Background
Family
Genomoviridae was accepted in 2016 by the ICTV [
1] as the first one including single-stranded DNA (ssDNA) viruses infecting fungi and insects. These viruses are classified in the phylum
Cressdnaviricota and the recently defined order
Geplafuvirales. They possess a circular ssDNA genome of 1.8–2.4 kb with two open reading frames (ORF) encoding a rolling-circle replication initiation protein (Rep) and a capsid protein (CP). Since its establishment, more than 1450 viral genomes (viral database NCBI) have been described inside the family. Genomoviruses have been classified in ten genus:
Gemycircularvirus,
Gemyduguivirus,
Gemygorvirus,
Gemykibivirus,
Gemykolovirus,
Gemykrogvirus,
Gemykroznavirus,
Gemytondvirus,
Gemyvongvirus and
Gemyptripvirus [
2].
Sclerotinia sclerotiorum hypovirulence-associated DNA virus 1 (SsHADV-1) was the first ssDNA virus demonstrated to infect fungi [
3]. SsHADV-1 has an ambisense circular genome of 2.2 kb with two ORFs, one in positive sense encoding a CP and the other in negative sense coding for the Rep. It also contains two intergenic regions, one of them including the origin of replication (Ori) TAATATTAT [
4]. SsHADV1 is encapsidated forming isometric viral particles that were able to infect
Sclerotinia sclerotiorum extracellularly and reduce fungal lesions in
Brassica napus. SsHADV1 was classified in the genus
Gemycircularvirus. Since its discovery, only two more new mycoviruses of the family
Genomoviridae have been described. Fusarium graminearum gemyptripvirus 1 (FgGMTV1) was the first described tripartite genomovirus composed by DNA-A, DNA-B and DNA-C coding for Rep, CP, and a hypothetical protein, respectively [
5]. FgGMTV1 has been classified within the new genus
Gemytripvirus (gemini-like myco-infecting tripartite virus) [
6]. Multipartitism has been hypothetically explained because of the formation of defective genomes which can be posteriorly, transreplicated and/or transencapsidated, generating new components in the genome [
7], which could be beneficial for viral evolution.
Botrytis cinerea ssDNA virus 1 (BcssDV1) [
8], Botrytis cinerea genomovirus 1 (BcGV1) [
9] and Botrytis cinerea gemydayirivirus 1 (BcGDV1) [
10] have been described infecting the necrotrophic fungus
Botrytis cinerea. They all had a monosegmented genome of approximately 1.7 Kb with at least one ORF encoding for the Rep that contained all conserved motifs of replication-associated proteins of this type of viruses (motifs I to III, GRS domain, Walker A and B, and motif C) [
11]. BcssDV1 was first identified infecting
B. cinerea field isolates from Spanish and Italian vineyards. It contained an ORF that coded for a 380 amino acid (aa) protein containing Rep conserved domains. AACAGTAC was proposed as the nonanucleotide of the Ori. BcGV1 was detected in
B. cinerea strains isolated from bean leaves in China. Its genome contained two ORFs: ORF1 coding for a 321 aa rep and ORF2 coding for a 129 aa hypothetical protein. These two ORFs were separated by two intergenic regions, a large intergenic region (LIR) and a small intergenic region (SIR), and inside LIR sequence, the nonanucleotide TAACAGTAC was marked as possible initiation site for viral DNA replication. BcGDV1 was identified in fungal isolates obtained from asymptomatic plants in New Zealand. It was composed by three ORFs: ORF1 was coding for a 321 aa protein, while ORF2 and ORF3 were overlapping, and encoded for hypothetical proteins of 124 and 97 aa, respectively. ORF1 and overlapping ORF2 and 3 were separated by LIR and SIR sequences, and CTATCAACAC was identified as the putative nonanucleotide sequence in the loop of the stem-loop structure in the Ori region. Reps sequences of BcssDV1, BcGV1 and BcGDV1 shared an identity of 98% among them, so according to demarcation criteria that no sequences from different species share > 78% aa identity in the Rep protein [
6], they should be considered as members of the same viral species.
