Toshihiko Hosoya

Dynamic predictive coding by the retina

Retinal ganglion cells convey the visual image from the eye to the brain. They generally encode local differences in space and changes in time rather than the raw image intensity. This can be seen as a strategy of predictive coding, adapted through evolution to the average image statistics of the natural environment. Yet animals encounter many environments with visual statistics different from the average scene. Here we show that when this happens, the retina adjusts its processing dynamically. The spatio-temporal receptive fields of retinal ganglion cells change after a few seconds in a new environment. The changes are adaptive, in that the new receptive field improves predictive coding under the new image statistics. We show that a network model with plastic synapses can account for the large variety of observed adaptations.

The habenula is crucial for experience-dependent modification of fear responses in zebrafish

The zebrafish dorsal habenula (dHb) shows conspicuous asymmetry in its connection with the interpeduncular nucleus (IPN) and is equivalent to the mammalian medial habenula. Genetic inactivation of the lateral subnucleus of dHb (dHbL) biased fish towards freezing rather than the normal flight response to a conditioned fear stimulus, suggesting that the dHbL-IPN pathway is important for controlling experience-dependent modification of fear responses.

Genetic Single-Cell Mosaic Analysis Implicates ephrinB2 Reverse Signaling in Projections from the Posterior Tectum to the Hindbrain in Zebrafish

The optic tectum is a visual center in vertebrates. It receives topographically ordered visual inputs from the retina in the superficial layers and then sends motor outputs from the deeper layers to the premotor reticulospinal system in the hindbrain. Although the topographic patterns of the retinotectal projection are well known, it is not yet well understood how tectal efferents in the tectobulbar tract project to the hindbrain. The retinotectal and the tectobulbar projections were visualized in a zebrafish stable transgenic line Tg(brn3a-hsp70:GFP). Using a single-neuron labeling system in combination with the cre/loxP and Gal4/UAS systems, we showed that the tectal neurons that projected to rhombomeres 2 and 6 were distributed with distinctive patterns along the anterior–posterior axis. Furthermore, we found that ephrinB2a was critically involved in increasing the probability of neurons projecting to rhombomere 2 through a reverse signaling mechanism. These results may provide a neuroanatomical and molecular basis for the motor command map in the tectum.

Mammalian Gcm genes induce Hes5 expression by active DNA demethylation and induce neural stem cells

Signaling mediated by Notch receptors is crucial for the development of many organs and the maintenance of various stem cell populations. The activation of Notch signaling is first detectable by the expression of an effector gene, Hes5, in the neuroepithelium of mouse embryos at embryonic day (E) 8.0–8.5, and this activation is indispensable for the generation of neural stem cells. However, the molecular mechanism by which Hes5 expression is initiated in stem-producing cells remains unknown. We found that mammalian Gcm1 and Gcm2 (glial cells missing 1 and 2) are involved in the epigenetic regulation of Hes5 transcription by DNA demethylation independently of DNA replication. Loss of both Gcm genes and subsequent lack of Hes5 upregulation in the neuroepithelium of E7.5–8.5 Gcm1−/−; Gcm2−/− mice resulted in the impaired induction of neural stem cells. Our data suggest that Hes5 expression is serially activated first by Gcms and later by the canonical Notch pathway.

Mismatched Decoding in the Brain

“How is information decoded in the brain?” is one of the most difficult and important questions in neuroscience. We have developed a general framework for investigating to what extent the decoding process in the brain can be simplified. First, we hierarchically constructed simplified probabilistic models of neural responses that ignore more than Kth-order correlations using the maximum entropy principle. We then computed how much information is lost when information is decoded using these simplified probabilistic models (i.e., “mismatched decoders”). To evaluate the information obtained by mismatched decoders, we introduced an information theoretic quantity, I*, which was derived by extending the mutual information in terms of communication rate across a channel. We showed that I* provides consistent results with the minimum mean-square error as well as the mutual information, and demonstrated that a previously proposed measure quantifying the importance of correlations in decoding substantially deviates from I* when many cells are analyzed. We then applied this proposed framework to spike data for vertebrate retina using short natural scene movies of 100 ms duration as a set of stimuli and computing the information contained in neural activities. Although significant correlations were observed in population activities of ganglion cells, information loss was negligibly small even if all orders of correlation were ignored in decoding. We also found that, if we inappropriately assumed stationarity for long durations in the information analysis of dynamically changing stimuli, such as natural scene movies, correlations appear to carry a large proportion of total information regardless of their actual importance.

Pathway‐dependent modulation by P2‐purinoceptors in the mouse retina

Adenosine trisphosphate (ATP) activates purinoceptors and acts as a neurotransmitter in the nervous system. In the retina, we previously reported that the immunohistochemical distribution of the subset of P2‐purinoceptors differs between the ON and OFF pathways. Here, we investigated whether ATP activates P2‐purinoceptors and modulates the physiological function of the mouse retina. We also examined if signal processing by P2‐purinoceptors is pathway specific. Results showed that ATP activated both ON‐ and OFF‐cholinergic amacrine cells. However, responses in OFF‐cholinergic amacrine cells were greater than those in ON‐cholinergic amacrine cells. Pharmacological studies in OFF‐cholinergic amacrine cells showed that the response of OFF‐cholinergic amacrine cells is mediated P2X2‐purinoceptors. Further, ATP increased γ‐aminobutyric acid (GABA)ergic inhibitory postsynaptic currents (IPSCs) in OFF‐ but not ON‐cholinergic amacrine cells. The increase in GABAergic IPSCs was mediated by P2‐purinoceptors. P2‐purinoceptor‐mediated signals suppressed OFF ganglion cells but activated ON ganglion cells. Our findings indicate that ATP physiologically modulates signal processing of the ON and OFF pathways in a pathway‐specific manner through P2‐purinoceptors.

