Deconstructing the Kingdoms
by Susan Goldhor
Ever since Linnaeus’ 1735 publication of the first edition of Systema Naturae, living creatures have been divided into kingdoms. Actually, Linnaeus divided both living and non-living creation into three kingdoms: animal, vegetable and mineral, thus giving rise to the immortal game, Twenty Questions. But for our purposes, we can state that there were two biological kingdoms, Plantae and Animalia, at least until 1866 when the Protista were added by Haeckel. Bacteria separated as a kingdom from the Protista in 1938 (although they were – confusingly – called “Monera”) and, spurred by this bold action, the Fungi separated from the Plantae in 1969 and, aided by a scientist named Whittaker, declared their independence, becoming the Fifth Kingdom. Subsequently (in 1977), bacteria split into Eubacteria and Archaebacteria, making six kingdoms in all.1
Of course, fungi have always been the poor step-children of classification schemes. Or, in some cases, the wicked stepsisters. And, most recently, as DNA studies confuse all our systems and change all our names, perhaps even the bastard offspring, born on the wrong side of the botanical blanket. After all, although Linnaeus classified them as plants, they never really fit into the Linnaean system which organized the plant kingdom on the basis of sex organs (number of). And here, I can’t resist a little sidetrack, so if you thought that that Mushroom, the Journal was a family publication, safe to leave out on the coffee table for the children and in-laws to peruse, please skip (or cut out) the following paragraph.
In her wonderful book The Brother Gardeners, Andrea Wulf points out that although American gardeners, naturalists and explorers were delighted by Linnaeus’ new classification method, which was simpler than any of the old ones, and did away with the requirement for expensive and rare texts, magnification, and years of study; the British were horrified at the raw sexuality of the system. Linnaeus divided plants into 23 classes according to the number of their male stamens, which he termed “husbands”, and then he further divided them into orders according to the number of female pistils, which he called “wives”. (There was also a 24th class of flowerless plants, called Cryptogamia, or “hidden marriage”, including mosses, algae, liverworts, lichens and, of course, fungi.)
However, as Wulf writes, “Linnaeus encountered even more reluctance than he had anticipated when trying to introduce his system into Britain…. Many botanists were scandalised to learn that plants made love in the flower head – the ‘bridal bed’ – which God had ‘adorned with such precious bedcurtains, and perfumed with so many sweet scents’. The calyx (the mostly green petal-like parts at the bottom of the flower head) Linnaeus had named the ‘bedroom’, and in a later publication he called it ‘the lips of the cunt’. Nor did it help that, according to Linnaeus, one wife frolicked with ‘six husbands’ in the lily’s flower head, while tulip trees enjoyed ‘Twenty males or more in the same marriage’. Oaks, pines and birch featured husbands and wives that lived in one house but had ‘different beds’, while marigolds and ash trees were even more promiscuous, since their flowers contained husbands that ‘live with wives and concubines’.
The whole plant world suddenly seemed to be involved in a horticultural orgy – even the twenty-fourth class of flowerless plant was drawn into ‘Clandestine Marriages’ where the “Nuptials were celebrated privately. Only a few flowers were ‘virtuous’ – like the canna, whose one stamen and pistil were in a monogamous relationship.” What the British would have thought of fungi, promiscuously offering billions of spores to the wind and possessed of too many sexes to count, boggles the mind. Luckily, these traits remained hidden. (Although not fungal nomenclature, which Linneaus did his best to bring out of the nursery and into the bedroom. He was the one who – unforgettably – christened the stinkhorn Phallus impudicus, and named the morel Phallus esculentus – a name that has disappeared into the maw of time; probably more for reasons of taxonomy than propriety, although perhaps because male taxonomists were upset by the pitting and wrinkling.)
Returning to the concept of six kingdoms, and my real message. (And to those wondering if there was one – read on!) I know that the professional classifiers have labored and argued over these kingdoms, and that the increase in number of kingdoms represents a huge investment of research by thousands of workers around the world. Moving from two kingdoms to five and, more recently, to six was not undertaken lightly, and represents big steps forward in our knowledge of the biota. But… I’d like to suggest that new findings in biology imply that, rather than expanding the number of kingdoms, we should be reducing it. Or at least (and perhaps more accurately), blurring the boundaries between kingdoms. In fact, I’d like to change the entire way we regard the biological world, which is neither so neat nor so discrete as the systematists and the concept of kingdoms would have us believe.
