JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M401960200 on April 20, 2004

J. Biol. Chem., Vol. 279, Issue 26, 27357-27364, June 25, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
279/26/27357    most recent
M401960200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wu, Z.
Right arrow Articles by Crabb, J. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wu, Z.
Right arrow Articles by Crabb, J. W.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Identification of CRALBP Ligand Interactions by Photoaffinity Labeling, Hydrogen/Deuterium Exchange, and Structural Modeling*

Zhiping Wu{ddagger}§, Azeem Hasan{ddagger}||, Tianyun Liu**, David C. Teller**, and John W. Crabb{ddagger}§{ddagger}{ddagger}§§

From the {ddagger}Cole Eye Institute and {ddagger}{ddagger}Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, Ohio 44195, the §Department of Chemistry, Cleveland State University, Cleveland, Ohio 44115, and the **Department of Biochemistry, University of Washington, Seattle, Washington 98195

Received for publication, February 23, 2004 , and in revised form, April 5, 2004.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Cellular retinaldehyde-binding protein (CRALBP) functions in the retinal pigment epithelium (RPE) as an acceptor of 11-cis-retinol in the isomerization step of the rod visual cycle and as a substrate carrier for 11-cis-retinol dehydrogenase. Toward a better understanding of CRALBP function, the ligand binding cavity in human recombinant CRALBP (rCRALBP) was characterized by photoaffinity labeling with 3-diazo-4-keto-11-cis-retinal and by high resolution mass spectrometric topological analyses. Eight photoaffinity-modified residues were identified in rCRALBP by liquid chromatography tandem mass spectrometry, including Tyr179, Phe197, Cys198, Met208, Lys221, Met222, Val223, and Met225. Multiple different adduct masses were found on the photolabeled residues, and the molecular identity of each modification remains unknown. Supporting the specificity of photo-labeling, 50% of the modified residues have been associate with retinoid interactions by independent analyses. In addition, topological analysis of apo- and holo-rCRALBP by hydrogen/deuterium exchange and mass spectrometry demonstrated residues 198–255 incorporate significantly less deuterium when the retinoid binding pocket is occupied with 11-cis-retinal. This hydrophobic region encompasses all but one of the photo-labeled residues. A structural model of CRALBP ligand binding domain was constructed based on the crystal structures of three homologues in the CRAL-TRIO family of lipid-binding proteins. In the model, all of the photolabeled residues line the ligand binding cavity except Met208, which appears to reside in a flexible loop at the entrance/exit of the ligand cavity. Overall, the results expand to 12 the number of residues proposed to interact with ligand and provide further insight into CRALBP ligand and protein interactions.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Cellular retinaldehyde-binding protein (CRALBP)1 is a 36-kDa water soluble protein with a high affinity binding pocket for retinoids uniquely associated with vision, namely 11-cis-retinal and 11-cis-retinol (1). Human CRALBP gene defects can either tighten or abolish retinoid interactions, which in turn can compromise substrate carrier interactions with 11-cis-retinol dehydrogenase and lead to several retinal pathologies (2). Targeted disruption of the CRALBP gene in mice impairs regeneration of both rod and cone visual pigments (3). These and other studies establish multiple functions for CRALBP in the retinal pigment epithelium (RPE), including roles as a major acceptor of 11-cis-retinol in the isomerization step of the rod visual cycle and as a facilitator of oxidation of 11-cis-retinol to 11-cis-retinal by 11-cis-retinol dehydrogenase (2–6). Ongoing proteomic studies support the existence of a RPE retinoid processing protein complex containing CRALBP (7). Direct CRALBP interactions have been demonstrated in vitro with 11-cis-retinol dehydrogenase (2) and with ERM (ezrin, radixin, moesin)-binding phosphoprotein 50 (EBP50), also known as sodium hydrogen exchanger regulator factor type 1 (NHERF-1) (8). Interactions with EBP50 have been suggested as a mechanism for localizing CRALBP to the apical RPE plasma membrane for export of 11-cis-retinal to the adjacent rod photoreceptor cells for visual pigment regeneration (8). The functions of CRALBP in tissues other than the RPE remain to be determined (i.e. in retinal Müller cells, ciliary epithelium, iris, cornea, pineal gland, and a subset of oligodendrocytes of the optic nerve and brain).

To better understand CRALBP visual cycle functions, which require rapid association and dissociation of retinoid, we are characterizing the structure of the ligand binding pocket. Ligand interactions in CRALBP are non-covalent and previous structure-function studies (2, 9, 10) have implicated eight residues as possibly interacting with retinoid (Fig. 1). In this report, the interaction between human recombinant CRALBP and retinoid were characterized by mass spectrometry using hydrogen/deuterium exchange and photoaffinity labeling with 3-diazo-4-keto-11-cis-retinal (DK-11-cis-retinal). This photoaffinity analogue of 11-cis-retinal has previously been used to map the movement of ligand within rhodopsin following sensitization with light (11, 12). The present results not only identify new components of the CRALBP ligand cavity but also reveal unpredictable protein modifications generated with this retinoid photoaffinity reagent.



View larger version (15K):
[in this window]
[in a new window]
  		FIG. 1.
The amino acid sequence of human rCRALBP. Amino acids previously associated with CRALBP ligand interactions are shown in bold font and include Trp165, Met208, Gln210, Lys221, Met222, Met225, Arg233, and Trp244 (2, 9, 10). Boxed residues denote the location of disease-causing mutations. Residues 1–119 and 314–316 (underlined) maybe be removed by limited proteolysis without disrupting retinoid binding (42). The His tag fusion sequence is shown in italics, and asterisks denote photoaffinity-modified residues identified in the present report.

 
To further evaluate ligand interactions proposed from photoaffinity labeling and other biochemical analyses, we have generated a structural model of the CRALBP ligand binding domain. CRALBP belongs to the CRAL-TRIO protein family (13) and shares 20–33% sequence identity with conserved regions of three cytosolic lipid-binding proteins for which high resolution crystal structures are available. Our model of CRALBP is based on the crystal structures of homologues human {alpha}-tocopherol transfer protein ({alpha}TTP) (14, 15), yeast Sec 14, a phosphatidylinositol-binding protein (16), and human supernatant protein factor (SPF) (17). The structural model supports ligand interactions with the photolabeled residues and provides further insight into CRALBP ligand and protein interactions.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Materials—11-cis-retinal was obtained from the NEI, National Institutes of Health. The photoaffinity reagent DK-11-cis-retinal was a generous gift from Professor Koji Nakanishi, Columbia University (11, 12).

Production of rCRALBP—Wild-type human recombinant CRALBP (rCRALBP) was expressed in Escherichia coli strain BL21(DE3)LysS with an N-terminal His tag and apo-rCRALBP was purified to apparent homogeneity by nickel affinity chromatography (10, 18). Unless otherwise noted, protein in the first five nickel affinity chromatography fractions (0.5 ml each) was pooled for the present structure-function study.

