Termite symbiotic systems: efficient bio-recycling of lignocellulose
- First Online:
- Cite this article as:
- Ohkuma, M. Appl Microbiol Biotechnol (2003) 61: 1. doi:10.1007/s00253-002-1189-z
- 1.6k Views
Termites thrive in great abundance in terrestrial ecosystems and play important roles in biorecycling of lignocellulose. Together with their microbial symbionts, they efficiently decompose lignocellulose. In so-called lower termites, a dual decomposing system, consisting of the termite's own cellulases and those of its gut protists, was elucidated at the molecular level. Higher termites degrade cellulose apparently using only their own enzymes, because of the absence of symbiotic protists. Termite gut prokaryotes efficiently support lignocellulose degradation. However, culture-independent molecular studies have revealed that the majority of these gut symbionts have not yet been cultivated, and that the gut symbiotic community shows a highly structured spatial organization. In situ localization of individual populations and their functional interactions are important to understand the nature of symbioses in the gut. In contrast to cellulose, lignin degradation does not appear to be important in the gut of wood-feeding termites. Soil-feeding termites decompose humic substances in soil at least partly, but little is known about the decomposition. Fungus-growing termites are successful in the almost complete decomposition of lignocellulose in a sophisticated cooperation with basidiomycete fungi cultivated in their nest. A detailed understanding of efficient biorecycling systems, such as that for lignocellulose, and the symbioses that provide this efficiency will benefit applied microbiology and biotechnology.
Lignocellulose is the predominant component of woody plants and dead plant material, as well as being the most abundant biomass on earth,especially in terrestrial ecosystems. Since utilization of lignocellulose as a biomass resource is often difficult due to the inefficiency of its degradation, it is beneficial to understand natural systems in which lignocellulose is efficiently decomposed. The natural process is also important in order to know more about the global carbon cycle. Microorganisms, chiefly fungi and bacteria, are thought to carry out lignocellulose decomposition, whereas soil invertebrates greatly enhance the activity of microorganisms by simple dispersion and crumbling of plant materials as well as their actual dissimilation. Termites are one of the most important of the soil insects that efficiently decompose lignocellulose with the aid of their associated microbial symbionts. Termites are said to dissimilate a significant proportion of the cellulose (74–99%) and hemicellulose (65–87%) components of lignocellulose they ingest. Due to their digestive ability and their huge abundance, termites have a tremendous ecological impact on the biorecycling of lignocellulose. They also greatly contribute to the physical and chemical modifications of habitats, particularly soils, analogous to the role of earthworms in temperate zones. Also, many animals prey on termites as a rich protein source (Abe et al. 2000).
Termites (order Isoptera) comprise a complex assemblage of diverse species, roughly divided into so-called lower and higher termites (Abe et al. 2000). Lower termites harbor a dense and diverse population of prokaryotes and flagellated protists (single-cell eukaryotes) in their gut. Higher termites comprise only one apical family (Termitidae) but more than three-quarters of all termite species. While they also harbor a dense and diverse array of prokaryotes, higher termites typically lack flagellated protists. The higher termites show considerable variation in their feeding behavior, which is not limited to xylophagy. Some feed exclusively on soil, presumably deriving nutrition from the humic compounds therein, and others cultivate and consume cellulolytic fungi.
Although a number of excellent reviews are available (Slaytor 1992; Breznak and Brune 1994; Varma et al. 1994; Brune 1998; Radek 1999; Brune and Friedrich 2000; Abe et al. 2000; König et al. 2002; Dyer 2002; Rouland-Lefèvre and Bignell 2002; Ohkuma 2002), powerful molecular biological tools and methods based on new technology continue to generate remarkable progress in the study of termites and their symbiotic microbes. Consequently, new aspects in our understanding of termite symbiotic systems have emerged. A discussion of those advances as well as of the perspectives for the future is the subject of this review.