Genome vs Biome
What got me started on this line of thought were publications issued at the end of the first decade of the 21st century, celebrating the genetic analysis of the human biome: all of the microbes that live in and on humans. Not the complete genetic sequencing of all of the thousand or so species that we harbor, but the complete genetic analysis of about five hundred of them, plus enough partial analysis to show that a thousand species were cohabiting with us. And not that each one of us is host to the entire thousand, but that each of us is home to about a hundred and sixty of these species, and the zoo in and on each of us has a somewhat different set of species. They’re teeming in our every crevice and on our every surface, inside and out. If you could take a picture of just them, you’d see each of us outlined inside and out in microbes. It would be an image of us, but missing everything that we think of as us. No one had guessed that we hosted so many species, and no one had guessed how many individual microorganisms each of us was carrying around, all those bottles of antimicrobial soap notwithstanding.
There’s been a lot of fanfare for the human genome, the complete sequencing of which celebrated its tenth anniversary in 2013, despite the fact that one sequencer noted that it was like “being given the best book in the world, but it’s in Russian, and it’s incredibly boring to read.”
When I was in college and grad school, which was around the era of cave painting, the molecular biologists were just emerging from the primeval ooze and proclaiming their superiority to what Harvard now calls “organismal” biology. Indeed, the reason why Harvard and many other schools have two biology departments is that the DNA folks made academic life untenable for their more classically minded colleagues. It’s sad but true that DNA has been affiliated from the start with scientists whose brilliance has been matched (and sometimes eclipsed) by their egos. The entire second half of twentieth century biology was dominated by a cabal of molecular biologists who insisted that DNA was the key to all things, and that the sequencing of our genetic code would lead to the discovery of those genes and mutations that caused disease, and thenceforth to the eradication of those diseases and, ultimately, although this was not specifically promised, the death of death.
Although this led to their work getting funded, it has not panned out. Perhaps future work will lead to effective gene therapy for the ills that plague us, but the bright promise of sequencing our genome has led largely to the hypertrophy (and sometimes death) of a series of much-hyped biotech firms. And one of the most intriguing recent findings (intriguing at least to me) is that our DNA may be incomplete. That is, if you could make a human out of purely human DNA, that human would be unable to survive in the world.
The analysis of the human biome has gotten much less PR than that of the human genome, which I guess is understandable, given our massively ethnocentric and anti-microbial views. And this despite the fact that our resident microbes contain about 150 times more non-redundant genes than we do (whatever “we” means). But here’s the part that really grabbed me: if you put any one of us into a blender that reduced us to our component cells and then analyzed each individual cell, only about one in ten of those cells would have that much-hyped human genome. The other nine cells would be those of bacteria, fungi, mites, and the rest of our tiny cohabitants. So… if only ten percent of our cells (and less than one percent of the genes in and around our bodies) are human, what are we? We’re not simply humans; we’re not just individuals – we’re ecosystems.
Just like the forest or the meadow, we’re the sum of all the species that compose us. And, just like the forest or the meadow, we need all those species. (Well… most of them. I must confess that there are a few pathogens and parasites among them. As with all ecosystems.) Without them, we’d be vitamin-deficient, malnourished, and sitting ducks for every germ and auto-immune disease that came our way. It’s not just that we need them to fight off specific pathogens; we need them for our immune systems to develop normally. It’s not just that they digest things we can’t; it turns out that their digestive enzymes and ours fit together like pieces of a puzzle. Now that we’re aware of them, new reports are emerging almost daily, showing that they affect us in wildly unexpected ways.
For example, not every child fed a protein-deficient diet develops kwashiorkor (protein deficiency malnutrition); of two identical twins, carrying the same genes, one may be protected by his gut bacteria. And, although this microbiome lives on our inner and outer surfaces (our GI tract, lungs, skin, etc.), every cell in our bodies (and the bodies of essentially all other eukaryotes, which includes plants and fungi, as well as protists) contains copies of an endosymbiont, called mitochondria, which supply us with energy in small useable packets. Mitochondria are believed to have originated as bacteria – probably Rickettsia relatives – just as chloroplasts are believed to have originated as cyanobacteria; this belief having as its basis the following facts: they contain their own DNA, which resembles that of bacteria; can only come from the division of other mitochondria (or chloroplasts); possess certain enzymes similar to those of bacteria; and have an innermost membrane which is strikingly similar to a bacterial cell wall (and not at all similar to an animal cell). But they have been part of us for so long now that neither of us could survive without the other.
