Referee 2's review (Jorg H. Fritz):

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication the manuscript has been revised in light of these comments and to address other editorial requirements.
**********

Review of
"Anti-HIV activity mediated by Natural Killer and CD8+ cells after Toll-like receptor 7/8 triggering" by Erika Schlaepfer and Roberto F. Speck

The manuscript by Erika Schlaepfer and Roberto F. Speck entitled ""Anti-HIV activity mediated by Natural Killer and CD8+ cells after Toll-like receptor 7/8 triggering" examines the activity of TLR 7/8 agonists for the induction of anti-HIV activity. The authors demonstrate clearly that depletion of dendritic cells (DC), NK (natural killer) and CD8+ cells from PBMC results in the reversal of TLR7/8 induced anti-HIV effects, while depletion of either NK or CD8+ cells alone was only partially effective. Although the authors demonstrate a critical role of soluble factors induced upon TLR7/8 triggering for the mediation of anti-HIV activity, they fail to pin-down a single soluble factor responsible for the potent effects.

Although the dataset presented by the authors is of interest, the manuscript is written in a poor uninspiring way. The authors fail to outline the novelty of their dataset, nor do the place it within the proper context of published literature. The presented data is reported to be novel, whilst many reports have outlined the anti-HIV activity upon TLR7/8 stimulation. The importance of TLR7/8 stimulation for anti-HIV activity in vivo has been recently shown by Wille-Reece et al (JEM, 203(5), 1249-59; PNAS, 102(42) 15190-94), demonstrating that TLR7/8 agonists influence the magnitude and quality of anti-HIV CD8+ T cell responses. It has been shown already that DC-derived factors, as well as DC-NK cell contacts and/or accessory cell-contacts are required for full-blown anti-HIV activity upon TLR7/8 stimulation (Alter G. Et al., JI, 178:7658-66; Gorski K.S. et al., Int. Immunol, 18:1115-26). Thus, it is not surprising or novel that depletion of DCs and NK cells reduces the TLR7/8 induced HIV-activity. Nevertheless, the author's finding that depletion of both NK cells and CD8+ T cells markedly reduces these effects is of great interest. It is disappointing that the author's neither emphasize this by discussing putative mechanisms of action, nor do they address the resulting open questions experimentally. I recommend that these aspects need to be considered after major revision of the manuscript for a putative publication in PLOSOne.


Specific comments:
Results & Figure1:
Figure1 is insufficiently labelled. It would be helpful for the reader to add a legend of the three different TLR agonists (names & TLR usage), and indicate in Figure 1A & B which HIV strain has been tested, as well as in Figure 1D & E which cell type(s) (tonsilar tissue-derived cells (TTDCs) or PBMCs, respectively) have been stimulated. The preparation of tonsilar tissue-derived cells is lacking in the Materials & Methods section. It would be of utmost importance to describe the procedure in more detail. Moreover, the authors should refer throughout the text that they stimulated tonsilar tissue-derived cells and not whole tonsilar tissue sections itself. How many tonsilar tissue-derived cells have been used for stimulation with TLR agonists and HIV infection assays? In the light of the author's observation that differences between PBMCs and TTDCs exist in their readouts, the lymphocyte and leukocyte composition of the TTDC preparation would be of interested to report (FACS analysis). Are the observed differences between PBMCs and TTDCs due to the percentage of DCs? They authors should report and document on that. Moreover, the methodology of the used WST-1 assay is not described. The percentage of early and late apoptotic cells before and after TLR triggering and HIV infection should be analyzed by FACS analysis using propidiumiodide and AnnexinV staining.
In the first sentence of the results section it would be helpful to explain the experimental setup in a more detailed way and explain what kind of working concentration was optimized. I also suggest to include the data not shown into the manuscript or remove it from the results section and put this into the discussion. The interpretative part in the legend to Figure1 ("R848 revealed.........") should be included in the result section.

Results, Figure2 & Table1:
The authors should indicate here the specific cell types expressing TLR7 and TLR8 present in total PBMCs and cite the literature accordingly. Although the authors indicate clearly that depletion of DCs almost completely abolishes the production of TLR7- and TLR8-induced cytokine and interferon production, they have not formally proven (e.g. by intracellular FACS analysis) that DCs are the main source of these cytokines. Intercellular contacts between DCs and other cell types could still be of great importance. This aspect needs to be carefully outlined here. It would be interesting to report if TLR stimulation of moMDCs induces the same set of cytokines than stimulation of whole PBMCs? This would be important to add to the dataset shown here. Moreover, the authors should indicate if and how they tested moMDCs for DC purity (e.g. CD14/CD1a staing) and should indicate the source of GM-CSF and IL-4 in the Material & Method section. Moreover, the method of cytokine measurement of ALL cytokines, interferons and chemokines should be described properly. What was the duration of PBMC-stimulation (Table1)?

