OUP user menu

Diversity and occurrence of Burkholderia spp. in the natural environment

Stéphane Compant, Jerzy Nowak, Tom Coenye, Christophe Clément, Essaïd Ait Barka
DOI: http://dx.doi.org/10.1111/j.1574-6976.2008.00113.x 607-626 First published online: 1 July 2008

Abstract

Both in natural and in managed ecosystems, bacteria are common inhabitants of the phytosphere and the internal tissues of plants. Probably the most diverse and environmentally adaptable plant-associated bacteria belong to the genus Burkholderia. This genus is well-known for its human, animal and plant pathogenic members, including the Burkholderia cepacia complex. However, it also contains species and strains that are beneficial to plants and can be potentially exploited in biotechnological processes. Here we present an overview of plant-associated Burkholderia spp. with special emphasis on beneficial plant–Burkholderia interactions. A discussion of the potential for utilization of stable plant–Burkholderia spp. associations in the development of low-input cropping systems is also provided.

Keywords
  • Burkholderia
  • pathogen
  • phytosphere
  • rhizosphere
  • endophytic bacteria
  • symbiotic bacteria

Introduction

Over the past two decades, research on Burkholderia species has been steadily expanding. Members of the genus Burkholderia are very abundant, occupying diverse ecological niches (Estrada-de los Santos, 2001; Coenye & Vandamme, 2003), including soil (van Elsas, 2002; Salles et al., 2002, 2004; Janssen et al., 2006) and hospital (Coenye & Vandamme, 2003) environments. Many members of the genus can cause infections in humans and animals (Coenye & Vandamme, 2003; Valvano et al., 2005, 2006). Nevertheless, in recent years, a growing number of Burkholderia strains and species have also been reported as plant-associated bacteria. Indeed, Burkholderia spp. can be free-living in the rhizosphere as well as epiphytic and endophytic, including obligate endosymbionts and phytopathogen (Coenye & Vandamme, 2003; Janssen et al., 2006). Several strains are known to enhance disease resistance in plants, contribute to better water management, and improve nitrogen fixation and overall host adaptation to environmental stresses (Coenye & Vandamme, 2003; Nowak & Shulaev, 2003; Compant et al., 2005a; Sessitsch et al., 2005; Ait Barka, 2006; Barrett & Parker, 2006; Janssen et al., 2006; Balandreau & Mavingui, 2007). These findings have stimulated a growing interest in using Burkholderia isolates in agriculture. Due to the fact that some species/isolates can be opportunistic or obligate pathogens causing human, animal or plant diseases, any development of agricultural and/or biotechnological applications using Burkholderia germplasm needs to include a stringent assessment of the potential risks (Coenye & Vandamme, 2003). This review summarizes the current knowledge of interactions between Burkholderia species and plants, and discusses the potential for utilization of some members of the genus in agriculture and biotechnology.

Overview of the genus Burkholderia

In 1942, Walter H. Burkholder described one of the first Burkholderia sp., Phytomonas caryophylli (Burkholder et al., 1942), later known as Pseudomonas caryophylli. In 1949, he also described a phytopathogenic bacterium, which caused rot in onion bulbs as reported by vegetable growers in New York State in the mid 1940s and gave it the species name ‘cepacia’, meaning derived from ‘onion’ (Burkholder et al., 1950); it was later known as Pseudomonas cepacia. Burkholderia spp. were for many years included in the genus Pseudomonas owing to its broad and vague phenotypic definition. However, rRNA–DNA hybridization analyses during the early 1970s indicated considerable genetic diversity among members of this genus (reviewed in Kersters et al., 1996), which was thus divided into five so-called rRNA homology groups (Palleroni et al., 1973). Subsequent genotypic analyses have confirmed that these five groups are only distantly related to each other. Consequently, Pseudomonas s.s. was restricted solely to homology group I, containing the type species, Pseudomonas aeruginosa (De Vos, 1985). In 1992, the seven species belonging to rRNA homology group II (Pseudomonas solanacearum, Pseudomonas pickettii, P. cepacia, Pseudomonas gladioli, Pseudomonas mallei, P. caryophylli) were transferred to the novel genus Burkholderia (Yabuuchi et al., 1992), which resides in rRNA superfamily III sensuDe Ley (1992) or subgroup β-3 of the Betaproteobacteria sensuWoese (1987). A considerable number of species have been included in the genus Burkholderia in recent years (Coenye & Vandamme, 2003). Many of these changes involve Burkholderia cepacia as several polyphasic taxonomic studies (Vandamme et al., 1997, 2000, 2002, 2003; Coenye et al., 2001b, c, d; Vermis et al., 2004) have indicated that strains identified phenotypically as B. cepacia represent a complex of several closely related genomic species or genomovars. This group, collectively referred to as the B. cepacia complex, currently consists of nine species, including B. cepacia (genomovar I), Burkholderia multivorans (genomovar II), Burkholderia cenocepacia (genomovar III), Burkholderia stabilis (genomovar IV), Burkholderia vietnamiensis (genomovar V), Burkholderia dolosa (genomovar VI), Burkholderia ambifaria (genomovar VII), Burkholderia anthina (genomovar VIII) and Burkholderia pyrrocinia (genomovar IX).

Many other species of Burkholderia have been described since the discovery of B. cepacia by W.H. Burkholder and there are currently more than 40 validly described species (see http://www.bacterio.cict.fr/b/burkholderia.html for an up-to-date overview). Most of these species interact with plants and can be phytopathogens, endosymbionts in phytopathogenic fungi as well as plant-associated insects, although many can be neutral or beneficial for plants and have an intimate association with their hosts.

Phytopathogens in the genus Burkholderia

As noted before, several species of the genus Bukholderia can induce plant diseases (Table 1). For example, the well-known pathogenic bacterium B. cepacia is involved in this type of interaction (Table 1). One disease caused by this Burkholderia species is onion rot mediated by the infection of onion leaves and bulbs (Burkholder, 1950), but other plants can also be infected (Table 1). Another phytopathogenic Burkholderia species, Burkholderia caryophylli, induces formation of bacterial wilt in various plant species, including Russel prairie gentian in Japan (Furuya et al., 2000). It was also in Japan that Burkholderia plantarii was first isolated in 1982 (Azegami et al., 1987). Burkholderia plantarii provokes seedling blight on rice as well as on koyawarabi (Onoclea sensibilis L.) (Azegami et al., 1987; Tanaka & Katoh, 1999). Burkholderia gladioli induces bacterial soft rot in onions, leaf-sheath browning and grain rot in rice, and leaf and corm diseases in gladiolus and iris species (Palleroni et al., 1984; Lee et al., 2005; Ura et al., 2006), and also infects many other plants (Table 1), suggesting that pathovars of B. gladioli have different host ranges. Two other known phytopathogenic Burkholderia species are Burkholderia glumae and Burkholderia androponis. Burkholderia glumae causes seedling and grain rot in rice and wilting symptoms in tomato, sesame (Sesamum indicum L.), perilla (Perilla frutescens), eggplant and hot pepper (Jeong et al., 2003) as well as in 20 other plant species (Table 1). In the case of B. andropogonis, more than 52 species of 15 families of unrelated monocotyledonous and dicotyledonous plants can be infected (Table 1). This later phytopathogenic bacterium is thus an important causal agent of stripe disease of sorghum (Andropogon sp.), leaf spot of clover (Trifolium sp.), Odontioda orchids, velvet bean (Stizolobium deeringianum), jojoba (Simmondsia chinensis), Amaranthus cruentus L. and of Bougainvillea sp., and in addition causes considerable economic loss of carnation (Dianthus caryophyllus) (Smith et al., 1911; Murai & Goto, 1996; Scortichini et al., 2001; Cother et al., 2004; Takahashi et al., 2004; Li & De Boer, 2005). It is possible that in the future other species will also be correlated with negative effects on plants. Indeed, the ecological role of some Burkholderia spp., such as Burkholderia glathei, Burkholderia graminis, Burkholderia phenazinium, Burkholderia caribensis, Burkholderia caledonica, Burkholderia hospita, Burkholderia terricola and Burkholderia saccharii, remains largely unknown at this time (Coenye & Vandamme, 2003; Vandamme et al., 2007a). Their ecological niches and roles on some plants will probably be identified in the near future and other members of the genus Burkholderia are likely to be revealed as phytopathogenic agents.