The association of gut cellulolytic protists with lower termites is a well-known example of mutual symbiosis. Early studies indicated that the presence of gut protists is critical to termite survival on a diet of wood or cellulose. The protists endocytose wood or cellulose particles into their food vacuoles and degrade cellulose to produce acetate, which is in turn absorbed by termites as their energy and carbon source. The gut protists are also important for the decomposition of ingested xylan (Inoue et al. 1997). However, this interpretation has been modified based on evidence for cellolytic enzymes of termite origin (Watanabe and Tokuda 2001). Endogeneous cellulases (endo-β-1,4-glucanase and β-glucosidase) of termite origin, which are excreted from the salivary glands or the mid-gut, have been identified and characterized in both lower and higher termites. Molecular analyses revealed that the endogeneous endoglucanases are members of the glycosyl hydrolase family 9 (GHF9) (Watanabe et al. 1998; Tokuda et al. 1999; Nakashima et al. 2002). Other GHF9 members related to those of termites have been found in cockroach and crayfish, suggesting the presence of an ancestral enzyme before the symbiotic relationship with gut protists was established in termites (Lo et al. 2000). Indeed, wood-feeding species of higher termites decompose cellulose efficiently despite the absence of gut protists. It has been demonstrated that endogeneous cellulolytic activity in the midgut meets the metabolic requirements of higher termites (Slaytor et al. 1997). In lower termites, substantial activity particularly against crystalline cellulose is found in the hindgut (Inoue et al. 1997; Itakura et al. 1997; Nakashima et al. 2002). Probably, the ingested cellulose can be partially degraded by the endoglucanase of termite origin, and the cellulose not hydrolyzed in the anterior portion of the gut then travels to the hindgut, where it can be endocytosed and fermented by the symbiotic protists in lower termites. The existence of this dual system in lower termites explains their capacity to assimilate wood glucan to an extent greater than 90%.
In contrast to the endogeneous cellulase, cellulases of protist origin have hardly been analyzed at the molecular level. This is because only a limited number of species have been axenically cultured. Consequently, there biochemical investigations are lacking. Recently, genes encoding cellulase homologs belonging to the GHF45 family were identified in gut protists using a culture-independent approach, i.e. consensus PCR and screening of a cDNA library (Ohtoko et al. 2000). All known members of GHF45 encode endoglucanase. The protist origin was confirmed by whole-cell in situ hybridization using oligonucleotide probes specific for regions conserved in some of the GHF45 sequences. Among members of GHF45, the cellulase sequences of gut protists were phylogenetically monophyletic, suggesting their recent diversification within the protistan taxa. Through exhaustive screening of the cDNA library of the gut protist population, GHF5 and GHF7 cellulases, including cellobiohydrolase homologs, xylanases, and β-glucosidases were identified (our unpublished data).
Both the cellulases of termite origin (belonging to GHF 9) and those of protist origin consist of a single catalytic domain and lack the ancillary domains such as cellulose-binding domains found in most microbial cellulases. As discussed by Watanabe and Tokuda (2001), termites grind and crunch their ingested material, which may enhance digestion by increasing the amount of surface that can be accessed by the cellulolytic enzymes. Heterologous expression of the cellulase genes, particularly those of protist origin, may help to characterize their encoded enzymatic properties. Whether any synergistic cooperation among the cellulases is present or not should be addressed in order to understand efficient cellulose decomposition by termite symbiotic systems.
A wide variety of bacteria and yeasts, including cellulolytic and hemicellulolytic ones, have been isolated from the termite gut (Prillinger et al. 1996; Schäfer et al. 1996; König et al. 2002; Wenzel et al. 2002). Furthermore, the termite guts is expected to be the source of novel microorganisms with wide-ranging industrial applications. However, the extent to which these microorganisms are actually involved in the degradation in situ remains to be elucidated.
Microbial ecology of gut symbiotic systems
Roles of gut prokaryotes
Microbial communities involved in anaerobic decomposition of polymerized carbon compounds in nature usually consist of distinct physiological groups of microbes. The final step of the decomposition involves microbes that consume electron equivalents, usually molecular hydrogen (H2), produced from intermediate steps. Efficient elimination of electron equivalents greatly enhances the decomposition itself. In the gut microbial fermentation of termites, CO2-reducing acetogenesis as an "H2 sink" reaction is one of the most characteristic but enigmatic features (Breznak 1994). Due to the thermodynamic deficiency of acetogenesis, methanogens usually dominate in most anoxic habitats. Although methanogenesis seems to be common in the gut of termites, acetogenesis dominates methanogenesis, particularly in wood-feeding termites (Brauman et al. 1992). Fermentation in the gut of wood-feeding termites is considered to be essentially homoacetogenic, and acetate produced by gut microbiota support up to 100% of the respiratory requirement of termites (Breznak 1994, and references therein). Therefore, reductive acetogenesis, as opposed to methanogenesis, is clearly beneficial for termite nutrition.
