|Home | About | Journals | Submit | Contact Us | Français|
Tyrosine-O-sulfation, a post-translational modification, is catalyzed by two independent tyrosylprotein sulfotransferases (TPSTs). As an initial step towards understanding the role of TPSTs in retinal function, this study was undertaken to determine the extent to which tyrosine-O-sulfation of proteins is utilized in the retina. A previously characterized anti-sulfotyrosine antibody was used to determine the presence and localization of tyrosine-O-sulfated proteins (TOSPs) in the retina. Using Western blot, RT-PCR and immunohistochemical analyses, we detected TOSPs in the retinas from diverse species, including frog, fish, mouse and human. Some of the variability in the observed sizes of retinal TOSPs in the mouse, at least, may result from differential patterns of glycosylation; however, there seem to be species-specific sulfated retinal proteins as well. TOSPs were detected in most of the retinal layers as well as in the retinal pigment epithelium from human and mouse. Several retinal TOSPs were detected in the inter-photoreceptor matrix, which is consistent with the secreted nature of some sulfated proteins. Transcripts for both TPST-1 and -2 were expressed in both the human and mouse retinas. These data show that retinal protein tyrosine-O-sulfation is highly conserved which suggest a functional significance of these proteins to retinal function and structure.
Until recently, tyrosine-O-sulfation (TOS) of proteins, a post-translational modification, had received little attention. TOS occurs in the trans-Golgi network (Baeuerle and Huttner, 1987) and involves the covalent transfer of a sulfate group from the universal sulfate donor 3′-phosphoadenosine 5′-phosphosulfate to a tyrosine moiety of the nascent target peptide (Lee and Huttner, 1983). The sulfate transfer is catalyzed by two independent tyrosylprotein sulfotransferases (TPST, 3′-phosphoadenylyl-sulfate:protein-tyrosine O-sulfotransferase, EC 18.104.22.168) (Lee and Huttner, 1983;Moore, 2003). The two TPST enzymes differ in their acidic pH optima, degree of stimulation by Mn2+ and Mg2+ (Mishiro et al., 2006) and have differential substrate affinities. Compared with TPST-2, TPST-1 exhibits significantly lower Km and Vmax for the majority of the substrates tested (Mishiro et al., 2006). This is corroborated by the finding that in comparison to TPST-1, TPST-2 has a high degree of substrate preference towards the thyroid stimulating hormone receptor (Sasaki et al., 2007). Furthermore, studies on the expression of TPST-1 and TPST-2 demonstrated significant variations in the levels of expression among different tissues, suggesting distinct physiological functions for TPST-1 and TPST-2 (Mishiro et al., 2006;Ouyang, Lane et al., 1998;Ouyang and Moore, 1998;Beisswanger et al., 1998;Nishimura and Naito, 2007).
Many of the tyrosine-O-sulfated proteins (TOSPs) are secreted and may become part of the extracellular matrix through either direct autocrine re-association or association with other members of the extracellular matrix. In the absence of TOS, the normally sulfated protein may have a retarded transport (Friederich et al., 1988) or have an altered ability to interact with other proteins (Costagliola et al., 2002;Hirata et al., 2004).
Although both TPSTs are thought to be expressed universally, there is no information available regarding protein TOS in the retina. In this report, TOS of proteins was examined in retinas from various species. We found that sulfation occurs in retinas from diverse species, including frogs, fish, and humans. However, the sizes of TOSPs varied among these species, suggesting that post-translational glycosylation may contribute to some of the observed size differences. In addition, TPST-1 and -2 transcripts were expressed in both the human and mouse retinas. Furthermore, immunohistologic analysis detected TOSPs in most retinal layers and the retinal pigment epithelium (RPE). Interestingly, the sulfated proteins in the RPE seem to be different than those present in the retina.
