Mouse models of COVID-19 recapitulate inflammatory pathways rather than gene expression

Abstract

How well mouse models recapitulate the transcriptional profiles seen in humans remains debatable, with both conservation and diversity identified in various settings. Herein we use RNA-Seq data and bioinformatics approaches to analyze the transcriptional responses in SARS-CoV-2 infected lungs, comparing 4 human studies with the widely used K18-hACE2 mouse model, a model where hACE2 is expressed from the mouse ACE2 promoter, and a model that uses a mouse adapted virus and wild-type mice. Overlap of single copy orthologue differentially expressed genes (scoDEGs) between human and mouse studies was generally poor (≈15–35%). Rather than being associated with batch, sample treatment, viral load, lung damage or mouse model, the poor overlaps were primarily due to scoDEG expression differences between species. Importantly, analyses of immune signatures and inflammatory pathways illustrated highly significant concordances between species. As immunity and immunopathology are the focus of most studies, these mouse models can thus be viewed as representative and relevant models of COVID-19.

SciScore for 10.1101/2022.02.24.481866: (What is this?)

Please note, not all rigor criteria are appropriate for all manuscripts.

Table 1: Rigor

Ethics	IACUC: Mouse work was approved by the QIMR Berghofer Medical Research Institute animal ethics committee (P3600, A2003-607). Euthanasia Agents: For intrapulmonary inoculations via the intranasal route, mice were anesthetized using isoflurane. Field Sample Permit: All infectious SARS-CoV-2 work was conducted in a dedicated suite in a biosafety level-3 (PC3) facility at the QIMR Berghofer MRI (Australian Department of Agriculture, Water and the Environment certification Q2326 and Office of the Gene Technology Regulator certification 3445).
Sex as a biological variable	Injected zygotes were transferred into the uterus of pseudo pregnant F1 females.
Randomization	not detected.
Blinding	not detected.
Power Analysis	not detected.
Cell Line Authentication	Contamination: Virus stocks were prepared in Vero E6 cells as described (29) and were checked for mycoplasma as described (56).

