Viral Populations, Evolution and Transmission

What survives when a viral population moves between hosts, tissues or species?

In one paragraph

Seven publications sit in this area, and they do not form a single program. What holds most of them together is a technique rather than a question. Varble 2014 built a library of more than one hundred influenza A viruses distinguished only by a neutral barcode carried in an engineered intergenic region of a split NS segment, which converts the composition of a viral population into a quantity that can be read by sequencing at any point in an infection. That platform was then used to measure how severely transmission narrows a population, to build a fitness landscape for a single viral gene across several hosts, and, in a collaborative study led by another laboratory, to follow an enteric virus through the tissues of a mouse. A second and separate line asks what genomic capacity allows a virus to escape a defined selective pressure, and answers it with homologous recombination in Aguado 2018 and with host mediated RNA editing in Uhl 2023. One publication, Guzmán-Solís 2021 on ancient viral genomes from Colonial Mexico City, connects to this corpus through personnel rather than through scientific lineage, and the record for that paper says exactly that.

The defining scientific question

The area question is what survives when a viral population moves. That question decomposes in two directions that the corpus treats separately.

The first is quantitative. An RNA virus carries standing variation because its polymerase lacks proofreading, and that variation is only usable if enough of it crosses each transition between hosts and tissues. Measuring how much crosses requires a way to read population composition without altering it, and the methodological centre of this area is that measurement problem rather than any one biological conclusion.

The second is about capacity rather than quantity. Given a pressure that no virus in a comparison has evolved to antagonise, which feature of a replication strategy decides whether escape is available at all? Aguado 2018 poses this explicitly for small RNA defence.

These two directions share a logic. In both, a defined and uniform pressure is applied to a population whose members differ in a controlled way, and the surviving composition is read by sequencing. They do not share a biological subject.

Origins

The platform originates in engineering that was not done for population work. Varble 2010 showed that segment 8 of influenza A virus could be reconfigured so that the overlapping NS1 and NEP reading frames occupy separate positions with a noncoding intergenic region between them, and that foreign sequence placed there is carried without measurable loss of replication or genome stability. Varble 2013 used barcode libraries in that segment as a neutral drift control in an in vivo screen, where the barcodes existed to show what happens to a population under no selection.

Varble 2014 inverted that role. If barcodes confer no advantage, then any change in their distribution is a direct report on population sampling, and the drift control becomes the measurement. That paper tested the neutrality assumption against wild type growth curves, against individually amplified clones and against duplicate sequencing of the same population before drawing conclusions from it.

The escape line has a separate origin in the laboratory's small RNA engineering. Benitez 2015 inserted perfectly complementary microRNA target sites into influenza A virus to reconstruct a slicing competent small RNA defence in mammalian cells, and it reported an escape asymmetry that framed everything after it. Across several designs, including one carrying a single target site, escape arose only by destroying guide production and never by altering the target sequence.

Major findings

Transmission narrows an influenza population severely, and the narrowing differs by route and recipient. Varble 2014 found no detectable bottleneck in canine or human cell monolayers, five to thirteen clones surviving amplification in embryonated chicken eggs, two to five clones in guinea pig recipients infected across cages, seven to twenty four clones in ferret contact recipients against seventy one to one hundred in donors, and as few as two clones after airborne transmission.

Two distinct kinds of narrowing were separated in the same study. Egg passage was accompanied by convergent hemagglutinin change at residues associated with avian receptor specificity, which the authors read as genetically driven selection during host adaptation. Mammalian transmission showed no such convergence, and three guinea pigs cocaged with one donor carried different barcode sets despite identical exposure, which places the restriction at the recipient and argues against viral genetics as its cause.

Phylogenetic position is a poor predictor of phenotype for the NS1 gene. Muñoz-Moreno 2019 competed 107 barcoded viruses carrying 56 natural NS1 sequences in a common backbone across dog cells, human cells, eggs and mice, and network analysis identified clusters sharing fitness profiles despite substantial amino acid divergence. Allele B NS1 viruses were overrepresented in every substrate, and Stat1 deficiency flattened much of the spread while Rag1 deficiency did not, placing the selective filter in early innate signalling.

Loss of population diversity in an enteric infection happens after replication rather than at the barrier. McCune 2020, led by the Pfeiffer laboratory, recovered all 135 barcodes of a coxsackievirus B3 library from the upper gastrointestinal tract at 7.5 and 19 hours, found three or fewer barcodes dominating every tissue by 48 to 72 hours, and used neutral red labelling to show that diverse members had genuinely replicated before diversity collapsed.

Escape from a uniform small RNA pressure tracks with recombination capacity. Aguado 2018 applied the same cassette to six viruses across four families and found negative sense viruses cleared by more than five logs while positive sense viruses recovered by precise excision. A poliovirus carrying the D79H polymerase substitution, which blocks homologous recombination, grew normally without the pressure but could not excise the cassette and was undetectable by passage four. That single substitution is the causal evidence that template switching rather than polarity is the requirement.

A virus that cannot recombine can still escape, but the change is not its own. Uhl 2023 held infections without passage and found a five target Sendai virus escaping at six to eight days in roughly eight percent of wells, through dense A to G editing confined to the target sites. ADAR1 knockout abolished escape in all 96 wells and adenoviral reconstitution restored it, which makes a host enzyme rather than a viral adaptation the relevant actor.

