Protection of the small intestine from harmful bacteria is a never-ending job. Paneth cells of the intestinal crypts function as prime producers of defensins in response to the microbial onslaught of the gut. antimicrobial peptides with strong sequence similarity to leukocyte defensins5. This is not unique to mice since defensin mRNAs are also abundant in human and rat Paneth cells6, 7. Like leukocyte defensins, natural or recombinant Paneth cell defensins display a broad spectrum of antimicrobial activity in vitro8, 9. In addition, type II phospholipase A2, an enzyme specialized in the lysis of bacterial phospholipids, is also secreted by Paneth cells7. The presence of multiple antibacterial proteins in the secretory granules of Paneth cells stimulated the hypothesis that these cells can sense bacteria entering the crypt, and attack these bacteria by releasing defensins and other antimicrobial substances. The conventional approach to testing the proposed function of Paneth cell defensins, targeted disruption (“knockout”) of their genes, was precluded by the large (> 20) number of similar genes encoding these peptides in the mouse. Luckily, defensins are first synthesized as larger inactive precursors and then activated by posttranslational proteolytic removal of an anionic peptide segment. If this step could be blocked, defensins would remain inactive and the effect of this deficit could be analyzed. The key processing enzyme turned out to be the metalloprotease matrilysin. Mice with a disrupted matrilysin gene could not activate most of their Paneth cell defensins, and showed decreased resistance to some intestinal infections10. However, at this point, the evidence for the antimicrobial role of Paneth cells still remained circumstantial as matrilysin was likely to have other host defense roles that could contribute to the weakened resistance of the mice lacking the enzyme. A dissonant note was sounded by a study in which transgenic mice were selectively poisoned as soon as they started making defensins. This was achieved by expressing a diphtheria toxin fragment or the large T-antigen of the simian virus under the control of a Paneth cell specific defensin promoter11. In these mice, Paneth cell development was delayed for the first 180 days of life, and the production of Paneth cell defensins was greatly diminished. Surprisingly, this did not affect the health of the mice, their weight, intestinal morphology or the distribution or amount of stainable microflora in the small intestine. To date, no adverse consequences of the ablation of Paneth cells in germ-free mice have been reported. Challenge studies are notoriously difficult to interpret, however, because host defense systems are highly redundant. Finding the right combination of microbes and test conditions that will expose the defect is laborious and no such studies have been reported yet. Ayabe et al. directly tested the ability of crypts and their Paneth cells to release antimicrobial substances and to kill bacteria2. After treatment with a calcium chelator, functioning crypts can be shaken loose from the underlying intestinal tissue and collected largely free of other cell types. Exposure to bacteria or their characteristic surface molecules stimulated