A healthy human retina from a 61-year-old Caucasian donor was obtained from the Illinois Eye Bank, Chicago, IL. A trephine with a 4-mm inside diameter was used to dissect the fovea from the rest of the tissue, which is considered the peripheral retina in this study. Both samples were frozen in liquid nitrogen before long-term storage at −70°C. After the removal of the retina, the human RPE, without differentiating foveal from peripheral, was peeled and frozen as above until use. The dog retina was a generous gift from Dr. Gustavo Aguirre (University of Pennsylvania, Philadelphia, PA). Squirrel, chicken, frog, toad, fish and rat retinas were a generous gift from Dr. Muna Naash (OUHSC, Oklahoma City, OK). Pig and cow retinas were obtained from local slaughter houses. The mouse retina was from an in-house inbred strain of wild-type (wt, (Xu et al., 2000)) mice. The rd/rd (C3H) mice were purchased from Charles River (Redfield, Arkansas). After the extraction of the mouse retina, eye cups were obtained and dissected in halves. Using a dissection microscope the RPE was then peeled. Mouse retinas and RPEs were frozen and stored as above.
All experiments were approved by the local Institutional Animal Care and Use Committees, and conformed to the National Institute of Health Guide for the Care and Use of Laboratory Animals and the guidelines of the Association for Research in Vision and Ophthalmology Resolution on the Use of Animals in Research.
RGC-5 (a rat ganglion cell line (Krishnamoorthy et al., 2001)), MC-1 (a rat Müller cell line (Sarthy, 1985)) and 661W cells (a mouse cone photoreceptor cell line (Tan et al., 2004)) were all grown as described before (Kanan et al., 2007). ARPE-19 cells were grown as described by Vogel et al. (Vogel et al., 2007).
The protocol for preparation of the soluble fraction of the inter-photoreceptor matrix (S-IPM) was a slight modification of that described before ((Johnson and Hageman, 1989;Hollyfield et al., 1990)). Briefly, 20 mouse retinas from in-house inbreed strain of mice (Xu et al., 2000) were extracted and incubated in PBS containing protease inhibitors (Complete mini, EDTA-free, Roche, Nutley, New Jersey) for 2 min. Then retinas were collected by mild centrifugation (200 g for 2 min in an Eppendorf microfuge, Westbury, NY) and the supernatant was centrifuged again at 20,000 g in the microfuge for 10 min. The resulting supernatant was considered as the S-IPM. The remainder of the sample was extracted as described below and named “IS-IPM+” since it contained the insoluble part of the IPM and the rest of the retina.
Western blots were processed as described before (Tan et al., 2001) and images were obtained using a Kodak image station (Carestream Molecular Imaging, Rochester, NY). Briefly, retinal protein extracts were prepared by homogenization of frozen tissue as has been described (Tan et al., 2001). Retinal extracts were combined with Laemmli sample buffer (Laemmli, 1970) containing 0.2 M Tris (pH 6.8), 1 mM EDTA, 4% (wt/vol) SDS, 20% (vol/vol) glycerol, 0.005% bromophenol blue, and 5% β-mercaptoethanol. After SDS-PAGE, gels were transferred to PVDF membrane (Immunoblot; Bio-Rad, Hercules, CA) and membranes were probed for sulfated tyrosines with the PSG2 antibody (Hoffhines al., 2006) at 200 ng/mL. Membranes were processed and imaged as detailed in (Tan et al., 2001). The PSG2 antibody was kindly provided by Dr. Kevin Moore (Oklahoma Medical Research Foundation, Oklahoma City, OK).
The specificity of the PSG2 antibody was demonstrated by Hoffhines et al. (Hoffhines et al., 2006) and confirmed by Western blot analysis (Supplemental Figure 1). Forty micrograms of total mouse retinal extracts were loaded into each of two lanes and subjected to SDS-PAGE. The proteins were transferred to PVDF membrane. The two lanes were separated by cutting the membrane; one lane was probed with PSG2 and the other was probed with PSG2 in the presence of 4 mM sulfotyrosine. As shown in Supplemental Figure 1, four of the protein bands detected with PSG2 (black arrows) were not detected when sulfotyrosine was included. However, a protein band at ~27 KDa (black arrowhead) was not competed out, suggesting that it is not a sulfated protein.
