iv

Keeping track of the literature isnt 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 cant afford to miss. animals. However, in heat-shocked animals, the peak of the DCMDs response which occurs close to the time when an object would be expected to collide with the locust is higher at 45C than at 25C, due to brief, high-frequency bursts of spikes. This does not happen in control animals. Possibly, the higher peak ring rate in heat-shocked animals is a compensation for the lower ring rate earlier in the response, and the differences earlier in the response could cause different behavioural outcomes in control and heatshocked animals. The properties of the DCMD neuron also change as the temperature rises: action potentials in DCMD diminish in height, although in heat-shocked animals action potentials are bigger than in controls. The membrane potential of the neuron is more hyperpolarized (more negative) in heatshocked animals, which also affects how quickly the neuron can recover and transmit the next action potential. In effect, heat shock makes the action potentials in DCMD more robust. Another factor affected by heat shock was an afterdepolarisation (ADP); a rise in the membrane potential seen in the neuron after the action potential. This ADP occurred more frequently and was bigger in heat-shocked animals at all temperatures than in control animals. Although the authors did not investigate the nature of the ADP, in other animals it has been shown to make neurons more excitable, meaning that they are more likely to spike and elicit a response. A similar mechanism could be operating in the heat-shocked DCMD. The effect of heat shock on the properties and response of DCMD is important because it helps reliably maintain the neurons response in the face of high temperatures. This makes it more likely that DCMD will trigger an escape response if a looming object is detected. So, rather than being overwhelmed by an extreme heat wave, a locust might be less likely to end up as lunch.
10.1242/jeb.01725

FEELING THE HEAT

A locust can make a tasty morsel for a hungry predator, so being able to notice potentially dangerous looming objects and trigger the correct response could save a locusts life. In locusts, a visual neuron called the descending contralateral movement detector (DCMD) spikes in response to looming objects and could therefore be the trigger for escape responses. Tomas Money and colleagues at Queens University in Canada knew that nervous systems in cold-blooded animals like locusts can be severely affected by temperature and can even fail at high temperatures. But the effects of very high temperature can be lessened if a creature has experienced a prior heat shock at a high, but sub-lethal, temperature. Money and his colleagues wondered if heat shock protects the DCMD neuron from soaring temperatures. If it does, then the behaviours that are guided by this neuron should be unaffected in heat-shocked locusts, even when the going gets hot. To test this, the authors divided locusts (Locusta migratoria) into two groups: control animals and those that received a heat shock at 45C for 3 h before experiments. Recording from the DCMD, the team explored how the neuron responded to looming objects when they kept the locusts at a range of temperatures from 25 to 45C. As they boosted the temperature, they saw that the number of spikes in control animals DCMD in response to a looming object decreased and the rst spike occurred later. But in heatshocked animals, the DCMDs response stayed the same across temperatures, which suggests that the prior heat shock had indeed protected the DCMD. Looking more closely at the DCMDs ring frequency during the early phase of a threatening objects approach, they saw that spike rate decreased with temperature, but to a lesser degree in heat-shocked THE JOURNAL OF EXPERIMENTAL BIOLOGY

Outside JEB

HEAT SHOCK

Money, T. G. A., Anstey, M. L. and Robertson, R. M. (2005). Heat stress-mediated plasticity in a locust looming-sensitive visual interneuron. J. Neurophysiol. 93, 1908-1919.

Laura Blackburn University of Cambridge lmlb2@cam.ac.uk

Outside JEB
v

REMOTE CONTROL

PASS THE REMOTE PLEASE

A goal of neuroscientists is to understand how the brain produces behaviour. The question of what is happening in the brain during behaviour is quite a brain-teaser. Neuroscientists usual approach is to passively listen to brain activity in freely behaving animals. But what if we could actively excite a specic group of neurons in an animals brain without using invasive electrodes? Lima and Miesenbock have developed a method to do just that; they have worked out how to stimulate a population of specic neurons in a fruit ys brain by remote control. To create their remote-controlled ies, Lima and Miesenbock genetically expressed phototriggers in particular cells in the central nervous system of the fruit y, Drosophila. When they shone a light on these ies, the phototrigger-expressing cells were excited, i.e. they red action potentials. In other words, the researchers could now stimulate a network of neurons in a freely behaving y and observe the behavioural consequences, simply by illuminating the ies. To validate their method, the authors expressed the phototriggers in a very simple circuit that triggers a stereotypical behaviour with an all or nothing response. They chose the giant bre system, which is responsible for a ys typical escape manoeuvre: leg extension, jumping and wing apping. They found that ies that had phototriggers expressed in giant bre neurons exhibited these typical escape responses upon illumination. But when they illuminated ies that did not have phototriggers expressed in the giant bre neurons, they saw that the ies didnt perform an escape response. Lima and Miesenbock also demonstrated quite dramatically that the visual systems response to illumination was not causing the escape response: they shone a light on decapitated ies and found that the

