key: cord-0313545-q3vc5yof authors: Silva, João M. F.; Nagata, Tatsuya; Melo, Fernando L.; Elena, Santiago F. title: Heterogeneity in the response of different subtypes of Drosophila melanogaster enteroendocrine cells to viral infections date: 2020-07-02 journal: bioRxiv DOI: 10.1101/2020.07.01.182352 sha: 6b953629c9b6a20637d16fcc3e09b6d83e851b68 doc_id: 313545 cord_uid: q3vc5yof Single cell RNA sequencing (scRNA-seq) offers the possibility to monitor both host and pathogens transcriptomes at the cellular level. Here, public scRNA-seq data from Drosophila melanogaster have been used to compare the differences in replication strategy and cellular response between two viruses, Thika virus (TV) and D. melanogaster Nora virus (DMelNV) in enteroendocrine cells (EEs). TV and DMelNV exhibited different patterns of replication and for TV, accumulation varied according to cell subtype. Cells infected with TV underwent down-regulation of genes that represent bottlenecks in the fruit fly interactome, while cells infected with DMelNV went through a down-expression of translation-related genes that represent both hubs and bottlenecks in the host interactome. In contrast, flies infected with DMelNV show only a systemic level down-regulation of bottleneck genes. Here, we use scRNA-seq to highlight the differences and commonalities between cellular response to TV and DMelNV and between cellular and systemic response to DMelNV. Enteroendocrine cells (EEs) play an important role in the control of many physiological processes. They are scattered throughout the gastrointestinal tract, making it the largest endocrine organ in 30 the body, and produce more than 20 neuropeptide hormones including allatostatines A, B and C (AstA, AstB and AstC, respectively), tachykinin (Tk), neuropeptide F (NPF), diuretic hormone 32 31 (DH31), and CCHamide-1 and -2 (CCHa1 and CCHa2, respectively) (Rehfeld, 2004 ; Furness 3 viral RNA in TV infected cells was always below 5% with only one exception (a single cell I-app), whereas for DMelNV, the percentage of viral RNA was up to ~80%. The percentage of 117 infected cells on each cluster also varied drastically between these viruses ( Figure 2C ). TV 118 infected a very low proportion of cells from subtypes located in the gastric region, and was found 119 on more than 60% of the cells of the I-p AstA cell-subtype located in the posterior region of the 120 gastrointestinal tract that is characterized by the expression of AstC, AstA and CCHa1. One-way ANOVA tests showed a significant influence of EE subtype on the expression level of 122 TV (P = 7.3´10 -4 ) but not of DMelNV (P = 0.12). Following these results, Tukey's HSD post 123 hoc tests revealed two well defined groups exhibiting, respectively, the highest and lowest levels (adjusted P = 9.7´10 -7 ), whereas a negative correlation was found for tetraspanin 42Ec (Tsp42Ec; 136 P = 0.014), CDC50 (P = 0.022) and CG14989 (P = 0.004). The most significant correlation was 137 found for mt:lrRNA, which plays a role in the formation of granular poles in germ line progenitor 138 cells (Iida and Kobayashi, 1998) . GO enrichment analyses were performed for nine and 79 genes 139 which were positively and negatively correlated, respectively, to the replication level of DMelNV 140 ( Figure 3 ). An overrepresentation in processes mostly related to translation was found in 141 negatively correlated genes but no overrepresentation was found for the few positively correlated 142 genes. For the midgut atlas dataset, the expression of 112 genes was found to be correlated to the 143 replication level of DMelNV, with 27 correlated genes found in both datasets (supplementary file In addition to the gene expression correlation analyses we also investigated differentially expressed genes (DEGs) according to cell infection status (infected or uninfected) and the in the second class, referred to as cell-subtype-specific response genes, we found 21 and 13 DEGs 156 for DMelNV and TV, respectively. These categories are not exclusive, and some DEGs are also The results obtained for DMelNV were only somewhat reproducible for the midgut atlas dataset, 183 which is not surprising given that new cell types were analyzed. Translation-related processes were enriched in both lists of generic response down-regulated genes and cell-subtype-specific response genes. Viral processes were also overrepresented in cell-subtype-specific responses ribosome entry site (IRES)-mediated translation