OOuu
ttss
iidd
ee  
JJEE
BB  
Outside JEB
iv
Keeping track of the literature
isn?t easy, so Outside JEB is a
monthly feature that reports the
most exciting developments in
experimental biology. Short
articles that have been selected
and written by a team of active
research scientists highlight the
papers that JEB readers can?t
afford to miss. 
THE DEADLY SCENT OF
FOOD!
It is a well-known phenomenon, from C.
elegans to lab rat to human, that severe
dietary restriction greatly extends life span.
Of course, the problem is that even if
severe dietary restriction didn?t extend life
span, it would certainly feel like it, as
eating becomes a joyless experience. This
is why ?big pharma? is trying to identify
drugs that interfere in the pathway, so that
we can eat as much as we like and still live
longer. How does the body sense the
reduction in nutrient quality, and what is
changed? Downstream elements of the
pathway seem to include insulin signalling,
but which sense is involved is less clear.
Recent work in C. elegans suggested a
possible role for taste in longevity, and so
Scott Pletcher and colleagues decided to
apply Drosophila genetics to the question,
because behavioural analyses of olfaction
and feeding are also relatively advanced in
D. melanogaster. As this tiny fly naturally
feeds on rotting fruit, it finds the volatile
odours generated by yeast to be particularly
appealing. However, a little of what you
fancy can actively do you harm, as we shall
see. The authors exposed flies on a severely
restricted diet to these strong-smelling,
volatile odorants from live yeast, or live
yeast itself. Both treatments significantly
reduced life span, showing that odorants
alone are capable of restricting life span.
Are these odorants simply toxic? Apparently
not, because flies fed on normal diet are
unaffected by the same odorants. So it
seems as if at least a part of the life
extension caused by dietary restriction is
attributable to odorant-mediated perception
of restricted nutrient availability.
If this were the case, then genetic blockade
of olfaction might be expected to extend
life span. In Drosophila, olfaction is
mediated by a family of 62 odorant
LO
NG
 LI
FE
SP
AN
receptors. One of these, Or83b, is broadly
expressed and is required for correct
targeting of other receptors. Mutants are
almost anosmic, meaning that they can?t
smell. Fully-fed female mutants of Or83b
lived over 50% longer than wild-type flies,
and males also lived longer, although the
difference was less spectacular. This
increased life span was ?rescued? back to
normal when a normal copy of Or83b was
expressed in Or83b mutants, showing that
the longevity effect was a direct effect of
Or83b mutation. In simple terms, these
results would suggest that the smell of rich
food restricts life span.
What happens when Or83b mutants are
exposed to dietary restriction? In fact, the
enhanced longevity effect is maintained
even under conditions of severe restriction.
This shows that olfaction is not a necessary
step for the effect of dietary restriction,
although there is obviously some
interaction between their sense of smell
and the effect of eating less. Presumably
the fly has some other means of detecting
nutritional quality of food, such as sensing
the level of nutrients in the haemolymph,
so the olfactory pathway is semi-redundant.
Nonetheless, the depressing message from
C. elegans and Drosophila is clear: if you
want to live longer, eat much less and don?t
enjoy it.
10.1242/jeb.000836
Sergiy Libert, S., Zwiener, J., Chu, X.,
VanVoorhies, W., Roman, G. and Pletcher, S.
D. (2007). Regulation of Drosophila life span by
olfaction and food-derived odors. Science 315,
1133-1137.
Julian A. T. Dow
University of Glasgow
j.a.t.dow@bio.gla.ac.uk
THE JOURNAL OF EXPERIMENTAL BIOLOGY
Outside JEB
v
HOVERING IN THE DARK
In spite of the huge selective advantages
that it can give an animal, powered flight
has evolved only four times in the history
of life on Earth: in insects, pterosaurs,
birds and bats. One of the things that
makes powered flight complex in animals
is that it has to be accomplished via
flapping, which generates inherently
unsteady aerodynamic forces as the wings
move back and forth. These challenges are
greatest for hovering animals, which
require an intricate feedback system to
remain in one place. In dragonflies, flight
stability is probably accomplished using
visual feedback from the eyes. In other
insect groups, one pair of wings has been
modified into a pair of club-shaped
structures called halteres. These structures
allow two-winged insects such as flies and
mosquitoes to detect rotational movements
of their bodies, which is critical for the
maintenance of flight stability in the face
of external perturbations like gusts of wind.
