Ординатура / Офтальмология / Английские материалы / Retinal Degeneration Disease_Hollyfield, Anderson, LaVail_1999

Figure 66.1. Structural Model of the CRALBP Ligand Binding Domain (residues 143-301). The twelve residues implicated in ligand interactions by biochemical analyses are labeled; 11-cis retinal occupies the ligand pocket. A. View parallel to the beta sheet. B. View perpendicular to that of A.

3.4. The rCRALBP Ligand-Binding Cavity from Topological Analyses

Hydrogen/deuterium (H/D) exchange detected by mass spectrometry was used to measure the effect of protein-bound 11-cis-retinal on the solvent accessibility of amide hydrogens.16 N-terminal residues 5-42, C-terminal residues 282-316, and residues 41-71, 80-94, 127-137 and 262-275 all incorporated more deuterium in holo-rCRALBP and therefore are more solvent exposed in the holo-protein. With the exception of residues 127-137, the central region of CRALBP (residues 111-197) incorporated low levels of deuterium with or without ligand and appears to be buried in both the apo and holo protein structures. The only region that incorporated significantly more deuterium in the absence of bound 11-cis- retinal encompasses amino acids 198-255 and most of the identified ligand cavity residues. The results are also consistent with an earlier lower resolution topological analysis of CRALBP using antibodies and proteases.18

4. CRALBP PROTEIN INTERACTIONS

4.1. CRALBP Interacts with 11-cis-Retinol Dehydrogenase (RDH5)

Early studies with crude extracts from bovine RPE microsomes support a substrate carrier interaction between CRALBP and 11-cis-RDH.4,19 Kinetic analyses performed with purified recombinant RDH5 and purified rCRALBP demonstrate a direct, functional interaction between rRDH5 and rCRALBP6 and are consistent with the notion that CRALBP affects the activity of RDH5 by ‘channeling’ retinoids to the enzyme. Immunoprecipitation experiments with the purified recombinant proteins and different anti-peptide CRALBP antibodies18 support a structural interaction in a C-terminal region of CRALBP (Figure 66.2). How retinoid is released from the high affinity CRALBP binding pocket for export from the RPE for visual pigment regeneration remains unresolved. Interactions with RDH5 may be involved in the release mechanism.

Figure 66.2. Immunoprecipitation of RDH5 with anti-CRALBP. Purified recombinant RDH5 and CRALBP were mixed together then exposed to different agarose bead-immobilized anti-peptide antibodies recognizing the CRALBP residues in parentheses. Coomassie blue stained SDS-PAGE is shown above and the Western analyses below with anti-RDH5 indicate only the N-terminally directed anti-K25Q pulled down rRDH5, supporting an interaction with a C-terminal region of CRALBP.

4.2. CRALBP Interacts with EBP50

An overlay assay was used to detect interactions of CRALBP with components of RPE microsomes.20 Interacting proteins were separated by 2D-PAGE and identified by LC MS/MS. Protein interactions were characterized by affinity chromatography, peptide competition, and recombinant expression of protein domains. CRALBP bound to a 54 kD protein in RPE microsomes identified as ERM (ezrin, radixin, moesin)-binding phosphoprotein 50 (EBP50). EBP50 is also known as NHERF-1 (sodium/hydrogen exchanger regulatory factor type-1). EBP50 was found in multiple adjacent 2D gel spots, suggesting phosphorylaton may play a role in regulating retinoid trafficking. CRALBP bound to both recombinant PDZ domains of EBP50 but not to the C-terminal ezrin-binding domain. In outer retina, EBP50 and ezrin were localized to RPE and Müller apical processes. CRALBP was distributed throughout both RPE and Müller cells including their apical processes. ERM proteins are multivalent linkers that connect plasma membrane proteins with the cortical actin cytoskeleton. EBP50 interacts with an N-terminal domain of the ERM proteins and binds other targets through its PDZ domains, thus contributing to an apical localization of target proteins. These results support a retinoid-processing complex in the apical RPE20 and are consistent with observations of retinoid processing proteins in reciprocal immunoprecipitations of RPE microsomes using antibodies to visual cycle proteins.21 The C-terminus of CRALBP interacts with the PDZ domains in EBP5020 and a separate C-terminal domain in EBP50 binds ezrin or another ERM family member. Notably, a fraction of RDH5 activity appears to be associated with the RPE plasma membrane.22 Interactions with EBP50 and ezrin may help localize the apical plasma membrane where interaction with RDH5 could facilitate the release of 11-cis-retinal from CRALBP for export to the photoreceptors for visual pigment regeneration.

