news and views
damaged areas, they give rise to a pool
of new myelin-producing cells (the oligodendroglial cells), and to new neurons
(Fig. 1). They also produce growth factors,
including ciliary neurotrophic factor, that
may provide a restorative milieu6. Most
importantly, the symptoms of mice improve
even when the precursor cells are delivered
to animals already suffering an attack of
paralysis. Disability wanes, and electrical
conductivity along nerve fibres increases.
The potential of strategies such as this
to treat neurological damage on a wide front
is impressive. Although some neurological
disorders, such as Parkinson’s and Huntington’s diseases, are confined to specific brain
regions, others, like multiple sclerosis and
Alzheimer’s disease, affect much broader
areas. When the damage is confined, a local
injection of neural precursors might be
beneficial. Similarly, in previous studies7,8,
precursors of myelin-generating cells have
been transplanted directly into demyelinated
brain regions. But this would be an impractical means of coping with the widespread
damage seen in multiple sclerosis, which
must be tackled differently. Until now, this
requirement seemed daunting, but the results
of Pluchino et al. put matters in a new light, by
showing that neural precursors can be injected
into the blood or spinal fluid and still find
their way to the many areas where they are
needed. One point of particular interest here
is that these cells hitch a ride into damaged
sites by using a4 integrin — the very molecule
that mobilizes the immunological attack2,4,5.
To give such cell-based strategies the best
possible chance, it will be imperative to
reduce the risk that newly formed myelin-
producing cells will be targeted in another
round of friendly fire2,9. But on both fronts
— in silencing the autoimmune attacks and
in repairing the brain damage — there is,
I believe, good reason to be optimistic.
Many attractive methods for dampening
the autoimmunity that is characteristic of
multiple sclerosis are under development.
These include broad-scale tolerization with
myelin-derived peptides2 and with genes
encoding myelin proteins2,3,9. They can perhaps be combined with well-known drugs
such as statins, which have recently been
shown10 to be extremely effective in suppressing autoimmunity. It should be feasible
to stop collateral damage. And once the
immune system has been made to surrender,
the molecules at fault can perhaps be turned
to help promote rehabilitation. If sufficient
numbers of human neural precursor cells can
be collected, and if we can work out how to
make these cells proliferate and differentiate,
then the results of Pluchino et al. might be
translated into a treatment that eliminates
collateral damage in multiple sclerosis.
■
Lawrence Steinman is in the Department of
Neurological Sciences and Neurology,
Interdepartmental Program in Immunology,
Stanford University, Stanford, California 94305, USA.
e-mail: [email protected]
1.
2.
3.
4.
5.
6.
7.
Pluchino, S. et al. Nature 422, 688–694 (2003).
Steinman, L. Nature Immunol. 2, 762–765 (2001).
Lock, C. et al. Nature Med. 8, 500–508 (2002).
Yednock, T. et al. Nature 356, 63–66 (1992).
Miller, D. et al. New Engl. J. Med. 348, 15–23 (2003).
Linker, R. et al. Nature Med. 8, 620–624 (2002).
Archer, D., Cuddon, P., Lipsitz, D. & Duncan, I. Nature Med. 3,
54–59 (1997).
8. Imaizumi, T., Lankford, K., Burton, W., Fodor, W. & Kocsis, J.
Nature Biotechnol. 18, 949–953 (2000).
9. Garren, H. et al. Immunity 15,15–22 (2001).
10. Youssef, S. et al. Nature 420, 78–84 (2002).
Plant biology
Mutual sanctions
Janet Sprent
The bacterium-filled nodules found on legumes represent a mutually
beneficial arrangement. But it is evidently one with sophisticated checks
and balances to ensure a fair deal for both partners in the marriage.
ome soil bacteria live in apparent
harmony with plant cells in a mutually
beneficial arrangement. The bacteria
can reduce nitrogen gas, ‘fixing’ it into forms
that the plants can use; in return, the plant
cells provide the bacteria with products
of photosynthesis. On page 722 of this
issue1, Lodwig and co-workers describe an
exchange-control system that enables the
two partners to share their resources without
either one becoming dominant.
The enzyme complex involved in nitrogen fixation is nitrogenase, which is ancient
and widespread among bacteria. Nitrogenase
can use a variety of substrates, but its main
role in today’s world is the production of
S
672
ammonia (NH3) from nitrogen gas (N2).
