The dopamine D4 receptor polymorphism affects the canine fearfulness
The dopamine D4 receptor polymorphism affects the canine fearfulness
Lee, Chaeyoung; Kim, Changhoon; Shin, Sooan; Shin, Daesung; Kang, Joohyun; Park, Chankyu
2008-01-01 00:00:00
Animal Cells and Systems 12: 77-83, 2008 The Dopamine D4 Receptor Polymorphism Affects the Canine Fearfulness 1 2 2 2 2 2, Chaeyoung Lee , Changhoon Kim , Sooan Shin , Daesung Shin , Joohyun Kang and Chankyu Park * Lab of Statistical Genetics, Institute of Environment & Life Science, Hallym University , Chuncheon 200-702, Korea; Department of Biological Sciences , Korea Advanced Institute of Science and T echnology , Daejon 305-701, Korea Abstra ct : The canine fearfulness is a behavioral trait known different types of objects were employed, i.e. stationary to have a genetic basis. This research analyzed genetic objects and objects with movable part. effects of the dopamine D4 receptor polymorphism on this In a subsequent study (Goddard and Beilharz, 1983 and behavior by postulating a mixed model of inheritance. 1985), a large battery of behavioral tests (thirty eight) was Genotyping for the three different repeat polymorphism developed. A factor analysis of the test scores obtained for found in the third exon of the receptor gene was carried out the progenies of four different breeds indicated that the for the population of the Korean native dogs. Four hundred fifty eight dogs with known pedigree were genotyped, and phenotypic variation in fear response could be described in 264 individuals were tested for their fear responses to an terms of 12 factors. A further statistical analysis identified experimenter, in which four different behavioral paradigms the three discriminant functions, implying three underlying were adopted. Since the results assessed by principal factor behavioral dimensions labeled as “general fearfulness”, analysis revealed a major factor explaining 69% of the total “fearfulness of objects”, and “ inhibited response to fear”. ph enoty p ic v a ri ance, the subsequ ent anal y s e s w e re conducted A heritability estimate for the subdimensions ranges from for this quantity . Analyses of the factor scores by estimating 0.47 to 0.80 (Goddard and Beilharz, 1985). their posterior means indicated that there is a fixed effect exerted by the three different repeat polymorphism found in The D4 dopamine receptor has been subjected to the D4 receptor as well as sex, in addition to unidentified behavioral studies in h u man an d mo use. Initially , the repeat polygenic effects. The phenotypic contribution of the D4 rd polymorphism found in the 3 exon of the receptor was genotype was roughly estimated to be about 2%, which is a reported to be associated with the novelty seeking trait in fraction of the tot al genetic ef fect s responsible for more than human s (Ebstein et al., 19 96; Be n j amin et al., 1996), which 20% of the total phenotypic variance. was supported by an observation with a knock-out mice Key words: dopamine receptor, fearfulness, dog, polymor- (Dulawa et al., 1999). The human polymorphism contains phism, behavior variable number of tandem repeats (VNTR) of 48-bp unit, normally occurring in 2-10 copies. Recently , it was reported that a particular allele with seven repeat has been selected Dog’s response of fear has been observed against various in human population (Ding et al, PNAS,2002). A similar objects, which can be measured as a behavioral phenotype. repeat polymorphism with variable u nit sizes of 27, 39, an d A c c o r d i ng t o Me lz ac k ( 195 2) , f our c r i t e r i a of f ear exp r e s s ion 51 bps was found in the canine D4 receptor (Niimi et al., were noted; (1) turning its head more than 90 degrees to 2001; Inoue-murayama et al., 2002; Jeoung et al., in avoid an object, (2) lowering its head and body as well as preparation), which is analogous to the primate counterpart flattening ears, (3) hiding in a place away from the object, in its location and structural characteristics. They are part of rd and (4) sitting at fa r end of the cage to maintain the furthest the 3 cytoplasmic region, exh ibiting any notable sequence distance from th e object. Th e last observation wa s regarded similarity to that of human except the proline-rich as an extreme criterion for fearfulness. In this test, two characteristics. As a study to elucidate functional consequence of genetic polymorphism found in the canine D4 gene, we examined * T o whom correspondence should be addressed. whether the variation at D4 locus, together with other T el: +82-42-869-2629; Fax: +82-42-869-2610 E-mail: ckpark@kaist.ac.kr polygenic effects, affect fearfulness of dogs. An analysis ANIMAL CELLS AND SYSTEMS Vo l. 12 No. 2 77 Chaeyoung Lee, Changhoon Kim, S ooan Shin, Daesung Shin, Joohyun Kang, and Chankyu P ark b ased o n the m i x e d m o d el r e ve ale d t h a t t he D 4 p o l y m orp hi sm monitoring the responses of dogs in the first cage, the with other unknown genes significantly influences canine experimenter moved to the fifth cage for another testing fearfulness. that was not closely located. Any unusual movement or loud noise was prohibited because they might be perceived MA TERIALS AND METHODSas threats by the dogs. T o measure fearfulness, dog’s responses to a stranger in Ani m a l four different test situations (O1-O4) were recorded. Our The animals tested in our study were chosen from a colony choice of the test objects is comparable to the one used by with several hundred dogs raised and maintained in an Melzack (1952) or to the FM6/12 criteria of Goddard et al. isolated farm. This Korean breed, Sapsari (Kim et al., (1985), which were taken to score the degree of avoidance 1998), tends to recognize humans as large and unfamiliar from the experimenter . Some of the responses were similar ob jects because they have fe w contacts with humans. Thus, to the ones observed by Malzack (1952), but a series of they tend to be cautious and express fear whenever they dog’s responses to each test object were not exactly encounter people except breeders who feed them. In this matched. Thus, the responses to each test object were study, we measured fearfulness behaviors of individual ranked independently as in Table 2. The first rank was dogs of ages from 6 months to 3 years. The sizes of canine given to the extreme fearfulness response and the next to pedigrees used in this analysis were summarized in Table 1.the second best, etc. The largest pedigree comprises 268 individuals in seven The actual tests were performed as the followings. First, generations. the experimenter examined the dog’s responses when the experimenter was standing outside the cage with the dog Genotyping of the repeat alleles by PCR amplification inside (the O1). Second, the experimenter walked in and Genomic DNAs from Sapsari were prepared by the stayed inside th e cage while examining the dog’ s respo nses standard method with a minor modification. The canine (O2). Third, the response was monitored when he or she specific primers, 5'-ggtgg ccgtg ccgct gagtt acaac cgcca-3' approached the dog (O3), and when the experimenter (forward) and 5'-cgacc accac gggca gcacc ctcata gcct-3' touched the dog with his or her hand (O4). The tests were (reverse), recognizing an imme dia te upstream region of TM4 conducted more than five times with an interval of more