Regioselective nitration of tryptophan by a complex between bacterial nitric-oxide synthase and tryptophanyl-tRNA synthetase

Bacterial nitric-oxide synthase proteins (NOSs) from certain Streptomyces strains have been shown to catalyze biosynthetic nitration of tryptophanyl moieties in vivo Nature 429, 79-82]. We report that the complex between D. radiodurans NOS (deiNOS) and an unusual tryptophanyl tRNA synthetase (TrpRS II) catalyzes the regioselective nitration of tryptophan (Trp) at the 4-position. Unlike, non-enzymatic Trp nitration, and similar reactions catalyzed by globins and peroxidases, deiNOS only produces the otherwise unfavored 4-nitro Trp isomer. Although deiNOS alone will catalyze 4-nitro-Trp production, yields are significantly enhanced by TrpRS II and ATP. 4-nitro-Trp formation exhibits saturation behavior with Trp (but not tyrosine) and is completely inhibited by the addition of the mammalian NOS cofactor tetrahydrobiopterin (H4B). Trp stimulates deiNOS oxidation of substrate L-arginine (Arg) to the same degree as H4B. These observations are consistent with a mechanism where Trp, and possibly adenyl-Trp formed by TrpRS II, binds in the NOS pterin site, participates in Arg oxidation, and subsequently becomes nitrated at the 4-position.


INTRODUCTION
The generation and fate of reactive nitrogen species (RNS) are important to a variety of physiological processes that include pathogen cell death, general progression towards disease states, and stimulation of regulatory pathways (1)(2)(3). For example, nitrated aromatic protein residues mark the involvement of RNS in neurodegenerative diseases, artherosclereosis, infections, inflammation and cancer (1)(2)(3). Globins and peroxidases have been shown to catalyze complex reactions that can result in nitration of aromatic amino acids (3)(4)(5)(6)(7). In mammals, a primary source of RNS is nitric oxide (NO), which is largely produced by the three isozymes of nitric-oxide synthase (NOS) (8). The benefits and repercussions of generating NO in a cellular environment are complex and far from well understood (1)(2)(3)(4)(5)(6)(7).
Whereas mammalian NOS-mediated nitration events are generally non-specific, a bacterial nitric-oxide synthase protein has recently been shown to participate in a specific biosynthetic nitration reaction (9). Some pathogenic Streptomyces produce a family of unusual phytotoxin dipeptides (derivatives of cyclo-[L-tryptophanyl-L-phenylalanyl]) called thaxtomins that contain a 4-nitro-tryptophanyl moiety (10). The transferable pathogenicity island that contains the genes responsible for thaxtomin biosynthesis also codes for a NOS (9). Genetic and isotope labeling studies have shown that the Streptomyces NOS nitrates thaxtomin (9). The chemical mechanism of a NOS-mediated nitration is intriguing because NO is unlikely to react directly wiith indole (11,12). Nevertheless, readily oxidized forms of NO, such as nitrosonium (NO + ), peroxynitrite (ONOO -), nitronium (NO 2 + ), or nitrogen dioxide (NO 2 ) actively nitrate aromatic amino acids (11,12).
Another link between tryptophan metabolism and bacterial NOS comes from our recent finding that in D. radiodurans, an unusual tryptophanyl tRNA synthetase (TrpRS II) (13). First, the amino acid is activated to the aminoacyl-5'-adenylate, which is then reacted with the 3'-acceptor stem of tRNA to form the aminoacyl-tRNA. Adenylation of aminoacids also precedes their condensation into natural products by non-ribosomal peptide synthases, such as the enzymes that produce thaxtomin (10

