In order to examine the role of neurogenesis during memory formation and extinction, adult male mice were subcutaneously injected with several doses of MAM (1, 2.5 and 5 mg/kg) for 2 weeks. Since MAM can induce cachexsia, we monitored the general health and weight of the injected mice [16
]. We did not detect any weight loss or fur deterioration in the MAM-treated group. Instead, retardation of weight gain was detected in the mice that were injected with 2.5 mg/kg MAM (vehicle vs. 2.5 mg/kg, p < 0.05, data not shown). However, the 2.5 mg/kg MAM-treated mice showed similar levels of moving distance in the open field test (data not shown).
After the open field test, the mice were sacrificed for BrdU immunohistochemistry. MAM-treated dentate gyrus showed normal morphology compared with the vehicle-treated dentate gyrus (Fig. ). However, in MAM-treated mice, the number of BrdU+ cells was decreased in a dose-dependent manner. MAM treatments reduced the number of BrdU+ cells in the dentate gyrus by 17%, 35% and 58% in 1 mg/kg, 2.5 mg/kg and 5 mg/kg, respectively (Fig. ; n = 3 except for 5 mg/kg, n = 1; *, p < 0.05; **, p < 0.01 compared with vehicle). Notably, two out of three mice in the 5 mg/kg group died during MAM treatment. Therefore, we selected 3 mg/kg as MAM dose for the following behavioral experiments.
Figure 1 MAM treatments reduced hippocampal neurogenesis but did not affect gross morphology. (A) Representative images show the dentate gyrus of mice that were injected with MAM for 2 weeks. Doses of MAM are indicated in each image. Cells were labeled with BrdU (more ...)
As shown in Fig. , the open field test, contextual fear conditioning and extinction training were performed serially. Finally, the proliferation of newborn neurons was examined with immunohistochemistry. In the open field test, there was no difference between groups in the distance moved (Fig. ; saline = 3,390 cm ± 218 cm, MAM = 3,616 cm ± 225 cm, n = 10 and 13, respectively, p > 0.05). This result shows that 3 mg/kg of MAM did not affect general locomotion of the mice. When the MAM-treated mice were trained with contextual fear conditioning, they showed a similar percentage of freezing level on the following day (Fig. ; saline = 58.9% ± 6.5%, MAM: 64.0% ± 2.7%, n = 10 and 13, respectively, p > 0.05). Furthermore, ANOVA analysis revealed that MAM had no effect on extinction (p > 0.05). These results suggest that the depletion of newborn neurons did not affect either the formation or the extinction of contextual fear memory. However, when we performed BrdU immunohistochemistry to confirm the depletion of newborn neurons by MAM treatment, we did not find a significant reduction in the total number of BrdU+ cells in the MAM-treated dentate gyrus (data not shown). Since the mice were decapitated 17 days after the BrdU injection, it is possible that the survival of newborn neurons in MAM-treated mice may be enhanced by maintaining the proper number of newborn neurons.
Figure 2 Reduced hippocampal neurogenesis did not impair locomotion or the extinction of contextual fear memory. (A) A schematic diagram showing the experimental procedure using MAM. (B) Total distance moved of the saline- and MAM-treated mice. (C) Percentage (more ...)
To clearly address the effects of neurogenesis deficiency on the formation and extinction of contextual fear memory, we used gamma-ray irradiation, which is known to ablate newborn neurons more completely than MAM treatment [17
]. The latter is known to only reduce newborn neurons by up to 35%. Only the hippocampal region was exposed to gamma rays, while other brain regions were shielded from irradiation [17
]. We performed behavioral experiments 3 months after the irradiation in order to avoid side effects such as inflammation by activation of the microglia [19
]. After behavioral experiments, brain slices were analyzed with immunohistochemistry. One gamma-ray application depleted the number of BrdU+ cells in the dentate gyrus more efficiently than MAM treatments. Irradiated dentate gyrus showed a reduction of newborn neurons by 78%, 75% and 89% in 10 Gy, 15 Gy and 20 Gy, respectively (Fig. ; sham = 1,988 ± 58; 10 Gy = 433 ± 70; 15 Gy = 499 ± 73; 20 Gy = 216 ± 35; n = 8–16; **, p < 0.001 compared with sham; *, p < 0.05 compared with 15 Gy). However, we found no abnormal morphological features in the dentate gyrus from irradiated group (Fig. ). Moreover, the total distance moved showed no difference between sham and 10 Gy (Fig. ; sham = 2,770 cm ± 112 cm; 10 Gy = 2,797 cm ± 226 cm, n = 9, p > 0.05).
Figure 3 Gamma-ray irradiation ablated hippocampal neurogenesis. (A) Each image shows the dentate gyrus of sham (upper) and gamma-ray irradiated (lower) mice. Proliferating cells were stained with BrdU (red). Nuclei were counterstained with DAPI (blue). Scale (more ...)
Figure 4 Low gamma-ray irradiation did not block the formation and extinction of contextual fear conditioning with a 24-hour interval between retrieval and extinction training. (A) Experimental procedures. (B) Total distance moved of sham and 10 Gy-irradiated (more ...)
To examine the effect of ablated neurogenesis on the formation and extinction of contextual fear memory, we used a procedure similar to that of the MAM experiments. There was a 24-hour interval between retrieval and extinction training (Fig. ). In consistent with the MAM results, 10 Gy irradiation had no effect on the formation and extinction of fear conditioning compared with the sham control (Fig. ; n = 3–6, p > 0.05).
Next, we speculated that if hippocampal neurogenesis is required for extinction, the incorporation of newborn neurons in the dentate gyrus circuit may be important because it could interfere with the flow of information that encodes contextual fear memory [20
]. Therefore, in the next experiment, we introduced a 2-week interval between retrieval and extinction training because some contents of the synaptic connection may be required for blocking the flow of old memory (Fig. ) [1
]. As shown in Fig. , 20 Gy-irradiated mice showed less freezing compared with sham or 10 Gy when they were reexposed to the conditioned context on the day of retrieval (Fig. , sham = 54.0% ± 2.5%; 10 Gy = 53.8% ± 3.9%; 15 Gy = 44.3% ± 4.8%; 20 Gy = 33.9% ± 3.4%; n = 8–16; *, p < 0.05 compared with sham or 10 Gy). However, gamma-ray application had no effect on extinction (Fig. ; n = 8–16, p > 0.05). These results are consistent with Fig. . Reduced neurogenesis did not impair the extinction of contextual fear memory. To demonstrate these results clearly, we further analyzed the data from gamma-ray irradiation experiments. The level of extinction was plotted against the number of BrdU+ cells in the dentate gyrus. However, we did not find any correlation between the level of extinction and the level of neurogenesis (Fig. ; A: r = 0.194, p > 0.05; B: r = -0.282, p > 0.05). These data suggest that hippocampal neurogenesis is not correlated with the extinction of contextual fear memory. Considering that MAM treatment and low gamma-ray (10 Gy and 15 Gy) irradiation did not affect freezing, these results suggested that newborn neurons should be reduced by approximately 89% in order to significantly inhibit the formation of contextual fear memory.
Figure 5 High gamma-ray irradiation blocked the formation but not the extinction of contextual fear memory with a 2-week interval between retrieval and extinction training. (A) Experimental procedures. (B) Percent freezing in gamma-ray-irradiated groups on the (more ...)
Figure 6 Level of extinction is not correlated with total number of BrdU+ cells. (A) and (B) Level of extinction is not correlated with total number of BrdU+ cells. Data in Figs. 4D and 5C were reanalyzed to examine the correlation between the level of extinction (more ...)