|Home | About | Journals | Submit | Contact Us | Français|
Stress hormones are thought to be involved in the etiology of depression, in part, because animal models show they cause morphological damage to the brain, an effect that can be reversed by chronic antidepressant treatment. The current study examined two mouse strains selected for naturalistic variation of tissue regeneration after injury for resistance to the effects of chronic corticosterone (CORT) exposure on cell proliferation and neurotrophin mobilization. The wound healer MRL/MpJ and control C57BL/6J mice were implanted subcutaneously with pellets that released CORT for 7 days. MRL/MpJ mice were resistant to reductions of hippocampal cell proliferation by chronic exposure to CORT when compared to vulnerable C57BL/6J mice. Chronic CORT exposure also reduced protein levels of brain derived neurotrophic factor (BDNF) in the hippocampus of C57BL/6J but not MRL/MpJ mice. CORT pellet exposure increased circulating levels of CORT in the plasma of both strains in a dose dependent manner although MRL/MpJ mice may have larger changes from baseline. The strains did not differ in circulating levels of corticosterone binding globulin (CBG). There were also no strain differences in CORT levels in the hippocampus, nor did CORT exposure alter glucocorticoid receptor or mineralocorticoid receptor expression in a strain dependent manner. Strain differences were found in the NMDA receptor, and BDNF I and IV promoter. Strain and CORT exposure interacted to alter tropomyosine-receptor- kinase B (TrkB) expression and this may be potential mechanism protecting MRL/MpJ mice. In addition, differences in the inflammatory response of matrix metalloproteinases (MMPs) may also contribute to these strain differences in resistance to the deleterious effects of CORT to the brain.
Depression is a debilitating disorder that effects 6.7% of the United States population within a given year (Kessler et al., 2005). Lifetime prevalence across cultures varies from 2-20% (Weissman et al., 1996). Currently 40% of patients experiencing depression do not respond to traditional treatments indicating a need for new treatments based on the underlying biology of vulnerability to depression (Culpepper, 2010). Tests of cortisol stimulation and feedback have been used to identify dysregulation of the hypothalamic-pituitary- adrenal axis (HPA axis) that occurs in half the population suffering from depression and other psychiatric disorders (Gold and Chrousos, 2002, Gillespie and Nemeroff, 2005). In mice, chronic exposure to elevated levels of the stress hormone corticosterone (CORT) has been used as a behavioral model of depression as it alters depression-associated behaviors (Gourley et al., 2008, Murray et al., 2008, David et al., 2009, Gourley et al., 2009).
One way that stress hormones may contribute to the etiology of depression is that they decrease cell proliferation (Cameron et al., 1998), dendritic arborization (Woolley et al., 1990) and the mobilization of neurotrophins in the hippocampus (Duman and Monteggia, 2006). Blocking glucocorticoid receptors can reverse detrimental effects of chronic stress on synaptic structure and function (Krugers et al., 2010) and prevent decreases in neurogenesis induced by exposure to CORT (Mayer et al., 2006). Chronic antidepressant treatment can also reverse decreased neurogenesis in both stress induced rodent models of depression (Malberg and Duman, 2003, Czeh et al., 2007) and in humans with depression (Boldrini et al., 2009) further indicating that decreases in neurogenesis can be used as a biomarker of a depressive like state. Stress hormones are known to decrease neurogenesis via the activation of NMDA receptors which are downstream of the glucocorticoid receptor (Cameron et al., 1998). In addition it is thought that changes in neurotrophin mobilization contribute to alterations in neurogenesis as stress decreases protein levels of a number of neurotrophins in the hippocampus including brain derived neurotrophic factor (BDNF) (Schmidt and Duman, 2007). BDNF depletion in the dentate gyrus has also been shown to attenuate the behavioral effects of antidepressants, although selective depletion of BDNF alone does not induce a depressive like behavioral phenotype (Adachi et al., 2008). These data suggest that neurotrophin mobilization most likely acts in concert with other stress pathways in the development of depression.
It has been proposed that both genetic and environmental factors contribute to the vulnerability to depression in humans (Caspi et al., 2003, Charney and Manji, 2004). As a model of genetic differences in response to stress hormones, we hypothesized that MRL/MpJ and C57BL/6J mice would differ in their resiliency and vulnerability to the deleterious effects of chronic elevations of CORT on cell proliferation and BDNF levels. MRL/MpJ mice exhibit enhanced wound healing and unusual regenerative capacities (Heber-Katz et al., 2004c). In addition, they are behaviorally responsive to antidepressants and show increased cell proliferation, cell survival and neurotrophin levels compared to C57BL/6J mice when exposed to chronic antidepressant treatment (Balu et al., 2009a). The current study compares the effects of chronic exposure to stress hormone CORT on cell proliferation, BDNF mobilization and stress hormone levels in MRL/MpJ and C57BL/6J mice. Then using a candidate gene approach based on published literature we examined a number of potential molecular mechanisms that may contribute to these strain differences in susceptibility to stress hormones.
