杏吧原创

PUT down this magazine, step outside, and pluck a leaf off the nearest tree.
Chances are you鈥檒l be holding in your fingers not just a part of one plant, but
an invisible collection of many species of fungi as well. Now turn around and
look at the nearest hedge or bush. You鈥檒l be laying eyes on a whole island of
biodiversity, and an intricate tangle of ecological interactions.

All around us, plants are not as they seem. Far from being just single
organisms, they鈥檙e entire symbiotic systems. Tiny fungi weave throughout their
leaves and stems in an inconspicuous embroidery of threadlike filaments. While
the presence of these so-called fungal endophytes (endo = inside, phyte = plant)
is common knowledge to experts, only now are scientists beginning to realise how
numerous and important they are. Most temperate-zone plants play host to dozens
of species, and researchers are just beginning to explore the species-rich
tropics (see 鈥淚t鈥檚 a jungle in there鈥�). 鈥淭here鈥檚 another world out there
hiding in leaves,鈥� says Phyllis Coley, an ecologist at the University of Utah in
Salt Lake City.

Of all the microbes living in and on plants, endophytic fungi may be the
least known. Their higher-profile neighbours include pathogenic fungi that cause
disease, mycorrhizal fungi that form associations with plant roots to help the
plants pull nutrients from the soil, and nitrogen-fixing bacteria in the roots
of some plants. What we do know about fungal endophytes suggests that they too
could play a vital role in the plants that house them鈥攁nd can offer big
benefits for people, too.

This fungal world within plant leaves went largely unappreciated until 1977,
when researchers fingered a grass endophyte as the culprit behind many livestock
poisonings. Neotyphodium coenophialum causes weight loss, heat stress,
gangrene, decreased milk production and pregnancy disorders in cattle and horses
that eat its host, tall fescue grass. Livestock munching their way through
perennial ryegrass containing Neotyphodium lolii develop neuromuscular
maladies that lead to tremors, staggering, weight loss, and sometimes death.
These problems occur worldwide, but are greatest in the US, where losses top
$1 billion annually.

Staggering cows

Starved and staggering cows attract a wealth of agricultural research
funding, so endophyte experts have focused on pasture grasses like fescue and
perennial ryegrass. Researchers have learned that some endophytes produce
toxins, such as the alkaloids ergovaline and peramine, that deter leaf-feeding
insects, seed-eating rodents or root-nibbling nematodes by slowing down their
growth or reproduction, or even killing them outright. Endophytes may also
protect plants against microbial pathogens by producing anti-microbial chemicals
or simply by squeezing them out. Some endophytes can even help their hosts
survive drought by making leaf pores close more quickly to boost water
efficiency. Others increase their hosts鈥� nutrient uptake, growth rate, seed
production, or germination rate, so they can beat the competition in the
struggle to survive. In return, the plant gives the fungus a home and plenty to
eat鈥� altogether a mutually beneficial exchange.

But the cosy relationship found in domestic pasture grasses may not be
typical of endophytes in the rest of the plant kingdom. Tall fescue is 鈥渁n
isolated anomaly of selective inbreeding shaped by intensive grazing by cows,鈥�
says Stanley Faeth, an ecologist at Arizona State University in Tempe. In tall
fescue鈥檚 wild relative, Arizona fescue, Faeth found no sign that endophytes ward
off herbivores like grasshoppers, leaf-cutting ants and cattle. The endophyte
helps the plant grow faster, but that鈥檚 about it, he says. Likewise, his earlier
work with student Kyle Hammon showed that endophytes didn鈥檛 stop caterpillars
feasting on oak trees.

In fact, results from the few studies of the ecological effect of endophytes
of wild plants are a decidedly mixed bag. A handful found that endophytes deter
herbivores, while a few others showed the fungi actually give herbivores a
helping hand by increasing the plant鈥檚 nitrogen content or hindering the
herbivores鈥� natural enemies. But in most cases we have no idea what endophytes
are doing inside their host plants, researchers admit. Perhaps they鈥檙e waiting
for leaves to die so they can suck nutrients from them. Perhaps they鈥檙e helping
the plant in some unknown way. Perhaps they鈥檙e actually latent pathogens biding
their time until an opportune moment to turn against their host. Faeth and
others argue that relationships between plants and their endophytes may
range鈥攅ven within a single plant鈥攆rom friendly cooperation to
outright antagonism, depending on the circumstances.

