| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Department of Veterinary and Biomedical Science,
Department of Animal Sciences,
Center for Biotechnology, University of Nebraska-Lincoln, Lincoln, Nebraska, USA
1Correspondence: Department of Veterinary and Biomedical Science, University of Nebraska-Lincoln, Beadle Center, Rm. E117, 1901 Vine St., Lincoln, NE 68588-0665, USA. E-mail: yzhou2{at}unl.edu
ABSTRACT
A polymorphic glucocorticoid receptor (GRQn) with an expanded CAG track and two silent mutations, when compared with the sequence of other isoform (GRwt), is found in two outbred mouse lines that were produced by selection for high (SH) or low (SL) stress response from high or low heat loss lines of mice, respectively. The GRQn allele, which is also found in 5 of 16 commonly used inbred mouse lines, had a much higher frequency in SL mice; the GRwt/wt was found only in the SH line. Both GRQn/Qn and GRwt/Qn mice had a much weaker corticosterone response to stress than the GRwt/wt mice. Assessment of open field activity revealed that GRQn/Qn and GRwt/Qn mice exhibited significantly lower velocity and locomotor activity, less time in the center, and much longer duration in corner zones than the GRwt/wt mice. The increased anxiety-type behaviors of the GRQn/Qn and GRwt/Qn mice were confirmed by the "elevated plus maze" test in which GRQn/Qn and GRwt/Qn mice spent significantly less time in the "open arm" and longer duration in the "closed arm," than GRwt/wt mice. These results suggest this GR polymorphism plays a role in complex mechanisms leading to lower corticosterone response to stress, and may also be associated with decreased locomotive and increased anxiety-type behaviors in mice.—Xu, D., Buehner, A., Xu, J., Lambert, T., Nekl, C., Nielsen, M. K., Zhou, Y. A polymorphic glucocorticoid receptor in a mouse population may explain inherited altered stress response and increased anxiety-type behaviors.
Key Words: corticosterone response • polyglutamine track • mouse model • open field assessment • elevated plus maze
MOLECULAR AND BIOCHEMICAL events associated with stress-induced neurological and behavioral abnormalities remain far from clear due to the complexity of genetic and environmental effects and interactions. Many studies used transgenic mice to target specific genes to determine their roles in physiological and behavioral alterations in response to stress. Alternative approaches to study stress-induced changes in relation to different genetic backgrounds have focused on the generation of animal models through long-term selection for specific stress-induced behaviors or on a systematic mutagenesis screening of mouse lines generated with point mutations of single genes. Such animal models with parallel lines from artificial selection within the same strains, in combination with microarray and quantitative trait loci (QTL) analysis, have become a useful tool for identification of genes associated with specific phenotypic characteristics, including stress susceptibility and anxiety-like behaviors (1
2
3
4)
.
One of the major stress-responsive/regulatory genes studied extensively for decades is glucocorticoid receptor (GR; also known as the NR3C1 gene), a key component in the hypothalamic-pituitary-adrenocortical axis (HPA) -associated central regulation of stress response and in mediation with a wide range of effects induced by glucocorticoid on cellular physiology in most systems of the body. The HPA, glucocorticoids, and stress are closely related factors influencing metabolism, energy balance, and mobility (5
, 6)
. The GR is also involved in modulation of aggregation and nuclear localization of polyglutamine protein in addition to its well-known role as a transcription regulator (7
, 8)
. Down-regulation of the GR in transgenic mice resulted in cognitive defects, elevated serum corticosterone concentrations, and altered open field behavior in response to stress in mice (9
, 10)
. It has also been shown that inhibition or blockade of GR mRNA expression prolongs responses to a novel stressor and may interfere with cognitive aspects of fear and anxiety, suggesting GR-mediated effects on information processing and behavioral adaptation. In humans, GR polymorphisms have a direct impact on neuroendocrine response to stress and may contribute to individual vulnerability for disorders associated with functional disruption of the HPA (11)
. These findings indicate that stress effects can vary significantly among individuals (animals or humans) with genetic differences associated with polymorphic GR isoforms.
