ScienceWeek June 29, 2007
Dan Agin
June 29, 2007
Vol. 11 - Number 25
--------------------------------
There is nothing like astronomy to pull the stuff out of man. His
stupid dreams and red-rooster importance: let him count the
star-swirls.
-- Robinson Jeffers (1887-1962)
--------------------------------
Contents (full text below):
1. Astronomy: Inside a Cosmic Train Wreck
2. Genetics: Evolutionary Insights from Sponges
3. Behavior: A Narrow Road to Cooperation
4. Neurophysiology: Stressful Pacemaking
5. Earth science: Silicon-Enhanced Core
6. Probabilistic Reasoning by Neurons
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
1.
Science 29 June 2007: Vol. 316. no. 5833, pp. 1852 - 1854 DOI:
10.1126/science.1139057
Astronomy: Inside a Cosmic Train Wreck
Paolo Coppi
When young galaxies crash into each other, the result is often
not a pretty sight. The violent gravitational forces of the
encounter rip apart the beautiful galactic spiral arms, and gas
and stars shoot out into intergalactic space at high velocity
(see the figure). Yet we are only realizing now that the most
important result of such an encounter is often not visible to us
at all.
Much of the gas in the collision is not flung out but instead
cools quickly, collapsing to the center of the system. Eventually
tens of billions of solar masses of gas can pile up into a region
only a few hundred light-years across. The gas becomes so dense
that it blocks most light and so compact that standard ground-
based telescopes cannot resolve the details of the collapse due
to blurring by Earth's atmosphere. The same density and
compactness that make the gas collapse so hard to study
observationally also make it hard to study theoretically. Two
papers in this issue begin to lift the veil on this unexplored
central region. On page 1877, Max et al. (1) report an advance in
ground-based imaging that permits us to directly observe black
holes in the densest areas of the collapse, and on page 1874,
Mayer et al. (2) present high-resolution simulations showing how
black holes in the colliding galaxies follow and respond to the
collapsing gas.
To penetrate the dense gas, Max et al. used a detector operating
at infrared wavelengths. To achieve high spatial resolution, they
used an adaptive optics technique in which the shape of the
telescope mirror is modified in real time to compensate for
jittering of the image due to atmospheric turbulence. This
combination enables Max et al. to present one of the highest-
resolution observations yet of the central, "nuclear" region of
the NGC 6240 galaxy merger, mapping out its distribution of
stellar light and unambiguously reconciling the different
estimates for the positions of the two supermassive black holes
that lurk there.
Mayer et al. present complementary theoretical calculations that
are some of the most realistic to date of the gas distribution at
the center of a merger. Although several important physical
effects, in particular the "feedback" of energy from the luminous
central stars and black holes back into the collapsing gas,
ultimately require better modeling, the calculation already seems
accurate enough to resolve a long-standing puzzle: Rather than
wander forever around the center of the merger, two black holes
in a system like NGC 6240 should quickly merge to emit a
potentially detectable blast of gravitational wave radiation.
Why is so much effort going into understanding what happens when
gas-rich galaxies, and in particular massive ones, collide?
Comparison of data from experiments such as the Wilkinson
Microwave Anisotropy Probe, which tells us what primordial
density fluctuations looked like, to data from galaxy surveys
like the Sloan Digital Sky Survey, which tells us what those
density fluctuations have evolved into today, strongly suggests
that we live in a universe where the matter density is dominated
by unknown massive particles that interact only gravitationally
("cold dark matter"), i.e., one where today's galaxies assembled
hierarchically, from mergers of smaller galaxies (3).
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
2.
Science 29 June 2007: Vol. 316. no. 5833, pp. 1854 - 1855 DOI:
10.1126/science.1144387
Genetics: Evolutionary Insights from Sponges
Michael W. Taylor, Robert W. Thacker, Ute Hentschel
Sponges (phylum Porifera) are among the most ancient of the
multicellular animals, or Metazoa, with a fossil record dating
back at least 580 million years (1). Found both in marine and
freshwater environments, they filter-feed by pumping water
through their bodies, which can contain a remarkable number of
microbial symbionts. Sponges lack many of the characteristics
typical of animals, but recent genomic studies--including the
report by Jackson et al. on page 1893 of this issue (2)--have
shown that they possess many major metazoan gene families.