Adult male C57BL/6J and MRL/MpJ mice (Jackson Laboratories, Bar Harbor, ME, USA) were 7-10 weeks of age at the beginning of all studies. C57BL/6J mice were used as they are a common laboratory strain and are the background strain used in a majority of genetic manipulations. MRL/MpJ mice demonstrate enhanced regenerative responses to tissue injury and have increased neurogenesis following chronic treatment with both fluoxetine and desipramine compared to C57BL/6 mice (Heber-Katz et al., 2004b, Balu et al., 2009a). Mice were group housed (4 to a cage) in polycarbonate cages and maintained on a 12-h light/dark cycle (lights on at 07:00 hours) in a temperature (22°C) and humidity-controlled colony. Animals were given ad libitum access to food and water. All procedures were conducted in accordance with the guidelines published in the NIH Guide for Care and Use of Laboratory Animals and all protocols were approved by the University of Pennsylvania Institutional Animal Care and Use Committee.
Experiment 1 was a dose response study (n=84) performed to examine the effects of CORT exposure on cell proliferation in MRL/MpJ and C57BL/6J mice. Strains were run individually in cohorts of 20-30 animals (n=5-6 per CORT dose) due to limitations in tissue processing and equipment access. Two cohorts were run for each strain. Blood and tissue samples for dose response ELISAs, including BDNF, CORT and corticosterone binding globulin (CBG) (n= 40-44) were taken from a single cohort of animals in each strain (Figure 1a). Experiment 2 examined the effects of a single dose of CORT (2 pellets) on mRNA expression in the hippocampus and hypothalamus (n=30-32). ELISAs for CORT in the hippocampus and blood were run on samples taken from these animals (Figure 1b).
Mice were anesthetized via inhalation of isoflurane and 5 mg, 21 day release placebo and CORT pellets (1, 2 or 4 pellets; Innovative Research of America, Sarasota, FL) were implanted subcutaneously with a trochar. MRL/MpJ mice weighed on average 35± 0.7 g whereas C57BL/6J mice weighed on average 26±3 g. For C57BL/6J mice the daily equivalent doses adjusted for body weight were as follows: 1 pellet= 9 mg/kg/ day, 2 pellets= 18 mg/kg/ day, 4 pellets= 36 mg/kg/day. For MRL/MpJ mice the daily equivalent doses were as follows: 1 pellet= 6 mg/kg/day, 2 pellets = 13 mg/kg/day, 4 pellets= 27 mg/kg/day. Mice were exposed to 7 days of constant CORT secretion prior to sacrifice.
On the seventh day of CORT exposure mice received a single injection of BrdU (200 mg/kg) and were sacrificed 24 hours later (n=84). Labeling of BrdU was measured in cells displaying the nuclear marker 7-aminoactinomycin D (7-AAD) by flow cytometry as previously described and validated (Balu et al., 2009a, Balu et al., 2009b). Mice were decapitated, their brains quickly removed, and the bilateral hippocampus was removed. The right hippocampus was analyzed for cell proliferation, as cell counts did not differ between hemispheres (Balu et al., 2009a). Tissue was placed in Hank's Balanced Salt Solution (HBSS, Gibco Grand Island, NY), and finely minced. Tissue was digested using an enzymatic cocktail (0.5 mL, 1 mg/mL papain, Roche Applied Sciences Indianapolis, IN; 0.1 M L-cysteine, Sigma St. Louis, MO) and incubated in a dry heat block at 37°C for 15 min. Hibernate-A (Brain Bits Springfield, IL) containing 10% heat-inactivated fetal bovine serum (FBS; Gibco Grand Island, NY) was added to stop the enzymatic digestion. Tissue was then mechanically triturated to form a single cell suspension and spun in a centrifuge at 2000 rpm for 5 min.
The supernatant was removed and the resultant cells were stained using the FITC BrdU Flow Kit (BD Biosciences San Jose, CA). Cells were initially fixed and permeabilized by resuspension in 100 μL of Cytofix/Cytoperm buffer. Cells were then washed in staining buffer (PBS, 3% FBS, 0.09% sodium azide), spun at 5000 rpm and aspirated. Cells were stored overnight in staining buffer. On day 2 of staining cells were spun at 5000 rpm, staining buffer was removed and cells permeabilized (10 min on ice), washed and refixed (5 min) and washed. The cells were then resuspended in 100 μL of DNAse (30 μg; stock from kit was diluted in DPBS (Ca2+/Mg2+ free) containing 0.1 mM CaCl2 and 10 mM MgCl2) in a dry heat block at 37°C for one hour to break down DNA into a single strand and allow bonding of the BrdU antibody. Following washing and spinning, the cells were labeled with 50 μL of FITC-conjugated anti-BrdU (1:50 dilution) in the dark at room temperature for 20 min. After the samples were washed, they were labeled with 20 μL of the nuclear marker, 7-AAD, at room temperature in the dark. The cells were then resuspended in staining buffer (PBS, 3% FBS, 0.09% sodium azide). Prior to analysis, cells were filtered through a cell strainer cap (30 μm) to remove debris. The data was collected on a BD FACS Canto system at the University of Pennsylvania Flow Cytometry Core Facility. Background signal was accounted for by staining tissue from animals that had not been injected with BrdU. All data was collected an analyzed using BD FACSDiva software (BD Biosciences, San Jose, CA).
