DOI: 10.1126/science.1243339, 544 (2014);343 Science

et al.Christi M. GendronPerception and Reward

Life Span and Physiology Are Modulated by SexualDrosophila

This copy is for your personal, non-commercial use only.

clicking here.colleagues, clients, or customers by , you can order high-quality copies for yourIf you wish to distribute this article to others

here.following the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

): March 17, 2014 www.sciencemag.org (this information is current as of

The following resources related to this article are available online at

http://www.sciencemag.org/content/343/6170/544.full.htmlversion of this article at:

including high-resolution figures, can be found in the onlineUpdated information and services,

http://www.sciencemag.org/content/suppl/2013/11/27/science.1243339.DC1.html can be found at: Supporting Online Material

http://www.sciencemag.org/content/343/6170/544.full.html#relatedfound at:

can berelated to this article A list of selected additional articles on the Science Web sites

http://www.sciencemag.org/content/343/6170/544.full.html#ref-list-1, 4 of which can be accessed free:cites 27 articlesThis article

http://www.sciencemag.org/content/343/6170/544.full.html#related-urls1 articles hosted by HighWire Press; see:cited by This article has been

http://www.sciencemag.org/cgi/collection/neuroscienceNeuroscience

subject collections:This article appears in the following

registered trademark of AAAS. is aScience2014 by the American Association for the Advancement of Science; all rights reserved. The title

CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theScience

on

Mar

ch 1

7, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

7, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

7, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

7, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

7, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

7, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

http://oascentral.sciencemag.org/RealMedia/ads/click_lx.ads/sciencemag/cgi/reprint/L22/968072652/Top1/AAAS/PDF-R-and-D-Systems-Science-140201/FENS-trave-grant.raw/1?xhttp://www.sciencemag.org/about/permissions.dtlhttp://www.sciencemag.org/about/permissions.dtlhttp://www.sciencemag.org/about/permissions.dtlhttp://www.sciencemag.org/about/permissions.dtlhttp://www.sciencemag.org/content/343/6170/544.full.htmlhttp://www.sciencemag.org/content/343/6170/544.full.htmlhttp://www.sciencemag.org/content/suppl/2013/11/27/science.1243339.DC1.html http://www.sciencemag.org/content/343/6170/544.full.html#relatedhttp://www.sciencemag.org/content/343/6170/544.full.html#relatedhttp://www.sciencemag.org/content/343/6170/544.full.html#ref-list-1http://www.sciencemag.org/content/343/6170/544.full.html#ref-list-1http://www.sciencemag.org/content/343/6170/544.full.html#related-urlshttp://www.sciencemag.org/content/343/6170/544.full.html#related-urlshttp://www.sciencemag.org/cgi/collection/neurosciencehttp://www.sciencemag.org/cgi/collection/neurosciencehttp://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/

evolutionary advantage of preserving limited re-sources for the offspring (30) or preventing com-petition from other males.

References and Notes1. D. Gems, D. L. Riddle, Nature 379, 723–725 (1996).2. T. Chapman, L. F. Liddle, J. M. Kalb, M. F. Wolfner,

L. Partridge, Nature 373, 241–244 (1995).3. L. A. Herndon et al., Nature 419, 808–814 (2002).4. C. F. Glenn et al., J. Gerontol. A Biol. Sci. Med. Sci. 59,

1251–1260 (2004).5. D. K. Chow, C. F. Glenn, J. L. Johnston, I. G. Goldberg,

C. A. Wolkow, Exp. Gerontol. 41, 252–260 (2006).6. M. D. McGee et al., Aging Cell 10, 699–710

(2011).7. J. Hodgkin, T. M. Barnes, Proc. Biol. Sci. 246, 19–24

(1991).8. K. C. Kiontke et al., BMC Evol. Biol. 11, 339 (2011).9. S. B. Pierce et al., Genes Dev. 15, 672–686

(2001).10. E. J. Tisdale, J. R. Bourne, R. Khosravi-Far, C. J. Der,

W. E. Balch, J. Cell Biol. 119, 749–761 (1992).11. T. J. Maures, E. L. Greer, A. G. Hauswirth, A. Brunet,

Aging Cell 10, 980–990 (2011).12. C. Jin et al., Cell Metab. 14, 161–172 (2011).

