In the presence of such regularities, the past can help predict the future. A way to do this is to use information from the past for building a statistical model of the environment (Winkler et al., 2009). The model is then used for predicting

the future and interpreting it. Indeed, numerous studies have demonstrated sensitivity of neural activity to the overall probability of a stimulus, an important characteristic of the statistical structure of stimulation sequences. Since their introduction as a tool for studying single neurons in the auditory system by Ulanovsky et al. (2003), oddball sequences have been used to study probability sensitivity in a number of animal models and at different levels of auditory pathway, including the inferior colliculus of rats (Malmierca Alectinib ic50 et al., 2009; Zhao et al., 2011), the auditory thalamus of mice (Anderson et al., 2009) and rats (Antunes et al., 2010), and auditory cortex of rats (Farley et al., 2010; Taaseh et al., 2011; von der Behrens et al., 2009).

These studies demonstrated that the probability of appearance of a stimulus affects the responses of many neurons VE-821 mw at least to the same degree as the physical characteristics of the stimulus such as its frequency. In fact, cortical responses to rare tones embedded in sequences of common tones are larger than expected from a model of adaptation in narrow frequency channels, suggesting the presence of true deviance sensitivity in auditory cortex (Taaseh et al., 2011). Oddball sequences are most commonly constructed by selecting the sounds essentially randomly given their probabilities. However, the statistical structure of the auditory environment is richer than that of such random sequences. For example, language and music incorporate sequential dependencies, so that the probability of a sound depends much more subtly on the recent auditory past. The goal of the current study

was to examine the many sensitivity of neuronal responses to statistical contexts that include sequential dependence. We contrasted neuronal responses to sequences in which the overall probability of the rare tone was identical but the rare tone itself was either randomly presented or appeared periodically among the common tones. If the periodic order can be recognized, periodic sequences should evoke less surprise, and therefore smaller neuronal responses. Our data, from intracellular and extracellular recordings in the auditory cortex of anesthetized rats, suggest that neurons are sensitive to the periodic order of presentations, even for periods of length 20 (rare tone probability of 0.05). We recorded responses in the left auditory cortex of halothane-anesthetized rats to sounds presented monaurally to their right ear. We used both intracellular recordings (n = 17 neurons in 16 rats) and extracellular recordings (n = 180 recording locations in 12 rats) to collect membrane potentials, local field potentials (LFPs), and multiunit activity (MUA).