On 2014 Mar 11, Daniel Haft commented:

The CXX repeat proteins described in this paper are almost certainly modified post-translationally, with the side chains of multiple Cys residues bridged to their preceding residues to form thiazole-type heterocycles. In retrospect, the paper should have avoided overconfident use of the term "bacteriocin," given the lack of evidence then that these heterocycle-containing natural products were toxins rather than, say, peptide pheromones. The term "ribosomally synthesized and post-translationally modified peptide," or RiPP, was introduced in 2013 (see PMID:23165928), and addresses the linguistic hole for natural products that resemble bacteriocins in structure but may have another function. In the absence of convincing evidence for toxin activity, the broader term RiPP should be used.

However, Chopra, et al. have just published a description of sonorensin, a member of the heterocycloanthracin family from a marine isolate, Bacillus sonorensis MT93. They purified the product to homogeneity, and found broad spectrum antibiotic activity, affecting both Gram positive and Gram negative bacteria. (see http://aem.asm.org/content/early/2014/03/03/AEM.04259-13.abstract). Consequently, it now seems likely that additional members of the protein family defined by TIGRFAMs entry TIGR03601, including heterocycloanthracin itself (from Bacillus anthracis), indeed are active as bacteriocins.

On 2014 Mar 11, Daniel Haft commented:

The CXX repeat proteins described in this paper are almost certainly modified post-translationally, with the side chains of multiple Cys residues bridged to their preceding residues to form thiazole-type heterocycles. In retrospect, the paper should have avoided overconfident use of the term "bacteriocin," given the lack of evidence then that these heterocycle-containing natural products were toxins rather than, say, peptide pheromones. The term "ribosomally synthesized and post-translationally modified peptide," or RiPP, was introduced in 2013 (see PMID:23165928), and addresses the linguistic hole for natural products that resemble bacteriocins in structure but may have another function. In the absence of convincing evidence for toxin activity, the broader term RiPP should be used.

However, Chopra, et al. have just published a description of sonorensin, a member of the heterocycloanthracin family from a marine isolate, Bacillus sonorensis MT93. They purified the product to homogeneity, and found broad spectrum antibiotic activity, affecting both Gram positive and Gram negative bacteria. (see http://aem.asm.org/content/early/2014/03/03/AEM.04259-13.abstract). Consequently, it now seems likely that additional members of the protein family defined by TIGRFAMs entry TIGR03601, including heterocycloanthracin itself (from Bacillus anthracis), indeed are active as bacteriocins.