Fig 1.
BLT-L mice effectively resolve SARS-CoV-2 infection.
(A) Representative schematic of BLT-L mice. Created with BioRender.com. (B) Fetal lung xenografts (fLX) of BLT-L mice were infected with 1x106 PFU of SARS-CoV-2 WA-1 isolate via subcutaneous intra-graft injection. Created with BioRender.com. (C) BLT-L mice were monitored for weight change over the course of 12 days post-infection (dpi). (D) Viral titer (log(PFU/mg)) in fLX at 2, 6, and 12 dpi as determined by plaque assay. (E-H) Immunohistochemistry for SARS-CoV-2 S and (I-L) N protein was performed on naïve fLX (E,I) and fLX at 2 dpi (F,J), 6dpi (G,K), and 12 dpi (H,L). Scale bar = 100 μM. Data are representative of two to three independent cohorts and donors. n = 6-10 per timepoint. Error bars represent mean ± standard error of the mean. One-way ANOVA analysis was performed. p-values are indicated. n.s. = non-significant.
Fig 2.
BLT-L mice resolve histopathological damage in fLX.
(A-E) Hematoxylin and eosin staining was performed on naïve fLX (A) and fLX at 2 dpi (B-E), 6dpi (F), and 12 dpi (G). Scale bar = 100 μM. C: Bronchiolar attenuation and denuding (blue arrows) correlate directly with the presence of SARS-CoV-2 spike protein. D: Alveolar spaces are multifocally filled with necrotic debris admixed with neutrophils and edema (blue hashes). Type II pneumocytes (ATII) are often denuded or attenuated in areas of inflammation that correlate directly with the presence of SARS-CoV-2 spike protein. Fibrin thrombi are routinely observed in neighboring parenchyma (black asterisks). E: Coagulative necrosis as evidenced by loss of cellular detail and generalized eosinophilia (above the black hashed line) with numerous regional fibrin thrombi (black asterisks). Although SARS-CoV-2 spike protein is not observed in the area of coagulative necrosis, viral antigens are located within adjacent tissues. (H) Histopathological scoring of naïve fLX and fLX at 2, 6, and 12 dpi. Data are representative of two to three independent cohorts and donors. n = 6 -10 per timepoint. Error bars represent mean ± standard error of the mean. One-way ANOVA analysis was performed. p-values are indicated. n.s. = non-significant.
Fig 3.
Cellular remodeling in fLX upon SARS-CoV-2 infection.
Single-cell RNA sequencing analysis was performed on single-cell suspensions of naïve fLX and fLX at 2 and 12 dpi. Naïve n = 3 fLX (9,605 cells), 2 dpi n = 2 fLX (4,405 cells), and 12 dpi n = 2 fLX (5,857 cells). (A) UMAP plot clustering of the human cell compartment of naïve fLX and fLX at 2 and 12 dpi. (B) Temporal annotation of the human clusters on the UMAP plot: naïve (black), 2 dpi (red), 12 dpi (blue). Subclusters unique to 2 dpi are annotated by A (T cell compartment), B (myeloid compartment) and C (mesenchymal compartment). (C) Frequency of each cell compartment determined by single-cell RNA-sequencing.
Fig 4.
Cellular compartmentalization of viral RNA during infection.
(A) Dot plots displaying the expression of SARS-CoV-2 viral RNA transcripts across the different human clusters and time points analyzed by scRNA-seq. (B) UMAP plot showing the distribution of SARS-CoV-2 viral RNA transcripts by density across all human cell clusters and time points analyzed by scRNA-seq. (C) UMAP plot showing the distribution of SARS-CoV-2 viral RNA reads across all human cell clusters and time points analyzed by scRNA-seq. Inset showing distribution of viral RNA reads in the mesenchymal cluster.
Fig 5.
Remodeling of the epithelial compartment of fLX during infection.
(A) UMAP plot subclustering of the human epithelial compartment in fLX across all time points analyzed by scRNA-seq. (B) Temporal annotation of the human epithelial subclusters on the UMAP plot: naïve (black), 2 dpi (red), 12 dpi (blue). The inset graph indicates the change in frequency of each compartment (AT2; black, Basal; pink, Secretory; teal, and AT1; purple) within the epithelial compartment. Red line indicates the change in the AT2 compartment over time. (C) Heatmap displaying the relative expression (from -1 to 1) of the top differentially regulated genes over the course of infection within the AT2 compartment.
Fig 6.
Macrophage-like cells emerge upon acute infection and are enriched in viral RNA.
(A) UMAP plot showing the subclustering of the human T-cell compartment of fLX across all time points analyzed by scRNA-seq. (B) Temporal annotation of the human T-cell subclusters on the UMAP plot: naïve (black), 2 dpi (red),2 dpi (blue). (C) UMAP plot (all time points) showing the distribution of SARS-CoV-2 viral RNA transcripts in the T-cell compartment. (D) GO-term analysis showing the most highly upregulated signaling pathways in MLC (subcluster 6). (E) Relative expression of T cells, macrophage, mesenchymal and antiviral markers within each of the T-cell subclusters (From 0 to 7; as labeled in panel A).
Fig 7.
CD163 + extravascular inflammatory monocytes arise during acute viral infection and are predominant antiviral mediators.
