MALDI Tissue Imaging of Ocular Lens α-Crystallin

Abstract

purpose. To apply MALDI (matrix-assisted laser desorption ionization) tissue imaging methods to obtaining a profile of the distribution of the lens α-crystallins and their modified forms in calf and mature bovine lenses.

methods. Frozen bovine lenses were cut equatorially at −12 °C to −20°C into 10- to 40-μm sections depending on lens age. Tissue sections were mounted onto MALDI sample plates by ethanol soft-landing to maintain tissue integrity. A two-layered matrix deposition method was used to improve mass spectral reproducibility across sections. Molecular images of the two subunits of α-crystallin and their modifications over approximately one-half of a single tissue section were reconstituted from mass spectral data sets acquired in 250-μm steps. Identification of protein truncation products and confirmation of phosphorylation distribution patterns were performed by reverse-phase liquid chromatography of soluble extracts from specific tissue regions followed by tandem mass spectrometry (LC/MS/MS).

results. Distinct distribution patterns were observed for the two subunits of α-crystallin and their modified forms. αA-crystallin showed extensive truncation across whole sections, especially in the nuclei, whereas αB-crystallin was observed to be relatively stable. Both αA-crystallin and αB-crystallin displayed the highest level of phosphorylation in the middle cortex region, a finding confirmed by LC/MS/MS analysis of dissected regions.

conclusions. A new imaging technique has been successfully applied to molecularly characterize the spatial distribution of lens proteins and their modifications in lens sections. The different distributions of α-crystallin revealed in this study provide new leads in the investigation of underlying physiological significance of the modified forms of the two α-crystallin subunits.

The mammalian lens is primarily composed of lens fiber cells which differentiate from a single layer of lenticular epithelial cells. As the lens grows, the newly differentiated fiber cells in the periphery elongate and compress the older cells toward the center of the lens, to form concentric fiber cell layers, the inner part of which is the nucleus and the outer part of which is the cortex. The nucleus and the cortex represent the old and the newly derived fiber cells with different cell ages. One of the unique properties of lens fibers is that the nuclei and other organelles of fiber cells are lost during cell differentiation providing a highly organized transparent medium for light transmission. ¹ An additional consequence of this process is the loss of protein turnover in the lens. In addition, the mammalian lens has an exceptionally high protein content making up approximately 30% to 35% of its dry mass. ² Among the various resident proteins expressed in the lens are the three major lens structural proteins, α-, β- and γ-crystallins, which constitute up to 90% of total protein in a lens. ³ Each crystallin exists as several homologous subunits. Because of the lack of protein turnover, lens proteins undergo various age-related posttranslational modifications that include truncation, deamidation, glycation, and phosphorylation. The modifications accumulate over the lifetime of mammalian species and possibly contribute to lens opacification and cataractogenesis by inducing water insolubilization and aggregation of lens proteins in aged and cataractous tissue. ⁴ The mechanism of how the lens protein modifications affect the protein function during lens development and lead to cataractogenesis is not completely understood. In this study, a new mass spectrometric method for tissue imaging was used to characterize the spatial distribution of lens proteins and their post-translational modifications with cell age in bovine lenses. 

Traditionally, immunohistochemistry has been a used technique commonly used for profiling protein distributions in a tissue. In this approach antibodies are needed to detect specific proteins and usually only one protein can be detected in a single experiment. Imaging mass spectrometry (IMS) of a tissue section is a new application of matrix assisted laser desorption/ionization mass spectrometry (MALDI-MS) for global proteome analysis. This technique was first introduced in 1997 by Caprioli et al. ⁵ as a new technology for biological investigation. IMS enables a detailed assessment of complex protein distribution patterns within a tissue by simultaneously generating two-dimensional molecular maps of hundreds of peptides and proteins directly from the surface of a tissue section. Moreover, the molecular maps display the relative abundance and spatial distribution of these molecules with high resolution. One important aspect of the application is the comparison of histologic features obtained from normal and diseased tissue sections with molecular images. To date, IMS has been used for molecular profiling of multiple tissue samples, such as rat pancreas, ⁵ rat pituitary, ⁵ mouse brain, ⁶ human brain tumor xenografts, ⁶ and mouse prostate, ⁷ among others. IMS of tissue sections is clearly an emerging technique of great potential for biological and medical investigation. 

In this report, we developed and applied MALDI tissue imaging methodology to measure the spatial distribution of bovine lens α-crystallin. Multiple forms, including truncated and phosphorylated forms, were imaged in a single experiment. This method holds great promise for determining protein changes in developing, aged, and diseased lenses. 

