Increasing scrutiny of the negative ecological and environmental effects of chemical surfactants is driving research into the production and use of biosurfactants that offer low toxicity and high biodegradability. One particular form, sophorolipids, are a strong contender owing to their good surface activity and high production by Starmerella bombicola. Currently, market entry is hindered by high costs linked to the use of first generation âhigh-gradeâ feedstocks including glucose and rapeseed oil, that also have significant carbon footprints associated with their use. Explorations into low-cost, eco-friendly alternative feedstocks are underway, but the current approach has by and large been unsuccessful, due in part to the limited approach currently used to screen feedstocks, where feedstocks are tested under generic conditions without consideration of their specific composition. Similarly, analytical methods used to identify and quantify sophorolipid variants vary greatly across the literature and are poorly documented and understood, with no clear comparison of their effectiveness. The work presented in this PhD thesis centres on the development of a feedstock screening platform for alternative feedstocks with Starmerella bombicola, focusing on a fair screening process with a comparison of the most suitable analytical methods. Firstly, a predictive tool that defines media conditions for a feedstock based on its composition was constructed. Secondly, appropriate methods for sample preparation, identification and quantification are explored and developed. The thesis is presented in a Journal (Alternative) Format and is organized into four main research articles. To identify suitable conditions for sophorolipid production, a central composite design of experiments was applied to alter the concentrations of glucose, rapeseed oil and nitrogen (cornsteep liquor/ammonium sulfate). Maximal production of sophorolipids was achieved at 39.5 g/L, improving on the 19.3g/L gained in a commonly cited literature composition. Glucose was not found to have a significant effect on sophorolipid production, which was linked to the production and subsequent consumption of glycerol via rapeseed oil breakdown during the fermentation. Inversely, nitrogen was found to be largely inhibitive to the production process, as increased levels diverted S. bombicola to growth of biomass over sophorolipid production. The work significantly improved the understanding of how S. bombicola utilises feedstocks in the process of sophorolipid production and novelly describes the role of nitrogen and glycerol in the production process. The subsequent predictive model created from the design of experiments results was applied to predict appropriate conditions for wood-derived sugar solutions provided by an industrial manufacturer, determining optimal conditions for production alongside a high-grade feed internal control for comparison. The potential of a C6 sugar rich feed was noted, with 44.3g/L final sophorolipid production, demonstrating a significant improvement compared to the use of pure glucose (29.00 g/L sophorolipid) at identical media concentrations. Investigations into the potential composition of the C6 sugar feeds found acetic acid consumption linked to the improved production levels, which has not been previously reported. Feedstock preparation methods were compared, with tangential flow filtration (TFF) and autoclaving, with TFF allowing an otherwise un-productive feedstock to achieve 40.15g/L sophorolipids, marking a novel method of preparation that can significantly improve the potential of feedstocks. To ensure accurate quantification of the sophorolipids produced by alternative feedstocks, three common analytical methods (anthrone assay, liquid:liquid extraction/gravimetric and high performance liquid chromatography) were compared on 75 unique fermentation broth samples with varying compositions of glucose, rapeseed oil and nitrogen. The HPLC was shown to be capable of high specificity and sensitivity (0.3 g/L detection limit) towards sophorolipids, whilst the liquid:liquid extraction/gravimetric method was not capable of detecting below 11.06g/L quantity, even when no sophorolipids were detected with the HPLC method, demonstrating the significant lack of specificity and co-extraction of non-sophorolipid materials from the fermentation broth. Similarly, the anthrone assay was shown to have no linearity (R2 = 0.129) and was found to cross-react with media components (rapeseed oil, cornsteep liquor, glucose) leading to over-estimation of the concentration of sophorolipids in the sample. Sample preparation to acquire a suitably representative sample was also explored, with the addition of ethanol leading to enhanced recovery (113.67% improvement) and appropriate pelleting of residual biomass. The novel improvements and comparison of the quantification techniques in this work marks a significant contribution to the current literature, allowing for researchers to better understand and select appropriate techniques for use. To ensure better understanding of the final material produced by the fermentation process, suitable identification techniques such as LC-MS must be able to distinguish between different structural variants with confidence. The development and application of a liquid chromatography mass-spectrometry (LC-MS) and tandem mass spectrometry (MS/MS) method was explored, identifying 37 variants, including 15 novel sophorolipid forms not reported in the literature. To identify these sophorolipid forms, the fragmentation characteristics of tentative candidate peaks under MS/MS were explored with the in-silico computational fragmentation tool MetFrag, identifying key structural fragments that can be used to identify different variants, including the details on distinguishing the positional isomers of C6â and C6ââ mono-acetylated forms. New fragmentation pathways were shown under positive and negative ionisation, marking the first addition to sophorolipid fragmentation in nearly 30 years. The elution order of sophorolipid congeners under liquid chromatography was explored, with a clear demonstration that acidic sophorolipids are capable of eluting after lactonic forms when other structural characteristics are applied, disproving the common belief found in the literature. The culmination of this research is the release of the first publicly available database of sophorolipid MS2 spectra, marking a significant tool to accelerate the development of sophorolipid products. Overall, the research findings contained in this thesis significantly advance the field of biosurfactant research by developing and proving the utility of a robust and comprehensive platform for screening alternative feedstocks for sophorolipid production. The integration of experimental design, predictive modelling addresses concerns around the utility of current screening approaches, providing a fair platform to compare alternative feedstocks. Similarly, the research and development of analytical methods provides an invaluable resource for researchers to utilise in order to allow greater accessibility and appropriate method selection.
| Date of Award | 17 Jan 2025 |
|---|
| Original language | English |
|---|
| Awarding Institution | - The University of Manchester
|
|---|
| Supervisor | Peter Martin (Supervisor) & James Winterburn (Supervisor) |
|---|