We recently carried out divergent genetic selection for altered corticosterone response to stress (high=SH and low=SL), starting from separate base populations of two mouse lines already characterized by genetic difference in heat loss, locomotor activity, and expression of hypothalamic genes (12
13
14
15)
. The initial selection process produced two mouse lines with genetic differences in activity and in levels of serum corticosterone response to thermal environment stress (16)
and to restraint stress (unpublished data). Further selection for stress response has expanded differences in corticosterone response to stress. We started to investigate expression of the major stress response regulatory genes associated with the HPA in the SH and SL mice and identified a much higher frequency in the SL line of a polymorphic form of GR with an expanded track of CAG or polyglutamines (polyQ) when compared with the SH mice. To date, little is known about this GR polymorphism and its phenotypic characteristics in mice. Here we report molecular, biochemical, and phenotypic behavioral characterizations of the polymorphic form of GR in mice.
MATERIALS AND METHODS
Mouse lines and genetic selection
All animal tests were carried out following a protocol approved by the Institutional Animal Care and Use Committee of the University of Nebraska-Lincoln and under NIH guidelines of the "Principles of Laboratory Animal Care" (Publications No. 80–23; revised in 1996).
The initial selection for high and low (SH and SL, respectively) corticosterone response to stress started from two existing selected mouse lines by sampling 25–30 males and 
15 females from each line. The existing selection lines of mice arose from a 4-way composite of four outbred stocks [CF-1, ICR, NIH, and CFW (Sw)] base population developed at the University of Nebraska followed by several generations of divergent selection for heat loss (13
, 14)
. Line SH was derived from the high heat-loss line, and line SL was derived from the low heat loss selection lines. In the SH and SL selections, we routinely kept 12 litters up to weaning and sampled up to 8 different litters to obtain enough numbers of males and females in each line to initiate the next selection cycle. Male mice were used for this study and were singly housed at age 16 wk (they were previously housed 3–4 mice per cage). Mice in the same line were divided into two groups: control (8–10 mice per line) and stress (15–18 mice per line). All selections of males for breeding were carried out based on the following criteria: behavior, body weight, and, most important, serum corticosterone concentration (the highest or lowest in the same line, SH or SL selection, respectively) in response to restraint stress. Females were drawn randomly representing all litters within a closed SH or SL line. Matings were then assigned within line to minimize inbreeding.
The fourth generation (G4) of mice resulted from the above selection and were divided into three genotypic groups based on differences in the GR alleles as determined by GR genotyping: 1) wild-type (WT) form (GRwt/wt) from the SH line, 2) polymorphic form (GRQn/Qn) of the SL line, and 3) heterozygous form (GRwt/Qn) from the SL line. Mice of each genotype were divided into two groups for further phenotypic measurements: control and restraint (single restraint).
We purchased DNA samples of 16 inbred lines from the Jackson Laboratory (Bar Harbor, ME, USA) to screen this polymorphic GR form among commonly used inbred strains.
Stress and sample collection
A 30 min restraint with a plastic restraining tube designed for 30–70 g body wt (31.75 mm inner diameter, 50–100 mm in length; Harvard Apparatus Inc., Cambridge, MA, USA) was used as the "stressor," and this was applied to all stress groups of mice during a comparable time of each day of stress application.
For genetic selection, all mice in the stress group were exposed to three sessions of restraints with a 4-day interval, and each session consisted of three consecutive days of restraint stress (30 min/day). Seven time points of blood sampling by a tail-tip bleeding method (40 µl blood collection within 2 min from each mouse at each time point) were used in all genetic selection cycles. This included immediately before and after the initial restraint (time 0, 30, and 50 min.), before the second session, and before and after the last restraint of the last session of stress (time 0, 30, and 50 min). Control blood samples were collected from both lines of mice at a comparable time of day. Data from corticosterone response to the initial restraint were used as the most important criteria in selecting mice for the next generation.
For characterization of mouse lines with different GR forms, two time points of tail bleed (30 and 50 min after start of restraint) were applied to mice of three GR genotypes. We found in our measurements, as also reported by others (17)
, that serum corticosterone peaked in mice between 20 and 50 min after initiation of stress. Control blood samples were collected from mice without exposure to restraint.
For preparation of total RNAs or proteins, whole brains were dissected from mice decapitated immediately after either the 30 min restraint or those taken from their cages (control), snap-frozen in liquid nitrogen, and stored at –80°C.
