In the early 1990âs, Kock and co-workers discovered acetylsalicylic acid (ASA)-sensitive oxylipins in yeasts. It was also reported that the site of production of these compounds may serve as important targets to control fungal infections. In 2004, researchers exposed another function for these oxylipins â they may act as lubricants during spore release from enclosed asci. Since oxylipin production in only a limited number of species representing Eremothecium was thus far studied, it became the aim of this project to further extend this study and to determine the type and distribution of 3- hydroxy (OH) oxylipins in the remaining species i.e. Eremothecium coryli, E. cymbalariae and E. gossypii. In addition, the possible functions of these oxylipins as well as ascospore shape and ornamentations were assessed. Finally, the antifungal activity of ASA was also investigated in this group of important plant pathogens as well as other yeasts. Eremothecium coryli is known to produce intriguing spindle-shaped ascospores with long and thin whip-like appendages. In this study, ultra structural studies using scanning electron microscopy, indicate that these appendages serve to coil around themselves and around ascospores causing spore aggregation. Furthermore, using immunofluoresence confocal laser scanning microscopy it was found that hydrophobic 3-OH oxylipins cover the surfaces of these ascospores. Using gas chromatography-mass spectrometry, only the oxylipin 3-OH 9:1 (a monounsaturated fatty acid consisting of a hydroxyl group on carbon 3) could be identified. Sequential digital imaging suggests that oxylipin-coated spindle-shaped ascospores are released from enclosed asci probably by protruding through an already disintegrating ascus wall. Using immunofluorescence microscopy and 3-OH oxylipin specific antibodies, it was possible to map the presence of these compounds also in other Eremothecium species. In E. cymbalariae, these oxylipins were found to cover mostly the spiky tips of narrowly triangular ascospores while in E. gossypii, oxylipins covered the whole spindle-shaped ascospore with terminal appendages. The presence of these oxylipins was confirmed by chemical analysis. When ASA, a 3-OH oxylipin inhibitor, was added to these yeasts in increasing concentrations, the sexual stage was found to be the most sensitive. Results suggest that 3-OH oxylipins, produced by mitochondria through incomplete β-oxidation, are associated with the development of the sexual stages in both yeasts. Strikingly, preliminary studies on yeast growth suggest that yeasts, characterized by mainly an aerobic respiration rather than a fermentative pathway, are more sensitive to ASA than yeasts characterized by both pathways. These data further support the role of mitochondria in sexual as well as asexual reproduction of yeasts and its role to serve as target for ASA antifungal action.

Although beer production is one of the oldest biotechnologies in the world, a major constraint in brewing remains controlling flocculation. Evidence points towards a possible role of lipids, associated with the cell surfaces, as a major factor responsible for flocculation. Therefore, the aim in this study became to evaluate the contribution of lipids, especially oxylipins, in the flocculation of Saccharomyces cerevisiae UOFS Y-2330. Saccharomyces cerevisiae UOFS Y-2330 was selected as a model, since it was found to demonstrate both Flo1 and NewFlo phenotype flocculation behaviour, when cultivated in different media. In a defined medium with glucose as a sole carbon source, this strain immediately flocculated strongly and lost this ability before stationary phase was reached. In a complex medium containing glucose, this yeast strongly flocculated towards the stationary growth phase without losing this ability during this phase. This inverse pattern may be ascribed to a switch in sensitivity of the yeast to flocculate in the presence of glucose as well as pH level, which may, in turn, influence the availability of calcium ions. In both media, matured cells produced protuberances upon flocculation as observed by electron and immunofluorescence microscopy, which may be involved in cell adhesion. This was followed by further investigations into the role of lipids over the growth cycle of this yeast. Here, it was uncovered that Sacch. cerevisiae UOFS Y-2330 does not only demonstrate inverse flocculation, but is also characterised by two different lipid turnover patterns. During Flo1 phenotype flocculation, this yeast showed two neutral lipid accumulating stages (i.e. at 8 h and from 12 h). This is probably triggered by flocculation, which may be regarded as a survival mechanism where cells accumulate especially neutral lipids as reserve energy source - a similar mechanism is probably