SEARCH

SEARCH BY CITATION

Keywords:

  • hippocampus;
  • neurogenesis;
  • post-traumatic stress disorder;
  • subgranular zone

Abstract

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Aim:  Inescapable shocks (IS) have been reported to reduce the number of 5-bromo-2′-deoxyuridine (BrdU)-positive cells in hippocampus. Antidepressants prevent this reduction, and the role of neurogenesis in depression is now suggested. It has been reported, however, that the number of BrdU-positive cells was not different between the rats that developed learned helplessness and those that did not. This suggests that reduction of neurogenesis does not constitute a primary etiology of depression. It has been previously shown that IS can cause various post-traumatic stress disorder (PTSD)-like behavioral changes in rats. The aim of the present was therefore to examined whether the reduction of BrdU-positive cells relates to any PTSD-like behavioral changes in this paradigm.

Methods:  Rats were given either inescapable foot-shocks (IS) or not shocked (non-S) treatment in a shuttle box on day 1 and received BrdU injections once daily during the first week after IS/non-S treatment. On day 14, rats treated with IS and non-S were given an avoidance/escape test in the shuttle box and dorsal hippocampal SGZ were analyzed by BrdU immunohistochemistry.

Results:  In accordance with previously reported results, IS loading resulted in fewer BrdU-positive cells in the hippocampal subgranular zone (SGZ). Furthermore, in the IS-treated group, the number of BrdU-positive cells in the hippocampal SGZ was negatively correlated at a significant level with several hyperactive behavioral parameters but not with hypoactive behavioral parameters. Earlier findings had indicated that chronic selective serotonin re-uptake inhibitor administration, which is known to increase hippocampal neurogenesis, restored the increase in hypervigilant/hyperarousal behavior but did not attenuate the increase in numbing/avoidance behavior.

Conclusion:  The regulatory mechanism responsible for the decreased proliferation and survival of cells in the hippocampus may be related to the pathogenic processes of hypervigilance/hyperarousal behaviors.

NEUROGENESIS HAS BEEN reported to continue postnatally in the subventricular zone and the hippocampal subgranular zone (SGZ) in the brain.1 Physiologically relevant stressors including psychosocial stress,2 intruder stress,3 and predator odor stress4 decrease neurogenesis in the mammalian hippocampus.5 It has also been reported that severe and inescapable electric foot shocks (IS),6 immobilization stress,7 and tail shocks8 reduce the number of 5-bromo-2′-deoxyuridine (BrdU)-positive cells in the hippocampus. Antidepressants prevent this reduction, and the role of neurogenesis in depression has been suggested.9 It has also been reported, however, that the number of BrdU-positive cells was not different between the rats that developed learned helplessness and those that did not.7 This suggests that the reduction of neurogenesis does not constitute a primary etiology of depression.

In post-traumatic stress disorder (PTSD) the immediate response to a life-threatening traumatic event is characterized by hypervigilance/hyperarousal. The response, however, concurrently includes hyporesponsivity that is characterized by a numbing of general responsiveness and avoidance of stimuli associated with the trauma. Various procedures have been adopted to generate traumatic stress stimuli in PTSD animal models;10 these include the predator odor exposure paradigm,11 electric tail shocks,12 and underwater trauma.13 A sequential stress paradigm with restraint stress, swim stress, and halothane exposure has also been used.14 We have considered that IS inducing learned helplessness behavior in rats corresponds to a life-threatening stressor that may cause PTSD in humans. At the same time, we also devised an avoidance/escape test (AET) in order to examine behavioral effects of IS in the same environment. We previously showed that IS cause PTSD-like behavioral changes in rats.15–17 In the present study we examined whether the reduction of BrdU-positive cells in rat hippocampal subregions correlate with any PTSD-like behavioral parameters in this paradigm.

METHODS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Animals

Young 6-week-old male Fischer 344 rats (Charles River, Tokyo, Japan) were used in all experiments. In order to minimize the number of rats used in the experiment, we used the Fischer rat strain in our animal model because they exhibit the most apparent behavioral changes.16 All animal treatments were approved by the local animal ethics committee of the National Defense Medical College (No. 05031, 06008, 06083). Animals were group-housed under controlled conditions (24°C room temperature, 55% humidity, and lights on from 0700 to 2000 h) with freely available food and water.