As well as SsHADV1, both FgGV1 and BcGDV1 showed hypovirulence traits in its respective hosts
F. graminearum and
B. cinerea [
10,
12]. FgGV1 infectious clones inoculated in protoplasts were proven to induce hypovirulence in
F. graminearum [
5] and BcGDV2 purified particles applied to the growing margins of a
B. cinerea virus-free strain resulted in a hypovirulent phenotype of the new infected strain in canola leaves. Additionally, BGDaV1 methylation studies proved that
B. cinerea host defense machinery targets the BGDaV1 genome down-regulating BGDaV1 RNA accumulation. This silencing results in a long-term mutualistic infection relation that makes both survive during viral infection [
10].
B. cinerea associated genomoviruses are potential biotechnological tools to understand virus-fungi-plant interactions and to provide solutions for the control of gray mould diseases. In fact,
B. cinerea is one of the most important and aggressive plant-pathogenic fungus worldwide [
13]. It infects a wide range of crops, in the field and in postharvest, in form of mycelia, spores or sclerotia for long periods of time. Chemical fungicides, as the main control method, tends to change to more sustainable solutions as mycovirus-derived control methods [
14].
In this work, we show the molecular characterization of several variants of BcssDV1 identified in different isolates of Italy and Spain. It is demonstrated that BcssDV1 genome is composed by four segments, instead of a single one as was previously published [
8]. Additionally, in silico alignment and prediction methods helped us to elucidate the putative functions of the hypothetical proteins in the interactions with their host.
Discussion
The full length genome of BcssDV1, a ssDNA virus belonging to family
Genomoviridae, with a tetrasegmented genome, has been characterized in
B. cinerea isolates of Italy and Spain. FgGV1 was the first described mycovirus with a multisegmented genome inside the family
Genomoviridae [
5]. Both FgGV1 and BcssDV1 possess a L-CR and a S-CR in all their genomic segments. Other ssDNA multisegmented viruses also exhibit these common regions in their segments, such as nanoviruses that contains CR-SL and CR-M or nanoviruses with CR-II [
33]. The BcssDV1 S-CR also encompasses the proposed Ori, where the stem-loop structure displays similar secondary structures for all genomic segments. While other multisegmented geminiviruses as tomato golden mosaic virus (TGMV) or bean golden mosaic virus (BGMV) also share the same stem-loop structures [
34], FgGV1 stem-loop structure differs in DNA-A and DNA-B with DNA-C.
BcssDV1 has been found in different regions of Italy and Spain with an overall incidence of 27.8%. From 98 samples of Italy and 150 of Spain, a total of 69 samples resulted positive by RT-PCR. Other mycoviruses infecting the fungus
B. cinerea showed similar or lower incidences: Botrytis virus F showed 14.3% (12/84) in New Zealand and worldwide [
35], Botrytis cinerea mitovirus 1 showed an overall incidence of 30% in different regions of Southern Spain (29/96) [
36] and Botrytis cinerea mymonavirus 1 was only detected in four out of the 508 (0.8%)
B. cinerea strains isolated from China [
37]. However, there are examples of higher incidence rates in mycoviruses infecting other fungal hosts, such as Hymenoscyphus fraxineus mitovirus 1 (HfMV1), which displayed an incidence rate of 86% in the population of Sargans (N = 103 isolates; Eastern Switzerland) and 95% in the population of Aigle (N = 109 isolates; Western Switzerland) [
38]. It is worth noting that BcssDV1 is not limited to Spain and Italy, as other variants have been identified in New Zealand (QLD98948.1) and China (QPB44148.1, UIX26077) [
9,
10]. Given its high incidence and dispersal, it is plausible to suggest that BcssDV1 may also infect
B. cinerea isolates worldwide in many different hosts.
Regarding mean pairwise genetic distances, DNA-C of BcssDV1 showed the highest mean pairwise identity (0.075 ± 0.007), while DNA-A displayed the lowest (0.011 ± 0.001). a partial sequence of Botrytis cinerea mitovirus 1 genome showed values of 0.023 ± 0.005 [
36], higher than the mean value of pairwise identity found in DNA-A but lower than those obtained for segments DNA-B, DNA-C and DNA-D. In contrast, the mean pairwise identity of Botrytis virus F sequences (0.162 ± 0.021) was significantly higher than the ones obtained for each of the genomic segments of BcssDV1 [
35]. The observed differences in incidence and nucleotide-level variability among these viruses can be attributed to their belonging to different viral families, possessing distinct types of genomes, and varying numbers of genomic segments.
