Animals

All animal procedures were approved by the 4^th Local Ethics Committee for Animal Experimentation in Warsaw (Permission Number: 39/2014). The procedures were carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and documented in compliance with the ARRIVE guidelines. All efforts were made to minimize suffering.

All animals used in the study were maintained in standard conditions in accordance with the requirements of the Directive 2010/63/EU of the European Parliament and of the Council on the protection of animals used for scientific purposes.

For imaging studies young adult male rats at the same age (8–9 week old) were sourced from the following three populations:

1. The Warsaw Wild Captive Pisula-Stryjek rats (WWCPS, n = 9), considered as a model of population for the wild Rattus norvegicus, were obtained from a colony maintained in the facility of the Institute of Psychology, Polish Academy of Sciences, Warsaw, Poland. The WWCPS rats were derived in 2006 from sample originating from five independent colonies of feral rats [20]. In order to prevent the development of domestication in the breeding colony, new genetic material from rats captured freshly in a variety of locations has been systematically introduced to the colony. To avoid the stress response to experiment only the second and third generation (F2-F3) of laboratory-reared WWCPS wild rats were selected.
2. Brown Norway BN/CrlCmd (BN, n = 8) originated from the inbred strain maintained at the Mossakowski Medical Research Centre, Polish Academy of Sciences, Warsaw, Poland (MMRC). The parent pairs were obtained in 1997 from the Charles River Laboratories, Germany.
3. Wistar Cmd:WI(WU) (WI, n = 8) originated from outbred stock maintained at MMRC. The parent pairs were obtained in 1987 from Charles River Laboratories, Germany.

Magnetic resonance images acquisition and processing

MR brain imaging was performed with a Bruker BioSpec 70/30 Avance III 7T system, equipped with a transmit cylindrical radiofrequency coil (8.6 cm inner diameter) and a rat brain dedicated receive-only array coil (2x2 elements) positioned over the animal’s head. The animals were anaesthetized with isoflurane (4% for induction, 1.5–2% in oxygen for maintenance, supplied through a mask), and placed on a dedicated rat scan bed in the prone position with the head fixed in a stereotactic apparatus with plastic retainers for the teeth and ears (Bruker, Germany). Physiological monitoring, including respiration rate and body temperature, was performed throughout the imaging sessions.

Structural transverse MR images covering the whole brain were acquired with the T2-weighted TurboRARE sequence (TR/TEeff = 5000/30ms, RARE factor = 4, spatial resolution = 125μm x 125μm x 500μm, 54 slices, no gaps, number of averages = 3). The structural images were exported in DICOM format (.dcm) and then converted to NifTI format (.nii).

For the segmentation of individual brain image structures we employed the in vivo rat atlas (created from images of the brains of 97 male Sprague-Dawley rats) and published by Schwarz et al. [21] that provides template image covering the whole brain and labelled images of 50 cortical and subcortical delineated brain structures based on the rat brain atlas by Paxinos and Watson [22]. To improve accuracy of structure delineation we transformed the Schwarz template image and the labelled image into tissue probability maps space associated with the template image provided by Valdes-Hernandez et al. [23] that is based on in vivo images of the rat brain.

The analysis of the image data was performed in the following steps (Fig 1):

1. Intensity correction. An intensity bias field correction was performed on the data using N4ITKBiasFieldCorrection, a tool distributed with the 3D Slicer software [24].
2. Image registration. After intensity, non-uniformity correction was implemented with the same origin of the coordinate system as in the template image was set for each individual image. Using tissue priors, each subject image was segmented using SPM8 into grey matter, white matter and cerebrospinal fluid. This method additionally provides forward and inverse transformation to the template space.
3. Structures labelling. The inverse transformations obtained from the image registration (step 2) were then used to warp labelled atlas images into the space of each individual subject image using the IBASPM toolbox for SPM5 [25].
4. Volumetric measurements. Absolute volumes for subject segmented structures and the whole brain volume were calculated by multiplying number of the voxels belonging to the structure by voxel volume [26, 27]. Scaled (normalized) volumes were obtained by dividing structure volume by whole brain volume.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. A simplified image processing pipeline.

High resolution T2-weighted images were intensity corrected and automatically segmented for white matter, grey matter and cerebrospinal fluid. In next step the images were transformed to brain template which allowed applying labels (ie. identification of brain structures defined in brain atlas). Back-transformation to original dimension allowed identification of brain structures in the original images and calculation of volumes of the pre-defined structures.

https://doi.org/10.1371/journal.pone.0215348.g001

The segmentation accuracy was validated by comparing the results of automated segmentation with the results of manual segmentation of the hippocampus. Manual segmentation was performed in 3D Slicer software (version 4.8.1, www.slicer.org). The volumes of automatically segmented hippocampi were compared with those of manually segmented hippocampi according to Badea et al. [28]. Volume difference (VD) was calculated as follows:

The olfactory bulb (and the accessory olfactory bulb) was manually segmented in 3D Slicer according to the Paxinos atlas [29] and Paxinos MRI/DTI atlas [30].