Table 2: Resources

Experimental Models: Cell Lines
Sentences	Resources
Viral titrations were performed at 5 days post-infection with a CCID50 assay using Vero E6 cells and serial dilution of supernatants from homogenized tissues as described previously (29).	Vero E6 suggested: RRID:CVCL_XD71)
Experimental Models: Organisms/Strains
Sentences	Resources
K18-hACE2 mice: K18-hACE2+/- mice were purchased from Jackson laboratories and were maintained in-house as heterozygotes by backcrossing to C57BL/6J mice (27, 28).	K18-hACE2 suggested: RRID:IMSR_GPT:T037657) K18-hACE2+/- suggested: RRID:IMSR_GPT:T037657)
The mACE2-hACE2 mouse line was maintained in-house as heterozygotes by backcrossing onto C57BL/6J mice.	C57BL/6J suggested: RRID:IMSR_JAX:000664)
Software and Algorithms
Sentences	Resources
Quality control of fastq files was performed using FastQC v0.11.9 (58).	FastQC suggested: (FastQC, RRID:SCR_014583)
Reads were trimmed to remove adapter content, size-selected to remove reads less than 36nt in length, and quality-filtered to remove reads with less than a Q20 Phred score within a sliding-window tetramer, using Trimmomatic v0.36 (60)	Trimmomatic suggested: (Trimmomatic, RRID:SCR_011848)
Processed reads were aligned to either the GRCm39 vM26 or GRCh38 v37 reference genome for mouse and human datasets, respectively, using STAR aligner v2.7.1a (61).	STAR suggested: (STAR, RRID:SCR_004463)
The number of reads mapping to SARS-CoV-2 was calculated using Samtools v1.9 (62).	Samtools suggested: (SAMTOOLS, RRID:SCR_002105)
Host gene expression was calculated using RSEM v1.3.1 (63) and differential expression was calculated using Bioconductor v3.13 (64) and EdgeR v3.34.0 (65) in R v4.1.0 (66)	RSEM suggested: (RSEM, RRID:SCR_013027) Bioconductor suggested: (Bioconductor, RRID:SCR_006442) EdgeR suggested: (edgeR, RRID:SCR_012802)
Mouse-human orthologues were extracted from the Ensembl database using BiomaRt v2.48.2 (67) in R.	Ensembl suggested: (Ensembl, RRID:SCR_002344) BiomaRt suggested: (biomaRt, RRID:SCR_019214)
The proportion of up- and down-regulated DEGs and scoDEGs shared between groups was calculated in R and plotted using ggVennDiagram v1.1.4 (69) in R.	ggVennDiagram suggested: None
Reciprocal gene set enrichment analysis: For each group, a log2 fold-change ranked gene list was produced using DESeq2 (70) with default settings.	DESeq2 suggested: (DESeq, RRID:SCR_000154)
A Gene Set Enrichment Analysis using GSEA v4.1.0 (40) with 100 permutations and the ‘no_collapse’ setting was used to test for enrichment of filtered orthoDEG sets within ranked gene lists.	GSEA suggested: (SeqGSEA, RRID:SCR_005724)
Pathway analysis: Pathway analysis was performed using Ingenuity Pathway Analysis (IPA) v65367011 (Qiagen) with default settings.	Ingenuity Pathway Analysis suggested: (Ingenuity Pathway Analysis, RRID:SCR_008653)
Data were plotted using pheatmap v1.0.12 (72) and ggplot2 v3.3.3 (73) in R.	pheatmap suggested: (pheatmap, RRID:SCR_016418) ggplot2 suggested: (ggplot2, RRID:SCR_014601)
Each network was then exported in tabular format and plotted using Cytoscape v3.8.2 (74).	Cytoscape suggested: (Cytoscape, RRID:SCR_003032)
Nnt genotyping: Mouse RNA-Seq data were interrogated for the presence of exon two and nine of the nicotinamide nucleotide transhydrogenase (Nnt) gene as described (52) using Repair and BBduk from the BBmap package v38.90.	BBmap suggested: (BBmap, RRID:SCR_016965)
Statistics: Statistics were performed using IBM SPSS Statistics for Windows, version 19.0.	SPSS suggested: (SPSS, RRID:SCR_002865)

Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:

There are clearly a number of limitations for this kind of analysis. Unavoidable is the issue of single copy orthologues, which comprised 63-77% of genes identified by RNA-Seq in lung tissues. This issue is less of a problem for pathway analyses when using programs such as IPA that accept both human and mouse gene nomenclature. The different sources of tissues and the different technologies used to generate gene expression data (Table 1) likely add to non-biological variability, although this was perhaps mitigated herein by combining multiple human and mouse studies. The large differences in viral loads between some groups (Figure 2A) would appear to play a role in the poor concordance in gene expression profiles, particularly for human groups and down-regulated genes. However, analyses of K18-hACE2 (high viral load) and mACE2-hACE (lower viral loads) argued that the difference in viral loads was not a major player in the poor overlap in up-regulated orthoDEGs for mouse vs. human groups. In summary, the analyses herein argue that overlap in orthoDEG expression in the lung tissues of hACE2-transgenic mice and humans after SARS-CoV-2 infection is generally poor. In contrast, the concordance in immune and inflammation pathways was high, arguing that the transgenic mouse models provide relevant and pertinent models in which to evaluate new interventions for SARS-CoV-2 and COVID-19.

Results from TrialIdentifier: No clinical trial numbers were referenced.

Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

Results from JetFighter: We did not find any issues relating to colormaps.

Results from rtransparent:

Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
No protocol registration statement was detected.

Results from scite Reference Check: We found no unreliable references.

Read the original source

Mouse models of COVID-19 recapitulate inflammatory pathways rather than gene expression

This article has been Reviewed by the following groups

Listed in

Abstract

Article activity feed