Next, Lima and Miesenbock investigated the role of dopaminergic neurons in the control of movement. Dopaminergic function is important for planned movement, and a loss of dopaminergic neurons leads to a Parkinsonian syndrome of impaired movement. The authors wished to examine the behavioural consequence of an acute increase in dopaminergic signalling. To do this, they expressed the phototriggers in ies dopaminergic neurons, illuminated the ies and watched the resulting ight patterns. Even though the ies ight speed did not change after illumination, they were more active, because the number of pauses between ight episodes and the length of these pauses decreased. Before illumination, ies would y along the edge of the container they were held in and rarely ventured through the centre. But when the researchers illuminated the ies, they saw that the insects began ying through the centre, often in crisscrossing or looped patterns. Clearly, stimulating ies dopaminergic neurons triggers more complex ight patterns, perhaps by initiating more adventurous exploration of the central arena. Identifying and stimulating functionally circumscribed but anatomically dispersed populations of neurons in moving animals has been difficult. Lima and Miesenbocks method might prove useful for the study of any behaviour that is controlled by a given circuit; i.e. most, if not all, behaviours. Courtship, mating, aggression, feeding, grooming, learning, sleep and wakefulness, and reward and punishment are just some potential contenders that the authors propose. Of course, this could also be a PhD supervisors dream: they would only need to press a button to make their students jump!
10.1242/jeb.01728

ZEBRAFISH: A NOVEL SYSTEM TO STUDY KIDNEY DISEASE

Acute renal failure is a serious disease with high mortality rates that have not dropped over the past 40 years. This lack of advancement reects the struggle to nd a suitable model system to study the disease; current mammalian models used to study acute renal failure have proved to be inadequate for a number of reasons. Surprisingly, zebrash offer many advantages over the typical rat or mouse model. Early larval zebrash kidneys possess the biological complexity inherent to kidneys of higher vertebrates. Unlike their mammalian counterparts, zebrash are translucent, facilitating microscopic observation along the entire length of their kidneys. Zebrash tolerate manipulation on a molecular level, and their large numbers of offspring make them an important tool for drug discovery and the generation of transgenic sh. Thus, zebrash might be a good model system to study acute renal failure. That isif zebrash can suffer from it in the rst place. To nd out whether zebrash can develop acute renal failure, Hentschel, Bonventre and their team injected zebrash embryos with the antibiotic gentamicin and the cancer therapeutic cisplatin, which are known to cause kidney damage in humans. But would they have the same effect in zebrash? Sure enough, when the team analysed the kidney tissue of the treated zebrash embryos, they saw morphological changes that are consistent with acute renal failure. Zebrash embryos treated with gentamicin or cisplatin also suffered from severe swelling in a time- and dosedependent manner, reecting an inability to regulate water, probably due to the loss of glomerular function. To assess this potential kidney damage, the team measured glomerular ltration rate in

Lima, S. Q. and Miesenbock, G. (2005). Remote control of behavior through genetically targeted photostimulation of neurons. Cell 121, 141-152.

Susan Sangha Otto-von-Guericke Universitaet Magdeburg susan.sangha@medizin.unimagdeburg.de

THE JOURNAL OF EXPERIMENTAL BIOLOGY

RENAL FAILURE

headless ies still initiated an escape response.

