Gluconacetobacter xylinus (Ga. xylinus) is a Gram negative bacterium which can produce an extracellular polysaccharide called bacterial cellulose. This material potentially has a wide range of applications in the food industry and other fields. In addition to fresh fruits, Kombucha, a Chinese traditional beverage fermented from tea and sugar, is also known to be one of the habitats of Ga. xylinus. Despite this, almost all previous studies on bacterial cellulose production have used Ga. xylinus strains isolated from fresh fruit sources and Kombucha strains of this species have not been extensively characterized for bacterial cellulose production. It is hypothesized that as is the subject of long-term sub-culture under static conditions, Ga. xylinus strains from Kombucha may have different cellulose producing characteristics than those typically encountered from other sources. Although bacterial cellulose has been applied in various fields, the use of bacterial cellulose thus far remains unexploited in the area of active package production. We, therefore, have undertaken this study to (i) characterize and improve cellulose production by a strain of Ga. xylinus (K3) previously isolated in our laboratory from Kombucha, (ii) develop a new application of bacterial cellulose as an active food packaging material and (iii) investigate the occurrence of spontaneous mutations in cellulose synthesis in Ga. xylinus K3 as this may impact on cellulose production. In order to characterize cellulose production by Ga. xylinus K3, cellulose yield in tea medium with both black tea and green tea, and in Hestrin-Schramm (HS) medium under both static and agitated culture, was initially compared. In tea medium the highest cellulose yield was obtained with green tea (~0.20 g l-1) rather than black tea (~0.14 g l- 1). The yield in HS was higher (~0.28 g l-1) but did not differ between static and agitated incubation. 1H NMR and 13C NMR spectroscopy indicated that the cellulose produced is pure (free of acetan) and has high crystallinity, respectively. A time course study of cellulose production showed that no cell-density dependent effect was present. Subsequently, cellulose yield was improved by changing types and levels of carbon source in HS medium under static culture. An improved yield of ~2.64 g l-1 was obtained with mannitol at 20 g l-1 and corn steep liquor (CSL) at 40 g l-1 in combination. In tea medium, tea at a level of 3 g l-1 gave the highest cellulose yield and the addition of 3 g l-1 of tea to HS medium increased cellulose yield to ~3.34 g l-1. The improvements in the culture medium were directed at producing a high yield of bacterial cellulose for development of a new application of this material in the active food packaging field. In order to produce an active packaging material, bacterial cellulose pellicles produced by Ga. xylinus K3 in the improved medium had nisin incorporated in them. The active bacterial cellulose films were used in a proof-of-concept study to attempt to control Listeria monocytogenes and other bacteria on the surface of vacuum-packaged Frankfurters. Nisin incorporation studies using bacterial cellulose membranes at various moisture conditions showed that nisin bound well to wet cellulose films, but not to dried cellulose films. It was also determined that the lowest nisin concentration required was 625 IU ml-1 and shortest time needed was 6 h for production of effective active cellulose films. The active cellulose films produced in these conditions did not, however, significantly reduce (P>0.05) L. monocytogenes populations on Frankfurters during refrigerated storage. As this finding was thought to be due to low concentrations of nisin, a higher concentration of nisin (2500 IU ml-1) was used and cellulose films were exposed to this nisin solution for 6 h. The resulting films significantly decreased (P<0.05) L. monocytogenes counts in Frankfurters by nearly 2 log units over 14 day of storage. Both the above mentioned films showed the similar effectiveness in reducing spoilage bacteria, as measured by a Total Aerobic Plate Count (TPC), on Frankfurters. In both cases TPC levels were observed to be approximately 3.3 log units lower than in control samples after 14 day of storage. The presence and nature of spontaneous mutations in cellulose production by Ga. xylinus K3 was subsequently studied because this phenomenon has lead to a reduction of cellulose yield in other Ga. xylinus strains. Previous studies have found that spontaneous mutant formation is dependant on the growth conditions (agitated or static) of the media but not the media itself. In order to examine the presence of mutants, Ga. xylinus K3 was grown in HS medium and improved HS medium under both static and agitated conditions. It was observed that this strain generated spontaneous mutants with lower cellulose production in basic HS medium and not in improved HS medium, regardless of whether they were static or agitated cultures. This indicated that the spontaneous mutation of Ga. xylinus K3 seemed to occur depending on the type of culture medium used. Sub-culturing a colony of the mutated strain into HS medium showed that the mutants did not revert and produced a sinking pellicle of bacterial cellulose. NMR spectroscopy indicated that cellulose produced by the mutant and the parent strain had similar structural features, but the mutant cellulose had fewer fibrils as observed by scanning electron microscopy (SEM). Comparison of cellulose production of the mutant and parent strain in HS medium revealed that mutant grew faster than the parent strain but produced less cellulose, indicating that there was a trade-off between cell growth and cellulose production in the mutant. On the other hand, the mutant grew at a similar rate and produced a similar cellulose yield as the parent strain in improved HS medium. At the molecular level, two-dimensional gel electrophoresis demonstrated that there were some differences in the protein patterns of the mutant and parent strains. Sequences of peptides derived from these proteins revealed that only the mutant produced deoxythymidine diphosphate (dTDP)-4-dehydrorhamnose 3,5-epimerase. This enzyme is involved in acetan synthesis and thus may reduce the cellulose yield of the mutant. In conclusion, our studies extensively characterized cellulose production from a Ga. xylinus strain from Kombucha for the first time. The results indicated that this strain had independent cellulose producing characteristics which were different from previous described strains and these characteristics may be associated with its source of origin. The strain is suitable for cellulose production under static culture and cellulose yield can be improved by supplementation of the growth medium with tea. Bacterial cellulose was produced in an extremely pure form and thus shows high potential applicability. One of the potential applications of bacterial cellulose developed in this project is its use as an active packaging material for inhibition of pathogenic bacteria on the surface of food products, a potential use of this material which has so far not been exploited. It was also found that growth of Ga. xylinus K3 in nutrient poor media allowed the development of lower-cellulose-producing mutants, indicating that spontaneous mutation of the strain was different from that of previous described strains. This phenomenon was clearly related to a reduction in cellulose yield. Potential reasons for differences between mutant and the parent strains have been determined. These results provided helpful information for developing bacterial cellulose production at an industrial scale.