4.3. Retinoid-Processing Proteins in the Apical RPE

To further test the hypothesis of a retinoid-processing complex in apical RPE, we pursued proteomic analysis of RPE apical processes and plasma membranes. RPE apical membranes were isolated from mouse eyecups on lectin-coated agarose beads. Morpho-

Figure 66.3. Hypothesis for a retinoid-processing complex in apical RPE. Left: A model representation of apical RPE process filled with actin fibers cross-linked to each other and to the plasma membrane. Right: Enlarged view of the model in the left with identified protein components labeled.

logical analyses of the bead-bound RPE apical membranes support a highly purified preparation. Proteins were recovered from the beads, separated by SDS-PAGE, gel bands excised, digested in situ with trypsin and proteins identified by LC MS/MS. CRALBP, EBP50, RDH5 and ezrin were among the proteins bound to the lectin beads.23 These findings further support the proposed retinoid-processing complex in the RPE. Many other proteins were identified from the apical RPE membranes. The information should provide insights into processes occurring at this critical interface, which are important for the support and maintenance of vision.24

5. CONCLUSIONS

Considerable progress has been made in understanding the biochemical basis and enzymology of the visual cycle and associated retinal diseases. However, many aspects of the visual cycle remain unclear. The retinoid isomerization chemistry is controversial because the responsible enzymes have yet to be isolated. Less is known about cone visual pigment regeneration but the process is different than that for rhodopsin and may involve 11-cis- retinoid synthesis in Müller cells. In the RPE, the interaction of cytosolic proteins with mem- brane-associated enzymes is largely uncharacterized and little is known about mechanisms of retinoid trafficking, and spatial-compartmental organization of visual cycle components. Identification of visual cycle proteins associated with CRALBP will promote a better understanding of these processes and will advance the development of therapies for visual disorders.

6. ACKNOWLEDGEMENTS

This study was supported in part by NIH grants EY6603, EY14239, EY015638, GM63020, EY01730, EY02317, an unrestricted award from Research to Prevent Blindness,

Inc. (University of Washington), a Research Center Grant from The Foundation Fighting Blindness, and funds from the Cleveland Clinic Foundation.

7.REFERENCES

1.T.D. Lamb & E.N. Pugh, Jr., Dark adaptation and the retinoid cycle of vision, Prog Retin Eye Res 23:307-80 (2004).

2.J.C. Saari, J. Huang, D.E. Possin, R.N. Fariss, J. Leonard, G.G. Garwin, J.W. Crabb & A.H. Milam, CRALBP is expressed by oligodendrocytes in optic nerve and brain, Glia 21:259-68 (1997).

3.J.A. Treadwell, K.B. Pagniello & S.M. Singh, Genetic segregation of brain gene expression identifies retinaldehyde binding protein 1 and syntaxin 12 as potential contributors to ethanol preference in mice, Behav Genet 34:425-39 (2004).

4.J.C. Saari, D.L. Bredberg & N. Noy, Control of substrate flow at a branch in the visual cycle, Biochemistry 33:3106-12 (1994).

