Advertisement
JBC

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


     

Originally published In Press as doi:10.1074/jbc.M410466200 on October 15, 2004

J. Biol. Chem., Vol. 280, Issue 1, 484-493, January 7, 2005
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/1/484    most recent
M410466200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
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 arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Breustedt, D. A.
Right arrow Articles by Skerra, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Breustedt, D. A.
Right arrow Articles by Skerra, A.
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?

The 1.8-Å Crystal Structure of Human Tear Lipocalin Reveals an Extended Branched Cavity with Capacity for Multiple Ligands*

Daniel A. Breustedt{ddagger}, Ingo P. Korndörfer{ddagger}, Bernhard Redl§, and Arne Skerra{ddagger}||

From the {ddagger}Lehrstuhl für Biologische Chemie, Technische Universität München, D-85350 Freising-Weihenstephan, Germany and §Institut für Molekularbiologie, Medizinische Universität Innsbruck, A-6020 Innsbruck, Austria

Received for publication, September 13, 2004 , and in revised form, October 5, 2004.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
In contrast with earlier assumptions, which classified human tear lipocalin (Tlc) as an outlier member of the lipocalin protein family, the 1.8-Å resolution crystal structure of the recombinant apoprotein confirms the typical eight-stranded antiparallel {beta}-barrel architecture with an {alpha}-helix attached to it. The fold of Tlc most closely resembles the bovine dander allergen Bos d 2, a well characterized prototypic lipocalin, but also reveals similarity with {beta}-lactoglobulin. However, compared with other lipocalin structures Tlc exhibits an extremely wide ligand pocket, whose entrance is formed by four partially disordered loops. The cavity deeply extends into the {beta}-barrel structure, where it ends in two distinct lobes. This unusual structural feature explains the known promiscuity of Tlc for various ligands, with chemical structures ranging from lipids and retinoids to the macrocyclic antibiotic rifampin and even to microbial siderophores. Notably, earlier findings of biological activity as a thiol protease inhibitor have no correspondence in the three-dimensional structure of Tlc, rather it appears that its proteolytic fragments could be responsible for this phenomenon. Hence, the present structural analysis sheds new light on the ligand binding activity of this functionally obscure but abundant human lipocalin.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Tear lipocalin (Tlc,1 also called Lcn1 or von Ebner's gland protein) was originally described as a unique major component of human tear fluid, comprising 15-33% of its protein content and secreted by the lacrimal glands (1, 2). Later on it was also found to be produced by a number of other secretory glands and tissues, including lingual glands, nasal mucosal glands, secretory glands of the tracheobronchial tract, sweat glands, mammary gland, adrenal gland, prostate, thymus, testis, and by corticotrophs of the pituitary gland (3-6).

cDNA cloning and sequencing revealed that this protein component is a member of the lipocalin superfamily with significant amino acid similarity to {beta}-lactoglobulin (7), one of the classical lipocalins. Because it lacks one of the three conserved amino acid motifs that are normally present in lipocalins, Tlc was initially classified as an outlier member of this protein family (8). However, a typical {beta}-barrel fold was predicted from a hydropathy profile of the amino acid sequence (7), circular dichroism analysis (9), and also from an extended site-directed tryptophan substitution and fluorescence study (10), revealing {beta}-sheet and {alpha}-helical composition similar to other lipocalins.

Nevertheless, Tlc possesses several biochemical activities that must be conferred by unique structural properties. As an ordinary feature of a lipocalin protein, it binds hydrophobic compounds; yet it exhibits an unusually broad ligand specificity, as compared with other members of this family. A large number of lipophilic substances belonging to different chemical classes, including fatty acids, fatty alcohols, phospholipids, glycolipids, cholesterol, retinol, arachidonic acid, as well as lipid peroxidation products, such as isoprostanes, were found as endogenous ligands of the protein isolated from tear fluid or from human NT2 precursor cells (5, 11, 12). Contrasting with other lipocalins, the ligand affinity correlates with the length of the hydrocarbon chain both for alkyl amides and fatty acids. Thus, Tlc binds most strongly the least soluble lipids (13).

By transporting or scavenging lipoid compounds, Tlc may function in the integrity of the tear film with its peculiar aqueous-lipid interface and, more generally, in the protection or repair of epithelial surfaces (5). Recently, Tlc was also shown to bind bacterial and fungal iron siderophore complexes, thus suggesting antimicrobial function (14). Furthermore, binding activity for the antibiotic drug rifampin was reported (15). Finally, Tlc accepts several synthetic ligands, such as 11-(((5-(dimethylamino)-1-naphthalenyl)sulfonyl)amino)undecanoic acid (13) or the spin label compound 16-doxylstearic acid (10).

The structural influence of the three Cys residues in Tlc has been investigated (9). Mass determination of tryptic fragments revealed that a single intramolecular disulfide bond joins Cys-61 with Cys-153 and that Cys-101 resides as a free cysteine (numbers corresponding to the mature polypeptide sequence after processing of the 18-residue signal peptide). The disulfide bridge is conserved in most lipocalins and connects the {beta}-barrel, usually strand D, with the C-terminal peptide segment (16). Reduction of this disulfide bridge in Tlc did not significantly affect protein secondary structure but had an influence on the ligand binding behavior (9).

Apart from its ligand binding properties, Tlc was shown in vitro to inhibit cysteine proteinases (17) and to act as a nonspecific endonuclease (18), both activities that are unusual among lipocalins. So far, no high resolution structural information was available for this lipocalin. In order to understand the molecular basis of its possibly multiple physiological functions, we expressed Tlc in Escherichia coli and crystallized the recombinant protein. Here we report the 1.8-Å crystal structure of Tlc in its apo state.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Vector Construction—The coding sequence for human tear lipocalin, comprising residues 5-157 of the mature LCN1 gene product (7), was amplified from plasmid pQE70/Lcn1 (19) via PCR according to a published procedure (20), using primers 5'-CCG CCT CAG ACG AGG AGA TTC AG-3' and 5'-GCT CCC TGG AGA GCA GGT TTC GC-3'. The unique amplification product was purified by agarose gel electrophoresis, phosphorylated with T4 polynucleotide kinase (New England Biolabs, Beverly, MA), and ligated with the expression vector pASK75-strepII (21) that had been cut with StuI and Eco47III and dephosphorylated using shrimp alkaline phosphatase (U. S. Biochemical Corp.). After transformation of E. coli XL1-Blue, the plasmid was isolated, and its composition was confirmed by restriction digest as well as double-stranded dideoxy sequencing (ABI PRISM 310 Genetic Analyzer, Applied Biosystems, Foster City, CA). On the resulting vector, which was designated pTlc2, the truncated Tlc was encoded as a fusion protein with the N-terminal OmpA signal peptide and with the C-terminal Strep-tag II of nine residues. The codon for the unpaired Cys-101 (numbering according to the mature full-length protein) was subsequently replaced by a Ser codon via site-directed mutagenesis (22) with the oligodeoxynucleotide 5'-AGC TCG CCC TCA GAG TAA AAG ATG TAG TG-3' and yielding the vector pTlc3, which was used throughout this study.

