Molecular Psychiatry (2006) 11, 1106–1112 & 2006 Nature Publishing Group All rights reserved 1359-4184/06 $30.00 www.nature.com/mp ORIGINAL ARTICLE Serotonin transporter gene variation impacts innate processing: acoustic startle response and emotional startle B Brocke1, D Armbruster1,JMu¨ller1, T Hensch1, CP Jacob2, K-P Lesch2, C Kirschbaum3 and A Strobel1 1Personality and Individual Differences, Institute of Psychology II, Dresden University of Technology, Dresden, Germany; 2Molecular and Clinical Psychobiology, Department of Psychiatry and Psychotherapy, University of Wu¨rzburg, Wu¨rzburg, Germany and 3Institute of Psychology I, Dresden University of Technology, Dresden, Germany

Anxiety-related behaviors are closely linked to neural circuits relaying fear-specific informa- tion to the . Many of these circuits, like those underlying processing of innate fear, are remarkably well understood. Recent imaging studies have contributed to this knowledge by discriminating more detailed corticoamygdalar associations mediating processing fear and anxiety. However, little is known about the underlying molecular mechanisms. We used the acoustic startle paradigm to investigate the impact of molecular genetic variation of serotonergic function on the acoustic startle response and its fear potentiation. Startle magnitudes to noise bursts as measured with the eye blink response were recorded in 66 healthy volunteers under four conditions: presenting unpleasant and pleasant affective pictures as well as neutral pictures, and presenting the startle without additional stimuli as a baseline. Subjects were genotyped for functional polymorphism in the transcriptional control region of the serotonin transporter gene (5-hydroxytryptamine transporter gene-linked region: 5-HTTLPR). Analyses of variance revealed a significant effect of 5-HTTLPR on overall startle responses across conditions. Carriers of the short (s) allele exhibited stronger startle responses than l/l homozygotes. However, we could not confirm our hypothesis of enhanced fear potentiation of the startle in s allele carriers. In conclusion, the results provide first evidence that the startle response is sensitive to genetic variation in the serotonin pathway. Despite some issues remaining to be resolved, the startle paradigm may provide a valuable endophenotype of fear processing and underlying serotonergic influences. Molecular Psychiatry (2006) 11, 1106–1112. doi:10.1038/sj.mp.4001908; published online 10 October 2006 Keywords: acoustic startle response; amygdala; 5-HTTLPR; fear; serotonergic system

Introduction or in primitive cortex that are directly activated by primary reinforcement and are assumed to be compar- Anxiety is associated with a spectrum of behaviors, able in all mammalian species.1,2 Most of the including temperamental traits, conditioned and behaviors associated with the withdrawal system, innate fear, anxiety disorders and depression. Anxiety that is, anxiety-related behaviors, are closely linked to disorders account for an essential part of all the neural circuits relaying fear-specific information to conditions that psychotherapists and psychiatrists are the amygdala.3 However, most of the current know- consulted for. It is thus important that we understand ledge about the neuronal substrates and molecular as much as possible how anxiety-related neuro- mechanisms of fear processing is based on animal circuits work. research. The advent of the genetic variance approach Basic such as fear and anxiety can be seen and genomic imaging has rendered it possible to as prototypes of primitive systems that developed to elucidate molecular genetic factors and neuronal ensure the survival of organisms. These prototypical circuitries underlying fear processing also in healthy systems involve approach to stimuli to enhance or humans. maintain species or withdrawal from dangerous A recent functional imaging study has underscored events. The survival systems comprise neural struc- the importance of the amygdala for processing fear tures and their connections within subcortical areas and anxiety in humans.4 A series of subsequent imaging studies successively discriminated function- Correspondence: Dr B Brocke, Personality and Individual ally divergent components of a more extensive Differences, Institute of Psychology II, Dresden University of amygdala-relayed corticolimbic circuit for fear pro- Technology, Dresden 01062, Germany. cessing in humans.5–7 More specifically, using E-mail: [email protected] 7 Received 6 January 2006; revised 31 August 2006; accepted 1 functional connectivity analyses, Pezawas et al. September 2006; published online 10 October 2006 identified two distinct regions within the perigenual 5-HTTLPR and acoustic startle response B Brocke et al 1107 anterior cingulate cortex (pACC), which are function- reticularis pontis caudalis (PnC) and spinal moto- ally connected with the amygdala: the rostral and neurons, the excitation of which gives rise to the caudal subgenual ACC, which were positively and acoustic startle response (ASR). Because the startling negatively correlated with amygdala activity, respec- stimuli of the ASR are aversive and able to induce a tively. Furthermore, both regions showed strong state of fear or anxiety,11 the ASR can be regarded as a positive connectivity with each other, suggesting that means to study innate fear responses. they form a feedback