It has been 70 years since Donald Hebb published his formalized theory of synaptic adaptation during learning. Hebb’s seminal work foreshadowed some of the great neuroscientific discoveries of the following decades, including the discovery of long-term potentiation and other lasting forms of synaptic plasticity, and more recently the residence of memories in synaptically connected neuronal assemblies. Our understanding of the processes underlying learning and memory has been dominated by the view that synapses are the principal site of information storage in the brain. This view has received substantial support from research in several model systems, with the vast majority of studies on the topic corroborating a role for synapses in memory storage. Yet, despite the neuroscience community’s best efforts, we are still without conclusive proof that memories reside at synapses. Furthermore, an increasing number of non-synaptic mechanisms have emerged that are also capable of acting as memory substrates. In this review, we address the key findings from the synaptic plasticity literature that make these phenomena such attractive memory mechanisms. We then turn our attention to evidence that questions the reliance of memory exclusively on changes at the synapse and attempt to integrate these opposing views.

自唐纳德·赫布（Donald Hebb）发表他的学习过程中突触适应的形式化理论以来已有70年了。赫布的开创性工作预示了随后几十年的一些重大神经科学发现，包括长期增强和其他持久形式的突触可塑性的发现，以及最近记忆在突触连接的神经元组装体中的驻留。关于突触是大脑中信息存储的主要场所的观点主导了我们对学习和记忆的潜在过程的理解。该观点已得到多个模型系统研究的大力支持，有关该主题的绝大多数研究证实了突触在存储器存储中的作用。然而，尽管神经科学界做出了最大的努力，我们仍然没有确凿的证据表明记忆存在于突触中。此外，已经出现了越来越多的非突触机制，它们也能够用作存储基板。在这篇综述中，我们讨论了突触可塑性文献中的关键发现，这些发现使这些现象成为有吸引力的记忆机制。然后，我们将注意力转向证据，这些证据质疑记忆对突触变化的依赖性，并试图整合这些对立的观点。