MATERIALS AND METHODS
Preparation of deiNOS and TrpRS II -DeiNOS, was cloned with a N-terminal His 6 -affinity tag, expressed in E. coli BL21(DE3) cells and purified by nickel-NTA affinity chromatography (Qiagen) as previously described 2 (14).
Notably, to improve solubility, deiNOS was eluted from the nickel column with 25 mM HEPES  (14)) and in the presence and absence of 30 µM ATP.
Non-enzymatic nitration of tryptophan -Hydrogen peroxide and nitrite were used to generate nitrating agents, which primarily include peroxynitrite (4,21). Hydrogen peroxide (10 mM in 0.1M potassium phosphate buffer, pH 6.9) was treated with sodium nitrite (10 mM in 0.1M potassium phosphate buffer, pH 6.9) and vortexed for about 1 min. The vortexed mixture was added to L-tryptophan (10 mM) and the reaction mixture was incubated at 37°C for about 1 hr and the products were evaluated using reverse-phase HPLC, mass-spectroscopy and UV-Visible spectroscopy.
Purification of nitration products -The reaction mixture was passed through Supelco's Bio Wide Pore C18-5 column (25 cm X 4mm) with pore size of 300 Å. A gradient was run using buffers A (0.07 % TFA in water) and B ( 0.07 % TFA in acetonitrile) and the products collected, NOS assays -Nitrite formation by deiNOS was monitored using the Griess reagents, whereas NO production was followed with the oxy-haemoglobin assay (14,22).  (Fig. 1A), massspectroscopy (Fig. 1B), and optical spectroscopy (Fig. 1B). Under all conditions tested, the reaction produced one nitro-Trp isomer, which elutes on a reverse-phase HPLC column at the same retention time as a synthetic 4-nitro-Trp standard (Fig. 1A). Isomers of nitro-Trp can be readily differentiated by their chromatographic behavior and their far-UV/visible absorption spectra (4,5). The product was further confirmed to be 4-nitro-Trp by its characteristic absorption spectrum (Fig. 1B) and mass spectrum (Fig. 1B), whose parent ion and fragments also match the 4-nitro-Trp standard. Consistent with previous reports of non-specific Trp nitration (4-6,21) reactions of Trp with nitrite and peroxide (a peroxynitrite generating system (4,21)) produced primarily a product whose much longer HPLC retention time and mass-spectrum identify it as N 1 -nitro-Trp (not shown), and a minor product (Fig. 1A) with a mass and elution profile consistent with the favored 6-nitro Trp isomer (4,5). Comparisons between HPLC elution areas of the 4-nitro-Trp product and absorption changes in the reaction mixture indicated that 4-nitro-Trp formation accounted for at least 80% of the absorbance change at 400 nm (∆A 400 ).

Production of 4-Nitro-tryptophan by deiNOS-TrpRS
No other small molecule products with far UV/visible absorption were observed in appreciable quantities. Thus, kinetics of Trp nitration were followed by monitoring (∆A 400 ). Addition of 50 U/mL of superoxide dismutase also did not significantly reduce the production of nitro-Trp by deiNOS in either the reaction with peroxide or with mNOS red .

Stimulation of 4-nitro formation by TrpRS II and ATP -
Addition of TrpRS II stimulates the production of 4-nitro Trp by a factor of 2-3 whether peroxide or NOS red /NADPH drives the reaction (Table 1). In parallel studies, we find that TrpRS II increases the affinity of deiNOS for Arg and stimulates nitrite production from NHA in oxygenated solution 2 . Surprisingly, the peroxide catalyzed Trp nitration reaction enhances further by the addition of ATP (Table 1). Consistent with greater incorporation of [NO] into the 4nitro-Trp product, the amount of nitrite produced decreases as the amount of 4-nitro-Trp increases when ATP is present. ATP has no effect on 4-nitro-Trp or [NO] formation by deiNOS alone. ATP binds tightly to TrpRS II 2 and is required for Trp-adenylation by the synthetase.
The effects of pterin on 4-nitro-Trp and NO production -  (8,18). Surprisingly, in the absence of H 4 B, but presence of Trp, nitrite is still produced from Arg at nearly equal amounts (despite 4-nitro-Trp also being formed) (Table 1). suggests that Trp binds in the deiNOS pterin site and furthermore can stimulate oxidation of Arg from this position. We were unable to detect NO production from deiNOS-TrpRS II in the presence of Trp by the oxy-haemoglobin assay; however, the amount of NO expected from the low deiNOS concentrations required to monitor oxy-haemoglobin conversion were at the detection limit. NO production from Trp is of interest because H 4 B-depleted mNOS will still produce nitrite in oxygenated solution but does so through nitroxyl (NO -), rather than NO formation (22,23).