Incidentally, although our bacteria are getting more attention and respect, we also host a human virome, which is just starting to be examined. Carl Zimmer, author of A Planet of Viruses (as well as the wonderfully terrifying Parasite Rex), speaking of retroviruses, like HIV, which insert their own genes into our chromosomes, has said, “it turns out that over tens, hundreds and millions of years, our ancestors have been picking up viruses like this, and they’ve been clogging our genomes. So now there are probably about 100,000 elements in the human genome that you can trace to a virus ancestor. It makes up about 8 percent, all told, of our genome. Just bear in mind that all of the genes that encode proteins only make up 1.2 percent of our genome. So we’re much more virus than human, really.”
So do we really fit into a nice neat kingdom Animalia? And how about plants? It’s one thing to separate them from their mycorrhizal fungal partners. A lot would die and most would weaken, but we might still have plants on earth (although probably not enough to supply us with sufficient oxygen). But suppose we took away their chloroplasts? Their mitochondria? Like us, they’d be dead within days if not hours. And how about all of their fungal endophytes – those tiny fungi that live inside the cells of plants?
We’re only now starting to learn a little about them, but we already know that those minute fungi (of which there might be hundreds in a single pine needle), protect their hosts against infection by fungal and other pathogens, synthesize the toxic compounds that keep many plants from being eaten by animal herbivores, and allow plants to survive in extreme environments, such as abnormally hot locations. No one has done the complete analysis (or perhaps they have, and I don’t know it), but I’d guess that plants are just like us animals, with more bacterial and fungal cells than plant cells in and on their bodies, and – like us – unable to survive if they were removed.
What about the fungi?
Well, how about fungi? (I bet you thought I’d never get around to them.) Well, fungi are both different and similar in this respect. Not only do they have mitochondria, they have their own microbiomes, with bacterial partners in and/or on their cells. We know less about them than we do about our own symbionts, since we look at ourselves first and foremost, and also because the bacterial symbionts of fungi are very small, and have proven difficult to culture. It occurred to me, during one of those hazy times when my brain was floating free of any real world tether, that some of the clever tricks attributed to plants, such as producing toxins or growing in extreme environments, which have recently been shown to be the work of the fungal endophytes inside the plant cells, might actually be carried out by bacterial endosymbionts of those fungi. What a terrific idea! However, like so many of my (and others’) brilliant ideas, it seemed to be yet one more example of what T.H. Huxley described as a beautiful theory killed by an ugly fact. While plants may host a myriad of fungal endophytes, fungi generally are far less hospitable to other kingdoms wishing entry; perhaps because of all the antibiotics they produce.
However, the fact is, I just hadn’t looked hard enough. As early as 2005, Partida-Martinez & Hertweck had published a letter in Nature showing that the dread pathogenic fungi in the genus Rhizopus. which cause Rice Seedling Blight, are actually the beneficiaries of endosymbiotic bacteria of the genus Burkholderia, and it is those bacteria that produce the toxic substance: rhizoxin. Rhizoxin is a complex molecule of the sort which can only be produced by fungi or bacteria, and the process by which the researchers showed that it was bacterial in origin, was equally complex and makes this paper interesting reading for those who enjoy scientific detective stories. Because Rhizopus is such a virulent pathogen, I had assumed that purely agricultural concerns were the driving force here; however, it turns out that rice crop protection was only one piece, the others being that this fungus is used in biotechnology and the production of fermented foods, and is regarded as a good candidate for cancer therapy because rhizoxin works as a powerful antimitotic agent.2
And, if you’re wondering how the fungus survives its symbiont, since fungi must undergo mitosis along with the rest of us, the authors asked themselves the same question; three years later they, along with others, published a paper (Schmitt et al.) showing that pathogenic Rhizopus species (but not the benign ones) have mutations in their tubulin (the organelles that direct mitosis) which render them immune to rhizoxin. Apparently, evolution has turned a pathogen into a symbiont, but directed its toxicity outwards, gaining sustenance for both itself and its host.
