ÐÓ°ÉÔ­´´

THE POLLEN GRAIN is one of evolution’s most ingenious inventions. Thrown
to the winds or smeared on a passing insect, its mission is to transport
a cargo of genes from stamen to stigma – a task at which it excels. Yet,
as victims of hay fever know only too well, its efficiency is often bought
at the expense of subtlety. Snuffling noses and smarting eyes bear witness
to an annual genetic downpour of gargantuan proportions.

Such liberality now looks as if it might cause a few headaches for genetic
engineers. Sooner or later, farmers will be turning their fields over to
engineered crops. When the crops flower they will distribute their pollen
in the time-honoured fashion. Researchers are now asking whether engineered
genes could escape via pollen and infiltrate plants growing nearby. Studies
of this and related issues will loom large in the forthcoming PROSAMO programme,
financed jointly by the Department of Trade and Industry, the Agricultural
and Food Research Council (AFRC) and a consortium of companies. PROSAMO
(which stands for planned release of selected and modified organisms) has
Pounds sterling 1.5 million at its disposal to share between research on
microbes and plants.

Unlike most animals, plants are notoriously relaxed about their reproductive
habits, showing scant regard for conventional barriers to breeding. The
nature of pollination, combined with a certain readiness to hybridise, means
that there is always a potential for exotic sexual union, both within and
between species . On the farm, hybridisation is most likely where a flowering
crop grows in proximity to wild relatives. These may either be native plants,
or weeds that escaped from the crop in previous years and have taken up
an independent existence. If pollen makes the journey from field to hedgerow,
it may sire hybrid seeds containing genes from the crop. Such effects are
not new; the advent of genetic engineering has merely brought them into
sharper focus.

Past ages must have witnessed many genetic escapes from crops into the
wild. Jack Harlan of the University of Illinois has studied the process
in sorghum, a cereal grown in the tropics. Sorghum fields are often plagued
by weeds, which arise through hybridisation between the cultivated plant
and its wild relatives. In many cases, the weeds bear a close resemblance
to the crop itself. Two processes enhance the similarity: a steady flow
of genes from the crop into weeds via pollen and a tendency on the part
of farmers to overlook the best mimics among the weeds.

Researchers believe that genetic leakage must occur in a wide range
of crops, such as oilseed rape, other brassicas, apples and sugar beet (which
is notorious for growing in the company of weeds). Yet there is a shortage
of well-documented cases. When the Royal Commission on Environmental Pollution
examined the matter recently, it found no evidence that traits such as resistance
to insect pests had spread from traditional crops into wild relatives.

Until research gets under way, ideas about the frequency of genetic
leakage from crops and its impact on wild populations will remain a matter
of guesswork. The trouble is that leakage depends on a host of factors:
whether the crop is allowed to flower, how far its pollen travels, the success
or otherwise of fertilisation, the extent to which seeds are dispersed,
their chances of survival and so on. Even if a gene does escape, its future
may be bleak if it handicaps its new host. Researchers now want to examine
each link in this chain, so that breeders and growers can take suitable
precautions once engineered crops arrive on the farm.

Of particular importance is the aerial transport of pollen grains. With
plants that are pollinated by insects, one approach is to follow the pollinators
as they forage, tracking their flights from flower to flower. Insects such
as honeybees and bumblebees are often loyal to a particular variety of flower
at any one time. Moreover, they prefer short trips between blooms, visiting
several flowers on a single plant perhaps, and then flying to a near neighbour,
a few metres away at most.

Yet because bees do not deposit all the pollen from one plant on the
next, genes tend to travel further than appearances suggest, according to
Barbara Schaal at Ohio State University. Schaal created an artificial population
made up of three varieties of a flower called the Texas bluebonnet (Lupinus
texensis), each of which carried a distinctive genetic marker. She confined
each variety to a particular section of the plot and allowed them to reproduce
normally. By analysing the seed, Schaal could deduce how far the marker
genes had travelled during pollination. While bees generally went in for
short hops of less than a metre, genes travelled twice that distance. Even
longer journeys are possible if insects other than bees work the flowers.
Butterflies, for example, travel up to 12 times as far as bees do between
successive stops.

