Enzymatic modification and potential bioactivity of diferuloyl- and feruloyl-substituted arabinoxylan

Corn bran is an abundant coprocessing stream of corn-starch processing, rich in highly substituted, di-feruloyl cross-linked arabinoxylan. The feruloyl moieties are generally attached at the O5 position of the arabinose of arabinoxylan chains. In addition, the arabinoxylan chain can be cross-linked by diferulic acid (diFA), mainly as 8-5’ diFA, 5-5’ diFA, 8-O-4’ diFA, 8-5’ benzofuran diFA and 8-8’ diFA. These di-feruloyl cross-links enhance the complexity of glucuronoarabinoxylan, making it more recalcitrant to enzymatic conversion, and thus constitute a hindrance to designing effective enzymatic upgrading of corn glucurono-arabinoxylan. Feruloyl esterase (FAE) (EC 3.1.1.73) is an enzyme that can catalyze the cleavage of the ester bonds between feruloyl moieties and arabinose to release ferulic acid (FA) from arabinoxylan. The most common FAEs belong to carbohydrate esterase family 1 (CE1) in the Cazy database, but recently some tannases (EC 3.1.1.20) have also shown good FAE activity. Study of the enzymatic release of diFAs is still limited, and previously reported enzymatic release of diFAs from complex arabinoxylan was low in efficiency and normally required the presence of endo-1,4-β-xylanase.

Arabinoxylan has been reported to have good prebiotic effect, which may be due to the complex structure of arabinoxylan. High feruloylation in arabinoxylan may promote growth of beneficial bacteria in the gut and promote production of short chain fatty acids (SCFAs). Hence, feruloylation of arabinoxylan is likely to beneficially modulate the human gut microbiota community. However, due to lack of well-defined feruloylated arabinoxylan, the understanding of the structure-function relationship between feruloylated arabinoxylan and prebiotic effect is limited.

Therefore, in this PhD Thesis work, one of the major aims was to discover novel FAEs that can cleave di-feruloyl cross-links from complex di-feruloyl cross-linked glucuronoarabinoxylan (FGAX), and to expand knowledge on FAE enzyme functions. The other aim was to investigate how soluble FGAX modulates the human gut microbiota in vitro and determine whether level of feruloylation of FGAX influences the potential prebiotic effects.

Several types of methyl hydroxycinnamate substrates have been widely used to study the specific activity and substrate specificity of FAEs. However, the results obtained in the present PhD study indicates that activity data measured on methyl hydroxycinnamate substrates do not accurately reflect the specific activity and substrate specificity of FAEs since different results were observed when using the corresponding arabinosyl hydroxycinnamate substrates. Furthermore, when kinetic parameters of release of FA by FAEs were measured on the soluble FGAX substrate, the results furthermore indicated the enzyme activity on methyl ferulate did not reflect the activity on the more realistic and complex arabinoxylan substrates. According to the kinetic parameters, fungal FAEs AnFae1, AoFae1 and MgFae1 were considered to be more efficient compared to the other studied FAEs.

To further examine the ability of FAEs to release diFAs, soluble FGAX substrate was used together with an HPLC-MS based analytical method. Although many FAEs were able to release detectable levels of diFAs after exhaustive enzymatic hydrolysis of FGAX, only AnFae1, AoFae1, MgFae1 and the two bacterial FAEs wtsFae1A and wtsFae1B could release diFAs with relatively good efficiency. The kinetics parameters of release of diFAs by these FAEs were measured with FGAX, and results indicated that AnFae1 and AoFae1 had a higher kcat for release of 8-5’ diFA and 5-5’ diFA compared to other studied FAEs. However, only MgFae1, wtsFae1A and wtsFae1B could release 8-O-4’ diFA and 8-5’ benzofuran diFA. AnFae1 and AoFae1 may have a preference for release of 8-5’ diFA and 5-5’ diFA, while MgFae1 preferentially releases 5-5’ diFA. wtsFae1A showed no significant discrimination among the four types of diFAs while wtsFae1B appear to preferentially catalyze release 8-5’ benzofuran diFA.

The structure-function analysis of FAEs described here may help in understanding different enzyme activities of the studied FAEs. The molecular docking of FAEs with di-feruloyl cross-linked substrate may contribute to understanding of the ability of FAEs to release diFA as well as perhaps help explain any possible preferential release of different diFAs by these enzymes. Additionally, the CBM domain in MgFae1, wtsFae1A and wtsFae1B may be involved in the release of diFAs.

Soluble FGAX samples were able to modulate the human gut microbiota community in vitro and produce SCFAs. To understand the effect of feruloylation level on prebiotic effects, FGAX was modified to produce a highly di-feruloyl cross linked FGAX sample and a non-feruloylated FGAX sample which were fermented in vitro. Despite a significantly high level of di-feruloyl cross-links in FGAX, fermentation of the highly cross-linked FGAX was not hindered and showed a prebiotic effect similar to that of the original FGAX, whereas non-feruloylated FGAX appeared to have arelatively lower prebiotic effect. All three kinds of FGAX samples could promote the relative abundance of some health- promoting gut bacteria while decreasing the relative abundance of harmful gut bacteria, which further indicates that FGAX is a prebiotic candidate.

In conclusion, the thesis reports several novel FAEs that can cleave di-feruloyl crosslinks from complex arabinoxylan substrates. Further, the present work indicates that FGAX has a prebiotic effect and that feruloylation of FGAX may also have an effect. The results obtained may help pave the way for upgrading of recalcitrant corn bran arabinoxylans and contribute to the application of FGAX in healthy food supplements.