Protein Production and Purification—Recombinant Tlc was produced in the E. coli K12 strain JM83 (23) harboring pTlc3 via secretion into the bacterial periplasm. Cultures (2 liters) were grown in LB medium (24) supplemented with 100 mg/liter ampicillin at 22 °C. Gene expression was induced at a cell density of A550 = 0.5 by adding 0.2 mg/liter anhydrotetracycline (21). After further shaking for 3 h, the cells were harvested by centrifugation, resuspended in 0.5 M sucrose, 1 mM EDTA, 100 mM Tris-HCl, pH 8.0, and kept on ice for 30 min. The resulting spheroplasts were sedimented by centrifugation, and the supernatant containing the recombinant protein was recovered. The protein extract was dialyzed against 150 mM NaCl, 1 mM EDTA, 100 mM Tris-HCl, pH 8.0, and applied to a column with immobilized engineered streptavidin (25) by using the same buffer. The recombinant Tlc was competitively eluted by application of 2.5 mM D-desthiobiotin in the chromatography buffer. Elution fractions were concentrated, applied to a Superdex 75 gel filtration column (Amersham Biosciences), equilibrated with 150 mM NaCl, 1 mM EDTA, 50 mM Tris-HCl, pH 8.0, as running buffer, and eluted in a homogeneous peak corresponding to the monomeric protein. The yield was ~1.5 mg of purified protein per 1 liter of E. coli culture.

SeMet-labeled Tlc was produced in the same manner by shutting down endogenous Met biosynthesis of E. coli in the presence of the amino acid derivative (26, 27). To this end, the culture was grown in M9 minimal medium (24) supplemented with 0.4% (w/v) glucose, 1 mM MgSO4, 200 mg/liter L-Pro, 10 mg/liter thiamin, and 100 mg/liter ampicillin at 22 °C. L-Lys, L-Thr, L-Phe (100 mg/liter each), L-Leu, L-Ile, L-Val (50 mg/liter each), and L-SeMet (25 mg/liter) were added at a cell density of A550 = 0.5. After 15 min of incubation, gene expression was induced by adding 0.2 mg/liter anhydrotetracycline. The cells were harvested after further shaking at 22 °C for 14 h, and recombinant Tlc was purified as above. Gel filtration proved to be dispensable due to the high purity after Strep-tag chromatography and to the higher yield ~2.5 mg of purified recombinant protein per 1 liter of E. coli culture, probably as a result of the longer induction period. SeMet incorporation was verified by matrix-assisted laser desorption ionization time-of-flight-mass spectrometry.

Crystallization and Data Collection—Crystals of native as well as SeMet-labeled Tlc were grown using the hanging drop vapor diffusion technique by mixing 3 µl of protein solution (10 mg/ml, dialyzed against 10 mM Tris-HCl, pH 8.0) with 1 µl of precipitant solution on a siliconized glass coverslip, followed by equilibration against 0.5 ml of the precipitant solution. Several crystals were obtained at 12 °C in the presence of 60% (w/v) ethylene glycol, 200 mM zinc acetate, 100 mM Tris-HCl, pH 7.0-7.9, after about 6 weeks. Protein crystals were harvested directly from the drop using Nylon loops (Hampton Research, Laguna Niguel, CA), cryo-protected by transfer to a drop of Paratone N (Hampton Research), thereby removing excessive mother liquor, and frozen in a 100 K nitrogen stream (Oxford Cryosystems, Oxford, UK). Native diffraction data (Table I) were collected at the SLS beamline PX1 (PSI, Villigen, Switzerland). A three-wavelength anomalous data set of the SeMet derivative (Table I) was collected at the synchrotron beamline BM14 (ESRF, Grenoble, France). Data were processed with MOSFLM and SCALA (28). The space group was determined as C2 (a = 76.4 Å, b = 61.4 Å, c = 50.8 Å, {beta} = 129.7° for the native data set) with one protein molecule in the asymmetric unit.


View this table:
[in this window]
[in a new window]
  		TABLE I
Data collection statistics

All parameters are as defined in MOSFLM and SCALA.

 
Structure Determination, Model Building, and Refinement—The selenium substructure of a crystal grown at pH 7.9 was solved with SnB (29) using anomalous peak data, whereby three of the four SeMet sites that had to be expected from the primary structure were identified. Phasing was carried out with SHARP (30) by using the three SeMet sites and the anomalous data from all three wavelengths (Table I) to a resolution of 1.8 Å. Solvent flattening was subsequently performed with SOLOMON (32) by using a solvent content of 45%. A partial structural model, comprising the {beta}-barrel and the {alpha}-helix but lacking residues in the loop regions and at both polypeptide termini, was calculated with ARPwARP (33) as implemented in SHARP. Missing residues were then added in several sessions of manual rebuilding with QUANTA (34), alternating with restrained refinement cycles using REFMAC 5 (28). During this process three coordinated zinc ions and two chloride ions were identified and added to the model. Final refinement was carried out with the native data set of a crystal grown at pH 7.0. At this pH, two previously missing residues (Asn-32 and Leu-33) in loop 1 appeared in the electron density map and were built accordingly. Water molecules were added where stereochemically plausible, and the 2Fo - Fc and Fo - Fc difference Fourier maps revealed densities of more than 1.0 and 3.5 {sigma}, respectively. The backbone dihedral angles fall mostly into the favored regions of the Ramachandran plot. The only exception is residue Leu-33, which is situated C-terminal to the disordered part of loop 1 and has only weak electron density.

Analysis of the Crystal Structure—The few missing residues at the tips of loop 2 (Leu-56 and Ile-57) and loop 4 (Leu-105 and His-106) were inserted in QUANTA (34) with plausible stereochemistry (using the "regularize range" option). In order to model the stretch of seven residues in loop 1 without clear electron density (Asp-25 to Met-31), the Protein Data Bank was searched for a loop with identical length and matching conformation of the well resolved flanking segments (C-{alpha} positions 21-23 and 33-35, respectively) using INSIGHT II (Accelrys, San Diego, CA). One nicely fitting peptide segment, corresponding to residues 117-125 of actophorin (PDB accession code 1CNU [PDB] ), was found and grafted onto the Tlc stump using least squares superposition of the flanking segments. Side chains were subsequently adjusted to match the Tlc sequence and oriented in standard conformation (35). The completed model was subjected to one round of refinement using REFMAC 5 with the occupancy of all three amended segments (residues 25-31, 56 and 57, and 105 and 106) set to 0.01, resulting in an R-factor of 19.6% and Rfree of 25.4%. There were no structural clashes due to the new residues. Notably, the modeled loop 1 partly adopted a similar orientation as the corresponding loop of the lipocalin Bda, though shorter by two residues (see text).

Graphical representations were prepared using MOLSCRIPT (36), RASTER3D (37), and PyMOL (www.pymol.org). Secondary structure assignments were made with DSSP (39) and extracted with MAKEMOLS (40). For use with MOLSCRIPT, electron densities were converted from CCP4 format (28) with MINIMAGE (41). Molecular surfaces were either prepared with PyMOL or calculated with GRASP (42) and exported for use in MOLSCRIPT with a modified version of UNGRASP (37). Superpositions were made with LSQMAN (43). Ligand complexes were modeled using INSIGHT II. The crystallographic coordinates of the apo-Tlc structure (without the modeled segments) have been deposited at the Protein Data Bank (44) under accession code 1XKI [PDB] .


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Crystallization, Structure Determination, and Quality of the Final Model—Recombinant human Tlc, comprising residues 5-157 of the mature protein (7, 19), with the free Cys-101 substituted by Ser, was fused to the OmpA signal peptide and secreted into the periplasm of E. coli, where processing, folding, and efficient disulfide bond formation can take place. One-step purification from the bacterial periplasmic extract was achieved by streptavidin affinity chromatography via the Strep-tag II (25), which had been fused to the C terminus. Tlc was finally isolated as a homogeneous monomeric protein by gel filtration.

Crystallization of Tlc was achieved at pH 7.9 with ethylene glycol as precipitant in the presence of zinc acetate. The crystal structure was determined by multiple wavelength anomalous diffraction phasing after in vivo labeling with selenomethionine (SeMet). Refinement against a data set for a crystal of the native protein, grown under similar conditions at pH 7.0, resulted in an R-factor of 18.9% and an Rfree of 25.4% with good stereochemistry (Table II). One molecule of the apoprotein was found in the asymmetric unit. The quality of the electron density map at a resolution of 1.8 Å allowed us to unambiguously build 128 of the 152 residues of the recombinant Tlc (neglecting the Strep-tag II).


