Potato Virus Y NIb Multifunctional Protein Suppresses Antiviral Defense by Interacting with Several Protein Components of the RNA Silencing Pathway

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is an interesting manuscript that identifies new activities and interactions of the potato virus Y (PVY) nuclear inclusion protein b (NIb), and is worthy of publication. However, I have identified several points which I believe need to be fixed, or at least clarified, prior to acceptance.

l.2          The title is overly generic, as the demonstrated activities and interactions have been shown only for PVY, and not for potyviruses in general. The KKK motif shown here to be an important nuclear localization signal (NLS) has also been found In some other potyvirus NIb sequences – of which four examples are noted at l.408-410 – but it is not noted whether other potyviruses were found to lack this motif. For example, a random selection of the important soybean mosaic virus reference sequence (NC_002634/NP_072165) reveals KKQ(134-136) in the context KGKKQD(133-138), and plum pox virus (NC_001445/NP_734347) has KKR(135-137) in the context GKKRDK(134-139). In consequence, it might be preferable to change the title to the more specific “Potato virus Y NIb multifunctional protein’, as the KKK NLS appears not to be a generic feature of potyvirus NIb.

l.8          The address for Delaware State is incomplete.

l.22       Is a span of 140 nt really ‘a hotspot’? Apart from covering about 20% of the GFP transcript length, Figure 5 shows a minimum of three distinct peaks.

l.59       It is not correct to state that all RNA viruses except influenza virus replicate in the cytoplasm; a) influenza is not the only member of the Orthomyxoviridae, but b) members of several genera of plant-infecting viruses in the Betarhabdovirinae (Rhabdoviridae) also replicate in the nucleus, including Alphanucleorhabdovirus, Betanucleorhabdovirus, Deltanucleorhabdovirus, Gammanucleovirus, and Dichorhavirus.

l.160 and panel A         Delete the extra ‘A’ from ‘NIb^{KK303/304AAA’}

l.164-167          It is not true that expression from construct NIb^Δ1/17 was almost undetectable, as a clearly visible band is present at the 14 dpi time point in Fig. 3B.

l.174    Note - from personal observation with a variety of GFP fusion proteins, a number of constructs show stronger fluorescence at 72 or even late than at 48 hours of infiltration.

l.186-189          It appears that Figure 4 was Figure 5 at the time the Figure legend was prepared, as the panel A-H descriptions are 5A-5H, rather than 4A-4H. Please correct.

l.198    By ‘early hotspots’, you presumably mean those in the 5’ portion of the transcript, as there is only a single 14 day timepoint in this experiment.

l.199-200         Note that the NIb^Del3x2 peak at ~nt 675 is higher than that for WT NIb – so the statement that values for NIb are higher is incorrect.

l.291    The C-terminal 6xHis tag is not described until the M&M at l.638 – so ‘below’ rather than ‘above’.

l.305-306         More accurately, ‘these complexes showed smaller molecular weight proteins, but no bands corresponding to the native proteins’.

Figure 9 legend.            Please indicate which molecular weight makers are used in this figure. Note that both sets of markers used in Figure 10 (and presumably that used here) are intended for use in SDS-PAGE rather than native protein gels. Notably, the HSP70-related bands seem to be smaller than HSP70 alone, let alone the expected size for the HSP70/NIb-HA complex. For HSP90, the fainter upper band appears more likely to be the HSP90/NIb-HA complex than the more prominent band, if the indicated molecular weight markers can be used to estimate native protein (or protein complex) sizes.

L348-349          If the M1 molecular weight markers in panel B can be used to estimate native protein  sizes, this statement may be correct, as the top two markers of the PageRuler Plus Prestained Protein Ladder are stated by the supplier to be 250 and130 kDa, respectively.