One cohort of each strain was used in the dose response study (Fig. 3a,b,c: BDNF, CORT plasma, CBG/ Experiment 1 n = 40-44) and all animals from the second experiment were used to measure CORT in the hippocampus and blood of animals exposed to placebo vs. 2 CORT pellets. (Table 1: n= 30-32). The left hippocampus was flash frozen in isopentane and placed at –80° C until analysis. The tissue was homogenized in 0.75 mL of lysis buffer (100 mM PIPES pH 7.0, 500 mM NaCl, 2 mM EDTA, 0.1% sodium azide, 2% bovine serum albumin, 0.2% Triton X-100, 5 μg/mL aprotinin, 0.1μg/mL pepstatin A, 0.5 μg/mL antipain). The homogenate was centrifuged at 13,000 rpm for 30 min at 4°C. The supernatant was removed and the amount of BDNF protein or CORT in the supernatant from each sample was measured in duplicate by ELISA following the manufacturer's instructions. BDNF, CORT, CBG protein levels were quantified using commercially available sandwich ELISA kit ; BDNF (Promega, Madison, WI), CORT (ImmunoDiagnostic Systems, Fountain Hills, AZ), and CBG (Kamiya Biomedical Company, Seattle, WA). Intra assay variability for the BDNF kit ranged from 2.2%- 8.8%, mean assay sensitivity was 15.6 pg/mL, Inter-assay variability was not measured by the manufacturer. Intra assay variability for the CORT kit ranged from 3.8%-6.6%, inter assay variability ranged from 7.5%- 8.6%; mean assay sensitivity was 0.55 ng/mL.Inter-assay variability and intra assay variability were not measured by the manufacturer for the CBG ELISA, assay sensitivity was 2.75 ng/mL. BDNF and CORT levels taken from tissue were normalized to the wet tissue weight. Additionally CORT was measured in plasma, normalized to placebo to allow for unbiased analysis of samples run on separate plates and compared as fold changes. Samples analyzed on the same plate are listed as ng/mL (fig 1b, experiment 2).Trunk blood was collected at time of sacrifice in EDTA treated tubes (Sarstedt, Numbrecht, Germany). Blood was centrifuged at 3000 rpm for 20 min and plasma was removed and stored frozen (–80° C) until analysis. Samples were diluted and run in duplicate following the manufactures instructions.
MRL/MpJ (n= 16) and C57BL/6J mice (n=16) were surgically implanted with 2 CORT (5 mg) or placebo pellets (Experiment 2, Fig 1b). After 7 days of pellet implantation animals were sacrificed. The unilateral right hippocampus was dissected out and mRNA was isolated using the RNAqueous -4PCR kit for Isolation of DNA-free RNA (Ambion, Austin, TX) following the manufactures’ instructions. RNA concentrations were measured and 150 ng/ul RNA was used as a template to synthesize c-DNA using the Superscript Vilo c-DNA Synthesis kit (Invitrogen, Carlsbad, CA). All reactions were performed with a master mix of SYBR green (Applied Biosystems, Austin, TX) and 300 nM primers (final concentration). Quantitative real-time polymerase chain reactions (qRT-PCR) were run using the Stratagene MX3000 and MXPro QPCR software. Cycling paramaters were as follows: 95°C for 10 mins, 40 cycles at 95°C (30 s) and 60°C (1 min), ending with a melting curve analysis to control for amplification. All reactions were performed in triplicate and the mean cycle threshold was used for analysis. The mRNA levels of target genes were normalized to the housekeeping gene with the closest cycle threshold using the 2ΔΔCt method. The two house keeping genes used were TATA binding protein (TBP) and Glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Primer sequences are available upon request.
Statistical analysis for cell proliferation data and qRT PCR were performed using 2-way ANOVAs (strain × CORT exposure) with Newman-Keuls used for post hoc analysis. For ELISAs in which strains were run on separate plates at separate times (experiment 1, dose response studies: BDNF and CORT) data were normalized to within strain placebos and calculated as a fold change. In experiment 2, two-way ANOVA compared circulating levels of CORT in the blood and brains of the two strains given placebo or 2 CORT pellets with both strains run on the same plate within the same assay. Statistica software (StatSoft, Tulsa, OK) was used to perform all analysis. Statistical significance was set at p < 0.05.
Exposure to 7 days of the stress hormone CORT reduced cell proliferation in the hippocampus of C57BL/6J mice at all doses but did not significantly reduce cell proliferation in MRL/MpJ mice at any dose as indicated by a significant interaction [F3,74 = 3.92, p < 0.05] (Fig. 2) Post hoc analysis indicated that MRL/MpJ mice treated with CORT at all doses had an average of 3-4-fold more cells than C57BL/6J mice treated with the same dose (p values < 0.001). There was also a significant main effect of CORT exposure [F3,74 = 18.04, p < 0.001 and of strain [F1,74 = 50.81, p < 0.001].
The contra-lateral lobe of the hippocampus from each cohort used for the cell proliferation study was examined for BDNF protein levels. ELISAs were run on separate strains at different times. To control for inter-assay variability BDNF levels were normalized to within strain placebo controls and reported as fold change. Two-way ANOVA indicated a significant interaction between CORT exposure and strain [F3,33 = 3.03, p < 0.05]. Post hoc analysis indicated that BDNF levels significantly differed between MRL/MpJ and C57BL/6J mice when implanted with 2 or 4 pellets (p values < 0.05) (Fig. 3a). C57BL/6J mice exposed to 4 CORT pellets had a significant reduction in BDNF levels compared to placebo pellets (p < 0.05). CORT exposure did not significantly alter BDNF levels in MRL/MpJ mice at any dose (p > 0.05).
CORT levels in the plasma were examined to determine whether there were strain differences in circulating levels of CORT as a result of pellet implantation (Fig. 3b). Samples from experiment 1 (dose response study) were run on separate plates at separate times. To control for inter-assay variability samples were normalized to within strain controls and reported as fold change. Two-way ANOVA indicated a significant main effect of CORT exposure [F3,36 = 109.05, p < 0.001]. Post-hoc analysis indicated that CORT levels were significantly elevated in animals implanted with 2 or 4 pellets compared to mice implanted with placebo or 1 pellet. There was also a significant difference in circulating levels of CORT between animals implanted with 2 or 4 pellets (p < 0.05). There was a significant main effect of strain [F3,20 = 6.25, p < 0.05], with MRL/MpJ mice showing a larger increase in CORT than C57BL/6J mice, but there was no significant interaction between CORT exposure and strain.
In experiment 2, plasma from C57BL/6J and MRL/MpJ mice implanted with 2 pellets or placebo were taken at the same time, and processed and measured on the same plate. Two-way ANOVA indicated that CORT was elevated in both strains [F1,26 = 75.78, p < 0.001] following implantation with 2 pellets but did not differ between strains [F1, 26 = 2.50, p > 0.05] or interact with strain [F1, 26 = 0.87, p > 0.05] (Table 1). Furthermore, hippocampal levels of CORT from the same animals were increased in both strains implanted with 2 pellets [F1,25 = 16.21, p < 0.001] but there was no significant main effect of strain or strain by CORT interaction (Table 1). CORT levels in the hippocampus were not measured following implantation of 1 or 4 pellets.
CBG levels were measured in plasma of MRL/MpJ and C57BL/6J mice exposed to placebo or 1,2 and 4 pellets of CORT (Fig. 3c) as the majority of CORT in the body remains in a bound state and CBG can regulate the availability of CORT to tissue (Breuner and Orchinik, 2002). All samples were analyzed on the same plate at the same time. Two-way ANOVA indicated that there was no significant strain by CORT exposure interaction or main effects of CORT exposure or strain [p > 0.05] on CBG levels in the plasma of mice.
Glucocorticoid (GR) mRNA expression was examined in the hippocampus (Fig. 4a/ Table 2) and hypothalamus (Fig. 4b/ Table 2) as a potential mechanism of strain differences in the effects of CORT on cell proliferation and neurotrophin mobilization. MRL/MpJ mice had significantly higher levels of GR expression in the hippocampus than C57BL/6J mice [F1,26 = 14.71, p < 0.001]. Exposure to 2 pellets of CORT significantly reduced GR expression in the hippocampus of both MRL/MpJ and C57BL/6J mice as demonstrated through a main effect of CORT exposure [F1,26 = 83.41, p < 0.01]. There was no strain by CORT exposure interaction [F1,26 = 0.11, p > 0.05]. Exposure to CORT for seven days significantly reduced GR expression in the hypothalamus of both MRL/MpJ and C57BL/6J mice [F1,27 = 9.14, p < 0.01] but there were no significant differences between strains [F1,27 = 1.76 , p > 0.05] and no strain × CORT exposure interaction [F1,27 = 0.55, p > 0.05].
As CORT can also bind to mineralocorticoid receptors (MR) and down-regulate neurogenesis through this mechanism (Chang et al., 2008), the effects of 7 days of CORT exposure on MR gene expression in the hippocampus (Fig. 4c/Table 2) and hypothalamus (Fig. 4d/ Table 2) were also examined. MRL/MpJ mice had higher expression levels of MR than C57BL/6J mice as indicated by a main effect of strain [F1,26 = 5.72, p < 0.05]. CORT pellet exposure decreased MR gene expression in the hippocampus of MRL/MpJ and C57BL/6J mice as indicated by a main effect of CORT [F1,26 = 10.28, p < 0.01], with no significant strain × treatment interaction[ F1,26 = 2.69, p > 0.05] MR expression in the hypothalamus did not differ between strains [ F1,27 = 0.09, p > 0.05] and was unaltered by CORT pellet exposure[ F1,27 = 0.78, p > 0.05] and there was no strain by CORT exposure interaction [F1,27 = 0.005, p > 0.05].
Reductions of hippocampal neurogenesis by stress are known to involve the NMDA receptor and its activation occurs downstream of the GR (Cameron et al., 1998). Gene expression of the NR1 and subunits for NR2A,NR2B and NR2C (Fig. 5a/ Table 2) were examined in the hippocampus to determine if CORT exposure altered mRNA levels in a strain-dependent manner. MRL/MpJ mice expressed higher NR1 levels of than C57BL/6J mice as indicated by a main effect of strain [F1,28 = 28.70, p < 0.001]. However, there were no effects of CORT exposure [F1,28 = 0.005, p > 0.05] and no strain × CORT exposure interaction [ F1,27 = 0.001, p > 0.05]. There were no strain differences [F1,26 = 1.16, p > 0.05] or main effects of CORT [F1,26 = 0.05, p > 0.05] or interaction [F1,26 = 0.15, p > 0.05] on the NR2A subunit. MRL/MpJ mice had higher expression of the NR2B subunit than C57BL/6J mice as indicated by a main effect of strain [F1,25 = 13.34, p < 0.01]. Although a main effect of CORT exposure indicated that CORT decreased NR2B expression in both strains, [F1,25 = 20.16, p < 0.001], there was no strain × treatment interaction[F1,25 = 0.53, p > 0.05] . MRL/MpJ mice also had higher expression of NR2C than C57BL/6J mice as indicated by a main effect of strain [F1,28 = 13.36, p < 0.01]. A main effect of CORT exposure indicated that CORT decreased NR2C expression in both strains [F1,28 = 4.43, p < 0.05] in the absence of a strain × treatment interaction[ F1,28 = 0.17, p > 0.05].
Tropomyosine-receptor- kinase B (TrkB) mediates the effects of the neurotrophin BDNF, on hippocampal plasticity (Krystal et al., 2009), neurogenesis (Li et al., 2008) and cell survival (Sairanen et al., 2005) . Neurotrophins are reduced following exposure to stress (Balu and Lucki, 2009; Duman and Monteggia, 2006). The TrkB receptor was examined as ELISA indicated there were potential strain differences in the effects of CORT exposure on BDNF levels. A significant interaction indicated that CORT exposure significantly reduced TrkB expression in C57BL/6J mice but not MRL/MpJ mice [F1,25 = 9.48, p < 0.01] (Fig. 5B/ Table 2). Post hoc analysis indicated that MRL/MpJ mice given placebo or CORT pellets had 1.2 and 1.7- fold more TrkB mRNA than C57BL/6J mice respectively. In addition there was a main effect of strain [F1,25 = 52.02, p < 0.001], MRL/MpJ mice had greater expression of the TrkB receptor than C57BL/6J. There was no significant main effect of CORT exposure [F1,25 = 0.36, p > 0.05]
To examine which BDNF promoters were involved in the strain differences on the effects of CORT on BDNF protein levels, BDNF gene expression at promoter I, IV and IX were examined. BDNF expression at promoter 1 (Fig. 5B/ Table 2) was elevated in both MRL/MpJ and C57BL/6J mice exposed to CORT as indicated by a main effect of CORT exposure [F1,28 = 13.19, p < 0.05]. A significant main effect of strain indicated that overall MRL/MpJ mice had higher expression of BDNF I than C57BL/6J mice [F1,?8 = 13.40 , p < 0.05], but there was no interaction between strain and CORT exposure [F1,28 = 0.72 , p > 0.05]. BDNF expression at promoter IV (Fig. 5B) was significantly lower in both strains of mice exposed to CORT as indicated by a main effect of CORT exposure [F1,28 = 7.44 , p < 0.05]. A significant main effect of strain indicated MRL/MpJ mice had higher levels of BDNF expression at promoter IV than C57BL/6J mice regardless of treatment [F1,28 = 20.86 , p < 0.05]. There was no interaction between CORT exposure and strain on BDNF expression at promoter IV [F1,28 = 0.73 , p > 0.05]. At BDNF promoter IX there was no interaction between CORT and strain [F1,28 = 0.05, p > 0.05] and no main effects of CORT [F1,28 = 1.09 , p > 0.05] or strain [F1,28 = 0.86, p > 0.05] on BDNF expression at promoter IX .
The inflammatory markers MMP2 and MMP3 (Fig 5C) were examined as they have been shown to contribute to strain differences in wound healing (Gourevitch et al., 2003, Hampton et al., 2004). There was a significant interaction between CORT exposure and strain on mRNA levels for MMP2 [F1,23 = 4.30, p < 0.05]. Post hoc analysis indicated that MRL/MpJ mice expressed 2.5-fold higher levels of MMP2 expression than C57BL/6 mice but this marker was unaltered following exposure to CORT. In addition there was a main effect of strain [F1,23 = 337.31, p < 0.001] but no main effect of CORT exposure [F1,23 = 0.08, p > 0.05]. A significant main effect of strain indicated that MRL/MpJ mice also expressed overall higher levels of MMP3 than C57BL/6J mice [F1,27 = 7.36, p < 0.05]. There was no main effect of CORT exposure on MMP3 expression [F1,27 = 0.04, p > 0.05], there was a trend towards an interaction between strain and CORT exposure [F1,27 = 3.17 p = 0.08]. Post-hoc analysis determined that CORT treated MRL/MpJ mice had 1.4-fold higher expression of MMP3 than CORT treated C57BL/6J mice (p < 0.05) whereas placebo mice did not differ by strain (p > 0.05).
Stress has repeatedly been shown to reduce cell proliferation and neurogenesis in a number of animal models of depression (Tanapat et al., 1998, Malberg and Duman, 2003, Dranovsky and Hen, 2006, Stranahan et al., 2006). Cell proliferation is reduced in humans suffering from depression, an effect that can be reversed with antidepressant treatment (Boldrini et al., 2009). CORT exposure can also reduce neurogenesis and this effect can be reversed by antidepressant treatment as well (Murray et al., 2008). The present study demonstrates for the first time the existence of strain differences in the effects of CORT manipulation on cell proliferation and neurotrophin mobilzation in mice. MRL/MpJ mice are resistant to the effects of chronic CORT exposure on cell proliferation and BDNF protein expression when compared to C57BL/6J mice. Strain differences in transcription for the NMDA receptor, TrkB receptor, ,BDNF protein and inflammatory response may contribute to these strain differences in proliferation and neurotrophin release, although functional relevance still needs to be addressed.
Two previous studies have examined strain differences in the relationship between CORT levels and cell proliferation in rodents. One study found that strain can effect cell proliferation and CORT levels in mice although a causal relationship was not examined (McCutcheon et al., 2008). An examination of neurokinin-1 receptor mice backcrossed on to a C57BL/6J or 129B6 backgrounds found differences in the CORT response of the two strains to acute restraint stress. In addition they found that the mutation increased neurogenesis in the 129B6 background but not the C57BL/6J background. A study in rats, reported strain differences in levels of cell proliferation between Sprague-Dawley and Lister-Hooded rats treated with fluoxetine along with corresponding differences in basal levels of CORT (Alahmed and Herbert, 2008). Sprague-Dawley rats had lower levels of CORT when data was collapsed across treatment and higher levels of cell proliferation. Removal of the adrenal glands increased cell proliferation equally in both strains indicating that CORT was necessary for this strain difference. Implantation of Sprague-Dawley rats with CORT pellets flattened the diurnal rhythm of CORT release and blocked a fluoxetine induced increase in cell proliferation, similar to effects that had previously been reported in Lister-Hooded rats although a direct comparison was not made. The current study suggests that differences in the effects of CORT pellet implantation on cell proliferation and BDNF levels between B6 and MRL mice are not due to strain differences in circulating levels of CORT or HPA activity as there were no interactions between CORT pellet exposure and strain, and no effects of strain or CORT pellet implantation on CBG levels.
Examination of CORT exposure in other strains of mice all show some reductions in cell proliferation similar to to the effects reported here in C57BL/6J mice. This suggests that the wound healer MRL/MpJ mouse has a novel response to CORT exposure. Seven days of exposure to 4 CORT pellets reduced cell proliferation in CD 1 mice after 14 days of implantation (Murray et al., 2008). Stressors such as restraint and foot shock that increase CORT reduce cell proliferation in BALB/c mice when animals are stressed in the absence of companions (Cherng et al., 2010).
Pellet implantation increased circulating levels of CORT in both strains, indicating that the stress hormone was equally effective in both strains. A possible confound exists that even though the animals received the same number of pellets, the doses varied between strains due to weight differences. MRL/MpJ mice weigh on average 9 g more than C56BL/6J mice and differences between weight-based doses ranged from 3 mg/kg/day (1 pellet) to 9 mg/kg/day (4 pellets). It is unlikely that this could account for the differences in CORT pellet exposure on cell proliferation found between MRL/MpJ and C57BL/6J mice as cell proliferation was altered at a dose (1 pellet) that did not significantly increase circulating CORT in either strain. Also regardless of the weight-based differences, the doses overlapped. For example, the four pellet dose in the MRL/MpJ mouse overlapped the 2 pellet dose in the C57BL/6J mouse. Furthermore, C57BL/6J mice did not receive more CORT than MRL/MpJ mice. In the dose-response study, MRL/MpJ mice actually had a significantly higher change from baseline CORT levels when data was collapsed across dose. However, in experiment 2, the strains did not differ in circulating levels of glucocorticoids in the hippocampus of mice given placebo or two CORT pellets. The possibility existed that CBG levels could contribute to strain differences in the effects of CORT exposure as more than 90% of CORT may exists in a bound state (Breuner and Orchinik, 2002). However, plasma levels of CBG were also not altered by CORT pellet implantation, strain or any interaction between these factors.