Molecular phylogenetic identification of gut symbionts
Although several hundreds of bacterial strains have been isolated and characterized (see König et al. 2002), studies of the gut symbiotic systems have often been hampered because of the difficulty of isolation and cultivation of a large number of gut microorganisms, including the predominant species within the gut community. Most of these species have been characterized only on the basis of their morphology. Meanwhile, culture-independent molecular approaches using small-subunit rRNA genes have enhanced our ability to assess naturally occurring microbial diversity. Such approaches have been applied to the analysis of the termite-gut microbial community, and have demonstrated that the majority of the gut community consists of phylogenetically diverse species that are yet-uncharacterized and thus unknown to microbiologists (Kudo et al. 1998; Ohkuma 2002).
The symbiotic flagellated protists in the termite gut belong to the phylum Parabasalia and the order Oxymonadida. These protist species are anaerobic and have no mitochondria in their cells. Instead, in the case of parabasalids, the cells harbor hydrogenosomes, which are thought to share the same symbiotic origin as mitochondria. Hydrogenosomes are anaerobic energy-generating organelles producing acetate, H2 and CO2. Molecular phylogenetic studies revealed that parabasalids are one of the most primitive groups of eukaryotes (Berchtold et al. 1995; Ohkuma et al. 2000; Keeling et al. 1998; Gerbod et al. 2000; Moriya et al. 2001). A molecular study also revealed that oxymonads are related to an excavating group of protists (the genus Trimastix) which has small membrane-bound organelles resembling hydrogenosomes in the cell, suggesting that oxymonads are not primitively amitochondriate but that they lost their mitochondria secondarily (Dacks et al. 2001).
Methanogens inhabiting the gut of termites have been characterized as isolates (Leadbetter and Breznak 1996; Leadbetter et al. 1998) and by culture-independent analyses of archaeal 16S rDNA (Fröhlich and König 1999b; Ohkuma et al. 1999b; Shinzato et al. 1999, 2001; Tokura et al. 2000; Brauman et al. 2001; Friedrich et al. 2001). Most of the sequences from lower termites were related to the genus Methanobrevibacter, whereas those from higher termites were related to the genus Methanomicrococcus or a not-yet-cultivated genus in the Methanomicrobiales in addition to the Methanobrevibacter. Archaeal 16S rDNAs affiliated with the Termoplasmales and with the Crenoarchaeota have also been identified from the guts of some termites (Shinzato et al. 1999; Friedrich et al. 2001).
As a typical culture-independent study, a gut bacterial community was analyzed by 16S rDNA clones in the lower termite Reticulitermes speratus (Ohkuma and Kudo 1996). Most of the clones were affiliated with spirochetes (Treponema-like), the bacteroides subgroup in the cytophaga-flavobacter-bacteroides (CFB) phylum, the low G+C gram-positive bacteria (Clostridium-like), and the Proteobacteria. However, the majority of the clones hadvery low sequence similarity to any known cultivated microorganisms. Furthermore, a novel bacterial group that could not be assigned into any recognized divisions of Bacteria was found as one of the dominant clones. Spirochetes and members of the bacteroides subgroup in the CFB phylum in the guts were analyzed and compared in several termite species (Berchtold and König 1996; Paster et al. 1996; Ohkuma et al. 1999a, 2002; Lilburn et al. 1999; Iida et al. 2000). It is evident that there is significant microbial diversity within the gut of a single termite species, and unique lineages consisting of phylotypes found only in the termite gut are often present. Closely related phylotypes rarely occur among termites. Given the existence of over 280 genera and more than 2,600 species of termites on Earth, the termite gut may be a rich reservoir of novel and diverse microorganisms.
