The study presented here consists in an original piece of work to better understand complex food packaging interactions. The majority of investigations on food polymer interactions related to orange juice and this provided a good base to our study (Literature reviews: cf. Chapters 1a and b). Additionally a rather remarkable finding in 1994 was that limonin, a trace bitter material found in some varieties of orange juice was rapidly absorbed by highly plasticised polyvinyl chloride (PVC plastisol) (Chapter 2). Several commercial absorbants are available for debittering, relying on limonin absorption on the large surface area of the highly porous absorbant pellets. However, the absorptive properties of the smooth plastisols apparently relied on a different mechanism. Limonin is a very large (470.5 g/mol) compound, but some preliminary experiments with another much smaller orange juice constituent d- of absorbates in plastisols, methods used earlier (Moisan 1980, Holland and Santangelo 1988) to measure solubilities and diffusion constants in packaging films could be advantageously used to survey these properties in a wide range of materials, including model compounds of various types, and a number of compounds which may be found in citrus juices (Chapters 3, 4 and 5). Experimentally, the method found most suitable was to use a ‘test film’ of pure plastisol which was wrapped tightly on both sides by a similar ‘supply film’ blended with 1 Molar test material (also called ‘absorbate’), setting up a concentration gradient. The inner test film was removed at regular intervals (minutes to hours) to measure (mainly by weighing) the uptake of the test reagent with time. Rather unexpectedly, it was found in a number of cases that the test film lost weight, either from the beginning, or after a period of time. Three main types of behaviour were identified: Type A lost weight from the beginning and over a long period of time, Type B gained weight initially and then lost weight, and Type C gained weight until a steady state was reached. Often the maximum, or near maximum, mass increase occurred within around 100 minutes, indicating a very rapid, liquid-like diffusion mechanism, in harmony with the rapid uptake of d-limonene and limonin. The major parameters of interest with these compounds are their diffusion rates and their solubilities, and in the presence of aqueous media (orange juice and other foodstuffs) the partition coefficient between the plastisol and water, which is related to the hydrophobicity function LogP for the compound. The major complicating factor in these measurements is the observation that the plasticiser materials themselves also migrate, in the reverse direction, because of the lower effective concentration in the supply film. This effect tends to be small, but is one explanation for the mass loss observed above, and cannot be ignored over the long term, nor in its practical applications to contamination in foods. There are many possible applications for the techniques described above. The removal or addition of compounds in food packaging itself is one. Upgrading foods, such as orange juice, commercially, is another. In many cases ‘scalping’ off-flavours or other minor components takes place exclusively through solid or liquid contact with the packaging. The removal from the headspace measured by the current gas permeation methods is irrelevant for the vast numbers of involatile, but easily diffusable compounds. For such compounds these novel applications are simple and rapid, require little specialised equipment, and fill a niche in the armoury of food and packaging chemists. / Doctor of Philosophy (PhD)

Limonin and naringin are the two major bitter compounds in Citrus paradisi (grapefruit), and tissue-specific patterns on their distribution are well-established. This study was undertaken to determine the distribution of these compounds within Duncan, Marsh, and Thompson Pink tissues using three-dimensional fruit dissection (600-900 samples per fruit) and highly specific radioimmunoassays for limonin and naringin quantification. Results from a GLM ANOVA showed that there was no radial distribution pattern of limonin or naringin accumulation in these fruit. There were statistically significant differences in the axial distribution of these compounds within the fruit tissues. The limonin concentration in flavedo, albedo, outer segment memberane, and juice vesicles increased toward the distal end of the fruit. Naringin concentrations in flavedo, albedo, and outer segment membranes tended to be higher in the center portion of the fruit. Limonoate A-ring monolactone levels are also reported.

The molecular mechanisms of putative anti-inflammatory nutrients, i.e., fish oil, curcumin and limonin, were investgated with respect to CD4+ T cell function. Initially, using a DO11.10 mouse model which exhibits a transgenic T cell receptor specific to OVA 323-339 peptide, we demonstrated that dietary fish oil suppresses antigen-specific Th1 clonal expansion in vivo. Following immunization, the accumulation of adoptively transferred transgenic cells in wild type recipient mouse lymph nodes was suppressed. In addition, cell division analysis by carboxyfluorescein succinimidyl ester (CFSE) revealed that both total cell number in lymph nodes as well as cell division were decreased by fish oil. Since n-3 polyunsaturated fatty acids (PUFA), active long chain fatty acids in fish oil, elicit favorable effects on a variety of cell types, e.g., anti-tumor effect on colonocytes, amelioration of coronary heart disease and anti-inflammatory effects involving T cells, B cells, dendritic cells and macrophages, we postulated that a fundamental mechanism of action may explain the multiple effects observed. In a series of experiments described herein, we demonstrated that n-3 PUFA alters the formation/location of membrane subdomains, referenced to as lipid rafts. Specifically, lipid raft formation at the immunological synapse (IS) in CD4+ T cells was suppressed following membrane enrichment with n-3 PUFA. The alteration of lipid rafts down-regulated the localization of select signaling proteins, including F-actin, PKC and PLC-1, and phosphorylation of PLC-1 at the IS. Consequently, CD4+ T cell proliferation was suppressed as assessed by CFSE analysis and radioactive thymidine incorporation. Phytochemicals have been used for chemopreventive and chemotherapeutic purposes. We examined the putative anti-inflammatory effects of curcumin (1%) and limonin (0.02%) with respect to CD4+ T cell function. Dietary curcumin and limonin suppressed NF-B activation in CD4+ T cells. In addition, CD4+ T cell proliferation was modulated by 2% curcumin. We further investigated the combined therapeutic potential of phytochemicals and fish oil, containing n-3 PUFA. Interestingly, fish oil and limonin together significantly (P<0.05) suppressed T cell proliferation, whereas feeding either fish oil or limonin alone showed little effect. In summary, our data indicated that dietary fish oil alters proximal signaling of T cells by perturbing lipid raft formation. Curcumin and limoin are capable of suppressing NF-B in T cells, thereby exhibiting a synergistic effect when combined with fish oil. Further studies are required to elucidate the relationship of dietary dose of active compoments with respect to mechanism of actions.