Behavioural Brain Research 157 (2005) 119–126

Research report

Increased sensitivity of the serotonergic system during the breedingseason in free-living American tree sparrows

Todd S. Sperrya,∗, Ignacio T. Moorea,b, Simone L. Meddlea,c,Z. Morgan Benowitz-Fredericksa, John C. Wingfielda

a Department of Biology, University of Washington, Box 351800, Seattle, WA 98195, USAb Department of Biology, Virginia Tech, Blacksburg, VA 24061-0406, USA

c Division of Biomedical and Clinical Laboratory Sciences, The University ofEdinburgh Medical School, George Square, Edinburgh EH8 9XD, UK

Received 9 October 2003; received in revised form 6 June 2004; accepted 17 June 2004Available online 23 July 2004

Abstract

how thisn mined thee s (a parrowsw d territoriali d accordingt late, days1 levelo dependentd urthermore,t strates thatf eason. Ther©

K

1

rascsd

ss todi-

essediron-pho-ualulti-andple,ringgive

0d

In order to understand the physiological role of serotonin in regulating aggressive behaviour it is important to understandeuromodulator acts within the context of a naturally fluctuating social and physical environment. To accomplish this, we exaffect of the selective serotonin reuptake inhibitor fluoxetine during the breeding season in free-living male American tree sparrowSpizellarborea) in Northern Alaska. During this time period males are maximally aggressive towards territorial intruders. Male, territorial sere injected with either vehicle or a 10 mg/kg dose of fluoxetine. One hour later, aggression was measured using a simulate

ntrusion. Depending upon when birds were sampled, the aggression scores for vehicle and fluoxetine treatments were groupeo the number of days after territorial behaviour was initiated. The three groups were: early, days 1–5; middle, days 6–10; and1–15. There was a significant overall difference between groups (F(5,36) = 5.18,P < 0.0015). Post hoc analysis demonstrated that thef aggression did not differ between the three groups of saline injected birds. However, fluoxetine injected birds showed a timeecrease in aggression. When compared to control birds the middle and late fluoxetine groups had significantly less aggression. F

he late group of fluoxetine treated birds were significantly less aggressive than the early group of fluoxetine birds. This demonree-living male ATSPs show a rapid change in their behavioural response to fluoxetine across the first 2 weeks of the breeding sapidity of the change in responsiveness suggests a dynamic sensitivity of the serotonergic system.

2004 Elsevier B.V. All rights reserved.

eywords:American tree sparrow; Serotonin; Aggression; Fluoxetine

. Introduction

In free-living vertebrates, both the expression of territo-ial aggression and the magnitude of an agonistic encounterre fluid, varying in response to external environmental andocial cues[61]. Among vertebrates, territorial aggression oc-urs within specific life history stages coincident with otherocial behaviour, for example within the context of the repro-uctive season to gain access to mates[24,39,52]or within

∗ Corresponding author. Tel.: +1 206 5437623; fax: +1 206 5433041.E-mail address:tsperry@u.washington.edu (T.S. Sperry).

the context of the nonbreeding season to maintain acceresources[39,57]. These life history stages follow a unirectional, predictable temporal sequence[24]. Life historystages are further divided into substages which are exprin response to the local and immediate social and envmental conditions and are defined by the current morlogical, physiological and behavioural state of the individ[24,63]. Aggressive behaviours may be expressed in mple substages, but territorial aggression may be sociallyphysically relevant in only specific substages. For examterritorial aggression in birds may be expressed only duthe initial stage of mate attraction and nest building and

166-4328/$ – see front matter © 2004 Elsevier B.V. All rights reserved.oi:10.1016/j.bbr.2004.06.015

120 T.S. Sperry et al. / Behavioural Brain Research 157 (2005) 119–126

way to mate defence and finally nest and chick guarding[30].In addition, territorial aggression will vary in response to thepresence of conspecific males, environmental conditions, thereproductive status of females, and the presence of offspring[61]. Thus, there is an inherent necessity for flexibility in theexpression of territorial aggression within substages.

There is a large body of evidence demonstrating thatsteroid hormones are involved with the regulation of territo-rial aggression[61,64]. For example, within breeding season,aggression is activated by testosterone[50,61]with circulat-ing plasma levels correlating with aggression only during so-cially unstable periods (i.e. the Challenge Hypothesis)[60].In addition to steroidal regulation of aggression, there aremultiple neuromodulators that are known to effect aggres-sive behaviour[40]. One of principle importance is serotonin(5-HT) which has been linked to the inhibition of aggressionacross all vertebrate classes[12,19,41,42,47,51]. The 5-HTsystem, including, but not limited to the 5-HT1A and 5-HT1Breceptors and the 5-HT transporter (SERT), is also regulatedby the steroids 17�-estradiol and testosterone[3,9,35]. The5-HT modulation of aggression is likely controlled throughthe 5-HT1A and 5-HT1B receptors[25,46,47,49].

To examine the activity of the serotonergic systemwithin the free-living Arctic breeding American tree sparrowSpizella arborea(ATSP), we measured the effect on territo-r toninr onalu lev-e ef lar5H hibitt -t ss thevt s,M ad-m sesm set-t

od-e thatc andt tage[ oralc s dur-i reeda musts atingo timec TSPsb s ap-p pons thes imizet rctic

[22]. Since we are interested in understanding the mecha-nisms for this rapid modulation, we examined the effect offluoxetine has on territorial behaviour during the first twoweeks of the breeding season in male ATSPs. This time pe-riod encompasses the initiation of territorial behaviour, mateacquisition, nest building (by females) and egg laying. To testthe hypothesis that 5-HT modulates aggressive behaviour, wemeasured aggression in free-living ATSPs using a simulatedterritorial intrusion (STI) (see[58]) 1 h following fluoxetineadministration.

