Rab11B is required for binding and entry of recent H3N2, but not H1N1, influenza A isolates

  1. Allyson H. Turner
  2. Sara A. Jaffrani
  3. Hannah C. Kubinski
  4. Deborah P. Ajayi
  5. Matthew B. Owens
  6. Madeline P. McTigue
  7. Conor D. Fanuele
  8. Cailey L. Appenzeller
  9. Hannah W. Despres
  10. Madaline M. Schmidt
  11. Jessica W. Crothers
  12. Emily A. Bruce

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

As an obligate intracellular parasite, influenza A virus (IAV) depends on host proteins to complete several important functions, including trafficking viral proteins throughout the cell. Previous work has established a critical role for the cellular vesicular trafficking protein, Rab11A, in transporting the viral genome segments to the site of budding at the plasma membrane. While the role of Rab11A in IAV assembly is relatively well understood, very little is known about the function of a closely related isoform (Rab11B) during influenza virus infection. We have shown that both Rab11A and Rab11B are required for successful IAV infection by current H1N1 or H3N2 isolates. Cells in which either Rab11A or Rab11B were depleted failed to efficiently produce virus, with significant reductions in infectious titer. Surprisingly, our data reveals that recent (2022) H3N2, but not H1N1, isolates failed to efficiently produce viral proteins in single-cycle infections when Rab11B (but not Rab11A) was depleted. Flow cytometry analysis suggests that the defect in protein production is driven by a reduction in the total number of infected cells, rather than a decrease in viral protein production at the single cell level. RNA-qPCR analysis of H3N2 virions bound to the cell surface showed a ∼50% decrease in virus binding the surface of cells depleted of Rab11B, but not Rab11A. These data suggest a novel role for Rab11B early in IAV infection, likely at the stage of viral binding, that is specific to H3N2 isolates.

Article activity feed

  1. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    Reviewer #1 (Evidence, reproducibility and clarity (Required)):

    • Summary:*

    • In this manuscript, Turner AH. et al. demonstrated the viral replication in cells depleting Rab11B small GTPase, which is a paralogue of Rab11A. It has been reported that Rab11A is responsible for the intracellular transport of viral RNP via recycling endosomes. The authors showed that Rab11B knockdown reduced the viral protein expression and viral titer. This may be caused by reduced attachment of viral particles on Rab11B knockdown cells.*

    • Major comments:*

    • Comment 1 Fig 2-4: The authors should provide Western blot results with equal amount of loading control (GAPDH). The bands shown in these figures lack quantifiability and are not reliable as data.*

    We have rerun these western blots with more equal loading, and included a second loading control (beta-actin) in addition to the GAPDH. These blots can be seen in new Figures 2 and 3, and the quantification against both GAPDH (Figure 2/3) as well as actin (Fig S2) is now included. We have also included additional biological replicates for Fig 2 B-D. These additional experiments have strengthened our conclusion that Rab11B is required for efficient protein production in cells infected with recent H3N2, but not H1N1, isolates.

    Comment 2 Fig 2-4: Why are the results different between Rab11B knockdown alone and Rab11A/B double knockdown? If the authors claims are correct, the results of Rab11B knockdown should be reproducible in Rab11A/B double knockdown cells.

    Prior literature indicates that the Rab11A and Rab11B isoforms can play opposing roles in the trafficking of some cargos (ie, with one isoform transporting a molecule to the cell surface, while the other isoform takes it off again). In this scenario, it is possible that removing both 'halves' of the trafficking loop can ablate a phenotype. However, since our double knockdown used half the amount of siRNA for each isoform (for the same total amount), it is also possible this observation is simply the result of less efficient knockdown. In order to distinguish between these possibilities we depleted Rab11A or Rab11B individually, with this same 'half dose' of siRNA (see new Figure S3). We observed that Rab11B was still robustly required for H3N2 viral protein production. These results suggest that Rab11A and Rab11B could be playing mutually opposing roles in this case, which is consistent with prior Rab11 literature.

    Comment 3 Fig 6: For better understanding, please provide a schematic illustration of experimental setting.

    We have added a new graphical overview to this figure (see new Figure 6A).

