Retrotransposon-Mediated Host Gene Capture, Complementation
Total Page:16
File Type:pdf, Size:1020Kb
1 Cascade evolution in poxviruses: retrotransposon-mediated host gene capture, 2 complementation and recombination 3 4 M. Julhasur Rahman1, Sherry L. Haller2, Jie Li3, Greg Brennan1 and Stefan Rothenburg1* 5 6 1 School of Medicine, University of California Davis, Department of Medial Microbiology and 7 Immunology, Davis, CA 95616, USA 8 2 University of Texas Medical Branch at Galveston, Department of Microbiology and 9 Immunology, Galveston, TX 77555, USA 10 3 Genome Center, University of California Davis, Davis, CA 95616, USA 11 *correspondence: [email protected]. ORCID iD:0000-0002-2525-8230 12 13 14 Keywords: horizontal gene transfer; LINE-1; retrotransposons; poxvirus; evolution, PKR; E3L 15 1 16 Abstract 17 18 There is ample phylogenetic evidence that many critical virus functions, like immune evasion, 19 evolved by the acquisition of genes from their hosts by horizontal gene transfer (HGT). However, 20 the lack of an experimental system has prevented a mechanistic understanding of this process. We 21 developed a model to elucidate the mechanisms of HGT into poxviruses. All identified gene 22 capture events showed signatures of LINE-1-mediated retrotransposition. Integrations occurred 23 across the genome, in some cases knocking out essential viral genes. These essential gene 24 knockouts were rescued through a process of complementation by the parent virus followed by 25 non-homologous recombination to generate a single competent virus. This work links multiple 26 evolutionary mechanisms into one adaptive cascade and identifies host retrotransposons as major 27 drivers for virus evolution. 28 29 30 2 31 Introduction 32 33 Horizontal gene transfer (HGT) is the transmission of genetic material between different 34 organisms. This process shapes the evolution and functional diversity of all major life forms, 35 including viruses (Keeling, 2009, Gilbert and Cordaux, 2017). The genomes of large DNA viruses, 36 including poxviruses and herpesviruses, contain numerous genes that likely have been acquired 37 from their hosts by HGT. Many of these captured genes are involved in immune evasion or 38 protection from environmental damage (Hughes and Friedman, 2005, Bratke and McLysaght, 39 2008, Bratke et al., 2013, Wilson and Brooks, 2011, Iyer et al., 2006). For example, viral homologs 40 of host interleukin 10 (IL-10) genes are among the best supported examples of these HGT events. 41 Viral IL-10 genes, which presumably provide a selective advantage through inhibition and 42 modulation of the antiviral response, appear to have been independently acquired by several 43 herpesviruses and poxviruses (Hughes and Friedman, 2005, Schonrich et al., 2017). 44 So far, the detection of HGT in large DNA viruses has relied on bioinformatic approaches, 45 primarily phylogenetic and sequence composition analyses. While these methods can detect 46 putative HGT they can do little to elucidate the frequency or mechanisms of HGT. The two most 47 likely mechanisms for HGT into DNA viruses are either a DNA-mediated process such as 48 integration via recombination or the help of DNA transposons, or an RNA-mediated integration, 49 such as gene transfer via retrotransposons or retroviruses (Schonrich et al., 2017, Sekiguchi and 50 Shuman, 1997). While it is unclear how most viral orthologs of host genes were acquired, the vIL- 51 10 gene in ovine herpesvirus 2 has retained the intron structure of the host gene, supporting a 52 mechanism of DNA-mediated HGT in this instance (Jayawardane et al., 2008). However, because 53 most herpesvirus genes and all known poxvirus genes lack introns, it is unclear how common 54 DNA-mediated HGT events may be. In fact, the taterapox virus genome contains a host-derived 55 short interspersed nuclear element (SINE), which is flanked by a 16 bp target-site duplication 56 (TSD), indicating that the long interspersed nuclear element-1 (LINE1, L1) group of 57 retrotransposons can also facilitate the transfer of host genetic material into poxviruses through an 58 RNA-mediated process (Piskurek and Okada, 2007). 