loop. Furthermore, the ASR can be potentiated by With regard to the molecular mechanisms under- presenting the acoustic startle stimulus in the lying fear processing, the efficacy of serotonergic presence of a cue that had previously been paired compounds in the treatment of depression and most with an unconditioned fear stimulus. There is con- anxiety disorders suggests serotonergic neurotrans- verging evidence that this fear-potentiated startle mission being one of the most important molecular (FPS) is crucially modulated by the amygdala.12 In factors modulating anxiety-related behavior.8 In line models of this amygdala modulation of the FPS, fear with this assumption, the results of the mentioned stimuli proceed from sense receptor systems to the genomic imaging studies suggest that carriers of the sensory thalamus and then to the amygdala complex short (s) allele of a functional polymorphism in the and to the sensory cortex, which also gives rise to a transcriptional control region of the serotonin trans- projection to the amygdala.13 From the amygdala, porter gene (5-hydroxytryptamine transporter gene- there are three basic efferent connections: projections linked region: 5-HTTLPR),9 which impairs serotonin from the central amygdala to the lateral hypothalamic reuptake, exhibit greater amygdala activity in re- area (autonomic system), projections to the midbrain sponse to fearful or threatening facial expressions central gray (freezing and escape behavior) and to the than individuals homozygous for the long allele (l/l).4 PnC, which modulates the startle .1,14 It has been This finding suggests that increased anxiety in shown that lesions of the amygdala block the carriers of the 5-HTTLPR s allele9 might reflect a expression of FPS.12 Hence, FPS is a valid tool for hyper-activity or responsiveness of their amygdalae. the study of amygdala-modulated conditioned fear However, recent studies5–7,10 have offered further responses. insights into this hyper-activity revealing that the 5- Concerning the role of serotonin in the modulation HTTLPR s allele has strong impact on the corticolim- of ASR and FPS, there is evidence that in rats bic feedback circuit described above. Carriers of the systemic administration of agonists of the serotonin 5-HTTLPR s allele showed a significant reduction of 1A receptor [5-hydroxytryptamine (5-HT1A) receptor] amygdala–pACC functional connectivity as compared augments ASR amplitudes15–18 and reduces or even to l/l homozygotes. More specifically, the findings blocks FPS.16,19 In contrast, evidence from human suggest that the amygdala overactivation in s allele studies is scattered and in part contradictory: while in carriers reflect a failure of downregulation of the one study, the selective serotonin reuptake inhibitor amygdala response. As a consequence, the anxiety (SSRI) citalopram increased the ASR,20 other studies proneness of s allele carriers might not be seen as a employing the SSRIs fluvoxamine21 or sertraline22 simple overactivity of the amygdala complex, but as a did not find evidence for altered ASR amplitudes. basic dysfunction in fear processing. Unfortunately, emotional startle modulation was not In the light of these findings, differential amygdala examined in these studies. In an emotional startle reactivity can be regarded as a most revealing paradigm, citalopram abolished FPS on unpleasant endophenotype for the study of genetic variation of pictures, but did not alter the ASR on neutral the serotonin system. However, (1) the particular pictures.23 In two further studies, tryptophan supple- nature of the underlying molecular mechanisms is not mentation, which stimulates synthesis of serotonin well understood up to now and (2) this endopheno- was found to reduce the ASR, but not to significantly type presently is restricted to imaging results without FPS in healthy women,24 whereas tryptophan including those parts of amygdala-relayed circuits, depletion resulted in an overall increase of the ASR which are related to clearcut and valid behavioral across emotional conditions in a group of recovered phenotypes. Therefore, more comprehensive and depressed subjects, but not in healthy controls.25 In differentiated amygdala-related endophenotypes of contrast, another study did not find evidence fear processing including valid behavioral pheno- for altered ASR amplitudes following tryptophan types are needed. depletion.26 Here we used the acoustic startle paradigm, which Taken together, the impact of serotonin on startle allows the measurement of learned and innate fear response modulation remains to be determined. responses. In this paradigm, sudden high-intensity However, the well-established role of the amygdala noise bursts evoke a startle response, which can be in FPS and the genomic imaging findings outlined measured by electromyographic (EMG) recordings above provide a strong rationale for the hypothesis from the orbicularis oculi muscle. The neuronal that 5-HTTLPR s allele carriers that have been pathways mediating this reflex and forming the demonstrated to exhibit a stronger amygdala activa- primary acoustic startle circuit consist of the sense tion in response to fearful stimuli would also show a receptor systems and the auditory nerve, the cochlear stronger FPS in the emotional startle paradigm nucleus, the ventrolateral lemniscus, the nucleus employed in the present study. Furthermore, in order