    Comment 4: It is necessary to test other siRNA sequences or perform a rescue experiment by expressing an siRNA-resistant clone in the knockdown cells. There seems to be an activation of host defense system, such as IFN pathways.

    In order to rule out the possibility of off-target effects we created a novel cell line that inducibly expresses a Rab11B shRNA sequence (see new Fig 4). This knockdown strategy used a completely different method (shRNA delivered by lentiviral vector vs transient transfection of siRNA), in a different cellular background (H441 "club like" cells vs A549 lung adenocarcinoma). This new depletion strategy showed that the Rab11B dependent H3N2 protein production phenotype is seen across multiple knockdown strategies and cellular backgrounds.

    **Referees cross-commenting**

    I agree with other reviewers' comments in part.

    Reviewer #1 (Significance (Required)):

    The authors propose a novel role for Rab11B in modulating attachment pathway of H3N2 influenza A virus by unknown mechanism. Although previous studies focus on the function of Rab11A on endocytic transport, the function and specificity of Rab11B has remained less clear. The findings may be of interest to a broad audience, including researchers in cell biology, immunology, and host-pathogen interactions. However, the study remains at a superficial level of analysis and does not lead to a deeper understanding of the underlying mechanisms.

    We agree with the reviewer that a strength of this manuscript is its multi-disciplinary nature, particularly with regard to advances in our understanding of Rab11B function. We have added a significant number of experiments and new figures to bolster the rigor and reproducibility of our findings. We have also added a new figure (Fig 7) that uses reverse genetics to map the Rab11B phenotype to the HA gene of the H3N2 isolate under study. By creating '7+1' reassortant viruses with the H3 HA or the N2 NA on a PR8 (H1N1) background (see Fig 7E-H) we were able to demonstrate that Rab11B is acting specifically on one of the HA-mediated entry steps. This provides additional mechanistic insight, by mapping the Rab11B-phenotype to a step at or prior to fusion. Fundamentally, we believe the novelty and rigor of our observation that recent H3N2 viruses enter through a different route than H1N1 isolates is worthy of observation in this updated form, so that the field can begin follow up studies.

    Reviewer #2 (Evidence, reproducibility and clarity (Required)): Summary: The authors compare the effect of RAB11A and RAB11B knockdown on replication of contemporary H1N1 and H3N2 influenza A virus strains in A549 cells (human lung epithelials cells). They find a reduction in viral protein expression for tested H3N2 but not for H1N1 isolates. Mechanistically they suggest that RAB11A affects virion attachment to the cell surface.

    Major comments: The provided data do not conclusively support the suggested mechanism of action and essential controls are missing to substantiate the authors claims: • Knockdown efficacy has to be confirmed on protein level, showing reduced levels of RAB11A and B by Western blot. This is a standard in the field. Off target effects cannot be avoided by RNAi approaches and are usually ruled out by using multiple siRNAs or by complementing the targeted protein in trans.

    We have verified knockdown efficacy at the protein level in new Fig 1A/B. However, due to the high degree of protein level conservation between Rab11A and Rab11B it is very difficult to develop isoform specific antibodies, and we were unable to obtain a Rab11B-specific antibody that can detect endogenous protein (despite testing 6 commercially available antibodies for specificity). Using an antibody that detects both 11A and 11B (Fig1A) we were able to observe very slight changes in the molecular weight of the Rab11 band(s) detected upon knockdown of 11A vs 11B (suggestive of the two isoforms running as a dimer, with Rab11A the lower band and Rab11B the upper band). Cells depleted of both isoforms simultaneously showed a near complete loss of signal. Using a Rab11A antibody (that we confirmed as specific) we were able to observe loss of the Rab11A signal in both the 11A and 11A+B knockdowns (Fig 1B).

    Viral titers should be presented as absolute titers not as % (here the labelling is actually misleading in all graphs indicating pfu/ml)

    This data is now shown in new Figure S1, where it is clear that the trends remain consistent across biological replicates. The axis labels of Fig 1D/E and Fig 3A have been corrected as requested to make clear we are normalizing to account for experiment-to-experiment variation in peak titer.