In a recent article in Science, Sanjay Sane
and colleagues ask the question of how
moths achieve flight stability in the
absence of both visual cues and halteres,
since moths fly in low-light conditions and
have two pairs of functional wings. In this
paper, they test the hypothesis that
hawkmoths (Manduca sexta), which are
excellent hoverers, use their antennae in
the same way that flies use their halteres.
This hypothesis predicts that moth
antennae, like halteres, should vibrate at a
rate that is comparable to the wing beat
frequency. This is exactly what they found
when they used high-speed video to
analyze the motion of moth antennae
during hovering. By recording from
neurons in the base of the antennae, they
also demonstrated that antennae are
capable of detecting the kinds of
mechanical perturbations that they would
experience as a result of bodily rotations
during hovering flight.
While these experiments showed that the
antennae could act as rotation sensors, they
did not demonstrate that they actually do.
To directly measure the contribution of the
antennae to flight stability, the researchers
amputated antennae from a group of moths
and measured their flight performance.
Moths lacking antennae were far more
likely to crash to the ground or into the
walls of the test chamber, which was
compelling evidence that the antennae
contribute to flight stability. Antennae are
endowed with a rich diversity of sensory
neurons that detect not only mechanical
input but also chemicals, humidity and
temperature. To test whether any of these
sensory systems is involved in flight
stability, the investigators re-attached
antennae with cyanoacrylate glue and
found that this rescued flight performance. 
These results lead to two important
conclusions. The first is that the odor,
humidity and temperature sensors are not
involved in flight stability, since the axons
from these structures were still severed in
moths with re-attached antennae, and these
moths could still fly. The second is that the
mechanical sensors used for flight stability
must be located at the base of each
antenna, below the point where they were
cut. Structures called Johnston?s organs are
located in this area and are known to detect
antennal movements associated with sound
perception, so it is likely that the
Johnston?s organs are also used in flight
stability in moths. These findings are
interesting because they suggest that
although the evolution of powered flight
has been a rare event, mechanisms of
achieving flight stability have evolved
several times in insects using a variety of
sensory structures including the eyes,
wings and antennae.
10.1242/jeb.000844
Sane, S. P., Dieudonn?, A., Willis, M. A. and
Daniel, T. L. (2007). Antennal mechanosensors
mediate flight control in moths. Science 315,
863-866.
Douglas Fudge
University of Guelph
dfudge@uoguelph.ca
LONG IN THE TOOTH
When a herbivore wears away its teeth to
nothing by years of chewing on abrasive
foliage, its outlook is not bright. Many
specialist herbivore species have
independently evolved long teeth, often as
part of an arms race with plants armouring
themselves with silica. It is then tempting
to suppose that longer teeth should lead to
longer life spans; however, recent work by
Vebj?rn Veiberg and colleagues provides
strong evidence that the reverse can be the
case, with longer life-span requirements
driving the evolution of longer teeth.
The research team studied two populations
of roe deer living wild in forests in France.
Both live in habitats dominated by oak and
beech, but the forests differ in climate and
productivity. Deer from the population
living in the less productive environment
tend to have a longer life expectancy,
probably due to caloric restriction. On the
other hand, those deer from the more
productive forests tend to follow the ?live
fast, die young? strategy. They are probably
more capable of producing more offspring
at a younger age but also suffer a higher
mortality rate.
The team wanted to test two hypotheses:
the ?tooth-wear? hypothesis, which relates
variation in tooth height to diet quality, and
the ?life-history? hypothesis, which relates
initial molar height to life expectancy.
They measured molar heights ? top of
tooth to bottom of enamel ? in 93 animals
found dead within the study areas and
correlated this with age. Veiberg?s team
show no difference in the rate of tooth
wear between the two populations, but deer
from the poor quality environment, with
long life expectancies, started adulthood
with longer molars despite being generally
smaller animals. Thus, differences in tooth
height are related to adaptations for
different life expectancies rather than diet
TO
OT
H 
W
EA
R
FL
IG
HT
 S
TA
BI
LI
TY
THE JOURNAL OF EXPERIMENTAL BIOLOGY
Outside JEB
vi
COCKROACHES GOING
ROUND THE BEND
For animals making their way through
complex environments, walking is not only
a question of moving their legs in rhythmic
patterns, but they also may need to avoid,
obstacles or to reach across gaps. Many
insects are exceptionally good at
maneuvering through complicated
environments such as the forest canopy,
where maintaining coordination is essential
since the penalties for poor performance
can be high: failing to find mates, food or,
worse still, becoming someone else?s food!