Rep of other members of the family
Genomoviridae display recognizable sequence similarity to Rep of member of the family
Geminiviridae. In fact, DNA-A variants demonstrate high pairwise identities between aa sequences and contain characteristic aa Gemini_AL1 Rep catalytic conserved domains. In contrast, DNA-B, coding for the CP, and DNA-C, coding for the HP1, present the lowest minimum identities at aa level. This variability may be explained by the low values of sequence similarity previously found among CP and other hypothetical proteins of viruses belonging to the family
Genomoviridae [
6]. The obtained results suggest that the four BcssDV1 genomic segments do not followed the same evolution patterns. For instance, it has been observed that for bipartite begomoviruses, DNA-A and DNA-B respond differentially to evolutionary processes, with DNA-B being more permissive to variation [
39]. Furthermore, evolutionary studies of FBNSV, formed by 8 segments, have demonstrated that each segment is subjected to variation and selection as an individual entity [
34].
Phylogenetic analysis at nt or aa level reveals limited clusterization based on country or different and distant regions inside Spain, with Italian and Spanish samples clustering in different groups and Spanish isolates from La Rioja with tendency of forming groups with isolates from Penedés. Similar patterns of clusterization based on geographic regions have been observed in full-length genomes of other multisegmented ssDNA viruses like faba bean necrotic stunt nanovirus, which also formed two distinct groups, including different regions of Iran [
40]. Additionally, begomoviruses have shown a clear geographic segregation of segments DNA-A and DNA-B, and some differences in the genetic structure of both [
39].
FgGV1 was composed by three segments DNA-A, DNA-B and DNA-C, encoding Rep, CP, and a hypothetical protein, respectively. While DNA-A and DNA-B of FgGV1 were mutually interdependent and sufficient for their replication, DNA-C enhances the accumulation level of viral DNA in infected fungi and facilitates transmission via conidia [
5]. Similar to FgGV1, DNA-A of BcssDV1, encoding Rep, contains the conserved domains of Gemini_AL1, while DNA-B encodes the hypothetical CP. BcssDV1 DNA-C and DNA-D both code for two hypothetical proteins without conserved motifs, which are presumed to be involved in different stages of the BcssDV1 life cycle. Tridimensional structure and function of the replication-associated protein has been deeply characterized in different viruses of family
Genomoviridae [
41‐
43].
The BcssDV1 Rep sequence not only contains Gemini_AL1 Rep catalytic conserved domains, but also exhibits structural homology with predicted tertiary conformation of Rep and DNA-binding domain of the replication initiation protein of members of the family
Geminiviridae, the primase-helicase of a DNA virus of the family
Papillomaviridae and an ATP-dependent protease ATPase. Geminiviruses encoded Rep plays an important role in the replication and transcription of the other viral genes and possesses DNA helicase and ATPase activity [
44]. Geminiviruses replicate inside the nucleus, and Rep contains NLS, and removal of the aa of the NLS reduce nuclear import and nuclear accumulation of Rep [
45]. BcssDV1 Rep (321 aa) also contains one NLS and spliced Rep (380 aa) contains three NLS, suggesting the importance of these NLS in the import and accumulation of the protein in the cell nucleus. Therefore, our results definitively confirm the role of BcssDV1 Rep in the life cycle of this mycovirus. Due to low conservation at sequence level of CP, HP1 and HP2 with other viral proteins, prediction of the tertiary structure of these proteins have allowed to propose their possible function in BcssDV1 life cycle. The BcssDV1 CP showed structural homology with CPs of ageratum yellow vein virus (AYVV) and faba bean necrotic stunt virus (FBNSV), other circular ssDNA viruses, confirming the function of DNA-B protein in the encapsidation of the mycovirus. AYVV is a plant geminivirus that form a two incomplete icosahedral particle, fused to form a geminate capsid [
28]. FBNSV is a nanovirus whose distinct genome segments are encapsidated individually in icosahedral particles [
46]. In the future, electron microscopy analysis will be conducted in order to analyse the structure of the BcssDV1 particles. Since the transcripts derived from BcssDV1 will be synthetized in the nucleus, it will be canonicals mRNA with 5’-CAP and 3’ poly-A. The first step in translation of 5′-capped host-cell mRNAs is the binding of the initiation factor eIF4F, that consists of three subunits. eIF4E binds to the cap, eIF4A is an RNA helicase and eIF4G connects the ribosome to the mRNA through eIF3 [