Outside JEB
vi
the treated zebrash embryos. This can be estimated by determining the excretion of a substance that is only ltered at the kidneys glomeruli and not signicantly secreted or reabsorbed by its tubules. Thus, they injected dextran or inulin, two substances that are ltered by the glomeruli but are neither secreted nor reabsorbed by kidney tubules, into zebrash bloodstreams. They had labelled both substances with a uorescent dye. To determine whether the kidneys were successfully ltering and excreting the uorescently labelled substances, the team monitored the decline in uorescence intensity over time. They measured uorescence intensity in zebrash hearts by taking uorescent microscopy images of individual sh immediately after dextran or inulin injection and 1, 5 and 24 h later. They found that the rate of decline in uorescence intensity was greatly reduced in gentamicin-injected sh, indicating that their damaged kidneys were not excreting the substances. The team observed a 75% and 67% reduction in dextran and inulin excretion, respectively. They went on to investigate whether treatments used successfully in mammals with gentamicinand cisplatin-induced kidney damage would also be effective in zebrash. They found that, like in mammals, the amino acid taurine prevents gentamicin-induced damage and the compound Ucf-101 prevents the effects of cisplatin in zebrash. And so, the old August Krogh/Claude Bernard adage that there is a perfect animal system for every biological problem comes to mind in this case. Gentamicin- or cisplatin-injected larval zebrash develop renal failure with characteristics typical of those in higher vertebrates and respond to treatments in the same manner as higher vertebrates. This nding consolidates their role as valuable and unique models for studying the pathophysiology of acute renal failure and for establishing novel therapies for use in humans. Who knew?
10.1242/jeb.01727

GLIDING ANTS

the ants revealed a large drop-off in the number making it back to the tree during a fall (10% vs 85%), suggesting the importance of visual stimuli in locating the trunk and controlling the direction of the glide. The scientists also quantied several gliding performance parameters by analyzing videos of ants dropped within a ight arena near a fabric-covered vertical column. In these trials, they found that the ants average glide speeds were 4.3 m s1 and glide angles averaged 75 (relative to the horizontal). Although this angle is steep, it is far from vertical and reects a much better glide than what Yanoviak and his colleagues calculated for an ant-shaped cylinder falling under similar airow conditions (i.e. at a comparable Reynolds number). It turns out that not all tree-dwelling ant species have (or use) this capacity for directed aerial descent. Instead, the behavior may be associated with groups that share certain traits, including arboreal nests, branch-tip foraging and evolutionary origins in ooded habitats. A broader comparative assessment of the distribution, control and mechanics of this gliding behavior is sure to reveal intriguing results. For now, simply pause to reect on natures brilliance in the form of wingless ants being led by their rears to safety in the canopies of tropical forests around the world.
10.1242/jeb.01726

LOOK OUT BELOW

Did you know that there are ants that actually jump from the trunks or branches of tall trees? Its easy enough to imagine ants being knocked out of their arboreal homes by heavy winds or pelting raindrops, but apparently certain species of canopy ant, such as Cephalotes atratus, voluntarily drop from great heights if approached by an imposing object. Stephen Yanoviak, Robert Dudley and Michael Kaspari recently reported good news for those of us who might be walking under such trees: these ants are unlikely to land on us, or any other part of the understorey for that matter. Instead, ants of this species exhibit directed descent during free fall and somehow generally manage to land on the trunk of the tree from which they fell rather than hitting the ground. To investigate the nature and extent of the ants control during their descent, Yanoviak and colleagues dropped 120 individual worker ants from branches of their resident trees and videotaped their fall trajectories. Falls typically consisted of three phases. First was a period of uncontrolled, vertical free fall, second was a shift in body positioning, and third was a relatively steep glide back toward the trunk, regardless of the animals original orientation when dropped. The scientists used markings on individuals to determine that they glide abdomen-rst. A remarkable 85% actually landed on the trunk and, after a few tumbles, managed to establish their footing. Moreover, within minutes, many of these individuals had climbed back up to the branch from which they had been dropped. As is the case with most things in biology, size plays an important role here. Smaller ants (inter- and intraspecically) do better than larger ants; they fall a shorter distance before landing back on the trunk of the tree from which they were dropped. Further experiments in which the scientists blinded THE JOURNAL OF EXPERIMENTAL BIOLOGY

Yanoviak, S. P., Dudley, R. and Kaspari, M. (2005). Directed aerial descent in canopy ants. Nature 433, 624-626.

