Transgenic mice expressing constitutive active MAPKAPK5 display gender-dependent differences in exploration and activity
© Gerits et al. 2007
Received: 28 September 2007
Accepted: 12 November 2007
Published: 12 November 2007
The mitogen-activated protein kinases, MAPKs for short, constitute cascades of signalling pathways involved in the regulation of several cellular processes that include cell proliferation, differentiation and motility. They also intervene in neurological processes like fear conditioning and memory. Since little remains known about the MAPK-Activated Protein Kinase, MAPKAPK5, we constructed the first MAPKAPK knockin mouse model, using a constitutive active variant of MAPKAPK5 and analyzed the resulting mice for changes in anxiety-related behaviour.
We performed primary SHIRPA observations during background breeding into the C57BL/6 background and assessed the behaviour of the background-bred animals on the elevated plus maze and in the light-dark test. Our results were analyzed using Chi-square tests and homo- and heteroscedatic T-tests.
Female transgenic mice displayed increased amounts of head dips and open arm time on the maze, compared to littermate controls. In addition, they also explored further into the open arm on the elevated plus maze and were less active in the closed arm compared to littermate controls. Male transgenic mice displayed no differences in anxiety, but their locomotor activity increased compared to non-transgenic littermates.
Our results revealed anxiety-related traits and locomotor differences between transgenic mice expressing constitutive active MAPKAPK5 and control littermates.
The mitogen-activated protein kinases, MAPKs for short, constitute cascades of signalling pathways involved in the regulation of several cellular processes that include cell proliferation, differentiation, apoptosis, motility and embryogenesis. MAPKs also influence the development of cancer, inflammatory processes and most important in this article, neurological processes [1–4].
The MAPK pathway exists of a four component module where a MAPK kinase kinase (MAP3K) activates a MAPK kinase (MAP2K) by phosphorylation. The latter subsequently phosphorylates and activates a MAPK. This occurs on two residues within the TXY activation motif. MAPKs can in turn phosphorylate downstream targets like transcription factors but also other kinases, referred to as MAPK-activated protein kinases, MAPKAPKs or MKs. This branch of the MAPKs consists of 5 subfamilies: the ribosomal S kinases (RSK), the mitogen- and stress activated protein kinases (MSK), the MAPK-interacting kinases (MNK), MK2 and MK3, and finally MK5.
Although the nomenclature suggests that MK2, MK3 and MK5 belong to the same branch of the MAPKAPKs, there exist considerable differences (in sequence and function) between MK2 and MK3 on one hand and MK5 on the other [3, 5]. To date, little is known about MK5 its biological role. To explore the biological function of such a protein, one can either delete the gene encoding for the protein (knockout or KO strategy), or use a knockin strategy (see further) in mice. Examination of the resulting phenotypical changes can reveal more about the proteins' function. In the case of MK5, the KO mouse model, displayed either no particular phenotype or embryonic lethality, depending on which genetic background the mice were bred [6, 7]. However, a recent report on a different MK5 KO model showed that MK5 deficiency enhanced DMBA-induced skin papillomas, which indicates a role for MK5 as tumour suppressor . An alternative approach to study the function of a protein is to mutate the endogenous gene or to introduce an exogenous copy of the gene that carries the mutation, resulting in the expression of a dominant negative or constitutive active form of the protein. This approach is called a knockin (KI) strategy. The mutant form of the protein may perturb cellular functions and thereby allow the deduction of the normal role for the genuine protein. Currently, only four MAPK models that use the KI strategy have been described. These include the kinases B-Raf, MEKK3, PAK1 and TAK1. No such models exist for any of the known MAPKAPKs .
In this report we describe for the first time a KI MAPKAPK mouse model which expressed a constitutive active variant of MK5. The resulting mice, called MK5 L337A transgenic mice, were bred onto a C57BL/6 background. With high brain expression levels for MK5 and the data from our primary SHIRPA protocol, which indicated a role in anxiety-related processes, we decided to conduct two validated anxiety tests (the elevated plus maze (EPM) and the light-dark box (LD)). Our results suggest an unexpected difference between male and female mice, as well as between transgenic (TG) and non-transgenic (NTG) mice in exploration and activity of the respective tests. Furthermore, these results indicate for the first time the involvement of a MAPKAPK in an anxiety-related context.
