Abstract
Background
The genetic and molecular mechanisms of tameness are largely unknown. A line of silver foxes (Vulpes vulpes) selected for non-aggressive behavior has been used in Russia since the 1960's to study the effect of domestication. We have previously compared descendants of these selected (S) animals with a group of non-selected (NS) silver foxes kept under identical conditions, and showed that changes in the brain transcriptome between the two groups are small. Unexpectedly, many of the genes showing evidence of differential expression between groups were related to hemoproteins.
Results
In this study, we use quantitative RT-PCR to demonstrate that the activity of heme related genes differ between S and NS foxes in three regions of the brain. Furthermore, our analyses also indicate that changes in mRNA levels of heme related genes can be well described by an additive polygenic effect. We also show that the difference in genetic background between the two lines of foxes is limited, as estimated by mitochondrial DNA divergence.
Conclusion
Our results indicate that selection for tameness can modify the expression of heme related genes in canid brain regions known to modulate emotions and behavior. The possible involvement of heme related genes in behavior is surprising. It is possible that hemoglobin modulates the behavior of canids by interaction with CO and NO signaling. Another possibility is that hemorphins, known to be produced after enzymatic cleavage of hemoglobin, are responsible for behavioral alterations. Thus, we hypothesize that hemoglobin metabolism can be a functionally relevant aspect of the domestic phenotype in foxes selected for tameness.
Similar content being viewed by others
Background
Domestication of animals can be described as the process by which wild animals adapt to coexistence with humans and to the environment we provide. Behavioral adaptation to the human environment is achieved through genetic changes occurring over generations together with environmental stimulation and experience during an animals lifetime [1]. A common aspect of early domestication is strong artificial selection for animals showing weak defensive behavior and low aggression towards humans [1], resulting in tame behavior.
The domestic dog (Canis familiaris) originated from the domestication of wild gray wolves (Canis lupus) at least 15,000 years ago [2, 3]. The molecular mechanisms responsible for the behavioral changes associated with domestication in dogs are not known. We have previously shown that the brain gene expression profile of domestic dogs has diverged from that observed in their wild ancestor the gray wolf, suggesting that behavioral selection during domestication has modified mRNA expression levels of genes with multiple functions [4]. However, from the comparison of domestic animals with their free-living wild ancestor it is not possible to separate phenotypic differences that have a genetic background from those that are caused by differences in the animal's external environment and experiences. We therefore decided to study domestication in another model where environmental conditions could be well controlled. In the 1960s D.K. Belyaev, the late director of the Institute of Cytology and Genetics in Novosibirsk, initiated a domestication experiment on farmed silver foxes, a color morph of the red fox (Vulpes vulpes), to test the hypothesis that physiological and morphological changes in the domestic dog could have resulted from selection for tameness alone [5]. In this experiment, that still continues, farmed foxes have been selected for non-aggressive behavior towards man for more than 40 generations. The selection has resulted in animals that not only are friendly towards humans, but also are skilled as dogs in communicating with people [6]. Moreover, developmental, morphological and neurochemical changes (including modulations of the serotonergic system and the activity of the hypothalamus-pituitary-adrenal (HPA) axis) are associated with the behavioral changes in the selected foxes [7–10].
We have previously compared the global brain gene expression of descendants of the selected (S) silver foxes with a group of non-selected (NS) foxes, raised under identical conditions in a Norwegian farm [11]. Our microarray results suggested that behavioral changes caused by selection for tameness may be associated with only limited changes in the brain transcriptome. Moreover, our results indicated that expression of several genes related to heme have been affected by behavioral selection [11]; as many as six of the 14 clones showing the largest expression differences coded for hemoproteins.
In our previous experiments, mRNA samples from foxes were cross hybridized to human cDNA microarrays. Although this is a well accepted method for comparing transcriptional profiles within species [12], it requires an additional step of identifying the orthologous genes in fox that were responsible for the signals. Few fox mRNA sequences are available in public databases. However, chromosomal organization is almost the same in dogs and red foxes [13, 14], and the sequence homology may be as high as 99.5% for individual genes [15]. Consequently, in this study we used dog sequences to design primers for quantitative real-time PCR (qPCR) to compare the mRNA levels of five heme-related canine sequences in the brains of the original sample set of S and NS foxes. To determine whether the mode of inheritance is primarily additive, we included animals which resulted from crossing the two original lines. In addition, we assess the divergence in genetic background between the S and NS lines, by comparing the left domain of the mitochondrial DNA (mtDNA) control region.