Reverse transcription, polymerase chain reaction (PCR) amplification, cloning, and sequence analysis
Total brain RNA from each mouse was obtained by using TriZol Reagent (Invitrogen, Carlsbad, CA, USA) according to the supplier’s protocol. Genomic DNA was obtained from 0.4 cm tail snips homogenized in liquid nitrogen with pestles and mortars, followed by a conventional extraction method. All the G4 SH and SL mice were genotyped (78 SH mice and 68 SL mice) by PCR using the following GR primers: forward 5'-CTG CTTCTCAGGCAGATTCC-3' and reverse 5'-TCC AGAAGCCGAAAGTCTGT-3'. This pair of GR primers was also used to screen GR types using DNA samples of common inbred strains of mice (purchased from Jackson Laboratory).
For reverse transcription (RT) -polymerase chain reaction amplification, 5 µg of total RNA was treated with DNase I (Ambion, Austin, TX, USA) and reverse transcribed with an oligo dT primer and Superscript RT III (Invitrogen). Synthesized cDNA (1 µl, equal to 250 ng total RNA) was then used as a template in standard PCR reactions and PCR products were run on a 2% agarose gel. An additional six pairs of primers were designed based on the published GR mRNA sequence to obtain the full-length cDNA (primer sequences not shown). A forward primer (5'-GACTTTGGTTTGGGAGTTACCTAAAG-3') was signed within the intron region upstream of Exon 2 for PCR using genomic DNA to confirm the first initiation site (ATG) of both GR forms. Products obtained by genomic and RT-polymerase chain reaction (RT-PCR) from five mice from each GR genotype were cloned and sequenced.
Corticosterone RIA
Serum corticosterone concentrations were determined by RIA and assayed in duplicate using a Corticosterone RIA Kit (MP Biomedicals, Orangeburg, NY, USA) following the protocol provided by the company. Intra-assay and interassay variability (coefficient of variation) was determined by using duplicates of control samples, with known corticosterone concentration provided with the assay kit and several previously used samples.
Immunochemical analysis of GR protein isoforms
Brain homogenates from GRwt/wt or GRQn/Qn mice were used for immunoprecipitation of GR proteins using a rabbit antibody (Ab) to GR (GR-M20, Santa Cruz). Detection and visualization of the immunoprecipitated GR proteins were carried out by means of the SDS-PAGE and Western blot method using a Bio-Rad Minielectrophoresis and Minitransfer system (Bio-Rad Laboratories, Hercules, CA, USA). The protein materials (IgG and GR protein; 15 µl) from immunoprecipitation were fractionated using electrophoresis on a 10% acrylamide SDS gel and transferred to PVDF membranes. After 1 h blocking with 3% BSA in Tris-buffered saline (TBS) containing 0.05% Tween-20 (TBS-T), membranes were incubated in TBS-T containing 1% BSA and rabbit anti-GR antibodies (GR-M20; 1:1000 dilution) for 1 h at room temperature. After three washes in TBS-T, the blots were incubated for 1 h in TBS-T containing horseradish-conjugated polyclonal anti-rabbit IgGs (1:1000 dilution; Jackson ImmunoResearch Lab, West Grove, PA, USA). The immunoprecipitated products were visualized using SuperSignal chemilluminescent detecting kit (Pierce, Rockford, IL, USA).
Behavioral assessments
"Open field" activity
Two groups from each GR genotype of mice (at least five mice per group) were used for the open field tests: stress (30 min restraint immediately before exposure to open field) and control, based on our published procedure (18)
. Each mouse was placed in a corner of an "open field" (a lighted area, 0.91 m x 0.91 m surrounded with 46 cm high wall made of dark plastic glass), allowed to roam the field for 5 min, and tested on two consecutive days. Activities of each mouse within 5 min in the open field were automatically recorded by an EthoVision image motion system (Noldus Information Technology Inc., Leesburg, VA, USA). The field was divided by software programming into nine squares containing one center and four corner zones. All data of given parameters, such as total distance traveled, velocity, time in the corner zones, and frequency/time in the center zone, were calculated by the same EthoVision program.
"Elevated plus maze"
Mice of each GR genotype were divided into two groups (control and restraint for 30 min), and each was tested for 5 min with an "elevated plus maze". The plus-shaped maze (made with black plastic glass) is elevated 30.5 cm above the floor and has two open arms and two "closed" arms, with 25.4 cm tall glass sides. Each arm is 30.5 cm long and 5 cm wide with a 5 x 5 cm "center zone." Each mouse was placed in the center zone facing the same open arm, and their behaviors were recorded and analyzed using the same EthoVision system described above.
