Final thesis
Gene expression in brains from red jungle fowl
(Gallus gallus) that differ in fear response
Markus Jöngren
LITH-IFM-Ex--07/1903--SEContents:
1. Abstract 1
2. Introduction 1
3. Materials and methods 5
3.1 Behavioural studies 5
3.1.1 Animals 5
3.1.2 General procedure 6
3.1.3 Ground predator test 6
3.1.4 Tonic immobility test 10
3.1.5 Aerial predator test 11
3.1.6 Statistical analysis 11 3.2 Genetic analysis 12 3.2.1 RNA isolation 12 3.2.2 Microarray analysis 13 3.2.3 Statistical analysis 13 4. Results 15
4.1 Ground predator test 15
4.2 Tonic immobility 15
4.3 Aerial Predator 15
4.4 Multivariate factor analysis 15
4.5 Genetic analysis 23
5. Discussion 28
6. Acknowledgements 31
1. Abstract
The fear response of two different captive populations of red jungle fowl (rjf, Gallus gallus) was measured in three different tests, a ground predator test, an aerial predator test and a tonic immobility test. The two populations originated from Copenhagen zoo (Cop) and Götala research station (Got) but had been kept together for four generations. Earlier generations had a confirmed difference in fearfulness where the Cop birds exhibited a higher degree of fear response than Got birds (Håkansson and Jensen, 2005; Håkansson et al., 2007). The most and least fearful birds of each sex and population were identified and used in a gene expression study. The midbrain regions from the candidate birds were collected and RNA was isolated from each brain. The RNA was then reversed transcribed to cDNA which was used in a cDNA microarray experiment. 13 significantly
differentially expressed genes were found between the fearful and non-fearful females. Among others were the neuroprotein Axin1, two potential DNA/RNA regulating proteins and an unknown transcript in the
Quantitative Trait Locus 1(QTL 1), a well studied QTL on chromosome one with substantial effect on both behaviour and morphology during
domestication (Schütz et al., 2002). This thesis succeeds in finding a
difference in gene expression between fearful and non fearful female rjf but not between males. It fails in identifying gene expression differences
between the two populations. Finally, the found differentiated genes suggest a potential molecular mechanism controlling the fear response in fowl. Keywords:Domestication, Fearfulnes, Microarray
2. Introduction
Fear, fear response, fearfulness and anxiety. Many words to try to describe the same set of behaviours shown by animals in the face of a threat, real or imagined. It is also one of the most important behaviours that changes during the domestication of a species, maybe the most important. An
ongoing Russian experiment that started in the 1950: s , reviewed by (Trut, 1999), showed that using the fear response of captive silver foxes as the sole selection trait, part of the offspring began to show similar behavioural and morphological phenotypes to the domesticated dog after only six generations of selection.
Further studies show that fear response differs greatly between
al., 2004). This is thought to be due to the difference in selection pressure on domesticated and wild animals (Price, 1999). Domesticated animals live in an environment protected from outer threats such as predation and starvation and are instead directly selected for production and sociality traits and
indirectly for non fearful behaviour. A wild animal held in captivity is constantly stressed by its confinement and the proximity of potential predators, humans. According to Price, the natural behaviours does not change in captivity but the thresholds for eliciting different behaviours differ (Price, 1999). In the case of the fear response, the threshold should increase. The effect will be that the animals with highest reproductive success during domestication are the animals with the most favourably changed fear
response.
To be able to measure fear, fear has to be defined. One possibility is to measure the physiological response, such as changes in heart rate and hormone levels (Macrí and Würbel, 2007), to a frightening stimulus. This gives detailed information on when the animal becomes frightened/stressed but is quite invasive which in itself could be a frightening experience and therefore affect the results. A less invasive method is to observe the animal behaviour during an exposure to a frightening stimulus(Evans et al., 1993; Keer-Keer et al., 1996; Bayly and Evans, 2003; Schütz et al., 2004;
Forkman et al., 2007; Håkansson et al., 2007; Ponder et al., 2007). The fear response differs from species to species, some run, others hide or play dead, while others threaten and more. The amplitude of the fear response also differs depending on the stimuli (Takahashi et al., 2005). This means that to measure fear one must first define what the fear response looks like in the organism of choice and then figure out what stimuli that elicits it, and to what extent the response is applied towards the stimuli.
Common animals used in fear studies are mice (Mus musculus)
(Mongeau et al., 2003; Ramos et al., 2003; Takahashi et al., 2005; Ponder et al., 2007), rats (Rattus norvegicus) (Jaako-Movits et al., 2005; Takahashi et al., 2005; Fendt, 2006; Macrí and Würbel, 2007) and, to some degree, fowl (Gallus gallus)(Schütz et al., 2001; Bayly and Evans, 2003; Schütz et al., 2004; Håkansson and Jensen, 2005; Forkman et al., 2007; Håkansson et al., 2007). In some cases (Takahashi et al., 2005; Håkansson et al., 2007; Ponder et al., 2007), fear is the main focus of the study, in other it is used as a tool to measure or define another behavioural or physiological trait
(Jaako-Movits et al., 2005; Macrí and Würbel, 2007). Indifferent of the purpose the means are identical, the animal is exposed to a frightening stimulus and the
reaction is recorded. The number and intensity of the fear related behaviours are then recorded to quantify, and measure, the fear response. Commonly used methods are open field tests (Ramos et al., 2003; Schütz et al., 2004; Forkman et al., 2007), novel object tests (Schütz et al., 2004; Forkman et al., 2007) ,predator exposure test (Bayly and Evans, 2003; Håkansson and
Jensen, 2005; Forkman et al., 2007; Håkansson et al., 2007) and restraints tests (Schütz et al., 2001; Forkman et al., 2007).
The genetical control of behaviour is not fully understood but within a species, it is thought to depend more on gene expression differences than genomic differences (Enard et al., 2002). Therefore, to identify the genetical component of the fear response one has to compare the gene expression differences between animals with high and low fear response. The most common method for global gene expression analysis is the use of cDNA microarrays with its possibility to measure the expression of several thousands of genes simultaneously (Allison et al., 2006).
On a microarray, thousands of cDNA probes are fixed to a surface. RNA transcripts, or their cDNA, from a sample are labelled with fluorescent dye and then hybridised to the probes on the array. Only transcripts that match the probes adhere to the array and the relative number of copies of each transcript can be estimated from the intensity of emission from each probe.
In a cDNA microarray experiment, both the probes and the targets consist of cDNA. mRNA from a sample is reverse transcribed to cDNA which in turn is labelled with a flourophore, often cy3 or cy5. Onto each array two differently labelled (cy3 and cy5) cDNA samples are hybridized. Depending on experimental design it could be two different biological
samples or a biological sample and a reference. Two images are taken of the array, one in the respective spectrum of the flourophores. These two pictures are in turn merged to create one composite image. This image constitutes the raw data for the gene expression analysis.
