, 1997; Trotter & Celebrini, 1999; Rosenbluth & Allman, 2002; Durand et al., 2010) and in the posterior parietal cortex (Andersen et al., 1985; Andersen, 1995; Xu et al., 2012) are modulated by gaze direction via gain control mechanisms (termed the

gain fields). However, modulation of neuronal responses by gaze-dependent EX 527 molecular weight gain fields cannot explain our results, because the spatiotopic learning effect completely transfers to untrained gaze directions as long as the trained stimulus relation remained unchanged (Zhang & Li, 2010). It has been proposed that the gain control mechanisms can be used to transform a retinotopic Tacrolimus cell line map into a spatiotopic one (Zipser & Andersen, 1988; Salinas & Thier, 2000; Pouget et al., 2002), whereby neuronal representation of a stimulus becomes independent of its retinal location. Neurons with such spatiotopic properties have been found in the parietal cortex (Galletti et al., 1993; Duhamel et al., 1997). Some imaging and psychophysical studies even suggest the presence of spatiotopic maps in visual cortical areas processing motion (Melcher & Morrone, 2003; d’Avossa et al., 2007; Crespi et al., 2011; Turi & Burr, 2012) and form (Melcher, 2005) information. However, several lines of evidence actually

argue for an absence, in the visual cortex, of any explicit spatiotopic representation that is independent of stimulus location on the retina (Gardner et al., 2008; Wenderoth & Wiese, 2008; Knapen et al., 2009, 2011; Morris et al., 2010; Ong & Bisley,

2011; Golomb & Kanwisher, 2012). Our finding that the learning-induced spatiotopic effect depended on the trained retinal location also argues against an explicit spatiotopic map for processing simple stimulus attributes. Instead, the retinotopic dependence of the spatiotopic learning effect and its orientation dependency suggest that the underlying spatiotopic processing is directly based on a retinotopic map; but how is this process accomplished? CYTH4 Considering that the first stimulus in our experiments was followed by a saccade, one might speculate that the spatiotopic learning effect and its retinotopic dependency might involve peri-saccadic updating of the visual representation of the first stimulus on a retinotopic map. Such a transient spatiotopic mechanism, which has been reported in the parietal (Duhamel et al., 1992; Merriam et al., 2003), frontal (Sommer & Wurtz, 2006) and even visual (Nakamura & Colby, 2002; Merriam et al., 2007) cortical areas, enables updating of a visual stimulus from one retinotopic location to another around saccadic eye movements.