Exploration of proteomic biomarkers and digital imaging analysis for Oryctes rhinoceros nudivirus infection

Oryctes rhinoceros nudivirus (OrNV) infects the larval stage of many coleopteran insects; however, the underlying mechanisms and biomarkers of infection are not fully characterised. In this study, an optimal culture condition was developed for OrNV replication and proteomic biomarkers were identified using comparative proteomic analysis. The highest level of viral copy number was observed in Sf9 cells treated with 450 mu M of H2O2 and 2% foetal bovine serum (FBS). Among the 48 identified proteins, 14 proteins were significantly modulated in 2% FBS and H2O2- treated OrNV-infected cells (F2V) as compared with 10% FBS treated non-infected cells (F10M). Network analysis revealed that SLC25A5, VDAC3, PHB2, and ANXA1 act as signature proteins for OrNV replication. Moreover, viral envelope glycoproteins, GRBNV_gp28-like and GrBNV_gp62-like proteins could be used as sensitive diagnostic signatures for OrNV infection. Furthermore, to conveniently identify the OrNV-infection in Allomyrina dichotoma larvae, an image classification model was trained by Google Teachable Machine, which distinguished images with accuracy rates of 91% and 86% for infected and non-infected larvae, respectively, at a learning rate of 0.001. This study demonstrated that Sf9 cell medium treated with 2% FBS and 450 mu M H2O2 is a permissible culture condition for OrNV replication. Proteomic signatures may be involved in the progression of viral infection. Additionally, a low-cost and non-invasive machine learning-derived digital imaging analysis may improve the prediction of OrNV infection in larvae.

Automated summary

In this study, proteomic biomarkers and an optimal culture condition for the replication of Oryctes rhinoceros nudivirus (OrNV) were identified using comparative proteomic analysis. Additionally, a machine learning-derived digital imaging analysis was developed to predict OrNV infection in larvae.