Поиск по сайту

Пишите нам: info@ethology.ru

Follow etholog on Twitter

Система Orphus

Кросс-культурный метод
Старые форумы
Не в тему

список статей

Provisional evidence that the arginine vasopressin 1a receptor gene is associated with musical memory

Roni Y. Granota, Yoav Frankelb, Vladimir Gritsenkoc, Elad Lererd, Inga Gritsenkoc, Rachel Bachner-Melmane, Salomon Israelb, Richard P. Ebsteince

1. Introduction

2. Methods

3. Results

4. Discussion



1. Introduction

In a seminal study, we demonstrated an epistatic association between two genes, the arginine vasopressin 1a receptor (AVPR1a) and the serotonin transporter (SLC6A4) HTTLPR promoter region repeats, and creative dance (Bachner et al., 2005). An association was also observed between these two genes and scores on the reward dependence factor in Cloninger's Tridimensional Personality Questionnaire, a measure of need for social contact and openness to communication. Music, like dance, serves as a social and expressive form of communication cross-culturally and shares with dance rhythmic and motor components as well as possible common evolutionary roots related to courtship displays or group bonding (Wallin, Merker, & Brown, 2001).

Arginine vasopressin (AVP) and its structurally related neuropeptide oxytocin (OXT) are released peripherally and centrally, the latter component through several structures, including parvocellular neurons in the paraventricular nucleus, medial amygdala, and the bed nucleus of the stria terminalis. Both neuropeptides have been found to be involved in social behaviors such as regulation of maternal behavior, male courtship, territorial defense, and paternal care in diverse species (Insel & Young, 2000). Importantly, they have also been found to be involved in processing and memory for social cues, including facial features, scent, and voice (Ferguson, Young, & Insel, 2002), as well as in social bonding as seen for example in the prairie vole (Hammock & Young, 2005). Indeed, transgenic mice with a null mutation in the AVPR1a gene lack the ability to recognize familiar conspecifics despite repeated exposure and despite normal general olfactory activity, normal spatial memory, and normal sensorimotor processing (Bielsky, Hu, Szegda, Westphal, & Young, 2004). These two neuropeptides have also been implicated in sensorimotor processes underlying acoustic communication in a number of species ranging from the teleost fish (Goodson & Bass, 2000) to the mustached bat (Prasada-Rao & Kanwal, 2004). Injection of OXT to female hamsters, for example, has been found to increase ultrasonic vocalizations presumably aimed at the establishment and maintenance of sociosexual relationships (Floody, Cooper, & Albers, 1998).

Depue and Morrone-Strupinsky (2005) outlined a neurobehavioral model of bonding in humans in which vasopressin and OXT serve an important role in perceptual and attentional processing of and memory for affiliative stimuli. Consistent with this model, intranasal administration of AVP in humans was found to increase agonistic responses in men and affiliative responses in women to same-sex models (Thompson, George, Walton, Orr, & Benson, 2006). In addition, higher levels of OXT were reported to be correlated with greater openness to social interactions, less aggression, and more trust in relation to age-matched control subjects (Kosfeld, Heinrichs, Zak, Fischbacher, & Fehr, 2005). Moreover, autistic children typically deficient in interpreting social cues have been shown to have lower OXT plasma levels (Modahl et al., 1998). A clear relationship has been established between AVP and memory, with a large number of studies reporting on its influence on spatial and social memory (Aarde & Jentsch, 2006). In contrast, its influence on verbal memory remains to be controversial (Perras, Droste, Born, Fehm, & Pietrowsky, 1997).

One mechanism through which OXT and AVP receptors may exert influence on social behavior is through interaction with dopaminergic neurotransmission in reward pathways (Depue & Morrone-Strupinsky, 2005), such as the nucleus accumbens and ventral pallidum—areas found to be active in humans in functional magnetic resonance imaging studies using various types of stimuli beyond classical rewarding stimuli, including pictures of people whom participants are in love with, beautiful faces, own baby's cry, and, most significantly for our study, music (Menon & Levitin, 2005). Another important neurotransmitter that has been found to interact with AVP in the hypothalamus to control communicative behavior is serotonin (Ferris et al., 2001, Galfi et al., 2005). The effects of serotonin (5-HT) on AVP and OXT secretion were studied in 13- to 14-day cultures of isolated rat neurohypophysial tissue. The AVP and OXT contents of the supernatant were determined by radioimmunoassay after 1 or 2 h of incubation. Significantly increased levels of AVP and OXT production were detected in the tissue culture media after 5-HT administration, depending on the 5-HT dose. These and other studies suggest that AVPR1a and SLC6A4 may also interact.