Amino Acid Analysis, Electrophoresis, and Protein Quantification— Phenylthiocarbamyl amino acid analysis was performed using an Applied Biosystems model 420H/130/920 automated analysis system (19). SDS-PAGE was performed according to Laemmli (20) on acrylamide gels using a Mini-Protein II system (Bio-Rad). rCRALBP was quantified by amino acid analysis or according to Bradford (21) using rCRALBP previously quantified by amino acid analysis as the standard reference protein (10).

Non-covalent Retinoid Labeling of rCRALBP—Purified apo-rCRALBP was labeled with retinoid (DK-11-cis-retinal or 11-cis-retinal) using 1.2-fold molar excess retinoid over rCRALBP and excess retinoid was removed by Sephadex G25 spin chromatography (10). Retinoid binding properties were characterized by scanning UV-visible spectroscopy with a Hewlett Packard 8453 diode array spectrophotometer (10). All retinoid labeling and analysis procedures were performed in red light. Bleaching was by exposure to ambient white light for 20 min at 4 °C.

Photoaffinity Labeling of rCRALBP—rCRALBP with noncovalently bound DK-11-cis-retinal in 50 mM Tris-HCl pH 7.2 containing 1 mM dithiothreitol-EDTA, 67% glycerol was placed in the wells of a 24-well tissue culture plate (Model 5243, Corning Inc.) in the dark. The wells were partially immersed in liquid nitrogen and photolysis was performed at 254 nm for 5 s to 20 min with an 18.4 watt UV lamp (Model UVG-11, UVP, Inc., Upland, CA) held 2 cm above the plate. Photolysis was terminated by transferring the preparation from the plate to a tube and immediately extracting unbound retinoid 3x with hexane (2:1, v/v). Glycerol was removed by Centricon centrifugation (Ultrafree 5-kDa cutoff, Millipore, Bedford, MA), the photolabeled protein (DK-rCRALBP) concentrated by vacuum drying, then redissolved in 100 mM ammonium bicarbonate pH 8.0 containing freshly deionized 8 M urea. Protein-bound retinal was then reduced to retinol with tritiated NaB[3H]4 (5:1 molar ratio) (22). The radioactive, covalently modified protein ([3H]DK-rCRALBP) was exhaustively dialyzed against 0.1% trifluoroacetic acid to remove excess NaB[3H]4.

To demonstrate covalent labeling, intact [3H]DK-rCRALBP was analyzed by RP-HPLC using an Applied Biosystem Model 120 HPLC system and chromatography fractions (1 min) were collected for subsequent radioactivity measurements. To determine the mass of intact [3H]DK-rCRALBP, LC ESMS was performed with a PerkinElmer Sciex API 3000 triple quadrupole electrospray mass spectrometer equipped with an Applied Biosystems Model 140D HPLC system, a 5-µ Vydac C18 column (300 Å, 1.0 x 150 mm), aqueous trifluoroacetic acid/acetonitrile solvents, and at a flow rate of 20 µl/min (2). For LC ESMS, nitrogen was used as the nebulization gas (40 psi) and the curtain gas. Mass spectra were acquired in positive ion mode over a scan range of 300–2000 m/z with 0.2 atomic mass unit steps, 0.5 ms dwell time, 80 volt orifice potential, and 5000 volts ion spray.

Identification of rCRALBP Photoaffinity-modified Amino Acids— [3H]DK-rCRALBP was reduced with dithiothreitol and alkylated with iodoacetamide in the presence of 8 M urea, 100 mM ammonium bicarbonate pH 8 (23). Carboxyamidomethyl [3H]DK-rCRALBP (5 nmol) was digested overnight with trypsin (Promega, 20:1, w/w) in 2 M urea, 25 mM ammonium bicarbonate, pH 8.0, at 37 °C. Tryptic digests were fractionated by RP-HPLC using the Applied Biosystem Model 120 HPLC system as described for the intact protein. One hundred chromatography fractions (~100 µl each) were collected, radioactivity was measured, and four separate radioactive pools were prepared. The radioactive pooled fractions and all other tryptic peptide fractions were dried in a SpeedVac, redissolved in 2% acetonitrile, 0.1% formic acid (15 µl) for LC MS/MS analysis with a quadrupole time-of-flight mass spectrometer equipped with a CapLC (QTOF2, Waters Corporation, Milford, MA). Each sample (10 µl) was injected onto a Pepmap C18 trapping column (300 Å, 0.3 x 1 mm, LC Packings Inc., San Francisco, CA). Peptides were gradient-eluted and separated on a C18 capillary column (300 Å, 75 µ x 5 cm, New Objective, Woburn, MA) at a flow rate of 270 nl/min after splitting the pump eluents after mixing (30:1). Solvent A consisted of 0.1% formic acid and 2% acetonitrile, and solvent B consisted of 0.1% formic acid and 98% acetonitrile. The QTOF2 mass spectrometer was operated in standard data-dependent acquisition mode using survey scans over 300–2000 m/z acquired in 1.0 s followed by MS/MS over 50–1900 m/z in 2.1 s and with 0.1 s between survey scans (23). The capillary temperature was set at 120 °C, the voltage at 2.5 kV, and a collision energy profile of 25–55% was used for peptide fragmentation with argon. MS/MS data were searched against SwissProt or NCBI sequence databases utilizing Proteinlynx Global Server version 1.0 and Proteinlynx version 3.5 (23). All MS/MS spectra of photoaffinity-modified peptides were manually examined to confirm sequences.

Hydrogen/Deuterium Exchange and LC ESMS Analysis of Intact rCRALBP—Purified apo- or holo-rCRALBP with bound 11-cis-retinal (100 pmol) was equilibrated in 5 mM phosphate buffer pH 6.9 by buffer exchange using a Centricon (Ultrafree, 5-kDa cutoff) at room temperature. The apo- and holo-proteins were diluted 20-fold with D2O 5 mM phosphate buffer pD 6.9 (200 µl) and incubated for 5–60 s at room temperature. HCl (0.1 M, 14 µl) was added to lower the pD to 2.5, and the samples were placed in dry ice to quickly quench hydrogen/deuterium (H/D) exchange. The optimum reaction time with deuterium was determined by LC ESMS analysis of the intact proteins with a PerkinElmer Sciex API 3000 mass spectrometer equipped with an Applied Biosystems Model 140D HPLC system, and a 5 µ Vydac C4 column (300 Å, 1.0 x 50 mm). For these analyses, aqueous trifluoroacetic acid/acetonitrile solvents were used at 0 °C, with a gradient of 30–60% acetonitrile over 5 min, and a flow rate of 30 µl/min (24). All H/D-exchanged samples were stored at -80 °C until analysis. Deuterium concentration measurements (pD) were from pH meter readings without correction for isotope effects. Holo-rCRALBP was maintained in the dark or under red illumination prior to LC ESMS or peptic digestion.

