3D reconstruction of SARS-CoV-2 infection in ferrets emphasizes focal infection pattern in the upper respiratory tract

The visualization of viral pathogens in infected tissues is an invaluable tool to understand spatial virus distribution, localization, and cell tropism in vivo. Commonly, virus-infected tissues are analyzed using conventional immunohistochemistry in paraffin-embedded thin sections. Here, we demonstrate the utility of volumetric three-dimensional (3D) immunofluorescence imaging using tissue optical clearing and light sheet microscopy to investigate host-pathogen interactions of pandemic SARS-CoV-2 in ferrets at a mesoscopic scale. The superior spatial context of large, intact samples (> 150 mm3) allowed detailed quantification of interrelated parameters like focus-to-focus distance or SARS-CoV-2-infected area, facilitating an in-depth description of SARS-CoV-2 infection foci. Accordingly, we could confirm a preferential infection of the ferret upper respiratory tract by SARS-CoV-2 and emphasize a distinct focal infection pattern in nasal turbinates. Conclusively, we present a proof-of-concept study for investigating critically important respiratory pathogens in their spatial tissue morphology and demonstrate the first specific 3D visualization of SARS-CoV-2 infection.

and Supplementary Movie S1). While there were some unspecific signals detectable 2 2 5 in the SARS-CoV-2 N-stained sample (individual green or magenta spots), they could 2 2 6 be clearly distinguished from specific SARS-CoV-2 detection by the absence of 2 2 7 colocalization (white) of the signals from either antibody ( Figure 2B and Figure S1).

8
Within the about 4-mm-thick URT sample (4 days post-infection), multiple 2 2 9 comparatively small SARS-CoV-2 infection hot spots were visualized ( Figure 2B).   colocalization of both independent antibody stainings ( Figure 3B and C, right side).  and Table 1). This increased spatial context within the nasal turbinate sample is   and C, red square) or using 3D rendering to virtually "fly through the sample" emphasizes the system's flexibility to switch from broad, mesoscopic overviews to 2 5 5 detailed, resolved close-ups. By maintaining the full infection environment, we were 2 5 6 able to establish quantifiable relations between the individual SARS-CoV-2 foci and 2 5 7 highlight the oligofocal infection pattern of SARS-CoV-2 in the URT of ferrets.  segmented SARS-CoV-2 infection foci. Linear distances were calculated either as 2 6 0 the distance between the center of two foci or as the shortest possible distance 2 6 1 between the edges of two foci. The area affected by SARS-CoV-2 infection was 2 6 2 measured by calculating the surface area of the segmented objects and dividing the 2 6 3 resultant value by two, thus, only accounting for the surface facing outwards. the optically cleared high-volume tissue sample to 1-mm thick slices using a tissue 2 7 1 matrix to achieve compatibility with the limited free working distances of CLSM 2 7 2 objectives ( Figure 1A).

7 3
Using the spatio-morphological information on the distribution of SARS-CoV-2  CoV-2 infection foci in the LRT, we looked at optically cleared high-volume lung and 2 8 6 tracheal samples. As before, some unspecific fluorescence signals could be seen in both lung (Figure suggested detection of debris-associated antigen, which was mostly likely inhaled 2 9 5 from the URT. Overall, while we were able to detect an 8.6x10 -5 mm 3 (86,000 µm 3 ) 2 9 6 spot of debris-associated antigen within a > 80 mm 3 volume, we did not identify 2 9 7 additional sites of infection within the LRT of ferrets, which corroborates preferential 2 9 8 replication of SARS-CoV-2 in the URT of ferrets. morphological context of ferret nasal turbinates ( Figure 4 and Table 1).

1 2
Here, we employed an ECi-based TOC approach [59] and adjusted it to visualize utility of a phase contrast x-ray tomography concept to investigate unstained lung PRO-3 for nuclear contrast and Eosin-Y for cytoplasmic/stromal contrast) to achieve blocks, specificity for SARS-CoV-2 was ensured via colocalization of either staining 3 2 7 (Figures 2, 3, 5, and 6) and absence of colocalization in naïve animals ( Figure S1).

2 8
Further optimization of the immunostaining protocol or the availability of SARS-CoV-3 2 9 2-specific antibodies will likely aid in reduction of background staining and 3 3 0 improvement of virus detection.

1
In addition to the specific 3D reconstruction of SARS-CoV-2 infection within its spatio- following accurate quantification of interrelated 3D parameters ( Figure 4 and Table 1) represents a pronounced advantage of 3D immunofluorescence imaging over conventional IHC. To achieve a somewhat comparable yet more artifact-prone 3D    foci in narrow areas of the URT might also have implications for the likelihood of comparison to nasal washes from ferrets and possibly other animal models.

5 7
Accordingly, a high degree of variation in viral copy numbers can be observed from high-volume imaging approach, we aimed to screen the tissue for rare SARS-CoV-2 positive structure inside a lung airway ( Figure 6B). However, spatial analysis turbinate epithelium (Figures 2 and 3), was detected above the airway epithelial 3 8 0 layer. This strongly suggested that the structure most likely represents aspirated LSFM approach to identify rare and highly localized pathogen-related events.