For an accurate expression analysis four components of the
experimental design must be considered when planning the experiment, for more input on the different components see (Allison et al., 2006) and (Yang and Speed, 2002).
• Design. The experimental layout must be optimised to yield as much information as possible from a minimum number of arrays.
• Pre-processing. Image processing and normalisation of the data set. The goal is to have the data as uniform within and between the arrays as possible.
• Inference and/or classification. What methods to use for the actual statistical analysis of the gene expression.
• Validation of findings. Validating your results, most often by confirming a small portion of the expression analysis with another method, e.g. Real-time PCR.
In a doctoral thesis (Lindberg, 2007) microarrays were used to investigate what changes in gene expression that have occurred between wolves and dogs during domestication. Ponder et al (Ponder et al., 2007) used microarrays in a study of the genetic background to anxiety related behaviour in mice and (Edenberg et al., 2005) used microarrays in a search of genes related to alcohol addiction, also in mice.
As shown by Schütz et al(Schütz and Jensen, 2001; Schütz et al., 2004) there are differences in fear response between red jungle fowl (rjf), the
ancestor of all domesticated fowls, and a domesticated production line of fowl (WL 13). Further (Håkansson and Jensen, 2005; Håkansson et al., 2007) showed that there are genetically inherited differences in fear response
between groups of red jungle fowl held in different captive environments. The most profound difference was found between birds from
Copenhagen Zoo (Cop) and Götala research station (Got). At Copenhagen zoo the birds lived in a semi natural environment, free roaming within the zoo and with a limited predation pressure. In Götala the birds were kept in indoor cages and let out in small enclosures on occasions. When the fear response was compared between the two populations Cop birds showed a higher fear response than Got birds in several tests. The results were the same when offspring from the two populations raised together were tested (Håkansson et al., 2007).
In this thesis the fourth generation of offspring from Copenhagen and Götala birds were tested in three behaviour tests measuring the fear response, a ground predator exposure, an aerial predator exposure and the tonic
immobility test. The first two tests are thought to measure the bird’s active fear of predation and the third their passive fear of handling.
From these results the fear response of each individual bird was
the gene expression in the midbrain region were conducted on the four most and least fearful birds of each sex and population. The results from the microarrays were analysed and compared in the statistical platform R.
It was hypothesised that genes related to fearful behaviour are
expressed in varying degrees in the brains of birds displaying different levels of fear response, resulting in two distinct phenotypes defined as fearful and non-fearful birds. From previous findings (Håkansson et al., 2007) it is suggested that Got birds are less fearful than Cop birds.
If the hypothesis holds then differences in fear related behaviour are due to differently expressed genes and such genes should be found. Further, the hypothesis states that there should be similarities between all fearful birds regardless of origin. The same goes for non-fearful birds.
3. Materials and methods 3.1 Behavioural studies
Three different behaviour studies were conducted. Two predator response tests and a tonic immobility test.
3.1.1 Animals
The animals originated from Copenhagen zoo and Götala research station in Skara. In Copenhagen zoo the birds were held in a semi-wild state with access to the entire zoo (seven HA) while the birds in Götala were kept under strictly controlled forms (Håkansson et al., 2007). The last four generations the two populations have been held together, the first three generations at Götala and the fourth at the Linköpings university research facility at Vreta. They were all held under the same conditions as the original Götala population.
The population originating from Copenhagen Zoo counted 30 animals, 15 males and 15 females. The Götala population consisted of 33 animals, 16 males and 17 females, making a total of 63 test birds. Three birds (One Got male, One Got female and one Cop male) died of natural causes during the time between the first and the last test lessening the number of test
individuals to 60. They were held in two home cages; with an equal number of males and females of both populations in each cage. The cages measured 2.53m (width) * 3.09m (length) * 3.01m (height) and they contained perches, two platforms at different levels, food trays, water nipples and nesting boxes.
The temperature in the home cages was ≥21°C. Food and water were provided ad libitum.
The two cages were positioned in front of each other and there were two additional cages in the room which both held other fowls not used in this study.
3.1.2 General procedure
The tests were conducted during March through May 2007, and took place at the Linköping University research facility at Vreta in different rooms
connected to the room housing the home cages of the birds.
The birds were caught in their home cages and carried to the prepared test arena. The test started immediately to minimize the handling of the birds. When the test was done the bird was released back into its home cage and the arena thoroughly cleaned.
All the birds had a personal identification number attached to their wings and were in addition marked with leg rings prior to testing, the later to make it easier to identify the birds on sight in their home cages.
At the time of the testing the test conductor had no knowledge of what population (Cop or Got) the birds originated from.
3.1.3 Ground predator test
The test was designed to measure the active fear of predation. The test arena (Fig. 1) had the external measures 0.95* 3.8*1.80m. The lower part (0.5 m) of the arena was made of wood and the rest (1.3 m) was made of chicken wire. In one end of the test arena was a platform (0.60* 0.95*0.42 m) on which the bird to be tested was placed. The bird was prevented from leaving the platform by a removable net.
Two sides of the test platform were covered with black plastic bags to prevent the birds from exploreing the environment outside the test platform in any direction other then the test arena, the third side was placed against an empty bird cage that prevented the bird from seeing the rest of the room.
There were two perches above the platform (39 and 65 cm) and food and water were freely available. On the other side of the arena was a wooden box with an opening to the arena covered with black plastic. Inside the box was a stuffed polecat, standing on a wagon. The wagon stood on a track which ran through the arena and ended under the platform and there was a rope attached to the wagon which followed the track through the arena and
out of the arena under the platform. A DVD-recorder was placed 2.2 m above the platform with a full view of the platform.
The test birds were placed inside the arena in darkness and the test started when the light was turned on. The test setup was controlled from outside the test room. The whole test was recorded.
The test started when the light was turned on, the bird was left alone for five minutes and then the faked predator attack started. The test conductor began pulling the rope attached to the predator model which then left its lair and advanced towards the fowl to finally disappear under the platform. The exposure lasted approximately 30 seconds. After the exposure the bird was left alone for five minutes and then the test was terminated.
Behaviours were recorded throughout the whole test and the DVD recordings were decoded using the Observer XT ™ software and a continuous sampling method. The behaviours were divided into two
categories, agitated and non agitated (Tab 1,) and then analyzed according to Tab. 2a-c . A third category (misc.) arose when it was observed that a bird could show both agitated and non agitated while perching. In addition each bird was scored a subjective reaction score ranging from one to five.
Tab 1: Ethogram, the behaviours that was recorded in the ground predator test
Behaviour State/event* Description
Exploring State The bird stands in or walks around the arena turning its head at a moderate speed in all directions (headmovements/s)
Alert State The bird stands perfectly still with its head/attention pointed towards the source of alarm. If the source moves, the bird keeps its attention on it, slowly turning its head. If no source can be identified, the bird stand still for 15-30 s and then begin to slowly turn its head, look in for the source of alarm.