59 The genus Poxviridae comprise many viruses of medical and veterinary importance, including 60 variola virus, the causative agent of human smallpox, vaccinia virus (VACV) the best-studied 61 poxvirus, which was used as the vaccine against smallpox, and monkeypox virus, an emerging 62 human pathogen. Poxviruses can enter a large number of mammalian cell types in a species- 63 independent manner; therefore, productive infection largely depends on the successful subversion 64 of the host immune response (Haller et al., 2014). More than 40% of poxvirus gene families show 65 evidence that they were host-derived, including many genes involved in immune evasion (Hughes 66 and Friedman, 2005, Bratke and McLysaght, 2008). This high frequency of host-acquired genes 67 suggests that poxviruses may represent ideal candidates to study the process of HGT. 68 In order to establish an experimental model of HGT, we exploited the dependence of VACV 69 replication on effective inhibition of PKR, which is an antiviral protein activated by double- 70 stranded (ds) RNA formed during infection with both DNA and RNA viruses. Activated PKR 71 phosphorylates the alpha subunit of eukaryotic translation initiation factor 2 (eIF2a), which 72 inhibits cap-dependent protein synthesis, thereby preventing virus replication (Wu and Kaufman, 73 1997). Because PKR exerts a strong antiviral effect, many viruses, including poxviruses, have 74 evolved inhibitors that can block PKR activation (Langland et al., 2006). Most poxviruses that 75 infect mammals contain two PKR inhibitors (Bratke et al., 2013), which are called E3 (encoded 76 by E3L) and K3 (encoded by K3L) in VACV. E3 is a dsRNA-binding protein, which prevents 3 77 PKR homodimerization and thereby inhibits PKR activation (Davies et al., 1993, Chang et al., 78 1992). K3L, which may itself be a horizontally transferred gene, possesses homology to the PKR- 79 binding S1 domain of eIF2a, and acts as a pseudosubstrate inhibitor of PKR (Beattie et al., 1991, 80 Dar and Sicheri, 2002). 81 In this work we have established a model system to study HGT from the host genome into VACV. 82 In all of the HGT events we detected, the captured genes showed multiple signatures that host 83 LINE-1 retrotransposons were involved in the gene capture. In some cases, integration of the host 84 gene occurred in essential virus genes. In those cases, complementation of both the parental virus 85 and the virus that acquired the gene was necessary for virus replication. After only a few rounds 86 of continued passaging, these complementing viruses recombined to generate single replication- 87 competent viruses. Taken together, retrotransposon-mediated HGT followed by complementation 88 and recombination describes an unexpected cascade of evolutionary processes to rapidly integrate 89 host genes into the viral genome. 90 91 Results 92 93 An experimental system to detect host gene capture by vaccinia virus 94 In order to detect HGT into VACV, we designed an experimental system that uses the selective 95 pressure imposed by PKR on VACV replication. In this model, we use a VACV strain (VC-R2) 96 that lacks both PKR antagonists and can therefore only replicate in PKR-deficient cells or in 97 complementing cells that express viral PKR inhibitors, such as VACV E3 or K3 (Hand et al., 98 2015). We stably transfected PKR-competent RK13 cells with E3L-tagged mCherry (mCherry- 99 E3L) (Figure 1A), selected an individual clone that showed robust mCherry-E3 expression by 100 fluorescence microscopy, and verified that it contained the entire expression cassette (Supporting 101 Figure 1). The PKR sensitive virus VC-R2 can infect these cells as efficiently as wild type VACV, 102 but not wild type RK13 cells. However, if during the course of infection VC-R2 acquired mCherry- 103 E3L from the host cell, the virus would be able to replicate in PKR-competent cells and fluoresce 104 red (Figure 1A). The expression cassette also contains the rabbit b-globin intron (Figure 1B) 105 enabling us to distinguish between direct DNA-mediated integration or indirect RNA-mediated 106 integration, the two most likely mechanisms for HGT into DNA viruses. We infected the RK13- 107 mCherry-E3L cell line with VC-R2 and titered the resulting viruses on permissive RK13+E3+K3 108 cells (Rahman et al., 2013). Then, we infected wild type RK13 cells, which are non-permissive for 109 VC-R2, with these viruses and isolated plaques that expressed both mCherry and EGFP (Figure 110 1A). Overall, we identified 20 recombinant viruses with unique integration sites (Figure 1D, E). 111 In total, we screened approximately 7.5 x 108 virions while optimizing this system. After 112 optimization, we screened 3.63 x 108 virions derived from a single 48-hour infection of RK13 cells 113 (multiplicity of infection (MOI) = 0.01). We screened this population in RK13 cells (MOI = 0.2) 114 and identified 16 unique HGT isolates. This indicates a detected transfer rate of mCherry-E3L into 115 the VACV genome of approximately 1 in 22.7 million virions. 116 117 Horizontally acquired genes show evidence of LINE-1-mediated retrotransposition 118 To map the integration sites, we plaque purified each HGT candidate three times and used both 119 Sanger sequencing of amplicons from inverse PCR (Ochman et al., 1988) and PacBio-based long- 120 read sequencing (Eid et al., 2009). The integration sites were distributed throughout the VACV 121 genome, and were found in intergenic spaces (three