There’s another aspect to the rhizoxin story, which is unexpected but demonstrates just how complicated mutualistic relations can be, and how much of a symbiont this particular Burkholderia bacteria has become. The researchers were able to “cure” the fungus of its bacterial infection. As expected, the cured Rhizopus was no longer pathogenic to rice. But it also was unable to reproduce. Turns out that the resident bacteria control spore production as well as pathogenicity.
If we divide the microbiome into bacterial endosymbionts, which dwell inside of host cells, and those that are on or merely associated with their hosts, (and – always excepting mitochondria – essentially all of our own microbiome consists of the latter), looking first at mycorrhizal fungi, it used to be thought that only endomycorrhizals had endosymbionts – and wasn’t that neat and easy to remember? Alas, the Rhizopus story was just one tale that blew that concept out of the water. As to the endosymbionts of endomycorrhizals (also known as arbuscular mycorrhizal fungi or AMFs), we know that they have endosymbiotic bacteria partly because we can actually see at least some of their bacterial partners under the microscope and partly because, even though no one has ever been able to do a genetic analysis of an AMF,3 we still know enough from ectomycorrhizal and other fungal genomes to be able to tell if a chunk of DNA is bacterial or fungal in origin.
These endobacteria – or, to be more accurate, bacteria-like organisms (BLOs) are obligate, and have been captive for so long that they have lost their cell membranes. These were discovered by a lab in Torino headed by Paola Bonfante, which announced their findings in a paper with the satisfactorily clear and symmetrical title, “An obligately endosymbiotic mycorrhizal fungus itself harbors obligately intracellular bacteria.”
Since this discovery, Naumann, Schussler & Bonfante have shown that these BLOs appear to be related to the Mollicutes, to reside in all AMFs and to have a common origin, apparently dating back 450 million years, when the first AMFs came ashore with land plants. They write, “Our results show a vertically inherited, monophyletic and globally distributed lineage of endobacteria thriving in AMF cytoplasm.”
This second bacterium was initially found residing in the cytoplasm of one particular AMF, Gigaspora margarita. It probably should not surprise us that it’s a Burkholderia strain. And it probably should not surprise us that it has since been found in other AMFs (Scutellospora isolates), although not in all. What is surprising – at least to me – is the sheer number of such endosymbionts, with as many as 106 microbial cells in a single spore – pretty impressive, even noting that AMFs have extremely large spores! (J. M Ruiz-Lozano & P. Bonfante)
Clearly some of the Burkholderia family have figured out how to live cushily. Bonfante’s Torino group had the bright idea of “curing” these fungi (at least in spore form) of their infection and comparing their pre-and post-cure metabolic products which, as we might expect, were significantly different, with the post-cure products showing signs of stress to the fungi and especially in pre-symbiotic growth, so that it experiences difficulty in hyphal lengthening and branching in response to plant root exudates. (Lumini et al.; Salvioli et al.)4
In fact, endobacteria are widespread
Although I have a soft spot for the AMFs, and would have been happy to give them unique status in the fungal endobacterial sweepstakes, it turns out that some EMFs harbor endobacteria as well. (To say nothing of non-mycorrhizals like Rhizopus.) There is, however, a difference: the bacteria (or at least the BLOs) are obligate and thus omnipresent in the AMFs, and even the Burkholderia in some AMFs are carried by spores from one generation to the next, while the original ectomycorrhizal poster child for endobacterial infection, Laccaria bicolor, exhibits fluctuating populations of its partner, Pseudomonas fluorescens, that appear to be environmentally acquired; possibly from MHBs (mycorrhizal helper bacteria) that have snuck inside from their usual placement as a biofilm on hyphae. Also, the pathogenic Rhizopus strains appear to be stable, so perhaps Burkholderia has a particularly firm grasp on permanence.
But this is a field where our knowledge base is increasing so rapidly that any assertion will probably fall – indeed, has probably already fallen, since it’s almost impossible for a non-specialist to keep up with the field. Cantharellus spp. harbor a variety of bacteria. This triggered an old memory of having read that it’s the bacteria in chanterelles that give them their long shelf life. And why don’t they harbor insects? Bacterial pesticides?
