The presence of grid structure was quantified by calculating, for

The presence of grid structure was quantified by calculating, for each cell, a grid score based on rotational symmetry in the cell’s spatial autocorrelogram IWR-1 chemical structure (Sargolini et al., 2006 and Langston et al., 2010). Cells were classified as grid cells if they had grid scores and spatial information scores that each exceeded the 95th percentile of grid scores and spatial information scores, respectively, from a shuffled distribution

for the respective age group (Figure 4B). Two out of 128 cells (1.6%) passed this dual criterion in the P16–P18 group (Figure 4C). The fraction was slightly but significantly larger than in the shuffled data, where 0.2% of the cells passed both criteria (Z = 3.3, p = 0.001). In the P19–P21 group, seven out of 185 cells (3.8%) passed the dual criterion (chance level: 0.2%–0.3%; Z = 8.1, p < 0.001). At subsequent ages, the percentage of grid cells increased slowly (all p < 0.001). The percentage of cells that passed the grid cell criterion was significantly larger in the adult group than in the entire group of young animals (P16–P36; Z = 9.02, p < 0.001). Cells that passed the criterion for grid cells showed a significant increase in grid scores

check details across age blocks (Figure 4D; F(7, 82) = 3.858, p = 0.001). The stability of the grid fields increased significantly with age (Figures 4E and 4F; within trials: F(7, 82) = 6.1, p < 0.001; between trials: F(7, 82) = 11.1, p < 0.001); as did the spatial discreteness of the firing fields (ANOVA for spatial coherence: F(7, 82) = 2.9, p < 0.01; spatial information: F(7, 82) = 2.3, p < 0.05). Head direction cells were present in all age groups, in agreement with previous studies (Langston et al., 2010 and Wills et al., 2010). Directional

modulation was expressed by the mean vector length of the cell’s firing rate. Cells were classified as head direction cells if the mean vector length exceeded the 95th percentiles of shuffled distributions for both directional information and mean vector length. Fifty-five out of 128 cells (43.0%) passed the criterion for head direction cells in the P16–P18 group. This fraction is significantly larger than in the shuffled data, where 0.9% either of the cells passed both criteria (Z = 49.0, p < 0.001). The percentage of head direction cells did not increase with age (P19–P21: 40.5%; P22–P24: 34.5%; P25–P27: 29.6%; P28–P30: 25.3%; P31–P33: 34.1%; P34–P36: 35.0%, and adult: 48.8%). Cells that passed the criterion for head direction cells showed a significant increase in mean vector length across age blocks (F(7, 424) = 4.3, p < 0.001). The stability of directional tuning increased significantly (within trials: F(7, 421) = 3.8, p < 0.001; between trials: F(7, 406) = 3.6, p = 0.001). The key finding of this study is that entorhinal border cells are already present when rat pups make their first navigational experiences. When rat pups leave the nest at the age of 2.

Understanding “how” the ventral pathway achieves this requires th

Understanding “how” the ventral pathway achieves this requires that we define one or more levels of abstraction between full cortical area populations and single neurons. For example, we hypothesize that canonical subnetworks of ∼40K neurons form a basic “building block” for visual computation, and that each such subnetwork has the same meta function. Even if this framework ultimately proves to be correct, it can only be shown by getting the many interacting “details” correct. Thus, progress will result from two synergistic lines of work. One line will use high-throughput computer simulations to

systematically explore the very large space of possible subnetwork algorithms, implementing each possibility as a cascaded, full-scale algorithm, and measuring performance in carefully considered benchmark object recognition tasks. A second ATR inhibitor line will use rapidly expanding systems neurophysiological data volumes and psychophysical performance measurements to sift through those algorithms for those that best explain the experimental data. Put simply, we must synergize the fields of psychophysics, systems neuroscience, and computer vision around the problem of object recognition. Fortunately, the foundations and tools are now available to make it so. J.J.D. was supported by the U.S. National Eye Institute (NIH