Apparatus

All behavioral tests were performed using a fully automated shuttle box (MED Associates, Albans, VM, USA) that consisted of two compartments (20 × 20 × 46 cm) with four pairs of infrared photocells above a grid floor, separated by a retractable door in an arched doorway measuring 11 cm × 9 cm. Scrambled foot-shocks were delivered through the grid floor on the side where a rat was present.

Experimental design

Rats were randomly divided into the IS (n = 16) and not shocked (non-S, n = 9) groups. An avoidance/escape test (AET) was performed on day 14 after the IS/non-S treatment (Fig. 1a). The time point of 14 days after IS was used in order to avoid an acute stress response including learned helplessness.18

image

Figure 1. (a) Experimental design and (b) definition of behavioral parameters during an avoidance/escape test (AET). BrdU, 5-bromo-2′-deoxyuridine; ITI, inter-trial interval; IS, inescapable foot shock; non-S, not shocked.

Download figure to PowerPoint

Procedure for inescapable shock

On day 1, rats in the shocked group were given 60 IS (0.8 mA, 15-s duration) with a random inter-trial interval (ITI) ranging from 7.5 to 22.5 s (average, 15 s) in the absence of any conditioned stimulus (CS). Rats in the non-S group were placed in the shuttle box for 30 min but did not receive any shocks.

Avoidance/escape task procedure

On day 14 the rats were given an AET consisting of 80 trials. Before testing, rats were given a habituation period for 5 min in the shuttle box, and could freely move to the other compartment through the gate in the absence of any shock. The number of crossings during habituation was recorded as ‘crossings during adaptation’, and the number of photocell interruptions in the compartment was also recorded as ‘movements during adaptation’. The first trial was initiated by presenting the signal light as the CS and administering foot shocks at 5 s after the CS. Five seconds after the onset of CS, the rats received a foot shock of 0.8 mA for a maximum length of 15 s that served as the non-condition stimulus (NCS). Each AET trial was divided into three segments: (i) a condition phase of 5 s; (ii) a non-condition phase of maximum length 15 s; and (iii) a random ITI of 15 ± 7.5 s. If the rat moved to the next compartment during CS presentation, ‘avoidance’ was recorded. If the animal moved to the other side when both CS and NCS were presented, ‘escape’ was recorded. The CS and NCS were terminated immediately after the rat moved to the next compartment. If the animal failed to respond, ‘error’ was recorded, and both CS and NCS were terminated at the end of the trial. Movements through the gate after the foot shock had been terminated and prior to initiation of the next signal light were recorded as ‘crossings during the ITI’ (Fig. 1b).

5-Bromo-2′-deoxyuridine injections

Survival of newly generated cells in the hippocampus was evaluated using the BrdU (Sigma, St Louis, MO, USA) labeling method. Rats received 50 mg/kg of i.p. BrdU injections dissolved in saline once daily during the first week after IS/non-S treatment (Fig. 1a) and were killed immediately after AET. This injection schedule was set to avoid the immediate influence of AET just before death6 and to examine the altered level of proliferation and survival of hippocampal cells.

Tissue preparation and BrdU immunohistochemistry

Rats were decapitated after anesthetization with an i.p. injection of 50 mg/kg sodium pentobarbital. Brains were removed and frozen directly in a cryocompound (Microedge, Tokyo, Japan). Sections of the brain (10 µm thick, 60 µm apart) were prepared from the dorsal hippocampus using a frozen microtome. Sections were transferred to MAS-coated slides (Matsunami, Osaka, Japan) and air dried. After fixation with 4% paraformaldehyde in 1× phosphate-buffered saline (PBS), the sections were incubated with 2 N HCl for 1 h at 37°C to denature the DNA. They were blocked for 1 h with 10% fetal bovine serum (FBS), and incubated overnight at 4°C with anti-mouse BrdU monoclonal IgG antibody (1:12.5; Roche, Mannheim, Germany) in 1× PBS containing 2.5% FBS. The sections were incubated with Cy3-labeled anti-mouse IgG antibody (1:200; Chemicon, Temecula, CA, USA) for 90 min at room temperature. The sections were mounted and coverslipped.