2. Materials and methods

2.1. Study area

This research was conducted during May and June of 2000and 2002 at the Toolik Lake Field Station on Alaska’s northslope of the Brooks Range, Alaska (68◦38′N, 149◦38′W).Male ATSPs were caught using mist nets combined withpre-recorded conspecific song playback. After capture, birdswere immediately removed from the net and measurementstaken for wing and tarsus length, weight and the lengthand width of the cloacal protuberance. Cloacal protuber-ance volume was obtained using the formula for a cylinder.I singa dw andg ed a7 .9%Np re-l to thet hilem d andh d andc xper-i linese An-i

2

achm singa terri-t ack.C ationw fer-e s froma e af-t thatt entf usede 0 mino by

ial aggression of acute treatment with the selective seroeuptake inhibitor fluoxetine. Fluoxetine blocks the neurptake pump for 5-HT causing increased extra cellularls in both the dorsal raphe nucleus[31] as well as in th

orebrain[23]. The increase in local, synaptic extra cellu-HT levels are then thought to act via the 5-HT1A and 5-T1B receptors to decrease 5-HT synthesis as well as in

he firing of serotonergic neurons[2]. The ability of fluoxeine to lower aggressive behaviour has been shown acroertebrate classes from fish[42], reptiles[12] and birds[51]o mammals[14,16,41]. In a previous study in song sparrowelospizamelodia morphna, we demonstrated that acuteinistration of fluoxetine (10 mg/kg) significantly decreaale-male agonistic behaviour in a controlled laboratory

ing without affecting general activity levels[51].Arctic breeding birds, such as the ATSP, are excellent m

ls with which to examine the endocrine mechanismsontrol the short-term regulation of territorial aggressionemporal shifts between substages within a life history s1]. Birds breeding in the arctic face considerable temponstraints due to the brief spring and summer seasonng which they must establish a territory, find a mate, bnd raise young. Importantly, the balance which malestrike between aggressive defence of resources and mr parental care is made more tenuous because of theonstraints within the breeding season. In the case of Areeding above the Arctic Circle the breeding season iroximately 28–35 days. Thus, a premium is placed upeed with the rapid onset of territorial aggression andubsequent movement through the substages to maxhe benefits and minimize the costs of breeding in the A

n addition, abdominal and furcular fat was assessed ustandard five-point scale (see[59]). Birds were bandeith United States Fish and Wildlife aluminium bandsiven a unique set of colour bands. Each bird receiv0�l subcutaneous injection of either saline vehicle (0aCl) or fluoxetine ((±)N-methyl-�-(4-[trifluoromethyl]-henoxy)-benzenepropanamine) (10 mg/kg) and then

eased at the capture site. The researchers were blindreatment groups until after the study was completed. Wales were targeted, some females were also capturead body morphometrics assessed, brood patch scorehecked the abdomen for the presence of an egg. The ements reported herein were performed under the guidestablished by the University of Washington Institute of

mal Care and Use Committee.

.2. Assessment of territorial aggression

One hour after injection the territorial aggression of eale was quantified using a STI. STIs were conducted ucaged conspecific decoy placed near the centre of the

ory combined with pre-recorded conspecific song playbare was taken to place the decoy away from the lochere the focal bird was previously captured. Two difnt playback tapes were used and each contained songsingle male repeated every 10 s with a 1 min silenc

er every 2 min of song. The tapes were alternated sohe playback used to catch the bird initially was differrom the one used during the STI and each treatmentqual numbers of each song type. The STI consisted of 1f focal observation with decoy and playback, followed

T.S. Sperry et al. / Behavioural Brain Research 157 (2005) 119–126 121

10 min of focal observation to measure the persistence of theresponse with no decoy present and no song playing (Per-sistence). During the STI, the number of songs, number offlights, closest approach, time spent within 5 m of the de-coy and the latency to respond to the decoy and song wereall recorded. None of the ATSPs were on adjacent territorieswithin each year. The ATSPs that did not respond were as-sumed to be present on their territories and were assigned aresponse latency of 600 s and a closest approach of 50 m. Theterritories had a maximum size of 100 m2 and the decoy wasplaced near the centre of the territory, thus 50 m was set asan approximate maximum distance from the decoy (personalobservation and[38]). For the vehicle injected group, 17 of18 males responded to the STI and were confirmed presenton the territory. For the fluoxetine treatment group 9 out of 19males responded to the STI and 11 of 19 were confirmed to bepresent during the STI. After each STI, an attempt was madeto find each of the non-responding males, although it wasnot possible to search each territory completely. The struc-ture of the territories consists of willow (Salixsp.) and dwarfbirch (Betula sp.) between 1 and 3 m high surrounded byopen tundra. The non-and low-responding fluoxetine treatedmales that were confirmed to be present on their territoriesshowed no indication of sedation as a result of the treatment.These birds were more secretive and remained quiet untild rri-t ownfi son,w esento toryt t doseo iour[

botht om-b mpo-n tP o-r nents PC1e uralm equal fourb STI.P e and

TF l com-p istencep

V C1

RTCNN

each of the variables loaded approximately equally onto PC1(Table 1).

2.3. Statistics

Body morphometrics were compared using at-test to com-pare between years and treatment groups, while fat scoresand wing length were analyzed using a Mann–Whitney ranksum test (α = 0.05). The aggression score (PC1) and persis-tence of aggression scores (PC1) were analyzed using two-factor ANOVAs to examine the effects of treatment, time af-ter the initiation of territoriality and sampling-year. Data arepresented as means± S.E.M. Statistics were computed us-ing JMP (SAS Institute Inc. Cary, NC, USA) and SigmaStat(SPSS, Chicago, IL, USA).

