Avirulence genes in plant pathogens did not evolve for that purpose but are a consequence of the occurrence in plants of surveillance mechanisms that detect highly specific features of only certain pathogen biotypes. Thus, while the functions of avirulence genes in pathogens are generally not known, it is assumed that their role in recognition by plants is gratuitous. If recognition of an avirulent pathogen occurs, the plant undergoes a complex active defense reaction called the hypersensitive response (HR) that restricts further pathogen development. The plant surveillance system is controlled by plant disease resistance genes that are thought to encode specific receptors. They recognize pathogen metabolites called specific elicitors. The formation of these elicitors is poorly understood, but recent evidence suggests that they are either the primary protein products of pathogen avirulence genes or cellular metabolites arising from their catalytic activity. We also know little about the structure and function of plant disease resistance genes, although recent progress has occurred in attempts to clone them (e.g., Bennetzen et al 1988). In contrast, several pathogen avirulence genes have been cloned and sequenced (Keen and Staskawicz 1988).

My task is to review the subject of second messengers in animals and its possible relevance for plants. Implicit in this title is the hope that useful guidance will be found in the results from the relatively well-studied animal regulatory systems as we begin to probe transmembrane signalling systems that operate in plants during pathogenic and symbiotic plant-microbe interactions. Some support for this position may be taken from arguments recently set forth by Janssens (1987). He has concluded that similarities are evident in signal transduction mechanisms in vertebrates, invertebrates and eukaryotic microbes, whereas signalling mechanisms in prokaryotes appear to be significantly different. Although much of the evidence on which these comparisons are based is still very preliminary, Janssens suggests that eukaryotic transduction mechanisms, along with many other properties of the eukaryotic cell, may well have originated with eukaryotic microbes. An extension of this reasoning suggests that higher plants would share these attributes as well. On the other hand, it may be well to remember that the signal transduction pathways we are considering in animal cells are mostly concerned with the regulation of basic cellular processes and the integration of metabolic activities within multicellular organisms, whereas we are considering in this Workshop signalling processes that probably evolved at a later stage during the interaction of two different organisms. We should not be surprised to see special features of the signalling pathways that operate in these cases.

Axenic legume root hairs typically remain straight. Early in the infection process, Rhizobium secretes compounds that induce various morphological changes in root hairs, including enhanced differentiation (leading to greater hair density per root surface area), elongation, various deformations including helices, branches, bulbs, and curlings including the 360° tight curl (so called “shepherd crooks”) at hair tips (Yao & Vincent 1969). It has been proposed that these deformations may represent altered synthesis and/or stabilization of the root hair cell wall during growth (for reviews, see Bauer 1981; Dazzo & Hubbell 1982; Halverson & Stacey 1986; Rolfe & Gresshoff 1988).

Plants respond to invasive microorganisms in a variety of ways (Bell 1981). Among these, the synthesis and accumulation of toxic molecules called phytoalexins is a response observed in a number of plant-pathogen interactions (Dixon 1986; Ebel 1986). Phytoalexins accumulate rapidly at the site of infection, and reach levels inhibitory to microbial growth quickly enough to play a role in defense against disease (Hahn et al 1985). Several studies have demonstrated that phytoalexin accumulation is limited, in incompatible (plant resistant) responses, to the immediate vicinity of the infection (Mansfield et al 1974; Moesta et al 1982; Mayama & Tani. 1982; Hahn et al 1985). These observations are important since phytoalexins are toxic to both plaqt and microbial cells (Smith 1982; Weinstein & Albersheim 1983).

The discoveries that complex carbohydrates are tissue-specific cell-surface antigens and the receptors for hormones, toxins, bacteria, and viruses have created considerable interest in complex carbohydrates. Furthermore, the complex carbohydrate portions of some glycoprotein hormones are required for their biological activity, and the complex carbohydrates of some glycoprotein enzymes keep the enzymes stable, while the carbohydrates of some glycoproteins direct the glycoproteins to their proper locations. Moreover, the immune response to glycoproteins is often directed to their carbohydrate side chains, and the carbohydrate chains have been shown to affect the activity and residence time of glycoprotein pharmaceuticals. Still another discovery that has generated considerable interest in complex carbohydrate research is that these molecules perform regulatory functions in plants and animals, often at the level of controlling gene expression.