View this table:
1

Phytopathogenic members of the genus Burkholderia and their plant hosts

SpeciesInfected plantReferences
B. andropogonisAmaranthus cruentus L.Bradbury et al. (1973)
Andropogon sp.Li & De Boer (2005)
Bougainvillea sp.Takahashi et al. (2004)
Dianthus caryophyllus L.
Dolichos lablab L.
Euchlaena mexicana Schrad.
Lespedeza sp.
Medicago sativa L.
Mucuna deeringiana (Bort) Merr.
Odontoglossum sp.
Phaseolus vulgaris L.
Simmondsia chinensis Link.
Sorghum bicolor (L.) Moench
Sorghum halepense (L.) Pers.
Saccharum officinarum L.
Stizolobium deeringianum Bort.
Trifolium pratense L.
Trifolium repens L.
Trifolium subterraneum L.
Vicia faba L.
Vicia sativa L.
Zea mays L.
B. caryophylliDianthus caryophyllus L.Saddler et al. (1994a)
Gypsophila paniculata L.
Helianthus annuus L.
Limonium sinuatum L.
B. cepaciaAllium cepa L.Saddler et al. (1994b)
Allium sativum L.
Cymbidium spp.
Dendrobium sp.
Lycopersicon esculentum Mill.
Paphiopedilum spp.
B. gladioli pv. alliicolaAllium cepa L.Saddler et al. (1994c)
Tulipa spp.
B. gladioli pv. gladioliAdiantum sp.Saddler et al. (1994d)
Asplenium nidus L.Ura et al. (2006)
Crocus spp.
Cyrtomium falcatum (L. f.) Presl.
Davallia fejeensis Hook.
Dendrobium spp.
Freesia refracta Jacq.
Freesia x hybrida Hort.
Gladiolus x hortulanus L.
Gladiolus x colvillei Sweet.
Iris spp.
Ixia maculata L.
Oryza sativa L.
Pteris ensiformis L.
Pellaea rotundifolia Hook
Platycerium bifurcatum Cav.
Pteris cretica L.
Tigridia pavonia (L. f.) D.C.
B. glumaeAndropogon virginicus L.Saddler et al. (1994e)
Arundinella hirta L.Jeong et al. (2003)
Beckmannia syzigachne (Steud) Fernald
Chloris gayana Kth.
Coix lacryma-jobi L.
Eleusine indica (L.) Gaertn.
Eleusine coracana (L.) Gaertn.
Eragrostis curvula (Schrad.) Nees
Eragrostis pilosa (L.) Beauv.
Lolium multiflorum Lam.
Lycopersicon esculentum Mill.
Oryza sativa L.
Panicum coloratum L.
Panicum dichotomiflorum Michaux
Panicum maximum Jacq.
Paspalum dilatatum Poiret
Paspalum distichum L.
Pennisetum alopecuroides (L.) Spreng.
Perilla frutescens (L.) Britton
Phleum pratense L.
Phragmites australis (Cav.) Steud
Sesamum indicum L.
Setaria viridis (L.) Beauv.
Solanum melongena L.
B. plantariiOnoclea sensibilis L.Tanaka & Katoh (1999)
Oryza sativa L.Azegami et al. (1987)

Endosymbiotic Burkholderia spp. in phytopathogenic fungi

It is clear that some phytopathogenic fungi can contain members of the genus Burkholderia as endosymbionts (Table 2). For example, Burkholderia fungorum has been isolated from the white-rot fungus Phanerochaete chrysosporium, which can induce diseases on trees (Coenye et al., 2001a). Similar to this association, Burkholderia sordidicola has been detected inside the phytopathogenic fungus Phanerochaete sordida, which inhabits fallen branches of hardwood trees (Lim et al., 2003). Based on 16S rRNA gene sequence similarity, B. fungorum and B. sordidicola are only distantly related (Lim et al., 2003), suggesting that associations between fungi and Burkholderia spp. have arisen on multiple occasions. However, it is important to note that although these species are associated with fungi, they are not responsible for diseases themselves. Rather, they interact with their hosts, which cause plant diseases. Nevertheless, some other Burkholderia species can produce phytotoxins. Recent studies have demonstrated that the ‘mycotoxin’ rhizoxin causing rice seedling blight, which was previously attributed to Rhizopus microsporus, is in fact secreted by a member of the genus Burkholderia that thrives as an endosymbiont in the fungus (Partida-Martinez & Hertweck, 2005; Scherlach et al., 2006). Rhizoxin inhibits mitosis in rice plant cells and effectively weakens or even kills the plant. Both the host and the symbiont benefit from nutrients derived from the decaying plant material (Partida-Martinez & Hertweck, 2005). The fungus profits from the biosynthetic capabilities of the endosymbiont in order to access nutrients and it may be possible that the fungus manipulates the bacterium to produce the toxin. This intriguing phytopathogenic alliance against rice seedlings represents an unprecedented example in which a fungus hosts a bacterial population for the production of a virulence factor (Partida-Martinez & Hertweck, 2005). In a subsequent study, Partida-Martinez (2007a) found that the endofungal bacteria are also responsible for the production of the ‘mycotoxin’ rhizonin. This endosymbiont of the plant-pathogenic fungus R. microsporus was identified as Burkholderia rhizoxinica sp. nov. (Partida-Martinez et al., 2007b). These authors, report Burkholderia endofungorum sp. nov. also as a bacterial endosymbiont of the plant-pathogenic fungus R. microsporus. The phylogenetic placement of these novel bacterial species, which form endosymbioses with zygomycota, gives further evidence of the broad ecological amplitude of the genus Burkholderia.

View this table:
2

Burkholderia spp. associated with pathogenic fungi

SpeciesFungiReferences
B. fungorumPhanerochaete chrysosporium Burds.Coenye et al. (2001a)
B. sordidicolaPhanerochaete sordida Karst. Höhn & LitschLim et al. (2003)
B. rhizoxinicaRhizopus microsporusPartida-Martinez et al. (2007a)
B. endofungorumR. microsporusPartida-Martinez et al. (2007b)

Endosymbiotic Burkholderia spp. in plant-associated insects

Some Burkholderia species are known as inhabitants associated with insects feeding on plants and also with insects that can be used in biocontrol (Table 3). For instance, the plant-associated Japanese common broad-headed bugs Riptortus clavatus and Leptocorisa chinensis contain bacterial endosymbionts (Kikuchi et al., 2005). These insects possess a number of crypts in the posterior region of the midgut, the lumen of which contains a large amount of bacterial cells.