Figure 10           Please add the sizes of the M1 and M2 marker proteins

l.361 and Figure 10    Please note that the bar indicating the lanes representing 10 dpi does not extend far enough to the left to encompass the second ‘NIb’ lane …

l.365    ‘co-expression’, not ‘co-expressed’

l.390-396, Figure 11 legend  If the results shown here are indeed protein complexes separated in Native PAGE, please indicate that clearly in the legend; Figure legends are expected to include sufficient detail that the figure can be understood without reference to the main text. Please also indicate the significance of the solid arrowheads to the right of panels A (AGO4) and D (HSP90); are these symbols used in place of the ‘X’ for artifacts indicated at l.396? If so, how were these features identified as artifacts?

L.398, 407, and 417  As noted at l.2, use of ‘potyvirus’ in the title is too generic when the activities and interactions have been demonstrated only for PVY, and the ‘KKK’ motif only noted to occur in four other potyviruses named in the text – and NOT found in the NIb of two of three other well-known potyviruses selected essentially at random. I therefore suggest that the Discussion should – at least in these beginning paragraphs, qualify ‘NIb’ in these three indicated lines as ‘the PVY NIb’ to recognize that although potentially occurring in multiple other potyviruses, the KKK motif is not a conserved feature of all other potyvirus NIb proteins.

l.457    The Discussion (section 3) skips directly here to section 5, omitting a Section 4.

l.520    The Materials and Methods should therefore be section 5.

l.521    ‘5.1’

Figure 12 and l.560-562          Why is the important newly-identified KKK motif NLS at residues 135-137 not indicated, in addition to the previously-identified N-terminal NLS, and alanine-substituted KK(303/304), GDD and DEEE?   Ideally the figure would be drawn closer to scale, so that the KK(303/304) would be closer to 58% of the length, and KKK(135-137) at about 26% of the length.

l.564    ‘5.2’

l.576    ‘5.3’

l.603    ‘5.5’

l.651-654          Is there any significance to the different font sizes used for the various author initials?

Author Response

We deeply appreciate the amount of time you took to critically review our paper. This has indeed enhanced the quality of our paper. We have highlighted the revisions yellow in the text.

This is an interesting manuscript that identifies new activities and interactions of the potato virus Y (PVY) nuclear inclusion protein b (NIb), and is worthy of publication. However, I have identified several points which I believe need to be fixed, or at least clarified, prior to acceptance.

 

l.2          The title is overly generic, as the demonstrated activities and interactions have been shown only for PVY, and not for potyviruses in general. The KKK motif shown here to be an important nuclear localization signal (NLS) has also been found In some other potyvirus NIb sequences – of which four examples are noted at l.408-410 – but it is not noted whether other potyviruses were found to lack this motif. For example, a random selection of the important soybean mosaic virus reference sequence (NC_002634/NP_072165) reveals KKQ(134-136) in the context KGKKQD(133-138), and plum pox virus (NC_001445/NP_734347) has KKR(135-137) in the context GKKRDK(134-139). In consequence, it might be preferable to change the title to the more specific “Potato virus Y NIb multifunctional protein’, as the KKK NLS appears not to be a generic feature of potyvirus NIb.

Thanks, title has been modified accordingly

 

l.8          The address for Delaware State is incomplete.

Thanks, full address added

 

l.22       Is a span of 140 nt really ‘a hotspot’? Apart from covering about 20% of the GFP transcript length, Figure 5 shows a minimum of three distinct peaks.

We agree that the ~560–700 nt region spans multiple peaks rather than a single narrow peak. We revised the text to describe three reproducible hotspot regions, explicitly noting that the 3′-proximal region (~560–700 nt) contains multiple local maxima (Abstract; Results 2.4; Figure 5 legend).

 

l.59       It is not correct to state that all RNA viruses except influenza virus replicate in the cytoplasm; a) influenza is not the only member of the Orthomyxoviridae, but b) members of several genera of plant-infecting viruses in the Betarhabdovirinae (Rhabdoviridae) also replicate in the nucleus, including Alphanucleorhabdovirus, Betanucleorhabdovirus, Deltanucleorhabdovirus, Gammanucleovirus, and Dichorhavirus.

