Molecular Vision 2004; 10:240-247 <http://www.molvis.org/molvis/v10/a30/> Received 12 December 2003 | Accepted 22 March 2004 | Published 29 March 2004

Download Reprint

Technical Brief

Preparation of planar retinal specimens: verification by histology, mRNA profiling, and proteome analysis

Gareth J. McKay,¹ Lucinda Campbell,¹ Mike Oliver,² Simon Brockbank,¹ David A. C. Simpson,¹ William J. Curry¹
 
(The first two authors contributed equally to this publication)
 
 

¹Ophthalmic Research Centre, Institute of Clinical Science, Queen's University of Belfast, Belfast, Northern Ireland; ²Waters, Manchester, United Kingdom

Correspondence to: D. A. C. Simpson, Ophthalmology and Vision Science, Queens University Belfast, Institute of Clinical Science, Royal Victoria Hospital, Belfast BT12 6BA, Northern Ireland; Phone: (44) 28 90632719; FAX: (44) 28 90330744; email: david.simpson@qub.ac.uk

Abstract

Purpose: Elucidation of the transcriptome and proteome of the normal retina will be difficult since it is comprised of at least 55 different cell types. However the characteristic layered cellular anatomy of the retina makes it amenable to planar sectioning, enabling the generation of enriched retinal cell populations. The aim of this study was to validate a reproducible method for preparing enriched retinal layers from porcine retina.

Methods: The thicknesses of the retinal photoreceptor, inner nuclear and ganglion cell, and fiber layers were determined by routine histology of cross sections of fresh whole retina mounted on polyvinylidene difluoride (PVDF) membrane. Dissected retina (5 mm²) was placed on PVDF membrane and a series of planar cryosections corresponding to the photoreceptor and inner nuclear layer were removed leaving the ganglion cell and fiber layer which was subsequently detached from the membrane. The retinal specimens were stored at -80 °C. Representative planar tissue sections were sonicated in ice-chilled 40 mM ammonium bicarbonate pH 7.9 and aliquots removed for RNA extraction. Quantitative RT-PCR was used to analyze the mRNA expression of genes indicative of specific retinal layers. Ammonium bicarbonate protein extracts were centrifuged, lyophilized and prepared for direct liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis using a Waters Q-Tof Ultima.

Results: Histological analysis established the parameters for planar cryosectioning: photoreceptor layer (69±1.8 μm), outer plexiform (11±0.6 μm), inner nuclear layer (28±0.5 μm), inner plexiform, ganglion cell and fiber layer (100±5.3 μm). Gene expression profiling provided an independent method for validating the respective retinal preparations. For example, glial fibrillary acidic protein (GFAP) was expressed up to 21 fold higher in the inner retinal "ganglion cell enriched" fraction than in the outer retinal "photoreceptor enriched" fraction. The pattern was reversed for blue cone opsin, which was expressed up to 24 fold higher in the "photoreceptor enriched" fraction. Endogenous protein fragments indicative of each layer were identified by mass spectrometry and de novo sequence data obtained.

Conclusions: Combined histological and mRNA expression profiling has confirmed the development of a reproducible method for generating validated porcine retinal layers enriched for specific cell types. Direct proteome analysis detected endogenous peptide fragments of characteristic retinal proteins. Further analysis of these enriched retinal cell preparations will facilitate a more selective investigation of the retinal transcriptome and proteome than studies of the intact retina.

Introduction

The value of global gene expression analysis has been demonstrated by microarray studies that have enhanced current knowledge of retinal biology and revealed genes involved in retinal degeneration [1,2]. Likewise, the development of mass spectrometry technologies interfaced with genome databases have enabled investigations of tissue and cellular proteomes including those of the retina to be initiated [3]. mRNA concentrations are indicative of the expression of their cognate proteins, but they are not an absolute measure and there may be major discrepancies between mRNA and protein concentrations [4]. In addition, mRNA studies cannot indicate the degree of post-translational modification, which influences protein function and stability. Direct analysis of protein expression patterns in the retina is therefore required to complement information derived from gene expression studies [3].