5.J.C. Saari, M. Nawrot, B.N. Kennedy, G.G. Garwin, J.B. Hurley, J. Huang, D.E. Possin & J.W. Crabb, Visual cycle impairment in cellular retinaldehyde binding protein (CRALBP) knockout mice results in delayed dark adaptation, Neuron 29:739-48 (2001).

6.I. Golovleva, S. Bhattacharya, Z. Wu, N. Shaw, Y. Yang, K. Andrabi, K.A. West, M.S. Burstedt, K. Forsman, G. Holmgren, O. Sandgren, N. Noy, J. Qin & J.W. Crabb, Disease-causing mutations in CRALBP tighten and abolish ligand interactions, J Biol Chem 278:12397-402 (2003).

7.A. Winston & R.R. Rando, Regulation of isomerohydrolase activity in the visual cycle, Biochemistry 37:20442050 (1998).

8.H. Stecher, M.H. Gelb, J.C. Saari & K. Palczewski, Preferential release of 11-cis-retinol from retinal pigment epithelial cells in the presence of CRALBP, J Biol Chem 274:8577-85 (1999).

9.M.A. Maw, B. Kennedy, A. Knight, R. Bridges, K.E. Roth, E.J. Mani, J.K. Mukkadan, D. Nancarrow, J.W. Crabb & M.J. Denton, Mutation of the gene encoding cellular retinaldehyde-binding protein in autosomal recessive retinitis pigmentosa, Nat Genet 17:198-200 (1997).

10.M.S. Burstedt, O. Sandgren, G. Holmgren & K. Forsman-Semb, Bothnia dystrophy caused by mutations in the CRALBP gene (RLBP1) on chromosome 15q26, Invest Ophthalmol Vis Sci 40:995-1000 (1999).

11.H. Morimura, E.L. Berson & T.P. Dryja, Recessive mutations in the RLBP1 gene encoding cellular retinalde- hyde-binding protein in a form of retinitis punctata albescens, Invest Ophthalmol Vis Sci 40:1000-4 (1999).

12.N. Katsanis, N.F. Shroyer, R.A. Lewis, J.C. Cavender, A.A. Al-Rajhi, M. Jabak & J.R. Lupski, Fundus albipunctatus and retinitis punctata albescens in a pedigree with an R150Q mutation in RLBP1, Clin Genet 59:424-9 (2001).

13.E.R. Eichers, J.S. Green, D.W. Stockton, C.S. Jackman, J. Whelan, J.A. McNamara, G.J. Johnson, J.R. Lupski & N. Katsanis, Newfoundland rod-cone dystrophy, an early-onset retinal dystrophy, is caused by splicejunction mutations in RLBP1, Am J Hum Genet 70:955-64 (2002).

14.J.W. Crabb, Z. Nie, Y. Chen, J.D. Hulmes, K.A. West, J.T. Kapron, S.E. Ruuska, N. Noy & J.C. Saari, Cellular retinaldehyde-binding protein ligand interactions. Gln-210 and Lys-221 are in the retinoid binding pocket, J Biol Chem 273:20712-20 (1998).

15.Z. Wu, Y. Yang, N. Shaw, S. Bhattacharya, L. Yan, K. West, K. Roth, N. Noy, J. Qin & J.W. Crabb, Mapping the ligand binding pocket in CRALBP, J Biol Chem 278:12390-6 (2003).

16.Z. Wu, A. Hasan, T. Liu, D. Teller, J.C. Saari, & J.W. Crabb, Identification of CRALBP Ligand Interactions by Photoaffinity Labeling, Hydrogen/Deuterium Exchange and Structural Modeling, J Biol Chem 279:2735727364 (2004).

17.C. Panagabko, S. Morley, M. Hernandez, P. Cassolato, H. Gordon, R. Parsons, D. Manor & J. Atkinson, Ligand specificity in the CRAL-TRIO protein family, Biochemistry 42:6467-74 (2003).

18.J.W. Crabb, V.P. Gaur, G.G. Garwin, S.V. Marx, C. Chapline, C.M. Johnson & J.C. Saari, Topological and epitope mapping of CRALBP from retina, J Biol Chem 266:16674-83 (1991).