View this table:
3

Burkholderia spp. in plant-associated insects

SpeciesInfected plantReferences
B. cenocepaciaAtta sexdens rubropilosa ForelSantos et al. (2004)
B. fungorumHarpalus pensylvanicus DeGeerLundgren et al. (2007)
Burkholderia spp.Leptocorisa chinensis DallasKikuchi et al. (2005)
Riptortus clavatus ThunbergKikuchi et al. (2005)
Tetraponera binghami Forelvan Borm (2002)

In a number of pentatomomorphan bugs, symbiont-free insects exhibit retarded growth, nymphal mortality and/or sterility (Fukatsu & Hosokawa, 2002). A dense and specific colonization of the midgut crypts by the symbiotic bacteria has been found in R. clavatus and L. chinensis in some natural populations (Kikuchi et al., 2005). The predominant 16S rRNA gene sequences obtained from different individuals and species of the bugs and homology searches in the DNA databases of the bacteria have recently revealed that the sequences of these insect symbionts showed highest levels of similarity (96–99%) to the sequences of Burkholderia spp. In situ hybridizations with specific oligonucleotide probes have also confirmed that some Burkholderia spp. are localized in the lumen of the midgut crypts (Kikuchi et al., 2005). This demonstrates that symbiosis can take place between insects and members of the genus Burkholderia and that it is a specific interaction, other than those described in the case of fungi. Interestingly, this kind of symbiosis has been described not only for R. clavatus and L. chinensis, but also for other Alydidae as some Burkholderia symbionts were detected in the cryptic midguts of all six other species of alydid bugs examined. Recents results have also indicated that R. clavatus acquires Burkholderia symbionts postnatally from the environment every generation, uncovering a previously unknown pathway through which a highly specific insect–microbe association is maintained (Kikuchi et al., 2007). It appears plausible that bugs belonging to the family Alydidae generally carry Burkholderia spp. in midgut crypts (Kikuchi et al., 2005), and that during insect evolution, specificity may have arisen between Burkholderia species and members of the family Alydidae. Molecular phylogenetic analyses have recently demonstrated that the Burkholderia communities of the bugs formed a well-defined monophyletic group, confirming this hypothesis of specific coevolution, although the group also contained several environmental Burkholderia strains (Kikuchi et al., 2005). Recently, the intermingled phylogenetic patterns of Alydidae have suggested the possibility that there was horizontal transmission of the symbiont between populations and species of the alydid bugs (Kikuchi et al., 2005), which is consistent with specific interactions between these organisms. However, it is important to note that insects other than those of the family Alydidae interact also with Burkholderia sp. Van Borm (2002) found that the microflora inhabiting a pouch-shaped organ at the junction of the midgut and the intestine of Tetraponera binghami ants partially consists of Burkholderia species. This is further proof that members of this bacterial genus can be endosymbionts in plant-associated insects. The authors also suggest that these symbionts may fix nitrogen. However, no experimental data were provided for this statement. Santos (2004) described a more sophisticated symbiosis between Atta sexdens rubripilosa ants and B. cenocepacia. These ants establish fungus gardens to grow a symbiotic fungus, Leucoagaricus gongylophorus. The ant-associated Burkholderia isolates secreted a potent antifungal agent that inhibits germination of conidia of the entomopathogenic fungi Beauveria bassiana, Metarhizium anisopliae, of a saprophytic Verticillium lecanii, and of a fungus garden species Escovopsis weberi, but does not affect Leucoagaricus gongylophorus (Santos et al., 2004). This could be an example of an intimate biological control association between ants and Burkholderia that allows the insect to culture a specific fungus, eliminating direct competition from other fungi.

The recently discovered colonization of the intestinal tract of plant-beneficial ground beetles (Harpalus pensylvanicus) by B. fungorum, which may contribute to fitness of the beetles, may also provide additional knowledge on the mechanisms of the control of plant and insect pests as recently discussed in Lundgren (2007).

Nonphytopathogenic and beneficial Burkholderia spp. associated with the phytosphere

Although some members of Burkholderia are either pathogenic or associated with some diseases of their hosts, the vast majority of the plant-associated Burkholderia species are nonpathogenic, and can be either neutral or beneficial to their hosts (De Costa & Erabadupitiya, 2005). Some of them are able to establish epiphytic and endophytic populations colonizing the exterior and interior of plant organs and form an intimate association with their hosts. This can lead to the stimulation of plant growth, nitrogen fixation and/or enhancement of the host's resistance to abiotic and biotic stresses (Sharma & Nowak, 1998; Estrada-de los Santos, 2001; Nowak & Shulaev, 2003; Compant et al., 2005a; Sessitsch et al., 2005; Ait Barka, 2006). Some members of the genus Burkholderia are also able to induce plant nodulation (van Oevelen, 2002; Elliott et al., 2007). For instance, Elliott (2007) demonstrated for the first time that two B. tuberum strains, STM678 and DUS833, can nodulate Cyclopia species and papilionoid legumes from widely different tribes. Agricultural industry is interested in exploring this phenomenon although more knowledge needs to be generated on the ecology of the bacteria and molecular determinants of the nodulation process (Li et al., 2002; De Costa & Erabadupitiya, 2005).

Burkholderia spp. associated with the rhizosphere

It is now well established that some Burkholderia sp. can inhabit the plant rhizosphere, i.e. soil on root surfaces and adjacent to the roots (Table 4). By colonizing the rhizosphere Burkholderia spp. can form neutral or beneficial associations with their hosts, although the nature of these interactions cannot always be defined. Associations involve various crops and species and can even arise under different geographical zones.

