Fig 1.
Replication proteins from Fny-CMV, but not LS-CMV, efficiently support the replication of satRNAs in trans-replication assays.
(A) Northern blotting analyses of the accumulation of CMV RNAs and sat-T1 in the infiltrated leaves. (B) Northern blotting analyses of the accumulation of RNA3 and its subgenomic RNA4. In the trans-replication assays, RNA3 from Fny-CMV (F3) or LS-CMV (L3) was replicated by the replicase of both CMV strains (F1a and F2a, L1a and L2a), in combination with the RNA silencing suppressor P19 from tomato bushy stunt virus. (C, D) Northern blotting analyses of the accumulation of sat-T1 (C) and seven wild-type satRNA strains (D). Sat-T1 and the other six satRNAs were replicated with the replication proteins of Fny-CMV or LS-CMV, along with P19. (E) Northern blotting analyses of viral RNAs and satRNAs in the co-infected plants. Seven satRNA strains were separately co-inoculated with LS-CMV on Nicotiana benthamiana plants via agroinfiltration. Both the trans-replication assays (A-D) and virus inoculation (E) were conducted by infiltrating the 5th true leaves of N. benthamiana plants with Agrobacterium cells. Total RNAs were extracted from the infiltrated leaves at 3 days post-agroinfiltration (DAPI) (A-D) or from the upper systemic leaves at 6 DAPI (E), followed by northern blot hybridization using the CMV- or satRNA-specific oligonucleotide probe. Mock plants were treated with the infiltration solution. Vector controls in (A, C-E) refer as to these samples infiltrated with Agrobacterium cells carrying the vector pCB301, in combination with sat-T1 (A and C), viral replication proteins (D) or LS-CMV (E). Asterisk (*) denotes the non-specific bands shown after over-exposure in (A) and (C). Ethidium bromide-stained ribosomal RNAs were used to assess the relative loading amounts of each RNA sample.
Fig 2.
Viral RNA(s) were essential for LS replication proteins to support satRNA replication, independent of translation products.
(A) Schematic diagrams of the non-coding genomic RNAs of LS-CMV or TAV. All of the mutants were generated using one-step site-directed mutagenesis, with substituted nucleotides highlighted in red. The untranslated regions corresponding to each open reading frame in these mutants are delineated by black dashed lines. (B-E) Northern blotting analyses of the accumulation of sat-T1, viral RNAs or their mutants in the trans-replication assays expressing LS replication proteins (L1a and L2a). In the trans-replication assays, sat-T1 was co-expressed with RNA1, RNA2Δ2b (lacking 2b expression) and RNA3 of LS-CMV (B) and its non-coding RNAs (ncL1, ncL2 and ncL3) (C), as well as non-coding RNAs of Fny-CMV (ncF1, ncF2, ncF3) (D) or TAV (ncT1, ncT2, ncT3) (E). The trans-replication assays were conducted in the 5th true leaves of Nicotiana benthamiana plants. Mock plants were treated with the infiltration solution. Vector controls refer as to these samples infiltrated with Agrobacterium cells carrying the vector pCB301. Total RNAs were extracted from the infiltrated leaves at 3 days post-infiltration, and subjected to northern blotting hybridization for detection of viral RNAs, their mutants, and positive-sense and negative-sense RNAs of sat-T1. The oligonucleotide probe targeting to the conserved sequence in the 3′ UTR of all CMV strains were used to detect the RNA mutants from LS-CMV (B and C), and RNAs 3 & 4 of Fny-CMV and TAV (D-E). RNAs 1 & 2 of Fny-CMV were detected using both RNA-specific oligonucleotide probe (D). The band intensity corresponding to sat-T1 in each sample was arbitrarily quantified, and the relative accumulation levels were presented below. Ethidium bromide-stained ribosomal RNAs were used to assess the relative loading amounts of the RNA samples.
Fig 3.
Replicable viral RNAs enhance satRNA replication in trans-replication assays.
Seven satRNA strains were separately expressed alone or co-expressed with the replication proteins from LS-CMV (L1a + L2a) or Fny-CMV (F1a + F2a), together with or without ncL3 (A) or ncF3 (B) in the 5th true leaves of Nicotiana benthamiana plants. The RNA silencing suppressor P19 was included in all the treatments. At 3 days post-agroinfiltration, total RNAs were extracted from the infiltrated leaves and subjected to northern blot hybridization for measuring the RNA3 mutants and both positive-sense and negative sense strands of these satRNA strains. Mock plants were treated with infiltration solution alone, while vector controls denote these samples infiltrated with Agrobacterium cells carrying the vector pCB301. The ratios of satRNA accumulation in the presence of ncF3 (+) relative to its absence (-) are shown below. Ethidium bromide-stained ribosomal RNAs served as the loading control.
Fig 4.
The TLS element is essential for viral RNAs to enhance satRNA replication in trans-replication assays.