of two picorna-like viruses from the family (Jannot et al., 2011) . The degree of DEGs were analyzed by comparing their degree distribution with that of the 212 complete fly interactome. Figure 6 and Figure S6A shows the power-law fit to the log-degree 213 distribution of the complete interactome network and subnetworks constructed using only DEGs 214 found in the EE and midgut atlas dataset, respectively. The values of the critical exponent, mean exponent estimated for the DEGs was significantly smaller than for the whole interactome only infection led to an under-expression of hub genes with a high number of interactions. Next, the indicates that while DMelNV infection leads to a down-regulation of both hub and bottleneck 224 genes, TV infection leads only to a down-regulation of bottleneck genes. The cellular response to DMelNV, but not the systemic response, is characterized by a down- it with our previously obtained results. It is important to note, however, that the cellular response 231 was obtained by comparing infected/uninfected cells but we are unaware whether uninfected cells 232 were exposed to viruses. Another limitation of this approach is that bulk RNA-seq of infected 233 flies will average the gene expression for all cell types, including infected and uninfected cells. Five and four genes were simultaneously down-and up-regulated, respectively, in both cellular 235 and systemic responses (supplementary file S1). Surprisingly, two genes, Hsc70-3 and Hsc70Cb, In contrast, we failed to detect any perturbation of hub genes in the interactome for TV infection. One reason why modulation of hub genes upon TV infection was not detected may be that we evidence for modulation of hub or bottleneck genes was found for cell-subtype-specific response and Tk are differentially expressed in cells infected by DMelNV or TV in a cell-subtype-specific Last, given the presence of unacknowledged viruses in public scRNA-seq data, we hypothesize 328 that viruses may be confounding factor in these kinds of experiments. In the data analyzed here, 329 viruses did not seem to cluster based on infection status (Figure 1 and Figure S2A ). However, Here, through analysis of in vivo scRNA-seq data, we show the similarities and differences in the To simplify the output, child terms were removed when the parent was also present, with the 513 exception of translation (GO:0006412) and neuropeptide signaling pathway (GO:0007218). All DEGs are available at supplementary file S1. conducted a step-wise analysis of the false-positive rate by randomly assigning uninfected cells 518 as infected and performing two-way ANOVA tests as described above. For each step, 100 two-519 way ANOVA tests were carried out with 10 to 500 cells randomly assigned as infected, using an 520 incremental step of 10. Note that the tests were executed on the lists of 1460 and 307 genes for 521 the EE and midgut atlas datasets, respectively, meaning that for each step, 146,000 and 30,700 522 two-way ANOVA tests were performed for each dataset. Generation of enteroendocrine cell diversity in 556 midgut stem cell lineages Network biology approach to complex 558 diseases Differential gene expression 560 related to Nora virus infection of Drosophila melanogaster Essential role of mitochondrially encoded large rRNA for 597 germ-line formation in Drosophila embryos The ribosomal protein RACK1 is required for microRNA function in both 600 C. elegans and humans A meta-analysis reveals the commonalities and differences in Arabidopsis thaliana 624 response to different viral pathogens Extreme heterogeneity of influenza virus 626 infection in single cells Single-cell analysis uncovers extensive biological noise in 628 poliovirus replication Cytoscape: a software environment for integrated models 631 of biomolecular interaction networks Comprehensive integration of single-cell data Convergent 643 evolution of argonaute-2 slicer antagonism in two distinct insect RNA viruses The discovery, 647 distribution, and evolution of viruses associated with Drosophila melanogaster Dual RNA-seq reveals viral 651 infections in asthmatic children without respiratory illness which are associated with changes in 652 the airway transcriptome Single-cell analysis of the impact of host cell heterogeneity on infection with foot-and-655 mouth disease virus Single-cell transcriptional 657 dynamics of flavivirus infection Massively parallel digital 660 transcriptional profiling of single cells