Analyses of bovine [5,6] and rat [7] retina have initiated the establishment of species-specific retinal proteomes. However, the complexity of the retina, which is composed of approximately 50 distinct cell phenotypes [8], will confound holistic proteome analysis; cell type and even more so, cell-specific proteomes, will be submerged amongst the total tissue protein. One potential approach to help reduce the complexity would be to analyze isolated retinal cell types. However, the resultant cellular proteomes become progressively less representative of the norm with increased post-isolation events [9]. Alternative strategies exploiting the laminar architecture of the retina employed vibratome preparation [5] and retinal "printing" [6]. Alternatively, focusing on abundant organelles such as mitochondria [10] and synaptosomes [11] facilitates the acquisition of selected retinal proteome data. The application of laser capture microdissection in other tissue types has demonstrated its capacity to define specific cellular proteomes [12,13].

Although numerous rodent models of human retinal disease exist, the small size of the retina restricts analysis of rare proteins and peptides. Porcine eyes share a similar size, architecture, and overall cone to rod ratio (albeit lacking a cone dominated macula) with the human eye [14,15]. Studies have demonstrated that the pig is a highly suitable model for the human pathology of retinitis pigmentosa [16]. The objective of this study was to demonstrate that it is possible to use cryosectioning to reproducibly generate planar porcine retinal specimens enriched for the cell types resident in the ganglion cell layer (GCL), inner nuclear layer (INL), and photoreceptor layer (PRL). Validation of this technique was achieved by histological measurement, mRNA profiling for genes indicative of specific retinal cell types, and proteomic analysis of endogenously generated peptide fragments.

Methods

Preparation of retinal samples

Porcine (White Landrace) eyes from animals slaughtered for human food consumption were collected within 30 min of death and placed on ice for transportation to the laboratory. Left and right eyes were identified, a single bisecting cut was made through the ora serrata generating the posterior chamber, the vitreous was removed and the posterior cup rinsed in chilled 0.1 M PBS containing 5 mM EDTA (Sigma, Gillingham, UK). The retina was manually detached from the retinal pigmented epithelium and placed, ganglion nerve fiber layer uppermost, in a Petri dish containing chilled 0.1 M PBS on an ice stage. Retinal specimens (5 mm²) were consistently recovered from the mid-peripheral retina and placed photoreceptor layer uppermost on a PVDF membrane support (VWR International, Poole, UK). The thickness of each retinal layer was assessed using fresh unfixed tissues rapidly mounted in Optimal Cutting Temperature (OCT) compound (VWR International Ltd., Poole, UK), prior to freezing (-25 °C). Transverse cryosections (10 μm) were air dried (30 min) on silanized slides (Superfrost Plus, VWR International) prior to histological staining with Toludine blue, with a wash in 0.1 M PBS prior to viewing using a Nikon Eclipse E400 microscope interfaced with Lucia 4.60 image analysis software (Laboratory Imaging, Praha, Czech Republic). Ten measurements of each retinal layer were completed from each of 4 retinal sections to establish the mean thickness of the retinal layers (Table 1).

PVDF planar-mounted retinal specimens (5 mm²) were placed on an OCT support generated by cryosectioning and frozen at -25 °C. Planar tissue specimens corresponding to the thickness of the retinal layers were collected and stored at -80 °C prior to analysis. Transverse sections were also prepared from selected planar specimens following removal of the PRL and INL; these specimens were stained histologically to further verify removal of the appropriate layer.

For each layer, tissue from two preparations was pooled and suspended in ice chilled 40 mM ammonium bicarbonate (pH 7.9), and sonicated (35 kHz) in an iced water bath for 15 min. Aliquots (40 μl) were removed for RNA extraction and the remaining supernatant was subjected to gentle agitation (18 h, 4 °C). Following centrifugation at 15000x g for 20 min the supernatant was decanted and the protein content estimated using the 2D Quant kit (Amersham Biosciences).

