PSY290 Lecture 6: Learning & Memory PDF

Important note This lecture will open with a short guest presentation by Sunny Choi The guest lecture will be recorded Slides for the guest lecture will not be posted, but a study guide will be provided Guest lecture material will be included on the final exam, but the questions ask...

Important note This lecture will open with a short guest presentation by Sunny Choi The guest lecture will be recorded Slides for the guest lecture will not be posted, but a study guide will be provided Guest lecture material will be included on the final exam, but the questions asked will be very closely related to the study guide questions 1 Lecture 6: Learning and Memory Paul Whissell, Ph.D. Behavioral Neuroscience (PSY290) Reference: Chapter 11 2 Overview Part 1: Neural Basis of Learning Neuroplasticity LTP Part 2: Neural Basis of Memory Types of memory The memory trace Representation of memory in the brain: focus on HPC + PFC Part 3: Manipulating Memory Enhancing, erasing and modifying memory in the laboratory 3 Neural basis of learning Learning is a relatively permanent change in behavior resulting from experience If the brain generates behavior, then learning should involve a relatively permanent change in the brain (structure and/or function) Today we will talk about how the brain changes with experience (neuroplasticity) Neuro = pertaining to the nervous system, plasticity = the quality of being easily shaped or molded 4 Sequence of events Learning stimulus Neuronal activity Intracellular signaling pathways Gene expression Protein expression Structural and functional changes 5 Neuroplasticity is ubiquitous 6 Neuroplasticity Baseline (When a Brain Activity during Long-term result Person is Doing Experience (e.g. (weeks to years later) Nothing Special) Studying) The areas that are engaged by an experience may change over time. 7 At the “micro” level… Synapses are modifiable. Changes in the strength of synapses might be important for many behaviors. 8 Hebb’s Postulate 1 Neuron A Neuron B When an axon of cell A is near enough to excite a cell B and repeatedly or persistently takes part in firing it, some growth process or metabolic change takes place in one or both cells such that A's efficiency, as one of the cells firing B, is increased Hebb. 1949. Organization of Behavior. 9 Synaptic strengthening Compelling evidence first obtained by Bliss + Lomo (1973) within the hippocampus They showed that high-frequency stimulation of synaptic connections lead to a persistent increase in their strength This phenomenon was called long-term potentiation (LTP) and has since become a favored model for learning 10 What is LTP? After this form of LTP, the same stimulation from Neuron A generates a bigger response (EPSP) in Neuron B. 11 Experimentally measuring LTP 12 Experimentally measuring LTP 3. Brief, large EPSP increase immediately after stimulation 4. Lasting, smaller but still significant EPSP increase long after stimulation 1. EPSP before stimulation 2. Tetanic Stimulation Field excitatory post-synaptic potentials (EPSPs) represent the collective response of a large population of neurons (i.e. many EPSPs). 13 Consequences of stimulation Presynaptic Glutamate Terminal 1. Activation of AMPA receptor NMDA by Glutamate 4. Mg2+ blockade receptor Mg2+ relieved AMPA receptor 7. Increased Receptor 3. Depolarization Expression 2. Cation (Na+) influx 5. Cation 6. Intracellular (Ca2+) influx Cascades Post-synaptic Terminal DNM This is one model of LTP. There are many others. 14 Consequences of stimulation Changes in protein expression occur likely due to epigenetic modifications (see L02) Neuronal stimulation in a learning episode might also activate transcription factors, which in turn regulate the expression of many genes Jarome and Lubin. 2014. Neurobiol Learn Mem. 15 Features of LTP Experimentally induced by high-frequency stimulation Conditions similar to those created by learning stimuli? Long-lasting (hours to weeks), which makes it a suitable candidate for long-term memories Correlated with memory in many animal models When strong, learning/memory tends to be strong When weak, learning/memory tends to be relatively poor Many forms (though post-synaptic is most studied) 16 LTP is a widespread phenomenon LTP is seen in many parts of the nervous system, including the cortex, striatum and spinal cord Activity-dependent variations in synaptic strength such as LTP may be a fundamental mechanism by which we acquire and modify all behaviors Even behaviors such as pain, motor learning and addiction may involve changes in synaptic strength 17 Is LTP all there is? Many other forms of experience-dependent plasticity exist, including long-term depression (LTD) LTD is a persistent reduction in synaptic strength and is generally induced by very prolonged weak stimulation LTD might serve a variety of functions Role in learning1-3 Resetting synapses (prevents saturation, allows re-use) Degradation of less useful synapses4 Piochon et al. 2014. Nat Comm. Kemp and Manahan-Vaughan. 