NEI R01EY014970-01), The Defense Advanced Research Projects Agency (DARPA), and the National Science Foundation (NSF). D.Z. was supported by an Accademia Nazionale dei PD0325901 price Lincei-Compagnia di San Paolo Grant, a Programma Neuroscienze grant from the Compagnia di San Paolo, and a Marie Curie International Reintegration Grant. N.R. was supported by the NIH NEI and a fellowship from the Alfred P. Sloan Foundation. “
“Chronic pain is a major public health problem, with epidemiological studies reporting about one fifth of the general population to be affected both in the USA and Europe (Breivik et al., 2006). The condition is debilitating and causes not only considerable personal suffering old but also enormous socioeconomic costs, estimated to reach an annual 60 billion U.S. dollars in lost productivity. It is a figure that only stands to increase with the aging populations

of the Western world (Krueger and Stone, 2008). In addition to these bleak statistics, pharmacological management of chronic pain conditions has seen only limited progress in the last decades. Despite the seemingly bewildering array of nonprescription analgesics being advertised and sold in dedicated drugstore aisles, treatment of pain is still very much dominated by two classical medications: opioids and nonsteroidal anti-inflammatory drugs. Only a handful of compounds acting on novel, distinct molecular targets have emerged since the 1960s, for instance gabapentinoids, TprVI agonists, or cannabinoids (Kissin, 2010). Many of these painkillers have serious side effects, such as neurotoxicity (e.g., TrpVI agonists) and addictive properties (e.g.

Next, we compared the laminar distribution of Lhx6+ cells in both

Next, we compared the laminar distribution of Lhx6+ cells in both mutants and we found that they had very similar defects: reduced numbers of Lhx6+ cells in Screening Library clinical trial the MZ and SVZ, and an increase in the number of Lhx6+ cells in the CP, especially in the lower CP ( Figures 3A″–3C″ and 3E). Thus, both Cxcr7 and Cxcr4 had similar functions in maintaining interneurons within

the MZ and SVZ migratory streams and in controlling the timing for interneuron invasion into the cortical plate. While Cxcr4−/− and Cxcr7−/− mutants share very similar interneuron laminar positioning phenotypes in the cortical plate, these deficits may arise from distinct alterations in migration dynamics. To further explore the migration behaviors of Cxcr4−/− and Cxcr7−/− interneurons in vivo, we performed real-time imaging of Lhx6-GFP+ cortices from control (Cxcr4+/− or Cxcr7+/−), Cxcr4−/−, and Cxcr7−/− embryos at E15.5. We first examined the transition from tangential to radial migration as interneurons migrated from either the MZ or the SVZ into the CP. In the MZ, Cxcr4−/− and Cxcr7−/− mutants demonstrated a 3-fold and 2.4-fold increase in the number of Lhx6-GFP+ interneurons C646 chemical structure switching from tangential to radial migration, respectively ( Figures 4A–4C and 4G; Movies S1–S3). In the SVZ, Cxcr4−/− mutants displayed a 2-fold increase in the number of Lhx6-GFP+

interneurons switching from tangential to radial migration and no effect was detected in the Cxcr7−/− mutants ( Figures 4D–4F and 4H; Movies S4–S6). Therefore, the laminar deficits from both mutants were mainly due to an increased number of Lhx6-GFP+ cells moving from either the MZ or the SVZ into the cortical plate. Next, we studied the tangential

migration rate of Lhx6-GFP+ cells that were maintained in the MZ and SVZ during the 20-hour live-imaging session. Compared to controls, Cxcr7−/− mutants exhibited a substantial decrease in tangential migration rate in the MZ and SVZ; and Cxcr4−/− mutants displayed a modest decrease in migration rate in the SVZ ( Figure 4I). Thus, Methisazone the decreased migration rate may underlie the reduced extent of interneuron migration into the dorsal cortex observed in both mutants during early embryonic stages. Finally, we explored the migration behaviors of Lhx6-GFP+ cells after they entered into the cortical plate. Cxcr4−/− and Cxcr7−/− mutants displayed major differences in interneuron motility and leading process morphology. Cxcr4−/− mutants exhibited a significant increase in the number of motile cells and in the leading process length of tangentially oriented cells, while Cxcr7−/− mutants showed a significant decrease in the number of motile cells and in the leading process length of both radially and tangentially oriented cells ( Figure 5 and Movies S7–S9).