Quantitative RT-PCR

RNA was extracted from 40 μl aliquots of each retinal sample using 1 ml of TRI reagent (Sigma) and an RNeasy kit (Qiagen, Crawley, UK) incorporating DNase treatment. Reverse transcription was performed with Superscript II reverse transcriptase (Invitrogen, Paisley, UK), primed with random hexanucleotides (Roche, Lewes, UK). Real time PCR was performed on a LightCycler (Roche) using a QuantiTect PCR kit (Qiagen) according to the manufacturer's instructions. All reactions were performed for 40 repeats of the following cycle: 94 °C, 15 s; 50 °C, 15 s; 72 °C, 20 s.

PCR primers were designed using Vector NTI software (Informax Inc., Oxford, UK) and used at a final concentration of 1 μM. No porcine sequence was available for neurofilament (NFL) gene and primers were therefore designed from conserved regions of human, murine and rat orthologs. In these cases the identity of the amplicon was confirmed by nucleotide sequencing. Primer sequences and relevant accession numbers are shown in Table 2.

Mass spectrometry and protein identification

40 mM ammonium bicarbonate extracts were lyophilized and reconstituted in trifluoroacetic acid (TFA)/water prior to direct analysis of endogenously generated peptide fragments by liquid chromatography-mass spectrometry (LC-MS) using a modular CapLC connected to a Nano flow Z-spray source of a Q-Tof Ultima API (Waters, Elstree, UK). Briefly, peptides were loaded and desalted on a 300 μm C₁₈ precolumn at a flow rate of 30 μl/min, eluted from the precolumn onto a 75 μm C₁₈ analytical column at a flow rate of 200 nl/min. The peptides were eluted using an increasing gradient of acetonitrile from 7% (v/v) to 45% (v/v) over 27 min. Mass spectra (MS) data were acquired on the Q-Tof Ultima API operating in data directed acquisition mode with the instrument selecting up to the three most intense peptides for MS/MS data acquisition. The instrument was calibrated with a fifth order calibration using selected ions from Glu-fibrinopeptide-B.

Data were processed automatically using the ProteinLynx Global Server version 2 (Waters) and the SWISSPROT protein database and Automod, de novo sequencing and BLAST searches [17] were performed. In addition, peak list files from ProteinLynx were submitted via Mascot [18] to an MS/MS Ion Search of Mass Spectrometry protein sequence DataBase (MSDB). Proteins identified and de novo sequenced peptides are shown in Table 3.

Results

Generation of retinal enriched specimens

Unfixed tissue was chosen for histological analysis to minimize potential architectural deformation that may have arisen following conventional fixation protocols designed to optimize tissue morphology. The ability of planar cryosectioning to remove sections highly enriched for specific retinal layers was readily demonstrated (Figure 1) and the mean thicknesses of the photoreceptor, outer plexiform, and inner nuclear layers of the mid-peripheral porcine retina were determined (Table 1). These data were used to generate consecutive cryosections of 75 μm that contained a predominance of photoreceptor and outer nuclear layer tissues and a 40 μm specimen enriched in cells derived from the inner nuclear layer and tissue from inner plexiform layer. The remaining tissue, comprising the remnant of the inner plexiform layer and the ganglion cell and nerve fiber layers, was peeled while frozen, from the PVDF support.

Quantitative RT-PCR

A quantitative measure of the enrichment of each sample was obtained by measuring the expression of genes specific for each layer by quantitative RT-PCR (Figure 2). Due to the small amount of tissue available for analysis, relative levels of rRNA in each sample were determined using primer pairs specific for 18S and 28S rRNA. The results obtained for each gene were almost identical (Figure 2) and as such were regarded as sufficient for sample normalization according to the criteria as described [19,20]. Subsequent gene specific amplifications were normalized according to rRNA levels (using an average value obtained from both 18S and 28S data). Rhodopsin (opsin 2, rod pigment) was on average approximately 2 and 6 fold increased in the PRL relative to the "INL" and "GCL" preparations, respectively. The enrichment for blue sensitive cone opsin in the "PRL" was even more marked at 3.5 and 13.5 fold, respectively. Enrichment of rhodopsin and opsin relative to whole retina was 1.5 and 4.5 fold, respectively. Another marker for rod photoreceptors, S-antigen (SAG), demonstrated an expression pattern similar to that of rhodopsin. Genes known to be expressed predominantly in the GCL, such as glial fibrillary acidic protein (GFAP), exhibited the predicted expression pattern (15.5 fold enrichment in the GCL relative to PRL).

