Mosquito-borne and sexual transmission of Zika virus (ZIKV), a TORCH pathogen, recently initiated a series of large epidemics throughout the Tropics. Animal models are necessary to determine transmission risk and study pathogenesis, as well screen antivirals and vaccine candidates. In this study, we modeled mosquito and sexual transmission of ZIKV in the African green monkey (AGM). Following subcutaneous, intravaginal or intrarectal inoculation of AGMs with ZIKV, we determined the transmission potential and infection dynamics of the virus. AGMs inoculated by all three transmission routes exhibited viremia and viral shedding followed by strong virus neutralizing antibody responses, in the absence of clinical illness. All four of the subcutaneously inoculated AGMs became infected (mean peak viremia: 2.9 log₁₀ PFU/mL, mean duration: 4.3 days) and vRNA was detected in their oral swabs, with infectious virus being detected in a subset of these specimens. Although all four of the intravaginally inoculated AGMs developed virus neutralizing antibody responses, only three had detectable viremia (mean peak viremia: 4.0 log₁₀ PFU/mL, mean duration: 3.0 days). These three AGMs also had vRNA and infectious virus detected in both oral and vaginal swabs. Two of the four intrarectally inoculated AGMs became infected (mean peak viremia: 3.8 log₁₀ PFU/mL, mean duration: 3.5 days). vRNA was detected in oral swabs collected from both of these infected AGMs, and infectious virus was detected in an oral swab from one of these AGMs. Notably, vRNA and infectious virus were detected in vaginal swabs collected from the infected female AGM (peak viral load: 7.5 log₁₀ copies/mL, peak titer: 3.8 log₁₀ PFU/mL, range of detection: 5–21 days post infection). Abnormal clinical chemistry and hematology results were detected and acute lymphadenopathy was observed in some AGMs. Infection dynamics in all three AGM ZIKV models are similar to those reported in the majority of human ZIKV infections. Our results indicate that the AGM can be used as a surrogate to model mosquito or sexual ZIKV transmission and infection. Furthermore, our results suggest that AGMs are likely involved in the enzootic maintenance and amplification cycle of ZIKV. Zika virus (ZIKV) is primarily maintained in an enzootic cycle involving nonhuman primates and mosquitoes, with epizootics and epidemics occurring when the virus is introduced into naïve populations of nonhuman primates or humans, respectively. While, the primary transmission mechanism of the virus is by the bite on an infected mosquito, ZIKV can also be sexually transmitted. In an effort to develop novel animal models to study ZIKV disease, and to better understand the role of nonhuman primates as amplification and maintenance hosts of ZIKV in nature, we modeled mosquito-borne and sexual transmission of ZIKV in the enzootic host, the African green monkey (AGM). Infection dynamics and neutralizing antibody responses in all three AGM ZIKV models (subcutaneous, intravaginal and intrarectal) in the absence of clinical illness–recapitulated reported generalized human disease course. Furthermore, we detected prolonged shedding with high viral loads and infectious virus in the vaginal swabs collected from an infected female AGM inoculated intrarectally. Notably, these results support limited human clinical evidence that ZIKV transmission can occur during female-to-male vaginal sexual acts, and furthermore indicate the existence of ZIKV super-spreaders. Finally, our results indicate sexual transmission of ZIKV could occur among infected nonhuman primates (e.g. Chlorocebus spp.) in Africa and may serve as a secondary transmission and maintenance mechanism in the absence of mosquito-to-nonhuman primate transmission.

Bulk RNA sequencing (RNA-seq) produces high-dimensional gene expression data where the number of measured features greatly exceeds the number of available samples, which challenges artificial intelligence (AI)-based modeling. In this setting, models are highly susceptible to overfitting and may fail to generalize across independent datasets when feature dimensionality is not adequately controlled. Feature (gene) selection is therefore essential for reliable inference, particularly when it is performed strictly within training data to prevent information leakage. This review examines how high dimensionality and limited sample size constrain AI-based analysis of bulk RNA-seq data and surveys feature selection strategies used to address these challenges. Emphasis is placed on statistically guided, training-only frameworks, with a focus on generalized linear models with quasi-likelihood F tests and magnitude–altitude scoring (GLMQL-MAS). Across published viral infection studies, GLMQL-MAS yields compact, interpretable gene sets that support robustness, reproducibility, and cross-dataset generalization in bulk transcriptomic modeling.

