Background
Defective RNAs (D RNAs), which have internal deletion of viral sequences, have been described for a variety of plant viruses [
1,
2]. Some D RNAs interfere with replication of the helper virus (called as defective interfering or DI RNAs) and affect symptom phenotypes, whereas others have little effect [
1,
3]. Previous reports have shown the generation of D RNAs or DI RNAs was a general biological process among multipartite RNA viruses of plants [
1,
4].
Beet Necrotic Yellow Vein virus (BNYVV), transmitted by zoospores of
Polymyxa betae, is a member of the genus
Benyvirus causing a worldwide sugar beet disease rhizomania [
5]. BNYVV contains four or five plus-sense single-stranded RNAs designated as RNA 1 to RNA 5 in descending order of molecular size, individually packaged into rod-shaped virions [
6]. RNA 1 and 2 encode "house-keeping" genes involve in replication, virion assembly, cell-to-cell movement, silencing suppression and vector transmission [
7‐
10]. Only RNA1 and RNA 2 are sufficient for virus replication in local lesion host and virus vascular movement in the systemic host such as
Spinacea oleracea or
N. benthamiana[
11‐
14]. RNA 3, RNA 4 and RNA 5 are functional in the natural infection process [
15]. p25 encoded by RNA3 is required for induction of classical rhizomania symptom on natural hosts and local lesion phenotype on leaves of local hosts, while the downstream N protein could induce necrosis outside of the context of a BNYVV infection [
16‐
18]. The core region from nt1033 to nt1257 of RNA3 is essential for the vascular movement of BNYVV in
Beta macrocarpa[
14]. RNA4 encoded p31 is associated with efficient vector transmission, virulence and RNA silencing suppression in roots [
19,
20]. Some BNYVV isolates contain RNA5 encoding a single 26-kDa protein that can also enhance symptom severity in a synergistic fashion with RNA3 [
21‐
27].
D RNAs usually contain in-frame deletions within one or more genes and in some cases may be associated with pathogenicity of virus [
2]. Several D RNAs have been described in BNYVV infections [
28]. During serial mechanical passages, D RNA-2a and D RNA-2b are generated by in-frame deletions in the read-through region of the coat protein, and the mutant strains could not be transmitted by the fungal vector, indicating that the read-through product is necessary for vector transmission [
28]. Deleted forms of BNYVV RNAs 3 and 4 have been detected in field isolates and induced by mechanical inoculation [
29]. In some isolates, RNA 3 and 4 are prone to be eliminated spontaneously, whereas they are usually persistent with shorten forms in others isolates [
29‐
32]. When the
in vitro transcripts of RNA 3 and 4 are inoculated on
Chenopodium quinoa leaves, they usually generate internal deletions spontaneously within only one or two mechanical passages [
33]. Besides the naturally occurring D RNAs, a series of artificial deletion mutants derived from BNYVV RNAs have been constructed and characterized to investigate the function of each gene on symptom expression of BNYVV [
17,
34‐
36].
BNYVV can infect
N.benthamiana systemically, causing severe or mild symptom [
11]. Early reports show that RNA4 encoded p31 is associated with severe symptom such as curling and stunting in
N. benthamiana, whereas RNA3 is non-related with these severe symptom [
20]. Furthermore, RNA3 of Japanese isolate O11 is not stable in the systemic infection, and usually eliminated during virus propagation in
N. benthamiana[
20]. However, the reason why the RNA3 is unstable in the systemic movement still remains to be determined.
In this paper, the Chinese isolate BN345 (RNAs 1, 2, 3, 4 and 5) and BN3 (RNAs 1 2 and 3) were used to infect
N. benthamiana[
37]. Unlike the eliminated RNA3 of reported isolate O11, the generation of internal deletion RNA 3 mutants (defined as D-RNA3s) occurred with a high frequency in the systemic infected leaves. Most interestingly, the D-RNA3s could cause more severe symptom than full-length RNA3. Sequencing of RT-PCR showed the D-RNA3s included three distinct deleted forms. The inoculation of
in vitro transcripts revealed that the D-RNA3s were very stable in the systemic movement, and induced the obvious necrotic symptom in the
N. benthamiana, which provided the first evidence showing the stability of viral RNAs could be improved by spontaneous deletion under the selective pressure of systemic plants. Besides, the severe necrotic along the vein of systemic leaves induced by D-RNA3s confirmed N protein is responsible for these symptom. Finally, the possibilities of D-RNA3s causing severe symptom in
