The absolute configuration and the enantiomeric excess of the product 6a were determined as (R)-(-)-6a and 94.0% ee by chemical correlation with (R)-(-)-2-methyl-2-phenylsuccinic acid (12) of known configuration as follows. Sulfenylation of the ketone (-)-6a obtained above with diphenyl disulfide followed by sodium borohydride reduction of 7 produced an aphenylthio alcohol 8. Alcohol 8 upon treatment with lead tetraacetate in toluene-acetic acid (4:1) at 0 OC for 8 h underwent an oxidative cleavage to give a thioacetal acetate 9. Hydrolysis of this acetate 9 with potassium hydroxide in methanol at room temperature gave a hemiacetal 10. Oxidation of this hemiacetal 10 with chromic acid in aqueous sulfuric acid-acetone at 0 OC produced 2-methyl-2-phenylsuccinic anhydride (11), which upon hydrolysis with potassium hydroxide in refluxing methanol gave (R)-(-)-12. The reaction sequences starting with ethyl methyl ketone (2b) were successfully executed in the same way. Addition of the a-carbanion of (Rs)-(+)-1 ((00% ee) to 2b gave (Ss)-3b in 72% yield (the diastereomers of 3b (ratio 3:2) were unseparable). The thermal rearrangement of (Ss)-3b thus obtained was carried out by treatment with p-toluenesulfonic acid in refluxing benzene for 3.5 h to furnish a cyclobutene derivative (Ss)-4b in 65% yield. Reduction of the sulfoxide (Ss)-4b with acetyl chloride followed by hydrolysis of the enol thioether (-)-5b produced (S)-(-)-2-ethyl-2-methylcyclobutanone (6b). The absolute configuration and the enantiomeric excess of the product 6b were determined as (S)-(-)-6b and 73.3% ee by transformation of 6b into 4-methyl-4-hexanolactone (13) of known configuration; Baeyer-Villiger oxidation of (-)-6b (H2O2-NaOH in aqueous methanol) followed by lactonization by heating in refluxing benzene with a catalytic amount of p-toluenesulfonic acid led to (S)-(-)-13. On the basis of the above experimental results, the asymmetric inductions in these thermal 1,2-rearrangements of 3a,b to 4a,b were determined to give 94.0% and 73.3% optical yields, respectively. From these stereochemical results, the mechanistic pathway for this asymmetric induction would be represented as follows. In the acid-catalyzed thermolysis, the carbonium ion 14 would be formed initially. The 1,2-migration of a carbon-carbon bond of the cyclopropane ring would occur via a transition state 15, and a new asymmetry would be induced at this stage. The degree of asymmetric induction would depend on the difference between the thermodynamical stability of 15a and 15b, that is, on the difference of the steric interference between R' or R2 and the lone pair or the oxygen atom of the chiral sulfoxide (Scheme III). The easy access to the starting chiral sulfoxide and the high degree of asymmetric induction in this thermal rearrangement represent a potentially great advantage for the construction of asymmetric quaternary carbons. Furthermore, this method provides a facile entry to chiral cyclobutane derivatives, which have usually been hard to access. Nosiheptide (1), a metabolite of Streptomyces actuosus, is a member of a broader class of highly modified, sulfur-rich peptide antibiotics, which also includes thiostrepton, micrococcin, the thiopeptins, and several other compounds. Compound 1 inhibits protein synthesis in Gram-positive bacteria by binding to the 50s ribosomal subunit; it is used as an animal-feed additive to increase weight gains. Nosiheptide contains several structural elements with biosynthetic origins that were of interest to us, notably a 2,3,4-trisubstituted indole, five thiazole rings, and a trisubstituted pyridine. Cultures of S. actuosus were fed labeled substrates after 32 h of growth, and 1 was extracted from the mycelium 48-72 h later and purified (precipitation from tetrahydrofuran/hexane, CH2Cl2/EtOH, HPLC) for radioactivity (Beckman LS 7500) or 13C NMR (Bruker WM-300, 7.1 T, Me2SO-d6) analysis. The results of the stable isotope experiments are summarized in Table I and Figure 1. The structure of the thiazole rings suggests origin from cysteine, which provides the sulfur, the nitrogen, C-5, C-4, and the attached carboxyl group; C-2 comes from the carboxyl group of another amino acid. Consistent with this notion, DL-[3-13C]Cysteine (97% 13C, 50 mg/L) labeled C-5 of the thiazole rings. C-3 of the D-cysteine moiety was also labeled but not C-3 of the dehydroalanine portion. The latter is not derived from alanine (1.3% specific incorporation for L-[u-14C]alanine vs. 37% for L-[U-14C]serine). Its origin from serine is confirmed by labeling of C-1 from DL-[1-13C]serine (99% 13C, 200 mg/600 mL) and C-3 from L-[3-13C]serine (93% 13C, 126 mg/600 mL). C-1 of serine, as expected, labeled the carboxyl carbons attached to C-4 of the five thiazole rings. One of these forms part (C-6) of the pyridine moiety. C-3 of L-serine labeled all the carbons enriched by C-3 of DL-cysteine except, notably, C-3 of the D-cysteine moiety. Since C-2 of the thiazole(4), the carboxyl carbon of this D-cysteine moiety, is labeled by DL-[1-13C]serine, one must conclude that the D-cysteine moiety is efficiently derived from D- but not L-serine, presumably via D-cysteine. The pyridine ring is formed, rather uniquely, from the carboxyl group of one cysteine (C-6) and from two molecules of serine, which are connected 'tail to tail', i.e., through their methylene carbons, to form the C-3/C-4 connectivity (Figure 2). Our initial working hypothesis for the origin of the indole moiety was cyclization of phenylalanine and methylation at C-3 and C-4. The observed labeling of both indole C-3' and C-4' by C-3 of serine is consistent with this idea. However, L-[methyl-13C]methionine (90% 13C, 200 mg/L) labeled exclusively C-4' and hence, only the -CH2O group at indole C-4 is derived by methylation. It is noted, however, that the 2-carboxyl group of the indole moiety was labeled extensively by C-1 of serine. Since the alanine side chain of tryptophan is derived from L-serine, this suggests that the indole moiety may arise by cyclization of tryptophan, connecting indole C-2 with the carboxyl group, followed by excision of the side-chain carbon atom 2 plus its attached nitrogen and methylation of indole C-4 (Figure 2). Consistent with this hypothesis, DL-[7a-13C]tryptophan (0.5 mmol/L, 10% and 8% specific incorporation), L-[methylene-14C]tryptophan (0.5 mmol/L, 7% and 12% specific incorporation) and [2-14C]indole (1 mmol/L, 13 and 8% spec. incorp.) were efficiently incorporated into 1. Nonincorporation of DL-4-methyl[methylene-14C]tryptophan suggests that methylation of the indole is not the first step in the reaction sequence. On the basis of the above results and reasonable extrapolations, one may speculate that 1 arises from a dodecapeptide H2N-L-Ser-L-Cys-L-Thr-(L?)-Thr-L-Cys-L-Glu-L-Cys-D-Cys-L-Cys-L-Ser-L-Cys-L-Ser-COOH through connection of the carbon atoms 3 of ser(3) and ser(12) and attachment of a (modified) tryptophan. This paper reports the first gas-phase observation and infrared spectral characterization of Cr(CO)4(C2H4)2. This complex is unstable and reacts with CO by dissociative substitution. We follow the kinetics of this process by time-resolved IR absorption spectrometry, extracting a unimolecular decay constant orders of magnitude larger than the reported solution value for Cr(CO)4(η4-butadiene), in an apparent conflict with elementary theory as cited above.