Behavioural significance of hippocampal theta oscillations : Looking elsewhere to find the 2 right answers 3 4 5

Abstract

23 The function of hippocampal theta oscillations has been subjected to constant speculation. 24 Dynamic coupling of theta field potentials and spiking activity between the hippocampus 25 and extra-hippocampal structures emphasizes the importance of theta-frequency 26 oscillations in global spike-timing precision in the brain. Recent advances in 27 understanding theta coupling between distant brain structures are discussed and explored 28 in this review. 29 30 Theta oscillations are prominent local field potential signals recordable in many parts of 31 the brain, particularly in the rodent hippocampus. Many support the notion that theta 32 oscillations, as well as other forms of brain oscillatory activity, may participate in 33 contextual binding and the exchange of information in the brain to allow emergent 34 properties such as complex behaviour and cognition to occur. Accumulating evidence has 35 shown that structures outside of the hippocampus exhibit locally generated theta 36 oscillations and these oscillations may undergo coupling with the hippocampus during 37 specific epochs of behaviour. However, very little effort has been focused on the cellular 38 correlates of such transitions. Over the past year, significant contributions from different 39 laboratories have provided insights to the importance of theta oscillations within and 40 beyond the hippocampus, and how theta rhythms mediate the exchange in information 41 with other regions of the brain in a behaviourally meaningful manner. 42 The seminal account of hippocampal theta oscillations by Green and Arduini 43 (1954) triggered the surge of interest on the subject. With more than five decades of 44 studying theta oscillations in the hippocampus, many more behavioural correlates have 45 been described (see Buzsaki 2005 for a review), yet there is no consensus regarding the 46 behavioural relevance of hippocampal theta oscillations. Robert Miller proposed that 47 theta oscillations are crucial in the coordination of behaviour through cortico48 hippocampal dialogues (Miller 1991). The evidence for cortico-hippocampal interactions 49 through theta oscillations was scarce at the time, with many accounts of recordings 50 outside of the hippocampus deemed to be solely volume-conducted from the 51 hippocampus itself. The interest in cortico-hippocampal interactions through theta 52 oscillations was re-kindled when Siapas and colleagues documented the entrainment of 53 prefrontal cortical neurons to ongoing hippocampal theta oscillations (Siapas et al. 2005), 54 which was later found to be correlated with decision making paradigm (Jones and Wilson 55 2005). At around the same time, behaviour-dependent theta coupling between the 56 hippocampus and many other brain regions, such as the amygdala (Seidenbecher et al. 57 2003), were also described in the literature. These observations have revitalized the quest 58 to understand the role of theta oscillations in the hippocampus, as well as their role in 59 other brain areas where theta field potential and entrained cell activity can be recorded. 60 In many studies it has been reported that extra-hippocampal theta oscillations can 61 occur independent of hippocampal oscillations, and vice-versa, in brief periods. These 62 observations suggest that brain regions which oscillate at theta frequencies do not do so 63 in an all-or-none fashion; therefore mechanisms must exist to control not only periods of 64 oscillatory synchrony in local circuits, but also how inter-regional oscillatory synchrony 65 waxes and wanes as a function of behaviour. For example, it is known that when rats 66 need to make a correct choice in a T-maze working memory task, theta oscillations in the 67 hippocampus and the prefrontal cortex can become synchronized, with prefrontal cell 68 activity also entrained to ongoing hippocampal theta oscillations (Jones and Wilson 69 2005). The described interaction has been implicated in the transfer of information from 70 the hippocampus to the neocortex as part of the consolidation process. However, it has 71 also been reported that hippocampo-prefrontal theta coherence can be observed during 72 spontaneous behaviours in the open field, increasing between behavioural transitions 73 from immobility, ambulation to rearing (Young and McNaughton 2009). So, does 74 hippocampo-prefrontal synchrony reflect some internal construct such as attention, or 75 does it reflect an experience-dependent process that establishes hippocampo-cortical 76 Hebbian ensembles? The recent study by Benchenane and colleagues (2010) set out to 77 answer these questions by recording ensemble and field potential activities from the 78 medial prefrontal cortex of the rat, while correlating these recordings