Molecular Psychiatry 5-HTTLPR and acoustic startle response B Brocke et al 1108 to further elucidate the role of serotonin in startle used as filler stimuli and was not used for startle data modulation, we examined whether 5-HTTLPR influ- evaluation. The timing of the startle probes during ences the overall ASR across emotional conditions. picture viewing was balanced across content cate- gories. Pictures were organized such that not more than three pictures of the same affective valence and Methods not more than three pictures with the same startle Subjects onset time could occur consecutively. Otherwise, Participants were 62 female and 21 male students at stimulus order was pseudo-randomized. Finally, in the University of Dresden. Of these, 11 subjects were each block, three acoustic startle probes were deli- excluded during data preprocessing because of vered in the intertrial interval to measure the baseline excessive EMG artifacts or because of virtually no startle reaction and to further decrease the predict- startle responses. Six of the remaining 72 subjects had ability of the startle stimulus. to be excluded before statistical testing because of outliers (see below Statistical analysis), leaving 66 Affective rating subjects (17 male) with a mean age of 21.5 years Evaluative judgements of pleasure and arousal were (s.d. = 3.4, range 18–37 years) for the final sample. All measured using the Self-Assessment Manikin participants confirmed that they were without any (SAM).29 The SAM valence scale shows a graphical lifetime psychiatric or neurological diagnosis or representation of a figure with expressions ranging treatment. Participants were informed about the aims from happy to unhappy, and the SAM arousal scale of the study, consented in the procedure and were displays a graphic figure with expressions ranging either paid 15 euro or received course credits. The from calm and relaxed to excited. Ratings of valence study design was approved by the Ethics Committee and arousal were made on nine-point scales. of the German Psychological Association. Physiological data collection and reduction Materials and design The eye blink component of the startle response was In the startle paradigm used here, acoustic startle measured by recording EMG activity over the orbicu- probes were delivered alone and during viewing of laris oculi muscle beneath the left eye, using two Ag- emotional pictures. The startle stimulus consisted of a AgCl electrodes with 4 mm inner diameter. A ground 95 dB sound pressure level (SPL), 50 ms burst of white electrode was attached to the left mastoid. Impedance noise with an instantaneous rise time and was level was kept below 10 kO. The raw EMG signal was presented binaurally over Eartone A3 Audiometric amplified by a SynAmps amplifier (NeuroScan Inc., Insert Earphones (Aearo Company, Indianapolis, IN, El Paso, TX, USA), sampled at 1000 Hz, filtered USA). Pictorial stimuli consisted of 48 affective (30–200 Hz band pass), rectified and integrated. pictures. Forty color pictures, consisting of 12 Responses to startle probes were defined as EMG pleasant, 12 neutral and 16 unpleasant scenes, were peak in a time window from 20 to 140 ms after probe selected from the International Affective Picture presentation. Trials with excessive EMG artifacts were System27 on the basis of their affective valence and excluded. arousal ratings by the normative sample. Eight additional unpleasant black and white pictures Procedure displaying angry or fearful faces were drawn from a After providing informed consent, subjects reclined standard set of pictures of facial affect.28 The picture in a comfortable chair and the physiological sensors series comprised 12 different semantic contents, were attached. The subject was instructed that a series including three pleasant (attractive men, attractive of affective pictures would be presented and that each women, erotic couples), three neutral (household picture should be viewed for the entire time it was on objects, kitchen objects, mushrooms) and six unplea- the screen. In addition, the subject was told that sant (attacking humans, attacking animals, mutilated occasional noises heard over the earphones could be bodies, contamination, angry faces, fearful faces). ignored. Then a series of 48 pictures was presented Each of the 12 contents included four different for 6 s each. Between each picture, the computer exemplars. Digitized versions of the pictures were screen displayed a fixation cross for a randomly displayed on a 17-inch computer screen at a distance generated variable interval, ranging from 11 to 24 s, of approximately 1.5 m from the subject’s head. Each in order to clear any associated with the picture was presented for 6 s and the pictures were previous image. After the picture series was finished, arranged in four blocks of 12 pictures. Each block the sensors were removed and subjects were famil- consisted of six unpleasant, three neutral and three iarized with the SAM rating procedure. All pictures pleasant content pictures. On nine of these 12 trials, were presented a second time in the same order. an acoustic startle probe was delivered during picture Subjects were told to view each picture as long as they viewing. Three of the four picture trials for each needed to make their ratings for valence and arousal, content category were accompanied by an acoustic and then to press the button to turn off the picture and startle probe, which was administered at 0.5, 1.5 or turn on the ratings. After picture offset, participants 2.5 s after picture onset. One picture in each content rated their subjective experience of valence and category was presented without a startle probe and arousal, using the computerized version of the SAM