Gary B. Gillis Mount Holyoke College ggillis@mtholyoke.edu

Hentschel, D. M., Park, K. M., Cilenti, L., Zervos, A. S., Drummond, I. and Bonventre, J. V. (2005). Acute renal failure in zebrash: a novel system to study a complex disease. Am. J. Physiol. 288, F923-F929.

M. Danielle McDonald RSMAS, University of Miami dmcdonald@rsmas.miami.edu

Outside JEB
vii

EXTENDED LIFESPAN

oxidative enzymes, where it converts the ROS hydrogen peroxide (H2O2) to water. The scientic team generated three groups of mice, either overexpressing catalase in peroxisomes (PCAT), the nucleus (NCAT) or the mitochondria (MCAT). They found that the PCAT and NCAT groups showed a slight increase in median lifespan but found no increase in maximum lifespan compared with wild-type (WT) mice. On the other hand, the MCAT mice lived 20% longer than the WT cohort of mice. It seems that removing ROS from mitochondria can extend a mouses lifespan. The team decided to take a closer look at the MCAT mice; they compared the physiological state of various organs during aging between WT and MCAT mice by carrying out a histopathology analysis. They noticed pronounced differences in histopathology between WT and MCAT old mice, notably in the heart. Indeed, many indicators of cardiac pathology were elevated in WT mice compared with MCAT animals, indicating that overexpression of catalase in heart mitochondria affords protection against age-related heart problems. To directly assess whether the healthier hearts of MCAT mice were due to increased protection from ROS, the authors measured the activity of the mitochondrial enzyme aconitase, which is rapidly inactivated in the presence of ROS. When the team added a pulse of H2O2 to the mitochondria, they saw that aconitase activity decreased signicantly in mitochondria from WT mice, whereas the drop was drastically smaller in mitochondria from MCAT mice. The team concluded that heart mitochondria from MCAT mice are more resistant to oxidative stress than those from WT mice.
2005 The Company of Biologists Limited

THE ENEMY WITHIN

In energy metabolism, there are two means of ATP production: glycolysis and oxidative phosphorylation in mitochondria. The fact that mitochondria produce roughly 15 times more ATP per mole of glucose than glycolysis usually makes them the ideal choice for ATP production. However, relying on mitochondrial metabolism has its costs; mitochondria are the main source of reactive oxygen species (ROS), molecules that can attack DNA, lipids and proteins in cells. It is thought that the accumulation of oxidative damage over time is responsible for aging; this has been called the free radical theory of aging. One implicit assumption of this theory is that reducing ROS levels in cells should extend lifespan. This is exactly what Samuel Schriner and collaborators tested. They studied the effect of overexpressing catalase, an ROSremoving enzyme, on lifespan in mice. Catalase is normally found in peroxisomes, organelles that contain

Finally, to show conclusively that ROS levels are lower in cells and mitochondria of MCAT mice, the team measured the level of a marker of oxidative damage to DNA called 8-hydroxydeoxyguanosine (8OHdG) in muscle and heart. They saw that 8-OHdG increased with age in skeletal muscle of control animals but did not increase in the MCAT group, indicating that MCAT mice indeed experience less oxidative damage, in their muscles at least. However, none of the mice hearts showed an age-related change in 8-OHdG levels. The team also noted a trend with age towards a smaller accumulation of mitochondrial DNA deletions, which are associated with oxidative damage, in skeletal muscle and heart of MCAT mice compared with WT animals. This study demonstrates that mitochondria are at the heart of the aging process and that limiting ROS production by mitochondria can extend lifespan. Given the negative correlation between standard metabolic rate and aging, this study will probably stimulate comparative physiologists to identify key intrinsic differences in mitochondrial ROS metabolism among species.
10.1242/jeb.01729

Schriner, S. E., Linford, N. J., Martin, G. M., Treuting, P., Ogburn, C. E., Emond, M., Coskun, P. E., Ladiges, W., Wolf, N., Van Remmen, H. et al. (2005). Extension of murine lifespan by overexpression of catalase targeted to mitochondria. Science doi: 10.1126/science.1106653

Julie St-Pierre Dana-Farber Cancer Institute and Harvard Medical School julie_stpierre@dfci.harvard.edu

THE JOURNAL OF EXPERIMENTAL BIOLOGY