Materials and Methods

Acetonitrile, ethanol, high-performance liquid chromatography (HPLC)–grade water, formic acid, and sinapinic acid (SA) were purchased from Sigma-Aldrich (St. Louis, MO). Bovine calf eyes were freshly collected in a local abattoir. The lenses were removed from the anterior region, snap frozen with liquid nitrogen, and stored at −80°C without the use of optimal cutting temperature (OCT) embedding medium. Mature bovine lenses were obtained from Pel-Freez Biologicals (Rogers, AR). All the lenses were stored at −80°C. Typically, a calf lens used in this study is 10 to 32 weeks old, whereas a mature lens is greater than 52 weeks old. 

Tissue Sectioning and Mounting

Frozen bovine lenses were decapsulated quickly on ice and then placed at −20°C for at least 2 hours. At the time of tissue sectioning, the lenses were mounted on the cutting stage of a microtome in a cryostat (HE 505; Microm, Walldorf, Germany) with the aid of OCT embedding medium and cut in parallel with the equatorial plane. The tissue was typically sliced into 10- to 20-μm sections at −20°C for calf lenses and 30 to 40 μm at −12°C for mature lenses. The sections close to or at the equatorial plane were collected for mass spectrometric analysis. Before tissue mounting, MALDI plates were coated with a thin, uniform layer of ethanol by dripping two droplets of anhydrous ethanol on the plates and swirling the plates. A tissue section was then swiftly transferred to the plate using forceps. The plate was warmed to room temperature and allowed to dry, typically taking 10 to 20 seconds. 

Matrix Deposition

The tissue sections on MALDI plates were first sprayed with several cycles of an acetonitrile-water (50:50, vol/vol) solution, with an air-brush sprayer (model 200; Badger Air-brush, Franklin Park, IL) resulting in a tightly bound section. After drying, the surface of tissue sections was preseeded with a thin layer of matrix by spraying several cycles of 15 mg/mL SA solution freshly prepared in ethanol-water (50:50, vol/vol). After drying, the sections were evenly sprayed with several cycles of 15 mg / mL SA matrix solution prepared in ethanol-water-formic acid (44:44:12, vol/vol/vol) solution until good test MALDI mass spectra were achieved across the whole section. 

MALDI Imaging

Mass spectrometric analyses were performed in the linear, positive mode at +25 kV accelerating potential on a time-of-flight mass spectrometer (Voyager DE-STR; Applied Biosystems, Inc. [ABI], Foster City, CA), which was equipped with a 337-nm N₂ laser capable of operating at a repetition rate of 20 Hz with optimized delayed extraction time. The instrument was used in its standard configuration, and the laser beam size was estimated to be approximately 100 μm in diameter. Mass spectral data sets were acquired over approximately one half of a single tissue section using the data acquisition software MMSIT (http://maldi-msi.org/software.htm/; Novartis Pharma AG, Basel, Switzerland) in the mass range of m/z 3,000 to 30,000 with a raster step size of 250 μm and 50 to 60 shots per spectrum. After data acquisition, protein images were reconstituted by importing the acquired image files into the image analysis software BioMap (http://maldi-msi.org/software.htm/; Novartis Pharma AG). Based on the acquisition parameters and imaged tissue size, each image is composed of 3000 to 4000 pixels. For comparison, all the images from a single section were plotted using the same reference color settings. 

Protein Identification by LC/MS

Lens tissues close to the equatorial plane from the imaged bovine lenses were collected and separated into equator, middle cortex, and nucleus regions, and the sample regions were pooled. The tissues were homogenized in 6 M urea, 10 mM Tris, 10 mM NaF, and 10 mM dithiothreitol (DTT) buffer and then acidified into 0.1% trifluoroacetic acid (TFA). The homogenates were loaded onto Sep-Pak C₁₈ cartridges, washed with 0.1% TFA and eluted with 95% acetonitrile containing 0.1% TFA. The eluted proteins were dried (SpeedVac; ThermoSavant, Holbrook, NY) and reconstituted in 10% acetonitrile, 0.1% TFA. Protein amounts equivalent to 500 μg of wet lens tissue were injected onto a C₄ column (2.1 mm inner diameter × 10 cm long) which was coupled to a linear ion trap mass spectrometer (LTQ; Thermo Finnigan, Mountain View, CA) and separated with mobile phase A (0.1% TFA) and B (acetonitrile, 0.1% TFA) at a flow rate of 200 μL/min using a gradient of 15% to 30% B over 30 minutes, 30% to 45% B over 90 minutes and 45% to 100% over 30 minutes followed by a re-equilibration step. The eluant was split, and only 10% was directed into the mass spectrometer. Mass spectra were acquired in the range of m/z 600 to 1600 with the dynamic exclusion function enabled. The two most intense peaks within m/z 750 to 1200 in each full MS scan were selected for MS/MS sequence analysis. Protein identification was accomplished using the MS-Seq program of ProteinProspector (http://prospector.ucsf.edu/ provided in the public domain by the University of California, San Francisco) using input parameters: (1) the protein mass determined from deconvoluted electrospray mass spectra and (2) the m/z and charge state values of consecutive sequence ions of the protein determined by manual interpretation. 