Molecular Psychiatry 5-HTTLPR and acoustic startle response B Brocke et al 1109 rating method.29 Finally, buccal cells for DNA extrac- covariates and 5-HTTLPR genotype (age: ANOVA, tion were collected using Catch-All Sample Collec- P = 0.475; gender: w2-test, P = 0.567). tion Swabs provided with the BuccalAmp DNA Extraction Kit (Epicentre, Madison, WI, USA). Parti- cipants were subsequently debriefed, given credit and Results thanked. Genotype frequencies The percentages of the 5-HTTLPR genotypes were Genotyping 47% (n = 31) for l/l, 44% (n = 29) for s/l and 9% (n =6) For genotyping, DNA was isolated from buccal cells for s/s. The genotypes were in Hardy–Weinberg using the BuccalAmp DNA Extraction kits and equilibrium (w2 = 0.05, P = 0.833). protocol. 5-HTTLPR genotypes were determined by polymerase chain reaction as described by Lesch et Affective ratings al.,9 followed by agarose gel size fractionation. Alleles As the majority of valence and arousal ratings were were designated according to their relative size: s (14 not normally distributed (Kolmogorov–Smirnov tests, repeats) and l (16 repeats). For statistical testing, s P < 0.20), the medians of the valence and arousal allele carriers (l/s and ss genotypes = S group; N = 35, ratings for the different picture categories were eight male, age mean 21.274.0 years) were compared compared with each other using the non-parametric to l/l carriers (L group; N = 31, nine male, age mean Wilcoxon tests for paired samples. The median 21.872.6 years). valence ratings for unpleasant, neutral and pleasant pictures were 2.5, 5.0 and 6.75, respectively, and the median arousal ratings were 4.5, 1.0 and 4.0, Statistical analysis respectively. All two-way comparisons for valence, All analyses were performed using SPSS for Windows and for arousal, respectively, were highly significant 12.0.1 (SPSS Inc., Chicago, IL, USA). In the sample of (all Pp0.003). 5-HTTLPR genotype groups did not the 72 subjects who passed data preprocessing, the 48 differ in valence and arousal ratings, although the startle variables (18 for unpleasant, nine for neutral, S group showed a tendency toward lower arousal nine for pleasant and 12 for baseline startle condition) ratings for unpleasant pictures (non-parametric U- were log-transformed because of the highly skewed tests, P = 0.085; all other PX0.414). distribution of the raw startle variables and the resulting deviation from the normal distribution 5-HTTLPR impact on acoustic startle and emotional (Kolmogorov–Smirnov tests, P < 0.20). Subsequently, startle response modulation the log-transformed variables were examined for ANOVA showed there were no significant main and outliers, especially because after log transformation, interaction effects of the covariates age and gender still 31% of the variables showed deviation from the (all PX0.124). There was a significant condition main 2 normal distribution. Outliers were identified using effect (F2.1,127.2 = 5.43, P = 0.005, Z = 0.08). Within- boxplots, which are well suited for the descriptive subjects contrast analyses revealed that the presenta- analysis of non-normally distributed variables. In tion of pleasant pictures resulted in significant boxplots, identification of outliers is based on the pleasure attenuation of the startle (PAS; pleasant vs 2 interquartile range (IQR, an equivalent of the s.d.). neutral condition: F1,62 = 7.39, P = 0.009, Z = 0.11), In SPSS and other statistical packages like SAS or whereas the presentation of unpleasant affective SYSTAT, outliers are defined as values below or above pictures did not result in significant fear potentiation

1.5*IQR from the 25th and the 75th percentile, of the startle (FPS; unpleasant vs neutral: F1,62 = 2.17, respectively. Subjects were selected for exclusion P = 0.146, Z2 = 0.03). Compared to baseline, the startle from further analyses if they accounted for at least response was marginally higher in the unpleasant 2 5% of all outliers. Six subjects (three L and three S condition (F1,62 = 3.26, P = 0.076, Z = 0.05), but in- subjects with very low startle magnitudes) were significantly enhanced in the neutral condition 2 identified who met this criterion and who jointly (F1,62 = 0.78, P = 0.380, Z = 0.01). accounted for over 90% of all the outliers. After Furthermore, there were no genotype-specific dif- exclusion of these subjects, all 48 log-transformed ferences in FPS or PAS as indicated by the absence of startle variables were normally distributed (Kolmo- a significant condition  5-HTTLPR interaction effect 2 gorov–Smirnov tests, PX0.30). For the remaining 66 (F2.1,127.2 = 1.14, P = 0.324, Z = 0.02). However, there subjects, the average startle magnitudes in the four was a significant 5-HTTLPR main effect on average conditions (baseline, unpleasant, neutral, pleasant) startle magnitudes across conditions (F1,62 = 4.28, were computed, were tested for univariate normality P = 0.043, Z2 = 0.07) with the S group showing higher (Kolmogorov–Smirnov tests, PX0.568) and were then overall startle magnitudes. Table 1 presents the exact entered into a repeated measures analysis of variance values of the mean startle magnitudes and standard (ANOVA) with condition as a within-subjects factor, errors of means and gives the results of the univariate 5-HTTLPR genotype as a between-subjects factor and comparisons of the 5-HTTLPR genotype groups age and gender as covariates. Greenhouse–Geisser separately for the four conditions. Figure 1 illustrates corrected degrees of freedom were used where the genotype differences both at the level of appropriate. There were no associations between the each individual’s startle magnitudes in the L and