Whereas free-living nitrogen-fixing bacteria use ammonia for their own growth,
those living in symbiosis with other organisms, such as in the nodules on roots of pea
and bean plants, normally hand it over to
their host in the form of ammonium ions, in
exchange for products of photosynthesis that
are used to provide the energy for nitrogen
reduction2. Why should these bacteria
behave so altruistically, when by so doing
they lose their own source of amino acids?
Lodwig et al. propose an answer. Using
plants of the garden pea, they induced the
formation of root nodules containing
either wild-type or mutant nitrogen-fixing
© 2003 Nature Publishing Group
bacteria (known collectively as rhizobia).
Through analysis of these nodules, they
could then separate and dissect the processes
of nitrogen reduction, assimilation of
ammonium into amino compounds, and
transport between the two partners. To the
authors’ surprise, the host plant cells could
not assimilate ammonium when they were
nodulated by bacterial mutants in which
amino-acid transport was blocked.
There were two aspects to this observation. First, mutants that could fix nitrogen at
rates comparable to the wild-type bacteria
could not pass the products on to the host
cell unless they were supplied with an amino
acid, probably glutamate, by the host cell.
Second, the host cell could not assimilate
ammonium from bacteria unless it was
also supplied with another amino acid,
aspartate. Lodwig et al. propose that these
two transport systems may have distinct
functions in symbiosis (see Fig. 4 on page
725). One serves to import glutamate from
plant to bacteria, and the other to export
aspartate from bacteria to plant. So each side
can impose a sanction on the other, by withholding a vital amino acid. If this circuit is in
place, bacteria can export ammonium and
ensure both their own amino-acid supply
and that of their host. Thus, both sides have a
strong interest in maintaining the marriage.
Before a host plant accepts bacteria into
this intimate association, an intricate dialogue occurs between the two partners,
which tests their mutual compatibility.
Events begin in the soil, when plants and rhizobia exchange signals. They proceed via
‘infection pathways’ and nodule development (Fig. 1 shows a variety of nodule
types). And they culminate in the formation
of symbiotic units such as those studied by
Lodwig et al.1. But does this courtship always
end in harmony? Unfortunately not.
Problems may occur at any stage, and two
are illustrated by the work of Lodwig et al.
First, with bacterial mutants that can induce
nodulation but cannot allow ammonium
assimilation, numerous small, ‘ineffective’
nodules result, typical of those sometimes
found in nature. In this case, host control
over the number of nodules produced3 is
depressed. Second, mutants that cannot
effectively use the host products of photosynthesis to fuel nitrogen fixation may store
those products in the form of the polymer
polyhydroxybutyrate (PHB). This polymer
accumulates naturally in bacteria of certain
nodules, most notably those of soybean, but
much less so in their close relatives, such as
Phaseolus vulgaris (French bean, navy bean)
or species of Vigna (cowpea, green gram).
Does this mean that soybean nodules and
their bacteria are less well matched? Or
does PHB have another function4? These
are just two of the questions raised by the
new results1.
More broadly, other issues arise when we
NATURE | VOL 422 | 17 APRIL 2003 | www.nature.com/nature
news and views
a
b
c
d
Figure 1 Nodule variety. a–c, The nodules formed by nitrogen-fixing bacteria come in various
forms, ranging from spherical, to branched and coralloid. The plants involved are, a, Centrosema
angustifolium, a tropical forage legume; b, Chadsia grevei, a shrub from Madagascar; and
c, Enterolobium cyclocarpum, a Brazilian tree. d, Nodules usually form on roots but on some
species, such as Aeschynomene sp. from Senegal, shown here, they occur on stems.
to the highly effective in delivering ammonia6. Similarly, a single strain of bacterium
may nodulate many genera and species of
legume. Bacteria that induce nodules are now
known to be far more heterogeneous than
once thought, with many having close relationships with plant or animal pathogens —
even to the extent of being members of
the same genus, as occurs, for example,
with species of Burkholderia and Ralstonia.
Symbiotic and pathogenic relatives may have
similar ways of avoiding their host’s defence
responses7. Genetic exchange between
bacteria in soil may lead to some species
losing the genes determining symbiosis and
nitrogen fixation, and others gaining these
genes8. We can expect many ‘new’ nodulating
bacteria to be found in the future.
When coupled with the impressive range
of techniques for studying whole genomes
One host
genotype
One bacterial
genotype
Simple, specific
signal exchange
Effective symbiosis
More than one
host genotype
One bacterial genotype
More than one
bacterial genotype
One host genotype
Complex, multiple
signal exchange
General trend with increasing latitude
Decrease in specificity, increase in promiscuity
look at the full landscape of nodulation
processes. Our detailed knowledge of nodulation comes from just a few species of the more
highly evolved legumes, mainly from temperate or sub-tropical regions. But legumes are
the third largest family of flowering plants,
and nodulation has arisen in them on several
separate occasions during evolution; many
woody species still lack this ability5. There are
thus wide variations in the specificity and
strength of the association with rhizobia,
especially in the tropics (Fig. 2).