Mass spectrometry

Positive identifications were determined for twelve different proteins according to the Mascot probability-based mowse scores (p<0.05). De novo sequence was obtained for a further 44 peptides and two proteins matched by BLAST searches (Table 3). Proteins from two of the genes for which mRNA expression analysis was performed were detected; GFAP was present in the GCL and S-antigen in the PRL and INL.

Discussion

The goal of proteomics is to define the total number of proteins and their post-translational derivatives expressed in an organism, tissue, cell, or subcellular organelle at a defined time. This task is not insurmountable but is greatly exacerbated by cellular heterogeneity which is particularly evident in the retina, which contains at least 55 defined cell types [8]. Therefore it would be useful to devise strategies to minimize cellular complexity. Ideally this could be achieved by rapid isolation of specific cell types; however it is not surprising that the impact of cellular isolation significantly modifies the resultant proteome [9,21]. Since the retina exhibits a defined layered anatomy it is amenable to laminar separation and previous studies have exploited this fact [5,7]. This study employed laminar cryosectioning of fresh frozen retina (-25 °C). Within any heterogeneous tissue such as the retina, the degree of enrichment of the preparations is central to the power to detect the transcriptome and proteome. Therefore, 5 mm² planar samples were employed to minimize the effect of inherent retinal curvature whilst maximizing the purity of the recovered samples. Three modes of evaluation were employed to measure the degree of enrichment of retinal layer specimens: histology, mRNA profiling, and protein fragment analysis.

Toludine blue histology of unfixed retina proved a simple, rapid, and effective visual mode to establish the parameters of section thickness and to determine the quality of planar sectioning. mRNA profiling employing genes characteristic of different retinal cell types offered a quantitative mode of analysis, which could be applied to an aliquot of the actual sample used in subsequent proteomic analyses. Relative rRNA levels were used as a measure of variation in total RNA concentrations between samples. The accuracy of rRNA quantification was demonstrated by the similarity of expression profiles obtained for 18S and 28S rRNA [19,20]. This approach enables the normalization of samples too small for conventional spectrophotometric quantification. Following normalization, the relative gene expression patterns observed for five retinal genes in three independent sets of samples was highly reproducible. This demonstrated the effectiveness of this normalization approach even for samples such as these, which had widely varying RNA concentrations due to differences in the thickness and cellular content of sections from different layers and the exact number of sections successfully collected from specific retinas.

Gene expression analysis revealed that, as would be expected, the photoreceptor-specific genes rhodopsin and SAG were both present at highest levels in the PRL preparation. Rhodopsin mRNA was not restricted to the PRL preparation because the samples are not comprised exclusively of one retinal layer, but are enriched for specific layers. This reflects the fact that the flat planes of the chosen sections do not coincide exactly with the boundaries of the layers. This would be impossible because the layers are integrated with one another, there are local variations in thickness, there is inherent retinal curvature, and some folding can occur during retinal preparation. The degree of enrichment could be increased by reducing the diameter and thickness of the sections, at the expense of reducing the amount of tissue recovered. Greater enrichment was observed in the PRL preparation for cone opsin than the rod specific genes. This is probably because cone photoreceptors form a smaller proportion of the total retina and are located toward the outer edge of the ONL, making them less likely to contaminate specimens collected from the INL preparation.

