Hokto Kazama

Insecticidal Cry proteins produced by the bacterial insect pathogen Bacillus thuringiensis possess conserved three domains (3D). Two other families of insecticidal proteins—Prb proteins from bacterial symbionts of entomopathogenic nematodes and a marine bacterial pathogen, and IPD113 proteins from ferns—are two-domain (2D) proteins structurally resembling domains I and II of Cry proteins. Despite the structural resemblance, the amino acid sequence identity among these families is only ∼10%, supporting the prevailing view that they evolved independently through convergent evolution. However, the possibility of remote homology and ancient horizontal gene transfer (HGT) has remained largely unexplored. Here, we present evidence supporting divergent evolution of these protein families. We found that proteins containing Cry-like two domains are distributed far more broadly across bacteria, eukaryotes, and archaea than previously recognized. Phylogenetic analyses separated the structurally similar proteins into five clades including Cry, IPD113, Prb, and 2D/3D-mixed clades. Domain-level trees revealed intermixing within domain II, with nematocidal Cry proteins frequently clustering with IPD113 and 2D/3D-mixed clades, suggesting domain swapping events across these clades. Conserved amino acid motifs and domain core structures across clades further support a shared evolutionary origin. Phylogenetic and structural analyses suggest a bacterial origin of the fern IPD113 gene and its horizontal transfer via fungal symbionts, as well as multiple HGT events into diverse terrestrial and marine eukaryotes and archaea. This work redefines the evolutionary landscape of the insecticidal protein families, revealing their unexpectedly wide taxonomic distribution, possibly shaped through bacterial ecological networks and close biological interactions such as symbiosis.

Large-scale neuronal measurements have revealed continual drift of neural representations without explicit learning, raising fundamental questions about how the brain maintains reliable perception. While representational drift has been studied both experimentally and theoretically, the underlying mechanisms and computational consequences remain unclear. Here, we systematically compare two drift mechanisms—fluctuations in synaptic weights versus membrane potential—using an analytically tractable neural network model with aligned drift magnitude and timescale. We found that both mechanisms preserve representational structure across time regardless of the drift magnitude. However, only synaptic fluctuations maintain discriminability between representations under large drift magnitude. We use a selectivity space to explain how these similar and differential effects on representational drift emerge from the dynamics of individual neuronal tuning. We further examine functional consequences of drift using reversal learning tasks where stimulus-reward associations switch. Both mechanisms enhance single stimulus learning by gradually removing the old associations, but only synaptic fluctuations substantially improved learning of multiple stimuli by preserving the representational structure during drift. Our results reveal distinct computations performed by specific fluctuation mechanisms and explain how representational drift can support both stable and adaptive behavior.

To navigate the environment and search for hosts, mosquitoes utilize multiple sensory cues including carbon dioxide (CO₂), visual, olfactory, and humidity cues. However, how mosquitoes shape their behavior by integrating these cues is poorly understood. Here we monitored the flight maneuvers of Aedes albopictus in a virtual reality environment where sensory cues were presented in open- or closed-loop. We found that CO₂ enhances optomotor responses to wide-field motion stimuli provided with various angular speeds and contrasts. Furthermore, CO₂ allowed more accurate tracking of a low contrast, visual object whose position was updated in closed-loop. We also found that olfactory preference is bidirectionally modulated by CO₂ depending on the value of odors: CO₂ enhances attraction to host odors whereas it enhances aversion to a repellent. This olfactory enhancement was further contingent on the humidity level. Together, our results highlight the context-dependent impact of CO₂ on cross-modal integration in mosquitoes.

Odors are intimately tied to the taste system to aid food selection and determine the sensory experience of food. However, how smell and taste are integrated in the nervous system to drive feeding remains elusive. We show in Drosophila that odors alone activate gustatory receptor neurons (GRNs), trigger proboscis extension reflex (PER), a canonical taste-evoked feeding behavior, and enhance food intake. Odor-evoked PER requires the function of sugar-sensing GRNs but not olfactory organs. Calcium imaging and electrophysiological recording show that GRNs directly respond to odors. Odor-evoked PER is mediated by the Gr5a receptor, and is bidirectionally modulated by olfactory binding proteins. Finally, odors and sucrose co-applied to GRNs synergistically enhance PER and food consumption. These results reveal a cell-intrinsic mechanism for odor-taste multimodal integration that takes place as early as in GRNs, indicating that unified chemosensory experience is a product of layered integration in peripheral neurons and in the brain.