N. benthamiana were discussed.
Discussion
For the multipartite RNA viruses, how the distinct functions of each RNA genome and protein comparatively facilitate the viral infection in different hosts and environment is an intriguing topic in the current studies. For accommodating to different conditions, the viruses usually prefer to replicate some necessary RNA, and discard the full-length or some part of other inessential genome RNA. For example, the RNA3 of isolate O11 was usually eliminated spontaneously in the
N. benthamiana plants [
20], and the RNA5 does not persist in many natural isolates [
5,
8]. Here, we examined the infection activity of the BNYVV RNA1, 2, 3, 4, 5 from Chinese isolate, and found the RNA1, 2, 4 successfully infected in the systemic leaves of host
N. benthamiana as reported previously [
20]. The RNA5 was almost under detected in the systemic leave whether co-infected with RNA1 2 3 4 or only with RNA1 2 (Figure
1), indicating the RNA5 was inessential factor at least on the
N. benthamiana, although the RNA5 could enhance the pathogenicity of BNYVV in some other hosts [
23,
37]. At present, it is unknown whether the 5' and 3' un-translated region (UTR) or the open reading frame (ORF) affects the infection of RNA5 on the
N. benthamiana. Different with instability of RNA3 from isolate O11 reported [
20], the RNA3 of Chinese isolate generated some internal deletion forms in most systemic leaves of infected
N. benthamiana. RNA3, 4, 5 could replicate with high efficiency in the local host
T. expansa, indicating the different stability of RNA3, 4, 5 in systemic host is resulted from the distinct systemic movement activity, rather than the replication efficiency.
The full-length RNA3 was eliminated spontaneously in some systemic leaves of
N. benthamiana, whereas the D-RNA3αwas more stable in the systemic movement (Figure
5). This paper provides an evidence to prove RNA3 could increase the systemic movement activity by spontaneously deletion on the
N.benthamiana. Indeed, the northern results about the RNA3 of isolate O11 also contain some deletion forms in the previous results [
20]. In addition to enhance the stability of RNA3 by the internal deletion, the inoculation assay in this study also demonstrated that one of D-RNA3s not only caused obvious necrotic spots on
T. expansa, but also induced severe necrotic in the systemic leave of
N. benthamiana. Therefore, the natural D-RNA3s, rather than full-length RNA3, acts as a new pathogenicity factor involving into the infection on the
N. benthamiana, in addition to the reported RNA4 [
20].
The sequence analysis of the natural D-RNA3α showed the deletion region (nt453-1057) led to a frame-shift mutation and abolished the synthesis of p25 protein (Figure
2A). In the genomic RNA3, the N protein with overlapping region with C terminal of p25 protein was confirmed to elicit a necrotic response through artificial cloning or expression by
Cauliflower mosaic virus[
17]. Here, the first four amino acids of N protein were deleted in the natural D-RNA3α. According to the previous results, the deletion of upstream sequence might activate the expression of N protein [
17], which would be responsible for the necrotic symptom. To confirm this prediction, the N
tr protein pre-terminated mutants were constructed and inoculated on the local host and systemic host. The mutant of D-RNA3αM1 did not elicit the necrotic response any more in the local host as the D-RNA3α, indicating the N
tr protein was related with necrotic symptom. However, when inoculated on the systemic host
N. benthamiana, the mutant of D-RNA3αM1, unlike the stable D-RNA3α, was unable to infect
N. benthamiana systemically, confirming that the deletion region in the D-RNA3αM1 was also related with systemic movement as reported previously [
14].
For confirming the function of N
tr protein in the systemic infection, the second mutant D-RNA3αM2 was constructed and successfully infected the
N. benthamiana systemically without inducing the necrotic response. The results showed here indicated that the N protein, acting as inducer of necrotic symptom, was under-expressed in the wild-type RNA3 probably due to inhibition by the expression of p25. Based on these finding, we proposed that when the N protein was required for inducing symptom or enhancing stability in
N. benthamiana, the wild-type RNA3 might adopt spontaneous deletion to improve the expression of N protein by sacrificing the temporarily non-functional p25 (Figure
5).