and the portion of TM6 were used for PCR amplification of than a week. The arithmetic means of the observed ranks, the D4 repeat sequence. The final reaction mixture contains excluding the maximum and minimum values, were used 75 mM Tris-Cl, pH 9.0, 15 mM (NH ) SO , 2.5 mM as the fearfulness scores (S1-S4) obtained by O1-O4 tests, 4 2 4 MgCl, 100mg/ml BSA, 200ng of template DNA, 0.4mMrespectively. All the scores were normalized. of each primers, 0.2 mM dNTPs, 2 M betain, and 2 units of In order to find the relationship among the fearfulness Taq polymerase in a total volume of 50 ml. PCR reaction scores for different test objects, correlations between the was performed with 1 cycle at 95 C (2 min), 35 cycles at scores obtained were assessed, followed by principal o o o 95 C (1 min)/62 C (1 min)/74 C (3 min), followed by ancomponent factor analysis with varimax rotation (SAS Inst. extension reaction at 72 C for 20 min. The PCR products Inc., Cary , NC, USA). The scores of principal components were electrophoresed on 2.0% agarose gel and scored by achieved by each dog were also normalized for easy sizes of 718, 769 and 1218 bp for alleles of small (S), lar gecomparison with the original scores. (L) and ultralong (U). Genetic model Fearfulness testThe study used a mixed inheritance model as shown below: A person never exposed to the dogs was considered a y = Xβ + Za + e, stranger. The home cage was fenced surrounding the area of 6 ×4 m, in which the test was carried out. For an entry ofwhere y represents a vector of observations for behavior experimenter, a door was installed on one side. Theperformance, a vector of sex and fixed genotypic effects, following protocol for behavioral test was initially set up.and a a vector of additive polygenic random effects with The experimenter who made observation did not use any the assumption of a~N(0, A σ ), where A is polygenic fragrances such as perfume and lotions. Also to minimize relationship matrix among animals in the pedigree (Quaas, an environmental disturbance regarding odor, windy days 1976). σ is additive polygenic variance component, and e were avoided. Furthermore, the experimenters spent half an is a random vector of residuals with the assumption of hour to get accustomed to odor surrounding the cage priore~N(0, I σ ), where I is identity matrix, and σ residual to an observation. Cages for observation were randomlyvariance component. X and Z are known incidence matrices selected. In order to avoid the influence by the previous relating the fixed and random effects, respectively , to their test, a careful consideration was given. For example, aftercorresponding observations. 78 ANIMAL CELLS AND SYSTEMS Vol. 12 No. 2 Genetics of Canine Fearfulness Parameter estimation receptor gene of the dog population, two primers were Inferences about unknown variance are based on theirdesigned to amplify the corresponding region of 0.7-1.2Kb marginal posterior distribution components in Bayesian in size. Based on the previous identification of three approaches, and the marginalization of the joint posterior different alleles (Fig. 1) by DNA sequencing, allelic distribution can be attained through Gibbs sampling variations in the Sapsari population were genotyped on (Tanner, 1993). The joint posterior distribution of all agarose gel electrophoresis (Materials and Methods). parameters in our model can be expressed by Bayes Among 458 animals tested in twelve pedigrees (Table 1), theorem: the three alleles of S, L, and U were found at the frequencies of .57, .36, and .07, respectively . The genotypes of SS, LL, 2 2 2 2 2 2 2 f(β, a, σ , |y σ) ∝ f(y|β, a, σ, σ ) f(a| σ )π(β)π() σ π() σ SL, SU, and LU among 458 dogs were found at the a e a e a a e frequencies of 0.31, 0.106, 0.458, 0.07, and 0.056. 2 2 2 2 where f(y|β, a, σ, σ )~N(Xβ +Za + I σ and f(a| σ )~N(0 , a e e a 2 2 2 2 A σ ). The π(β), π( σ ), and π( σ ) are the priors for β, , σ Correlations between the behaviors assessed by a a e a and σ , respectively. For the priors, uniform distribution factor analysis was assumed for the fixed effects, and inverse GammaMeasurements of canine fearfulness were conducted for the distribution was assumed for the variance components, test objects or situ atio ns (T able 2). V arious responses of the because the application of the Gibbs sampling with flat dogs to each test were recorded and converted into scores priors for variance components may yield a theoretically as described in the Materials and Methods section. Thus, improper posterior distribution (Hobert and Casella, 1996). each dog has four fearfulness scores, S1 through S4, as Furthermore, very small shape parameters for the inversemeasured in Table 2. They were subject to a factor analysis, Gamma distributions were taken due to a lack of prior from which two major components, F1 and F2 presumably information, and Gibbs sampling was applied to parameter associated with fear response, were drawn, each explaining estimation. It enabled random samples to be drawn with 69 and 17% of the total phenotypic variance, respectively. desired posterior distributions. Figure 2 displays normal distribution of dogs exhibiting Then, initial values were randomly assigned for all dif ferent degrees of F1 or F2 fearfulness, shown here as the unknowns. Sampling from the full conditional posterior normalized scores. The F1 factor is almost equally loaded distributions and updating the distribution based on theonto the four fearfulness scores (0.86, 0.86, 0.86, and 0.72), sample were repeated to the conver gence. Algorithms from while F2 factor in varying degrees ( −0.34, −0.32, 0.12, and Numerical Recipes by Press et al. (1992) were used to 0.65). The result indicates that the general fearfulness is generate the samples. Programs from FSPAK by Misztal represented by F1. An interpretation about F2, however, (1990) were used to deal with sparse matrix inversion. remains to be determined. This result provides a basis for Gibbsit program by Raftery and Lewis (1995) was used to further analysis of th e fearfulness mainly in terms of the F1 determine warming-up periods and thinning intervals. Thescores. Gibbs sampler was run 102,000 rounds, and the first 2000 rounds were discarded as a warming-up period. A Estimations of fixed effects, variance components, and their correlations conservative thinning interval of 100 rounds was used to retain sampled values that reduced lag correlation among Based on the mixed model of inheritance (Materials and thinned samples. The point estimate used in the current Methods), variance components were inferred from the study was the posterior mean estimate, calculated as themarginal posterior distributions. Variances due to polygenic mean of the conditional expected values of the parameters and environmental effects were estimated, as shown in in post warming-up rounds. Table 3, which are statistically significant ( p < 0.05). The polygenic effect ( σ ) on F1 is approximately 20% of the total phenotypic variance, while the environmental effect RESUL TS ( σ ) was about 70%. Therefore, about 10 % of the Allele frequencies of the D4 repeat polymorphism remaining phenotypic variance might be attributed to the The repeat region of the canine D4 receptor was found infixed effects by sex and D4 genotype. rd rd the 3 exon of D4 gene, including the 3 cytoplasmic The posterior mean estimates