Endogenous Caulimovirids: Fossils, Zombies, and Living in Plant Genomes
Vassilieff, Geering, Choisne, Teycheney, Maumus · 2023
공개 학술지 Biomolecules 에 실린 총설 논문입니다(2023년). CC BY 4.0 으로 공개된 글이고, 그림 · 표 · 참고 문헌 표시를 뺀 본문을 실었습니다.
출처 Vassilieff et al. (2023), Biomolecules · 라이선스 CC BY 4.0 · 원문
이 장은 원문만 있습니다. 문장별 해석은 앞 장부터 차례로 붙입니다.
4장 · Endogenous Caulimovirids: Fossils, Zombies, and Living in Plant Genomes — Integration, Genomic Distribution, and Expression of ECVs
Mechanisms of Integration
The mechanism(s) by which caulimovirid DNA is captured in plant genomes remain(s) speculative. However, there are two widely held hypotheses, the first implicating repair mechanisms for dsDNA breaks in chromosomes and the second involving the integration machinery of LTR retrotransposons.
ECV loci are often flanked by TA dinucleotide simple repeats ((TA)n). These simple repeats are palindromic in nature and can form stable non-B DNA structures (e.g., cruciform and hairpin structures) that stall the replication fork and lead to double-strand breaks (DSBs) in the chromosomes through local fragility or from the action of DNA nucleases. The integration bias of ECVs toward (TA)n repeats may thus result from the insertion of caulimovirid DNA as filler DNA during the repair of these DSBs, by non-homologous end joining (NHEJ) or microhomology-mediated recombination.
ECVs have been frequently observed to form tandem arrays and multicopy rearrangements, sometimes leading to integration hotspots (see below). Such structures could arise directly from the integration of concatemers of viral DNA or from homologous recombination either between existing ECVs or between an ECV and the episomal form of a cognate virus. Interestingly, a recent study has demonstrated the capability of the fission yeast Tf1 LTR retrotransposon to transpose through an integrase-independent pathway. In this system, Tf1 transposition occurs by homologous recombination mediated by the single-strand annealing protein Rad52 and not by canonical homologous recombination that would require Rad51, resulting predominantly in insertions within existing copies of related elements. Rad52 homologs were identified in gymnosperms, monocots, and dicots and could promote the integration of ECVs by a similar homologous recombination mechanism.
An alternative path for the integration of ECVs could involve retrotransposons through recombination between a caulimovirid genome and an LTR retrotransposon, followed by the integration of the resulting chimera. Such a mechanism has been suggested for the integration of badnaviruses in the genome of Musa balbisiana and that of bornaviruses and flaviviruses in the genomes of mammals and mosquitoes, respectively. It has also been suggested that caulimovirid integration could involve the hijacking of a transposon-derived integrase.
How Caulimovirid EVEs Have Invaded Plant Genomes
The tendency of ECVs to integrate into hotspots and their abundance raises the question of their accumulation and maintenance in plant genomes. In principle, each ECV results from a single integration event, except those resulting from segmental duplications, whole genome duplications, or natural hybridization. In many plant genomes, there is evidence of recent integration events, reflected by the presence of highly similar copies of ECVs. For example, de Tomas and Vicient reported over 30 clusters of identical ECVs in angiosperm genomes spanning different plant genera, suggestive of very recent integration events. Most of these clusters correspond to petuviruses and florendoviruses for which no episomal form has been reported. At least two hypotheses could explain this situation: either (i) these viruses became extinct recently, or (ii) they are still extant and cause asymptomatic or unnoticed infection; hence, they were never isolated. The first hypothesis would involve massive virus extinction events, which cannot be ruled out a priori. Although most extant members of the Caulimoviridae cause disease symptoms, asymptomatic viral infections are produced by some badnaviruses such as taro bacilliform virus and are commonplace in plants; therefore, the second hypothesis cannot be ruled out either.
Alternatively, repetitive ECV could result from endogenous proliferation, which would not require an episomal stage: viral dsDNA would rather be produced from replication-competent EVEs or in trans by complementation of defective EVEs by episomal counterparts. Such a situation exists for the endogenous retrovirus group HERV-K (HML2) in primates, for which proliferation likely results from the reinfection of germline cells. In plants, genes encoding caulimovirid MP homologs are widespread in the genomes of euphyllophytes. Their constitutive expression could facilitate cell-to-cell movement of caulimovirid virions and promote iterative infection of plant reproductive tissues, resulting in repetitive ECVs.