The anti-TPST-1 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA or from Abnova Corporation, Taipei, Taiwan) and anti-TPST-2 antibodies (Abcam, Inc., Cambridge, MA) were used at 1 μg/ml. The Santa Cruz anti-TPST-1 antibody (sc-25031) is a goat anti-peptide antibody raised against a peptide mapping near the N-terminus of human TPST-1. The Abnova anti-TPST-1 antibody (H00008460-A01) is a mouse anti-partial recombinant GST-tagged human TPST-1. The Abcam anti-TPST-2 antibody (ab59958) is a rabbit anti-human TPST-2 antibody raised against a peptide mapping to the N-terminus conjugated to KLH.
Peptide:N-Glycosidase F (PNGaseF) was obtained from ProZyme (San Leandro, CA) and reactions with PNGaseF were carried out as recommended by supplier.
Total RNA was isolated from mouse and human retinas or the MCF-7 human breast cancer cell line by using the Trizol reagent (Invitrogen, Carlsbad, CA). Complimentary DNA was synthesized using Superscript III (Invitrogen). Genomic DNA contamination was prevented by treatment of the samples with RQ1 RNAse free DNAse (Promega, Madison, WI) and verified by the absence of amplification in samples that were not reverse transcribed to cDNAs. Amplified samples were separated on 1% agarose gels in the presence of ethidium bromide and alongside the 1 Kb plus DNA ladder (Invitrogen). Amplification of the house-keeping gene hypoxanthine-guanine phosphoribosyltransferase (HPRT) was used as a control as described before (Kanan et al., 2008).
Several cell lines, including MCF-7, were tested for the presence of sulfated proteins by using Western blot analysis with the PSG-2 antibody (data not shown). Reverse transcription (RT)-PCR confirmed that the MCF-7 cell line had both TPSTs present, thus it was chosen as a positive control.
Human Tpst1 Forward 5′ CCACAGACTGAGCAAGTGGA 3′
Reverse 5′ CAATACACTTAAACAGGCACACG 3′
Product size 483 bp
Human Tpst2 Forward 5′ CATGGAGGTAGGCAAGGAGA 3′
Reverse 5′ CCTTTCAGATTGGCTGGTGT 3′
Product size 465 bp
Mouse Tpst1 Forward 5′ AGCTGGCTTTGACCTGAACA 3′
Reverse 5′ TTAGGAGGATTGGCGTATGG 3′
Product size 402 bp
Mouse Tpst2 Forward 5′ AGTGACACAGTCCTGCACCA 3′
Reverse 5′ TGCTTGGGTTAGTCCAGCTT 3′
Product size 424 bp
Sections of a human retina were a generous gift from Dr. Joe Hollyfield (Cleveland Clinic, Cleveland, OH). The eye was obtained from a 57-year-old Caucasian male donor who had no history of vision problems. Eyes were fixed in 4% paraformaldehyde in 0.1 M phosphate. Sections were processed as described previously (Al-Ubaidi et al., 1997). Mouse retinal sections, from an in-house inbred strain of wild-type (wt, (Xu et al., 2000)) mice, were prepared and processed as described in Al-Ubaidi et al. (Al-Ubaidi et al., 1997). Sections were incubated with 1μg/ml PSG2 antibody.
To determine whether any retinal proteins are sulfated and whether sulfation of retinal proteins is biologically relevant, the degree of conservation of TOS as a post-translational modification was investigated (Figure 1). Western blot analysis of retinal extracts from various species probed with PSG2, the anti-sulfotyrosine antibody (Hoffhines et al., 2006), revealed the presence of sulfated proteins in the retinas from all species tested. These proteins ranged in size from 30 to 250 kDa (Figure 1). The retinas from human, squirrel and chicken had the most complex patterns of sulfated proteins. However, each species has at least one major and multiple minor sulfated proteins. The human fovea and peripheral retina exhibited similar patterns of TOSPs, suggesting that sulfated proteins are present in both rods and cones (Figure 1C). However, it is difficult to judge from only observed migration on SDS-PAGE whether the proteins (e.g. at 50KDa (white asterisk)) observed in both the fovea and the periphery are the same. Since TOSPs were detected in the retinas of animals that are evolutionally millions of years apart, this may suggest a strong biological significance for protein tyrosine sulfation.