View this table:
[in this window]
[in a new window]
  		TABLE II
Refinement statistics

 
The N-terminal residues Ala-5 to Gln-11 and the C-terminal residues starting from Glu-151, including the Strep-tag II (residues 158-166), were disordered. The disulfide bond joining Cys-61 and Cys-153 was also not fully resolved in the electron density and hence was omitted from the model. Residues Asp-25 to Met-31, which form the long first loop connecting {beta}-strands A and B at the open end of the {beta}-barrel and partially cover the calyx in other lipocalins (e.g. bovine dander allergen Bos d 2 (45)), were largely disordered. Furthermore, small pieces of electron density were missing at the tips of loop 2 connecting strands C/D (Leu-56 and Ile-57) and of loop 4 connecting strands G/H (Leu-105 and His-106).

Adjacent residues had elevated temperature factors but could be built unambiguously. Hence, as in other lipocalins the {beta}-barrel of Tlc appears rather rigid, whereas extended loops and extraneous terminal segments of the polypeptide chain seem to be more flexible, probably also due to a lack of stabilizing crystal contacts. For purposes of illustration, the missing loop segments at the entrance to the ligand pocket were modeled with plausible stereochemistry (for details see "Experimental Procedures") and included in the Tlc structure described here.

Structure of Tear Lipocalin—Tlc exhibits the typical lipocalin fold, consisting of an eight-stranded, strictly antiparallel {beta}-barrel that forms a central cavity. The cavity opens to one end of the barrel, where four loops (loops 1-4) connect neighboring {beta}-strands (dubbed A-H) in a pairwise manner (Fig. 1). A C-terminal {alpha}-helix (residues Leu-127 to Ala-136), which is followed by an almost extended peptide segment (strand I), packs against one side of the {beta}-barrel. The N-terminal peptide segment of Tlc covers the bottom of the barrel between the loops connecting strands B/C and F/G and thereby buries Trp-17, one of the few residues conserved in all lipocalins (8).



View larger version (54K):
[in this window]
[in a new window]
  		FIG. 1.
Three-dimensional structure of human Tlc. A, ribbon diagram, side view: Tlc exhibits the typical lipocalin fold with an eight-stranded antiparallel {beta}-barrel forming a central cavity at its open (upper) end and an {alpha}-helix attached to it. The N and C termini of the polypeptide chain, the eight {beta}-strands A-H, the extraneous {beta}-strand I, and the four loops at the open end of the {beta}-barrel are labeled. Loop 1 connecting {beta}-strands A/B (Asp-25 to Met-31), loop 2 connecting {beta}-strands C/D (Leu-56 and Ile-57), and loop 4 connecting {beta}-strands G/H (Leu-105 and His-106) are partially disordered in the crystal structure; the modeled regions are depicted in gray. A short stretch with local 310-helical conformation at the tip of loop 3 connects {beta}-strands E/F. B, complexation of three Zn(II) ions (magenta), in conjunction with two chloride ions (green), by Tlc side chains (magenta). C, top view into the calyx of Tlc (the molecule shown in A rotated by ~90° about a horizontal axis): the {beta}-barrel is ellipsoid rather than circular, and loop 1 partially covers the cavity. D, surface representation of Tlc, in a roughly similar orientation as in C, including the modeled loop segments (colored according to the electrostatic potential: positive charge, blue; negative charge, red). The opening of the calyx measures about 10 Å in diameter and is lined with negative charges, most prominently Glu-34 and Asp-80. A second entry to the large cavity, located between loops 1 and 4, is indicated by an arrow on the left.

 
Three of the loops at the open end of the {beta}-barrel are rather long and partially disordered in the crystal structure, hence allowing access to the central cavity even for voluminous ligands. In particular, the {Omega}-type loop 1 with its 13 residues is one of the largest loops among the prototypic lipocalins that have been structurally characterized so far (16). On its N-terminal side, this loop is likely to form a short {alpha}-helix comparable with those observed in Bda and Blg (Fig. 2). Loops 2 and 4 appear as {beta}-hairpins, whereas the tip of the rather short loop 3 connecting {beta}-strands E/F (cf. Fig. 1A) locally assumes the conformation of a 310-helix (residues Asp-80 to Gly-82).



View larger version (47K):
[in this window]
[in a new window]
  		FIG. 2.
Superposition between Tlc, Bda, and Blg. A, stereoview of the superposition of Tlc (red, modeled regions in magenta) with Bda (yellow, PDB accession code 1BJ7 [PDB] ), and Blg (blue, PDB accession code 1BEB [PDB] ). Structurally conserved {beta}-barrel residues (as defined in B) are colored in different shades of gray. r.m.s. deviations relative to Tlc for the 58 C-{alpha} positions of these {beta}-barrel residues are 1.5 and 1.7 Å for Bda and Blg, respectively. Loop 1 of Tlc, connecting {beta}-strands A/B, is by two and three residues shorter than the corresponding loops of Bda and Blg, respectively. B, sequence alignment of Tlc, Bda, and Blg. All three amino acid sequences were aligned based on structural similarities. Equivalent {beta}-barrel residues (as defined for Bda and Blg in Ref. 16) that were used for superposition are boxed. Secondary structure elements are indicated as interpreted by DSSP (39): {beta}-strands A-I are depicted with green background and {alpha}-helices with an orange background. In case of Tlc, {beta}-strands and {alpha}-helices are also marked as green arrows and orange cylinders, respectively. Residues His-1 to Leu-4 of Tlc, which were not present in the recombinant protein, as well as the Cys-101 -> Ser replacement and the C-terminal Strep-tag II are shown as lowercase letters. Residue Asp-158 of native Tlc was substituted by Ala as linker to the Strep-tag II.

 
Even though Tlc represents one of the smallest members of the lipocalin family, with 158 residues for the native, mature polypeptide (cf. Fig. 2), it exhibits an unusually large cavity inside its {beta}-barrel. Nevertheless, the {beta}-barrel itself, with the three loops that tie it up at its closed end and with the {alpha}-helix attached to it, appears conformationally rigid and is nicely superimposable to the supersecondary structure of other lipocalins.

Most interestingly, apart from the wide mouth formed approximately at the middle between the loops 1 to 4 there is a second, smaller entry to the cavity of Tlc, which appears between {beta}-strands A and H (Fig. 1D). The main entrance of the calyx-like pocket measures about 10 Å in diameter and is lined with negatively charged residues (Glu-34 and Asp-80) as well as hydrophobic residues. The cavity itself exhibits a bifurcated shape, is about 15 Å deep, and its inside is slightly positively charged (Fig. 3). This positive charge arises from the side chains of residues His-86 and Lys-114 together with the dipoles of main chain amide groups of up to 10 residues in the vicinity.



View larger version (88K):
[in this window]
[in a new window]
  		FIG. 3.
Illustration of the Tlc cavity and apparent shape complementarity with cognate ligands. A, inner accessible surface of the cavity, colored according to the electrostatic potential (positive charge, blue; negative charge, red), with the polypeptide backbone shown in gray. The ligand pocket is about 15 Å deep, bifurcated, and positively charged at the bottom. The large cavity is lined with basic and hydrophobic residues, most notably Met-39 and Phe-99, which separate the two lobes. B, the antibiotic drug rifampin (stick model) can be nicely docked into the cavity (here shown as translucent inner surface) and almost fully occupies the pocket. C and D, retinoic acid docked into the cavity in two different orientations. Whereas the {beta}-ionone ring of the ligand resides approximately at the same position within one of the lobes at the bottom, its aliphatic side chain with the terminal carboxylate group can take two different directions, one pointing toward the small entry between strands A and H (C) and one pointing toward the central opening of the {beta}-barrel (D).