3. Results

3.1. Initiation of territorial behaviour

In 2000, the ground was completely snow covered until 5June and because they are ground nesters the ATSPs did notinitiate territorial behaviour until after the snow melted fromthe breeding territories. In this year, the birds were caughtb s nos iourw the1 24tha werec oft

3

2000a sl= htwP wass

TA r thiss

W

T

WA

C

isturbed, after which they flew to another part of their teory. Based upon the treated birds behaviour and their kndelity to their territories during the early breeding seae concluded that the non-responding males were prn their territories during the STI. Furthermore, labora

ests in song sparrows demonstrated that an equivalenf fluoxetine had no adverse effects on locomotive behav

51].In the analysis of agonistic behaviour, the data from

reatment groups for all five behavioural variables were cined into a single aggression score using a principal coent (PC) analysis without rotation[36]. We interpret the firsC scores (PC1) (seeTable 1) as a measure of over all territ

ial aggression and use each individual principal compocore as a proxy for an individual’s level of aggression.xplained 66% of the variance within the five behavioeasures and each of the measures had approximately

oading onto PC1. PC scores were also generated forehavioural measures for the persistence portion of theC1 for these measures explained 54% of the varianc

able 1actor loadings and variance explained by PC1 for the five behaviouraonents of the STI and the three behavioural components of the persortion of the STI

ariable STI PC1 Persistence P

esponse latency −0.47ime within 5 m 0.38 0.31losest approach −0.48 −0.54umber of flights 0.43 0.57umber of songs 0.46 0.53

l

etween June 6th and 15th. However, in 2002 there wanow cover when the birds arrived, thus territorial behavas initiated almost immediately after their arrival on9–21st May. In 2002, birds were sampled between Maynd June 4th. Therefore, the data sets for the 2 yearsombined along a time line with respect to the initiationerritorial aggression.

.2. Body morphometrics

There were no significant differences between yearsnd 2002 in the wing length (U(15,27) = 299,P= 0.546), tarsu

ength (t(1,37) = 1.74,P = 0.09) or in the fat score (U(15,27)270.5,P = 0.176) (Table 2). However, the average weigas significantly higher in 2000 than in 2002 (t(1,40) = 6.94,< 0.001) and the volume of the cloacal protuberance

ignificantly greater in 2000 than in 2002 (t(1,40) = −7.33,

able 2comparison of the body morphometrics of the male ATSPs used fo

tudy

2000 (n) 2002 (n) (α = 0.05)

ing length(mm)

72.0± 0.7 (15) 72.6± 0.2 (27) P = 0.546

arsus length(mm)

23.4± 0.08 (15) 23.1± 0.1 (24) P = 0.09

eight (g) 20.5± 0.25(15) 18.5± 0.86 (27) P < 0.001verage fatscore

0.6± 0.1 (15) 0.81± 0.1 (27) U = 0.176

loacalprotuberance(mm3)

1242.4± 42.1 (15) 743.9± 44.8 (27) P < 0.001

122 T.S. Sperry et al. / Behavioural Brain Research 157 (2005) 119–126

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

Agg

ress

ion

Sco

r

Number of Days after the Initiation of Territorial Behaviour

e

FluoxetineSaline

A

A

A

CB, C

A, B

+1-5 +6-10 +11-15

(4)(4) (7)

(10) (5)

(7)

Fig. 1. Effect of fluoxetine and the number of days after initiation of terri-torial behaviour in male ATSP on the overall aggression scores exhibited inresponse to a territorial challenge 1 h after injection. Data are presented asmeans ± S.E.M. for aggression scores for fluoxetine and vehicle injectedbirds. It is important to note that the PC scores represent relative levels ofaggression, with negative numbers referring to lower levels of aggressionnot negative aggression. Different letters represent statistically significantdifferences at P < 0.05.

P < 0.001) (Table 2). There were no significant differencesbetween the treatment groups within either year for any bodymeasurement (data not shown).

3.3. Territorial aggression

We examined the effect of fluoxetine on territorial ag-gression in male ATSPs immediately after the birds had es-tablished and begun to actively defend breeding territories.Depending upon when birds were sampled, the aggressionscores for vehicle and fluoxetine treatments were groupedaccording to the number of days post-initiation of territorialbehaviour. The three groups were: early, days 1–5; middle,days 6–10; and late, days 11–15. There was a significant dif-ference between groups (F(5,36) = 5.18, P< 0.0015) (see Fig.1), but there was no significant difference in aggression scoresbetween sampling years (F(1,36) = 0.069,P< 0.794). Post hocanalysis demonstrated that the level of aggression did not dif-fer between the three groups of saline injected birds. On theother hand, fluoxetine injected birds demonstrated a signifi-cant decrease in aggression with respect to the control birds

Table 3Scores for each of the behavioural measures observed during the 10 min. STI and thand fluoxetine (10 mg/kg) injected ATSPs (average ± S.E.M.)

STI

NNLCT

in the middle and late groups of birds. Furthermore, the lategroup of fluoxetine treated birds showed significantly lowerlevels of aggression compared to the fluoxetine treated birdssampled between days 1 and 5. The birds sampled betweendays 6 and 10 showed an intermediate level of aggressionthat was not significantly different from either the early orlate groups. See Table 3 for the averages for each of the fivebehavioural measures used to generate the PC scores.

3.4. Persistence of territorial aggression

The persistence of aggression scores were analyzed in anidentical manner to the STI aggression scores. There were nosignificant differences in persistence of aggression betweenthe six treatment groups (F(5,36) = 1.47, P < 0.223), nor wasthere a difference in aggression between the sampling years(F(1,36) = 0.017,P< 0.898). However, the data do demonstratea similar pattern to that seen for the STI aggression scores(Fig. 1).