    Reduction of viral protein expression goes hand in hand with a reduction in GAPDH. While this is accounted for in the quantification a general block of protein expression cannot be ruled out since the stability of house keeper proteins and viral proteins might be different. Testing multiple house keeping proteins could overcome this issue.

    We have included a second loading control (beta-actin) in addition to the GAPDH for new Figure 2 and 3. The quantification of viral protein production compared to beta actin is now included in new Fig S2. We have also included additional biological replicates for Fig 2 B-D. These additional experiments have strengthened our conclusion that Rab11B is required for efficient protein production in cells infected with recent H3N2, but not H1N1, isolates.

    The FACS data in Fig 5 are not convincing. The previous figures showed modest reduction in viral protein expression and the fluorescence is indicated here on a logarithmic scale. Quantification and indication of mean fluorescence intensity from the same data would be a better readout to convincingly show that less cells are infected.

    We have reanalyzed the existing data to quantify the geometric mean of viral protein expression in the infected cell populations (new Figure 5D, E). This analysis shows no significant difference in geometric mean of HA (Fig 5D) or M2 (Fig 5E) expression between cells treated with NT, 11A or 11B siRNA. This additional analysis strengthens our original conclusion that when Rab11B is knocked down, fewer cells get infected, but those that do produce the same level of viral proteins.

    During the time of addition experiment in Fig 6, the authors are testing for HA/M2 positive cells after 16h of infection. This is a multicycle scnario so in a second round they would measure the effect of knockdown in absence of amonium chloride. Shorter infections up to 8h with higher MOI would overcome this problem.

    By maintaining cells in ammonium chloride throughout the infection we are preventing endosomal acidification at any point in the infection period, so this experiment should be measuring solely the effect of one round of infection. The 16 hr timepoint was chosen to allow for optimized staining and analysis of samples by flow cytometry, within the available hours of the flow cytometry facility.

    Standard error of mean is not an appropriate way of representing experimental error for the provided results and should be replaced by SD. Correct labeling of axis with units is required.

    We have updated the axes throughout the manuscript as requested. We have obtained additional statistical expertise (reflected in the updated author list) regarding the issue of SD vs SEM. Standard deviation (SD) would show a measure of the spread of the data, however the full distribution can be clearly seen as we plotted every individual data point. Standard error of the mean (SEM) is a measure of confidence for the mean of the population which takes into account SD and also sample size. SEM is not obvious to estimate by eye in the same way as SD, and we feel is more helpful to the reader to understand how likely the two population means differ from each other on a given graph.

    Minor comments: • The authors show a rescue of viral replication upon double knockdown of RAB11A and B. Maybe this is just a consequence of inefficient knockdown since only half of the siRNAs were used?

    In order to determine if this was the case we depleted Rab11A or Rab11B individually, with this same 'half dose' of siRNA (see new Figure S3). We observed that Rab11B was still robustly required for H3N2 viral protein production. These results suggest that Rab11A and Rab11B could be playing mutually opposing roles in this case (ie, Rab11B transporting a molecule to the surface, while Rab11A recycles it off), which is consistent with prior Rab11 literature.

    Specific experimental issues that are easily addressable. • Are prior studies referenced appropriately? • Are the text and figures clear and accurate? • Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

    Reviewer #2 (Significance (Required)): Significance The authors claim an H3N2 specific dependency on RAB11B for early steps of infection. While this is per se interesting the provided data do not fully support the claims and lack a mechanistic explanation. What is the difference between H1 and H3 strains (virion shape, HA load per virion, attachment force of H1 vs H3). The readouts used are not close enough to the events with regards to timing and could be supported by established entry assays in the field.