In general, though, the endophytes in grasses tend to be friendlier than
those in woody plants. Most grass endophytes infect all of their host plant and
are 鈥渧ertically鈥� passed on to the next generation in the seeds, while most woody
plant endophytes occupy local spots within the plant, and are transmitted
鈥渉orizontally鈥� from one plant to another by spores. Evolutionary biologists
think vertical transmission encourages cooperation, while horizontal
transmission fosters antagonism
(see Diagram).

Studies by Christopher Schardl of the University of Kentucky support this
scenario. Schardl and his team compared the DNA sequences of several beneficial
endophytes in the genus Neotyphodium with those of related groups of
fungi. Each Neotyphodium species is more closely related to species in
the genus Epichlo毛than to other Neotyphodium species,
they found, which implies that each Neotyphodium species evolved
independently from Epichlo毛. In fact, many Neotyphodium
species appear to have arisen through hybridisation. Most Epichlo毛
fungi are horizontally transmitted pathogens that stop their hosts from
producing seeds. The hybrid lineages apparently allowed their hosts to reproduce
and began spreading vertically, through the host鈥檚 seeds. With that, they
acquired a life-or-death stake in their hosts鈥� reproduction, and eventually
evolved to be their hosts鈥� best friends.

But a two-way interaction isn鈥檛 the end of the story, because the endophytes鈥�
toxins do more than just ward off herbivores. They also influence the growth,
survival, reproduction, behaviour and population dynamics of small mammals,
insects and their parasitoids. For example, female prairie voles that feed on
endophyte-infected fescue reach sexual maturity later than animals on similar
diets without endophytes, according to a recent study by Keith Clay, an
ecologist at Indiana University in Bloomington, and his colleagues. The
endophytes may also spur the voles to move to new sites. Because rodents and
other common herbivores fill crucial ecological roles, endophytes are likely to
have 鈥渁 dramatic and reverberating effect through whole ecosystems,鈥� says
Clay.

Endophytes can also bolster their host鈥檚 competitive ability and alter
dominance hierarchies among plants. For example, endophytes help tall
fescue鈥攁 European native that is already a noxious invading species in the
eastern US鈥攖o squeeze out other species and decrease biodiversity. When
Clay and his student Jenny Holah followed experimental plots of infected and
non-infected fescue for four years, they found that infected plots ended up with
40 per cent fewer species than non-infected plots, and that in infected plots
fescue made up 90 per cent of the biomass, versus less than 60 per cent in
endophyte-free plots (Science, vol 285, p 1742). 鈥淲e鈥檝e got to think
about what we鈥檙e doing,鈥� says Clay. 鈥淲e鈥檝e spread fescue all over. It may be the
next generation鈥檚 kudzu.鈥� Clay鈥檚 home state of Indiana has recently brought in
regulations requiring people who plant fescue to use only endophyte-free
varieties.

But endophytes also offer potential benefits. Biologists are exploring ways
to maximise endophytes鈥� anti-insect effects while minimising their impact on
livestock. Endophytes that make crops more resistant to pests or pathogens could
be effective biocontrol agents. Endophytic ryegrass in New Zealand has had some
success in controlling a weevil pest, for instance. And in Panama, Allen Herre
of the Smithsonian Institution and Elizabeth Arnold of the University of Arizona
in Tucson are searching the leaves of cacao plants for endophytes that might
protect against diseases that attack chocolate crops.

Designer fungi

Some researchers seeking to improve grasses for livestock are searching for
and breeding desirable endophyte strains and combining endophytes and hosts in
novel pairings. Others are trying to identify the genes responsible for
endophytes鈥� effects and modify them to create strains that will do what we want
them to do. Schardl鈥檚 lab, for instance, is attempting to alter genes for
alkaloid synthesis to produce strains non-toxic to livestock. Charles Bacon of
the US Department of Agriculture in Athens, Georgia, and his team aim to use
genetic engineering and more traditional techniques to improve drought
resistance and nitrogen uptake of grasses. But before these techniques can be
applied widely, says Bacon, we need to know what genes do what鈥攁nd such
information is limited for fungi.