19.J.C. Saari & L. Bredberg, Enzymatic reduction of 11-cis-retinal bound to cellular retinal-binding protein,

Biochim Biophys Acta 716:266-72 (1982).

20.M. Nawrot, K. West, J. Huang, D.E. Possin, A. Bretscher, J.W. Crabb & J.C. Saari, Cellular retinaldehydebinding protein interacts with ERM-binding phosphoprotein 50 in retinal pigment epithelium, Invest Ophthalmol Vis Sci 45:393-401 (2004).

21.S.K. Bhattacharya, Z. Wu, Z. Jin, L. Yan, M. Miyagi, K. West, M. Nawrot, J.C. Saari & J.W. Crabb. Proteomic Approach to Identification of a Mammalian Visual Cycle Protein Complex. in Experimental Biology 2002 Vol. 16: A14 (FASEB J, New Orleans, LA, 2002).

22.N.L. Mata, E.T. Villazanna & A.T.C. Tsin, Colocalization of 11-cis retinyl esters and retinyl ester hydrolase activity in retinal pigment epithelium plasma membrane, Invest Ophthal Visual Sci 39:1312-19 (1998).

23.V.L. Bonilha, S.K. Bhattacharya, K.A. West, J.S. Crabb, J. Sun, M.E. Rayborn, M. Nawrot, J.C. Saari & J.W. Crabb, Support For a Proposed Retinoid-Processing Protein Complex in Apical Retinal Pigment Epithelium, Exp Eye Res 79:419-22 (2004).

24.V.L. Bonilha, S.K. Bhattacharya, K.A. West, J. Sun, J.W. Crabb, M.E. Rayborn & J.G. Hollyfield, Proteomic Characterization of Isolated Retinal Pigment Epithelium Microvilli, Mol Cell. Proteomics 3:1119-27 (2004).

A group of small GTPases in the Ras superfamily were shown to interact specifically with PDEd. Most of these GTPases have a common feature, that is, a CAAX box motif in their C-terminal amino acid sequence which signals posttranslational modification resulting in cleavage of -AAX polypeptides followed by carboxymethylation and prenylation (prenylthioether formation) of the cysteine residue. The last residue of the CAAX box specifies whether the added prenyl chains are either farnesyl (C15 moieties) or geranylgeranyl (C20 moieties). Prenylation is a step necessary for the targeting of proteins to membrane, and interaction of PDEd with prenylated proteins relies on the prenyl lipid chain. The binding of PDEd to PDE in vitro is mediated by its prenylated C-terminal (Cook et al., 2000). The y2h screening in our lab identified rhodopsin kinase (GRK1) as another interacting member. Mammalian GRK1s are farnesylated proteins and the prenyl group is required for interaction with PDEd. We also showed that PDEd interacted with GRK7, a cone pigment kinase believed to be important for the recovery of photoresponse in cone photoreceptors in some species (Chen et al., 2001; Weiss et al., 2001; Weiss et al., 1998). GRK7 is a geranylgeranylated protein. Table 67.1 provides a summary of a subset of proteins which can interact with PDEd.

3. PDEd BINDS ISOPRENYL CHAINS IN THE ABSENCE OF

C-TERMINAL CYSTEINE

Given that PDEd interacts with multiple isoprenylated proteins and that isoprenylation is a requisite for interaction, we sought direct evidence for the interaction between PDEd and prenyl chain alone. By conjugating a fluo-rescent probe to either farnesyl and geranylgeranyl isoprenoid chains, we used a technique called FRET (Fluorescence Resonance Energy Transfer) to demonstrate isoprenoid binding to PDEd (Zhang et al., 2004). FRET (reviewed by (Centonze et al., 2003)) revealed a significant increase in fluorescence emission from one of the fluorescent probes suggesting interaction between PDEd and the iso-

prenoid chain. By varying ligand concentrations, the binding constants for farnesyl and geranylgeranyl chains were determined to be 0.70 mM, and 19.06 mM, respectively.