Although much is known about how
insects control the movements of a single
limb, how they maintain coordination
between limbs whilst negotiating obstacles
or gaps is less clear. To do this, insects
must take sensory information obtained by
rhythmic movements of the antennae or
forelimbs as well as by the visual system
and use this information to plan leg
movements during walking, which is
difficult to investigate experimentally. 
Angela Ridgel and colleagues at Case
Western Reserve University in Cleveland
used brain lesions in cockroaches, Blaberus
discoidalis, to investigate the role of one
enigmatic brain region that has been
implicated in controlling limb coordination
during obstacle avoidance and turning: the
central body complex (CBC). Following
24?h recovery after surgery, the team
assessed the cockroaches? behavior in two
setups designed to make them turn. Using
a U-shaped arena, they could assess more
general turning and obstacle-avoidance
behavior, while using a tethered setup in
which a wall was moved to simulate an
obstacle allowed them to examine the
detailed movements of the legs.
Lesions made with a razor blade, which
severed links between the left and right
sides of the CBC or connections to and
from one side of the brain and the CBC,
had severe effects on the turning behavior
of the cockroaches in the U-shaped arena.
Over half the turns made by these animals
were abnormal, including turning in
circles, turning towards stimulation of one
antenna but away from stimulation of the
other or simply running into the wall. The
team also made foil lesions at several
regions within or closely associated with
the CBC and likely to be involved in
walking. In many cases, cockroaches with
these lesions failed to turn and ran straight
into the wall. This showed that lesions
made with a razor blade mainly disrupt
left?right coordination ? the cockroaches
still turn but do so incorrectly ? whereas
foil lesions prevent the cockroaches turning
at all. More detailed analysis in tethered
cockroaches showed that many of the
lesions made with foil or razor blades
prevented cockroaches from producing the
appropriate leg movements to turn and
avoid an obstacle.
As the authors point out, this study
represents only the beginning of studies
aimed at finding out how the CBC and its
closely associated brain regions contribute
to limb coordination during complex
behaviors. Their results already emphasize
the importance of communication between
the two sides of the CBC for turning in the
right direction and avoiding obstacles and
of communication between the CBC and
other parts of the brain in allowing the
cockroaches to turn at all. The next step for
Ridgel and colleagues may be to delve
further into the CBC?s circuits in the brain,
where they may find answers to many of
the questions the present study raises such
as how the CBC integrates inputs from the
left and right antennae to adjust the motor
programs that control walking and initiate
turning. These experiments are likely to
prove challenging but have great potential
for providing new insights into insect
motor control.
10.1242/jeb.000869
Ridgel, A. L., Alexander, B. E. and Ritzmann,
R. E. (2007). Descending control of turning
behavior in the cockroach, Blaberus discoidalis.
J. Comp. Physiol. A 193, 385-402.
Jeremy E. Niven
University of Cambridge and
Smithsonian Tropical Research Institute
jen22@hermes.cam.ac.uk
nivenj@si.edu
abrasiveness or rate of chewing. Deer from
poor habitats have to survive long enough
to reproduce effectively despite their low-
calorie diet; increased initial tooth height
evolves in response to this longer life-span
strategy, supporting the life-history
hypothesis.
While it makes some sense that individuals
from a population with greater life
expectancies have to be better built, this
study is remarkable in that it isolates life
expectancy from other factors exceptionally
well, which is challenging to show in wild
populations of relatively large mammals.
Previous work by Juan Carranza and
colleagues, published in Nature in 2004
(vol. 432, p. 215), related tooth wear to life
expectancy and suggested a relationship
between life expectancy and initial tooth
height in red deer. The males truly do ?live
fast, die young?: not only do they have
smaller teeth than females despite their
considerably larger size but they also
expend huge amounts of energy competing
with other males to mate with females.
Hence, they eat much more and wear their
teeth down faster, making it difficult to
attribute tooth height to life expectancy.
Therefore, Carranza and colleagues
couldn?t reject the tooth-wear hypothesis.
Veiberg?s work on roe deer, by contrast,
compares populations with very similar
tooth wear rates, so that the tooth-wear
hypothesis can be cautiously rejected.
So, what we have here is in nice agreement
with ageing evolutionary theory, but it is
also vaguely disturbing. That such a simple
metric as initial adult tooth height might
give away an organism?s ?design life-span?,
or effective ?warranty period?, has
interesting implications. What if insurance
companies could determine a similar
metric for humans? Instead of looking at
indicators of risk for some specific failure
(such as heart disease), they could look at
an apparently innocent trait and calculate
our allotted life spans. 
