Edited by: Jessica Lee Oliva, Monash University, Australia
Reviewed by: Carolyn Walsh, Memorial University of Newfoundland, Canada; Lesley J. Rogers, University of New England, Australia; Lauren E. Thielke, Oregon State University, United States
*Correspondence: Enikő Kubinyi
This article was submitted to Comparative Psychology, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Meeting humans is an everyday experience for most companion dogs, and their behavior in these situations and its genetic background is of major interest. Previous research in our laboratory reported that in German shepherd dogs the lack of G allele, and in Border collies the lack of A allele, of the oxytocin receptor gene (OXTR) 19208A/G single nucleotide polymorphism (SNP) was linked to increased friendliness, which suggests that although broad traits are affected by genetic variability, the specific links between alleles and behavioral variables might be breed-specific. In the current study, we found that Siberian huskies with the A allele approached a friendly unfamiliar woman less frequently in a greeting test, which indicates that certain polymorphisms are related to human directed behavior, but that the relationship patterns between polymorphisms and behavioral phenotypes differ between populations. This finding was further supported by our next investigation. According to primate studies, endogenous opioid peptide (e.g., endorphins) receptor genes have also been implicated in social relationships. Therefore, we examined the rs21912990 of the OPRM1 gene. Firstly, we found that the allele frequencies of Siberian huskies and gray wolves were similar, but differed from that of Border collies and German shepherd dogs, which might reflect their genetic relationship. Secondly, we detected significant associations between the OPRM1 SNP and greeting behavior among German shepherd dogs and a trend in Border collies, but we could not detect an association in Siberian huskies. Although our results with OXTR and OPRM1 gene variants should be regarded as preliminary due to the relatively low sample size, they suggest that (1) OXTR and OPRM1 gene variants in dogs affect human-directed social behavior and (2) their effects differ between breeds.
To investigate the comparative biology of human behaviors and uncover the genetic background of behavior disorders, in the past decade several research groups have studied the effects of gene variants on dog behaviors (Hall and Wynne,
Oxytocin is an evolutionarily highly conserved neuropeptide that plays an important role in various complex social behaviors, such as social cognition, trust, attachment, and sociability (Donaldson and Young,
When examining the genetic background of the oxytocin receptor gene, dog breeds differ from wolves (
On a genetically varied sample, Ottenheimer-Carrier et al. (
The oxytocin system interacts with the opioid system that modulates reward, motivation, emotional responses, cognition, nociception, and autonomic functions. In females, opioids inhibit oxytocin release, especially at mu and kappa receptors (Vuong et al.,
Several other studies implicate a role for endogenous opioid peptides (e.g., endorphins) in forming stable social relationships, such as pair bonding and attachment [e.g., in prairie voles (Burkett et al.,
Morphine was the first chemical discovered to bind to mu opioid receptors. Genetic variations in the mu 1 opioid receptor genes (OPRM1) may primarily explain individual variation in the development of social relations in humans [parent-child interaction (Copeland et al.,
In captive rhesus macaques, functional polymorphisms in the OPRM1 gene have been identified that are associated with social behaviors. Infants carrying the G allele of the C77G SNP exhibited higher levels of attachment behavior and higher distress to separation from their mothers, and they spent more time with their mothers upon reunion than individuals homozygous for the C allele. C/G infants were also less likely to interact with other individuals in the group (Barr et al.,
In dogs, low doses of exogenous opiates have been found to significantly reduce distress vocalization and activity in socially isolated puppies (Panksepp et al.,
In this study, we present data on genetic associations between the oxytocin and mu opioid receptor gene variants and greeting behavior in dogs. Dog breeds and breeding lines differ in several aspects of their social behavior toward humans (for a review see Mehrkam and Wynne,
Previous findings support the latter, more complex scenario. Hejjas et al. (
The above findings suggest that with some divergence, gene-behavior associations might be similar between breeds. However, the study of Kis et al. (
Certainly, the allele frequencies that are typical for a breed could modify the association patterns. Nevertheless, the question remains: how generalizable are gene-behavior associations in dogs?
In this study, we used the Greeting Test behavioral data collected by Kis et al. (
We also ran two pilot studies, which are presented in the
One hundred and three border collies (1–12-years-old; mean age ±
None of the subjects were closely related, i.e., littermate and parent-offspring relationships were excluded.
For assessing the allele frequencies of the opioid receptor genes, 42 gray wolves (male: 45%) living in Hungarian, Serbian, Austrian, and German zoos and parks were also genotyped but did not participate in the Greeting Test.