Thanks, we have changed the sentence to “Most RNA viruses, including potyviruses, replicate in the cytoplasm”

 

l.160 and panel A         Delete the extra ‘A’ from ‘NIb^{KK303/304AAA’}

Thanks, we have deleted the extra A

 

l.164-167          It is not true that expression from construct NIb^Δ1/17 was almost undetectable, as a clearly visible band is present at the 14 dpi time point in Fig. 3B.

Thanks, we have excluded NIb^Δ1/17 from mutants that were barely detectable

 

l.174    Note - from personal observation with a variety of GFP fusion proteins, a number of constructs show stronger fluorescence at 72 or even late than at 48 hours of infiltration.

Thanks for the information. If we are unable to detect expression, we will examine at later times.

l.186-189          It appears that Figure 4 was Figure 5 at the time the Figure legend was prepared, as the panel A-H descriptions are 5A-5H, rather than 4A-4H. Please correct.

Thanks, corrected

 

l.198    By ‘early hotspots’, you presumably mean those in the 5’ portion of the transcript, as there is only a single 14 day timepoint in this experiment.

We clarified the wording to avoid any temporal implication and now refer to the 5′-proximal hotspots (≈175 and ≈320–330 nt) rather than “early hotspots” (Results 2.4).

 

l.199-200         Note that the NIb^Del3x2 peak at ~nt 675 is higher than that for WT NIb – so the statement that values for NIb are higher is incorrect.

We corrected this description. The revised text now states that within the broader ~560–700 nt hotspot region, a prominent local maximum near ≈675 nt is higher in NIbDel3x2 (Results 2.4).

 

l.291    The C-terminal 6xHis tag is not described until the M&M at l.638 – so ‘below’ rather than ‘above’.

Thanks, we changed to described in Materials and Methods.

 

l.305-306         More accurately, ‘these complexes showed smaller molecular weight proteins, but no bands corresponding to the native proteins’.

Modified accordingly.

 

Figure 9 legend.            Please indicate which molecular weight makers are used in this figure. Note that both sets of markers used in Figure 10 (and presumably that used here) are intended for use in SDS-PAGE rather than native protein gels. Notably, the HSP70-related bands seem to be smaller than HSP70 alone, let alone the expected size for the HSP70/NIb-HA complex. For HSP90, the fainter upper band appears more likely to be the HSP90/NIb-HA complex than the more prominent band, if the indicated molecular weight markers can be used to estimate native protein (or protein complex) sizes.

Thanks, the marker is: PageRuler™ Plus Prestained Protein Ladder, 10 to 250 kDa

 

L348-349          If the M1 molecular weight markers in panel B can be used to estimate native protein  sizes, this statement may be correct, as the top two markers of the PageRuler Plus Prestained Protein Ladder are stated by the supplier to be 250 and130 kDa, respectively.

We have added the protein sizes

 

Figure 10           Please add the sizes of the M1 and M2 marker proteins

We have added the protein sizes

 

l.361 and Figure 10    Please note that the bar indicating the lanes representing 10 dpi does not extend far enough to the left to encompass the second ‘NIb’ lane …

Thanks, we fixed it

 

l.365    ‘co-expression’, not ‘co-expressed’

Corrected

 

l.390-396, Figure 11 legend  If the results shown here are indeed protein complexes separated in Native PAGE, please indicate that clearly in the legend; Figure legends are expected to include sufficient detail that the figure can be understood without reference to the main text. Please also indicate the significance of the solid arrowheads to the right of panels A (AGO4) and D (HSP90); are these symbols used in place of the ‘X’ for artifacts indicated at l.396? If so, how were these features identified as artifacts?

We are sorry to indicate that we interchanged (A) AGO4 and (B) HSP70; this error was corrected and the arrowheads explained.

 

L.398, 407, and 417  As noted at l.2, use of ‘potyvirus’ in the title is too generic when the activities and interactions have been demonstrated only for PVY, and the ‘KKK’ motif only noted to occur in four other potyviruses named in the text – and NOT found in the NIb of two of three other well-known potyviruses selected essentially at random. I therefore suggest that the Discussion should – at least in these beginning paragraphs, qualify ‘NIb’ in these three indicated lines as ‘the PVY NIb’ to recognize that although potentially occurring in multiple other potyviruses, the KKK motif is not a conserved feature of all other potyvirus NIb proteins.