Results
Estimation of mtDNA divergence between selected and non-selected farm foxes
To estimate the genetic variability and degree of differentiation between S, NS and wild foxes we compared 333 base pair (bp) sequences of the left domain of the mtDNA control region. These sequences were complemented with others corresponding to wild red foxes deriving from public databases (GenBank). The comparison of the mtDNA sequences were made for a total of 29 NS, 23 S, 23 Cross, 12 wild foxes and an additional 26 wild foxes retrieved from databases. Five different mitochondrial DNA haplotypes were observed in farm foxes, three of them in the NS and two in the S foxes. All hybrids (Cross) shared one of the S haplotypes as the females for line crosses always were from the S line and the mtDNA is maternally inherited. All sequences observed in farm foxes were very similar. Their pair-wise sequence divergence was lower than 3% in all cases and all clustered in one clade in a neighbor-joining tree (Figure 1). The genetic divergence between all farm fox haplotypes was similar to that observed in wild foxes sampled in a geographically limited region in Sweden. The four haplotypes observed in this wild population occupy different positions in the neighbor-joining tree. This indicates that the differences observed among farm animals (either selected or not) are comparable to the differences observed within one single natural population, and that S and NS lineages are closely related, with their haplotypes intermixed within one clade.
Neighbor-joining phylogenetic tree based on mtDNA control region. Sequences corresponding to the non-selected (NS) foxes (□), selected (S) foxes (■) and wild foxes from one population (*). Sequences of other wild foxes obtained from GenBank are indicated by a dot. Bootstrap support is indicated when higher than 50%.
Sequence similarity of heme-related genes between humans and canids
Since the expressed sequences in foxes should be very similar to sequences in dogs, we identified dog mRNA species that could have yielded one or more of the signals from the heme-related human clones on the arrays in our previous study [11] (Table 1). These dog mRNA species corresponded to five predicted dog genes, including four hemoglobins, namely, the beta (HBB) and epsilon (HBE) chains (located on canine chromosome 21), the alpha (HBA) and zeta (HBZ) chains (ESTs that mapped to canine chromosome 6), and the heme binding protein 1 gene (HEBP1) (located on canine chromosome 27). The canine mRNA sequence of HBB [GenBank accession number: LOC480784] and HBE [GenBank: LOC485256] were similar, as evident from a pair-wise blast (364 identities of 450 bp, e-value = 5e-55), and consequently, hybridization of either of the two mRNA species in foxes could have yielded the signals after hybridization to three human clones: 242594, 292528 and 239611. Similarly, the two canine ESTs, similar to human HBA [GenBank: CN003912] and HBZ [GenBank: DN366534], showed a high sequence similarity (207 identities of 270 bp, e-value = 4e-33), and either of the sequences expressed in fox could have yielded the signals on the two human clones 207558 and 248463. Finally, we could only identify one canine gene, [GenBank: LOC477690] which showed sequence similarities with the human clone 730346, encoding HEBP1.
Expression levels of hemoglobin related genes are modified by selection for tameness
To examine which of the canine heme-related genes were differentially expressed in brains of selected silver foxes, and whether the mRNA levels showed evidence of a polygenic additive mode of inheritance, we designed primers based on canine sequences and quantified the mRNA levels corresponding to HBB, HBE, HBA, HBZ and HEBP1 in three regions of the brain in 9 S foxes, 17 NS and 12 Cross and 12 wild. Three hemoglobin genes, HBE, HBB and HBA, and the heme binding protein HEBP1 showed significantly lower mRNA levels in S foxes than in NS foxes in all brain regions (Figure 2). The response was similar in all three brain regions and there were no significant interactions between origin and region (Table 2). The difference between S and NS foxes was 3.1-fold for HBE, 2.3-fold for HBB, 2.2-fold for HBA and 1.5-fold for HEBP1 (Table 3). The mRNA levels of the foxes resulting from the crosses did not provide any strong evidence for a composite dominance effect (p-values > 0.13 for all genes, Table 3), and thus the major variation in mRNA levels between S and NS foxes appears to be due to a composite additive genetic effect. However due to the limited sample sizes, this study was underpowered to detect a moderate dominance effect. The power to detect a dominance effect between the two parental lines (d = a) were 0.21, 0.47, 0.40, and 0.19 for HBA, HBB, HBE and HEBP1, respectively. Consequently, it is possible that a true dominance effect of this size may have been overlooked.
Gene expression levels of heme related genes in fox brain. Average mRNA levels of hemoglobin alpha (HBA), beta (HBB) and epsilon (HBE) chains, and heme binding protein (HEBP1) are shown for four groups of foxes, Vulpes vulpes. Wild = wild red foxes (N = 12), NS = non selected farm foxes (N = 17), S = farm foxes selected for tame behavior (N = 9), Cross = off-spring from crosses and back crosses of NS and S foxes (N = 12). Least square means and standard errors are given.
