Author: Zosia Anna Marie Zielinski
Publisher:
Published: 2021
Total Pages:
ISBN-13:
DOWNLOAD EBOOK →
Oxygen causes the degradation of virtually all hydrocarbons over time through a process called autoxidation: it's the reason engine oil breaks down, plastics get brittle over time, and butter turns rancid. Autoxidation, in many ways, is what causes humans to degrade as well; the process has been linked to aging, cancer, cardiovascular disease, asthma, macular degeneration, and neurodegenerative conditions like Alzheimer's, Parkinson's, and multiple sclerosis. Unsurprisingly, the body has many natural antioxidant defences to combat oxidative stress, and many so-called "superfoods" are dense in antioxidants such as: vitamins A, C, and E, found in fruits and vegetables like leafy greens, citrus, and avocado; the (poly)phenolic compounds found in berries, green tea, and red wine; and the organosulfur compounds found in garlic and onions. As such, there has been significant interest in this area of research. Our understanding of the kinetics and mechanisms of autoxidation, and the antioxidants that slow the process, has advanced significantly over the past few decades; however, some questions remain unanswered, and some widely-accepted beliefs warrant revisitation. One such widely-accepted belief was that the autoxidation of cholesterol yields only a single regiosomeric hydroperoxide product (the cholesterol 7-hydroperoxide). We have shown, as outlined in Chapters 2 and 3, that the mechanism of cholesterol autoxidation is far more complex than previously appreciated. Indeed, cholesterol 4-, 6-, and 7-hydroperoxides are produced in the free-radical process, and an additional regiosomeric product, cholesterol 5-hydroperoxide, is produced when autoxidation occurs in the presence of a good H-atom donor. We have also demonstrated that the pathogenic secosterol compounds linked to a number of degenerative diseases can arise from the Hock fragmentation of cholesterol 5- or 6-hydroperoxide, and heretofore uncharacterized secosterols may arise from a similar transformation in cholesterol 4- and 7-hydroperoxide. This, taken together with the well characterized pathogenic potential of such cholesterol-derived electrophiles, is excellent evidence for a link between cholesterol autoxidation and degenerative disease, without invoking high-energy oxidants like singlet oxygen or ozone. Our mechanistic study of cholesterol autoxidation also led to the discovery of some surprising stereoelectronic and quantum mechanical effects. In Chapter 3, we show that cholesteryl acetate, a model for the esterified cholesterol particularly abundant in lipoproteins (e.g. LDL and HDL), autoxidizes at 4 times the rate of free cholesterol, which is more abundant in lipid bilayers. We also characterized the pathway of peroxyl radical addition to cholesterol, resulting in cholesterol epoxides. This pathway could be favoured by deuterating the allylic positions in cholesterol to suppress H-atom abstraction by the peroxyl radical, resulting in a deuterium kinetic isotope effect of 20. This evidence of quantum mechanical tunnelling in the H-atom transfer pathway, along with some recent reports of such an effect in other lipids, led us to further explore the impact of tunnelling on H-atom transfer reactions more generally. Using a series of computational model systems, as well as an experimental model for the simplest unsaturated lipid, oleic acid, we demonstrated that tunnelling in C-H abstractions by peroxyl radials is likely significant across the board, so long as the barriers to such reactions are sufficiently high. We also suggest that oleate-derived epoxides may have been overlooked as (minor) products of its autoxidation, and the epoxides derived from plasmalogen lipids may be even more significant than the corresponding hydroperoxide products. Throughout our work, we have supplemented experimental evidence with computational investigations. A common theme to the computations has been the influence of secondary orbital interactions in H-atom transfers to peroxyl radicals, when the substrate has a secondary orbital donor (lone pair or p-system) adjacent the position bearing the labile H-atom, which can interact with the internal peroxyl oxygen atom. Our work comparing such an effect in sulfenic and selenenic acids (RSOH vs. RSeOH), as outlined in Chapter 5, is some of the best experimental evidence for the influence of secondary orbital interactions. The hindered triptyceneselenenic acid reacted only 18-fold more slowly with peroxyl radicals than the corresponding triptycenesulfenic acid, despite its O-H bond being 9 kcal/mol stronger! Furthermore, we show that the reaction of peroxyl radicals with unhindered selenenic acids is predicted to be over an order of magnitude faster than their sulfenic acid counterparts, owing to the influence of proper geometrical alignment on the secondary orbital interaction, which is hindered with the bulky substituent. The influence of secondary orbital interactions is further explored in Chapter 6, where we evaluate the feasibility of quantifying or predicting the effects thereof on H-atom transfer reactivity across more diverse classes of compounds. While the many confounding stereoelectronic factors present in these transition states makes such an endeavour difficult, this work has underscored the significance of computations in helping to explain seemingly anomalous experimental evidence.