Molecular Psychiatry 5-HTTLPR and acoustic startle response B Brocke et al 1110 Table 1 Mean startle magnitudes (s.e.m.) for total sample Discussion and 5-HTTLPR genotype groups Our genetic variance approach demonstrates that Sample N Condition differential serotonin transporter function influences the overall startle response across valence conditions. Baseline Unpleasant Neutral Pleasant Carriers of the 5-HTTLPR s allele exhibited stronger startle responses than l/l homozygotes. However, we Total 66 1.02 (0.07) 1.14 (0.07) 1.13 (0.07) 0.79 (0.09) could not confirm or test our hypothesis that s allele L 31 0.89 (0.10) 0.99 (0.10) 1.00 (0.10) 0.61 (0.13) carriers show a stronger enhancement of the startle S 35 1.14 (0.09) 1.29 (0.10) 1.26 (0.10) 0.97 (0.12) response during the presentation of unpleasant LvsSa P = 0.068 P = 0.035 P = 0.069 P = 0.040 pictures (FPS), because there was no significant fear potentiation in our sample. That means, albeit s allele Abbreviations: L = l/l genotype; S = l/s and s/s genotype. carriers showed stronger startle responses in all aDetermined using univariate analysis of variance with age valence conditions, they showed the same pattern of and gender as covariates. emotional startle response modulation as l/l homo- zygotes. This leaves the crucial question unanswered to which extent the observed differences in the startle response are due to 5-HTTLPR influences at the level of the amygdala. A more straightforward explanation for higher startle responses of s allele carriers might be seen in an influence of 5-HTTLPR on serotonergic function at the level of the or even at the level of motoneurons.30 It is yet an open question whether the amygdala complex is not only involved in the emotional modulation of the startle response but also in the processing of startling stimuli per se. If the response signal within the usual time window (20–140 ms) integrates several serial projections from different brain nuclei to the PnC,31 an early involvement of the amygdala complex in the processing of startling stimuli seems possible.13 While in the model of a primary acoustic startle circuit, the PnC is the most important brainstem site for the evocation of the ASR, other brain nuclei than the PnC also play a role in mediating the ASR.32,33 The startling stimuli of the Figure 1 5-HTTLPR and the acoustic startle response. (a, b) ASR themselves are aversive and can induce a state of Boxplots of log startle magnitudes in the four conditions in fear or anxiety.11 Therefore and with regard to the the L group (a, N = 31) and in the S group (b, N = 35), finding that field potentials specifically related to the with gray lines indicating each individual’s profile of 34 startle responses; the bars within the box give the median, ASR were recorded in the basolateral amygdala, it is the boxes’ length reflect the interquartile ranges and the suggestive that the amygdala is to some extent also whiskers indicate the range of observations; (c) log startle involved in the modulation of the ASR. This view is magnitudes and standard errors of means in the four supported by the finding that the excitatory post- conditions stratified for 5-HTTLPR genotype groups. synaptic potentials (EPSP) recorded intracellularly from PnC neurons show multiple peaks that occur at constant latencies,31 which suggests excitatory input the S group (Figure 1a and b) and at the level of the from multiple afferents. Moreover, although amygdala group comparison (Figure 1c). lesions have been demonstrated to block FPS,35 there Finally, to exclude the possibility that these results is also albeit inconclusive evidence for an impaired are due to possible influences of affective ratings and overall startle response following amygdala le- especially the S group’s tendentially lower arousal sion.36,37 Hence, we cannot rule out that 5-HTTLPR ratings for negative pictures, confounding effects of genotype-specific amygdalar processing of startling these variables were examined by residualizing stimuli and affective pictures influenced the startle the startle magnitudes for valence and arousal response differences observed in the present study. ratings (dummy-coded) using linear regression. If, as recent genomic imaging findings suggest, 5- When entering the residualized values into an HTTLPR s allele carriers exhibit a stronger amygdala analysis of variance with only the picture conditions activity during presentation of fearful or threatening (because for baseline startle, no affective ratings stimuli,4–7 this stronger amygdala responsiveness to were available), the results essentially remained the the threatenting startle stimuli may lead to a stronger same (5-HTTLPR Â condition interaction: P = 0.385; amygdala innervation of the PnC, and hence to a 5-HTTLPR main effect: P = 0.049). higher startle response.

