Differential response of hippocampal subregions to stress and learning

Abstract

The hippocampus has two functionally distinct subregions-the dorsal portion, primarily associated with spatial navigation, and the ventral portion, primarily associated with anxiety. In a prior study of chronic unpredictable stress (CUS) in rodents, we found that it selectively enhanced cellular plasticity in the dorsal hippocampal subregion while negatively impacting it in the ventral. In the present study, we determined whether this adaptive plasticity in the dorsal subregion would confer CUS rats an advantage in a spatial task-the radial arm water maze (RAWM). RAWM exposure is both stressful and requires spatial navigation, and therefore places demands simultaneously upon both hippocampal subregions. Therefore, we used Western blotting to investigate differential expression of plasticity-associated proteins (brain derived neurotrophic factor [BDNF], proBDNF and postsynaptic density-95 [PSD-95]) in the dorsal and ventral subregions following RAWM exposure. Lastly, we used unbiased stereology to compare the effects of CUS on proliferation, survival and neuronal differentiation of cells in the dorsal and ventral hippocampal subregions. We found that CUS and exposure to the RAWM both increased corticosterone, indicating that both are stressful; nevertheless, CUS animals had significantly better long-term spatial memory. We also observed a subregion-specific pattern of protein expression following RAWM, with proBDNF increased in the dorsal and decreased in the ventral subregion, while PSD-95 was selectively upregulated in the ventral. Finally, consistent with our previous study, we found that CUS most negatively affected neurogenesis in the ventral (compared to the dorsal) subregion. Taken together, our data support a dual role for the hippocampus in stressful experiences, with the more resilient dorsal portion undergoing adaptive plasticity (perhaps to facilitate escape from or neutralization of the stressor), and the ventral portion involved in affective responses.

Figure 1. CUS and learning were both stressful.

Animals that underwent CUS did not gain weight over the 2-week period of stressor exposure, whereas control animals did (A). Exposure to the CUS paradigm raised corticosterone levels, as did learning in the RAWM (B). Note, however, that learning did not further elevate corticosterone in stressed animals. *significantly different from baseline, † significantly different from Post CUS control.

Figure 2. CUS facilitated long-term spatial memory in the RAWM.

Escape latencies did not differ between control and stressed animals during the acquisition trials (1–12), or on the short-term memory trial (30 min) (A). However, stressed animals took significantly less time to locate the hidden platform on the long-term memory trial (24 hrs). A similar pattern was seen for errors made during search (B). * significantly different from control.

Figure 3. Stress most severely affected neurogenesis in the ventral dentate gyrus.

Compared with controls, rats in the CUS group showed decreased proliferation (A), survival (B) and neuronal differentiation (C) in the dentate gyrus. This effect was most pronounced in the ventral, compared to the dorsal, sub-region († indicates significant difference between subregions). * significantly different from control.

Figure 4. A stressful spatial navigation task differentially affected protein expression in the dorsal and ventral subregions.

Expression of mature BDNF was not significantly changed by RAWM exposure in either the dorsal or ventral dentate gyrus (A). In contrast, proBDNF was significantly increased in the dorsal dentate, and significantly decreased in the ventral (C). PSD-95 was unchanged in the dorsal, but significantly increased in the ventral dentate (C). * significantly different from control.