operative when cells enter stationary growth. Contrary to Flo1 behaviour, this strain in NewFlo phenotype mode demonstrates only a single lipid accumulation phase i.e. when cells enter stationary growth, which coincides with the increase in flocculation. In addition, an increase in phospholipids was experienced during active growth in both flocculation behaviours, probably as a result of active membrane production. These results prompted us to investigate the possible role of oxylipins present on the cell surfaces during the flocculation process. It was found that some strains of Sacch. cerevisiae (include strains used in fermentation processes) produce short chain (mainly 8 carbon) oxylipins and not potent inflammatory long chain (20 carbon) oxylipins such as prostaglandins. When aspirin was added to cultures of Sacch. cerevisiae UOFS Y-2330, flocculation was significantly inhibited as well as the production of 3-hydroxy (OH) 8:0 thereby linking flocculation and this oxylipin. Furthermore, no traces of 3-OH 8:0 could be detected before flocculation onset in this yeast. Next, the involvement of these oxylipins in co-flocculation was assessed. According to the lectin-theory, the yeast Schizosaccharomyces pombe lacks the specific receptors necessary to facilitate co-flocculation with Sacch. cerevisiae species. In this study we demonstrate oxylipin associated co-flocculation between Sacch. cerevisiae UOFS Y-2330 and S. pombe strains using differential cell staining, immunofluorescence and ultrastructural studies. Using a 3-OH oxylipin specific antibody coupled to a fluorescing compound, 3-OH oxylipins were found to be present on the cell surfaces of Sacch. cerevisiae and S. pombe. The presence of 3-OH oxylipins was confirmed using gas chromatography-mass spectrometry. Whether these 3- OH oxylipins play a role in affecting co-flocculation of Sacch. cerevisiae with S. pombe cells through possibly entropic-based hydrophobic interactions and/or hydrogen bonds still needs to be verified. Studies on the physiological, genetic as well as colloidal aspects of flocculation using this model strain may lead to important new insights in this fascinating phenomenon as well as applications in industry.

Biohydroxylation reactions are catalyzed by various types of hydroxylating enzymes (Ayala and Torres, 2004) which include dioxygenases, lipooxygenases as well as CYP450 monooxygenases. These particular hydroxylation reactions have several advantages over chemical synthesis. Several microorganisms including yeasts have the ability to hydroxylate various substrates. The exploitation of microbial hydroxylations for the production of industrially useful products such as pharmaceuticals is a more recent development (Holland et al., 2000). Yeasts from the genera Schizosaccharomyces, Pichia, Saccharomyces and Yarrowia have all been used to express foreign CYP450 genes (Mukarami et al., 1990; Nthangeni et al., 2004) since they offer an advantage especially when a eukaryotic environment is required for the functional expression of the heterologous gene (Blanquet et al., 2003). A recent evaluation of several yeasts revealed that Y. lipolytica is, a highly attractive alternative host for secretion and expression cloning (Muller et al., 1998; Juretzek et al., 2001). However, a literature search on successful expression of CYP450s in Y. lipolytica yielded only six cases. Three of these were done in our laboratory. In most of the reported cases, the recombinant CYP450 activities were never evaluated in terms of whole cell biotransformations. It was therefore the aim of this study to evaluate Y. lipolytica as a recombinant whole-cell biocatalyst for hydroxylation reactions by using available Y. lipolytica strains overexpressing different CYP450s which were (i) CYP1A1 coding for polyaromatic hydrocarbon hydroxylase (ii) CYP53B1 coding for benzoate para-hydroxylase (iii) CYP52F1 coding for alkane hydroxylase and (iv) CYP557A1 coding for putative fatty acid hydroxylase. Hydroxylase activities of the genetically engineered strains were compared with activities in wild type yeasts expressing the relevant CYP450s. A variety of substrates used for biotransformation reactions included, acetanilide, benzoic acid , phenylnonane, trans-cinnamic acid, 4-nitrophenyl octyl ether and 4-nonyloxybenzoic acid. Experiments using Y. lipolytica overexpressing CYP1A1 illustrated the limitation of using Y. lipolytica for the biotransformation of substrates such as AA since the endogenous enzymes degraded this substrate within only 12 h after substrate addition. In an attempt to distinguish the activities of the putative fatty acid hydroxylase and the alkane hydroxylase overexpressed in Y. lipolytica from the endogenous CYP450s, 4-nitrophenyl octyl ether, 4-nonyloxybenzoic acid and phenylnonane were used as substrates. 