Results

MALDI imaging experimental workflows include several important aspects that affect mass spectrometric signal quality. Appropriate sample preparation and treatment, including tissue collection, freezing, sectioning, mounting, and matrix selection/deposition, is the prerequisite for successful generation of reproducible and high-quality mass spectra for MALDI imaging. Detailed practical aspects of IMS strategy and methodology have been reviewed in several publications. ^{8 9 10 11 12 13} With the modified sample preparation method implemented in the present study, distinct distribution patterns were observed with lens fiber cell age in the different tissue regions for the two subunits of α-crystallin and their modified forms. 

Tissue Sectioning and Mounting

In this experiment, bovine lenses were cut equatorially and imaged to compare the distribution patterns of lens proteins in a single tissue section. Unlike other tissues, such a tissue section is composed of lens fibers with remarkably different cell ages. It has been shown that it is difficult to maintain lens tissue section integrity during sectioning and transferring from a microtome blade to a MALDI plate ¹⁴ using the commonly used tissue sectioning and mounting protocol. ¹¹ This is exemplified by the fact that under normal sectioning conditions, the sectioned tissue slices frequently fragment in the nucleus region on cutting or mounting on the MALDI plates. To keep tissue sections as intact as possible, we used a modified tissue sectioning and mounting protocol in this study. To prevent the tissue sections from breaking during cutting, different temperatures and cutting thicknesses were applied to lenses of different ages and were found to help maintain tissue sections intact. As described earlier, calf lenses were optimally cut into 10- to 20-μm sections at approximately −20°C, while the older lenses were typically cut into 30- to 40-μm sections at approximately −12°C. The optimal tissue section temperature and thickness depended on the actual age of an individual lens. By this approach, most of the tissue sections were kept intact during cutting except for the oldest lenses. To prevent the tissue sections from cracking or tearing when placed on the MALDI plates, a soft-landing step was used to assist in tissue mounting by forming a thin layer of ethanol on the plate surface. The thin layer of ethanol acted as a soft support for the thin tissue sections, preventing any observable cracking or tearing. It should be noted that the thickness of ethanol layer should be strictly controlled, because a thick ethanol layer resulted in the tissue section’s floating on the plate and taking a longer time to dry. 

Matrix Application and Tissue Imaging

Soft-landing successfully kept the tissue sections intact during transfer to MALDI plates, but most tissue sections were only loosely attached to the plates after soft-landing, creating the possibility of being jettisoned from the plate on the application of matrix solution. Therefore, a fixation step was used to make the tissue sections bind to the plates tightly. Spraying the sections with two or three cycles of acetonitrile-water (50:50) was found to bind the tissue to either gold or stainless steel MALDI plates. Next, to generate uniform and reproducible spectra for MALDI imaging, we used a two-layered method to apply matrix onto the surface of a tissue section. First, a thin layer of SA was seeded uniformly on the tissue surface by spraying several cycles of the matrix solution prepared in acid-free ethanol-water (50:50) solution. Each spray cycle barely wetted the tissue surface. After this step, no peptide or protein signals were observed in MALDI mass spectra. Finally, the tissue sections were sprayed with matrix solution containing an optimized concentration of formic acid. Formic acid helped to dissolve the proteins on the tissue surface and to incorporate them into the matrix crystals. Preliminary experiments showed that a matrix solution containing 10% to 15% formic acid favored the generation of abundant protein ion signals from the two subunits of α-crystallin and their degradation products. In this study, 12% formic acid was used to image the two subunits of α-crystallin and their degradation products selectively. Matrix solution prepared in solvents containing 0.3% to 0.5% TFA did not generate reasonable quality spectra over the entire lens tissue sections, although this formulation has been successfully used in other tissues. ¹² This is probably due to the effect of different lens cell ages on the protein solubility in different regions in a section. The two-layered matrix application step was observed to improve the reproducibility of mass spectra by reducing ion intensity variability across the whole tissue section. 