In fact, numerous symbionts were isolated from the gut and described as novel species. Many studies involving pure cultures of gut symbionts have contributed to an understanding of their roles. Although spirochetes have long been uncultivated, two strains of the Treponema branch were recently isolated from the termite gut as CO2-reducing acetogens (Leadbetter et al. 1999). Some of the spirochetes, including the gut isolates, were shown to have potential nitrogen fixation activity (Lilburn et al. 2001). The findings imply important roles for symbiotic spirochetes in the nutrition of host termites. As mentioned by Breznak (2002), a sustained effort to obtain the "not-yet-cultured majority" in culture will greatly enhance our understanding of environmental microbiology.
Highly structured gut micro-environments
The termite gut is, although very small, a highly structured microenvironment with physicochemically distinct microhabitats, rather than a simple anoxic fermentor (Brune 1998; Brune and Friedrich 2000, and references therein). Studies using microelectrodes have shown the presence of steep gradients of oxygen and hydrogen within the gut. The oxygen gradient drives a continuous influx of oxygen into the gut periphery, rendering a large proportion of the gut microoxic, and hydrogen accumulates in the anoxic lumen. These gradients significantly impact microbial activities within the gut. A significant proportion of reducing equivalents produced by gut fermentation is consumed by O2 reduction either directly or indirectly in the gut (Tholen and Brune 2000). Indeed, many isolates including lactic acid bacteria and sulfate-reducing bacteria exhibit high rates of O2 reduction (Kuhnigk et al. 1996; Tholen et al. 1997; Bauer et al. 2000).
From the gut of the lower termite Reticulitermes flavipes, distinct types of methanogen species have been isolated as three novel species of the genus Methanobrevibacter (Leadbetter and Breznak 1996; Leadbetter et al. 1998). They utilize H2 plus CO2 but use other substrates poorly, and seem to be tolerant against O2. These species are morphologically similar to the cells colonizing densely on or near the gut epithelium, whereas methanogens located in other portions of the gut are rare in this termite species. The preference of acetogenesis in wood-feeding termites rather than methanogenesis may be explained by a difference in the distribution of the respective microbial populations and by the presence of the gradient of hydrogen partial pressures in the gut (Brune 1998). Methanogens sometimes occur on and within the cells of symbiotic protists. The endobiotic methanogens of gut protists have been identified in other termite species as yet-uncharacterized novel phylotypes of the genus Methanobrevibacter, which are phylogenetically distinct from phylotypes of methanogens attached to the gut epithelium of the same termite species (Fröhlich and König 1999b; Tokura et al. 2000).
The various associations between gut protists and prokaryotes indicate that the gut symbiotic system accumulates many kinds of symbiotic interactions (Berchtold et al 1999; Fröhlich and König 1999a, b; Radek 1999; Dolan 2001; Dyer 2002; Ohkuma 2002), and that these protist-associated prokaryotes probably play an important role in efficient decomposition. Studies of methanogens and spirochetes (Iida et al. 2000; Tokura et al. 2000) revealed that gut microbes are not evenly dispersed but occupy distinct micro-niches within the termite gut. Such a spatial organization of microbial populations as well as the physicochemical gradients in the gut suggest the importance of studying microecology, in which microbial function have to be linked with both their localization and their microhabitats (Brune and Friedrich 2000).
An advantage of functional marker genes
Culture-independent analyses of rDNA sequences have opened a window to investigate the diversity and composition of natural microbial communities, avoiding the largely unrepresentative nature of microbial cultivation. However, in general, the physiological properties of individual microbial populations in a community usually cannot be predicted solely on the basis of the rRNA sequences. Metabolic functions of microorganisms within specific phylogenetic groups can be inferred for only some members, such as methanogens. Under such circumstances, genes encoding metabolically important enzymes are useful as molecular markers. The cellulase genes identified from gut protists serve as an example of the analysis of functional genes. Also, the gene nifH, encoding dinitrogenase reductase, allowed the characterization of potential nitrogen-fixing symbionts without the need for their cultivation, and a remarkable diversity of nifH genes in the gut community was thereby detected (Ohkuma et al. 1999c).
