potential and pitfalls of trying to extend symbiotic ... · we briefly review this research and...

Plant and Soil 174: 225-240, 1995. (~) 1995 Kluwer Academic Publishers. Printed in the Netherlands.

Potential and pitfalls of trying to extend symbiotic interactions of nitrogen-fixing organisms to presently non-nodulated plants, such as rice

EJ . de Bru i jn 1'2'3, Y. J ing 1'4 and E B . D a z z o 2'3 1MSU-DOE Plant Research Laboratory, 2Department of Microbiology, 3 NSF Center for Microbial Ecolog); Michigan State University, E. Lansing, M148824, USA and 4Institute of Botany, Chinese Academy of Sciences, Beijing, China

Key words." endophytic microbes, endo-symbiosis, Rhizobium, rice, root hypertrophies, ultrastructure

Table of contents

Abstract

Introduction

Plant-microbe signalling, nodulation, and symbiotic nitrogen fixation

The oxygen paradox and other potential problems related to microbial nitrogen fixation in novel physiological environments

Nodulation of cereals: What has been tried, what has been published, and is nodulation really necessary to achieve useful levels of nitrogen fixation

Induction of hypertrophies on rice roots and their colonization by microbes

Materials and methods

Induction and microscopic examination of "nodule-like" structures on rice roots

Molecular methods for strain identification

Results and discussion

Abnormal structures on rice roots grown in China

Abnormal structures on rice roots grown at Michigan State University

Acknowledgements

References

Abstract

It has been a long-standing goal in the field of biological nitrogen fixation to extend nitrogen-fixing symbioses to presently non-nodulated cereal plants, such as rice. A number of researchers have recently described the induction of "nodule-like" structures on the roots of cereals primarily by rhizobia, in either the presence or absence of plant cell-wall-degrading enzymes or plant hormones. We briefly review this research and discuss the potential problems associated with the introduction of nitrogen-fixing microbes in novel physiological environments, such as rice roots. The results of experiments carried out in China on the induction of "nodule-like" structures on rice roots by rhizobia are highlighted. In addition, we present preliminary results of a series of experiments designed to repeat and evaluate these results using a variety of microscopic techniques and molecular genetic approaches.

226

Introduction

In modern agriculture, the replenishment of soil nitrogen most commonly involves extensive application of chemical fertilizers (Peoples et al., 1995), an approach that suffers from several serious drawbacks, including high costs and severe negative environmental impacts. These considerations have stimulated research on alter- natives, such as biological nitrogen fixation. Biological nitrogen fixation has been estimated to contribute more than 170 million tons of fixed nitrogen to the biosphere (Earl and Ausubel, 1993). Eighty percent of the stable, biologically fixed nitrogen is a direct result of the symbiotic interaction of members of the Rhizobiaceae and some actinomycetes with leguminous, as well as certain non-leguminous plants.

In the nodules induced on the roots and stems of the host plant, nitrogen fixed by rhizobia or actinomycetes is directly incorporated into nitrogen-rich organic compounds such as amides, amino acids, or ureides and proteins (Schubert, 1986). In contrast to this efficien- cy, nitrogen fixed by strictly free-living or associa- tive diazotrophs can be rapidly lost by denitrifcation reactions in the biosphere (rhizosphere). Therefore, a considerable amount of research effort has been invest- ed in elucidation of the molecular basis of symbiotic plant-microbe interactions. Two long-term objectives of these studies have been to further improve the effi- ciency of symbiotic nitrogen fixation and to investi- gate the potential of extending this beneficial process to presently non-nodulated plants, especially cereals. Recently, a number of claims have been made with regard to the latter, some of which may have been somewhat overstated (Simon Moffatt, 1990). In fact, several of the reports on the supposed induction of (root) "nodules" on nonlegumes, including rice, have been controversial, despite their endorsement in the popular scientific press. This situation has led to con- cern among scientists in the field of symbiotic nitrogen fixation not to be unrealistic and repeat the mistakes of the early 1970s, when the transfer of the entire package of nitrogen-fixation (ni33 genes from a free- living diazotroph into plants was widely claimed as "within reach". It has also led to the organization of an International Rice Nodulation Group (Cocking and Davey, 1991) and a small international conference on the potential and prospects of nodulation and nitrogen fixation in cereals such as rice (Khush and Bennett, 1992).

In this paper, we will first review very briefly the present state of knowledge about plant-microbe sig-

nailing and the induction of nitrogen-fixing stem and root nodules on legume plants. Second, we will sum- marize some of the crucial biological problems that would need to be solved to establish the process of symbiotic (endophytic) nitrogen fixation in plants such as cereals. Third, some of the published reports about "nodulation" of cereals will be critically reviewed and the need for the induction of proper nodule structures on presently non-nodulated plants for the purpose of nitrogen fixation will be questioned. Fourth, we will discuss some recent preliminary results from experiments designed to examine a previous report (Li et al., 1991) describing the formation of hypertrophies and nodule-like structures by selected microbes on rice roots, and the presence of microbes in these structures.

Plant-microbe signalling, nodulation, and symbiotic nitrogen fixation

The induction of nitrogen-fixing root and stem nodules on leguminous plants by soil bacteria belonging to the Rhizobiaceae involves fine-tuned interactions between the two symbiotic partners, including multiple regulatory signals that go back and forth between the bacterium and the plant to coordinate expression of gene sets in both partners. These signals include plant factors that induce rhizobial nodulation (nod) genes, rhizobial Nod factors that are essential for early stages of nodule ontogenesis, and hormones and other regulatory factors involved in symbiotic nitrogen fixation (for reviews see de Bruijn and Downie, 1991; D6nari6 and Cullimore, 1993; Fisher and Long, 1992; Hirsch, 1992). The first step in the plant-microbe signalling pathway leading up to nodulation is the production/secretion of flavonoids, chalcones, conjugat- ed isoflavonoids and betaines by the plant host (see Peters and Verma, 1990; Phillips, 1992; Phillips et al., 1994). These primary signal compounds are respon- sible for the induction of the nodulation (nod) genes of the microbial symbiont, most of which are silent in free-living rhizobia. This induction process involves the specific activation of regulatory factors in the rhizobia (NodD proteins) by particular representatives of the group of plant factors listed above. These activated NOdD proteins, in turn, serve as transcriptional activators of other nod genes (see Fischer and Long, 1992).