2007. Trends Neurosci. Piochon et al. 2014. Nat Comm. Wiegert and Oertner. 2013. PNAS. 18  These changes are in animals. Do we have any evidence for changes in humans? 19 Taxi Drivers Learning Streets What type of neuroplasticity might underlie learning this complex spatial information? 20 Taxi Drivers + the HPC Woollett and Maguire. 2011. Curr Biol. Maguire et al. 2000. PNAS. 21 What types of changes occur? Large scale (many cells) Neuronal changes Visible in neuroimaging (micro) (macro) GM = Gray matter WM = White matter 22 1. Fox et al. 2014. Neurosci Biobehav Rev. Functional differences In this study, a person was gradually taught a new language (from Session 1 to Session 3) As they learned this language, they begun to exhibit a neural response to abnormal sentences in that language Osterhout et al. 2008. J Neuroling. 23 Review: Sequence of events Learning stimulus Neuronal activity Intracellular signaling pathways Gene expression Protein expression Structural and functional changes 24 Part 2: Memory 25 Memory Process whereby information is stored, consolidated + retrieved Several types (sensory, short-term and long-term) 26 (Re)constructing Memory Memory is not passively retrieved but actively assembled We are not ‘seeing the past’ but instead building a mental representation of it This process is influenced by our goals, expectations, knowledge and schemas This process is inaccurate 27 Memory reconsolidation 28 Memory in the brain Your brain has ~86 billion cells A given memory may be mapped on to a subset of these cells (memory trace/engram) The engram is believed to include the cells that were active during the original experience “Neurons that fire together, wire together” Stronger connections might reflect stronger memories Recall of the memory might involve re-activation of the original engram 29 All neurons Engram 1 These engrams do not overlap. This is not necessarily the case in the real brain! Engram 2 30 Why only some cells in the trace? Cells vary in excitability and plasticity over time, with some being highly excitable and highly plastic at a particular time The excitability and plasticity at the time of experience is critical Cells more excitable/plastic at the time of experience are more likely to be included in the engram ‘Winner take all’ model 31  Most excitable cells at Time 1 All neurons Engram for Experience 1 32 If this theory is correct, then varying neuronal excitability and plasticity should affect the composition of the engram. How can we test this? 33 From earlier… Events that change behavior (i.e. induce learning) due so by altering gene expression Gene expression is regulated by transcription factors, which are activated by events associated w/learning One transcription factor of interest is CREB (cyclic AMP-response element binding protein) 34 Focusing on CREB Cellular events associated w/learning activate CREB CREB then activates other genes, altering overall protein expression If we artificially increase CREB expression, we might affect the chance a neuron is included in the engram 35 Experimental support Neurons overexpressing CREB are more active during fear memory training Killing CREB-overexpressing neurons after the fear training impairs the fear memory 36  If two memories (M1, M2…) are acquired around the same time, they are likely to involve similar neurons. 37 Linked and non-linked memories Possible mechanism by which we recall related memories at once? There are other implications too – Josselyn and Tonegawa. 2020. Science. 38 Consequences of linked memories Because these two memories involve the same cells, changing one memory changes the other! Yetton et al. 2019. Front Human Behav. 39 Where are memory traces found? 