All mammalian retinas used in the preparation of the Western blot (Figure 1) expressed TOSP(s) between 100 and 150 kDa in size, suggesting that these protein(s) may be common in most mammals. However, other sulfated proteins were detected only in certain species. These bands may represent the same proteins with differences in the degree and pattern of glycosylation that would influence the apparent molecular mass as observed on the Western blot. Consistent with this possibility, several TOSPs appeared as diffuse bands (Figure 1). To determine the contribution of glycosylation to the observed molecular mass of sulfated proteins, mouse retinal extracts were subjected to PNGaseF treatment before separation by SDS-PAGE. As shown in Figure 2A, PNGaseF treatment increased the mobility of both of the high molecular weight sulfated proteins. Interestingly, the peptides after PNGaseF treatment did not collapse to sizes similar to that observed in other species, suggesting that the variability in the sizes of TOSPs in the mouse retina is not resulting from differential glycosylation patterns.
The separation of the human retinal sample shown in Figure 1 into foveal and peripheral fractions was intended to determine whether there are rod- or cone-specific TOSPs. To determine whether some cell-specific TOSPs were present in the retina and since it is difficult to fractionate the different cell types from the retina in quantities sufficient for Western blot analysis, three established retinal cell lines were used. Extracts from RGC-5 (a rat ganglion cell line (Krishnamoorthy et al., 2001)), MC-1 (a rat Müller cell line (Sarthy, 1985)), and 661W (a mouse cone photoreceptor cell line (Tan et al., 2004)), were used. The mouse retina and all three cell lines expressed a major common protein band of approximately 125 kDa that was recognized by the PSG2 antibody (arrow, Figure 2B). However, there were minor cell-type specific protein bands (asterisks in each lane, Figure 2B) that were not observed when total retinal extracts were used. This is probably due to the relatively low contribution of their respective cell types to total retinal proteins. Furthermore, the lower band (immediately above 100 kDa) observed in total retinal extracts does not appear in the cell line lanes (Figure 2B), suggesting that it is produced by cell types other than cones, Müller or ganglion cells. Of specific interest is the appearance of TOSPs in extracts from 661W cells (Figure 2B) that may be a cone-specific product. Alternatively, the expression of these proteins may have been altered as a result of the growth of these cells in culture.
To further investigate the differential distribution of TOSPs in different cell types of the retina, extracts from one-month-old rd/rd retinas were analyzed. The rd/rd retina, which has a mutation in the β-subunit of phosphodiesterase, exhibits rod photoreceptor degeneration as early as postnatal day (P) 7 and complete rod photoreceptor loss after P21. Therefore, any TOSP band(s) that is observed in extracts from wt retinas but disappears in extracts from degenerated rd/rd retinas should be a rod-specific, sulfated protein. There is a TOSP band that is shared by wt and rd/rd retinas (black arrowhead, Figure 2C). However, two differences were observed when these extracts were probed with PSG2. The first is the presence of a band in wt retinal extracts that was absent from rd retinal extracts (asterisk, Figure 2C). Since the rod photoreceptors have degenerated by one month of age in the rd retina, and since they constitute ~60% of the total number of cells in the retina, it is possible to assume that this TOSP is expressed in rods. Alternatively, this protein may be produced by inner retinal cells and may have been lost during the re-organization of the retina as a consequence of the degeneration. The second difference is the appearance of a protein band in extracts from rd/rd retinas (black arrow, Figure 2C), which may be expressed by one of the remaining retinal cell types and only becomes apparent because of the increased contribution of those cell types after the disappearance of the rods.
Some TOSPs are known to be secreted and may or may not associate with the cells that produce them. To determine whether any TOSPs are secreted from retinal cells, 661W conditioned medium (661W CM) was collected, concentrated and used in Western blot analyses. Concentrated unconditioned medium (un-CM) was used a control. As seen in Figure 3A, probing with PSG2 detected two high molecular weight, secreted protein bands in the conditioned medium. The lower band was present in the total retinal extract as well as in extracts from 661W cells (Figure 2B). However, the higher band (asterisk, Figure 3A) is not observed in either retinal or 661W extracts (Figure 2B), suggesting that this is a secreted protein and a possible member of the extracellular matrix. Absence of that protein in the retinal extract may be a reflection of the small contribution of cones to the total number of cells in the retina. There may also be a third protein band found at ~90 KDa (white asterisk, Figure 3A) that seems diffused. The lower band that was seen in the 661W extracts at ~70 KDa (Figure 2B) was not found likely due to the distortion of the gel by the large amounts of albumin in the medium (gray arrowhead, Figure 3A). Finally, the lower band (black arrow, Figure 3A) that appears in the 661W CM, is also found in the un-CM lanes and is probably a component of the fetal bovine serum.