We have qualified as PVY encoded NIb

 

l.457    The Discussion (section 3) skips directly here to section 5, omitting a Section 4.

Corrected, thanks

 

l.520    The Materials and Methods should therefore be section 5.

Corrected

 

l.521    ‘5.1’

Corrected

 

Figure 12 and l.560-562          Why is the important newly-identified KKK motif NLS at residues 135-137 not indicated, in addition to the previously-identified N-terminal NLS, and alanine-substituted KK(303/304), GDD and DEEE?   Ideally the figure would be drawn closer to scale, so that the KK(303/304) would be closer to 58% of the length, and KKK(135-137) at about 26% of the length.

We have drawn to scale and added the new NLS

 

l.564    ‘5.2’

Corrected

 

l.576    ‘5.3’

Corrected

 

l.603    ‘5.5’

Corrected

 

l.651-654          Is there any significance to the different font sizes used for the various author initials?

Font size corrected

 

 

Reviewer 2 Report

Comments and Suggestions for Authors

This study investigates the role of the NIB protein encoded by Potato virus Y (PVY) in suppressing RNA silencing. It also examines the interaction of NIB with several proteins involved in the RNA silencing pathway. While the work presents novel findings, some experiments require further clarification, and the presentation of the research flow and data could be improved. Suggestions are provided below.

• The description of NIB mutants, including Figure 12, should be presented from the beginning of result section.
• Figure 3A, how about the blot for 10 dpi?
• Figure 4, subcellular localization observation requires nuclear marker/staining.
• Figure 5, provide the data for total read numbers of GFP siRNA of NIB and NIB mutant samples after normalization with total small RNA reads (read/per million).
• Figure 7 and 8, these data are not essential for this study, can be moved to supplementary data.
• Figure 9, protein interaction should be verified with additional assays such as in vitro pull down, yeast two-hybrid, BIFC assays.
• Figure 11, The data are not conclusive. It is possible the low expression of NIB mutants affects the results of planta interaction assay. For this kind of in planta assay the expression of silencing suppressor (such as p19) is required. Alternatively, other assays such as in vitro pull-down assay may be better.
• Line 59-60, other RNA viruses such as nucleorhabdoviruses also replicate in nucleus.
• Line 73-74, P1 Serine Protease of Cucumber Vein Yellowing Ipomovirus was reported to function as RNA silencing suppressor.
• Line 181-182, some NIB mutants that retain nuclear localization have also lost suppression activities, so at this point it is still premature to conclude that nuclear localization is required for NIb suppression of RNA silencing.
• Line 194, change “mapping” to “mapped”
• Line 210-211, “because antisense strands are preferentially loaded into AGO-containing effector complexes.”. This statement is not correct.
• Figure 9, marked the full size (expected size) of the protein bands and also the smaller bands.

Author Response

We deeply appreciate the amount of time you took to critically review our paper. This has indeed enhanced the quality of our paper. We have highlighted the revisions yellow in the text.

 

This study investigates the role of the NIB protein encoded by Potato virus Y (PVY) in suppressing RNA silencing. It also examines the interaction of NIB with several proteins involved in the RNA silencing pathway. While the work presents novel findings, some experiments require further clarification, and the presentation of the research flow and data could be improved. Suggestions are provided below.

• The description of NIB mutants, including Figure 12, should be presented from the beginning of result section.

Much appreciated, we have included important mutations in the introduction in the paragraph before Results.

 

• Figure 3A, how about the blot for 10 dpi?

We agree that another blot can be added, however, here we wanted to show that GFP is transcribed and therefore loss of GFP is posttranscriptional, which the blot at 7 dpi shows.

 

• Figure 4, subcellular localization observation requires nuclear marker/staining.

We agree that the nucleus could be stained, however, using confocal microscopy we could clearly see the nucleus. We have provided additional images in Supplementary Figure S1, and this includes z-stack imaging showing that HGP localizes in the nuclear membrane but not in the nucleus.

