Barbara B. Lippincott

Plasmid and chromosomal determination of Agrobacterium adherence specificity

To initiate tumours Agrobacterium tumefaciens must adhere to a wound site via a process which requires considerable specificity on the part of the bacterium (Lippincott & Lippincott, 1969; Whatley et al., 1976). Because this site binding ability and the presence of a large plasmid (Zaenen et al., 1974) are both essential for tumourigenicity, it was of interest to determine if site binding ability was a plasmid-borne trait. Whatley et al. (1976) showed that lipopolysaccharide (LPS) of the outer membrane of the Agrobacterium cell envelope is the bacterial adherence component. Two avirulent strains of Agrobacterium derived from virulent strains and reported to be plasmid-free were found to retain site binding ability (Whatley et al., 1976). It was proposed, therefore, that chromosomal DNA determined the specificity of LPS essential for adherence (Whatley et al., 1976). Van Larebeke et al. (1975) reported that introduction of a virulence plasmid into Agrobacterium radiobacter strain s1005 converted this strain to a tumourigenic state. However, whole cells and LPS of strain s1005 lack the ability to adhere to tumour initiation sites (Lippincott & Lippincott, 1969; Whatley et al., 1976). Thus, either the strains of s1005 tested in the two laboratories differed in their site binding ability, or the presence of the plasmid in s1005 must affect LPS structure. To resolve these possibilities we determined the localization of genes governing adherence by testing non-site binding strains to which a virulence plasmid was added and additional site binding strains after curing of all plasmids.

The Genus Agrobacterium

The genesis of the genus Agrobacterium, a member of the family Rhizobiaceae (Buchanan and Gibbons, 1974), can be traced to the discovery by Smith and Townsend (1907) of the causal agent of crown gall diseases of plants (now called Agrobacterium tumefaciens; Allen and Holding, 1974) and, some years earlier, by Beijerinck and van Delden (1902) of the common nonphytopathogenic soil inhabitant (now called Agrobacterium radiobacter; Allen and Holding, 1974). Establishment of the genus Agrobacterium by Conn (1942) followed from the dismantling of the genus Phytomonas, into which a great array of phytopathogenic agrobacteria, corynebacteria, pseudomonads, and xanthomonads had been lumped (this Handbook, Chapter 4).

Microbial Adherence in Plants

The importance of adherence in the establishment of plant—microbe interactions and the alteration of plant metabolism has only recently begun to be appreciated and to be a direct subject of investigation. A search for factors responsible for specificity in host—pathogen, host—symbiont interactions and for the mechanisms involved in the induction of disease symptoms and host defense reactions provided the initial incentive for most of the work which comprises this developing field. A few reviews have covered certain aspects of this problem, e.g., Albersheim and Anderson-Prouty (1975), Marshall (1976), Raa et al. (1977), Bauer (1977), Sequeira (1978), Beringer et al (1979) and Dazzo (1979). The present review will focus primarily on interactions of micro-organisms with seed plants; adherence interactions involving micro-organisms with algae, yeasts and fungi or between different portions of the plant life cycle are outside the scope of this review.

Timing of events in crown-gall tumor development on pinto bean leaves☆

Abstract Tumors initiated on primary Pinto bean leaves by Agrobacterium tumefaciens may be counted as early as 48 hours after inoculation. The number of tumors per leaf increases linearly until day 7 or 8 and by day 9 is complete. Tumor growth (logarithm of mean tumor diameter) shows an approximate parallel relation to tumor appearance, diminishing after day 8 to one-fifth the original growth rate. Tumors appearing early or late exhibit similar growth curves; the decrease in growth rate, however, occurs at smaller tumor sizes, the later the tumor is first observed on the leaf. The number of plant cells initially participating in tumor formation may be estimated from these data to be 9 cells for the average tumor. Infected leaves are most sensitive to the inhibition of tumor formation by elevated temperature treatments when the treatments are applied 9–10 hours after inoculation and are completely resistant after 42 hours. Conversion of half the potential tumor sites to tumor sites may take place within 12 hours after inoculation, and by 42 hours all potential sites are converted. A portion of these tumors may be initiated immediately after inoculation. The number of viable A. tumefaciens cells recoverable from inoculated leaves decreases to less than 1% of the original inoculum within the first day following infection. A minimum of 10 4 to 10 5 bacteria per leaf is reached at 4–6 days after inoculation. On the average, there are about 1500 bacteria at this minimum point per tumor initiated. The number of recoverable bacteria is not obviously related to tumor formation.