13. R. A. Butcher, M. Fujita, F. C. Schroeder, J. Clardy,Nat. Chem. Biol. 3, 420–422 (2007).

14. P. Y. Jeong et al., Nature 433, 541–545 (2005).15. J. Srinivasan et al., Nature 454, 1115–1118

(2008).16. P. T. McGrath et al., Nature 477, 321–325 (2011).17. E. Z. Macosko et al., Nature 458, 1171–1175

(2009).18. A. H. Ludewig et al., Proc. Natl. Acad. Sci. U.S.A. 110,

5522–5527 (2013).19. Y. Izrayelit et al., ACS Chem. Biol. 7, 1321–1325

(2012).20. L. A. Perkins, E. M. Hedgecock, J. N. Thomson,

J. G. Culotti, Dev. Biol. 117, 456–487 (1986).21. B. van Swinderen, L. B. Metz, L. D. Shebester,

C. M. Crowder, Genetics 161, 109–119 (2002).22. A. D. Cutter, Mol. Biol. Evol. 25, 778–786 (2008).23. D. Gems, D. L. Riddle, Genetics 154, 1597–1610

(2000).24. J. Apfeld, C. Kenyon, Nature 402, 804–809 (1999).25. J. Alcedo, C. Kenyon, Neuron 41, 45–55 (2004).26. S. Libert et al., Science 315, 1133–1137 (2007).27. C. I. Bargmann, H. R. Horvitz, Science 251, 1243–1246

(1991).28. A. H. Ludewig, F. C. Schroeder, WormBook 2013,

1–22 (2013).

29. P. Ren et al., Science 274, 1389–1391 (1996).30. T. B. Kirkwood, Nature 270, 301–304 (1977).

Acknowledgments: We thank members of the Brunetlaboratory, S. Kim, A. Fire, and A. Villeneuve for helpfulsuggestions and J. Lim and S. Zimmerman for critical readingof the manuscript. We thank N. Kosovilka and the Protein andNucleic Acid Facility facility for the microarray experimentsand T. Stiernagle from the Caenorhabditis Genetics Center.Supported by R01AG031198, DP1AG044848, the GlennFoundation for Medical Research (A.B.), postdoctoral fellowshipF32AG37254 (T.J.M.), T32HG000044 and the Helen HayWhitney Foundation (L.N.B.), Stanford Dean’s Fellowship (BAB),R01GM088290 (F.C.S.), and T32GM008500 (Y.I.).

Supplementary Materialswww.sciencemag.org/content/343/6170/541/suppl/DC1Materials and MethodsFigs. S1 to S3Tables S1 and S2Movies S1 to S6References (31–37)

2 August 2013; accepted 11 November 2013Published online 28 November 2013;10.1126/science.1244160

Drosophila Life Span and PhysiologyAre Modulated by SexualPerception and RewardChristi M. Gendron,1* Tsung-Han Kuo,2* Zachary M. Harvanek,1,3 Brian Y. Chung,1

Joanne Y. Yew,4,5 Herman A. Dierick,2 Scott D. Pletcher1

Sensory perception can modulate aging and physiology across taxa. We found that perceptionof female sexual pheromones through a specific gustatory receptor expressed in a subsetof foreleg neurons in male fruit flies, Drosophila melanogaster, rapidly and reversiblydecreases fat stores, reduces resistance to starvation, and limits life span. Neurons thatexpress the reward-mediating neuropeptide F are also required for pheromone effects.High-throughput whole-genome RNA sequencing experiments revealed a set of molecularprocesses that were affected by the activity of the longevity circuit, thereby identifyingnew candidate cell-nonautonomous aging mechanisms. Mating reversed the effects ofpheromone perception; therefore, life span may be modulated through the integrated actionof sensory and reward circuits, and healthy aging may be compromised when the expectationsdefined by sensory perception are discordant with ensuing experience.