Quantification of BrdU-labeled cells

Quantification of BrdU-labeled cells was carried out by counting the total number of BrdU-labeled cells in the SGZ, granule cell layer (GCL), and hilus. Sampling of BrdU-positive cells was carried out throughout the dorsal hippocampus (from −2.1 to −4.8 mm relative to the bregma).19 To stereologically estimate the cell numbers, all BrdU-labeled cells were counted in every fifth section (nine sections per rat) at 100× magnification on a DP70 immunofluorescence microscope (Olympus, Tokyo, Japan) by an experimenter blinded to the study. BrdU-positive cells that were within two cell body widths of the edge of the GCL were considered to be located in the SGZ; cells located two cell widths away were regarded as being in the hilus.20

Statistical analysis

All statistical analysis was done using SPSS 9.0 (SPSS, Chicago, IL, USA). Data are presented as mean ± SEM. The results of experiments with two groups were compared using the unpaired Student's t-test. The average avoidance latency, escape latency, and number of avoidances and crossings during the ITI in each block of 10 trials were analyzed using repeated measures analysis of variance (anova). If a significant difference was detected, pairwise comparisons were performed. The correlation coefficient was evaluated using simple regression analysis.

RESULTS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Avoidance latency

Both IS (n = 16) and non-S (n = 9) groups displayed a gradual shortening of the avoidance latency (within-subjects effects: IS, F7,17 = 15.5, P < 0.001; non-S, F7,17 = 4.1, P < 0.01; Fig. 2a). There was decreased avoidance latency in the IS group when compared to the non-S group (between-subjects effects: F1,23 = 5.9, P < 0.05; Fig. 2a). The interaction of IS × blocks was also significant (F7,161 = 2.9, P < 0.01; Fig. 2a).

image

Figure 2. Behavioral alterations of the inescapably shocked (IS; ●, n = 16) and not shocked (non-S; ○, n = 9) groups during the avoidance/escape test (AET). (a) Average avoidance latency in every 10 trials; (b) average escape latency in every 10 trials; (c) number of avoidances; (d) number of crossings during the inter-trial interval (ITI); (e) total number of avoidances, crossings during the ITI, and crossings during adaptation (▪, IS, n = 16; □, non-S, n = 9); (f) number of movements during adaptation. All data are presented as mean ± SEM. Significant differences in pairwise comparisons between both groups is shown by the asterisks. *P < 0.05, **P < 0.01, and ***P < 0.001 vs non-S group. (g) Correlation between the number of crossings during adaptation and movements during adaptation (r = 0.91), and (h) correlation between the numbers of crossings during the ITI and avoidance (r = 0.648). The correlation coefficient (r) was evaluated using simple regression analysis. **P < 0.01, ***P < 0.001.

Download figure to PowerPoint

Escape latency

Only the IS group had a gradual shortening of the escape latency (within-subjects effects: IS, F7,17 = 18.1, P < 0.001; non-S, F7,17 = 2.5, P > 0.05; Fig. 2b). There was no significant main effect of IS on the average escape latency (between-subjects effects: IS, F1,23 = 0.1, P > 0.05; Fig. 2b). A difference, however, in the escape latency between the IS group and non-S group was supported by the significant interaction of IS × blocks (F7,161 = 7.2, P < 0.001; Fig. 2b). In the IS group the escape latency was significantly prolonged in the first block but was drastically shortened thereafter when compared to the non-S group.

Avoidance response

The total number of avoidances was significantly higher for IS rats, (t-test; IS, 39.7 ± 3.3; non-S, 26.4 ± 4.4; t23 = 2.4, P < 0.05; Fig. 2e). Both groups had a gradual increase in the number of avoidances (within-subjects effects: IS, F7,17 = 14.5, P < 0.001; non-S, F7,17 = 5.7, P < 0.01; Fig. 2c). The number of avoidances for IS rats was significantly different to that for non-S rats (between-subjects effects: F1,23 = 5.8, P < 0.05; Fig. 2c). The interaction of IS × blocks was also significant (F7,161 = 3.3, P < 0.01; Fig. 2c).

Other crossings and movements

The total number of crossings during the ITI for IS rats was significantly higher than that for non-S rats (t-test; IS, 42.5 ± 5.1; non-S, 22.1 ± 3.6; t23 = 2.7, P < 0.05; Fig. 2e). Both groups had significant changes in the number of crossings during the ITI (within-subject effects: IS, F7,17 = 6.9, P < 0.001; non-S, F7,17 = 5.1, P < 0.01; Fig. 2d). The number of crossings during the ITI, however, tended to increase only for IS rats. The number of crossings during the ITI was significantly increased for IS rats when compared to non-S rats (between-subjects effects: F1,23 = 7.5, P < 0.05; Fig. 2d). The interaction of IS × blocks was also significant (F7,161 = 7.1, P < 0.001; Fig. 2d). There was a positive correlation between the numbers of avoidances and crossings during the ITI in the IS group (IS, n = 16, F1,14 = 10.1, t15 = 3.1, r = 0.64, P < 0.01; non-S, n = 9, F1,7 = 2.2, t8 = 1.5, r = 0.49, P > 0.05; Fig. 2h).