To estimate heme-related gene expression in blood cells, we also performed qPCR on dog blood, where all five examined transcripts were detected. Thus to evaluate the potential effect of blood perfusion on HBE and HEBP1 expression levels, we assumed that the average HBA and HBB levels (HB) could represent the maximum effect of blood perfusion caused by e.g. a systemic stress response. The HBE levels clearly increased with the normalized HB expression (p-value < 0.0001). However, even in presence of a perfusion effect S foxes had a 1.8-fold lower HBE mRNA levels than NS foxes (p-value = 0.031). The normalized HB level did not significantly affect the expression of HEBP1 (p-value = 0.18). Still if the perfusion effect was included in the statistical model, the difference in HEBP1 levels were reduced to 1.3-fold, suggesting that part of the HEBP1 difference between the S and NS foxes may be linked to the average HB mRNA levels. These results suggest that, while we cannot exclude a perfusion effect, part of the heme-related gene alterations is independent of this effect.
When we compared farmed foxes (S, NS, Cross) versus wild animals, we noted that HBE was the only gene to have mRNA levels that differed significantly between these two groups, with 2.7-fold higher levels in the wild foxes as compared to farmed (Table 3). This difference is difficult to explain as changes in environment and time post-mortem are confounded with the differences between the groups. Possible differences in HBZ levels could not be reliably quantified with our qPCR protocol because its mRNA levels were below the detection limit for almost half of the investigated samples.
Discussion
We have previously shown that dramatic behavioral changes caused by selection for tameness may be associated with only limited changes in the brain transcriptome. In this study, we first evaluated whether differences in genetic background between S and NS farm foxes (rather than genetic differences due to selection for tameness) may be responsible for the observed expression differences, by estimating the mtDNA divergence between S and NS lines. Our results showed that the mtDNA lineages present in S and NS lines were similar to each other and did not cluster in reciprocally monophyletic groups. Although both NS and S foxes were likely to have reduced genetic diversity due to founding effects and subsequent drift, our results suggest that they are not likely to derive from wild populations that had been isolated from each other for a long time [16]. In fact, all farmed silver foxes are said to have originated from animals captured in Canada over a Century ago [17]. This could explain their differentiation from wild red foxes from Sweden and implies that S and NS foxes have a common origin.
Our previous microarray results suggested that heme-related genes in particular, show pronounced expression differences between S and NS foxes. Thus, in this study we used qPCR to quantify mRNA levels of five heme-related genes in S, NS, and foxes resulting from crossing the two parental lines, to determine whether we could find any evidence of an additive genetic effect underlying transcriptional differences between S and NS animals. We found that four of the five investigated genes, HBA, HBB, HBE and HEBP1, showed decreased expression levels in foxes selected for tameness compared to non-selected farm foxes. If multiple loci controlling gene expression have been under selective pressure, then we would expect that the expression variation between the two experimental populations (and the crosses between them) should primarily follow an additive mode of inheritance. For three of the genes, HBA, HBE and HEBP1, an additive genetic component did indeed appear to explain the observed expression variation. Thus, our results suggest that a selective pressure for tameness may result in direct or indirect modulation of heme and hemoglobin metabolism. However, for the fourth gene, HBB, the mRNA levels in foxes resulting from crossing and back-crossing the parental lines exceeded those of both the S and NS lines, and a simulation analysis revealed that moderate deviations from additivity could not be ruled out for any of the investigated genes given the limited power of the experiment.
The comparisons with animals in the wild are more difficult to interpret. However, from the comparison of S foxes with wild foxes it is clear that the abundance of each individual transcript does not predict tameness independently of context. These results are not surprising since we have previously shown that environmental effects together with adaptation to life in captivity have a much larger influence on brain gene expression in foxes, than selection for tameness [11]. It is possible that mRNA levels of heme related genes also are affected by the living conditions of foxes, and that mRNA levels in wild foxes have been further modified by mRNA degradation (since the post mortem interval in wild foxes was unknown). Thus strict interpretation of the comparisons of mRNA levels in wild and farmed foxes is not possible, due to the confounding effect of selection with systematic differences in environmental variables. Our results thus demonstrate the need for rigorous control of living environment and experimental conditions when studying the evolution of the brain transcriptome.
Four of the five investigated genes encode hemoglobin subunits. The HBA, HBB, HBE and HBZ genes encode globins (alpha, beta, epsilon and zeta) which, together with iron-containing heme groups, form hemoglobin subunits [18]. The fifth gene, HEBP1, encodes a monomeric and soluble protein that binds heme in both man and mouse [19]. HEBP1 is induced during erythroid differentiation and the protein is likely to be involved in utilization of heme for synthesis of heme proteins [20]. The coordinated response to selection for tameness in these genes suggests that heme metabolism and hemoglobin levels may have been modulated in the brains of tame animals.