View this table:
4

Rhizospheric Burkholderia spp. and their natural plant hosts

SpeciesPlant hostsReferences
B. ambifariaPisum sativum L.Richardson et al. (2002)
Saccharum officinarum L.Omarjee & Balandreau (2005)
Zea mays L.Fiore et al. (2001)
Richardson et al. (2002)
Alisi et al. (2005)
Ramette et al. (2005)
B. caledonicaSaccharum officinarum L.Omarjee & Balandreau (2005)
Vitis vinifera L.Coenye et al. (2001a)
B. cepaciaBetula sp.Richardson et al. (2002)
Equisetum sp.Richardson et al. (2002)
Gossypium sp.Parke & Gurian-Sherman (2001)
Pisum sativum L.Parke & Gurian-Sherman (2001)
Quercus sp.Richardson et al. (2002)
Senecio vulgaris L.Richardson et al. (2002)
Saccharum officinarum L.Omarjee & Balandreau (2005)
Triticum aestivum L.Richardson et al. (2002)
Oryza sativa L.Trân Van (1994)
Zantedeschia sp.Richardson et al. (2002)
Zea mays L.Ramette et al. (2005)
Fiore et al. (2001)
Bevivino et al. (2002)
B. cenocepaciaZea mays L.Balandreau et al. (2001)
Fiore et al. (2001)
Alisi et al. (2005)
Ramette et al. (2005)
B. dolosaZea mays L.Ramette et al. (2005)
B. graminisSaccharum officinarum L.Omarjee & Balandreau (2005)
Zea mays L.Viallard et al. (1998)
Triticum aestivum L.Viallard et al. (1998)
B. kururiensisLycopersicon esculentum Mill.Caballero-Mellado et al. (2007)
B. multivoransZea mays L.Ramette et al. (2005)
B. pyrrociniaZea mays L.Bevivino et al. (2002)
Alisi et al. (2005)
Ramette et al. (2005)
B. sacchariSaccharum officinarum L.Omarjee & Balandreau (2005)
B. silvatlanticaSaccharum officinarum L.Perin et al. (2006a)
Zea mays L.Perin et al. (2006a)
B. stabilisZea mays L.Bevivino et al. (2002)
Ramette et al. (2005)
B. tropicaLycopersicon esculentum Mill.Caballero-Mellado et al. (2007)
Saccharum officinarum L.Reis et al. (2004)
Zea mays L.Reis et al. (2004)
B. unamaeCoffea spp.Caballero-Mellado et al. (2004)
Lycopersicon esculentum Mill.Caballero-Mellado et al. (2007)
Saccharum officinarum L.Caballero-Mellado et al. (2004)
Zea mays L.Caballero-Mellado et al. (2004)
B. vietnamiensisCoffea arabica L.Estrada-de los Santos (2001)
Oryza sativa L.Trân Van (1994), Gillis (1995)
Zea mays L.Trân Van (1994)
Estrada-de los Santos (2001)
B. xenovoransCoffea sp.Goris et al. (2004)
Lycopersicon esculentum Mill.Caballero-Mellado et al. (2007)

Since 1990, different Burkholderia spp. have been found in the rhizosphere of various crops (Table 4) worldwide. For instance, the B. cepacia complex has been isolated from the rhizosphere of maize in the US (Hebbar et al., 1992; Ramette et al., 2005), of rice (Trân Van, 1994), pea (Parke et al., 1990), cotton (Parke & Gurian-Sherman, 2001), as well as of oak trees, common groundsel, horsetail, arum lily and wheat in the UK (Richardson et al., 2002). The rhizosphere is now considered as a reservoir for many other members of the genus Burkholderia (Berg et al., 2005). The nonpathogenic species B. graminis is a common rhizobacterium of corn, pasture and wheat in Australia and France (Viallard et al., 1998). Burkholderia unamae is associated with maize, sugarcane and coffee (Caballero-Mellado et al., 2004), B. ambifaria with pea in the USA (Richardson et al., 2002), B. silvatlantica is a common inhabitant of the rhizosphere of maize and sugarcane grown in Brazil (Perin et al., 2006a) and B. caledonica has been isolated from the rhizosphere of vine (Vitis sp.) in Scotland (Coenye et al., 2001a). Plant hosts can also be colonized by more than one species of the genus Burkholderia. Burkholderia vietnamiensis, which was first described in the rice rhizosphere in Vietnam (Trân Van, 1994; Gillis et al., 1995), was isolated along with other Burkholderia species from the rhizosphere of maize and coffee in Mexico (Estrada-de los Santos, 2001). Furthermore, a recent study on Sphagnum rhizosphere of peat bogs of the boreal and tundra zones of Russia, Canada and Estonia discovered similar associations between these lower plants and several Burkholderia spp. (Belova et al., 2006). To date such associations have been reported for over 30 plant species (Table 4).

Different soil/crop management practices and the land-use patterns also affect diversity of Burkholderia spp. in the rhizosphere of plant crops (Salles et al., 2006a, b). It is thus possible that the diversity of Burkholderia spp. colonizing crop plants in different geographical areas could be linked to human activity. Various nonpathogenic species can therefore interact within the plant rhizosphere and thus more beneficial plant–microbial associations can be formed. It is obvious that farmers can take benefits from these interactions. Although the largest Burkholderia communities comprise species that are nonpathogenic to plants (Salles et al., 2006b), increasingly intensive agricultural practices may unknowingly accelerate proliferation of human pathogens.

The endophytic Burkholderia spp. and their plant hosts

The rhizosphere is now considered as the major source of plant endophytes (Hallmann et al., 1997, 2001; Sturz et al., 2000; Compant et al., 2005a; Gray & Smith, 2005). Some rhizosphere bacteria are able to enter the root tissue, transcend the endodermis barrier, cross from the root cortex to the vascular system, and subsequently establish endophytic populations in various organs such as vegetative and reproductive organs. This has been described in a wide range of plants, including both monocotyledonous and dicotyledonous plants, and for various bacterial genera (Lodewyckx et al., 2002). Members of the genus Burkholderia are common endophytes (Table 5). For instance, Burkholderia phytofirmans (Sessitsch et al., 2005) has been isolated from onion roots in Canada (Nowak & Shulaev, 2003), and B. cepacia from branches of a Citrus sp. cultivated in Brazil (Araujo et al., 2002) and from roots of rice in India (Singh et al., 2006). Recent analysis of N2-fixing bacteria associated with roots of maize and coffee plants grown under field conditions has revealed several novel diazotrophic bacterial species belonging to the genus Burkholderia that may form both epiphytic and endophytic populations. These include B. vietnamiensis (Estrada-de los Santos, 2001; Estrada et al., 2002), Burkholderia tropica (Reis et al., 2004). Burkholderia cenocepacia was also isolated from the inner tissues of wheat, lupine and maize in a study conducted in France and Australia (Balandreau et al., 2001), B. silvatlantica, from surface-disinfected leaves of sugar cane in Brazil (Perin et al., 2006a), and B. unamae from maize and sugarcane also from Brazil (Caballero-Mellado et al., 2004). Endophytic Burkholderia sp. have even been found in gymnospermae as described by Bal & Chanway (2000) who isolated Burkholderia pyrrocinia from the stems of lodgepole pine and other Burkholderia spp. from stems of western red cedar grown in Canada. This demonstrates the wide host range of Burkholderia spp. and that plant internal tissues can therefore be considered as reservoirs for Burkholderia spp.

View this table:
5

Endophytic Burkholderia spp. and their natural plant hosts

SpeciesPlant hostsReferences
B. cepaciaCitrus sinensis (L.) OsbeckAraujo et al. (2002)
Oryza sativa L.Singh et al. (2006)
B. cenocepaciaTriticum aestivum L.Balandreau et al. (2001)
Lupinus sp.Balandreau et al. (2001)
Zea mays L.Balandreau et al. (2001)
B. gladioliCoffea sp.Vega et al. (2005)
Glycine max (L.) Merr.Kuklinsky-Sobral et al. (2005)
B. phytofirmansAllium cepa L.Sessitsch et al. (2005)
Oryza sativa L.Muthukumarasamy et al. (2007)
Shagnum spp.Belova et al. (2006)
B. pyrrociniaPinus contorta Dougl.Bal & Chanway (2000)
B. silvatlanticaSaccharum officinarum L.Perin et al. (2006a)
B. tropicaAnanas comosus (L.) Merr.Cruz et al. (2001)
Saccharum officinarum L.Reis et al. (2004)
Zea mays L.Reis et al. (2004)
B. unamaeSaccharum officinarum L.Caballero-Mellado et al. (2004)
Perin et al. (2006b)
Zea mays L.Caballero-Mellado et al. (2004)
Perin et al. (2006b)
B. vietamiensisZea mays L.Estrada-de los Santos (2001)

Vegetative plant tissues are not the only plant organs colonized by these Betaproteobacteria. Recent evidence suggests that some Burkholderia sp. can also thrive as endophytes in both vegetative and reproductive organs of their hosts. This includes B. tropica, which was isolated from the stems and fruits of pineapple in Brazil (Cruz et al., 2001), and B. gladioli from roots, stems, seeds and berries of coffee, together with other Burkholderia strains residing inside the pulp and seeds of the fruits of this plant (Vega et al., 2005). The above reports demonstrate both the wide host range for Burkholderia spp. and the potential for their spread via agronomic practices and trade in agricultural commodities.

