On the basis of the results of labeling experiments, endogenous valine, glycine, and methionine are considered to be the precursors of 2-methoxy-3-isopropylpyrazine (MIPP) produced by Pseudomonas perolens. It is postulated that the feeding of [2-13C]pyruvate results in the labeling of C-2 and C-3 of valine and C-2 of glycine, which are subsequently incorporated into MIPP. The feeding of [3-13C]pyruvate results in the labeling of the two methyl groups of valine, C-1 and C-2 of glycine, and the methyl group of methionine. 13C NMR spectra of the MIPP produced by P. perolens in these feeding studies were consistent with the proposed biosynthetic route to MIPP and known metabolic pathways. Methyllabeled methionine arises from [2-13C]glyoxylate formed from [3-13C]pyruvate via the Krebs cycle and glyoxylate shunt. * Contribution No. 220 from the Natural Products Laboratory. This investigation was supported in part by Research Grant GM-13246 from the National Institutes of Health.

Capsaicin (N-(4-hydroxy-3-methoxybenzyl)non-írans-6-enamide) and 6,7-dihydrocapsaicin are found in approximately equal amounts in Capsicum frutescens. These alkaloids, isolated from the fruits of the plant 2 weeks after the administration of DL-phenyl-3-14C-alanine, were labeled solely on the methylene group of their vanillylamine residues. The administration of DL-tyrosine-3-14C afforded alkaloids with very low activity, l-Methionine-methyl-14C afforded radioactive alkaloids, and degradation of the dihydrocapsaicin indicated that all the activity was located on the methoxy group. No activity was found in the alkaloids when DL-mevalonic-2-14C acid or DL-leucine-l-14C were fed to the plant. However, radioactive capsaicin resulted when L-valine-U-14C was fed. Essentially all the activity was located at C-7, -8, -9, and -10 of the 8-methyl-6-nonenic acid moiety of the alkaloid. Capsaicin(1) and its dihydro derivative 2 occur in the fruit of Capsaicin species2 and are responsible for the hot taste of the peppers. In Capsicum annum Kosuge and coworkers3 found capsaicin and its dihydro derivative to be present in the ratio of 7:3. More recently Bennett and Kirby4 reported the following composition of the phenolic components of C. annum:(1) An account of this work was presented at the 154th National Meeting of the American Chemical Society, Chicago, 111., Sept 1967. This investigation was supported by a research grant, GB-3696, from the National Science Foundation.(2) S. Kosuge, Y. Inagaki, and K. Uehara, Nippon Nogei Kagaku Kaishi, 32, 578 (1958). These authors assigned the names capsaicin I and II to compounds 1 and 2, however we prefer the more descriptive names.

Illdole-3-aldehyde, indole-3-carboxylic acid and its ethyl ester were reduced by excess lithium alumir~um hydride to skatolc. The expected reduction product, 3-hydroxyrnethylindole, was obtained by the action of sodi~1111 hydroxide on grarnine rnethiodide. I t and its alkyl ethers were readily reduced to skatole. 3-Hydroxy~nethylindole ur~dernrent self-co~ldensation to 3.3'-di-indolylmethane in neutral and alkalille media, and with acidic reagents was collverted to a n oxygen-free polymeric substance. T h e mechanism of these reactions and of the hydrogeriolysis is discussed.In the course of the investigation of various indole alkaloids anomalous products were obtained on reduction of 3-substituted indole derivatives with lithium aluminum hydride. Thus the action of this reagent on simple indole compounds was investigated with the rather surprising result that indole-3-carboxylic acid, its ethyl ester, and indole-3-aldehyde were reduced in ether solution under very mild conditions to skatole. I t was also found that 3-acetylindole was reduced to 3-ethylindole.There have been a few isolated reports of the hydrogenolysis of acids and carbonyl compounds with lithium aluminum hydride to methyl or methylene groups. Witkop (40) found that an o-aminoketone (spirocyclopentane-1,2'-pseudoindoxyl) was reduced to an oxygen-free amine. Conover and Tarbell (7,8) showed that various benzene and thiazole acids and carbonyl derivatives underwent hydrogenolysis when there was an amino -group ortho or para to the group being reduced. Jones and Kornfeld (15) reported the hydrogenolysis of a carbethoxy group to a methyl in a pyridine derivative. Many of these reductions, however, were carried out for a prolonged time and a t a high temperature, and it was possible to isolate the expected alcohol by employing less vigorous conditions. But with 3-carbethoxyindole, even when less than the theoretical amount of lithium aluminum hydride to cause reduction to the alcohol was employed, the only products isolated were skatole and unreduced ester. I t was thus of interest to prepare by another method the 3-hydroxymethylindole which was apparently very readily reduced b17 lithium aluminum hydride. 'Manuscript

All the experimental evidence is consistent with the 1-methyl-A 1-pyrrolinium salt (30) being a precursor of the tropane nucleus. The various biosynthetic routes which have been established for its formation are summarized in Fig. 4. Initially it was thought that the double bond in the iminium salt could isomerize between its C-2 and C-5 positions, which would result in the scrambling of any isotopic label (e.g. 14C