 

• Figure 5, provide the data for total read numbers of GFP siRNA of NIB and NIB mutant samples after normalization with total small RNA reads (read/per million).
  We thank the reviewer for noting the need for explicit totals and replicate clarity. In the revised manuscript we corrected an error in the description of the small RNA sequencing dataset: the analysis includes six libraries total (three biological replicates per treatment), not eighteen as inadvertently stated. All occurrences of the incorrect n have been corrected in the Results, figure legends, and Methods. We added Table S1, reporting per-library totals of GFP-mapping reads and normalization to total adapter-trimmed small-RNA reads (reads per million; RPM), and we reference Table S1 in both the Results text and the Figure 5 legend.

 

• Figure 7 and 8, these data are not essential for this study, can be moved to supplementary data.
  We understand the suggestion and agree these data are supportive rather than the primary endpoint. However, we have retained Figures 7–8 in the main text because they provide an important mechanistic control: viral suppressors can act at distinct steps (e.g., affecting siRNA accumulation/biogenesis versus Argonaute/RISC function), and the NIb‑mapping siRNA profiles help interpret our GFP‑siRNA results in that framework. Also, removing these figures will scramble to whole document, including original images that have been provided to the Editor. We there request to include them in the text.

 

• Figure 9, protein interaction should be verified with additional assays such as in vitro pull down, yeast two-hybrid, BIFC assays.

Yes, indeed, it would have been easier and less labor intensive to use Y2H, and indeed BIFC. However, we quickly got into trouble because the key interactions result in degradation of the complex. We therefore suggest that the immunoblots show us that these proteins interact, especially given that protein complexes isolated using immobilized metal affinity chromatography (IMAC) with Ni-NTA magnetic beads and then detected with both anti-HA and anti-HIS antibodies.

 

• Figure 11, The data are not conclusive. It is possible the low expression of NIB mutants affects the results of planta interaction assay. For this kind of in planta assay the expression of silencing suppressor (such as p19) is required. Alternatively, other assays such as in vitro pull-down assay may be better.

Thanks for the observation. Indeed, over time mutants are silenced because they do not suppress the silencing machinery, unlike wild-type NIb. As shown in Figure 10A, early on there is expression of all constructs. This outcome only confirms that wild-type NIb is a suppressor of the silencing machinery.

 

• Line 59-60, other RNA viruses such as nucleorhabdoviruses also replicate in nucleus.

Absolutely, and thanks for the observation. We have changed the text to to “Most RNA viruses, including potyviruses, replicate in the cytoplasm”

 

• Line 73-74, P1 Serine Protease of Cucumber Vein Yellowing Ipomovirus was reported to function as RNA silencing suppressor.

Much appreciated, we have updated to include P1 of Cucumber Vein Yellowing Ipomovirus 

 

• Line 181-182, some NIB mutants that retain nuclear localization have also lost suppression activities, so at this point it is still premature to conclude that nuclear localization is required for NIb suppression of RNA silencing.

We agree. And note that the mutants that localized in the nucleus and showed “weak” suppression are NIb^{GDD351/353AAA}, and NIb^{DEEE491/494AAA}. This suggests that nuclear localization is not enough, that other factors are in play. We suggest that it is highly likely that these interactions occur in the nucleus or it starts in the nucleus.

 

• Line 194, change “mapping” to “mapped”

Thanks, this was changed

 

• Line 210-211, “because antisense strands are preferentially loaded into AGO-containing effector complexes.”. This statement is not correct.
  We agree and have revised the statement. Argonaute complexes retain one strand of a Dicer-generated duplex as the guide, while the passenger strand is discarded; guide selection depends on duplex properties and AGO sorting preferences rather than a blanket preference for antisense strands. We revised the text accordingly (Results 2.4).

 

• Figure 9, marked the full size (expected size) of the protein bands and also the smaller bands.

Much appreciated, we have added the sizes of the interacting proteins, as well as marked the cleaved products and the uncleaved product.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

no additional data was added