Burkholderia spp. as endosymbionts of beneficial endophytic fungi

It is now well established that plant roots colonized by mycorrhizal fungi offer an excellent ecological niche for bacteria. Some rhizosphere bacteria adhere tightly to fungal hyphae (Bianciotto et al., 2001), and can penetrate intracellular locations within the mantle and Hartig net of ectomycorrhizal roots (Nurmiaho-Lassila et al., 1997), whereas others are directly associated with root surfaces (Bianciotto et al., 2001). The results of Timonen (1998) have suggested that the genus Burkholderia may be a component of this community. Burkholderia spp. found as fungal endosymbionts are listed in Table 6. An association of Burkholderia with ectomycorrhizal fungi has been demonstrated, with Suillus variegatus and Tomentellopsis submollis in two Corsican Pinus nigra by Izumi (2007). Garbaye & Bowen (1989) isolated Burkholderia species from inside the ectomycorrhizal mantle that had stimulatory effects on the mycelial growth of Rhizopagon luteolus, and, in some cases, enhanced formation of the mycorrhizal association between Pinus radiata and R. luteolus. Burkholderia spp. have also been reported as inhabitants of several members of the Gigasporaceae (Bianciotto et al., 2000; Ruiz-Lozano & Bonfante, 2000). Although these have been recently reassigned to a new genus as ‘Candidatus Glomeribacter gigasporarum’ (Bianciotto et al., 2003) and do not formally belong to the genus Burkholderia, their close relationship with the latter genus suggests that similar interactions involving Burkholderia species could be possible. This is confirmed by experiments of coinoculation of some Burkholderia spp., i.e. B. vietnamiensis, B. cepacia and B. pseudomallei, with Gigaspora decipiens, in which it was demonstrated that these species can be endosymbionts inside the arbuscular fungus (Levy et al., 2003). Although this is based on laboratory experiments, it is possible that future work will lead to the discovery of Burkholderia spp. inside Gigasporaceae, as this bacterial genus seems to be widely distributed in various fungi.

View this table:
6

Burkholderia spp. as endosymbionts of beneficial endophytic fungi

SpeciesFungiReferences
B. cepaciaGlomus etunicatum Beck.Andrade et al. (1997)
G. intraradices Schenck & SmithAndrade et al. (1997)
G. mosseae Nicol. & Gerd.Andrade et al. (1997)
Burkholderia spp.Suillus variegatus SwartzIzumi et al. (2007)
Tomentellopsis submollis SvrckIzumi et al. (2007)
Rhizopogon luteolus Fr.Garbaye & Bowen (1989)

Andrade (1997) have discovered that B. ‘cepacia’ was a ubiquitous inhabitant of the hyphosphere of three arbuscular fungi tested (Glomus etunicatum, Glomus intraradices and Glomus mosseae). None of the other bacterial species was as widely distributed. This finding demonstrates that some bacteria belonging to the genus Burkholderia, including potential human pathogens, are common inhabitants of endomycorrhiza, i.e. Glomeraceae. Although interactions between Burkholderia spp. and fungi other than Glomeraceae have not been detected so far, it is likely that other Burkholderia–fungus interactions will be discovered in the future.

Burkholderia spp. can thus make an intimate association with beneficial fungi that interact with plants and this is particularly interesting with regard to symbiosis. However, interest regarding symbiosis of Burkholderiales with a host have largely involved plant nodulation.

Burkholderia spp. and plant nodulation

Studies of plant–bacteria symbiotic associations have primarily focused on the interaction between leguminous plants and gram-negative Alphaproteobacteria belonging or closely related to the genus Rhizobium (Hirsch et al., 2001). Burkholderia spp. have also recently been isolated from nodules of some plant species and are now recognized as effective symbionts associated with roots of Fabaceae and leaves of Rubiaceae (Table 7).

View this table:
7

Burkholderia spp. and plant nodulation

SpeciesNodulated plantReferences
Root endosymbiontsFabaceae
B. cepacia complexDalbergia spp.Rasolomampianina et al. (2005)
B. caribensisMimosa diplotrichaChen et al. (2003a)
Mimosa. pudica L.Chen et al. (2003a)
B. mimosarumMimosa spp.Chen (2005a, b, 2006)
B. nodosaMimosa spp.Chen (2005a, b, 2007)
B. phymatumMachaerium lunatum (L. f.) DuckeVandamme et al. (2002)
Mimosa invisa Mart.Elliott et al. (2007)
Mimosa pudica L.Elliott et al. (2007)
Mimosa pigraBarrett & Parker (2005, 2006)
Mimosa castaBarrett & Parker (2005, 2006)
B. tuberumAspalathus carnosa BergiusVandamme et al. (2002)
Cyclopia spp.Elliott et al. (2007)
Leaf endosymbiontsRubiaceae
Candidatus Burkholderia calvaPsychotria calva HiernVan Oevelen (2002)
‘Candidatus Burkholderia kirkiiPsychotria kirkii HiernVan Oevelen (2002)
Candidatus Burkholderia nigropunctataPsychotria nigropunctata HiernVan Oevelen (2004)

