A retrieval-specific mechanism of adaptive forgetting in the mammalian brain

Abstract

Forgetting is a ubiquitous phenomenon that is actively promoted in many species. How and whether organisms’ behavioral goals drive which memories are actively forgotten is unknown. Here we show that processes essential to controlling goal-directed behavior trigger active forgetting of distracting memories that interfere with behavioral goals. When rats need to retrieve particular memories to guide exploration, it reduces later retention of other memories encoded in that environment. As with humans, this retrieval-induced forgetting is competition-dependent, cue-independent and reliant on prefrontal control: Silencing the medial prefrontal cortex with muscimol abolishes the effect. cFos imaging reveals that prefrontal control demands decline over repeated retrievals as competing memories are forgotten successfully, revealing a key adaptive benefit of forgetting. Occurring in 88% of the rats studied, this finding establishes a robust model of how adaptive forgetting harmonizes memory with behavioral demands, permitting isolation of its circuit, cellular and molecular mechanisms.

Great apes selectively retrieve relevant memories to guide action

“…This indicates that great apes’ cognitive executive function repertoire includes resolution of memory competition. Suffering from, yet resolving, memory interference is considered a hallmark of flexible retrieval in humans 50 , 51 . Interference results from retrieval competition that occurs when two memories, a relevant and an irrelevant one, are cued by the present retrieval situation.…”

Does retrieving a memory insulate it against memory inhibition? A retroactive interference study

“…The generalisation of retrieval-induced forgetting to independent test cues unrelated to practiced items has been interpreted as an indication that the forgetting reflects inhibition of the competing trace itself and not an interference process specific to the original cue-target association (see Anderson, 2003;Anderson & Spellman, 1995;Weller et al, 2012, for discussions). Interestingly, retrieval-induced forgetting and its key theoretical properties (strength independence, cue independence, retrieval specificity) now also have been demonstrated in another species, with reversible lesions to the rodent prefrontal cortex selectively abolishing retrieval-induced forgetting (Bekinschtein et al, 2018). These findings point to a causal role of prefrontal inhibitory control processes in this forgetting phenomenon.…”

“…These findings point to a causal role of prefrontal inhibitory control processes in this forgetting phenomenon. Correspondingly, human neuroimaging evidence not only has found a role of the prefrontal cortex in resolving competition (Kuhl et al, 2007;Wimber, Alink, Charest, Kriegeskorte, & Anderson, 2015), but it also has shown this contribution to decline over successful target retrievals, as competition is resolved and competitors are forgotten (see also Bekinschtein et al, 2018, for evidence of this in rodents).…”

“…Retrieval plays a fundamental role in shaping the fate of experience in memory, both dramatically enhancing the retention of retrieved content and disrupting retention of traces that evoke competition during retrieval (e.g., Bekinschtein et al, 2018). Recently, there has been increasing interest in how the benefits and costs of retrieval might interact with one another, leading to the important question: Does retrieving a memory render it resilient to retrieval-based forgetting processes that might be applied to the retrieved memory on later occasions?…”

“…This "darker side" of retrieval, known as retrieval-induced forgetting, is believed to be produced, in part, by inhibitory control processes that isolate the desired trace in memory and that ultimately shape how accessible memories are. Taken together, retrieval's positive and negative effects suggest that this process shapes the state of memory adaptively, according to patterns in its use (e.g., Bekinschtein, Weisstaub, Gallo, Renner, & Anderson, 2018).…”