Improved Lymphocyte Function-associated Antigen-1 (LFA-1) Inhibition by Statin Derivatives
2004; Elsevier BV; Volume: 279; Issue: 45 Linguagem: Inglês
10.1074/jbc.m407951200
ISSN1083-351X
AutoresGabriele Weitz‐Schmidt, Karl Welzenbach, Janet Dawson, Joerg Kallen,
Tópico(s)Atherosclerosis and Cardiovascular Diseases
ResumoThe integrin lymphocyte function-associated antigen-1 (LFA-1) (αLβ2; CD11a/CD18) plays an important role in leukocyte migration and T cell activation. LFA-1 is inhibited by the cholesterol-lowering drug lovastatin, which binds to an allosteric site of the αL I domain termed the lovastatin site (L-site). Here we report for the first time the x-ray structures of the LFA-1 I domain complexed with derivatives of lovastatin optimized for LFA-1 inhibition. This analysis identified two new subpockets within the L-site occupied by chemical groups of the statin derivatives but not by lovastatin itself. Occupancy of these L-site subpockets led to distinct conformational changes in LFA-1, which were detectable by an epitope-monitoring assay. We utilized this assay to demonstrate improved LFA-1 inhibition in human blood in vitro and in blood samples from treated animals ex vivo. Moreover, we demonstrate that the novel lovastatin-derived LFA-1 inhibitor LFA878 exhibits potent anti-inflammatory effects in carrageenan-induced rat paw edema. In summary, the findings reported here extend the understanding of LFA-1 inhibition at the molecular level, allow for the identification and design of LFA-1 inhibitors of further enhanced potency, and support the expectation that LFA-1 inhibitors binding to the L-site will be of therapeutic value in treating inflammatory diseases. The integrin lymphocyte function-associated antigen-1 (LFA-1) (αLβ2; CD11a/CD18) plays an important role in leukocyte migration and T cell activation. LFA-1 is inhibited by the cholesterol-lowering drug lovastatin, which binds to an allosteric site of the αL I domain termed the lovastatin site (L-site). Here we report for the first time the x-ray structures of the LFA-1 I domain complexed with derivatives of lovastatin optimized for LFA-1 inhibition. This analysis identified two new subpockets within the L-site occupied by chemical groups of the statin derivatives but not by lovastatin itself. Occupancy of these L-site subpockets led to distinct conformational changes in LFA-1, which were detectable by an epitope-monitoring assay. We utilized this assay to demonstrate improved LFA-1 inhibition in human blood in vitro and in blood samples from treated animals ex vivo. Moreover, we demonstrate that the novel lovastatin-derived LFA-1 inhibitor LFA878 exhibits potent anti-inflammatory effects in carrageenan-induced rat paw edema. In summary, the findings reported here extend the understanding of LFA-1 inhibition at the molecular level, allow for the identification and design of LFA-1 inhibitors of further enhanced potency, and support the expectation that LFA-1 inhibitors binding to the L-site will be of therapeutic value in treating inflammatory diseases. Lymphocyte function-associated antigen-1 (LFA-1) 1The abbreviations used are: LFA, lymphocyte function-associated antigen; I domain, inserted domain; ICAM, intercellular adhesion molecule; L-site, lovastatin site; mAb, monoclonal antibody; MIDAS, metal ion-dependent adhesion site; REMA, reporter epitope monitoring assay; RT, room temperature. (αLβ2; CD11a/CD18) is a α/β heterodimeric receptor belonging to the β2 integrin subfamily. LFA-1 resides on leukocytes in a low affinity, non-ligand binding state. The integrin requires activation by divalent metal cations or intracellular signals in order to bind to its major counter receptor, intercellular adhesion molecule (ICAM)-1, expressed on both endothelial cells and leukocytes (1Springer T.A. Cell. 1994; 76: 301-314Abstract Full Text PDF PubMed Scopus (6414) Google Scholar). As a member of the β2 integrin family LFA-1 contains an inserted (I) domain located between the β-sheets 2 and 3 of a seven-bladed β-propeller region on the αL chain (2Huang C. Lu C. Springer T.A. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 3156-3161Crossref PubMed Scopus (60) Google Scholar). Within the I domain a divalent cation coordination site (named metal ion-dependent adhesion site or MIDAS) interacts directly with the Glu-34 of ICAM-1 (3Shimaoka M. Xiao T. Liu J.H. Yang Y. Dong Y. Jun C.D. McCormack A. Zhang R. Joachimiak A. Takagi J. Wang J.H. Springer T.A. Cell. 2003; 112: 99-111Abstract Full Text Full Text PDF PubMed Scopus (424) Google Scholar). The β2 chain of LFA-1 contains an I-like domain that is known to regulate the activity status of the αL I domain during LFA-1 activation. In contrast to the αL I domain, the β2 I-like domain does not participate directly in ligand binding (4Lu C. Shimaoka M. Zang Q. Takagi J. Springer T.A. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 2393-2398Crossref PubMed Scopus (170) Google Scholar, 5Yang W. Shimaoka M. Chen J. Springer T.A. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 2333-2338Crossref PubMed Scopus (56) Google Scholar). LFA-1 plays a major role in lymphocyte homing and leukocyte trafficking in inflammation (1Springer T.A. Cell. 1994; 76: 301-314Abstract Full Text PDF PubMed Scopus (6414) Google Scholar). In addition, as a component of the immunological synapse, LFA-1 is involved in T cell activation upon antigen recognition (6Grakoui A. Bromley S.K. Sumen C. Davis M.M. Shaw A.S. Allen P.M. Dustin M.L. Science. 1999; 285: 221-227Crossref PubMed Scopus (2563) Google Scholar). These multiple functions make LFA-1 a promising therapeutic target in inflammatory and immunological diseases. Indeed, clinical studies suggest that anti-LFA-1 antibody therapy may be beneficial in bone marrow and solid organ transplantation (7Cavazzana-Calvo M. Bordigoni P. Michel G. Esperou H. Souillet G. Leblanc T. Stephan J.L. Vannier J.P. Mechinaud F. Reiffers J. Vilmer E. Landman-Parker J. Benkerrou M. Baruchel A. Pico J. Bernaudin F. Bergeron C. Plouvier E. Thomas C. Wijdenes J. Lacour B. Blanche S. Fischer A. Br. J. Haematol. 1996; 93: 131-138Crossref PubMed Scopus (38) Google Scholar, 8Hourmant M. Bedrossian J. Durand D. Lebranchu Y. Renoult E. Caudrelier P. Buffet R. Soulillou J.P. Transplantation. 1996; 62: 1565-1570Crossref PubMed Scopus (125) Google Scholar). The most convincing clinical data to date has been reported for a humanized anti-LFA-1 mAb administered to patients with chronic plaque psoriasis (9Lebwohl M. Tyring S.K. Hamilton T.K. Toth D. Glazer S. Tawfik N.H. Walicke P. Dummer W. Wang X. Garovoy M.R. Pariser D. N. Engl. J. Med. 2003; 349: 2004-2013Crossref PubMed Scopus (565) Google Scholar). However, the use of biologicals poses significant limitations in clinical practice, e.g. the need for parenteral application. Thus the availability of orally active low molecular weight inhibitors of LFA-1 would be highly desirable. High throughput screening has identified various low molecular weight inhibitors of the LFA-1/ICAM-1 interaction (10Shimaoka M. Springer T.A. Nat. Rev. Drug. Discov. 2003; 2: 703-716Crossref PubMed Scopus (296) Google Scholar). Based on their mode of action, these inhibitors can be divided into two major classes. The α I allosteric inhibitors have been shown to bind to a hydrophobic pocket near the C-terminal helix α7 of the αL I domain but distant from the MIDAS (10Shimaoka M. Springer T.A. Nat. Rev. Drug. Discov. 