Buccal samples were collected from subjects in a non-invasive way, with cotton swabs from the inner surface of the cheek. Genomic DNA was isolated as described in a previous study (Kotyuk et al.,
We investigated the three OXTR SNPs studied by Kis et al. (
Mu opioid receptor gene (OPRM1) rs21912990 SNP (NCBI,
The genotyping of kappa opioid receptor gene (OPRK1) rs23478162 SNP is described in the
A female, unfamiliar experimenter greeted the dog, who was kept on leash by the owner (Figure
Sequence of video frames from a Greeting Test.
In an open, undisturbed area, the owner stood in place while holding the dog on a 1.5–2 m leash. The dog was allowed to move freely within the range of the stretched leash and was not corrected or rewarded for any behavior. A female experimenter (unfamiliar to the dog) approached the dog in a friendly manner (verbally greeted the owner and the dog and smiled). When the dog acted “friendly” (moved toward the experimenter with affiliative behaviors) or showed neutral behavior, the experimenter stepped toward the dog and patted its head, back, and shoulders. Then she stepped 1 m sideways within reach of the leash and waited 2–3 s in order to check whether the dog followed her. If the dog showed aggressive behavior (e.g., barking or growling), the experimenter stayed out of reach of the leash, crouched, and tried to call the dog. If the dog approached the experimenter and was not aggressive, she followed the protocol above. If it was not possible to approach the dog, the test was terminated in 30 s.
In the study of Kis et al. (
Previous studies have shown that the “Latency of approaching” and “Latency of following” variables of the Greeting Test correlated with approaching the experimenter in other contexts (separation from the owner, calling after threatening, Kis et al.,
SNPs deviating from Hardy-Weinberg equilibrium or with minor allele frequency below 0.05 were removed. SPSS 21.0 for Windows was used for all statistical analyses. General Linear Models (GLM) were used to test the effects of age (as covariate), sex, and SNP (with SNK
The descriptive statistics of the greeting score are shown in Table
Total number of participants in the Greeting Test, greeting score frequencies (%), and means for the three dog populations.
Border collie | 103 | 5.8 (6) | 16.5 (17) | 77.7 (80) | 1.72 (0.57) |
German shepherd dog | 104 | 2.9 (3) | 11.5 (12) | 85.6 (89) | 1.83 (0.45) |
Siberian husky | 96 | 8.3 (8) | 26.0 (25) | 65.6 (63) | 1.57 (0.64) |
Table
OXTR SNP genotypes (GT), number of individuals by genotype (
AA | 6 | 7 | AA | 1 | 1 | AA | 1 | 1 | |
AG | 28 | 31 | AT | 7 | 7 | AG | 8 | 8 | |
GG | 57 | 63 | TT | 88 | 92 | GG | 94 | 91 | |
Allele freq | G | 0.78 | T | 0.95 | G | 0.95 | |||
0.96 | 3.25 | 2.61 | |||||||
0.33 | 0.07 | 0.11 | |||||||
AA | 12 | 12 | AA | 40 | 40 | AA | 38 | 38 | |
AG | 48 | 49 | AT | 46 | 46 | AG | 49 | 49 | |
GG | 37 | 38 | TT | 15 | 15 | GG | 14 | 14 | |
Allele freq | G | 0.63 | T | 0.38 | G | 0.38 | |||
0.35 | 0.09 | 0.08 | |||||||
0.55 | 0.77 | 0.78 | |||||||
AA | 17 | 22 | AA | 0 | 0 | AA | 0 | 0 | |
AG | 23 | 30 | AT | 1 | 1 | AG | 12 | 15 | |
GG | 37 | 48 | TT | 74 | 99 | GG | 68 | 85 | |
Allele freq | G | 0.63 | T | 0.99 | G | 0.93 | |||
9.94 | 0.00 | 0.53 | |||||||
0.00 | 0.95 | 0.47 |
We did not find an age or sex effect in any of the populations. One gene-behavior association was detected in Siberian huskies, and the results of the association analysis are presented in Figure
Association between OXTR 19208A/G genotypes and greeting scores (mean + SE) in Siberian huskies (
The OPRM1 rs21912990 SNP was not in linkage disequilibrium with the OXTR SNP-s in either breed (
The genotype distribution, allele frequencies and Hardy-Weinberg equilibriums are presented in Table
OPRM rs21912990 genotype (GT), number of individuals by genotype (
CC | 27 | 33 | |
CT | 47 | 58 | |
TT | 7 | 9 | |
Allele freq | T | 0.38 | |
4.51 | |||
0.03 | |||
CC | 11 | 20 | |
CT | 27 | 50 | |
TT | 16 | 30 | |
Allele freq | T | 0.55 | |
0.00 | |||
0.95 | |||
CC | 68 | 78 | |
CT | 18 | 21 | |
TT | 1 | 1 | |
Allele freq | T | 0.11 | |
0.02 | |||
0.87 | |||
CC | 31 | 74 | |
CT | 11 | 26 | |
TT | 0 | 0 | |
Allele freq | T | 0.13 | |
0.95 | |||
0.33 |
Association analysis of behavioral and genetic data on the OPRM1 rs21912990 SNP was conducted in all three dog breeds. There was no association in Siberian huskies [