The rhizosphere is a zone of intense microbial activity where the successful establishment of symbiotic or pathogenic organisms can have a dramatic impact on plant health. Little is known about production and regulation of plant factors that influence the rhizosphere environment. Components of exudates from roots can influence the gene expression and growth of soilborne microorganisms (for ex Stachel et al 1985, Peters et al 1986, reviewed in Curl & Truelove 1986), and can act as chemoattractants (for ex Ashby et al 1988, Caetano-Anolles et al 1988, Hawes et al 1988). The primary sources of such exudates in healthy plants of some species are secretory cells from the root cap (Oades 1978). Sloughed root cap (SRC) cells, which can be isolated nondestructively (Hawes and Wheeler 1982), and can be cultured (Hawes and Pueppke 1986), have been shown to survive for some time in the rhizosphere independently of the root (Vermeer and McCully 1982). Not unexpectedly, carbon-rich SRC cells can support the growth of bacteria in culture (Hawes and Pueppke 1989), and have been shown to act as nuclei for microbial colonies in soil. However, a series of studies in my laboratory suggest that the impact of these cell populations may go beyond the nonspecific stimulation of microbial growth. In studies with several fungal and bacterial pathogens, I have found that the cells from different plant species and genotypes exhibit dramatic selectivity in binding, chemotaxis, and susceptibility to infection (Goldberg et al 1989, Hawes 1983, 1989, Hawes & Pueppke 1985, 1986, 1989, Hawes et al 1989, Hawes & Wheeler 1982). I am using Agrobacterium tumefaciens as a model system to test the hypothesis that the expression of such properties in SRC cells may act to regulate microbial populations in the rhizosphere.

Among microorganisms that cause disease on plants, fungi are the most prevalent. Some fungi are pathogenic to only a single host species and spend their whole life cycles, except for resting structures, on that plant. Others have broad host ranges and may even be able to grow as saprophytes separate from a host plant. In either case plant pathogenic fungi must have unique traits that distinguish them from strict saprophytes and allow for their intimate association with higher plants.

Linear α-l,4-linked oligogalacturonides of chain lengths between 10 and 13 have been shown to elicit phytoalexin accumulation and lignification in plants, but shorter oligogalacturonides do not have this effect (Hahn et al 1981; Nothnagel et al 1983; Jin & West 1984; Davis et al 1986; Robertsen 1986; De Lorenzo et al 1987b; Cervone et al 1989). Phytoalexin elicitor-active oligogalacturonides are thought to be released from the plant cell wall by the action of two groups of microbial pectic-degrading enzymes, the endopolygalacturonases and the endopectate lyases. These enzymes release elicitor-active oligogalacturonides from polygalacturonic acid in vitro but also quickly depolymerize the elicitor-active chains, converting them into shorter, elicitor-inactive molecules (Bruce & West 1982; Nothnagel et al 1983; Cervoneet al 1987). We will discuss two mechanisms by which the action of fungal endopolygalacturonases and bacterial endopectate lyases may be regulated such that the formation of elicitor-active oligogalacturonides is favored.

The hypersensitive response is a commonly observed phenomenon in plant pathology (Tomiyama et al 1979). A plant, when challenged by a microbe that is either a species that is not a pathogen of that plant or an incompatible race of a species that is a pathogen of other cultivars of the plant, commonly exhibits a hypersensitive resistance response, which is phenotypically observed as the rapid necrosis of one or a few plant cells at the site of attempted infection. This necrosis is presumed to result in processes that limit the spread of the pathogen.

Bacteria of the genus Rhizobium fix nitrogen in symbiotic association with leguminous plants. The establishment of symbiosis probably requires many signalling interactions between the bacteria and the plant. The bacterial nod genes are induced by plant flavones (Peters et al 1986) and are involved in synthesis of a compound which has been reported to induce plant cell division (Schmidt et al 1988). In our laboratory, we have studied three classes of Rhizobium meliloti mutants which may be involved in signalling interactions. These are mutants affected in exopolysaccharide synthesis, trpE(G) mutants, and a group of TnphoA insertion mutants.

