Nitric oxide (NO) is a potent intercellular signal in mammals that mediates key aspects of blood pressure, hormone release, nerve transmission and the immune response of higher organisms 1 , 2 , 3 , 4 . Proteins homologous to full-length mammalian nitric oxide synthases (NOSs) are found in lower multicellular organisms 5 . Recently, genome sequencing has shown that some bacteria contain genes coding for truncated NOS proteins; this is consistent with reports of NOS-like activities in bacterial extracts 6 , 7 . Biological functions for bacterial NOSs are unknown, but have been presumed to be analogous to their role in mammals. Here we describe a gene in the plant pathogen Streptomyces turgidiscabies that encodes a NOS homologue, and we reveal its role in nitrating a dipeptide phytotoxin required for plant pathogenicity 8 . High similarity between bacterial NOSs indicates a general function in biosynthetic nitration; thus, bacterial NOSs constitute a new class of enzymes 9 , 10 , 11 . Here we show that the primary function of Streptomyces NOS is radically different from that of mammalian NOS. Surprisingly, mammalian NO signalling and bacterial biosynthetic nitration share an evolutionary origin.

There are relatively few bacterial diseases of roots, in comparison to those of aerial plant tissues. Numerous species and pathovars of Pseudomonas,Erwinia and Xanthomonas are important pathogens of leaf and stem tissue on dozens of plant families but these bacterial genera only infrequently attack roots or other underground plant structures. In contrast, there is a growing list of Streptomyces species that are very effective root pathogens. These filamentous, Gram-positive bacteria can cause scab, rot and gall diseases of plant roots and other underground plant structures. The best known pathogenic Streptomyces species is S. scabiei. Horizontal transfer of pathogenicity genes among diverse scab-causing streptomycetes appears to explain the emergence of several new plant pathogens over the last half century. It is proposed that the ability to penetrate plant tissue is essential for successful root infection as there are few natural openings in roots. In contrast, leaves have many natural openings that allow bacteria access to the interior tissues. Thaxtomin, a phytotoxin produced by many plant pathogenic streptomycetes, appears to aid penetration of developing plant tissues by inhibiting primary cell wall development.

Streptomyces scabies and S. acidiscabies, causal agents of potato scab, produced the phytotoxin thaxtomin A and, to a lesser extent, other thaxtomins in oatmeal broth (OMB). Two nonpathogenic strains, Streptomyces lividans TK 24 and Streptomyces sp. 84-05, did not produce thaxtomins in OMB. Though thaxtomin A was produced by S. scabies strain 87-22 in potato broth, fresh potato starch, and some commercial starch preparations, production was much greater in OMB than in other media tested. Thaxtomin A was not produced in Luria broth or tryptic soy broth. Production of thaxtomin A by 87-22 was suppressed by 0.5% glucose in OMB but was stimulated by up to 5.0% glucose in oatmeal agar. Streptomyces scabies strain 87-22 produced 4.25 mu g thaxtomin A per milliliter of OMB while strain 84-34 produced 0.17 pg per milliliter of OMB. Streptomyces acidiscabies strains 84-110 and 90-25 produced similar amounts of thaxtomin A, 2.65 and 4.45 mu g per milliliter of OMB, respectively. Strain 87-22 was much more virulent on tubers of Chippewa, a scab-susceptible potato cultivar, than was strain 84-34. A tuber slice bioassay was useful for detection of thaxtomins in culture media and for identifying pathogenic Streptomyces strains.