Fourth, we used two control BAC mouse lines with normal CTG/CAG r

Fourth, we used two control BAC mouse lines with normal CTG/CAG repeats to demonstrate that disease pathogenesis in BAC-HDL2 mice, including NI formation, Vismodegib concentration is dependent on the CTG/CAG repeat expansion. Finally, by genetic silencing of the JPH3 sense strand in BAC-HDL2-STOP mice, preventing expression of CUG-containing transcripts, we provided definitive genetic evidence that the expression

of the HDL2-CAG transcript alone can lead to the formation of polyQ-containing NIs and manifestation of motor deficits. Taken together, our analyses of the series of HDL2 mouse models provide an important mechanistic insight that the expression of a novel expanded polyQ protein could play a critical role in HDL2 pathogenesis in vivo. Prior to this study, the expression of an expanded antisense CAG-containing transcript or a novel polyQ protein had not been demonstrated by using postmortem patient brain tissues (Holmes et al., 2001 and Margolis et al., 2005). There are several possible explanations to account for the more sensitive detection of the CAG transcript

and polyQ-expanded protein in HDL2 mice than in patients. First, unlike the postmortem brain tissues, the HDL2 mouse brains used for the studies do not exhibit robust neurodegeneration selleck compound that may lead to the loss of neurons expressing mutant CAG transcripts at high levels. Second, the sequence difference between the human mutant JPH3 transgene and murine wild-type locus in the same HDL2 mouse permits the easier detection of the mutant antisense transcripts. Third, the much longer polyQ repeat in the BAC-HDL2 mouse model (120Q, as compared to 40–50Q in patients) also permits more sensitive detection of the small amount of soluble mutant polyQ protein with antibodies that provide signal strength in western blots depending on the polyQ length (e.g., 3B5H10).

Because the brain regional and subcellular Adenosine distribution of the NIs are remarkably similar between HDL2 mice and patients, it would strongly argue that the molecular pathogenesis for such NIs in HDL2 mice and patients is also likely to be similar. Future studies with additional patient samples and more sensitive detection methods may be needed to demonstrate HDL2-CAG transcript and protein in the patient brains. Although our study is focused on the mechanistic investigation of the novel antisense CAG repeat transcript and its polyQ-containing protein product, our study does not exclude a contribution of other potential mechanisms to aspects of HDL2 pathogenesis, such as partial loss of JPH3 function or CUG repeat RNA gain-of-function toxicity (Rudnicki et al., 2007).

, 2000 and Thannickal et al , 2000) This finding was quite selec

, 2000 and Thannickal et al., 2000). This finding was quite selective, as the MCH neurons, which are intermingled with the orexin cells, were completely spared, and it probably represented cell loss rather than downregulation of orexin expression as there was concomitant loss of other markers (dynorphin and neuronal activity-related pentraxin) of the orexin cell population (Crocker et al., 2005). The loss of orexins is not due to a simple genetic abnormality, as orexin deficiency is acquired during young adulthood, Selleckchem Screening Library and the vast majority of people with narcolepsy do not have

mutations of the genes encoding the orexin peptides or their OX1 or OX2 receptors (Olafsdóttir et al., 2001 and Peyron et al., 2000). However, because about 90% of people with narcolepsy have human leukocyte antigen DQB1∗0602 (Mignot et al., 2001), researchers have hypothesized that the loss LY2157299 of orexin neurons may be immune-mediated (Lim and Scammell, 2010 and Scammell, 2006). It has recently been proposed that, at least in some individuals, an autoimmune attack on the orexin neurons may be related to antibodies to Tribbles homolog-2, a protein produced by the orexin neurons and other cells in

the brain (Cvetkovic-Lopes et al., 2010 and Kawashima et al., 2010). Several models have been proposed to explain how loss of the orexin neurons results in severe sleepiness. One popular hypothesis is that individuals with narcolepsy may be more sensitive to homeostatic sleep drive as, after a period of sleep deprivation, they fall asleep faster than normal (Tafti et al., 1992a and Tafti et al., 1992b). Mice lacking orexins also tend to fall asleep very quickly after being deprived of sleep, but they recover the lost sleep at a normal rate and to the same extent as wild-type mice (Mochizuki et al., 2004). Thus, orexin deficiency hastens

the transition to sleep, but the accumulation and expression of homeostatic sleep drive appears normal in mice and people with narcolepsy (Khatami et al., 2008 and Mochizuki et al., 2004). Another potential explanation is that circadian waking drive is impaired in narcolepsy. However, this too seems unlikely as mice lacking orexins have normal circadian rhythms of wake and NREM sleep when housed Suplatast tosilate in constant darkness (Kantor et al., 2009 and Mochizuki et al., 2004). A better explanation may be that impaired orexin signaling causes behavioral states to become unstable (Figure 5). In fact, this idea was first raised by Broughton over 20 years ago as narcoleptic people and animals have great difficulty remaining awake, but they also have fragmented sleep and many more transitions between all states (Broughton et al., 1986). This breakdown in the ability to produce cohesive wake and sleep states is consistent with a destabilized switching mechanism.