4. PDEd INTERACTS WITH Arl2 AND Arl3

Y2h screening also identified PDEd interacting proteins other than isoprenylated proteins, such as Arl2 and Arl3 (Hanzal-Bayer et al., 2002). So far, Arl2 and Arl3 are the only non-isoprenylated proteins shown capable of PDEd interaction. Arl2 and Arl3 are also small GTPases belonging to the Arf family in the Ras superfamily. Arf family members play important roles in intracellular vesicular trafficking processes. Arl proteins differ from Arfs in that Arl proteins do not possess co-factor activity of cholera toxin-catalyzed ADPribosylation of Gas subunits. And unlike in Arfs, the glycine residue at position 2 in some Arls cannot be myristoylated (Sharer et al., 2002). The binding of PDEd to Arl3 depends on the guanine nucleotide species bound to Arl3, and b-sheet interaction. PDEd only interacts with GTP-bound Arl3, but not GDP-bound Arl3, suggesting PDEd is one of the effectors of Arl3 GTPase (Linari et al., 1999a).

5. STRUCTURE OF PDEd

BLAST searching generated the first clue about the structure PDEd, which suggested that PDEd has relatively weak sequence similarity to the C-terminal part of RhoGDI (Nancy et al., 2002). RhoGDI can bind Rho GTPase through interaction with its C-terminal isoprenyl group. Shortly after the identification of sequence homology, the crystal structure of PDEd, cocrystallized with Arl2/GTP, was published (Hanzal-Bayer et al., 2002). It confirmed the striking structural similarity between PDEd and RhoGDI, although their primary sequences have very low homology (Figure 67.1). Both of these two proteins have a hydrophobic pocket formed between two b-propellers. This structure further confirmed that PDEd is a prenyl binding protein. Although the binding of PDEd to prenylated proteins usually depends on the structure of the prenyl group, the protein-protein interaction may also be important for their tight binding. One example for this notion is that excess amount of prenyl compounds can only partially compete off the binding between PDEd and GRK7. This additional protein-protein interaction could be responsible for the binding specificity of PDEd, a possible reason why both PDEd and RhoGDI only bind a subset of isoprenylated proteins. As for RhoGDI, it only interacts with proteins of the Rho GTPase family, whereas PDEd interacts with a larger number of polypeptides (PDEa, PDEb, GRK1, and GRK7 in photoreceptors).

6. CELLULAR LOCALIZATION OF PDEd IN PHOTORECEPTORS

Several results have been published regarding localization of photoreceptor PDEd leading to controversy. The first result of immunocytochemistry of PDEd in retina indicated that PDEd is localized to the outer segments of rods, but not cones (Florio et al., 1996). Later, PDEd was shown in dark-adapted retina to be distributed throughout the whole cell, from the synaptic termini to outer segments of rod and cone photoreceptors (Zhang et al.,

Figure 67.1. Structural homology of RhoGDI and PDEd. The PDEd sequence was aligned with the C-terminal portion of RhoGDI using ClustalW. Cylinders indicate two a-helices in PDEd, and arrows depict nine b-sheets, emphasizing the structural similarity (adapted from (Hanzal-Bayer et al., 2002). Note that sequence similarity is rather limited. Residues shaded yellow represent contacts with prenyl groups inside the binding pocket.

2004). Recent data demonstrated that PDEd in light-adapted bovine retina is localized primarily to the connecting cilium (Norton et al., 2004), a structure bridging the inner segment to the outer segment. The connecting cilium is a structure that controls the transport of phototransduction components from the inner segments, where biosynthesis takes place, to the outer segment. The localization of PDEd to the connecting cilium is consistent with the concept that PDEd may be involved in the intracellular trafficking of isoprenylated phototransduction proteins.