for the fixed effects due to region of the receptor . This region consists of a novel amino genotype and sex were obtained (Fig. 3), which are acid sequence preceded by approximately 243 amino acids statistically significant ( p < 0.05) for both factors deduced from the N-terminus that are homologous in the mammalian previously . A female is more fearful than a male when the D4 receptors. The block of repeated sequence is followed fixed effect was assumed to be caused only by sex by a homologous region of 89 amino acids extended to the difference. The SS and SU groups are the least fearful, C-terminus. To detect various repeat alleles in the D4 while SL and LU groups are the most fearful groups with ANIMAL CELLS AND SYSTEMS Vo l. 12 No. 2 79 Chaeyoung Lee, Changhoon Kim, S ooan Shin, Daesung Shin, Joohyun Kang, and Chankyu P ark Fig. 1. Repeat p atterns and variations found in the dop ami ne D4 receptor of Sa p s a r i. The basic units with 39, 51, and 39-27-51 base pairs are shown with their nucleotide and amino acid sequences. A variation in their sequences is also designated. In our population of Sap s ari, three different alleles (S, L, and U) were found with different repeat lengths. Table 1. Size and genetic complexity of pedigrees used LL group in between, when the dogs were grouped according to their genotypes disregarding their sexes. The Maximum generations Number of pedigrees Number of dogs differences in fearfulness among genotypes are not caused 7 1 268 by sexual bias, since females are predominant in all 4 3 68 genotype groups. The contribution of D4 genotypes to the 3 8 122 total fixed effect was roughly estimated to be one fifth by Total 12 458 analyzing the variances of fixed effects in each sex, which The number of the dogs included in the principal factor analysis was 264 was based on the assumption that D4 genotype and sex after omitting untested dogs. effects are independent in fearfulness response. Table 2. Measurement of canine fearfulness Role of experimenter Rank O1 O2 O3 O4 2 2 2 2 1Excreting Excreting Excreting Excreting 2 Hiding Hiding Hiding Trembling/rigid 3 Becoming agitated Becoming agitated Running away Becoming rigid 4 Sitting quietly Sitting quietly Sitting down plump Startled 5 Wandering Wandering Sitting quietly Mumbling 6 Approaching Approaching Approach/sitting Sitting quietly 7 Licking hands Dashing gladly Dashing gladly Licking 8 Dashing gladly Lying on its back O1: Outside the cage, O2: Standing still inside the cage, O3: Approaching the dog, and O4: T ouching the dog. Includes urination. 80 ANIMAL CELLS AND SYSTEMS Vol. 12 No. 2 Genetics of Canine Fearfulness Fi g. 2. Sc ore di s tri but i ons of t he fir s t and t he s econd f ac tors . T he s i z e of bi n is 0. 2. T he s haded bars repres ent the number of obs er v ed dogs in each bin and the line represents fitted curve with standard normal distribution. Fig. 3. Fixed effects on F1 and F2 scores of D4 genotypes and sex difference. The filled and empty bars are for the F1 and F2 scores, respectively . Error ranges are indicated by intervals. Although the analyses were primarily focused on the F1 factor score, the F2 results were also included for a comparison. In general, fixed ef fect s of the F2 score are less significant than that of the F1 score. Genot ypes of SS/SU, LL, LU , and SL wi th r egard to F1 and thos e of SS/ LL/LU, SL/S U wi th regard to F2 eac h r epres ent a group, s howing s i gnif i c ant ( p < 0 . 05) dif f erenc es among them. Sex effect s were significant ( p < 0.01) on both F1 and F2. Additive and dominance effects for the dopamine D4 Correlations within polygenic and environmental effects receptor locus were also assessed (Table 3). The additive were assessed for all pairs of the scores (Table 4). The effect of the L allele is negative, while the effect of the S polygenic and environmental correlations for S1, S2, S3, allele is positive. Over-dominance observed between the and S4 are positiv e and ran ging fro m 0.37 to 0.44 an d from alleles, i.e., that the estimates of dominance effect between 0.42 to 0.67, respectively, indicating that the original the S and L alleles are much larger, suggests that variables are better correlated with one another in po lygenic heterozygous individuals ( SL) are more fearful than effects as well as in fixed effects. Correlations between F1 homozygous ones ( SS or LL). More analysis on allelic and F2 are weak, in dicating that they are fairly indep en dent interaction could not be made due to an absence ofeven in the extracted components. individuals with UU genotype. ANIMAL CELLS AND SYSTEMS Vo l. 12 No. 2 81 Chaeyoung Lee, Changhoon Kim, S ooan Shin, Daesung Shin, Joohyun Kang, and Chankyu P ark Table 3. Additive and dominance effects of dopamine D4 receptor type of object, a modification of the test, e.g. an approaching to an animal, was attempted to incorporate the results of the AdditiveDominanceVariance Fac tors 2 2 previous tests (Melzack, 1952). SL SL σ σ a e Although more recent analysis of the canine fear response F10.69-0.15-2.49(L)0.20±0.050.70±0.16 (Goddard and Beiharz, 1985) included a test of dogs to a F2-0.17-0.220.99(S)0.12±0.040.82±0.25 human object with an interval of 6 months, scored as the Alleles in parentheses are regarded as the dominant ones. The additive FM6 and FM12 behaviors, they were classified into a nd d omin an ce ef fec t s of U alle le were n ot e stima ted d ue to an a bse nc e of individual with UU genotype. All the numbers were multiplied by ten. different categories other than the one named ‘fearfulness of object’. It seems odd that the FM12 behavior, the most T able 4. Correlations within polygenic (upper off-diagonal elements) similar to ours, is in the group called ‘general fearfulness’ and environmental (lower off-diagonal elements) estimates among which comprises a variety of behaviors, all involving a the six fearfulness traits sound cue. On the other hand, the object fearfulness defined S1S2S3S4F1F2 by their third d iscriminant function was shown to be higher S10.440.400.370.580.37 in females than in males (Goddard and Beiharz, 1985). S20.670.380.390.630.33 Although the linkage analysis searching for major genes S30.580.540.410.590.32 involved in the fear and anxiety responses has yet to S40420.490.510.490.30 identify a specific gene in mouse (Flint et al., 1995, Dulawa F10.650.660.620.600.15 et al., 1999), the complex nature of these phenotypes may involve a number of genes with minor contributions. As F2 0.40 0.37 0.37 0.32 0.17 1 demonstrated in this study, various behaviors expressing Correlations were inferred based on the posterior mean estimates of poly- genic effects and residuals. canine fearfulness can be explained in part by the genetic diversity found in the D4 locus. In the analytical model, maximum effects of D4 locus was considered because the DISCUSSION estimates might be confounded with the effects of other loci We found that genetic polymorphism at the dopamine D4 that are linked to the D4 locus. Nonetheless, we expected receptor influenced fearfulness in an outbred population of that major contribution would be due to the variation in the a canine breed, Sapsari. The same receptor involved in the dopamine D4 receptor since the estimates in the current dopaminergic pathway of human has been implicated in study were obtained for the D4 locus regardless