Statistical analysis

We used STATISTICA for Windows (version 10, StatSoft, Inc., Tulsa, OK, USA, www.statsoft.com). Levene’s test revealed significant differences in variances for scaled brain volume and several brain structures (retrosplenial cortex, accumbens shell, bed nucleus stria terminalis, corpus callosum, mesencephalic region, raphe; S3 Table), therefore non-parametric tests were chosen. Significances of differences between groups were evaluated with the Kruskal-Wallis test followed by U Mann-Whitney test. Critical levels of significance were corrected for multiple comparisons with Bonferroni method. For overall significance level α = 0.05 and testing three (m = 3) hypothesis, then we tested each individual hypothesis at α = 0.05/3 = 0.01(6) according to the equation:

Domestication-related changes in brain size

The rats from all populations were in the same maturation state, ie. young adults. However, the weights of the animals were significantly different as expected. The differences in weight between different rat strains are well known and our weight records are in good agreement with those provided by professional animal suppliers [36, 37].

Brain volume was bigger in rats with greater body weight, ie. Wistar>BN>WWCPS. However, brain to body weight ratio was similar in BN and WWCPS rats but smaller in Wistar. Brain size usually increases with body size, but most domesticated animals often have a greater body weight and relatively smaller brain volume than their wild counterparts [38–41]. Some of the structural brain differences may be associated with inbreeding and albinism. While, Wistar are the one of the oldest albinotic laboratory outbred stock, BN is an inbred strain that received a significant genetic contribution from wild rat [42]. The laboratory strains also differ among themselves in behavior. The small differences in brain volume between WWCPS and BN rats may result from the much shorter time of BN domestication than WI rats.

As brain size in animals generally increases with body weight, it could be assumed that body weight and brain size are tightly related. For this reason evolutionary and allometric studies in animals frequently base on relative brain size rather than absolute brain size [41, 43]. Because of the difference in the total brain size, comparison of “absolute” volumes of the segmented brain structures was not very informative: virtually all brain structures were bigger in Wistar rats than in WWCPS. To detect possible differences in the brain proportions within the animal groups, we have presented volumes of the segmented brain structures as a percentage of total brain volume.

Domestication-related changes in regional brain morphology

Comparing brain structures in wild-captive rats with laboratory rats we detected, a reduction in the brain volume to body weight ratio in WI and noticed that the changes in proportions of particular brain structures do not appear homogeneously but rather in a mosaic fashion. Some cortical structures seem to be more affected by domestication than other areas developed at earlier phylogenetic stages. We analyzed 50 segmented brain regions based on the MRI Schwarz atlas and extensively studied the structures which could be coupled with biobehavioral profile changes during the domestication process such as lower aggression, socio-positive maternal behavior and cognitive abilities, as well as less hormonal stress responses [44]. These changes are caused by systematic selection made by humans and adaptations to the exposure to strictly defined artificial environments with standardized diet, low level of external stimuli or without threat from the predators.

The relative total cortex volume in laboratory rats was smaller than in wild rats. After cortex segmentation there were a reduction in almost all of cortical structures volume in albino Wistar rats and in few in BN. In particular we observed significant differences in some functionally related parts of cortex, especially in the area responsible for processing sensory information (visual, retrosplenial, parietal and auditory cortex) or controlling movements (motor cortex). These areas are essential for the survival of animals in the wild and were significantly decreased in laboratory rats. Similarly, decrease in total cortex and areas involved in processing of the sensory information and motor system was found both in early domesticated mammals (eg. guinea pig Cavia apera f. porcellus, pig Sus scrofa, llama Lama glama, sheep Ovis ammon) and in the relatively recently domesticated species (eg. gerbil Meriones unguiculatus or mink Mustela vison energumenos) [40, 44]. Only the entorhinal cortex which is considered to be periallocortex and is involved in spatial learning in rodents was significantly larger in WI rats compared to wild rats.

Changes in structures related to sensory processing and motor functions

Many laboratory strains of this species are albinotic and develop various ocular pathologies resulting in impaired visual acuity [45] as well as structural abnormalities in the connections of the occipital cortex [16]. Importantly, abnormalities in the functioning of the visual system are manifested already in very young animals (eg. 1.5 months old Wistar rats) [46]. We observed the largest reduction in the relative volume of the visual cortex in albino Wistar rat when compared with both pigmented rats (WI and WWCPS about 9% and 4% between WI and BN) whereas, there was no difference between WWCPS and BN. Also, the retrosplenial and temporal association cortex which border and have functional connections with the visual cortex presented a smaller volume in WI then in pigmented rats. Our study confirms that changes in visual acuity in albinotic Wistar rats lead to structural changes in the vision-related areas. This diminution of visual structures could be a remarkable example for conspecific variability and brain plasticity. This fact should be considered during selecting animal models for neurological studies, especially for behavioral tests associated with vision.

We detected only slight differences in somatosensory cortex (that is involved in processing of proprioception and touch, including whiskers sensory system) volume between BN and WWCPS. However the sensory association areas, such as, the parietal cortex that integrates sensorimotor information among various modalities (visual, auditory, whiskers) was significantly decreased in laboratory rats. The relative volume of the of auditory area in cerebral cortex was also significantly smaller in the BN than in WWCPS. Such a notable reduction was not observed in albino WI rats. The principal auditory relay, the medial geniculate nucleus was even slightly enhanced in WI.