In humans, hemoglobin alpha- and beta transcripts are known to be expressed in both leukocytes and reticulocytes of circulating blood [21, 22]. In the normal human adult, alpha- and beta chains account for more than 98% of the total globin chains in blood [21]. As hemoglobin alpha- and beta also are the two dominating globins in fox blood [23], it is possible that the increased mRNA levels of these genes in NS foxes could be a manifestation of an increased perfusion in the brain as a direct consequence of a stronger stress response to human handling. However, HBA and HBB differences between S and NS foxes could only partly account for the observed decreased levels of HBE and HEBP1 in selected foxes. Thus, although a systemic response to stress may have contributed to the observed differences in levels of hemoglobin transcripts, our results suggest that selection may also have direct effects on the heme metabolism in the brain of tame animals. Possible consequences of an altered heme metabolism in the brain on tame behavior are discussed below.
Although involvement of hemoglobin genes in behavior is surprising, several observations support this possibility. First, a gene involved in behavior should be expressed in brain cells, and both hemoglobin α and β are expressed in neurons and glial cells [24]. Second, besides the well known role as an oxygen transport molecule, hemoglobin also binds carbon monoxide (CO) and nitric oxide (NO) with high affinity, and acts as an extra-cellular scavenger of these signaling molecules in the mammalian brain [20]. Evidence from in vivo studies demonstrates that CO produced from the enzymatic cleavage of heme, has important biological functions in the control of the stress response of mammals [25, 26]. Nitric oxide also modulates the neuroendocrine responses and behavior that is linked to serotonin and the HPA-axis [27–29]. A third evidence for a potential role of hemoglobin in behavior is based in the knowledge that β-, κ-, δ- and ε-chains of hemoglobin also serve as the main precursors for hemorphins, a class of endogenous "non-classical" opioid peptides [30]. Hemorphins are naturally occurring peptides in brain, plasma and cerebrospinal fluid, with high binding affinity to μ- and δ-opioid receptors [30]. Although the function of hemoglobin in the brain is not known, hemorphins induce release of β-endorphin and dynorphin, endogenous opioids that are released in response to acute stress or pain [31, 32]. From these results we postulate that hemorphins may also modulate stress responses in the fox. Similarly, it has been suggested that hemorphins produce dampening of exploratory and aggressive behavior in mice [33, 34].
We have previously showed that genetic changes in behavior caused by selection for tameness may be associated with only limited changes in the brain transcriptome [11]. In this study, we present evidence indicating that selection for tameness has modulated heme and hemoglobin metabolism in the brain, and that the (additive) mode of inheritance of mRNA expression is similar to that of tame behavior. It can not be ruled out that part of the observed response of heme related transcripts is the result from a systemic stress response. However, our results indicate that selection has also had direct effects on the heme metabolism in the brain of tame animals, and we hypothesize that heme-related genes are directly, or indirectly, involved in behavior related to tameness.
Experimental Procedures
Animals
Samples from two groups of farm silver foxes (Vulpes vulpes), selected (S) and non-selected (NS) for tame behavior, were used in this study. The selected (S) foxes were descendants from two male and three female Russian silver foxes that were imported to Norway in 1996 from the Institute of Cytology and Genetics at the Russian Academy of Sciences (Novosibirsk, Russia) and belonged to a line of foxes that had been selected for tame behavior for more than 40 generations [8]. The non-selected (NS) animals consisted of Norwegian farm-bred silver foxes that had not been selected for tame behavior. We also included animals from crossing (Cross) the two parental lines(F1) and fromback-crossing the F1 animals to the NS line. The silver foxes were kept at a farm belonging to the Norwegian University of Life Science (Ås, Norway) and housed in cages of standard dimensions in accordance with the regulations of the Norwegian Ministry of Agriculture and Food. Foxes of the S and NS lines were randomly located in cages within the same building, were treated in the same way, had equal amount and kind of interaction with humans and were fed simultaneously. The foxes were euthanized in a random order for all groups during two consecutive days during January 2004, and all brains were removed and frozen to -80°C within 20 minutes after death. Samples of wild red foxes (wild) from central Sweden were used for comparison. These animals were shot during fox hunts in February 2004 and submitted by hunters to the Swedish National Veterinary Institute for necropsy, where brains were collected and frozen to -80°C. This resulted in a wider variation in postmortem intervals, which might have affected transcript levels, along with several other environmental differences (e.g. living conditions, diet, method used to kill them). Sex and age of individual foxes used in this study can be found in Table S1 in the Supplementary Material (see Additional file 1). All foxes studied died for other reasons than their participation in this study.