Sponges are thus invaluable systems for studying the evolution of
metazoans and their interactions with microorganisms.
Furthermore, their highly stable skeletons are of interest to
materials scientists.
Biomineralization is an important feature of metazoan life.
Animals including vertebrates, insects, mollusks, and sponges use
minerals [such as calcium carbonate, iron, and silica] to form
skeletal structures such as bones, seashells, and coral reefs
(3). Biocalcification arose among many metazoan lineages during
the "Cambrian explosion," between 530 and 520 million years ago,
when the ancestors of today's animals first appeared in the
fossil record. Did these lineages share the same gene(s) for
biocalcification, or did multiple independent evolutionary events
give rise to the ability to biocalcify? Recent studies, including
that by Jackson et al., are beginning to provide an answer to
this question.
Jackson et al. use the Indo- Pacific sponge Astrosclera willeyana
to show that the last common ancestor of the metazoans possessed
a precursor to the alpha-carbonic anhydrases. This gene family is
used by animals today in a range of processes including ion
transport, pH regulation, and biomineralization (4). By
integrating molecular techniques ranging from protein sequencing
to gene expression, the authors identified a group of closely
related alpha-carbonic anhydrase sequences in A. willeyana. These
sequences are similar to those recovered from a whole-genome
project on another sponge, Amphimedon queenslandica (5).
Together, the sponge alpha-carbonic anhydrases form a sister
group to those of all other metazoans.
Jackson et al. confirm that at least one of the proteins from A.
willeyana--the Astrosclerin-3 enzyme--possesses alpha-carbonic
anhydrase activity. Expression of this protein in Escherichia
coli yielded activity comparable to that of a highly active
bovine alpha- carbonic anhydrase. Furthermore, the A. willeyana
genes coding for these anhydrases were only expressed in the
outer portion of the sponge, where the calcareous skeleton is
first deposited. Collectively, these data indicate that the
alpha-carbonic anhydrase gene family originated from a single
ancestral gene. This gene subsequently underwent multiple
independent gene-duplication events in other sponges and
eumetazoans, yielding the striking structural complexity and
diversity we see today among biocalcifying animals.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
3.
Science 29 June 2007: Vol. 316. no. 5833, pp. 1858 - 1859 DOI:
10.1126/science.1144339
Behavior: A Narrow Road to Cooperation
Robert Boyd and Sarah Mathew
In every human society, from small-scale foraging bands to
gigantic modern nation states, people cooperate with each other
to solve collective-action problems. They share food to ensure
against shortfalls, risk their lives in warfare to protect their
group, work together in building canals and fortifications, and
punish murderers and thieves to maintain social order. Because
collective action benefits everyone in the group, whether or not
they contribute, natural selection favors non-contributors. So,
why do people contribute? Everyday experience suggests that
people contribute to avoid being punished by others.
But this answer raises a second question: Why do people punish?
From an evolutionary perspective, this question has two parts:
First, how can contributors who punish avoid being replaced by
"second-order" free-riders who contribute but do not incur the
cost of punishing? There has been much work on this topic lately,
and plausible solutions have emerged (1-5). However, these
solutions are not much good unless we can solve the second
problem: How can punishment become established within populations
in the first place? On page 1905 of this issue, Hauert et al.
provide the first cogent answer to this question (6).
Surprisingly, they find that punishment can become established if
there are individuals who neither produce collective benefits nor
consume collective benefits produced by others.
In previous models of the evolution of collective action,
individuals in a group can either contribute and benefit from the
public good (i.e., cooperate), or not contribute and benefit
(i.e., defect). In the absence of punishment, defection wins.
However, if punishment is possible and punishers are common, it
does not pay to defect. But punishment is costly to impose. A
rare punisher in a group of defectors suffers an enormous
disadvantage from having to punish everyone in the group. This
means that in very large populations, punishment can sustain
cooperation when punishment is common, but punishing strategies
cannot increase in numbers when they are rare (i.e., invade a
population of defectors). In a finite population, random chance
affects the number of each type that reproduce, and the resulting
stochastic fluctuations allow punishers to eventually invade a
population of defectors, even though selection favors defectors.
