Malondialdehyde (MDA).

Quantifying levels of malondialdehyde (MDA) is performed to assess lipid peroxidation in a wide range of disciplines from food preservation [13] to medicine [14]. In ecological studies, plasma MDA levels are commonly used to estimate the overall oxidative damages experienced by an individual [5,7,10,15]. Traditionally, MDA levels were quantified using the thiobarbituric acid reactive substances (TBARS; [16]), an assay that is still commonly used in animal physiology [17–19]. Thiobarbituric acid reacts with tissues and create a colored adduct that can be quantified by spectrophotometry.

Although being a relatively simple assay, TBARS has been readily criticized for its lack of specificity to MDA. Thiobarbituric acid can react with a wide range of components present in the plasma, such as aldehydes, sugars and urea [20–22]. Thus, external factors like diet or nitrogen elimination can confound assay values, leading to an overestimation of MDA levels up to 3 folds [23,24]. A greater specificity to MDA can be achieved with the TBARS assay by injecting the reacted sample into a high-performance liquid chromatograph (HPLC) [15,23,25,26]. However, HPLC methods for plasma MDA detection are mainly developed in biomedical studies where blood sampling volume is not a limiting factor, requiring plasma volumes ranging from 50 μL to 100 μL [23,27,28]. Therefore, approximately 200–400 μL of whole blood would have to be collected from an individual to enable replicated measurements of MDA levels. This amount of blood may require sampling procedures that are too invasive for conservation studies or for longitudinal studies of smaller species, preventing the measurement of multiples assays on a single sample [5].

Also, the MDA generated in the cell quickly binds to amino acids and proteins due to its high reactivity, making it present in both free and bound states in tissues. Free MDA readily reacts with TBA while bound MDA must return to its free form before reacting. Although most bound MDA becomes free during the derivatization process [27], it appears crucial to thoroughly liberate all the bound MDA since it may represents about 90% of all the MDA present in a sample [26]. Adding an additional alkaline hydrolysis step allows a more complete and uniform release of bound MDA, thus yielding better measurements [26,29].

Moreover, high-throughput methods can be necessary for larger-scale studies, since MDA levels in stored plasma samples [25] or processed samples [27,30] can degrade over time. Therefore, optimizing the retention of the compound while using the shortest possible runtime could allow to process hundreds of injections per day on a single HPLC system.

Ferric reducing/antioxidant power (FRAP).

The ferric reducing/antioxidant power (FRAP) assay has been used extensively in various disciplines ranging from agronomy and nutrition [31,32] to ecological studies [8,9,17]. Antioxidants in the samples reduce ferric-tripyridyltriazine (Fe³⁺-TPTZ) complex to ferrous tripyridyltriazine (Fe²⁺-TPTZ), producing a blue colour. The coloration intensity is proportional to the reducing capacity of the mainly non-enzymatic antioxidants in the plasma, hence providing the overall antioxidant capacity. The FRAP assay is a simple, quick and inexpensive test which already requires fairly small blood volumes [17,19,33,34].

As for other antioxidant capacity test like the TAS/TEAC assay [35], FRAP values are strongly correlated with uric acid levels, which could bias the antioxidant power estimate [33]. It may be difficult to decipher whether fluctuations in uric acid levels are due to other incidental metabolic processes, like an increase of amino acid catabolism, or a change in antioxidant regulation [35,36]. Furthermore, differential selective pressure among species may have operated to induce variation in concentration and dynamics of uric acid [36]. In recent studies, its level is often assayed in parallel with antioxidant assays and statistically accounted for [37,38]. However, this approach requires the use of an extra blood assay kit which may add measurement errors to the data and reduce the amount of blood available to other assays.

An approach to remove uric acid from a plasma sample has been previously proposed by pre-treating the plasma with great excess of uricase to specifically eliminate uric acid [39]. This approach spares the need of using an extra blood assay kit but high plasma volumes are usually required (e.g. 200 μL; [40]. Here we optimized the amount of uricase added to fully remove all uric acid from the samples and we modified the incubation and assay protocol for smaller blood volumes.