4-nitrophenyl octyl ether proved to be expensive and less sensitive to TLC, GC and GC-MS analyses. It has been used in other studies because it yielded p-nitrophenol which can be assayed colourimetrically by measuring absorbance at 420 nm. However, in our experiments, intermediates accumulated that were not completely transformed into p-nitrophenol. Further biotransformation experiments were carried out using 4- nonyloxybenzoic acid as the substrate. Biotransformation experiments were done using Y. lipolytica strains with intact and partially disrupted -oxidation pathway overexpressing CYP52F1 and CYP557A1. Additional experiments were carried out using wild type Y. lipolytica W29, R. minuta and R. retinophila strains. The results demonstrated that, the wild type Y. lipolytica W29 demonstrated the highest specific hydroxylase activity when 4-nonyloxybenzoic acid was used as the substrate. The main limitation here was the inability to selectively induce the overexpressed CYP450 genes alone without the background endogenous CYP450 activity. Due to the limitations above, the next strain used was Y. lipolytica TVN91 (overexpressing benzoate-para hydroxylase from R. minuta). In this case, the host strain lacked the specific CYP450 to perform the same hydroxylation reaction. The substrate used here was BA. Different growth and induction conditions were evaluated to optimize benzoate-para-hydroxylase activities. A comparison of the hydroxylase activities indicated that the activity of the recombinant Y. lipolytica strain overexpressing the CYP53B1 was about 30 times slower than that of the wild type R. minuta from which the gene was cloned. Continuous addition of stearic acid resulted in the best activity with Y. lipolytica TVN91, because the hydroxylase activity was maintained for a longer duration. When PN and CA were used to evaluate substrate transport limitation, the results demonstrated that substrate transport was not limiting and the specific hydroxylase activity was not increased. PN was initially converted to BA before hydroxylation to form pHBA. These results further demonstrated that hydroxylase activity of PN was much faster than that of BA. The results from the bioreactor study demonstrated that an improved aeration and mixing led to an increase in the benzoate para-hydroxylase activity. The possibility of using a chemically defined media (YNB) supplemented with yeast extract and casamino acid for biotransformation was also demonstrated. The results of this study also demonstrated that it is possible to use harvested recombinant cells for biotransformation without significant loss of activity. This makes it possible to study in detail the kinetics of the overexpressed CYP450s. It was, however apparent that the hydroxylase activities were significantly increased by both aeration and cell concentration.

The main aim of this study was to determine whether data, obtained from FCP and using semi-naïve panellists (experienced in descriptive techniques, but no previous experience with cheese), could express the perceptions of South African consumers on Cheddar cheese attributes. Firstly a panel of 220 consumers was asked to indicate their level of acceptance on a nine-point hedonic scale for overall acceptance of 15 Cheddar cheese products, in two locations in South Africa. The 15 Cheddar cheese samples included five retail Cheddar cheeses from four dairy companies, four cheese samples from two culture houses and six experimental cheeses. The ageing period of the 15 cheeses ranged from 60 to 180 days, all being mild Cheddars, except one mature Cheddar cheese. Significant differences (p < 0.05) occurred amongst consumer demographics for consumer acceptability. The number of cheeses showing a significant difference for the different main effects was: gender (two); income (one); population group (three); age (eight); and location (five). For the overall liking attribute, there was a significant difference in acceptance (p < 0.05) between the 15 Cheeses. Fisherâs Least Significant Difference test at a 5% significance level was performed to determine which cheeses differed significantly from one another for overall liking/acceptance. The most liked cheese sample had the highest mean value of 7.16, was aged for 60 days and yellow in colour. The least acceptable cheese had a mean value of 4.75, was aged for 180 days and was white. Free choice profiling (FCP) was carried out in order to investigate how semi-naïve consumers (who had experience in descriptive work and received minimal training on Cheddar cheese) described and perceived different Cheddar cheese samples. This method allowed participants to use their own attributes to describe and quantify the food product. The study used 15 different