    We have provided additional discussion of the differences between H1s and H3s, including sialic acid binding preferences and changes in the HA-sialic acid avidity (lines 76-84). Notably, we have included a new assay (new Fig 7) that provides additional mechanistic insight into the observation that recent H3N2 but not H1N1 isolates depend on Rab11B early in infection. Using reverse genetics we were able to map the Rab11B phenotype to the HA gene of the H3N2 isolate under study. By creating '7+1' reassortant viruses with either the H3 HA or the N2 NA on a PR8 (H1N1) background (see Fig 7E) we are able to demonstrate that Rab11B is acting specifically at one of the HA-mediated entry steps. This excludes several non-HA dependent steps early in the life cycle (uncoating, RNP transport to the nucleus, nuclear import), thus providing additional confirmation that Rab11B acts at one of the earliest steps in the viral life cycle (and by definition, at or prior to fusion). Fundamentally, we believe the novelty and rigor of our observation that recent H3N2 viruses enter through a different route than H1N1 isolates is worthy of observation in this updated form, so that the field can begin follow up studies.

    Reviewer #3 (Evidence, reproducibility and clarity (Required)):

    Manuscript Reference: RC-2025-03007 TITLE: Rab11B is required for binding and entry of recent H3N2, but not H1N1, influenza A isolates Allyson Turner, Sara Jaffrani, Hannah Kubinski, Deborah Ajayi, Matthew Owens, Madeline McTigue, Conor Fanuele, Cailey Appenzeller, Hannah Despres, Madaline Schmidt, Jessica Crothers, and Emily Bruce

    Summary Here, Turner et al. build upon existing knowledge of Influenza A virus (IAV) dependence on the Rab11 family of proteins and provide insights into the specific role of Rab11B isoform in H3N2 virus binding and entry. The introduction is clearly written and provides sufficient background on prior research involving Rab11. It effectively identifies the current gap in knowledge and justifies the investigation of more clinically relevant, circulating strains of IAV. The methods section provides sufficient detail to ensure reproducibility. Similarly, the discussion is well structured, aligns with the introduction, and thoughtfully outlines relevant follow-up experiments. The authors present data from a series of experiments which suggest that the reduced H3N2 infection and viral protein production in Rab11B-depleted cells is due to impaired virus binding. While the evidence supports a Rab11B-specific phenotype in the context of H3N2 infection, we recommend additional experiments (outlined below), to further validate and strengthen these findings. These would help solidify the mechanistic link between Rab11B depletion and the observed phenotype for H3N2 strains of IAV.

    Major comments Figure 1. (B) & (C) The authors normalise viral titers to the non-targeting control (NTC) siRNA set at 100. While this approach allows for relative comparisons, we recommend including the corresponding raw PFU/ml values, at least in the supplementary materials. This will better illustrate the biological significance of gene depletion and variability of the results.

    We have included the raw PFU/mL values in new Figure S1, while peak viral production varied by biological replicate (pasted below, with each biological replicate having a differently shaped data point). While the depletion-induced trends are clearly visible across biological replicates, normalization to average titer in the NT condition for each replicate allows for cleaner visualization.

    In addition, the current protocol uses a high MOI (1), and a relatively short infection period (16 hours) to capture single-cycle replication. However, to better assess the impact of gene knockdown on virus production and spread, we suggest performing a multicycle replication assay using a lower MOI (e.g, 0.01-0.001) over an extended time period, such as 48 hours before titration, provided that cell viability under these conditions is acceptable.

    We appreciate this suggestion and repeatedly attempted to carry out a multicycle growth curve to obtain this data. Unfortunately, out of four independent biological replicates we attempted, we were only able to maintain cell viability and adherence in one biological replicate (shown below). We have not included this data in the revised manuscript due to the limited replicates we were able to obtain, though we can add it in a further revision if the reviewer feels it is warranted.

    Figure 7. (B) & (C) The authors present interesting data showing that siRNA-mediated depletion of Rab11B reduces virion binding of a recently circulating strain of H3N2, but not H1N1, suggesting a subtype-specific role. However, we strongly recommend complementing this assay with a single-cell resolution approach such as immunofluorescence detection of surface-bound viruses through HA staining and image quantification. This would allow the authors to directly assess virion binding per cell and visualise the phenotype, strengthening the mechanistic insight on H3N2 binding in Rab11B-depleted cells. Furthermore, the data, particularly for H1N1 (Figure 7.C), shows substantial variance, which suggests a suboptimal assay sensitivity and limits the strength of the conclusion that the knockdown does not affect H1N1 binding, this limitation may be overcome by implementing the above experimental suggestion.

