seen at SG4, SG3, and SG2 in older lenses, little or no SG1 was present in the cortex at any age. 

To determine whether particular crystallins were involved in the increase in protein density at SG1, the iTRAQ (Invitrogen, Carlsbad, CA) procedure ¹² was used on samples isolated from the lens core. The following crystallin polypeptides increased in 71- and 72-year-old lenses compared with levels detected in 49- and 50-year-old lenses: αA, 12%; αB, 50%; βA3, 64%; βA4, 23%; βB1, 67%; βB2, 16%; γS, 68%; γB, 61%; γC, 66%; and γD, 61%. It is clear that all the crystallins increased significantly in SG1 from the two older lenses when compared with SG1 from two younger lenses, with the only exception αA crystallin, which was almost unchanged. To investigate this process in detail, all SG regions will have to be studied as a function of age. Details of the specific crystallins and other proteins involved will be reported separately. 

Discussion

This study has documented remarkable age-related changes to the proteins and membranes in the center of human lenses. These were particularly pronounced in the decade between age 40 and 50, though significant changes were detected before the fifth decade and in later years. 

The results are consistent with large-scale binding of crystallins to fiber cell membranes at middle age. Previous work by Spector et al. ^{13 14} suggested an association of proteins with membranes in cataract lenses. In our study, the same changes seemed to occur to the inner and core regions but were delayed by approximately one decade in the inner region. 

In the years leading up to the fifth decade of life, the major changes detected were the conversion of crystallins to HMWt proteins, which we found to sediment at a slightly higher density (i.e., SG6a), and an apparent increase in the amount of protein binding to the fiber cell membranes at SG5. The formation of HMWt protein is thought to be largely a reflection of the chaperone action of α-crystallin in binding denaturing proteins in the lens. ^{15 16 17} The apparent increase in binding of proteins to the cell membranes at SG5 we observed might reflect an alternative fate for the denatured proteins, which could interact with lipids, extrinsic proteins, and intrinsic proteins of the membranes, possibly by hydrophobic bonds. Alternatively the data may be consistent with HMWt protein binding to lens membranes, as has been documented by Cobb and Petrash ¹⁸ and others. ¹⁹ Binding to membranes at SG5 was documented by densitometric scanning for protein (Fig. 4b) , but the increase was not apparently sufficient to cause a significant increase in density in our gradient system. 

In the fourth decade, as the staining intensity of SG5 continued to increase, higher density bands (e.g., SG4; Fig. 4c ) began to appear. Protein staining and lipid analyses suggested that SG3 and SG4 may be transient intermediates in which the extent of binding of proteins to the fiber cell membranes increase such that the density of the complex causes it to sediment lower in the sucrose gradient. SG3 and SG4 peak at about age 50 and then decline. When SG3 and SG4 are reaching their maxima, the higher density bands at SG1 and SG2 appear. 

After age 50, the visual appearance of the lens density gradients does not alter greatly, but changes to individual bands can still be measured. For example, the intensity of the SG1 band continues to increase with age, perhaps because of ongoing conversion of SG3 and SG4 to this higher density component. 

The most detailed investigations in this study were undertaken on the lens core, which is equivalent to the fetal nucleus. ¹ However, the same changes took place, albeit at a later time, to the infantile nucleus (inner region) that is formed later in life. This may reflect the time that the membranes and proteins have been present in the lens. It is unlikely that variation in the composition of crystallins in the two regions is responsible for the delay because two-dimensional gels of these are almost identical. ¹  

The correlation of the time scales for the loss of total soluble protein and the decrease in “native membrane” as a function of age of the lens core is summarized in Figure 7a . It is clear from a comparison of Fig. 7awith Fig. 7bthat the higher density bands SG1 and SG2 only become prominent once the decreases in SG5 and SG6/6a are essentially complete. 

A probable explanation for these large changes in the center of the lens can be constructed around the known depletion of the major lens crystallin, alpha crystallin. A finite amount of this chaperone protein is present in the nuclear region at birth. As crystallins and other lens proteins denature over a period of decades, alpha crystallin binds to them, forming HMWt complexes. By age 40, all the alpha crystallin has been consumed ^{4 20 21} ; however, the major drivers of protein denaturation (heat, ^{4 22} deamidation, ^{23 24} truncation, ²⁵ racemization ^{26 27} ) continue. Indeed, after age 50, the levels of some posttranslational modifications in the lens center increase dramatically. ^{28 29}  

Our data suggest that when HMWt protein levels increase, there is also a significant increase in the binding of proteins to the cell membranes (i.e., SG5). HMWt may be implicated in this binding. ^{18 19} After age 40, when no chaperone is available in the nucleus, the degree of binding of proteins to membranes is much greater, leading to the formation of higher density bands. Initially SG3 and SG4 form, but as the process of protein denaturation and aggregation continues in the lens center in the absence of alpha crystallin, the higher density bands at SG1 and SG2 form. The results of iTRAQ analysis of SG1 indicate an increase in the abundance of all the crystallin classes in the older lenses with αA and βB2 changing the least. 