4. Discussion

The current study demonstrates that free-living, male AT-SPs show a rapid change in their behavioural response to flu-osriwasdrrsrsots(eao

Saline

umber of songs per STI 27.8 ± 4.7umber of Flights per STI 11.5 ± 3.5atency to respond (s) 151.8 ± 37.0losest approach to decoy (m) 8.3 ± 3.4ime spent within 5 m of decoy (s) 117.3 ± 40.8

e subsequent 10 min persistence portion of the STI for the vehicle (saline)

Persistence

Fluoxetine Saline Fluoxetine

6.0 ± 3.0 15.1 ± 4.5 4.4 ± 2.91.3 ± 0.5 2.7 ± 0.6 0.68 ± 0.6

406.5 ± 55.832.1 ± 4.7 22.0 ± 4.6 35.0 ± 4.854.3 ± 30.7 0 14.5 ± 11.8

xetine during the first 2 weeks of the breeding season. At thetart of the breeding season the birds settle on territories andapidly begin defending it against conspecific intruders. Dur-ng the first 5 days of territorial behavior, the males injectedith fluoxetine did not differ in their aggressive response toSTI when compared to control birds. The birds begin to

how an increased response to fluoxetine between 6 and 10ays after initiating territorial behaviour. After 11 days theesponse to fluoxetine is both significantly different from theesponse of fluoxetine treated during the first 5 days of thetudy as well as all of the control birds. This suggests that theesponse to fluoxetine is dynamic and that as the breedingeason progresses there is likely a change in the sensitivity ofne or more components of the serotonergic system that con-ributes to the increasing efficacy of the drug. In addition, thistudy also re-affirms our laboratory findings in song sparrowsMelospiza melodia morphna) which found that both fluox-tine and the 5-HT1A receptor agonist 8-OH-DPAT reducedggressive behaviour [51]. The time course and dose of flu-xetine used in the laboratory study were equivalent to that

T.S. Sperry et al. / Behavioural Brain Research 157 (2005) 119–126 123

used in the present study. However, the current study high-lights the importance of field studies since the dynamic natureof the serotonergic system is not present under the controlledlaboratory conditions. To our knowledge this is the first studyto examine the serotonergic control of aggressive behaviourin a free-living song bird. In a similar study, Perreault etal. [42] show that fluoxetine reduces territorial aggression inblue-head wrasse, Thalassoma bifasciatum, in a comparablemanner in both field and laboratory studies (Fig. 2).

The rapidity of life-history changes during the breedingseason make male ATSPs an excellent model for examiningrapid changes in the neuroendocrine control of territorial ag-gression. In response to the brief arctic breeding season, AT-SPs arrive on their breeding grounds in Northern Alaska andquickly shift from migratory behaviour to territory establish-ment and mate acquisition. This rapid initiation of territorialbehaviour occurred during the 2002 field season. However,in 2000 the start of the breeding was delayed because thebreeding grounds were snow-covered. The ATSPs foragedin flocks waiting to commence pair formation and territoryestablishment once the snow melted. The males and femalesform pairs within hours once males establish a territory. Itis likely that the differences in body morphometrics betweenthe 2 years is a result of these dramatic differences in weatherpatterns. In 2000, the additional time the males spent foragingpctw

ott

Frslsd

and personal observation). There is a 2–3-week-period duringwhich females are fertile and males are maximally aggressivetowards other males and plasma testosterone (T) is at peaklevels [1,22]. Thus, during the time period in which this studywas conducted, the ATSPs were experiencing maximum lev-els of T and were displaying maximum levels of aggressivebehaviour towards conspecific males. Importantly, this studyshows that during this time the behavioural response to flu-oxetine is rapidly changing and suggests that despite highlevels of aggression these birds may have an increasinglyactive serotonergic system.

4.1. Mechanisms of fluoxetine action

Within this study, we used fluoxetine as a neuroendocrineprobe to examine the activity of the serotonergic system infree-living birds. Fluoxetine as well as other selective sero-tonin reuptake inhibitors have been previously been shown tobe good tools to assess the activity of the serotonergic system[26,43]. In contrast to these other studies, we used territorialaggression as the endpoint rather than a hormonal responseto assess serotonergic activity.

It is important to note that while we used aggression as anendpoint to measure the activity of the serotonergic systemit does not necessarily mean that we are measuring a directetipooHwgodtinuhpFcresott

icflee

rior to pair formation allowed them to gain weight and theloacal protuberance to grow larger. Since the cloacal pro-uberance is an organ for sperm storage, a delay in breedingill increase its volume [48].The focus of this current study was upon the first 2-weeks

f territorial aggression, a time period which encompasseshe establishment of territories, acquisition of a mate, initia-ion of nest-building and laying of the first sets of eggs ([38]

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

Per

sist

ence

of

Number of Days after the Initiation of Territorial Behaviour

Agg

ress

ion

Sco

re

FluoxetineSaline

+1-5 +6-10 +11-15

(4)(4) (7)

(10) (5)

(7)

ig. 2. Effect of fluoxetine and the number of days after initiation of ter-itorial behaviour in male ATSP on the overall persistence of aggressioncores exhibited during the 10 min immediately following a territorial chal-enge. Data are presented as means ± S.E.M. for persistence of aggressioncores for fluoxetine and vehicle injected birds. There were no significantifferences between treatment groups.

