Factors affecting production of an antilisterial bacteriocin by Carnobacterium piscicola strain A9b in laboratory media and model ®sh systems B. Himelbloom 1 , L. Nilsson 2 and L. Gram 2 1 University of Alaska Fairbanks, Fishery Industrial Technology Center, AK, USA, and 2 Danish Institute for Fisheries Research, Technical University of Denmark, Lyngby, Denmark 730/1/01: received 17 January 2001, revised 21 March 2001 and accepted 5 April 2001 B. HIMELBLOOM, L. NILSSON AND L. GRAM. 2001. Aims: To investigate factors in¯uencing bacteriocin production and bacteriocin stability of the bioprotective culture Carnobacterium piscicola strain A9b. Methods and Results: Maximum activity was obtained in MRS7 broth (MRS adjusted to pH 7á2), with or without glucose. No bacteriocin was produced in APT broth when a low inoculum level (0á001%) was used. In contrast, inoculum level did not in¯uence bacteriocin production in BHI and MRS7 without glucose. Bacteriocin production in APT was induced by the presence of an extracellular compound present in the sterile, ®ltered, cell-free supernatant ¯uid of a stationary-phase culture. Increasing concentrations of NaCl (2±7%) reduced bacteriocin production and maximum cell density of C. piscicola A9b when grown in cooked ®sh juice at 4°C. Conclusions: Media composition, inoculum level and sodium chloride concentration affected production. Signi®cance and Impact of the Study: The in¯uence of NaCl on bacteriocin production may negate the inhibitory effect of C. piscicola A9b against Listeria monocytogenes in salty foods. INTRODUCTION Listeria monocytogenes is isolated frequently from freshly- produced, cold-smoked salmon (Eklund et al. 1995; Rùrvik et al. 1995; Jùrgensen and Huss 1998). Due to the psycho- trophic and halotolerant nature of L. monocytogenes, there are concerns about the risk of consuming contaminated cold- smoked salmon. Inoculation trials show that this pathogen grows to considerable cell numbers during cold storage (Rùrvik et al. 1991; Guyer and Jemmi 1991; Rùrvik and Yndestad 1991; Hudson and Mott 1993; Nilsson et al. 1997). Although cold-smoked salmon has never been implicated in human listeriosis, cases are reported from the consumption of other lightly preserved, ready-to-eat ®sh products such as smoked mussels, cold-smoked rainbow trout and gravad trout (Ericsson et al. 1997; Brett et al. 1998; Miettinen et al. 1999). As L. monocytogenes is a common micro-organism in the environment, and since the cold-smoking process does not contain a listericidal step, it is impossible for all commercial processors to produce Listeria-free, cold-smoked salmon. Government and industry have focused on introducing critical control points that can prevent the growth of L. monocytogenes in ready-to-eat ®sh products such as cold- smoked salmon. This hazard may be controlled by using puri®ed bacteriocins or live protective bacterial cultures (Nilsson et al. 1997, 1999; Duffes et al. 1999a, b). However, the listericidal or listeristatic ef®ciency of bacteriocins and protective cultures in cold-smoked salmon may be in¯u- enced by physico-chemical and microbiological factors such as NaCl (typically 3±6% in the water phase), temperature and cell density of the protective culture. The effect of NaCl on bacteriocin production and activity varies, and may be enhanced (Thomas and Wimpenny 1996; Parente et al. 1998; Uguen et al. 1999) or reduced (De Vuyst et al. 1996; Casla et al. 1996; Bouttefroy et al. 2000). Previously, the selection and application of Carnobacterium piscicola strains as protective cultures for the control of L. monocytogenes in cold-smoked salmon was described Correspondence to: L. Nilsson, Danish Institute for Fisheries Research, Department of Seafood Research, Sùltofts Plads, c/o Technical University of Denmark, bldg. 221, DK-2800 Lyngby, Denmark (e-mail: lni@dfu.min.k). ã 2001 The Society for Applied Microbiology Journal of Applied Microbiology 2001, 91, 506±513