Functions of gamma-band synchronization in cognition: from single circuits to functional diversity across cortical and subcortical systems Conrado A. Bosman, 1,2 Carien S. Lansink 1,2 and Cyriel M. A. Pennartz 1,2 1 Cognitive and Systems Neuroscience Group, Center for Neuroscience, Faculty of Science, Swammerdam Institute for Life Sciences, University of Amsterdam, Postal Box 94216, 1090, GE Amsterdam, The Netherlands 2 Research Priority Program Brain & Cognition, University of Amsterdam, Amsterdam, The Netherlands Keywords: behavioral control, evolution, gamma-band synchronization, memory, selective attention, visual perception Abstract Gamma-band activity (30–90 Hz) and the synchronization of neural activity in the gamma-frequency range have been observed in different cortical and subcortical structures and have been associated with different cognitive functions. However, it is still unknown whether gamma-band synchronization subserves a single universal function or a diversity of functions across the full spectrum of cognitive processes. Here, we address this question reviewing the mechanisms of gamma-band oscillation generation and the func- tions associated with gamma-band activity across several cortical and subcortical structures. Additionally, we raise a plausible expla- nation of why gamma rhythms are found so ubiquitously across brain structures. Gamma band activity originates from the interplay between inhibition and excitation. We stress that gamma oscillations, associated with this interplay, originate from basic functional motifs that conferred advantages for low-level system processing and multiple cognitive functions throughout evolution. We illustrate the multifunctionality of gamma-band activity by considering its role in neural systems for perception, selective attention, memory, motivation and behavioral control. We conclude that gamma-band oscillations support multiple cognitive processes, rather than a sin- gle one, which, however, can be traced back to a limited set of circuit motifs which are found universally across species and brain structures. Introduction Neuronal oscillations have been observed since the advent of elec- troencephalographic recordings (Berger, 1929; Buzs aki, 2006). Oscillations have been traditionally grouped in several frequency bands and these bands have been associated with different brain functions and mechanisms of generation (Buzs aki & Draguhn, 2004). There is broad consensus that, in general, different types of oscillation denote different brain activity states and that oscillatory fluctuations across time are representative of the dynamic interplay between different cell types in various cortical and subcortical cir- cuits (Buzs aki, 2006). Here, we will focus on a particular frequency band denoted as gamma (30–90 Hz). The concept of phase synchro- nization indicates the presence of a consistent phase relation (i.e. phase coupling) between two neuronal signals that oscillate at a par- ticular frequency. Gamma-band synchronization (GBS) has attracted considerable interest throughout recent years because mechanistic roles have been proposed for GBS in phase coding, perceptual inte- gration and flexible routing of information in the visual system (Fries, 2009; Vinck et al., 2013a), and furthermore because of its widespread occurrence in several cortical and subcortical structures (Fries, 2009; van der Meer et al., 2010; Wang, 2010). Nevertheless, the question whether GBS plays a role in neural cir- cuit dynamics and cognitive functions is still a matter of contro- versy. Although GBS has been found in several species and many brain regions, a few studies have failed to find it in situations where it would have been predicted to occur, such as in visual perception tasks (Thiele & Stoner, 2003; Roelfsema et al., 2004; Palanca & Deangelis, 2005; Lima et al., 2010). Subsequently, some concerns over the relevance of GBS for coding and communication were raised (Shadlen & Movshon, 1999; Merker, 2013). Other studies confirmed the presence of GBS, but also found strong dependencies of gamma-band amplitude and peak frequency on stimulus proper- ties that are deemed less relevant for computations subserving per- ceptual integration and information routing (Gieselmann & Thiele, 2008; Ray & Maunsell, 2010; Jia et al., 2013). Because of these dependencies, it has been argued that GBS cannot be a reliable mechanism for neural coding and dynamic routing of information (Ray & Maunsell, 2010; Jia et al., 2013). This seemingly contradic- tory evidence has led some authors to consider gamma-band oscilla- tions an essential mechanism arising from neural circuitry, and consider their ubiquity as reflecting fundamentally applicable com- putational principles (Fries, 2009; Kopell et al., 2010; Vinck et al., 2013a). In contrast, others have argued that GBS is an epiphenome- non arising from neural circuit organization as a non-functional Correspondence: Conrado A. Bosman, 1 Cognitive and Systems Neuroscience Group, as above. E-mail: C.A.BosmanVittini@uva.nl Received 22 January 2014, revised 18 March 2014, accepted 3 April 2014 © 2014 Federation of European Neuroscience Societies and John Wiley & Sons Ltd European Journal of Neuroscience, Vol. 39, pp. 1982–1999, 2014 doi:10.1111/ejn.12606