, 2005 and Tank et al , 1988), which have been attributed to prop

, 2005 and Tank et al., 1988), which have been attributed to propagating dendritic calcium spikes. While regenerative events have been recorded from proximal smooth dendrites both in vivo (Fujita, 1968 and Kitamura and Häusser, 2011) and in vitro (Davie et al., 2008 and Llinás and Sugimori, 1980), the variability of CF calcium transients measured in distal spiny branchlets suggests that calcium spikes may not always

occur at distal sites. The amplitude of the CF calcium signal is modulated by the somatic holding potential (Wang et al., 2000 and Kitamura and Häusser, 2011), by dendritic field depolarization (Midtgaard et al., 1993), by synaptic inhibition of the dendrites (Callaway et al., 1995 and Kitamura and Häusser, 2011), and by the activity of Idelalisib PFs (Brenowitz and Regehr, 2005 and Wang et al., 2000). The mechanisms underlying these modulations remain unknown. Purkinje cells express a high density of P/Q-type (Usowicz et al.,

1992) and T-type AZD6244 mouse (Hildebrand et al., 2009) calcium channels. P/Q-type channels sustain propagating high-threshold dendritic calcium spikes (Fujita, 1968, Llinás et al., 1968 and Llinás and Sugimori, 1980). In contrast, T-type channels are involved in local spine-specific calcium influx during PF bursts (Hildebrand et al., 2009). Purkinje cell dendrites also express a variety of voltage-gated potassium channels, but their roles in the regulation of dendritic calcium electrogenesis are poorly understood (Etzion and Grossman, 1998, Llinás and Sugimori,

1980, McKay and Turner, 2004 and Womack and Khodakhah, 2004). Here, we used random-access no multiphoton (RAMP) microscopy to monitor the calcium transients induced by CF stimulation (CF-evoked calcium transients [CFCTs]) at high temporal resolution to unambiguously distinguish between subthreshold calcium transients and calcium spikes. We show that calcium spike initiation and propagation in distal spiny branchlets are controlled by activity-dependent mechanisms. CFCTs were mapped optically in Purkinje cell smooth and spiny dendrites using RAMP microscopy (Otsu et al., 2008). At repetition rates close to 1 kHz, the peak of Fluo-4 (200 μM) fluorescence transients was well resolved (Figure S1 available online). Using dual indicator quantitative measurements (see Experimental Procedures), we found that the amplitude of the CFCT (Figures 1A and 1B) decreased with distance from the soma (Figure 1C). In individual spiny dendrites, CFCT amplitude decreased linearly as a function of the distance from the parent dendritic trunk (Figure 1D) by −1.4% ± 0.4% μm−1 (±SD) for spines (r = −0.26, p < 0.001; n = 157 of 14 cells), and −1.5% ± 0.4% μm−1 for spiny branchlet shafts (r = −0.36, p < 0.001; n = 114 of 14 cells). In proximal compartments (<50 μm from soma), fluorescence transients averaged 0.023 ± 0.008 ΔG/R (±SD) in spines (n = 15, 5 cells), 0.020 ± 0.008 ΔG/R in spiny branchlets (n = 19, 7 cells), and 0.014 ± 0.008 ΔG/R in smooth dendrites (n = 25, 10 cells).

Nonetheless, any individual cell of a given type can dynamically

Nonetheless, any individual cell of a given type can dynamically alter its precise molecular profile and corresponding physical and electrical properties in response to a variety of external cues (Curran and Morgan, 1985 and Greenberg et al., 1985). Hence, although all cells of the same type stably express a common suite of genes, individual BIBW2992 mouse members of a cell type may vary in the precise profile of genes expressed depending upon context and activity. We argue also that this ground state is determined shortly after cells exit from their last mitotic

cycle and that the execution and stabilization of neuronal gene expression programs require local events that occur in the final stages of maturation during what are commonly referred to as “critical periods” of development. Furthermore, although it is apparent that cell types can be defined molecularly, an understanding of the nervous system cannot be reached without comprehensive data regarding the circuits in which