(A) Schematic diagrams of L3, ncL3 and their derivatives. The 3′ UTR of CMV RNAs is divided into three regions: a variable region (VR) at the 5′ end, a conserved TLS at the 3′ end, and a highly conserved region (CR) separating them. Deleted sequences in the constructed mutants were indicated by dashed lines. The TLS in ncL3 was substituted with the TLS of BMV, PSV, TAV, TMV, or TYMV, to generate six chimeric ncL3 mutants. (B-D, F) Northern blotting analyses of the accumulation of sat-T1, L3 and its mutants in the 5th true leaves of Nicotiana benthamiana transiently expressing LS replication proteins (L1a+L2a) and the RNA silencing suppressor P19. The mutants of L3 or ncL3 tested in these experiments are shown above. (E) Northern blotting analyses of TLS, RNA4 (L4) and its noncoding version (ncL4) of L3, as well as both polarities of sat-T1 in the trans-replication assays. In this replication assay, the TLS, L4 and ncL4 were provided separately in trans via agroinfiltration. It is worth mentioning that the probe used to detect RNA3 and its subgenomic RNA4 in (C) & (D) is the digoxin-labeled oligonucleotide complementary to the sequence spanning from nt 1200 to 1333 of L3. The digoxin-labeled oligonucleotide probe used to detect these RNAs in (E) is complementary to the sequence positioning at nt 2128–2167 of L3. Mock plants were treated by infiltration solution alone. The relative accumulation levels of RNA3, and positive-sense or negative-sense RNA of sat-T1 in (B-D, F) are shown below. Ethidium bromide-stained ribosomal RNAs served as the loading control.
Fig 5.
Complete replication of viral RNAs is required for enhancing satRNA replication in trans-replication assays.
(A) Schematic diagrams of ncL3 and its derivative ncL3-ΔVRΔCRΔ5U. The green blocks denote the substitution of the 5′ UTR with the sequence of M13R(-48), and the orange blocks denote the deletion of both VR and CR in the 3′ UTR. (B) Northern blotting analyses of the accumulation of sat-T1, ncL3 and its derivatives ncL3-ΔVRΔCR and ncL3-ΔVRΔCRΔ5U. In the trans-replication assay expressing LS replication proteins (L1a+L2a), sat-T1 was co-expressed with ncL3 or its derivatives, as well as the vector (pCB301) control in the 5th true leaves of Nicotiana benthamiana plants. At 3 days post-infiltration, total RNAs were extracted from the infiltrated leaves and subjected to northern blot hybridization. The relative accumulation levels of RNA3, and positive-sense or negative sense RNAs of sat-T1 are shown below. Ethidium bromide-stained ribosomal RNAs served as the loading control. (C) Strand-specific RT-PCR for detecting the negative-sense RNAs of ncL3 and its derivative in the RNA samples shown in panel (B). RNA samples were digested with TURBO DNase to remove DNA, and subjected to RT reactions with the primer ncL3a-F. The RT products were amplified using the primer pair ncL3a-F and ncL3b-R. In parallel, total RNAs were amplified directly in PCR reactions without the RT process, indicated with RT(-). PCR products were separated in a 1% agarose gel and observed under UV light after ethidium bromide staining. “M” denotes the DL2000 DNA ladder.
Fig 6.
The recruitment element Box-B is indispensable for viral RNAs to enhance satRNA replication.
(A) Schematic diagrams of L3 and its derivatives. The sequence from positions 47–96 nt in the 5′ UTR was removed from L3, to create L3-Δ5UII, as shown in the rectangle with green dashed lines. The Box-B motif in the IGR was removed from L3 to generate the mutant L3-ΔBoxB, as shown in the rectangle with blue dashed lines. The stem-loop C in the TLS was mutated with the nucleotides (AGUCAC) to create the mutant L3-mSLC, as shown in the rectangle with brown dashed lines. (B) Northern blotting analyses of the accumulation of RNA3 and sat-T1 in the presence of LS replication proteins (L1a+L2a). In the trans-replication assays, sat-T1 was co-expressed with L3, the aforementioned mutants, or the vector pCB301 in the 5th true leaves of Nicotiana benthamiana plants. At 3 days post-infiltration, total RNAs were extracted from the infiltrated leaves and subjected to northern blot hybridization. Mock plants were treated by infiltration solution alone. Ethidium bromide-stained ribosomal RNAs were used to ensure the equal loading for all RNA samples. The relative accumulation levels of sat-T1 or RNA3 are depicted in the graph on the right. Mean values with standard errors were calculated from three independent biological experiments.
Fig 7.
Both replicase components of LS-CMV are responsible for the defect in supporting satRNA replication.
(A-C) Northern blotting analyses of the accumulation of L3 (A), sat-T1 (B), or both together (C). L3, sat-T1, or both were co-expressed with the replication proteins of Fny or LS, or their heterologous recombinants (F1a + L2a, L1a + F2a) in the 5th true leaves of Nicotiana benthamiana plants. Co-expression of sat-T1 with F1a or F2a served as negative controls in (B). At 3 days post-infiltration, total RNAs were extracted from the infiltrated leaves and subjected to northern blot hybridization. Mock plants were treated by infiltration solution alone. The relative accumulation levels of L3, and positive-sense or negative sense RNAs of sat-T1 were presented below. Ethidium bromide-stained ribosomal RNAs were used to assess the relative loading amounts of the RNA samples.
