Divergent synthesis of N -hydroxy-l-indospicine, the carbon isostere of N -hydroxy-l-arginine, and N -hydroxy-l-homoarginine from l-glutamate

The carbon isostere of N -hydroxy-L-arginine, N -hydroxy-L-indospicine, and N -hydroxy-L- homoarginine, were prepared by divergent synthesis as possible NOS substrates, from L-glutamate


Introduction
Biologically important nitric oxide (NO) 1 is biosynthesized by nitric oxide synthases (NOSs) during the conversion of L-arginine (1) into L-citrulline (3) through N-hydroxy-L-arginine (L-NOHA) (2). 2 We have reported that N-hydroxyagmatine (4), a decarboxylated NOHA, caused the relaxation of cardiovascular muscle, albeit less effectively than does NOHA itself 3 and that NO was evolved when N-hydroxyphenethylguanidine (5) was subjected to photo-sensitized oxidation in a model study for biological NO generation. 4 In this paper we present the divergent synthesis of N-hydroxy-L-indospicine (L-NOHI) (6), the carbon isostere of L-NOHA, and Nhydroxy-L-homoarginine (L-homoNOHA) (7), as alternative NOS substrates, from a Lglutamate derivative.

Results and Discussion
We planned to synthesize L-NOHI (6) and L-homoNOHA (7) from the L-glutamate derivative 8, through the norvaline derivative 9 as a common intermediate, as shown in Scheme 1.When we started our synthetic study, Feldman et al. 5 had just reported the preparation of L-indospicine itself from L-glutamate using a similar synthetic strategy, in which a protected L-NOHI 19 (see Scheme 3) was also prepared.Thus, we decided to follow their basic method for our purpose.After modification of the reported method, 6 hydroxynorvaline 15 was prepared from γ-benzyl (Bzl) N-butoxycarbonyl (Boc)-L-glutamate (12) by the three steps of esterification with N,Ndimethylformamide di-t-butylacetal, catalytic hydrogenation over Pd-C, and reduction with NaBH 4 after conversion into a mixed anhydride with ethyl ortho-chloroformate.The alcohol 15 was then converted into its mesylate 5 16 as a common precursor for the target compounds (Scheme 2).
At first, we focused on the preparation of L-NOHI (6) (Scheme 3).Feldman et al. 5 had reported the preparation of the cyanophenylsulfonyl hexanoate 17 from the mesylate 16 through an iodide.Although smooth conversion (>86%) of 16 to the iodide had been reported, we obtained an unsatisfactory result (24% yield).Examination of the reaction conditions led to an improvement of the yield to 63%, when ethanol was used as a solvent in place of acetonitrile.
However, a proline derivative -a cyclized product -was always produced as a side product, in ca 40% yield.Trials for the protection of the nitrogen function of 16 as a diBoc group failed.Fortunately, direct elongation of 16 with phenylsulfonylacetonitrile in THF in the presence of sodium hydride gave a desired cyanophenylsulfonyl hexanoate 17, quantitatively.Feldman et al. 5 reported successful desulfonylation of 17 with aluminum amalgam in a mixed solvent of THF-MeOH-H 2 O; however, in our case, no reaction was observed, even after modification of the reaction conditions.Thus, we examined the conditions for desulfonylation of 17 using either sodium amalgam or Raney nickel.Treatment with 5% sodium amalgam in a mixture of THF and MeOH (5:1) gave a nitrile derivative 18, albeit in moderate yield (28%).Oximation of 18 with hydroxylamine afforded a protected NOHI, 19, in nearly the same yield (30%) as reported. 5eprotection of 19 with hydrogen chloride in dioxane smoothly afforded a desired NOHI (6) hydrochloride as a colorless viscous oil, which was characterized by spectroscopic means such as HRMS and NMR.The optical rotation, [α] 589 18 , showed +13.1 (c 2.1x10 -3 , MeOH).Next, we tried to synthesize L-homoNOHA (7) from the mesylate 16 (Scheme 4).After conversion of 16 into a nitrile 20 by a conventional method, and reduction of the nitrile function to an amine, one method was examined under several conditions (NaBH 4 in the presence of CF 3 COOH or CoCl 2 , diborane, or catalytic hydrogenation using 5% Rh-Al 2 O 3 or PtO 2 as catalyst).A desired lysine derivative was produced when 20 was hydrogenated over 5% Rh-Al 2 O 3 in 10% NH 3 -EtOH under a pressure of 2.5 kg/cm 2 .Since isolation of the lysine had failed, the crude product was treated with cyanogen bromide without isolation to give the N-cyanamide 21 in 13% yield.

Conclusions
In conclusion, we have synthesized L-NOHI (6) and L-homoNOHA (7) from the L-glutamate derivative 12 through the hydroxynorvaline derivative 16 as a common synthetic precursor.The overall yields were 5% and 4.9%, respectively.NOHI has also been synthesized in a racemic form, 7,8 and the synthesis 8 of L-homoNOHA (7) from a lysine derivative was reported just before our work was finished.Thus, their possibilities as NOS substrates have been examined and weak activity had been observed on racemic NOHI, whereas L-homoHONA (7) showed significant NO production.Our products synthesized here will be tested independently to act as NOS substrates in our estimation system using aortic rinds. 3perimental Section General Procedures.All melting points were measured on a micro melting-point hot stage (Yanagimoto) and are uncorrected.IR spectra were recorded on a JASCO FT/IR-300E spectrophotometer.