with simultaneously 79 recorded field potentials from the intermediate/ventral hippocampus, which directly 80 projects to the medial prefrontal cortex. The authors were able to show that a Y-maze 81 based choice task evokes hippocampo-prefrontal synchrony in the central arm, prior to 82 the decision making point, as shown previously (Jones and Wilson 2005). They were able 83 to further the observation by demonstrating that when a new rule had to be learned, theta 84 coherence between the hippocampus and prefrontal cortex increased. The increased 85 synchronization is reflective of synchronization at the single cell level, where the phase 86 preference and entrainment increased after the acquisition of the new rule. Importantly, 87 not all cells displayed this increased entrainment to theta oscillations. The selective 88 population of pyramidal cells that were entrained by theta oscillations appear to form a 89 functional ensemble that displays increases in correlated activity during theta coherence, 90 increased probability of correlated firing after learning a new rule, and preference of co91 reactivation during sleep. To examine how hippocampal inputs entrain prefrontal theta 92 coherence at the single cell level, the authors used parameters based on spike waveforms 93 they recorded to classify whether the recorded cells were presumed principal cells or 94 interneurons. Consistent with the idea that interneurons shape local network activities, the 95 authors found that the presumed prefrontal interneurons always fired action potentials at 96 the same hippocampal theta phase during periods of high or low theta coherence. Instead, 97 the presumed pyramidal cells converged to the same hippocampal theta phase preference 98 in response to transition from low to high theta coherence, suggesting that the increased 99 spike-to-field coupling was due to an increase of pyramidal cell entrainment. 100 If the increase in theta coherence is related to an increased entrainment of 101 presumed principal cells but not local interneurons, is the increased pyramidal 102 entrainment due to increased synaptic input from the hippocampus? Or is another 103 mechanism at play that can account for such an increase? Since the behavioural task in 104 the study is reward-based, Benchenane et al. (2010) hypothesized that the reward-related 105 release of dopamine may drive theta neural synchrony. Remarkably, dopamine injection 106 in anesthetised animals brought upon the same changes observed in behaving animals 107 when they transitioned into the decision area – increased theta synchrony and the 108 entrainment of pyramidal cells through phase reorganization without modification of 109 interneuron response. With virtually no change in theta power recorded from either 110 structure, it is perhaps reasonable to assume that dopamine did not change the net 111 synaptic currents at theta frequencies in either structure. Then, it is likely that dopamine112 mediated changes are modifying how pyramidal cells respond to interneuronal 113 modulation but not increasing afferent synaptic inputs, since there was no detectable 114 increase in theta entrainment of local interneurons and no changes in local theta 115 oscillation power. 116 In this single study, Benchenane and colleagues were able to demonstrate that 117 increases in theta oscillation coupling between the hippocampus and the prefrontal cortex 118 can be experience-dependent, and are gated by the neuromodulator dopamine. 119 Hippocampal input to the prefrontal local interneurons appears to provide theta rhythmic 120 inhibition, and the selective synchronization of specific principal cell ensembles during 121 periods of increased dopamine. This selective recruitment of principal cells during 122 decision making is strengthened by re-activation during slow-wave sleep, providing 123 support for the formation of a cortico-hippocampal Hebbian ensemble. 124 As mentioned, the hippocampus and the prefrontal cortex are only two of many 125 brain structures that exhibit theta oscillation and entrained cell activities. The amygdala 126 also exhibits theta oscillations that becomes coherent with hippocampal theta oscillations 127 during conditioned freezing in fear conditioning paradigms (Seidenbecher et al. 2003). 