Construction of the vector
We mutated the first NotI site (NotI820) of pCMV-β (Clontech) to a KpnI site using following primers : forward 5'-GCT-GCG-GAA-TGG-TAC-CCG-CGG-CCG-CAA-TTC-3' and reverse 5'-GAA-TTG-CGG-CCG-CGG-GTA-CCA-TTC-CGC-AGC-3'. To create the same site for the insert for ligation, we amplified the insert from pcDNA-HA-MK5 (described in ) using following primers: forward 5'-CGG-CGG-GGT-ACC-ATG-TAT-GAT-GTT-CCT-G-3' and reverse 5'-CTA-TAG-AAT-AGC-GGC-CGC-TAG-ATG-CAT-GC-3'. The β galactosidase fragment was cut out using restriction enzymes KpnI and NotI and replaced with the cDNA sequence of HA-tagged MK5. The subsequent plasmid was sequenced and the leucine 337 of MK5 mutated to an alanine, using primers: forward 5'-CAG-GCG-CAT-GCC-GAG-CAG-GCG-GCA-AAC-ATG-AGG-ATC-3' and reverse 5'-GAT-CCT-CAT-GTT-TGC-CGC-CTG-CTC-GGC-ATG-CGC-CTG-3'. This single amino acid substitution has previously been shown to render MK5 constitutive active . For microinjection the plasmid was linearized with SphI to generate a 2.5 kb fragment. Transient transfection of pCMV-MK5 L337A into PC12 and COS cells confirmed expression of the transgene by RT-PCR and immunoblotting with anti-HA and anti-MK5 antibodies.
The resulting mice were genotyped by PCR using Jumpstart (Invitrogen). Following primers were used to specifically detect the transgene alone: forward 5'-GAG-CTG-GTT-TAG-TGA-ACC-GTC-3' and reverse 5'-CTT-TAT-CTG-TGA-ATC-CAC-GGC-CAT-TC-3'. The forward primer is complementary with the CMV promoter, and the reverse primer is complementary to MK5 sequences, allowing specific detection of the transgene. To validate the quality of genomic DNA, a genomic fragment of β-globin was amplified using the β-globin forward 5'-CCA-ATC-TGC-TCA-CAC-AGG-ATA-GAG-AGG-GCA-GG-3' and reverse 5'-CCT-TGA-GGC-TGT-CCA-AGT-GAT-TCA-GGC-CAT-CG-3'. The former led to a 1.6 kb fragment, the β globin primer set generated 500 bp fragments. PCR conditions were: 5 min. at 95°C; then for 35 cycles: 95° for 30 sec, 60°C for 30 sec, 72°C for 1 min; finally at 72°C for 10 min.
The transgenic mouse project was approved by the "Forsøksdyrutvalget" (Laboratory Animal Department) and the "Sosial og Helsedirektorat (Social and Health Directorate) and performed in accordance with the Norwegian Animal Welfare Act and the European Community Council Derivative. Animals were maintained on a 12 h light/dark cycle (light on at 08:00) in a humidity, air and temperature (20–22°C) controlled environment, and tested during the light cycle. The animals were provided with food and tap water ad libitum. At weaning (21 days ± 1 day), the offsprings were earpunched and genotyped.
Mice were sacrificed with CO2. The F7 generation (all the same age) used for the anxiety tests was sacrificed at the same moment, whereas animals during background breeding were sacrificed over different time spans for reasons of fighting wounds, tumours, old age or place restraints. After asphyxiation, perfusion fixation was first performed with PBS, then 4% formaldehyde in PBS. Organs (liver, spleen, reproductive organs, heart, stomach, brain, lungs) were removed and stored overnight in 4% formaldehyde in PBS, and transferred to storage buffer containing 0.5% formaldehyde in 200 mM Hepes (pH 7.2) the next day. Normal tissue architecture was examined by hematoxyline-eosine staining on brain, heart and spleen slices.
A modified primary SHIRPA protocol was performed on the background bred animals . Briefly, mice were observed during a five-minute period in a jar for spontaneous activity, body position, respiration and tremor. Then mice were transferred to the arena (a rat cage of 59.5 – 38 – 20 cm) where transfer arousal, palpebral closure, piloerection, gait, pelvic elevation, tail elevation, touch escape were monitored. We checked the following in or above the arena: trunk curl, limb grasping, visual placing, grip strength, pinna reflex, corneal reflex, toe pinch, wire manoeuvre, skin colour, lacrimation, salivation, provoked biting, negative geotaxis and vocalization.