To our knowledge this is the first report to use quantification of mRNA levels to validate sample enrichment prior to proteomic analysis. This approach could be adapted by varying cryosection thicknesses and analysing alternative genes to focus on different or more specific areas of the retina. By extension to DNA quantification it would be possible to define preparations enriched for nuclear or plexiform layers.

The present proteome research platform of choice is 2D electrophoresis followed by tryptic digestion, MS analysis, and database interrogation. However, several new approaches utilizing direct MS-based technologies have been employed to enhance both defined [22] and global proteome analysis [23,24]. This study employed direct proteome analysis of endogenously generated protein fragments using a Waters Q-Tof Ultima API system and readily detected proteins characteristic of the defined retinal layers. Additional unidentified endogenously generated peptide fragments were also detected. This mode of direct proteome analysis of unresolved protein extracts is comparable to shotgun proteome studies previously employed in the analysis of Saccharomyces cerevisiae [25,26] and human lens tissue [22].

Whilst the choice of pig provides abundant tissue for protein analysis, the paucity of publicly available porcine protein and nucleotide sequence creates a problem. The identification of proteins from the MS data must rely on the conservation of sequence with homologous proteins from other organisms which are present in the databases. The feasibility of this approach was demonstrated by glutamine synthetase which was identified both by matching of peptides to the porcine protein (AAO64254) and to orthologous proteins from multiple organisms including rodents, cows, dogs and humans. In all, it was possible to identify twelve different proteins from which peptides were derived. The validity of the protein identifications is supported by previous reports of their expression in the retina and the presence of predicted proteins in each of the layers. The photoreceptor-specific protein S-antigen [27-29] is highly expressed, explaining its positive identification in the PRL and WR. Its presence in the INL preparation must reflect "contamination" of this sample with photoreceptor tissue. S-antigen mRNA expression, as expected, was highest in the PRL, but was also high in the WR and INL (Figure 2). Low mRNA expression in the GCL correlated with the lack of detection of S-antigen protein in this sample. Likewise, GFAP protein was identified only in the GCL, correlating with mRNA expression >10 fold higher in this layer relative to the INL or PRL. Haemoglobin beta chain was identified at a high concentration in the GCL, but also in the INL and WR, but not the PRL; this correlates with the distribution of vessels within the retina.

The amino acid glutamate is the major excitatory neurotransmitter in the retina and is released by photoreceptors, bipolar cells, and ganglion cells [30]. To prevent neurotoxicity, glutamate must be rapidly removed from the extracellular space by glial cells [31], in which it is metabolized to the non-toxic amino acid glutamine by the enzyme, glutamine synthetase (GS; EC 6.3.1.2). GS would therefore be expected to be highly expressed throughout the retina and indeed is a major spot on 2D gels of whole retinal protein extracts [32]. The detection of GS in all the retinal layers confirms its high level of ubiquitous expression.

The small number of significant similarities revealed by BLAST searches between the de novo sequenced peptides and known proteins highlights the problem of lack of porcine sequences. However, this will be resolved with ongoing sequencing efforts and data will be re-analyzed at a later date. No known biologically active peptides were identified and the physiological relevance of the observed peptides remains to be elucidated.

Conclusion

This study has demonstrated that use of a large animal model facilitates the procurement of defined intra-retinal regions for proteome analysis. The application of histological analysis allied with mRNA profiles provides a reliable method of validating tissue samples for subsequent proteomic analysis. Sufficient protein was extracted for LC-MS/MS analysis and despite the complexity of the GCL, INL, and PRL preparations, it was possible to detect the most highly expressed polypeptides and identify a proportion of the proteins from which they were derived. However, to investigate the more weakly expressed proteins characteristic of specific retinal cell populations, simpler protein preparations will be required. These could be provided by adaptation of the presented methods to use, for example, thinner cryosections. Moreover, the ability to extract meaningful data from even smaller amounts of starting material is important to achieve clinical applicability, especially within the field of retinal dysfunction [3]. This approach will provide the opportunity for detailed investigations of how the proteome is altered during retinal degeneration.

