Knowledge of proteins expressed during retinogenesis is pivotal to gaining a better understanding of the events that drive retinal neurogenesis and synaptogenesis, both factors are central to the generation of the mature retina and to its maintenance throughout adulthood. This basic information may also provide vital clues to direct the development of stem cell therapies to treat a spectrum of sight threatening degenerative retinal diseases.
This investigation demonstrated the feasibility of employing proteome analysis of the embryonic and post hatch chick retina to generate spatio-temporal maps of protein expression and to identify a number of proteins that were developmentally regulated. 2D PAGE resolved on average 1514 protein spots per gel in the pH range 4–7. Although wide range IPG strips such as pH 3–10 would give a wider pH separation range on a 2D gel, there will be a trade off with resolution. A pH 4–7 strip will only resolve proteins with a pI between 4 and 7, however, it will provide much better spot resolution as it separates the proteins over the same distance as a pH 3–10 strip, proteins may therefore be resolved much more clearly on a narrower pH range than they would be on a wider one.
Identification of proteins was performed using MALDI and ESI MS. In order to make a positive protein identification the matched peptides were always related back to the raw spectra to check the matched peptides were the most intense peaks in the spectra. The search results were then also evaluated by the Mowse score. In Table 2 proteins identified by MALDI all had Mowse scores greater than 79 and were significant (p < 0.05), scores lower than this (proteins identified by ESI MS), although not significant in terms of the Mowse score were still the best matches. Because the Mowse score depends on the number of sequences in the database at the time of the search, species with small database sizes (chicken at that time had a small database size) the Mowse score-based approach was not the most appropriate to determine the significance of a match. In an ideal situation the best match and the most significant match would be the correct match, however, Mowse score significance is highly dependent on the data quality, and, the data generated by ESI MS/MS was not as high a quality as that produced by MALDI, there were less digested fragments and the equipment seemed to be less sensitive, there were therefore less mass values to search the database with. In these cases matches were critically assessed, they were compared to existing 2D gel databases and assessed by correspondence of the experimental pI and molecular weight to the pI and molecular weight based on the spot's position within the gel. These proteins were also compared to proteins that had previously been identified from our lab from porcine retinal 2D gels.
The majority of protein spots on the gels displayed conserved expression densities between the ages studied, however from E12 to P33 approximately 198 protein spots increased ≥ 1.5 fold (p < 0.05), and, 234 decreased ≥ 1.5 fold (p < 0.05). Expression changes were evident in a large population of proteins below the 1.5 fold threshold, but these proteins were not subjected to further analysis in this study as they fell outside the p < 0.05 significance level.
Haniu et al. (2006)  introduced the concept 'proteomic trajectory mapping', to profile the spatiotemporal kinetics of protein expression in the mouse retina. Likewise, in this study proteins were placed into four groups based upon this basic classification system. Accordingly, proteins that decreased in expression from E12 to P33 (i.e. highly expressed during early retinogenesis) were termed 'juvenile' 'J' type proteins, 4 of the identified proteins belonged to this group. The second group of proteins, 'adult' 'A' type, gradually increased in expression during retinogenesis, 4 of the identified proteins were more highly expressed in the later stages of retinal development. The third group of proteins were termed 'transient' 'T' type and 1 identified protein showed significant changes in expression midway through retinogenesis at some point between E12 and P33, and, finally 7 proteins that showed no detectable significant variation from E12 to P33 were termed 'constitutive' 'C' type.
Mizukami et al. (2008)  studied chick retina at three embryonic ages and identified two of the proteins also identified in the present study, the differences in proteins identified between this study and that of Mizukami et al. (2008)  may reflect the fact that a different method of protein extraction was employed in the present study and a narrower pH range (pH 4–7) was used which allows for much sharper resolution of proteins with pIs falling within this pH range.
Nucleophosmin 1 (NPM1) was classified as a J-type protein, two spots on the 2D gels were identified as NPM1, the horizontal resolution along the acidic to basic axis was indicative of some form of as yet unidentified post translational modification (PTM). NPM1 is a member of a family of nuclear chaperone proteins that are implicated in aiding the proper assembly of nucleosomes and maintaining chromatin structure, both key events in preserving cellular function [30, 31]. Previous studies have shown that NPM1 knockout causes embryonic lethality between E11.5 and E12.5  and knockout animals displayed many developmental abnormalities including complete absence of eyes and deficient brain organogenesis . The present study demonstrated that NPM1 was highly expressed at early stages of retinogenesis and exhibited a steady decline in expression with retinal maturation. This observation allied with the NPM1 knockout murine data  would suggest this protein may play an important role in the early stages of ocular and retinal development.