Statistical analysis
All data were analyzed using a SAS mixed procedure with ANOVA and orthogonal contrasts (SAS Institute, Inc., Cary, NC, USA). Models varied for different phenotypic measures. Fixed effects were line (SH vs. SL), GR genotype within line, stress treatment (restraint or control), and day (repeated measurement for mice in the open field collections). Random effects included mouse within-line GR genotype and residual for the repeated measurements. Orthogonal contrasts were calculated, and these were used to identify the type of gene action of the GR genotypes: 1) GRwt/Qn vs. the average of GRwt/wt and GRQn/Qn to test for the presence of dominance, and 2) GRwt/wt vs. GRQn/Qn to test for the presence of an additive effect. Results are presented as least-squares means ± SE, and differences between the genotypes with a probability value (P) of <0.05 were considered statistically significant.
RESULTS
The SH and SL mouse lines have marked genetic difference in corticosterone response to stress
The SH and SL lines of mice exhibited significant differences in serum corticosterone levels in response to restraint stress. As shown in Fig. 1
, the third generation (G3) of SH mice exhibited a much greater corticosterone response to the initial restraint stress than the SL line (overall comparison between the two lines, F value=68.82; P value<0.0001). Such a difference was also observed between the two lines after exposure to repeated restraints (data not shown). During the selection phase, we chose five mice with the highest (within the SH line) and lowest (among the SL mice) levels of serum corticosterone as sires to produce the next generation.
|
A polymorphic form of GR (GRQn) with an expanded track of polyglutamines has a much higher frequency in the SL line of mice
We began to analyze the expression of major stress-response regulatory genes in the brain tissues obtained from G3 mice, including GR, to investigate whether alteration of the HPA regulation is associated with the differences observed in corticosterone responses to stress. Surprisingly, we observed differences in the size of the RT-PCR products (Fig 2
A) in different mice using specific primers for the GR coding sequences within Exon 2. The PCR products were cloned, and sequence analysis revealed a polymorphic form of GR with an expanded track of 16 CAG repeats (Fig 2B
), which yields a polyQ track (GRQn; GenBank Accession No., DQ504162) when compared with the WT form (the mouse NR3C1 gene; Accession No., NM_008173) with 8 glutamines (GRwt). Using additional primers based on the published sequences, we cloned full-length cDNA of both forms of GR and sequenced several clones obtained from independent RT-PCR using different mouse brain RNA. Sequence data indicated that both forms of GR have the same initiation site. However, the GRQn has two silent mutations in contrast to GRwt: position 222, A 
G and position, 468, T C (444 for the WT GR), as shown in Fig 2B
. The two GR isoforms were further confirmed as to protein level by immunoprecipitation, followed by Western blot analysis using GR-specific antibodies, which revealed the GR isoforms with different molecular mass as shown in Fig 2C
.
|
Because we produced the G4 SH and SL mice before discovering the existence of the GR polymorphism among the SH and SL mice, we screened all G4 mice by genomic PCR using the same primers covering the coding region containing the CAG repeats of the GR gene. Analysis of the PCR products of the GR isoforms to identify GR genotypes are summarized in Table 1
, with percentages of the different genotypes in the SH and SL lines. Frequency of the GRQn form was much greater in SL line (GRQn frequency=0.84, GRwt frequency=0.16), whereas frequency of the GRwt form was much higher in SH mice (GRwt frequency=0.91, GRQn frequency=0.09).
|
The distribution pattern of the two forms of GR suggests an association of GRQn with low corticosterone response to stress. To confirm this hypothesis, we measured serum corticosterone in response to a 30 min restraint among three GR genotypes of G4 male mice: GRwt/wt (from SH line), GRQn/Qn (from SL line) and GRwt/Qn (from SL line). While no difference in levels of serum corticosterone concentration was seen in control groups of the three GR genotypes, as shown in Fig. 3
, GRwt/wt mice showed significantly greater corticosterone response at both given time points than the GRQn/Qn and GRwt/Qn mice exposed to restraint stress (F=70.43, P<0.0001). However, the latter two GR genotypes had similar corticosterone levels in response to stress. This indicates that the GRQn is probably involved, if not directly, in altered corticosterone response to restraint stress.