ffect on the neural control of aggression. It is possible thathe fluoxetine induced decrease in aggressive behaviour is anndirect effect as a result of activation of the hypothalamic-ituitary-adrenal (HPA) axis [7]. 5-HT is a known modulatorf glucocorticosteroid release potentially acting at all levelsf the HPA axis [10]. Interestingly, under stress conditions 5-T can depress the maximum release of glucocorticosteroids,hereas under basal conditions 5-HT can increase plasmalucocorticosteroid levels [20]. However, ATSPs, like manyther animals [62] resist the effects of environmental stressorsuring the reproductive season and are behaviourally insensi-ive to exogenous administration of corticosterone [1]. Thus,n this study a rise in corticosterone as a result of the combi-ation of a brief handling stress and fluoxetine treatment isnlikely to explain the observed decrease in aggressive be-aviour. However, activation of the HPA axis also has theotential to alter the sensitivity of the serotonergic system.or example, after an acute restraint stress there was a de-rease in mRNA for SERT in rat brain tissue within 1 of 6aphe nuclei that were examined [56]. However, to our knowl-dge there has not been an examination of the effects of acutetress on SERT protein levels or on SERT activity. This leavespen the possibility that corticosterone feedback could con-ribute to the fluoxetine induced behavioural change seen inhis study.

The predominant mechanism of action for fluoxetine is vianhibition of the neuronal uptake pump for 5-HT leading to in-reased synaptic levels of 5-HT [15]. Acute administration ofuoxetine, either subcutaneous or intra-peritoneal, increasesxtra-cellular levels of central 5-HT [11,31]. Significantly el-vated levels of 5-HT are measured at 30 min and maximum

124 T.S. Sperry et al. / Behavioural Brain Research 157 (2005) 119–126

stimulation occurs within 60 min at a dose of 10 mg/kg in rats[31]. The increase in 5-HT occurs primarily in the hind-brainwithin the raphe nuclei with 5-HT levels also increasing inthe hippocampus and frontal cortex [23,31]. The magnitudeof the 5-HT release correlates with the number of SERT [31].

The antiaggressive properties of fluoxetine [16] are theresult of increased 5-HT acting via the 5-HT1A and 5-HT1Breceptors [25,46,47,49]. 5-HT1A receptors, located both pre-and post-synaptically, are linked to G-protein-gated ion chan-nels where their activation decreases neuronal firing [21].5-HT1B receptors are located pre-synaptically on both sero-tonergic and non-serotonergic neurons where their activa-tion inhibits neurotransmitter release [2,21]. As a result of itsmechanism of action, an acute injection of fluoxetine can beconsidered a general 5-HT agonist.

4.2. Possible targets of modulation

We do not know from this study the mechanism that re-sults in the observed effect on aggressive behaviour, but thechange in fluoxetine efficacy could be occur at several dif-ferent levels within the serotonergic system. There is consid-erable evidence showing that 5-HT levels are regulated bygonadal steroids, in particular, T, 17�-estradiol (E2) and pro-gesterone (P4) (reviewed in [3]). The principle enzymes andrpodaafics

wstalr[[t

rflriirnaat

a significantly greater increase in extra cellular 5-HT duringthe dioestrous phase. This was in conjunction with higherextra cellular 5-HT levels in the diestrous phase when com-pared to oestrous females [18]. This study demonstrates twoimportant parallels with our current study. First, the increased5-HT activity is associated with an increased responsivenessto paroxetine. Second, this drug’s actions are dynamic anddependent upon the state, or substage, of the animal.

4.3. Serotonin and aggression

5-HT has been demonstrated to be an important modu-lator of aggressive behaviour, with the historical consensusbeing that 5-HT levels are inversely correlated with aggres-sion [19,41]. For example, a decrease in 5-HT function cor-relates with impulsive aggression and/or violence in primates[8,13,19]. While the relationship between 5-HT and aggres-sion holds true across a wide variety of species, there are anumber of studies that demonstrate that high levels of ag-gressive behaviour also co-occur with high levels of 5-HT.For example, in free-living lizards (Sceloporus jarrovi) [32]and in laboratory held lizards (Anolis carolinensis) [28] 5-HTactivity is higher immediately after an aggressive encounterin the more aggressive males. Studies in rats and mice pro-vide further evidence that an active serotonergic system isaaf5frclhilasp

asmthshinigbtbsn

eceptors that directly influence 5-HT levels include trypto-han hydroxylase, the rate-limiting enzyme in the synthesisf 5-HT [4,29,53], monoamine oxidase-A, the enzyme thategrades 5-HT after its release [17], SERT and the 5-HT1And 5-HT1B receptors. Since each of these targets is regulated,t least in part, by gonadal steroids, each is a potential siteor regulation during the rise in testosterone associated withnitiation of territorial behaviour [3,45]. Therefore, this mayontribute to the rapid increase in the efficacy of fluoxetineeen in this study.

The modulation of either the 5-HT1A or 5-HT1B receptorsill affect the degree to which fluoxetine reduces aggres-

ive behaviour. As described earlier, these two receptors arehought to be the endpoint through which fluoxetine reducesggressive behaviour [25,46,47,49] and both are regulated, ateast in part, by gonadal steroids [35,49]. T can either increaseeceptor number [35] or change the affinity of the receptors6]. This raises the possibility that T can increase aggression61,64] concomitant with an increase in 5-HT1A receptor ac-ivity.

The regulation of SERT proteins is dynamic with rapidegulation at the cell-surface [5] with levels of the protein re-ecting levels of 5-HT [3,31]. This is in large part due to theegulation of SERT by 5-HT. This leads to a scenario where anncrease in serotonergic activity is coupled with an increasen SERT activity [5,31]. Steroids have also been shown toegulate SERT mRNA and protein levels, but do so in a man-er that differs between species and brain regions [3,33]. Instudy that is analogous to our current study, Gundlah et

l. [18] showed that across the oestrous cycle of female ratshe selective serotonin reuptake inhibitor paroxetine caused

lso positively associated with aggressive behaviour. Withinwild house mice strain (Musmusculus domesticus) selected

or low aggression there are significantly fewer postsynaptic-HT1A receptors when compared to a line of mice selectedor high aggression [27]. Similarly, in a strain of wild-typeats (Rattus norveticus) a highly aggressive line showed in-reased 5-HT1A activity compared to a line demonstratingow aggression [54]. Further study within this strain of ratsas demonstrated that 5-HT neuronal activation is increasedn the highly aggressive strain compared to that found in theow aggression strain [55]. Thus, the relationship betweenggression and serotonin is certainly more complex than aimple linear relationship between 5-HT activity and the ex-ression of aggressive behaviour.