However, it can take a very long time for this to occur, and
thus, most of the time there is no punishment and no cooperation.
Hauert et al. provide a way out of this dilemma. They introduce a
strategy that simply opts out of collective action. These
"nonparticipants" neither contribute to the collective good nor
consume the benefits, but instead pursue some solitary activity.
Surprisingly, this innovation allows punishment to increase when
rare. To see why, consider a population of defectors. Hauert et
al. assume that nonparticipants get a higher payoff than
defectors who attempt to free-ride when there are no cooperators
in their group. Therefore, nonparticipants invade the defectors.
Now, consider a population of all nonparticipants. Hauert et al.
assume that two contributors working together can produce a
higher payoff than a nonparticipant working alone. This means
that rare contributors invade nonparticipants. Once contributors
are common, defectors invade, and the cycle continues. The three
strategies oscillate endlessly (7).
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
4.
Nature 447, 1059-1060 (28 June 2007) | doi:10.1038/4471059a;
Published online 27 June 2007
Neurophysiology: Stressful Pacemaking
Bruce P. Bean
In Parkinson's disease, dopamine-secreting neurons die - perhaps
because unrelenting calcium entry during spontaneous electrical
activity puts them under unusual pressure.
At a conference on calcium-channel-blocking drugs in the 1980s, a
neurologist friend told me of colleagues who kept pills of the
calcium-channel blocker nimodipine handy, planning to ingest one
immediately should they suffer a stroke. Showing even more faith
in the drug, an executive from the pharmaceutical company
sponsoring the conference said that he took an unprescribed
nimodipine pill every morning with his cereal, on the assumption
that it was better not to wait.
Nimodipine and other dihydropyridine drugs were originally
developed to treat high blood pressure; they relax vascular
smooth muscle by blocking calcium entry. Subsequent off-label use
to treat stroke was based partly on the idea that these drugs
might also minimize neuronal death resulting from excessive
calcium entry following oxygen deprivation. But despite the
enthusiasm of my colleagues, controlled clinical trials1 failed
to show any beneficial effects of nimodipine for treating
ischaemic stroke. Now, however, a study by Chan et al.2 (page
1081 of this issue) has raised the exciting possibility that
these or similar calcium-channel-blocking drugs might provide a
strategy to treat Parkinson's disease by their effect on
particular neurons - those that act by secreting the
neurotransmitter dopamine (dopaminergic neurons).
The critical event in Parkinson's disease is death of
dopaminergic neurons in a region of the brain known as the
substantia nigra (SNc). Progressive loss of these neurons
produces devastating symptoms, including tremors and loss of
voluntary movement. Despite intensive research, it is not known
what the causative pathological events are, nor why they
selectively damage dopaminergic neurons. Moreover, current
treatment cannot slow disease progression, only ameliorating its
symptoms3.
Most commonly, activity in a neuron is triggered by
neurotransmitter released by other neurons and diffusing across
the gap, or synapse, between the neurons. But dopaminergic
neurons in the SNc are spontaneously active even without synaptic
input, firing action potentials at about 1-2 hertz (Fig. 1a).
Such autonomous 'pacemaking' activity is seen in many types of
neuron and requires ion channels that can open at membrane
potentials lower than the threshold for firing action potentials.
The electrical current entering the cell through these channels
then depolarizes the membrane to the threshold for action
potentials.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
5.
Nature 447, 1060-1061 (28 June 2007) | doi:10.1038/4471060a;
Published online 27 June 2007
Earth science: Silicon-Enhanced Core
Tim Elliott
What elements, besides iron, make up Earth's core? Discrepancies
in the isotopic ratios found in rocks from Earth's mantle and in
undisturbed meteoritic material indicate strongly that one answer
is silicon.
Earth's core is its most inaccessible part. New information on
its composition - such as that revealed by Georg et al. on page
1102 of this issue1 - is therefore a hard-won prize. The core is
known to be made predominantly of iron, but geophysical estimates
of its density require that a lighter element makes up some 10%
of its mass. Hydrogen, carbon, oxygen, sulphur and silicon have
all been fingered as culprits, but which elements are actually
involved remains controversial.