Hypochlorous acid shock capacity assay (HASC).

In biological studies, the HASC assay is commonly referred to by its kit name OXY-adsorbent test^TM from Diacron International (Grosseto, Italy). This total antioxidant test is popular in animal ecological studies [5,41] and it is also used in agronomy [42,43] and neurophysiology [44]. The test works by incubating a plasma sample with an excess of hypochlorous acid, a strong oxidant. The hypochlorous acid oxidizes any reduced components in the plasma, then a chromogenic dye is added to quantify the remaining hypochlorous acid. The original concentration of hypochlorous acid minus the remaining quantity provides the total antioxidant capacity of the sample. The HASC assay is not correlated with uric acid levels [11]. Because of its general mechanism and the non-specificity of the pro-oxidant reagent, this assay allows most of the antioxidant mechanism to mitigate plasma oxidation. However, hypochlorous acid is not the substrate of the antioxidant enzymes, which are not contributing to its neutralisation. Therefore, the HASC assay measures total non-enzymatic antioxidant capacity. Comparative studies between HASC and FRAP have found no correlation [11] or weak positive correlation [45], suggesting that the two assays are complementary and may generate a more complete picture of the total antioxidant capacity when used together [11,45].

OXY-adsorbent test^TM requires small plasma/serum volumes (10 μL) which allows to simultaneously measure other oxidative stress or physiological components. However, with large sample sizes and/or longitudinal studies, these kits can prove to be expensive. Because of the simplicity of the assay, we propose a low cost in-house version of this assay that is specifically adapted for small plasma volumes.

MDA reagents and system settings.

HPLC grade acetonitrile was purchased from VWR Int. (Mississauga, ON, Canada). A license to procure and possess 2-thiobarbituric acid was obtained from Health Canada (authorization #43305.06.17). The 2-thiobarbituric acid was purchased from EMD Millipore (Billerica, MA, USA). Malondialdehyde bis (diethylacetal), butylated hydroxytoluene, trichloroacetic acid and trifluoroacetic acid were purchased from Sigma-Aldrich (Oakville, ON, Canada). All other reagents used were of analytical grade unless otherwise noted. Ultrapure water was made in-house using a Barnstead NANOpure water system (Resistivity ≥ 18.2 MΩ • cm^-1).

MDA was determined using an Agilent Technologies 1100 HPLC system consisting of an autosampler (G1313A), a binary pump (G1312A), a thermostatted column compartment (G1316A) and a variable wavelength detector (G1315A) (Santa Clara, CA, USA). The “OpenLab Chromatography Data System” software (Agilent Technologies) was used to control the instrument and collect the data. Analytical separations were performed using a Varian ODS2 250x4.6 mm column at 30.0 ^oC. The mobile phase was an isocratic mixture of 0.1% TFA in water: acetonitrile (70:30 V/V) at a flow rate of 1.00 mL min^-1. The backpressure of the system was 165±5 BAR. The injection volume was 20 μL. Detection was at a wavelength of 532 nm. The runtime of the method was 4 min and the retention time of the TBA₂-MDA adduct was 3.00±0.05 min.

MDA analysis for 20 μL of plasma.

The MDA quantification protocol was adapted from [23]. We added 2 μL of butylated hydroxytoluene (0.1M) in anhydrous ethanol and 5 μL of NaOH (2.0 M) to a 20 μL aliquot of plasma or standards (see below), before being placed in a shaking dry block incubator set at 60 ^oC for 30 min and 400 RPM. To precipitate proteins, 100 μL of aqueous trichloroacetic acid (15% V/V) was added. The samples were then briefly vortexed, placed in an ice bath for 5 minutes, then centrifuged for 10 min at 14,000•g and 4 ^oC. The supernatant (100 μL) was transferred to a screw-cap 500 μL microcentrifuge tube and 50 μL of 2-thiobarbituric acid (0.375% w/v) as a solution in HCl (0.25 M) was added. The tubes were capped, vortexed, and placed in a dry bath incubator at 100 ^oC for 60 min. They were then cooled on ice and centrifuged for 5 minutes at 14,000•g and 4°C, before the resulting derivatized solution (60 μl) was transferred to a 250 μL microinsert HPLC vial for analysis.