Studies in primates provide additional lines of evidencend also point to a key link between 5-HT, aggression andocially productive behaviour. In male, free-ranging rhesusacaques (Macaca mulatta) low levels of serotonergic ac-

ivity correlate with highly aggressive but inappropriate be-aviour [19]. During the breeding season these highly aggres-ive animals have lower reproductive success than males thatave higher levels of serotonergic activity and are aggressiven a contextually appropriate manner [34]. Interestingly, theegative relationship between 5-HT activity and aggressions not present during the breeding season when inter-male ag-ression is at its peak, but there is a strong positive correlationetween grooming and 5-HT activity [34]. The negative rela-ionship between 5-HT activity and competent socio-sexualehaviour in macaques [34] is also mirrored in other primatetudies that have examined the relationship between domi-ance and 5-HT. In response to increasing brain 5-HT levels

T.S. Sperry et al. / Behavioural Brain Research 157 (2005) 119–126 125

social dominance was enhanced in vervet monkeys (Cerco-pithecus aethiops sabaeus) [44] as well as in humans [37].Thus, it is evident that in primates low 5-HT tends to lead toinappropriate behaviour whereas males with higher seroton-ergic activity tend to be dominant and socially and reproduc-tively successful.

Together, the studies in lizards, rodents, primates and nowbirds all suggest that an active and possibly an enhanced sero-tonergic system may be necessary for displaying high levelsof behaviourally appropriate aggression. Clearly, there aretwo sides to the relationship between aggression and 5-HT.On the one hand, dysregulation of the serotonergic system canlead to increased aggressive behaviour that is frequently ex-pressed in a contextually inappropriate manner. On the otherhand, there is a body of research that suggests that an activeserotonergic system is necessary to support high levels of ap-propriate aggression. Thus, the direct action of 5-HT is todecrease aggressive behaviour. However, in the broader con-text the serotonergic system may provide a mechanism foran animal to effectively manage aggressive behaviour withina complex social environment. This leads to the hypothesisthat the rapid change in the behavioural response to fluoxetineseen in male ATSPs may provide a mechanism to manage themultiple and sometimes conflicting behavioural roles malesmust undertake during the breeding season such as mate at-trtuaf

A

asIS

R

[6] Bonson KR, Johnson RG, Fiorella D, Rabin RA, Winter JC. Seroton-ergic control of androgen-induced dominance. Pharmacol BiochemBeh 1994;49:313–22.

[7] Carrasco GA, Van de Kar LD. Neuroendocrine pharmacology ofstress. Eur J Pharmacol 2003;463:235–72.

[8] Coccaro EF, Kavoussi RJ, Hauger RL. Serotonin function and an-tiaggressive response to fluoxetine: a pilot study. Soc Biolog Psych1997;42:546–52.

[9] Cologer-Clifford A, Simon NG, Richter ML, Smoluk SA, Lu S. An-drogens and estrogens module 5-HT1A and 5-HT1B agonist effectson aggression. Physiol Beh 1999;65:823–8.

[10] Contesse V, Lefebvre H, Lenglet S, Kuhn J-M, Delarue C, Vaudry H.Role of 5-HT in the regulation of the brain-pituitary-adrenal axis:effects of 5-HT on adrenocortical cells. Can J Physiol Pharmacol2000;78:967–83.

[11] Czachura JF, Rasmussen K. Effects of acute and chronic administra-tion of fluoxetine on the activity of serotonergic neurons in the dorsalraphe nuclei of the rat. N-S Arch Pharmacol 2000;362:266–75.

[12] Deckel AW. Behavioral changes in Anolis carolinensis followinginjection with fluoxetine. Behav Brain Res 1996;78:175–82.

[13] Fairbanks LA, Melega WP, Jorgensen MJ, Kaplan JR, McGuireMT. Social impulsivity inversely associated with CSF 5-HIAA andfluoxetine exposure in vervet monkeys. Neuropsychopharmacology2001;24:370–8.

[14] Ferris CF, Melloni RH, Koppel G, Perry KW, Fuller RW, DelvilleY. Vasopressin/serotonin interactions in the anterior hypothala-mus control aggressive behaviour in golden hamsters. J Neurosci1997;17:4331–40.

[15] Fuller RW. Uptake inhibitors increase extracellular serotonin con-centration measured by brain microdialysis. Life Sci 1994;55:163–7.

[

[

[

[

[

[

[

[

[

[

[

[

raction, pair bonding, mating and parental behaviour. Thisaises the question of whether the dynamic nature of the sero-onergic system in ATSPs is an adaptation to cope with thenique arctic environment or a more universal adaptation tollow successful trade-offs between aggressive territorial de-ence and other affiliative behaviours.

cknowledgements

We thank N. Owen-Ashley, L. Hayward and H. Wada forssistance in the field and D. Wacker and W. Goyman fortatistical assistance. This work was supported by a Nationalnstitutes of Health training grant (TSS) and the Nationalcience Foundation (JCW).

eferences

[1] Astheimer LB, Buttemer WA, Wingfield JC. Corticosterone treat-ment has no effect on reproductive hormones or aggressive be-haviour in free-living male tree sparrows Spizella arborea. HormBehav 2000;37:31–9.