128 Specifically, it was demonstrated that hippocampal and lateral amygdala theta 129 oscillations became highly correlated during the presentation of the conditioned fear 130 stimulus, which elicited freezing behaviour. Do the hippocampo-amygdalar theta 131 interactions mirror those observed between hippocampo-prefrontal interactions? Popa 132 and colleagues examined hippocampal, amygdalar and prefrontal theta field potential 133 synchrony in a fear conditioning paradigm (Popa et al. 2010). Recordings were made 134 before and after the conditioning, across different behavioural states, including sleep. The 135 authors found intermittent coherent theta oscillations and field potential entrained theta136 rhythmic spiking across all structures examined. Correlating theta coherence across all 137 examined structures with the strength of footshock conditioning, it was shown that 138 increases in coherence were selectively increased during paradoxical sleep between the 139 amygdala and the hippocampus or the prefrontal cortex, but not between the 140 hippocampus and the prefrontal cortex. Using a multivariate approach, Popa and 141 colleagues applied Granger causality analysis to examine possible dynamic interactions 142 between coherent theta oscillations recorded from the structures of interest. Granger 143 causality analysis is a way to estimate how much predictive value one signal holds over 144 another; hence, it is a statistical method that provides clues to the directionality of 145 interactions between areas. By using this approach, the authors reported that, as a general 146 rule, hippocampal theta oscillations seem to drive theta oscillations in the basolateral 147 amygdala and the prefrontal cortex. However, when the directionality of theta interaction 148 during paradoxical sleep was examined based on the amount of freezing during fear 149 recall tests across all animals, it appeared that the hippocampus preferentially drives the 150 basolateral amygdala, which then in turn drives the prefrontal cortex at theta frequencies 151 without detectable correlates between hippocampo-prefrontal interactions. 152 The results from Popa et al. (2010) show the consolidation of fear-related memory 153 may be preferentially cemented during paradoxical sleep rather than slow wave sleep, 154 and such interaction may be initiated from the hippocampus, relayed through the 155 amygdala and finally routed to the prefrontal cortex. However, only theta frequency 156 interactions were investigated in this study; therefore, it is possible that hippocampo157 prefrontal interactions during sleep may occur selectively during sharp wave/ripple 158 activities, which may also contribute to the consolidation of long-term fear memories 159 (Benchenane et al. 2010; Quinn et al. 2008). These results also suggest that apart from 160 dopamine, other neuromodulators and/or mechanisms may be involved in gating theta 161 synchrony in general, since at least during paradoxical sleep, increased hippocampo162 amygdalar or amygdalo-prefrontal theta synchrony can increase independent of 163 hippocampo-prefrontal interactions which appear to be dependent on dopamine. 164 As the two discussed studies have shown, to truly understand the functional 165 significance of hippocampal theta oscillations in the context of brain circuitries, it is 166 important to understand how oscillatory coupling is regulated and controlled amongst all 167 components of a functional circuit. To achieve this, novel techniques to perturb 168 neuromodulator transmission and ways to simultaneously sample from as many structures 169 from a functional circuit is necessary. Dzirasa and colleagues have pioneered this 170 approach by developing transgenic animals with impaired neuromodulator transmission 171 and obtaining multi-site recordings using a multi-electrode array. In their most recent 172 report, a noradrenergic depletion in freely moving mice was achieved by injecting a drug 173 that inhibits catecholamine synthesis into a transgenic line that lacked noradrenaline 174 transporter (Dzirasa et al. 2010). Using this approach, the authors were able to reduce the 175 availability of noradrenaline in the animal to <5% compared to wild-type animals. With 176 this manipulation and recording from ten interconnected structures in the mesolimbic 177 circuitry, the authors show complex, bi-directional changes in single unit firing rates, as 178 well as theta and delta oscillatory coherence between the recorded areas. These observed 179 changes were coupled with behavioural anomalies, such as hyperactivity and stereotypy. 180 Remarkably, treatment with noradrenergic precursors or catecholamine re-uptake 181 inhibitors attenuated some of the behavioural abnormalities, as well as partially reversing 182 delta/theta coherence changes in select mesolimbic circuits examined brought upon by 183 the noradrenergic depletion challenge. Although the relationship between single-cell and 184 field potential activities in these areas and the changes induced by noradrenergic 185 depletion were not examined, the study demonstrates that it is possible to probe the role 186 of modulators in gating coupled activities simultaneously in many distant but functionally 187 interacting circuits. 188 Collectively, these studies suggest that neuromodulators play a crucial role in 189 gating theta-related coupling between brain structures across different behavioural states. 