Once the rhizobial nod genes are activated, their gene products are involved in the synthesis of a class of compounds (Nod factors), which serve as the first

227

return signal from the bacteria to their host plant. This class of compounds has been shown to consist of lipo- oligosaccharides of varying lengths, carrying different side groups and substitutes that play major roles in conferring host specificity (see D~nari6 and Cullimore, 1993). The Nod factors do not only mediate recognition between the proper rhizobial species and its natural host(s), but also serve as specific morphogens by initiating root hair curling, and cortical cell division in the target root cells (see DEnari6 and Cullimore, 1993; Vijn et al., 1993). Similar to the plant derived first signals that activate rhizobial (nod) gene expression, the second rhizobial signal molecules (Nod factors) induce specific gene expression in the cognate host plant (see D6nari6 and Cullimore, 1993; Vijn et al., 1993). Gen- erally, the plant genes specifically expressed during the formation and functioning of nodules are referred to as nodulin genes (Van Kammen, 1984). They can be divided into several classes, depending on their time point of induction during nodule formation, but the most prominent classes have been named "early and late nodulin genes" (Nap and Bisseling, 1990), and are clearly activated in different ways (see de Bruijn and Schell, 1992). Purified Nod factor has been found to activate early nodulin genes such as Enod5 and Enod 12 in the host plant (Horvath et al., 1993), further support- ing the proposed role of the Nod factor as an essential signal molecule in nodule ontogeny.

In addition to lipo-oligosaccharide signal molecules, other metabolites have been suggested to transfer microbe-plant signals and to play a role in nodulation. These include hormones (see Cooper and Long, 1994; Bruijn et al., 1992; Hirsch, 1992) and compounds such as N-acetyl glutamic acid (Hollingsworth et al., 1991) and diglycosyl diacylglycerol (Orgambide et al., 1994). With regard to the former, the hormone cytokinin has been proposed to play an important role in signalling required for nodulation, since it activates the early nodulin gene Enod2 (Dehio and de Bruijn, 1992) and permits otherwise nodulation deficient rhizobia (lacking the common nod genes) to induce nodulation on plants (Cooper and Long, 1994). The N-acetyl glutamic acid and diglycosyl diacylglycerol metabolites made by rhizobia were found to elicit root hair deformation and have mitogenic activity on host root cells (Hollingsworth et al., 1991; Orgambide et al., 1994).

Once the nodule is developed and the infecting rhizobia have been released into the plant cell cytoplasm and differentiated into a bacteroid state, the nitrogen fixation process is activated. This activation process

is mediated by micro-aerobic induction of the nitrogen fixation (nif/fix) genes (David et al., 1990), via the oxygen sensing protein FixL, a hemoprotein, (Gilles- Gonzales et al., 1992) in most cases examined (see Fis- cher 1994). Concurrent with the onset of nitrogen fixation, a group of late nodulin genes are activated, such as the genes encoding the oxygen carrier protein leghemoglobin (Ib genes) and genes involved in carbon and nitrogen assimilation (see Nap and Bisseling, 1990). The regulation of the late nodulin genes may involve another type of bacterial signal, since these genes are expressed predominantly in the infected cells of the nodule (de Bruijn and Schell, 1992; Lauridsen et al., 1993; Szczyglowski et al., 1994). The physical presence of (membrane-enclosed) bacteria in the cytoplasm of the infected cells appears to be required and a rhizobial DNA-binding protein interacting with a specific cis-acting element in the plant lb promoter has been described (de Bruijn et al., 1994; Welters et a1.,1993) that may be involved in infected-cell-specific expression of the lb genes. If this hypothesis is proven to be true, it would constitute another interesting example of microbe-plant (trans-kingdom) signalling in nodulation.

These studies make it clear that the process of plant- microbe signalling in nodulation is highly complex and that we must develop a clear, basic understanding of the nature and expression of rhizobial and plant genes involved in the nodulation and symbiotic nitrogen fixation process before we attempt to design experiments to extend the symbiotic interaction to presently non- nodulated plants (see de Bruijn and Downie, 1991).

The oxygen paradox and other potential problems related to microbial nitrogen fixation in novel physiological environments

The reduction of dinitrogen to ammonia catalyzed by the enzyme nitrogenase requires large amounts of ATP (up to 40 ATPs/N2 reduced). Rhizobia gener- ate this ATP via oxidative phosphorylation, requiring a constant 02 supply. However, nitrogenase is irre- versibly denatured at 02 concentrations exceeding 10 nM (Appleby, 1984). This "oxygen paradox" repre- sents an important biological problem when consider- ing the extension of symbiotic interactions to presently non-nodulated plants. In legume nodules, several mechanisms have evolved to deal with the oxygen problem, including the creation of an apparent oxygen barrier in the nodule parenchyma and the high-level synthesis of the oxygen carrier protein leghemoglobin

228

(Appleby, 1984; Hirsch, 1992). Although some evidence exists for the presence of hemoglobin-type genes in non-legumes (see Bogusz et al., 1988), it remains unclear whether or how the proper physiological con- ditions for nitrogen fixation can be created in plant tissues not normally infected by nitrogen-fixing organisms.