Pepsin Digestion—Prior to H/D exchange analyses of rCRALBP peptides, the pepsin fragmentation pattern of rCRALBP (100 pmol) was determined (25). For these analyses, pepsin digestion was performed under conditions that limited back exchange of deuterium for hydrogen, namely 0 °C, pH 2.5, pepsin:rCRALBP 2:1 (w/w), and 5 min of digestion time (24). Immediately following digestion, peptides were subjected to LC MS/MS analysis using the above QTOF2 mass spectrometer equipped with an Applied Biosystems Model 140D HPLC system, a 5-µ Vydac C18 column (300 Å, 1.0 x 50 mm), aqueous formic acid solvents at 0 °C and gradient elution of 2–60% acetonitrile over 60 min. The QTOF2 mass spectrometer was operated with capillary and cone voltages set to 3 kV and 45 V, respectively, and in the standard data-dependent acquisition mode as described above. In addition, the "include function" was employed to sequence minor peptides, which were not detected by the standard analysis program (26). Peptides were identified using Proteinlynx Global Server version 1.0 and Proteinlynx version 3.5 to search an in-house sequence data base containing about 1000 protein entries including human rCRALBP.

Hydrogen/Deuterium Exchange and LC MS/MS Analysis of rCRALBP Peptides—H/D exchange, pepsin digestion, and LC MS/MS of peptic peptides were performed as described above with the following additions. Experiments were performed in duplicate for both apo-and holo-rCRALBP using 20-s deuterium incubation times and with protein isolated only in nickel affinity chromatography fraction 3. To minimize back-exchange of deuterium, a shorter HPLC gradient of 10–40% acetonitrile over 15 min was used and much of the HPLC system was maintained at 0 °C, including the solvent reservoirs, tubing, injector port, and C18 column (24). Deuterium incorporation in specific peptides was determined from the mass difference between the non-deuterated and deuterated peptide after correction for loss of deuterium during pepsin digestion and LC MS/MS analysis. Deuterium incorporation differences <5% were considered within experimental error of the methodology (25). To estimate deuterium loss during digestion and analysis, fully deuterated samples were prepared by overnight incubation in D2O at 60 °C then analyzed for m100% in Equation 1 (27),

(Eq. 1)
where m = the mass of deuterated peptide, m0% = the mass of nondeuterated peptide and m100% = the mass of the fully deuterated peptide. The number of incorporated deuterium in peptides was calculated using Equation 2,

(Eq. 2)
where N is the number of exchangeable peptide amide hydrogens.

Structural Modeling—The sequence alignment of {alpha}TTP, SPF, and Sec14 presented by Hendrickson and co-workers (15) was used as a guide for locally written programs to superimpose the ligand binding domains of these structures (28). These programs select residues for equivalence if the distance is within a cutoff threshold. In this method, the structures match well within the aligned region but may deviate outside those matched regions. In order to make a model of the ligand binding domain (CRALBP residues 143–301), the {alpha}TTP backbone and the residues that match in three structures were superimposed. With occupancy of 0.5 for {alpha}TTP and 0.25 for each of SPF and Sec14, the structures were placed in a box, and the space group assigned as P1. The CCP4 program sfall (29) was used to Fourier transform the three structures, and the format converted to that of XtalView (30). The resolution chosen was to 2.8 Å. The structure factors and phases from the transform of the three structures were used as a probability density. The CRALBP residues of the alignment were fitted into this probability density using XtalView. The primitive CRALBP model was refined to the probability density using CNS (31). The R-factor from this refinement was 0.30 and R-free was 0.33.

The CRALBP model was constructed in the ligand-bound form. The retinal schiff base was taken from the rhodopsin coordinates (1hzx [PDB] ) and converted to 11-cis-retinal. The 11-cis-retinal was oriented in the model by trial such that after refinement, the aldehyde group was solvent inaccessible (4, 9, 32). In MOE (33) the model was flooded with water molecules, energy minimized and dynamics calculated for up to 1 ns. Following the dynamics (CHARMm22, 300 K, at constant temperature and volume) the energy was minimized to finish the process (28). Model figures were drawn with Molscript (34) and Raster3d (35).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
rCRALBP Binds DK-11-cis-Retinal in the Ligand Binding Pocket—To determine whether rCRALBP binds the photoaffinity reagent in a similar manner as endogenous ligands, the purified apoprotein was incubated with either DK-11-cis-retinal or 11-cis-retinal, excess unbound retinoid removed and UV-visible spectral analyses performed. Spectral analyses of the holoprotein complexes are shown in Fig. 2, along with the molecular structures of the two ligands and absorbance spectra of the free retinoids. DK-11-cis-retinal exhibits a bathochromic absorbance shift from 380 to ~408 nm when complexed with human rCRALBP, yielding a chromophore maxima between that of the holoprotein with bound 11-cis-retinal ({lambda}max = 425 nm) or 9-cis-retinal ({lambda}max = 400 nm). Upon bleaching of CRALBP bound DK-11-cis-retinal with white light, the absorption maxima of the photoaffinity reagent shifts to ~380 nm because of the production of free, all-trans-retinoid. These results support a specific interaction between DK-11-cis-retinal and the rCRALBP ligand binding pocket.



View larger version (33K):
[in this window]
[in a new window]
  		FIG. 2.
rCRALBP binds photoaffinity reagent DK-11-cis-retinal. A, the structures and UV-visible spectra of 11-cis-retinal and 3-diazo-4-keto-11-cis-retinal (DK-11-cis-retinal) in ethanol are shown. B, UV-visible spectra are shown for rCRALBP complexed with DK-11-cis-retinal and 11-cis-retinal in 50 mM Tris-HCl, pH 7.2, 1 mM dithiothreitol-EDTA. rCRALBP exhibits a chromophore maximum at 408 nm with bound DK-11-cis-retinal and at 425 nm with bound 11-cis-retinal. Light isomerizes the 11-cis bond to the all-trans configuration, changing the conformation of the binding pocket and shifting the absorbance maxima to ~380 nm because of the production of unbound ligand.