Sensory perception can modulate agingand physiology in multiple species (1–6).In Drosophila, exposure to food-basedodorants partially reverses the anti-aging effectof dietary restriction, whereas broad reduction inolfactory function promotes longevity and altersfat metabolism (2, 4). Even the well-known rela-

tion between body temperature and life span mayhave a sensory component (7, 8).

To identify sensory cues and neuronal cir-cuitry that underlie the effects of sensory percep-tion on aging, we focused on the perception ofpotential mates. Social interactions are prevalentthroughout nature, and the influence of socialcontext on health and longevity is well knownin several species, including humans (9). Suchinfluences include behavioral interactions withmates and broader physiological “costs of repro-duction,” which often form the basis for evolu-tionary models of aging (10, 11).

In Drosophila, the presence of potential matesis perceived largely through nonvolatile cutic-ular hydrocarbons, which are produced by cellscalled oenocytes and are secreted to the cutic-ular surface, where they function as pheromones(12, 13). To test whether differential pheromone

exposure influenced life span or physiology, wehoused “experimental” flies of the same geno-type with “donor” animals of the same sex thateither expressed normal pheromone profiles orwere genetically engineered to express phero-mone profiles characteristic of the opposite sex(Fig. 1A). Donor males with feminized pheromoneprofiles were generated by targeting expressionof the sex determination gene, transformer, to theoenocytes [via OK72-GAL4 or Prom-E800-GAL4(14) (fig. S1)], whereas masculinization of femaleflies was accomplished by expressing tra-RNAi ina similar way (15). This design allowed manipu-lation of the experimental animals’perceived sexualenvironment without introducing complications as-sociated with mating itself.

In Drosophila, sensory manipulations can af-fect life span, fat storage [as determined by base-line measures of triacylglyceride (TAG)], andcertain aspects of stress resistance (2, 4). Wefound that flies exposed to pheromones of theopposite sex showed differences in these pheno-types. Experimental male flies exposed to maledonor pheromone had higher amounts of TAG,were substantially more resistant to starvation,and exhibited a significantly longer life span thangenetically identical male siblings exposed to fe-male donor pheromone (Fig. 1, B to D). Femalesexhibited similar phenotypes in response to maledonor pheromone, but the magnitude of the ef-fects was smaller (fig. S2). Subsequent experimentswere therefore focused on males.

The characteristics of pheromone exposurewere indicative of a mechanism involving sen-sory perception. Effects were similar in severalgenetic backgrounds, including a strain recent-ly collected in the wild (fig. S3), and were large-ly unaffected by cohort composition (fig. S4).Pheromone-induced phenotypes were detectedafter as little as 2 days’ exposure to donor ani-mals (Fig. 1, B and C), persisted with longer ma-nipulations (Fig. 1D), and were progressively

1Department of Molecular and Integrative Physiology andGeriatrics Center, Biomedical Sciences and Research Build-ing, University of Michigan, Ann Arbor, MI 48109, USA. 2De-partment of Molecular and Human Genetics, Baylor Collegeof Medicine, Houston, TX 77030, USA. 3Medical ScientistTraining Program, Taubman Medical Library, University ofMichigan, Ann Arbor, MI 48109, USA. 4Temasek Life SciencesLaboratory, National University of Singapore, Singapore 117604.5Department of Biological Sciences, National University ofSingapore, Singapore 117543.

