Referee 2's review (Jürgen Brojatsch):

This paper demonstrates that the quinone methide triterpene celastrol is a potent inhibitor of LT-mediated killing of the murine macrophage cell line RAW264.7. The paper shows that celastrol is more efficient than the commonly used proteasome inhibitor MG132 in blocking cytopathic effects triggered by LT. The authors demonstrate that celastrol blocks LT-mediated processing of IL-18 and inflammasome activation, thereby blocking cytolysis. However, celastrol did not block early events such as cleavage of MAPKK1, indicating that it blocks late events in LT killing. This was further supported by findings that demonstrated that celastrol blocks LT killing, even when added as late as 1.5 hours post-LT exposure. These findings suggest that celastrol prevents LT killing by blocking proteasome activity in RAW cells. The findings are presented in clear fashion, and the paper is well written. The inhibitor described in this manuscript has clear potential as a therapeutic, and it remains to be shown whether it is able to prevent morbidity and mortality of mice challenged with LT or anthrax spores.

There were a few minor issues that should be addressed by the authors before publication:

The authors should mention in the abstract that their findings suggest that celastrol blocks LT killing by its ability to inhibit proteasome activity. Figure 2 suggests that celastrol is more efficient than the proteasome inhibitor MG132 in blocking proteasome activity. Celastrol treatment prevented LPSmediated IkB degradation, while MG132 showed comparitively little effect on IkB levels. Moreover, levels of ubiquitinated proteins increased drastically following celastrol treatment indicating that the inhibitor prevents the degradation of ubiquitinated proteins in RAW cells, while levels of ubiquitinated proteins increased only marginally following MG132 treatment. These results are surprising, as MG132 is a powerful proteasome inhibitor that usually completely prevents proteasome activity at concentrations even below 10 ug/ml used by the authors. The authors should determine the sensitivity of RAW cells to the proteasome inhibitors celastrol and MG132. A titration of both inhibitors and their effect on LT killing and accumulation of ubiquitinated proteins should be presented. These experiments would indicate the efficiency of each inhibitor.
Subsequent inhibition studies should be performed at celastrol and MG132 concentrations that are able to block more than 90% of LT killing and accumulation of ubiquitinated protein. In case MG132 fails to block proteasome activity in RAW cells, alternative proteasome inhibitors should be considered.

Also, the authors did not indicate whether the presented results show representative results. The authors should indicate how often experiments were repeated, and whether representative results are presented. The authors demonstrated that celastrol and MG132 block proteasome activity in distinct ways. While MG132 blocked the cleavage of fluorogenic substrates by isolated 20S and 26S proteasomes, celastrol was inefficient in blocking the cleavage of these substrates. This indicates that both proteasome inhibitors block the proteasome in different ways. This is surprising, as celastrol is very efficient in blocking degradation of ubiquitinated proteins, and presumably blocks the activity of the 20S proteasomes. The authors should discuss this potential discrepancy.

The authors demonstrate that celastrol prevents IL-18 cleavage in LT-treated RAW cells. However, only minimal levels of processed IL-18 are visible in RAW cells following LT treatment. This result is surprising, as caspase-1 activation is essential for LT killing. We consider this a key experiment, and suggest that the authors perform the experiment using either J774A.1 macrophages or primary bone marrow derived macrophages from a susceptible mouse. These cells are more susceptible to LT killing and show higher levels of caspase-1 activation than RAW cells. Additionally, a new caspase-1 detection kit is available that has drawn praise from several groups investigating inflammasome activation. FAMYVAD-FMK is useful at detecting even low levels of active caspase-1 in murine cells, and kits are available from Immunochemistry Technologies, Bloomington MN. This caspase-1 detection kit provides an alternative robust caspase-1 assay.

----------
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.