The same uncertainty applies to other physiological requirements for endosymbiotic or endophytic nitrogen fixation, such as high levels of energy- reducing equivalents, mechanisms to rapidly assimi- late the ammonia produced by dinitrogen reduction, and ways to avoid the plant defense response. Again, we must carry out extensive basic physiological studies on these processes before attempting to introduce nitrogen-fixing organisms into novel environments, such as rice roots.

Nodulation of cereals: What has been tried, what has been published, and is nodulation really necessary to achieve useful levels of nitrogen fixation ?

Although much basic knowledge about the molecular basis of plant-microbe interactions and the phys- iology of nitrogen fixation/assimilation is still lacking, researchers worldwide have started to explore the possibility of extending the nitrogen-fixing symbiosis of rhizobia and legumes to presently non- nodulated plants. Several recent reports describe various approaches used to induce nodule-like structures on the roots of rice and other cereals (see Kennedy and Tchan, 1992; Khush and Bennett, 1992; Simon Mof- fatt, 1990). The infection of "para-nodules", induced by chemicals such as 2,4-D on wheat roots, by rhizobia and free-living diazotrophs has been reported and suggested to be a potential vehicle to achieve nitrogen- fixation in cereals (Bender et al., 1990; Kennedy and Tchan, 1992; Tchan et al., 1991). In fact, nitrogen fixation by Azospirillum brasilense and Azorhizobi- um caulinodans in para-nodules induced on wheat roots has been reported (Chen, 1993; Chen et al., 1993; Tchan et al., 1991) and the effect of a variety of microorganisms on the formation of 2,4 D- induced nodule-like structures on wheat roots has been examined (Ridge et al., 1992). Moreover, Christansen- Weniger and Vanderleyden (1994) have shown that ammonium-excreting Azospirillum can become established intracellularly in 2-4-D induced paranodules of maize.

The induction of nodule-like structures on and invasion of rice, maize, and wheat seedling roots by

Parasponia and Aeschynomene rhizobia, either spon- taneously or after the treatment of roots with cell-wall- degrading enzymes, have also been described (A1- Mallah et al., 1989; Cocking et al., 1990, 1992). It has been proposed that these approaches, involving "nodulation in non-legume crops by rhizobia" may lead to advances in obtaining "nitrogen from the air for nonlegume crops" (Cocking and Davey, 1991).

A Rhizobium strain has been constructed containing a plasmid carrying a nodD allele able to respond to signals produced by rice roots. This strain induces infected nodule-like structures on rice seedlings at a low frequency (Rolfe and Bender, 1990).

Jing et al. (1990) have reported the induction of pseudonodules on barley roots by Rhizobium astragali under a permanent magnetic field. Moreover, a high- frequency induction of infected nodule-like structures on rice roots by Sesbania rhizobia has been reported (Jing et al., 1992; Li et al., 1991). In fact, acetylene reduction by "rice root nodules" infected by rhizobia and a positive effect of rhizobial infection of rice plants on growth and yield has been suggested (Jing et al., 1990, 1992), and the presence of infection threads and bacteroids in the infected "nodule-like structures" has been proposed (Jing et al., 1992; Li et al., 1992).

Several of these reports are controversial and have not yet been independently confirmed. In some cases, the structures reported have been called "nodules"; without proper experimental support they should prob- ably not even be considered "nodule-like structures". In fact, many of the structures reported appear to be modified lateral roots (Cocking and Davey, 1991; Ridge et al., 1992; see also Kennedy and Tchan, 1992, and below). With regard to nitrogen fixation in these infected structures or positive effects on plant growth, very few of the reported studies have employed 15N- based techniques or have included the proper controls or statistical analyses. Only in the case of wheat "para- nodules", has 15N been used to confirm N2 fixation (Yu et al., 1993). The ultrastructural analyses of infected cells via various microscopic techniques have also been dubious in several cases; an intact cytoplasm cannot be seen in many of the photographs presented, making it difficult to exclude saprophytic infections of broken cells, to discern whether the infected cells are intact, or to establish whether the infecting bacteria are truly (and stably) maintained intracellularly.

Thus, a lot of unresolved issues remain with regard to these "nodulation" experiments, which need to be carefully (re)examined. The more general ques- tion could also be raised whether one actually needs

229

Fig. 1. Rice roots grown in China and containing dark “nodule-like structures” as previously described (Li et al., 1991).

Fig. 2. Stereomicrographs showing the unusual dark hemispheric structures (arrows) attached to the rice root surface. (A) Combined transmitted/incident illumination; (B) darkfield illumination. Bar scale is 1 mm.

230

Fig. 3. Brightfield micrographs of the unusual hemispheric structures attached to the rice root surface; 2 pm sections stained with alkaline toluidine blue. (A) Rice root in cross-section. (B) Higher magnification of the interface between the external hemispheric structures and the root epidermis. Note the septate hyphae (arrows) penetrating the root epidermis and extending into the root cortex. (C) Rice root in longitudinal section, showing the external hemispheric dome structure and colonization in the underlying root cortex. Bar scales are 100 pm in (A), 50 pm in (B, C).

nodules or nodule-like structures on cereal roots to achieve symbiotic/endophytic nitrogen fixation (see also Kennedy and Tchan, 1992; Ladha et al., 1993). It may, indeed, be sufficient to identify a stable endophyte of rice roots and, if necessary, to engineer this microbe to efficiently fix nitrogen and excrete the fixed nitrogen for use by the plant. This endophyte may not need to be stably maintained intracellularly, like a rhizobial endosymbiont; it may be sufficient if it colonizes the plant root intercellularly, as long as it is able to evade the plant defense responses (Khush and Bennett, 1992; Quispel, 1992).

