The integration of deep learning and theories of reinforcement learning (RL) is a promising avenue to explore novel hypotheses on reward-based learning and decision-making in humans and other animals. Here, we trained deep RL agents and mice in the same sensorimotor task with high-dimensional state and action space and studied representation learning in their respective neural networks. Evaluation of thousands of neural network models with extensive hyperparameter search revealed that learning-dependent enrichment of state-value and policy representations of the task-performance-optimized deep RL agent closely resembled neural activity of the posterior parietal cortex (PPC). These representations were critical for the task performance in both systems. PPC neurons also exhibited representations of the internally defined subgoal, a feature of deep RL algorithms postulated to improve sample efficiency. Such striking resemblance between the artificial and biological networks and their functional convergence in sensorimotor integration offers new opportunities to better understand respective intelligent systems.

Given the increasing aging population and the rising prevalence of musculoskeletal diseases due to obesity and injury, urgent research is needed to formulate new treatment alternatives, as current options remain inadequate. Viruses can exacerbate arthritis and worsen symptoms in patients with pre-existing osteoarthritis. Over the past decade, the chikungunya virus (CHIKV) has emerged as a significant public health concern, especially in Asia and South America. Exploring natural products, such as berberine, has shown promise due to its anticatabolic, antioxidative, and anti-inflammatory effects. However, berberine’s low stability and bioavailability limit its efficacy. We hypothesized that encapsulating berberine into a proniosome gel, known for its ease of preparation and stability, could enhance its bioavailability and efficacy when applied topically, potentially treating CHIKV infection. Our investigation focused on how varying berberine loads and selected excipients in the proniosome gel influenced its physical properties, stability, and skin permeability. We also examined the biological half-life of berberine in plasma upon topical administration in mice to assess the potential for controlled and sustained drug release. Additionally, we analyzed the antioxidant stress activity and cell viability of HaCaT keratinocytes and developed a lipopolysaccharide-stimulated cell culture model to evaluate anti-inflammatory effects using pro-inflammatory cytokines. Overall, the research aims to transform the treatment landscape for arthritis by leveraging berberine’s therapeutic potential.

Osteoarthritis is a common and chronic degenerative joint disease that leads to persistent symptoms such as pain, stiffness, swelling, and fatigue, all of which significantly impair mobility and reduce quality of life. Viral infections can further worsen joint inflammation in individuals with pre-existing osteoarthritis. Among these, the chikungunya virus (CHIKV) has become an increasing public health concern, particularly in Asia and South America, where post-viral arthritis frequently mimics or intensifies chronic osteoarthritic symptoms. Current treatments rely mainly on analgesics, corticosteroids, and biologics, which often provide only temporary relief and can cause systemic side effects. Poor bioavailability and limited drug penetration into joint tissues also remain major obstacles to effective long-term therapy. To address these challenges, this study focused on developing a proniosome-based topical gel capable of delivering therapeutic agents locally and in a sustained manner to inflamed joints. The formulation incorporated berberine hydrochloride, Coptis chinensis extract, and glucosamine, selected for their anti-inflammatory and cartilage-protective properties. Non-ionic surfactants and co-lipids were optimized to enhance skin permeation, improve vesicle stability, and support controlled release. Preclinical evaluations were conducted using both osteoarthritis and CHIKV-induced arthritis in vivo models. The optimized proniosome gel demonstrated strong therapeutic efficacy, including reductions in inflammatory markers, improved cartilage preservation and better subchondral bone structure. Importantly, no hepatic or renal toxicity was observed. These findings suggest that the proniosome-based gel offers a safe and effective topical approach for arthritis management, providing localized and sustained drug delivery while minimizing systemic exposure.

Reward prediction errors (RPEs) quantify the difference between expected and actual rewards, serving to refine future actions. Although reinforcement learning (RL) provides ample theoretical evidence suggesting that the long-term accumulation of these error signals improves learning efficiency, it remains unclear whether the brain uses similar mechanisms. To explore this, we constructed RL-based theoretical models and used multiregional two-photon calcium imaging in the mouse dorsal cortex. We identified a population of neurons whose activity was modulated by varying degrees of RPE accumulation. Consequently, RPE-encoding neurons were sequentially activated within each trial, forming a distributed assembly. RPE representations in mice aligned with theoretical predictions of RL, emerging during learning and being subject to manipulations of the reward function. Interareal comparisons revealed a region-specific code, with higher-order cortical regions exhibiting long-term encoding of RPE accumulation. These results present an additional layer of complexity in cortical RPE computation, potentially augmenting learning efficiency in animals.