2003; 2: 703-716Crossref PubMed Scopus (296) Google Scholar, 11Kallen J. Welzenbach K. Ramage P. Geyl D. Kriwacki R. Legge G. Cottens S. Weitz-Schmidt G. Hommel U. J. Mol. Biol. 1999; 292: 1-9Crossref PubMed Scopus (234) Google Scholar, 12Weitz-Schmidt G. Welzenbach K. Brinkmann V. Kamata T. Kallen J. Bruns C. Cottens S. Takada Y. Hommel U. Nat. Med. 2001; 7: 687-692Crossref PubMed Scopus (967) Google Scholar, 13Liu G. Link J.T. Pei Z. Reilly E.B. Leitza S. Nguyen B. Marsh K.C. Okasinski G.F. von Geldern T.W. Ormes M. Fowler K. Gallatin M. J. Med. Chem. 2000; 43: 4025-4040Crossref PubMed Scopus (208) Google Scholar, 14Last-Barney K. Davidson W. Cardozo M. Frye L.L. Grygon C.A. Hopkins J.L. Jeanfavre D.D. Pav S. Qian C. Stevenson J.M. Tong L. Zindell R. Kelly T.A. J. Am. Chem. Soc. 2001; 123: 5643-5650Crossref PubMed Scopus (126) Google Scholar, 15Wattanasin S. Albert R. Ehrhardt C. Roche D. Sabio M. Hommel U. Welzenbach K. Weitz-Schmidt G. Bioorg. Med. Chem. Lett. 2003; 13: 499-502Crossref PubMed Scopus (32) Google Scholar, 16Crump M.P. Ceska T.A. Spyracopoulos L. Henry A. Archibald S.C. Alexander R. Taylor R.J. Findlow S.C. O'Connell J. Robinson M.K. Shock A. Biochemistry. 2004; 43: 2394-2404Crossref PubMed Scopus (32) Google Scholar). By binding to this pocket, the LFA-1 inhibitors are thought to block a downward axial displacement of helix α7, which is critical for I domain activation (10Shimaoka M. Springer T.A. Nat. Rev. Drug. Discov. 2003; 2: 703-716Crossref PubMed Scopus (296) Google Scholar). In addition to this first group of inhibitors, a second group of LFA-1 inhibitors has been recently identified that are thought to bind to the MIDAS of the β2 I-like domain, thereby affecting both the β2 I-like and αL I domain (17Welzenbach K. Hommel U. Weitz-Schmidt G. J. Biol. Chem. 2002; 277: 10590-10598Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar, 18Shimaoka M. Salas A. Yang W. Weitz-Schmidt G. Springer T.A. Immunity. 2003; 19: 391-402Abstract Full Text Full Text PDF PubMed Scopus (127) Google Scholar). These inhibitors, termed α/β I-like allosteric inhibitors, appear to disrupt signal transmission between the LFA-1 I domain and the I-like domain, resulting in a default low affinity conformation of the I domain (10Shimaoka M. Springer T.A. Nat. Rev. Drug. Discov. 2003; 2: 703-716Crossref PubMed Scopus (296) Google Scholar, 18Shimaoka M. Salas A. Yang W. Weitz-Schmidt G. Springer T.A. Immunity. 2003; 19: 391-402Abstract Full Text Full Text PDF PubMed Scopus (127) Google Scholar). The 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor lovastatin belongs to the class of α I allosteric inhibitors that interact with the hydrophobic pocket near the C-terminal helix α7 of the I domain (11Kallen J. Welzenbach K. Ramage P. Geyl D. Kriwacki R. Legge G. Cottens S. Weitz-Schmidt G. Hommel U. J. Mol. Biol. 1999; 292: 1-9Crossref PubMed Scopus (234) Google Scholar, 12Weitz-Schmidt G. Welzenbach K. Brinkmann V. Kamata T. Kallen J. Bruns C. Cottens S. Takada Y. Hommel U. Nat. Med. 2001; 7: 687-692Crossref PubMed Scopus (967) Google Scholar). Because lovastatin was the first molecule shown to bind to this site, the site was termed the lovastatin site (L-site) (12Weitz-Schmidt G. Welzenbach K. Brinkmann V. Kamata T. Kallen J. Bruns C. Cottens S. Takada Y. Hommel U. Nat. Med. 2001; 7: 687-692Crossref PubMed Scopus (967) Google Scholar). A series of lovastatin derivatives has since been synthesized that inhibit LFA-1 more potently than lovastatin itself and do no longer affect the activity of 3-hydroxy-3-methylglutaryl coenzyme A reductase (12Weitz-Schmidt G. Welzenbach K. Brinkmann V. Kamata T. Kallen J. Bruns C. Cottens S. Takada Y. Hommel U. Nat. Med. 2001; 7: 687-692Crossref PubMed Scopus (967) Google Scholar, 17Welzenbach K. Hommel U. Weitz-Schmidt