40 The search for the trace If memory was localized to a particular region, damage to that region should impair memory Lashley studied the effects of cortical lesions on memory While the extent of overall cortical damage was associated with memory loss, no particular region was critically important Suggests no single localization 41 The search for the trace However, Lashley’s experiments only considered: the cortex, little address of subcortical areas (e.g. striatum) one type of memory (i.e. maze performance) Later studies did indeed show an interesting role of certain brain structures in memory, though in a more complex way One key brain area involved in memory is the hippocampus 42 HM and the Hippocampus Hippocampus and adjoining areas surgically removed to treat epilepsy Had anterograde amnesia from the point of injury 43 Amnesia 44 Types of memory IMPAIRED IN HM 45 LTM – Declarative memory Episodic = A person’s unique memory of an event from their perspective “REMEMBERING” Semantic = General knowledge that anyone could know (almost like trivia) “KNOWING” 46 Standard Memory Consolidation Recent events: Remote events: Hippocampus involved Hippocampus NOT involved Frankland and Bontempi. 2005. Nat Rev Neurosci. 47 Problems with SCT Recent data conflicts with the central proposal of standard consolidation theory (i.e. that remote memories reside in the cortex) Other patients with hippocampal damage (not HM) showed retrograde amnesia (sometimes ~3 years before injury) Why would this retrograde memory loss occur if all remote memories were in the cortex? To address these issues, a new theory was proposed 48 Multiple Trace Theory (c. 1997) Each time a rich, detailed (i.e. episodic) memory is recalled, the hippocampus lays down a new trace of it T1 reactivation T2 reactivation T3 reactivation 49 SMCT vs. MTT 50 Summary – HPC in Memory The hippocampus is critical for memory acquisition, but is likely a gateway site rather than a storage site Hippocampal involvement changes over time; initially being high and declining over time According to consolidation theory, the hippocampus is involved in recent but not remote memories According to multiple-trace theory, the hippocampus is important for recent and some remote episodic memories 51 The hippocampus is not required for all forms of memory. 52 Types of memory PARTIALLY INTACT IN HM 53 Recognition memory processes Linked to the perirhinal cortex In animals, perirhinal cortex lesions impair recognition 54 Visual versus Spatial memories Perirhinal cortex - Recognition Hippocampus - Space 55 Depending upon the type of memory and the age of the memory, different brain regions might be involved. 56 Some big questions… YES and YES. Can you erase a Can you implant memory? modify a memory? https://blog.ted.com/8-classic-movies-about-memory-manipulation-and-how-they-inspired-real-neuroscience/ 57 Part 3: Manipulating Memory 58 How? Active Memory (in use) Reconsolidation Learning Short-term Long-Term Stimulus Memory Memory Acquisition Consolidation 59 1 – Improving memory Activation of certain brain areas (e.g. entorhinal cortex) using electrodes can facilitate spatial memory Suthana et al. 2012. NEJM. 60 1 – Improving memory Neuroprosthesis is the use of direct electrical stimulation (via implanted devices) to affect behavior and perhaps enhance memory1,2 Non-invasive measures (e.g. TMS) could work too Effects are subtle3 Focus need not be limited to memory networks and could target other systems (e.g. attentional networks) Post-training drug treatment might also work (a consolidation effect)4 1. Berger et al. 2011. JNE. 2. Hampson et al. 2018. JNE. 3. Bengemann et al. 2020. Psychol Med. 4. LaLumiere et al. 2003. J Neurosci. 61 2 – Erasing a fear memory engram Josselyn and Tonegawa. 2020. Science. 62 3 – Modifying memories Our memory for aversive experiences is generally good (certain details anyway) However, aversive memories can be highly disruptive to our lives (as in post-traumatic stress disorder/PTSD) Can we treat PTSD by altering our aversive memories? If we appreciate how memory processing works, there may be an opportunity for intervention 63 3 – Modifying memories Drug treatment Active Memory (in use) Reconsolidation Recall/Retrieval Learning Short-term Long-Term Stimulus Memory Memory 64 3 – Modifying memories Arousing memories (e.g. of terrifying events) are notably long-lasting* One reason could be that the physiology of arousal enhances memory storage Specifically, adrenergic signaling during arousing states might facilitate the storage of arousing memories If this is true, would blockage of adrenergic signaling might change the characteristics of memory 65 3 – Modifying memories For more on adrenaline, revisit L01-02 (stress) A β-adrenergic receptor blocker (propanolol) applied during reactivation reduces PTSD symptoms Brunet et al. 2018. Am J Psy. 66