In this context, it is interesting to note that stable endophytic diazotrophs that contribute substantially to

plant growth via nitrogen fixation have been described for other plants, such as sugar cane (Acetobucter diu- zotrophicus; see Boddey et al., 1995; Dobereiner et al., 1993). Moreover, a strain of the nitrogen-fixing bacterium Azoarcus has been described that colonizes and spreads systemically in grasses such as rice (Hurek et al., 1994). This bacterium was found to invade roots inter- and intracellularly, to enter and spread through the xylem, and to increase rice yield, although the latter phenotype could not be ascribed to nitrogen fixation (Hurek et al., 1994). Therefore, Azoarcus cannot be classified as an endophytic diazotroph, as defined by Dobereiner et al. (1993) and Quispel(1992). It may nevertheless serve as an interesting model organism in

231

Fig. 4. Brightfield micrographs of the unusual hemispheric structures detached from the rice root; 2 pm sections stained with alkaline toluidine blue. (A) Section through the hemispheric structure. (B) Higher maaification of the central zone showing the sacks with several ornamented cells (&row) inside. Ba; scales are lb0 pm in (A, B). - -

studying the persistence of non-pathogenic bacteria in experiments and the molecular analysis of some of the plant tissues in general and rice root colonization in strains employed for these experiments will be present- particular (Hurek et al., 1994). ed here.

Induction of hypertrophies on rice roots and their colonization by microbes Materials and methods

The high-frequency induction of nodule-like structures on rice roots by rhizobia isolated from Sesbania cannabina nodules (see Fig. 1) and the presence of rhizobial infection (producing thread-like structures) and peribacteroid-like membranes in plant host cells have been reported recently (Li et al., 1991). “Nodulation percentages” of up to 66% on the Chinese rice vari- eties Lianjiang, Juefu and Jiangxi 80074 were reported, reisolated bacteria were shown to contain nifand nod genes, and were classified by nutritional requirements in order to show that they were closely related to the input bacteria. They were also found to be able to re-nodulate S. cannabina (Li et al., 1991).

Induction and microscopic examination of “nodule-like” structures on rice roots

This research by Li and colleagues has been considered of particular interest and worthy of further inves- tigation (Khush and Bennett, 1992; Simon Moffatt, 1990). Therefore, experiments were initiated by Dr Jing to re-examine the induction of nodule-like structures described by Li et al. (1991). He used a variety of microscopic techniques to obtain similar structures at Michigan State University (MSU) on a rice variety shown to be highly responsive to bacterial infection (Zhong Xi 8408; Y Jing, unpubl. results) when inoculated with some of the same strains previously used (Li et al., 1991). The preliminary results of these

Rice variety Zhongxi 8408 was grown in China (Jing et al., 1992). Roots bearing the previously described dark “nodule-like -structures” (Li et al., 1991; Fig. 1) were excised, preserved in formalin, and transport- ed to MSU. Root segments were examined directly by stereomicroscopy with transmitted, incident, and darkfield illumination. Short root segments with these hemispherical domes still attached were excised while viewing under the stereomicroscope and processed for plastic embedding using standard methods of spec- imen preparation, followed by combined light and transmission electron microscopy (Dazzo, 1982). Oth- er formalin-preserved root segments with the dome- shaped structures still attached were processed directly for examination by scanning electron microscopy (Umali-Garcia et al., 1980), and laser scanning confocal microscopy after staining with acridine orange (Daze0 et al., 1993).

In other experiments, rice plants (20-30 seedlings per pot) were grown at MSU in 25-30 cm diameter pots with either sterile sand or soil mixtures in Convi- ron growth chambers under 15 hr day-’ of light (80%

232

Fig. 5. Scanning electron micrographs of the unusual hemispheric structures attached to the rice root. Note the encrusted surface layer, the internal hyphal network, and the sacks containing several spherical cells with reticulate surface ornamentations. Bar scales are 1000 pm in (A), 100 pm in (B, C), and 10 brn in (D).

humidity), and inoculated with fresh cultures of rhizobial strains A201, R201 or A3, originally isolated from Sesbania cannabina (Y Jing, unpubl.). Periodically a subset of the plants were uprooted and cleaned by run- ning tap water. These fresh roots were examined directly by stereomicroscopy for the presence of abnormal hypertrophies. Root segments containing these structures were excised and processed in three ways for microscopic examination. Some root segments were cleared under vacuum with sodium hypochlorite (40% Chlorox solution), stained with methylene blue solution, and examined by brightfield light microscopy (Hollingsworth et al., 1989; Truchet et al., 1989). Oth- er subsamples of root were fixed, dehydrated, critical-

point dried, sputter coated, and examined by scanning electron microscopy (Umali-Garcia et al., 1980), or fixed, dehydrated, embedded in plastic, and thin- sectioned for combined light and transmission electron microscopy (Dazzo, 1982).

Molecular methods for strain identification

Rep-PCR-mediated genomic fingerprinting was used to identify and compare bacterial isolates used for rice root infections and bacteria reisolated from the rice root hypertrophies, as described (de Bruijn, 1992). Nodulation (nodABC) and nitrogen fixation (nifHDK) gene probes of Rhizobium meliloti (pRmSL42; kindly

233

Fig. 6. Laser scanning micrographs of a small hemispherical strocture attached to the rice root. (A) Top view by transmitted light microscopy. (B) Computer reconstructed composite 2” section of a side view examined in the epifluorescence confocal mode. Shown are surface (top) and internal (below) colonization by bacteria which are stained with acridine orange. Bar scales are 50 km in (A), 10 pm in (B).

provided by S Long) and Azorhizobium caulinodans (pRS2; Elmerich et al., 1982), respectively, were used in Southern hybridization experiments in order to veri- fy the presence of the equivalent genes in the original or reisolated bacterial strains, as described by Pawlowski et al. (1987).