Cheddar cheeses available in South Africa, analyzed by ten consumers in three replications. The data were analyzed by using generalized Procrustes analysis. The FCP procedure generated between 21 and 42 attributes, with an average of 35, including 16 descriptors on the attribute aroma, 14 descriptors on the attribute texture/appearance, 15 descriptors on the attribute mouthfeel, 20 descriptors on the attribute taste, 18 descriptors on the attribute aftertaste and nine descriptors on the attribute afterfeel. Rubbery texture and sweet, buttery and Gouda taste and aftertaste attributes were some of the important attributes that separated the cheeses in the study. The results from the descriptive profiling method suggested that the FCP method, which is less expensive and time consuming, is an appropriate technique when used with semi-naïve assessors. The relationship between consumer acceptability and descriptive sensory attributes of cheddar cheese were determined by using preference mapping (PM). The results from the PM indicated that two major (but very similar) consumer clusters, showed higher acceptance for cheeses with more âyoung/undevelopedâ attributes. The âidealâ Cheddar cheese, for the South African consumer, can therefore be described as having the following attributes: an aroma characterized by âsweetmilkâ, âGouda-likeâ and âbutteryâ attributes; a âshinyâ appearance and ârubberyâ texture; a mouthfeel characterized by ârubberyâ, âfattyâ, âtackyâ and âsoft- and/or hardnessâ attributes; a âGoudaâ, âbutteryâ, âsweetâ taste; a âGoudaâ, âbutteryâ, âsweetâ aftertaste; and an afterfeel that can be described as âfatty coatingâ and âoily/fattyâ. Results from this study seem to support and confirm speculations of some cheese specialists in the dairy industry who have perceived that Cheddar cheese, manufactured and available in the last few years, have showed/exerted more Gouda-like sensory attributes than typical known Cheddar-like attributes. Therefore, the results from the study indicated that data, obtained from FCP and using semi-naïve panellists, successfully expressed the perceptions of South African consumers on Cheddar cheese attributes.

Y. lipolytica has the ability to utilise hydrophobic hydrocarbons as carbon sources. It is also an attractive host for heterologous expression of cytochrome P450 (CYP) genes. Y. lipolytica strains with CYP genes cloned under control of two different promoters, pPOX2 and pICL were used in this study. The purpose of this project was to detect the effect of cloned alkane hydroxylases in Y. lipolytica. Alkylbenzenes and alkylcyclohexanes were used to compare the hydroxylase activities of the genetically engineered strains with control strains. Butylbenzene and hexylbenzene were transformed to phenylacetic acid while pentylbenzene, heptylbenene and nonylbenzene yielded benzoic acid as product. Butylcyclohexane and limonene were transformed to cyclohexylacetic acid and perillic acid, respectively, as major products. The activity towards hexylbenzene was highest. Phenylhexanoic acid and phenylbutanoic acid were also for the first time observed as intermediates in the biotransformation of hexylbenzene. Y. lipolytica strains expressing alkane hydroxylases under pPOX2 were induced with oleic acid and harvested. A strain with multiple copies of a proven alkane hydroxylase, cloned, had in one experiment higher activity than the other strains, towards both hexylbenzene and nonylbenzene. However, these results could not be confirmed, because in subsequent experiments the resting cells had very low activities. Ethanol and sodium acetate were used as inducers in the experiments conducted with Y. lipolytica strains with CYP557A1, a putative alkane and/or fatty acid hydroxylase, cloned under pICL. The substrates were added directly to the cells. With single addition of ethanol, the strain with cloned CYP557A1 had in one whole cell experiment with butylbenzene as substrate higher activity than the control strains. With multiple additions of sodium acetate the strain with cloned CYP557A1 showed higher activity in two shake flask experiments when hexylbenzene was used as a substrate. However, when ethanol and sodium acetate were used as inducers the alkylbenzenes were often consumed without the equivalent formation of detectable products, complicating the interpretation of results. This also happened in bioreactor experiments. Biotransformation of alkylcyclohexanes also did not demonstrate the effect of the cloned gene, since the activities of the test strain with CYP557A1 cloned under pICL and the control strain were similar. Alkylcyclohexanes are not promising substrates to distinguish between hydroxylase activity of the cloned genes as they are very volatile and activity towards them is relatively low.