Acknowledgements

This work was supported by grants from the British Retinitis Pigmentosa Society and The R & D Office of the Northern Ireland Health and Social Services Central Services Agency to DS and JC. The authors wish to thank Mr. Terence Archibald, Stevenson's Meats, Cullybacky, NI for provision of porcine tissue and Mr. Ronald Hunter, Department of Agriculture and Rural Development, NI for DNA sequence analysis. Dr. M. Oliver is an employee of Waters (UK).

References

1. Livesey FJ, Furukawa T, Steffen MA, Church GM, Cepko CL. Microarray analysis of the transcriptional network controlled by the photoreceptor homeobox gene Crx. Curr Biol 2000; 10:301-10.

2. Kennan A, Aherne A, Palfi A, Humphries M, McKee A, Stitt A, Simpson DA, Demtroder K, Orntoft T, Ayuso C, Kenna PF, Farrar GJ, Humphries P. Identification of an IMPDH1 mutation in autosomal dominant retinitis pigmentosa (RP10) revealed following comparative microarray analysis of transcripts derived from retinas of wild-type and Rho(-/-) mice. Hum Mol Genet 2002; 11:547-57.

3. Poulaki V. Microdissection, microarrays and proteomics: a new approach to the study of eye diseases. Graefes Arch Clin Exp Ophthalmol 2003; 241:527-8.

4. Gygi SP, Rochon Y, Franza BR, Aebersold R. Correlation between protein and mRNA abundance in yeast. Mol Cell Biol 1999; 19:1720-30.

5. Nishizawa Y, Komori N, Usukura J, Jackson KW, Tobin SL, Matsumoto H. Initiating ocular proteomics for cataloging bovine retinal proteins: microanalytical techniques permit the identification of proteins derived from a novel photoreceptor preparation. Exp Eye Res 1999; 69:195-212.

6. Matsumoto H, Komori N. Ocular proteomics: cataloging photoreceptor proteins by two-dimensional gel electrophoresis and mass spectrometry. Methods Enzymol 2000; 316:492-511.

7. Cavusoglu N, Thierse D, Mohand-Said S, Chalmel F, Poch O, Van-Dorsselaer A, Sahel JA, Leveillard T. Differential Proteomic Analysis of the Mouse Retina: The Induction of Crystallin Proteins by Retinal Degeneration in the rd1 Mouse. Mol Cell Proteomics 2003; 2:494-505.

8. Masland RH. Neuronal diversity in the retina. Curr Opin Neurobiol 2001; 11:431-6.

9. Alge CS, Suppmann S, Priglinger SG, Neubauer AS, May CA, Hauck S, Welge-Lussen U, Ueffing M, Kampik A. Comparative proteome analysis of native differentiated and cultured dedifferentiated human RPE cells. Invest Ophthalmol Vis Sci 2003; 44:3629-41.

10. Morel V, Poschet R, Traverso V, Deretic D. Towards the proteome of the rhodopsin-bearing post-Golgi compartment of retinal photoreceptor cells. Electrophoresis 2000; 21:3460-9.

11. Jimenez CR, Eyman M, Lavina ZS, Gioio A, Li KW, van der Schors RC, Geraerts WP, Giuditta A, Kaplan BB, van Minnen J. Protein synthesis in synaptosomes: a proteomics analysis. J Neurochem 2002; 81:735-44.

12. Craven RA, Totty N, Harnden P, Selby PJ, Banks RE. Laser capture microdissection and two-dimensional polyacrylamide gel electrophoresis: evaluation of tissue preparation and sample limitations. Am J Pathol 2002; 160:815-22.

13. Banks RE, Dunn MJ, Forbes MA, Stanley A, Pappin D, Naven T, Gough M, Harnden P, Selby PJ. The potential use of laser capture microdissection to selectively obtain distinct populations of cells for proteomic analysis--preliminary findings. Electrophoresis 1999; 20:689-700.