Earlier studies demonstrated that similar to other bacteria species, B. cepacia is able to facilitate infection of plant roots by nodulating Frankia spp., probably through root hair deformation (Knowlton et al., 1980). Knowlton & Dawson (1983) reported that the B. cepacia strain could increase the number of nodules up to fourfold, although its presence was not an absolute prerequisite for the nodulation per se. More recently, Rasolomampianina (2005) reported for the first time that a strain belonging to the B. cepacia complex induces efficient nodules on a wild tropical legume from Madagascar, Dalbergia spp. Burkholderia tuberum was isolated from root nodules of Aspalathus carnosa in South Africa, and B. phymatum from Machaerium lunatum in French Guiana (Moulin et al., 2001; Vandamme et al., 2002). The occurrence of the B. cepacia strain in plant nodules raises concern that leguminous plants can be a source of this potential pathogen. Several Burkholderia strains were also isolated from Abarema macradenia and Pithecellobium hymenaefolium in Panama and Costa Rica (Barrett & Parker, 2005, 2006). The same finding was recently described in a study of nodulation of four Papillonaceae (Vigna sp., Clitoria sp., Crotalaria sp. and Centrosema sp.) in which nodules contained both rhizobial species and Burkholderia species (J. Balandreau, pers. commun.). The most intensive research on plant–Burkholderia spp. interactions has involved mimoisoid legumes. Burkholderia caribensis has been isolated from Mimosa diplotricha and from Mimosa pudica in Taiwan (Chen et al., 2003a) and B. phymatum from Mimosa invisa and M. pudica in Papua New Guinea (Elliott et al., 2007), M. pigra, Mimosa casta and M. pudica, in Panama and Costa Rica (Barrett & Parker, 2005, 2006), M. pudica in India (Pandey et al., 2005), and from Mimosa spp. in Taiwan, Venezuela and Brazil (Chen et al., 2005a, b). Recently, new species were also described such as B. mimosarum from M. pigra nodules in Taiwan (Chen et al., 2006) and B. nodosa isolated from Mimosa scabrella nodules in Brazil (Chen et al., 2007). We may conclude that the genus Burkholderia exhibits affinity for colonization of mimosoid legumes, probably as a result of an intimate coevolution between the bacteria and this plant taxon. Although members of the genus Burkholderia and other Betaproteobacteria have been isolated from leguminous root nodules, their direct involvement in the formation of root nodules and nitrogen fixation has been documented only recently. Nodules formed by B. tuberum STM678 or DUS833 on Cyclopia genistoides were effective, typically indeterminately, with an invasion zone and a nitrogen-fixing zone (Elliott, (2007). The nitrogenase Fe-protein (nifH protein) confirmed the presence of B. tuberum bacteroids expressing nitrogenase enzyme complex in the host cells of Cyclopia nodules (Elliott et al., 2007). The authors also demonstrated that B. phymatum can nodulate a large number of Mimosa species, in a wide range of geographical zones and taxa (Elliott et al., 2007). The nodulating Burkholderia species contain symbiosis-essential genes such as nod and nif and exhibit significant acetylene reduction activity (Chen et al., 2003a, b, 2005a, b; Barrett & Parker, 2005, 2006; Elliott et al., 2007), which demonstrate direct involvement in plant nodulation. By contrast, Rasolomampianina (2005) were not successful in detecting a nodA-like gene in B. cepacia strain STM 1424. This may indicate that the nodA gene sequence of strain STM1424 is either absent or distant from any known nodA sequence.

However, rhizobia-like nodulation ability of some Burkholderia species has been recently confirmed using light and electron microscopy to monitor the pathway of the bacteria transformed with the green fluorescent protein (Chen et al., 2003b, 2005a, b; Elliott et al., 2007) (Fig. 1).

1

Confocal photomicrographs of a section of Mimosa pudica nodules infected with Burkholderia phymatum STM815GFP (a,b). Scale bars: (a) 500 μm, (b) 5 μm. Pictures provided by Dr Geoffrey N. Eliott (School of Life Sciences, University of Dundee, UK).

The symbiotic association of Burkholderia spp. was also found in leaf nodules in some genera of tropical angiosperms belonging to Rubiaceae, Myrsinaceae and Dioscoreaceae (Silver et al., 1963; Centifanto & Silver, 1964; Lersten & Horner, 1967; Whitmoyer & Horner, 1970; Miller et al., 1983). This phenomenon has been known since 1902 (Zimmermann et al., 1902), but only recently was this attributed to the genus Burkholderia. In Rubiaceae, known leaf endosymbionts of Psychotria spp. include ‘Candidatus Burkholderia calva’ (van Oevelen, 2004), ‘Candidatus Burkholderia kirkii’ (van Oevelen, 2002), and ‘Candidatus Burkholderia nigropunctata’ (van Oevelen, 2004). Further studies on this plant bacterial association have demonstrated that plant tissue cultures of Psychotria sp. without the bacteria had distorted leaves, stunted growth and eventually died. This demonstrates an obligate association between Burkholderia spp. and Psychotria spp. (van Oevelen, 2003). The endosymbiosis between Burhkolderia spp. and their host plant is unique as it lasts throughout the plant's entire life cycle and the bacteria can be transmitted to the next generation via seeds, without the need for external infection (von Faber, 1912; Miller et al., 1990).

Use of Burkholderia spp. for biocontrol of plant diseases and plant and stimulation of plant growth

Biocontrol properties of some Burkholderia sp.

Most of the beneficial Burkholderia species cited above have been studied for biological control. As previously described, Burkholderia spp. can be free-living and endophytic, forming mutualistic and symbiotic associations that promote growth and health status of plants. Several Burkholderia species are considered to be beneficial in the natural environment. These species are well known for their biological and metabolic properties, which can be exploited for biological control of fungal diseases in plants but also for bioremediation and plant growth promotion (Govan et al., 1996; Holmes et al., 1998; Parke & Gurian-Sherman, 2001; Perin et al., 2006b). Many species belonging to the genus Burkholderia have the ability to produce compounds with antimicrobial activity (Hu & Young, 1998; Kang et al., 1998) and can potentially be used as biocontrol agents of phytopathogenic fungi. This was well demonstrated with B. cenocepacia, B. cepacia, B. ambifaria, B. pyrrocinia, B. vietnamiensis and B. phytofirmans strains towards Pythium aphanidermatum, Pythium ultimum, Fusarium sp., Phytophthora capsici, Botrytis cinerea and/or Rhizoctonia solani (Parke et al., 1990; Parke et al., 1991; Hebbar et al., 1992, 1998; McLoughlin et al., 1992; Bowers & Parke, 1993; King & Parke, 1993; Heydari & Misaghi, 1998; Ait Barka, 2000, 2002; Cain et al., 2000; Heungens & Parke, 2000; Parke & Gurian-Sherman, 2001; Singh et al., 2006). Many Burkholderia species are also able to inhibit growth of other bacteria, protozoa (Cain et al., 2000) and nematodes (Meyer et al., 2000). Their ability to suppress plant diseases was observed in many different crops, such as corn, sweet corn, cotton, grapevine, pea, tomato and pepper. Burkholderia strains form dominant populations on mosses and show high antagonistic potential against fungal pathogens (Opelt & Berg, 2004). Recently, two new species, Burkholderia bryophila and Burkholderia megapolitana, isolated from moss and showing antifungal activity against phytopathogens as well as plant-growth-promoting properties were reported by Vandamme (2007b). Similar to other plant growth-promoting bacteria (PGPB) or plant growth-promoting rhizobacteria (PGPR), the beneficial effects of Burkholderia sp. involve diverse mechanisms of action (Fig. 2), including rhizosphere competence determining their population density on root surface, secretion of allelochemicals, including antibiotics and siderophores, competition for nutrients, as well as induced systemic resistance (van Loon, 1998; Baldani et al., 2000; Welbaum et al., 2004; Compant et al., 2005b). Thus, several strains of Burkholderia produce very efficient low-molecular-weight iron-chelating compounds, called siderophores (Bevivino et al., 1994). These compounds are implicated in antibiosis against plant pathogens through iron sequestration under iron-limiting conditions. Ornibactins are the predominant siderophores produced by Burkholderia strains (Meyer et al., 1995). Nevertheless, several other siderophores are also produced by Burkhlderia strains such as cepaciacheline and cepabactine (Barelmann et al., 1996; Meyer et al., 1995).

2

Beneficial properties of Burkholderia spp. for agricultural improvement.

Recently, Pandey (2005) reported that an 1-aminocyclopropane-1-carboxylate (ACC) deaminase-containing endophyte belonging to Burkholderia sp. exhibited antagonistic activity against Rhizoctonia solani and Sclerotinia sclerotiorum. Furthermore, ACC deaminase can act synergistically with other mechanisms of biocontrol in reducing symptom development without having an effect on the population density of the pathogen (Wang et al., 2000).