of their various neurological phenomena including locomotion, direct or indirect effects. A possible concern on the emotional stability , and neuroendocrine function (Civelli et analytical model was that the genotypes of D4 might be al., 1993). Indeed, pharmacological modulation of thisconsidered as random effects. In other words, the genotypes pathway resulted in numerous neurological and behavioral could be determined at random, each drawn from infinite consequences. In spite of these pharma co -medical evidences, population of possible genotypes. Yet in our case, we recent studies of genetic association in human, primarily treated the D4 genotypes as fixed effects since the limited focused on the repeat polymorphism (Wong et al., 2000;number of the genotypes was observed. Lung et al., 2002; Schinka et al., 2002 ) and the 120-bp The statistical genetics approach introduced here to duplication in the 5'-UTR (McCracken et al., 2002) lend reveal an involvement of the D4 polymorphism in canine supports to both positive and negative conclusions. In most fearfulness might provide a sensitive tool for finding a cases, the phenotype concerned was either the novelty-genetic association in multi-factorial trait. Indeed, our result seeking personality trait or the attention deficit/hyperactivity obtained with the canine model strongly supports the role disorder (ADHD). In addition, none of the genome-wide of D4 receptor in determining personality trait, which is screening for related traits including ADHD (Fisher et al.,still controversial in human. 2002) found a linkage to the D4 locus. Fearfulness is a personality trait ex pressing an emotional ACKNOWLEDGMENTS state of an organism, which is also associated with anxiety . C. Lee and C. Kim contributed equally to this work. In animal, a fear response can be elicited by various types This work was supported by grant from the KOSEF. of stimuli associated with learning, ecological history, and signals from their social group. Apparently, the type of REFERENCES stimulus we employed for testing dogs belongs to a category of novel object, i.e. an unexposed stranger. A similar Benjamin J, Li L, Patterson C, Greenberg BD, Murphy DL, and behavioral paradigm was adopted in the systematic study of Hamer DH (1996) Population and familial association canine fear (Melzack, 1952), in which significant difference between the D4 dopamine receptor gene and measures of Novelty Seeking. Nat Genet 12: 81-84. was noted in various dog breeds. Although we used a single 82 ANIMAL CELLS AND SYSTEMS Vol. 12 No. 2 Genetics of Canine Fearfulness Civelli O, Bunzow JR, and Grandy DK (1993) Molecular of the dopamine receptor D4 exon III repetitive region in diversity of the dopamine receptors. Ann Rev Pharmacol several species of the order carnivora. J V et Med Sci 64: 747- T o xi col 33: 281-307. 7 49. Ding Y, Chi H, Grady DL, Morishiwa A, Kidd JR, Kidd KK, Kim KS, Lee SE, Jeong HW, and Ha JH (1998) The complete Flodman P, Spence MA, Schuck S, Swanson JM, Zhang Y , nucleotide sequence of the domestic dog ( Canis familiaris ) and Moyzis RK (2002) Evidence of positive selection actingmitochondrial genome. Mol Phylogen Evol 1 0 : 2 10- 220 . at the human dopamine receptor D4 gene locus. Proc Natl Lung F, Tzeng D, and Shu B (2002) Ethnic heterogeneity in allele Ac a d Sc i U S A 99: 309-314. variation in the DRD4 gene in schizophrenia. Schizophr Res Dulawa SC, Grandy DK, Low MJ, Paulus MP, and Geyer MA57: 239-45 (1999) Dopamine D4 receptor-knock-out mice exhibit Me lzack R (1 952) I r r a t i on al f ear s in t h e do g. Can J P s ych 6: 1 41- reduced exploration of novel stimuli . J Neurosci 19: 9550- 1 47. 95 56. Misztal I (1990) Restricted maximum likelihood estimation of Fisher EF , Francks C, McCracken JT , McGough JJ, Marlow AJ, variance components in an animal model using sparse matrix MacPhie IL, Newbury Low DF , Crawford LR, Palmer CGS, i nve rsio n and a sup e r c omp u ter . J Da i r y S c i 73: 163-172. Woodward JA, Del’Homme M, Cantwell DP, Nelson SF, Niimi Y, Inoue-Murayama M, Kato K, Matsuura N, Murayama Y, Monaco AP , and Smalley SL (2002) A genomewide scan for Ito S, Momoi Y, Konno K, and Iwasaki T (2001) Breed loci involved in attention-deficit/hyperactivity disorder . Am J differences in allele frequency of the dopamine receptor D4 Hum Genet 70: 1183-1196. gene in dogs. J Hered 92: 433-436. Flint J, Corley R, DeFries JC, Fulker DW , Gr ay JA, Miller S, and Press WH, Teukolsky SA, Vetterling WT, and Flannery BP Collins AC (1995) A simple genetic basis for a complex nd (1992) Numerical recipes in FORTRAN. 2 ed. New York, psychological trait in laboratory mice. Scie nce 269: 1432- U S A: Camb r i dg e Univ er si ty P r e ss. 14 35. Quaas RL (1976) Computing the diagonal elements and inverse Goddard ME and Beiharz RG (1983) Genetics of traits which of a large numerator relationship matrix. Bi omet rics 32: 949- determine the suitability of dogs as guide-dogs for the blind. 9 53. Appl Anim Ethol 9: 299-315 Raftery AE and Lewis SM (1995) Implementing MCMC. In: Goddard ME and Beilharz RG (1985) A multivariate analysis of Markov Chain Monte Carlo in Practice, (ed) (Gilks WR, the genetics of fearfulness in potential guide dogs. Behavior Richardson S, andSpiegelhalter DJ). London, UK, Chapman Gen e ti cs 15: 69-89. and Hall, pp 641-649. Ebstein RP , Novick O, Umansky R, Priel B, Osher Y , Blaine D, Schinka JA, Letsch EA, and Crawford FC (2002) DRD4 and Bennett ER, Nemanov L, Katz M, and Belmaker RH (1996) novelty seeking: Results of meta-analyses. Am J Med Genet Dopamine D4 receptor (D4DR) exon III polymorphism 1 14: 6 43- 648 . associated with the human personality of Novelty Seeking . Tanner MA (1993) Tools for Statistical Inference: methods for the N a t Ge n e t 12: 78-80. exploration of posterior distributions and likelihood functions, Harville DA (1974) Bayesian inference for variance components rd 3 ed. New York, USA , Springer-Verlag. using only error contrasts. Biometrika 61: 383-385. W ong AH, Buckle CE and V an T ol HHM (2000) Polymorphisms Hobert JP and Casella G (1996) The effect of improper priors on in dopamine receptors: what do they tell us? Eur J Pharmacol Gibbs sa mpling in hierarchica l lin ear mixed mo dels. J A m S t at 410: 183-203. Assoc 91: 1461-1473. Inoue-Murayama M, Matsuura N, Murayama Y, Tsubota T, [Received March 10, 2008; accepted May 19, 2008] Iwa s aki T , Ki tag a w a H, a n d Ito S (20 02) Se que nc e Co m p a r is on ANIMAL CELLS AND SYSTEMS Vo l. 12 No. 2 83
http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.pngAnimal Cells and SystemsTaylor & Francishttp://www.deepdyve.com/lp/taylor-francis/the-dopamine-d4-receptor-polymorphism-affects-the-canine-fearfulness-5z8CL7HgG5
The dopamine D4 receptor polymorphism affects the canine fearfulness
The dopamine D4 receptor polymorphism affects the canine fearfulness
Abstract
Abstract The canine fearfulness is a behavioral trait known to have a genetic basis. This research analyzed genetic effects of the dopamine D4 receptor polymorphism on this behavior by postulating a mixed model of inheritance. Genotyping for the three different repeat polymorphism found in the third exon of the receptor gene was carried out for the population of the Korean native dogs. Four hundred fifty eight dogs with known pedigree were genotyped, and 264 individuals were tested for their...
