Phylloplane Bacterium Reveal Altered Nickel Metabolism

Mutation at either of two genetic loci (Eu2 or Eu3) in soybean (Glycine max [L.] Merr.) results in a pleiotropic elimination of the activity of both major urease isozymes. Surprisingly, the phenotype of a phylloplane bacterium, Methylobacterium mesophilicum, living on the leaves of eu2/eu2 or eu3-el/eu3-el mutants is also affected by these plant mutations. The bacteria isolated from leaves of these soybean mutants have transient ureaseand hydrogenase-deficient phenotypes that can be corrected by the addition of nickel to free-living cultures. The same bacterium growing on wild-type soybeans or on urease mutants eul-sun/ eul-sun or eu4/eu4, each deficient in only one urease isozyme, are urease-positive. These results suggest that the bacterium living on the eu2/eu2 oreu3-eI/eu3-eI mutant is unable to produce an active urease or hydrogenase because it is effectively starved for nickel. We infer that mutations at Eu2 or Eu3 result in defects in nickel metabolism but not in Ni2+ uptake or transport, because eu2/eu2 and eu3-el/eu3-el mutants exhibit normal uptake of 'NiCI2. Moreover, wild-type plants grafted on mutant rootstocks produce seeds with fully active urease, indicating unimpeded transport of nickel through mutant roots and stems. Mutations in the expression of the urease isozymes in soybean fall into three general classes (12). Class I mutations result in the loss or modification of the abundant embryospecific urease. All class I mutations isolated thus far identify a single genetic locus, Eul, of which eul-sun is the beststudied allele. Because Eul alleles affect either the level of urease transcripts or the nature of urease protein, it was concluded that Eul contains the gene encoding the embryospecific urease (15). The sole class III mutation, eu4 (20), affects the so-called ubiquitous isozyme normally present in all tissues ofthe plant. Cosegregation between eu4 and a restriction fragment length polymorphism marker revealed by a urease genomic subclone is consistent with Eu4 being the structural gene of the ubiquitous urease (J.D. Griffin, J.C. Polacco, unpublished observations). ' Supported by the Missouri Agricultural Experiment Station and by the National Science Foundation (grants DCB 8718314 and DCB 8804778). This paper is journal contribution 11,524 from the Missouri Agricultural Experiment Station. Unlike class I and class III mutations, each ofwhich affects the expression of a single urease isozyme, class II mutations are pleiotropic, resulting in the loss of all urease activities in the plant. Class II mutations define two genetic loci, Eu2 and Eu3 (16). Although the production of inactive ureases in eu2 and eu3 mutants suggests that they are defective in maturation processes common to both isozymes, the function of the two genes is not known. A study of urease mutants made us aware of the association ofMethylobacterium mesophilicum with soybean tissues (13). Maternal tissues of eu4/eu4 plants are generally urease-negative but show 15 to 40% of the wild-type activity in whole tissue assays of callus (20) and expanding unifoliate leaves (our unpublished observation). This activity is absent in class II mutants and does not appear to be due to "leaky" expression of the eu4 allele, because the urease activity in eu4/eu4 callus and leaf does not resemble the ubiquitous isozyme by several biochemical and immunological criteria. Nor does it appear to be due to low-level Eul expression in nonembryogenic tissue, because eu4/eu4, eul-sun/eul-sun double mutants also express it. To test the possibility that a bacterium is responsible for this basal eu4/eu4 urease activity, we ground leaf tissue and callus in sterile water and streaked the macerates on rich (LB2) and defined (AMS) media (ref. 8, formulations 1065 and 784, respectively). All plates showed bacterial growth after several weeks at 28C. LB plates inoculated with leaf macerate, as expected, showed a variety ofbacterial and fungal contaminants. Surprisingly, LB plates inoculated with apparently axenic callus and AMS plates inoculated with either leaf or callus shared what appeared to be a single bacterial species. This PPFM has been assigned to the genus Methylobacterium (10), formerly Pseudomonas spp. (2, 5, 6, 14) among other designations. In soybean, PPFMs are present on mature leaves in populations as high as 1 x I05 colony-forming units/g fresh weight (J. Dunleavy, personal communication). In young leaves, we have observed up to 4 or 5 x I05 colony-forming units/g fresh weight. The bacterium is difficult to remove by normal treatments for surface sterilization, e.g. soaking in 10% bleach or 70% ethanol. Indeed, every leafand leaf-derived callus sample tested contained the bacterium. In work presented here, we examined whether PPFMs 2Abbreviations: LB, Luria broth; AMS, ammonium mineral salts; PPFM, pink-pigmented facultative methylotroph.