Biology subjects

Castellanos, A. A.

Publications and source records attributed to Castellanos, A. A..

7 recordsLinked to original sources

bioRxiv2026

From receptor binding to biogeography: Multi-scale prediction of filovirus hosts in bats

Forecasting zoonotic risk requires identifying which host species are biologically susceptible to infection, yet susceptibility is rarely predicted using frameworks that integrate molecular mechanisms with macroecology. Filoviruses, a diverse group of bat-associated viruses that include Ebola and Marburg viruses, illustrate this challenge: viral entry depends on interactions between viral glycoproteins and the host receptor NPC1, and host ecology and distribution determine opportunity of viral entry. Additionally, receptor sequence data used for informing viral entry are available for only a small fraction of bat species. Here, we extend virus-specific susceptibility prediction across the global diversity of bats by integrating experimentally measured and physicochemically inferred virus-receptor binding strengths with phylogenetic, ecological, and environmental data. Using boosted regression models trained on binding assay labels, we generate predictions of NPC1-mediated binding strength for more than 1,300 bat species. Predicted susceptibility is strongly structured by evolutionary relationships, with high binding concentrated in particular bat lineages, but is further differentiated within clades by morphology, life-history strategy, and environmental context. Strikingly, macroevolutionary structure alone recovers interaction patterns originally derived from amino acid-level physicochemistry, indicating that information about receptor-mediated compatibility is recoverable from host evolutionary history and ecological traits. Predicted high binding strength extends well beyond historically recognized outbreak regions, suggesting that the fundamental host range of filoviruses may be substantially broader than their currently realized distribution. By scaling receptor biology to global host diversity, this multi-scale framework expands mechanistic susceptibility forecasting beyond species with available molecular data and provides a generalizable approach for integrating molecular and ecological information in zoonotic prediction.

ecology↗

bioRxiv2025

Aminopeptidase N is a receptor for hedgehog merbecoviruses

Merbecoviruses, closely related to the highly pathogenetic Middle East Respiratory Syndrome Coronavirus (MERS-CoV), circulate in hedgehogs throughout Europe and Asia, raising concerns about zoonotic transmission to humans and domestic animals. Unfortunately, how these viruses enter host cells remains unknown, hindering experimental studies. Here, we tested known coronavirus receptor orthologues from European hedgehogs (Erinaceus europaeus) and identified Aminopeptidase N (APN) as an entry receptor for hedgehog merbecoviruses. We confirm this result with single-cycle pseudotype and replication-competent virus experiments as well as protein binding assays. A screen of 30 mammalian APN orthologues reveals restricted cross-species receptor use. Cryo-electron microscopy analysis of the viral glycoprotein-receptor complex shows a unique interface distinct from known coronavirus spike:APN interactions, providing a molecular basis for species barriers. These findings expand the known range of receptor use not only within merbecoviruses but also betacoronaviruses, improving our understanding of betacoronavirus receptors, and informing risk assessments for viral emergence.

microbiology↗

bioRxiv2024

Integrating predictors of host condition into spatiotemporal multi-scale models of virus shedding

Understanding where and when pathogens occur in the environment has implications for reservoir population health and infection risk. In reservoir hosts, infection status and pathogen shedding are affected by processes interacting across different scales: from landscape features affecting host location and transmission to within-host processes affecting host immunity and infectiousness. While uncommonly done, simultaneously incorporating processes across multiple scales may improve pathogen shedding predictions. In Australia, the black flying fox (Pteropus alecto) is a natural host for the zoonotic Hendra virus, which is hypothesized to cause latent infections in bats. Re-activation and virus shedding may be triggered by poor host condition, leading to virus excretion through urine. Here, we developed a statistical modeling approach that combined data at multiple spatial and temporal scales to capture ecological and biological processes potentially affecting virus shedding. We parameterized these models using existing datasets and compared model performance to under-roost virus shedding data from 2011-2014 in 23 roosts across a 1200-km transect. Our approach enabled comparisons among multiple model structures to determine which variables at which scales are most influential for accurate predictions of virus shedding in space and time. We identified environmental predictors and temporal lags of these features that were important for determining where reservoirs are located and multiple independent proxies for reservoir condition. The best-performing multi-scale model delineated periods of low and high virus prevalence, reflecting observed shedding patterns from pooled under-roost samples. Incorporating regional indicators of food scarcity enhanced model accuracy while incorporating other stress indicators at local scales confounded this signal. This multiscale modeling approach enabled the combination of processes from different ecological scales and identified environmental variables influencing Hendra virus shedding, highlighting how integrating data across scales may improve risk forecasts for other pathogen systems.