Results and discussion

Abnormal structures on rice roots grown in China

Examination of root segments by stereomicroscopy using combined transmitted/incident illumination revealed various brown, hemispherical structures attached to the root surface (Fig. 2A). The same roots viewed by darkfield stereomicroscopy also indicated the presence of several thin, transparent filaments (not root hairs) protruding externally from these dome- shaped structures, and various scattered, dark lesions of the root epidermis that appear to have undergone hypersensitive responses (Fig. 2B). Brightfield light microscopy of thin-sectioned tissue clearly indicated that these dome-shaped structures developed external to the root and did not involve a hypertrophy of the root epidermis, cortex, or endodermis itself (Fig. 3Aj. Further examination of these cross-sections at higher magnification showed that filamentous, septated cells indicative of fungal hyphae originating from the external dome-shaped structure did penetrate the otherwise intact root epidermis and the underlying root cortex (Fig. 3B).

Root longitudinal sections revealed more extensive colonization of the root cortex just beneath the external dome structures (Fig. 3C). These dome-shaped external structures disintegrated during attempts to clear them with dilute sodium hypochlorite solution. However, they remained intact when pealed off the root and were processed for plastic embedding. Thin- sections of the isolated, larger (presumably older) hemispherical structures revealed an external, acellu- lar encrusted layer, an underlying extensive network of cells, and a central zone containing several large sacks each containing multiple, spherical, thick-walled cells resembling spores with prominent surface ornamentations and internal toluidine-blue staining granules (Fig. 4A,B).

An even clearer indication of the histological organization of these structures was obtained by scanning electron microscopy of their exposed cut surface. This method revealed in greater detail the attachment of these structures to the root, the outer encrusted protec- tive layer, the internal network of filamentous hyphae, and the centrally located sacks containing multiple spherical spore-like cells with reticulate surface orna- mentation (Fig. 5A-D). The presence of bacteria within small, hemispherical structures attached to the root was examined by staining with acridine orange solution followed by laser scanning microscopy in the epifluorescence confocal mode. Optical sections revealed a morphologically diverse community of bacteria dis- tributed on the surface and within these hemispherical structures attached to the mot (Fig. 6A,B).

It was not possible to match unequivocally the mor- phology of these hemispherical structures on rice roots

234

Fig. 7. Unusual modified lateral root meristems on rice viewed by (A) darkfield stereomicroscopy of fresh root tissue; (B) brightfield light microscopy after clearing in sodium hypochlorite and staining with methylene blue. Bar scales are 300 pm in (A), 100 pm in (B).

to any known fungus. Based upon interpretation of photomicrographs by three mycologists who specialize in soil fungi, the following possibilities were suggested: endogomes of chlamydospores of endomycorrhizal fungi; ascocarps of an ascomycete, and a chytrid. It remains a possibility that more than one fungus occu- pies these hemispheric structures on rice roots, and in addition, they are certainly colonized by a diversity of bacteria.

Abnormal structures on rice roots grown at Michigan State University

Three of the strains used by Li et al. (1991) were also used to infect rice plant roots at MSU. At a low frequency (O.l-O.S% of infected plants) different types of small protruding structures, often present in clusters, were observed on the rice roots. One structure resembling the hypertrophies found on the rice roots in Beijing was identified. A second class of structures

235

Fig. 8. Scanning electron micrographs of unusual modified lateral root meristems on rice roots, (A, B) Longitudinal sections of modified lateral meristems cut perpendicular to the parent rice root axis shown in cross-section. (C, D) Bacterial endorhizosphere colonization of the rice parent root and modified lateral root meristem. Bar scales are 100 pm in (A, B), 10 pm in (C, D).

was found to consist of stunted (modified) lateral roots, and these structures were found to be colonized by bacteria.

Stereomicroscopic examination of fresh roots infected by the strains described by Li et al. (1991) revealed infrequent clusters of hemispherical and dome-shaped “pseudonodules” protruding from the surface of thin, young rootlets (Fig. 7A). In contrast to the fungus colonized, hemispheric structures described above, the hemispheric structures illustrated in Fig- ure 7A were derived from the rice root itself. Bright- field microscopy of cleared and stained root segments revealed that these rounded protrusions were short, thick, lateral root meristems with a central vascular

system connected to the vascular system of the parent rootlet and an abnormal cortical hypertrophy (Fig. 7B). The histology of these modified lateral roots was confirmed by scanning electron microscopy of the longitudinal cut surface sectioned perpendicular to the parent rootlet axis (Fig. 8A,B).

Scanning electron microscopy at higher magnification revealed an extensive bacterial colonization of the rice endorhizosphere, extending into the vascular system of the parent rice root and these modified lateral roots (Fig. 8C,D). Combined light and transmission electron microscopy confirmed the presence of bacteria in association with these modified lateral root meristems, including their extensive coloniza-

236

tion of the rhizoplane, entry through cracks created at relate to the finding of modified lateral roots of rice col- emergence of these modified root structures, intercel- onized by Bradyrhizobium from Parasponia (Cocking lular dissemination and colonization, and colonization et al., 1992) and genetically engineered A. legumi- of dead internal plant cells (Fig. 9A-J). These results nosarum bv. trifolii (Rolfe et al., 1992). However,

237

Fig. 9. Combined light microscopy (A-F) and transmission electron microscopy (G-J) of modified lateral root meristems on rice roots, showing bacterial colonization of the rhizoplane, crack entry, intercellular dissemination, and colonization of dead plant cells. Bar scales are 100 /em ia (A), 20 pm in (B-F), 5 pm in (G-H), 1 pm in (I), and 0.5 pm in (J).

in contrast to the latter two reports, we have not yet found evidence of a bacterial endosymbiotic state (as would be suggested by membrane-enclosed bacteria within intact host cell cytoplasm). Further studies on plants grown under microbiologically controlled con- ditions will be necessary to address this issue. In any event, we can already conclude that modified lateral root meristems develop on healthy rice plants with-

out added hormones and that these are invaded by a substantial population of endorhizosphere-colonizing bacteria without the host exhibiting obvious symptoms of disease.