In rats, smell plays an important role in recognition of their offspring and partners or predators and searching for food. In domesticated animals, the components of the olfactory system (in particular the olfactory bulbs) are usually reduced. [43, 47].

In our study the olfactory bulbs (OB) were significantly smaller in laboratory pigmented BN strain than in WWCPS rats but not in the albino Wistar rat. Other structures involved in the sense of smell such as olfactory nuclei and olfactory tubercles and entorhinal cortex were relatively bigger in WI than WWCPS but not in the BN rat.

The larger olfactory- and auditory-related brain structures in albino rats than in pigmented animals could be a result of compensation for the visual impairment, i.e. activation the mechanisms of cross-modal neuroplasticity.

In the natural environment wild animals have to escape from predators, explore large areas to find food, shelter or mates. Laboratory rats show less exploratory behavior than their wild conspecifics [48]. The motor cortex which is the area of the cerebral cortex involved in the planning, commanding and controlling voluntary movements were significantly smaller in laboratory rats when compared to wild rats. Laboratory rats had also noticeably smaller parietal and retrosplenial cortex, structures which has connection with motor cortex and play a role in navigational motor planning. For example, it was shown that destruction of a part of the retrosplenial cortex resulted in worse spatial movement parameters such as running speed in rats [49]. The reduction of motor areas in the cortex was also noted in other domesticated mammals [50]. It could be explained by breeding in restricted space with limited ability to perform motor activities.

Changes in structures possibly related to emotions and stress reaction

Artificial selection during domestication leads to consolidation of traits that are more advantageous and desirable for humans. Namely, animals are selected which present less aggression toward people and other animals, more sociable behavior, high fecundity, and diminished stress responsiveness caused by crowding in cages and restriction of movement. Our study shows significant differences in the relative volume of the entorhinal cortex, amygdala, hippocampus and nucleus accumbens in laboratory rats compared to their wild counterparts. These structures are involved in the regulation of various functions including memory and emotional response such as aggression, fear or motivation. The most significant differences in these structures were observed between WWCPS and Wistar rats. Notably, Wistar stock is characterized by extremely docile behavior and high fertility.

Hippocampus was relatively larger in both laboratory strains than in wild rats. Dorsal hippocampus, which in rodents is involved mainly in spatial memory [51], was slightly bigger. However, the most significant differences were found between the WI and WWCPS rats in relative volume of the ventral hippocampus. This structure, responsible for learning and memory function, is also involved in modulation of anxiety and aggression [52]. Similarly, in farm-bred foxes (Vulpes vulpes) selected for tameness Huang et al. showed increased neurogenesis within ventral hippocampus which consequently would lead to enlargement of hippocampal size [53].

Generally, the amygdala has definitely increased in size during evolution process [54]. Significantly higher relative volume in amygdala was noted in both laboratory strains. Similarly, slight increase of relative volume was found, by Kruska for the amygdala volume, in domesticated guinea pig when compared to wild cavies [47]. The increased of relative volume was also shown in WI rats in the core and shell of nucleus accumbens. These structures are involved in the cognitive processing of reward, aggression and sexual behavior in rats. The most pronounced differences were observed in the accumbens shell which is strongly connected to amygdala.

Changes in structures related to memory and learning

Until recently, it was considered that due to domestication the brain size is reduced and animals lose a lot of their cognitive and learning abilities. However, a recent study on spatial learning with use the Morris water maze shows that some domestic animals such as Guinea pig or Mongolian gerbil [44, 55] have equivalent or even better learning performance than their wild ancestors. These findings are similar to those on rats, in which the laboratory rats might also outperform the wild rats in spatial learning and memory tasks [56, 57]. Domesticated animals including laboratory rats seem to adapt faster to man-made environment and better cooperate with the experimenter, whereas wild animals exhibit excessive emotional responses and they are more susceptible to stress during the test. For this reason wild animals can be worse in solving spatial memory tasks in laboratory conditions. On the other hand, they are more effective in complicated tasks that require a higher cortical coordination (wild cavies, gerbils, foxes, dogs) [58, 59].

In our study we find that relative brain volume was smaller in laboratory WI rats. Particularly some cortical regions involve in the cognitive and memory functions including visuospatial orientation and planning the future such as the retrosplenial, temporal associations and parietal association cortex were decreased considerably. Furthermore, phylogenetically older structures such the hippocampus (archeocortex) which play a crucial role in learning and memory processes (especially spatial memory) in rodents, were changed but not as much as the phylogenetically younger cortex structure. Only the enthorhinal cortex (periallocortex) which cooperate with hippocampus by registering information on grid maps of the surrounding area had extraordinarily bigger volume in WI rats. Relative total volume of the hippocampus was significantly bigger in both laboratory strains compared to WWCPS. The differences were observed in the anterodorsal, posterior and ventral hippocampus and subiculum. In particular, the most pronounced difference was observed in the relative volume of the ventral hippocampus between WWCPS and both the laboratory strains.