ecology↗

bioRxiv2024

Understanding Ecological Systems Using Knowledge Graphs: An Application to Highly Pathogenic Avian Influenza

Ecological systems are complex. Representing heterogeneous knowledge about ecological systems is a pervasive challenge because data are generated from many subdisciplines, exist in disparate sources, and only capture a subset of important interactions underpinning system structure, resilience, and dynamics. Knowledge graphs have been successfully applied to organize heterogeneous data systematically and to predict new linkages representing unobserved relationships in complex systems. Though not previously applied broadly in ecology, knowledge graphs have much to offer in an era of global change when system dynamics are responding to rapid changes across multiple scales simultaneously. We developed a knowledge graph to demonstrate the methods utility for ecological problems focused on highly pathogenic avian influenza (HPAI), a highly transmissible virus with a broad animal host range, wide geographic distribution, and rapid evolution with pandemic potential. We describe the development of a graph to include a wide range of data related to HPAI including pathogen-host associations, animal species distributions, and human population demographics, using a semantic ontology that defines relationships within the data and between datasets. We use the graph to perform a set of proof-of-concept analyses validating the method and identifying new relationships and features of HPAI ecology, underscoring the generalizable value of knowledge graphs to ecology including their utility in revealing previously known relationships between entities and generating testable hypotheses in support of a deeper mechanistic understanding of ecological systems.

ecology↗

bioRxiv2022

Phylogenetic and biogeographical traits predict unrecognized hosts of zoonotic leishmaniasis

The spatio-temporal distribution of leishmaniasis, a parasitic vector-borne zoonotic disease, is significantly impacted by land-use change and climate warming in the Americas. However, predicting and containing outbreaks is challenging as the zoonotic Leishmania system is highly complex: leishmaniasis (visceral, cutaneous and muco-cutaneous) in humans is caused by up to 14 different Leishmania species, and the parasite is transmitted by dozens of sandfly species and is known to infect almost twice as many wildlife species. Despite the already broad known host range, new hosts are discovered almost annually and Leishmania transmission to humans occurs in absence of a known host. As such, the full range of Leishmania hosts is undetermined, inhibiting the use of ecological interventions to limit pathogen spread and the ability to accurately predicting the impact of global change on disease risk. Here, we employed a machine learning approach to generate trait profiles of known zoonotic Leishmania wildlife hosts (mammals that are naturally exposed and susceptible to infection) and used trait-profiles of known hosts to identify potentially unrecognized hosts. We found that biogeography, phylogenetic distance, and study effort best predicted Leishmania host status. Traits associated with global change, such as agricultural land-cover, urban land-cover, and climate, were among the top predictors of host status. Most notably, our analysis suggested that zoonotic Leishmania hosts are significantly undersampled, as our model predicted just as many unrecognized hosts as unknown hosts. Overall, our analysis facilitates targeted surveillance strategies and improved understanding of the impact of environmental change on local transmission cycles. Author summaryLeishmaniasis is a zoonotic, vector borne disease of poverty with a high burden throughout the Americas: within Latin America there are an estimated 58,500 new cases per year and 54,050 years of life lost due to disability. Although the World Health Organization has targeted leishmaniasis for elimination and control by 2030, the disease remains a persistent threat. Across the Americas, particularly in Central America, the southeastern United States, and perimeters of the Amazon Basin, risk of infection is increasing in geographic extent and elevation. While it is known that Leishmania parasites, the causative agent of leishmaniasis, are maintained in the environment via a mammalian host, the full suite of wildlife hosts has yet to be documented, which significantly hinders control efforts. Here, we use machine learning and ecological and evolutionary trait profiles of known hosts to identify unrecognized potential wildlife hosts of Leishmania. We identify 136 mammals in the Americas that are likely to be exposed to and infected by zoonotic Leishmania in the wild. The high number of unrecognized potential hosts emphasizes a need to better invest in studying the ecological epidemiology of leishmaniasis. The study provides information and tools to support targeted intervention and management of this important poverty-associated disease.