MafB interacts with Gcm2 and regulates parathyroid hormone expression and parathyroid development.

Serum calcium and phosphate homeostasis is critically regulated by parathyroid hormone (PTH) secreted by the parathyroid glands. Parathyroid glands develop from the bilateral parathyroid‐thymus common primordia. In mice, the expression of transcription factor Glial cell missing 2 (Gcm2) begins in the dorsal/anterior part of the primordium on embryonic day 9.5 (E9.5), specifying the parathyroid domain. The parathyroid primordium then separates from the thymus primordium and migrates to its adult location beside the thyroid gland by E15.5. Genetic ablation of gcm2 results in parathyroid agenesis in mice, indicating that Gcm2 is essential for early parathyroid organogenesis. However, the regulation of parathyroid development at later stages is not well understood. Here we show that transcriptional activator v‐maf musculoaponeurotic fibrosarcoma oncogene homologue B (MafB) is developmentally expressed in parathyroid cells after E11.5. MafB expression was lost in the parathyroid primordium of gcm2 null mice. The parathyroid glands of mafB+/− mice were mislocalized between the thymus and thyroid. In mafB−/− mice, the parathyroid did not separate from the thymus. Furthermore, in mafB−/− mice, PTH expression and secretion were impaired; expression levels of renal cyp27b1, one of the target genes of PTH, was decreased; and bone mineralization was reduced. We also demonstrate that although Gcm2 alone does not stimulate the PTH gene promoter, it associates with MafB to synergistically activate PTH expression. Taken together, our results suggest that MafB regulates later steps of parathyroid development, that is, separation from the thymus and migration toward the thyroid. MafB also regulates the expression of PTH in cooperation with Gcm2.

Periodic Organization of a Major Subtype of Pyramidal Neurons in Neocortical Layer V

A major question in neocortical research is the extent to which neuronal organization is stereotyped. Previous studies have revealed functional clustering and neuronal interactions among cortical neurons located within tens of micrometers in the tangential orientation (orientation parallel to the pial surface). In the tangential orientation at this scale, however, it is unknown whether the distribution of neuronal subtypes is random or has any stereotypy. We found that the tangential arrangement of subcerebral projection neurons, which are a major pyramidal neuron subtype in mouse layer V, was not random but significantly periodic. This periodicity, which was observed in multiple cortical areas, had a typical wavelength of 30 μm. Under specific visual stimulation, neurons in single repeating units exhibited strongly correlated c-Fos expression. Therefore, subcerebral projection neurons have a periodic arrangement, and neuronal activity leading to c-Fos expression is similar among neurons in the same repeating units. These results suggest that the neocortex has a periodic functional micro-organization composed of a major neuronal subtype in layer V.

Estimating Receptive Fields from Responses to Natural Stimuli with Asymmetric Intensity Distributions

The reasons for using natural stimuli to study sensory function are quickly mounting, as recent studies have revealed important differences in neural responses to natural and artificial stimuli. However, natural stimuli typically contain strong correlations and are spherically asymmetric (i.e. stimulus intensities are not symmetrically distributed around the mean), and these statistical complexities can bias receptive field (RF) estimates when standard techniques such as spike-triggered averaging or reverse correlation are used. While a number of approaches have been developed to explicitly correct the bias due to stimulus correlations, there is no complementary technique to correct the bias due to stimulus asymmetries. Here, we develop a method for RF estimation that corrects reverse correlation RF estimates for the spherical asymmetries present in natural stimuli. Using simulated neural responses, we demonstrate how stimulus asymmetries can bias reverse-correlation RF estimates (even for uncorrelated stimuli) and illustrate how this bias can be removed by explicit correction. We demonstrate the utility of the asymmetry correction method under experimental conditions by estimating RFs from the responses of retinal ganglion cells to natural stimuli and using these RFs to predict responses to novel stimuli.

Lattice system of functionally distinct cell types in the neocortex

The basic modules of the neocortex The fundamental organization of excitatory and inhibitory neurons in the neocortex is still poorly understood. Subcerebral projection neurons, a major excitatory cell type in neocortical layer 5, form small cell clusters called microcolumns. Maruoka et al. examined large regions of mouse brain layer 5 and observed that thousands of these microcolumns make up a hexagonal lattice with a regular gridlike spacing. The other major layer 5 excitatory cell class, cortical projection neurons, also form microcolumns that interdigitate with those of the subcerebral projection neurons. Microcolumns received common presynaptic inputs and showed synchronized activity in many cortical areas. These microcolumns developed from nonsister neurons coupled by cell type–specific gap junctions, suggesting that their development is lineage-independent but guided by local electrical transmission. Science, this issue p. 610 Neocortical layer 5 is composed of microcolumns containing specific cell types that form a brainwide hexagonal lattice system. The mammalian neocortex contains many cell types, but whether they organize into repeated structures has been unclear. We discovered that major cell types in neocortical layer 5 form a lattice structure in many brain areas. Large-scale three-dimensional imaging revealed that distinct types of excitatory and inhibitory neurons form cell type–specific radial clusters termed microcolumns. Thousands of microcolumns, in turn, are patterned into a hexagonal mosaic tessellating diverse regions of the neocortex. Microcolumn neurons demonstrate synchronized in vivo activity and visual responses with similar orientation preference and ocular dominance. In early postnatal development, microcolumns are coupled by cell type–specific gap junctions and later serve as hubs for convergent synaptic inputs. Thus, layer 5 neurons organize into a brainwide modular system, providing a template for cortical processing.