    We have made substantial efforts to include this data, but were ultimately unable to include this assay due to technical difficulties in implementation (NA stripping caused cells to lift off coverslips, difficulties in antibody sensitivity and specificity, among other issues). We also piloted single cell-based flow cytometry assays to attempt to measure signal from bound virions, but were unable to achieve sufficient differentiation between mock and bound samples with the antibodies we could obtain. However, we have included a new experimental approach that is able to genetically map the 11B-dependent phenotype to the HA gene, thus providing additional mechanistic insight and confirming that Rab11B acts on one of the earliest steps in the viral life cycle (prior to or at fusion).

    Minor comments General The authors should state which statistical test was used for each dataset in the respective figure legends.

    This information is now included in each figure legend.

    Figure 1. Suggest changing Y axis title to PFU/ml [relative to NTC]

    We have changed the axis titles of normalized data to "PFU as % of NT" throughout.

    The co-depletion of Rab11A and Rab11B appears to be less efficient than individual knockdowns, based on RT- qPCR data (Figure 1.A). It is possible that the partial 'rescue' phenotype observed in Figures 2-4 is due to incomplete knockdown, rather than a true biological interaction. This possibility should be acknowledged.

    In order to distinguish between a partial 'rescue' and inefficient knockdown, we depleted Rab11A or Rab11B individually, with the same 'half dose' of siRNA used in the double knockdown (see new Figure S3). We observed that Rab11B was still robustly required for H3N2 viral protein production. These results suggest that Rab11A and Rab11B could be playing mutually opposing roles in this case, which is consistent with prior Rab11 literature, rather than simply inefficient knockdown.

    Furthermore, knockdown efficiency is assessed only at the mRNA level. To strengthen the conclusions, the authors are encouraged to provide western blot data confirming protein-level depletion of Rab11A and Rab11B, particularly in the double knockdown condition. This would help clarify whether co-transfection of siRNAs affect the efficiency of each individual knockdown at the protein level.

    We have verified knockdown efficacy at the protein level in new Fig 1A/B. However, due to the high degree of protein level conservation between Rab11A and Rab11B it is very difficult to develop isoform specific antibodies, and we were unable to obtain a Rab11B-specific antibody that can detect endogenous protein (despite testing 6 commercially available antibodies for specificity). Using an antibody that detects both 11A and 11B (Fig1A) we were able to observe very slight changes in the molecular weight of the Rab11 band(s) detected upon knockdown of 11A vs 11B (suggestive of the two isoforms running as a dimer, with Rab11A the lower band and Rab11B the upper band). Cells depleted of both isoforms simultaneously showed a near complete loss of signal. Using a Rab11A antibody (that we confirmed as specific) we were able to observe loss of the Rab11A signal in both the 11A and 11A+B knockdowns (Fig 1B).

    Figure 6. (A) & (B) are missing error bars, particularly the Rab11B knockdown data points.

    Error bars are plotted in each graph, but due to very limited experimental variation these error bars are too small to appear on the graph (11B points in Fig 6B, D).

    Figure 7. If including any repeats in the binding assay, authors are encouraged to use appropriate controls in each experiment such as exogenous neuraminidase treatment or sialidase treatment.

    When attempting to establish a microscopy based binding assay we included exogenous neuraminidase in each experiment. Unfortunately, the combination of glass coverslips and treatment with exogenous neuraminidase at incubation times sufficient to strip virus also removed cells from the coverslips.

    Reviewer #3 (Significance (Required)):

    General assessment: Provides a conceptual advancement of subtype specific receptor preferences.

    Advance: The study raises interesting observations regarding influenza virus subtype differences in cell surface receptor binding, in a Rab11B-dependent manner.