Results of the genetic association analysis for the two herding breeds and OPRM1 rs21912990 SNP are shown in Figure
Association between OPRM1 rs21912990 genotypes and greeting scores (mean + SE) in Border collies (
In Border collies OPRM1 rs21912990 T/T homozygotes tended to have lower scores than C/T and C/C dogs [
In German shepherd dogs, OPRM1 rs21912990 T/T homozygotes had lower scores than C/T heterozygotes [
In this study, we assessed the potential association of oxytocin and opioid receptor gene SNPs with greeting an unfamiliar human in three dog populations: pet Border collies, pet German shepherd dogs, and racing Siberian huskies. The genes were selected as candidates because they have been implicated in the development of complex social behaviors in mammals (e.g., Copeland et al.,
We found that Siberian huskies that lacked the C allele of the OXTR 19208A/G SNP had higher scores in the Greeting test, i.e., they approached a friendly unfamiliar woman more frequently in a non-aggressive way. The same polymorphism was linked to the friendliness behavioral scale in the study of Kis et al. (
In the Kis et al. (
In conclusion, although the OXTR gene seems to be associated with the broad dimension of human directed social behavior, the precise variables of this dimension may or may not be related to the specific polymorphisms of the gene in different breeds. Thus, the relationship between behavior and genes is complex and breed specific. This is not surprising if we consider that (1) behavior traits are usually associated with many genetic variants, each of which has only a subtle effect on the behavior, thus gene-gene and gene-environment interactions could easily obscure them; (2) the relationship between genetic variations and allele frequencies are different between isolated dog populations with different ancestries; (3) alleles may not directly affect the phenotype, but are correlated with a causative allele located nearby—and the relationship between locations differ between breeds.
In the present study, we investigated not only OXTR, but another gene, which has been implicated in social behavior, the OPRM1. The allele frequencies of the rs21912990 SNP were similar in Siberian huskies and gray wolves, but differed from that of Border collies and German shepherd dogs, which might reflect their genetic relatedness and/or more similar social behavior.
The OPRM1 rs21912990 SNP tended to associate with greeting behavior in Border collies and was also significantly linked to this behavior in German shepherd dogs. T/T homozygotes approached the unfamiliar human less frequently than other dogs. This result is in harmony with findings in rhesus macaques (Barr et al.,
Although the results should be regarded as preliminary due to the relatively low sample size, our results highlight how the social behavior of dogs toward humans is influenced by the oxytocin and opioid system, but the links between SNP-s and behavioral variables might differ by breeding populations.
Non-invasive studies on dogs are currently allowed to be done without any special permission in Hungary by the University Institutional Animal Care and Use Committee (UIACUC, Eötvös Loránd University, Hungary). The currently operating Hungarian law “
The idea for the paper was conceived by EK and ÁM. The experiments were designed by EK, MB, ZR, and MS. The experiments were performed by EK, MW, MB, DK, and EP. The data were analyzed by EK. The paper was written by EK, MB, and MS.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We thank Borbála Turcsán and Lisa Wallis for useful comments on a previous version of the manuscript. We are grateful for Judit Bánffyné Vas, Ildikó Brúder, Anita Illés, Borbála Turcsán, and Anna Vargáné Kis, who helped in the behavioral testing of the dogs. This project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (Grant Agreement No. 680040), the Hungarian Academy of Sciences (MTA 01 031) and the János Bolyai Research Scholarship (for EK), the Hungarian Scientific Research Fund (K100695; ANN107726), and the WWTF (CS11-026).
The Supplementary Material for this article can be found online at:
1Position number is determined according to the most recent genome version, this explains the change of nomenclature compared to Kis et al. (
2See above.