Inducible nitric oxide synthase: Identification of amino acid residues ...

Proc. Natl. Acad. Sci. USA Vol. 92, pp. 11514-11518, December 1995 Biochemistry

Inducible nitric oxide synthase: Identification of amino acid residues essential for dimerization and binding of tetrahydrobiopterin HEARN J. CHO*t, EMIL MARTIN*t, QIAO-WEN XIE*, SHIGERU SASSAt,

AND

CARL NATHAN*

*Beatrice and Samuel A. Seaver Laboratory, Department of Medicine, Cornell University Medical College and the tRockefeller University, New York, NY 10021

Communicated by Seymour J. Klebanoff University of Washington, Seattle, WA, September 5, 1995 (received for review July 17, 1995)

Nitric oxide synthases (NOSs) require tetABSTRACT rahydrobiopterin (BH4) for dimerization and NO production. Mutation analysis of mouse inducible NOS (iNOS; NOS2) identified Gly-450 and Ala-453 as critical for NO production, dimer formation, and BH4 binding. Substitutions at five neighboring positions were tolerated, and normal binding of heme, calmodulin, and NADPH militated against major distortions affecting the NH2-terminal portion, midzone, or COOH terminus of the inactive mutants. Direct involvement of residues 450 and 453 in the binding of BH4 is supported by the striking homology of residues 448-480 to a region extensively shared by the three BH4-utilizing aromatic amino acid hydroxylases and is consistent with the conservation of these residues among all 10 reported NOS sequences, including mammalian NOSs 1, 2, and 3, as well as avian and insect NOSs. Altered binding of BH4 and/or L-arginine may explain how the addition of a single methyl group to the side chain of residue 459 or the addition of three methylenes to residue 453 can each abolish an enzymatic activity that reflects the concerted function of 1143 other residues.

low-spin state (17). In the present study, we have identified residues critical for the binding of BH4 within one of the longest regions of conservation among reported NOS sequences (18-20).

MATERIAL AND METHODS

The radical nitric oxide (NO) has emerged as an important signaling and cytotoxic molecule in metazoan physiology (1-4). NO is synthesized from L-Arg, oxygen, and NADPH by variably regulated isoforms of NO synthase (NOS). Products of the three known mammalian NOS genes (4) include two (NOS1 and NOS3) that are constitutively expressed and activated by binding calmodulin in response to elevated Ca2+ and one [inducible NOS (iNOS); NOS2] that is activated transcriptionally and binds calmodulin without an elevation of Ca2+ (5). These complex enzymes attach at least six molecules besides calmodulin (3, 6): L-Arg, heme, tetrahydrobiopterin (BH4), FMN, FAD, and NADPH. For only two of these cofactors or cosubstrates have residues important for binding been identified through mutational analysis. These are Cys184, which controls the binding of heme in human NOS3 (7) and corresponds to Cys-415, which controls the binding of heme in rat NOS1 (8, 9), and residues 1121-1144 in mouse iNOS, which contribute to the binding of NADPH (10). The binding site for BH4 is of particular interest for two reasons. (i) BH4 is essential for formation of stable NOS dimers in purified enzyme preparations (11, 12) and in cells (T. Billiar, personal communication), and dimerization is obligatory for generation of NO (11, 12). (ii) The BH4 binding site is likely to share, abut, or allosterically alter the hemecontaining catalytic site, since L-Arg and BH4 each increase the affinity for the other in NOS1 (13), L-Arg affects the Soret peak and CO binding properties of the heme in NOS1 and NOS2 (8, 9, 14,15), BH4 protects NOSl's heme from NO (16), and BH4 deficiency converts much of the heme in NOS1 and NOS2 from a pentacoordinate high-spin to a hexacoordinate