 
The remaining side chains that line the pocket are hydrophobic in nature, particularly residues Phe-99 and Met-39, which separate the two lobes at the bottom of the calyx. Although the cavity is accessible to solvent, the lower end and, most notably, the two lobes at its bottom are free of ordered water molecules in the crystal structure. This may be due to the absence of suitable H-bond donor/acceptor groups in order to fix the polar molecule (46).

Implications for Ligand Binding—There are three bound zinc ions in the crystal structure of Tlc (Fig. 1B). Zn-H2 is complexed by His-96 and Glu-125 together with two chloride ions. Zn-H3 is coordinated by His-92 and Asp-129, also located in the region around the N terminus of the {alpha}-helix, as well as by two water molecules. In contrast, Zn-H1 is involved in a crystal packing contact between two Tlc molecules and complexed by the side chains of Glu-73 and Lys-76 of one polypeptide and Glu-102 of a symmetry-related molecule. The fourth coordination site is occupied by a water molecule.

All three bound Zn(II) ions exhibit the typical tetrahedral coordination geometry, and their distances to the closest side chain atoms of Tlc (Table III) are in accordance with the published parameters for zinc-binding sites (47). Notably, when the Zn(II) that had been used in the precipitant solution (see "Experimental Procedures") was substituted with Mg(II), no crystal growth was observed. However, preliminary binding studies between the recombinant Tlc and Zn(II) revealed a rather weak dissociation constant of ~1.33 mM (determined as half-maximal metal ion concentration to achieve saturation during CD titration at 212 nm).2


View this table:
[in this window]
[in a new window]
  		TABLE III
Interatomic distances of the zinc-binding sites

 

The unusual shape of the binding pocket of Tlc prompted us to investigate possible modes of interaction with two of its known ligands, retinoic acid and rifampin (7, 15). Unfortunately, corresponding co-crystallization trials remained unsuccessful. However, molecular docking with rifampin (Cambridge Structural Database entry RIFAMP) clearly revealed that the bulky molecule occupies almost the entire pocket and that it fits the cavity with the two lobes astonishingly well (Fig. 3B). In contrast, two possible binding modes were identified for retinoic acid (Cambridge Structural Database entry VITAAC01), which is significantly smaller than rifampin and has a more extended shape. In both orientations, the {beta}-ionone ring of retinoic acid can be placed into the larger of the two lobes at the bottom of the cavity, whereas the terminal carboxylate group is alternatively arranged at the two different exits (Fig. 3, C and D).

Structural Comparison with Other Lipocalins—A search through the PDB using the DALI server (48) revealed that Tlc (without the modeled loop regions) is structurally most closely related to the bovine dander allergen Bos d 2 (PDB accession code 1BJ7 [PDB] (45); Z score = 15.1), one of the prototypic lipocalins (16). Mouse major urinary protein (PDB accession code 1MUP [PDB] (49); Z score = 14.2), rat epididymal retinoic acid-binding protein (PDB accession code 1EPA [PDB] (50); Z score = 14.1), and bovine {beta}-lactoglobulin (PDB accession code 1BEB [PDB] (51); Z score = 13.7) showed lower similarity. Because of its comparably high sequence homology of 23%, the latter lipocalin was employed previously as a template structure for the homology modeling of Tlc (10). Even though Tlc and Bda display merely 17.8% sequence identity (Fig. 2B), superposition of the 58 C-{alpha} atoms of the {beta}-barrel (16) gave rise to a low r.m.s. deviation of 1.5 Å, whereas a corresponding value of 2.3 Å was calculated for all 128 superimposable C-{alpha} atoms.

Despite the high structural similarity of the {beta}-barrel and {alpha}-helix, major differences were found in the loop regions at the open end of the lipocalin calyx (Fig. 2A). Loop 1, which is largely disordered in Tlc but well defined in the Bda crystal structure (45), runs almost across the opening of the calyx, thereby controlling accessibility of the ligand-binding site. The spatial orientation of the segments flanking this loop, which could be built with confidence in the Tlc crystal structure, together with the comparable loop lengths in both lipocalins (with two residues less in Tlc than in Bda; cf. Fig. 2B) indicate that the loop might adopt a similar although more flexible conformation in Tlc.

In addition, the C-{alpha} position of residue Gly-59 at the tip of loop 2 is located ~8 Å further away from the central axis of the {beta}-barrel in Tlc than the homologous residue Gly-61 of Bda. Notably, the well resolved loop 3, which is flipped toward the cavity in Bda, points outward in Tlc, resulting in an 8-Å displacement of the C-{alpha} positions of the corresponding residues Gly-84 and Gly-82. As a consequence, the entrance to the cavity is wide open in Tlc, whereas it is closed in Bda. Apart from the small gap in the electron density of Tlc, the conformation of loop 4 is similar in both lipocalins. Most interestingly, at the bottom of the {beta}-barrel the short loops connecting the {beta}-strands B/C and F/G, respectively, are shifted away from the central axis in Tlc compared with Bda. Hence, another surface region with concave shape appears opposite to the conventional ligand pocket in this lipocalin.

The structure of Blg, which has served as a homologue of Tlc so far (7, 10), is highly similar to that of Bda. Again, the {beta}-barrel and {alpha}-helix superimpose well with those of Tlc (Fig. 2A), whereas the loop regions of Blg resemble more closely those of Bda. Loops 1-3 are found in almost the same conformation as in Bda, whereas loop 4 is shifted away from the {beta}-barrel axis in Blg compared with both Bda and Tlc. The N-terminal peptide segment, which is disordered in Tlc and in Bda, is well defined in Blg and runs across the bottom of the {beta}-barrel as observed for most other lipocalins (16).

Comparison of Tear Lipocalin with Protease Inhibitors—Previous experimental evidence indicated that full-length Tlc as well as a synthetic peptide comprising residues Leu-3 to Ala-21 exert inhibitory activity toward the cysteine proteinase papain (17, 52). Three sequences were identified within the first 17 N-terminal residues of Tlc that are homologous to the motifs known to be essential for the inhibitory activity of cystatins and the closely related stefins (53). Although the sequence motifs are more than 40 residues apart in the primary structures of these well characterized inhibitors, they are spatially close in their three-dimensional fold, thus forming a complementary interface with the catalytic cleft of the proteinase.

To investigate any conformational similarities between the N-terminal segment of Tlc and the cystatin-type proteinase inhibitor fold, the responsible motifs of chicken egg-white cystatin (PDB accession code 1CEW [PDB] (54)), stefin A (PDB accession code 1NB5 [PDB] (55)), and stefin B (PDB accession code 1STF [PDB] (56)) were superimposed onto each other and with the N-terminal segment of Tlc (Fig. 4). Although mutual superposition of the different cystatin-type inhibitors revealed a clear pattern of conserved main chain and side chain orientation, the proposed inhibitory peptide of Tlc did not exhibit reasonable structural similarity. In particular, Trp-17, which was considered functionally homologous to the central tryptophan residue of the cystatins in the third sequence motif and occurs there exposed at the tip of a hairpin loop, is buried in the hydrophobic core of Tlc. In fact, this amino acid is even part of the Gly-Xaa-Trp signature characteristic for all lipocalin structures (8) and therefore is not likely to participate in an inhibitor-proteinase interaction, at least as long as the native fold is adopted.



