It seems natural to picture viruses as individual microscopic entities, but is there a more accurate way to think about them? In the first of this three-part series, Stephan Guttinger presents the case for a process view of viruses.
When hearing or reading about viruses in the news we are usually confronted with the picture of a particle: a thing, often coloured in red, that can infect us and thereby cause harm.
It is a powerful picture that provides us with a well-defined enemy. It also presents researchers with a well-defined target: the virus is often conceptualised as a “molecular machine” that can be isolated, disassembled, and analysed.1 The insights from this analysis can then be used to design molecules that should block the machine’s functioning.
In this three-part series of posts, I will discuss a range of findings from virology that point to fundamental flaws in the machine/thing view of viruses. A number of approaches in virology suggest that viruses should not be understood as things but rather as processes, a position John Dupré and I have discussed in more detail here. Such a shift in perspective has important implications for how we think about viruses and how we design sustainable strategies to deal with them.
In this first post, I want to look at how findings from virology support a process- rather than a thing-based view of viruses. In the second part, I will discuss why choosing a process perspective matters for scientific practice, using the example of how to deal with future viral pandemics. In the third part, I will turn to more technical philosophical issues and discuss how virology can inform debates about dispositions, intrinsic properties, and process ontology.
The thing-view of viruses: its power and underlying assumptions
Part of what makes the thing-view so attractive is its ability to provide scientists with a clear starting point for their investigations. There is a phenomenon (e.g., a viral infection) and a well-defined “thing” that is somehow responsible for it. The task is to figure out how the latter is connected to the former.
One reason the thing-view provides a “clear” starting point is hidden in the word “well-defined”: a critical assumption of the thing-view is that entities not only have distinct boundaries, but that they also have certain properties that belong to them and only them; there are intrinsic features that a thing possesses, independently of what is going on around it. This introduces a clear distinction between the thing and its environment.
Among other things, this distinction allows researchers to deal more easily with a messy world. Things in everyday life don’t behave in just one way. Mechanical systems such as cars might run better or worse depending on the ambient temperature or air pressure. Organisms can develop different shapes or features, depending on the conditions under which they develop. Viruses can be more or less virulent (i.e. able to cause disease), depending on the host organism they are infecting. All this change and plasticity can severely complicate the research process. But the thing-view can cut through this mess, as it can ground the observed diversity in the intrinsic features or the “essences” of things. These unchanging properties define what the thing can do or what it looks like in different circumstances.2
The assumption of intrinsic properties also has important methodological implications, as it justifies the analysis of viruses and their parts in isolation. Scientists can, for instance, remove a single viral protein from its biological context, analyse its core features, such as its structure, and use this information to design a drug that blocks its functioning in vivo. This is the targeted approach to drug development mentioned earlier.
This targeted approach might sound straightforward in theory. In practice it has a mixed record so far. Over the 50 years of its application (roughly) there have been some successes, such as treatments for herpes simplex (Aciclovir) or the combination therapies used to suppress HIV expression. But in most cases, the antiviral drugs that are available on the market seem to have limited effectiveness, even if they target well-known and common viruses such as influenza (a good example of this is Oseltamivir, also known as Tamiflu). On top of the limited effectiveness, viral strains can quickly develop drug-resistance, a problem that also affects some of the existing successful treatments.
Viruses: processes rather than things?
There are many explanations for why the targeted approach has had limited success in virology. Inefficient delivery and severe side-effects of drugs are factors that have to be mentioned here. But an important part of the problem, some researchers suggest, is also that viruses operate in a manner that is significantly different from what the thing-view would imply.
An example that can help illustrate these differences is the reproductive success of viruses. According to the thing-view, reproductive success is explained by the reproductive machinery of the virus (e.g., the viral enzymes used to replicate its genome) and by how well it is adapted to a specific context. A well-suited machinery will mean higher reproductive success, which means higher relative fitness. Over time, the fittest virus will come to dominate the population, as it outperforms its competitors. The thing-view emphasises the intrinsic features of the individual viral particles and how these features shape the virus’ behaviour and success in different contexts.
New insights into how viruses develop and multiply inside infected hosts, however, paint a different picture of viral fitness. Developments in DNA and RNA sequencing technology have allowed researchers to assess the in vivo diversity of viral populations in more detail. These and other studies have shown that for many viruses, including influenza, HIV, or hepatitis, the viral population within an organism represents a highly diverse and, importantly, dynamic system. Rather than forming a collective of identical particles, these viruses form what researchers now call a “mutant cloud” or “swarm”.3 It is such a cloud or swarm that develops and replicates within an infected organism.
The cloud nature of viruses matters functionally: through interactions between the cloud’s members, and between the cloud and its larger context, the virus obtains new behavioural features. The mutant swarm can, for instance, pool or recombine the genetic resources its diverse members contain and use this to quickly respond to changes in its environment. For instance, the heterogeneity of the population allows some viruses to evade the pressures exerted by antiviral drugs or cellular defence mechanisms.
The diversity and the dynamics of the mutant cloud can therefore lead to an increased (or in some cases decreased) fitness, compared to what individual “viral machines” could achieve. Understanding and measuring the reproductive success of a virus, therefore, depends on factoring in the relational and dynamic nature of the cloud, rather than just assessing the molecular “Bauplan” of the individual viral particles.