During the light cycle, mice were transferred to the testing room 20 minutes prior to the anxiety tests. The Elevated Plus Maze (EPM) was constructed as previously described [12, 13]. Mice (2 months old) were placed in the centre of the maze facing one of the open arms. We videorecorded each mouse for 5 minutes. During this time, the amount of entries into each arm was counted, the time spent in each arm and the centre time were measured and the amount of fecal boli was recorded.
The Light-Dark Box (LD), was constructed by using a rat cage (59.5 – 38 – 20 cm), spray painted black for one third of the area (walls and floor) and the floor of the 2/3 part was spray painted white [12, 14]. The closed area contained a dark roof top above the dark area and a separation wall. In the middle of the wall at floor level there was an opening (7.5 – 7 cm) through which the mouse could pass on to the other compartment. Mice were placed in the middle of the white area with their backs towards the opening in the wall, and videorecorded for 10 minutes. After each individual mouse test, we removed excretion from both the maze and the LD and rinsed with water.
We videorecorded each mouse for both tests with an IIDC Point Grey research camera connected to a standard personal computer running Linux 2.6.18. For the EPM we used a Medion lens able to acquire the full size of the elevated maze. For the light-dark test we relied on a Canon 25 mm television lens. Coriander version 1.03 grabbed images in full colour at a resolution of 1024 – 768 pixels and streamed them to V4L2 frame buffer at 15 frames per second. Mencoder version 1.0rcl-4.1.2 recorded and encoded the videostreams. After video acquisition, all images were cropped to the area under investigation. We removed the beginning of each videorecording where the mouse was placed on the maze. Time points where the mouse fell off the maze (n = 4NTG) were removed from the video stream. For the EPM, we used only the first 4500 frames (5 minutes). For the light-dark test, we used the first 9000 frames (10 minutes).
We aligned all movies using 4 calibration points and a standard affine texture mapping algorithm, without oversampling. For each video, we calculated the background as the modus of the video stream. Subtracting this background from each frame made it possible to create 'yes/no' images. These images reflect the mouse presence at specific time points. To measure the mouse-activity, we calculated the difference between each frame and the previous frame and limited the result to binary values.
The association values between mice and the number of times they vocalized, bit or jumped were calculated using contingency table analysis by means of the algorithm given in . The significance levels were calculated using Yates correction  since all variables could only take on two values. The probabilities that these values could be a random effect was calculated (p-value) using a standard Chi-square distribution with 1 degree of freedom. The probability that the experiment can be repeated and will give the same value was calculated (p-rep) according to .
The EPM and light-dark test were first analyzed as described in Methods. The numerical output was grouped into female/male transgenic/non-transgenic mice. For each group the standard deviation and mean were calculated. The difference of the means was calculated as well as the standard deviation of the difference. A one tailed T-test was performed to assess the significance of each measured difference. When the variation of the two groups was the same we relied on the homoscedastic T-test, otherwise we used the heteroscedastic T-test. We refrained from performing an ANOVA because most of the data did not show a Gausian distribution.
To avoid impact of potential differences in mouse surface area of the different groups, we calculated the mean surface area for each group and accounted for them in the further analysis. The differences in surface area were insignificant.
Construction of the MK5 L337A mice
Positive founders and subsequent generations of transgenic mice were identified by PCR with CMV and MK5 primers as described in Materials and Methods. To assure that the lack of HA-MK5 DNA in the PCR (Figure 2B and 2C) was not the result of insufficient DNA or low DNA quality, β-globin primers were used to amplify β-globin DNA sequences. For the background breeding, we crossed three female founders with male C57BL/6 WT mice to produce the first generation of transgenic mice. Subsequent transgenic generations were bred to background with WT C57BL/6 mice for 7 generations, whereby each generation was screened for transgenic positive offsprings by PCR (Figure 2B and 2C). The expression of the transgene was verified by RNA isolation from mouse heart tissue (Figure 2D). We found the highest transgene expression in line B and decided to work with this line.
General observations of MK5 L337A mice during background breeding
Overview of defensive behaviour of mice during background breeding.
