Dopamine-dependent oscillations in frontal cortex index "start-gun" signal in interval timing.

Abstract

Although perceiving the passage of time is a basic building block of cognitive processes and behavior such as expecting relevant events to happen, the neural underpinnings of interval timing are not well-understood as yet (van Wassenhove, 2009; Allman and Meck, 2012; Merchant et al., 2013). From the neurobiological point of view, it has been established that dopamine impacts interval timing (e.g., Meck, 1986, 1996; Allman and Meck, 2012). However, a link between pharmacological manipulations and their impact on neurophysiological signals has been rarely investigated. The leading neurobiologically plausible model of interval timing that considers both components is the Striatal Beat Frequency (SBF) model (Mattel and Meck, 2004; Buhusi and Meck, 2005; van Rijn et al., 2014). The SBF model relies on the neuromodulatory dynamics of the thalamo-cortico-striatal loops. Although currently most of the interactions in the SBF model are assumed to be unidirectional, Mattel and Meck (2004) also acknowledged the possibility of feedback from the cortex to the neurons in the VTA as well as from striatal neurons to both the cortex and the substantia nigra pars compacta. These potential feedback mechanisms are unaddressed in the SBF model. However, their implementation would allow for more accurate description of clock speed and memory updating mechanisms on a trial to trial basis (W. Meck, personal communication, May 15, 2015). Nevertheless, the SBF assumes that time is coded by the coincidental activation of striatal spiny neurons with cortical oscillators (CO). Numerical implementations of the SBF model utilizes the phase, or amplitude, of the CO that are envisioned to oscillate at various frequencies giving rise to different amplitude patterns over time as illustrated in Figure ​Figure1.1. Hence, at a given time point a specific amplitude pattern of the CO can be encoded by striatal spiny neurons. Crucially, the SBF model assumes that, at the onset of the to-be-timed interval, the phases of CO are reset by a burst of dopaminergic input from the ventral tegmental area (VTA, Mattel and Meck, 2004). Further, the SBF model contends that the initial dopamine-triggered phase-resetting of CO by VTA plays the role of a “start-gun” that initiates timing. This “start-gun” signal forces a whole set of CO to start from the same phase, allowing for coincidence detection to read the state of CO, and code for a particular duration over multiple trials. Importantly, the more accurate the phase-reset at the onset of the to-be-timed interval—that is, the proportion of CO that is reset to the same phase—the larger the phase synchronization and the oscillatory power of ongoing oscillations (Canavier, 2015), and as such reduces variability in memory representation of to-be-timed interval (see Figure ​Figure1;1; Ng et al., 2011). Within this framework more precise phase reset should be associated with an increase in timing accuracy. Note that timing accuracy can be seen as a peak latency of a response distribution, or kurtosis of a response distribution. According to the SBF model the peak latency and kurtosis of response distribution can be accounted for by different features of the model. The peak latency is modulated by frequency range of CO whereas kurtosis is accounted for by accuracy of initial phase reset (Oprisan and Buhusi, 2014). As such the width shows consistency of memory representation estimated over a number of trials. What is referred to here is the accuracy as the width of the response distribution that is associated to initial reset in terms of the SBF model. Therefore, the width of the response distribution should be affected by consistency of phase reset. However, such covariation between the precision of the “start-gun” represented by the modulation of oscillatory synchrony in any neural population and timing accuracy has not yet been tested directly.