Many interactions lack the close coevolution of host and bacteria that leads to
the highly effective recognition and developmental processes evident in pea and most
temperate species. A single host species may
be nodulated by several different genera and
species of bacteria5, with bacteria inside the
nodules varying from the essentially parasitic
Wide variation in number
of nodules produced,
size and longevity of
nodules, and effectiveness
of nitrogen fixation
Figure 2 No fixed relationship. Interactions between soil bacteria and host legumes vary widely in
their specificity, and in the strength of the association and its results. The most widely studied
interactions are the highly specific ones found in advanced legumes of a particular subfamily (the
Papilionoideae)5. But the less specific associations found in many legumes from all subfamilies may
be more common. There is an overall trend in specificity and likelihood of nodulation with latitude
(and, to some extent, altitude): the higher the latitude, the more specific the relationship between the
host plant and the bacterium.
NATURE | VOL 422 | 17 APRIL 2003 | www.nature.com/nature
© 2003 Nature Publishing Group
100 YEARS AGO
The Corporation of the City of London is
rightly taking part in the crusade against
tuberculosis. It has for many years instituted
legal proceedings against farmers, butchers
and meat-salesmen for sending tuberculous
meat into the City markets, or for exposing
the same for sale. Since it would appear
that in some cases such offences may have
been due to ignorance, the Public Health
Department has issued a circular describing
the indications of tuberculosis in the
carcase, and the symptoms of the disease
in the living animal.
ALSO...
Reuter reports that an eruption of the
volcano Del Tierra Firme (Columbia), near
Galera de Zamba, occurred on March 22 by
which the village of Tiojo was destroyed.
Brightly illuminated clouds, giving rise to the
appearance of flames, were seen above the
volcano on the night of March 24 by ships
passing sixty miles off the coast.
From Nature 16 April 1903.
50 YEARS AGO
At all periods, mankind has danced to get rid
of surplus nervous emotion — to obtain
release. During the First World War a United
States hospital unit took over a British
general hospital soon after the Germans had
launched mustard-gas attacks. The sights
and sounds were particularly distressing and
the nurses, new to war conditions, in many
cases became hysterical, though doing their
duties magnificently. When the matron
organized dances, the nervous tension was
released and the troubles ceased… Even in
prehistoric times it would seem that dancing
had a place in the various cults. Both in
ancient Egypt and in Greece dancers are
shown in the pictures of religious festivals;
again, we read in the Old Testament how
David danced before the Lord… Naturally,
then, when Europe became Christian,
dancing was absorbed into the new cultus,
though the Church naturally looked on it with
disfavour, and from time to time attempts
to exclude it were made. Nevertheless, it
was not only the populace who frequently
expressed their religious emotions by
dancing; the clergy and choir, too, sometimes
danced during the services. The Easter
dances before the high altar in the cathedral
at Seville are well known and still take place,
and many less famous though equally
ancient ones still happen in churches or
churchyards at certain times of the year.
From Nature 18 April 1953.
673
news and views
of legumes and other organisms9, and the
detailed literature on the various steps in
nodulation2, highly targeted work such as
that of Lodwig et al. will deepen our understanding of how nitrogen-fixing symbioses
function. If this is extended to other legumes
and other nodulating bacteria, exciting
prospects are raised for answering questions
ranging from why some legumes cannot
nodulate to what distinguishes a pathogen
from a symbiont. Above all, perhaps, given
their agricultural importance, a better
understanding of tropical legumes will assist
the management of nitrogen fixation in
those areas of the world that need it most. ■
Janet Sprent is emeritus professor in the School
of Life Sciences, University of Dundee,
Dundee DD1 4HN, UK.
e-mail: [email protected]
1. Lodwig, E. M. et al. Nature 422, 722–726 (2003).
2. Perret, X., Staehelin, C. & Broughton, W. J. Microbiol. Mol. Biol.
Rev. 64, 180–201 (2000).
3. Downie, J. A. & Parniske, M. Nature 420, 369–370 (2002).
4. Lodwig, E. & Poole, P. CRC Rev. Plant Sci. 22, 37–78 (2003).
5. Sprent, J. I. Nodulation in Legumes (Royal Botanic Gardens,
Kew, 2001).