Tissue samples and isolation of nucleic acids
Amygdala, frontal lobe and hypothalamus were morphologically identified with the support of a pathologist at the Swedish National Veterinary Institute. Samples from the three regions were processed as previously described [4, 11]. Briefly, frozen tissue was homogenized in TRIzol®, and RNA was extracted according to the manufacturers' instructions (Invitrogen™, Life Technologies). About 10–30 μg of tissue were digested over night at 37°C in 500 μl of Laird's buffer (0.1 M Tris, 5 mM EDTA, 0.2 M NaCl, 7 mM SDS, adjusted to pH 8.5) with 0.3 mg proteinase K. DNA was extracted using a modified phenol/chloroform protocol [35]. For some samples, DNA was extracted from the TRIzol® leftovers after RNA isolation using the PureGene® DNA kit (Gentra systems). The concentrations of extracted RNA and DNA were measured using a NanoDrop® ND-1000 instrument (NanoDrop Technologies, USA).
MtDNA control region amplification and sequencing
For the mtDNA analysis, DNA was extracted from 29 NS, 23 S, 23 Cross (16 F1 + 7 backcrosses) and 12 wild foxes. The left domain of the mitochondrial control region was PCR-amplified and sequenced using the primers, Thr-L 15926 5'-CAA TTC CCC GGT CTT GTA AAC C-3' and DL-H 16340 5'-CCT GAA GTA GGA ACC AGA TG-3' [36]. The amplification was done in a reaction volume of 10 μl, which included 1 × HotStar buffer, 0.5 μM of each primer, 0.2 mM dNTP, 2.25 mM MgCl2, 0.45 U HotStarTaq® and 20 ng DNA. A negative control was included in the PCR reaction to check for contamination. The PCR profile included an initial denaturation step at 95°C for 7 min followed by 35 cycles of amplification (denaturation at 95°C for 1 min, annealing at 50°C for 2 min and extension at 72°C for 1 min and 30 s) and a final extension at 72°C for 10 minutes. PCR products were then purified with ExoSAP-IT (Amersham Biosciences). Sequencing of the purified PCR products was done on a MegaBACE™ 1000 (Amersham Biosciences) following protocols and using chemistry recommended by the manufacturer. The forward and reverse sequences for each fragment were aligned with Sequencher 4.1.4 (Gene Codes Corporation, Ann Arbor, MI) and consensus sequences were constructed.
Phylogeny reconstruction
A neighbor-joining tree was built with sequences corresponding to the farm foxes and wild foxes. Additionally to these, 26 red fox sequences were obtained from public databases as well as one arctic fox (Alopex lagopus) that was used as outgroup. A neighbor-joining phylogenetic tree was obtained with PAUP 4.0b10 [37] using the Kimura 2-parameter model of sequence evolution, with a gamma shape parameter a = 0.5. The support for the tree topology was evaluated with 1000 bootstrap pseudo-replicates.
Real-Time RT-PCR assay
Messenger-RNA was isolated from amygdala, hypothalamus and frontal lobe from a total of 50 foxes, including 17 NS, 9 S, 12 Cross (6 F1 + 6 backcross) and 12 wild foxes as described previously [11]. In brief, each sample was extracted with PolyATtract® mRNA Isolation Systems (Promega) and converted to cDNA using TaqMan® Gold RT-PCR Kit (Applied Biosystems). Preparations of 21 μl qPCR reactions using SYBR® Green PCR Core Reagents kit (Applied Biosystems) were added to plates where 4 μl of all 150 samples (50 individuals × 3 brain regions) were randomly distributed on two 96-well plates according to a balanced incomplete block design with respect to origin (S, NS, Cross and wild), sex and tissue. The sequences of the human clones corresponding to HBG1, HBG2, HBE, HBZ/HBA1, HBZ, and HEBP1 present in the arrays used for the study of Lindberg et al. (2005) were used for BLAST searches of the dog genome, and the homologous dog sequences were used to design qPCR primers. Primers were also designed for the reference genes GAPDH (Glyceraldehyde-3-phosphate dehydrogenase) and ACTB (Actin, beta); all primer sequences are available in the Supplementary Material (see Additional file1). The qPCR reactions were run on an ABI PRISM® 7000 Sequence Detector System (Applied Biosystems) using the default thermal profile; 50°C for 2 minutes and 95°C for 10 minutes, followed by 40 cycles of amplification with denaturation at 95°C for 15 seconds and annealing/extension at 60°C for 1 min.
