恐懼記憶形成的神經機制: 海馬、前額皮質與依核的互動

Abstract

記憶痕跡是否形成及儲存於單一區域一直是個重大爭議。本實驗室過去的研究顯示情緒記憶形成需要依賴不同的神經結構共同運作。本論文研究利用抑制型躲避學習作業和情境恐懼制約兩種情緒記憶作業，探討在記憶穩固歷程中，大白鼠之海馬、前額皮質以及依核三者的互動情形。實驗1發現在抑制型躲避學習作業的訓練後，立即將正腎上腺素（0.2 μg）注射到背側海馬中，會促進大白鼠的記憶穩固歷程；若同時使用二丁卡因抑制前額皮質內側的功能，則正腎上腺素引發的促進記憶效果會被阻斷。同樣地，當背側海馬被二丁卡因抑制時會阻斷由正腎上腺素（0.1 μg）在前額皮質內側中所造成的促進記憶效果。實驗2在訓練後立即將正腎上腺素（2.0 μg）注射到背側海馬或前額皮質內側中，能夠促進情境恐懼制約作業之記憶穩固；過低（0.2 μg）或過高（5.0 μg）的劑量並不會對記憶表現造成影響，且使用正腎上腺素2.0 μg於訓練後六小時再注射到背側海馬或前額皮質內側中，亦不影響記憶。實驗3發現以二丁卡因抑制前額皮質內側，可阻斷正腎上腺素（2.0 μg）在背側海馬所導致的記憶增進效果。同樣地，抑制背側海馬的功能，亦逆轉了前額皮質內側中由正腎上腺素（2.0 μg）所造成的促進記憶效果。可見此二結構之間的互動對於情境恐懼記憶的形成十分重要。實驗4顯示閾下劑量的正腎上腺素同時注射到背側海馬與前額皮質內側，會顯著增進抑制型躲避學習記憶，但對於情境恐懼制約作業並無影響。實驗5進一步發現在抑制型躲避學習作業的訓練後，以二丁卡因抑制依核活動，正腎上腺素在背側海馬或前額皮質內側便無法促進記憶。由此推測在抑制型躲避學習的訊息處理過程中，海馬和前額皮質的訊息是匯送到依核進行處理。然而在情境恐懼制約作業中，抑制依核並不阻斷正腎上腺素注射在任一結構所造成的記憶增進效果。因此，在該作業中海馬與前額皮質兩者的訊息可能經由其他神經結構形成振盪的神經迴路。以上結果顯示大白鼠在形成恐懼記憶時，需要海馬與前額皮質的互動，但在操作式制約學習與古典條件化學習中，互動模式有所不同。

Evidence from our laboratory has shown that affective memory processing engages many brain regions. However, how information flows within this widely distributed neural circuit is unclear. The hippocampus (DH), medial prefrontal cortex (mPFC) and nucleus accumbens (NAc) are implicated in processing affective memory. This study examined interaction among these regions during memory formation in two affective learning tasks. Male Wistar rats with chronic cannulae implanted in the DH, mPFC or NAc were trained on a one-trial step-through inhibitory avoidance task or a contextual fear conditioning task. Shortly after training, norepinephrine (NE) was given into one of the targets and 4% lidocaine was infused into another. Retention was tested 24 hrs after training. The results in inhibitory avoidance showed that infusion of 0.2 μg NE after training into DH enhanced memory and this effect was blocked by concurrent intra-mPFC infusion of lidocaine. Likewise, the memory enhancing effect of 0.1 μg NE infused into the mPFC was blocked by concurrent intra-DH infusion of lidocaine. In contextual fear conditioning, we found immediately posttraining infusion of 2.0 μg, but not other doses, of NE into DH or mPFC increased retention. However, the same treatment given 6 hrs after training had no effect on memory. These data suggest that NE infused into the DH and mPFC had dose- and time-dependent effects in memory consolidation of contextual fear conditioning. Furthermore, in contextual fear conditioning, the memory enhancing effect of NE given to the DH was blocked by simultaneously infusing lidocaine into mPFC and vice versa. These findings suggest that during memory formation the DH and mPFC either form a reverberating circuit or send convergent inputs to a third target, such as the NAc. The latter possibility was favored by the findings in the inhibitory avoidance task that NE at sub-enhancing doses induced robust enhancement if simultaneously given into both regions. In contrast, results from the contextual fear conditioning showed that infusing sub-enhancing doses of NE into DH and mPFC simultaneously induced no additive effect. It implies that in this task information flow may form a reverberating circuit. A following-up experiment showed that the memory enhancing effect of NE infused into the DH or mPFC was blocked by inactivating the NAc with lidocaine in the inhibitory avoidance task, but not in the contextual fear conditioning task. These findings, taken together, suggest that NE in mPFC and DH during an inhibitory avoidance task modulated memory formation processes via the convergent inputs to the NAc, but they formed a reverberating circuit without engaging the NAc during formation of contextual fear memory.