Agitated State The bird stands in or walks around the arena with its attention focused on the source of agitation. Its movements are fast and distinct. The head movements are especially characteristic, it quickly and distinctly moves its head from side to side to try to get a good view of the threat.
Escape attempt State The bird panics and tries to escape any way possible, vocalising loudly.
Perching high/low
State The bird sits or walks on one of the perches above the arena.
Vocalisation Event Alarm call
Feeding/Drinking Event The bird puts its head in the feeding tray or the water bowl.
* All states are mutually exclusive except perching which could be coupled with another state.
Tab. 2b: Behaviour analysis during exposure Analysed behaviours.
Agitated:
Time to first alert Total time spent alert Time to first agitated Total time spent agitated Time to first vocalisation. Number of vocalisations. Time to first escape attempt Number of escape attempts Total duration of escape attempts Non agitated:
Total time spent exploring
Total time spent feeding or drinking Misc:
Total time spent perching Tab. 2 a: Behaviour analysis pre exposure Analysed behaviours.
Agitated:
Total time spent alert Total time spent agitated Number of escape attempts Total duration of escape attempts Number of vocalisations
Non agitated:
Total time spent exploring
Total time spent feeding or drinking Misc:
Total time spent perching
Tab 2a-c The ground predator behaviour analysis. The behaviours were analysed differently depending which part of the test that was analysed.
One indicated that the bird did not react. Two that the bird became agitated and vocalised but did not escape. Three indicated single escape attempts lasting no longer than ~sec. Four several escape attempts lasting together ~5-20 sec. Five indicated a constant escape attempt lasting from ~20 sec. to the whole exposure. For each population and gender the mean and standard error of each analyzed behaviour and the reaction score was calculated. These were then compared between the populations and genders.
3.1.4 Tonic immobility test
The TI test is thought to measure the passive fear of handling (Håkansson and Jensen, 2005; Forkman et al., 2007). The bird is placed on its back in a wooden cradle with its head outside the cradle. A light pressure is applied to its chest with the test conductor’s hand. After 10 seconds the hand is
removed. If the bird does not move in any way, TI has successfully been induced. After TI was induced the time to the first head movement and to the rightening of the bird was measured. The test was terminated if the bird lay motionless for 10 minutes and the bird was marked for 600 seconds.
If TI was not induced in three attempts the test was terminated and the bird was marked a maximum value, 5, to separate it from those birds which were actually induced on the third attempt.
The number of induction attempts, time to first head movement and time to rightening was recorded. As in the ground predator test, the means
Tab. 2c: Behaviour analysis post exposure Analysed behaviours.
Agitated:
Total time spent alert Total time spent agitated Number of escape attempts Total duration of escape attempts Number of vocalisations
Non agitated:
Total time spent exploring
Total time spent feeding or drinking Misc:
Total time spent perching Time until agitated stops.
and standard errors were calculated and compared with respect to gender and population.
3.1.5 Aerial predator test
The data from the AP test was taken from Håkansson et al (Håkansson and Jensen, 2007) generation four. The test procedure was similar to the ground predator test. The test bird was placed within, and confined to a test arena in which food and water were provided. Agitated and non agitated behaviours were recorded for five minutes using a one-zero sampling method with ten second intervals. After five minutes a hawk model flew over the arena once which took approximately 1.5 seconds. The test bird’s reaction to the
exposure was scored on a scale from zero to three where zero indicated no reaction to the exposure and three a strong reaction to the exposure. After the exposure behaviours were recorded for five minutes.
The data, which were divided into agitated and non agitated behaviours, were then transformed into an agitation index (AI) described by Håkansson above. For each occurrence of an agitated behaviour the animal scored one point on the AI and for each non agitated behaviour the animal lost one point. A fearful animal would score high on the AI while a non fearful animal
would score low.
A comparison of the AI each animal scored before and after the exposure was made. This number showed how the bird changed its behaviour after the exposure. A positive number indicated that the bird became more agitated after the exposure while a negative number indicated that it became calmer. If the comparison scored a zero, then the bird did not change its behaviour with respect to fearfulness.
The means and standard errors of the AI’s before and after the predator exposure, the AI comparison and the reaction score was calculated with respect to population and gender.
3.1.6 Statistical analysis
To quantify the fear response of each bird a multivariate factor analysis was performed (Cattel, 1978; Svartberg and Forkman, 2002) based on the
variance of selected behaviours. From the ground predator test the reaction score of the exposure, the comparison of time spent on alert, exploring, agitated or escaping and the comparison of the number of alarm calls before and after the exposure. From the tonic immobility test the number of
movement and time to rightening. From the aerial predator test the reaction score during the exposure and the comparison of AI before and after
exposure were used.
The behaviours were chosen to be as representative as possible of the bird’s behaviour during the tests and to contribute to as much as possible to the factor analysis.
A multivariate factor analysis is a statistical tool used to identify
underlying correlations within a data set of sufficient size, containing a large amount of variation. The result of the factor analysis is a number of factors that together explain the variance within the data. The first factor explains the most of the variance, the second explains the second most and so on. The underlying factor sought in this study was the first one and the aim was to find a factor that explained as much as possible of the variance within the data.
Each bird was scored based on the first factor and 16 candidates, the two most and least fearful of each population and gender, were chosen for genetic analysis. This experimental design made it possible to compare genetic differences between population, gender and fear phenotype
separately with 8 birds in each group in every comparison. Unfortunately one of the fearful Cop males died before he could be collected and there was no other fearful Cop male to replace him with. Therefore he was replaced with one intermediately fearful Cop male, for the population analysis, and a fearful Got male, for the phenotypic analysis, giving a total of 17 birds.
3.2 Genetic analysis 3.2.1 RNA isolation
The birds were decapitated and the midbrain region collected. The midbrain regions consists of the thalamus, hypothalamus, pituitary, mesencephalon, pons, nucleus tracus solitarii (NTS) and medulla oblongata. To protect the RNA from degradation the brain samples were then placed in liquid N2 within 10 min of the time of death and then stored in -80°C until the time of homogenization.
The frozen samples were homogenized in TRI solution from Ambions TRI Reagent kit . The TRI solution protected the RNA from degradation but as a security measure the homogenate was stored at -80°C.
The RNA was isolated from the midbrain samples, using the TRI Reagent kit. The lowest accepted concentration of isolated RNA was 1300
ng/µl. The purity of the RNA samples was controlled using an Agilent 2100 Bioanalyzer and the Agilent 6000 Nano kit. The minimum RNA integrity number accepted was 8.0.
3.2.2 Microarray analysis
The microarray was the Chicken 14K cDNA microarray described by Fitzsimmons et al (Fitzsimmons et al., 2006), an array based on a testis and brain library from rjf and white leghorn containing 13907 cDNA clones of which 135 were an internal control (GAPDH, see Fitzsimmons above).
