Dispatch R194

Heather L. Eisthen1 andRufus Isaacs2

Animals and plants in terrestrialand aquatic environments use anarray of chemical defences toavoid being eaten [1–3].Chemicals may be effectivebecause they are toxic, like thecantharidins released by blisterbeetles, or because the predatorperceives them as noxious, thewell-known example being skunkodours. In these interactions, thepotential predator suffers a directnegative impact. Increasingly,more complex modes ofchemical defence are beingdiscovered that reduce predationthrough indirect mechanisms. Inthis issue of Current Biology,Kicklighter et al. [4] report

evidence for a new type ofindirect defence in which the seahare Aplysia releases chemicalsthat stimulate the predator toengage in other behaviours,diverting it from successfulpredation of the prey.

Aplysia faces a number ofpotential predators, includinganemones, teleost fishes, crabsand lobsters. When threatened orattacked, an Aplysia will turn awayfrom the apparent direction ofattack, withdraw its mantle and, attimes, squirt clouds of secretionsfrom ink and opaline glands as ittries to escape (Figure 1). Neuralcircuits, ionic currents,neurotransmitters andneuromodulators involved in inkand opaline release have been thesubject of considerable

investigation [5,6]. The origin andmetabolic pathway involved inpigment production, as well as thecell types involved in pigmentstorage and release, have beendescribed [7,8]. Thus, within thefield of chemical ecology, Aplysiaserves as an excellent model forunderstanding the cellular andmolecular mechanisms underlyingchemical defence againstpredation.

The paper by Kicklighter et al.[4] expands our understanding ofAplysia’s chemical defences byexamining the effects of ink andopaline on one predator, the spinylobster Panulirus. Although Aplysiaink is an effective deterrent againstthe predatory anemoneAnthopleura xanthogrammica [9],Kicklighter et al. [4] found thatremoving the ink gland had noeffect on the rate of successfulpredation by lobsters. Removingthe opaline gland, however,dramatically decreased theprobability that an Aplysia wouldescape during an encounter with alobster.

Behavioural observationsindicated that secretions fromboth ink and opaline glands causelobsters to groom theirantennules and mouthparts,suggesting that the viscoussecretions may interfere with thesensory hairs on the lobsters’appendages. In addition,secretions from the two glandshad distinct effects on thelobsters’ behaviour: ink glandsecretions elicited behavioursthat are generally associated withforaging and feeding in lobsters,such as digging in the substrateand moving the forelegs to themouthparts, whereas opalinegland secretions sometimes ledlobsters to try to escape bytailflipping.

Decades of research usingspiny lobsters as model animalsfor understanding transductionand coding in chemosensory

When threatened, sea hares secrete ink and opaline. This mixturehas now been shown to act on peripheral chemosensory neuronsof spiny lobsters, stimulating feeding-related behaviours as adeceptive method of avoiding predation.

Figure 1. Defensive secretion deterring lobster attack.

Captured video frame of a sea hare releasing defensive secretions while being attackedby a spiny lobster. (Image courtesy of Paul M. Johnson.)

Dispatches

Integrative Biology: Sea Hares Saved by a Delicious Distraction

systems [10] have establishedthat the chemosensory neuronsin Panulirus are extremelysensitive to amino acids, ureaand ammonium, chemicals thatare all associated with foraging.Kicklighter et al. [4] found thatthese substances are abundantin both opaline and inksecretions: opaline contains highlevels of taurine, a powerfulfeeding stimulant for lobsters,and ink contains large amountsof ammonium as well asmoderate amounts of taurine.Consistent with the view thattaurine stimulates feeding, theauthors found that artificialmixtures of the majorcomponents of ink and opalinestimulated grabbing andingestive behaviours by lobsters.Interestingly, natural opaline didnot, suggesting that it contains afeeding deterrent in addition tothe components described here.

Finally, Kicklighter et al. [4]recorded from singlechemosensory receptor neuronsin the lateral antennule andsecond maxilliped to examineresponses to ink and opaline, aswell as individual componentsand mixtures of the components.Analysis of the across-neuronpattern of activity in response tonatural ink and opaline as well asthe mixtures suggests that thesestimuli evoked responses similarto those of a typical food item,shrimp. The authors interprettheir results as indicating thatAplysia use a combination ofchemical defence mechanisms,including an aversive secretionand secretions that foul thelobsters’ antennules andmouthparts. More significantly,they propose a previouslyundescribed defence mechanism:Aplysia produce a cloud ofviscous chemicals that to alobster smells like food, andwhile the lobster is distractedtrying to eat the phantom prey,the Aplysia has an opportunity toescape.