In contrast to the growing body of information regarding brain function and the brain anatomy of music processing, the underlying neurochemistry is largely unknown. Only a small number of studies have looked into levels of dopamine, adrenocorticotropic hormone, cortisol, prolactin, growth hormone, and β-endorphins as a function of emotional responses to music (e.g., Gerra et al., 1998). The relationship between music and serotonin was examined in only one study that suggested that unpleasant music may lead to changes in serotonin levels, possibly through stress induction (Evers & Suhr, 2000). None of these studies looked into the relationship between music and AVP, although OXT has been postulated to be an important neurochemical underlying the ability attributed to music to establish group bonding (Freeman, 2001).

Given the clear relationship between music and dance as well as the important role of AVP in social and sexual behaviors and in memory, we examined through an exploratory study whether the AVPR1a and SLC6A4 genes are also associated with musical memory by testing a group of participants on a battery of musical memory tests. Phonological working memory tests served as a basis for comparison. The reader is referred to two review articles that discuss for nonspecialists the main strategies used in establishing associations between genes and complex human traits (Balding, 2006, Lewis, 2002). Population-based studies simply examine the dependent variable (in this study, scores on memory tests) grouped by genotype (independent variable) and test for significance. A possible conundrum in population-based studies is the presence of subpopulations of subjects who are characterized by different allele frequencies. Their presence would confound the results, and any association may therefore be caused by these hidden groups and not by a true association between genes and phenotypes (Hamer & Sirota, 2000). Family-based association tests (Spielman, McGinnis, & Ewens, 1993) that test the transmission of alleles to the affected offspring (the trait can be either categorical or continuously distributed) from a heterozygote parent were designed to resolve this problem. For a biallelic marker Aa, A and a are equally transmitted to the offspring, giving a 50:50 ratio. However, if subjects are chosen for a particular phenotype and the marker is contributing to individual differences in this phenotype (the sample is biased by selecting offspring with a certain phenotype), it is expected that there will be a deviation from the 50:50 ratio. This test was extended to multiple markers such as the AVPR1a repeats by Sham and Curtis (1995). The analysis methods remain to be similar, but problems arise when rare alleles lead to sparse contingency tables that cannot be analyzed by χ2 statistics. More sophisticated statistical methods then need to be used, such as Monte Carlo simulation methods.

2. Methods

Eighty-two university students (21 males and 61 females) with little or no musical training participated in the study. The study was approved by the hospital institutional review board. All participants were fully informed of the study's procedure, provided their signed consent, and were paid for participating.

Participants were administered an extensive battery of tests (total duration, ∼2.5 h) using an innovative take-home CD-ROM. In all, there were four tests of melodic memory, two tests of rhythmic memory, and phonological tests: the interval subtest from the Montreal Battery of Evaluation of Amusia (MBEA), which tests for sensitivity to interval size while contour and key are preserved (Peretz, Champod, & Hyde, 2003); the melodic imagery subtest from Gordon's Musical Aptitude Test, which requires participants to judge whether a musical response is a variant of a reference musical phrase (Gordon, 1965); the tonal memory test from Seashore Musical Talents (Seashore, Lewis, & Saetveit, 1960), in which participants are required to keep track of the order of pitches and indicate in which of those has a change been inserted within the comparison pattern; an adapted version of the Distorted Tunes Test (DTT), which tests for sensitivity to changes inserted in well-known songs, found to be clearly related to genetic factors (Drayna, Manichaikul, De Lange, Snieder, & Spector, 2001); the rhythmic memory test from Seashore Musical Talents, which includes only rhythmic information (i.e., drumbeats); and the rhythmic imagery test from Gordon's Musical Aptitude Test, which requires participants to compare pairs of melodic phrases in which melodic information is retained but rhythm may change. The phonological tests included a newly devised test in which participants indicate whether two strings of increasing length in a foreign language (Arabic, which has no direct semantic representation for Hebrew speakers) are the same or different (differing items include a change of vowel; e.g., “kul_kitab—kul_kitib”) and a nonword test in which participants were asked to repeat auditorily presented lists of increasing length of phonologically legal nonwords.

All participants (and their parents) were genotyped for the AVPR1a and the SCL6A4 promoter region repeats.

Scores in most tests were based on percentages of correct responses. In the DTT, only melodies familiar to a participant were scored, unless the deviating tone or tones strongly violated musical syntax, in which case unfamiliar melodies were also included in the score. The score in the phonological tests was based on the number of correctly recalled nonwords, regardless of their temporal order. Because of technical difficulties in the recording process, some of the recorded responses in the nonword test were not clear and hence the number of participants for this test was 65, as compared with 82 in the remaining tests.

Variations in the selected candidate genes (i.e., alleles) were measured and tested for an association with memory scores. Genotyping and statistical methods are described in detail in our previous investigation (Bachner et al., 2005, Bachner-Melman et al., 2005, Modahl et al., 1998). Scores on the four tests of musical memory were adjusted by regressing scores on years of training and testing for a genetic association with the saved residuals (SPSS) to control for the confounding effects of individual differences in musical training. Test scores were grouped into top third and bottom third bins to implement the conditional analysis (that requires categorical phenotypes); the top third scores were designated as affected, whereas the bottom two thirds were designated as nonaffected.

3. Results

As found in previous studies (for a review, see Shuter-Dyson, 1999), scores on some of the music tests were moderately intercorrelated (Table 1), supporting our assumption that the various tests tapped into somewhat different aspects of musical memory. Consistent with recent studies on the relationship between musical and phonological abilities (e.g., Ho, Cheung, & Chan, 2003), a correlation was found between two musical tests (Gordon rhythm) and the MBEA and the phonological test.

Table 1.

Pearson's correlations between various music tests and between musical and nonmusical memory tests (n=82)a

Gordon Melody Seashore Rhythm Seashore Melody MBEA DTT Phonological Nonword
Gordon Rhythm 0.46 (<.001) 0.46 (<.001) 0.45 (<.001) 0.55 (<.001) 0.35 (.001) 0.23 (.031) NS
Seashore Melody 0.47 (<.001) 0.43 (<.001) 0.48 (<.001) 0.41 (<.001) NS NS
MBEA NS 0.38 (<.001) 0.48 (<.001) 0.38 (<.001) 0.35 (.001) NS

In the phonological test, n=65.

We next investigated single-locus associations between AVPR1a (RS1 and RS3) and SLC6A4 (HTTLPR) repeat regions with musical as well as nonmusical memory using population-based (grouping test scores by allele) and family-based (testing for biased transmission of alleles from a heterozygous parent toward explaining individual differences in test scores) methods as applied in the UNPHASED program (Dudbridge, 2003, Koeleman et al., 2000). As shown in Table 2, a trend (not significant after correction: pT=trend for multiple testing 8 phenotypes×3 markers=24) for a single-locus association was observed between SLC6A4 (HTTLPR) and AVPR1a (RS1 and RS3) promoter repeats with scores on one or more of the musical and phonological memory tests: Gordon Melody and RS3; Gordon Rhythm and RS3; or MBEA and RS3. A two-locus haplotype analysis that used genetic information from both AVPR1a promoter repeats (RS1 and RS3) also showed a trend for association with scores on the Gordon Melody, MBEA, Seashore Rhythm, and the phonological tests. However, only the association with the scores of the MBEA and Seashore Rhythm was significant after Bonferroni correction (pB=pT×24) for multiple testing.

Table 2.

Association between AVPR1a and SLC6A4 promoter region repeats and scores on musical as well as nonmusical tests

Test Single-locus association Two-locus association
HTTLPR RS1 RS3 Haplotypes RS1 and RS3 Haplotype conditional on HTTLPR
Gordon Melody
QTPHASE (population based) results
LRS 0.04 8.84 21.62 52.71 89.69
df 1 7 11 30 47
pT .834 .264 .027 .006a .0002a (pB=.0048)
ETDT (family based) results
LRS 0.56 5.92 14.4 39.02 55.49
df 1 6 10 27 36
pT .452 .431 .155 .063 .019
Gordon Rhythm
QTPHASE (population based) results
LRS 0.26 0.97 24.56 39.28 76.29
df 1 7 11 30 45
pT .605 .995 .010 .119a .002a (pB=.048)
ETDT (family based) results
LRS 3.989 9.91 20.87 17.37 21.40
df 1 6 11 11 9
pT .045 (L) .128 .034 .097 .010
QTPHASE (population based) results
LRS 0.02 8.50 11.99 49.57 103.34
df 1 7 11 24 47
pT .878 .289 .364 .002a (pB=.048) 4.13e−006a (pB=.000099)
ETDT (family based) results
LRS 3.98 9.91 20.87 17.37 74.80
df 1 6 11 11 44
pT .045 (L) .128 .034 .097 .003a (pB=.072)
Seashore Melody
QTPHASE (population based) results
LRS 0.021 6.29 5.14 34.45 61.55
df 1 6 11 28 45
pT .882 .391 .176 .186 .051
ETDT (family based) results
LRS 4.32 3.052 11.2 37.12 62.02
df 1 6 12 30 43
pT .037 (L) .802 .506 .173 .030
Seashore Rhythm
QTPHASE (population based) results
LRS 0.88 3.65 22.48 54.98 94.48
df 1 7 11 31 45
pT .347 .818 .020 .005a 2.265e−005a (pB=.00054)
ETDT (family based) results
LRS 0.029 9.02 21.2 52.34 66.87
df 1 6 11 27 38
pT .864 .171 .031 .002 (pB=.048) .003 (pB=.072)
QTPHASE (population based) results
LRS 0.11 2.940 10.74 40.34 81.37
df 1 7 10 28 47
pT .743 .890 .378 .061a .001a (pB=.024)
ETDT (family based) results
LRS 0.85 1.82 12.09 32.20 57.667
df 1 6 10 26 38
pT .354 .935 .279 .186a .021a
QTPHASE (population based) results
LRS 2.99 4.08 18.54 40.34 60.70
df 1 6 13 28 25
pT .083 .665 .138 .061a 8.34e−005a (pB=.002)
ETDT (family based) results
LRS 1.22 2.82 8.25 44.67 26.04
df 1 6 10 25 13
pT .268 .830 .603 .009 .016
QTPHASE (population based) results
LRS 0.32 0.87 3.29 31.51 45.03
df 1 5 9 23 28
pT .567 .971 .951 .110a .021a
ETDT (family based) results
LRS 3.25 3.93 10.41 49.92 82.62
df 1 7 12 33 50
pT .071 (L) .786 .580 .029 .003a (pB=.072)

Logistic regression was also used to extend the TDT to multiallelic markers such as the AVPR1a repeat regions, giving the ETDT (Sham & Curtis, 1995).

LRS indicates likelihood ratio statistic (UNPHASED); df, degrees of freedom; (L), the long SLC6A4 promoter region long allele, which is preferentially transmitted.

p values are global p values (UNPHASED) that consider all alleles in each repeat region (RS1 and RS3) and take into account one level of multiple testing. Results for individual repeats are not shown. pT (trend) values were not corrected for overall multiple testing, whereas pB values were corrected for overall multiple testing using a very conservative application of the Bonferroni correction because we ignored the moderate correlations between musical and nonmusical phenotypes and the moderate linkage disequilibrium between RS1 and RS3 (8 musical/nonmusical tests×3 genetic markers=24) by multiplying p values by 24.


EM haplotype estimation.

In the last column of Table 2, the results are further extended to include an additional analysis of the original two markers to a third marker, the SLC6A4 HTTLPR 44 bp insertion/deletion, using the method of Koeleman et al. (2000). The Extended Transmission/Disequilibrium Test (ETDT) is adapted to test for an effect at a secondary locus or marker conditional on the association of a candidate locus (AVPR1a RS1 and RS3 repeats) with musical memory in case–parent triads. Considering gametic haplotypes of a candidate (or established) locus (AVPR1a) and a neutral marker (HTTLPR), it is expected that haplotypes with identical alleles at the candidate disease locus but different alleles at the marker have equal transmission probabilities. ETDT transmission probabilities can be estimated and, in the adaptation, can be tested for equality using a likelihood ratio test (i.e., Conditional ETDT). A significant difference in the transmission of haplotypes identical at the candidate locus but different at the secondary locus provides evidence for the involvement of either the secondary locus or a locus in linkage disequilibrium with it. UNPHASED conditional analysis that implements Gene×Gene tests revealed a highly significant association between AVPR1a repeats grouped by the HTTLPR short or long promoter repeat and several measures of musical and phonological memory. A number of these associations remained significant after the conservative Bonferroni correction for multiple testing, especially for the Gordon Melody, MBEA, Seashore Rhythm, and nonword tests. Importantly, population-based and family-based analyses showed an association.