View larger version (47K):
[in this window]
[in a new window]
  		FIG. 4.
Structural comparison between Tlc and proteinase inhibitors. A, the first ordered residues at the N terminus of Tlc, Asp-12 to Tyr-18, with 2Fo - Fc electron density map at 1.8-Å resolution contoured at 1.0 {sigma}. B, superposition of the inhibitory motif at the N terminus of Tlc (colored red) with the inhibitory sites of chicken egg-white cystatin (PDB accession code 1CEW [PDB] ), stefin A (PDB accession code 1NB5 [PDB] ), and stefin B (PDB accession code 1STF [PDB] ), all depicted in gray. Only five (Asp-12 to Gly-15 and Trp-17) of a total of 10 residues that comprise the three inhibitory motifs of Tlc are structurally defined in the present crystal structure and were used for the comparison. In the three proteinase inhibitors the side chain orientations of Gln-53 to Gly-57 (same residue numbers in all three molecules) as well as of residue Trp/Gly/His-104 are closely similar, whereas the presumed inhibitory peptide segment of Tlc clearly does not adopt a cystatin-like pattern. C, superposition of the N terminus of Tlc, residues Val-13 to Tyr-18 (colored red), with Phe-21 to Tyr-26 of triabin (yellow), Phe-27 to Ser-32 of staphostatin B (green), and Leu-11 to Val-16 of Inh (blue), i.e. the segments colored light gray in D. The Gly-X-Trp motif of residues 15-17 of Tlc, which is conserved in the {beta}-barrel structures of lipocalins, is also present in these three proteinase inhibitors. Yet it is not involved in their differing inhibitory mechanisms. D, crystal structures of the three known proteinase inhibitors with an eight-stranded {beta}-barrel fold: triabin (PDB accession code 1AVG [PDB] , yellow), staphostatin B (PDB accession code 1PXV, green), and Inh of E. chrysanthemi (PDB accession code 1SMP [PDB] , blue). Side chains of residues interacting with the respective target proteinase are depicted in a ball-and-stick representation in dark gray. The conserved stretch of six residues close to the bottom of each {beta}-barrel, shown in greater detail in C, is colored light gray.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
As predicted based on its amino acid sequence (7), Tlc exhibits the typical lipocalin fold, comprising an eight-stranded, antiparallel {beta}-barrel with a C-terminal {alpha}-helix. The crystal structure reveals a remarkably large ligand cavity inside the {beta}-barrel and hence provides insight into some of the unusual features of Tlc as member of the lipocalin family, especially its broad ligand specificity. The high promiscuity for a range of different ligands is reflected by the peculiar bifurcated shape of the central cavity, which is about 15 Å deep. With a mouth of roughly 10 Å in diameter and an even wider bottom, the binding site of Tlc cannot only accommodate slim fatty acid molecules but also much bulkier compounds like cholesterol and rifampin.

A three-dimensional model of Tlc was previously constructed on the basis of an extended site-directed tryptophan substitution and fluorescence study (10), in which each native amino acid, except Gly-117, was sequentially exchanged. Measurement of the fluorescence emission maxima for each Trp mutant revealed characteristic periodicities as expected for eight {beta}-strands and altogether three {alpha}-helical segments. These regions were subsequently aligned with the corresponding secondary structural elements of Blg, and a homology model of Tlc was constructed. All in all, this homology model is in good agreement with the crystal structure described here, whereby the lengths of the {beta}-strands and of the C-terminal {alpha}-helix deviate by no more than 3 residues from the lengths determined on basis of the crystallographic data using DSSP (39).

Furthermore, perturbations in the periodicities of the fluorescence emission peaks within the secondary structure elements allowed the correct prediction of the position of the {alpha}-helix with respect to the {beta}-barrel. Residues Val-113 and Leu-115 of {beta}-strand H as well as Phe-130 within the C-terminal {alpha}-helix showed less exposure to solvent than expected and were therefore placed at the interface between the {alpha}-helix and {beta}-strand H. These findings could be confirmed in the crystal structure, where Phe-130 is located almost in the middle of the {alpha}-helix and packs against the {beta}-barrel, contacting the two aliphatic side chains in strand H. Its C-{alpha} position lies 7.5 and 6.0 Å distant to the C-{alpha} positions of Val-113 and Leu-115, respectively.

Two other spectroscopic perturbations in the {beta}-strands A (at Lys-20 and Ala-21) and F (at Ile-89 and Arg-90) were interpreted as {beta}-bulges and, indeed, the side chains of these neighboring residues are all solvent-exposed in the crystal structure, with concomitant irregularities in the {beta}-strand backbone. Most interestingly, the three residues following the {beta}-bulge in strand A (Met-22 to Val-24) form a short antiparallel {beta}-sheet with the short strand I (Ser-143 to Leu-145) that is part of the extended C-terminal polypeptide segment, which is found in most lipocalins and attaches to one side of the {beta}-barrel.

The predicted short loop 3 connecting {beta}-strands E/F also becomes apparent from the crystal structure. Yet, its tip points outward from the {beta}-barrel axis, in contrast with the situation in Blg (cf. Fig. 2A). For residues Pro-29 to Thr-37 as part of loop 1 and residues Glu-9 to Ser-14 close to the N terminus of Tlc, {alpha}-helical and 310-helical conformation, respectively, was predicted. These assumptions could not be confirmed in the crystal structure, even though it must be pointed out that some of the residues in these regions are not well resolved in the electron density.

Determination of the dissociation constants between various Tlc mutants and 16-doxylstearic acid was used to identify residues that play a role in ligand recognition and are therefore likely to be located inside the calyx (10). In these experiments, three alanine residues (Ala-51, Ala-66, and Ala-86) were found that could not be replaced by tryptophan without significantly diminishing ligand affinity. A fourth residue, Ala-79, was furthermore proposed to play a role in this context but did not exhibit a similar experimental effect. In the crystal structure, all four Ala residues, as well as another two residues (Cys-101 and Gly-103) that were previously found to interact with a ligand of Tlc (57), are indeed located within the cavity.

Apart from that, energy minimization of the homology model for Tlc led to the assumption that Tyr-97 is exposed in the ligand pocket (10). In contrast, in the crystal structure this residue appears with its polar hydroxyl group on the outside of the {beta}-barrel. Perturbation of the alternating periodicity around Met-22 in {beta}-strand A was explained by its close distance to the side chain of Lys-114. The authors claimed that the positive charge of this residue induces the unexpected red shift in the fluorescence of the Met-22 -> Trp mutant and demonstrated that in the double mutant Met-22 -> Trp/Lys-114 -> Cys the expected periodic behavior is reinstalled. Indeed, in the crystal structure Met-22 and Lys-114 directly neighbor each other on {beta}-strands A and H, respectively, with their side chains lying almost in parallel.

Most interestingly, in the Trp substitution study all but one of the Tlc mutants, i.e. Gly-117 -> Trp, were purified as dimeric proteins. Inspection of the crystal structure revealed that this residue is located within the hydrophobic core of the protein. The introduction of another bulky residue is probably not tolerable there because it would disturb the dense packing of the surrounding side chains. However, in the production and purification of the recombinant wild-type Tlc employed here for protein crystallization, no signs of homodimerization appeared. Also, no tight packing contacts with a neighboring molecule were observed in the Tlc crystal so that an explanation for the previously described Tlc dimerization remains obscure. The unpaired Cys-101 residue does not seem to play a role in this respect because in contrast with other lipocalins that display a free Cys residue for covalent complex formation, such as NGAL (58), its side chain is buried sterically inaccessible inside the binding pocket of Tlc.

In the presently described crystal structure of the apo-form of human Tlc, a stretch of seven residues in loop 1, the tips of loop 2 and loop 4, as well as the disulfide bond linking the C terminus of the polypeptide chain to loop 2 were not defined in the electron density map. Adjacent residues showed high temperature factors, indicating that the whole loop region at the entrance to the cavity exhibits enhanced mobility. Previous investigation of the structural changes in Tlc upon ligand binding via CD spectroscopy revealed that ligand complexation induces the formation of additional {beta}-sheet structure and of a more rigid state of the protein (59). Thus, it is conceivable that the loops at the open end of the {beta}-barrel can at least to some degree adapt to different ligands.

