Referee 1's Review:

The authors have used IVIAT to identify interesting B. anthracis antigens that appear during a model infection. This work identifies some interesting new antigens that may be relevant to therapeutics and/or shed light on anthrax pathophysiology.

The authors outline the limitations of their approach on pages 12-13. This discussion is very useful. In my view, useful data is acquired in spite of these limitations, as the authors argue. However, these limitations are important in interpreting the data. For example, since, in the infection spores are not used and cells do not need to disseminate via the lymphatics, certain classes of virulence factors won't be identified. Given this, I would suggest that the authors at least consider moving this discussion to the beginning of the paper, so that the reader can think about the data in this light.

P. 3, l. 6 from the bottom: Explicitly indicate that these two amidases were found by IVIAT.

P. 8, middle: The authors are correct that protein levels could be controlled post-transcriptionally. However, in the absence of some documentation of this, the possibility of a methodological problem is still on the table.

l. 9: The word "induction" is a little confusing, as it suggests evidence of an active regulatory process. If the authors mean that the protein levels increased based on proteomic analysis, that should be stated.

P. 9, l. 8: Insert "by IVIAT" after "BA4074".

Most of the content of the discussion of proteins identified is useful, but could be written in a more compact manner.

P. 9, middle: It is surprising that SleB was found. The authors should discuss how this might have happened. Does sleB appear to have a vegetative promoter?

P. 11, bottom: These experiments are interesting, especially in light of the species-specificity. In considering the data, it would be helpful for the authors to clarify whether there are any caveats given that the enzymes are applied externally when, presumably, their natural action is from the inside.

P. 11, bottom; P. 12, top: It is worth considering adding some additional details on the pH dependencies for the two enzymes. Does only one enzyme have an optimal pH? Perhaps more importantly, is anything known about the dependency of enzyme activity on growth phase of the target organisms? The growth phases used in this experiment is not entirely clear. For all the species, it seems that the cells could be in either lag phase or early exponential phase. This should be specified, based on typical growth curves that are probably already in hand. Since cells in an infection will be at a variety of growth phases, and the ability of the autolysins to act on the envelope may differ significantly at different growth phases, the relevance of the data need to be interpreted in light of this potential dependency.

P. 12, l. 5-6 from the bottom: how does codon bias impact on the library? An additional, very minor point: I think a number of younger readers may fail to understand the impact of partial digestion on an expression library, at first glance. Perhaps, reference a cloning manual.

P. 14, bottom: A reference for the details of the aerosol infection protocol should be provided.

**********
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.