 
Photoaffinity Labeling of rCRALBP—To covalently label rCRALBP, the DK-11-cis-retinal-rCRALBP complex was irradiated with UV light and bound retinal was reduced with tritiated sodium borohydride to provide a radioactive marker at the sites of modification. Preliminary experiments (Supplemental Fig. 1) using 20 min of photolysis as described for photolabeling rhodopsin (11) yielded about 4 mol% 3H incorporation in rCRALBP and significant nonspecific protein modification (data not shown). Photolysis times of 5–40 s yielded a relatively constant ~1 mol% 3H incorporation (Supplemental Fig. 1) therefore 5 s photolysis was used for preparative procedures to minimize nonspecific labeling. Notably, UV-visible spectra of the preparation following 5 s photolysis at 254 nm but prior to reduction with NaB[3H]4 were essentially the same as after bleaching with white light for 20 min (Fig. 2B), suggesting that photolysis isomerized most of the ligand to intact all-trans-retinoid. RP-HPLC analysis of [3H]DK-rCRALBP (Fig. 3) demonstrated that 3H radioactivity coeluted with rCRALBP, supporting covalent modification of the protein. This result also indicated that all unbound retinoid and radio-activity was removed by hexane extraction and exhaustive dialysis during sample preparation. The major mass observed by LC ESMS for [3H]DK-rCRALBP (Mobs = 39,114 ± 2) was of the unmodified protein (Mcalc = 39,110) consistent with the low photolabeling efficiency. SDS-PAGE and amino acid analyses of [3H]DK-rCRALBP showed no evidence of protein fragmentation (results not shown).



View larger version (20K):
[in this window]
[in a new window]
  		FIG. 3.
RP-HPLC of intact 3H-photoaffinity-labeled rCRALBP. RP-HPLC absorbance and radioactivity profiles are shown for rCRALBP (5 µg, 128 pmol) with bound DK-11-cis-retinal following photolysis and reduction with NaB[3H]4. Chromatography conditions: 5-µ Vydac C18 column (300 Å, 1 x 150 mm), aqueous trifluoroacetic acid/acetonitrile solvents, gradient elution as indicated, 100 µl/min, and 1-min fractions. These results support covalent photoaffinity labeling of rCRALBP.

 
Identification of Photoaffinity Modification Sites in rCRALBP—To identify photoaffinity modified amino acid residues, [3H]DK-rCRALBP (5 nmol) was digested with trypsin, peptides fractionated by RP-HPLC (Fig. 4), and all chromatography fractions were analyzed by LC MS/MS. Ten peptides were detected in the four radioactive pools and were found to contain a total of eight modified residues, each with a different adduct mass addition (summarized in Table I and shown by individual mass spectra in Supplemental Fig. 2). Mass spectrometric analyses of all other chromatography fractions accounted for the entire sequence of rCRALBP (Supplemental Table I), and no other modified peptides were identified.



View larger version (30K):
[in this window]
[in a new window]
  		FIG. 4.
RP-HPLC of tryptic peptides from 3H-photoaffinity-labeled rCRALBP. RP-HPLC absorbance and radioactivity profiles are shown for 3H photoaffinity-labeled-rCRALBP (195 µg, 5 nmol) following tryptic digestion. Chromatography conditions: 5-µ Vydac C18 column (2.1 x 150 mm), aqueous trifluoroacetic acid/acetonitrile solvents, gradient elution as indicated, 100 µl/min, and 1-min fractions. Recovery of radioactivity in fractions A–D was ~60% of the total amount applied.

 


View this table:
[in this window]
[in a new window]
  		TABLE I
Photoaffinity-modified amino acids in rCRALBP The above eight photoaffinity-modified residues were identified by LC MS/MS analysis. Peptides accounting for 100% of the rCRALBP sequence were found in the chromatography fractions (Fig. 4) with modified residues only found in radioactive fractions. Four residues have been independently associated with the ligand binding pocket using other methods.

 
Identification of the rCRALBP Ligand Binding Pocket by H/D Exchange—The exposure of rCRALBP to deuterium resulted in a time-dependent increase in the mass of both apo-rCRALBP and rCRALBP complexed with 11-cis-retinal (Supplemental Fig. 3). After 20 s in D2O, the number of deuterium incorporated in apo-rCRALBP increased to 108 and in holo-rCRALBP to 121. After 60 s in D2O, no difference in H/D exchange between apo and holo-rCRALBP was detectable and the protein masses did not increase further (Supplemental Fig. 3). Therefore, a D2O exposure time of 20 s was chosen as appropriate for topological analysis. The reproducibility of the methodology was evaluated by LC ESMS analysis of five different intact apo-rCRALBP samples (100 pmol each) following 20 s in D2O, and the average deuterium incorporation determined to be 106.6 ± 1% relative S.D. (data not shown).

Unambiguous identification of peptic peptides covering most (92%) of the CRALBP sequence was established by tandem mass spectrometry prior to H/D exchange (Supplemental Table II). Also prior to topological analyses, the reproducibility of deuterium incorporation in six different peptides was evaluated by LC ESMS analysis of five separate preparations of apo-rCRALBP (100 pmol each) following 20 s in D2O and peptic digestion. In these analyses, the average deuterium incorporation per peptide exhibited a mean relative S.D. of ~2.6% (data not shown). Comparison of deuterium incorporation in specific peptides of the rCRALBP with and without bound ligand is summarized in Fig. 5 and itemized in detail in Supplemental Table II. The central portion of rCRALBP (residues 111–197) incorporated relatively low levels of deuterium (~30–47%) and exhibited little difference in H/D exchange between the apo- and holo-protein. The N and C termini of both structures exhibited moderate to high levels of deuterium exchange (~54–82%), suggesting that these regions are solvent exposed in both apo- and holo-rCRALBP. Ligand-dependent incorporation differences were found associated with N-terminal residues 5–42 and C-terminal residues 282–316, which incorporated significantly more deuterium in the holoprotein (Fig. 5) and therefore are more solvent-exposed in the holoprotein. Residues 41–71, 80–94, 127–137, and 262–275 also incorporated more deuterium in holo-rCRALBP. In contrast, residues 198–255 incorporated significantly less deuterium in holo-rCRALBP, indicating that this region is less accessible to aqueous solvent when the ligand binding pocket is occupied. Notably CRALBP residues 197–255 contain 10 of 12 amino acids proposed to interact with ligand (2, 9, 10), including 7 of 8 photolabeled residues (Table I).



View larger version (45K):
[in this window]
[in a new window]
  		FIG. 5.
Comparison of hydrogen/deuterium exchange in apo- and holo-rCRALBP. A, comparison of deuterium incorporation per peptide is shown in apo-rCRALBP (striped bars) and holo-CRALBP (solid bars). B, the difference in deuterium incorporation between holo- and apo-rCRALBP per peptide (i.e. [D%]holo - [D%]apo) is shown. The hydrophobic region encompassing residues 197–255 contains all but one of the photolableled residues and is less solvent accessible with bound 11-cis-retinal.

 
Structural Model of the CRALBP Ligand Binding Cavity—A stereo diagram of the structural model determined for CRALBP with bound ligand is shown in Fig. 6. The positioning of 11-cis-retinal within the ligand cavity of the model was determined by multiple energy minimization and dynamics calculations such that the aldehyde group was inaccessible to the aqueous environment (4, 9, 32). Within the model, distances were calculated between ligand and all proposed components of the ligand cavity from previous biochemical analyses (Table II). Except for Met208, Arg233, and Trp244, the residues in Table II line the binding cavity and either the side-chain or the main chain atoms are available for reaction with the photoactivated radical. In the model, residues Met208, Arg233, and Trp244 are more distant from bound ligand and may be associated with the entry or release of retinoid from the binding pocket.



