An enzyme-focused approach to create malic acid biosensors for winemaking

The tart-tasting malic acid is one of the key flavour determinants in wine. In winemaking, malic acid levels within grapes are carefully monitored prior to harvesting, as well as during secondary fermentations that aim to reduce its concentration. Commonly employed methods of measuring malic acid are expensive, consume large amounts of reagents, and require specialist laboratories and personnel to perform. Best exemplified by blood glucose monitors, enzyme-based amperometric biosensors are a solution to each of these problems, being easy to use, low-cost, and able to be integrated into miniature, portable devices. For measuring malate, these sensors use a malate oxidoreductase enzyme, which catalyses the oxidation of malate and concomitant reduction of a co-factor. When paired with an electrode, this electron transfer reaction can be transduced into a measurable electrical current. Approximately 40 enzyme-based, amperometric malate sensors have been described in the peer-reviewed literature since the first described in 1980. Much of the work into improving devices has focused on the use of specialised materials or chemical additives to enhance electron transfer to the electrode. However, in general, little attention has been given to the choice of malate oxidoreductase used, with most devices using one of three commercially available enzymes. I hypothesised that the selection of a bespoke enzyme can be used to obtain desirable performance characteristics and afford novel utility in a biosensor. As the proof of principle, my choice of enzyme aimed to solve an issue that is particularly prevalent in the malate biosensing literature: the linear ranges of devices are poorly suited for the needs of winemakers. As these ranges are often extremely narrow, a winemaker must perform a careful and accurate dilution series to enable measurements. A device with a wider linear range would be more convenient for the end user. Candidate malate oxidoreductases, characterised in the peer-reviewed literature, were selected based on their reported Michaelis-Menten parameters. I specifically focused on the Michaelis constant (KM), a parameter that governs the hyperbolic relationship between increase in substrate concentration and enzyme saturation, theorising that an enzyme with a large KM would yield a device with a wide linear range. Seven candidate enzymes were expressed, purified, and then characterised using spectrophotometric assays. The R181Q mutant of the malic enzyme from Ascaris suum was taken forward for the project. This enzyme has the highest KM for malate of all characterised malate oxidoreductases and a mechanism to modulate the Michaelis-Menten parameters, including KM, by adding different concentrations of ammonium to assays. An amperometric device was built using this enzyme in solution, on a gold screen-printed electrode. Six electron mediators were tested for their ability to increase the rate of electron transfer from the reduced co-factor, NADH, to the electrode surface. The aim was to produce the highest signal with the lowest operating potential, though it was ultimately found that many were incompatible with the sensor assay, being either insoluble, unstable, or the cause of electrode fouling. Conversely, 2,6-dichlorophenolindophenol proved compatible with the other assay components and gave relatively large signals at the relatively low potential of +150 mV. The device created reports on malic acid concentrations up to 200 mM, the maximum concentration seen in grapes, and has the widest linear range of any enzyme-based malate biosensor described to date. The addition of ammonium to assays was used to increase the sensitivity of the device. However, two other small molecules were also found that were capable of modulating the sensor response: Mn2+ and citrate. These three compounds, added in various combinations, were used to generate three different calibration curves. In this way, a novel type of biosensor was created with the previously undescribed utility of a linear range versus sensitivity trade-off: the sensor could be dynamically made more sensitive if required, making it better at resolving small differences in malic acid, though each increase in sensitivity came with the reduction in the linear range. The device holds value in being operationally flexible to accommodate the needs of winemakers seeking to measure both large and small differences in malic acid. The device was tested by applying a range of Pinot Noir and Pinot Gris grape juice taken from different stages of the winemaking process of grape and wine samples, with measurements being compared to two commercial malic acid testing kits. The sensor consistently produced current readings 5- to 6-fold higher than expected, based on the results of the commercial kits. This was shown to be due to grape juice and wine interfering with both the enzyme catalysed reaction and the mediator, resulting in non-specific current. Many of the compounds found within wine were tested for interfering effects, with ascorbic acid being identified as a potent interferent capable of reducing the mediator. Attempts to pre-treat real samples with chemical additives to remove ascorbic acid were unsuccessful, though an approach using an ascorbate oxidase remains to be tested. This work represents a successful proof of the concept that careful selection of a bespoke enzyme can yield tremendous benefits to an amperometric biosensor, both in terms of the sensor performance and the addition of previously unseen utility. The presence of comprehensive databases, coupled with the current convenience of gene synthesis, suggests that the field of malate-sensing has much to gain from the search for new and better enzymes to be incorporated into devices.