they are embedded, their connectivity, and their activity patterns in response to appropriate external stimuli. Only then can we begin to achieve Perifosine solubility dmso the ultimate goal of providing an understanding of the contributions of discrete cell types to behavior. In a general sense, the number of cell types present within a given substructure of the nervous system reflects the computational complexity of its functions. In simple organisms or in the context of the peripheral

nervous system (Garcia-Campmany next et al., 2010 and Arber, 2012), the contributions of many specific cell types to behavior have been studied in great detail, and, in most cases, the reasons for their specialization are apparent. For example, specific sensory and motor neuronal classes with distinct anatomical and electrophysiological properties make up simple motor circuits that generate fixed action patterns (Schiff et al., 1999). Local neuron types modulate or generate rhythmic behaviors, allowing these cell types to execute discrete functions (Bargmann and Marder, 2013 and Goulding and Pfaff, 2005). This general model may apply for even more complex circuits with a relatively large number of identifiable cell types. It is believed that a nearly complete accounting for all cell types present in the mammalian retina places the number at around 60 discrete types (Masland, 2012). Although the precise functions of each of these cell types are not known, the fact that they are tiled across the retina suggests that each of them contributes to specific aspects of visual perception. A particularly clear recent example of this idea comes from studies of the JamB retinal ganglion cell population in which the anatomy, physiology, receptive fields, and distribution of JamB cells are all tailored for their ability to perceive upward motion (Kim et al., 2008).

70, p < 0 05) or SRT (decode probability correct = 0 62, p < 0 05

70, p < 0.05) or SRT (decode probability correct = 0.62, p < 0.05) before a coordinated movement. Therefore, beta-band LFP activity reflects a population of neurons whose firing rate reliably predicts the RT of coordinated eye-hand movements but not saccades made alone. Neurons which do not participate in the coherent beta-band LFP activity do not predict RT of

either movement type. Beta-band activity may reflect the coordinated control of reach and saccade RTs together. We have shown that beta-band Erastin molecular weight spiking and LFP activity varies with both SRT and RRT across a population of sites, but this is not necessarily sufficient to demonstrate that the control of saccade and reach RTs occurs together. Activity at some sites may be involved in controlling one effector, while activity at different sites may control the other effector. To link beta-band activity to the coordinated control of movement timing, we examined whether selectivity for both saccade and reach RTs is present in activity at the same sites. We determined RT selectivity by grouping LFP power during trials with the slowest 33% of RTs and LFP power during trials with the fastest 33% of SRTs and computing a z-score using random permutations (see Experimental Procedures) and found

that RT selectivity does exist for both movements at the same sites (Figure 6A). At 15 Hz, LFP activity was significantly selective for both SRT and RRT at 10/72 sites (14%; p < 0.01, Binomial test). In comparison, LFP activity at 45 Hz was selective for both RTs at only 2/72 sites (3%; p = 0.88. Kinase Inhibitor Library concentration Binomial test. Figure 6B). The strength of the effect at single sites is limited by the number of trials available for analysis. When we restrict our analysis to recording sites with at least 135 trials per direction

and task, 30% of recording sites were significantly selective for both SRT and RRT in the beta-band. We found a high degree of correlation between SRT selectivity and RRT selectivity in both the beta-band (R = 0.65 at 15 Hz) and the gamma-band (R = 0.41 at 45 Hz). many Thus, LFP activity at each recording site predicts the RT of both the saccade and the reach in a similar manner, with the strongest effects present in the beta band. These data suggest that if changes in beta-band power change the RT for both movements, beta-band activity could coordinate movement timing. If beta-band power reflects the joint control of movement RTs, variations in the level of beta-band power could give rise to correlations in the behavioral RTs, and lack of power variation could lead to a reduction or even elimination in the RT correlations. To test this prediction, we calculated the relationship between saccade and reach RTs across groups of trials when beta-band power is relatively constant (see Experimental Procedures).