The number of crossings during adaptation for IS rats was significantly lower than for non-S rats (t-test; IS, 3.0 ± 0.9; non-S, 13.1 ± 1.0; t23 = −6.8, P < 0.001; Fig. 2e). The movements during adaptation measured by photocell interruption were also lower in IS rats than in non-S rats (t-test; IS, 29.5 ± 7.7; non-S, 96.7 ± 7.9; t23 = −5.6, P < 0.001; Fig. 2f) and were significantly correlated with the number of crossings during adaptation in IS rats (IS, n = 16, F1,14 = 68.5, t15 = 8.2, r = 0.91, P < 0.001; non-S, n = 9, F1,7 = 0.7, t8 = 0.8, r = 0.30, P > 0.05; Fig. 2g). No significant difference was found in the number of errors (t-test; IS, 0.1 ± 0.1; non-S, 0.0 ± 0.0; t15 = 1.4, P > 0.05).

BrdU-labeled cells in the hippocampus

The number of BrdU-positive cells in the SGZ, GCL, and hilus of the dorsal hippocampus were counted (Fig. 3a,b). In comparison to the non-S rats, the IS rats had an approximately 27% decrease in the number of BrdU-positive cells in the SGZ (t-test; IS, 8679 ± 319; non-S, 11 866 ± 290; t23 = −6.6, P < 0.001; Fig. 3c). There was no significant difference, however, in the number of BrdU-positive cells between groups both in the GCL (t-test; IS, 1245 ± 182; non-S, 1050 ± 115; t23 = 0.7, P > 0.05; Fig. 3c) and in the hilus (t-test; IS, 1923 ± 124; non-S, 2083 ± 88; t23 = −0.8, P > 0.05; Fig. 3c).

image

Figure 3. Immunofluorescence microscopy of a representative coronal section through the rat dorsal hippocampus. BrdU-positive cells were typically seen in the hippocampal subgranular zone (SGZ) between the granule cell layer (GCL) and the hilus. Immunohistochemistry was carried out 7 days after the last 5-bromo-2′-deoxyuridine (BrdU) injection and 14 days after (a) no shock (non-S; bar, 100 µm) or (b) inescapable foot shock (IS; bar, 100 µm) treatment. (c) Total number of newly generated and surviving cells in the SGZ, GCL and hilus (▪, IS, n = 16; □, non-S, n = 9). All data are presented as mean ± SEM. ***P < 0.001 vs non-S group. The number of BrdU-positive cells in the SGZ was plotted against behavioral parameters including (d) avoidances; (e) crossings during the inter-trial interval (ITI); (f) crossings during adaptation; and (g) movements during adaptation. The correlation coefficient (r) was evaluated on simple regression analysis. *P < 0.05, **P < 0.01.

Download figure to PowerPoint

Correlation between the number of BrdU-positive cells in the dorsal hippocampus and behavioral parameters

In the IS group the number of BrdU-positive cells in the SGZ was inversely correlated with the total number of avoidances (IS, n = 16, F1,14 = 8.5, t15 = −2.9, r = −0.61, P < 0.05; Fig. 3d), but no significant relationship was observed for non-S rats (non-S, n = 9, F1,7 = 0.1, t8 = −0.4, r = −0.15, P > 0.05; Fig. 3d). Similarly, in the IS group, the number of BrdU-positive cells in the SGZ was also inversely correlated with the total number of crossings during the ITI (IS, n = 16, F1,14 = 13.2, t15 = −3.6, r = −0.69, P < 0.01; Fig. 3e), whereas no relationship was found for non-S rats (non-S, n = 9, F1,7 = 0.7, t8 = −0.8, r = −0.30, P > 0.05; Fig. 3e). In both groups the number of BrdU-positive cells in the SGZ did not correlate with either the number of crossings during adaptation (IS, n = 16, F1,14 = 0.1, t15 = 0.2, r = −0.07, P > 0.05; non-S, n = 9, F1,7 = 0.72, t8 = 0.8, r = 0.30, P > 0.05; Fig. 3f) or the number of movements during adaptation (IS, n = 16, F1,14 = 0.0, t15 = −0.1, r = 0.01, P > 0.05; non-S, n = 9, F1,7 = 0.4, t8 = −0.6, r = −0.24, P > 0.05; Fig. 3g).