Animal Cells and Systems 12: 77-83, 2008 The Dopamine D4 Receptor Polymorphism Affects the Canine Fearfulness 1 2 2 2 2 2, Chaeyoung Lee , Changhoon Kim , Sooan Shin , Daesung Shin , Joohyun Kang and Chankyu Park * Lab of Statistical Genetics, Institute of Environment & Life Science, Hallym University , Chuncheon 200-702, Korea; Department of Biological Sciences , Korea Advanced Institute of Science and T echnology , Daejon 305-701, Korea Abstra ct : The canine fearfulness is a behavioral trait known different types of objects were employed, i.e. stationary to have a genetic basis. This research analyzed genetic objects and objects with movable part. effects of the dopamine D4 receptor polymorphism on this In a subsequent study (Goddard and Beilharz, 1983 and behavior by postulating a mixed model of inheritance. 1985), a large battery of behavioral tests (thirty eight) was Genotyping for the three different repeat polymorphism developed. A factor analysis of the test scores obtained for found in the third exon of the receptor gene was carried out the progenies of four different breeds indicated that the for the population of the Korean native dogs. Four hundred fifty eight dogs with known pedigree were genotyped, and phenotypic variation in fear response could be described in 264 individuals were tested for their fear responses to an terms of 12 factors. A further statistical analysis identified experimenter, in which four different behavioral paradigms the three discriminant functions, implying three underlying were adopted. Since the results assessed by principal factor behavioral dimensions labeled as “general fearfulness”, analysis revealed a major factor explaining 69% of the total “fearfulness of objects”, and “ inhibited response to fear”. ph enoty p ic v a ri ance, the subsequ ent anal y s e s w e re conducted A heritability estimate for the subdimensions ranges from for this quantity . Analyses of the factor scores by estimating 0.47 to 0.80 (Goddard and Beilharz, 1985). their posterior means indicated that there is a fixed effect exerted by the three different repeat polymorphism found in The D4 dopamine receptor has been subjected to the D4 receptor as well as sex, in addition to unidentified behavioral studies in h u man an d mo use. Initially , the repeat polygenic effects. The phenotypic contribution of the D4 rd polymorphism found in the 3 exon of the receptor was genotype was roughly estimated to be about 2%, which is a reported to be associated with the novelty seeking trait in fraction of the tot al genetic ef fect s responsible for more than human s (Ebstein et al., 19 96; Be n j amin et al., 1996), which 20% of the total phenotypic variance. was supported by an observation with a knock-out mice Key words: dopamine receptor, fearfulness, dog, polymor- (Dulawa et al., 1999). The human polymorphism contains phism, behavior variable number of tandem repeats (VNTR) of 48-bp unit, normally occurring in 2-10 copies. Recently , it was reported that a particular allele with seven repeat has been selected Dog’s response of fear has been observed against various in human population (Ding et al, PNAS,2002). A similar objects, which can be measured as a behavioral phenotype. repeat polymorphism with variable u nit sizes of 27, 39, an d A c c o r d i ng t o Me lz ac k ( 195 2) , f our c r i t e r i a of f ear exp r e s s ion 51 bps was found in the canine D4 receptor (Niimi et al., were noted; (1) turning its head more than 90 degrees to 2001; Inoue-murayama et al., 2002; Jeoung et al., in avoid an object, (2) lowering its head and body as well as preparation), which is analogous to the primate counterpart flattening ears, (3) hiding in a place away from the object, in its location and structural characteristics. They are part of rd and (4) sitting at fa r end of the cage to maintain the furthest the 3 cytoplasmic region, exh ibiting any notable sequence distance from th e object. Th e last observation wa s regarded similarity to that of human except the proline-rich as an extreme criterion for fearfulness. In this test, two characteristics. As a study to elucidate functional consequence of genetic polymorphism found in the canine D4 gene, we examined * T o whom correspondence should be addressed. whether the variation at D4 locus, together with other T el: +82-42-869-2629; Fax: +82-42-869-2610 E-mail: ckpark@kaist.ac.kr polygenic effects, affect fearfulness of dogs. An analysis ANIMAL CELLS AND SYSTEMS Vo l. 12 No. 2 77 Chaeyoung Lee, Changhoon Kim, S ooan Shin, Daesung Shin, Joohyun Kang, and Chankyu P ark b ased o n the m i x e d m o d el r e ve ale d t h a t t he D 4 p o l y m orp hi sm monitoring the responses of dogs in the first cage, the with other unknown genes significantly influences canine experimenter moved to the fifth cage for another testing fearfulness. that was not closely located. Any unusual movement or loud noise was prohibited because they might be perceived MA TERIALS AND METHODSas threats by the dogs. T o measure fearfulness, dog’s responses to a stranger in Ani m a l four different test situations (O1-O4) were recorded. Our The animals tested in our study were chosen from a colony choice of the test objects is comparable to the one used by with several hundred dogs raised and maintained in an Melzack (1952) or to the FM6/12 criteria of Goddard et al. isolated farm. This Korean breed, Sapsari (Kim et al., (1985), which were taken to score the degree of avoidance 1998), tends to recognize humans as large and unfamiliar from the experimenter . Some of the responses were similar ob jects because they have fe w contacts with humans. Thus, to the ones observed by Malzack (1952), but a series of they tend to be cautious and express fear whenever they dog’s responses to each test object were not exactly encounter people except breeders who feed them. In this matched. Thus, the responses to each test object were study, we measured fearfulness behaviors of individual ranked independently as in Table 2. The first rank was dogs of ages from 6 months to 3 years. The sizes of canine given to the extreme fearfulness response and the next to pedigrees used in this analysis were summarized in Table 1.the second best, etc. The largest pedigree comprises 268 individuals in seven The actual tests were performed as the followings. First, generations. the experimenter examined the dog’s responses when the experimenter was standing outside the cage with the dog Genotyping of the repeat alleles by PCR amplification inside (the O1). Second, the experimenter walked in and Genomic DNAs from Sapsari were prepared by the stayed inside th e cage while examining the dog’ s respo nses standard method with a minor modification. The canine (O2). Third, the response was monitored when he or she specific primers, 5'-ggtgg ccgtg ccgct gagtt acaac cgcca-3' approached the dog (O3), and when the experimenter (forward) and 5'-cgacc accac gggca gcacc ctcata gcct-3' touched the dog with his or her hand (O4). The tests were (reverse), recognizing an imme dia te upstream region of TM4 conducted more than five times with an interval of more and the portion of TM6 were used for PCR amplification of than a week. The arithmetic means of the observed ranks, the D4 repeat sequence. The final reaction mixture contains excluding the maximum and minimum values, were used 75 mM Tris-Cl, pH 9.0, 15 mM (NH ) SO , 2.5 mM as the fearfulness scores (S1-S4) obtained by O1-O4 tests, 4 2 4 MgCl, 100mg/ml BSA, 200ng of template DNA, 0.4mMrespectively. All the scores were normalized. of each primers, 0.2 mM dNTPs, 2 M betain, and 2 units of In order to find the relationship among the fearfulness Taq polymerase in a total volume of 50 ml. PCR reaction scores for different test objects, correlations between the was performed with 1 cycle at 95 C (2 min), 35 cycles at scores obtained were assessed, followed by principal o o o 95 C (1 min)/62 C (1 min)/74 C (3 min), followed by ancomponent factor analysis with varimax rotation (SAS Inst. extension reaction at 72 C for 20 min. The PCR products Inc., Cary , NC, USA). The scores of principal components were electrophoresed on 2.0% agarose gel and scored by achieved by each dog were also normalized for easy sizes of 718, 769 and 1218 bp for alleles of small (S), lar gecomparison with the original scores. (L) and ultralong (U). Genetic model Fearfulness testThe study used a mixed inheritance model as shown below: A person never exposed to the dogs was considered a y = Xβ + Za + e, stranger. The home cage was fenced surrounding the area of 6 ×4 m, in which the test was carried out. For an entry ofwhere y represents a vector of observations for behavior experimenter, a door was installed on one side. Theperformance, a vector of sex and fixed genotypic effects, following protocol for behavioral test was initially set up.and a a vector of additive polygenic random effects with The experimenter who made observation did not use any the assumption of a~N(0, A σ ), where A is polygenic fragrances such as perfume and lotions. Also to minimize relationship matrix among animals in the pedigree (Quaas, an environmental disturbance regarding odor, windy days 1976). σ is additive polygenic variance component, and e were avoided. Furthermore, the experimenters spent half an is a random vector of residuals with the assumption of hour to get accustomed to odor surrounding the cage priore~N(0, I σ ), where I is identity matrix, and σ residual to an observation. Cages for observation were randomlyvariance component. X and Z are known incidence matrices selected. In order to avoid the influence by the previous relating the fixed and random effects, respectively , to their test, a careful consideration was given. For example, aftercorresponding observations. 78 ANIMAL CELLS AND SYSTEMS Vol. 12 No. 2 Genetics of Canine Fearfulness Parameter estimation receptor gene of the dog population, two primers were Inferences about unknown variance are based on theirdesigned to amplify the corresponding region of 0.7-1.2Kb marginal posterior distribution components in Bayesian in size. Based on the previous identification of three approaches, and the marginalization of the joint posterior different alleles (Fig. 1) by DNA sequencing, allelic distribution can be attained through Gibbs sampling variations in the Sapsari population were genotyped on (Tanner, 1993). The joint posterior distribution of all agarose gel electrophoresis (Materials and Methods). parameters in our model can be expressed by Bayes Among 458 animals tested in twelve pedigrees (Table 1), theorem: the three alleles of S, L, and U were found at the frequencies of .57, .36, and .07, respectively . The genotypes of SS, LL, 2 2 2 2 2 2 2 f(β, a, σ , |y σ) ∝ f(y|β, a, σ, σ ) f(a| σ )π(β)π() σ π() σ SL, SU, and LU among 458 dogs were found at the a e a e a a e frequencies of 0.31, 0.106, 0.458, 0.07, and 0.056. 2 2 2 2 where f(y|β, a, σ, σ )~N(Xβ +Za + I σ and f(a| σ )~N(0 , a e e a 2 2 2 2 A σ ). The π(β), π( σ ), and π( σ ) are the priors for β, , σ Correlations between the behaviors assessed by a a e a and σ , respectively. For the priors, uniform distribution factor analysis was assumed for the fixed effects, and inverse GammaMeasurements of canine fearfulness were conducted for the distribution was assumed for the variance components, test objects or situ atio ns (T able 2). V arious responses of the because the application of the Gibbs sampling with flat dogs to each test were recorded and converted into scores priors for variance components may yield a theoretically as described in the Materials and Methods section. Thus, improper posterior distribution (Hobert and Casella, 1996). each dog has four fearfulness scores, S1 through S4, as Furthermore, very small shape parameters for the inversemeasured in Table 2. They were subject to a factor analysis, Gamma distributions were taken due to a lack of prior from which two major components, F1 and F2 presumably information, and Gibbs sampling was applied to parameter associated with fear response, were drawn, each explaining estimation. It enabled random samples to be drawn with 69 and 17% of the total phenotypic variance, respectively. desired posterior distributions. Figure 2 displays normal distribution of dogs exhibiting Then, initial values were randomly assigned for all dif ferent degrees of F1 or F2 fearfulness, shown here as the unknowns. Sampling from the full conditional posterior normalized scores. The F1 factor is almost equally loaded distributions and updating the distribution based on theonto the four fearfulness scores (0.86, 0.86, 0.86, and 0.72), sample were repeated to the conver gence. Algorithms from while F2 factor in varying degrees ( −0.34, −0.32, 0.12, and Numerical Recipes by Press et al. (1992) were used to 0.65). The result indicates that the general fearfulness is generate the samples. Programs from FSPAK by Misztal represented by F1. An interpretation about F2, however, (1990) were used to deal with sparse matrix inversion. remains to be determined. This result provides a basis for Gibbsit program by Raftery and Lewis (1995) was used to further analysis of th e fearfulness mainly in terms of the F1 determine warming-up periods and thinning intervals. Thescores. Gibbs sampler was run 102,000 rounds, and the first 2000 rounds were discarded as a warming-up period. A Estimations of fixed effects, variance components, and their correlations conservative thinning interval of 100 rounds was used to retain sampled values that reduced lag correlation among Based on the mixed model of inheritance (Materials and thinned samples. The point estimate used in the current Methods), variance components were inferred from the study was the posterior mean estimate, calculated as themarginal posterior distributions. Variances due to polygenic mean of the conditional expected values of the parameters and environmental effects were estimated, as shown in in post warming-up rounds. Table 3, which are statistically significant ( p < 0.05). The polygenic effect ( σ ) on F1 is approximately 20% of the total phenotypic variance, while the environmental effect RESUL TS ( σ ) was about 70%. Therefore, about 10 % of the Allele frequencies of the D4 repeat polymorphism remaining phenotypic variance