(A) UMAP plot showing subclustering of the human myeloid compartment of fLX across all time points analyzed by scRNA-seq. (B) Temporal annotation of the myeloid subclusters on the UMAP plot: naïve (black), 2 dpi (red), 12 dpi (blue). iMO subcluster is indicated with a blue circle. (C) Frequency of each myeloid subcluster by timepoint. (D-F) UMAP plot showing the expression of MRC1 (D), CD163 (E), and CD4 (F) across myeloid subsets. (G) Relative expression of highly upregulated ISGs and inflammatory markers in each of the myeloid subclusters. (H) UMAP plot showing the distribution of SARS-CoV-2 viral RNA transcripts in the myeloid compartment across all time points analyzed by scRNA-seq. (I) Violin plots showing the expression level of differentially expressed genes between SARS-CoV-2 positive and negative iMO (subcluster 3). (J) Heatmap displaying the relative expression (from -1 to 1) of the top differentially regulated genes over the course of infection within the patrolling monocyte (0) compartment. (K) Multiplex fluorescent immunohistochemistry of naïve fLX and fLX at 2 and 12 dpi. The images below represent 20x magnification of the indicated area (red). Images are representative of three independent fLX. CD163: teal, CD4: yellow, CD3: orange, PPARγ: green, Dapi: gray. Two representative images. Scale bar = 50 μM.
Fig 8.
Mesenchymal and endothelial signatures of infection resolution.
(A) UMAP plot subclustering of the human mesenchymal compartment in fLX across all time points analyzed by scRNA-seq. (B) Temporal annotation of the mesenchymal subclusters on the UMAP plot: naïve (black), 2 dpi (red), 12 dpi (blue). (C) Frequency of mesenchymal subclusters. (D) GO-term analysis showing cellular pathways linked to significantly upregulated genes within the activated fibroblasts subcluster. (E-F) UMAP plot showing CXCL10 (E) and TIMP1 (F) expression within all human mesenchymal subclusters and across all time points. Activated fibroblasts are indicated by a red circle in A and B. (G) UMAP plot subclustering of the human endothelial compartment of fLX across all time points analyzed by scRNA-seq. (H) Temporal annotation of the endothelial subclusters on the UMAP plot: naïve (black), 2 dpi (red), 12 dpi (blue). (I) List of upregulated (red) and downregulated (blue) genes specific to the activated endothelial subcluster (Cluster 2) over the other endothelial subclusters (p.adj ≤ 0.05). (J) Heatmap displaying the relative expression (from -1 to 1) of the top differentially regulated genes over the course of infection within the venous endothelial compartment.
Fig 9.
Systemic depletion of CD4 + cells results in persistent infection of fLX.
(A) BLT-L mice were administered 200 μg of anti-CD3e (OKT3), anti-CD4 (OKT4), anti-CD8 (OKT8) or IgG2a isotype. Created with BioRender.com. (B) Viral titer (log10(PFU/mg)) of fLX collected from BLT-L mice treated with isotype, OKT3, OKT4, or OKT8 antibody at 12 dpi. (C) Immunohistochemistry for SARS-CoV-2 N protein of fLX collected from antibody-treated BLT-L mice at 12 dpi. Images are representative of two independent cohorts and donors. (D) Multiplex immunohistochemistry of fLX collected from antibody-treated BLT-L mice at 12 dpi. Dapi = gray, CD3 = teal, CD4 = red. Scale bar = 50 μm. Images are representative of two independent cohorts and donors. (E-F) Multiplex fluorescent immunohistochemistry (E) and CD163 + area quantification (AQ) analysis (F) of fLX following infection resolution (12 dpi, isotype-treated), or during acute (2 dpi) and persistent infection (12 dpi, OKT4-treated). Images are representative of two independent cohorts and donors. CD163: teal, CD4: yellow, CD3: orange, PPARγ: green, Dapi: gray. Scale bar = 50 μM.
Fig 10.
MHC-I expression in fLX collected from naïve mice, antibody-treated mice and acutely infected mice.
MHC-I staining of fLX tissue sections collected from naive, isotype-treated (12 dpi), OKT4-treated (12 dpi), and acutely infected mice (2 dpi). Scale bar = 100 μM. Images are representative of two independent fLX for each experimental condition.
Fig 11.
CD4 + cell depletion results in signatures of chronic infection.
(A-C) Cytokine quantification (A: CCL2, B: CCL3, C: CXCL10) in the serum of naïve, infected (2 dpi and 12 dpi isotype-treated or not) and OKT4-treated BLT-L mice (12 dpi). (D) Violin plot showing expression level of CXCL10 per cell and within each human lineage. (E) UMAP plot showing CXCL10 expression within all human lineages in naive fLX and at 2- and 12 dpi. Locations of the mesenchymal, endothelial, myeloid and iMO clusters at 2 dpi are indicated. (F) Quantification of CCL19 in the serum of naïve, infected (2 dpi and 12 dpi isotype-treated or not) and OKT4-treated BLT-L mice (12 dpi). (G) Histopathological scoring of fLX collected from isotype- and OKT4-treated BLT-L mice (12 dpi). Error bars indicate mean ± Standard error of the mean. One-way ANOVA, t-test. p-values are indicated on graphs.
Fig 12.
Speculative model of SARS-CoV-2 infection resolution in human lung tissues.
During effective viral resolution and tissue repair (left panel), myeloid cells are recruited to the site of infection and differentiate into CD163 + inflammatory monocytes (iMO). iMO produce antiviral and inflammatory signals and exert phagocytic activity on viral particles. Additionally, CD3-expressing macrophage-like cells act as dead-end viral reservoirs, facilitating viral clearance, and elicit robust cytotoxic antiviral responses that overall potentiate infection resolution. In parallel, activated endothelial cells and fibroblasts contribute to the recruitment of myeloid cells and inflammatory response, respectively. Collectively, these events lead to viral clearance and infection resolution. In the context of persistent infection (left panel), absence of effective CD4 + monocyte recruitment prevents infection resolution. Created with BioRender.com.