Molecular Psychiatry 5-HTTLPR and acoustic startle response B Brocke et al 1111 A further result of the study is the finding that Acknowledgments while startle responses of the total group were attenuated when processing pleasant affective pic- We thank N Steigerwald for excellent technical tures, they were not potentiated in the unpleasant assistance in DNA sample processing and genotyping condition as compared to the neutral condition. This and U Buhss for excellent work in processing and is in part contrary to our expectations, but this result analyzing the EMG data. This work was supported by as well as the enhancement of the startle response the Deutsche Forschungsgemeinschaft (KI 537/20-2 even in the neutral condition compared to baseline is and KFO 125/1-1 to CPJ and KPL). to some extent compatible with the usual findings.38–40 It has been observed that different sub-categories of References neutral pictures produce very heterogeneous effects on the magnitude of the startle response, sometimes 1 Bradley MM, Lang PJ. Measuring emotion: behavior, feeling, and surmounting the magnitude of the startle potentiation physiology. In: Lane R, Nadel W (eds). Cognitive Neuroscience of 40 Emotion. Oxford University Press: New York, 2000, pp 242–276. through categories of unpleasant pictures. Likewise, 2 Konorski J. Integrative Activity of the Brain: An Interdisciplinary in the present study a much weaker startle response Approach. University of Chicago Press: Chicago, 1967. was observed following viewing kitchen objects than 3 Davis M, Whalen PJ. The amygdala: vigilance and emotion. following viewing other neutral pictures. It may be Mol Psychiatry 2001; 6: 13–34. assumed that the processing of the neutral pictures 4 Hariri AR, Mattay VS, Tessitore A, Kolachana B, Fera F, Goldman D, Egan MF et al. Serotonin transporter genetic variation and the gives rise to additional projections from the associa- response of the human amygdala. Science 2002; 297: 400–403. tive cortex to the amygdala (nucleus centralis) and 5 Canli T, Omura K, Haas BW, Fallgatter A, Constable RT, Lesch KP. then to the PnC.3 Nevertheless, the result that the Beyond affect: a role for genetic variation of the serotonin enhancement of the response in the neutral condition transporter in neural activation during a cognitive attention task. Proc Natl Acad Sci USA 2005; 102: 12224–12229. does not differ in height from that in the unpleasant 6 Heinz A, Braus DF, Smolka MN, Wrase J, Puls I, Hermann D et al. condition is not in line with the ‘classical’ findings. Amygdala–prefrontal coupling depends on a genetic variation of Future studies on emotional startle modulation the serotonin transporter. Nat Neurosci 2005; 8: 20–21. should therefore pay special attention to picture 7 Pezawas L, Meyer-Lindenberg A, Drabant BA, Verchinski BA, selection. Munoz KE, Kolachana BS et al. 5-HTTLPR polymorphism impacts human cingulate–amygdala interactions: a genetic susceptibility Taken together, the startle response proved to be mechanism for depression. Nat Neurosci 2005; 8: 828–834. sensitive for genetic variation of serotonergic function 8 Nemeroff CB, Owens MJ. Treatment of mood disorders. Nat and hence provides a valuable endophenotype of fear Neurosci 2002; 5(Suppl): 1068–1070. processing and underlying serotonergic influences. 9 Lesch K-P, Bengel D, Heils A, Sabol SZ, Greenberg BD, Petri S et al. Association of anxiety-related traits with a polymorphism in the Taking into account earlier findings on the impact of serotonin transporter gene regulatory region. Science 1996; 274: 5-HTTLPR on brain function, it can be subsumed that 1527–1539. 