Researchers have devised some ingenious techniques for mapping the movements
of pollen. In a recent test, James Thomson and Barbara Thomson at the State
University of New York, Stony Brook, made use of an unusual species from
the lily family, Erythronium grandiflorum. This plant exists in two forms,
one with red pollen and the other with yellow pollen. The Thomsons planted
a flower of the red type in the middle of a stand of yellows, with the result
that pollination became colour coded. The advance of the red tide confirmed
that bumblebees generally travel short distances, but it also revealed the
occasional lengthy trek of as much as 35 metres. The researchers conclude
that most flowers in their population were producing some pollen that travelled
a long distance.

Mick Crawley is pondering the problem at Imperial College’s field station
at Silwood Park in Berkshire. ‘A very, very high proportion of pollen travels
a negligible distance,’ says Crawley, ‘but the proportion that does travel
a distance travels a great distance.’ Such effects are hard to describe
mathematically, but traditional plant breeders have understood them for
decades. They know that if a plant variety has to be kept genetically pure
– if growers are multiplying it up for general use, for example – then they
have to grow it at a certain distance from any nearby relatives. For sugar
beet or brassicas, the distance involved has to be at least a kilometre
if contamination is to be avoided.

Norman Ellstrand of the University of California at Riverside vividly
demonstrated the effect with radish, a plant pollinated by honeybees. Ellstrand
began by analysing the genetic make-up of a population of wild radish. After
the plants had flowered, he made a similar analysis of their seed. Some
seeds had a genetic profile impossible to derive from the local plants,
indicating that pollen must have made the long journey from a separate population.
Some pollen had travelled at least a kilometre. Considering his findings
in the light of genetic engineering, Ellstrand concludes that the escape
of engineered genes into natural populations is likely for crops that have
local wild relatives. Ellstrand recommends a series of precautions for preventing
the spread of genes through pollen. These include physical isolation, altering
the timing of flowering so that it no longer coincides with the flowering
season of nearby wild relatives, or growing strains that cannot produce
viable pollen.

Given the foraging habits of bees and their fidelity to a particular
spot, it is difficult to understand how they can ferry genes over long distances.
One suggestion is that they pick up pollen from one another inside the hive.
They might then carry it back to their own patch. It is not unusual to find
three or four different brands of pollen grain on a bee’s back. Such grains
could survive for between 24 and 48 hours, a state of affairs that promotes
adventurous pollination.

With plants that are pollinated by wind, most pollen is again deposited
near to home, but a small amount travels much farther. Urban hay fever is
one consequence. The vast majority of this airborne pollen will never contact
a receptive flower, so from a genetic point of view it is inert. Nevertheless,
gene flow can be impressive in wind-pollinated plants such as the grasses.

In a classic study two decades ago, Tom McNeilly of the University College
of North Wales, Bangor, examined the vegetation of an old copper mine at
Drws y Coed, Gwynedd. Grass of the species Agrostis tenuis is normally no
lover of copper, but at Drws y Coed a strain that tolerates copper thrives
on the polluted soil. Elsewhere the tolerant grass is at a slight disadvantage,
so it tends to yield to the normal form. McNeilly sampled plants from a
series of sites in and around the mine, which, like most spots in Wales,
generally enjoys a bracing wind from the west. Upwind of the mine, he found
a sharp cutoff in tolerance to copper across the mine’s boundary. Downwind,
the story was entirely different. Even at a distance of 150 metres, tolerant
grass was growing on normal soil. Any slight local disadvantage to the tolerant
form was more than outweighed by the endless volley of genes fired from
grasses living on the mine.

As well as studying flows of genes in nature, researchers are also looking
more directly at pollen from genetically engineered crops. At the AFRC’s
Institute of Plant Science Research (IPSR) in Cambridge, Phil Dale has arranged
an intriguing experiment. Half a kilometre from the nearest ordinary potatoes
is a square plot measuring 60 metres on a side. At its heart is a bed of
genetically engineered potatoes. These plants contain two extra genes, both
of which came from bacteria. One of the bacterial genes renders the potato’s
cells resistant to an antibiotic called kanamycin; the other simply advertises
its presence by making a distinctive enzyme. (Neither gene is agriculturally
useful, except as an experimental tool.) Surrounding the engineered plants
are ordinary potato plants and a range of species from the potato’s family,
the Solanaceae, which includes tobacco, petunia and wild flowers such as
henbane, bittersweet and black nightshade.

As the summer advances and the potatoes flower, Dale and his colleagues
track the movements of bees and other insects visiting the site. At the
end of the season, they will harvest berries from all the surrounding plants
and gauge their ability to germinate and grow in the presence of kanamycin.
Because kanamycin is normally poisonous to plant cells, success will mean
that pollen from the engineered potatoes has fertilised the species in question,
producing hybrid seed carrying the gene for resistance to the antibiotic.