The experiment had a referential design (Yang and Speed, 2002;
Allison et al., 2006; Fitzsimmons et al., 2006) where each array was treated with a cy5 labelled sample from one bird and a cy3 labelled reference
containing pooled mRNA from the midbrain region of ~30 birds. The reference was collected from the test birds and from birds used in another study.
The RNA labelling, cDNA synthesis, microarray hybridization and the image processing were done at the Royal Institute of technology (KTH) using the same methods as Fitzsimmons et al (Fitzsimmons et al., 2006).
The image processing of the microarrays yielded 17 image files in the TIFF format with corresponding GPR-files, a format which describes the TIFF-files numerical, making it possible to use them in a statistical analysis.
3.2.3 Statistical analysis
The GPR files were loaded into the R statistical platform, a free software for statistical analysis, and analysed using the same packages and methods (filtration, normalization and gene expression analysis) as Fitzsimmons et al (Fitzsimmons et al., 2006). Tab 3 shows the versions of R and the different packages used with the exception of two filters, filterB2SD and filterMtoM, which could not be used due to incompatibility between the versions of R and the different packages. After filtration approximately 60% of the spots on each array remained for expression analysis.
Gene expression was compared between each sex (M vs Fm), population (Got vs Cop) and fear response (NF vs F). Further, each gender was split with respect to fear response and each population split with respect to gender. The gene expression patterns from the different subgroups were compared according to fig 2.
The expression analyses were carried out using an empirical Bayesian method described by (Fitzsimmons et al., 2006) and (Lindqvist et al., 2007) Bayesian statistics are used as significance measure in large datasets where the false discovery rate (FDR) would provide a large number of false
positives because of the large amount of data. For further details see (Efron et al., 2001; Vaince et al., 2006).
The analysis yielded a significance value (B-value) and a measure of fold change (M.value, M = 2log of fold change). The significance
delimitations were set to B > 0 and M ≥ 1 , M ≤ -1 and are based on those Fig 2a)
Fearfull Females Nonfearfull Females
Fearfull Males Nonfearfull Males
Fig 2b)
Cop Females Got Females
Cop Males Got Males
Tab3. The R statistical platform and the different packages used. Software Version web-site
R 2.1.1 http://www.R-project.org Bioconductor 1.6 http://www.bioconductor.org KTH-package 0.4.5 http://www.biotech.kth..se./molbio/ microarray/pages/kthpackagetransfer.html limma 1.9.6 http://bioinfo.wehi.edu.au.limma statmod 1.2.0 http://www.statsci.org/r/ rimage 0.5.7 http://cran.r-projects.org/doc/packages/rimage.pdf R.classes 0.6.2 http://www.maths.lth.se/help/R/R.classes aroma 0.8.5 http://www.maths.lth.se/help/R/aroma
Fig 2 Subgroup comparisons within a) gender and phenotype and b) population and gender
used by Fitszimmons et al (Fitzsimmons et al., 2006) And Lindqvist et al (Lindqvist et al., 2007)
When the probes of the transcripts with significantly differentiated expression had been identified, the corresponding GeneBank id was
collected from the annotation file of the microarray. The specific transcripts were identified using the Entrez cross-database search tool(Entrez, 2007) and BLAST (BLAST, 2007) at the NCBI homepage.
4. Results
4.1 Ground predator test
As shown in fig. 3a-f ) all the birds displayed high levels of fear response. Although only two behaviours are shown, the pattern was found in all the data gathered from the Ground predator test.
Fig. 4a-f displays the variance within the datasets used in the multivariate factor analysis.
4.2 Tonic immobility
In the tonic immobility test there was a significant (p =0.002) difference between the two populations, Cop birds were more prone to lay motionless on their backs (fig. 5a) then Got birds. As for the number of induction attempts needed and time to righting, there were no differences (fig. 5b-c).
The variance within each group was reasonably high so the data could be used in the factor analysis (fig. 6a-c).
4.3 Aerial Predator
As with the Ground Predator test, no significant differences could be found between the two populations (fig. 7a-b), but there was sufficient variance for the factor analysis (fig. 8a-b).
4.4 Multivariate factor analysis
The analysis yielded an underlying factor that explained 21% of the variance among the males and almost 27% among the females (tab 4a-b). Both the males and the females had four factors with Eigen values > 1 and together they explained 66.8% of the male variance and 75.0% of the female variance. The behaviours loaded differentially with respect to gender (tab 5a-b).
Among the males the behaviours that loaded highest on the factor was the number of Tonic immobility induction attempts (0.72) , the difference in
time trying to escape before and after the ground predator exposure (0.63), time to righting in the tonic immobility test (-0.63) and the time to first movement in the tonic immobility test. The males also showed two very low loading behaviours; the difference in time spent alert (0.10) and exploring (-0.1) before and after exposure in the ground predator test.
Fig 3a Change in agitated behaviour G C 25.0 20.0 15.0 10.0 5.0 0.0 S e c. G C 25.0 20.0 15.0 10.0 5.0 0.0 S e c.
Fig 3b Change in alertness
G C 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 S ec . G C 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 S ec .
Fig 3c Change in time trying to escape G C 2 1.5 1 0.5 0 -0.5 S ec . G C 2 1.5 1 0.5 0 -0.5 S ec .
Fig 3d Change in exploratory behaviour G C -20.0 -30.0 -40.0 -50.0 -60.0 -70.0 -80.0 S e c. G C -20.0 -30.0 -40.0 -50.0 -60.0 -70.0 -80.0 S e c.
Fig 3.Comparison of the population means of the different behaviours form the ground predator test used in the factor analysis. Populations are represented on the X-axis (C = Cop, G = Got) The Y-axis depends on the studied
behaviour. The differences were calculated subtracting the frequency of the behaviour before the exposure from the frequency of the behaviour after the exposure. The error bars represent +-1SE.
a) Change in time spent displaying the agitated behaviour after exposure. b) Change in time spent on alert after the exposure.
c) Change in time spent escaping after the exposure.
d) Change in exploratory behaviour after the exposure
e) Change in vocalisation frequency after the exposure. f) The reaction score
Fig 3e Change in vocalisation
G C 50 25 0 -25 -50 N u m b e r of v oc . G C 50 25 0 -25 -50 N u m b e r of v oc .