Distribution of Caulimovirid EVEs in Plant Genomes
Only a few studies have attempted to quantify the contribution of ECVs to angiosperm genomes, one being that of Kim et al., who estimated that caulimovirid sequences represent 0.86% and 1.02% of the Capsicum annuum and Capsicum chinensis genome assemblies, respectively. The abundance of ECVs in plant genomes is very variable, ranging from trace amounts to above 0.5% of the genome content for Jatropha curcas, Amborella trichopoda, Citrus clementina, Vitis vinifera, and Beta vulgaris, and above 1% in Ricinus communis and Solanum melongena.
To provide a rough estimate of the relative abundance of ECVs across tracheophytes, Diop et al. counted the number of ECV reverse transcriptase (ECRT) sequences in various plant genomes and normalized this data according to the size of the plant genome. With a few exceptions, they found ECRTs in all the genomes of the ferns, gymnosperms, and angiosperms they analyzed, albeit at highly variable densities. Overall, ECRTs were relatively abundant in gymnosperms (over 500 copies per genome) and less abundant in monocots than in dicots, among which some botanic orders (e.g., Solanales and Malpighiales) tend to display high ECRT densities. The highest densities were registered in Citrus sinensis (sweet orange) and Ricinus communis (castor bean) at 2.3 and 2 ECRTs per Mb, respectively, compared to an average density of 0.2 ECRT/Mb across the 62 seed plant species that were analyzed.
The distribution of ECVs along host chromosomes has been investigated in several ECRT-rich angiosperm genomes. In silico and cytological analysis showed that ECVs are present in all chromosomes of Citrus maxima, Medicago truncatula, Beta vulgaris, Solanum lycopersicum, Petunia hybrida, and Fritillaria imperialis L.. In the genome of Capsicum annuum, which contains about 1,500 ECRTs, their distribution was found to be homogeneous along all chromosomes. However, ECVs often tend to cluster in hotspots, which can be unevenly distributed across chromosomes. For example, hotspots of petuvirus EVEs were observed in most, if not all, centromeres of the giant genome of monocotyledonous species Fritillaria imperialis, whose estimated size is ~42 Gbp. This situation could result from a recent host swap since the genome of closely related Fritillaria species is devoid of petuvirus EVEs. ECVs were also found to be a component of centromeric repeats in citrus hybrids, suggesting that they could play a role in the centromeres of different plant species.
In a majority of plant genomes, ECVs were found to be frequently located in close proximity to transposable elements (TEs). In the genomes of Beta vulgaris, Petunia hybrida, and Solanum lycopersicum, they localize preferentially in gene-poor, TE-rich pericentromeric regions. This preferential distribution could result from genome-wide selection against deleterious events, such as EVE integrations, which are more likely to occur in gene-rich regions where integrants have the potential to impact both the structure and expression pathways of host genes. Therefore, as for TEs, heterochromatin/pericentromeric regions may represent “safe havens” for ECVs, where they can escape purge mechanisms and their retention is neutral to the host and can last over extended evolutionary periods.
Replication-Competent ECVs
Replication-competent ECVs have been reported only for endogenous banana streak viruses (eBSVs) in banana, endogenous petunia vein-clearing virus (ePVCV) in petunia, and endogenous tobacco vein-clearing virus (eTVCV) in Nicotiana edwardsonii. They are the only currently known examples of infective EVEs in plants, although it is likely that others will be discovered. Activation by biotic and abiotic stresses triggers their expression, leading to spontaneous infections; therefore, they are often referred to as replication-competent EVEs. Replication-competent eBSVs were discovered in 1999, but their complex structure was elucidated only in 2013. Chabannes et al. discovered that infective eBSVs of three BSV species, BSOLV, BSGFV, and BSIMV, are present as allelic forms in Musa balbisiana genomes: eBSOLVs and eBSGFVs are di-allelic, and only one allele is infective, whereas eBSIMV is mono-allelic, although there are minor sequence differences between alleles, and both alleles are infective. The structure and/or infective nature of eTVCVs and ePVCVs were unraveled in 2000 and 2003, respectively. The mechanisms of activation of infective ECVs are still poorly understood; however, wounding, temperature differences, water stress, grafting, and cell culture have been identified as activating stresses, and activation has been observed only in interspecific hybrids. Nicotiana edwardsonii is a hybrid between Nicotiana glutinosa, whose genome harbors infective eTVCVs, and Nicotiana clevelandii. Likewise, Petunia hybrida results from a cross between Petunia inflata and Petunia axillaris subsp. axillaris, whose genomes both harbor infective ePVCVs. Finally, activation of infective eBSVs occurs in triploid (AAB) or tetraploid (AAAB) interspecific banana hybrids resulting from crosses between Musa balbisiana (B genome), which harbors infectious eBSVS, and Musa acuminata (A genome).