    Audience: Influenza virologists, respiratory virologists

    Expertise: Virus entry, Virus cell biology

  2. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    Title: Rab11B is required for binding and entry of recent H3N2, but not H1N1, influenza A isolates

    Allyson Turner, Sara Jaffrani, Hannah Kubinski, Deborah Ajayi, Matthew Owens, Madeline McTigue, Conor Fanuele, Cailey Appenzeller, Hannah Despres, Madaline Schmidt, Jessica Crothers, and Emily Bruce

    Summary

    Here, Turner et al. build upon existing knowledge of Influenza A virus (IAV) dependence on the Rab11 family of proteins and provide insights into the specific role of Rab11B isoform in H3N2 virus binding and entry. The introduction is clearly written and provides sufficient background on prior research involving Rab11. It effectively identifies the current gap in knowledge and justifies the investigation of more clinically relevant, circulating strains of IAV. The methods section provides sufficient detail to ensure reproducibility. Similarly, the discussion is well structured, aligns with the introduction, and thoughtfully outlines relevant follow-up experiments. The authors present data from a series of experiments which suggest that the reduced H3N2 infection and viral protein production in Rab11B-depleted cells is due to impaired virus binding. While the evidence supports a Rab11B-specific phenotype in the context of H3N2 infection, we recommend additional experiments (outlined below), to further validate and strengthen these findings. These would help solidify the mechanistic link between Rab11B depletion and the observed phenotype for H3N2 strains of IAV.

    Major comments

    Figure 1. (B) & (C)

    The authors normalise viral titers to the non-targeting control (NTC) siRNA set at 100. While this approach allows for relative comparisons, we recommend including the corresponding raw PFU/ml values, at least in the supplementary materials. This will better illustrate the biological significance of gene depletion and variability of the results. In addition, the current protocol uses a high MOI (1), and a relatively short infection period (16 hours) to capture single-cycle replication. However, to better assess the impact of gene knockdown on virus production and spread, we suggest performing a multicycle replication assay using a lower MOI (e.g, 0.01-0.001) over an extended time period, such as 48 hours before titration, provided that cell viability under these conditions is acceptable.

    Figure 7. (B) & (C)

    The authors present interesting data showing that siRNA-mediated depletion of Rab11B reduces virion binding of a recently circulating strain of H3N2, but not H1N1, suggesting a subtype-specific role. However, we strongly recommend complementing this assay with a single-cell resolution approach such as immunofluorescence detection of surface-bound viruses through HA staining and image quantification. This would allow the authors to directly assess virion binding per cell and visualise the phenotype, strengthening the mechanistic insight on H3N2 binding in Rab11B-depleted cells. Furthermore, the data, particularly for H1N1 (Figure 7.C), shows substantial variance, which suggests a suboptimal assay sensitivity and limits the strength of the conclusion that the knockdown does not affect H1N1 binding, this limitation may be overcome by implementing the above experimental suggestion.

    Minor comments

    General

    The authors should state which statistical test was used for each dataset in the respective figure legends.

    Figure 1.

    Suggest changing Y axis title to PFU/ml [relative to NTC] The co-depletion of Rab11A and Rab11B appears to be less efficient than individual knockdowns, based on RT- qPCR data (Figure 1.A). It is possible that the partial 'rescue' phenotype observed in Figures 2-4 is due to incomplete knockdown, rather than a true biological interaction. This possibility should be acknowledged. Furthermore, knockdown efficiency is assessed only at the mRNA level. To strengthen the conclusions, the authors are encouraged to provide western blot data confirming protein-level depletion of Rab11A and Rab11B, particularly in the double knockdown condition. This would help clarify whether co-transfection of siRNAs affect the efficiency of each individual knockdown at the protein level.

    Figure 6.

    (A) & (B) are missing error bars, particularly the Rab11B knockdown data points.

    Figure 7.

    If including any repeats in the binding assay, authors are encouraged to use appropriate controls in each experiment such as exogenous neuraminidase treatment or sialidase treatment.

    Significance

    General assessment: Provides a conceptual advancement of subtype specific receptor preferences.

    Advance: The study raises interesting observations regarding influenza virus subtype differences in cell surface receptor binding, in a Rab11B-dependent manner.