Fig 3f Reaction score
G C 4.2 4.1 4 3.9 3.8 3.7 3.6 3.5 R e ac tio n sc o re G C 4.2 4.1 4 3.9 3.8 3.7 3.6 3.5 R e ac tio n sc o re
Fig 4a GP Comp Agitated 200.0 0.0 -200.0 25 20 15 10 5 0 N um be r of B ird s 25 20 15 10 5 0 C G P op ( G = G ot, C = C op ) Sec 200.0 0.0 -200.0 25 20 15 10 5 0 N um be r of B ird s 25 20 15 10 5 0 C G P op ( G = G ot, C = C op ) Sec
Fig 4b GP Comp alert
200.0 0.0 -200.0 8 6 4 2 0 N u m be r of B ir ds 8 6 4 2 0 C G P op ( G = G o t, C = C o p) Sec 200.0 0.0 -200.0 8 6 4 2 0 N u m be r of B ir ds 8 6 4 2 0 C G P op ( G = G o t, C = C o p) Sec
Fig 4d GP Comp Exploring
200.0 0.0 -200.0 6 4 2 0 N um b er of B ir ds 6 4 2 0 C G P op ( G = G o t, C = C o p) Sec. 200.0 0.0 -200.0 6 4 2 0 N um b er of B ir ds 6 4 2 0 C G P op ( G = G o t, C = C o p) Sec. Fig 4c Gp Comp Escape att.
15 10 5 0 -5 -10 25 20 15 10 5 0 N um be r of B ird s 25 20 15 10 5 0 C G P op ( G = G o t, C = C op ) Sec. 15 10 5 0 -5 -10 25 20 15 10 5 0 N um be r of B ird s 25 20 15 10 5 0 C G P op ( G = G o t, C = C op ) Sec.
Fig 4e GP Comp vok 500 0 -500 12.5 10.0 7.5 5.0 2.5 0.0 N u m be r of B ird s 12.5 10.0 7.5 5.0 2.5 0.0 C G P o p ( G = G o t, C = C op ) Number of voc. 500 0 -500 12.5 10.0 7.5 5.0 2.5 0.0 N u m be r of B ird s 12.5 10.0 7.5 5.0 2.5 0.0 C G P o p ( G = G o t, C = C op ) Number of voc.
Fig 4f GP reaction score
6 5 4 3 2 1 0 -1 Reaction Score 12.5 10.0 7.5 5.0 2.5 0.0 N um b er of B ird s 12.5 10.0 7.5 5.0 2.5 0.0 C G P o p ( G = G ot, C = C o p) 6 5 4 3 2 1 0 -1 Reaction Score 12.5 10.0 7.5 5.0 2.5 0.0 N um b er of B ird s 12.5 10.0 7.5 5.0 2.5 0.0 C G P o p ( G = G ot, C = C o p)
Fig 4 The distribution of birds performing each behaviour described in fig 3. The distributions were used as a measure of variance within each behaviour and population.
a) Distribution of change in time spent displaying the agitated behaviour after exposure.
b) Distribution of change in time spent on alert after the exposure. c) Distribution of change in time spent escaping after the exposure.
d) Distribution of change in exploratory behaviour after the exposure
e) Distribution of change in vocalisation frequency after the exposure. f) Distribution of reaction scores
Fig 5c Righting time G C 425 400 375 350 325 300 S ec . G C 425 400 375 350 325 300 S ec .
Fig 5a Time to first movement
G C 30 25 20 15 10 5 S ec . G C 30 25 20 15 10 5 S ec .
Fig 5b Induction attempts
G C 2.6 2.4 2.2 2 N um b er of in d. A tt em pt s G C 2.6 2.4 2.2 2 N um b er of in d. A tt em pt s
Fig 5. The population means of the behaviours recorded in the tonic immobility test. The X-axis displays populations.
a) Time to first movement,
b) Number of induction attempts. c) Righting time
Fig 6b Induction attempts 6 5 4 3 2 1 0 Number of attempts 12 10 8 6 4 2 0 N um b er of B ird s 12 10 8 6 4 2 0 C G P o p ( G = G ot, C = C op ) 6 5 4 3 2 1 0 Number of attempts 12 10 8 6 4 2 0 N um b er of B ird s 12 10 8 6 4 2 0 C G P o p ( G = G ot, C = C op )
Fig 6c Righting time
500 300 100 0 Sec 12.5 10.0 7.5 5.0 2.5 0.0 N um b er of B ird s 12.5 10.0 7.5 5.0 2.5 0.0 C G P op ( G = G ot, C = C op ) 500 300 100 0 Sec 12.5 10.0 7.5 5.0 2.5 0.0 N um b er of B ird s 12.5 10.0 7.5 5.0 2.5 0.0 C G P op ( G = G ot, C = C op )
Fig 6a Time to first movement
80 60 40 20 0 12 10 8 6 4 2 0 N um b er of B ir ds 12 10 8 6 4 2 0 C G P op ( G = G ot, C = C op ) Sec . 80 60 40 20 0 12 10 8 6 4 2 0 N um b er of B ir ds 12 10 8 6 4 2 0 C G P op ( G = G ot, C = C op ) Sec .
Fig 6 The distribution of birds performing the behaviours described in fig 5. The distributions were used as a measure of variance within each behaviour and population.
a) Time to first movement,
b) Number of induction attempts. If a bird scored five, tonic immobility was not induced.
c) Righting time. A bird that scored 600 seconds never righted itself.
Fig 7a Reaction score G C 2.2 2.1 2.0 1.9 1.8 1.7 1.6 R ea ct io n S co re G C 2.2 2.1 2.0 1.9 1.8 1.7 1.6 R ea ct io n S co re Fig 7b Change in AI G C 22.0 20.0 18.0 16.0 14.0 12.0 10.0 A I d iff er en ce G C 22.0 20.0 18.0 16.0 14.0 12.0 10.0 A I d iff er en ce
Fig 7 The population means of the behaviours recorded in the aerial predator test. The X-axis displays populations..
a) Reaction score
b) Change in AI after the exposure. The change was calculated the same way as the changes in fig 3
Fig 8a AP reaction score
4.000 2.000 0.000 Reaction Score 12 8 4 0 N um be r of b ird s 12 8 4 0 C G P op ( G = G ot, C = C op ) 4.000 2.000 0.000 Reaction Score 12 8 4 0 N um be r of b ird s 12 8 4 0 C G P op ( G = G ot, C = C op )
Fig 8b AP AI change after exposure
75 50 25 0 -25 AI difference 10 8 4 0 N u m be r of b ird s 10 8 4 0 C G P o p ( G = G ot, C = C o p) 75 50 25 0 -25 AI difference 10 8 4 0 N u m be r of b ird s 10 8 4 0 C G P o p ( G = G ot, C = C o p)
Fig 8 The distribution of birds performing the behaviours described in fig 7. The distributions were used as a measure of variance within each behaviour and population.
a) Reaction score
In the female analysis the highest loading behaviours were the difference in time spent alert before and after exposure in the ground
predator test (-0.75), the differences in number of vocalisations in the ground predator test (0.69) and the reaction score on the exposure in the ground predator test(-0.60). The lowest loading behaviours were the tonic immobility time to righting (0.25) and the aerial predator reaction score (0.34).
As shown by fig. 9 The mean fear factor score for Cop and Got birds did not differ significantly and both the genders were evenly distributed along the fear factor (fig. 10)
4.5 Genetic analysis
As expected, there were significant differences in gene expression pattern between the genders (fig 11a tab 6). Of the nine significantly differentially expressed transcripts six were different versions of Wpkci-8, a gene on the W chromosome with unknown function, two were transcripts similar to Adseverin (chromosome 2) and the immunoglobulin lambda chain
(chromosome 15) and the last one an unknown transcript from an unknown chromosome. All nine of them have been found in earlier expression
analysis of the gender differences using the Chicken14K cDNA microarray. The most differentially expressed (tab 6) transcripts were the six
Wpkci-8 copies followed by the unknown transcript, adseverin and the immunoglobulin lambda transcript.
When the two populations were compared, no differentally expressed genes could be found.(fig 11b).
As for the comparison of the fearful and non-fearful birds one transcript was found that differed significantly between fearful and non-fearful birds (fig 11c), namely CN216922 described in tab 7, but when they were compared with respect to gender the fearful females differed significantly from the non fearful females (fig 11d). The males on the other hand did not differ at all (fig 11e).
There were 13 significantly differently expressed transcripts between the fearful and non-fearful females (B > 0; M ≥ 1, M ≤ -1 ) and one
transcript that tended towards significance (B > 0; M = 0.95). See Tab 7 Among the differentiated transcripts were found genes similar to RNA and DNA regulating proteins (CN237554 and CN220264), genes coding for metabolic proteins (CN224985, CN231217 and CN220996), a gene coding
for a neurological active protein (BU409223) and transcripts with poorly understood or unknown function.
tab 4a Component Eigenvalue % of Variance Cumulative % 1 2,3368 21,243 21,243 2 2,2517 20,47 41,713 3 1,6436 14,942 56,655 4 1,1159 10,145 66,799 5 0,9677 8,7973 75,597 6 0,7975 7,2502 82,847 7 0,6985 6,3504 89,197 8 0,5542 5,038 94,235 9 0,393 3,5729 97,808 10 0,1384 1,2578 99,066 11 0,1027 0,9339 100 tab 4b)
Component Eigenvalue % of Variance Cumulative %
1 2,9485 26,805 26,805 2 2,3755 21,595 48,4 3 1,838 16,709 65,109 4 1,0856 9,8695 74,978 5 0,8224 7,476 82,454 6 0,5652 5,1379 87,592 7 0,5043 4,5843 92,177 8 0,3331 3,0282 95,205 9 0,2876 2,6146 97,819 10 0,2052 1,8658 99,685 11 0,0346 0,315 100
Tab. 5a) Factor contribution Behavior Comp.1 TI Ind. Attempts 0,717 TI Time to movement -0,601 TI Righting time -0,629 LP Reaction score 0,215 LP AI comparison. 0,497 MP Reaction score 0,500 Comp. time alert 0,097 Comp. time exploring -0,100 Comp. # of voc. -0,253 Comp. time agitated -0,205 Comp. time escaping 0,630
Tab.5b) Factor contribution
Behavior Comp. 1 TI Ind. Attempts -0,523 TI Time to movement 0,460 TI Righting time 0,247 LP Reaction score 0,345 LP AI comparison. 0,433 MP Reaction score -0,600 Comp. time alert -0,750 Comp. time exploring 0,450 Comp. # of voc. 0,687 Comp. time agitated 0,539 Comp. time escaping 0,419 Tab5. Each behaviour’s contribution to the fear factor (factor 1) of a) males and b) females. Fig 9b. G C 0.9 0.6 0.3 0.0 -0.3 -0.6 F ea rf ac to r sc or e G C 0.9 0.6 0.3 0.0 -0.3 -0.6 F ea rf ac to r sc or e Fig 9a. G C 0.6 0.3 0.0 -0.3 -0.6 F ea r fa ct or sc or e G C 0.6 0.3 0.0 -0.3 -0.6 F ea r fa ct or sc or e
Fig 9 The mean population score on the fear factor, populations are depicted on the X-axis. a)Display the male score and b) the female score
Fig. 10b, Females 3.0 2.0 1.0 0.0 -1.0 -2.0 -3.0 Factor score 6 4 2 0 N u m be r of B ird s 6 4 2 0 C G P o p ( G = G ot, C = C op ) 3.0 2.0 1.0 0.0 -1.0 -2.0 -3.0 Factor score 6 4 2 0 N u m be r of B ird s 6 4 2 0 C G P o p ( G = G ot, C = C op )
Fig. 10a Males
3.0 2.0 1.0 0.0 -1.0 -2.0 -3.0 Factor score 6 4 2 0 N um be r of B ird s 6 4 2 0 C G P op ( G = G ot, C = C op ) 3.0 2.0 1.0 0.0 -1.0 -2.0 -3.0 Factor score 6 4 2 0 N um be r of B ird s 6 4 2 0 C G P op ( G = G ot, C = C op )
Fig 10 Distribution on the fear factor. a) males b)females
Tab 6 Transcripts that differed between genders
GenBankID Chromosome Annotation M B
CN234245 W Wpkci (WPKCI-8) -4,35 43,28 CN225924 W Wpkci (WPKCI-8) -2,91 40,51 CN217425 W Wpkci (WPKCI-8) -3,47 39,04 CN221959 W Wpkci (WPKCI-8) -2,33 36 CN221776 W Wpkci (WPKCI-8) -1,24 30,17 CN225974 Un Un -1,21 28,01 CN217761 W Wpkci (WPKCI-8) -1,76 13,9
CN232964 2 Similar to Adseverin (Scinderin)
(RCJMB04_23d8) 1,174 9,397
CN225453 15 Similar to immunoglobulin lambda
Fig 11a Gender differences
Fig 11c Fearful/non-fearful birds
Fig 11b Population differences
Fig 11d Fearful/non-fearful females
Fig 11e fearful/non-fearfull males Fig 11a-e: Gene expression
differences. The X-axis displays the M values and the Y- axis the B-values. To be significant differently expressed B ≥ 0, M ≥1, M ≤ -1. The delimitations are shown as lines in the graphs. The transcripts of interest are found in the uppermost squares to the left and right.
a) Gender differences b) Population differences c) Fearful/non-fearfull birds d) Fearfull/non-fearfull females e) fearfull/non-fearfull males
5. Discussion
In this thesis it has been shown that there are differences in gene expression between brains from fearful and non fearful female rjf, but not between fearful and non fearful males. Further, in contradiction to the prediction, only small differences in fear response and no differences in gene expression was found between two populations of rjf in which previous generations have shown significant differences in fear response (Håkansson and Jensen, 2005; Håkansson et al., 2007). Finally, significant differences in gene expression were found between male and female rjf.
The ground predator test was constructed for this study with the aerial predator test described in Håkansson et al 2005 as model. The test worked as planned but was too effective, almost all the test birds showed a high to very high fear response. As the test was design to separate fearful from
non-Tab 7. Candidate genes. A positive M-value indicates that the transcript was upregulated in non-fearful females compared to fearful females. A negative M-value that it is down regulated in non-fearful females compared to fearful females.
