During my PhD I spent many an hour in a dark room, staring down a microscope, manually counting cells infected with hepatitis C virus (see our earlier post). As I robotically clicked away at the tally counter my thoughts would often stray; after the inevitable plans for what I would eat for dinner, I would start thinking about the microscopic events that had led to the pattern of infection I could see through the eyepiece.
Each group of infected cells that I counted had arisen from a single virus particle infecting a single cell. But I was curious: how many viruses took part in the experiment, how many attached to a cell, how many succeeded in entering a cell? It occurred to me that the whole process was likely to be probabilisitic; there are multiple steps a virus particle has to take towards productive infection and it was possible that the majority of virus particles fail at some stage and fall by the wayside. It also occurred to me that to investigate this would require some maths, alas, I am no Carl Friedrich Gauss. My thoughts went no further.
Seven years pass, I am at virology meeting having presented some of our work on virus entry. Over breakfast Chris Illingworth enquires "have you ever considered mathematical modelling of virus entry". By comparison to me, Chris is Carl Friedrich Gauss; I bit his hand off. We collaborated over the course of 18 months to design experiments and build a mathematical framework to understand the early stages of virus entry by hepatitis C virus. The products of this work have just been uploaded to bioRxiv (and submitted for publication, watch this space...).
Working with Chris has proven to be very rewarding. We were able to start with data generated by basic virology assays, build a mechanistic framework to explain that data and then use mathematical modelling to test that framework and make new predictions. I won't provide great detail in this post (you can read it for yourself), however, I can start to answer some of the questions posed by my younger self: HCV entry is extremely inefficient, of the virus particles that manage to attach to the cell surface only ~25% make it past the early stages of virus entry, of these, only ~5% make it to productive infection. This discovery may explain the, so-called, 'bottleneck' in HCV transmission between people; sequencing data indicates that only a 1-4 particles are responsible for establishing new infections.
On another matter, the molecular events of HCV entry are targeted by potent neutralising antibodies. We are hoping that our mechanistic model will help us to better understand how HCV evades these antibodies, and how we may be able to design a vaccine to prevent virus entry.
We recently contributed to a Wellcome Open Research article describing a new panel of monoclonal antibodies against the tetraspanin CD81. This was in collaboration with my PhD supervisor Prof. Jane McKeating (who first got me hooked on virus entry).
I think CD81 is cool. It is a small transmembrane protein found on the cell surface of lots of different cell types. Like most tetraspanins, it acts as a molecular scaffold to organise events at cellular membranes. CD81 is multifunctional, for example: it is necessary for proper B-cell receptor signalling (i.e. no CD81 = messy immune response); it is important for sperm-egg fusion (i.e. making babies); CD81 is also critical for the entry of both hepatitis C virus and malaria sporozoites. The list goes on.
My admiration of this protein only grew when the crystal structure was published recently. I don't have time to go in to detail, but for a CD81 nerd it was pretty wild stuff!
Jane's team started work on developing a range of anti-CD81 antibodies whilst I was still working in her lab. A lot of people worked to characterise the antibodies, in particular the lab manager Ke Hu. The paper summarises our major findings and, effectively, advertises these great antibodies as tools for basic research of tetraspanin biology.
In the Grove Lab we have a few of projects in which CD81 is centre stage. These antibodies are proving invaluable!
Viruses are masters of evolution, allowing them to outstrip immune responses and rapidly adapt to new hosts. This is one of the secrets of their success. However, this genetic plasticity may, in some cases, pose a problem.
Some viruses, such as HIV and HCV, establish lifelong infections and many years may pass before they have the opportunity to transmit to a new host. In this situation, can a virus become too adapted to the specific environment of their current host?
We discuss the various scenarios in which viruses may become 'short sighted'; being so well adapted to one host that they become a 'rusty' at transmitting to new hosts. We also propose the possible mechanisms viruses may have to avoid this short-sightedness.
I really enjoyed collaborating with Katrina to put this article together. It proved to be quite a mental workout (evolutionary biology is not in my comfort zone), but the writing process has given me a different perspective on persistent viral infections and how the host environment may shape their evolution.
This work was supported by the Wellcome Trust, Royal Society, The Natural Environment Research Council and The European Research Council
Image courtesy of Frances Cacino.
...this was fine (once I'd gotten over the unexpected nausea), it got me data and I enjoyed the moments when I peered down at my experiment and realised maybe, just maybe, I'd discovered something. However, manual counting of cells is slow, tedious and prone to error; so when I set up the Grove Lab I decided to find another way.
To be clear, the cells I'm talking about here are those infected with hepatitis C virus and we identify them by immunofluoresent labelling. By quantifying the number of infected cells we can measure viral replication; this is a cornerstone technique in many of our experiments.
During my postdoc at the LMCB I got a comprehensive training in microscopy and image analysis, and when I returned to the HCV field I put these new skills to use by designing Infection Counter, an ImageJ plugin for automated quantification of infected cells. This approach uses images of cells that have been labelled for viral antigen and cellular nuclei using DAPI; the plugin first approximates the location of each cell using the nuclei then scores them as positive or negative for viral antigen. It's a simple analysis, but very robust. By exploiting the plate-reading capabilities of our microscope we can easily image and quantify hundreds of samples every week with relatively little effort. A significant improvement on me, feeling sick, clicking away on a tally counter for hours upon end.
We recently published a description of Infection Counter in Viruses, where we give a detailed account of how it works and provide data to validate the technique. The plugin is free to download so we encourage y'all to give it go on your virus of choice.
This work was done in collaboration with the Henriques Lab at the LMCB and the Towers Lab in Infection and Immunity. It was supported by the Wellcome Trust, Royal Society and Medical Research Council.