What are the likely consequences of the age-dependent increased interaction of presumably denatured proteins with fiber cell membranes, which becomes increasingly apparent in the fifth decade of life? Changes in the physical properties of the lens could be expected, and this is indeed what is observed experimentally. For example, stiffness of the center of the lens increases dramatically with age, particularly so in the center after age 40 to 50. ^{4 30} This may be the main reason for presbyopia, though experiments on extracted lipids indicate changes in membrane fluidity with age, ³¹ suggesting that additional lens components may play a role. Interestingly, previous studies noted marked biochemical changes in the human lens between ages 40 and 60. For example, free calcium increases significantly after age 50, ^{32 33} which can lead to changes in protein structure. ³⁴ Of particular interest to this study, calcium can also induce the formation of lenticular HMWt protein aggregates ³⁵ and also bind sphingolipids. ³⁶ During this period, too, significant changes have been observed in protein solubility in the lens nucleus. ⁴  

It is likely that if large-scale binding of proteins to the cell membranes occurs in the fifth decade, it may disrupt the proper functioning of membrane pores by occluding the openings of the channels on the cytoplasmic side of the cell. In separate studies, a pronounced decrease in the diffusion of small molecules within the lens was detected in older lenses. A barrier to the diffusion of water ³⁷ and glutathione ³⁸ from lens cortex to nucleus develops at middle age. These two small molecules are thought to use different membrane pores; in the case of water, aquaporin 0 is used, whereas glutathione is thought to move from cell to cell by way of connexons. Therefore, it was difficult at first to envisage how both these membrane pores could be affected at the same age. If the occlusion of both types of pores is the result of large-scale binding of crystallin aggregates to the fiber cell membranes, then this apparent anomaly is resolved. Gradient centrifugation studies on the barrier region of the lens are under way. 

Although it seems likely that the age-related changes we observed in protein sedimentation were caused by membrane binding, we cannot yet eliminate the possibility that proteins bind to cell membranes during the extraction process or that protein aggregation leads to “clumping” of membrane fragments. Experiments are under way to examine these possibilities. In data yet to be published, significant age-related changes in membrane lipids have been observed in these lens regions; hence, this factor may also contribute to the changes we observed. The results of this study extend those obtained using whole lenses by Chandraseker and Cenedella ⁸ and suggest that much of the so-called insoluble protein detected in older human lenses may arise from the interaction of protein aggregates with fiber cell membranes. It is clearly important to characterize the mechanism of interaction of these protein complexes with cell membranes and to unravel the consequences of this age-dependent binding for the functioning of the lens. 

 Supported by National Institutes of Health Grant EY013570 and National Health and Medical Research Council (NHMRC) Grant 512334. RJWT is an NHMRC Senior Research Fellow.

 Submitted for publication February 18, 2009; revised April 16, 2009; accepted August 4, 2009.

 Disclosure: M.G. Friedrich, None; R.J.W. Truscott, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Comparison of sucrose density gradients patterns as a function of age in the lens (a) core and (b) inner region. Protein was visualized with Coomassie blue.

Figure 2.

 

SEC HPLC analysis of sucrose gradient fractions (a) SG6 and (b) SG6a from a 46-year-old lens core.

Figure 3.

 

The content of protein in SG6a as determined by densitometry from the core (•) and inner (○) regions as a function of lens age; 24 cores and 15 inner regions were examined.

Figure 4.

 

Change in protein density as a function of age. Amount of protein in (a) SG6, (b) SG5, (c) SG4, (d) SG3, (e) SG2, and (f) SG1 as determined by densitometry from the core (•) and inner (○) regions as a function of lens age; 24 cores and 15 inner regions were examined.

Figure 5.

 

(a) Comparison of the protein pattern (as judged by photographs of the centrifuge tubes) with lipid content in the sucrose gradients as a function of age in the core region. Membrane sedimentation was assessed by detection of sphingolipids (right) and cholesterol (left). Sphingolipid data are shown as a percentage of total SM recovered from SG1 to SG5. Cholesterol data are a percentage of total recovered from SG1 to SG6. (b) Mass spectrum of phospholipids in SG5 from a 22-year-old lens. The positive ion tandem mass spectra of precursors of the phosphocholine cation of m/z 184 of a 22-year-old SG5. SM (d18:0/16:0) m/z 705, SM (d18:1/22:0) m/z 787, SM (d18:0/24:1) m/z 815, SM (d18:0/14:0) m/z 677, SM (d18:1/16:0) m/z 703, SM (d18:1/24:1) m/z 813 used for determination of phospholipid content. Note the internal standard at SM (d18:0/12:0) m/z 649.6.

Figure 6.

 

The content of sphingolipids detected at (a) SG4, (b) SG3, (c) SG2, and (d) SG1 as a function of age in the core region.

Figure 7.

 

(a) Amount of sphingolipid in SG5 (•) and water-soluble protein (SG6 and SG6a) (□) in the core of the human lens as a function of age. (b) Amount of sphingolipid (•) and protein (□) in SG1 and SG2 combined.

The authors thank Jane Deeley for help with lipid analysis and Raj Devasahayam of the Sydney Lions Eye Bank for providing the human lenses.