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

In the present study we found a gradual shortening in the avoidance latency and a gradual increase in the number of avoidances during AET performance. We also showed that changes were more pronounced in the IS group. Time-dependent shortening of escape latency was significant in the IS group. In addition, a time-dependent increase in the number of crossings during the ITI was observed in the IS group. Two explanations are possible for the time-dependent enhancement of hyperactive behavior in the IS group. One is that learning related to avoidance and escape from stress is accelerated in the IS group. The other is that hyperarousal in IS rats is gradually potentiated during AET. Previous studies reported that exposure to inescapable stress selectively enhanced aversive memory processing21 and also induced a greater avoidance response.22

We have postulated that decreased numbers of movements during adaptation and crossings during adaptation are indicative of numbing/avoidance, and that increased number of crossings during the ITI and avoidances are indicative of the hypervigilance/hyperarousal during PTSD.15–17 In the present study we confirmed that in the IS group, two parameters (numbing/avoidance behaviors and hypervigilant/hyperarousal behaviors) are significantly correlated with each other. The efficacy of paroxetine, a selective serotonin re-uptake inhibitor (SSRI), in normalizing hypervigilant/hyperarousal behaviors has been demonstrated.15 The facts that SSRI are therapeutically effective for PTSD23 and that arousal symptoms are the first to be ameliorated24 form the predictive validity for the present animal model. We considered that this behavioral paradigm satisfies three validities of the animal model. Facial validity is satisfied by hyperarousal, avoidance, and hypersensitivity to trauma-related stimuli; predictive validity by the effect of SSRI on behaviors; and constructive validity by long-term effect of traumatic stress.

In this behavioral paradigm, learned helplessness was demonstrated by longer escape latencies, in addition to fewer avoidances and crossings during the ITI in the IS group during the first block of the first 10 AET trials. It was previously reported that there was no relationship between reduced neurogenesis in the hippocampal SGZ and the intensity of learned helplessness.7

In the present study, after confirming IS-mediated reduction of the number of BrdU-positive cells in the hippocampal SGZ, a direct relationship with behavioral changes was investigated. BrdU was therefore administered for 7 days after IS loading, and the influence of AET was minimized6 before the animals were killed. We consider that the number of BrdU-positive cells reflects both the proliferation and survival rate of newly generated cells in the hippocampus. Microscopy indicated that the number of BrdU-positive cells in the SGZ was significantly and negatively correlated with hypervigilant/hyperarousal behavior but not with numbing/avoidance behavior in the IS group. This finding suggests that decreased number of BrdU-positive cells in the SGZ after IS might be relevant to the PTSD-like hyperactive behavior in rats.

Only 4–10 days are reportedly required for mitotic SGZ cells to extend axons into the CA3 pyramidal cell layer.25 Other reports, however, have shown that a 2–3-week period is required for newly divided cells in the SGZ to express neuronal markers,26 and a 4-week period is thought to be necessary for newly proliferated cells to mature and function as granule neurons.27 Therefore, we do not consider that the reduced number of SGZ BrdU-positive cells directly relate to increased hypervigilant/hyperarousal behavior at 2 weeks after IS. It is possible that changes in the microenvironment, including altered levels of hippocampal glucocorticoids,28 neurotrophins,29 interleukins,30 and altered serotonergic neurotransmission31 that collectively decrease the proliferation and survival of newly generated cells in the SGZ, could also influence hypervigilant/hyperarousal behavior by altering the functional activity of the existing neuronal system.

The other important implication of the present study is that numbing/avoidance and hypervigilance/hyperarousal may follow different pathogenic processes, and is supported by our previous finding that a 2-week administration of SSRI ameliorated only hypervigilance/hyperarousal.15 A recent report demonstrated hippocampal involvement in processing of ambiguous cues related to a non-condition stimulus.32

In the present study we observed a time-dependent increase in hypervigilance/hyperarousal that was negatively correlated with the number of BrdU-positive cells in the hippocampal SGZ. We therefore consider that the hippocampal microenvironment influencing the neurogenic activities is related to PTSD hyperarousal symptoms. One major theme to be clarified in future studies is the altered glucocorticoid secretion that can induce hippocampal vulnerability in the stress paradigm.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES