Myoglobin was found in the nitrogen-fixing cyanobacterium Nostoc commune. This cyanobacterial myoglobin, referred to as cyanoglobin, was shown to be a soluble hemoprotein of 12.5 kilodaltons with an amino acid sequence that is related to that of myoglobins from two lower eukaryotes, the ciliated protozoa Paramecium caudatum and Tetrahymena pyriformis. Cyanoglobin is encoded by the glbN gene, which is positioned between nifU and nifH-two genes essential for nitrogen fixation-in the genome of Nostoc. Cyanoglobin was detected in Nostoc cells only when they were starved for nitrogen and incubated microaerobically.

No abstract

Ferrous-carbonyl complexes of the soluble Pseudomonas putida cytochrome P-450,,, and a denatured form (P-420) of this enzyme, as well as cytochromes P-450 and P-448 in liver microsomes from rats pretreated with phenobarbital and 3-methylcholanthrene, have been studied by infrared spectroscopy. The FeCO bonding was examined in order to probe the spatial relationship between the dioxygen-and substrate-binding sites and to determine whether the presence of a unique axial ligand bound trans to carbon monoxide could be responsible for the red-shifted Soret absorbance band maximum at 450 nm. The d-camphor(K+)-bound cytochrome P-450,,, yielded a single infrared absorbance band for the heme-bound carbonyl (uc-, 1940 cm-') having a bandwidth a t half-height ( A v l p ) of 13 ern-', while the camphor-free enzyme gave rise to two stretching frequencies of equal area a t 1963 and 1942 cm-' ( A u l / 2 11-12 and 19-21 cm-', respectively). Addition of d-camphor and a monovalent metal ion (K+) to the camphor-free ferrous carbonyl-enzyme converted the infrared spectrum back to that of the original camphor-bound enzyme. The area of the 1940-cm-I band was found to equal that of the combined areas of the 1963-and 1942-cm-l bands. Conversion of the native enzyme to a denatured form (P-420) yielded a w-, 1966 cm-' with A u l p 23 1 nfrared spectroscopy is a powerful technique for probing the oxygen binding site of heme proteins. Carbon monoxide has been the ligand of choice for such studies because the C-0 stretching frequency is much more easily observed in aqueous media than is that of the 0-0 stretching frequency (Maxwell and Caughey, 1978) and because its substitution for 0 2 is assumed to leave the structure of the protein unaltered. The infrared stretching frequency is related to the C-0 bond energy which is very sensitive to differences in both bond type and environment. The present study was initiated to directly probe, by infrared spectroscopy, the dioxygen-binding site of cyto-

Aeroplysinin-1 (1) and the structurally related dienone 2 were cytotoxic to Ehrlich ascites tumor (EAT) cells and HeLa tumor cells in the microculture tetrazolium (MTT) and clonogenic assays. Both compounds are bromotyrosine derivatives, isolated from the marine spong Aplysina aerophoba. As the effective concentrations in the MTT assay were lower than in the clonogenic assay, 1 and 2 are able to cause growth inhibition as well as actual cell death in these cell lines. With an IC50 value of 8.2 microM (MTT assay, 2-h incubation, EAT cells), 1 was the more toxic compound. When the cells were depleted of glutathione by pretreatment with buthionine sulfoximine, they were significantly more sensitive toward 1 and 2 in the MTT assay. A dose-enhancement factor as high as 11.8 was found in EAT cells after 2-h incubation with 2. Using electron paramagnetic resonance we were able to measure free radical formation of 1 and 2, yielding the semiquinone structures 3 and 4, respectively, in a culture medium with tumor cells. It is concluded that free radicals are, at least in part, responsible for the cytotoxicity of 1 and 2. This conclusion is in line with expectations derived from the chemical structures of both compounds.

No abstract