Because promoter repeat length is an important control element in AVPR1a regionally specific brain expression and determines social behavior in the vole (Hammock & Young, 2005), we were prompted to compare repeat length and musical as well as nonmusical memory scores (Fig. 1, Fig. 2). A weak correlation was observed between scores on the DTT and total RS1 and RS3 repeat length (r=−0.200; one-sided p=.037; n=80) as well as scores on the MBEA (r=−0.199; one-sided p=.038; n=80).

View full-size image.

Fig. 1. Bivariate correlation between AVPR1a RS1 and RS3 repeat length (base pairs) and scores on the DTT.

View full-size image.

Fig. 2. Bivariate correlation between AVPR1a RS1 and RS3 repeat length (base pairs) and scores on the MBEA.

4. Discussion

This is the first study to suggest that common polymorphisms contribute to the molecular genetic architecture of the musical brain in Homo sapiens. Significantly, our study suggests that polymorphisms related to the AVPR1a and SLC6A4 HTTLPR promoter region repeats are jointly associated with short-term memory for music, although they are in no way specific to music processing. In humans, musical parameters such as pitch and rhythm underlie prosodic cues, which carry much of the social and emotive information in speech. In fact, Juslin and Laukka (2003) argued that similar patterns of acoustic cues underlie coding of basic emotions in speech and in music performance, consistent with an evolutionary view of vocal expression. As such, music is not only a human universal language but also observed across the animal world, including birds, whales, and dolphins, as well as some primate groups. Interestingly, our study is not the first to point to some commonality in human and animal communication on the molecular level. Such commonality has already been found in the Foxp2 transcription factor, which is expressed in the brain of zebra finch birds during vocal learning and which was also found to be important in mouse ultrasonic vocalizations and, significantly, also in human speech (Haesler et al., 2004). Given the role of vasopressin in social behavior, the association found in our study between musical memory and vasopressin could serve to support evolutionary accounts postulating a social adaptive role in music, such as mother–infant communication, sexual selection, group cohesion, and even early protolanguage (e.g., Wallin et al., 2001).

These findings are clearly provisional. They require replication in a larger and independently recruited sample as well as converging evidence from human and animal brain studies. Nonetheless, they do point to a fascinating possibility that we are now further exploring.


Aarde & Jentsch, 2006 1.Aarde SM, Jentsch JD. Haploinsufficiency of the arginine-vasopressin gene is associated with poor spatial memory performance in rats. Hormones and Behavior,. 2006;49:501–508. CrossRef

Bachner et al., 2005 2.Bachner R, Christian D, Zohar AH, Constantini N, Lerer E, Hoch S, et al. AVPR1a and SLC6A4 gene polymorphisms are associated with creative dance performance. PLOS Sciences,. 2005;1/3:394–403.

Bachner-Melman et al., 2005 3.Bachner-Melman R, Zohar AH, Bacon-Shnoor N, Elizur Y, Nemanov L, Gritsenko I, et al. Linkage between vasopressin receptor AVPR1a promoter region microsatellites and measures of social behavior in humans. Journal of Individual Differences,. 2005;26(1):2–10.

Balding, 2006 4.Balding DJ. A tutorial on statistical methods for population association studies. Nature Review. Genetics,. 2006;7(10):781–791.

Bielsky et al., 2004 5.Bielsky IF, Hu SB, Szegda KL, Westphal H, Young LJ. Profound impairment in social recognition and reduction in anxiety-like behavior in vasopressin V1a receptor knockout mice. Neuropharmacology,. 2004;29:483–493.