[2] Barton CL, Hutson PH. Inhibition of hippocampal 5-HT synthesisby fluoxetine and paroxetine: evidence for the involvement of both5-HT1A and 5-HT1B/D autoreceptors. Synapse 1999;31:13–9.

[3] Bethea CL, Lu NZ, Gundlah C, Streicher JM. Diverse actions ofovarian steroids in the serotonin neural system. Front Neuroendocr2002;23:41–100.

[4] Bethea CL, Mirkes SJ, Shively CA, Adams MR. Steroid regulationof tryptophan hydroxylase protein in the dorsal raphe of macaques.Biol Psych 2000;47:562–76.

[5] Blakely R, Bauman AL. Biogenic amine transporters: regulation influx. Curr Opin Neurobiol 2000;10:328–36.

16] Fuller RW. The influence of fluoxetine on aggressive behaviour. Neu-ropsychopharmacology 1996;14:77–81.

17] Gundlah C, Lu NZ, Bethea CL. Ovarian steroid regulation of mon-amine oxidase-A and B mRNAs in the macaque dorsal raphe andhypothalamic nuclei. Psychopharmacology 2002;160:271–82.

18] Gundlah C, Simon LD, Auerbach SB. Differences in hypothalamicserotonin between oestrous phases and gender: an in vivo microdial-ysis study. Brain Res 1998;785:91–6.

19] Higley JD, Mehlman PT, Poland RE, Taub DM, Vickers J, SuomiSJ, et al. CSF testosterone and 5-HIAA correlate with different typesof aggressive behaviours. Biol Psych 1996;40:1067–82.

20] Hoglund E, Balm PHM, Winberg S. Stimulatory and inhibitory ef-fects of 5-HT1A receptors on adrenocorticotropic hormone and cor-tisol secretion in a teleost fish, the Arctic charr (Salvelinus alpinus).Neurosci Lett 2002;324:193–6.

21] Hoyer D, Hannon JP, Martin GR. Molecular, pharmacological andfunctional diversity of 5-HT receptors. Pharmacol Biochem Beh2002;71:533–54.

22] Hunt K, Wingfield JC, Astheimer LB, Buttemer WA, Hahn TP.Temporal patterns of territorial behaviour and circulating testos-terone in the lapland longspur and other actic passerines. Am Zool1995;35:274–84.

23] Invernizzi R, Bramante M, Samanin R. Role of 5-HT1A receptors inthe effects of acute and chronic fluoxetine on extracellular serotoninin the frontal cortex. Pharmacol Biochem Behav 1996;54:143–7.

24] Jacobs JD, Wingfield JC. Endocrine control of life-cycle stages:a constraint on response to the environment. Condor 2000;102:35–51.

25] Joppa MA, Rowe RK, Meisel RL. Effects of serotonin 1A and 1Breceptor agonists on social aggression in male and female Syrianhamsters. Pharmacol Biochem Behav 1997;58:349–53.

26] Kojima H, Terao T, Iwakawa M, Soya A, Inoue N, Shiraishi Y, etal. Paroxetine as a 5-HT neuroendocrine probe. Psychopharmacology2003;167:97–102.

27] Korte SM, Meijer OC, de Kloet ER, Buwalda B, Keijser J, SluyterF, et al. Enhanced 5-HT1A receptor expression in forebrain regionsof aggressive house mice. Brain Res 1996;736:338–43.

126 T.S. Sperry et al. / Behavioural Brain Research 157 (2005) 119–126

[28] Korzan WJ, Summers TR, Summers CH. Monoaminergic ac-tivities of limbic regions are elevated during aggression: in-fluence of sympathetic social signalling. Brain Res 2000;870:170–8.

[29] Lu NZ, Shlaes TA, Dziennes SE, Lyle SE, Bethea CL. Ovariansteroid action on tryptophan hydroxylase protein and serotonin com-pared to localization of ovarian steroid receptors in midbrain guineapigs. Endocrine 1999;11:257–67.

[30] Lynn SE, Hayward LS, Benowitz-Fredericks ZM, Wingfield JC.Behavioural insensitivity to supplementary testosterone during theparental phase in the chestnut-collared longspur Calcarius ornatus.Anim Behav 2002;63:795–803.

[31] Malagie I, Trillat A-C, Jacquot C, Gardier AM. Effects of acutefluoxetine on extracellular serotonin levels in the raphe: an in vivomicrodialysis. Eur J Pharmacol 1995;286:213–7.

[32] Matter JM, Ronan PJ, Summers CH. Central monoamines in free-ranging lizards: differences associated with social roles and territo-riality. Brain, Behav Evol 1998;51:23–32.

[33] McQueen JK, Wilson H, Sumner BEH, Fink G. Serotonin transproter(SERT) mRNA and binding densities in male rat brain affected bysex steroids. Mol Brain Res 1999;63:241–7.

[34] Mehlman PT, Higley JD, Fernald BJ, Sallee FR, Suomi SJ, LinnoilaM. CSF 5-HIAA, testosterone, and sociosexual behaviours in free-ranging male rhesus macaques in the mating season. Psych Res1997;72:89–102.

[35] Mendelson SD, McEwen BS. Testosterone increases the concentra-tion of [3H]8-hydroxy-2-(di-n-propylamino)tetralin. Binding at 5-HT1A receptors in the medial preoptic nucleus of castrated malerats. Eur J Pharmacol 1990;181:329–31.

[36] Moore IT, Wada H, Perfito N, Busch DS, Hahn TP, Wing-

[

[

[

[

[

[

[

[

[

[

[47] Saudou F, Amara DA, Dierich A, LeMeur M, Ramboz S, Segu L, etal. Enhanced aggressive behaviour in mice lacking 5-HT1B receptor.Science 1994;265:1875–8.