190 The flexibility and control of inter-regional theta coupling mediated by neuromodulators 191 may underlie the fast dynamics of consciousness and cognition. Of course, the 192 assumption that field potential level coupling is meaningful is based on its usefulness as a 193 measure of local ensemble activity. Volume conduction can be problematic when 194 recording from neighbouring brain areas or in the vicinity of a strong oscillatory current 195 source (Sirota et al. 2008); hence it is of great importance to demonstrate a clear and 196 robust spike-to-field relationship to demonstrate the relevance of local field potential as a 197 second order measure of local ensemble activity (e.g. Popa et al. 2010), especially if such 198 relationships have not been demonstrated previously. Even when these relationships are 199 established, it is important to elucidate what activity changes are brought upon the 200 different cell types in the recorded ensemble in order to understand the cellular relevance 201 of oscillatory coupling in behaviour (e.g. Benchenane et al. 2010). 202 These recent advances in hippocampal theta oscillation research have allowed us 203 to truly integrate and understand the function of the hippocampus as part of the brain, 204 instead of an omnipotent integrator that is assumed to carry out all higher order functions 205 of the brain. The characterization of new brain regions that also oscillate at theta 206 frequencies is a growing field of research. It is likely that as more studies on behavioural 207 correlates of theta coupling across structures emerge, more behavioural correlates will be 208 found for increased theta coherence between the same structures, as has been done in the 209 hippocampus. Then, efforts must be focused on examining how theta-coupling are gated, 210 modulated and coordinated amongst structures involved in order to understand how brain 211 states emerge. Of equal importance, understanding how the activity of different cell types 212 change in relation to the ongoing field potential as a function of the strength of inter213 regional coupling will reveal the role of local computation during whole-brain 214 synchronization. The papers reviewed here represent the first important step towards the 215 goal of not only understanding the significance of the hippocampal oscillations, but also 216 technical and theoretical advances that could provide a clearer picture as to how 217 information is exchanged and processed in the brain. 218 219 Benchenane K, Peyrache A, Khamassi M, Tierney PL, Gioanni Y, Battaglia FP, and 220 Wiener SI. Coherent Theta Oscillations and Reorganization of Spike Timing in the 221 HippocampalPrefrontal Network upon Learning. Neuron 66: 921-936, 2010. 222 Buzsaki G. Theta rhythm of navigation: link between path integration and landmark 223 navigation, episodic and semantic memory. Hippocampus 15: 827-840, 2005. 224 Dzirasa K, Phillips HW, Sotnikova TD, Salahpour A, Kumar S, Gainetdinov RR, 225 Caron MG, and Nicolelis MA. Noradrenergic control of cortico-striato-thalamic and 226 mesolimbic cross-structural synchrony. J Neurosci 30: 6387-6397, 2010. 227 Green JD, and Arduini AA. Hippocampal electrical activity in arousal. J Neurophysiol 228 17: 533-557, 1954. 229 Jones MW, and Wilson MA. Theta rhythms coordinate hippocampal-prefrontal 230 interactions in a spatial memory task. PLoS Biol 3: e402, 2005. 231 Miller R. Cortico-Hippocampal Interplay and the Representation of Contexts in the 232 Brain. Heidelberg: Springer-Verlag, 1991. 233 Popa D, Duvarci S, Popescu AT, Lena C, and Pare D. Coherent amygdalocortical 234 theta promotes fear memory consolidation during paradoxical sleep. Proc Natl Acad Sci 235 U S A 2010. 236 Quinn JJ, Ma QD, Tinsley MR, Koch C, and Fanselow MS. Inverse temporal 237 contributions of the dorsal hippocampus and medial prefrontal cortex to the expression of 238 long-term fear memories. Learn Mem 15: 368-372, 2008. 239 Seidenbecher T, Laxmi TR, Stork O, and Pape HC. Amygdalar and hippocampal theta 240 rhythm synchronization during fear memory retrieval. Science 301: 846-850, 2003. 241 Siapas AG, Lubenov EV, and Wilson MA. Prefrontal phase locking to hippocampal 242 theta oscillations. Neuron 46: 141-151, 2005. 243 Sirota A, Montgomery S, Fujisawa S, Isomura Y, Zugaro M, and Buzsaki G. 244 Entrainment of neocortical neurons and gamma oscillations by the hippocampal theta 245 rhythm. Neuron 60: 683-697, 2008. 246 Young CK, and McNaughton N. Coupling of Theta Oscillations between Anterior and 247 Posterior Midline Cortex and with the Hippocampus in Freely Behaving Rats. Cereb 248 Cortex 19: 24-40, 2009. 249 250 251

Cite this paper

@inproceedings{Young2011BehaviouralSO, title={Behavioural significance of hippocampal theta oscillations : Looking elsewhere to find the 2 right answers 3 4 5}, author={Calvin K. Young and Robert F. Miller}, year={2011} }