Examination of mRNA levels for the glucocorticoid receptor were similar and decreased to the same degree in hippocampus and hypothalamus of both strains suggesting that negative feedback on gene expression were not different between the strains. However, differences in corticotropin releasing factor (CRF) and adrenocorticotropic hormone (ACTH) have not been examined and may still contribute to these strain differences. Mineralocorticoid receptor mRNA levels were also decreased in both strains by exposure to CORT but were not altered in the hypothalamus. Given that there were no differences in transcription, it seems unlikely that there were strain differences in transactivation, however the possibility is not ruled out that there could be indirect strain differences through transrepression of NF-kappa B and other transcription factors (Scheinman et al., 1995, Wu et al., 2004). Stress likely acts on multiple pathways to alter neurogenesis, as cytokines have been shown to alter neurotrophins and neurogenesis through the NF-Kappa B pathway in addition to the direct effects of glucocorticoids (Koo and Duman, 2008, Koo et al., 2010).
The effects of adrenal steroids on neurogenesis are thought to be dependent on the downstream effects of the NMDA receptor. Administration of the NMDA antagonist MK-801 can block the effects of CORT exposure on cell proliferation and application of NMDA can block the increase in cell proliferation induced by adrenalectomy (Cameron et al., 1998). Therefore NMDA receptor subunits were examined to determine whether they contributed to strain differences in vulnerability to stress. The NMDA receptor is an ionotrophic glutamate receptor formed from a heteromeric assembly of a NR1 and NR2 subunit resulting in a voltage dependent receptor that and requires co-activation of both gylcine and glutamate (Cull-Candy et al., 2001, Prybylowski and Wenthold, 2004). The NR1 binds glycine and the NR2 subunit binds glutamate; a few NMDA receptors also have a NR3 subunit that binds gylcine and may modulate the activity of NR1 and 2 but occur rarely (Nacher et al., 2007, Low and Wee, 2010).
The present study examined the effects of 7 days of CORT exposure on NR 1 and NR 2a-c. NR 2d was not examined as it is only found on a subset of inter-neurons in the hippocampus and does not occur on granule or pyramidal cells (Standaert et al., 1996). NR1 and the 2b and 2c subunits were expressed at higher levels in MRL/MpJ mice than C57BL/6J mice. Exposure to two CORT pellets for 7 days did not alter NR1 or 2a levels in either strain. As animals mature, the majority of NMDA receptors switch from NR2b to NR2a assembly, however NR2b subunits are present on proliferating cells and newborn neurons in the adult hippocampus (Nacher et al., 2007). Gene expression of the 2b and 2c subunits were reduced in both strains exposed to glucocorticoids although there were higher levels in MRL/MpJ mice than C57BL/6J mice even with the stress induced reductions. It should be noted that NR2b expression may result in negative effects upon proliferation as application of a NR2b specific antagonist led to higher levels of cell proliferation and greater cell survival in vivo under un-stressed conditions (Hu et al., 2008). This would argue against the higher levels of NR2b found in the MRL/MpJ mouse as being beneficial. As the authors also found in this study that MK-801 in vitro decreases cell proliferation whereas others have reported it to block the effects of CORT on cell proliferation in vivo (Cameron et al., 1998), it is possible that NMDA subunit effects on neurogenesis may differ between stressed and unstressed conditions. The NR2c subunit has been characterized as occurring on proliferating astrocytes in the adult hippocampus and does not co-localize with the mature neuronal marker NeuN (Karavanova et al., 2007).
The current study demonstrates strain differences in the effects of CORT on BDNF protein levels. BDNF levels were lower in C57BL/6J mice exposed to 2 or 4 CORT pellets than in MRL/MpJ mice. Within the hippocampus BDNF is made predominantly within granule cells (Binder et al., 2001) and the interesting possibility exists that differences in cell proliferation could affect BDNF transcription although generally it is more accepted that changes in BDNF levels lead to changes in neurogenesis (Sairanen et al., 2005, Li et al., 2008). Recent data has indicated that NMDA receptor activation alters transcription of BDNF at promoters I and IV via chromatin remodeling (Tian et al., 2010). CORT pellet exposure reduced expression of the BDNF IV promoter in C57BL/6J mice but not MRL/MpJ mice. BDNF IV has been shown to be involved with activity dependent translation of BDNF (Koppel et al., 2010, Gomez-Pinilla et al., 2011). Transgenic mice lacking BDNF promoter IV transcription show depression-like behavior including increased immobility in the tail-suspension test, increased anhedonia in the sucrose preferences test and learned helplessness behavior (Sakata et al., 2010). Furthermore, BDNF promoter IV is suppressed by hyper-methylation in the post-mortem brains of human suicides (Keller et al., 2010) suggesting a relationship between reductions in the expression of this promoter and susceptibility to stress. BDNF I activity was also altered by stress. Both strains had increased levels of transcription when exposed to CORT, but the MRL/MpJ mice had higher overall levels of BDNF I expression. An examination of a single restraint stress in rats found that expression of the BDNF promoter I and IV were both reduced 2 hours after the stressor but not 24 hours after stress (Fuchikami et al., 2009) and may reflect differences between acute stress and chronic CORT exposure. Interestingly, BDNF promoter I expression is increased by NMDA receptor activation (Tian et al., 2010) and is unregulated by contextual fear learning in rats (Lubin et al., 2008) suggesting that its up regulation may be involved in how an organism learns about threatening stimuli in the environment.