The inter-photoreceptor matrix (IPM) is a unique milieu that surrounds the outward extension of the photoreceptors. It is essential for the trafficking of nutrients and retinoids between the retina and the RPE, retinal attachment, photoreceptor alignment, and the cellular interactions involved in outer segment (OS) shedding and phagocyotsis by the RPE (Hollyfield, 1999). To prepare IPM fractions, freshly isolated retinas from mice were rinsed in PBS (pH 7.2) and the eluted proteins constituted the soluble fraction (S-IPM) while the remainder contained the insoluble fraction (IS-IPM+) of the IPM and the remainder of the retina. Western analysis using PSG2 was performed on both fractions and total retinal extracts from wt mice. As shown in Figure 3C, two major protein bands between 100 to 150 kDa were recognized by the PSG2 antibody in the S-IPM. The source of these proteins could be either the retina or the RPE. As expected, the TOSPs detected in total retinal extracts seem to be a combination of those observed in both the S-IPM and the IS-IPM+. It is clear that the top band is retained in the IS-IPM+, suggesting that this protein may be associating with the plasma membrane after being secreted. Other minor TOSPs were also detected in the S-IPM. One TOSP (arrowhead, Figure 3C) seems enriched in the S-IPM fraction while the other two (asterisks, Figure 3C) are not detected in total retinal extracts. The latter two protein bands may have become obvious due to their increased contribution to the total protein content in the S-IPM. However, comparing the Western blot probed with PSG2 (Figure 3C) to the Coomassie Blue stained sister gel (Figure 3B) clearly shows that only a minor fraction of the proteins in the IPM underwent TOS.
To determine whether any of the TOSPs observed in the S-IPM originated from the RPE, Western analysis was performed on extracts from mouse RPE, mouse S-IPM, and mouse retina (Figure 4). By only using the criterion of co-migration, we determined that the S-IPM contained TOSPs that likely originated from the retina itself (Figure 4, black arrows). Likewise, S-IPM also contained a TOSP that seemed to originate from the RPE (Figure 4, white arrowhead). The rest of the proteins that appear in the RPE lane seem to be unique to the RPE (asterisks, Figure 4). Minor amounts of the protein identified by an asterisk in Figure 4 appear in the S-IPM lane. Since extracted retinas are always contaminated by small amounts of RPE, it is likely that the presence of this protein species in the retina and S-IPM lanes is solely due to the contamination of these samples by the RPE.
To determine whether the human RPE differentially expresses a subset of TOSPs, Western analysis was performed on protein extracts from human retina, human RPE, and the ARPE-19 cell line using the PSG2 antibody. The human RPE expressed a different subset of TOSPs than the retina as judged by migration on SDS-PAGE (Figure 5). Compared to human RPE cells, ARPE-19 cells expressed a different subset of TOSPs (notice the two lower molecular weight species present in the extracts of adult human RPE, Figure 5). However, there is a protein band at ~125 KDa (arrow, Figure 5) that seems to be present in both retina and RPE. Since the ARPE-19 cells are not fully differentiated (day 5) this observation suggests that TOSPs may be developmentally regulated, at least in the RPE. Alternatively, these cells may have changed their characteristics as a result of their passage in culture.