The structure of Tlc is most similar to Bda, a lipocalin with yet unknown biological function produced by the sweat glands of domestic animals. In this protein the loop regions are more rigid, and some minor structural differences with Tlc are found. Loop 1 partially forms an {alpha}-helix in Bda and runs almost straight across the entrance to the ligand-binding site. Also, loops 2 and 3 are shifted toward the central axis of the {beta}-barrel in Bda compared with Tlc. As result, the ligand pocket is essentially obstructed in this allergen, and in fact no endogenous ligands, except of two water molecules, were found inside its cavity (45).

Unexpectedly, two zinc-binding sites were identified in the crystal structure of Tlc, together with a third one that occurs at a crystal packing contact. Spectrophotometric titration experiments revealed that there is a measurable affinity for Zn(II) in the low millimolar range. Hence, the zinc binding effect may not just be an artifact of crystallization in the presence of 200 mM Zn(II), even though the apparent dissociation constant seems to be above physiological levels. Whereas Zn(II) binding activity was not yet described for Tlc, zinc-responsive promoter elements were identified in the LCN1 gene (60). Furthermore, it was shown that gene expression is induced in knockout mice for MTF-1 (metal-responsive transcription factor 1), a highly conserved protein involved in the transcriptional response to heavy metal exposure (61).

In addition to its broad ligand binding function, which is a more or less typical feature of lipocalins, Tlc was reported to inhibit the proteolytic activity of papain. Van't Hof et al. (17) could show that Tlc inhibits papain to the same extent as cystatin S, an inhibitor also present at high concentration in tear fluid. However, the target proteinase of cystatin S has not yet been identified, and its inhibitory activity toward papain is much lower than that of cystatin C. In the following experiments with recombinant Tlc, and cystatin C serving as positive control, its inhibitory function was confirmed, but it also became evident that inhibition of papain occurs only in a narrow concentration range (52).

Still, a synthetic 18-amino acid peptide derived from the N terminus of Tlc exhibited comparable proteinase inhibitory activity as a peptide that contained the three inhibitory motifs of cystatin S linked by single Ala residues, although both peptides inhibited papain only at much higher concentration than their respective parent proteins (17). In fact, at increased papain concentration, Tlc is used as a substrate (52). N-terminal sequencing revealed that two cleavage sites lie within the N-terminal Leu-Leu-Ala-Ser motif (residues 3-6 of Tlc), but not in the conserved Gln-X-Val-X-Gly motif (residues 11-15 of Tlc), as is typical for cystatin-type inhibitors.

Structural comparison revealed (Fig. 4B) that although in various cystatins the three inhibitory sequence motifs form a contiguous interface that binds to the catalytic cleft of the target proteinase (also reviewed in Ref. 53), the N-terminal segment of Tlc cannot be superimposed onto this structural element in a meaningful way. Although the N-terminal Leu-Leu-Ala-Ser sequence and the first residue of the Gln-X-Val-X-Gly motif are not resolved in the electron density and probably not conformationally ordered in Tlc, the third sequence motif containing Trp-17 is buried within its hydrophobic core. Because the Trp motif seems to be of minor importance for the inhibitory activity of cystatin-derived peptides (62), it is possible that the flexible N terminus of Tlc acts in a peptide-like manner. This would also explain why Tlc is cleaved at higher papain concentration within one of the inhibitory motifs, similarly as the cystatin-derived peptides but in contrast to the full-length cystatins.

Notably, there are several other proteinase inhibitors that rely on an eight-stranded {beta}-barrel fold similar to Tlc and for which the mode of interaction with their respective target enzymes is known in detail (Fig. 4, C and D). Triabin (PDB accession code 1AVG [PDB] (63)), a 142-residue protein from the saliva of the blood-sucking insect Triatoma pallidipennis, prevents blood clot formation by inhibiting the serine proteinase thrombin via binding to its exosite I. Unlike the prototypic lipocalin fold, {beta}-strands A-D are arranged in an up-up-down-down topology, whereas the following {beta}-strands are antiparallel.

Recently, the structure of staphostatin B of Staphylococcus aureus, the inhibitor of the cysteine proteinase staphopain B, was elucidated (PDB accession code 1PXV (64)). In contrast to Tlc with its eight strictly antiparallel {beta}-strands, neighboring strands A and B of staphostatin B have parallel topology. The inhibitory site of staphostatin B, which was co-crystallized with its target proteinase, is located in the loop connecting the two C-terminal {beta}-strands. Residues 97-101 of staphostatin B interact with the catalytic cleft of staphopain. Most interestingly, this peptide stretch does not bind in a substrate-like manner but in an inverted (antiparallel) orientation.

Like Tlc, the metalloproteinase inhibitor Inh (PDB accession code 1SMP [PDB] (65)) from Erwinia chrysanthemi folds into a strictly antiparallel {beta}-barrel. This protein was shown to inhibit the 50-kDa proteinase from Serratia marcescens and its presumed physiological role is the protection of periplasmic proteins against proteolysis (66). In the crystallized complex with this proteinase, the five N-terminal residues of the inhibitor extend away from the {beta}-barrel and bind to the catalytic cleft of the enzyme. Faint electron density for residue Ser-1 of the inhibitor, together with N-terminal sequencing, revealed that this residue is cleaved by ~50% in the proteinase-inhibitor complex.

In contrast with Tlc, none of these proteinase inhibitors possesses a cavity that appears suitable for ligand binding, rather the interior of each {beta}-barrel is densely packed with mostly hydrophobic side chains. Furthermore, they lack the C-terminal {alpha}-helix typical for lipocalins, and due to their varying {beta}-barrel topologies, they exhibit structural differences in the loop regions. However, comparison with Tlc also reveals some similarities. The C-{alpha} positions of the {beta}-barrel residues superimpose astonishingly well, resulting in r.m.s. deviations of 1.6, 2.1, and 2.2 Å for triabin, staphostatin B, and Inh, respectively.

Moreover, the Gly-Xaa-Trp motif, which is strictly conserved in the lipocalins, is also found in these three inhibitors at approximately the same position (Fig. 4C), close to the bottom of the {beta}-barrel with the Trp side chain buried in the hydrophobic core. Indeed, triabin and Inh have recently been ascribed to the calycin protein superfamily (31). Nevertheless, a structural explanation for the inhibitory activity of Tlc toward a thiol proteinase is not evident from comparison with either of these three proteins (Fig. 4D).

Even though experimental evidence exists that the N terminus of Tlc can act as a weak inhibitor of a thiol-proteinase in vitro, this probably does not correspond to a relevant physiological function. There are different, more potent cystatins present in the same secretory fluids and tissues, which, in contrast to Tlc, have also been isolated from tear samples in complex with various proteinases (38). Consequently, the major function of Tlc seems to lie in the binding of multiple physiological or pathological compounds. In order to gain further insight into the mode of interaction between this lipocalin and its presumed ligands, co-crystallization experiments are currently under way.


   FOOTNOTES
 
The atomic coordinates and structure factors (code 1XKI) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

* 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

Supported by Austrian Science Foundation (FWF) Grant P14850 [GenBank] . Back

|| To whom correspondence should be addressed: Lehrstuhl für Biologische Chemie, Technische Universität München, D-85350 Freising-Weihenstephan, Germany. Tel.: 49-8161-71-4351; Fax: 49-8161-71-4352; E-mail: skerra{at}wzw.tum.de.