G. J. Biol. Chem. 2002; 277: 10590-10598Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar). Little is known about the molecular basis leading to improved LFA-1 inhibition by these compounds. Moreover, most of the studies addressing LFA-1 inhibition by low molecular weight compounds via the L-site have been performed in assays involving the isolated I domain, purified LFA-1, or LFA-1 expressed on cell lines. Direct proof for L-site occupancy by lovastatin or statin derivatives in a more physiological setting is lacking. Utilizing x-ray crystallography, we investigate here for the first time the molecular basis for improved LFA-1 inhibition by the previously described statin-based inhibitor LFA703 (12Weitz-Schmidt G. Welzenbach K. Brinkmann V. Kamata T. Kallen J. Bruns C. Cottens S. Takada Y. Hommel U. Nat. Med. 2001; 7: 687-692Crossref PubMed Scopus (967) Google Scholar, 17Welzenbach K. Hommel U. Weitz-Schmidt G. J. Biol. Chem. 2002; 277: 10590-10598Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar) and the newly synthesized compound LFA878. Furthermore, we show that improved LFA-1 inhibition via the L-site can be detected in vitro in human blood and ex vivo in blood from treated animals. Our results in carrageenan-induced rat paw edema further substantiate the expectation that inhibitors targeting the L-site may have therapeutic potential as anti-inflammatory agents. Crystallization and Structure Determination of the αL I Domain/LFA878 Complex—For the αL I domain, the construct comprising residues Lys127–Gly311 was used, and purification was performed as reported previously (11Kallen J. Welzenbach K. Ramage P. Geyl D. Kriwacki R. Legge G. Cottens S. Weitz-Schmidt G. Hommel U. J. Mol. Biol. 1999; 292: 1-9Crossref PubMed Scopus (234) Google Scholar). The complex αL I domain/LFA878 was prepared by adding the ligand (100 mm solution in Me2SO) to the protein solution (12.7 mg/ml 10 mm MgSO4) in a 3:1 molar ratio. Crystals with maximal dimensions up to 0.4 mm could be grown by the hanging drop vapor diffusion method. Two microliters of αL I domain/LFA878 were mixed with an equal volume of well solution (0.085 m sodium citrate, pH 5.6, 0.17 m ammonium acetate, 15% glycerol, and 25.5% (w/v) polyethylene glycol 4000) and allowed to equilibrate against 1 ml of the latter at 20 °C. Diffraction data were collected to 2.1 Å at 20 °C on a Mar345 imaging plate detector at the Swiss-Norwegian beamline of the European Synchrotron Facility (Grenoble, France). After data processing with the HKL software suite (19Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref PubMed Scopus (38617) Google Scholar) and the CCP4-package (20Collaborative Computational Project No 4 Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (19797) Google Scholar), initial phases were obtained by molecular replacement with X-PLOR, Version 3.1 (21Bruenger A.T. X-PLOR Version 3.1: A System for X-ray Crystallography and NMR. Yale University Press, New Haven, CT1992Google Scholar) using our coordinates of LFA-1 I domain/lovastatin (11Kallen J. Welzenbach K. Ramage P. Geyl D. Kriwacki R. Legge G. Cottens S. Weitz-Schmidt G. Hommel U. J. Mol. Biol. 1999; 292: 1-9Crossref PubMed Scopus (234) Google Scholar) as a search model. Manual rebuilding was done with O (22Jones T.A. Zou J.-Y. Cowan S.W. Kjeldgaard M. Acta Crystallogr. Sect. A. 1991; 47: 110-119Crossref PubMed Scopus (13014) Google Scholar) and refinement with REFMAC5 (23Murshudov G.N. Vagin A.A. Dodson E.J. Acta Crystallogr. Sect. D Biol. Crystallogr. 