Two mechanisms have been proposed to explain the release of episomal viral genomes from ECVs. They depend on the structure and location of the ECVs in the host genome. Endogenous PVCVs are integrated into tandem arrays in the genome of Petunia hybrida. They form head-to-tail concatemers, allowing transcription from the promoter of the first copy to the polyadenylation site of the next, resulting in transcripts of complete viral genomes that would be infectious. In Musa balbisiana, replication-competent eBSVs display complex rearrangements. Fragmented viral sequences are often duplicated in the same or opposite directions. A homologous recombination (HR)-based model has been proposed to explain the release of a banana streak GF virus (BSGFV) genome from its endogenous counterpart, eBSGFV-7, which requires two recombination events. The first one between two inverted repeat sequences would produce a complete linear copy of the infectious eBSV with its three ORFs in the same orientation. The second one would excise a viral circular molecule resembling the cognate BSV genome, from which infection could arise. It is important to note that the insertion of infective ePVCVs and eBSVs is supposed to be relatively recent, preventing the accumulation of mutations and sequence decay that generally make EVEs nonfunctional.
Epigenetics and Silencing
The epigenetic status of ECVs has only been addressed in a few angiosperms. Like most repetitive DNA elements, ECVs are subjected to DNA methylation and histone modifications (i.e., H3K9me2), which are hallmarks of transcriptionally silent heterochromatin environments. Furthermore, in silico mining of small RNA (sRNA) datasets and actual sequencing experiments have shown that a heterogenous population of 18–30 nt sRNAs that are homologous to ECVs are present at variable densities in a large range of angiosperms. This pool of sRNAs is probably composed of subcategories acting in different silencing pathways, including transcriptional gene silencing (TGS) and post-transcriptional gene silencing (PTGS). For instance, the predominant function of plant 24-nt sRNAs is to guide the canonical RNA-directed DNA methylation (RdDM) pathway, while 21–22 nt sRNAs are usually involved either in non-canonical RdDM pathway or in PTGS.
Upon viral infection, plants produce a diversity of virus-derived sRNA as a defense mechanism to restrict viral replication and limit its spread (reviewed in). In plants, Jakowitsch et al. hypothesized that tobacco-endogenous pararetroviruses (TEPRVs) contribute to resistance against the cognate exogenous viruses, and Mette et al. provided indirect experimental evidence supporting this hypothesis. Furthermore, it was demonstrated in insects that gene silencing targeting a non-retroviral EVE results in the inhibition of the replication of a closely related exogenous virus via gene silencing mechanisms. Therefore, sRNA produced from ECVs could also play a central role in the defense against cognate viruses.
Recently, Valli et al. reported the presence of a subset of ECVs with high 22-nt sRNA mapping density in the Solanum lycopersicum genome. These regions, called TSAs (transcriptionally competent siRNA areas), correspond to rearranged ECVs with inverted repeat (IR) conformations. TSAs can thus produce transcripts that are prone to the formation of hairpin structures known to promote the production of 22-nt siRNA by dicer-like protein 2 (DCL2). A similar situation was reported from the genome of soybean, in which inverted repeats derived from endogenous cucumber mosaic virus sequences (family Bromoviridae) have the potential to form a hairpin structure promoting the production of sRNAs. Another study showed that caulimovirid- and TE-derived IR in soybean fuel the production of DCL2-dependant 22-nt sRNA. Furthermore, in their analysis of sRNA landscapes in interspecific Solanum hybrids, Lopez-Gomollon et al. found a predominant deregulation in loci corresponding to ECVs, affecting mainly the levels of 22-nt sRNA as compared to 21-nt and 24-nt species. Remarkably, Valli et al. further demonstrated that TSA-derived 22-nt sRNA can trigger the degradation of transcripts carrying the complementary sequence through PTGS. Altogether, these results suggest that 22-nt sRNA produced from TSAs, and more generally from EVE-derived IRs, could play a central role in the defense against EVE amplification and/or against infection by cognate viruses through silencing by non-canonical RdDM or through PTGS.
이 글은 위 출처의 원문을 그 라이선스(CC BY 4.0)에 따라 옮겼습니다. 참고문헌 목록과 인용 번호, 그림 · 표 · 각주 · 수식 · 부록은 빼고 절별로 나누어 실었으며, 그 밖의 문장은 원문 그대로입니다. 원 저작자와 발행 기관은 이 화면을 만들거나 검토하지 않았고, 이곳의 해석과 낱말 풀이는 공식 번역이 아닙니다.