47]. In plant potexviruses and carlaviruses CP interaction with this factor is necessary but not sufficient for facilitating the accumulation of these type of viruses [
48]. BcssDV1 CP showed structure homology with a eukaryotic translation initiation factor 4-e binding protein, suggesting that could be involved in the interaction with eIF4E and possible in the accumulation of BcssDV1 inside its fungal host. The CP of geminiviruses contains NLS and is involved in the nucleocytoplasmic shuttling of the viral DNA [
45], however BcssDV1 CP does not contain NLS. Notably, in this work, seven hypothetical conserved motifs were detected in CP sequences of viruses of the family
Genomoviridae and other families, suggesting their potential importance in the initial stages of viral infection, could be through the interaction with a membrane cellular receptor, in viral particle assembly, in nucleocytoplasmic movement of proteins or DNA, or as a pathogenicity determinant as has been described for other geminiviruses [
45]. Discovering new viruses of the family
Genomoviridae and performing directed mutagenesis of these specific regions would help to elucidate their significance in viral structure and the role of the CP in interactions with the host. For BcsssDV1 DNA-C, HP1, structural homology was found with human ephrin type a receptor 2 transferases and signalling proteins of ephrin type-b, involved in regulation of cell–cell adhesion and motility in cancer cells [
49]. In TYLCV1, the AC4 has a role in complementing the function of viral movement and degree of severity of the symptoms [
45], then, HP1 of BcssDV1 may be contributing to movement and pathogenesis between the mycovirus and host cells. Interestingly, it was also found structural homology of DNA-C with an apoprotein with two sequence domains, a putative DNA-binding domain and DNA/RNA helicase domain. This protein is a negative regulator of G1/S transition of mitotic cell cycle. Then, since ssDNA mycovirus need to inhibit the DNA damage checkpoint, resulting in cell cycle progression while also stimulating DNA replication by preventing programmed cell death [
45], BcssDV1 HP1 could be involved in the interaction with the DNA, maintaining separated the viral DNA duplex formed in the nucleus to facilitate the mycoviral replication cycle, instead to be related with the regulation of the cell cycle. Among all the analysed segments of BcssDV1, segment DNA-C was the most diverse among all variants and contains hypothetical motifs related to viral movement proteins. It is plausible that DNA-C represents the most susceptible segment within the viral genome for mutation and evolution, thereby facilitating future adaptation to yet undiscovered hosts. Considering this hypothesis, it is conceivable that HP1 shares functions with the movement protein of Bean golden mosaic virus (BGMV) and TGMV, both of which are considered essential factors in host adaptation, alongside other genes [
50]. Furthermore, the description of new host of Sclerotinia sclerotiorum hypovirulence-associated DNA virus 1 (SsHADV-1) has been confirmed in
Penicillium olsonii [
51], raising the possibility of BcssDV1 infecting other ascomycetes.
In DNA-D HP2, structural homology was observed with the kinetochore heterodimer from yeast, which is involved in cell cycle, and with cell division cycle-related kinase. Results also supported homology with a cell division cycle related protein kinase (CDC7) which activity is critical for the G1/S transition of the replication cell cycle. Control and regulation replication cell cycle are some of the most important parts in the viral cycle. For instance, AC4 protein of TYLC1 acts as a viral “oncogene” stimulating cell DNA replication [
45]. Additionally, HP2 also exhibited structural homology with RNA polymerase i-specific transcription initiation factor of
Saccharomyces cerevisiae [
52]. These factors play an essential role in cap-bearing cellular and viral RNAs, from the nucleus to the cytoplasm [
49]. It could be that BcssDV1 HP2 acts stimulating replication cell cycle to favour BcssDV1 replication inside the nucleus. As another possible hypothesis, HP1 and HP2, which seem to have antagonistic effects on cell division cycle, could play a combined role in regulating it to favour replication of BcssDV1. Construction of a synthetic BcssDV1 would allow to determine the real function of their predicted proteins and study the mycoviral life cycle and its interaction with the fungal cell.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.