1997; 53: 240-255Crossref PubMed Scopus (13914) Google Scholar). Five percent of the unique data were randomly selected for the calculation of Rfree. The two complexes in the asymmetric unit were refined without non-crystallographic restraints. The final refinement model includes two αL I domains (residues Gly128–Ile309), two LFA878 ligands, two magnesium ions, 148 water molecules, and has good stereochemistry (overall G value = 0.20; see Table I for crystal data, data collection, and refinement statistics) (24Laskowski R.A. MacArthur M.W. Moss D.S. Thornton J.M. J. Appl. Crystallogr. 1993; 26: 283-291Crossref Google Scholar).Table IX-ray data collection and refinement statisticsαl I domain/LFA878αl I domain/LFA703Crystal dataSpace groupP212121P43212Unit cell dimensions (Å)a = 71.9; b = 77.4; c = 92.3a = b = 116.7; c = 82.7No. complexes/asymmetric unit22Intensity data processingResolution (highest resolution bin) (Å)8-2.1 (2.17-2.10)8-2.2 (2.28-2.2)Rsym (%)aRsym=Σ|I-〈I〉|/ΣI, where I is observed intensity, and 〈I〉 is average intensity obtained from multiple observations of symmetry related reflections.6.7 (23.6)9.4 (28.6)No. of measurements123,226213,739No. of unique reflections29,63528,795Completeness (%)98.8 (99.8)99.8 (100.0)〈I/σ(I)〉22.5 (6.5)22.7 (7.8)Refinement statisticsResolution (highest resolution bin) (Å)8-2.1 (2.15-2.10)8-2.2 (2.25-2.20)Rcryst/Rfree (%)bRcryst=Σ‖Fobserved - Fcalculated‖/Σ Fobserved · Rfree is Rcryst, calculated by using 5% of the data, chosen randomly, and omitted from refinement.19.1/21.3 (23.3/27.7)16.1/20.6 (18.2/27.3)〈B〉 for protein/ligand (Å2)34.3/43.134.2/37.4Rmsd bond lengths (Å)cRmsd is root mean square deviation.0.0140.018Rmsd bond angles (°)cRmsd is root mean square deviation.1.381.51a Rsym=Σ|I-〈I〉|/ΣI, where I is observed intensity, and 〈I〉 is average intensity obtained from multiple observations of symmetry related reflections.b Rcryst=Σ‖Fobserved - Fcalculated‖/Σ Fobserved · Rfree is Rcryst, calculated by using 5% of the data, chosen randomly, and omitted from refinement.c Rmsd is root mean square deviation. Open table in a new tab Crystallization and Structure Determination of the αL I Domain/LFA703 Complex—The complex αL I domain/LFA703 was prepared as for LFA878. Crystals with maximal dimensions up to 0.3 mm could be grown by the hanging drop vapor diffusion method. One microliter of αL I domain/LFA703 was mixed with an equal volume of well solution (0.17 m ammonium acetate, pH 4.6, 0.085 m sodium acetate, 15% glycerol, and 26% (w/v) polyethylene glycol 4000) and allowed to equilibrate against 0.7 ml of the latter at 20 °C. Diffraction data were collected to 2.2 Å at 20 °C on a Mar345 imaging plate detector at the Swiss-Norwegian beamline of the European Synchrotron Facility (Grenoble, France). Data processing, rebuilding, and refinement were done as for LFA878. The final refinement model includes two αL I domains (residues Gly128–Ile309), two LFA703 ligands, two magnesium ions, 278 water molecules, and has good stereochemistry (overall G value = 0.02; see Table I). LFA-1 Binding Assays—The cell-free LFA-1/ICAM-1 binding assay quantifies the binding of biotinylated recombinant ICAM-1 to immobilized LFA-1. The HUT78/ICAM-1 adhesion assay quantifies the binding of fluorescently labeled HUT78 cells to immobilized recombinant ICAM-1. Both assays were performed as described previously (11Kallen J. Welzenbach K. Ramage P. Geyl D. Kriwacki R. Legge G. Cottens S. Weitz-Schmidt G. Hommel U. J. Mol. Biol. 1999; 292: 1-9Crossref PubMed Scopus (234) Google Scholar). Anti-LFA-1 mAbs Binding to Jurkat Cells—The effect of the compounds on the conformational status of LFA-1 was tested as described previously (17Welzenbach K. Hommel U. Weitz-Schmidt G. J. Biol. Chem. 2002; 277: 10590-10598Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar). Briefly, Jurkat cells were pre-incubated with the compounds at a final concentration of 50 μm in Tris-buffered saline containing 150 mm NaCl, 2 mm MgCl2, 2 mm