Silk occupies a unique position as a textile fibre with a rare combination of beauty and strength. Production and processing of silk is labour intensive which leads to high cost and limited production of the silk fibre. Unfortunately the high cost of silk makes it unaffordable for many consumers; therefore mixed yarn fabrics could be constructed in order to lower the price of the fabric, without changing the unique properties of the silk negatively. The aim of this study was to evaluate and compare the properties of Gonometa postica silk fabric with the properties of mixed yarn fabrics consisting of Gonometa postica silk weft on a wool warp, and Gonometa postica silk weft on an acrylic warp. This is done in order to determine which of the wool or the acrylic create a more suitable mixed yarn fabric with the Gonometa postica silk. Standard methods were used to evaluate the abrasion resistance (ASTM 4966), tensile strength and elongation (ISO 13934), stiffness (BS 3356 ), crease recovery (AATCC 66), fabric thickness (BS 2544), dimensional change (AATCC 99) and moisture regain (ASTM 2654). Analysis of variance supported the interpretation of the results of the tests. The Gonometa postica silk textile fabric has relatively good abrasion resistance, with a mean value of 28 750 rubs necessary to break two yarns. The Gonometa postica silk weft/wool warp test fabric showed very good abrasion resistance with a mean value of 51 000 rubs required to break two yarns. And the Gonometa postica silk weft/acrylic warp test fabric also showed relatively good abrasion resistance, although it was lower than the other test fabrics with a mean value of 25 197 rubs needed to break two yarns. The Gonometa postica silk test fabric had the largest weight loss, while the Gonometa postica silk weft/wool warp fabric had the smallest weight loss. Tensile strength and displacement were measured and the Gonometa postica silk fabric had the highest mean maximum load necessary to break the silk weft yarns of 492.317 N and the mean displacement at maximum load was 39.048 mm. The Gonometa postica silk weft/acrylic warp fabric had the lowest mean maximum load that the silk weft yarns could carry before break at 347.910 N and the displacement was 34.465 mm. The bending lengths of all the samples were small enough to indicate that it has good draping qualities, considering the thickness of the fabrics. The Gonometa postica silk weft/wool warp fabric showed the best crease recovery especially in the warp direction (145°), while the Gonometa postica silk fabric had the worst crease recovery especially in the warp direction (128°). The Gonometa postica silk weft/Gonometa postica warp fabric was thinner than the Gonometa postica silk weft/wool warp fabric and the Gonometa postica silk weft/acrylic warp fabric. The moisture regain of the Gonometa postica silk fabric was found to be 13%, while the Gonometa postica silk weft/wool warp fabric had a moisture regain of 11% and the Gonometa postica silk weft/acrylic warp fabric had a moisture regain of 8.6%. The Gonometa postica silk weft/wool warp fabric and the Gonometa postica silk weft/acrylic warp fabric had no shrinkage in the warp directions. The Gonometa postica silk fabric showed more shrinkage in the warp direction than in the weft direction. No residual shrinkage was found. This lead to the conclusion that the wool would be the best fibre to mix with the Gonometa postica silk as it enhanced some of the properties of the silk, without influencing the properties negatively.