The previous results have reported various D-RNA3s from different hosts [
8,
33,
35]. However, the natural D-RNA3s from systemic host
N. benthamiana firstly reported in this paper was different with the previous sequences. Extensive experiments will be needed to determine the similarities and divergence of the natural deletions undergoing in different hosts, as well as the meaning of their emergence.
Methods
Plants and virus isolates
T. expansa, N. benthamiana, and C.
amaranticolor were grown at 24 ± 1 °C under a 16 h light and 8 h dark regimen. In this studies, the BN12(RNAs 1+2), BN3(RNAs 1+2+3), BN345 (RNAs 1+2+3+4+5) were respectively derived from isolates BNYVV-Hu0, Hu3, and Hu as described previously [
37]. BN4 (RNA1+2+4), BN3α (RNA1+2+3α), BN3αM (RNA1+2+3M1 or M2), BN125 (RNA1+2+5) were the mixture of total RNA from the inoculated leaves of
T. expansa by BN12 and
in vitro transcripts of each genome.
Construction of full-length and defective RNA3 infectious cDNAs
The full-length and the defective RNA3s cDNA clones were amplified by RT-PCR from total RNA of isolate BN3 and BN3A. The forward primer 5'- CG
GAATTC TAATACGACTCACTATAG AAATTCAAAA TTTACCATTA - 3' (
Eco RI site in italics and T7 promoter sequence underlined) and reverse primer 5'-CC
TCTAGA T
(26) GTCAATACACTGACAGAGAA -3' [
Xba I site in italics and oligo (dT) tract shown by T
(26)] were used. The PCR products were purified and cloned into the pMD19-T vector to obtain a full-length cDNA clone named pMDR3 and three D RNAs cDNA clones, pMDRNA3α, pMDRNA3β and pMDRNA3γ (Figure
2A). Mutant RNA3αM (Figure
2A and
2B.) was produced by PCR-based, oligonucleotide directed mutagenesis within the pMDRNA3α clone. All constructions were sequenced and then used for transcription.
In vitro transcription and inoculation
Infectious plasmids were linearized by Xba I and used by run-off transcription at 37°C for 2 hr with a T7 RNA polymerase kit as described by the manufacturer (Promega). The freshly synthesized transcriptions were mixed with total RNA of T. expansa infected by isolate BN12, the mixture supplemented with an equal volume of inoculation buffer (50 mM glycine, 30 mM K2HPO4, 1% bentonite, 1% celite, pH 9.2) were rubbed onto T. expansa, C. amaranticolor or N. benthamiana leaves. Local lesions generally appeared at 5-8 days post inoculation, while systemic symptom of N. benthamiana appeared at 12-14 dpi.
RNA extraction and detection
Inoculated
T. expansa leaves and systemic infected
N.benthamiana were harvested at 7 and 14 dpi, respectively. Total RNA was extracted for RT-PCR detection and Northern hybridization, as described previously [
38]. Probes were appropriate 32
P-labeled cDNA specific for RNA1 (nt5815-6531), RNA2 (nt145-714), and RNA3 (nt445-1102) sequences, respectively.
The following primers were used for RT-PCR: for detection of RNA 2, the forward primer BN81 (5'-CGATGTCGAGTGAAGGTAGATA-3', nt145 to 164) and the reverse primer BN80 (5'- CTATTGTCCGGGTGGACTGG -3', complementary to nt962 to 712), for detection of RNA3, the forward primer BN78 (5'-GTGATATATTAGGCGCAGTTTATG-3', nt450 to 473) and the reverse primer BN77 (5'-TCATTATCATCAACACCGTCAG-3', complementary to nt1080 to 1101), for detection of RNA 4, the forward primer BN209 (5'-CTGATGGAGAGATATG-3', nt384 to 339) and the reverse primer BN210 (5'-CTAATCGTGATAAAAGACAAACCA-3'complementary to nt1205 to 1228) and for detection of RNA 5, the forward primer BN225 (5'-GATGGATATTGATCATTGTATG -3', nt459 to 480) and the reverse primer BN226 (5'-TCCACAATCATTATCATGAT-3', complementary to nt1124 to 1143).
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
YW carried out most of the experiments and wrote the manuscript. HF anticipated the inoculation and the construction of D form RNA3s. XW and ML provided useful advice. CH, DL and JY conceived of the study and participated in its design and coordination. All authors read and approved the final manuscript.