MnCl2, and 0.5% bovine serum albumin, pH 7.4 (assay buffer) at room temperature (RT) for 20 min followed by the addition of the anti-LFA-1 mAbs. After 25 min at RT, cells were washed two times with assay buffer and counterstained with Alexa Fluor® 488 goat anti-mouse IgG (H+L) conjugate (Molecular Probes, Leiden, Netherlands) diluted 1:175 in assay buffer for 30 min at RT. After a washing step, antibody binding was immediately analyzed by flow cytometry on a FACScan (BD Biosciences). Mean fluorescence intensities (geometric mean) were calculated using the CellQuest software. Mean fluorescence intensities were corrected for background staining by subtracting the mean fluorescent intensity of the appropriate isotype-matched negative controls. Mean fluorescence intensities of the solvent controls were set as 100%. Inhibition of anti-LFA-1 mAb binding induced by inhibitor treatment was expressed as a percentage of these controls. Whole Blood LFA-1 Reporter Epitope Monitoring Assay (LFA-1 REMA)—The LFA-1 inhibitors were dissolved and diluted in Me2SO. Heparinized human blood samples or blood samples from different animal species (198 μl) were mixed with the compound solution or Me2SO (2 μl) and incubated for 25 min at RT. The compound containing blood samples (90 μl) were then transferred to 96-well microtiter plates. The fluorescein isothiocyanate-conjugated mAb R7.1 (BioSource, Camarillo, CA) and the isotype control mAb were diluted in phosphate-buffered saline, and 10 μl was added to the samples (final concentration was 2–3 μg of mAb per milliliter of blood). After 25 min at RT, the blood samples (100 μl) were transferred to tubes containing 1 ml of FACS lysing solution (BD Biosciences), mixed, and incubated for 10 min in the dark at RT. Then the samples were centrifuged at 200 × g for 5 min. The supernatants were removed, and the pellets were washed two times in 1 ml of Tris-buffered saline (pH 7.4) containing 0.5% bovine serum albumin and then resuspended in 150 μl of the same buffer. Bound antibody was detected by flow cytometry gating the major leukocyte subpopulations according to their light scatter properties. In each sample, 20,000 lymphocytes were counted. In control experiments, unlabeled mAb TS2/4 was diluted and added to the blood samples (final concentration was 2.5 μg of mAb per milliliter of blood). After lysis and washing steps the Alexa Fluor® 488 goat anti-mouse IgG conjugate diluted 1:150 in Tris-buffered saline, pH 7.4, containing 0.5% bovine serum albumin was added. The samples were incubated for 30 min, and bound mAb TS2/4 was quantified by flow cytometry as described. Ex Vivo Rabbit LFA-1 REMA—Female Russian dwarf rabbits received 0.1 ml/kg Prequillan (10 mg/ml) subcutaneously. Compounds dissolved in a mixture of cremophor EL and ethanol (2:1) (w/w) and then diluted further with 5% glucose (1:3) (v/v) were administered intravenously by bolus injection (right ear; 1.5 ml per rabbit). Blood samples (200 μl) were taken from the left ear at indicated time points and placed into microtubes containing heparin. The blood samples were stored on ice until analysis. Ninety microliters of each sample were transferred to microtiter plates, and REMA analysis was performed as described above. Baseline values were determined in blood samples taken 15 min before compound administration. The animal experiments were conducted in accordance with the animal experimentation guidelines and laws laid down by the Swiss Federal and Cantonal Authorities. Carrageenan-induced Paw Edema in Rats—Male OFA rats were treated orally with LFA878 (3, 10, and 30 mg/kg), diclofenac (3 mg/kg), or the respective vehicle controls (vehicle for LFA878 was a mixture of cremophor EL and ethanol (2:1) (w/w) diluted 1:3 (v/v) with 5% glucose; vehicle for diclofenac was Tween 80/0.75% methylcellulose). The anti-LFA-1 mAb WT.1 (NatuTec, Frankfurt, Germany) was diluted in phosphate-buffered saline and administered at a concentration of 2.5 mg/kg. One hour later the rats received a 100-μl intra-plantar injection of a 1% (w/v) carrageenan solution in 0.9% saline in the hind paw, and the volume of the paw was measured by plethysmography. The paw volume measurements were repeated at 3 and 5 h after injection of the carrageenan. Percentage inhibition of paw swelling at 3 and 5 h was calculated by reference to vehicle-treated animals (0% inhibition). The animal experiments were conducted according to the animal experimentation guidelines and laws laid down by the Swiss Federal and Cantonal Authorities. X-ray Structures of the Complexes αL I Domain/LFA703 and αL I Domain/LFA878 Identify Novel Subpockets within the L-site—The compounds LFA703 and LFA878 are two examples of a series of lovastatin-derived inhibitors that inhibit LFA-1 function more potently than lovastatin itself (Fig. 1). Both compounds blocked LFA-1 mediated binding to ICAM-1 with IC50 values in the low nanomolar range in cell-free and cell-based in vitro assays, whereas lovastatin was only active in the micromolar range (Fig. 1). To determine the structural basis for this improved activity, we solved the crystal structures of the I domain portion of LFA-1 bound to LFA703 or LFA878 at resolution limits of 2.2 and 2.1 Å, respectively (Fig. 2). The crystal form for LFA878 is very similar to the one reported previously for lovastatin (with a local dyad relating the two complexes in the asymmetric unit) (11Kallen J. Welzenbach K. Ramage P. Geyl D. Kriwacki R. Legge G. Cottens S. Weitz-Schmidt G. Hommel U. J. Mol. Biol. 1999; 292: 1-9Crossref PubMed Scopus (234) Google Scholar), whereas the form for LFA703 is new and different (no local dyad relating the two complexes). Interestingly, for LFA703 the C-terminal residues of the I domain monomers form parallel β-sheet interactions with the respective residues from a neighboring molecule, in contrast to the antiparallel β-sheet interactions in the LFA878 or lovastatin complexes. This finding demonstrates the structural adaptability of the linker region after the helix α7 and suggests possible structural changes occurring upon the activation of LFA-1. Moreover, a new type of interaction with the MIDAS site is observed in the crystal form for LFA703. The Mg2+ in the MIDAS itself is coordinated in an identical way as that seen for the other αL I domain/statin-based inhibitor complexes or the uncomplexed αL I domain (25Qu A. Leahy D.J. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 10277-10281Crossref PubMed Scopus (290) Google Scholar). In all the latter cases (including the complex with LFA703) the I domain is in its low affinity form, which is characterized (among other features) by a direct coordination of Mg2+ by Asp239. The direct coordination is thought to reduce the electrophilicity of Mg2+ and, thus, to weaken the interaction with Glu34 of ICAM-1 (3Shimaoka M. Xiao T. Liu J.H. Yang Y. Dong Y. Jun C.D. McCormack A. Zhang R. Joachimiak A. Takagi J. Wang J.H. Springer T.A. Cell. 2003; 112: 99-111Abstract Full Text Full Text PDF PubMed Scopus (424) Google Scholar). This hypothesis is now corroborated with the new tetragonal crystal form of the αL I domain/LFA703 complex, which does not show a direct interaction but only a water-mediated, indirect interaction between the Mg2+ (protomer B) and a glutamate residue (Glu218) from a neighboring protomer A (Fig. 2F).Fig. 2X-ray structures of the complexes αL I domain/LFA703 and αL I domain/LFA878. A, stereo image of the L-site for αL I domain/LFA703. I domain residues