14. Chandler MJ, Smith PJ, Samuelson DA, MacKay EO. Photoreceptor density of the domestic pig retina. Vet Ophthalmol 1999; 2:179-184.

15. Gerke GC Jr, Hao Y, Wong F. Topography of rods and cones in the retina of the domestic pig. Hong Kong Med J 1995; 1:302-8.

16. Petters RM, Alexander CA, Wells KD, Collins EB, Sommer JR, Blanton MR, Rojas G, Hao Y, Flowers WL, Banin E, Cideciyan AV, Jacobson SG, Wong F. Genetically engineered large animal model for studying cone photoreceptor survival and degeneration in retinitis pigmentosa. Nat Biotechnol 1997; 15:965-70.

17. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol 1990; 215:403-10.

18. Perkins DN, Pappin DJ, Creasy DM, Cottrell JS. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 1999; 20:3551-67.

19. Thellin O, Zorzi W, Lakaye B, De Borman B, Coumans B, Hennen G, Grisar T, Igout A, Heinen E. Housekeeping genes as internal standards: use and limits. J Biotechnol 1999; 75:291-5.

20. Goidin D, Mamessier A, Staquet MJ, Schmitt D, Berthier-Vergnes O. Ribosomal 18S RNA prevails over glyceraldehyde-3-phosphate dehydrogenase and beta-actin genes as internal standard for quantitative comparison of mRNA levels in invasive and noninvasive human melanoma cell subpopulations. Anal Biochem 2001; 295:17-21.

21. Hauck SM, Suppmann S, Ueffing M. Proteomic profiling of primary retinal Muller glia cells reveals a shift in expression patterns upon adaptation to in vitro conditions. Glia 2003; 44:251-63.

22. Skold K, Svensson M, Kaplan A, Bjorkesten L, Astrom J, Andren PE. A neuroproteomic approach to targeting neuropeptides in the brain. Proteomics 2002; 2:447-54.

23. MacCoss MJ, McDonald WH, Saraf A, Sadygov R, Clark JM, Tasto JJ, Gould KL, Wolters D, Washburn M, Weiss A, Clark JI, Yates JR 3rd. Shotgun identification of protein modifications from protein complexes and lens tissue. Proc Natl Acad Sci U S A 2002; 99:7900-5.

24. Romijn EP, Krijgsveld J, Heck AJ. Recent liquid chromatographic-(tandem) mass spectrometric applications in proteomics. J Chromatogr A 2003; 1000:589-608.

25. Link AJ, Eng J, Schieltz DM, Carmack E, Mize GJ, Morris DR, Garvik BM, Yates JR 3rd. Direct analysis of protein complexes using mass spectrometry. Nat Biotechnol 1999; 17:676-82.

26. Wolters DA, Washburn MP, Yates JR 3rd. An automated multidimensional protein identification technology for shotgun proteomics. Anal Chem 2001; 73:5683-90.

27. Mirshahi M, Boucheix C, Collenot G, Thillaye B, Faure JP. Retinal S-antigen epitopes in vertebrate and invertebrate photoreceptors. Invest Ophthalmol Vis Sci 1985; 26:1016-21.

28. Reid DM, Loeffler KU, Campbell AM, Forrester JV. Electron immunocytochemical localization of retinal S-antigen with a rat monoclonal antibody. Exp Eye Res 1987; 45:731-45.

29. Singh AK, Kumar G, Shinohara T, Shichi H. Porcine S-antigen: cDNA sequence and expression in retina, ciliary epithelium and iris. Exp Eye Res 1996; 62:299-308.

30. Wassle H, Boycott BB. Functional architecture of the mammalian retina. Physiol Rev 1991; 71:447-80.

31. Whetsell WO Jr, Shapira NA. Neuroexcitation, excitotoxicity and human neurological disease. Lab Invest 1993; 68:372-87.

32. Curry WJ, Healy A, Wallace A, Miller V. Porcine neural retinal proteomics. ARVO Annual Meeting: 2003 May 4-May 9; Fort Lauderdale, FL.