However, the mere presence of a gene does not always mean that the encoded biological activity is being expressed. In fact, mRNA analysis by reverse transcriptase (RT)-PCR in the gut community points to an important concept in molecular microbial ecology: that not only the presence of a gene but also its expression should be characterized so as to evaluate real microbial activity. It was demonstrated that only a few genes among the diverse nifH genes are preferentially transcribed in the gut community (Noda et al. 1999, 2002). An analysis of gene organization indicated that the preferentially transcribed nitrogenase genes encode the alternative nitrogenase system (Noda et al. 1999). The typical nitrogenase contains molybdenum (Mo) as a cofactor, whereas alternative nitrogenases are Mo-independent. A quantitative evaluation, which is important when PCR is used since PCR often introduces some biases, suggests that the availability of Mo is a key to the preference (Noda et al. 1999). Some evolutionary trends in the diazotrophic inhabitants during termite evolution have also been described (Ohkuma et al. 1999c). Studies using additional functional genes and especially in situ monitoring of mRNA at the cellular level are needed to understand microbial activity in termite symbiotic systems.
Lignin degradation in the gut
Lignin is a heterogeneous and irregular arrangement of phenylpropanoid polymer that resists chemical or enzymatic degradations, and thus it protects cellulose. Lignin is extremely recalcitrant and is mineralized in an obligately aerobic oxidative process carried out by appreciably only few taxa. It has been thought that lignin degradation is a rate-limiting step of carbon recycling in lignocellulose-rich ecosystems. There is no convincing evidence of degradation of polymeric lignin to any significant extent in the gut of wood-feeding termites (Breznak and Brune 1994, see references therein). Lignins radiolabeled naturally and synthetically were mineralized only to a limited extent, suggesting that high-molecular-weight core lignin is not degraded at all. Probably, increasing the number of accessible surfaces of wood by mastication and long retention of wood particles in the gut passage are adequate for hydrolyzing cellulose and hemicellulose without significant degradation of core lignin.
On the other hand, mineralization of lignin-derived monomeric aromatic compounds in vivo and in gut homogenates has been demonstrated (Kuhnigk et al. 1994; Brune et al. 1995; Kuhnigk and König 1997). Oxygen is indispensable for ring cleavage and complete degradation of the aromatic compounds in gut homogenates. Numerous isolates that aerobically degrade lignin monomers and dimers have been obtained from the guts of various termites. Since oxygen penetrates the gut epithelium, these aromatic-compound-degrading bacteria presumably reside on or near the gut periphery. Under anaerobic conditions, the tested compounds were only modified, such as demethylation and side-chain reduction (Kuhnigk et al. 1994).
Microbial activity of metabolizing lignin and its components is one of the plausible evolutionary origins for the degrading pathway of aromatic xenobiotics and/or environmental pollutants, such as polychlorinated biphenyl (PCB). Indeed, PCB-degrading Rhodococcus strains were isolated from R. speratus and were genetically studied (Maeda et al. 1995; Kosono et al. 1997). Although their significance in situ remains to be elucidated, these studies suggest the usefulness of termite symbiotic systems as screening sources for a desired metabolic trait.
Soil-feeders constitute approximately 50% of all known species of termites and dominate the termite assemblages in tropical forests (Abe et al. 2000). They greatly impact the cycling oforganic matter and nutrients, and influence both structural and physicochemical properties of soils. Microbial activity is enhanced through the gut passage of soil, probably contributing further processing of organic matter. However, soil-feeding termites have hardly been studied, mainly due to difficulty in sampling and their short life in the laboratory. The substrates degraded by soil-feeders is still obscure, presumably polysaccharides, polyaromatic compounds, and polypeptides immobilized by tannins of plant origin. In contrast to wood-feeding termites, the diet of soil-feeders appears to be low in carbohydrates and high in polyphenolic and humic compounds. Feeding experiments with synthetic humic acids showed that peptidic components of humic substances are selectively digested, whereas aromatic components are not (Ji et al. 2000). Stable-isotope analysis (Tayasu 1998) and 13C NMR (Hopkins et al. 1998) may provide further information.
The gut of soil-feeding termites is highly compartmentalized, and there are extremely alkaline compartments, exceeding pH 12, in the anterior hindgut (Bignell and Eggleton 1995; Brune and Kühl 1996). The high alkalinity enhances chemical oxidation and solubilization of soil organic matter (Kappler and Brune 1999), probably renderingit accessible to digestion. Some microbial constituents and activities are unevenly distributed in these compartments (Schmitt-Wagner and Brune 1999; Tholen and Brune 1999). Soil-feeders as well as fungus-growers (see below) emit more methane than wood-feeders but do not fix N2 in significant amounts because their food (soil or fungi) contains substantial amounts of nitrogen sources.