Arthropod-borne viruses such as Alphaviruses and Flaviviruses are the causative agents of severe human disease, highlighted by fatal encephalitis and neurological sequelae in survivors. The lack of FDA-approved vaccines and therapeutics that can prevent or treat these infections results in a significant global disease burden. An important unmet need to address this capability gap is the need for affordable, scalable, clinically relevant human-based neuronal models to study neuroinvasive viruses and evaluate therapeutic options. Here, we described the application of a human neuronal precursor cell model, LUHMES (Lund human mesencephalic) cells, that can be differentiated into dopaminergic midbrain neurons and used to study virus infections. In this study, we demonstrated the susceptibility of LUHMES cells to infection by three arthropod-borne neurotropic viruses: Venezuelan equine encephalitis virus, dengue virus serotype 2, and West Nile virus. We also demonstrated how the model may be applied to evaluate potential therapeutic options using an FDA-approved small molecule, Omaveloxolone. Finally, we analyzed host cell responses to infection and treatment using gene expression and phospho-signaling analyses. These findings highlight the value of this model to interpret the pathogenic mechanisms of neurotropic viral infections and evaluate potential therapeutic intervention strategies in a clinically relevant in vitro human neuronal model.

Monkeypox virus (MPXV) is the causative agent of mpox disease in humans. The virus comprises two clades, Central African clade I and West African clade II, with case fatality rates of ~11 and ~4%, respectively. Since the discovery of mpox disease in 1970, the virus has been restricted to Africa. However, in 2022, a previously unrecognized subclade IIb caused the largest global outbreak of mpox disease with a case fatality rate of ~0.2%. The difference in virulence of MPXV subclades in human infection warrants further investigation; however, one critical limitation is the lack of susceptible small animal models. In this study, we investigated the susceptibility of four murine models, including CAST-EiJ and three immunocompromised models (C57BL/6 Ifnar ^−/− , C57BL/6 Ifngr ^−/− , and C57BL/6 Ifnar ^−/− / Ifngr ^−/− ) to MPXV clade IIa (WR 7-61 and US-2003) and IIb (MA-2022) isolates. All four mouse models were susceptible to clade IIa infection, leading to severe disease marked by decreased body temperature, weight loss, and lethality. In contrast, clade IIb infection produced minimal to mild disease at similar doses in all four murine models. The clade IIb isolate produced severe disease (40% lethality) at only the highest dose (8.0 log ₁₀ PFU) in the most susceptible immunocompromised mouse model, C57BL/6 Ifnar ^−/− / Ifngr ^−/− . This is the first demonstration of lethal disease with clade IIb in a murine model. In addition, these data show that clade IIa is ~100 to 100,000-fold more virulent than clade IIb and provides three additional murine models for investigating MPXV infection and pathogenesis. IMPORTANCE: Mpox is an emerging human disease caused by four distinct MPXV subclades (Ia, Ib, IIa, and IIb). Despite genetic similarities, the case fatality rate varies considerably between the subclades: Ia (~11%), Ib and IIa (~4%), and IIb (~0.2%). Since 2022, multiple mpox outbreaks have occurred due to previously unrecognized subclades, leading to the declaration of two public health emergencies by the World Health Organization. This unprecedented global spread, coupled with the variation in the severity of human disease, underscores the importance of research into the pathogenesis of emerging MPXV subclades. However, a critical limitation is the lack of suitable small animal models. This study identifies three additional murine models susceptible to MPXV clade II infection and demonstrates significant virulence differences between clades IIa and IIb. These models will enable a rapid characterization of previously unrecognized subclades and will facilitate countermeasure development.