It’s even been suggested that their enticing apricot-like aroma may be due to bacteria. Chanterelles seem to take up bacteria from their rhizosphere, albeit selectively, and with somewhat different populations in different anatomical parts. Truffles do the same, with the caveat that researchers (Antony-Babu et al.) have suggested that bacteria known to break down sulfur compounds may be responsible for the short life of the famed truffle aroma. The microbiome giveth and the microbiome taketh away. I think it was when I learned that truffles harbor a wide variety of bacteria, including Rhizobium spp. that are probably supplying them with nitrogen, that I backed off and wouldn’t read the paper. (Barbieri et al. – feel free to peruse it.) And no further papers, either — at least for now. I have to draw the line somewhere.
The AMFs are probably the best place to look for fungal oddities, being oddities themselves in ways that deserve a separate essay. AMs are obligate symbionts with plants, unlike the EMs that can survive without a photosynthetic partner, but there is one AM outlier, the extremely primitive ancestral member, Geosiphon pyriformis, which instead of forming a symbiosis with a green plant, utilizes cyanobacteria as endosymbionts, holding them in a special cup-shaped compartment, the symbiosome.
If you’ve never heard of Geosiphon, do not despair; these are basically microscopic fungi living at the surface of nutrient-poor soils, and unlikely to grab our attention. However, the study of their endosymbionts led to the discovery of the first sugar transporter to be characterized in the AMs, and caused a stir in some (highly specialized) circles. Their Darmstadt-based discoverer, Arthur Schussler (there are also a number of co-workers) wrote, “The Geosiphon symbiosis is the only known fungal endosymbiosis with cyanobacteria…” at the start of a chapter published in 2005 in a Springer volume, “In vitro Culture of Mycorrhizas”. But this struck me as an odd statement, since there is a varied and numerous group of fungi which have cyanobacteria as endosymbionts: the lichen-forming fungi. True, most lichens capture green algae as endosymbionts, but still… Geosiphon has more to recommend it than its cyanobacteria; its symbiosomes have also been shown to contain betaproteobacteria belonging to the Oxalobacter. Their role was unknown as of my last perusal of the literature.
Lichens, which have been the go-to organism for symbiosis because of their association with photobionts, also offer complex associations with endosymbiotic bacteria, which are emerging as being incredibly varied and carrying out an unexpectedly wide range of (as yet imperfectly understood) functions. The lichen-bacteria communities deserve an article to themselves (and have received several in the technical literature). Suffice it to say that a 2011 paper by Bates et al., which looked at four samples each of four species of lichens, all growing near each other on the same rock faces, found a total of 657 distinct bacterial phylotypes. (All samples had been extensively sterilized, and soil samples yielded different bacterial communities.) These lichen thallus bacteria included Actinobacteria, Firmicutes, Gammaproteobacteria, Bacteroidetes, Planctomycetes, and Verrucomicrobia representatives. I confess that I found this massive zoo of lichen endobymbionts overwhelming and this (along with the truffles) may have been the point where I threw up my hands in despair of keeping up with the field.
Recently, an entire new class of fungal endosymbionts has come under scrutiny: viral endosymbionts in endophytic fungi. This started in 2007 when Marquez et al. discovered that the thermal tolerance imparted by Curvularia protuberata to panic grass was actually dependent upon a virus (clearly but somewhat tediously named “Curvularia thermal tolerance virus” or CThTV). Since then, it has become clear that fungal endophytes frequently contain viral endosymbionts, and we can look forward to more news from this front. Please note that I was correct in my original wild surmise that traits such as toxicity or thermal tolerance, which had been credited to endophytic fungi, could indeed be due to endosymbionts of the endophytes! I just hadn’t gone far enough.
Fungal endosymbionts aside, there is also a large body of relatively recent work on mycorrhizal helper bacteria (MHB). These soil-dwelling organisms are apparently ubiquitous: Frey-Klett et al. state that “MHB have been found every time they have been looked for, under very different environmental conditions and in various plant-mycorrhizal fungus symbioses. . .” They appear to share the ability to increase the rate and/or amount of mycorrhizal infection. This has been shown for both ecto- and endomycorrhizal fungi, and for helper as well as pathogenic fungi. (Those wishing a review of this entire subject should download Frey-Klett and coworkers’ paper.)