Fig 8.
The replication proteins of Fny-CMV and LS-CMV exhibit significant differences in their ability to recruit positive-sense RNA of sat-T1.
(A) The hairpin structures containing the Box-B sequence in blue and the mutated Box-B (mBox-B) in red. mF3 denotes the F3 mutant, in which the Box-B was substituted with mBox-B. mF3-T1(+) and mF3-T1(-) are mF3 derivatives with positive-sense or negative-sense RNA of sat-T1 inserted between the CP and 3′ UTR in mF3, respectively. A fragment of the GUS gene (337 nt), equivalent in size of sat-T1, was introduced into F3 or mF3, resulting to the creation of F3-gus or mF3-gus, respectively. (B) The accumulation levels of F3 and mF3 in the trans-replication assay. Either F3 or mF3 was co-expressed with LS replication proteins and P19 in the 5th true leaves of N. benthamiana plants. Mock plants were treated with infiltration solution. At 3 days post-infiltration, total RNAs were extracted separately from three infiltrated leaves for each treatment and subjected to northern blotting analyses. The relative accumulation levels of F3 and mF3 are shown below as the mean values with standard errors from three independent biological samples. (C) Determination of the replication activities of mF3-T1(+), mF3-T1(-), and the controls F3-gus and mF3-gus. These four F3 derivatives was co-expressed with the P19 suppressor and the replication proteins of LS-CMV (upper panel) or Fny-CMV (the lower panel) in the 5th true leaves of Nicotiana benthamiana plants. Mock plants were treated with infiltration solution. At 3 days post-infiltration, total RNAs were extracted separately from three infiltrated leaves for each treatment, and analyzed by northern blot hybridization. The relative accumulation levels with standard errors shown below were calculated from three independent biological samples. “UD” denotes the undetectable level. Ethidium bromide-stained ribosomal RNAs were used as a loading control for normalization of the relative accumulation levels.
Fig 9.
The 1a protein of Fny-CMV, but not LS-CMV, displays co-localization in nuclei with MCP-YFPnls.
(A) Schematic diagrams of the DNA constructs used in the developed method for analyzing the interaction of CMV 1a with sat-T1. The bacteriophage MS2 stem-loop structure containing the binding site of MS2 CP (MCP), as shown in the rectangle with red dashed lines. Six copies of MS2 stem-loop (6×MS2) were linked at the 5′ end of sat-T1 or an equal-size fragment of the Gus gene, to create 6×MS2-satT1 and 6×MS2-Gus, respectively. The coding sequence of MCP was fused with YFP, followed by a nuclear localization signal, generating MCP-YFPnls. F1a and L1a were tagged with a copy of mCherry at their C terminus. (B) The subcellular localization of mCherry-tagged 1a proteins and YFPnls-tagged MCP. These proteins were individually expressed with p19 in the leaves of Nicotiana benthamiana plants, and subjected to fluorescence visualization using a laser confocal microscopy at 2 days post-agroinfiltration (DAPI). (C) Subcellular distributions of F1a-mCherry and L1a-mCherry when co-expressed with MCP-YFPnls and either 6×MS2-satT1 or 6×MS2-Gus. The 6th true leaves of 3-weeks old N. benthamiana plants were infiltrated with the mixture of Agrobacterium cells to express the fusion proteins and RNAs as indicated. Fluorescence was visualized at 2 DAPI. The green color represents the fluorescence signal omitted from the fusion protein MCP-YFPnls, and the red color indicates the signals from either F1a-mCherry or L1a-mCherry. Two arrows in yellow indicate the nuclear accumulation of F1a-mCherry, when co-expressed with MCP-YFPnls and 6×MS2-satT1, but not with 6×MS2-Gus. The scale bars denote 20 μm.
Fig 10.
A proposed model of satRNA replication stimulated by viral RNAs in plants.
During the initial replication stage, CMV replication proteins localize to tonoplast and remodel it to create viral replication organelles (VROs). This process involves the recruitment of satRNAs, viral RNAs, or both to assemble viral replication complexes (VRCs). Notably, VROs could be free of both CMV and satellite RNAs, as viral replication proteins themselves can form VRO-like spherules, as reported previously [38]. VRCs assembled with satRNAs exhibit lower replication activity (indicated by VROs enclosed in green circles), whereas those formed with viral RNAs exhibit high activity (indicated by VROs enclosed in red circles). These highly active VRCs replicate not only viral RNAs, but also satRNAs when satRNAs are recruited alongside viral RNAs into the same VROs. However, considering that LS replication proteins have limited capability for satRNA recruitment, some VROs may lack satRNAs (indicated by VROs enclosed in gray circles) in the absence of viral RNAs. Consequently, VRCs formed in the presence of satRNAs alone would produce less satRNAs compared to those formed in the presence of both viral and satellite RNAs. Following the initial replication, more satRNAs, along with viral RNAs produced during the initial replication, participate in the creation of new VROs. This is expected to contribute to the enhanced proliferation of satRNAs with the assistance of viral RNAs.