View larger version (68K):
[in this window]
[in a new window]
  		FIG. 6.
Structural model of the CRALBP ligand binding domain. Stereo diagrams of the ligand binding domain of CRALBP are shown in the ligand-bound, lid-closed conformation (see text for details). The backbone ribbon is rainbow-colored from red at the N-terminal Tyr143 to violet at the C-terminal (Ala301). The twelve residues implicated in ligand interactions by biochemical analyses are colored according to sequence number. In the current model, the ligand (11-cis retinal) is colored red, with the ionone ring toward the bottom of the figure and the aldehyde group toward the top, near Trp165. Limited proteolysis studies support the existence of a separate N-terminal domain in the intact protein (42). The interface between the N-terminal and ligand binding domains of CRALBP remains to be determined. A, viewed parallel to the {beta}-sheet. B, viewed perpendicular to the {beta}-sheet, a 90° rotation from that of A.

 


View this table:
[in this window]
[in a new window]
  		TABLE II
Distance of 11-cis-retinal from selected residues in the CRALBP model of the ligand binding domain Dave is the average distance of all atoms of the residue from the retinal. Dmin is the minimum distance, and Dmax is the maximum distance. All distances are in Angstrom units. S.D. is the standard deviation of the distance.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
The visual cycle is the enzymatic retinoid-processing pathway involved in regenerating bleached rod and cone visual pigments (for reviews see Refs. 1, 6, 36–39). In mammalian rods, photoisomerization converts rhodopsin-bound 11-cis-retinal to all-trans-retinal, which is reduced to all-trans-retinol within the photoreceptor outer segment. all-trans-Retinol leaves the rod photoreceptor cells and enters the adjacent RPE, where it is esterified by lecithin:retinol acyltransferase, converted to 11-cis-retinol by an isomerohydrolase, and oxidized to 11-cis-retinal by one or more short chain dehydrogenases. 11-cis-Retinal then leaves the RPE and re-enters the photoreceptor cell where it regenerates rhodopsin, completing the rod visual cycle. Less is known about cone visual pigment regeneration, but the process is different than for rhodopsin and may involve 11-cis-retinoid synthesis in Müller cells, which also express CRALBP (40). In the RPE, CRALBP functions as an 11-cis-retinol acceptor for the isomerohydrolase and as a substrate carrier for 11-cis-retinol dehydrogenase (2–6). These functions require the rapid association and release of retinoid from the CRALBP ligand binding pocket. Previous studies have demonstrated high CRALBP binding affinity for 11-cis-retinal (Kd ≤ 21 nM) and that the protein preferentially binds 11-cis-retinal over 11-cis-retinol (2, 9, 41). Toward a better understanding of CRALBP function, we are characterizing the structure of the retinoid binding pocket.

Structure-function studies first localized the CRALBP retinoid binding domain to C-terminal residues 120–313 by limited proteolysis (42). In subsequent work, the lack of covalent interactions between rCRALBP and retinoid was demonstrated and residues Gln210 and Lys221 associated with retinoid binding by a combination of protein chemical modifications, site-directed mutagenesis, and UV-visible and fluorescence spectroscopy (9). More recently using similar methods combined with heteronuclear single quantum correlation NMR and enzymatic assays with purified recombinant 11-cis-retinol dehydrogenase, rCRALBP residues Trp165, Met208, Met222, Met225, and Trp244 were found to influence retinoid binding and substrate carrier function. Most, if not all of these residues were found to undergo ligand-dependent conformational changes (10). Also, disease-causing mutation R233W was found to significantly tighten retinoid affinity and to change the rCRALBP retinoid binding pocket conformation, suggesting residue 233 as part of or in close proximity to the retinoid binding cavity (2). In the present study, the CRALBP retinoid binding pocket has been further characterized by photoaffinity labeling, high resolution topological analysis, and structural modeling.

Photolabeling of the CRALBP Ligand Binding Pocket— CRALBP exhibits highly stereoselective ligand interactions in 1:1 stoichiometry with 11-cis-retinol, 11-cis-retinal, and 9-cis-retinal, resulting in a red shift in ligand absorbance relative to the unbound ligand (41). When complexed with CRALBP, the photoaffinity reagent DK-11-cis-retinal also exhibits a bathochromic absorbance shift, supporting a specific interaction with the ligand binding pocket. The chromophoric maximum of human rCRALBP-bound DK-11-cis-retinal (~408 nm) approximates that of bovine CRALBP complexed with 9-cis-retinal (405 nm) (41). Photolabeling of rCRALBP with DK-11-cis-retinal yielded covalent incorporation (1–4 mol%) consistent with the estimated labeling efficiency of rhodopsin with this analogue (0.7–1.5 mol%) (12). To help stabilize the protein conformation, photolysis of rCRALBP was performed near -196 °C in 67% glycerol, but in contrast to the 20 min rhodopsin photolabeling procedure (11, 12), photolysis was limited to 5 s to achieve specific labeling. Even with only 5 s of photolysis, most of the DK-11-cis-retinal ligand appears to be isomerized to all-trans-retinoid for which rCRALBP has no affinity (41). About 1 mol% incorporation of photo-label was obtained in 5 nmol of rCRALBP (~195 µg) and a total of eight photoaffinity-modified rCRALBP residues were identified by mass spectrometry. The specificity of the labeling is supported by independent analyses (9, 10) associating ligand interactions with four of the photo-adducted amino acids, namely Met208, Lys221, Met222, and Met225. Photolabeled residues Tyr179, Phe197, Cys198, and Val223 are proposed as newly identified components of the retinoid binding pocket and supported further by topological analyses and a structural model.

An unexpected outcome of photoaffinity labeling with DK-11-cis-retinal was that each of the modified residues in rCRALBP exhibited different adduct masses, and more than one mass addition per residue was detected for Met222 and Val223. Photoaffinity labeling with diazo reagents involves photogeneration of a carbene group, which inserts into nearby polypeptide. A mass addition of 296 was expected after removal of the diazo group and covalent attachment of the otherwise intact retinoid analogue, however this adduct mass was not observed. Diazo reagents are highly labile as are retinoid reagents and the chemistry of the present photolabeling remains unclear. We speculate that upon UV-irradiation of rCRALBP bound DK-11-cis-retinal, the resulting carbene radical moves freely throughout the highly conjugated polyene structure of the retinoid analogue. We suspect fragmentation of the photolabel occurs at the time of attachment to the protein and assume that prior to insertion, the photoisomerized, intact ligand begins to diffuse out of the binding pocket. Alternatively, the modifications in DK-rCRALBP may occur in a manner similar to photolabeling proteins directly with [3H]retinoic acid (43, 44), another poorly understood mechanism. Hypothetical structures for four of the ten observed photo-adducts in rCRALBP are presented in Supplemental Fig. 4; however, the identity and molecular origin of all the modifications remain unknown. The sites modified in rhodopsin with radioactive DK-11-cis-retinal and in the ligand binding pocket of cellular retinoic acid-binding protein with all-trans-retinoic acid (44) were identified by Edman degradation. Their molecular identities have yet to be determined (11, 12). Despite these uncertainties, the photoaffinity modifications in rCRALBP are localized within the retinoid binding cavity or at the entrance/exit of the pocket as shown by structural modeling.