GenBankID Chrom. Anotation M B
CN237554 Un Similar to Small nuclear ribonucleoprotein D2 M. musculus and H. sapiens
PREDICTED: similar to small nuclear ribonucleoprotein D2 R. norvegicus
1,119 9,159
CN224985 9 Transferrin (TF) 1,112 6,178
CN221461 15 Ig rearranged light-chain mRNA C-region 1,991 5,931
CN220620 26 Un 1,092 4,165
CN216922 1 Avian leukosis virus strain ev-3 -1,81 4,13 CN231217 15 Bone morphogenetic protein 1 (BMP1) -0,95 3,273 CN230959 1 Avian leukosis virus strain ev-3 (QTL 1) -1,06 2,962 CN220996 1 Glutathione S-transferase kappa 1
(GSTK1)
-1,16 2,67
BU409223 14 Axin 1 (AXIN1) -1,66 2,593
CN219930 15 Transcribed locus, strongly similar to XP_415224.1 similar to Nur77
downstream protein 2 [Gallus gallus]
1,309 2,444
CN221990 32 Bromodomain and PHD finger containing, 3 (BRPF3)
1,559 1,264
CN223361 1 Un 1,68 1,202
CN220264 Un Similar to pol (LOC770705) -1,84 1,172
fearful birds it has to be revised if it is to be used in that context. Possible alterations could be to let the predator model disappear before it reaches the bird, the platform could contain more and/or higher perches, the ethogram could be revised to include more detailed fear related behaviours or the predator could move in the vicinity of the bird but not directly towards it. In it’s existing form, the test was unable to separate the two populations but generated enough variation for the factor analysis.
Despite significant population differences in earlier generations no population difference was found in any of the behavioural tests or in the gene expression comparisons. The differences are thought to originate from the difference in captive environment, but for the last generations they have been kept under the same conditions. This should lead to a homogenisation of the fear responses between the two populations. For further thoughts on this subject see (Håkansson and Jensen, 2007). The fear factor was very different with respect to genders. There was a great difference in which behaviours that loaded high on the factor; the difference in time that each bird spent on alert before and after the exposure of the ground predator was the highest loaded behaviour among the females but the lowest among the males. There were also inversions between some behaviours, the number of induction attempts was highly loaded in both genders but positive with respect to males and negative with respect to females.
Overall there is a higher loading on all the behaviours among the
females which could explain why the female fear factor explains more of the variance than the male one. On the other hand, there are more extreme
values among the males and quite a large number of behaviours do not score high at all on the male fear factor. A possible explanation is that the
behaviour tests give a fairly accurate quantification of fear response with respect to females but not to males which would explain a higher amount of randomness in the male fear factor.
If this is the case it also explains why no gene expression difference was found between the fearful and nonfearful males (fig 11e).
Of the differently expressed transcripts the most interesting were:
• A transcript similar to Small nuclear ribonucleoprotein D2 (CN237554), a RNA regulating protein in M. musculus and H. sapiens. For a review of small ribonucleotides see (Schümperli and Pillai, 2004)
• Transferrin (CN224985), a major Fe transporter (Moos et al., 2007). • Ig rearranged light-chain mRNA C-region (CN221461), a part of the
immune system.
• Part of the Avian leukosis virus strain ev-3 (CN230959), this transcript exists in one repeat in each end of the virus sequence, making it a possible transposable element. It is also found in the chicken genome in the middle of the Quantitative trait loci 1 (Qtl1), a Qtl with documented effects on the behaviour, including fear
response, and morphology of chicken during domestication (Schütz et al., 2002; Schütz et al., 2004). For a review of QTLs and how they affect behaviour se Flint (Flint, 2003).
• Glutathione S-transferase kappa 1 (GSTK1) (CN220996), a predicted membrane transport protein involved in cellular detoxification. (LI et al., 2005)
• Axin 1 (BU409223), a protein involved in the wnt pathway which have been connected to the formation of the CNS, morphognesis of most organs, neural signal transduction and with cancer development (Salahshor and Woodgett, 2005).
• Similar to pol (CN220264), a predicted DNA polymrase.
• Bone morphogenetic protein 1 (BMP1) (CN231217), a metabolic protein involved in skeletal formation.
Taken together these transcripts have the possibility to control large proportion of the animal’s molecular and neurological signalling pathways either directly in the neuron (Axin) and molecular pathways (GSTK1, transferrin and BMP1) or indirect through regulation of transcription and translation (possibly Small nuclear ribonucleoprotein D2, pol and/or
whatever regulating function carried by the avian leukosis virus fragment). These all all assumptions and nothing can be said for certain. First of all the expression analysis most be verified through a quantitative method, e.g. Real Time PCR. But if these results are verified then there glimpses a potential molecular pathway controlling the fear response.
The results in this thesis support the hypothesis that there are differences in gene expression between fearful and non-fearful birds in female rjf but they fail to support the hypothesis with respect to male rjf. Further, the results fail to find any differences in gene expression between Cop and Got birds and finds only minor differences in behaviour. Finally the
genes that were differently expressed between the fearful and non-fearful females indicate a potential molecular pathway controlling the fear response.
6. Acknowledgments
I would like to thank my supervisors Per Jensen, Jennie Håkansson and Daniel Nätt for all input and support, and for giving me free hands to do what ever I liked with this thesis.
I would also like to thank the Applied Ethology group at Linköpings university, the animal keepers at Vreta and the microarray staff at KTH.
7. References
Allison, D.B., Xiangqin, C., Page, G.P., Sabripour, M., 2006. Microarray data analysis: from disarray to consolidation and consensus. Nature Reviews Genetics 7, 55-65.
Bayly, K.L., Evans, C.S., 2003. Dynamic changes in alarm call structure: a strategy for reducing conspicousness to avian predators? Behaviour 140 353-369.
BLAST, N., 2007. NCBI blast http://www.ncbi.nlm.nih.gov/blast/Blast.cgi. Cattel, R.B., 1978. The scientific use of factor analysis in behavioral and life sciences. Plenum press, New York and london.
Edenberg, H.J., Strother, W.N., McClintick, J.N., Tian, H., Stephens, M., Jerome, R.E., Lumeng, L., Li, T.-K., McBride, W.J., 2005. Gene expression in the
hippocampus of inbred alcohol-preferring and -nonpreferring rats. Genes, Brain and Behavior 4, 20-30.
Efron, B., Tibshirani, R., Storey, J.D., Tusher, V., 2001. Empirical Bayes Analysis of a Microarray Experiment. Journal of the American Statistical Association 96, 1151-1160.
Enard, W., Khaitovich, P., Klose, J., Zöllner, S., Heissig, F., Giavalisco, P., Nieselt-Struwe, K., Muchmore, E., Varki, A., Ravid, R., Doxiadis, G.M., Bontrop, R.E., Pääbo, S., 2002. Intra- and Interspecific Variation in Primate Gene
Expression Patterns. Science 296, 340-343.