Transfer of foreign genes

If hybridisation fails, as seems likely in crosses between potato and
other British species, the team will try to identify the cause. The pollen
may fail to penetrate the stigma, for example, or the seed might not develop
properly; everything depends on the mechanisms that operate in a particular
case. Dale reasons that if hybridisation breaks down at a late stage, there
is a greater chance that some hybrids will eventually appear. ‘The further
it gets through the process,’ he says, ‘the more likely it is that that
rare event will happen.’

Much more common will be pollination of the ordinary potatoes by their
engineered counterparts. Such risks are also under consideration by George
Mackay and colleagues at the AFRC’s Scottish Crops Research Institute at
Invergowrie. Mackay’s approach is to look for exchange of genes between
varieties bred in the conventional way, rather than planting engineered
potatoes in the field. One idea might be to grow a variety that resists
attack by nematode worms close to a susceptible variety. If cross-pollination
takes place, seed from the susceptible form might acquire resistance to
nematodes.

As far as British agricultural crops are concerned, genes are most likely
to escape from crops into wild plants (rather than to other cultivated varieties)
in sugar beet and brassicas, especially oilseed rape. At the University
of Birmingham, Brian Ford-Lloyd is hoping to start experiments on sugar
beet. Instead of working with an engineered variety of the crop, his idea
is to study wild and cultivated beets and search for any genetic similarities
that might signal a past exchange of genes.

In France, at the University of Paris-Sud, Pierre-Henri Gouyon is looking
into the likelihood of genes escaping from oilseed rape. There are at least
two potential sinks for such genes: rape that has escaped from cultivation
into the hedgerows and wild relatives such as Brassica campestris, sometimes
called the wild turnip. Gouyon’s research focuses on a gene that governs
resistance to a herbicide called atrazine. Breeders have introduced this
gene into rape, both by conventional methods and by a technique called somatic
fusion. (This involves forcing cells from two plants to join together in
a test tube and then to develop into a new plant.) Whatever the origin of
resistant rape, genetic escapes could generate a population of weeds that
resist herbicides.

The gene that gives resistance to atrazine resides not in the nucleus
of the cell, but in its chloroplasts, the cellular sugar factories. Although
many plants inherit chloroplasts only from their female parent, inheritance
through pollen is not unknown. Gouyon is investigating whether pollen grains
from oilseed rape can transmit chloroplasts down the generations. Another
route by which such a gene could escape is through seed. ‘The cultivated
plant could receive some pollen from the weed,’ explains Gouyon, ‘then some
hybrid seed could be produced and some of this seed could be dispersed.’
Either way, the upshot would be hybrids between weed and crops, carrying
genes for resistance to atrazine. Gouyon is currently surveying the hybrid
plants around fields of rape to see if the genes inside their chloroplasts
bear any resemblance to those in the rape crop. Similarities might signal
past escapes of genes. To complement these studies, Gouyon and Chris Gliddon
in Bangor, are devising mathematical ways of predicting the spread of escaped
genes in wild populations. According to Gliddon, the rate of spread depends
critically on the nature of the population in question, whether it is split
into small subgroups, for example. ‘It’s also going to be very dependent
on the amount of long-distance gene flow that there is,’ explains Gliddon,
‘and this is the area where not only is there no evidence, but it’s almost
impossible to conceive of trying to measure it.’

Gliddon also points to the importance of what he calls the citadel effect.
‘If a released gene can reach a fairly high frequency in a local population,
it will have formed what I would call a citadel, from which it can then
emigrate. Unless you wipe out that citadel – and that’s very difficult –
it can continue to reinfect the species that are available.’

Another approach, currently under test by Michael Wilkinson at the Scottish
Crops Research Institute, is to study the nature of the pollen produced
by genetically engineered plants. Some researchers believe that such pollen
could be at a disadvantage compared to ordinary pollen, simply because it
contains extra genes. Any such handicap would be compounded if these genes
were permanently switched on, so wasting valuable resources. Wilkinson is
comparing the speed with which pollen grains from engineered and nonengineered
plants germinate and grow.