Interestingly, it is not just the workings of the cloud as a whole that rely on relations and context. The way in which the cloud is generated out of single particles is itself not an intrinsic feature of “the” virus particle. Virus-encoded factors certainly play a central role in this process. It is, for instance, well-known that RNA-based viruses, such as influenza or HIV, employ an error-prone process to replicate their genomes. This process is one of several factors involved in the generation of genetic diversity within a viral population. However, it has been shown that the low fidelity of genome replication is only to a small extent guided by the viral machinery and “its” features. The majority of the high mutation rates of viruses such as HIV can be attributed to the influence of cellular factors and processes. The virus’ diversity is therefore, in part at least, defined by the larger systems within which the cloud develops and moves. It is not some sort of disposition that could be ascribed to the virus and its intrinsic feature.
The shift in focus that these findings demand – from isolated particles with intrinsic properties to dynamic, de-localized systems – also leads researchers to new ways of thinking about antiviral treatments. Rather than only studying the atomic structures of viral particles, some scientists are now looking for ways of interfering with the dynamics of the mutant cloud. Twisting the system in ways that increases the mutation rate can, for instance, unbalance the cloud in such a way that can lead to the extinction of the virus. Such “lethal mutagenesis” might also be used by organisms as a natural defence mechanism against certain viruses.
Whilst such process-based interventions will encounter their own challenges, the focus on cloud dynamics is a strategy that seems to be more in line with how viruses behave in infected organisms, and that might eventually lead to the development of more effective and sustainable antiviral treatments.
The developments discussed above are only part of the changes in virology that suggest the need to move from a thing-view to a more process-based understanding of viruses. Other developments, such as the discovery of viral epitranscriptomics, suggest that even at the level of the single viral genome a more relational and dynamic view is needed to understand basic features such as viral genome stability or gene expression. I will return to some of these issues in the second post of this three-part series, where I will look at how a process-based view of viruses can re-shape the debate about how we should deal with the risk of future pandemics.
Part 2 →
Stephan Guttinger is a philosopher of biology based at LSE’s Centre for Philosophy of Natural and Social Science. Apart from looking at viruses and process ontology, he is also interested in how biologists create trustworthy knowledge, and how conceptual changes in the life sciences affect broader debates about science and health policy.
This is a great and really fascinating post! Most biological populations, changing over time, can have a “process-view”, as opposed to a “thing-view”. A viral population changes very rapidly, of course, but other than that, it seems like the “process-like” elements are very similar to what you would get in a population of any other kind. Your description of genetic diversity in a mutant swarm sounds similar to that in a species or population. And the cellular influences on viral genetics seem similar the the way to environments influence the biology of the organisms that have evolved and consequently develop in them. Do you think there are particular reasons the process-view is more appropriate for viruses, that don’t also apply to other species or biological populations?
Thank you so much for taking the time to comment. I think you raise a great point and I fully agree with you: what is being described here for viral populations also applies to other biological entities and populations. I personally think viruses are a particularly powerful example to make the case for a process ontology. But the view certainly applies more broadly.
There are in fact a number of authors who have applied process thinking to organisms and ecosystems. A recent collection of essays on the topic can be found in the book ‘Everything Flows’, edited by Dan Nicholson and John Dupré. The book was published by Oxford University Press in 2018 and is Open Access, meaning it can be downloaded for free from the OUP website. There are no specific chapters on viruses in that book, but a number of very interesting essays that deal with organisms (and process ontology more generally).
I have been a researcher in viral immunopathology for 40 years, studying herpesviruses, hepatitis B and HIV amongst others. I have always been frustrated with how widely-accepted research dogmas associated with various aspects of viruses don’t fit with real-world observations. For instance, we know about all the elaborate mechanisms that herpesviruses like CMV, EBV and HSV have for evading the immune system, and yet a functional immune system is so effective at controlling these micro-organisms. Your article is the most profound thing I’ve read since I started in this field. Thanks.
Thank you so much for your kind words. I am glad you enjoyed reading the post. The point you make about the disconnect between established research dogmas and real-world behaviour/observations is actually something I will come back to again in the second post of this series. Also, the viruses that you are studying, and in particular herpesviruses, represent another fascinating part of the viral realm that deserves much more attention in philosophy of biology (I think).
Stephan Guttinger this is an amazing work from the both perspective you are using, biological and philosophical. Thanks for this
absurd distinction, all “things” are processes the relevant point for virology and clinical treatments is that we can isolate viruses from their environs in ways which allow us to assess them in terms of their relevant internal characteristics.
I fully agree – all ‘things’ are processes. And the laboratory is an important space in which some aspects of viruses can and should be assessed.
But I also think that if we fully buy into a process view of viruses and other ‘things’ then this has important implications for how we think about ‘internal characteristics’ of viruses and for what can be done with an isolate-and-assess approach. I aim to reflect more on this in the second post of this series and am interested in hearing your thoughts on it!
This is how all living populations are. ‘Mutant clouds’ is the state of being for the an entire ‘species’ of organism.
This is pretty much the preview of all evolution, Even within a species much of the genes within are a spectrum of mutants that loosely resemble one another.
Even within single genes of those mutants there are some domains that stay the same and others that change more readily. I’d *hope* that virologists already take this approach. But then I think vocabulary falls flat in the face of the enormous amount of information goes into different gene combinations. It’s definitely an interesting topic.