and ribbon representation are in white, LFA703 in cyan (oxygens are red, and nitrogens are blue). The substituted naphthyl group of LFA703 occupies a region (the naphthyl subpocket, colored magenta) formed mainly by Val130, Thr231, Val233, and Ile255. B, stereo image of the L-site for αL I domain/LFA878 (coloring as for LFA703). The veratryl group of LFA878 occupies a region (the veratrylsubpocket, colored magenta) formed mainly by Tyr257, Glu284, and Phe285. A comparison with the I domain/LFA703 complex shows that Glu284 has dramatically changed its side-chain conformation. C, Fo – Fc electron density (contour level 3σ, 8–2.2 Å) before LFA703 was inserted into the model. Superposed is the final model of LFA703 (carbons are cyan/yellow, oxygens are red, and nitrogens are blue). The naphthyl group adopts two alternate conformations. D, Fo (contour – Fc density level 3 electron σ, 8–2.1 Å) before LFA878 was inserted into the model. Superposed is the final model of LFA878 (coloring as for LFA703). E, superposition of LFA703 (carbons are yellow), LFA878 (carbons are cyan), and lovastatin (carbons are white) using Cα atoms of the respective αL I domains. The decalin moieties occupy practically identical positions. F, MIDAS site for αL I domain/LFA703. Final 2Fo – Fc electron density (contour level 3σ, 8–2.2 Å) is shown in blue; carbons are yellow, oxygens are red, nitrogens are blue, water molecules are white, and the Mg2+ ion is cyan. Selected hydrogen bonds are shown as black lines (dashed lines for the interaction with neighboring molecule). Because of the direct coordination by Asp239, the electrophilicity of Mg2+ is reduced so that a glutamate (Glu218, colored magenta) from a neighboring molecule is only interacting indirectly (via a water molecule) with the Mg2+ in the MIDAS site. Single letter amino acid abbreviations are used with position numbers throughout the figure.View Large Image Figure ViewerDownload (PPT) Both LFA703 and LFA878 were found to bind to the L-site of the αL I domain (Fig. 2, A and B) and displayed a clear difference electron density before, present in the models (Fig. 2, C and D). The decalin moieties of the statin derivatives adopt very similar positions within the respective x-ray structures even though different interactions between the I domain monomers occur at the carboxyl termini of the different crystal forms (Fig. 2E). The main contacts between LFA703 and the L-site (distance cutoff of 4.2 Å between non-hydrogen atoms) are mediated by the residues Gly128, Val130, Leu132, Phe153, Tyr166, Thr231, Val233, Ile235, Ile255, Tyr257, Glu284, Phe285, Lys287, Leu298, Glu301, Leu302, Lys305, and Ile306. Interestingly, the carbonyl oxygen of the isobutyl-ester moiety attached to the decalin ring of LFA703 contributes a water-mediated hydrogen bond to both NZ-Lys287 and OE1- or OE2-Glu284. The substituted naphthyl of LFA703 adopts two alternate conformations (for which the naphthyl groups are in the same plane), allowing the hydroxy-methylene group of the naphthyl in one conformation to form a hydrogen bond with OG1-Thr231 (Fig. 2, A and C). This region within the L-site is not reached by LFA878 or lovastatin and constitutes a new L-site subpocket, designated here the "naphthyl subpocket," formed mainly by the side chains of Val130, Thr231, Val233, and Ile255. A least squares superposition of the αL I domains in the complexes with LFA703 and lovastatin shows that the protein backbones are similar (e.g. the root mean squares deviation for the Cα-atoms of residues Gly128–Lys304 from the respective protomers A is 0.36 Å, and from the respective protomers B it is 0.52 Å) but display significant changes at the carboxyl-terminal residues Lys305–Ile309, which are involved in
Referência(s)