During background breeding we also observed the following occurrences : five Staphylococci infections (2TG vs 3NTG), indicating no particular differences in sensitivity towards Staphylococci infection between TG and NTG mice. The -only- two tumours we discovered, we found in transgenic mice. Both were benign adenomas and originated from the liver and skin (results not shown). Although C57BL/6 can develop tumours spontaneously, we did not detect any in the NTG mice. We also observed at least one closed eyelid in 2 FTG, 7 FNTG mice, and 1 MNTG mouse. The C57BL/6 strain displays congenital defects in approximately 10% of the WT mice, including eye defects. The cause of the eye defect lies most likely in deficiencies in lens development. These results indicate that although the genetic background is the same for all the groups, FNTG mice are more sensitive to these eye defects than MNTG, MTG and FTG mice. In addition, we observed 4 cases of obesity (>40 g) during old age, 3 of them were founder mice and included the founder mouse for line B (which we selected for further analysis based on high TG expression levels) and 1 was an F1 TG mouse of line B. Finally, C57BL/6 mice are genetically predisposed to hydrocephaly (1–4%), we observed 2 cases in NTG mice and none in TG mice. Other C57BL/6 traits which we observed as well: high wheel activity, high incidence of tail rattling and high locomotor activity.
Differences in open arm exploration on the EPM
Overview of differences between trangenic and nontransgenic mice in body weight and behaviour on the EPM and in the LD box.
Average Weigth (g)
Stdev weigth (g)
Surface area (px/ms)
Elevated plus maze
# Head dips
Time open (%)
Time closed (%)
Time center (%)
Light dark box
Mouse activity (px/ms. fr)
We performed the EPM first, because of its high sensitivity towards previous experience [18, 19]. Traditionally, the nose dips on the open arms, the time spent in the open/closed compartments and the number of entries into open/closed arms are recorded. Table 2 illustrates these parameters for each gender in TG and NTG mice. In our case, we find on average 6.86 (p = 0.03) more head dips in FTG mice than in FNTG littermates, whereas there is no significant difference in head dips between MTG and MNTG mice. The MK5 L337A female mice spend on average 8.65% (p = 0.09) more time in the open arms compared to the FNTG, whereas there is no significant difference between MNTG and MTG mice. We found no statistically significant differences between FNTG, FTG, MNTG and MTG mice on the number of entries into the open arm of the EPM.
Taken together, for NTG and TG mice, we found no differences in the number of entries into the open arms, but an increased number of head dips and time spent on the open arm of the EPM for TG compared to NTG mice. When subdivided into genders, we found that FTG mice displayed an increased amount of head dips on the open arm and resided longer in the open arm compared to littermate controls, but did not differ in the number of open arm entries. Additionally, our image analysis allowed us to detect that FTG mice also ventured further into the open arm compared to FNTG controls. MNTG and MTG mice did not differ in open arm head dips or time spent in the open arm, but MNTG explored further into the open arm compared to MTG. These results indicate a difference between FTG and FNTG in open arm exploration on the EPM, in favour of FTG being less anxious. It also demonstrates a gender difference between female and male mice in their exploration of the open arm which would have escaped us if we had not distinguished between gender.
Mouse presence probability and activity in the closed arm on the EPM
Detailed analysis of the mouse presence probability and activity profiles on the elevated plus maze
As shown in the previous section, FNTG mice displayed a higher activity from the middle of the closed arm towards the end, as indicated by a smaller white box in Figures 3A and 3B. Similarly, in Figure 5B, we see that FNTG mice display a higher activity from the middle of the closed arm to the end (vertical downward arrow). The activity is accompanied by a higher mouse presence in the middle of the closed arm, but resembles that of FTG mice near the end of the closed arm. This indicates that FNTG are more active and more present in the centre area of the closed arm compared to the FTG mice. The black ellipse reflects that mice actively investigate a small shadow cast at the end of the closed arm, but this exploration does not last long enough to influence the total mouse presence.
The MTG mice are more present in the first part of the open arm compared to MNTG mice as indicated by Figure 3C and 3D and by the horizontal open arrow in Figure 5C. However, the remainder of the mouse presence probability of MTG resembles that of MNTG, as illustrated in Figure 5C.
Summary of the results.
% time open
% time closed
% time open
% time closed
In summary, FTG mice displayed an increased presence in the open arm, and a lower presence and activity in the middle of the closed arm, when compared to FNTG mice. MTG are more present in the first part of the open arm, but resemble the MNTG for the rest of the open arm. In the closed arm, MTG are more present in the middle of and more active over the entire closed arm compared to MNTG mice. They also investigate the shadow area more profoundly, resulting in increased presence and activity.
