Single-particle measurements reveal damage to filamentous influenza virions during laboratory handling

Most laboratory strains of influenza virus produce near-spherical virions, but clinical isolates also produce extended filaments whose biophysical properties are understudied. Most functional studies of filamentous influenza viruses do not include data on the concentration or lengths of the virions, making it hard to interpret their sometimes contradictory results. Furthermore, anecdotal reports suggest that filaments are damaged during routine laboratory handling. Therefore, to understand filament function we require a tool to assess the number and dimensions of filaments in a sample and an assessment of how filaments respond to standard handling procedures. We initially sought to analyse filament populations using negative stain particle counting, but found that this was low-throughput and could not detect particles longer than 10 µm. Instead, we used confocal microscopy with semi-automated image analysis. This allowed a high-throughput, quantitative analysis of length distributions in filament populations. Using this, we assessed the effects of pipetting, vortexing, sonicating, clarification and freezing on filaments. Most procedures did not appreciably alter filament dimensions. Pipetting and vortexing both slightly reduced filament numbers, but their effects were only appreciable after extended treatment. In contrast, freezing substantially reduced the number and median length of filaments, as well as creating ‘kinks’ in filaments which suggest damage to the capsid. We conclude that confocal microscopy can provide the basic measurements needed to interpret functional studies of filamentous strains. Using this approach, we found that freezing filaments causes previously unappreciated damage, which should be considered when planning further research.