Figure 6 shows the results of immunohistochemical (IHC) labeling of TOSPs in sections of human and mouse retinas with the PSG2 antibody. In the human sections (Figure 6A-C), labeling was observed in the RPE, photoreceptor outer segments (OS), photoreceptor inner segments (IS), photoreceptor outer nuclear layer (ONL), outer plexiform layer (OPL), inner nuclear layer (INL), ganglion cell layer (GCL) and inner limiting membrane (white arrowhead, Figure 6B&C). There seems to be an area clearly labeled between the OS and RPE that contained TOSPs (green arrow and inset, Figure 6C). These TOSPs could either be in the tips of the OS adjacent to the RPE or in the RPE processes that happened to separate upon section processing. Alternatively, this protein(s) could be in the IPM. Finally, the sclera (Sc) was clearly labeled in the presence of the antibody. In the absence of the PSG2 antibody (Figure 6A), labeling is observed in a blood vessel (white arrow) and to a lesser degree in the Sc.
To compare the cellular localization of TOSPs in the human retina to that of the mouse, IHC analysis was performed on wt mouse retinal sections using PSG2 antibody. As shown in Figure 6E, PSG2 antibody recognized retinal TOSPs in the Sc, RPE, OS, IS, ONL, INL, and GCL. Minor signal was observed in OPL and IPL. Although a stronger signal was observed in OPL and IPL of the human retina, the pattern of distribution of sulfated proteins was similar between both human and mouse retinas.
It is clear that human and mouse retinas express several TOSPs. However, since there are two enzymes that can sulfate proteins, experiments were performed to determine whether both of these enzymes are present in retinas of both species. Two sets of primers, one for each of the two enzymes in each species, were designed across introns to eliminate the potential amplification from genomic DNA. Furthermore, all RNA samples were treated with DNase and some PCR amplifications were performed in the absence of reverse transcription (−RT) (Figure 7). In that case only primer dimers were observed in the Tpst2 lanes (Figure 7). RNA extracted from MCF-7, the human breast cancer cell line, was used as a positive control (Figure 7) as described in the Experimental section. Although the primers for TPST-2 were less efficient in generating a robust product, it is clear that the human and mouse retinas contained transcripts for both enzymes (Figure 7). The identities of the amplified products were confirmed by subcloning and sequencing (data not shown). Since RT-PCR was not quantitative in nature (e.g. real time) it is not possible to conclude that there are more Tpst1 transcripts that Tpst2 in the human retina solely based on the intensity of the amplified products. Detection of TPST-1 and/or TPST-2 in the retina by Western or IHC analyses using commercially available antibodies was unsuccessful due to the lack of specificity of the antibodies (data not shown).
As the first step towards understanding the role of TOS in retinal function, an antibody that recognizes sulfotyrosine residues in proteins was used to determine the presence and localization of retinal TOSPs. This study demonstrated the evolutionary conservation of retinal TOS across diverse species. Some TOSPs seem to be species-specific as determined by the pattern of migration on SDS-PAGE (Figure 1). Several retinal TOSPs were observed in the IPM, which is consistent with the secreted nature of TOSPs. Furthermore, transcripts for both TPST-1 and -2 were expressed in human and mouse retinas.
Identification of retinal TOSPs will be instrumental to deciphering the role of sulfation in retinal function. Furthermore, it will be interesting to determine the retinal substrates for each of the two TPSTs and determine whether those targets are preferentially sulfated by one of the two TPSTs. It is also of interest to know the fate of retinal TOSPs when one or both of the TPSTs is absent. Freiderrich et al. (Friederich et al., 1988) showed that when the sulfation of the Drosophila melanogaster protein vitellogenin II was inhibited, the transport of the protein was retarded (Friederich 1988). However, inhibition of sulfation of P-selectin glycoprotein ligand-1 (PSGL-1), glycoprotein Ibα, factor V, factor VIII, C4, and type III and V collagens did not affect their surface expression or secretion (for reviews see (Monigatti et al., 2006;Moore, 2003)). Instead, sulfation of these proteins is important for their protein-protein interactions. This conclusion was strongly supported by the co-crystallization of sulfated proteins with their ligands. It has been shown that when thrombin and hirugen were co-crystallized, all three sulfated tyrosines were involved in an extensive intermolecular hydrogen-bonded network with α-thrombin (Skrzypczak-Jankun et al., 1991;Chen et al., 1995). Sulfation of tyrosine 385 on the thyroid stimulating hormone is required for its high-affinity binding and activation of the receptor, but not by thyroid-stimulating autoantibodies (Costagliola et al., 2002). Similarly, tyrosine sulfation of PSGL-1 is required for high affinity binding to P-selectin (Hirata et al., 2004;Rodgers et al. 2001;Pouyani and Seed, 1995;Sako et al., 1995;Wilkins et al., 1995). Sulfation has also been shown to play a role in optimal receptor-ligand interactions such as chemokine/chemokine receptor binding, optimal proteolytic processing of gastrin, and proteolytic activation of extracellular proteins such as factor V and VIII (Moore, 2003). However, the only published evidence showing that lack of sulfation in humans can lead to a phenotype is from cases of mild-to-moderate hemophilia A that result from missense mutations in the factor VIII gene. These mutations result in a Tyr1680Phe substitution in the von Willebrand factor binding site at the junction of the B and A3 domain (Jacquemin et al., 2000;Higuchi et al., 1990;Moore, 2003). It was later shown that optimal binding to von Willebrand factor requires sulfation of Tyr1680 in factor VIII (Leyte et al., 1991).