Despite what Hollywood would have us believe, we can't sample
Earth's core directly. Instead, we must assess its composition by
indirect means. One such method is mass balance: the budget of
elements not in the silicate-dominated, outer portions of the
Earth must be in the core (Fig. 1). For many elements of
interest, the well-mixed, convecting mantle is the only
significant reservoir in these outer layers. Thus, if we know the
composition of the mantle and of the Earth as a whole, the make-
up of the core can be calculated from the difference.
That might sound straightforward. Earth's bulk composition is
estimated from analyses of 'undifferentiated' meteorites, thought
to be made of the same primordial material from which Earth
originally formed. And the composition of the mantle, normally
hidden beneath a veneer of crust, can be determined from rare
fragments fortuitously exposed at the surface. But significant
chemical variability in both the meteorite and mantle samples
makes compositional estimates of the bulk Earth and mantle
uncertain - and constraints on the make-up of the core poorer
still.
Georg et al.1 investigate the possible presence of silicon in
Earth's core using a novel isotopic, rather than elemental, mass
balance. Different isotopes of an element show the same chemical
behaviour, but form bonds of slightly different strengths. This
can result in mass differences (fractionations) between reactants
and products in chemical reactions. Sometimes, such mass
differences are very small. Georg et al. show that this is
apparently so for silicon isotopes in silicate melts and
minerals. They find that the silicon isotope ratios - expressed
as delta30Si, or parts per thousand difference in the ratio
30Si/28Si relative to a reference standard - of a range of
mantle-derived silicates are very similar. The silicon isotopic
composition of the silicate portion of Earth is thus well
defined.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
6.
Nature 447, 1075-1080 (28 June 2007) | doi:10.1038/nature05852
Probabilistic Reasoning by Neurons
Tianming Yang & Michael N. Shadlen
Our brains allow us to reason about alternatives and to make
choices that are likely to pay off. Often there is no one correct
answer, but instead one that is favoured simply because it is
more likely to lead to reward. A variety of probabilistic
classification tasks probe the covert strategies that humans use
to decide among alternatives based on evidence that bears only
probabilistically on outcome. Here we show that rhesus monkeys
can also achieve such reasoning. We have trained two monkeys to
choose between a pair of coloured targets after viewing four
shapes, shown sequentially, that governed the probability that
one of the targets would furnish reward. Monkeys learned to
combine probabilistic information from the shape combinations.
Moreover, neurons in the parietal cortex reveal the addition and
subtraction of probabilistic quantities that underlie decision-
making on this task.
Decision-making is a complicated process that is often based on
more than one source of evidence. The brain needs to combine
these sources to maximize the chance of achieving a correct
decision or to achieve another related goal. Recent advances in
neuroscience are beginning to expose the neurobiological
mechanisms that underlie simple decisions1, 2, 3, 4, 5, 6. It has
been demonstrated that, when the outcome of a decision is an eye
movement, a neural correlate of the evolving decision can be
recorded in brain areas associated with high level motor planning
and attention allocation7, 8, 9, 10, 11. More specifically,
neurons in the lateral intraparietal area (LIP) have been shown
to accumulate sensory information provided by earlier visual
cortex when a decision is being formed8, 9, 12, 13. The mechanism
mimics statistical decision processes that accrue evidence
sequentially in the form of a log likelihood ratio (logLR) that
favours one outcome over another4, 14, 15. Therefore, it has been
hypothesized that a neuronal substrate of probability integration
exists in area LIP16.
To test this hypothesis, we trained two monkeys to perform a
probabilistic categorization task (Fig. 1a). The task was adapted
from the well-known weather-prediction task17, 18 used to study
human learning and memory. In each trial the monkey viewed four
highly discriminable shapes; these were sampled randomly (with
replacement) from a set of ten possible shapes. The shapes were
added successively to the display over four half-second epochs.
The monkey then made an eye movement to either a red or a green
target to receive a reward. Reward was not guaranteed, but was
instead governed by a random process based on the combination of
preset weights (w) that were assigned to the ten shapes {w1, w2,
..., w10}. The sum of the four weights associated with the shapes
shown in a trial established the log of the odds that reward
would accompany a red or a green choice (see Methods Summary).
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=