Real-Time RT-PCR data analysis
The qPCR data was analyzed with the following mixed ANCOVA model:
![](http://media.springernature.com/full/springer-static/image/art%3A10.1186%2F1744-9081-3-18/MediaObjects/12993_2007_Article_86_Equ1_HTML.gif)
where y jkli refers to the mRNA expression in lth brain region of fox jk, P to the ith PCR-plate, O to the jth group of origin (wild, S, NS, Cross), F the kth fox within group j, R to the lth region of the brain, OR to the interaction between the jth group and the lth region, G to the reference gene expression in region l of fox jk and β to the regression coefficient. Initially, both age and sex were included in the statistical analysis but the effect of these parameters was not shown to be significant. When included as potential confounders, neither age nor sex affected the expression difference between S and NS foxes, and consequently these variables were excluded from the final model. In the analysis, the differences between origin (S, NS, Cross, wild) is assessed using the between-animal variability, whereas the effects of region and the interaction between region and origin is assessed using the variability of samples within animal. We used three contrasts to: 1) compare the average mRNA levels of S and NS foxes (corresponding to an additive genetic effect), to 2) test the dominance deviation by comparing the mRNA levels of the foxes resulting from crosses to the weighted average of the parental lines (5/8 NS + 3/8 S, which represents the expected value for the crossed foxes under an additive model), and 3) to test the difference between wild foxes and the average of all farmed foxes. To account for variation in mRNA quantity, the geometric mean of the expression of the two reference genes ACTB and GAPDH [38] were included as the covariate (G) in the model [39]. The analysis was carried out in Proc Mixed, SAS v8.2. A modified version of the above linear model was used for power analysis of a genetic model including a composite additive (a) and a composite dominance (d) effect. That is, mRNA levels of the target genes were simulated, replacing "O" value in the formula above with "a" for S foxes, "-a" for NS foxes, "d" for F1 animals, and "0.5 d" for backcrosses. The interaction terms OR were dropped from the model as there were little evidence for any interactions between origin and region (Table 2). For the simulations, observed reference genes expression was used together with estimated effects for: "a", "P", "R", "Owild", and "β," and the dominance effect "d" was set to equal "a". The random effects "Fjk" and "εijkl", were drawn from two normal distributions (0, σF) and (0, σε) respectively, again using estimates of the between and within animal variance components from our data. For each gene 10 000 simulations were carried out and the power to detect an additive effect was defined as the fraction of simulations in which the contrast S versus NS was significant at α = 0.05. The power to detect dominance was defined as the fraction of simulations in which the contrast Cross vs. (5/8 NS + 3/8 S) was significant at α = 0.05.
To examine if an increased blood perfusion could be the main driver of the observed mRNA differences between S and NS foxes, we assumed that the average HBA and HBB levels (HB) may reflect the maximum effect of blood perfusion. Since the average HB level was correlated to the reference gene expression (r2 = 0.52, slope = 1.0 p-value < 0.0001), we normalized the HB levels prior to the analysis. This was done by regressing the average HB expression onto the total reference gene expression and storing the residuals as the normalized HB expression. This orthogonalized HB expression is positive when the amount of HB is larger than expected from the reference gene expression, and is negative when the average HB amount is lower than expected from the reference gene expression. The HB mRNA level was then included as an additional covariate in the original model, and thus the expression difference of HBE and HEBP1 was evaluated taking the maximum perfusion effect into account.
References
Price EO: Behavioral development in animals undergoing domestication. Appl Anim Behav Sci. 1999, 65: 245-271. 10.1016/S0168-1591(99)00087-8.
Vila C, Savolainen P, Maldonado JE, Amorim IR, Rice JE, Honeycutt RL, Crandall KA, Lundeberg J, Wayne RK: Multiple and ancient origins of the domestic dog. Science. 1997, 276: 1687-1689. 10.1126/science.276.5319.1687.
Savolainen P, Zhang YP, Luo J, Lundeberg J, Leitner T: Genetic evidence for an East Asian origin of domestic dogs. Science. 2002, 298: 1610-1613. 10.1126/science.1073906.
Saetre P, Lindberg J, Leonard JA, Olsson K, Pettersson U, Ellegren H, Bergstrom TF, Vila C, Jazin E: From wild wolf to domestic dog: gene expression changes in the brain. Brain Res Mol Brain Res. 2004, 126: 198-206. 10.1016/j.molbrainres.2004.05.003.
Trut L: Early canid domestication: The farm fox experiment. Am Sci. 1999, 87: 160-169. 10.1511/1999.2.160.
Hare B, Plyusnina I, Ignacio N, Schepina O, Stepika A, Wrangham R, Trut L: Social cognitive evolution in captive foxes is a correlated by-product of experimental domestication. Curr Biol. 2005, 15: 226-230. 10.1016/j.cub.2005.01.040.
Popova NK, Voitenko NN, Kulikov AV, Avgustinovich DF: Evidence for the involvement of central serotonin in mechanism of domestication of silver foxes. Pharmacol Biochem Behav. 1991, 40: 751-756. 10.1016/0091-3057(91)90080-L.
Trut LN, Naumenko EV, Belyaev DK: Change in the pituitary-adrenal function of silver foxes during selection according to behavior. Sov Genet. 1974, 8: 585-591.
Gulevich RG, Oskina IN, Shikhevich SG, Fedorova EV, Trut LN: Effect of selection for behavior on pituitary-adrenal axis and proopiomelanocortin gene expression in silver foxes (Vulpes vulpes). Physiol Behav. 2004, 82: 513-518. 10.1016/j.physbeh.2004.04.062.