The poultry industry suffers severely every year due to bacterial and viral infections. Hence, great effort has been incorporated into isolating and identifying the different microorganisms that afflict poultry with their disease. The efficiency of the vaccines provided against such infections is only as good as the research of the responsible bacterium or virus. There are cases, however, when âunknownâ viruses create havoc within industry. This is mainly due to the a) virulent nature of the virus, b) frequency of infection and c) lack of knowledge about the virus. The unknown virus may be a truly novel virus or an existing virus that has mutated whilst replicating. It is thus important to refine the procedure of isolation and identification of the poultry viruses in order for outbreaks of an unknown virus to be dealt with quickly and efficiently. In this study, the isolation and passaging of two unknown virus samples as Newcastle disease virus (NDV) but failed to react with ND-specific PCR primers and did not haemagglutinate red blood cells were investigated These samples were two of a group of viruses submitted as NDV to the Veterinary Biotechnology Laboratory of the University of the Free State by the Onderstepoort Veterinary Institute. The aim of this study was to investigate into the identity of the two unknown virus samples, D1446/95 and 834/05, using various methods of identification. Procedures performed on this virus include:(presumed to be Newcastle propagation in embryonated SPF eggs and primary chicken embryo fibroblast (CEF) cell cultures; Disease Virus), purification of virus samples for ultrastructure identification (ultrastulu studies as by transmission electron microscopy (TEM) and negative staining;) serology work as per haemagglutination and mean death time (MDT) studies as well as restriction transcriptase polymerase chain reaction (RT-PCR) work. The virus samples were cultivated in embryonated SPF eggs via the allantoic cavity route. The embryo morality was observed and recorded, and the allantoic fluid containing the virus was successfully harvested. CEF primary cell cultures that were prepared in-house were inoculated with the individual samples separately. The virus samples showed cytopathic effects (CPE) after inoculation and incubation of the cell cultures. These CPE indicated that the virus samples infected the CEF monolayer and could be grown on CEF cell cultures as well. However, the cultivation of the virus samples in SPF eggs was preferred due to the difficulty that the primary cell cultures presented in a laboratory that is not directed at cell culture production and use. Through haemagglutination (HA) tests the virus samples were found to be unable to haemagglutinate chicken red blood cells (RBCs). It is speculated through this result that the virus samples are not classical NDV. However, it is entirely possible that the viruses agglutinate the RBCs of other mammalian or avian species. It is possible that these viruses may be evolved from classical NDV into, for example, Goose paramyxovirus The ultrastructure of the virus sample D1446/95 as observed through transmission electron microscopy, found an enveloped virus that measures about 100 nm to 150 nm in diameter. Although the ultrastructure of the virus does not conclusively identify the virus, it helps in preliminarily grouping it to other known viruses with similar ultrastructure and eliminating it from groups of viruses that have obvious differences in ultrastructure. Investigation into the virulence of the virus samples using the mean death time (MDT) indicated that the virus sample D1446/95 was moderately virulent with an MDT of 71,75 hours and the sample 834/05 was highly virulent with an MDT of 60 hours. As NDV ranges from being highly virulent (velogenic strains) to having low virulent (lentogenic strains) the MDT results were compared to the standard of the MDT of NDV. The standard of MDT for NDV states that velogenic strains take less than 60 hours to kill; mesogenic strains kill between 60 to 90 hours and lentogenic strains take longer than 90 hours to kill the SPF embryos. After the extraction of RNA from the harvested virus sample using TRIzol reagent, the viral RNA of D1446/95 and 834/05 were amplified using RT-PCR, primer sets designed based on Goose paramyxovirus ZJ1 and the Access RT-PCR System (Promega). The RT-PCR products were run and observed on 1% gels using gel electrophoresis. The RT-PCR products of the two virus samples resulted in multiple bands for the two virus samples that indicated that they are probably not classical NDV, but may possibly be NDV of other species perhaps even of goose NDV. However, this is still to be investigated further. Thus, through this study it was found that the unidentified virus samples D1446/95 and 834/05 are most likely not classical NDV, but rather could be a variation of classical NDV that may be branching off into paramyxovirus of other poultry or mammalian species as was found in the study of Goose paramyxovirus.