*These authors contributed equally to this work.†Corresponding author. E-mail: spletch@umich.edu

31 JANUARY 2014 VOL 343 SCIENCE www.sciencemag.org544

REPORTS

reversed when female donor pheromone was re-moved (Fig. 1, E and F, and fig. S5). Pheromoneeffects appeared not to be mediated by aberrantor aggressive interactions with donor flies, be-cause (i) we did not observe significant differencesin such behaviors and (ii) continuous, vigorousagitation of the vials throughout the exposureperiod, which effectively disrupted observed be-haviors, had no effect on the impact of donorpheromone (fig. S6). Furthermore, exposure ofexperimental males to the purified female pher-omone 7,11-heptacosadiene (7,11-HD) producedphysiological changes in the absence of donoranimals (mean survival time during starvation,51.1 T 1.7 hours and 45.4 T 1.2 hours for controland 7,11-HD exposure, respectively; P = 0.007,log-rank test).

To explore the sensorymodality throughwhichdonor pheromone exerts its effects,we testedwhetherthe broadly expressed olfactory co-receptorOr83b,whose loss of function renders flies largely un-able to smell (16), was required for pheromoneeffects.Or83bmutant flies and controls exhibitedsimilar changes in starvation resistance (fig. S7)in response to donor pheromone, indicating thatolfaction was not required. To test whether taste

perception was involved, we used flies mutantfor the gene Pox neuro (Poxn), a null mutationthat putatively transforms all chemosensory neu-rons into mechanosensory neurons. Drosophilataste neurons are present in the mouthparts anddistributed on different body parts, including thewings, legs, and genitals, which allow sensationby contact.When thePoxn nullmutation is coupledwith a partially rescuing transgene, PoxnDM22-B5-DXB, flies are generally healthy but gustatoryperception is eliminated in the labelum, the legs,and the wing margins (17). PoxnDM22-B5-DXB fliesshowed no pheromone-induced changes in star-vation resistance, TAG amounts, or life span (Fig. 2,A to C). However, the responses of Poxn mutantflies that carried a transgene that restores tastefunction to the legs and wing margins [but notlabelum;PoxnDM22-B5-Full1 (17)] were similar tothose of control flies (Fig. 2, A to C). Thus, theeffects of pheromone exposure appear to be me-diated by taste perception through gustatory neu-rons outside of the mouthparts.

To identify specific gustatory receptors andneurons that might mediate the pheromone ef-fects, we tested candidate pheromone receptors.Of the mutants that we examined, only flies

that carried a loss-of-function mutation in thegene pickpocket 23 (ppk23) were resistant to theeffects of pheromone exposure (fig. S8). Furtheranalysis verified that ppk23 was required for theeffects of pheromone exposure on starvation re-sistance, TAG amounts, and life span (Fig. 2, Dto F). Silencing ppk23-expressing neurons onlyduring exposure to donor males by expressing atemperature-sensitive dominant-negative allele ofthe dynamin gene shibire (via ppk23-GAL4;UAS-shits) also eliminated the differential response topheromones (Fig. 3A). In male Drosophila, thetranscription factor fruitless (fru) is expressed withppk23 in pheromone-sensing neurons located in theanimals’ forelegs (18), and silencing fru-expressingneurons during exposure (via fru-GAL4;UAS-shits)abrogated pheromone effects (Fig. 3B). Consistentwith a requirement for these neurons, we foundthat surgical amputation of the forelegs, but notinjury alone, was sufficient to reproducibly elim-inate the effects of pheromone exposure (Fig. 3Cand fig. S9). Moreover, acute targeted activa-tion of ppk23-expressing neurons by means ofa temperature-sensitive TRPA1 channel (ppk23-GAL4;UAS-TRPA1) was sufficient to mimic theeffects of female pheromone without exposure

P

Control

ppk23 > shi ts

Female

FemaleMale

Male

Control

fru > shi ts

Female

FemaleMale

Male

Sur

vivo

rshi

p

Sur

vivo

rshi

p

Sur

vivo

rshi

p

Starvation Time (Hours) Starvation Time (Hours) Starvation Time (Hours)

Sur

vivo

rshi

p

Starvation Time (Hours)

Sur

vivo

rshi

p

Age (Days)

Control

Foreleg Removed

Female

FemaleMale

Male

BA C

FD E

ppk2

3>

T

RPA1

Cont

rol

ppk23>TRPA1Control

ppk23>TRPA1Control

P=0.002P

(Fig. 3, D to F). Together, these data indicate thatpheromone-sensing neurons in the foreleg of themale fly that express the gustatory receptor ppk23and the transcription factor fruitless influencestress resistance, physiology, and life span inresponse to perception of female pheromones.