The rCRALBP Ligand Binding Cavity Revealed by H/D Exchange—We also used H/D exchange detected by mass spectrometry to measure the effect of protein-bound 11-cis-retinal on the solvent accessibility of amide hydrogens (27). Comparison of H/D exchange in intact apo- and holo-rCRALBP revealed about 10% more deuterium incorporation in the holoprotein, suggesting a more open, solvent exposed conformation for the protein with bound ligand. Comparison of H/D exchange in rCRALBP peptides indicated that the N and C termini are solvent exposed in both the apo- and holoproteins but more so with bound ligand. N-terminal residues 5–42, C-terminal residues 282–316, and residues 41–71, 80–94, 122–137, and 262–275 all incorporated more deuterium in holo-rCRALBP and therefore are more solvent-exposed in the holoprotein. 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 holoprotein structures, perhaps forming the core of the folded protein. The only region of rCRALBP that incorporated significantly more deuterium in the absence of bound 11-cis-retinal encompasses amino acids 198–255. This region contains most of the residues currently associated with ligand interactions by photoaffinity labeling and other biochemical analyses. These results are consistent with an earlier topological analyses using antibodies and protease that supports holo-rCRALBP residues 1–30, 100–124, and 257–285 as being exposed and residues 30–99 and 176–229 as being inaccessible or buried (45). The exposed regions of CRALBP constitute potential functional domains for interaction with other proteins and indeed the docking site for EBP50 has recently been shown to reside in the C-terminal four residues (8).

Structural Modeling of the CRALBP Retinoid-binding Domain—The homology between CRALBP and other CRAL-TRIO family members ({alpha}TTP, SPF, Sec14) spans ~185 amino acids and includes their respective ligand-binding pockets (13). These CRALBP homologues are all cytosolic lipid-binding proteins and high resolution structures for all three have been determined by x-ray crystallography. Sec14 catalyzes exchange of phosphotidylinositol and phosphatidylcholine between membrane bilayers and is essential for yeast Golgi secretory function (16). {alpha}TTP mediates the stereo-selective transfer of {alpha}-to-copherol into nascent VLDL in hepatocytes and plays a major role in maintaining human plasma {alpha}-tocopherol levels (14, 15). SPF appears to be involved in cholesterol biosynthesis and squalene epoxidation in the liver but its natural ligand(s) and physiological role are unknown (17). The crystal structures for {alpha}TTP were determined with and without {alpha}-tocopherol in the binding pocket, yielding similar structures except for a solvent-exposed "lid" or "mobile lipid-exchange loop" (residues 198–221) that limits access to the ligand binding cavity (14). Large conformational changes are required for entry or exit of ligand from the {alpha}TTP binding pocket (15), consistent with the ligand-dependent conformational changes observed in CRALBP by NMR (10, 46). SPF was crystallized with an unidentified ligand in the binding pocket and the determined structure resembles the liganded conformation of {alpha}TTP with the lid in the closed position (17). The structure of Sec14 was determined with {beta}-octylglucoside in the binding cavity and resembles the unliganded conformation of {alpha}TTP, with the lid open and part of the hydrophobic binding pocket exposed.

CRALBP retinoid binding pocket components are located within a highly homologous region of sequence containing the ligand binding cavities of other CRAL-TRIO family members (Fig. 7). Notably, five of the CRALBP residues we have associated with the retinoid binding pocket (Trp165, Tyr179, Phe197, Met222, and Met225) align directly with components identified in the ligand cavities of the CRAL-TRIO crystal structures and are very close to the ligand in our structural model (average distance ~4.4 Å). Four other residues biochemically associated with CRALBP ligand interactions (Cys198, Gln210, Lys221, and Val223) align in relatively close proximity to ligand cavity components in the crystal structures (average model distance from ligand ~7.6 Å). However, CRALBP residues Met208, Arg233, and Trp244 also have been biochemically implicated in ligand interactions (2, 10) but are more distant from ligand in the model (average model distance from ligand ~14.3 Å). Notably, the lipid exchange loop or lid region in the {alpha}TTP and SPF structures is strongly conserved in CRALBP residues 241–264, which contains CRALBP residue Trp244. In the CRALBP ligand-containing model, Trp244 extends into the solvent and makes substantial contact with Tyr245 (3.4 Å), which in turn contacts Met208. In the open structure of Sec14, the residues equivalent to Met208 and Trp244 (Ile184 and Phe219) are widely separated due to the opening of the binding cavity by movement of the helices. The central, green helix in Fig. 6B moves to the left, and the blue vertical helix moves to the right in the Sec14 open structure. From comparison of the crystallographic structures with the CRALBP model, we propose that Met208 and Trp244 are not in intimate contact with ligand but rather are located at the entrance/exit to the ligand cavity and involved in the conformational changes necessary for ligand binding and release. Accordingly, the observed photolabeling of Met208 probably occurred as the isomerized ligand was exiting the binding cavity.



View larger version (31K):
[in this window]
[in a new window]
  		FIG. 7.
Ligand binding cavity sequences in CRAL-TRIO proteins. Residues shown on black background in {alpha}TTP (14, 15), SPF (17), and Sec14 (16) have been associated with ligand interactions in their respective crystal structures. Residues on black in the CRALBP sequence have been associated with ligand interactions based on photoaffinity labeling, H/D exchange, and other biochemical measurements (2, 9, 10). The lid or mobile lipid exchange loop first recognized in {alpha}TTP (14) is underlined, as are the homologous domains in SPF (17) and CRALBP.