One of the most fascinating, and still largely unexplained, aspects of Rhizobium/Bradyrhizobium- legume interaction is the host specificity exhibited. Investigations of Rhizobium meliloti and R. leguminosarum bv. viciae and trifolii have shown that host specificity is determined by the respective nodD gene as well as unique host specificity genes (reviewed in Long 1989). Against this background of work with various Rhizobium species, comparatively little is known about the determinants of host specificity in Bradyrhizobium species. Clearly, the specificity of the nodD gene product for specific isoflavone inducers from the plant plays an important role in legume infection by Bradyrhizobium (Kosslak et a1 1988; Banfalvi et al 1988). In addition, a few loci have been identified by mutagenesis in Bradyrhizobium that appear to affect nodulation on only particular host species (Bender et al 1987; Nieuwkoop et al 1987; Hahn and Hennecke 1988).

The process of root nodule formation on legumes, induced by Rhizobium, can be looked upon as a sequence of several distinct steps. These steps have been defined by cytological studies on developing wild-type root nodules, and by analyses of nodules formed by either plant or bacterial mutants (Vincent 1980). Nowadays attachment of bacteria, root hair deformation and curling, induction of cortical cell division, infection thread formation, nodule development, bacterial release from infection threads, bacteroid development and effective nitrogen fixation are recognized as successive steps in root nodule formation (Vincent 1980). The multistep nature of root nodule formation has led to the hypothesis that at several stages in the Rhizobium-plant interaction signal molecules from either symbiontic partner are involved in inducing a process in the other partner Identification of the different bacterial and plant signals and analysis of the mode of action of each seperate compound would then significantly enlarge our knowledge about the establishment of symbiosis.

Oligo- and polygalacturonides can signal the induction of the expression of defensive genes in tissues of plants from at least 8 families (Bishop et al 1984; West et al 1984; Darvill and Albersheim 1984). In dicotyledonous plants polygalacturonides are major constituents of the primary cell wall and oligomeric fragments are released in response to pathogen or insect attacks. The nature of the polygalacturonic acid (PGA) oligomers that are released in vivo, and how they interact with target cells to effect gene expression, is poorly understood. Among the chemicals produced in response to oligo- and polygalacturonides are proteins that inhibit the serine class of proteinases (Bishop et al 1981; Ryan 1987). The biochemical basis for signalling processes that regulate the synthesis of proteinase inhibitors in response to insect attacks has been a major focus of this laboratory for several years (Bishop et al 1981; Ryan 1987).

Our long term goals are to study the interaction of Rhizobium and plants at the molecular and cellular levels. Our system of study is Rhizobium meliloti and its host plants such as alfalfa (Medicago sativa L.). The approaches we use are primarily to combine genetic analysis of nodulation with microscopic observation of normal and mutant bacteria during interaction with plants. Our recent work includes the definition of two new nodulation genes, and the further characterization of the network for regulation of nodulation genes in Rhizobium meliloti. One theme found in both these studies is the complexity arising from multiple gene copies in R. meliloti.

Considerable information has accumulated about the mechanisms by which animal cells perceive extracellular signals. In sharp contrast, very little is known about cellular signal perception in plant cells. The purpose of this article is to examine signal perception mechanisms in animal cells and to highlight general principles that may prove applicable to the study of plant systems. I will focus on the receptors in target cells that specifically bind the signal molecule(s). This is the initial event in the signal cascade that eventually leads to changes in cellular metabolism (e.g., gene expression) in targeted cells. Secondary messengers, which are often involved in the signal transduction process subsequent to the binding of signal molecules to receptors, are the focus of another review (see article by C. A. West, this volume).

Although plants are continuously exposed to a variety of potentially pathogenic microorganisms, successful infections are rare. Plants utilize a diverse array of defense mechanisms to prevent microbial infections. Some of these defense strategies involve constitutively expressed physical and chemical barriers that provide a first line of defense against potential pathogens (Mansfield 1983). Other defense mechanisms are specifically induced upon attempted infection. These induced defense responses include the synthesis of polyphenolic lignins and hydroxyproline-rich glycoproteins which are incorporated into plant cell walls, causing the walls to be more resistant to microbial invasion? the synthesis of the hydrolases (β-1,3- glucanase and chitinase, which may inhibit fungi by degrading their cell walls; and the synthesis and accumulation of antimicrobial compounds called phytoalexins (Bell 1981, Collinge and Slusarenko 1987, Darvill and Albersheim 1984, Hahlbrock and Scheel 1987).