Alchemist's Laboratory Showing Original Apparatus of the 16th and 17th Century. London. Courtesy Wellcome Trust.

Read the Reflection, written 3 June 2021, below the following original Transmission.

Numerous questions emerge when considering the nature and relevance of viruses: What exactly are they? Are they even alive? How and when did viruses first evolve, and are viruses unavoidable consequences of biological replicators? Why are they so diverse in genomic architecture, yet so limited in lifestyles? Why are there so many emerging viruses, and what are the ecological and genetic drivers of that emergence? Why do they become pathogenic, and what is the nature of the complex molecular interaction they establish with their hosts? 

Viruses are complex systems spanning orders of magnitude in size. The study of their behavior and structure, particularly using multidisciplinary frameworks, has revealed a number of universal patterns of organization. RNA viruses display high mutation rates that push them to the edge of disorder, where high instability, but also adaptability, occurs. As with many phenomena in virology, this edge is well described by phase transitions — where small changes in mutation rates can lead to large changes in structure — analogous to a liquid transitioning to a gas with a small shift in temperature. 

Viruses have influenced evolution at all levels of biological organization, from cells and organisms to populations, ecosystems, and even the entire biosphere. Their dynamics involve nonlinear phenomena, tipping points, and self-organization. Viruses offer unique experimental and theoretical windows into the origin and evolution of complex systems. 

For 30 years, I have combined experimental evolution, molecular genetics, systems biology, molecular epidemiology, mathematical modeling, and computer simulations to pursue the enigma of the virus, using different animal and plant systems. In this search for basic knowledge, we have acquired many small pieces of information that, together, delineate a picture of how viruses evolve. This picture offers insight for our current pandemic.

We now know that the mutations required for an emerging virus to adapt to a novel host come with a cost, and we know that this cost is relaxed if the virus circulates among different host types. We have also learned that high mutation rates allow RNA viruses to achieve the highest possible fitness and replication rates. And, we have identified the molecular mechanisms by which some viruses, such as influenza A, infect a wide range of hosts, whereas others, like mumps, are highly specialized to particular hosts. To generalist viruses, hosts are more or less the same on the cellular level. The range of hosts for these viruses all have common, highly conserved, elements in their genetic regulatory networks. Specialist viruses, on the other hand, interact with unique, or non-conserved, elements within their host’s cells. 

The emergence and pandemic spread of SARS-CoV-2 has raised many questions, hitherto of interest to fundamental, theoretical science, to the level of immediate practice. The pressure to give rapid support to authorities in our public health systems and the urgent need to find new specific antiviral treatments has moved us to rapidly repurpose the lab and to focus our research toward more urgent needs. 

For example, our know-how in virus detection and purification can now be applied to diagnostics. By moving our quantitative PCR machines from the lab bench into biohazard secure rooms and adopting protective measures, our research protocols transform into medical diagnostic procedures. 

Our past experimental and mathematical analyses of the interactions between viruses and so-called defective interfering particles (DIPs) — basically, the parts of a virus that are incapable of self-replication — are inspiring a research project aimed at finding new SARS-CoV-2 antivirals. Because DIPs interfere with the replication, accumulation, and transmission of the wild-type virus, we can think of them as antivirals. DIPs offer several benefits over conventional antiviral approaches: They are transmissible in the presence of the full virus and hence can spread with it to target infected cells; DIP-mediated protection is effective immediately, unlike classic vaccines that work via priming the adaptive innate immune pathways; and DIPs cannot replicate and are transcriptionally defective in the absence of the full virus, thus limiting possible side effects. 

DIPs are just one example of a potential new antiviral inspired by the analysis of viruses as complex replicative systems. My colleagues and I are sure that many other biomedical applications will emerge from this multidisciplinary perspective. Basic science is not just future applied science; it is our applied reserve, available to be repurposed when society calls.

Santiago Elena
Instituto de Biología Molecular y Celular de Plantas
Santa Fe Institute

T-011 (Elena) PDF

Read more posts in the Transmission series, dedicated to sharing SFI insights on the coronavirus pandemic.

Listen to SFI President David Krakauer discuss this Transmission on episode 29 of our Complexity Podcast.



June 3, 2021


In my Transmission, “A complex systems perspective of viruses offers insight for controlling SARS-CoV-2 and future emerging viruses,” I described how many years of experience in basic and theoretical virology research placed us in the situation to quickly move into applied SARS-CoV-2 research. The diverse range of problems included the biological effect of mutations circulating among infected people and the identification of defective interfering particles (DIPs) that can be selected and further engineered as potential antiviral agents.