Depue & Morrone-Strupinsky, 2005 6.Depue RA, Morrone-Strupinsky JV. A neurobehavioral model of affiliative bonding: Implications for conceptualizing a human trait of affiliation. Behavioral and Brain Sciences,. 2005;28(3):313–350. CrossRef

Drayna et al., 2001 7.Drayna D, Manichaikul A, De Lange M, Snieder H, Spector T. Genetic correlates of musical pitch recognition. Science,. 2001;291:1969–1972. MEDLINE | CrossRef

Dudbridge, 2003 8.Dudbridge F. Pedigree disequilibrium tests for multilocus haplotypes. Genetic Epidemiology,. 2003;25(2):115–121. MEDLINE | CrossRef

Evers & Suhr, 2000 9.Evers S, Suhr B. Changes of the neurotransmitter serotonin but not of hormones during short time music perception. European Archives of Psychiatry and Clinical Neuroscience,. 2000;250:144–147. MEDLINE | CrossRef

Ferguson et al., 2002 10.Ferguson JN, Young LJ, Insel TR. The neuroendocrine basis for social recognition. Frontiers in Neuroendocrinology,. 2002;23:200–224. MEDLINE | CrossRef

Ferris et al., 2001 11.Ferris CF, Rasmussen MF, Messenger T, Koppel G. Vasopressin-dependent flank marking in golden hamsters is suppressed by drugs used in the treatment of obsessive–compulsive disorder. BMC Neuroscience,. 2001;2(1):10. MEDLINE | CrossRef

Floody et al., 1998 12.Floody OR, Cooper TT, Albers HE. Injection of oxytocin into the medial preoptic–anterior hypothalamus increases ultrasound production by female hamsters. Peptides,. 1998;19(5):833–839. MEDLINE | CrossRef

Freeman, 2001 13.Freeman W. A neurobiological role of music in social bonding. In:  Wallin NL,  Merker B,  Brown S editor. .The origins of music. .Cambridge, MA: MIT Press; 2001;p. 411–424.

Galfi et al., 2005 14.Galfi M, Radacs M, Juhasz A, Laszlo F, Molnar A, Laszlo FA. Serotonin-induced enhancement of vasopressin and oxytocin secretion in rat neurohypophyseal tissue culture. Regulatory Peptides,. 2005;127(1–3):225–231. MEDLINE | CrossRef

Gerra et al., 1998 15.Gerra G, Zaimovic A, Franchini D, Palladino M, Giucastro G, Reali N, et al. Neuroendocrine responses of healthy volunteers to ‘techno-music’: Relationships with personality traits and emotional state. International Journal of Psychophysiology,. 1998;28:99–111. MEDLINE | CrossRef

Goodson & Bass, 2000 16.Goodson JL, Bass AH. Forebrain peptides modulate sexually polymorphic vocal circuitry. Nature,. 2000;403(17):769–772. MEDLINE | CrossRef

Gordon, 1965 17.Gordon E. Musical Aptitude Profile. Chicago: Riverside Publishing; 1965;.

Haesler et al., 2004 18.Haesler S, Wada K, Nshdejan A, Morrisey EE, Lints T, Jarvis ED. FoxP2 expression in avian vocal learners and non-learners. Journal of Neuroscience,. 2004;24(13):3164–3175. CrossRef

Hamer & Sirota, 2000 19.Hamer D, Sirota L. Beware the chopsticks gene. Molecular Psychiatry,. 2000;5(1):11–13. MEDLINE | CrossRef

Hammock & Young, 2005 20.Hammock EA, Young LJ. Microsatellite instability generates diversity in brain and sociobehavioral traits. Science,. 2005;308(5728):1630–1634. CrossRef

Ho et al., 2003 21.Ho YC, Cheung MC, Chan AS. Music training improves verbal but not visual memory: Cross-sectional and longitudinal explorations in children. Neuropsychology,. 2003;17(3):439–450. MEDLINE | CrossRef

Insel & Young, 2000 22.Insel TR, Young LJ. Neuropeptides and the evolution of social behavior. Current Opinion in Neurobiology,. 2000;10:784–789. MEDLINE | CrossRef

Juslin & Laukka, 2003 23.Juslin PN, Laukka P. Communication of emotion in vocal expression and music performance: Different channels, same code?. Psychological Bulletin,. 2003;129(5):770–814. MEDLINE | CrossRef

Koeleman et al., 2000 24.Koeleman BP, Dudbridge F, Cordell HJ, Todd JA. Adaptation of the Extended Transmission/Disequilibrium Test to distinguish disease associations of multiple loci: The Conditional Extended Transmission/Disequilibrium Test. Annals of Human Genetics,. 2000;64(Pt. 3):207–213. MEDLINE

Kosfeld et al., 2005 25.Kosfeld M, Heinrichs M, Zak P, Fischbacher U, Fehr E. Oxytocin increases trust in humans. Nature,. 2005;435/2:673–676.

