Thesis (Ph. D.)--Columbia University, 1927. / Vita. Bibliography: p. [29-30].

Gram-positive bacteria are part of the normal flora of fish from different aquatic environments. They are mesophilic bacteria and demonstrate optimum growth at ambient temperature. In the sub-tropics, marine fish are caught from seas at temperatures of 16 to 34C, they are usually not iced and are handled at ambient temperature. It was hypothesized that under these conditions Gram-positive bacteria will be abundant in sub-tropical marine fish and will have roles in the spoilage of fish. A review of literature showed that there is a gap in understanding the Gram-positive bacterial populations in sub-tropical marine fish. This is partly due to the fact that the selective media used for isolating Gram-positive bacteria have limitations. Ecological and speciation studies have revealed that the ecology and speciation of many Gram-positive bacteria have not been clearly elucidated. The effect of ambient storage on the individual genera and species of Gram-positive bacteria in fish has been rarely studied. The spoilage potential of Gram-positive bacteria of marine fish origin has not been clearly determined. Therefore, the main aims of this study were to isolate Gram-positive bacteria from fresh and ambient-temperature-stored sub-tropical marine fish, speciate the isolates and study the spoilage potential of the isolates. The practical components of this study were conducted in four parts. The first part dealt with validation of tryptone soya agar with 0.25% phenylethyl alcohol (PEA-TSA) to enumerate Gram-positive bacteria. The second part enumerated Gram-positive bacteria from the muscles, gills and gut of Pseudocaranx dentex (Silver Trevally), Pagrus auratus (Snapper) and Mugil cephalus (Sea Mullet) stored at 25C for 15 hours using PEA-TSA. The third part dealt with the speciation of the isolates using appropriate methods such as polymerase chain reaction, 16S rRNA gene sequence, the VITEK JR system and conventional biochemical methods. In the fourth part, the isolates were assayed qualitatively for their ability to produce volatile sulphur compounds (VSC), reduce trimethylamine oxide and decarboxylate histidine, lysine and ornithine at mesophilic temperature, 32C. Initial studies indicated that PEA-TSA significantly (P< 0.05) reduced the total aerobic bacterial count of fish whereas control Gram-positive bacteria were not affected (P> 0.05). Gram-positive aerobic bacterial counts (GABC) significantly (P< 0.05) increased in the muscles and gills during ambient storage for 15 hours. Within each species, no significant (P> 0.05) differences were found in GABC between muscles and gills. Moreover, there were no significant differences (P> 0.05) in GABC between fish species during storage. In total, 390 bacteria were isolated from the fresh and stored fish; 339 isolates (87%) were found to be Gram-positive. Two hundred and sixty-six isolates (78%) of Gram-positive bacteria were identified to fall into 13 genera, namely Staphylococcus, Micrococcus, Bacillus, Virgibacillus, Brevibacillus, Corynebacterium, Streptococcus, Enterococcus, Aerococcus, Exiguobacterium, Carnobacterium, Vagococcus and Sporosarcina and 30 species. In fresh fish, Staphylococcus epidermidis and Micrococcus luteus were the most frequent isolates. The effect of storage at 25C for 15 hours resulted in a change of Gram-positive bacterial populations; while S. epidermidis, S. xylosus and Bacillus megaterium were no longer present, S. warneri, B. sphaericus, Brevibacillus borstelensis, Enterococcus faecium and Streptococcus uberis increased. Three species, E. faecium, Str. uberis and B. sphaericus, were the most prevalent at the end of storage. Micrococcus luteus and S. warneri were the most prevalent isolates from Pseudocaranx dentex, but E. faecium and Str. uberis were the most frequently isolated from Pagrus auratus and Mugil cephalus. With respect to different parts of the fish body, E. faecium, Str. uberis and B. sphaericus were the most frequent isolates from the muscles, E. faecium, Str. uberis from the gills and M. luteus from the gut. Among the 228 isolates examined, Br. borstelensis 73, Br. borstelensis 291, Str. uberis 339, Vagococcus fluvialis 31 and Vag. fluvialis 132 produced VSC from sodium thiosulphate, cysteine and methionine. However, strains varied in sulphur source utilization. Exiguobacterium acetylicum 5, Exiguobacterium spp. 191, Carnobacterium spp. 338, Br. borstelensis 73, Br. borstelensis 291, Str. uberis 30, Str. uberis 339, Vag. fluvialis 31 and Vag. fluvialis 132 reduced TMAO. No histidine decarboxylase activity was found in the Gram-positive bacterial species tested. Lysine and ornithine were decarboxylated mainly by different strains of S. warneri, S. epidermidis and M. luteus. During ambient storage of fish, the frequency of lysine-decarboxylating bacteria increased and became more diverse after 5 hours of storage. Among fish species examined, the frequencies of lysine- and ornithine-decarboxylating bacteria were higher and more diverse in Pseudocaranx dentex than in Pagrus auratus and Mugil cephalus. This study found that Gram-positive bacteria were abundant and diverse in sub-tropical marine fish; however, their frequencies were affected by fish habitat and fish body part. Ambient temperature storage determined which Gram-positive bacterial species were dominant. With the exception of one isolate of S. aureus, Gram-positive bacteria isolated from sub-tropical marine fish caught from unpolluted water were not potential pathogens. The study also showed that Gram-positive bacteria had greater ability to decarboxylate lysine and ornithine than to produce VSC or reduce TMAO, and the spoilage potential of a bacterial species was a strain-dependent behaviour. This is a significant study as it is the first study on this aspect sub-tropical marine fish. It validated a selective medium that can be used to enumerate most Gram-positive bacteria from a marine environment. Most of the Gram-positive bacterial species from sub-tropical marine fish identified in this study were documented for the first time. The effects of ambient storage and the spoilage potential of Gram-positive bacteria from sub-tropical marine were clearly elucidated.