To examine brain circuits that may functionin transducing pheromone perception, we selec-tively expressed UAS-shits to block synaptic trans-mission in various neuroanatomical regions withthe goal of disrupting the physiological effects ofdonor pheromone exposure. The effects were ab-rogated whenUAS-shits was driven in neurons char-acterized by expression of neuropeptide F (NPF,as represented by npf-GAL4) (fig. S10). Furtheranalysis verified that pheromone-induced changesin starvation resistance and TAG abundance werelost after silencing of npf-expressing neurons(Fig. 4A). Consistent with a possible role in trans-ducing pheromone information, npf expressionwas significantly increased by 30% in experimen-tal males after exposure to feminized donormales(fig. S11), and activation of npf-expressing neu-rons was sufficient to decrease life span in theabsence of pheromone exposure (Fig. 4B).

NPF may function as a mediator of sexualreward in Drosophila (19), and its mammalian

counterpart, neuropeptide Y (NPY), has beenassociated with sexual motivation and psycho-logical reward (20, 21). We tested whether theeffects of pheromone perception might be res-cued by allowing males to successfully mate withfemales. Neither a small number of conjugalvisits with virgin females nor housing with wild-type females in a 1:1 ratio was sufficient to ame-liorate the effects of pheromone exposure (fig.S12). In this context, decreased longevity may bea consequence of pheromone perception and notof mating itself. MaleDrosophila are willing andable to copulate up to five times in rapid suc-cession before requiring a refractory period (22).We found that supplementing donor cohorts withan excess of mating females (in a 5:1 ratio) wassufficient to significantly reduce the effects onmortality and TAG caused by female donor pher-omone early in life (Fig. 4C and fig. S13). Thebenefits of mating on age-specific mortality de-creased with age, which suggests that aging mayreduce mating efficiency or may diminish effec-tive mating reward.

To identify how sexual perception and rewardmay alter physiological responses in peripheraltissues, we usedwhole-genomeRNA sequencing(RNA-seq) technology to examine changes in gene

expression. We found 195 genes with significant-ly different expression (using an experiment-wise error rate of 0.05) in control male flies thatwere exposed to feminized or control donormales for 48 hours. Nearly all (188/195 = 96%)of the changes appeared to be due to pheromoneperception, because they were not observed inidentical experiments using ppk23 mutant flies(table S1). Males exposed to female pheromonesdecreased the transcription of genes encodingodorant-binding proteins and increased the tran-scription of several genes with lipase activity(Fig. 4D). A significant enrichment was observedin secreted molecules, which includes genes en-coding proteins that mediate immune and stressresponses. Many of these genes and pathwayswere highlighted in a recent meta-analysis of geneexpression changes in response to stress andaging (23).

The activities of insulin and target of rapamycin(TOR) signaling, which modulate aging acrosstaxa, increase sexual attractiveness in flies (24).Our demonstration that perception of sexual char-acteristics is sufficient to modulate life span andphysiology suggests that aging pathways in oneindividual may modulate health and life span inanother (fig. S14). These types of indirect genetic

Control

npf > shi ts

Female

FemaleMale

Male

Ln(M

orta

lity

Rat

e)

Age (Days)

A B C

D

Feminized Male

Fem. Male+FemaleCont. Male+Female

Control Male

Donor Flies

P=0.02

Sur

vivo

rshi

p

Starvation Time (Hours)