5-HTTLPR differentially modulates multiple stages of 10 Hariri AR, Drabant BA, Munoz KE, Kolachana BS, Mattay VS, Egan information processing, ranging from (1) fast MF et al. A susceptibility gene for affective disorders and the as demonstrated here and (2) early sensory processing response of the human amygdala. Arch Gen Psychiatry 2005; 62: 146–152. as shown in recent event-related potential (ERP) 11 Leaton RN, Cranney J. Potentiation of the acoustic startle response 41–43 studies over (3) short- and middle-latency stages by conditioned stimulus paired with acoustic startle stimulus in of more conscious emotional processing as observed rats. J Exp Psychol Anim Behav Process 1990; 16: 279–287. in the imaging studies mentioned above, to (4) later 12 Davis M, Falls WA, Campeau S, Kim M. Fear-potentiated startle: a neural and pharmacological analysis. Behav Brain Res 1993; 58: processes of cognitive response control and error 175–198. processing as evidenced in the ERP studies of 13 LeDoux JE. Emotion circuits in the brain. Annu Rev Neurosci Fallgatter et al.,44,45 who were the first to report an 2000; 23: 155–184. association between 5-HTTLPR and prefrontal cor- 14 Davis M. Neurobiology of fear responses: the role of the amygdala. tex–limbic excitability in a Go-NoGo task and an J Neuropsychiatry Clin Neurosci 1997; 9: 382–402 (review). 15 Svensson L, Ahlenius S. Enhancement by the putative 5-HT error-processing task. As different kinds of the startle receptor agonist 8-OH-2-(di-n-propylamino)tetralin of the acoustic paradigm afford to examine information processing at startle response in the rat. Psychopharmacology (Berlin) 1983; 79: several stages, the startle paradigm can be seen as 104–107. an integrative tool to further our understanding of 16 Kehne JH, Cassella JV, Davis M. Anxiolytic effects of buspirone and gepirone in the fear-potentiated startle paradigm. Psycho- the serotonergic influences on complex behavior. pharmacology (Berlin) 1988; 94: 8–13. Future research into the role of 5-HTTLPR in startle 17 Nanry KP, Tilson HA. The role of 5HT1A receptors in the modulation should be supplemented by imaging modulation of the acoustic startle reflex in rats. Psychopharma- techniques,46 which may result in a more specific cology (Berlin) 1989; 97: 507–513. elucidation of the serotonergic modulation of differ- 18 McQueen DA, Overstreet DH, Ardayfio PA, Commissaris RL. Acoustic startle, conditioned startle potentiation and the effects of ent kinds of fear-processing and related neuronal 8-OH-DPAT and busprione in rats selectively bred for differences circuits and behavioral phenotypes. Conversely, in 8-OH-DPAT-induced hypothermia. Behav Pharmacol 2001; 12: genomic imaging data on neuronal circuits under- 509–516. lying anxiety processing might be validated by 19 Mansbach RS, Geyer MA. Blockade of potentiated startle respond- ing in rats by 5-hydroxytryptamine1A receptor ligands. Eur J behavioral data of the startle paradigm and be related Pharmacol 1988; 156: 375–383. to complementary neuronal circuits, which are 20 Liechti ME, Geyer MA, Hell D, Vollenweider FX. Effects of MDMA remarkably well understood. (ecstasy) on prepulse inhibition and habituation of startle in

Molecular Psychiatry 5-HTTLPR and acoustic startle response B Brocke et al 1112 humans after pretreatment with citalopram, haloperidol, or 33 Koch M. The neurobiology of startle. Prog Neurobiol 1999; 59: ketanserin. Neuropsychopharmacology 2001; 24: 240–252. 107–128. 21 Phillips MA, Langley RW, Bradshaw CM, Szabadi E. The effects of 34 Ebert U, Koch M. Acoustic startle-evoked potentials in the rat some antidepressant drugs on prepulse inhibition of the acoustic amygdala: effect of kindling. Physiol Behav 1997; 62: 557–562. startle (eyeblink) response and the N1/P2 auditory evoked 35 Funayama ES, Grillon C, Davis M, Phelps EA. A double response in man. J Psychopharmacol 2000; 14: 40–45. dissociation in the affective modulation of startle in humans: 22 Quednow BB, Kuhn KU, Stelzenmueller R, Hoenig K, Maier W, effects of unilateral temporal lobectomy. J Cogn Neurosci 2001; 13: Wagner M. Effects of serotonergic and noradrenergic anti- 721–729. depressants on auditory startle response in patients with 36 Angrilli A, Mauri A, Palomba D, Flor H, Birbaumer N, Sartori G, major depression. Psychopharmacology (Berlin) 2004; 175: di Paola F. Startle reflex and emotion modulation impairment 399–406. after a right amygdala lesion. Brain 1996; 119: 1991–2000. 