ecology↗

bioRxiv2022

Data-Driven Predictions of Potential Leishmania Vectors in Latin America

The incidence of vector-borne diseases is rising as deforestation, climate change, and globalization bring humans in contact with arthropods that can transmit pathogens. In particular, incidence of American Cutaneous Leishmaniasis (ACL), a parasitic disease transmitted by sandflies, is increasing as previously intact habitats are cleared for agriculture and urban areas, potentially bringing people into contact with vectors and reservoir hosts. Previous evidence has identified dozens of sandfly species that have been infected with and/or transmit Leishmania parasites. However, there is an incomplete understanding of which sandfly species transmit the parasite, complicating efforts to limit disease spread. Here, we apply machine learning models (boosted regression trees) to leverage biological and geographical traits of known sandfly vectors to predict potential vectors. Additionally, we generate trait profiles of confirmed vectors and identify important factors in transmission. Our model performed well with an average out of sample accuracy of 86%. The models predict that synanthropic sandflies living in areas with greater canopy height, less human modification, and within an optimal range of rainfall are more likely to be Leishmania vectors. We also observed that generalist sandflies that are able to inhabit many different ecoregions are more likely to transmit the parasite. Our results suggest that Psychodopygus amazonensis and Nyssomia antunesi are unidentified potential vectors, and should be the focus of sampling and research efforts. Overall, we found that our machine learning approach provides valuable information for Leishmania surveillance and management in an otherwise complex and data sparse system. Author SummaryAmerican Cutaneous Leishmaniasis (ACL) is a neglected parasitic disease transmitted by sandflies in Latin America. There is an incomplete understanding of which sandfly species transmit the parasite, complicating efforts to limit disease spread. In this study, the authors created a database of sandfly traits, then used predictive models to determine important factors in disease transmission and how different climate and environmental variables affect ACL transmission. The models suggest that transmission occurs at the interface between domestic habitats and well-preserved forests. The authors also generate predictions of which sandflies might be transmitting the parasite that are not known vectors at the time, specifically Psychodopygus amazonensis and Nyssomia antunesi. This new knowledge can lead to a better understanding of the system of transmission, and can point to possible hotspots of the disease. The analysis can also help direct researchers to areas of interest for sampling studies, as well as specific sandflies to focus their effort on.

ecology↗

bioRxiv2021

Predicting the zoonotic capacity of mammal species for SARS-CoV-2

Back and forth transmission of SARS-CoV-2 between humans and animals may lead to wild reservoirs of virus that can endanger efforts toward long-term control of COVID-19 in people, and protecting vulnerable animal populations that are particularly susceptible to lethal disease. Predicting high risk host species is key to targeting field surveillance and lab experiments that validate host zoonotic potential. A major bottleneck to predicting animal hosts is the small number of species with available molecular information about the structure of ACE2, a key cellular receptor required for viral cell entry. We overcome this bottleneck by combining species ecological and biological traits with 3D modeling of virus and host cell protein interactions using machine learning methods. This approach enables predictions about the zoonotic capacity of SARS-CoV-2 for over 5,000 mammals -- an order of magnitude more species than previously possible. The high accuracy predictions achieved by this approach are strongly corroborated by in vivo empirical studies. We identify numerous common mammal species whose predicted zoonotic capacity and close proximity to humans may further enhance the risk of spillover and spillback transmission of SARS-CoV-2. Our results reveal high priority areas of geographic overlap between global COVID-19 hotspots and potential new mammal hosts of SARS-CoV-2. With molecular sequence data available for only a small fraction of potential host species, predictive modeling integrating data across multiple biological scales offers a conceptual advance that may expand our predictive capacity for zoonotic viruses with similarly unknown and potentially broad host ranges.

ecology↗