Referee 2's Review (Orit Gat):

The study describes application of IVIAT to screen for Bacillus anthracis immunogenic proteins, expressed during infection. The screen generated a list of ten targets. Eight additional targets were bioinformatically selected according to function (cell wall biogenesis). RT-PCR performed on organs following challenge of mice, validated that four genes, in addition to the toxin gene, were indeed expressed during infection. Two of the cell wall hydrolases were found to specifically act upon B. anthracis cells. The authors suggested that cell wall hydrolases may represent novel potential therapeutic targets against anthrax.

The study addresses an important issue, it expands the existing data concerning in vivo expression of B. anthracis proteins and I recommend its publication following the corrections detailed below.

Major Comments:
1. Results, p. 8 line 3 from bottom to p. 11 line 11. This section discusses bioinformatics data concerning IVIAT-identified ORF products and does not belong under the title "RT-PCR…". The whole section should be transferred to page 7 following "Table 1a". P. 9 line 11. BA3767 was not identified by IVIAT but rather bioinformatically-selected. This should be clearly stated in the text. Furthermore, the data concerning all bioinformatically-selected ORFs should be removed from its original location and presented at the end of this section. P. 9 lines 13-20. Description of ORF BA2748 should be transferred following the description of the other immunoreactive proteins with similar cell wall activities. P. 11 lines 8-9. Delete "Our application … domains, and" since this does not add any information and additional ORFs from the screen are not discussed as well.
2. P. 12 line 16 ("our study has a number of limitations….." to end of section. (1) First part of the sentence describes issues not relevant to those described in the second part. The second half presents a limitation of any library screening process, not necessarily that presented in the manuscript. (2) This is an inherent characteristic in any genomic screen. (3) and (4) could be a complication when too many targets are identified – which is obviously not the case. (3) This is not relevant to the screen described here since there are no cross-reactions. (4) This is also not a limitation but rather an advantage, considering the fact that it prevents the loss of relevant genes. Anyways, it does not seem that the system is sensitive to such differences. (5) This point is well taken. Sera against B. anthracis-derived antigens expressed during infection are difficult to prepare. The serum described should be termed, more appropriately, hyperimmune. Preparing convalescent sera against B. anthracis entails treatment of infected animals with bacteriostatic antibiotics or using partially attenuated strains, such as pagA-deleted mutant. The sentence "We, however, chose…. Available" is not relevant and should be omitted. (6) This point is not part of IVIAT.
In summary, this section should be revised. A major drawback of the IVIAT is the use of an E. coli library. When E. coli is chosen as a host for cloning, proteins which are either membrane-associated or secreted are notoriously under-represented due to toxicity problems (caused by effects of secretion stress). In a screen that issues proteins which are mainly extracellular – this is a major complication. A second problem is the serum used for screening. If the total anti-B. anthracis antigen titer is not high enough, the serum will not react well with antigens, especially in the presence of PA which is an incomparable stronger immunogen and its sero-reactivity may mask other weak immunoreactions.
3. P. 13 last paragraph. A former recent study aiming at B. anthracis proteins that possess antigenic properties and are expressed during infection also supports a prominent role for enzymes that participate in cell wall biogenesis as targets for diagnostic and therapeutic intervention (Gat et al., 2006. Infect. Immun. 74:3987-4001). In the current screen, several of the antigens detected were already documented (BA314, pXO2-8, BA485, BA2805, BA3737, BA1952, BA5427). The screen described here confirms that published before and therefore, the previous genomic screen should be quoted and discussed properly.

Minor Comments:
1. Introduction, p. 5, line 10. Change "host cell membranes" to "specific host cell receptors".
2. Results, p. 7, line 3. Table 1a lists 12 proteins, rather than 10, as indicated in the text. However, since BXB0048 and BA2125 were not validated, the authors may consider removing them from the table. Furthermore, their presence does not add any important information.
3. P. 7, lines 8-12. Is there any reason to believe that immunoreactivity will be unique only to ORF products found in the screen? This statement should be revised. "Since 4-5 (?) out of 10 identified seroreactive ORFs fall into the category of cell-wall hydrolases, we further evaluated bioinformatically-selected paralogs of two IVIAT-identified … Out of 8 (?) ORFs selected, all (?) were found to be immunoreactive (Table 1b)".
4. P. 7, lines 15-16. Change to "… expression of the IVIAT-identified ORF products, we measured mRNA levels of their respective genes".
5. P. 7, line 21. How were bacteria grown?
6. P. 9, line 11. Delete "and prophage holin BA3768".
7. P. 9, line 18. Change "late in sporulation" to "late during sporulation"
8. P. 10 line 18. Change "domains" to "domain", "they are" to "it is"
9. P. 12 lines 5-6. What is the meaning of the numbers?
10. Table 1a. Change "Location of ORF" to "Gene encoded by";
BA314: add "substrate-binding protein of an" to "ABC transporter"
BA314: change "unknown location" to "extracellular, cell-associated"
BA669: add "ribose-binding protein of an"
BA669: change "unknown location" to "extracellular, cell-associated"
pXO2-8: add ", cell-wall"
BA4073, BA4606, BA698 and BA4074: according to their assigned functions are predicted as "extracellular, cell-associated" or "unknown" rather than "cytoplasmic"
BXB0048: delete "pXO2 encoded protein".
11. Table 1b. Change "Location of ORF" to "Gene encoded by";
BA3737 (SLH), pXO2-42 (SLH), BA1952 (NlpC/P60) and BA5427 (NlpC/P60): keep uniform predicted location, "extracellular, cell-associated".
12. P. 15, line 5 from bottom. Add "immuno-" to "detected _reactive clones using…"
13. P. 15, line 2 from bottom. Change "immuno-reactive" to "positive".
14. P. 16, line 13. Change to "bioinformatically selected paralogs".
15. Animal care and treatment should be detailed.
16. P. 29, legend to Figure 2. Revise. "B. anthracis species-specific autolysis effects following addition of exogenous N-acetyl….BA2446. … (A) Reduction in A595 optical density… (B) Reduction in colony forming units … (C,D) … of B. cereus (C) or B. anthracis (D). ". Omit the following words, "B. cereus, B. subtilis….are shown (C)." since this data does not appear in the figure.

**********
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.