It is interesting to observe that a sulfated protein becomes obvious in extracts of rd/rd retinas immediately after the completion of the degenerative process. One explanation could be that this protein may be produced by the remaining cells in the rd/rd retina and only becomes obvious after the rods degenerated. Alternatively, it may be expressed in response to the re-organization that occurs in the rd/rd retina after degeneration of the rods. At present, it is difficult to favor either of the two possibilities. Similar situation may be present with the sulfated proteins expressed in ARPE-19 versus the fully differentiated RPE. Some of the sulfated proteins expressed in ARPE-19 cells are different from those expressed in differentiated RPE. Since ARPE-19 cells are not fully differentiated (day 5) this observation may suggest that TOSPs may be developmentally regulated, at least in the RPE. Alternatively, these cells may have changed their characteristics as a result of their passage in culture.
Another interesting observation is the strong label for TOSPs observed between the OS and RPE in the human retina (green arrow, Figure 6C). These sulfated proteins can be in the tips of the OS adjacent to the RPE or in the RPE processes that happened to separate upon section processing or in the IPM. Regardless of where these proteins are located, this suggests a role for these proteins in RPE-OS interactions and a potential role in the shedding process.
Although only a small subset of mostly large molecular weight TOSPs are observed on Western blots of total retinal extracts, other sulfated protein species are obvious when cell lines representing different retinal cell types were used. This suggests that there are several less-abundant TOSPs present in the retina.
In summary, this study demonstrates for the first time that protein sulfation occurs in the retina and suggests a biological significance based on evolutionary conservation. Future studies will focus on identifying the sulfated proteins by proteomics and determining whether the lack of sulfation could lead to retinal diseases.
Supplemental Figure 1. Specificity of the PSG2 anti-sulfotyrosine antibody as determined by competition Western blot analysis. Forty micrograms of total mouse retinal extracts were loaded into each of two lanes and subjected to SDS-PAGE. Following transfer to PVDF membrane, the two lanes were separated and one was probed with PSG2 and the other was probed with PSG2 in the presence of 4 mM sulfotyrosine. Small arrows point to sulfated retinal proteins that are competed out with sulfotyrosine while a ~27KDa protein is observed in presence of sulfotyrosine (arrowhead) suggesting that, although identified by the antibody, it is not sulfated.
The authors are grateful to Dr. Kevin L. Moore for providing the PSG2 antibody and for his critical comments on the manuscript. The authors thank Dr. Shannon Conley for her critical evaluation of the manuscript. None of the authors have any commercial interests.
The project described was supported by a grant from the Foundation Fighting Blindness (MRA) and partially supported by grants from the National Center For Research Resources [P20RR017703] and the National Eye Institute [P30EY12190] and [R01EY14052] (MRA), [R01EY018137] (MRA), Hope For Vision (MRA), Reynolds Oklahoma Center on Aging (MRA), and the Knights Templar Eye Foundation, Inc. (YK). The content is solely the responsibility of the authors and does not necessarily represent the official views of NIH or any of its institutes.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.