The genus Chryseobacterium was first proposed in 1994 and consisted of six species, it has since rapidly expanded and currently consists of 36 different species. Most of these species have been proposed within the last 3 years. Bacteria belonging to the genus Chryseobacterium are aerobic, non-sporulating, non-motile, Gram-negative rods which produce a yellow pigment. Chryseobacterium species have been found in a variety of environmental, industrial and clinical surroundings. Several species have already exhibited pathogenic characteristics for different animals and for humans as well. The first aim of this study was to evaluate the potential pathogenesis of 14 Chryseobacterium species by evaluating a variety of known virulence factors (enzymes), to evaluate their antimicrobial resistance patterns as well as to determine their resistance to four commercially available disinfectants. When the production of specific enzymes by the 14 Chryseobacterium species were evaluated, it was found that C. daecheongense, C. joostei, C. shigense, C. soldanellicola, C. taeanense, C. vrystaatense together with C. indologenes which was regarded the standard organism, can be regarded potentially pathogenic. The species most resistant to all of the antimicrobials evaluated, were C. indologenes, C. indoltheticum and C. joostei. The most susceptible to the antimicrobials was C. balustinum. It was also found that it would be best to treat an infection of Chryseobacterium with antimicrobials of the quinolone class or cephem class. The penicillin class would seem to be the least effective. The disinfectant resistance was determined by determining the MIC for each tested disinfectant, with a higher MIC exhibiting more resistance. The species with the highest resistance were C. gleum, C. indologenes and C. joostei. The disinfectants tested were for use either on hard surfaces or on wounds (personal use). From the results obtained disinfectants for use on hard surfaces are more effective than those meant for personal use. In conclusion it can be said that C. gleum, C. indologenes and C. joostei may be regarded as the most pathogenic of the 14 Chryseobacterium species evaluated. Chryseobacterium joostei may be less of a threat as it was first isolated from raw milk, which had not yet undergone pasteurization and as such, is less likely to come into contact with the general population. It would seem that C. gleum and C. indologenes however come into contact with people more often as they have been found in or on ill or wounded individuals on several different occasions. The second aim of the study was to evaluate possible degradation of aflatoxin B1 by the 14 Chryseobacterium species. After statistical analysis, it was concluded that no Chryseobacterium species was able to degrade aflatoxin. This study, however contributed to the knowledge on the significance of these bacteria.

A great deal of research in cheese technology is devoted towards the manipulation of glycolysis, lipolysis and proteolysis in order to accelerate ripening times or to improve flavour. There is increasing evidence that certain yeast species contribute to flavour and texture development during ripening of certain cheeses, including Cheddar cheese. The addition of yeast cultures as adjunct cultures may accelerate ripening, or lead to a faster development of flavour and taste. This is brought about by an increase in primary proteolysis of the caseins as well as the break down of amino acids into flavour compounds. The respective enzymes involved seem to be proteases with plasmin or plasmin activator activity, and glutamate dehydrogenase. This research reports on the inclusion of yeasts as adjuncts in the processing of Cheddar cheese. Yeasts that expressed both enzyme activities were added as single adjuncts, while combinations of yeasts with singular enzyme activity were employed. Changes during ripening were monitored by sensory and biochemical analyses. The latter includes proteolysis, lipolysis and enzyme activity.

1. SUMMARY Bleach plant effluent obtained from the alkali extraction stage (Eo) of a dissolving sulphide pulp mill was chemically characterised by various methods. It was necessary to examine the chemical nature of the bleach plant effluent to devise effective treatment methods. Routine chemical analyses of effluent included: alkali-lignin, COD and pH determinations. Additional measurements involved physico-chemical techniques such as adsorption spectra analysis and determination of colour intensities of the wastewater. Effluent used during the biological treatments was studied in greater detail and analysis also included: AOX (absorbable organic chlorine), chlorophenol, gel permeation chromatography, nitrogen, phosphate and sugar determinations. Effluent collected at different periods from the bleach plant of the pulp mill usually differed significantly in chemical composition. Effluent employed during biological treatment studies contained trace amounts of nitrogen as well as carbohydrates however, no ortho phosphate was detected in the wastewater. Overall, the pH of the various effluents studied was in the alkaline range, colour intensities varied between 5 500 PCU to as high as 15 700 PCU and COD levels of 7 061 to 14 567 mg/l were measured. These methods provided crucial information that was essential for the planning of experiments conducted during this study. Furthermore, it was also used to measure the efficiency of the treatments methods applied.