The outlined beneficial properties of Burkholderia spp. indicate that they can be potentially useful in agriculture as a substitute for, or a complementary agent with, chemicals in the management of plant diseases. However, the use of organisms that may be opportunistic human pathogens is subject to debate. For example, the US Environmental Protection Agency has severely limited biotechnological applications of members of the B. cepacia complex (Vandamme et al., 2007a). Searching for other Burkholderia sp. with biocontrol properties has been an alternative and the potential of some new species is actually under investigation. For example, B. phytofirmans PsJN directly inhibits growth of Botrytis cinerea (Ait Barka, 2000, 2002) and can induce plant defence responses (Compant et al., 2005b, 2007), leading to systemic protection of the host plant toward this phytopathogenic fungus (S. Compant et al., unpublished data). Once activated, induced systemic resistance (ISR) mediated by plant growth-promoting rhizobacteria (PGPR-ISR) is maintained for prolonged periods against multiple pathogens; even if populations of the inducing bacteria decline over time (van Loon, 1998; van Loon, 2007), ISR induced by microorganisms is nonetheless of particular interest for agriculture.

Resistance against abiotic stress

Mechanisms of ISR triggered by Burkholderia spp. could also enhance resistance of the plant against abiotic diseases (Fig. 2). For review of the enhancement of heat stress tolerance of potato and transplant stress tolerance of tissue culture plantlets induced by Burhholderia phytofirmans strain PsJN see Nowak & Shulaev (2003). Recently, Ait Barka (2006) also reported that in vitro inoculation of Vitis vinifera L. cv. ‘Chardonnay’ explants with B. phytofirmans strain PsJN increased grapevine growth and physiological activity at low temperature (4 °C), improving the plant's ability to withstand cold stress. The beneficial effect of the endophytic bacterium may be through its induction of the synthesis of PR proteins and phenolics, which reduce the development of symptoms, and also through the prevention of a set of reactions that produce the symptoms of chilling injury (Ait Barka, 2006), although more knowledge will further describe the exact mechanisms of resistance (A. Theocharis et al., unpublished data). By contrast, and because of the freezing injury of plants that increased with increasing population sizes of Ice+ bacteria, Ait Barka (2006) have suggested that preemptive competitive exclusion of Ice+ bacteria with naturally occurring nonice nucleation-active bacteria such B. phtofirmans strain PsJN could be an effective and practical means of frost control.

Use of Burkholderia spp. for direct stimulation of plant growth

Some members of the genus Burkholderia can also directly stimulate plant growth. For example, Burkholderia spp. inoculation of crops such as maize and sorghum have increased root and shoot biomass (Chiarini et al., 1998; Bevivino et al., 2000). Similar to other beneficial bacteria, the growth promotion caused by Burkholderia sp. is probably due to the combination of several mechanisms that involve both plant and bacterial partners (Fig. 2; Govindarajan et al., 2006). Rhizobacteria facilitate plant growth either by (1) preventing the proliferation of pathogenic organisms or by (2) directly stimulating plant growth by either helping in the acquisition of nutritional resources such as nitrogen, phosphorus or iron, or by providing plant hormones such as auxin or cytokinin, or lowering plant ethylene levels through the action of the enzyme ACC deaminase (Glick et al., 1999; Vessey et al., 2003). As demonstrated for B. vietnamiensis (Trân Van, 2000), production of phytohormones and/or fixation of atmospheric nitrogen are common causes of plant growth promotion induced by this genus. Estrada-de los Santos (2001) have also isolated new diazotrophic Burkholderia species from coffee and maize plants, providing further proof of the wide range of Burkholderia species to fix atmospheric nitrogen.

Research on the mechanisms involved in plant growth promotion by Burkholderia sp. have also demonstrated that some members of this genus (e.g. B. phytofirmans strain PsJN) can express a quinolinate phosphoribosyl transferase involved in plant growth promotion (Wang et al., 2006) as well as a high level of ACC deaminase (Sessitsch et al., 2005). This later enzyme hydrolyses ACC (the immediate precursor of the plant hormone ethylene) to ammonia and α-ketobutyrate, resulting in decreased ACC levels within the plant with the concomitant reduction of plant ethylene levels (Glick et al., 2007). However, it is important to take into account that ACC deaminase activities or their associated genes can also be present in phytopathogenic members of the genus Burkholderia. Recently, Blaha (2006) have thus demonstrated the presence of the ACC deaminase-encoding gene acdS in a wide range of Burkholderia species including PGPR, phytopathogens and even opportunistic human pathogens. For instance, acdS was present in B. tropica, B. caledonica, in the Gladiolus sp. pathogen B. gladioli, the onion pathogen B. cepacia and the carnation pathogen B. caryophylli. In addition, the gene was also present in B. cepacia strain LMG16656, an opportunistic human pathogen implicated in the death of several patients (Govan et al., 1993), as well as in the true pathogen B. mallei (Nierman et al., 2004).

Rhizospheric and endophytic establishment of some Burkholderia sp. in nonnatural hosts

Some studies have reported that certain Burkholderia spp. can thrive as epi- and endophytes in plants other than the original hosts. The B. phytofirmans strain PsJN, originally isolated from surface-disinfected roots of onion in symbiosis with the arbuscular fungus Glomus vesiculiferum, can readily establish rhizospheric and endophytic populations in grapevine (Fig. 3), potato, tomato, cucumber, watermelon and chickpea (Frommel et al., 1991; Ait Barka, 2000; Nowak & Shulaev, 2003; Compant et al., 2005b; Sessitsch et al., 2005; Compant et al., 2007, 2008). The use of endophytic bacteria for plant disease management has great potential as the bacteria can act as direct bioprotectants of specific microenvironmental niches as well as induce plant metabolism to respond faster to encountered stresses (reviewed in Nowak & Shulaev, 2003). Furthermore, endophytes are more effective than free-living bacteria colonizing root and leaf surfaces as they form closer associations with the plant hosts (Conn et al., 1997; Chanway et al., 2000). Interestingly, research on B. phytofimans strain PsJN and grapevine has indicated that the bacterium can be translocated to young berries following soil inoculation (Compant et al., 2008).

3

Confocal photomicrographs (a) and 3D reconstruction (b) of grapevine root (a) and stem (b) inoculated with Burkholderia phytofirmans strain PsJN tagged with gfp showing gfp-bacteria inside root and stem xylem vessels. Scale bars: (a) 20 μm, (b) 15 μm. Pictures by Drs Stéphane Compant and Hervé Kaplan (University of Reims, France).

Genomics of the genus Burkholderia

Although much still remains to be learned about the genomics of this group of microorganisms, analyses of the first genome sequences that became publicly available have indicated that Burkholderia spp. can have large genomes (>8 Mbp; Holden et al., 2004; Nierman et al., 2004; Chain et al., 2006). Previously, the total genome size of members of the B. cepacia complex was estimated to range from 4 to 9 Mbp, more than twice the size of Escherichia coli (Parke & Gurian-Sherman, 2001). Interestingly, large genomes have been described as disproportionately enriched in regulation and secondary metabolism genes and depleted in protein translation, DNA replication, cell division and nucleotid metabolism genes in comparison with small-sized genomes (Konstantinidis & Tiedje, 2004). This may explain why species containing large genomes dominate in environments where resources are scarce but diverse, such as in soils (Konstantinidis & Tiedje, 2004) or in plants.