Evans, C.S., Macedonia, J.M., Marler, P., 1993. Effects of apparent sice and speed on the response of chickens , Gallus gallus, to computer- generated simulations of aerial predcators. Animal Behaviour 46, 1-11.
Fendt, M., 2006. Exposure to Urine of Canids and Felids, but not of Herbivores, Induces Defensive Behavior in Laboratory Rats. Journal of Chemical Ecology 32, 2617–2627.
Fitzsimmons, C.J., Westring, A., Savolainen, P., Lundeberg, J., Hallböök, F., Jensen, P., Andersson, L., 2006. Altered patterns of gene expression in the hypothalamus and thalamus regions of White Leghorn and red jungelfowl chickens. Doctoral thesis, Uppsala university, paper III.
Flint, J., 2003. Analysis of Quantitative Trait Loci That Influence Animal Behavior. Journal of Neurobiology 54, 46–77.
Forkman, B., Boissy, A., Meunier-Salaün, M.-C., Canali, E., Jones, R.B., 2007. A critical review of fear tests used on cattle, pigs, sheep, poultry and horses.
Physiology & Behavior 92 (2007) 340–374 92 340–374.
Håkansson, J., Bratt, C., Jensen, P., 2007. Behavioural differences between two captive populations of red jungle fowl (Gallus gallus) with different genetic
background, raised under identical conditions Applied Animal Behaviour Science 102, 24–38.
Håkansson, J., Jensen, P., 2005. Behavioural and morphological variation between captive populations of red junglefowl (Gallus gallus) – possible implications for conservation. Biological Conservation 122, 431–439.
Håkansson, J., Jensen, P., 2007. A longitudinal study of antipredator behaviour in four successive generations of two populations of captive red junglefowl. Part of doctoral thesis, Linköping university university, paper IV.
Jaako-Movits, K., Zharkovsky, T., Romantchik, O., Jurgenson, M., Merisalu, E., Heidmets, L.-T., Zharkovsky, A., 2005. Developmental lead exposure impairs contextual fear conditioning and reduces adult hippocampal neurogenesis in the rat brain. International Journal of Developmental Neuroscience 23 627-635. Keer-Keer, S., Hughes, B.O., Hocking, P.M., Jones, R.B., 1996. Behavioural comparison of layer and broiler fowl: measuring fear responses. Applied Animal Behaviour Science 49, 321-333.
LI, J., XIA, Z., DING, J., 2005. Thioredoxin-like domain of human k class
glutathione transferase reveals sequence homology and structure similarity to the y class enzyme. Protein Science 14, 2361–2369.
Lindberg, J., 2007. Exploring Brain Gene Expressioni Animal Models of Behaviour. Doctoral thesis, Uppsala university.
Lindqvist, C., Janczak, A.M., Nätt, D., Baranowska, I., Lindqvist, N., Wichman, A., Lundeberg, J., Lindberg, J., Torjesen, P.A., Jensen, P., 2007. Transmission of Stress-Induced Learning Impairment and Associated Brain Gene Expression from Parents to Offspring in Chickens. PLoS ONE 2, e364.
Macrí, S., Würbel, H., 2007. Effects of variation in postnatal maternal environment on maternal behaviour and fear and stress responses in rats. Animal Behaviour 73, , 171-184.
Mongeau, R., Miller, G.A., Chiang, E., Anderson, D.J., 2003. Neural Correlates of Competing Fear Behaviors Evoked by an Innately Aversive Stimulus. The
Journal of Neuroscience 23, 3855–3868.
Moos, T., Rosengren Nielsen, T., Skjørringe, T., Morgan, E.H., 2007. Iron trafficking inside the brain. Journal of Neurochemistry 103, 1730–1740.
Ponder, C.A., Kliethermes, C.L., Drew, M.R., Muller, J., Das, K., Risbrough, V.B., Crabbe, J.C., Gilliam, T.C., Palmer, A.A., 2007. Selection for contextual fear conditioning affects anxiety-like behaviors and gene expression. Genes, Brain and Behavior 6, 736–749.
Price, E.O., 1999. Behavioral development in animals undergoing domestication. Applied Animal Behaviour Science 65, 245–271.
Ramos, A., Correia, E.C., Izídio, G.S., Brüske, G.R., 2003. Genetic Selection of Two New Rat Lines Displaying Different Levels of Anxiety-Related Behaviors. Behavior Genetics 33, 657-668.
Salahshor, S., Woodgett, J.R., 2005. The links between axin and carcinogenesis. Journal of Clinical Pathology 58 225-236.
Schümperli, D., Pillai, R.S., 2004. The special Sm core structure of the U7 snRNP: far-reaching significance of a small nuclear ribonucleoprotein. Cellular and Molecular Life Sciences 61, 2560–2570.
Schütz, K., Kerje, S., Carlborg, Ö., Jacobsson, L., Andersson, L., Jensen, P., 2002. QTL Analysis of a Red Junglefowl 3 White Leghorn Intercross Reveals Trade-Off in Resource Allocation between Behavior and Production Traits. Behavior Genetics 32, 423-433.
Schütz, K.E., Forkman, B., Jensen, P., 2001. Domestication effects on foraging strategy, social behaviour and different fear responses: a comparison between red jungelfowl (Gallus gallus) and a modern layer strain. Applied Animal
Behaviour Science 74, 1-14.
Schütz, K.E., Jensen, P., 2001. Comparison of Red Junglefowl (Gallus gallus) and Two Domesticated Breeds of Poultry. Ethology 107, 753-765.
Schütz, K.E., Kerje, S., Jacobsson, L., Forkman, B., Carlborg, Ö., Andersson, L., Jensen, P., 2004. Major Growth QTLs in Fowl Are Related to Fearful Behavior: Possible Genetic Links Between Fear Responses
and Production Traits in a Red Junglefowl × White Leghorn Intercross. Behavior Genetics 34, 121-130.
Svartberg, K., Forkman, B., 2002. Personality tratis in the domestic dogs (Canis familiaris). Applied Animal Behaviour Science 79, 133-155.
Takahashi, L.K., Nakashima, B.R., Hong, H., Watanabe, K., 2005. The smell of danger: A behavioral and neural analysis of predator odor-induced fear.
Neuroscience and Biobehavioral Reviews 29 1157–1167.
Trut, L.N., 1999. Early Canid Domestication: The Farm-Fox Experiment. American Scientist 87, 160-169.
Vaince, F., Bona, J., Fathallah-Shaykh, H.M., 2006. Microarray Data Analysis: Current Practices and Future Directions. Current Pharmacogenomics 4, 209-218. Yang, Y.H., Speed, T., 2002. Design issues for cDNA microarray experiments. NATURE REVIEWS GENETICS 3, 579-588.