−1 0 0.5 1 1.5Fold-Change

Obp

9bO

r83b

Obp

83a

Obp

28a

lush

Obp

83a

Odorant/Pheromone Binding

a5Ls

p2

Obp

99b

lush

IM4

jeb

CG

1330

9

Cht

4

Acp

76a

CH

3422

7

Ag5

r

IM1,

IM2

CG

7017

Obp

83a

Obp

83a

Acp

36D

E

PG

RP

-SB

1

TotC

upd3

Obp

28a

Lsp1

β

Est

-6*

Est

-6*

IM1,

IM2

IM23 Dpt

Npl

p3

Mur

18b

CG

1582

8Im

pL2

verm

iota

Tryu

serp

Dpt

B

Dro

Extracellular/Secreted Molecules

OR

MD

L

Pep

ckC

G10

924

Glycolysis/Gluconeogenesis

IM4

PG

RP

-SB

1

IM1,

IM2

upd3

IM23 Drofon

IM1,

IM2

TotC Dpt

Dpt

B

Immune Response

CG

8773

CG

8773

CG

3134

3

CG

5527

CG

9505

CG

3119

8C

G58

49

CG

1328

3

CG

7631

CG

1525

5

Metallopeptidase Activity

CG

1108

9

ade5

ade3

IMP Biosynthesis

Ugt

86D

e*U

gt86

De*

CG

9363

Ugt

86D

e*

Drug Metabolism

Sur

vivo

rshi

p

Age (Days)

UAS-TRPA1

npf>TRPA1npf-GAL4

CG

1035

7

CG

5966

CG

2772

CG

6296

Lipase Activity

P=0.50

P=0.001

P=0.005

Fig. 4. Aging and physiology are modulated by neuralmechanisms of expectation and reward. (A) Inhibition ofnpf-expressing neurons abrogates differences in starvationcaused by pheromone exposure. N = 43 and 45 for control(npf-GAL4 only) flies exposed to male or female donor phero-mones, respectively (P = 0.005, log-rank test). N = 48 and 47for treatment (npf-GAL4; UAS-shits) flies exposed to male orfemale donor pheromones, respectively (P = 0.50 by log-ranktest). Inset is as described in Fig. 2. (B) Activation of npf-expressing neurons causes decreased longevity in the absenceof pheromone exposure. npf-GAL4;uas-dTRPA1 males (N = 239)exhibit significantly shorter life span relative to UAS-dTRPA1 only(N = 235; P ≤ 0.001, log-rank test) and npf-GAL4 only (N = 179;P ≤ 0.001, log-rank test) male transgene controls. (C) Mortalityrates are reduced when males exposed to female donorpheromone (dashed black line) are given access to excessfemales (dashed red line; P = 0.02 through 20 days of age byAalen regression). Cohorts consisted of five experimentalmales together with (i) 30 control donor males (solid black),(ii) 30 feminized donor males (dashed black), (iii) 5 feminizeddonor males + 25 females (dashed red), or (iv) 5 control donor males + 25females (solid red); 20 replicate cohorts, totaling 100 experimental flies,were measured for each treatment. (D) Significantly enriched Gene Ontology

pathways and functions whose genes are differentially regulated after pher-omone exposure. See table S1 for a complete list of genes with significantchanges in expression.

www.sciencemag.org SCIENCE VOL 343 31 JANUARY 2014 547

REPORTS

effects have the potential to be influential agentsof natural selection (25). Imbalances of expec-tation and reward may therefore have broadeffects on health and physiology in humans andmay represent a powerful evolutionary force innature.

References and Notes1. J. Apfeld, C. Kenyon, Nature 402, 804–809

(1999).2. S. Libert et al., Science 315, 1133–1137 (2007).3. N. J. Linford, T. H. Kuo, T. P. Chan, S. D. Pletcher,

Annu. Rev. Cell Dev. Biol. 27, 759–785 (2011).4. P. C. Poon, T. H. Kuo, N. J. Linford, G. Roman,

S. D. Pletcher, PLOS Biol. 8, e1000356 (2010).5. E. D. Smith et al., BMC Dev. Biol. 8, 49 (2008).6. J. Alcedo, C. Kenyon, Neuron 41, 45–55 (2004).7. S. J. Lee, C. Kenyon, Curr. Biol. 19, 715–722

(2009).8. R. Xiao et al., Cell 152, 806–817 (2013).9. R. M. Sapolsky, Science 308, 648–652 (2005).