    Audience: Influenza virologists, respiratory virologists

    Expertise: Virus entry, Virus cell biology

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    Summary:

    The authors compare the effect of RAB11A and RAB11B knockdown on replication of contemporary H1N1 and H3N2 influenza A virus strains in A549 cells (human lung epithelials cells). They find a reduction in viral protein expression for tested H3N2 but not for H1N1 isolates. Mechanistically they suggest that RAB11A affects virion attachment to the cell surface.

    Major comments:

    The provided data do not conclusively support the suggested mechanism of action and essential controls are missing to substantiate the authors claims:

    • Knockdown efficacy has to be confirmed on protein level, showing reduced levels of RAB11A and B by Western blot. This is a standard in the field. Off target effects cannot be avoided by RNAi approaches and are usually ruled out by using multiple siRNAs or by complementing the targeted protein in trans.
    • Viral titers should be presented as absolute titers not as % (here the labelling is actually misleading in all graphs indicating pfu/ml)
    • Reduction of viral protein expression goes hand in hand with a reduction in GAPDH. While this is accounted for in the quantification a general block of protein expression cannot be ruled out since the stability of house keeper proteins and viral proteins might be different. Testing multiple house keeping proteins could overcome this issue.
    • The FACS data in Fig 5 are not convincing. The previous figures showed modest reduction in viral protein expression and the fluorescence is indicated here on a logarithmic scale. Quantification and indication of mean fluorescence intensity from the same data would be a better readout to convincingly show that less cells are infected.
    • During the time of addition experiment in Fig 6, the authors are testing for HA/M2 positive cells after 16h of infection. This is a multicycle scnario so in a second round they would measure the effect of knockdown in absence of amonium chloride. Shorter infections up to 8h with higher MOI would overcome this problem.
    • Standard error of mean is not an appropriate way of representing experimental error for the provided results and should be replaced by SD. Correct labeling of axis with units is required.

    Minor comments:

    • The authors show a rescue of viral replication upon double knockdown of RAB11A and B. Maybe this is just a consequence of inefficient knockdown since only half of the siRNAs were used?
    • Specific experimental issues that are easily addressable.
    • Are prior studies referenced appropriately?
    • Are the text and figures clear and accurate?
    • Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

    Significance

    The authors claim an H3N2 specific dependency on RAB11B for early steps of infection. While this is per se interesting the provided data do not fully support the claims and lack a mechanistic explanation. What is the difference between H1 and H3 strains (virion shape, HA load per virion, attachment force of H1 vs H3). The readouts used are not close enough to the events with regards to timing and could be supported by established entry assays in the field.

  4. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    Summary:

    In this manuscript, Turner AH. et al. demonstrated the viral replication in cells depleting Rab11B small GTPase, which is a paralogue of Rab11A. It has been reported that Rab11A is responsible for the intracellular transport of viral RNP via recycling endosomes. The authors showed that Rab11B knockdown reduced the viral protein expression and viral titer. This may be caused by reduced attachment of viral particles on Rab11B knockdown cells.

    Major comments:

    Comment 1 Fig 2-4: The authors should provide Western blot results with equal amount of loading control (GAPDH). The bands shown in these figures lack quantifiability and are not reliable as data.

    Comment 2 Fig 2-4: Why are the results different between Rab11B knockdown alone and Rab11A/B double knockdown? If the authors claims are correct, the results of Rab11B knockdown should be reproducible in Rab11A/B double knockdown cells.

    Comment 3 Fig 6: For better understanding, please provide a schematic illustration of experimental setting.

    Comment 4: It is necessary to test other siRNA sequences or perform a rescue experiment by expressing an siRNA-resistant clone in the knockdown cells. There seems to be an activation of host defense system, such as IFN pathways.

    Referees cross-commenting

    I agree with other reviewers' comments in part.

    Significance

    The authors propose a novel role for Rab11B in modulating attachment pathway of H3N2 influenza A virus by unknown mechanism. Although previous studies focus on the function of Rab11A on endocytic transport, the function and specificity of Rab11B has remained less clear. The findings may be of interest to a broad audience, including researchers in cell biology, immunology, and host-pathogen interactions. However, the study remains at a superficial level of analysis and does not lead to a deeper understanding of the underlying mechanisms.

  5. Version published to 10.1101/2025.04.23.650275 on bioRxiv