Fig 1.
Viral loads and immunohistochemistry of viral antigen in lungs of BA.1 infected K18-hACE2 mice.
a Lung tissue titers at the indicated days post infection with BA.1 (limit of detection per mouse ≈ 2 log10CCID50/g). b Viral read counts per million reads, determined by RNA-Seq for K18-hACE2 mice infected with BA.1 or inoculated with UV-inactivated BA.1 virus. Statistics by Kolmogorov Smirnov exact tests. c Viral genome copy number determined by RTqPCR for K18-hACE2 mice infected with BA.1 or an original strain isolated SARS-CoV-2QLD02. Statistics by Kolmogorov-Smirnov exact test. d Immunohistochemistry of BA.1-infected lungs at 2 dpi, stained with an anti-spike monoclonal antibody. Staining of bronchial epithelia cells (dark brown) can be seen top left and bottom right (left image), with enlargements of these two areas shown in the right 2 images. e As for d showing staining of cells that surround alveoli with morphology consistent with that of pneumocytes (alveolar epithelial cells). * alveolar air sacs.
Fig 2.
Histology of lungs after BA.1 infection of K18-hACE2 mice.
a H&E staining of lung tissue sections at the indicated days after infection/inoculation. Green dotted lines encircle overt foci of cellular infiltrates. b Quantitation of cellular infiltrates. All cellular infiltrates were marked as in “a” and the summed infiltrate areas expressed as a percentage of the lung section area for each mouse. Statistics by Kolmogorov Smirnov exact tests. (For BA.1 infected mice, data for 2, 5 and 10 dpi are not significantly different, while 30 dpi shows a significant reduction from 5 and 10 dpi by t tests, p = 0.0.019 and p = 0.039, respectively). c Whole H&E stained lung sections from day 5 after infection/inoculation, illustrating reduced white space in lungs from infected mice. d QuPath digital analysis of white space in H&E stained whole lung sections such as those shown in “c”. Statistics by t test.
Fig 3.
Cellular infiltrates and types identified by RNA-Seq of lungs for 2–30 dpi.
a Z scores for the indicated annotations from Ingenuity Pathway analysis (IPA) Diseases or Function analysis plotted against dpi. (N.I. not identified; for 66 dpi there were insufficient DEGs for IPA analysis). b Using cellular deconvolution (SpatialDecon) and the gene expression matrices from the “Mouse Cell Atlas (MCA) Lung Cell expression matrix” and the “NanoString (NS) Immune Cell Family expression matrix”, cell types whose abundance scores were significantly different on day 5 were identified (q values provided). Blue text indicates cell type associated with inflammation resolution. c As for b but for day 10. Blue text indicates cell type associated with inflammation resolution. We assume “colon” T reg are well annotated in the NS matrices (and are thus readily identified), and represent tissue T regs, rather than implying that colonic T regs have migrated to the lungs.
Fig 4.
Inflammatory pathways identified by RNA-Seq of lungs for 2–30 dpi.
a Heat maps of cytokine UpStream Regulator (USR) z scores and significance (p values, all p<0.05) determined by IPA. White boxes means no cytokine USR annotation was returned for 30 dpi by IPA. RNA-Seq identified too few DEGs on 66 dpi for IPA analysis. b IPA cytokine USRs from BA.1 infected K18-hACE2 mice (2 dpi) were compared with those previously reported for the more severe infection of K18-hACE2 mice by an original strain isolate [55]. Differences in cytokine z scores for 2 dpi (with a cutoff filter of >0.2 and < -0.2) are plotted. BA.1 infected lungs show slightly increased z scores for some cytokine signatures associated with reduced disease severity (green) and slightly reduced z scores for some cytokine signatures associated with increased disease severity (red).
Fig 5.
RNA-Seq data for day 66 dpi, with DEGs clustered by broad functions. The whole gene list was also interrogated by GSEAs using gene sets from MSigDB and Blood transcription modules.
Fig 6.
Genes whose expression correlated with dpi.
a Using the whole gene lists, genes whose expression showed a statically significant positive correlation with dpi were identified. After removing genes that also positively correlated with dpi in the UV-inactivated BA.1 inoculated control groups, 2786 genes remained. The 2786 genes were analyzed by Enrichr and IPA (for the latter the r value was entered in the log2 expression column). Significant results clustered into two categories ‘Ciliated epithelial cell’ and ‘Protein synthesis’. Of the 2786 genes, 59 were RPL (ribosomal protein) genes. b Using the whole gene lists, genes whose expression showed a statically significant negative correlation with dpi were identified. After removing genes that also negatively correlated with dpi in the UV control groups, 1996 genes remained. The 1996 genes were analyzed by IPA as above, with some top annotations shown. c Normalized expression data (VST ‐ Variance Stabilizing Transformation) for five of the 1996 genes are plotted against dpi, with linear regression lines and 95% confidence intervals. Pearson correlation coefficients (r), coefficients of determination (r2) and corrected significances, FDR (q), are provided for each gene. d Using cellular deconvolution (as in Fig 3B and 3C), cell types whose abundance scores significantly positive correlated with dpi were identified. Linear regression lines and 95% confidence intervals, and Pearson correlation coefficients (r), coefficients of determination (r2) and corrected significances, FDR (q) are provided. (Full data sets are provided in S6 Table).