BDNF binds to and starts a molecular cascade at the TrkB receptor that can affect neurogenesis (Li et al., 2008), therefore TrkB receptor mRNA levels were examined. Exposure to two pellets of CORT for 7 days reduced TrkB expression in the hippocampus of C57BL/6 mice but did not decrease TrkB expression in MRL/MpJ mice. Chronic CORT exposure has previously been reported to reduce TrkB protein levels in the hippocampus and frontal cortex of mice (Kutiyanawalla et al., 2011). In humans with mood disorders and schizophrenia, TrkB receptor mRNA has also been reported as reduced in post-mortem hippocampal tissue (Thompson Ray et al., 2011) indicating that alterations in the expression of this receptor are related to psychiatric illness.
In addition to direct regulation of neurotrophin mobilization through the BDNF pathway, these growth factors can be cleaved by enzymes such as matrix metalloproteinases (MMPs) involved in modification of the extra-cellular environment (Benekareddy et al., 2008). While MMPs have traditionally been investigated for their role in the inflammatory response, cell death and neurodegenerative disorders (Rosenberg, 2009), they can also have beneficial impacts and have been implicated in synaptic and structural plasticity (Benekareddy et al., 2008). The gelatinases MMP2 and 9 are involved with neurogenesis, angiogenesis, axon regeneration and remyelination in addition a role in apoptosis (Rosenberg, 2009). In vitro, the inhibition of MMPs can reduce cell proliferation and differentiation of human stem cells, and in vivo a relationship exists between increased activation of the MMPs and neurogenesis following forebrain ischemia in gerbils (Wojcik et al., 2009). Earlier work demonstrated strain differences between MRL/MpJ and C57BL/6J mice in the expression of MMPs in cutaneous tissue and their inhibitors which played a causative role in wound healing (Heber-Katz et al., 2004a). Therefore we compared the effects of CORT treatment on the expression MMP2 and MMP3 in the hippocampus between strains. While MMP 2 and 9 have similar functions and have been implicated in neurogenesis, MMP3 plays a role only in degrading the extracellular matrix. By examining these two members of this extensive family we attempted to infer whether MMP expression played a specific role in the effects of strain differences in response to CORT on cell proliferation or if in general there was just greater overall expression of this family. While expression of both proteinases were elevated in MRL/MpJ mice compared to controls, only MMP2 demonstrated a strain-dependent response after CORT treatment. CORT treated MRL/MpJ mice had 2-fold higher expression of MMP2 than their C57BL/6J counterparts. Given the role MMP2 has been proposed to play in neurogenesis and angiogenesis (Rosenberg, 2009, Wojcik et al., 2009), and may contribute to the strain differences that we report on cell proliferation.
The MRL/MpJ mouse has an unusual capacity for wound healing which is due to differences in its genetic profile from C57BL/6J mice (Heber-Katz et al., 2004b). The MRL/MpJ mouse also demonstrates greater increases in hippocampal cell proliferation, neurotrophin generation and anxiolytic behavioral effects after chronic antidepressant drug treatments than C57BL/6J mice (Balu et al. 2009a). The present study is the first to show that MRL/MpJ mice are also resistant to the detrimental effects of stress hormone exposure. Exposure to 7 days of CORT treatment increased plasma and tissue levels of CORT and resulted in similar transcriptional negative feedback in the hippocampus and hypothalamus of both strains. However, MRL/MpJ mice were resistant to the effects of chronic CORT pellet exposure on cell proliferation and BDNF protein levels in the hippocampus. The possibility remains that negative feedback through ACTH and CRF may still contribute to these strain differences in the effects of on hippocampal plasticity as they were not directly tested. Changes in transcriptional activity were small and a more through approach such as micro-array or RNA sequencing is probably needed to find novel mechanisms. Using a candidate gene approach the current study reports that CORT exposure had different effects on pathways involved in neuronal plasticity in MRL/MpJ and C57BL/6J mice including the NMDA receptor, the TrkB receptor, activity dependent BDNF transcription and MMPs. It is likely that these transcriptional pathways act in concert to protect MRL/MpJ mice from the deleterious effects of stress.
This research was supported by USPHS grants RO1-MH86599 and T32-MH14654. We are grateful for the valuable assistance of Dr. Julie Blendy and Dr. Jill Turner with parts of this research.
Conflicts of Interest
The authors have no conflicts of interest to report.