Fatty Acid Binding Protein 7 (FABP7/B-FABP) was identified and its detection was consistent with previous studies reporting maximal expression at E7/E9 in the chick retina, with a general decrease in expression thereafter [33–36]. The down-regulation of this protein has also been previously reported in postnatal chick retina . FABPs are involved in signal transduction, lipid trafficking toward specific metabolic pathways, and fatty acid transport and metabolism during neuronal development [37–39]. FABPs are reported to be important in the development of the characteristic lipid compositions of neural tissue  and studies suggest that FABP expression in embryonic chick retina may be linked to the demand for polyunsaturated fatty acids during retinogenesis and that these molecules may sequester fatty acids in preparation for neurite outgrowth as retinal cells differentiate .
Stathmin 1 was classified as a J-type protein and showed decreased expression with retinal development. Previous studies detected elevated expression in the nervous system where it may be involved in the regulation of developmental cell proliferation and differentiation, and plasticity [39, 40]. It has also been reported to exhibit high expression at early stages of retinal development which decreased with retinal maturation [29, 35].
The present study detected an increase in B-CK expression from ED12 through to P33 (A-type protein), it is plausible that it may be involved in retinal cell proliferation, migration and differentiation. High B-CK levels were previously detected in the chick retina, with intense immunostaining evident in the GCL during early development (E2–E5) and in differentiating photoreceptor cells at E18 . Retinal expression increased until hatching; this was followed by a slight decline with constitutive expression during adulthood .
Platelet-activating factor (PAF) acetylhydrolase, isoform 1b, alpha 1 subunit exhibited a slight increase from E12 to E13, then gradually decreased into adulthood and was classified as a J-type protein. It is an ether-linked phospholipid derived from the metabolism of membrane phospholipids . Previous studies detected PAF expression in the chick retina in response to specific neurotransmitter stimuli [45–47]. However, the exact role of PAF remains unclear; it may be involved in intercellular signaling  and/or neuronal differentiation .
Beta-synuclein exhibited an A-type profile; its expression increased 8.6 fold from E12 to P33. Synucleins are small, soluble proteins primarily expressed in neural tissue  and in certain tumours . This family of molecules has been extensively studied due to their involvement, particularly alpha-synuclein, in neurodegenerative pathologies such as Alzheimer's disease . Alpha, beta and gamma synuclein are present in the retina and intense beta-synuclein has been detected in the IPL .
DJ1 protein displayed a gradual increase in expression form E12 to P33, and was classified an A-type protein. Haniu et al.  identified this protein in the mouse retina where it exhibited constitutive expression. This observation may reflect important differences in the ratios of cell types expressed in the developing cone-enriched chick versus the rod-dominant murine retina. DJ-1 has been implicated in the control of cell survival  and altering p53 activity , it was also found to be highly expressed in neuronal cells and to a lesser extent in non-neuronal cells in the murine brain . Previous studies have reported that DJ-1 plays an antioxidant role by protecting against oxidative damage and in the prevention of apoptosis [56, 57]. Presently there is no reported association between DJ1 and retinal function, although it is likely to be an important retinal protein based on previous reports of its role as an antioxidant and in the prevention of mitochondrial-induced cell death.
Capping protein (actin filament muscle Z line) a T-type protein is reported to control the addition of actin subunits to actin filaments and it also nucleates actin polymerization in vitro . The increase in capping protein expression at E17 coincides with the morphogenesis of the photoreceptor outer and inner segments; however, further analysis would be required to substantiate linkage.
Additional proteins, alpha and beta-tubulin, gamma-actin, PREDICTED: similar to TGF-beta interacting protein 1, dimethylarginine dimethylaminohydrolase, triosphoaphate isomerase, enhancer of rudimentary homologue were also identified as constitutive type proteins, i.e. their expression did not significantly change during development.
Analysis of any developmentally driven tissue that exhibits significant changes in cell number with the concomitant acquisition of distinct cellular phenotypes ensures that its study is inherently complex and the data generated must be interpreted with caution. Nonetheless, this study has demonstrated the potential of proteomics to profile protein expression in the developing chick retina and has generated data on a number of developmentally regulated proteins that may be intrinsic to retinal development and maturation. In order to begin to understand and treat such debilitating diseases as RP we must first unravel the events driving normal retinogenesis. The identification of proteins expressed during retinogenesis and the analysis of their expression levels throughout retinal maturation will only increase our knowledge of the subject, and, with the possibility of stem cell therapy as a means to treat RP, knowledge of protein expression during critical stages of retinal development raises the possibility of directing cell type differentiation and integration by manipulating trophic factors and protein levels to minimize rejection. The continued generation of substantive mass spectrometric data for additional protein populations that exhibit fold changes during development is an absolute necessity and will enable the establishment of a comprehensive retinal development proteome database. With the establishment of such databases we can then begin to construct a clearer picture of such processes as retinogenesis which may ultimately increase our understanding of such complex developmental mechanisms.