I’m not sure that the MHBs are truly comparable to our very own microbiomes. After all, we offer hundreds (maybe thousands) of microniches, ranging from dandruff particles to eyelashes, teeth to underarms, and crotches to belly buttons (innies, of course). They are ours. But the MHBs may be creatures of the soil; happy to help a nearby fungus to make contact with a root – for a price – but not really dependent upon it. As to the currency in which that price is paid, in an inverse parallel to the relationship between the fungus and its plant partner, the fungal hypha appears to use a sugar (trehalose) to attract and assist the bacteria and the bacteria gives its fungal partner thiamine (vitamin B1). In both cases, growth and development are assisted by this mutualistic exchange. As to where the fungus gets the trehalose, fungi cleverly convert plant sugars to trehalose, a form which can’t be repossessed by the plant.
Indeed, the only kingdom that seems to consist solely of the organisms that we think it consists of is that of the archaeobacteria. (The eubacteria do have resident viruses.) And I fear that the only reason I think this is because I don’t know any better.
Whither the Kingdoms?
Superorganism! Chimera! Biome! Ecosystem! We may not be as clear inhabitants of the Kingdom Animalia as we used to think, but we live in bodies that are far more exciting and cosmopolitan and complicated than we ever dreamed. And we are all linked. We contain, in and on us, members of at least three other kingdoms. No longer can we regard bacteria or fungi or protists or viruses either as exotic and bizarre or as pure pathogens. They are in our bodies. They are in our cells. They’re in our chromosomes! To paraphrase Pogo, “We have met the kingdoms and they are us.” So please forgive me for not writing this column exclusively about fungi. Because I’m high on the concept that the walls that separated us from them have crumbled.
Antony-Babu et al. 2013. “Black truffle-associated bacterial communities during the development and maturation of Tuber melanosporum ascocarps and putative functional roles”. Environ. Microbiol. doi:10.1111/1462-2920.12294.
Barbieri, E. et al. 2005. “New evidence for bacterial diversity in the ascoma of the ectormycorrhizal fungus Tuber borchii Vittad.” FEMS Microbiol Letters 247(1): 23-35.
Bates, S.T. et al. 2011. “Bacterial communities associated with the lichen symbiosis”. Appl. Environ. Microbiol. 77(4): 1309-1314.
Bianciotto, V. et al. 1996. “An obligately endosymbiotic mycorrhizal fungus itself harbors obligately intracellular bacteria”. Applied and environmental microbiology 62 (8), 3005-3010.
Frey-Klett, P., J. Garbaye & M. Tarkka. 2007. “The mycorrhiza helper bacteria revisited.” New Phytologist 176 (1): 22-36.
Lackner, G. et al. 2009. “Global Distribution and Evolution of a Toxinogenic Burkholderia–Rhizopus Symbiosis”. Appl. Environ. Microbiol. 75(9): 2982-2986. doi:10.1128/AEM.01765-08
Lumini, E. et al. 2007. “Presymbiotic growth and sporal morphology are affected in the arbuscular mycorrhizal fungus, Gigaspora margarita cured of its endobacteria”. Cellular Microbiology 9(7): 1716-1729.
Marquez, L.M. et al. 3007. “A virus in a fungus in a plant: three-way symbiosis required for thermal tolerance”. Science 63: 545-558.
Naumann et al. 2010. “The obligate endobacteria of arbuscular mycorrhizal fungi are ancient heritable components related to the Mollicutes”. The ISME Journal (2010) 4, 862–871; doi:10.1038/ismej.2010.21
Partida-Martinez, LP, & Hertweck C. 2005. “Pathogenic fungus harbours endosymbiotic bacteria for toxin production”. Nature 437: 884–888.
Partida-Martinez, L.P. et al. 2007. “Endosymbiont-Dependent Host Reproduction Maintains Bacterial-Fungal Mutualism”. Curr. Biology 17(9): 773-777. doi:10.1016/j.cub.2007.03.039
Ruiz-Lozano, J.M. & P. Bonfante. 2000. “A Burkholderia strain living Inside the Arbuscular Mycorrhizal Fungus Gigaspora margarita possesses the vacB gene, which is involved in host cell colonization”. Microb. Ecol. 39: 137-144.
Salvioli, A. et al. 2010. “Endobacteria affect the metabolic profile of their host Gigaspora margarita, an arbuscular mycorrhizal fungus”. Environ. Microbiol. 12: 2083-2095.
Schmitt et al. 2008. “Evolution of host resistance in a toxin-producing bacterial–fungal alliance”. The ISME Journal (2008) 2, 632–641; doi:10.1038/ismej.2008.19