10. L. Partridge, N. H. Barton, Nature 362, 305–311(1993).

11. L. Partridge, D. Gems, D. J. Withers, Cell 120, 461–472(2005).

12. J. C. Billeter, J. Atallah, J. J. Krupp, J. G. Millar,J. D. Levine, Nature 461, 987–991 (2009).

13. J. F. Ferveur, Behav. Genet. 35, 279–295 (2005).14. J. F. Ferveur et al., Science 276, 1555–1558 (1997).

15. M. P. Fernández et al., PLOS Biol. 8, e1000541(2010).

16. M. C. Larsson et al., Neuron 43, 703–714 (2004).17. W. Boll, M. Noll, Development 129, 5667–5681

(2002).18. R. Thistle, P. Cameron, A. Ghorayshi, L. Dennison,

K. Scott, Cell 149, 1140–1151 (2012).19. G. Shohat-Ophir, K. R. Kaun, R. Azanchi,

H. Mohammed, U. Heberlein, Science 335,1351–1355 (2012).

20. S. P. Kalra, J. T. Clark, A. Sahu, M. G. Dube, P. S. Kalra,Synapse 2, 254–257 (1988).

21. M. Heilig, Neuropeptides 38, 213–224 (2004).22. M. Ashburner, K. Golic, R. S. Hawley, Drosophila:

A Laboratory Handbook (Cold Spring Harbor LaboratoryPress, Cold Spring Harbor, NY, ed. 2, 2004).

23. G. Landis, J. Shen, J. Tower, Aging 4, 768–789(2012).

24. T. H. Kuo et al., PLOS Genet. 8, e1002684 (2012).25. J. W. McGlothlin, A. J. Moore, J. B. Wolf, E. D. Brodie 3rd,

Evolution 64, 2558–2574 (2010).

Acknowledgments: We thank the members of the Pletcherlaboratory for Drosophila husbandry, N. Linford for commentson the revision, P. J. Lee for figure illustration, and membersof the Dierick and Pletcher laboratories for suggestions onexperiments and comments on the manuscript. Supportedby NIH grants R01AG030593, TR01AG043972, andR01AG023166, the Glenn Foundation, the AmericanFederation for Aging Research, and the Ellison MedicalFoundation (S.D.P.); Ruth L. Kirschstein National Research

Service Award F32AG042253 from the National Institute onAging (B.Y.C.); NIH grant T32AG000114 (B.Y.C.); NIH grantsT32GM007863 and T32GM008322 (Z.M.H.), a Glenn/AFARScholarship for Research in the Biology of Aging (Z.M.H.); NSFgrant IOS-1119473 (H.A.D.); and the Alexander von HumboldtFoundation and Singapore National Research Foundation grantRF001-363 (J.Y.Y.). This work made use of the DrosophilaAging Core of the Nathan Shock Center of Excellence in theBiology of Aging, funded by National Institute on Aging grantP30-AG-013283. RNA-seq expression data are provided intable S1. The funders had no role in study design, datacollection and analysis, decision to publish, or preparationof the manuscript. The authors declare that they have nocompeting interests. C.M.G., T.-H.K., Z.M.H., and S.D.P.conceived and designed the experiments; C.M.G., T.-H.K.,Z.M.H., B.Y.C., J.Y.Y., H.A.D., and S.D.P. performed theexperiments; C.M.G., T.-H.K., Z.M.H., B.Y.C., J.Y.Y., andS.D.P. analyzed the data; and C.M.G., T.-H.K., J.Y.Y., H.A.D.,and S.D.P. wrote the paper.