might be attributed to the The repeat region of the canine D4 receptor was found infixed effects by sex and D4 genotype. rd rd the 3 exon of D4 gene, including the 3 cytoplasmic The posterior mean estimates for the fixed effects due to region of the receptor . This region consists of a novel amino genotype and sex were obtained (Fig. 3), which are acid sequence preceded by approximately 243 amino acids statistically significant ( p < 0.05) for both factors deduced from the N-terminus that are homologous in the mammalian previously . A female is more fearful than a male when the D4 receptors. The block of repeated sequence is followed fixed effect was assumed to be caused only by sex by a homologous region of 89 amino acids extended to the difference. The SS and SU groups are the least fearful, C-terminus. To detect various repeat alleles in the D4 while SL and LU groups are the most fearful groups with ANIMAL CELLS AND SYSTEMS Vo l. 12 No. 2 79 Chaeyoung Lee, Changhoon Kim, S ooan Shin, Daesung Shin, Joohyun Kang, and Chankyu P ark Fig. 1. Repeat p atterns and variations found in the dop ami ne D4 receptor of Sa p s a r i. The basic units with 39, 51, and 39-27-51 base pairs are shown with their nucleotide and amino acid sequences. A variation in their sequences is also designated. In our population of Sap s ari, three different alleles (S, L, and U) were found with different repeat lengths. Table 1. Size and genetic complexity of pedigrees used LL group in between, when the dogs were grouped according to their genotypes disregarding their sexes. The Maximum generations Number of pedigrees Number of dogs differences in fearfulness among genotypes are not caused 7 1 268 by sexual bias, since females are predominant in all 4 3 68 genotype groups. The contribution of D4 genotypes to the 3 8 122 total fixed effect was roughly estimated to be one fifth by Total 12 458 analyzing the variances of fixed effects in each sex, which The number of the dogs included in the principal factor analysis was 264 was based on the assumption that D4 genotype and sex after omitting untested dogs. effects are independent in fearfulness response. Table 2. Measurement of canine fearfulness Role of experimenter Rank O1 O2 O3 O4 2 2 2 2 1Excreting Excreting Excreting Excreting 2 Hiding Hiding Hiding Trembling/rigid 3 Becoming agitated Becoming agitated Running away Becoming rigid 4 Sitting quietly Sitting quietly Sitting down plump Startled 5 Wandering Wandering Sitting quietly Mumbling 6 Approaching Approaching Approach/sitting Sitting quietly 7 Licking hands Dashing gladly Dashing gladly Licking 8 Dashing gladly Lying on its back O1: Outside the cage, O2: Standing still inside the cage, O3: Approaching the dog, and O4: T ouching the dog. Includes urination. 80 ANIMAL CELLS AND SYSTEMS Vol. 12 No. 2 Genetics of Canine Fearfulness Fi g. 2. Sc ore di s tri but i ons of t he fir s t and t he s econd f ac tors . T he s i z e of bi n is 0. 2. T he s haded bars repres ent the number of obs er v ed dogs in each bin and the line represents fitted curve with standard normal distribution. Fig. 3. Fixed effects on F1 and F2 scores of D4 genotypes and sex difference. The filled and empty bars are for the F1 and F2 scores, respectively . Error ranges are indicated by intervals. Although the analyses were primarily focused on the F1 factor score, the F2 results were also included for a comparison. In general, fixed ef fect s of the F2 score are less significant than that of the F1 score. Genot ypes of SS/SU, LL, LU , and SL wi th r egard to F1 and thos e of SS/ LL/LU, SL/S U wi th regard to F2 eac h r epres ent a group, s howing s i gnif i c ant ( p < 0 . 05) dif f erenc es among them. Sex effect s were significant ( p < 0.01) on both F1 and F2. Additive and dominance effects for the dopamine D4 Correlations within polygenic and environmental effects receptor locus were also assessed (Table 3). The additive were assessed for all pairs of the scores (Table 4). The effect of the L allele is negative, while the effect of the S polygenic and environmental correlations for S1, S2, S3, allele is positive. Over-dominance observed between the and S4 are positiv e and ran ging fro m 0.37 to 0.44 an d from alleles, i.e., that the estimates of dominance effect between 0.42 to 0.67, respectively, indicating that the original the S and L alleles are much larger, suggests that variables are better correlated with one another in po lygenic heterozygous individuals ( SL) are more fearful than effects as well as in fixed effects. Correlations between F1 homozygous ones ( SS or LL). More analysis on allelic and F2 are weak, in dicating that they are fairly indep en dent interaction could not be made due to an absence ofeven in the extracted components. individuals with UU genotype. ANIMAL CELLS AND SYSTEMS Vo l. 12 No. 2 81 Chaeyoung Lee, Changhoon Kim, S ooan Shin, Daesung Shin, Joohyun Kang, and Chankyu P ark Table 3. Additive and dominance effects of dopamine D4 receptor type of object, a modification of the test, e.g. an approaching to an animal, was attempted to incorporate the results of the AdditiveDominanceVariance Fac tors 2 2 previous tests (Melzack, 1952). SL SL σ σ a e Although more recent analysis of the canine fear response F10.69-0.15-2.49(L)0.20±0.050.70±0.16 (Goddard and Beiharz, 1985) included a test of dogs to a F2-0.17-0.220.99(S)0.12±0.040.82±0.25 human object with an interval of 6 months, scored as the Alleles in parentheses are regarded as the dominant ones. The additive FM6 and FM12 behaviors, they were classified into a nd d omin an ce ef fec t s of U alle le were n ot e stima ted d ue to an a bse nc e of individual with UU genotype. All the numbers were multiplied by ten. different categories other than the one named ‘fearfulness of object’. It seems odd that the FM12 behavior, the most T able 4. Correlations within polygenic (upper off-diagonal elements) similar to ours, is in the group called ‘general fearfulness’ and environmental (lower off-diagonal elements) estimates among which comprises a variety of behaviors, all involving a the six fearfulness traits sound cue. On the other hand, the object fearfulness defined S1S2S3S4F1F2 by their third d iscriminant function was shown to be higher S10.440.400.370.580.37 in females than in males (Goddard and Beiharz, 1985). S20.670.380.390.630.33 Although the linkage analysis searching for major genes S30.580.540.410.590.32 involved in the fear and anxiety responses has yet to S40420.490.510.490.30 identify a specific gene in mouse (Flint et al., 1995, Dulawa F10.650.660.620.600.15 et al., 1999), the complex nature of these phenotypes may involve a number of genes with minor contributions. As F2 0.40 0.37 0.37 0.32 0.17 1 demonstrated in this study, various behaviors expressing Correlations were inferred based on the posterior mean estimates of poly- genic effects and residuals. canine fearfulness can be explained in part by the genetic diversity found in the D4 locus. In the analytical model, maximum effects of D4 locus was considered because the DISCUSSION estimates might be confounded with the effects of other loci We found that