A molecular genetic analysis of the bacterial strains used by Li et al. (199 1) and bacteria reisolated from the MSU structures using rep-PCR genomic fingerprinting (de Bruijn, 1992) and hybridization with rhizo-

238

bial nitrogen fixation (ni~ and nodulation (nod) gene probes, was also initiated and revealed a number of discrepancies with the published results (Li et al., 1991).

For example, experiments with the nodABC probe revealed distinct hybridization patterns with the original rhizobial isolate from Sesbania cannabina (H18; Li et al., 1991), but the strains A201, A301, R101, and R20 (Y Jing, unpubl.), which were reisolated from nodule-like structures induced on rice roots in China (equivalent to isolates Rrl , Rr2; Li et al., 1991) failed to hybridize with this probe (data not shown). Simi- lar results were obtained with the nifHDK probe (data not shown), suggesting that at least some of the reisolated bacterial strains are not (directly) derived from the parental strain, as suggested previously (Li et al., 1991). These results were also confirmed by the rep- PCR genomic fingerprinting studies, which showed that although some of the reisolates reported by Li et al. (1991) or isolated in subsequent experiments in China (Y Jing, unpubl, results) shared common fingerprinting patterns, in most cases they appeared to be unrelated to one another and to the parental Rhizobium strain from Sesbania cannabina (Maria Schneider, Y Jing and F J de Bruijn, unpubl, results).

Unfortunately, attempts to reisolate bacteria from the hypertrophies (nodule-like structures) generated at MSU were unsuccessful, due to a too severe surface sterilization protocol. Therefore, the relationship of the bacteria seen via the microscope in these structures (see above) to the strain used to inoculate the plants could not be established. These studies suggest that the phenomenon of colonization of rice root hypertrophies reported by Li et al. (1991) and here must be interpret- ed with extreme caution and may be more complex than initially proposed. More experimentation will be required to prove Koch's postulate and to make the process of rice root infection more reproducible.

In conclusion, although the formation of hypertrophies on rice roots that are infected/colonized by microbes has been documented, little evidence sup- ports their designation as "nodules" (Li et al., 1991) or even nodule-like structures thus far. Moreover, the high frequency of their induction reported by Li et al. (1991) could not be reproduced, and the nature of the strains inducing the colonized hypertrophies and their relationship to the original strains remain unclear. Nevertheless, non-saprophytic colonization of rice root endorhizospheres can be observed in this system, pro- viding clear incentive for further studies in this important area of research.

Acknowledgements

We thank Joann Palma, H Stuart Pankratz, and Joanne Whallon for technical assistance; Jon Stoltzfus for comments on the manuscript; Maria Schneider for technical assistance; Karen Bird for excellent editorial assistance; and J K Ladha, Gary Stacey, Ron Petersen, Barbara Hetrich and Amy Rossman for helpful dis- cussions. The experimental work discussed here was supported by a grant from USAID (RN-5600-G-00- 2073-00) to FJdB and Dr J K Ladha (IRRI, The Philip- pines).

References

AI-MaUah M K, Davey M R and Cocking E C 1989 Formation of nodular structures on rice seedlings by rhizobia. J. Exp. Bot. 40, 473-478.

Appleby C A 1984 Leghemoglobin and Rhizobium respiration. Annu. Rev. Plant Physiol. 35,443--478.

Bender G L, Preston L, Bamard D and Rolfe B G 1990 Formation of nodule-like structures on the roots of the non-legumes rice and wheat. In Nitrogen Fixation: Achievements and Objectives. Eds. P M Gresshoff, L E Roth, G Stacey and W E Newton. p 825. Chapman and Hall, New York.

Boddey R M, de Oliveira O C, Urquiaga S, Resi V M, de Olivares F L, Baldani V L D and DObereiner J 1995 Biological nitrogen fixation associated with sugar cane and rice: contributions and prospects for improvement. Plant and Soil 174.

Bogusz D, Appleby C A, Landsmann J, Dennis E S, Trinick N J and Peacock W J 1988 Functioning hemoglobin genes in non- nodulating plants. Nature 331,178-180.

Chen T-W 1993 Nitrogen fixation of Azorhizobium in artificially induced root para-nodules in wheat. Sci. China 35, 1463-1469.

Chen T-W, Scherer S and Btiger 1993 Nitrogen fixation of Azorhi- zobium in artificially induced root para-nodules in wheat. In Advances in Molecular Genetics of Plant-Microbe Interactions. Eds, E W Nester and D P S Verma. pp 593-606. Kluwer Academic Publishers, Dordrecht.

Christansen-Weniger C and Vanderleyden J 1994 Ammonium- excreting Azospirillum sp. become intracellulariy established in maize (Zea mays) para nodules. Biol. Fertil. Soils 17, 1-8.

Cocking E C and Davey M R 1991 Nitrogen from the air for nonlegume crops. Chem., London 18 Nov, 831-835.

Cocking E C, AI-Mallah M K, Benson E and Davey M R 1990 Nodulation of non-legumes by rhizobia. In Nitrogen Fixation: Achievements and Objectives. Eds. P M Gresshoff, L E Roth, G Stacey and W E Newton. pp 813-823. Chapman and Hall, New York.