There have been several suggestions for the roles of the symbiotic fungi in termite nutrition: (1) decomposition of lignin; (2) supply of cellulase and xylanase to act synergistically with the enzymes produced by the termite; and (3) concentration of nutrients such as nitrogen for the termite. Much attention has been devoted to the second suggestion, also known as the "acquired enzyme hypothesis" (Rouland-Lefèvre and Bignell 2002 and reference therein), but some researchers have questioned a part of this hypothesis. Since endogenous cellulases have been recognized in termites, as described above, it is difficult to make generalizations as to the significance of the acquired fungal cellulase in cellulose digestion in fungus-growing termites.
The well-coordinated cooperation between the termites and the fungi enables efficient utilization of lignocellulose (Hinze et al. 2002). So-called old workers forage outside the nest and collect plant litter. In the nest, young workers masticate and ingest the collected plant litter which passes rapidly through the termite gut without digestion. The resulting fecal pellets (primary feces) are pressed together to form a sponge-like structure, the fungus comb, whose matrices support the growth of the symbiotic fungi. The fungi form mycelia, and white, round asexual conidial structures called fungus nodules. The termite nest is a favorable environment for the growth of Termitomyces as its humidity and temperature are controlled. It has been reported that the lignin content progressively decreases as the fungus comb matures (Hyodo et al. 2000). It has also been shown that in vitro digestibility of cellulose in a mature fungus comb is approximately three-fold higher than in a newly formed one. Young workers usually consume fungus nodules, whereas old workers consume old senescent combs to produce final feces. However, final feces are rarely found in the nest of fungus-growing termites, suggesting the highly efficient decomposition and the complete mineralization of plant litter. These observations support the finding that symbiotic fungi have the ability to degrade lignin, which makes cellulose more easily degraded by the cellulase produced by the termite.
Basidiomycetes, which cause white-rot decay and thus are called white-rot fungi, are able to degrade lignin in wood efficiently. Termitomyces are classified into white-rot fungi. Lignin degradation by white-rot fungi has been extensively studied, and results revealed that three kinds of extracellular phenoloxidases, namely, lignin peroxidase (LiP), manganese peroxidase (MnP), and laccase (Lac), are responsible for initiating the depolymerization of lignin. The expression pattern of these enzymes depends on the organism. It is necessary to characterize ligninolytic activity of Termitomyces in order to understand the nature of this symbiotic relationship. In addition to lignin, white-rot fungi are able to degrade a variety of environmentally persistent pollutants, and thus their application is of interest not only in industrial processes but also in bioremediation (Pointing 2001).
Recently, the molecular phylogenetic relationship of Termitomyces species has been examined based on DNA sequences of nuclear rRNA genes, the internal transcribed spacer region ITS1–5.8S–ITS2, and partial large-subunit rRNA (Rouland-Lefèvre et al. 2002; Taprab et al. 2002). In congruence with the taxonomy based on morphology, Termitomyces appears to be monophyletic in members of the order Agaricales. For certain genera of termites, such as African Macrotermes and Pseudacanthotermes, the hypothesis of a termite-fungus coevolution is seemingly acceptable. However, the relationship of the symbiotic Termitomyces with the host termites and with their locality are apparently complex, supporting intricate evolutionary mechanisms by which the termite acquired its symbionts.
The efficiency of biorecycling of lignocellulose by termites is attributed to symbioses with microbes expressing a variety of function that termites do not possess. Recent application of novel technology and molecular methods has greatly enhanced our knowledge of these symbioses. However, detailed knowledge is lacking, because the relationships between termites and microbes as well as among microbes probably include a variety of functional interactions, which the symbiotic systems have accumulated and optimized during their evolution. Therefore, it is necessary to understand the mechanisms of these interactions at both the cellular and the molecular level. Indeed, several symbiotic systems of termites should be studied and compared so as to understand their evolution. Since many manufacturing techniques are simulations of a natural processes, these studies will help us not only to manipulate an existing system, consisting of multiple processes, but also to create new combinations of different organisms having desired functions.
The work of my group was supported by grants to the Bioarchitect Research Program and the Eco Molecular Science Research Program from RIKEN.