Comparative Response of Pylaisiella selwynii to Agrobacterium and Rhizobium Species

Thirty-five strains of Agrobacterium and five strains of Rhizobium were tested for their ability to induce developmental changes in the moss Pylaisiella selwynii. The avirulent A. radiobacter (seven strains) and nontumorigenic strains of A. tumefaciens (four strains) had relatively little effect on moss development. Virulent strains of A. tumefaciens (17 strains), A. rubi (two strains), and A. rhizogenes (five strains) significantly altered development of the moss. Bud formation was induced by most strains, and these developed into normal gametophores, abnormal gametophores, or callus masses, depending on the strain of Agrobacterium. Agrobacterium rhizogenes and three strains of A. tumefaciens induced formation of numerous rhizoid-like filaments and only occasional gametophores. All five species of Rhizobium tested induced bud formation and normal gametophore development. The most effective of these strains, R. leguminosarum C56, required contact between moss and bacterium, similar to that required by A. tumefaciens. This strain also competed with virulent A. tumefaciens for attachment sites on the moss filaments, suggesting Agrobacterium and Rhizobium adhere to the moss via similar bacterial wall components. The rhizoid-inducing substance produced by A. rhizogenes strain TR7 readily passed through filters separating bacterium and moss in a parabiotic chamber, indicating it is released into the medium by the bacterium. The correlation between higher plant infectivity and the ability to induce developmental changes in P. selwynii suggests this system may offer unique possibilities for the study of Agrobacterium and Rhizobium infectivity.

Two octopine dehydrogenases in crown-gall tumor tissue

Abstract Extracts from four crown-gall tumor tissue culture lines, originally induced by two octopine-type strains of Agrobacterium on three plant species, converted l -arginine-[5- 3 H] to a compound which co-migrated with octopine on electrophoresis. Synthesis showed dependence on added pyruvate and reduced pyridine nucleotide. Both NADH and NADPH were active and mixtures of the two coenzymes, when tested with Vinca strain W1 tumor extracts, were more effective than either coenzyme at comparable concentrations. Addition of an NADH-consuming enzyme system to reaction mixtures containing NADPH had little effect on this activity. Products formed by Vinca rosea strain W1 tumor extracts and Phaseolus vulgaris strain B6 tumor extracts in reaction mixtures containing pyruvate plus NADH or NADPH co-eluted with unlabeled octopine on ion exchange chromatography. The product from the Vinca reaction mixtures co-migrated with an octopine standard in three TLC systems. Permanganate treatment of the enzymatically formed tritiated product and of unlabeled octopine gave compounds with R f , similar to arginine and γ-guanidinobutyric acid, the products expected from permanganate degradation of octopine. The Vinca W1 extracts catalyzed the oxidative cleavage of octopine, with the formation of arginine, in the presence of NAD or NADP. Two octopine dehydrogenases were concluded to be present in these tissues, one dependent on NAD, the second on NADP.

Agrobacterium tumefaciens: Thermal Inactivation of Tumor-Inducing Ability

Treating cultures of Agrobacterium tumefaciens at 39� to 48�C reduces the infectivity of the bacteria without necessarily affecting viability. Destruction of the capacity to initiate tumor growth follows first-order kinetics from which rate constants for thermal inactivation are derived. From these rates, values for heat of activation of 56.7 kcal mole-1 and entropy of activation of 107 cal mole-1 deg-1 are obtained. A particular protein or nucleoprotein active in the process of infection may be inactivated by the treatment.

Lysopine and Octopine Promote Crown-Gall Tumor Growth in vivo

Growth of tumors induced on primary leaves of bean plants by Agrobacterium tumefaciens was increased by the addition of lysopine and octopine. A detectable response was observed when as little as 1 microgram of these compounds was added per leaf, and the mean volume of the tumors was increased two- to threefold when greater amounts were applied. The specificity of the response and the unique association of these compounds with the tumors suggest that endogenous lysopine and octopine contribute to the growth characteristics of these tumors.

Plasmid pSa causes loss of LPS-mediated adherence in Agrobacterium

Summary: Suppression of virulence in Agrobacterium caused by introduction of the IncW R plasmid pSa into cells containing a Ti plasmid is accompanied by loss of site adherence in the pinto bean infection assay and by loss of site adherence on the part of LPS isolated from these strains. When cured of the pSa plasmid, infectivity and site adherence are restored. This indicates that LPS produced by pSa-containing agrobacteria is sufficiently modified that it will not support site adherence, the initial step of the infection process.

Bacteria Isolated from Moss and Their Effect on Moss Development

Bacteria from four species of mosses from different geographical locations were isolated either on media selective for agrobacteria or on nutrient broth. Many isolates induced developmental changes in Pylaisiella selwynii protonemal filaments comparable to those obtained with agrobacteria. No bacteria induced crown gall or hairy root symptoms on beans, tomato, or Kalanchoe, although a few promoted callus growth on potato tuber and also promoted callus and roots on carrot disks, similar to Agrobacterium rhizogenes. Many isolates bind to specific Agrobacterium attachment sites in wounds; five of seven tested isolates competed with virulent Agrobacterium tumefaciens for these sites on bean leaves. No isolate could unequivocally be assigned to the genus Agrobacterium by several physiological and biochemical tests used to classify Agrobacterium. These results demonstrate that bacteria adhering to moss in nature may affect protonemal growth and gametophore initiation and thus may be an important ecological factor in regulating normal life cycle events at this stage of moss development.