Understanding influenza virus pandemic emergence through multimodal microscopy of infected lungs

Session

Session 6 - Imaging Across Scales

Authors

Calum Bentley-Abbot (2), Ryan Devlin (1), Jack McCowan (1), Anna Sims (2), Colin Loney (2), Ed Roberts (1), Edward Hutchinson (2)

Affiliations

1. Cancer research UK Scotland Centre
2. MRC-University of Glasgow Centre for Virus Research

Keywords

Influenza A virus, Co-infection, Respiratory disease, Light sheet microscopy

Abstract text

Introduction

One of the biggest imaging challenges in the study of disease is resolving microscopic events occurring within complex organs. The development of lightsheet microscopy (LiSM) has enabled the imaging of whole organs without the need for dissection, enabling researchers to study mesoscale processes, such as disease progression and pathology, within the anatomical context. However, most lightsheet systems cannot compete with the resolution offered by modern confocal microscopes and fail to reveal the mechanisms governing these events at the cellular level.

Viral infections cause some of the most serious diseases, and the study of these pathologies can benefit greatly from a multimodal approach. This effect is particularly well characterised for influenza A virus (IAV) infections. The proliferation of IAV throughout the lung during infection can only be clearly understood through whole lung mesoscale imaging. However, our in vitro studies show that interactions between IAVs result in novel cellular effects at the microscale. Following infection, there is a short window during which a cell can be infected by a second virus, after which secondary infection is cut off – a process known as superinfection exclusion (SIE).¹ As an IAV infection spreads, SIE divides the infected tissue into a patchwork of microscopic domains, restricting co-infection to a narrow band of cells between these regions. This could be of fundamental importance to influenza biology, as co-infected cells enable IAVs to exchange genetic material, resulting in new virus strains that cause influenza pandemics.

Although we have demonstrated that SIE applies to infections in vivo, we do not yet understand how this cellular process plays out across the 3D surfaces of the respiratory tract to control IAV co-infections. Here, we show how combining multiple imaging modalities, including LiSM, confocal microscopy and single molecule fluorescent in situ hybridisation (smFISH), can build a detailed model of how the genetic diversity of this major viral pathogen is controlled during the infection of a complex organ.

Methods

Using a mouse model of infection we employed lightsheet and confocal microscopy to study infected lungs across a range of length scales. Mice were infected with a mixture of IAVs encoding distinct fluorescent tags, enabling us to visualise viral interactions at sites of infection. Separately, mice were infected with two genetically distinct IAV strains (A/Puerto Rico 8/1934 and A/X31), lungs were sectioned and confocal microscopy and smFISH was used to characterise viral co-infection events. 

Results and Discussion

We applied LiSM to study the progression of disease in the lungs of mice co-infected with two fluorescently tagged IAVs. These studies provide the first mesoscale, whole organ insights into the progression of disease during IAV infection of the lung (Fig 1). During early infection (3 days post inoculation) viral populations are restricted to the larger airways immediately proximal to the trachea (Fig 1A), while by 6 days post inoculation these initial sites of infection are largely cleared, with the majority of the viral load found in minor peripheral airways (Fig 1B).

Figure 1: LiSM showing the progression of disease in the lungs of mice co-infected with IAV. Mice were co-infected with two fluorescently tagged viruses (green and red). Orange regions represent co-infected lesions. Some green autofluorescence from the tissue has been left in for spatial context. A: 3 days post inoculation. B: 6 days post inoculation. 

Interestingly, when studying co-infection in vivo using LiSM we see that the majority of lesions are co-infected, in contrast to the effects of SIE in vitro, where a patchwork of distinct microdomains is typical (Fig 2A).¹ This prompted us to investigate these lesions at higher resolution using thick section confocal microscopy. This revealed that regions which appear co-infected at the mesoscale actually comprise tightly constrained, singly-infected lesions at the microscale (Fig 2B). This resolves the observations made at different scales and shows that, as predicted in vitro, SIE restricts co-infection to a small number of cells between competing lesions in vivo (Fig 2C and D).

Figure 2: High resolution imaging of co-infected regions reveals SIE as observed in vitro. A: SIE between two fluorescent viral populations, green and magenta, in vitro.¹ B: Thick section confocal data showing SIE between fluorescent virus strains, magenta and orange, in the airway of a mouse. C: Using Imaris, a surface was rendered showing co-infected cells (cyan). D: Co-infected cell surface shown with the original image to demonstrate the restriction of co-infection to the periphery of lesions.

Having shown that SIE spatially restricts co-infection within the lung, we asked whether these co-infected cells enable genetic exchange between adjacent viral lesions. To do this, we designed smFISH probes targeting gene segments from two IAV strains. Specifically, our system targets the two gene segments encoding the IAV surface glycoproteins. These glycoproteins are the main targets for immune responses against influenza, and influenza pandemics typically result from IAVs exchanging these gene segments. Having validated these probes in a series of in vitro experiments, we are now employing this system in vivo to visualise the exchange of gene segments in co-infected cells for the first time. These findings will allow us to explore the variations in co-infection dynamics over the time course of infection, and the biological and anatomical conditions which promote genetic exchange between IAVs.

Conclusion

Here, we have employed a multi-length scale approach to better understand the mechanisms governing viral co-infection of the lower respiratory tract. Using LiSM we carried out the first visualisation of the progression of IAV disease in a whole, non-dissected lung, revealing the localisation of co-infected viral lesions throughout infection. Further study of these lesions using confocal microscopy enabled us to visualise the microscale mechanisms governing viral population dynamics on the cellular level, revealing highly constrained co-infected regions between viral populations. Our current smFISH studies will expand on this work, revealing how interactions between competing viruses within these co-infected interstitial regions contribute to pandemic strain emergence. Only by employing this multi-modal approach are we able to build a more complete understanding of influenza A virus pathology which will help to better inform public health and disease intervention across length scales.

References

 1 Sims A, Tornaletti LB, Jasim S, Pirillo C, Devlin R, Hirst JC, et al. (2023) Superinfection exclusion creates spatially distinct influenza virus populations. PLoS Biol 21(2): e3001941. https://doi.org/10.1371/journal.pbio.3001941