Supplementary Materialswww.sciencemag.org/content/343/6170/544/suppl/DC1Materials and MethodsFigs. S1 to S14Table S1References (26–28)

16 July 2013; accepted 31 October 2013Published online 28 November 2013;10.1126/science.1243339

Savanna Vegetation-Fire-ClimateRelationships Differ Among ContinentsCaroline E. R. Lehmann,1,2* T. Michael Anderson,3 Mahesh Sankaran,4,5

Steven I. Higgins,6,7 Sally Archibald,8,9 William A. Hoffmann,10 Niall P. Hanan,11

Richard J. Williams,12 Roderick J. Fensham,13 Jeanine Felfili,14 Lindsay B. Hutley,15

Jayashree Ratnam,4 Jose San Jose,16 Ruben Montes,17 Don Franklin,15

Jeremy Russell-Smith,15 Casey M. Ryan,2 Giselda Durigan,18 Pierre Hiernaux,19

Ricardo Haidar,14 David M. J. S. Bowman,20 William J. Bond21

Ecologists have long sought to understand the factors controlling the structure of savannavegetation. Using data from 2154 sites in savannas across Africa, Australia, and South America,we found that increasing moisture availability drives increases in fire and tree basal area, whereasfire reduces tree basal area. However, among continents, the magnitude of these effects variedsubstantially, so that a single model cannot adequately represent savanna woody biomassacross these regions. Historical and environmental differences drive the regional variation inthe functional relationships between woody vegetation, fire, and climate. These same differenceswill determine the regional responses of vegetation to future climates, with implications forglobal carbon stocks.

Savannas cover 20% of the global land sur-face and account for 30% of terrestrial netprimary production (NPP) and the vast ma-jority of annual global burned area (1–3). Savannaecosystem services sustain an estimated one-fifthof humans, and savannas are also home to mostof the remaining megafauna (1). Tropical savannais characterized by the codominance of C3 treesand C4 grasses that have distinct life forms andphotosynthetic mechanisms that respond differ-ently to environmental controls (4). Examplesinclude the differing responses of these func-tional types to temperature and atmospheric CO2concentrations, predisposing savannas to altera-

tions in structure and extent in the coming cen-tury (4–6).

Tropical savannas are defined by a contin-uous C4 herbaceous layer, with a discontinuousstratum of disturbance-tolerant woody species(7). Although savanna tree cover varies greatly inspace and time (8, 9), the similarities in structureamong the major savanna regions of Africa,Australia, and South America have led to anassumption that the processes regulating veg-etation structure within the biome are equiva-lent (10, 11). Current vegetation models treatsavannas as a homogenous entity (12, 13). Recentstudies, however, have highlighted differences

in savanna extent across continents (14, 15), andit remains unknown how environmental driversinteract to determine the vegetation dynamicsand limits of the biome (10, 14, 15).

We sought universal relationships betweensavanna tree basal area (TBA, m2 ha−1), a keymetric of woody biomass within an ecosystem,and the constraints imposed by resource availa-bility (moisture and nutrients), growing condi-tions (temperature), and disturbances (fire).Ecologists have devoted considerable effort tothe identification of universal relationships to de-scribe the structure and function of biomes (16).However, it has not been clear whether such re-lationships exist. Any such relationships maybe confounded by the unique evolutionary andenvironmental histories of each ecological set-ting (11).

Across Africa and Australia, TBA scales sim-ilarly with rainfall, but the intercepts and the 95thquantile differ substantially (Fig. 1, A to C). Onaverage, at a given level of moisture availability,TBA is higher in Africa and lower in Australia.However, in South America there is almost norelationship between rainfall and TBA, which isprobably in part attributable to the narrow rangeof rainfall that savanna occupies on this continent(Fig. 2). Further, across the observed range ofrainfall, the upper limits of TBA increase linearlywith effective rainfall for Australian savannas(Fig. 1B) but show a saturating response inAfrican and South American savannas (Fig. 1, Aand C). When TBA is used to estimate above-ground woody biomass (AWB) (17), the largedifferences in intercepts between Africa andAustralia are reduced but substantial differencesin the limits remain (fig. S1, A to C). By con-

31 JANUARY 2014 VOL 343 SCIENCE www.sciencemag.org548

REPORTS