[48] Sax A, Hoi H. Individual and temporal variation in cloacal pro-tuberance size of male bearded tits (Panurus biarmicus). Auk1998;115:964–9.

[49] Simon NG, Cologer-Clifford A, Lu S, McKenna SE, Hu S.Testosterone and its metabolites modulate 5HT1A and 5HT1Bagonist effects on intermale aggression. Neurosci Biobehav Rev1998;23:325–36.

[50] Simon NG, McKenna SE, Lu S-F, Cologer-Clifford A. Developmentand expression of hormonal systems regulating aggression. Ann NY Acad Sci 1996;794:8–17.

[51] Sperry TS, Thompson CK, Wingfield JC. Effects of acute treatmentwith 8-OH-DPAT and fluoxetine on aggressive behaviour in malesong sparrows (Melospiza melodia morphna). J Neuroendocrinol2003;15:150–60.

[52] Stamps J. Territorial behaviour: testing the assumptions. Advancesin the Study of Behaviour 1994;23:173–232.

[53] Tsai C-L, Wang L-H. Effects of gonadal steroids on the serotoninsynthesis and metabolism in the early developing tilapia brain. Neu-rosci Lett 1999;264:45–8.

[54] van der Vegt BJ, de Boer SF, Buwalda B, de Ruiter AJH, de JongJG, Koolhaas JM. Enhanced sensitivity of postsynaptic-1A recep-tors in rats and mice with high trait aggression. Physiol Behav2001;74:205–11.

[55] van der Vegt BJ, Lieuwes N, van de Wall EHEM, Kato K, Moya-Albiol L, Martınez-Sanchis S, et al. Activation of serotonergic neu-rotransmission during the performance of aggressive behaviour inrats. Behav Neurosci 2003;117:667–74.

[

[

[

[

[

[

[

[

[

field JC. Territoriality and testosterone in an equatorial popula-tion of rufous-collared sparrows Zonotrichia capensis. Anim Behav2004;67:411–20.

37] Moskowitz DS, Pinard G, Zuroff DC, Annable L, Young SN. Theeffect of tryptophan on social interaction in everday life: a placebo-controlled study. Neuropsychopharmacology 2001;25:277–89.

38] Naugler CT. American tree sparrow: Spizella arborea. In: Gill FB,editor. The Birds of North America; 1993. p. 1–12.

39] Nelson RJ. An Introduction to Behavioral Endocrinology. Sunder-land, MA: Sinauer Association Inc.; 1995 (611 pp.).

40] Nelson RJ, Chiavegatto S. Molecular basis of aggression. TRENDSNeurosci 2001;24:713–9.

41] Olivier B, Mos J, van Oorschot R, Hen R. Serotonin recep-tors and animal models of aggressive behaviour. Pharmacopsych1995;28:80–90.

42] Perreault H, Semsar K, Godwin J. Fluoxetine treatment decreasesterritorial aggression in a coral reef fish. Physiol Behav 2003;79:719–24.

43] Raap DK, Van de Kar LD. Selective serotonin reuptake inhibitorsand neuroendocrine function. Life Sci 1999;65:1217–35.

44] Raleigh MJ, McGuire MT, Brammer GL, Pollack DB, Yuwiler A.Serotonergic mechanisms promote dominance acquisition in adultmale vervet monkeys. Brain Res 1991;559:181–90.

45] Rubinow DR, Schmidt PJ, Roca CA. Estrogen-serotonin interac-tions: implications for affective regulation. Biol Psych 1998;44:839–50.

46] Sanchez C. Interaction studies of 5-HT1A receptor antagonists andselective 5-HT reuptake inhibitors in isolated aggressive mice. EurJ Pharmacol 1997;334:127–32.

56] Vollmayr B, Keck S, Henn FA, Schloss P. Acute stress decreasesserotonin transporter mRNA in the raphe pontis but not in otherraphe nuclei of the rat. Neurosci Lett 2000;290:109–12.

57] Wingfield JC. Control of territorial aggression in a changing envi-ronment. Psychoneuroendocrinology 1994;19:709–21.

58] Wingfield JC. Regulation of territorial behaviour in the sedentarysong sparrow Melospiza melodia morphna. Horm Behav 1994;28:1–15.

59] Wingfield JC, Farner DS. Avian endocrinology-field investigationsand methods. Condor 1976;78:570–3.

60] Wingfield JC, Hegner RE, Dufty AMJ, Ball GF. The “chal-lenge hypothesis” : theoretical implications for patterns of testos-terone secretion, mating systems, and breeding strategies. Am Nat1990;136:829–46.

61] Wingfield JC, Jacobs JD, Soma K, Maney DL, Hunt K, Wisti-Peterson D, et al. Testosterone, aggression, and communication: eco-logical bases of endocrine phenomena. In: Konishi M, Hauser M, ed-itors. The Biology of Communication. Cambridge, MA: MIT Press;1999. p. 255–83.

62] Wingfield JC, Sapolsky RM. Reproduction and resistance to stress:when and how. J Neuroendocrinol 2003;15:711–24.

63] Wingfield JC, Silverin B. Ecophysiological studies of hormone-behaviour relations in birds. In: Pfaff DW, Arnold AP, Etgen AM,Fahrbach SE, Rubin RT, editors. Hormones, Brain and Behaviour.San Diego: Academic Press; 2002. p. 587–647.

64] Wingfield JC, Whaling CS, Marler P. Communication in vertebrateaggression and reproduction: the role of hormones. In: Knobil E,Neill JD, editors. The Physiology of Reproduction. New York: RavenPress; 1994. p. 303–42.