Reviewer 2's Review

-----

This manuscript by Smith et al examines the replication of murine gammaherpes-68 (MHV-68) virus in DC and assesses the effect of infection on antigen presentation by DC to immune cells. The authors address an important point by determining if MHV-68 replication in DC is latent or lytic and how this affects antigen presentation. Nevertheless the manuscript needs further improvement in both experimental design and data presentation in order to strengthen the conclusions.

In order to determine if DC are latently or lytically infected, the authors try to correlate GFP expression with capsid distribution by confocal microscopy (figures 2-4). Alongside of the confocal pictures, it would be more informative and clearer to present a table with number and % of cells (immature/mature DC) that are positive for GFP and capsid markers so that we understand how frequently DCs are infected and the proportion of latent and lytic infection.

Furthermore the data suggest that immature but not mature DC are infected by MHV-68. It would be useful to confirm these findings by a different approach. For instance the authors could separate immature and mature DC (or induce maturation of immature DC in vitro), infect both populations and determine that only immature DC are infected.

Confirming the infection or absence of infection in these DC populations by MHV-68-specific PCR would strengthen the conclusions.

Interestingly K3 downmodulates the surface expression of H2 molecules in infected DC which results in a lesser recognition of infected cells by epitope-specific CTL. Is this downmodulation observed for all classes of H2 molecules in DC or just for the H2Db molecule that restricts the p56 epitope? Furthermore if the surface expression of H2 is reduced by half a log, one would expect a difference in the recognition of target cells pulsed with low concentration of peptides by epitope-specific CTL. It seems necessary to repeat peptide titration with lower peptide concentration.
If the authors want to compare the effect of K3 on lytically infected cells and on unfractionated DC, they should look at the same epitope in both experiments (figure 5-6 with MHV-68 p56 epitope vs figure 7 where OVA presentation is monitored).

The manuscript is built like a rebuttal to Flano et al (J Immunol 2005). Since the experimental systems used in this manuscript and the JI article are so different, I am not sure that this is the best way to present the data.
Moreover the description of the results sometime lacks important details. Figure 2B and 3B, use gM- eGFP rather than eGFP. In the description of the results of figures 6 and 7, indicate the H2 or MHC-II molecules restricting the epitopes.

-----

N.B. These are the general comments made by the reviewer when reviewing this paper in light of which the manuscript was revised. Specific points addressed during revision of the paper are not shown.