Imaging of α-Crystallin and Its Degradation

The protein content in mammalian lenses is dominated by a relatively small number of abundant lens crystallins. Among them, α-, β- and γ-crystallins constitute up to 90% of the total protein weight in a lens. ³ Under the present experimental conditions, the MALDI mass spectra of lens tissue samples are overwhelmed by ion signals from the two subunits of α-crystallin, whereas the other crystallins showed weak signals, thus making it possible to image the two α-crystallin subunits and their modified forms. Figures 1A and 1Bshow the test mass spectra acquired from the equatorial regions of a calf lens in a 16-μm-thick section and a mature lens in a 35-μm-thick section, respectively. In both calf and mature lens equatorial tissue sections, only a few ion signals corresponding to singly and doubly charged intact αA- crystallin (MW 19,832 Da, [M+H]⁺ = m/z 19,833, [M+2H]²⁺ = m/z 9917) and αB-crystallin (MW 20,079 Da, [M+H]⁺ = m/z 20,080, [M+2H]²⁺ = m/z 10,040.5) were observed in the spectra indicating little protein degradation in the very young lens fiber cells. Figure 1C and 1Dshow the test spectra from the nuclear regions of the above calf and mature lens tissue, respectively. There are significantly more ion signals displayed in the spectra from the nuclear regions than from the equatorial regions, and most of these signals were identified and found to originate from the two subunits of α-crystallin (discussed later). By acquiring MALDI mass spectra across the tissue in 250-μm steps, we were able to plot the spatial distribution maps of all the observed peptides and proteins in the test spectra. Figure 2shows images of a tissue section from a calf lens and the representative images of intact αA-crystallin (MW 19,832) and its major degradation products. The image of intact αA-crystallin shows the distribution patterns of this protein and clearly indicates that this protein undergoes much more degradation in the center of the lens than in the periphery, whereas most of the degradation products have complementary distribution patterns to the parent molecule. The ion at m/z 11,993, which was identified by MS/MS to be an truncated fragment, residues 1-101 at N as shown in Figure 3 , displayed the widest distribution and the most abundant signal among all the observed truncated fragments in the calf lenses, indicating that N-101 is one of the major truncation sites in the protein in the young lens of bovine species. This major truncation product was previously observed as such using standard biochemical methods. ¹⁵ Other major degradation products of αA-crystallin imaged in Figure 2include truncated backbone fragments: residues 1-171 (MW 19,658) truncated at the C terminus of P, 1-168 at S (MW 19,403), 1-151 at D (MW 17,571), 1-101 at N (MW 11,992) 1-80 at F (MW 9,588), 1-65 at R (MW 7,740), 1-58 at D (MW 7,011), 1-54 (MW 6,583) at R, and 1-50 (MW 6,079) at Q. Figure 3shows the electrospray MS/MS spectrum of the [M+12H]¹²⁺ ion at m/z 1000.4, corresponding to the [M+H]⁺ ion at m/z 11,993 in the MALDI mass spectrum, which was fragmented in a linear ion trap instrument. All significant fragments could be assigned to the sequence of αA-crystallin 1-101 giving unambiguous confirmation of this structure and the truncation site. 

Figure 4shows the MALDI images of intact αB-crystallin (MW 20,079) and some of its identified major degradation products: residues 1 to 170 (MW 19,583) truncated at the C terminus of T, 1 to 73 (MW 8,675) at N and 1 to 41 (MW 4,982) at D. Figure 5shows the electrospray tandem mass spectrum of the [M+19H]¹⁹⁺ ion at m/z 1031.7 supporting its assignment as αB-crystallin 1-170 (MW 19,583) and confirming truncation at T-170. Compared with the distribution pattern of αA-crystallin, intact αB-crystallin displayed a distinctly different distribution pattern. It produced abundant signals in both equator and nucleus regions, except in the innermost part of the lens with a slightly less intense distribution in the middle cortex region, suggesting αB-crystallin is more stable in the lens than is αA-crystallin. The observed weaker signals in the innermost part of the lens section were likely due to protein degradation in this region, because the identified degradation products of αB-crystallin were more prevalent in the inner nucleus than in other regions. The weaker signals of intact αB-crystallin in the middle cortex can be explained by the distribution of its phosphorylation forms. 