From a purely scientific perspective, this last year has been an exciting time of discoveries for virologists and evolutionary biologists. The amount of data accumulated about SARS-CoV-2 origins, biology, and epidemiology has been growing daily with no precedent in the history of biomedical sciences. In fact, the process of developing several working vaccines in a record time, moving them from lab assays to clinical trials and, finally, beginning to vaccinate the population (of rich countries) on a very wide scale must be qualified as a herculean task. Nonetheless, many challenges still remain. Some are undoubtedly associated with the virus biology, but others are popping up as a consequence of the existing economic, sociological, and educational inequalities across countries. Both are tightly linked.

As an RNA virus, SARS-CoV-2 has an error-prone replication machinery that results in the generation of a large number of genetic variants within infected individuals. Not surprisingly, we have been witnessing the sweeping of variants of concern (VOC in the World Health Organization jargon) into the virus population worldwide. These VOCs so far all affect the spike (S) protein. S interacts with the ACE2 cellular receptor protein and facilitates entrance into the target cells. By now, many readers will have heard of the UK (lineage B.1.1.7 or Alpha), South African (lineage B.1.351 or Beta), Brazil (lineage P.1 or Gamma), and even the most recently described Indian (lineage B.1.617.2 or Delta) and Californian (lineage B.1.427 or Epsilon) variants. All these variants show increased transmissibility, induce stronger symptoms, and may even jeopardize the effectiveness of vaccines. Luckily, most of these variants share a limited repertoire of mutations (e.g., L452R, E484K and E484Q, N501Y, D614G, P681R), suggesting that there are only a few evolutionary paths that the virus can take to adapt from its likely bat reservoir into the new human host. And here is where a first challenge appears. Once the virus is well adapted to human hosts, and as the frequency of immunized hosts increases until we reach herd immunity levels, a new selective force will enter into play: immune pressure. Host immunity will favor the virus escape variants that would probably fix a different panoply of mutations, not necessarily in the S protein but in other proteins that form the capsid.

Unfortunately, vaccination campaigns are not progressing at a similar pace in all countries, with 75% of all vaccines dispensed in only ten countries. This heterogeneity creates new challenges. First, and most important, it creates large reservoirs of susceptible hosts in which the virus continues evolving and adapting and, eventually, generating standing genetic variation that might contain future VOCs. Second, the effectiveness of current vaccines would diminish as new VOCs escaping from vaccine-elicited antibodies sweep into the viral population across the world. Third, obviously vaccine failures would impose the necessity of periodically reformulating the vaccines and re-vaccinating people. If we, as a society, really want to get back into our “normal” luxury lives, we must push our governments to enter into vaccination programs for the less wealthy countries and set up effective surveillance systems of reservoirs and intermediate animal hosts that may allow for predicting future waves of the virus. We will likely not eradicate SARS-CoV-2, and this is far from being our goal, but we must learn to live with it as we do now (more or less) with influenza A virus. We should learn a number of lessons from this pandemic and prepare ourselves for the next one, which will for sure occur.

Another challenge is to develop new and more efficient antiviral treatments. In my Transmission, I mentioned the possible usage of DIPs as an anti-SARS-CoV-2 evolving and adaptable therapy. In 2021 we have actively explored this possibility. We now have compelling evidence that large amounts of defective viral genomes (DVG) are found in multiple patients, with some forms being preferentially accumulated during the infection process, and pervasively found in different individuals. This suggests that DVGs may play some yet unknown role in virus replication and accumulation, as well as in its interaction with the host’s immune response. We are now synthesizing in vitro some of the most commonly found DVG variants and testing their interfering role (as DIPs). In addition, we are also developing machine learning algorithms to associate the presence and abundance of these DVGs with clinical symptoms of COVID-19, generation of neutralizing antibodies, and the progression and duration of infection.

Finally, in my previous Transmission I also advocated for a multidisciplinary approach to tackling the COVID-19 pandemic. This year has shown that this is, indeed, an absolute necessity. As mentioned above, astonishing amounts of different types of data are being produced that are not well integrated into a common predictive framework. Multilayer network approaches would provide the theoretical framework and mathematical tools to integrate information from the most elemental level of virus biology (i.e., the frequency and nature of genetic variants within infected individuals), the distributions of variants among individuals from the same population, the movements of individuals among populations and even between countries (despite quarantine measures), up to transport networks at continental levels. Objects in each layer should incorporate information relevant for this particular layer (e.g., clinical symptoms or even social networking (mis)information in the layer connecting individuals). Together, this multilayer integrative approach would provide a fresh look at all of the accumulated data and help us to better understand the ongoing processes and perhaps to predict future behaviors.

Read more thoughts on the COVID-19 pandemic from complex-systems researchers in The Complex Alternative, published by SFI Press.