Mutagenesis and Transfection. Mouse iNOS expression plasmid piNOSL8 (10) was subjected to PCR mutagenesis to derive A4531 and C451A variants, and site-specific mutagenesis was used to derive A447I, G450A, P452A, and P461A. Mutations were confirmed by DNA sequencing. Human epithelial 293 cells (American Type Culture Collection) were transiently transfected by using the calcium phosphate method. Twenty-eight to 36 h later, iNOS activity was measured in cell lysates as described (10) and iNOS protein expression was monitored by SDS/PAGE (7.5% gel) and immunoblot analysis with rabbit IgG against pure wild-type mouse iNOS (21). The same IgG was used for all immunoblots and immunoprecipitations described below unless otherwise stated. Blots were developed with goat anti-rabbit IgG conjugated with horseradish peroxidase as detected by enhanced chemiluminescence (10). To produce stable transfectants, 293 cells were cotransfected with a mutant iNOS construct and the pcDNA3 vector (Invitrogen) bearing the neomycin-resistance gene. Cells growing in G418 (GIBCO; 600 ,ug/ml) were cloned by limiting dilution and screened for iNOS production by immunoblot analysis. Native iNOS was induced in RAW 264.7 macrophages activated by interferon y and bacterial lipopolysaccharide as described (11) and was used interchangeably with recombinant iNOS from piNOSL8 to serve as the wildtype control. Characterization of Mutant iNOSs. Cell lysate (100 ,ul) from transient transfectants was subjected to size-exclusion chromatography in buffer A (40 mM Bis-Tris propane, pH 7.2/50 mM NaCl/1 mM L-Arg/3 mM dithiothreitol/2 ,uM BH4/2 p.M FAD) on TSK (North Bend, WA) G3000SW or Superdex 200 columns (Pharmacia). Fractions were analyzed by SDS/PAGE and immunoblot analysis. To detect heme, iNOS was immunoprecipitated from transient transfectant lysates and the protein A-Sepharose-bound complexes were boiled in 2 M oxalic acid. This reaction generates from heme a stoichiometric amount of protoporphyrin IX, which gives a characteristic two-banded fluorescence emission spectrum with peaks at 606 and 662 nm (excitation, 400 nm) (22). To detect BH4, iNOS was partially purified from lysates of 293 cells stably transfected with constitutively expressed G450A and A4531 mutant iNOS genes by affinity chromatography on 2', 5'-ADP Sepharose (see below) (11) in the absence of exogenous BR4 and a 100-,lI sample was treated with 5 p.l of 0.1 M HCI and 5 p.l of 0.1 M KI/12 at room temperature for 1 h in the dark. Ascorbic acid (10 p.l, 0.1 M) was added, and

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Abbreviations: NO, nitric oxide; NOS, NO synthase; iNOS, inducible NOS; DHFR, dihydrofolate reductase; BH4, tetrahydrobiopterin. tH.J.C. and E.M. contributed equally to this work.

11514

Biochemistry: Cho et al. the precipitate was removed by centrifugation. The supernatant was subjected to C18 reverse-phase HPLC (Waters 600E), eluting pterins were detected by fluorescence as described (23), and peak areas were compared to those of standards. Binding of calmodulin was assessed by immunoprecipitating iNOS from lysates of transient transfectants (1.5 mg of protein), SDS/PAGE (15% gel), and immunoblot analysis with anti-calmodulin monoclonal antibody (0.1 ,tg/ml) (5). Binding of NADPH was assessed by affinity chromatography (10). Lysates of transient transfectants (2-4 mg of protein) were applied to a 2', 5'-ADP Sepharose column equilibrated with buffer A. Nonspecifically bound proteins were washed out with buffer A containing 0.5 M NaCl and the column was reequilibrated with buffer A containing 0.12 M NaCl. Proteins able to bind to NADPH were eluted with buffer A supplemented with 8 mM NADPH and examined by SDS/PAGE and immunoblot analysis.

Proc. Natl. Acad. Sci. USA 92 (1995)

11515

A I

AA

A

'. 7 t

tt

t

active mutants

iNOS

481

aRgGCPADW WI VPPvSGS TPVFHQEM nYvLsP

inactive mutants

A

I

B ,.J 4

CD °

Lo

<