7. SUMMARY AND HYPOTHETICAL MODEL OF PDEd FUNCTION IN PHOTORECEPTORS

Several independent assays strongly suggest that a photoreceptor polypeptide, termed “PDEd”, is not an authentic PDE subunit, but a general prenyl binding protein (PrPB). Biochemically identified binding partners include rhodopsin kinase (GRK1), cone pigment kinase (GRK7), and the PDE catalytic subunits (PDEa,b), all of which are prenylated at their C-termini. In addition to prenylated partners, PDEd interacts with small GTPases like Arl2/3 which are not known to be prenylated. An important function of PDEd is to solubilize prenylated proteins which are normally membrane-associated.

Based on currently available information, we propose a model in which newly synthesized prenylated protein (GRK1) interacts with PDEd, and transports to the photoreceptor connecting cilium and the outer segment in conjunction with Arl2 charged with GTP. Hydrolysis of GTP may cause discharge of GRK1 into the OS. Arl2-GDP is recycled to Arl2-GTP by an unknown GEF most likely present in the inner segment and recombines again with PDEd for another round of transport.

Inner Segment

Arl-

GTP

PDEd GRK1

Endoplasmic Reticulum

Figure 67.2. Hypothetical model of photoreceptor PDEd in transport of rhodopsin kinase (GRK1) from the inner segment to the outer segment. PDEd binds to GTP-bound Arl2/3, and the resulting protein complex interacts with the tail of isoprenylated GRK1 which then dissociates from the ER (site of protein isoprenylation). The Arl2/3 and PDEd complex with GRK1 may diffuse to the connecting cilium, but more likely associate with membraneous carrier structure. At the cilium, a yet-identified GAP protein may catalyze the hydrolysis of GTP bound to ARl2/3 to GDP. Then the protein complex falls apart, and GRK1 is released after passage through the cilium. GTP-bound Arl2/3 are regenerated by an unidentified guanine nucleotide exchange factor (GEF), similarly as light activated R* promotes exchange of GDP by GTP bound to transducin.

8. FUTURE DIRECTION

Generation of a photoreceptor PDEd knockout will be extremely useful for revealing the in vivo function of PDEd in rod and cone photoreceptors. Our preliminary experiments suggested that the mouse lacking PDEd gene had reduced levels of PDE and rhodopsin kinase, as well as shortened rod outer segments, suggesting the importance of PDEd for the stability of its binding partners (Zhang et al., unpublished results). In the future, it will be important to study the roles Arl2 and Arl3 may play in transport of phototransduction components.

9. REFERENCES

Anant, J. S., Ong, O. C., Xie, H., Clarke, S., O’Brien, P. J., and Fung, B. K. K. (1992). In vivo differential prenylation of retinal cyclic GMP phosphodiesterase catalytic subunits. J. Biol. Chem. 267:687-690.

Baehr, W., Devlin, M. J., and Applebury, M. L. (1979). Isolation and characterization of cGMP phosphodiesterase from bovine rod outer segments. J. Biol. Chem. 254:1669-11677.

Beavo, J. A., Conti, M., and Heaslip, R. J. (1994). Multiple cyclic nucleotide phosphodiesterases. Mol. Pharmacol. 46:399-405.

Centonze, V. E., Sun, M., Masuda, A., Gerritsen, H., and Herman, B. (2003). Fluorescence resonance energy transfer imaging microscopy. Methods Enzymol. 360:542-560.

Chen, C. K., Zhang, K., Church-Kopish, J., Huang, W., Zhang, H., Chen, Y. J., Frederick, J. M., and Baehr, W. (2001). Characterization of human GRK7 as a potential cone opsin kinase. Mol. Vis. 7:305-313.

Cook, T. A. and Beavo, J. A. (2000). Purification and assay of bovine type 6 photoreceptor phosphodiesterase and its subunits. Methods Enzymol. 315:597-616.