Genomes of Burkholderia members seem to be shaped for flexibility, plasticity and versatility (Holden et al., 2004; Nierman et al., 2004; Chain et al., 2006). For example, analysis of the B. mallei genome has revealed the presence of >12 000 simple sequence repeats, and variation in these repeats within key genes may provide this organism with a mechanism for generating antigenic variation (Nierman et al., 2004). In the B. pseudomallei genome, Holden (2004) discovered 16 genomic islands (GIs), making up 6.1% of the entire genome. These islands were variably present in different B. pseudomallei isolates but absent from B. mallei. The GIs are characterized by a different G+C content or dinucleotide frequency signature of the DNA compared with other parts of the genome. Several GIs contain genes associated with mobile genetic elements, such as insertion sequence (IS) elements. Some GIs appear to be prophages while others encode large numbers of hypothetical proteins with no database matches. Thus, variability of some genome areas responsible for diseases can arise between different members of the genus Burkholderia.

While comparing the genome sequence of B. xenovorans and a B. cepacia complex isolate, Chain (2006) noticed only a moderate conservation in gene content (i.e. only 44% of all genes were conserved between these two species). There is also considerable intra-species genomic variation in B. xenovorans and the genome size may vary up to 2.3 Mbp (from 7.4 to 9.3 Mbp) within this species. This can be largely explained by more than 20% of sequences acquired by means of lateral gene transfer (Chain et al., 2006). However, the importance of lateral gene transfer and loss in shaping Burkholderia genomes is not only obvious from the large intraspecies variation observed in B. xenovorans (Chain et al., 2006) and B. cepacia complex (Parke & Gurian-Sherman, 2001) genome sizes. For instance, 43% of the transcriptional differences between two B. pseudomallei strains (K96243 and Bp15682) could be attributed to genes that were differentially present between K96243 and Bp15682, demonstrating the importance of lateral gene transfer or gene loss events in contributing to pathogen diversity at the gene expression level (Ou et al., 2005). In general, a large fraction of the differences in gene content within species is associated with bacteriophage and transposase elements, revealing an important role of these elements during bacterial speciation (Konstantinidis & Tiedje, 2005). Environmental behaviours of strains and species of the genus Burkholderia may have modulated the genome sizes and content of these proteobacteria, which are thus capable of flexibility, plasticity and versatility and make thus them adaptable to various microbial niches.

Postgenomic analyses have recently provided further clues to the evolution and contents of the complex genomes of Burkholderia spp. Using microarrays to study large-scale genomic variation in the two closely related pathogenic species B. mallei and B. pseudomallei and the closely related nonpathogenic species B. thailandensis, Ong (2004) have shown that some ORFs deleted between the three species were associated with diverse cellular functions, including nitrogen and iron metabolism and quorum sensing. Kim (2005) have also showed that some B. mallei genes, upregulated in a mouse infection model (genes potentially associated with virulence), tend to be less conserved between B. mallei and B. thailandensis than genes involved in basic metabolism, which provides evidence of differences between pathogenic and nonpathogenic members of the genus Burkholderia.

The present advances in sequencing of Burkholderia genomes such as for various strains of B. ambifaria, B. cepacia, B. cenocepacia, B. dolosa, B. glumae, B. mallei, B. multivorans, B. phymatum, B. phytofirmans, B. pseudomallei, B. thailandensis, B. vietnamiensis and B. xenovorans (http://www.ncbi.nlm.nih.gov/, http://genome.jgi-psf.org/ or http://pathema.tigr.org/Burkholderia/beta/) will undoubtely contribute to a better understanding of the versatility of these organisms, and may also lead to the development of new management practices to control their pathogenicity and/or to harness their benefits. This will provide further information regarding differences between pathogenic and nonpathogenic members via genome analysis, which have only just begun to be explored.

Conclusions and future prospects

Burkholderia spp. are among the most abundant bacteria in the environment. The plasticity of their genomes and their capacity to adapt to changing conditions allow them to colonize diverse environmental niches. Several species of the genus are classified as human, animal or plant pathogens but others may interact with host plants, resulting in beneficial effects. These can potentially be utilized as powerful pesticides in control of soil-borne diseases. Some Burkholderia can also be used as biofertilizers, either by fixing nitrogen or by releasing iron or phosphorus from rock phosphates, to benefit crops cultivated in low-fertility soils, thus becoming ecologically and/or economically important. However, widespread use of Burkholderia as biofertilizers would increase human exposure to these potentially hazardous bacteria.

The main concern associated with the broader use of Burkholderia isolates in production agriculture and biotechnological applications is that it is presently unclear what makes some Burkholderia isolates virulent. Several putative virulence and/or transmission markers have been identified in B. cenocepacia strains, i.e. the cblA gene encoding giant cable pili and the esmR gene as an epidemic strain marker regulator (Mahenthiralingam et al., 1997), which is part of a genomic island (Baldwin et al., 2004). Nevertheless, the role of these marker genes is still unclear as virulent and/or transmissible strains without the markers have also been identified (LiPuma et al., 2001; Coenye & LiPuma, 2003). More recently, Perin (2006b) identified cblA and esmR genes among clinical and environmental isolates of B. cenocepacia and other species of the B. cepacia complex, although these markers have not been found in any of the tested isolates of plant-associated diazotrophic Burkholderia spp. If confirmed on a larger scale, this finding may provide an opportunity for broad utilization of Burkholderia spp. diazotrophs, endophytes in particular, in the production of perennial crops for bioenergy and other uses under low-input production systems. Newly discovered Burkholderia spp., including B. bryophila and B. megapolitana isolated from moss, exhibit similar potential for biocontrol and plant growth-promoting capacity (Vandamme et al. 2007b). Both species are distant from the B. cepacia complex.

Progress in genome sequencing of Burkholderia species at the DOE Joint Genome Institute (e.g. http://genome.ornl.gov/microbial/bphy_psjn), Sanger Institute, J. Craig Venter Institute, and laboratories collaborating with these research centres will undoubtedly generate new knowledge on their genetic composition, molecular and functional adaptation to different environments, and pathogenicity. This information is critical before the ecological and metabolic potential of Burkholderia spp. can be harnessed for plant production, medicinal and/or industrial processes. Clarification of the mechanism of horizontal gene transfer that occurs in some microhabitats would also be a major step in the development of a regulatory framework for the environmental release of these bacteria.

Acknowledgements

We would like to thank Dr Jacques Balandreau for providing information on the genus Burkholderia. We are also grateful to Dr Geoffrey N. Eliott (School of life Sciences, University of Dundee) for providing confocal images of cells of B. phymatum STM 815 in Mimosa pudica nodules, to Dr Hervé Kaplan (University of Reims, France) for helping with confocal scanning of shoots and roots of Vitis vinifera L. inoculated with B. phytofirmans strain PsJN and 3D reconstruction, and to Dr Euan K. James (School of Life Sciences, University of Dundee) for a critical reading of the manuscript. Special thanks also to Philippe Laporte (University of Reims, France) for the verification of Latin names of plants, insects and fungi. This work was partly supported by a grant from Europôl-Agro (Reims, France).

Footnotes

  • Editor: Eva Top

References

View Abstract