Bioinformatic analysis of PD 1 checkpoint blockade responsive immune microenvironment in severe inuenza infection

： 13 Background: The programmed cell death 1 (PD-1)/PD-1 ligand 1 (PD-L1) signaling pathway is 14 significantly upregulated in severe influenza virus infection, which impairs the immune system 15 and causes increased tissue inflammation and damage. Blocking this signaling pathway will 16 reduce the damage, lower the virus titer in lung tissue, and alleviate the symptoms of infection to 17 promote recovery. The aim of this study was to identify the key factors and regulatory mechanisms 18 in the PD-1 checkpoint blockade – responsive immune microenvironment in severe influenza 19 infection. 20 Methods: A BALB/c mouse model of severe influenza A/H1N1 infection was constructed, and 21 whole-transcriptome sequencing of mice treated with PD-1 checkpoint blockade before severe 22 A/PR8(H1N1) influenza infection and IgG2a isotype control before infection were performed. 23 Subsequently, the differential expression of nucleic acids between these two groups was analyzed, 24 followed by functional interaction prediction analysis to investigate gene-regulatory circuits. 25 Results: In total, 84 differentially expressed (dif) mRNAs, 36 dif-microRNAs (miRNAs), 90 26 dif-lncRNAs (long noncoding RNAs), and 22 dif-circRNAs (circular RNAs) were found in PD-1 27 antagonist treated A/PR8(H1N1) influenza infection lung compared with the controls (IgG2a 28 isotype control treated before infection). In spleens between the above two groups, 45 dif-mRNAs, 29 36 dif-miRNAs, 57 dif-lncRNAs, and 24 dif-circRNAs were identified. Direct function 30 enrichment analysis of dif-mRNAs and dif-miRNAs showed that these genes were mainly 31 involved in myocardial damage related to viral infection, mitogen activated protein kinase (MAPK) 32 signaling pathways, RAP1 (Ras-related protein 1) signaling pathway, and Axon guidance. Finally, 33 595 interaction pairs were obtained for the lungs and 462 interaction pairs for the spleens were 34 obtained in the competing endogenous RNA (ceRNA) complex network, in which the 35 downregulated mmu-miR-7043-3p and Vps39-204 were enriched significantly. 36 Conclusions: The present study provided a basis for the identification of potential pathways and 37 hub genes that might be involved in the PD-1 checkpoint blockade – responsive immune 38 microenvironment in severe influenza infection. 39

the reference, and the modified software mirdeep2 and sRNA-tools-cli were used to obtain the 112 potential miRNA and draw the secondary structures. The software miREvo and mirdeep2 were 113 integrated to predict novel miRNAs. We followed the following priority rule: Known miRNA > 114 rRNA > tRNA > snRNA > snoRNA > repeat > gene > NAT-siRNA > gene > novel miRNA > 115 ta-siRNA to make every unique small RNA mapped to only one annotation. The known miRNAs 116 used miFam.dat (http://www.mirbase.org/ftp.shtml) to look for families; novel miRNA precursors 117 were submitted to Rfam (http://rfam.sanger.ac.uk/search/) to look for Rfam families. Predicting 118 the target genes of the miRNAs was performed using miRanda. Differential expression analysis 119 was performed using the DESeq R package (1.8.3) with a P-value of 0.05 set as the threshold. The 120 P-values was adjusted using the Benjamini & Hochberg method. 121 Gene Ontology (GO) enrichment analysis was used on the target gene candidates of the 122 differentially expressed miRNAs. We used KOBAS software to test the statistical enrichment of 123 the target gene candidates in KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways. 124

Library preparation and sequencing for lncRNAs 125
A total of 3 μg RNA per sample was used as input material to construct sequencing libraries, 126 which were generated using the rRNA-depleted RNA by NEB Next® Ultra™ Directional RNA 127 Library Prep Kit for Illumina®. The clustering of samples was performed on a cBot Cluster 128 Generation System using TruSeq PE Cluster Kit v3-cBot-HS (Illumina), the libraries were 129 sequenced on an Illumina Hiseq 4000 platform and 150 bp paired-end reads were generated.

Data analysis of lncRNAs 131
Clean data were obtained by removing reads containing adapter or poly-N sequences. An index 132 of the reference genome was built using bowtie2 v2.2.8 and paired-end clean reads were aligned to 133 the reference genome using HISAT2 v2.0.4. The mapped reads of each sample were assembled 134 using StringTie (v1.3.1) in a reference-based approach. We used phyloFit to compute phylogenetic 135 models for conserved and non-conserved regions and then submitted the model and HMM 136 transition parameters to phyloP to compute a set of conservation scores of lncRNAs and coding 137 genes. We clustered the genes from different samples using weighted gene co-expression network 138 analysis (WGCNA) to search for common expression modules and then analyzed their function 139 through functional enrichment analysis. Transcripts with P-adjust < 0.05 were assigned as 140 differentially expressed. GO enrichment analysis and KEGG pathway enrichment analysis were 141 performed as above. 142

Competing Endogenous RNA Network Construction 231
According to the dif-lncRNA-dif-miRNA pairs and dif-miRNA-dif-mRNA pairs, differentially 232 expressed lncRNAs and mRNAs regulated by the same miRNA were screened. In total, 77 233 lncRNA-miRNA-mRNA interactions in lungs were finally obtained (Figure 8  and subsets, and axon guidance molecules were proven to have a large pentapeptide overlap, thus 328 immune cross-reactivity between influenza HA and axon guidance molecules is possible 28-30 . PD-1 329 signaling inhibits Rap guanine nucleotide exchange factor 1 (RAPGEF1 also known as C3G) 330 phosphorylation by utilizing SHP-1/2 (also known as protein tyrosine phosphatase non-receptor 331 25 type 6 and type 11), and reduced levels of phosphorylated C3G result in reduced RAP1 activation 332 and adhesion to intercellular adhesion molecule 1 (ICAM-1) to inhibit T-cell adhesion. Several 333 studies suggested that sepsis-induced upregulation of PD-1 has an impact on the motility and 334 migratory capacity of T lymphocytes by regulating classical inhibitory motif recruitment, 335 activation of the phosphatases SHP-1/2, and signaling through RAP1 31 . 336 Additionally, we identified the significant role of downregulated mmu-miR-7043-3p and 337 Vps39-204 in the ceRNA network. Increased expression of mmu-miR-7043-3p was proven to be 338 one of remarkable miRNA signatures of myocardial reductive stress, which is associated with 339 cardiac hypertrophy 32 . Future mechanistic studies are needed to determine the role of 340 miR-7043-3p in PD-1/PD-L1 pathway-associated viral damage in severe influenza infection. 341 VPS39 is a member of the vacuolar tethering complex that promotes late endosome formation, and 342 evidence has shown that silencing VPS39 can increase the proliferation of aged human T cells and 343 memory responses of lysosome-defective T cells in a mouse viral infection model 33 , and thus 344 might play important roles in antiviral immunity. 345

Conclusions 346
In conclusion, this study explored the molecular mechanism of the PD-1 checkpoint Vps39-204 were most significantly enriched by PD-1 blockade. However, this study was limited 353