 
Residue Arg233 in CRALBP corresponds to {alpha}TTP residue Lys190 (Fig. 7), which is in a positively charged surface region that has been proposed as a protein interaction site possibly involved in ligand release (15). Using {alpha}TTP as a template, it is possible to predict that R233 is intimately involved in contacts with the conserved RAR sequence (residues 100–102) of the N-terminal domain of CRALBP. Additionally, a positively charged groove of residues can be identified in the CRALBP model involving residues Arg150, Lys152, Arg155, Arg233, Lys235, His238, Lys258, Lys260, Arg264, and possibly His268. Topological analysis support the solvent accessibility of most of these residues, namely, residues 147–157 and 262–275 incorporated more deuterium in holo-rCRALBP and residues 257–285 are more antibody accessible in holo-rCRALBP (45). Furthermore, in the ligand binding domain model, the Arg233 side chain is close to residues Gln193, Ile194, and Asn195. Disease-causing CRALBP mutation R233W appears to disrupt the conformation of the positive surface cluster of residues and may also result in stronger interactions between Trp233 and uncharged residues 193–195, which perhaps contributes to the tighter affinity for ligand exhibited by the R233W mutant (2). Disease-causing CRALBP mutation R150Q also likely disrupts the structure of this critically charged surface region, resulting in an insoluble protein (47).

In summary, we have identified eight photoaffinity modified residues in CRALBP, four of which were previously proposed as components of the retinoid binding pocket (Met208, Lys221, Met222, and Met225) and four of which appear to represent newly identified components (Tyr179, Phe197, Cys198, and Val223). This expands to 12 the number of residues proposed as being involved in CRALBP ligand interactions. Topological analysis by H/D exchange and mass spectrometry localized the CRALBP ligand binding cavity to residues 198–255 and identified solvent-exposed regions of holo-rCRALBP that may contain interaction sites with other visual cycle proteins. Structural modeling corroborated the ligand cavity location of 7 of 8 photolabeled residues and two other previously proposed components. Modeling also implicated photolabeled Met208 and Trp244 as being located at the entrance/exit of the ligand cavity and Arg233 to be in another exposed region possibly associated with the mechanism of ligand release. Perhaps 11-cis-retinol dehydrogenase facilitates 11-cis-retinal release by binding in the Arg233 region since the enzyme is known to interact directly with CRALBP (2, 4) and has been detected in the RPE plasma membrane (48). How retinoid is released from the high affinity CRALBP binding pocket remains an unresolved but important issue for understanding how 11-cis-retinal is exported from the RPE for visual pigment regeneration. These molecular details of visual cycle interactions require further investigation and determination of the three-dimensional structure of CRALBP.


   FOOTNOTES
 
* This study was supported in part by National Institutes of Health Grants EY6603, EY14239, GM63020, a Research Center Grant from The Foundation Fighting Blindness, and funds from the Cleveland Clinic Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

The on-line version of this article (available at http://www.jbc.org) contains Supplemental Figs. 1–4 and Supplemental Tables I and II. Back

This work was submitted in partial fulfillment of the requirements for a doctoral degree in chemistry from Cleveland State University. Back

|| Present address: University of Florida, General Clinical Research Center, Gainesville, FL 32610-0322. Back

§§ To whom correspondence should be addressed: Cole Eye Institute (i31), Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, Ohio 44195. Tel.: 216-445-0425; Fax: 216-445-3670; E-mail: crabbj{at}ccf.org.

1 The abbreviations used are: CRALBP, cellular retinaldehyde-binding protein; DK-11-cis-retinal, 3-diazo-4-keto-11-cis-retinal; DK-rCRALBP, rCRALBP photolabeled with DK-11-cis-retinal; H/D exchange, hydrogen/deuterium exchange; LC ESMS, liquid chromatography electrospray mass spectrometry; LC MS/MS, liquid chromatography tandem mass spectrometry; QTOF, quadrupole time of flight; RPE, retinal pigment epithelium; rCRALBP, recombinant CRALBP; RP-HPLC, reverse phase-high performance liquid chromatography. Back