|
The GRwt/Qn and GRQn/Qn mice exhibit hypoactivity and anxiety-type behavior
We used the same G4 control mice and those restrained only once used for the corticosterone assays described above to assess their open field activities. Both the GRQn/Qn and GRwt/Qn mice had a similar velocity, which was lower than the GRwt/wt mice (overall difference: F=13.4, P=0.001; Fig. 4
A). Under control conditions, GRwt/Qn mice had the lowest velocity on both test days. The 30 min restraint caused a decrease in velocity in both the GRQn/Qn and GRwt/Qn mice that was much lower than that of GRwt/wt mice (F=7.97, P=0.008). The GRQn/Qn and GRwt/Qn both traveled less (overall difference: F=14.21, P=0.0007) than the WT mice, especially the ones exposed to the restraint stress, which had the shortest distance traveled among all the groups in the day 1 open field test (Fig. 4B
). Again, the GRQn had a dominant effect on behavioral response to stress (F=4.97, P=0.033). The differences in locomotor activity can be clearly distinguished by comparing traveling patterns in the open fields (representative images are shown in Fig. 4C
.
|
The differences in distance traveled between either the GRQn/Qn or GRwt/Qn mice and the GRwt/wt were due not only to their lower locomotor activity, but also to the time they stayed in the corner zones (all four corners included) without much movement, as compared with the WT (overall difference: F=17.08, P=0.0003), especially those exposed to restraint (60–80% of time remaining in corner zone, Fig. 5
A). We noted that most of the GRwt/Qn stayed in the starting corner or moved out of the starting corner for only a few seconds (Fig. 5B
; like the one shown in the lower middle panel in Fig. 4C
); with only one exception, they never traveled to the center zone (Fig. 5C
). This indicates that the restraint stress resulted in a prolonged period of stay in the starting corner for mice of GRwt/Qn and GRQn/Qn genotypes, but not the WT mice. In addition, the GRQn/Qn and GRwt/Qn mice had significantly lower center activities (represented as the frequency and time in the center zone) than those of the GRwt/wt mice, but these increased when preexposed to the 30 min restraint (Fig. 5C, D
). No difference was seen between GRQn/Qn and GRwt/Qn genotypes of the restraint group. Furthermore, it appeared that only the GRwt/Qn had a stronger response to novelty stress in terms of time in the start corner between day 1 and day 2 (Fig. 5B
), and had the lowest (or zero) frequency of visiting or time spent in the center zone (Fig 5C, D
).
|
The lower locomotor activity and increased anxiety-type behaviors in mice with the polymorphic GR were further confirmed by the elevated-plus-maze assessment. The GRQn/Qn and GRwt/Qn mice spent significantly less time than the GRwt/wt mice in the open arms (Fig. 6
A; F=52.15, P<0.0001), and much longer in the closed arms (Fig. 6B
; F=6.7, P=0.016). In addition, the mice with WT GR showed much higher locomotor activity, traveling more than the GRQn/Qn and GRwt/Qn mice (Fig. 6C
; F=75.83, P<0.001). A stronger response to restraint stress was observed in the GRwt/Qn mice, which appeared to have lower mobility and spent more time in the center zone (data not shown), resulting in a nonstatistically significant decrease in time in closed or open arms, but much less distance moving (F=6.53, P<0.014). Taken together with data from the open field test (time in corner zones, the much lower activity of both the GRQn/Qn and GRwt/Qn mice in the center zones), this suggests a possible linkage of this polymorphic form of GR with the hypoactivity and anxiety-like behaviors.
|
Identification of the polymorphic form of GR in commonly used inbred lines of mice
Using the same GR primers covering the track of CAG repeats, we screened 16 inbred mouse strains by genomic PCR. Five of the examined inbred lines of mice have the polymorphic form of GR (ST/bJ, SWR/J, C57BL/6J, C57BR/cdJ, and C57L/J), whereas the other 11 strains have the WT form (129x1/SvJ, A/J, AKR/J, BALB/cJ, C3H/HeJ, C58/J, I/LnJ, NOR/Lt, RF/J, PERA/Rk, RIIIS/J). Among the five strains of mice with the GRQn, ST/bJ line has a different origin from the other four strains. The C57BL/6J, C57BR/cdJ, and C57L/J share the same origin with a genetic linkage to Asian Mus musculus, including the SWR strain (for detailed description of cross origin, see http://www.informatics.jax.org/external/festing/mouse/docs/C57BL.shtml).
