Referee 1's review (Yoram Vodovotz):

The authors have previously established a methodology (referred to as a "resolution map") that consists of various inflammation-associated indices. The authors carried out studies in a mouse model of zymosan-induced acute peritonitis, using their resolution map concept, aimed at understanding the impact of anesthetics on inflammation in a systems fashion. They examined the effects of a local anesthetic (lidocaine) in a mouse model of peritonitis. They carried out a systematic proteomics analysis, comparing lidocaine to a volatile anesthetic, isoflurane, and suggested that differences between these two types of anesthetic could be observed using this analysis. For example, they showed that lidocaine was associated with increased numbers of infiltrating neutrophils (which the authors is posit is due to decreased apoptosis of these cells), while isoflurane does not affect neutrophil numbers. The authors also suggest that clinically used doses of lidocaine inhibit macrophage ingestion of apoptotic PMN, blocking their removal and resolution. The authors conclude that both local and volatile anesthetics impact endogenous resolution program(s), altering specific resolution indices and selective cellular/molecular components in resolution. They further conclude that isoflurane enhances whereas lidocaine impairs timely resolution. This study represent a major effort at a systems-level understanding of a central process in the inflammatory response.

A few issues need to be addressed by the authors, which would greatly enhance the quality and relevance of the manuscript:

1. In the Introduction, the authors state that "Glucocorticoids... enhance resolution by stimulating macrophage uptake of apoptotic PMN [Ref 15]," but they ignore the long history of studies concerning suppressed wound healing in the settings of glucocorticoid administration (e.g. Anstead, Adv. Wound Care. 11:277. 1998). This statement should therefore either be revised or removed.
2. Results: The authors hypothesize that the increases that they observe in S100 proteins reflect the dynamics of neutrophil infiltration. They mention a reference with regard to the role of S100 proteins in stimulating the recruitment of inflammatory cells in inflammatory settings, but they should also mention the broader role these proteins play as "alarm/danger" signals that propagate inflammation (e.g. Froell et al. J. Leukoc. Biol. 81:28. 2007).
3. Results: The data in Figs. 4A and B are depicted as "% inhibition of zymosan-stimulated exudates", as opposed to showing the actual pg/ml levels of cytokines in control, zymosan, and zymosan + lidocaine-treated exudates (see point #8 below). Even with this reasonable normalization of the data, the data exhibit some variability in the inflammatory response to zymosan. Are any of the changes in cytokines and chemokines statistically significant? If not, additional repeats of this key experiment may be necessary.
4. Results: Fig. 4C depicts TGF-b1 levels, which in the text the authors state come from the mouse zymosan model of inflammation. However, the figure legend states that Fig. 4C should depict a panel of cytokines in human whole blood, which appears to referring to the data in Fig. 4B. This should be corrected.
5. Results: In Fig. 4, did the authors examine both active and latent TGF-b1?
6. Results: The authors do not show any comparison in the levels of cytokines and chemokines in exudates from mice treated with zymosan + isoflurane vs. the data shown in Fig. 4 (zymosan + lidocaine). These data would be useful for comparison of the two anesthetics. In fact, a revised Fig. 4 in which raw data for the following conditions
7. Results: The authors refer to a Supplemental figure 1, which was not shown in the manuscript provided for review.
8. Results: Have the authors confirmed the elevation of S100 and other proteins, determined by LC/MS, using another method (e.g. western blot)?
9. The main point that the authors need to reconcile in this manuscript concerns the paradoxical beneficial effects of anesthetics like lidocaine (e.g. reference 29 and the authors' own data regarding the early reduction of inflammation in the setting of zymosan + lidocaine) vs. the detrimental effects (increased presence of leukocytes, elevation of pro-inflammatory cytokines at late time points) so amply demonstrated by the authors. This point may be best addressed by obtaining data on more traditional markers of damage (e.g. systemic levels of liver transferases) or healing (e.g. histology). Indeed the authors' focus on a single phase of the response (resolution), while carried out admirably in a systems fashion, inexorably leads to only limited conclusions (i.e., about the effects of these anesthetics on the resolution phase only). This one additional step (showing data on the ultimate healing outcomes under various experimental conditions) would likely address not only the above-mentioned controversy but also make the present paper substantially more relevant. The authors also need to be careful not to be perceived as showing only data that appear to support a particular hypothesis with regard to the effects of a given anesthetic; showing data comparing the effects of lidocaine vs. isoflurane on cytokines (see point #6) would help avoid this perception.
10. Methods: the authors should affirm that their mouse work was carried out following approval by the Institutional Animal Care and Use Committee, and that all relevant guidelines were followed.
11. Methods: since the authors show data that were derived from human subjects, they should affirm that the studies were carried out following approval by their Institutional Review Board, that all relevant guidelines were followed, and they should also state the methods used.
12. Minor: In the Discussion, the authors state that "It is noteworthy that, during an inflammatory disease state, a complex network of interactions between different cytokines is likely to occur. The timing of cytokine release and the balance between pro- and anti-inflammatory cytokines is likely to contribute to the overall outcome and severity." Computational simulations have in fact been carried out to address these complex interactions in the setting of acute inflammation in vivo (e.g. Chow et al, Shock 24:74. 2005), and thus should be cited.
13. Minor: Perhaps this reflects the conversion of figures to the PDF format, but the figure legend fonts appear difficult to read. A standard font (e.g. arial/Helvetica) would improve these figures.

**********
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.