Spatial and Temporal K-Complex Variability Helps Elucidate Its Potential Role in Memory Consolidation
Material below summarizes the article Distribution, Amplitude, Incidence, Co-Occurrence, and Propagation of Human K-Complexes in Focal Transcortical Recordings, published on September 2, 2015, in eNeuro and authored by Rachel A. Mak-McCully, Burke Q. Rosen, Matthieu Rolland, Jean Régis, Fabrice Bartolomei, Marc Rey, Patrick Chauvel, Sydney S. Cash, and Eric Halgren.
The K-complex represents one of the largest events recorded in the human cortex. It occurs during non-REM sleep as an isolated downstate, which may arise spontaneously or be evoked in response to a sensory stimulus. In this study, we sought to characterize the basic but essential spatial and temporal dynamics of the K-complex across the human cortex using bipolar stereoencephalographic (SEEG) recordings.
Previous human scalp EEG and average reference intracranial studies recorded the largest K-complexes in frontal midline areas. K-complexes also appeared in these recordings to occur synchronously across the cortex. Disentangling where the recorded signals are generated in the cortex, however, remains a limitation of these techniques. Therefore, in order to address where K-complexes are locally generated in the human cortex, we analyzed adjacent bipolar contact pairs that spanned the grey matter along an SEEG electrode. We demonstrate in this study that this analysis technique measures the locally generated cortical signal.
The slow oscillation also characterizes non-REM sleep and arises as a series of large amplitude upstates and downstates. Neurophysiologically, the K-complex and slow oscillation downstate are identical. Unlike the synchrony seen with K-complexes, previous human scalp EEG studies of the slow oscillation showed that slow oscillation downstates sweep across the cortex from anterior to posterior areas. A unique feature of the K-complex is that it is an isolated event, rather than an oscillation, and it is therefore possible to pinpoint the start and end of the event in order to examine propagation patterns across the cortex.
In order to distinguish K-complexes from slow oscillations, we used a two-part detection process. K-complexes were manually marked and analyzed separately. However, in order to investigate the true extent and spread of K-complex activity, these manually-marked K-complexes were also used as templates to detect K-complex-like activity. Template-detected K-complexes were smaller in amplitude and had more variance preceding them than the manually-marked K-complexes.
While K-complexes appear to be a fundamental cortical state because they are recorded all over the cortex, they arise with a large degree of variability from K-complex to K-complex. We find that K-complexes vary in terms of their amplitude and how often they occur in a particular cortical region. In contrast to previous studies, our results indicate that anterior prefrontal regions generate significantly fewer K-complexes, as well as significantly smaller amplitude K-complexes. Overall, however, K-complex amplitude and occurrence rate vary across the cortex.
K-complexes also variably arise over small or large areas of cortex. Manually-marked K-complexes occur in a single recorded location 76 percent of the time and only one percent of the time in all recorded locations. When template-detected K-complexes are analyzed in addition to the manually marked K-complexes, only 15 percent now occur in a single location while 27 percent occur in all locations. Overall, the number of locations participating in any given K-complex can vary between a small and large number of cortical locations.
In contrast to findings in scalp EEG showing an anterior to posterior sweep of slow oscillation downstates, our results indicate that K-complexes propagate in many different directions across the cortex. We find evidence, however, of a strong propagation towards lateral temporal cortex.
Additionally, K-complexes traditionally characterize non-REM stage two sleep, while slow oscillations traditionally characterize non-REM stage three sleep. We demonstrate that overall, the characteristics of K-complexes in stages two and three are very similar, except that stage three K-complexes occur more frequently than those in stage two. Our results therefore included K-complexes detected in both stages two and three.
The high degree of variability that we observe in these focal recordings of K-complexes in the human cortex indicates that the K-complex is a heterogenous event. Compared with prior interpretations of the K-complex as a synchronous event, or the slow oscillation downstate as a propagating event from anterior to posterior sites on the scalp, this heterogeneity lends important implications for the functional role of the K-complex. The K-complex is thought to coordinate the re-activation of cortical activity patterns that underlie memory consolidation. Memories are highly variable, and so they must be associated with equally variable cortical patterns. Our finding that K-complexes can variably involve small to large areas of cortex and then propagate in different directions is thus consistent with its proposed role in memory consolidation during sleep.
Distribution, Amplitude, Incidence, Co-Occurrence, and Propagation of Human K-Complexes in Focal Transcortical Recordings. Rachel A. Mak-McCully, Burke Q. Rosen, Matthieu Rolland, Jean Régis, Fabrice Bartolomei, Marc Rey, Patrick Chauvel, Sydney S. Cash, Eric Halgren. eNeuro September 2015, 2 (4) DOI: 10.1523/ENEURO.0028-15.2015