Harri M: Behavioural and physiolocigal differences between silver foxes selected and not selected for domestic behaviour. Anim Welf. 2003, 12: 305-314.
Lindberg J, Bjornerfeldt S, Saetre P, Svartberg K, Seehuus B, Bakken M, Vila C, Jazin E: Selection for tameness has changed brain gene expression in silver foxes. Curr Biol. 2005, 15: R915-6. 10.1016/j.cub.2005.11.009.
Gilad Y, Rifkin SA, Bertone P, Gerstein M, White KP: Multi-species microarrays reveal the effect of sequence divergence on gene expression profiles. Genome Res. 2005, 15: 674-680. 10.1101/gr.3335705.
Sargan DR, Yang F, Squire M, Milne BS, O'Brien PC, Ferguson-Smith MA: Use of flow-sorted canine chromosomes in the assignment of canine linkage, radiation hybrid, and syntenic groups to chromosomes: refinement and verification of the comparative chromosome map for dog and human. Genomics. 2000, 69: 182-195. 10.1006/geno.2000.6334.
Yang F, O'Brien PC, Milne BS, Graphodatsky AS, Solanky N, Trifonov V, Rens W, Sargan D, Ferguson-Smith MA: A complete comparative chromosome map for the dog, red fox, and human and its integration with canine genetic maps. Genomics. 1999, 62: 189-202. 10.1006/geno.1999.5989.
Kukekova AV, Trut LN, Oskina IN, Kharlamova AV, Shikhevich SG, Kirkness EF, Aguirre GD, Acland GM: A marker set for construction of a genetic map of the silver fox (Vulpes vulpes). J Hered. 2004, 95: 185-194. 10.1093/jhered/esh033.
Avise J: Phylogeography: The History and Formation of Species. 2000, Cambridge, Massachusetts, Harvard University Press
Balcom AB: Fox Farming in Prince Edwards Island: A Chapter in the History of Speculation. Q J Econ. 1916, 30: 665-681. 10.2307/1884238.
Feldman BF, Zinkl JG, Jain NC: Schalm's Veterinary Hematology 5th edition. 2000, Philadelphia, Lippincott Williams & Wilkins
Jacob Blackmon B, Dailey TA, Lianchun X, Dailey HA: Characterization of a human and mouse tetrapyrrole-binding protein. Arch Biochem Biophys. 2002, 407: 196-201. 10.1016/S0003-9861(02)00471-X.
Taketani S, Adachi Y, Kohno H, Ikehara S, Tokunaga R, Ishii T: Molecular Characterization of a Newly Identified Heme-binding Protein Induced during Differentiation of urine Erythroleukemia Cells. J Biol Chem. 1998, 273: 31388-31394. 10.1074/jbc.273.47.31388.
Bonafoux B, Lejeune M, Piquemal D, Quere R, Baudet A, Assaf L, Marti J, Aguilar-Martinez P, Commes T: Analysis of remnant reticulocyte mRNA reveals new genes and antisense transcripts expressed in the human erythroid lineage. Haematologica. 2004, 89: 1434-1438.
Feezor RJ, Baker HV, Mindrinos M, Hayden D, Tannahill CL, Brownstein BH, Fay A, MacMillan S, Laramie J, Xiao W, Moldawer LL, Cobb JP, Laudanski K, Miller-Graziano CL, Maier RV, Schoenfeld D, Davis RW, Tompkins RG: Whole blood and leukocyte RNA isolation for gene expression analyses. Physiol Genomics. 2004, 19: 247-254. 10.1152/physiolgenomics.00020.2004.
He C, Braunitzer G: Carnivora: the primary structure of the hemoglobin from the silver fox (Vulpes vulpes var., Canidae). Biol Chem Hoppe Seyler. 1991, 372: 43-48.
Ohyagi Y, Yamada T, Goto I: Hemoglobin as a novel protein developmentally regulated in neurons. Brain Res. 1994, 635: 323-327. 10.1016/0006-8993(94)91455-9.
Mancuso C, Ragazzoni E, Tringali G, Liberale I, Preziosi P, Grossman A, Navarra P: Inhibition of heme oxygenase in the central nervous system potentiates endotoxin-induced vasopressin release in the rat. J Neuroimmunol. 1999, 99: 189-194. 10.1016/S0165-5728(99)00112-5.
Steiner AA, Colombari E, Branco LG: Carbon monoxide as a novel mediator of the febrile response in the central nervous system. Am J Physiol. 1999, 277: R499-507.
McCann SM, Karanth S, Kimura M, Yu WH, Rettori V: The role of nitric oxide (NO) in control of hypothalamic-pituitary function. Rev Bras Biol. 1996, 56 Su 1 Pt 1: 105-112.