23 Harmer CJ, Shelley NC, Cowen PJ, Goodwin GM. Increased 37 Kettle JW, Andrewes DG, Allen NB. Lateralization of the startle positive versus negative affective perception and memory in reflex circuit in humans: an examination with monaural probes healthy volunteers following selective serotonin and norepinephr- following unilateral temporal lobe resection. Behav Neurosci ine reuptake inhibition. Am J Psychiatry 2004; 161: 1256–1263. 2006; 120: 24–39. 24 Murphy SE, Longhitano C, Ayres RE, Cowen PJ, Harmer CJ. 38 Vrana SR, Spence EL, Lang PJ. The startle probe response: a new Tryptophan supplementation induces a positive bias in the measure of emotion? J Abnorm Psychol 1988; 97: 487–491. processing of emotional material in healthy female volunteers. 39 Lang PJ, Bradley MM, Cuthbert BN. Emotion, attention, and the Psychopharmacology (Berlin) 2006; 187: 121–130. startle reflex. Psychol Rev 1990; 97: 377–395. 25 Hayward G, Goodwin GM, Cowen PJ, Harmer CJ. Low-dose 40 Bradley MM, Codispoti M, Cuthbert BN, Lang PJ. Emotion and tryptophan depletion in recovered depressed patients induces motivation I: defensive and appetitive reactions in picture changes in cognitive processing without depressive symptoms. processing. Emotion 2001; 1: 276–298. Biol Psychiatry 2005; 57: 517–524. 41 Gallinat J, Senkowski D, Wernicke C, Juckel G, Becker I, Sander T 26 Phillips MA, Oxtoby EK, Langley RW, Bradshaw CM, Szabadi E. et al. Allelic variants of the functional promoter polymorphism of Effects of acute tryptophan depletion on prepulse inhibition the human serotonin transporter gene is associated with auditory of the acoustic startle (eyeblink) response and the N1/P2 cortical stimulus processing. Neuropsychopharmacology 2003; 28: auditory evoked response in man. J Psychopharmacol 2000; 14: 530–532. 258–265. 42 Strobel A, Debener S, Schmidt D, Hunnerkopf R, Lesch KP, Brocke 27 Lang PJ, Bradley MM, Cuthbert BN. International Affective Picture B. Allelic variation in serotonin transporter function associated System (IAPS): Instruction Manual and Affective Ratings. Techni- with the intensity dependence of the auditory evoked potential. cal Report A-4. Center for Research in Psychophysiology, Am J Med Genet B 2003; 118: 41–47. University of Florida: Gainesville, FL, 1999. 43 Hensch T, Wargelius HL, Herold U, Lesch KP, Oreland L, Brocke B. 28 Ekman P, Friesen WV. Pictures of Facial Affect. Consulting (in press) Further evidence for an association of 5-HTTLPR Psychologists Press: Palo Alto, CA, 1976. with intensity dependence of auditory-evoked potentials. Neuro- 29 Lang PJ. Behavioral treatment and bio-behavioral assessment: psychopharmacology 2006; 18: 2047–2054. computer applications. In: Sidowski JB, Johnson JH, Williams TA 44 Fallgatter A, Jatzke S, Bartsch A, Hamelbeck B, Lesch KP. (eds). Technology in Mental Health Care Delivery Systems. Ablex: Serotonin transporter promoter polymorphism influences topo- Norwood, NJ, 1980, pp 119–137. graphy of inhibitory motor control. Int J Neuropsychopharmoco- 30 Rekling JC, Funk GD, Bayliss DA, Dong XW, Feldman JL. Synaptic logy 1999; 2: 115–120. control of motoneuronal excitability. Physiol Rev 2000; 80: 45 Fallgatter AJ, Herrmann MJ, Roemmler J, Ehlis AC, Wagener A, 767–852. Heidrich A et al. Allelic variation of serotonin transporter function 31 Lingenho¨hl K, Friauf E. Giant neurons in the rat reticular modulates the brain electrical response for error processing. formation: a sensorimotor interface in the elementary acoustic Neuropsychopharmacology 2004; 29: 1506–1511. startle circuit? J Neurosci 1994; 14: 1176–1194. 46 Anders S, Lotze M, Erb M, Grodd W, Birbaumer N. Brain activity 32 Yeomans JS, Frankland PW. The acoustic startle reflex: neurons underlying emotional valence and arousal: a response-related and connections. Brain Res Brain Res Rev 1996; 21: 301–314. fMRI study. Hum Brain Mapp 2004; 23: 200–209.

Molecular Psychiatry