DISCUSSION
The SH and SL lines of mice were produced by selection for high and low stress responses (mainly based on the serum corticosterone levels) from a baseline population of two mouse lines with genetic differences in heat loss (13)
, and maintained the characteristics of their parent lines, such as greater locomotor activity in the SH line with high heat loss (12)
. However, the SH and the SL lines showed genetic difference in corticosterone response to restraint stress. It has been shown that alteration of levels of corticosterone or corticotrophin-releasing hormone expression plays an important role in regulation of metabolism and locomotor activity (17)
. The difference in neuroendocrine activity between the SH and SL lines of mice observed in this study may also be a key contributor to the genetic variation in body fatness and locomotor activity between the two lines. The two lines of mice generated through selection for high and low corticosterone response to stress resulted unintentionally in an apparent separation of two lines with different allelic combinations of two forms of GR: the wild form (GRwt) and a polymorphic form with an expanded track of CAG repeats/polyQ (GRQn). It is known that GR polymorphisms or down/up-regulation of GR expression can have marked effects on neuroendocrine activities, metabolism, and physiological and psychological behaviors (for recent reviews, see refs. 5
, 8
). The high frequency of the polymorphic GR in the SL line of mice observed in this study suggests an association of the GRQn, possibly in combination with effects of other genes, with the weaker corticosterone responses to restraint stress. The similarity in both neuroendocrine activity and assessed behaviors between the GRQn/Qn and GRwt/Qn genotypes indicates that the GRQn has mostly a dominant effect on corticosterone and behavioral responses to stress in heterozygous mice. An important question to be answered is whether those differences between mice with different GR forms may also be attributed to other genes associated with genetic selection of the SH and SL lines or original lines with genetic variation of heat loss. We are currently generating all three GR genotypes within both SH and SL lines to measure the expression of the HPA-axis-associated genes (such as corticotrophin-releasing factor, mineralocorticoid receptor, and ACTH) and behavioral differences in response to stress.
Molecular and biochemical events for actions of this polymorphic form of GR are unknown at this time and remain to be determined by in vitro and in vivo studies. It has been shown in vitro that rat GR polymorphic forms with different lengths of CAG repeats have similar ligand binding affinity to dexamethasone and corticosterone (19)
. However, when the length of CAG repeats are altered, the mutant form of GR exhibited significantly higher steroid binding affinity (20)
, suggesting a possible domain-domain interaction between the polyQ and the ligand binding regions within the GR. In addition, the GR serves not only as the well-known transcription regulator, but also as a modulator in mechanisms of aggregation and nuclear localization of other polyQ proteins (7)
. In our case, the RIA data for serum corticosterone levels suggest that the GRQn, which has 16 CAG repeats and yields an inframe 16 glutamine track, may have greater transactivation efficacy than the GRwt form. The GRQn proteins may have a faster translocation rate to the nuclear region than the GRwt form, as we observed (data not shown) in different brain regions of several brain samples from mice without exposure to restraint. Although in vitro experiments have shown a rapid formation of polyQ aggregates detectable within 1–4 min (21
, 22)
, the kinetic behavior of polyQ-containing proteins in vivo is unknown. It remains to be investigated in vitro whether the polyQ expansion in combination with silent mutations alters the properties of the GR itself in steroid binding, transactivation, translocation, or aggregation in addition to its function in mediating other polyQ-containing proteins (6
, 7)
. Our initial microarray data showed an 3-fold lower expression of HSP70 and a HSP-associated chaperone (Stch) in the brain of control GRQn/Qn mice than that of GRwt/wt (data not shown), suggesting that the GRQn protein itself or with other proteins may form nuclear aggregation due to a decreased degree of prevention of polyQ protein aggregation by these molecular chaperones as reported by others (23
24
25
26)
.