, 1983) Phase shifts could also arise from a single pacemaker (e

, 1983). Phase shifts could also arise from a single pacemaker (e.g., medial septum) through delays emerging from a chain of unidirectionally linked groups of neurons (Ermentrout and Kleinfeld, 2001). Another mechanism that could provide delays would be a chain of oscillators residing within the pacemaker itself (e.g., the septal area) and the phase shift observed in the hippocampus would be a reflection of the phase-delayed septal outputs. While this latter solution

cannot be fully excluded, it would require a complex temporal coordination of the hippocampal and entorhinal neurons in different regions and layers with appropriate delays. We hypothesize that traveling theta waves arise from a network of “weakly coupled” (Kopell and Ermentrout, 1986) intrahippocampal Lumacaftor ic50 and matched entorhinal cortex see more oscillators. In support of this hypothesis,

both the CA3 recurrent system and in vitro slices of the CA3 region can generate theta oscillations (Konopacki et al., 1987; Kocsis et al., 1999). In addition, delays with similar magnitudes documented here have been reported in the isolated CA3 region in vitro (Miles et al., 1988). The importance of weakly coupled oscillators in traveling waves is illustrated by the spinal cord activity of the lamprey during swimming (Kopell and Ermentrout, 1986; Cohen et al., 1992). The swim rhythm arises from intersegmental coordination of spinal cord oscillators, connected by local connections with short delays. The dominance of forward swimming is secured by the faster oscillators in the frontal end of the cord (Grillner et al., 1995). By analogy, the oscillation frequencies of place cell assemblies decrease progressively along the septotemporal axis of the hippocampus (Jung et al., 1994; Maurer et al., 2005; Kjelstrup et al., 2008; Royer et al., 2010) and theta oscillating cell

groups are coupled by delays (Geisler et al., 2010). Similarly, the oscillation frequencies of medial entorhinal cortex neurons decrease progressively in the dorsoventral direction (Giocomo et al., 2007), providing a frequency match between corresponding of entorhinal and hippocampal neurons. Due to the delays, the faster but transient assembly oscillators produce a slower global rhythm, expressed by the coherent LFP oscillation in the entire length of the hippocampus and entorhinal cortex (Geisler et al., 2010). The progressively decreasing excitability of pyramidal neurons along the long axis (Segal et al., 2010) might further contribute to the dominantly septotemporal spread of activity. Finally, septally projecting long-range interneurons (Dragoi et al., 1999; Tóth et al., 1993; Jinno et al.

05 (Figure S5A) For models constrained by both lineages, 59/66 c

05 (Figure S5A). For models constrained by both lineages, 59/66 cross-validation correlations ABT-737 nmr were significant at a threshold of p < 0.005 (Figure S5B). For 33 of these

neurons, we further explored the relationship between axial and surface tuning with an additional test (Figures 6B–6D) based on one high response medial axis stimulus, one intermediate response stimulus, and one low response stimulus. Medial axis structure was preserved while surface shape was substantially altered. For some neurons, responses to a given medial axis structure remained largely consistent across surface alterations (Figure 6B). In contrast, most neurons showed strong sensitivity to surface alterations (Figures 6C and 6D). The distribution of surface sensitivity (as measured by invariance to surface selleck products changes; Figure 6E, horizontal axis) was continuous. Even for neurons with substantial surface sensitivity (toward the left of the plot), tuning for medial axis structure remained consistent (as measured by correlation between axial tuning patterns across the different surface conditions; Figure 6E, vertical axis). The full set of 59 significant composite models (constrained by both lineages) is depicted in Figure 7. In each case, the model is projected onto

one high response stimulus from each of the two medial axis lineages (left and right), with the original shaded stimuli shown below. We identified a wide array of medial axis tuning configurations, ranging from 1–12 components, and including single and double Y/T junctions. In most cases (48/59), the surface templates were at least partially associated with the same object fragments described by the medial axis templates. Surface configuration tuning also varied widely, and this was substantiated by surface models identified for the 45 neurons studied with two surface lineages (Figure S6). It is important to note that, while these tuning templates were often complex, they did not define the entire global structure of

high response stimuli. In fact, high response stimuli varied widely in global shape, both within and between stimulus lineages (Figures 1, 4, 5, 7, and 8). Thus, individual IT neurons do not appear to represent global shape, at least in the domain of novel, abstract objects Adenosine studied here. Rather, novel objects must be represented by the ensemble activity of IT neurons encoding their constituent substructures. Object shape in three dimensions is inferred from 2D image features, including shading and 2D occlusion boundary contours (Koenderink, 1984). Many IT neurons appear to encode inferred 3D object shape (Janssen et al., 2000a and Janssen et al., 2000b), rather than low-level image features, since IT shape tuning remains consistent across dramatic changes in 2D shading patterns (produced by altered lighting direction) and is strongly diminished or abolished by removing depth cues (Yamane et al., 2008).