1 The abbreviations used are: Tlc, human tear lipocalin; Bda, bovine dander allergen Bos d 2; Blg, {beta}-lactoglobulin; PDB, Protein Data Bank; r.m.s., root mean square; SeMet, selenomethionine. Back

2 A. Skerra, C. Cirl, and D. A. Breustedt, manuscript in preparation. Back


   ACKNOWLEDGMENTS
 
We gratefully acknowledge the help of G. Reil for mass spectrometric analysis of the SeMet derivative and the EMBL Grenoble Outstation, in particular M. Walsh, for support of measurements at the ESRF under the European Union "Improving Human Potential Programme."



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Josephson, A. S., and Lockwood, D. W. (1964) J. Immunol. 93, 532-539[Abstract/Free Full Text]
  2. Bonavida, B., Sapse, A. T., and Sercarz, E. E. (1969) Nature 221, 375-376[CrossRef][Medline] [Order article via Infotrieve]
  3. Scalfari, F., Castagna, M., Fattori, B., Andreini, I., Maremmani, C., and Pelosi, P. (1997) Comp. Biochem. Physiol. 118, 819-824
  4. Lacazette, E., Gachon, A. M., and Pitiot, G. (2000) Hum. Mol. Genet. 9, 289-301[Abstract/Free Full Text]
  5. Redl, B. (2000) Biochim. Biophys. Acta 1482, 241-248[CrossRef][Medline] [Order article via Infotrieve]
  6. Wojnar, P., Dirnhofer, S., Ladurner, P., Berger, P., and Redl, B. (2002) J. Histochem. Cytochem. 50, 433-435[Abstract/Free Full Text]
  7. Redl, B., Holzfeind, P., and Lottspeich, F. (1992) J. Biol. Chem. 267, 20282-20287[Abstract/Free Full Text]
  8. Flower, D. R. (1996) Biochem. J. 318, 1-14[Medline] [Order article via Infotrieve]
  9. Glasgow, B. J., Abduragimov, A. R., Yusifov, T. N., Gasymov, O. K., Horwitz, J., Hubbell, W. L., and Faull, K. F. (1998) Biochemistry 37, 2215-2225[CrossRef][Medline] [Order article via Infotrieve]
  10. Gasymov, O. K., Abduragimov, A. R., Yusifov, T. N., and Glasgow, B. J. (2001) Biochemistry 40, 14754-14762[CrossRef][Medline] [Order article via Infotrieve]
  11. Glasgow, B. J., Abduragimov, A. R., Farahbakhsh, Z. T., Faull, K. F., and Hubbell, W. L. (1995) Curr. Eye Res. 14, 363-372[Medline] [Order article via Infotrieve]
  12. Lechner, M., Wojnar, P., and Redl, B. (2001) Biochem. J. 356, 129-135[CrossRef][Medline] [Order article via Infotrieve]
  13. Gasymov, O. K., Abduragimov, A. R., Yusifov, T. N., and Glasgow, B. J. (1999) Biochim. Biophys. Acta 1433, 307-320[CrossRef][Medline] [Order article via Infotrieve]
  14. Fluckinger, M., Haas, H., Merschak, P., Glasgow, B. J., and Redl, B. (2004) Antimicrob. Agents Chemother. 48, 3367-3372[Abstract/Free Full Text]
  15. Gasymov, O. K., Abduragimov, A. R., Gasimov, E. O., Yusifov, T. N., Dooley, A. N., and Glasgow, B. J. (2004) Biochim. Biophys. Acta 1688, 102-111[Medline] [Order article via Infotrieve]
  16. Skerra, A. (2000) Biochim. Biophys. Acta 1482, 337-350[CrossRef][Medline] [Order article via Infotrieve]
  17. van't Hof, W., Blankenvoorde, M. F., Veerman, E. C., and Amerongen, A. V. (1997) J. Biol. Chem. 272, 1837-1841[Abstract/Free Full Text]
  18. Yusifov, T. N., Abduragimov, A. R., Gasymov, O. K., and Glasgow, B. J. (2000) Biochem. J. 347, 815-819[Medline] [Order article via Infotrieve]
  19. Holzfeind, P., Merschak, P., Rogatsch, H., Culig, Z., Feichtinger, H., Klocker, H., and Redl, B. (1996) FEBS Lett. 395, 95-98[CrossRef][Medline] [Order article via Infotrieve]
  20. Skerra, A. (1992) Nucleic Acids Res. 20, 3551-3554[Abstract/Free Full Text]
  21. Skerra, A. (1994) Gene (Amst.) 151, 131-135[CrossRef][Medline] [Order article via Infotrieve]
  22. Geisselsoder, J., Witney, F., and Yuckenberg, P. (1987) BioTechniques 5, 786-790
  23. Yanisch-Perron, C., Vieira, J., and Messing, J. (1985) Gene (Amst.) 33, 103-119[CrossRef][Medline] [Order article via Infotrieve]
  24. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY
  25. Skerra, A., and Schmidt, T. G. (2000) Methods Enzymol. 326, 271-304[Medline] [Order article via Infotrieve]
  26. Van Duyne, G. D., Standaert, R. F., Karplus, P. A., Schreiber, S. L., and Clardy, J. (1993) J. Mol. Biol. 229, 105-124[CrossRef][Medline] [Order article via Infotrieve]
  27. Gassner, N. C. (1998) Structural, thermodynamic and kinetic analysis of T4 lysozyme variants with methionine cores. Ph.D. thesis, University of Oregon
  28. CCP4 (1994) Acta Crystallogr. Sect. D Biol. Crystallogr. 50, 760-763[CrossRef][Medline] [Order article via Infotrieve]
  29. Weeks, C. M., and Miller, R. (1999) J. Appl. Crystallogr. 32, 120-124[CrossRef]
  30. de la Fortelle, E., and Bricogne, G. (1997) Methods Enzymol. 276, 472-494[CrossRef]
  31. Flower, D. R., North, A. C., and Sansom, C. E. (2000) Biochim. Biophys. Acta 1482, 9-24[CrossRef][Medline] [Order article via Infotrieve]
  32. Abrahams, J. P., and Leslie, A. G. W. (1996) Acta Crystallogr. Sect. D Biol. Crystallogr. 52, 30-42[CrossRef][Medline] [Order article via Infotrieve]
  33. Lamzin, V. S., and Wilson, K. S. (1993) Acta Crystallogr. Sect. D Biol. Crystallogr. 49, 129-149[CrossRef][Medline] [Order article via Infotrieve]
  34. Oldfield, T. J. (2001) Acta Crystallogr. Sect. D Biol. Crystallogr. 57, 82-94[CrossRef][Medline] [Order article via Infotrieve]
  35. Ponder, J. W., and Richards, F. M. (1987) J. Mol. Biol. 193, 775-791[CrossRef][Medline] [Order article via Infotrieve]
  36. Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, 946-950[CrossRef]
  37. Merrit, E. A., and Bacon, D. J. (1997) Methods Enzymol. 277, 505-524[Medline] [Order article via Infotrieve]
  38. Sack, R. A., Sathe, S., Beaton, A., Kozinski, M., Bogart, B., Lew, G., Sharma, S., and Upponi, A. (2004) Exp. Eye Res. 78, 371-378[CrossRef][Medline] [Order article via Infotrieve]
  39. Kabsch, W., and Sander, C. (1983) Biopolymers 22, 2577-2637[CrossRef][Medline] [Order article via Infotrieve]
  40. Holm, L. (1995) Protein Data Bank Newsletter 71, 6-9
  41. Arnez, J. G. (1994) J. Appl. Crystallogr. 27, 649-653
  42. Nicholls, A., Sharp, K. A., and Honig, B. (1991) Proteins 11, 281-296[CrossRef][Medline] [Order article via Infotrieve]
  43. Kleywegt, G. J., Zou, J. Y., Kjeldgaard, M., and Jones, T. A. (2001) in International Tables for Crystallography (Rossmann, M. G., and Arnold, E., eds) Vol. F, pp. 353-367, Kluwer Academic Publishers Group, Dordrecht, Netherlands
  44. Berman, H. M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T. N., Weissig, H., Shindyalov, I. N., and Bourne, P. E. (2000) Nucleic Acids Res. 28, 235-242[Abstract/Free Full Text]
  45. Rouvinen, J., Rautiainen, J., Virtanen, T., Zeiler, T., Kauppinen, J., Taivainen, A., and Mantyjarvi, R. (1999) J. Biol. Chem. 274, 2337-2343[Abstract/Free Full Text]
  46. Levitt, M., and Park, B. H. (1993) Structure (Lond.) 1, 223-226[Medline] [Order article via Infotrieve]
  47. Harding, M. M. (2001) Acta Crystallogr. Sect. D Biol. Crystallogr. 57, 401-411[CrossRef][Medline] [Order article via Infotrieve]
  48. Holm, L., and Sander, C. (1993) J. Mol. Biol. 233, 123-138[CrossRef][Medline] [Order article via Infotrieve]
  49. Bocskei, Z., Groom, C. R., Flower, D. R., Wright, C. E., Phillips, S. E., Cavaggioni, A., Findlay, J. B., and North, A. C. (1992) Nature 360, 186-188[CrossRef][Medline] [Order article via Infotrieve]
  50. Newcomer, M. E. (1993) Structure (Lond.) 1, 7-18[Medline] [Order article via Infotrieve]
  51. Brownlow, S., Morais Cabral, J. H., Cooper, R., Flower, D. R., Yewdall, S. J., Polikarpov, I., North, A. C., and Sawyer, L. (1997) Structure (Lond.) 5, 481-495[Medline] [Order article via Infotrieve]
  52. Wojnar, P., van't Hof, W., Merschak, P., Lechner, M., and Redl, B. (2001) Biol. Chem. 382, 1515-1520[CrossRef][Medline] [Order article via Infotrieve]
  53. Bode, W., and Huber, R. (2000) Biochim. Biophys. Acta 1477, 241-252[CrossRef][Medline] [Order article via Infotrieve]
  54. Bode, W., Engh, R., Musil, D., Thiele, U., Huber, R., Karshikov, A., Brzin, J., Kos, J., and Turk, V. (1988) EMBO J. 7, 2593-2599[Medline] [Order article via Infotrieve]
  55. Jenko, S., Dolenc, I., Guncar, G., Dobersek, A., Podobnik, M., and Turk, D. (2003) J. Mol. Biol. 326, 875-885[CrossRef][Medline] [Order article via Infotrieve]
  56. Stubbs, M. T., Laber, B., Bode, W., Huber, R., Jerala, R., Lenarcic, B., and Turk, V. (1990) EMBO J. 9, 1939-1947[Medline] [Order article via Infotrieve]
  57. Gasymov, O. K., Abduragimov, A. R., Yusifov, T. N., and Glasgow, B. J. (2000) Protein Sci. 9, 325-331[Medline] [Order article via Infotrieve]
  58. Goetz, D. H., Willie, S. T., Armen, R. S., Bratt, T., Borregaard, N., and Strong, R. K. (2000) Biochemistry 39, 1935-1941[CrossRef][Medline] [Order article via Infotrieve]
  59. Gasymov, O. K., Abduragimov, A. R., Yusifov, T. N., and Glasgow, B. J. (1998) Biochim. Biophys. Acta 1386, 145-156[CrossRef][Medline] [Order article via Infotrieve]
  60. Holzfeind, P., and Redl, B. (1994) Gene (Amst.) 139, 177-183[CrossRef][Medline] [Order article via Infotrieve]
  61. Lichtlen, P., Wang, Y., Belser, T., Georgiev, O., Certa, U., Sack, R., and Schaffner, W. (2001) Nucleic Acids Res. 29, 1514-1523[Abstract/Free Full Text]
  62. Lalmanach, G., Hoebeke, J., Moreau, T., Brillard-Bourdet, M., Ferrer-Ditt Martino, M., Borras-Cuesta, F., and Gauthier, F. (1993) J. Protein Chem. 12, 23-31[CrossRef][Medline] [Order article via Infotrieve]
  63. Fuentes-Prior, P., Noeske-Jungblut, C., Donner, P., Schleuning, W. D., Huber, R., and Bode, W. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 11845-11850[Abstract/Free Full Text]
  64. Filipek, R., Rzychon, M., Oleksy, A., Gruca, M., Dubin, A., Potempa, J., and Bochtler, M. (2003) J. Biol. Chem. 278, 40959-40966[Abstract/Free Full Text]
  65. Baumann, U., Bauer, M., Letoffe, S., Delepelaire, P., and Wandersman, C. (1995) J. Mol. Biol. 248, 653-661[CrossRef][Medline] [Order article via Infotrieve]
  66. Letoffe, S., Delepelaire, P., and Wandersman, C. (1989) Mol. Microbiol. 3, 79-86[Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J BiochemHome page
K. Inoue, N. Yagi, Y. Urade, and T. Inui
Compact Packing of Lipocalin-type Prostaglandin D Synthase Induced by Binding of Lipophilic Ligands
J. Biochem., February 1, 2009; 145(2): 169 - 175.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
S. Shimamoto, T. Yoshida, T. Inui, K. Gohda, Y. Kobayashi, K. Fujimori, T. Tsurumura, K. Aritake, Y. Urade, and T. Ohkubo
NMR Solution Structure of Lipocalin-type Prostaglandin D Synthase: EVIDENCE FOR PARTIAL OVERLAPPING OF CATALYTIC POCKET AND RETINOIC ACID-BINDING POCKET WITHIN THE CENTRAL CAVITY
J. Biol. Chem., October 26, 2007; 282(43): 31373 - 31379.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
A. Eichinger, A. Nasreen, H. J. Kim, and A. Skerra
Structural Insight into the Dual Ligand Specificity and Mode of High Density Lipoprotein Association of Apolipoprotein D
J. Biol. Chem., October 19, 2007; 282(42): 31068 - 31075.
[Abstract] [Full Text] [PDF]