Trp nitration by deiNOS exhibits saturation behavior -
The dependence of 4-nitro Trp formation by deiNOS on Trp concentration exhibits saturation behavior with an observed Michaelis constant (K M ) of 2.8 +/-0.6 mM (Fig. 2). Addition of TrpRS II increases the effective V max 2-3 times and decreases K M to 1.5 +/-0.5 mM (Fig. 2).
Such saturation kinetics also indicate that 4-nitro-Trp production involves Trp binding to deiNOS. A different concentration dependence would be expected for reaction of free Trp with a diffusing RNS generated by deiNOS. We also tested inhibition of Trp nitration by the peroxynitrite scavanger cysteine (21). Although 1 mM cysteine completely prevented nitro-Trp formation from nitrite and peroxide, 4-nitro-Trp formation by deiNOS decreased only slightly (15%) in the presence of cysteine (Table 1). Tyrosine, the more reactive amino acid for nitration (24), is also nitrated by deiNOS, but with lower turnover numbers than Trp. At the Tyr concentrations we could evaluate, the reaction did not saturate and appeared to be roughly first order (Fig. 2).

DISCUSSION
The regioselective nitration reaction catalyzed by deiNOS that produces only 4-nitro-Trp is in marked contrast to non-enzymatic and heme protein catalyzed Trp nitration reactions.
Agents capable of nitrating aromatic aminoacids include peroxynitrite, nitrous acid, nitrite and hydrogen peroxide, nitric oxide and dioxygen, and acyl nitrates (4,11,12,24  Saturation behavior of the nitration reaction with Trp concentration also supports Trp binding to deiNOS. Neither saturation kinetics nor comparably high product yields are observed with Tyr, which is generally at least as reactive towards nitrating agents as Trp (24).

Inhibition of 4-nitro Trp formation by H 4 B indicates that rates saturate with Trp concentration
because Trp binds in the pterin site. In bsNOS, H 4 B accelerates decay of the heme ferrous-oxy species by presumably acting as a rapid electron donor to heme, akin to its role in mammalian NOS catalysis (8,17,18,27). The dependence of deiNOS on reduced pterins (15) or Trp for [NO] formation from Arg could imply that deiNOS maintains this mechanism, but that Trp can substitute for H 4 B as an electron donor. This is surprising, given that the reduction potential of the Trp radical (1.0 V, (28)) is expected to be much higher than that for H 3 B (0.15 -0.3 V, (29) and references therein). Nonetheless, the protein and or protonation environment could modulate these potentials (27) and indeed, NOSs do provide a negative electrostatic field at the pterin site that will faciliate oxidation of cofactors (18). Alternatively, Trp could be oxidized by the compound I-like oxidation state of NOS, which follows reaction of two-electron reduced oxygen (or peroxide) at the heme center. Peroxidases are capable of this reaction (4), but it is difficult to rationalize why Trp stimulates nitrite production from deiNOS-NOS red if Trp participation in the reaction is limited to steps after oxygen activation. with the radical proceeded prior to [NO] reaction with oxygen, a 4-nitroso-Trp would be a likely initial product (Fig. 3). As C-nitroso-Trp may be unstable to both oxidation and nitrosomigration (24,30), it is not surprising no such species was observed.
TrpRS II activates deiNOS for production of [NO] and for production of 4-nitro-Trp. We have found that these proteins tightly associate in vitro and in cell lysates of D. radiodurans 2 .
Significantly, TrpRS II activation of deiNOS nitration increases in the presence of ATP. TrpRS II requires ATP to adenylate Trp prior to tRNA charging. deiNOS itself does not interact with ATP, nor does ATP affect nitro-Trp or [NO] production of deiNOS alone. Thus, 5'-adenyl-Trp, produced by TrpRS II may be a preferred substrate for deiNOS nitration (Fig. 3). The 5' adenyl-Trp is unstable to hydrolysis and difficult to isolate as an intermediate (13). More productive binding of 5'-adenylate compared to Trp may explain the apparent activation seen with ATP.
Alternatively, ATP binding to TrpRS II may influence deiNOS in a manner that stimulates Trp nitration. The ability of TrpRS II and ATP to enhance the deiNOS nitration reaction underscores the functional coupling of these two enzymes.
Many enzymes involved in secondary metaboilsm, such as non ribosomal peptide synthases, polyketide synthases, terpene synthases and taxadiene synthases have relatively slow turnover numbers that often range between 0.001-0.