Although much progress hasbeen made in describing themolecular mechanisms ofolfaction in the last 10 years, ourunderstanding of the ways theolfactory system functions toprocess biologically relevant

information in a naturalisticcontext is still in its infancy. Manystudies of electrophysiologicalresponses to odorants use singlecompounds, generally off-the-shelf chemicals that have noinherent biological significance forthe organism being examined.Further, in many studies usingmixtures, the components areselected based solely on theirability to evoke responses from alarge subset of neurons. Oneproblem with this approach hasbeen made clear by a study withtiger salamanders [11] whichshowed that an odorant can elicitstrong neural responses from theolfactory system even though theanimal does not appear toperceive the chemical.

The use of artificial odorants iswidespread partly because itfacilitates stimulus control andreplicability, but also becausethere are few examples in whichthe components and ratios ofbehaviourally relevant mixtureshave been identified. Thus, thework by Kicklighter et al. [4]provides an important set of dataabout the chemical constituentsof two stimulus mixtures, ink andopaline, that evoke well-describedbehavioural responses fromlobsters. Future studies examiningneural responses to thesemixtures should contribute greatlyto our understanding of the linksbetween olfactory system functionand behaviour.

On a broader scale,understanding the stimuli andneural mechanisms underlyinginteractions between predatorsand prey has proved a dauntingtask, as it requires detailedinvestigation of two or morespecies, and has been achievedin only a handful of cases.Perhaps the best example comesfrom research on theecholocation signals used byforaging bats, as well as themeans by which noctuid mothsdetect these cues and use themto avoid predation [12]. Within thechemical senses, investigationsinto the tritrophic interactionsamong plants, caterpillars andpredatory wasps [13] haveidentified the chemical signal incaterpillar saliva that stimulatesplants to release odours in

response to caterpillar feeding[14,15]. These odor blendsstimulate upwind flight byparasitic wasps which then locateand attack the herbivore [16],thus reducing the degree of injuryto the plant. Such indirectdefence mechanisms suggestcomplex evolutionary originswhich may be found to convergeacross systems, once furtherinvestigations unravel the basis ofsuch interactions.

The new work by Kicklighteret al. [4], combined with previousstudies of Aplysia, providesinsight into the mechanismsinvolved in the production,release, processing, andbehavioural response to defensivesecretions used by Aplysia toavoid predation by Panulirus, andthus has the potential to serve asa new model for examining theneuroethology of predator-preyinteractions.

References1. Blum, M.S. (1981). Chemical Defenses of

Arthropods (New York: Academic Press).2. Hay, M.E., and Fenical, W. (1988). Marine

plant-herbivore interactions: the ecologyof chemical defense. Ann. Rev. Ecol.Syst. 19, 111-145.

3. Eisner, T., and Meinwald, J. eds. (1995).Chemical Ecology: The Chemistry ofBiotic Interaction (Washington, DC: Natl.Acad. Press).

4. Kicklighter, C.E., Shabani, S., Johnson,P.M., and Derby, C.D. (2005). Sea haresuse novel antipredatory chemicaldefenses. Curr. Biol. 15, this issue.

5. Byrne, J.H. (1980). Neural circuit forinking behavior in Aplysia californica. J.Neurophysiol. 43, 896-911.

6. Tritt, S.H., and Byrne, J.H. (1980). Motorcontrols of opaline secretion in Aplysiacalifornica. J. Neurophysiol. 43, 581-594.

7. Coelho, L., Prince, J., and Nolen, T.G.(1998). Processing of defensive pigmentin Aplysia californica: Acquisition,modification and mobilization of the redalgal pigment r-phycoerythrin by thedigestive gland. J. Exp. Biol. 201,425–438.

8. Prince, J., Nolen, T.G., and Coelho, L.(1998). Defensive ink pigment processingand secretion in Aplysia californica:concentration and storage ofphycoerythrobilin in the ink gland. J.Exp. Biol. 201, 1595-1613.

9. Nolen, T.G., Johnson, P.M., Kicklighter,C.E., and Capo, T. (1995). Ink secretionby the marine snail Aplysia californicaenhances its ability to escape from anatural predator. J. Comp. Physiol. A176, 239–254.

10. Ache, B.W. (2002). Crustaceans asanimal models for olfactory research. InCrustacean Experimental Systems inNeurobiology, K. Wiese, ed. (Berlin:Springer), pp. 189-199.

11. Dorries, K.M., White, J., and Kauer, J.S.(1997). Rapid classical conditioning ofodor response in a physiological modelfor olfactory research, the tigersalamander. Chem. Senses 22, 277-286.

12. Fullard, J.H. (1998). Sensory coevolution

Current Biology Vol 15 No 6R195

Current Biology Vol 15 No 6R196

of moths and bats. In ComparativeHearing: Insects. Springer Handbook ofAuditory Research, R.R. Hoy, A.N.Popper and R.R. Fay, eds. (New York:Springer), pp. 279-326.

13. Paré, P.W., and Tumlinson, J.H. (1999).Plant volatiles as a defense againstinsect herbivores. Plant Physiol. 121,325-332.

14. Turlings, T.C.J., Alborn, H.T., Loughrin,J.H., and Tumlinson, J.H. (2000).

Volicitin, an elicitor of maize volatiles inthe oral secretion of Spodoptera exigua:its isolation and bio-activity. J. Chem.Ecol. 26, 189-202.