Imaging of Phosphorylated α-Crystallin

Reversible phosphorylation is regarded as one of the most important posttranslational modifications of mammalian proteins for regulating their cellular and physiological functions. The two subunits of α-crystallin have been reported to be phosphorylated in vivo at S-122 on bovine αA-crystallin, ¹⁶ and at three sites on bovine αB-crystallin: S-19, S-45, and S-59. Two additional phosphorylation sites on αB-crystallin have been also reported. ^{17 18} Phosphorylation of α-crystallin has been reported to reduce molecular chaperone activity ¹⁹ and to decrease the oligomeric size of α-crystallin. ²⁰ In the MALDI tissue-imaging experiment, two ions observed at m/z 19,913 and m/z 20,160 are shifted by 80 mass units from the predicted masses of αA- and αB-crystallin and are assigned to single phosphorylation on the two subunits, respectively. The ion observed at m/z 20,240 (see Fig. 7 ), 160 mass units above the unmodified αB-crystallin signal at m/z 20,080 is assigned to double phosphorylation of αB-crystallin. The images of these phosphorylated proteins together with their corresponding nonphosphorylated forms are shown in Figure 6 . Comparing the images shown in Figures 6A and 6B , phosphorylated αA-crystallin has a similar distribution pattern to its nonphosphorylated form across imaged areas except for a slightly weaker distribution in the very outer equator region. Figures 6Cto 6Eshow the MALDI images of non-, singly, and doubly phosphorylated αB-crystallin. In contrast to the distribution pattern of phosphorylated αA-crystallin, phosphorylated αB-crystallin shows a different pattern from the nonphosphorylated form. The αB-crystallin shows a relatively low phosphorylation level in both nucleus and outer cortex regions, while in the middle cortex, αB-crystallin undergoes more single and double phosphorylation. This finding is evident when comparing extracted MALDI spectra from specific regions as shown in Figure 7A . To confirm whether the protein phosphorylation patterns of α-crystallin from the MALDI images reflects the actual difference of protein phosphorylation level in the tissue, a complementary LC/MS method was used to characterize the protein phosphorylation present in soluble extracts from dissected equatorial cortex, middle cortex, and nucleus regions. Figures 7B and 7Cshow the deconvoluted electrospray ionization (ESI)/MS masses of HPLC-separated αA-crystallin and αB-crystallin, respectively. The deconvoluted masses of the nonphosphorylated and phosphorylated forms reflect the relative phosphorylation level determined by LC/MS. The results are consistent with the distribution patterns displayed by MALDI imaging again demonstrating increased phosphorylation in the middle cortex region. 

MALDI tissue images were acquired on mature lens tissue sections in this experiment (data not shown). The two subunits of α-crystallin and their modified forms showed very similar distribution patterns to those from calf lenses except for the increased protein degradation observed in mature lenses than in calf lenses as shown in the MALDI mass spectra of Figure 1 . 

Discussion

α-Crystallin is one of the most abundant structural proteins in the lens as well as a major member of the small heat shock proteins in mammalian lens. It exists as large heterogeneous aggregates (approximately 800 kDa), which are composed of the two crystallin subunits, αA and αB, both of which play an essential role in maintaining the transparency of mammalian lens. In addition, α-crystallin was found to act as a molecular chaperone preventing nonspecific aggregation of denatured proteins. The chaperone activity can be altered by truncation ^{21 22 23} and phosphorylation. ^{19 20 24 25}  

The feasibility of a new MALDI tissue imaging technique applied to lens α-crystallin and its modifications is demonstrated in this study using refined tissue cutting and mounting and matrix deposition protocols. Indeed, any signals observed in the MALDI spectra can be spatially mapped to the tissue sections. Comparison of the MALDI images generated over the same tissue section from a single lens indicated that the lens proteins degrade with cell age, a well-characterized phenomenon. αA-Crystallin undergoes extensive backbone truncations in both calf and mature lenses and the degradation dramatically increases from the periphery to the center of the lenses. The images of αB-crystallin showed that this subunit is relatively stable in the lens, although its degradation products have distribution patterns similar to those from αA-crystallin (i.e., more degradation occurred in the central part of the lens). One of the major degradation products of αB-crystallin with the five C-terminal amino acid residues deleted has been reported to decrease its chaperone activity. ²² The decreased chaperone activity in truncated αB-crystallin may cause the insolubilization of many proteins in the lens, which is likely to lead to the progression of cataract formation. 