For calibration, a standard stock solution was prepared daily by diluting 1.00 mL of 1,1,3,3-tetraethoxypropane (TEP) (25 μL into 1.00 mL of anhydrous ethanol; 100 mM) by a factor of 1:100 with ultrapure water. Standards were prepared through serial dilution of the stock standard (1 mM) with ultrapure water to obtain concentrations of 20, 15, 10, 5, 2.5, 1.25, and 0 (blank) μM TEP. Quality control (QC) samples were prepared by pooling 5 μL of each derivatized sample in a single HPLC vial. The QC sample was injected every twenty sample injections, and again at the end of the series run. Standards were measured from triplicate injections. Samples were prepared in duplicate and measured from duplicate injections.

System settings.

All spectrophotometric assays for microplate analyses were conducted in standard flat bottomed 96 well plates (P/N 82.1581 Sarstedt AG & Co. KG, Nümbrecht, Germany) Analyses was carried out on an iMark Filter absorbance microplate reader with computer control and data analysis by Microplate Manager 6 Software (Bio-Rad Laboratories Canada Ltd., Mississauga, ON).

Uric acid independent FRAP analysis for 5 μL of plasma.

The FRAP assay protocol was adapted from [39]. The amount of uricase added was reduced to 5 μL of 1U/mL after calculating the expected physiological range of uricase in plasma and the theoretical rate at which the uric acid could be removed by uricase. 1U/mL uricase was compared to 10U/mL of uricase as used by [39], with no observable difference in FRAP levels.

A 5 μL aliquot of plasma or standards (see below) was mixed with 5 μL ultrapure water-containing uricase (1.00 U/mL) in a well of a 96 well plate and incubated for 5 min at room temperature on a 2D shaker. Then, 200 μL of FRAP reagent was added to each well. After a 30 min incubation at room temperature with regular shaking in the spectrophotometer, absorbance at 593 nm was read vs. reagent blank. Calibration curves were generated from a daily prepared standard stock FeSO₄ solution (30mM), serially diluted to 3.0 mM, 2.0 mM, 1.0 mM, 0.5 mM and 0 mM (blank) using ultrapure water. FRAP reagent was freshly prepared each day by mixing 20.0 mL acetate buffer (pH 3.6, 300 mmol/L), 2.00 mL 2,4,6-Tris(2-pyridyl)-s-triazine (TPTZ) (10 mmol/L) in HCl (40 mmol/L) and 2.00 mL FeCl3 (20 mmol/L). All analyses were conducted in duplicate.

HASC analysis for 1 μL of plasma.

An aliquot of plasma was added to a well of a 96 well plate containing 100 μL of ultrapure water and the plate was briefly vortexed on a 2D shaker. Then, 100 μL of oxidant solution (for samples) or standard solutions (see below) were added, followed by a 10 min incubation at room temperature on a 2D shaker. After that, 20 μL of a stock N,N-dimethyl-p-phenylenediamine (NPD) solution (83 μL diluted to 50mL in anhydrous ethanol) was added to every well, followed by a 1 min incubation on the 2D shaker, prior to reading the absorbance of the wells at 515 nm.

NPD solution was freshly prepared each day and protected from daylight. The oxidant solution was prepared daily by diluting 10% sodium hypochlorite by a factor of 500 and adjusting the pH to 6.2 using 0.6 M sulfuric acid. The concentration of hypochlorous acid prepared was determined by measuring the absorbance at 292 nm in a 1 cm pathlength quartz cuvette and using the Beer-Lambert law with an absorptivity coefficient of 350 L mol^-1 cm^-1. Calibration curves were generated by serial dilutions of the oxidant solution to factors of 2X, 4X, 8X, 16X, 32X with ultrapure water. The reagent blank was prepared using ultrapure water. All sample measurements were run in duplicate.