, 2011 and Goossens et al., 2006), with the two motion types being known to involve distinct circuitries both at single-cell as well as at regional levels (Duffy and Wurtz, 1995, Gu et al., 2008, Morrone et al., 2000, Royden and Vaina, 2004 and Zhang et al., 2004). A comparison of results from prior pursuit studies using 3D flow stimuli with our findings suggests that partly distinct neural substrates support the integration of pursuit eye movements with 2D planar motion versus 3D expansion flow (Morrone et al., 2000 and Royden and Vaina, 2004). Although our results are compatible with the presence of distinct functional

units responsive to heading either in retinal or in head-centered frames of reference find more in V5/MT, MST, V3A, and V6 (Arnoldussen et al., 2011, Chukoskie and Movshon, 2009 and Ilg et al.,

2004), they indicate drastic imbalances across regions in context of planar motion integration. Our results show that V3A and V6 are heavily involved in the integration of planar motion signals with eye movements, whereas previous human studies have not reported systematic regional differences for pursuit integration during heading-related forward motion (Arnoldussen et al., 2011). One reason why distinct neural substrates may be involved in integrating extraretinal AZD8055 molecular weight signals with planar retinal motion or with more complex retinal motion types could, in theory, be explained by the following reasoning. An efference copy most likely only contains information about planar speed—this can in principle be integrated with retinal planar speed signals directly, without further computations. As soon as any other motion component (such as 3D forward flow, or other types of relative motion) is contained within Sodium butyrate retinal motion, the calculations would likely become

more complex, involving for example an initial estimation (or parsing) of the planar component embedded in the complex motion, followed by its comparison with the efference speed signal. Because V6 is highly specialized for both, 3D flow processing (Cardin and Smith, 2010, Cardin and Smith, 2011 and Pitzalis et al., 2010) and, as shown here, for 2D planar objective motion estimation, it is a good candidate region for the aforementioned function of parsing 2D signals from complex stimuli containing 3D and 2D motion cues. The results of our experiment 3 (Figure 6) are consistent with this, though at uncorrected levels, extending the previous literature in suggesting that V6 has access to 2D planar velocity in complex stimuli also containing 3D flow, allowing it to discriminate self-induced from objective 2D planar motion components even in complex stimuli. The putative human VPS homolog, identified here based on its general motion response, anatomy, and previous studies (Lindner et al., 2006 and Trenner et al.