Information from Lay-Language Summaries is Embargoed Until the Conclusion of the Scientific Presentation
863—Animal Learning and Memory: Cortical and Hippocampal Circuits V
Wednesday, November 13, 2013, 1:00 pm - 5:00 pm
863.09: BOLD responses associated with hippocampal ripples in the rat brain
Location: Halls B-H
*O. ESCHENKO1, M. BESSERVE1,2, Y. MURAYAMA1, H. EVRARD1, M. BEYERLEIN1, A. OELTERMANN1, N. LOGOTHETIS1,3; 1Physiol. of Cognitive Processes, Max Planck Inst. for Biol. Cybernetics, Tuebingen, Germany; 2Max Planck Inst. for Intelligent Systems, Tuebingen, Germany; 3Ctr. for Imaging Sciences, Biomed. Imaging Institute, The Univ. of Manchester, Manchester, United Kingdom
Abstract Body: Hippocampal ripples, brief high-frequency oscillations, occur during behavioral states that are not associated with active sensory processing. The ripple event represents a simultaneous burst of a large neuronal population that is synchronized across the entire hippocampus. Reactivation of neuronal ensembles that were active during learning predominantly occurs during ripples. The number of ripples is increased after learning and this increase is predictive for memory recall. Ripple suppression is unfavorable for memory consolidation. Ripples have been suggested to provide a neurophysiological substrate for ‘off-line’ memory consolidation by facilitating synaptic plasticity within the learning-associated neuronal ensembles. The neuronal activity in other brain regions that is time-locked to hippocampal ripples may underlie a cross-regional information transfer. We exploited the methodology allowing simultaneous extracellular recording combined with fMRI. An anesthetized rat was fixed in the MRI scanner and MRI-compatible linear electrode array was placed with electrode contacts in cortex, hippocampus, and thalamus using a custom-made movable drive. Spontaneous whole-brain BOLD activity was acquired along with multi-site electrophysiological recording. The ripple events were detected and classified off-line using a custom software. The time series of BOLD responses were extracted for each voxel according to the event-triggered design, where the ripple onset was used as an event, and the statistical maps were generated indicating the voxels with positive and negative BOLD responses. The voxels were subsequently grouped according to the anatomical brain regions by co-registration of the functional images with the digital rat brain atlas. The positive BOLD response was detected within the direct proximity to the ripple recording site in the CA1 region of hippocampus. The most of the hippocampal volume was also co-activated. In addition, a number of cortical regions including sensory and associative cortices contained a substantial proportion of voxels showing positive BOLD responses. Several brain regions consistently showed negative BOLD responses. These included many of the thalamic nuclei, neuromodulatory nuclei of the midbrain and brain stem and cerebellum. The fMRI findings were further confirmed by electrophysiological recordings in multiple brain areas. Our results identify a brain network that possibly supports hippocampal-dependent memory consolidation. Besides, hippocampal ripples may cause a transient inhibition within competing functional networks to enable more efficient intra brain region communication.
Lay Language Summary: The brain never rests but it continuously processes information from the external and internal milieu. In order to process diverse information efficiently, multiple brain regions must work synergistically, like the musicians playing a symphony in an orchestra. But who is the orchestra director in the brain? We have recently discovered an amazing optimization principle of information transfer in the brains of primates and rodents. Specifically, we described the brain activation/deactivation patterns that create a condition of minimal interference for communication between two brain regions that are critical for creating episodic memories. Our findings provide new insights into the system mechanisms of memory formation, which can be studied in more detail in the future. New memories are highly susceptible to interference and therefore require time to stabilize. Periods of calmness or sleep are commonly considered to be beneficial for memory stabilization. It is also known that hippocampus and cortex are critical for transition of temporary, labile memories to stable representations. But what is the neural basis of this hippocampal-cortical dialog? To gain these insights, we used a novel multimodal methodology called “neural event-triggered functional magnetic resonance imaging” (NET-fMRI). This new methodology combines electrophysiological recording with fMRI of the entire brain to map widespread neuronal networks that are activated by local, structure-specific neural events. Electrophysiological recording in hippocampus allowed us to detect specific activity patterns, known as ripples, that reflect synchronous firing of 50-100,000 neurons representing recent experiences. The simultaneous fMRI enabled visualization of the activity of the entire brain associated with ripples. A prevailing memory-theory posits that new experiences are temporarily stored in hippocampus and they are subsequently reactivated during ripples to strengthen the cortico-cortical connections underlying long-term memory. Of obvious interest is the brain-state enabling such spatiotemporally specific interactions, involving hippocampus and a great number of other cortical or subcortical structures. We applied NET-fMRI to identify the brain areas that modulated their activity in relationship to ripples. We demonstrated that ripples are associated with cortical activations that occur concurrently with activity suppression in other brain structures. Interestingly, structures were suppressed whose activities could, in principle, interfere with the hippocampal-cortical communication. The suppression of activity in the thalamus, for instance, reduces signals related to sensory processing, while the suppression of the basal ganglia, the pontine region and the cerebellar cortex may reduce signals related to other memory systems, such as that underlying procedural learning, for example riding a bicycle. Capacities such as perception, attention, learning and memory are emerging properties of a complex system such as the brain and can be best investigated using multimodal methodologies. The present study demonstrates the advantage of using such multimodal methodology as the NET-fMRI. Our findings offered revealing insights into the large-scale organization of memory, a cognitive capacity emerging from the activation of widespread neural networks which were impossible to study in depth using conventional methodology. Ripples are characteristic neural events, and their use as an fMRI ‘trigger’ lead to data-driven empirical discovery. Yet, a causal relationship between the trigger event and the network activity changes should be affirmed with caution to avoid misinterpretation of the functional significance of such individual neural events. The state of widespread networks probably depends on a large number of variables (for example, activity changes in individual structures, or changes in inter-structure correlations), a subset of which may be eventually characterized following intensive experimentation. However, events in isolation are likely to be indicators rather than effectors of any cognitive capacity. In conclusion, it is difficult to overstate the importance of this new methodology as the vast majority of neurological failures actually reflect dysfunctions of large-scale neural networks.
Neuroscience 2013 (43rd annual meeting of the Society for Neuroscience)Exit