This GR polymorphism exhibited some effects on behavioral response to stress, as supported by results from open field tests that mice with the homozygous form of GRwt alleles exhibited greater locomotor activity (velocity and total distance traveled) than either GRQn/Qn or GRwt/Qn mice. Such differences in response to restraint between the GR genotypes were also observed using the elevated plus maze: a reduced activity for mice with the polymorphic GR but increased time spent by the WT mice in the open arm. In addition, mice with the polymorphic form of GR displayed a marked increase in anxiety-like behavior, much longer time in corner zones (especially the starting corner), and significantly lower or near-zero frequency of visits to and shorter time staying in the "bright-lit" center than the GRwt/wt mice (see Fig. 5
for details). Surprisingly, the control GRwt/Qn mice showed a much greater degree of anxiety-like behavior, with lowest frequency and time in the center zone on the first day. None entered the center zone during the second day of test, relative to both the GRQn/Qn and GRwt/wt mice. This suggests that the coexistence of both forms of GR may result in different physiological and biochemical effects on neuroendocrine regulation in response to restraint and/or novelty stress, which inserted a counter effect on anxiety-like behavior, with decreased "fear" factor of entering the center zone.
It has been suggested that GR is also a key modulator for emotional liability, including anxiety and cognitive aspects of fear (27)
. Disruption of GR function associated with the HPA axis due to prolonged elevation of glucocorticoids may result in impaired cognition and anxiety/depression related disorders (28)
. Both enhanced and reduced expression of GR or GR knockout in mice resulted in significant alteration of neuroendocrine activity and behavioral changes (27
, 29)
. However, corticosterone elevation may not always yield the same physiological effects because its receptors, such as GR, can act differently through either transactivation (DNA binding) or transrepression (interaction with other transcription factors), or both (30
31
32)
.
Genetic screening for the polymorphic form of GR in commonly used mouse strains led to the identification of five inbred lines with the GR allele of the expanded track of CAG repeat/polyglutamines. Four of the five identified strains have a genetic linkage to Asian mouse origin related to the SWR mice. Similar to what we observed in the present study, the SWR mice also reported having the phenotypic characteristics of hypoactivity and high avoidance behaviors (33
, 34)
, whereas the ST/bJ mice is know to have low metabolic rate (35)
similar to the parental line of the SL mice we developed. However, studies have shown that mice of the C57 family have high open field activity (locomotor and exploratory) compared with other inbred strains, including those with the GRwt (A, AKR, BALB, C3H, RF) or with the GRQn (ST and SWR) (33)
. This suggests that the polymorphic form is not a direct or dominant factor in regulating locomotor activity and exploratory/anxiety-type behaviors. It is still unclear whether the different GR forms have different efficacies in the central regulation of neuroendocrine response to stress. In our case, the relationship between low corticosterone and altered behavior in mice with one or both alleles of GRQn deserves additional investigation, since it is likely a total genomic effect involving a set of stress-responsive genes. The increased anxiety-like behaviors in the GRwt/Qn mice observed in the present study may also be due to the coexistence of the two forms of GR, which may have different rates of translocation and/or transactivation, thus post-transcription regulation in response to stress. We are currently investigating differential gene expression associated with the three GR genotypes, with or without exposure to restraint stress. These mouse lines with genetic variations in GR allelic combination accompanied by marked differences in phenotypic characteristics will provide a useful model for better understanding the central regulation of neuroendocrine response to stress and molecular events behind the anxiety-like behavior in relation to this GR polymorphism.
ACKNOWLEDGMENTS
This report is a contribution of the University of Nebraska Agricultural Research Division, Lincoln, NE (Journal Series No. 14652). This research was supported in part by funds provided through the Hatch Act, the Charles J. Millard Trust Fund, the Layman Award, and the UCARE Award of the University of Nebraska-Lincoln. We thank Ben Miller and Laura Meyerle for their research assistance.
Received for publication February 24, 2006. Accepted for publication June 2, 2006.
REFERENCES
Related Articles
This article has been cited by other articles:
|
|
 |
|
  P. Mormede, M.-P. Moisan, and W. E. Crusio Does a polymorphic glucocorticoid receptor explain inherited altered stress response and increased anxiety-type behaviors in a mouse population? FASEB J, January 1, 2008; 22(1): 5 - 6. [Full Text] [PDF]
|
|
|
|
|
|
Y. Zhou, S. Zheng, W. Sprout, J. McDonald, and M. K. Nielsen Additional evidence showing an additive effect of glucocorticoid receptor polymorphisms on anxiety-type behavior, stress response, and body weight in a population of mice with low heat loss background FASEB J, January 1, 2008; 22(1): 6 - 8. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