Cocking E, Srivastava J, Kothari S and Davey M 1992 Invasion of nonlegume plants by diazotrophic bacteria. In Nodulation and Nitrogen Fixation in Rice: Potentials and Prospects. Eds. G Khush and J Bennett. pp 119-121. IRRI, Manila.

David M, Daveran M-L, Batut J, Dedieu A, Domergue O, Ghai J, Hertig C, Boistard P and Kahn D 1988 Cascade regulation ofnif gene expression in Rhizobium meliloti. Cell 54, 671-683.

Dazzo F 1982 Leguminous root nodules. In Experimental Micro- bial Ecology. Eds. R Burns and J Slater. pp 431--466. Blackweli Scientic Publications, Oxford, UK.

239

Dazzo E Mateos P, Orgambide G, Philip-Hollingsworth S, Squar- tini A, Subba-Rao N, Pankratz H S, Baker D, Hollingsworth R and Whallon J 1993 The infection process in the Rhizobium- legume symbiosis and visualization of rhizoplane microorganisms by laser scanning confocal microscopy. In Trends in Micro- bial Ecology. Eds. R Guerrero and C Pedros-Alio. pp 259-262. Spanish Society for Microbiology, Barcelona.

De Bruijn F J 1992 Use of repetitive (repetitive extragenic palindromic and enterobacterial repetitive intergenic consensus) sequences and the polymerase chain reaction to fingerprint the genomes of Rhizobium meliloti isolates and other soil bacteria. Appl, Environ. Microbiol. 58, 218002187.

De Bruijn F J, Chen R, Dehio C, Goel A, Reiser S and Szczyglowski K 1992 Hormones and nodule formation: cytokinin induction of the Sesbania rostrata early nodulin gene Enod2. In Nodulation and Nitrogen Fixation in Rice: Potential and Prospects. Eds. G S Khush and J Bennett. pp 33-40. IRRI, Manila.

De Bruijn F J, Chen R, Fujimoto S Y, Pinaev A, Silver D and Szczyglowski K 1994 Regulation of nodulin gene expression. Plant and Soil 161, 59-68.

De Bruijn F J and Do~vnie J A 1991 Biochemical and molecular studies of symbiotic nitrogen fixation. Curr. Op. Biotech. 2, 184- 192.

De Bruijn F J and Schell J 1992 Regulation of plant genes specifically induced in developing and mature nitrogen-fixing nodules: Cis- acting elements and tram-acting factors. In Control of Plant Gene Expression. Ed. D P S Verma. pp 241-258. CRC Press, Boca Raton, Florida.

Dehio C and de Bruijn F J 1992 The early nodulin gene SrEnod2 from Sesbania rostrata is inducible by cytokinin. Plant J. 2, 117-128.

D6nari6 J and Cullimore J 1993 Lipo-oligosaccharide nodulation. factors: a new class of signaling molecules mediating recognition and morphogenesis. Cell 74, 951-954.

DObereiner J, Reis V M, Paula M A and Olivares F 1993 Endophyt- ic diazotrophs in sugar cane, cereals and tuber plants. In New Horizons in Nitrogen Fixation. Eds. R Palacios, J Mora and W E Newton. pp 671-676. Kluwer Acad. Publ., Dordrecht, Nether- lands.

Earl C D and Ausubel F M 1983 The genetic engineering of nitrogen fixation. Nutr. Rev. 41, 1--6.

Elmerich C, Dreyfus B, Reysset G and Aubert J P 1982 Genetic analysis of nitrogen fixation in a tropical fast-growing Rhizobium. EMBO J. 1,499-503.

Fischer H-M 1994 Genetic regulation of nitrogen fxation in rhizobia. Microbiol. Rev. 58, 352-386.

Fisher R F and Long S R 1992 Rhizobium-plant signal exchange. Nature 357, 655-660.

Gilles-Gonzalez M A, Ditta G S and Helinski D R 1991 A baemo- protein with kinase activity encoded by the oxygen sensor of Rhizobium meliloti. Nature 350, 170-172.

Hirsch A M 1992 Tansley review no. 40: Developmental biology of legume nodulation. New Phytol. 122, 211-237.

Hollingsworth R, Squartini A, Philip-Hollingsworth S and Dazzo F 1989 Root hair deforming and nodule initiating factors from Rhi- zobiura trifolii. In Signal Molecules in Plants and Plant-Microbe Interactions. Ed. B Lugtenberg. pp 387-393. Spfinger-Vedag, Berlin.

Horvath B, Heidstra R, Lados M, Moerman M, Spaink H E Prom~ J- C, van Kammen A and Bisseling T 1993 Lipo-oligosaccharides of Rhizobium induce infection-related early nodulin gene expression in pea root hairs. Plant J. 4, 727-733.

Hurek T, Reinhold-Hurek B, van Montagu M and Kellenberger E 1994 Root colonization and systemic spreading of Azoarcus sp. strain BH72 in grasses. J. Bacteriol. 176, 1913-1923.

Jing Y, Li G, Jin G, Shah X, Zhang B, Guan C and Li J 1990a Rice root nodules with acetylene reduction activity. In Nitrogen Fixation: Achievements and Objectives. Eds. P M Gresshoff, L E Roth, G Stacey and W E Newton. p 829. Chapman and Hall, New York.

Jing Y, Li G and Shah X 1992 Development of nodulelike struc- tare on rice roots. In Nodulation and Nitrogen Fixation in Rice: Potentials and prospects Eds. G Khush and J Bennett pp 123-126. IRRI, Manila, Phillippines.

Jing Y, Zhang B T and Shan X Q 1990b Pseudonodules formation on barley roots induced by Rhizobium astragali. FEMS Microbiol. Lett. 69, 123-128.

Kennedy I R and Tchan Y-T 1992 Biological nitrogen fixation in non-leguminous field crops: Recent advances. Plant and Soil 141, 93-118.

