Mpox Evolution: Has the Current Outbreak Revealed a Pox on “U”?

Abstract

The current mpox outbreak, caused by the monkeypox virus (MPXV), was declared a Public Health Emergency of International Concern in July 2022. There have now been more than 80 000 cases in over 100 countries. Historically, the virus caused relatively rare zoonotic infections, with episodic and limited endemic infections in Central (clade I) and West (clade IIa) Africa. The current global outbreak is due to an unprecedented person-to-person spread of the clade IIb virus (Figure 1A). Recently, genomic studies on the outbreak strain have shown unexpected footprints of mutations accumulating on the viral genome. This work has brought to light an intriguing possible connection between MPXV and APOBEC3 enzymes, mutators that change cytosine (C) bases to uracil (U) in DNA as an antiviral strategy [1, 2].

Figure 1.

Mpox and APOBEC3s. A, A schematic version of a stylized monkeypox virus (MPXV) phylogenetic tree is shown. The virus, which typically replicates with high fidelity, has shown APOBEC3-related mutagenesis along the clade IIb lineage (purple). Phylogenetic distance is not shown to scale and the sequenced virus isolates highlighted under clade IIb only represent a timeline. B, The introduction of C-to-U mutations by A3 enzymes can result viral restriction or viral evolution, defined by characteristic mutation patterns.

Open in new tabDownload slide

MPXV is a poxvirus in the same genera as variola, the causative agent of smallpox. Poxviruses are double-stranded DNA (dsDNA) viruses, with distinctive replication cycles [3]. Upon infection, poxviruses establish cytoplasmic replication “factories” where their large (approximately 200 kB) genomes are amplified in concatemers that are resolved into unit genomes to yield the next infectious virions. As with most DNA viruses, replication occurs with high fidelity, resulting in slower rates of evolution than those observed with small RNA-based viruses such as HIV (approximately 10 kB). The large MPXV genome encodes for the enzymatic machinery responsible for cytoplasmic replication, along with factors modulating host responses, including interferon-associated responses.

Having long been a focus of HIV researchers, APOBEC3 enzymes (A3s) are single-stranded DNA (ssDNA) deaminases targeting intermediates in the retroviral lifecycle [4]. As U does not belong in DNA, C-to-U mutations induce mechanisms that can often degrade viral genomes or garble their coding sequences. In viruses targeted by A3s, characteristic mutational footprints can be readily ascertained by two features. First, mutations occur preferentially in a 5′-TC sequence context on either strand. Second, given that ssDNA is often exposed in patches during replication, mutations typically cluster close together. When reading dsDNA, these features combine to result in signatures of clustered TC-to-TT or GA-to-AA changes (Figure 1B).

The importance of A3s is highlighted by the rapidity of their evolution in mammalian genomes. While the mouse genome harbors only a single A3, the family has undergone rapid expansion with seven human A3s (A3A to A3H), hypothesized to specialize for specific viruses or escape viral countermeasures, such as the HIV anti-A3 protein Vif. These host and viral factors are in constant flux. For example, human A3G restricts HIV but not SIV, while rhesus A3G restricts SIV but not HIV. A single mutation in either host factor can expand restriction to the other virus. These observations support the notion that A3s may be important in restriction of zoonotic viruses. Notably, although retroviruses have been studied in greatest depth, there is also strong precedent for A3s acting on DNA viruses, some of which harbor anti-A3 factors akin to HIV Vif.

While A3s offer powerful defense, they can also be viewed as a “double-edged sword,” given the risks posed by mutation as an antiviral strategy. Viruses that remain replication competent after mutation can occasionally escape immune restriction, acquire drug resistance, or even potentially leap between species [4]. A3s can also pose a risk to our genomes, as their mutational footprints have been commonly detected in various cancers [5].

Against this backdrop comes the surprising finding of mutations in the transmitted MPXV virus consistent with A3 activity. When compared with close relatives, the outbreak strain has >50 mutations, with C-to-U changes in a 5′-TC context on one or the other strand of the dsDNA genome accounting for approximately 90% of these changes [1]. Although not highlighted in initial reports, closer examination suggests mutational clustering in a manner consistent with A3 targeting of ssDNA intermediates in the viral lifecycle. Strikingly, clade I and clade IIa do not show the same pattern of mutations [2]. Thus, these data support an unexpected and specific relationship between the MPXV outbreak strain and A3 mutators.

While offering a useful signal for tracking viral spread, these mutations also raise an array of questions regarding the associated biology. Clustering and sequence context offers strong support for the hypothesis that A3s are acting on this clade IIb MPXV. Notably, although clade IIb sequences are available from as early as 1971, the pattern of A3 mutations appear to arise in strains that were sequenced around 2017, from which point forward they have continued to accumulate at what appears to be a highly accelerated pace (Figure 1A). The reason why clade I and IIa virus, along with older clade IIb viruses, do not readily show these patterns remains a mystery. On deeper reflection, there are hints of a possible longer-standing interaction. Poxvirus genomes are highly A-T rich, potentially reflecting prior A3 activity. The viral genome also includes its own DNA repair enzyme that can remove uracil from DNA. If the outbreak lineage, but not historical strains, are being targeted by A3s, what may have changed? Extended human-to-human transmission is likely relevant. Specifically, the mode of transmission through sexual contact may now be commonly exposing the virus to cell types that may express a different repertoire of A3s. Furthermore, poxviruses are known inducers of interferon responses and harbor factors that can modulate these responses. As some human A3s are interferon inducible, it is tempting to speculate that the viral-induced host response may have been modulated in a manner that now elicits A3 activity on viral genomes.

Are there functional implications to the A3-associated mutations? From the host side, it remains unclear whether these A3 signatures reflect true antiviral activity. While one prior study suggested that A3s may not restrict poxviruses, only a subset of A3s were examined [6]. It is plausible that one of the other human A3s may have antiviral effects or that a viral factor could be counteracting A3s. Similarly, while the reservoir for MPXV is not fully established, it is also possible that A3s from other mammalian hosts may have contributed to the initial suite of mutations observed in the outbreak strain. Moving to the viral perspective, whether the mutations reflect adaptive evolution or whether they could simply be “bystanders” is a critical question, ripe for study.

Taken together, although the current MPXV outbreak appears to be waning, we are also clearly just at the start when it comes to understanding an unexpected new dimension of mutation and adaptation at the host-pathogen interface.

Financial support. No financial support was received for this work.

References

Author notes