   ACKNOWLEDGMENTS
 
We thank Prof. Koji Nakanishi for the photoaffinity reagent 3-diazo-keto-11-cis-retinal, Prof. Larry Sayre for insights regarding photo-adducts, Dr. Z. Zhang for the Magtran computer program for H/D exchange analyses, and Prof. Ronald E. Stenkamp for preparing the stereo figures of the structural model. We also thank Prof. John C. Saari for valuable discussions and for reviewing the manuscript prior to publication.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Saari, J. C. (1994) in The Retinoids: Biology, Chemistry, and Medicine (Sporn, M. B., Roberts, A. B., and Goodman, D. S., eds) 2nd Ed., pp. 351-385, Raven Press, New York
  2. Golovleva, I., Bhattacharya, S., Wu, Z., Shaw, N., Yang, Y., Andrabi, K., West, K. A., Burstedt, M. S., Forsman, K., Holmgren, G., Sandgren, O., Noy, N., Qin, J., and Crabb, J. W. (2003) J. Biol. Chem. 278, 12397-12402[Abstract/Free Full Text]
  3. Saari, J. C., Nawrot, M., Kennedy, B. N., Garwin, G. G., Hurley, J. B., Huang, J., Possin, D. E., and Crabb, J. W. (2001) Neuron 29, 739-748[CrossRef][Medline] [Order article via Infotrieve]
  4. Saari, J. C., Bredberg, D. L., and Noy, N. (1994) Biochemistry 33, 3106-3112[CrossRef][Medline] [Order article via Infotrieve]
  5. Stecher, H., Gelb, M. H., Saari, J. C., and Palczewski, K. (1999) J. Biol. Chem. 274, 8577-8585[Abstract/Free Full Text]
  6. Winston, A., and Rando, R. R. (1998) Biochemistry 37, 2044-2050[CrossRef][Medline] [Order article via Infotrieve]
  7. Bhattacharya, S. K., Wu, Z., Jin, Z., Yan, L., Miyagi, M., West, K., Nawrot, M., Saari, J. C., and Crabb, J. W. (2002) FASEB J. 16, A14 (Abstr. 16.12)
  8. Nawrot, M., West, K., Huang, J., Possin, D. E., Bretscher, A., Crabb, J. W., and Saari, J. C. (2004) Investig. Ophthalmol. Vis. Sci. 45, 393-401[Abstract/Free Full Text]
  9. Crabb, J. W., Nie, Z., Chen, Y., Hulmes, J. D., West, K. A., Kapron, J. T., Ruuska, S. E., Noy, N., and Saari, J. C. (1998) J. Biol. Chem. 273, 20712-20720[Abstract/Free Full Text]
  10. Wu, Z., Yang, Y., Shaw, N., Bhattacharya, S., Yan, L., West, K., Roth, K., Noy, N., Qin, J., and Crabb, J. W. (2003) J. Biol. Chem. 278, 12390-12396[Abstract/Free Full Text]
  11. Borhan, B., Souto, M. L., Imai, H., Shichida, Y., and Nakanishi, K. (2000) Science 288, 2209-2212[Abstract/Free Full Text]
  12. Souto, M. L., Um, J., Borhan, B., and Nakanishi, K. (2000) Helvetical Chimica Acta 83, 2617-2628[CrossRef]
  13. Panagabko, C., Morley, S., Hernandez, M., Cassolato, P., Gordon, H., Parsons, R., Manor, D., and Atkinson, J. (2003) Biochemistry 42, 6467-6474[CrossRef][Medline] [Order article via Infotrieve]
  14. Meier, R., Tomizaki, T., Schulze-Briese, C., Baumann, U., and Stocker, A. (2003) J. Mol. Biol. 331, 725-734[CrossRef][Medline] [Order article via Infotrieve]
  15. Min, K. C., Kovall, R. A., and Hendrickson, W. A. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 14713-14718[Abstract/Free Full Text]
  16. Sha, B., Phillips, S. E., Bankaitis, V. A., and Luo, M. (1998) Nature 391, 506-510[CrossRef][Medline] [Order article via Infotrieve]
  17. Stocker, A., and Baumann, U. (2003) J. Mol. Biol. 332, 759-765[CrossRef][Medline] [Order article via Infotrieve]
  18. Crabb, J. W., Carlson, A., Chen, Y., Goldflam, S., Intres, R., West, K. A., Hulmes, J. D., Kapron, J. T., Luck, L. A., Horwitz, J., and Bok, D. (1998) Protein Sci. 7, 746-757[Abstract]
  19. Crabb, J. W., West, K. A., Dodson, W. S., and Hulmes, J. D. (1997) in Current Protocols in Protein Science, Unit 11.9, Supplement 7 (Coligan, J. E., Ploegh, H. L., Smith, J. A., and Speicher, D. W., eds) pp. 11.09.01-11.09.42, John Wiley & Sons, Inc., New York
  20. Laemmli, U. (1970) Nature 227, 680-685[CrossRef][Medline] [Order article via Infotrieve]
  21. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254[CrossRef][Medline] [Order article via Infotrieve]
  22. Garwin, G. G., and Saari, J. C. (2000) Methods Enzymol. 316, 313-324[CrossRef][Medline] [Order article via Infotrieve]
  23. Crabb, J. W., Miyagi, M., Gu, X., Shadrach, K., West, K. A., Sakaguchi, H., Kamei, M., Hasan, A., Yan, L., Rayborn, M. E., Salomon, R. G., and Hollyfield, J. G. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 14682-14687[Abstract/Free Full Text]
  24. Hasan, A., Smith, D. L, and Smith, J. B. (2002) Biochemistry 41, 15876-15882[Medline] [Order article via Infotrieve]
  25. Wang, L., Pan, H., and Smith D. L. (2002) Mol. Cell. Proteomics 1, 132-138[Abstract/Free Full Text]
  26. Miyagi, M., Sakaguchi, H., Darrow, R. M., Yan, L., West, K. A., Aulak, K. S., Stuehr, D. J., Hollyfield, J. G., Organisciak, D. T., and Crabb, J. W. (2002) Mol. Cell Proteomics 1, 293-303[Abstract/Free Full Text]
  27. Zhang, Z., and Smith, D. L. (1993) Protein Sci. 2, 522-531[Abstract]
  28. Behnke, C. A., Yee, V. C., Le Trong, I., Pedersen, L. C., Stenkamp, R. E., Kim, S.-S., Reeck, G. R., and Teller, D. C. (1998) Biochemistry 37, 15277-15288[CrossRef][Medline] [Order article via Infotrieve]
  29. CCP4 (1994) Acta Crystallogr. Sect. D 50, 760-763[CrossRef][Medline] [Order article via Infotrieve]
  30. McRee, D. (1999) J. Struct. Biol. 125, 156-165[CrossRef][Medline] [Order article via Infotrieve]
  31. Brunger, A., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunsatleve, R. W., Jiang, J.-S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren G. L. (1998) Acta Crystallogr. Sect. D 54, 905-921[CrossRef][Medline] [Order article via Infotrieve]
  32. Saari, J. C., and Bredberg, L. (1982) Biochim Biophys Acta 716, 266-272[Medline] [Order article via Infotrieve]
  33. MOE (2003), 2003.02 Ed., Chemical Computing Group Inc., Montreal H3A 2R7, Canada
  34. Kraulis, P. (1991) J. Appl. Crystallogr. 24, 946-950[CrossRef]
  35. Merritt, E., and Bacon, D. J. (1997) Methods Enzymol. 277, 505-524[Medline] [Order article via Infotrieve]
  36. Crouch, R. K., Chader, G. J., Wiggert, B., and Pepperberg, D. R. (1996) Photochem. Photobiol. 64, 613-621[Medline] [Order article via Infotrieve]
  37. Saari, J. C. (2000) Invest. Ophthal. Visual Sci. 41, 337-348[Free Full Text]
  38. McBee, J. K., Palczewski, K., Baehr, W., and Pepperberg, D. R. (2001) Prog. Retin Eye Res. 20, 469-529[CrossRef][Medline] [Order article via Infotrieve]
  39. Rando, R. R. (2001) Chem. Rev. 101, 1881-1896[CrossRef][Medline] [Order article via Infotrieve]
  40. Mata, N. L., Radu, R. A., Clemmons, R. C., and Travis, G. H. (2002) Neuron 36, 69-80[CrossRef][Medline] [Order article via Infotrieve]
  41. Saari, J. C., and Bredberg, D. L. (1987) J. Biol. Chem. 262, 7618-7622[Abstract/Free Full Text]
  42. Chen, Y., Johnson, C., West, K., Goldflam, S., Bean, M. F., Huddleston, M. J., Carr, S. A., Gabriel, J. L., and Crabb, J. W. (1994) in Techniques in Protein Chemistry V (Crabb, J. W., ed) pp. 371-378, Academic Press, San Diego
  43. Bernstein, P., Choi, S., Ho, Y.-C., and Rando, R. R. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 654-658[Abstract/Free Full Text]
  44. Chen, G., and Radominska-Pandya, A. (2000) Biochemistry 39, 12568-12574[CrossRef][Medline] [Order article via Infotrieve]
  45. Crabb, J. W., Gaur, V. P., Garwin, G. G., Marx, S. V., Chapline, C., Johnson, C. M., and Saari, J. C. (1991) J. Biol. Chem. 266, 16674-16683[Abstract/Free Full Text]
  46. Luck, L. A., Barrows, S. A., Venters, R. A., Kapron, J. T., Roth, K. E., Paradis, S. G., and Crabb, J. W. (1997) in Techniques in Protein Chemistry VIII (Marshak, D. R., ed) pp. 439-448, Academic Press, San Diego
  47. Maw, M. A., Kennedy, B., Knight, A., Bridges, R., Roth, K. E., Mani, E. J., Mukkadan, J. K., Nancarrow, D., Crabb, J. W., and Denton, M. J. (1997) Nat. Genet. 17, 198-200[CrossRef][Medline] [Order article via Infotrieve]
  48. Mata, N. L., Villazanna, E. T., and Tsin, A. T. C. (1998) Invest. Ophthal. Visual. Sci. 39, 1312-1319[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us