Even if genes from crops do leak into the wild via pollen, there is
no guarantee that they will spread. Everything depends on how they affect
their hosts. Until recently, most biologists agreed that genes from cultivated
varieties had little to offer wild plants. ‘Most of the genes that we select
for, or have selected for in the past, would be of no use to wild relatives,’
says Wilkinson. ‘Cultivars are monsters, ill adapted to the natural environment.’
Whether the same can be said of the new generation of engineered genes is
unclear.

A related issue concerns the ecological effects of engineered crops
themselves. This will form a second focus of the PROSAMO programme under
Crawley at Silwood Park. ‘The most interesting part of PROSAMO isn’t the
pollen flow,’ says Crawley. ‘There are so many incompatibility mechanisms
that the probability of producing hybrid plants is exceptionally small.
The much more interesting ecological question is: does making crops transgenic
make them any more likely to become an ecological nuisance of some kind
– for example, invaders of natural habitats or more persistent weeds of
natural habitats?’ The Royal Commission on Environmental Pollution recently
concluded that, in theory, small genetic changes could transform a well-behaved
crop into a potential weed. Engineering a salt-tolerant crop, for example,
might run the risk of creating a new estuarine pest. Potatoes resistant
to blight, frost and herbicides could become serious weeds.

Leakage of genes from crops into the wild has probably been going on
since people first began to till the soil. So genetically engineered plants
do not present a new problem, merely a variant on an old one. Against this
background, many researchers believe that the origin of a crop variety,
conventionally bred or engineered, is not itself all that important. When
it comes to reckoning risks, what is important are the genes that the plant
happens to have on board, not how they got there. Critics disagree, arguing
that engineered crops can carry genes from stunningly diverse sources –
unrelated plants, bacteria, viruses or animals. However benign such genes
appear, leakage into the wild could present them, at least in theory, with
all manner of new evolutionary opportunities. Outwitting the roving pollen
grain should ensure that they stay put.

* * *

From errant gene to aggressive alien

WHEN pollen from one species lands on the flowers of another, a battery
of mechanisms comes into play to block a fruitful union. Occasionally, however,
the result is a seed that can grow into a fully fledged hybrid plant. Hybrids
are often sterile, but this may not halt genetic adventure.

For any plant to be fully fertile, its chromosomes must pair up before
separating and packing into the sex cells. Hybrids inherit chromosomes from
two different sources, so pairing tends to be difficult. Yet some hybrids
are able to overcome this problem by the simple expedient of acquiring double
the normal number of chromosomes. The chromosomes can then pair happily,
like with like, and fertility returns. The result can be a new species,
breeding independently of its parents.

This process has played an important part in the evolution of flowering
plants. Occasionally the impact can be dramatic. When two species of cord
grass hybridised in Southampton Water around 1870, the result was a sterile
hybrid. Doubling of chromosomes turned it into Spartina anglica, an aggressive
resident of mudflats. Wheat and cotton are examples of crops that have arrived
as a result of hybridisation and duplication of chromosomes.

The tale of the Oxford ragwort offers a recent and striking example
of genetic wanderlust. The story begins in the 17th century with the introduction
from Italy of Senecio squalidus to the Oxford Botanic Garden. In the 1790s,
the ragwort went over the wall, so to speak, and began to spread. It is
now north of the River Tay in Scotland – and still on the move. On the way,
it has spawned a new species and altered the genetics of a native flower,
according to Richard Abbott and his colleagues at the University of St Andrews.

The new species arose between 1910 and 1930. It formed when the Oxford
ragwort crossed with the common groundsel, Senecio vulgaris. The resulting
hybrid was sterile to begin with, until it doubled its complement of chromosomes.
Once fertile, it achieved the status of a new species, the Welsh ragwort
S. cambrensis. At present, the new species appears confined to parts of
north Wales and Edinburgh. Studies by Abbott and his colleague, Paul Ashton,
confirm that the new species has evolved independently in the two areas.

Sterile hybrids between Oxford ragwort and groundsel are not confined
to these zones. Neither are they completely sterile.

Every so often they manage to produce some viable sex cells. When this
happens, they can cross successfully with groundsel. The issue of this union
is a form of groundsel with a hint of Oxford ragwort. The plant shows its
exotic pedigree in two main ways. First, its flowers are not the standard
compact shape: they have some protruding rays like those of the ragwort.
Secondly, these unusual flowers are much less likely to fertilise themselves
than are those of regular groundsel. In effect, the hybrid is acting as
a genetic go-between, channelling genes from ragwort to groundsel, where
they are influencing the most fundamental traits.

Stephen Young is a science writer based in Wales.