One of the interesting findings in this study is that the two subunits of α-crystallin have distinct distribution patterns with respect to their phosphorylation states. Phosphorylated αA-crystallin showed a distribution pattern similar to its nonphosphorylated forms except in the equator region, whereas αB-crystallin showed more phosphorylation in the middle cortex region than other regions. The mechanism underlying the difference of distribution patterns of the two subunits together with their phosphorylation needs further investigation. 

In situ approaches to MALDI imaging of lens proteins offers several advantages over the traditional MALDI-MS approaches: No protein extraction and separation steps, minimal sample preparation, and visible spatial distribution of lens proteins and their modifications over the tissue section are observed directly from the surface of tissue sections. Compared to traditional immunohistochemistry, no antibodies against the specific protein are needed and many protein products can be imaged in one experiment. Because MALDI imaging of proteins is based on their respective masses, the different truncated forms of a single protein could be imaged simultaneously. Although the sample preparation protocol used in the present study enhances the MALDI signals for α-crystallins, preliminary experiments indicate that altering solvent conditions allows other lens crystallins to be observed and imaged. The nature of this selectivity will be the subject of future investigation. As demonstrated in the present study, the MALDI tissue imaging approach provides a powerful means for better understanding of the processing of lens proteins during differentiation and aging. 

Conclusions

MALDI imaging provides for the first time a new tool to discern the spatial distribution of the two α-crystallin subunits and their multiple modifications in a single experiment, indicating MALDI tissue imaging can be applied to study lens development and aging / cataractogenesis. Refined sample preparation procedures are needed to image additional proteins in lens, such as membrane proteins. It is expected that the application of MALDI tissue imaging could be extended to the analysis of human lenses. 

 Supported by National Eye Institute Grants EY13462 (KLS) and Vision Core Grant R24 EY14793.

 Submitted for publication December 1, 2005; revised February 1, 2006; accepted April 26, 2006.

 Disclosure: J. Han, None; K.L. Schey, 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.

 

MALDI-MS profiles obtained from the equator (A) and nucleus (C) of a calf lens in 16-μm-thick section cut equatorially at −20°C and from the equator (B) and nucleus (D) of a mature lens in 35-μm-thick section cut at −12°C. Matrix: 15 mg/mL SA, prepared in ethanol-water-formic acid (44:44:12, vol/vol/vol).

Figure 2.

 

(A) Photograph of a 16-μm-thick calf lens section mounted on a MALDI plate, taken after two-layered matrix deposition. (B) The MALDI image of intact αA-crystallin. (C–K) are the images of identified degradation products of αA-crystallin. The sequences (1–x) and measured molecular weights of the proteins imaged are indicated below each image. (L) The color reference table used to reconstitute these images.

Figure 3.

 

Electrospray tandem mass spectrum of the [M+12H]⁺¹² ion at m/z 1,000.4, corresponding to the [M+H]⁺ ion observed at m/z 11,993 in the MALDI spectrum. This truncated αA-crystallin fragment, 1-101, shows the most extensive distribution over the lens equatorial sections.

Figure 4.

 

MALDI-MS images of (A) intact αB-crystallin and its major degradation products in a calf lens: (B) residues 1-168 (MW 19,583), (C) residues 1-73 (MW 8,675), (D) residues 1-41 (MW 4,982).

Figure 5.

 

Electrospray tandem mass spectrum of the [M+9H]⁺⁹ ion at m/z 1,031.7, corresponding to the [M+H]⁺ ion observed at m/z 19,584 in the MALDI spectrum. This fragment was identified to be a major degradation fragment (residues 1-170) of αB-crystallin with five residues lost.

Figure 6.

 

MALDI-MS images of the phosphorylated forms of αA- and αB-crystallin in a calf lens. (A) Nonphosphorylated and (B) phosphorylated αA-crystallin. (C) Nonphosphorylated, (D) singly phosphorylated and (E) doubly phosphorylated αB-crystallin.

Figure 7.

 

Comparison of phosphorylation levels of αA- and αB-crystallin by MALDI-MS and LC/MS. (A) MALDI mass spectra extracted from tissue images from the respective equator, middle cortex, and nucleus regions. (B) The deconvoluted electrospray ionization masses of αA-crystallin and its phosphorylation form and (C) the deconvoluted masses of αB-crystallin and its phosphorylated forms from LC/MS analysis of soluble extract of dissected lens regions.

The authors thank the Medical University of South Carolina Biological Mass Spectrometry Facility for use of their equipment.