6. Burdon, J. J., Gibson, A. H., Searle, S. D., Woods, M. J. &
Brockwell, J. J. Appl. Ecol. 36, 398–408 (1999).
7. Roop, R. M. II et al. Vet. Microbiol. 90, 349–363 (2002).
8. Van Elsas, J. D., Turner, S. & Bailey, M. J. New Phytol. 157,
525–537 (2003).
9. Trevaskis, B. et al. Comp. Funct. Genom. 3, 151–157
(2002).
Earth science
Roots of the matter
B. L. N. Kennett
How far down does the ancient continental material that constitutes
Earth’s ‘tectosphere’ extend? Fresh interpretation of the behaviour of
seismic waves helps in reconciling previous estimates.
ver the past three decades there has
been vigorous debate over how thick
the continents can be — that is, the
depth to which the rigid crust and upper
mantle reach before meeting convecting
mantle that can flow and drive tectonic
O
motion. On page 707 of this issue1, Gung
and colleagues add new seismological interpretations that go some way to explaining the
differing views.
The oldest continental rocks are more
than 3.8 billion years old and there are
extensive regions of continents, known as
‘shields’, that are older than 1 billion years.
In contrast, the oldest oceanic material is
only about 200 million years old, because
of the cycle in which oceanic crust is created
at mid-ocean ridges and subsequently
destroyed as material is returned to depth in
subduction zones (particularly around the
Pacific). The preservation of old material
as the continents move across the Earth’s
surface due to the relative motions of the
tectonic plates is related to what lies beneath:
samples brought to the surface through
various eruptive processes indicate that
there is a significant difference between the
continental and oceanic environments.
Based on information from heat flow,
geochemistry and the relative delay times of
seismic waves in different settings, Jordan2
proposed the ‘tectosphere’ model, in which a
zone moves with the motion of the plate
lying beneath the old continental shields and
would be expected to be about 400 km thick.
More recent assessments of heat-flow data
and geochemistry favour a zone no thicker
than 250 km. A thickness of 200–250 km is
also consistent with investigations of how
the Earth has responded to the removal of
the load caused by glaciers, and with regional
seismological studies using seismic surface
waves. But many seismological models of
three-dimensional structure based on global
observations would favour a zone extending
Materials science
Graphite has diverse applications,
ranging from pencils to electronic
devices and nuclear reactors.
Defects — displacements of carbon
atoms — can occur in its structure,
which may be beneficial in
electronics, but could be dangerous
in old-style, air-cooled nuclear
reactors such as that pictured.
There, the defects store energy
and can lead to fire. Quantummechanical computer simulations by
Rob Telling and colleagues (Nature
Materials doi:10.1038/nmat876;
2003) now show that these defects
may be structurally more complex
than previously thought.
The structure of graphite itself
is essentially simple. It consists of
layered sheets known as graphene,
each sheet being formed from a
planar array of carbon atoms.
Defects can form, for example,
through the irradiation of graphite in
nuclear reactors. This can induce a
carbon atom to leave a sheet,
674
forming a ‘vacancy defect’. Until
now, it had been assumed that the
remaining atoms in the sheet are
unaffected, and retain a planar
configuration.
But the simulations by Telling
and colleagues show that the
planar state would actually be
unstable, and that the atoms
surrounding the vacancy are more
likely to be displaced out of the
plane, very unlike the situation in
a flawless sheet. The authors
propose that if displaced atoms
in two sheets were near to each
other, a covalent bond could form
between them, effectively bridging
the gap.
A vacancy defect can also
create another type of flaw, in which
the removed carbon atom positions
itself between two neighbouring
sheets, forming an ‘interstitial’.
Again, covalent bonds could form.
Because the interaction between
sheets is usually quite weak — the
BETTMANN/CORBIS
Mind the graphite gap
distance between them is 3.35 Å,
some two-and-a-half times greater
than the distance between atoms
within a sheet — bonding between
them has implications for the
properties of the structure.
Although these simulations
challenge current ideas about the
nature of graphite defects, they are
not inconsistent with some of the
accepted evidence; they simply
provide another explanation for the
© 2003 Nature Publishing Group
results. Further research is needed
to corroborate Telling and
colleagues’ theory, but, if validated,
the new understanding may help to
make the decommissioning of old
nuclear reactors safer, and could
pave the way to a whole new set
of materials based on carbon
nanotubes — effectively, rolled-up
graphene sheets — which are
already provoking great interest and
a wealth of research.
Jane Morris
NATURE | VOL 422 | 17 APRIL 2003 | www.nature.com/nature