Summary Odors are intimately tied to the taste system to aid food selection and determine the sensory experience of food. However, how smell and taste are integrated in the nervous system to drive feeding remains elusive. We show in Drosophila that odors alone activate gustatory receptor neurons (GRNs), trigger proboscis extension reflex (PER), a canonical taste-evoked feeding behavior, and enhance food intake. Odor-evoked PER requires the function of sugar-sensing GRNs but not olfactory organs. Calcium imaging and electrophysiological recording show that GRNs directly respond to odors. Odor-evoked PER is mediated by the Gr5a receptor, and is bidirectionally modulated by olfactory binding proteins. Finally, odors and sucrose co-applied to GRNs synergistically enhance PER and food consumption. These results reveal a cell-intrinsic mechanism for odor-taste multimodal integration that takes place as early as in GRNs, indicating that unified chemosensory experience is a product of layered integration in peripheral neurons and in the brain.

Sleep is a state of reduced behavioural responsiveness to the external world. Lowered arousal is thought to arise from sensory gating in the brain, yet there is compelling evidence that stimuli can still be processed to some degree. Which aspects of sensory information are processed during sleep remains largely unknown. Here, we perform comprehensive recordings of activity from multiple neuron types during sleep and wakefulness in the Drosophila olfactory association center. We find that baseline activity of dopamine neurons (DANs) is distinct between active, rest and sleep states, enabling a decoder to predict when the fly is asleep. Optogenetic inhibition of DANs acutely induces sleep and lowers odor-evoked arousal, suggesting causality between DAN activity and behavior. We then examine odor processing during sleep, and find that whereas coding of odor identity in Kenyon cells (KCs) remains intact, coding of odor value in DANs is altered. Specifically, DANs encoding positive (attractive) and negative (aversive) values change their odor responses in an opposing manner, producing more aversive odor representations during sleep. Consistent with these changes in neural activity, sleeping flies awoken by odors show reduced approach and increased avoidance behaviour. Taken together, our data reveal that the brain does not simply suppress the propagation of sensory signals during sleep; rather, it utilizes different cell types to retain or modify specific aspects of sensory information. The aversive odor evaluation in DANs may have ethological benefits as it suppresses the motivation to pursue attractive odors during sleep and primes defensive action when aroused by the scent of potential dangers.

Developing a safe and potent repellent of mosquitoes applicable to human skins is an effective measure against the spread of mosquito-borne diseases. Recently, we have identified that hydrophobic solutions such as low viscosity polydimethylsiloxane (L-PDMS) spread on a human skin prevent mosquitoes from staying on and biting it. This is likely due to the ability of L-PDMS in wetting mosquito legs and exerting a capillary force from which the mosquitoes attempt to escape. Here we show three additional functions of L-PDMS that can contribute to repel Aedes albopictus, by combining physicochemical analysis and behavioral assays in both an arm cage and a virtual flight arena. First, L-PDMS, when mixed with topical repellents and applied on a human skin, enhances the effect of topical repellents in reducing mosquito bites by efficiently transferring them to mosquito legs upon contact. Second, L-PDMS applied to mosquito tarsi compromises visual object tracking during flight, exerting an influence outlasting the contact. Finally, L-PDMS applied to mosquito tarsi acts as an aversive reinforcer in associative learning, making mosquitoes avoid the conditioned odor. These results uncover a multifaceted potential of L-PDMS in altering a sequence of mosquito behaviors from biting a human skin, visual object tracking following takeoff, to the response to an odor linked with L-PDMS.

Updating the value of sensory cues through experiences is a critical element of adaptive behavior. Although dopaminergic neurons (DANs) achieve this by driving associative learning, whether they contribute to assessment of sensory values outside the context of association remains largely unexplored. Here we show in Drosophila that DANs in the mushroom body encode the innate value of odors and constantly update the current value by inducing plasticity during olfactory navigation. Simulation of neuronal activity in a network based on the connectome data reproduced the characteristics of DAN responses, proposing a concrete circuit mechanism for computation. Notably, odors alone induced value- and dopamine-dependent changes in the activity of mushroom body output neurons, which store the current value of odors, as well as the behavior of flies steering in a virtual environment. Thus, the DAN circuit known for discrete, associative learning also continuously updates odor values during navigation in a non-associative manner.