Lewis, 2002 26.Lewis CM. Genetic association studies: Design, analysis and interpretation. Brief Bioinformatics,. 2002;3(2):146–153.

Menon & Levitin, 2005 27.Menon V, Levitin D. The rewards of music listening: Response and physiological connectivity of the mesolimbic system. Neuroimage,. 2005;28(1):175–184. MEDLINE | CrossRef

Modahl et al., 1998 28.Modahl C, Green L, Fein D, Morris M, Waterhouse L, Feinstein C, et al. Plasma oxytocin levels in autistic children. Biological Psychiatry,. 1998;43:270–277. Abstract | Abstract + References | Full-Text PDF (761 KB) | MEDLINE | CrossRef

Peretz et al., 2003 29.Peretz I, Champod S, Hyde K. Varieties of musical disorders: The Montreal Battery of Evaluation of Amusia. Annals of the New York Academy of Sciences,. 2003;999:68–75.

Perras et al., 1997 30.Perras B, Droste C, Born J, Fehm H, Pietrowsky R. Verbal memory after three months of intranasal vasopressin in healthy old humans. Psychoneuroendocrinology,. 1997;22(6):387–396. Abstract | Abstract + References | Full-Text PDF (615 KB) | MEDLINE | CrossRef

Prasada-Rao & Kanwal, 2004 31.Prasada-Rao PD, Kanwal JS. Oxytocin and vasopressin immunoreactivity within the forebrain and limbic-related areas in the mustached bat, Pteronotus parnellii. Brain Behavior and Evolution,. 2004;63(3):151–168. MEDLINE | CrossRef

Seashore et al., 1960 32.Seashore CE, Lewis D, Saetveit JC. Manual of instruction and interpretations for the Seashore Measures of Musical Talents. New York: The Psychological Corporation; 1960;.

Sham & Curtis, 1995 33.Sham PC, Curtis D. An extended transmission/disequilibrium test (TDT) for multi-allele marker loci. Annals of Human Genetics,. 1995;59(Pt. 3):323–336. MEDLINE | CrossRef

Shuter-Dyson, 1999 34.Shuter-Dyson R. Musical ability. In:  Deutsch D editors. The psychology of music. 2nd ed.. San Diego: Academic Press; 1999;p. 627–651.

Spielman et al., 1993 35.Spielman RS, McGinnis RE, Ewens WJ. Transmission test for linkage disequilibrium: The insulin gene region and insulin-dependent diabetes mellitus (IDDM). American Journal of Human Genetics,. 1993;52(3):506–516. MEDLINE

Thompson et al., 2006 36.Thompson RR, George K, Walton JC, Orr SP, Benson J. Sex-specific influences of vasopressin on human social communication. Proceedings of the National Academy of Sciences,. 2006;103(20):7889–7894.

Wallin et al., 2001 37.In:  Wallin LN,  Merker B,  Brown S editor. (2001).The origins of music. Cambridge, MA: M.I.T; 2001;.

a Department of Musicology, The Hebrew University, Jerusalem, Israel

b Department of Cognitive Studies, The Hebrew University, Jerusalem, Israel

c S. Herzog Memorial Hospital, Jerusalem, Israel

d Neurobiology, The Hebrew University, Jerusalem, Israel

e Scheinfeld Center, Department of Psychology, The Hebrew University, Jerusalem, Israel

Corresponding authors. Roni Granot is to be contacted at Department of Musicology, The Hebrew University, Jerusalem 91905, Israel. Tel.: +972 2 5883953; fax: +972 2 5883944; 0544 486880 (mobile). Richard P. Ebstein, Scheinfeld Center for Genetic Studies in the Social Sciences, Department of Psychology, The Hebrew University, Jerusalem 91905, Israel, and S. Herzog Memorial Hospital, Givat Shaul, Jerusalem 91351, Israel. Tel.: +972 2 5316855; fax: +972 2 5316853; 0523 822810 (mobile).

 This study was partially supported by a Hebrew University R&D Young Researcher Grant awarded to Roni Y. Granot.

PII: S1090-5138(07)00047-5