De Oliveira RM, Del Bel EA, Guimaraes FS: Effects of excitatory amino acids and nitric oxide on flight behavior elicited from the dorsolateral periaqueductal gray. Neurosci Biobehav Rev. 2001, 25: 679-685. 10.1016/S0149-7634(01)00050-1.
Chiavegatto S, Nelson RJ: Interaction of nitric oxide and serotonin in aggressive behavior. Horm Behav. 2003, 44: 233-241. 10.1016/j.yhbeh.2003.02.002.
Nyberg F, Sanderson K, Glamsta EL: The hemorphins: a new class of opioid peptides derived from the blood protein hemoglobin. Biopolymers. 1997, 43: 147-156. 10.1002/(SICI)1097-0282(1997)43:2<147::AID-BIP8>3.0.CO;2-V.
Davis TP, Gillespie TJ, Porreca F: Peptide fragments derived from the beta-chain of hemoglobin (hemorphins) are centrally active in vivo. Peptides. 1989, 10: 747-751. 10.1016/0196-9781(89)90107-1.
Guillemin R, Vargo T, Rossier J, Minick S, Ling N, Rivier C, Vale W, Bloom F: beta-Endorphin and adrenocorticotropin are selected concomitantly by the pituitary gland. Science. 1977, 197: 1367-1369. 10.1126/science.197601.
van Daal JH, Jenks BG, Crusio WE, Lemmens WA, van Abeelen JH: A genetic-correlational study of hippocampal neurochemical variation and variation in exploratory activities of mice. Behav Brain Res. 1991, 43: 65-72. 10.1016/S0166-4328(05)80053-1.
Tordjman S, Carlier M, Cohen D, Cesselin F, Bourgoin S, Colas-Linhart N, Petiet A, Perez-Diaz F, Hamon M, Roubertoux PL: Aggression and the three opioid families (endorphins, enkephalins, and dynorphins) in mice. Behav Genet. 2003, 33: 529-536. 10.1023/A:1025774716976.
Sambrook JF: Molecular cloning: A Laboratory Manual. 1989, New York, Cold Spring Harbor Laboratory Press, 2nd ed
Vila C, Amorim IR, Leonard JA, Posada D, Castroviejo J, Petrucci-Fonseca F, Crandall KA, Ellegren H, Wayne RK: Mitochondrial DNA phylogeography and population history of the grey wolf canis lupus. Mol Ecol. 1999, 8: 2089-2103. 10.1046/j.1365-294x.1999.00825.x.
Swofford DL: PAUP* Phylogenetic Analysis Using Parsimony and Other Methods. 1998, Sunderland, MA, Sinauer Associates
Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, Speleman F: Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 2002, 3: RESEARCH0034-10.1186/gb-2002-3-7-research0034.
Bond BC, Virley DJ, Cairns NJ, Hunter AJ, Moore GB, Moss SJ, Mudge AW, Walsh FS, Jazin E, Preece P: The quantification of gene expression in an animal model of brain ischaemia using TaqMan real-time RT-PCR. Brain Res Mol Brain Res. 2002, 106: 101-116. 10.1016/S0169-328X(02)00417-5.
Acknowledgements
We are grateful for the help of Erik Ågren at the Swedish National Veterinary Institute (Uppsala, Sweden) for providing the samples of wild red foxes. Ulf Olsson gave valuable advice for the statistical analysis and Birgitte Seehuus assisted in collecting samples from the silver foxes. This work was supported by grants from the Swedish Research Foundation (621-2002-4258) from The Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (217-2004-2012), and a strategic grant from the Swedish University of Agricultural Sciences.
Author information
Authors and Affiliations
Corresponding author
Additional information
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JL participated in sample collection and preparation, carried out the expression analysis and participated in drafting of the manuscript. SB participated in sample collection and preparation, and carried out the mtDNA analysis. MB coordinated the sample collection. CV participated in the design of the study and the phylogeny reconstruction. EJ conceived of the study and participated in its design, coordination and manuscript drafting. PS participated in the design of the study, performed the statistical analysis and coordinated the preparation of the manuscript. All authors read and approved the final manuscript.
Electronic supplementary material
12993_2007_86_MOESM1_ESM.pdf
Additional File 1: Primer sequences and additional data on individual foxes. The file can be viewed with Adobe Reader [40] (PDF 54 KB)
Authors’ original submitted files for images
Below are the links to the authors’ original submitted files for images.
Rights and permissions
Open Access This article is published under license to BioMed Central Ltd. This is an Open Access article is distributed under the terms of the Creative Commons Attribution License ( https://creativecommons.org/licenses/by/2.0 ), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
About this article
Cite this article
Lindberg, J., Björnerfeldt, S., Bakken, M. et al. Selection for tameness modulates the expression of heme related genes in silver foxes. Behav Brain Funct 3, 18 (2007). https://doi.org/10.1186/1744-9081-3-18
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/1744-9081-3-18