Dans un premier temps, ce travail a concerné la réalisation de microparticules chargées en agent antimicrobien suivant la technique de microencapsulation par évaporation de solvant en émulsion simple. Différentes morphologies ont été obtenues avec des microparticules éloignées du standard lisse, démontrant des cicatrices et des défauts, de la rugosité ou encore des trous. Les paramètres ainsi que les mécanismes physico-chimiques responsables des dégradations morphologiques ont été identifiés et discutés. Il a été démontré que les paramètres de formulation tels que la masse et masse molaire du polymère ou encore la présence de tensioactifs ainsi que les paramètres du procédé tels que la force et la vitesse de cisaillement modifient l'état de surface finale des microparticules. Ce travail a notamment prouvé qu'il existe une compétition entre la cinétique d'évaporation du solvant et la vitesse de coalescence des gouttelettes d'émulsion qui est à l'origine des dégradations morphologiques. Suite à cette étude, les microsphères résultantes contenant de l'alcool phényléthylique ont été enduites à la surface du conditionnement primaire polyoléfine sous forme de films minces de différentes épaisseurs grâce à la technique de revêtement par immersion. L'introduction de microparticules au sein du liant ralentit la diffusion de l'agent antimicrobien en augmentant le nombre de matrices polymériques à traverser pour atteindre le milieu extérieur. La réalisation de telles couches a permis d'obtenir des libérations sur des périodes supérieures à au moins trois mois ce qui est 15 fois plus important que celles obtenues pour l'agent antimicrobien non encapsulé. Ce travail de thèse a également étudié l'activité antimicrobienne de l'alcool phényléthylique au sein d'une émulsion. Il a été mesuré le partage de l'alcool phényléthylique entre les phases aqueuse, huileuse et micellaire de l'émulsion. Les résultats obtenus ont permis de développer un modèle mathématique calculant la fraction en agent antimicrobien libre présent en solution aqueuse. Ce dernier a été corrélé à des dosages de l'émulsion et des mesures microbiologiques utilisant les cinq souches microbiennes du challenge test sur 14 jours. Ainsi, il a été démontré que les calculs permettent de prédire la concentration en conservateur nécessaire afin d'assurer la protection antimicrobienne des formulations. Cette étude a notamment prouvé que la quantité d'alcool phényléthylique nécessaire à la conservation des formulations est respectivement 1,6 et 4,3 fois plus importante dans une solution micellaire et une émulsion par rapport à une solution aqueuse / First, this work focused on the formulation of microparticles loaded with antimicrobial agent using the emulsion/solvent evaporation method. Several morphologies have been obtained with nonsmooth microparticles characterized by scars and defects, roughness and holes. The parameters and the physico-chemical mechanisms involved in these morphological deteriorations have been identified and discussed. It has been shown that the formulation and processing parameters as the polymer mass and molar mass, the surfactant as well as the speed and shear rate of the propeller play a key role in the final microparticles surface states. This study proved that there is a competition between solvent evaporation and the coalescence of emulsion droplets which is responsible for the morphological degradations. Following this study, the resulting microspheres loaded with phenylethyl alcohol were dispersed in a binder and coated as thin films of various thicknesses by the dip-coating method at the polyolefin surface. It has been measured that the use of microparticles slows the antimicrobial agent diffusion by increasing the number of polymeric matrices that have to be crossed in order to reach the external medium. Such thin films resulted in an antimicrobial agent delivery up to 3 months which is 15 times higher than the delivery obtained for the non-encapsulated antimicrobial agent. The antimicrobial activity of the phenylethyl alcohol in an emulsion has also been investigated. The phenylethyl alcohol partition between the water phase, the oil phase and the micellar phase of an emulsion has been measured. These results led to the development of a mathematical model calculating the fraction of free antimicrobial agent present in the aqueous phase. It has been correlated with emulsion dosages and microbiological measurements using the five microorganisms of the challenge test during 14 days. It has been demonstrated that calculations enable the prediction of the antimicrobial agent concentration needed to ensure the antimicrobial protection. In particular, this work proved that the phenylethyl alcohol quantity necessary for antimicrobial protection is respectively 1.6 and 4.3 times higher for a micellar solution and an emulsion compared to an aqueous solution