Endophytes can also produce extremely useful chemical compounds. For example,
the cancer-fighting compound taxol was originally derived from the Pacific yew.
Unfortunately, the yew is rare and produces taxol in minuscule amounts, and
synthesising taxol in the lab is prohibitively expensive. In 1993 Gary Strobel
of Montana State University in Bozeman and his colleagues made a startling
discovery: it isn鈥檛 just the yew tree that produces taxol, but also a fungal
endophyte living within its needles. Other experts were sceptical of Strobel鈥檚
finding, but his work since then has been extensive and convincing. Amazingly,
he and others have found not only that multiple endophytes in various yew
species produce taxol, but that other fungi in wholly unrelated plants do, too
(New 杏吧原创, 8 April, p 21). Strobel鈥檚 top performer so far is the
endophyte Pestalotiopsis microspora. One strain from the Himalayan yew
produced 1000 times as much taxol as the species from Pacific yew. Best of all,
because all these endophytes produce taxol in lab cultures, it鈥檚 possible that
this important drug could be produced in large quantities without felling any
trees.

Since taxol has antifungal properties, it may help keep pathogens at bay in
the damp environments favoured by yews and rainforest plants. Strobel is excited
about the pharmaceutical potential of the biochemically diverse genus
Pestalotiopsis, which is so widespread in the tropics he calls it 鈥渢he
E. coli of the rainforest鈥�. But he feels a growing sense of urgency. 鈥淲hen
you see the demise of rainforest,鈥� he says, 鈥渋t鈥檚 not just the trees going, but
all the bugs associated with them.鈥�

In Panama, too, Coley and her colleagues are screening endophytes for
potential drugs. She says her team will collect, isolate and screen anything
they find. 鈥淚t鈥檚 in the fishing-expedition stage,鈥� she says. 鈥淔ungal endophytes
are such a new arena that we don鈥檛 want to rule anything out yet, because we
don鈥檛 know enough about them.鈥� That鈥檚 how you need to operate when you鈥檙e
probing one of the last rich, unexplored ecosystems on Earth鈥攖he inside of
a leaf.

In 1991 David Hawksworth, a mycologist at the Royal Botanic Gardens, Kew,
estimated the world鈥檚 fungal diversity at 1.5 million species (equal to all
known living organisms). Many thought his estimate radical. But today, some
mycologists have proposed figures as high as 13 million. Nearly half of those
may be endophytes.

Virtually every plant species researchers have examined has fungal
endophytes鈥攖ypically, scores of them. The numbers in temperate-zone plants
are impressive enough (for example, 40 to 70 species in each of 11 European tree
species, and a whopping 400 in fescue), but the highest diversities may be in
the tropics. One study found 189 species from six palms in Australia and Borneo,
and Greg Gilbert of the University of California at Santa Cruz has isolated 19
鈥渕orphotypes鈥� from a single leaf of a shrub from Panama. Elizabeth Arnold of the
University of Arizona and her colleagues are examining 400-plus morphotypes
isolated from the leaves of two Panamanian tree species.

Researchers use the term 鈥渕orphotype鈥� to describe fungi that are similar to
one another and different from others but where firmer species identification is
impossible. It鈥檚 inherently difficult to recognise fungal species in a tangled
mass of featureless filaments, so unless a specimen produces spores, mycologists
can do little more than record traits such as colour, texture, growth rate and
colony shape.

Such difficulties confuse estimates of global diversity. Even worse, no one
knows whether different host plant species harbour unique sets of fungi or
largely overlapping ones. George Carroll of the University of Oregon in Eugene
thinks that as more DNA sequence data becomes available, many endophytes now
thought to be specialists on single host plants will turn out to be generalists
that also occur in other species. If he鈥檚 right, global diversity estimates may
be too high.

However, the early returns of sequence data from the tropics suggest the
problem may not be dire. 鈥淭here are mismatches in both directions,鈥� Gilbert
says鈥攕ome forms that look alike are genetically distinct and others that
look different are genetically identical. He has found that 75 per cent of his
fungal morphotypes from leaf litter on the forest floor look genetically
distinct. Now he and Arnold are each examining whether the same holds true for
endophytes.