10.1242/jeb.000877
Veiberg, V., Mysterud, A., Gaillard, J.-M.,
Delorme, D., Laere, G. V. and Klein, F. (2007).
Bigger teeth for longer life? Longevity and
molar height in two roe deer populations. Biol.
Lett. 3, 268-270.
James Usherwood
Royal Veterinary College
jusherwood@rvc.ac.uk
TU
RN
IN
G 
CO
NT
RO
L
THE JOURNAL OF EXPERIMENTAL BIOLOGY
Outside JEB
vii
SENSATIONAL SHELLS
In contrast to the vast majority of
vertebrates, which usually die within
minutes when deprived of oxygen,
freshwater turtles are extremely resistant to
oxygen deprivation, or anoxia. However,
not all freshwater turtles are equally
capable of surviving anoxia. For instance,
the champion survivor, the Western painted
turtle, can recover from up to 100?days of
anoxia at winter temperatures (3?5?C),
whereas the red-eared slider turtle
succumbs after around 45?days of anoxia at
the same temperature. Why such
differences in anoxia survival time exist
between turtle species remains a mystery,
although Donald Jackson?s team at Brown
University has been working feverishly on
elucidating what factors may account for
the differences in anoxia survival times.
In their recent paper, the team surmises
that, since all freshwater turtles are
resistant to anoxia per se, differences in
anoxia survival time among turtle species
must be due to how effectively a turtle can
extend the time before the changes in
physiology that accompany anoxia become
irreversible and lethal. For example, turtles?
fuel reserves, such as liver glycogen,
become depleted and the harmful acidic
end products of anaerobic metabolism,
lactate and H+, accumulate. Therefore, the
greater anoxia tolerance of some turtles
compared with others could be due to more
anoxia-tolerant species exhibiting traits that
prolong the time until glycogen reserves
are exhausted and/or a critically acidic pH
is reached.
Innovatively, a turtle?s shell is not just
protective armour but also reduces the
harmful effects of H+ accumulation during
anoxia exposure. Specifically, shells release
calcium, magnesium and sodium
carbonates into the extracellular fluid in
exchange for lactate to supplement
extracellular buffering of H+. Thus,
Jackson?s team hypothesized that some of
the difference in anoxia tolerance between
turtle species could be due to differences in
buffering characteristics of their shells.
To test this, the team made a number of
measurements to test how well the shells of
different turtle species could contribute to
acid buffering. First, they compared shell
mineral composition and total shell CO2
concentration, which is related to the
amount of carbonate in the shell, to tell
them how many ions were potentially
available for buffering. They found that
shells from the more anoxia-tolerant
painted and snapping turtles had more
mineralised shells and more shell CO2 than
shells from the less anoxia-tolerant map,
musk and red-eared slider turtles,
suggesting that anoxia-tolerant turtles had
more ions available for buffering.
Next, the team tested if the amount of
buffer chemicals released from the shell
differed among species. They measured the
amount of acid they needed to add to
powdered shell samples before the solution
pH could no longer be maintained constant
at pH 7 and found that more acid was
needed for shell samples from more
anoxia-tolerant species, indicating that
these shells could release more buffer.
Finally, the team tested whether there were
differences in the shells? ability to
accumulate lactate. They incubated pieces
of shell in a standard lactate solution for a
set time period, finding that shells from
anoxia-tolerant turtles accumulated more
lactate.
Because shells of more anoxia-tolerant
turtle species contain more minerals and
CO2, release more buffer chemicals and
accumulate more lactate than those of less
anoxia-tolerant turtles, the team concludes
that differences in shell composition and
buffering properties likely account for
some of the reported differences in anoxia
tolerance among freshwater turtles. The
superior shell-buffering mechanisms of the
more anoxia-tolerant turtles probably slow
the development of highly acidic
conditions during anoxia, prolonging
survival time. 
10.1242/jeb.000851
Jackson, D. C., Taylor, S. E., Asare, V. S.,
Villarnovo, D., Gall, J. M. and Resse, S. A.
(2007). Comparative shell buffering properties
correlate with anoxia tolerance in freshwater
turtles. Am. J. Physiol. 292, R1008-R1015.
Jonathan A. W. Stecyk
University of Oslo
jonathan.stecyk@imbv.uio.no
AN
OX
IA
 S
UR
VI
VA
L
? 2007 The Company of Biologists Limited
THE JOURNAL OF EXPERIMENTAL BIOLOGY