African associated viral lineages were present in early Colonial Mexico City. Guzmán-Solís 2021 recovered three parvovirus B19 genomes of genotype 3 and one hepatitis B virus genome of sub-genotype A4 from dental remains, in individuals whose nuclear ancestry, mitochondrial haplogroup and enamel strontium ratios indicate West African birth, with radiocarbon dates in the period of heaviest forced transport.

How the work evolved

The barcode platform moved outward rather than deeper. Between 2014 and 2020 it was applied to a new gene in a study led by the García-Sastre laboratory and to a new virus, route and barrier in a study led by the Pfeiffer laboratory, and in both cases the tenOever contribution is the method and its supervision rather than the discovery. Within the laboratory itself the transmission bottleneck work was not returned to. Varble 2014 identified the recipient as the site of restriction and left the barrier that performs the sampling unidentified, and no later publication in this corpus takes that up. This is a case where a line of work was taken up by others and not continued here, and the honest account says so.

The escape line did progress inside the laboratory, across three papers in eight years. Benitez 2015 established that the pressure works and that escape happens on the guide side. Aguado 2018 made the pressure comparative and identified recombination as the determinant. Uhl 2023 returned to the case Aguado 2018 had left as a dead end, on the argument that negative sense RNA viruses do exist in hosts with functional antiviral RNA interference, and found a host mediated route out by changing the passage regime. That is a genuine sequence in which each paper is set up by the last.

The paleovirology paper does not participate in either development. It has no predecessor and no successor in this corpus.

Principal publications

Varble 2014 is the methodological centre of the area and the source of the barcoded influenza library. Aguado 2018 converts a comparative observation into a causal one with a single polymerase substitution. Uhl 2023 relocates escape from the virus to the host. Muñoz-Moreno 2019 and McCune 2020 are the two extensions of the platform by other laboratories. Benitez 2015 supplies the pressure the escape line depends on and is otherwise primarily a paper about reconstructing small RNA defence. Guzmán-Solís 2021 is a collaborative ancient DNA study, valuable in itself and unconnected to the rest.

Connections to other areas

Three of these publications are shared with the Small RNA Biology and the Limits of Antiviral Silencing area, where Benitez 2015, Aguado 2018 and Uhl 2023 are filed under reconstructing antiviral RNA interference. In that context the same experiments serve an argument about why vertebrates use interferon rather than small RNA defence, which is a different question from the one asked here. Muñoz-Moreno 2019 is shared with the Influenza Genome Regulation and Replication area, where NS1 and interferon antagonism rather than fitness mapping are the subject. The split NS segment that makes the barcode platform possible belongs to the laboratory's influenza engineering line, running through Varble 2010, and barcode libraries as a drift control belong to the in vivo screening line running through Varble 2013 and Benitez 2015 on RNAi screening.

Current implications

The bottleneck measurements bear on pandemic risk assessment. If airborne transmission can be founded by as few as two genomes, then a phenotype requiring several mutations and present at low frequency in a donor is unlikely to be carried into a new host intact, and Varble 2014 connects this to why an airborne transmissible H5N1 has not emerged despite repeated human infections. Muñoz-Moreno 2019 presents its landscape approach as a way to flag strains whose NS1 supports broad host tropism, which is a screening rather than a predictive claim.

The escape results bear on design. Any attempt to attenuate a virus or to build biocontainment on microRNA targeting inherits the failure modes these papers mapped, and they differ by virus class. A positive sense virus will excise the target region, a self targeting design will lose its guide, and a negative sense virus may be rescued by host editing that the designer does not control.

Open questions

The barrier in the recipient that performs the transmission sampling has not been identified, and Varble 2014 leaves open whether an additional selection step occurs at secretion. How viral fitness is maintained across repeated severe bottlenecks is unresolved, with purifying selection and host heterogeneity offered as alternatives. For McCune 2020 the anatomical source of the monopolizing population, the reason a subpopulation takes over an entire animal, the route of dissemination within twenty minutes and the basis of its dependence on interferon receptor status are all stated by the authors as unresolved. Muñoz-Moreno 2019 establishes no molecular mechanism for any individual fitness difference and offers its CPSF30 and TRIM25 proposals as an untested framework. Uhl 2023 does not resolve how ADAR1 comes to edit those particular sequences, notes that the rarity of escape puts biochemistry out of reach, and reports that an equivalently targeted influenza A virus showed neither editing nor escape, which is unexplained. Across the area, every measurement rests on engineered viruses carrying inserted sequence in cultured cells and animal models, so generalisation to unmodified viruses in natural infection remains an extrapolation.

Publications referenced

• Influenza A Virus Transmission Bottlenecks Are Defined by Infection Route and Recipient Host
• Engineered Mammalian RNAi Can Elicit Antiviral Protection that Negates the Requirement for the Interferon Response
• Homologous recombination is an intrinsic defense against antiviral RNA interference
• Viral Fitness Landscapes in Diverse Host Species Reveal Multiple Evolutionary Lines for the NS1 Gene of Influenza A Viruses
• Rapid Dissemination and Monopolization of Viral Populations in Mice Revealed Using a Panel of Barcoded Viruses
• Ancient viral genomes reveal introduction of human pathogenic viruses into Mexico during the transatlantic slave trade
• ADAR1 Biology Can Hinder Effective Antiviral RNA Interference
• Engineered RNA viral synthesis of microRNAs
• An In Vivo RNAi Screening Approach to Identify Host Determinants of Virus Replication
• In Vivo RNAi Screening Identifies MDA5 as a Significant Contributor to the Cellular Defense against Influenza A Virus

Explore the themes

Connected discoveries

Publications in this area