genetic polymorphism at the dopamine D4 that are linked to the D4 locus. Nonetheless, we expected receptor influenced fearfulness in an outbred population of that major contribution would be due to the variation in the a canine breed, Sapsari. The same receptor involved in the dopamine D4 receptor since the estimates in the current dopaminergic pathway of human has been implicated in study were obtained for the D4 locus regardless of their various neurological phenomena including locomotion, direct or indirect effects. A possible concern on the emotional stability , and neuroendocrine function (Civelli et analytical model was that the genotypes of D4 might be al., 1993). Indeed, pharmacological modulation of thisconsidered as random effects. In other words, the genotypes pathway resulted in numerous neurological and behavioral could be determined at random, each drawn from infinite consequences. In spite of these pharma co -medical evidences, population of possible genotypes. Yet in our case, we recent studies of genetic association in human, primarily treated the D4 genotypes as fixed effects since the limited focused on the repeat polymorphism (Wong et al., 2000;number of the genotypes was observed. Lung et al., 2002; Schinka et al., 2002 ) and the 120-bp The statistical genetics approach introduced here to duplication in the 5'-UTR (McCracken et al., 2002) lend reveal an involvement of the D4 polymorphism in canine supports to both positive and negative conclusions. In most fearfulness might provide a sensitive tool for finding a cases, the phenotype concerned was either the novelty-genetic association in multi-factorial trait. Indeed, our result seeking personality trait or the attention deficit/hyperactivity obtained with the canine model strongly supports the role disorder (ADHD). In addition, none of the genome-wide of D4 receptor in determining personality trait, which is screening for related traits including ADHD (Fisher et al.,still controversial in human. 2002) found a linkage to the D4 locus. Fearfulness is a personality trait ex pressing an emotional ACKNOWLEDGMENTS state of an organism, which is also associated with anxiety . C. Lee and C. Kim contributed equally to this work. In animal, a fear response can be elicited by various types This work was supported by grant from the KOSEF. of stimuli associated with learning, ecological history, and signals from their social group. Apparently, the type of REFERENCES stimulus we employed for testing dogs belongs to a category of novel object, i.e. an unexposed stranger. A similar Benjamin J, Li L, Patterson C, Greenberg BD, Murphy DL, and behavioral paradigm was adopted in the systematic study of Hamer DH (1996) Population and familial association canine fear (Melzack, 1952), in which significant difference between the D4 dopamine receptor gene and measures of Novelty Seeking. Nat Genet 12: 81-84. was noted in various dog breeds. Although we used a single 82 ANIMAL CELLS AND SYSTEMS Vol. 12 No. 2 Genetics of Canine Fearfulness Civelli O, Bunzow JR, and Grandy DK (1993) Molecular of the dopamine receptor D4 exon III repetitive region in diversity of the dopamine receptors. Ann Rev Pharmacol several species of the order carnivora. J V et Med Sci 64: 747- T o xi col 33: 281-307. 7 49. Ding Y, Chi H, Grady DL, Morishiwa A, Kidd JR, Kidd KK, Kim KS, Lee SE, Jeong HW, and Ha JH (1998) The complete Flodman P, Spence MA, Schuck S, Swanson JM, Zhang Y , nucleotide sequence of the domestic dog ( Canis familiaris ) and Moyzis RK (2002) Evidence of positive selection actingmitochondrial genome. Mol Phylogen Evol 1 0 : 2 10- 220 . at the human dopamine receptor D4 gene locus. Proc Natl Lung F, Tzeng D, and Shu B (2002) Ethnic heterogeneity in allele Ac a d Sc i U S A 99: 309-314. variation in the DRD4 gene in schizophrenia. Schizophr Res Dulawa SC, Grandy DK, Low MJ, Paulus MP, and Geyer MA57: 239-45 (1999) Dopamine D4 receptor-knock-out mice exhibit Me lzack R (1 952) I r r a t i on al f ear s in t h e do g. Can J P s ych 6: 1 41- reduced exploration of novel stimuli . J Neurosci 19: 9550- 1 47. 95 56. Misztal I (1990) Restricted maximum likelihood estimation of Fisher EF , Francks C, McCracken JT , McGough JJ, Marlow AJ, variance components in an animal model using sparse matrix MacPhie IL, Newbury Low DF , Crawford LR, Palmer CGS, i nve rsio n and a sup e r c omp u ter . J Da i r y S c i 73: 163-172. Woodward JA, Del’Homme M, Cantwell DP, Nelson SF, Niimi Y, Inoue-Murayama M, Kato K, Matsuura N, Murayama Y, Monaco AP , and Smalley SL (2002) A genomewide scan for Ito S, Momoi Y, Konno K, and Iwasaki T (2001) Breed loci involved in attention-deficit/hyperactivity disorder . Am J differences in allele frequency of the dopamine receptor D4 Hum Genet 70: 1183-1196. gene in dogs. J Hered 92: 433-436. Flint J, Corley R, DeFries JC, Fulker DW , Gr ay JA, Miller S, and Press WH, Teukolsky SA, Vetterling WT, and Flannery BP Collins AC (1995) A simple genetic basis for a complex nd (1992) Numerical recipes in FORTRAN. 2 ed. New York, psychological trait in laboratory mice. Scie nce 269: 1432- U S A: Camb r i dg e Univ er si ty P r e ss. 14 35. Quaas RL (1976) Computing the diagonal elements and inverse Goddard ME and Beiharz RG (1983) Genetics of traits which of a large numerator relationship matrix. Bi omet rics 32: 949- determine the suitability of dogs as guide-dogs for the blind. 9 53. Appl Anim Ethol 9: 299-315 Raftery AE and Lewis SM (1995) Implementing MCMC. In: Goddard ME and Beilharz RG (1985) A multivariate analysis of Markov Chain Monte Carlo in Practice, (ed) (Gilks WR, the genetics of fearfulness in potential guide dogs. Behavior Richardson S, andSpiegelhalter DJ). London, UK, Chapman Gen e ti cs 15: 69-89. and Hall, pp 641-649. Ebstein RP , Novick O, Umansky R, Priel B, Osher Y , Blaine D, Schinka JA, Letsch EA, and Crawford FC (2002) DRD4 and Bennett ER, Nemanov L, Katz M, and Belmaker RH (1996) novelty seeking: Results of meta-analyses. Am J Med Genet Dopamine D4 receptor (D4DR) exon III polymorphism 1 14: 6 43- 648 . associated with the human personality of Novelty Seeking . Tanner MA (1993) Tools for Statistical Inference: methods for the N a t Ge n e t 12: 78-80. exploration of posterior distributions and likelihood functions, Harville DA (1974) Bayesian inference for variance components rd 3 ed. New York, USA , Springer-Verlag. using only error contrasts. Biometrika 61: 383-385. W ong AH, Buckle CE and V an T ol HHM (2000) Polymorphisms Hobert JP and Casella G (1996) The effect of improper priors on in dopamine receptors: what do they tell us? Eur J Pharmacol Gibbs sa mpling in hierarchica l lin ear mixed mo dels. J A m S t at 410: 183-203. Assoc 91: 1461-1473. Inoue-Murayama M, Matsuura N, Murayama Y, Tsubota T, [Received March 10, 2008; accepted May 19, 2008] Iwa s aki T , Ki tag a w a H, a n d Ito S (20 02) Se que nc e Co m p a r is on ANIMAL CELLS AND SYSTEMS Vo l. 12 No. 2 83
To get new article updates from a journal on your personalized homepage, please log in first, or sign up for a DeepDyve account if you don’t already have one.
All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.