Taken together, we found that FTG and MTG mice explored the light compartment longer compared to their NTG littermates. However, in contrast to FTG and MNTG mice, FNTG and MTG mice showed a significantly increased activity in the light area. This may indicate that the increased activity in MTG compensates for the increased time spent in the open compartment. It also confirms a reduced anxiousness in FTG mice, because they resided longer in the light compartment, despite decreases in their activity.
Since the constitutive expression of a protein with aberrant function in an animal model may provide insight into the biological functions of this protein we constructed transgenic mice that expressed constitutive active MK5, MK5 L337A . MK5 L337A mice were fertile and displayed no obvious physical or morphological differences compared to control littermates. Initial observations from the SHIRPA analysis suggested differences in anxiety-related behaviour. Subsequently, these mice were subjected to two validated anxiety tests: the elevated plus maze and the light-dark box test.
Difference between males and females
This study revealed more differences between NTG and TG mice when we subdivided both groups per gender. Our data imply that while there exist small differences between NTG and TG mice, these do not necessarily represent the situation per gender in NTG and TG groups. Additionally, we observed opposite reactions between males and females and between NTG and TG mice for several parameters. For example, FNTG mice are more active in the light compartment of the LD compared to FTG mice, whereas MTG mice also displayed increased activity compared to MNTG mice (see Tables 2 and 3).
Since the endogenous MK5 gene resides on murine chromosome 5 and the transgenic MK5 L337A gene is inherited in an autosomal manner (our unpublished results), it is unlikely that we can attribute the observed gender differences to a sex-linked transgene. However, it remains possible that MK5 targets a substrate involved in regulatory processes that differ between genders. For example, our microarray data suggest that upon increased expression of MK5 L337A DAZAP2 becomes downregulated. The function of DAZ-associated protein2 (DAZAP2) remains unelucidated but studies on its family member DAZAP1 indicate that the gene is expressed during sperm maturation, where it shuttles between the nucleus and the cytoplasm during the different stages [21–23]. In addition, DAZ itself is involved in the repression of genes that induce female development and is exclusively found on the Y-chromosome and hence essential for male development. This may partially explain why MTG mice resemble FNTG in behaviour in the closed arm of the EPM, but it does not explain why MNTG mice resemble FTG under these circumstances.
In addition, although MTG mice seem less anxious in the LD, the EPM data do not support such a behaviour. The discrepancy in behaviour between the results of the two tests could be attributed to slightly different forms of anxiety tested in the EPM versus the LD. However, a more likely reason for this difference lays in the increased activity of the MTG mice in the LD test. Since this test does not automatically compensate for differences in activity, as the EPM does, an open field test could help determine whether MTG presence in the open compartment is linked to increased activity. On the other hand, we did not observe any differences in the number of entries, or an increased activity on the open arm on the EPM for MTG mice (Figure 5C) .
Anxiety-related behavioural processes
Our results indicate that FTG mice on the EPM display more head dips on the open arm and reside there longer compared to FNTG, MNTG and MTG mice (see Table 2). A similar tendency occurs with the MTG mice in the light compartment of the LD. This behaviour, the increased presence in the light compartment, implies reduced anxiousness [20, 24]. Therefore, these findings could indicate a role for MK5 in reduced anxiety-related processes that may affect females more than males.
Anxiety-related differences on the EPM and in the LD have been described in other mouse models, particularly in relation to serotonin (5HT) signalling and availability. A recent report by Drapier et al revealed that inhibition of 5-HT and noradrenaline by pharmacological compounds resulted in decreased time spent in the open arm of the EPM or decreased exploration in the open arm, only when 5-HT reuptake was inhibited . These results confirm earlier studies with pharmacological compounds against the SERT or KO models of the gene for SERT, which all augmented anxious behaviour, regardless of postnatal stress [13, 26–28].
Anxiety-related processes in relation to gender have also been described in a double KO mouse model of the serotonin transporter (SERT) and the brain-derived growth factor receptor (BDNFR), referred to as sb mice. Male sb mice show increased anxious behaviour, indicated by less time spent in and less frequent entry into the open arm, while the mutation did not affect anxiousness in female sb mice . Although our results indicated no equal division in gender responses on the EPM. Another explanation for our results may lie in gender differences in thresholds. For example, in 5-HT1A KO mice, male KO mice displayed an increased impulsiveness, as measured by a shorter centre time and shorter distances traveled within the centre. Female KO mice remained unaffected and the researchers attributed these effects to differences in threshold levels between the genders . Similarly, expression of constitutive active MK5 may lower the threshold for female mice in exploration of the maze, whereas males may be more resistant.
In a KO mouse model of SERT alone, administration of 5HT1A receptor antagonists produced a gender-independent anxiolytic effect on the EPM thereby indicating subtle but persistent perturbations in 5HT homeostasis . Since we also observe rather small but significant and persistent perturbations, this may imply a role for MK5 in 5HT regulatory mechanisms. Even more, the upstream activator for MK5, p38, becomes activated by stress stimuli and stimuli that activate the A3 adenosine receptor in a PKG dependent and independent way. This p38 activation results in increased activity of SERT and thereby causes increased reuptake of 5HT [30, 31]. If MK5 intervenes in this signalling cascade it could do so by phosphorylating cytoskeletal proteins involved in transport of the SERT from vesicles to the membrane, or proteins associated with the SERT, thereby affecting their affinity for the transporter. Further experiments are necessary to confirm a role for MK5 in anxiety-related processes and if such a role exists, MK5 could become an interesting target for drug exploration.
As to date, there exist few reports on the involvement of MAPKs in anxiety, and none on the involvement of MAPKAPKs. However MAPKs do play a role in fear-related processes [32–34]. It would be interesting to explore further whether our observations comprise a fear-related component or not.
In contrast to the data from the LD and the closed arm of the EPM, we did not observe an increased activity of FNTG and MTG mice on the open arms of the maze (Figure 5B and 5D and Table 2). One might argue that the lower activity due to the lower presence for FTG in the closed arm of the maze compensates for the increased presence and activity in the open arm. However, FTG are equally active as FNTG in the open arm of the maze (Figure 5A). Differences in locomotion can be often attributed to alterations in dopaminergic systems. This is illustrated by the KO model for tyrosine hydroxylase (TH), the enzyme catalyzing the rate-limiting step during the synthesis of dopamine, where mice exhibit deficiencies in locomotor activity . MK5 has been shown to phosphorylate TH in vitro [36–38]. However, since no in vivo role has been documented, we require further exploration of this pathway to be able to couple MK5 to the regulation of dopamine synthesis and locomotion.
Other mouse models that differ in locomotor activity comprise mice that contain a deficient G-protein coupled inwardly rectifying potassium channel (GIRK) or an ATP-dependent K+channel ( ). In the KO model, the mice resided less in the centre and closed arm of the EPM, and longer in the open arm. In addition, they showed increased activity in the home cage but reduced activity in novel environments [39, 40]. For GIRK deficiency, the mice also displayed decreased anxiety, but in addition, these mice displayed increased locomotion . These studies suggest the possibility to alter both anxiety-related processes and locomotor activity at the same time. This is interesting for our results since we also observed differences in activity and anxiety. Even more, our microarray analysis reported that expression of MK5 L337A altered the expression of solute carriers and channels. It will be interesting to explore these options for MK5 as a regulator in neurological processes.
In this study we observed gender differences in the behaviour of transgenic mice expressing a constitutive active form of MK5. We found that female transgenic mice (FTG) displayed less anxious behaviour on the elevated plus maze and in the light-dark test compared to their littermate controls (FNTG). Male transgenic mice displayed similar anxiety profiles to male non-transgenic mice, but displayed increased activity in the closed arm of the maze. A behaviour that contrasted to FTG versus FNTG mice. In addition, our image analysis method also allowed us to detect that FTG explored further into the open arm on the elevated plus maze. Since both anxiety-related processes and locomotor activity can be regulated by differences in neurotransmitter metabolism, this may indicate for the first time a role for a MAPKAPK in such a process.
We would like to thank Jill Andersen and Hans van den Bout at the Norwegian Transgenic Centre in Oslo for the microinjection and Marijke van Ghelue at Medical Genetics at the University of Tromsø for advise. The histological slides were prepared and analysed by Lars Uhlin Hansen at the University Hospital of Tromsø. Finally, we acknowledge the routine maintenance of the mice by the staff of the animal facility at Tromsø University. This work was supported by grants of the Norwegian Research Council (641220 S5228) and the Norwegian Cancer Society (project A01037/004). NG is a recipient of the grant from the Norwegian Research Council.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.