Marburg hemorrhagic fever (MHF) is a fatal infectious disease caused by Marburg virus (MARV) infection, and MARV has been identified as a priority pathogen for vaccine development by the WHO. The glycoprotein (GP) of MARV mediates viral adhesion and invasion of host cells and therefore can be used as an effective target for vaccine development. Moreover, DNA vaccines have unique advantages, such as simple construction processes, low production costs, and few adverse reactions, but their immunogenicity may decrease due to the poor absorption rate of plasmids. Lysosome-associated membrane protein 1 (LAMP1) can direct antigens to lysosomes and endosomes and has great potential for improving the immunogenicity of nucleic acid vaccines. Therefore, we constructed a DNA vaccine based on a codon-optimized MARV GP (ID MF939097.1) fused with LAMP1 and explored the effect of a LAMP targeting strategy on improving the immunogenicity of the MARV DNA vaccine. ELISA, ELISpot, and flow cytometry revealed that the introduction of LAMP1 into the MARV DNA candidate vaccine improved the humoral and cellular immune response, enhanced the secretion of cytokines, and established long-term immune protection. Transcriptome analysis revealed that the LAMP targeting strategy significantly enriched antigen processing and presentation-related pathways, especially the MHC class II-related pathway, in the candidate vaccine. Our study broadens the strategic vision for enhanced DNA vaccine design and provides a promising candidate vaccine for MHF prevention.

[This corrects the article DOI: 10.3389/fimmu.2023.1075419.].

The proposed method in this paper proposes an end-to-end unsupervised semantic segmentation architecture DMSA based on four loss functions. The framework uses Atrous Spatial Pyramid Pooling (ASPP) module to enhance feature extraction. At the same time, a dynamic dilation strategy is designed to better capture multi-scale context information. Secondly, a Pixel-Adaptive Refinement (PAR) module is introduced, which can adaptively refine the initial pseudo labels after feature fusion to obtain high quality pseudo labels. Experiments show that the proposed DSMA framework is superior to the existing methods on the saliency dataset. On the COCO 80 dataset, the MIoU is improved by 2.0, and the accuracy is improved by 5.39. On the Pascal VOC 2012 Augmented dataset, the MIoU is improved by 4.9, and the accuracy is improved by 3.4. In addition, the convergence speed of the model is also greatly improved after the introduction of the PAR module.