Khush G S and Bennett J (Eds.) 1992 Nodulation and Nitrogen Fixation in Rice: Potential and Prospects. IRRI Press, Manila,

Ladha J K, Trol-Padre A, Reddy K and Ventura W 1993 Prospects and problems of biological nitrogen fixation in rice production: A critical assessment. In New Horizons in Nitrogen Fixation. Eds. R Palacios et al. pp. 677-682. Kluwer Academic Publishers, Dordrecht.

Lauridsen P, Franssen H, Stougaard J, Bisseling T and Marcker K A 1993 Conserved regulation of the soybean early nodulin Enod2 gene promoter in determinate and indeterminate transgenic root nodules. Plant J. 3,483-492.

Li G, Jing Y, Shan X, Wang H and Guan C 1991 Identification office nodules that contain Rhizobium bacteria. Chin. J. Bot. 3, 8-17.

Orgambide G G, Hollingsworth R I and Dazzo F B 1992 Structural characterization of a novel diglycosyl diacylglyceride glycolipid from Rhizobium trifolii ANU 843. Carbohydr. Res, 233, 151-159.

Orgambide G G, Philip-Hollingsworth S, Hollingsworth R I and Dazzo F B 1994 Flavone-enhanced accumulation and symbiosis- related biological activity of a diglycosyl diacylglycerol membrane glycolipid from Rhizobium leguminosarum biovar trifolii. J. Bacteriol. 176, 4338-4347.

Pawlowski K, Ratet P, Schell J and de Bruijn F J 1987 Cloning and characterization of nifA and ntrC genes of the stem nodulating bacterium ORS571, the nitrogen fixing symbiont of Sesbania rostrata: Regulation of nitrogen fixation (n/l) genes in the free living versus symbiotic state. Mol. Gen. Genet. 206, 207-219.

Peoples M B, Herridge D F and Ladha J K 1995 Biological nitrogen fixation: an efficient source of nitrogen for sustainable agriculture? Plant and Soil 174.

Peters N K and Verma D P S 1990 Phenolic compounds as regulators of gene expression in plant-microbe interactions. Mol. Plant- Micr. Interact. 3, 4--8.

Phillips D A 1992 Flavonoids: Plant signals to soil microbes. In Phenolic Metabolism in Plants. Ed. H A Stafford. pp 201-231. Plenum Press, New York, USA.

Phillips D A, Dakora F D, Sande E, Joseph C M and Zon J 1994 Synthesis, release, and transmission of alfalfa signals to rhizobial symbionts. Plant and Soil 161, 69-80.

Quispel A 1992 A search for signals in endophytic microorganisms. In Molecular Signals in Plant-Microbe Communications. Ed. D P S Verma. pp 475-491. CRC Press, Boca Raton, FL.

Ridge R W, Bender G L and Rolfe B G 1992 Nodule-like structures induced on the roots of wheat seedlings by the addition of the synthetic auxin 2,4-dichloropbenoxyacetic acid and the effects of microorganisms. Aust. J. Plant Physiol. 19, 481-492.

Rolfe B G and Bender G L 1990 Evolving a Rhizobium for nonlegume nodulation. In Nitrogen Fixation: Achievements and Objectives. Eds. P M Gresshoff, L E Roth, G Stacey and Vlr E Newton. pp 829. Chapman and Hall, New York.

240

Rolfe B, Ride K and Ridge R 1992 Rhizobium nodulation of non- legumes. In Nodulution and Nitrogen Fixation in Rice: Potentials and Prospects. Eds. G Khush and J Bennett. pp 83-86. IRRI, Manila.

Schubert K R 1986 Products of biological nitrogen fixation in higher plants: synthesis, transport, and metabolism. Ann. Rev. Plant Physiol. 37, 539-574.

Simon Moffatt A 1990 Nitrogen-fixing bacteria find new partners. Science 250, 910--912.

Szczyglowski K, Szabados L, Fujimoto S Y, Silver D and de Bruijn F J 1994 Site-specific mutagenesis of the nodule-infected-cell expression (NICE) element and the AT-rich element (ATRE- BS2*) of the Sesbania rostrata leghemoglobin glb3 promoter. Plant Cell 6, 317-332.

Tchan Y T, Zeman A M M and Kennedy I R 1991 Nitrogen fixation in para-nodules of wheat roots by introduced free-living diazotrophs. Plant and Soil 137, 43-47.

Truchet G Camut S, DeBill F, Ororico R and Vasse J 1989 The Rhizobium-legume symbiosis: two methods to discriminate between nodules and other root derived structures. Protoplasma 149, 82-88.

Umali-Garcia M, Hubbell D Gaskins M and Dazzo F 1980 Associa- tion of Azospirillum with grass roots. Appl. Environ. Microbiol. 39, 219-226.

Van Kammen A 1984 Suggested nomenclature for plant genes involved in nodulation and symbiosis. Plant Mol. Biol. Rep. 2, 43-45.

Vijn L, das Neves L, van Kammen A, Franssen H and Bisseling T 1993 Nod factors and nodulation in plants. Science 260, 1764- 1765.

Welters P, Metz B, Felix G, Palme K, Szczyglowski K and de Bruijn F J 1993 Interaction ofa rhizobiai DNA-binding protein with the promoter region of a plant leghemoglobin gene. Plant Physiol. 102, 1095-1107.

Yu D, Kennedy I R and Tchan Y T 1993 Verification of nitrogenase activity (C2H2 reduction) in Azospirillum populated, 2,4- dichlorophenoxyacetic acid induced root structures of wheat. Aust. J. Plant Physiol. 29, 187-195.

potential and pitfalls of trying to extend symbiotic ... · we briefly review this research and...

Documents