Figure3&4:
The applied measurements should be properly described in the results section AND the legend. Moreover, the authors should specify clearly also in the result section which cell type they depleted by CD56 beads. Introductory statements like "B cells are implicated in the anti-HIV effects" are detrimental to the clarity of the shown data and should be discussed with sufficient citation within the discussion. In addition it should be indicated what the expression of CD107a stands for as well as an explanation of how the cytolytic activity was assessed. How was the level of expression (inhibitory and activating receptors) analyzed? mRNA or protein level? This information as well as the data should be included! The whole paragraph starting with "Furthermore, the activating NK receptor, ....." needs to be explained why these experiments have been done AND what they indicate. Moreover, the data should be included in the manuscript!

The reported differences of NKG2D and TRAIL and the observation of that TRAIL blocking antibodies do not reverse the anti-HIV effects need to be discussed, since it has been shown that HIV turn pDCs into TRAIL-expressing killer pDCs, thereby down-regulating HIV coreceptors upon TLR7-induced IFN- production (Hardy AW et al., PNAS, 104(44), 17453-58). Since the authors do not see that TRAIL-blocking antibodies reverse TLR7/8 induced anti-HIV effects, wouldn't the authors data suggest that TRAIL-induction on pDCs by IFN-a is not directly relevant and sufficient for anti-HIV activity - like suggested by Hardy AW et al.?

Figure5, Table2&3:
Again, the rational as well as the experimental procedures of Figure 5 are insufficiently described. Contradictory procedures are described in the Material & Method section and the Figure legend. Were PBMCs now depleted for DCs and DC precursors, like described in the legend, or only depleted for CD4+ T cells, like described in the Material & Methods section? The authors need to clarify in detail: rationale, experimental procedure and describe their results properly; all this is lacking. What would depletion of DCs in this assay add to the dataset? Moreover, by depletion of CD4+ T cells from PBMCs from the Transwell-input population, how can the authors be sure that CD4+ T cells are not required in transmitting a DC-derived signal after TLR stimulation to CD8+ T cells and/or NK cells? These aspects need to be considered when describing and interpreting these sets of experiments.

Is Table2 an extended investigation of the soluble mediators released after TLR stimulation of PBMCs? If so it would be best to fuse these into one Table! Discrepancies in the levels of IL-12 exist between Table1 and Table2. Is this because of the duration of stimulation or is IL-12 meant to be IL-12p70? That needs to be clarified! Moreover, all these issues need to be outlined in the Result section! Moreover, it is not clear how many donors were tested here; the clarity of Table2 needs to be improved!
Again, Table3 is insufficiently labelled, and the experimental procedures are poorly described. In light of the fact that the levels of IFN- induced by the three compounds are within a comparable range, how do the authors explain that discrepancies between the three different TLR compounds and the IFN-R blocking data? It would be of utmost importance to measure cytokine-chemokine-interferon production from the same PBMCs with which the antibody-blocking experiments have been done. The slight effects of blocking, IL-12/IL-15/IL-18 has to be discussed. Another important issue which needs to be clarified; how did the author ensure that their blocking antibodies were efficient?

Figure6:
The description of results placed within the Figure legend needs to be moved into the result section. The Figure should be labelled in more detail and the experimental procedure explained properly.

Materials & Methods
.) The detailed culture conditions for MoMDC should be specified; (how many cells/well/volume; have cells been fed only once,.........)
.) The purity of TLR agonists needs to be specified, since synergistic effects of unclean preparations have been described.
.) Blocking antibodies: clones for IFN-/R, MIP3 and SDF-1 need to be specified.
.) The source and culture conditions of 293T, K562 and Hela cells needs to be indicated.
.) Flow cytometry antibodies; the clones used of each antibody should be indicated;
.) Cytokine measurement; which subunit(s) of IL-12 was analyzed? IL-12p40 and/or IL-12p70 - this needs to be clarified.