15. Rose, U.S.R., and Tumlinson, J.H. (2004).Volatiles released from cotton plants inresponse to Helicoverpa zea feedingdamage on cotton flower buds. Planta218, 824–832.

16. Rose, U.S.R., Lewis, W.J., andTumlinson, J.H. (1998). Specificity of

systemically released cotton volatiles asattractants for specialist and generalistparasitic wasps. J. Chem. Ecol. 24,303–319.

Departments of Zoology1 andEntomology2, Michigan State University,East Lansing, Michigan 48824, USA.

DOI: 10.1016/j.cub.2005.03.006

J. Allan Downie

Only bacteria contain thenitrogenase enzyme that canreduce N2 to ammonium and so,during nitrogen-limited growth,some plants enter into asymbiotic interaction withnitrogen-fixing bacteria, whichprovide the plants withammonium. A major problem inmaintaining a high rate ofnitrogen fixation, however, is thatthe bacterial nitrogenaseenzymes are very oxygensensitive, but at the same timerequire high levels of ATP to drivethe reaction. So ideally, thebacteria require a high flux ofoxygen to enable high rates ofATP synthesis, whilstsimultaneously maintaining a lowfree oxygen environment toprevent inactivation ofnitrogenase by oxygen. Theseparadoxical requirements are metby the formation of root nodulesin which legumes provide anappropriate niche for rhizobia, thebacteria that differentiate intoforms, known as bacteroids,which fix nitrogen in nodules. Acrucial part of this niche is thepresence of plant haemoglobinsin the cytoplasm of the plant cellscontaining the bacteroids.

If you dig up the roots of alegume such as pea or bean andcut into one of the nodules on theroot you will see that it has ablood-red colour (Figure 1, top).

This is due to the high levels ofhaemoglobins, referred to asleghaemoglobins because theyare always found in legumenodules [1]. As they report in thisissue of Current Biology, Ott et al.[2], using an RNA interference(RNAi) approach to silence theexpression of the three nodule-expressed leghaemoglobin genesin the legume Lotus japonicus,have now demonstrated thatleghaemoglobins really areessential for symbiotic nitrogenfixation in legume root nodules.

Whereas animal haemoglobinsin blood facilitate oxygen transferbetween cells and organs,leghaemoglobins function in amanner more analogous toanimal myoglobin [1], whichfacilitates oxygen transfer withinthe cytoplasm to mitochondria.Leghaemoglobins, however, canhave a twenty-fold higher affinityfor oxygen than myoglobin [3].The oxygen-bindingcharacteristics ofleghaemoglobins are unusual inthat they have an extremely fastO2 association rate and arelatively slow O2 dissociationrate [1], and so can buffer thefree oxygen concentration ataround 7–11 nM.

Calculations based on the levelsof oxygenation and concentrationof leghaemoglobin in the nodulecytoplasm suggest that theconcentration of leghaemoglobin-bound oxygen is around 70,000

times higher than the free oxygenconcentration [1]. This provides asubstantial buffering capacity thatwill be important for providing ahigh flux of oxygen for bacterialrespiration. But the low levels offree-oxygen pose a challenge forbacterial respiration, and thenitrogen-fixing bacteroids dealwith this by inducing a symbiosis-specific cytochrome oxidase witha very high affinity for oxygen [4].

Given these observations on thebiochemistry and physiology ofnitrogen fixation in nodules, it hadbeen anticipated that silencing ofleghaemoglobin expression wouldaffect symbiotic nitrogen fixation,as observed by Ott et al. [2]. Whatwas not anticipated, however,was the absence of bacterialnitrogenase in the bacteria withinthe nodules of the plants lackingleghaemoglobin. Although themeasured levels of free oxygen innodules of the plants lackingleghaemoglobin were somewhathigher than those seen innitrogen-fixing nodules, there wasstill a low oxygen environment,particularly in the deeper layers ofthe nodules [2]. So it seemsunlikely that the complete lack ofnitrogenase expression cansimply be explained by the lack ofa low oxygen environment, whichcould be observed in at leastsome parts of the nodules.

How then can we explain thelack of induction of bacteroidnitrogenase in these nodules?Perhaps some form of rampedinduction could be required.Possibly a rapid induction of thebacteroid nitrogenase and high-affinity oxidase might cause aproblem if it resulted in a high rateof oxygen consumption that couldnot be sustained in the absence ofthe leghaemoglobin-oxygen buffer.Conversely, if the leghaemoglobinwas induced before thespecialised bacteroid oxidase then

Legume Haemoglobins: Symbiotic Nitrogen FixationNeeds Bloody Nodules

How do plants create an environment in which symbiotic bacteriacan reduce enough N2 to provide the plant with sufficientammonium for growth? Gene silencing has now been used to showthat legume haemoglobins are crucial.