Home page
 Microbiol. Mol. Biol. Rev.Home page
M. Miethke and M. A. Marahiel
Siderophore-Based Iron Acquisition and Pathogen Control
Microbiol. Mol. Biol. Rev., September 1, 2007; 71(3): 413 - 451.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
K.-i. Miyazono, M. Watanabe, J. Kosinski, K. Ishikawa, M. Kamo, T. Sawasaki, K. Nagata, J. M. Bujnicki, Y. Endo, M. Tanokura, et al.
Novel protein fold discovered in the PabI family of restriction enzymes
Nucleic Acids Res., March 19, 2007; 35(6): 1908 - 1918.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
O. K. Gasymov, A. R. Abduragimov, P. Prasher, T. N. Yusifov, and B. J. Glasgow
Tear Lipocalin: Evidence for a Scavenging Function to Remove Lipids from the Human Corneal Surface
Invest. Ophthalmol. Vis. Sci., October 1, 2005; 46(10): 3589 - 3596.
[Abstract] [Full Text] [PDF]

Home page
 IOVSHome page
H. Saaren-Seppala, M. Jauhiainen, T. M. T. Tervo, B. Redl, P. K. J. Kinnunen, and J. M. Holopainen
Interaction of Purified Tear Lipocalin with Lipid Membranes
Invest. Ophthalmol. Vis. Sci., October 1, 2005; 46(10): 3649 - 3656.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/1/484    most recent
M410466200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
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 arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Breustedt, D. A.
Right arrow Articles by Skerra, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Breustedt, D. A.
Right arrow Articles by Skerra, A.
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?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2005 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement