Non-invasive determination of uric acid in human saliva in the diagnosis of serious disorders

Chromatography

Depending on the type of detector, HPLC is highly sensitive and reproducible analytical technique enabling the separation of various analytes and biomarkers. For the determination of salivary UA HPLC is usually combined with sensitive detection techniques such as mass spectrometry (MS), electrochemical (ED), and fluorescence detection. The separations were most often carried out using reverse phase mechanism. The stationary phase of choice was octadecyl bonded silica (C₁₈) with a particle size of 5 μm. The HPLC columns had a length of 150–250 mm and an internal diameter of 2.0–4.6 mm [4], [10], [17], [56]. Poroshell EC-_C18 column (3.0 × 50 mm, 2.7 µm) were also used [55]. The mobile phases were phosphate buffer with pH adjusted using trichloroacetic acid and sodium citrate in combination with methanol and used in isocratic elution [4], [10], [16], [17] although a gradient elution was also applied [55]. UA is a polar analyte with pKa equaling 7.25 and logP=1,544 that enable achieving good retention in acidic mobile phases with no or very small percentage of an organic modifier to be used. The newest version of HPLC, the ultra-high-performance liquid chromatography (UHPLC), was not reported for the determination of salivary UA yet. This highly efficient method relies on use of columns packed with sub 2 μm particles that expedite the analysis time. The HPLC methods used for the determination of UA in saliva are collected Table 2. In general, all the methods required a complex sample pre-treatment technique and were used with small cohorts of patients.

Inoue et al. investigated detection methods allowing the determination of salivary UA using reverse phase HPLC. and compared ultraviolet (UV), amperometric (Ampero-ED), and coulometric (Coulo-ED) detection. A C₁₈ column (150 × 2.0 mm, 5 μm), isocratic elution with potassium phosphate buffer (74 mmol/L, pH 3.0) mobile phase, a flow rate of 0.2 mL/min, an injection volume of 20 μL, and a total run time of 10 min were used. The limits of quantification (LOQ) using UV detection at 284 nm were 0.6 μmol/L, while it was 0.01 μmol/L for Ampero-ED and 0.02 μmol/L for Coulo-ED. Figure 2 demonstrates that HPLC-ED methods were significantly more sensitive than their UV counterpart. The authors also compared the Ampero-ED with single-electrode and Coulo-ED with multi-electrode arrays. The linear relationship in both electrochemical detectors was good in a range of 60–6,000 nmol/L and the correlation coefficient was over 0.999 [4]. This result demonstrated that ED for determination of salivary UA was useful, sensitive, and fast. However, samples acquired from only six healthy volunteers were tested. An extended cohort of participants needs to be tested in the future to confirm applicability of this method in clinical practice. Ampero-ED was also used by Honeychurch who achieved a good limit of detection (LOD) of 0.0059 μmol/L. The detection response was linear in a range of 0.02–16.65 μmol/L for UA and the coefficient of variation was 8.8% [17]. The general concepts of connection HPLC with ED were nicely described in the book by Flanagan et al. [57].

Figure 2:

HPLC chromatograms of 6 mmol/L UA standard using UV detection at 284 nm (A) and amperometric electrochemical detection at 1600 mV (B).

These data were obtained by C18 column (150 × 2.0 mm, 5 μm), isocratic elution with potassium phosphate buffer (74 mmol/L, pH 3.0) as the mobile phase, a flow rate 0.2 mL/min, an injection volume of 20 μL, and a total run time of 10 min. Adapted with permission from [4]. Copyright (2003) J. Chromatogr. B.

MS is becoming a routine analytical technique used in clinical laboratories typically including electrospray ionization technique in combination with quadrupole mass analyzer. This ionization technique was also applied for the MS determination of UA in both positive (+) [55] and negative-ion modes (−) [15]. Perelló et al. used the latter in the selected ion monitoring mode with anion exchange liquid chromatography for determination of salivary UA [16]. These authors applied a 20-fold diluted standard solution to provide better corrections of the matrix effects. The calculated LOQ was 0.42 μmol/L. Burger et al. analyzed 8 saliva samples using an alternative procedure based on enzymatic-photometric method [58]. The regression graph followed equation y=0.969 x + 0.923, R²=0.967 where y represented concentration obtained using the present method and x concentration determined with its enzymatic-photometric counterpart. Both methods were statistically comparable to the graph y=x at a 95% confidence level showing a good agreement between these methods [16]. Interestingly, they achieved a similar LOQ of 0.6 μmol/also L using UV detection [4]. These results suggest that UV, enzymatic, and MS detection have comparable sensitivity for determination of salivary UA.

Liu et al. developed a very fast, sensitive, and accurate HPLC method with MS/MS detection for determination of UA in saliva using an internal standard ¹⁵N labeled uric acid [55]. Saliva obtained from 28 healthy volunteers across the age, gender, and body mass index spectrum was tested. The UA determination range was 0.06–29.8 μmol/L. UA concentrations in saliva diluted with water in the range 5:7 were 46.8 ± 18.2 μmol/L.

Although all the HPLC-MS methods exhibited high sensitivity, until now only a handful of studies used MS detection for determination of UA in saliva. This is likely due to missing HPLC-MS hardware in clinical laboratories.

Capillary electrophoresis

CE is frequently used method for the determination of UA in trace analysis because it requires only small sample volumes in the range of nanoliters to femtoliters, the sample pre-treatment time is usually short, column efficiency is high, and CE is in general environmentally friendly [7]. The main advantage of CE in the determination is the negative charge of UA at a pH value of about 3.9 [59]. Rapid separations are feasible because high voltage and short capillaries can be applied in short capillaries [7]. A voltage of 30 kV was most often applied between the ends of a fused-silica capillary and samples injected electrokinetically. For the determination of UA in saliva, CE was coupled to a variety of detectors including UV [50] and mostly used Ampero-ED [13], [14], [15], [48]. A list of CE methods used is shown Table 3.

For example, Xing et al. used Ampero-ED with fused-silica capillary (57 cm long, 25 μm i.d.) and the borate buffer (20 mmol/L, pH 9.24) as the background electrolyte (BGE). Samples were injected electrokinetically applying 10 kV for 8 s [48]. A copper wire of 0.45 mm in diameter was used as the working electrode. The applied potential to the working electrode was maintained at +0.655 V with respect to saturated calomel electrode (SCE). These authors monitored effect of aerobic exercise on creatinine and UA levels in saliva and urine to determine the muscle growth and the intensity control of the exercise. Saliva collected before and after 2 h long exercise of medium intensity was analyzed. The linear range for UA was 0.59–2,380 μmol/L and LOD 0.86 μmol/L. However, the linear range is disputable since the authors published the first point in the concentration range lower than LOD. UA content in saliva increased after aerobic exercises from 166.2 ± 14.5 to 249.3 ± 17.8 μmol/L providing a valuable tool to study muscle growth and intensity control in aerobic exercises. No formal statistics analyses were presented.

Gonzalez et al. also focused on the effect of aerobic exercise on UA in saliva determined using enzymatic method [60]. This study confirmed again an increase in the content of UA after exercise. The exercise appeared to inhibit oxidative stress in saliva. On the contrary, Franco-Martínez et al. measured the UA levels in saliva using enzymatic method and did not revealed any changes after the effort sequences including 60 m dash [61].

Guan et al. utilized Ampero-ED detection with fused-silica capillary (75 cm long, 25 µm i.d., 360 µm o.d.) and 80 mmol/L borate buffer pH 7.6–8.2 as BGE [13]. Samples were injected electrokinetically applying 14 kV for 6 s. Working electrode was carbon disk with a diameter of 300 μm and the applied potential +0.950 mV (vs. SCE). The determination range for UA was 0.5–500 μmol/L and the LOD 0.33 μmol/L. This study focused on determination of UA and co-existing analytes xanthine, hypoxanthine, and ascorbic acid in the diagnosis of gout. Applicability of this method was tested on three healthy volunteers and three patients suffering from gout. The salivary UA concentration found for the patients was higher compared to the healthy controls [13]. No statistics was shown due to the insufficient cohort of tested individuals.

These authors later refined the original electrophoretic method [15] while determining salivary UA and p-aminohippuric acid in patients suffering from renal disease. The determination range for UA was 1–200 μmol/L and LOD 0.66 μmol/L. The UA levels in patients with a mean value of 272 μmol/L (n=3) was several times higher than that found for healthy volunteers (mean value=92.16 μmol/L, n=3). Although the sample pre-treatment was simple and rapid, use of Ampero-ED required long analysis times of about 20 min thus making these methods unsuitable for large sample series.

Capillary zone electrophoresis (CZE) was also utilized for determination of UA. Tanaka et al. used fused-silica capillary (33 cm long, 50 μm i.d., 365 μm o.d.), sodium phosphate buffer pH 6.5 as BGE, the hydrodynamic injection for 6 s at 5 kPa, the applied voltage −10 kV, and a low sensitivity UV detection at 214 nm [50]. The determination range for UA was between 23.7 and 1,189.7 μmol/L and LOD 13.68 μmol/L. They monitored oxidative and nitrosative stress in four healthy volunteers via the determination of daily variations in nitrite, nitrate, thiocyanate, and UA in saliva probed at 2 h intervals during the day. Elevated UA concentrations were observed at 11:00 a.m. Stability of the collected saliva before and after deproteinization was also assessed. The contents of the anionic components remained stable over a period of 48 h in the refrigerator and for 4 h at room temperature. Sample extracts ready for injection were stable for 48 h at room temperature [50]. The daily variations were determined without explicit consideration of both endogenous and exogenous factors such as smoking and eating during the day. These results indicate that the stability of salivary UA in both groups, i.e. before and after deproteinization for mere 48 h may not be sufficient for clinical studies. Extension of this period to weeks or even months would be useful for giving the clinical laboratories enough time to execute the analyses.

Capillary ion electrophoresis (CIE) with UV detection was used for determination of UA by Mori et al. [62] using fused silica capillaries (60 cm long, 75 μm i.d., 360 µm o.d.) coated with poly(vinyl alcohol) and 10 mmol/L phosphate buffer pH 7 BGE containing hexamethonium dichloride as the ion-pairing agent. The reproducibility of the method was simply enhanced by conditioning with BGE. The poly(vinyl alcohol) coating minimized protein adsorption and was stable over a wide range of pH 2.5–9.5. The constant voltages applied were 15 kV for electro-osmotic flow and −15 kV for anion separation. This approach allowed determination of UA in saliva in 13 min. The UA determination range was 0.1–1 mmol/L and LOD 3.42 μmol/L. The UA recoveries of 10-times diluted concentrations (n=3) using standard addition method were within the range of 86–108%. Anions in saliva collected after an overnight fasting from 12 healthy volunteers comprising nine males (four smokers and five-non-smokers) and three women (all nonsmoker) were determined. The concentration of UA in non-smoker male saliva was 0.110 mmol/L while it was 0.138 mmol/L for smokers. These values were about 2-times higher than 0.072 mmol/L observed for nonsmoker females. The cohort of patients in this study was again small and did not include smoking females. The direct injection of human saliva in the capillary was used to circumvent the low sensitivity of UV detection resulting from small injection volumes.

Trends focusing on miniaturization of analytical devices can be traced in recent years resulting in significant advantages such as a decrease in the separation time, high performance and throughput, reduced consumption of solvents, system integration, and multiplexing compared to techniques using conventional size devices.

For example, Chu et al. introduced miniaturized CE with Ampero-ED for the determination of UA in saliva and urine [14]. Feasibility of the device as an alternative to conventional CE was evaluated and confirmed its potential for the fast diagnosis of gout. The system was constructed at a 25 × 100 × 2 mm poly(methyl methacrylate) (plexiglass) base plate. A fused-silica capillary (8.5 cm long, 25 μm i.d., 360 μm o.d.) was used as the separation device and borate buffer pH 5.8–6.6 as BGE were used for the separation. BGE pH 6.0 was found optimal for the analysis. Four analytes including UA, xanthine, hypoxanthine, and ascorbic acid were separated within 3 min. BGE had to be replaced every 60 min to maintain the stability of the miniaturized CE-Ampero-ED system. Samples were injected electrokinetically applying 30 kV. A carbon disc with a diameter of 300 μm served as the working electrode and its potential was kept at +0.40 V (vs. Ag/AgCl reference electrode). The UA determination range was 2.97–1,189.68 μmol/L and LOD 0.27 μmol/L [14]. These results were in accordance with the previous CE-Ampero-ED study [13]. The miniaturized CE was designed for fast diagnosis of gout, but application of this system for that diagnostics was not confirmed. Only one patient with gout was tested who had UA level by higher 260.58 μmol/L than compared to that of a healthy volunteer. However, microchip CE is so far not generally available in diagnostic laboratories.

Enzymatic and optical instrumental methods

Enzymatic colorimetric assay kits were also used for the determination of UA and are usually the technique of choice for salivary UA analysis in clinical studies diagnosing metabolic syndrome [43], renal diseases [25], OSCC [63], and other oral diseases [23]. The principal advantages are simplicity, often high speed, no requirements for complex hardware, and commercial availability. Unfortunately, some of these assays suffer from cross reactions, each kit typically allows for analysis of only a single analyte, and some assays have inadequate sensitivity (e.g. 4.16 μmol/L [24]) in biological matrices such as renal and dental calculi [10]. The high costs speak against the use in large sample series.

The majority of kits for the determination of UA relies on the reaction catalyzed by enzyme uricase specific for UA. Back in 1941, Bulger and Johns proposed a method for detecting UA levels in human body using uricase [58]. UA reduced ferricyanide in the protein-free filtrate at pH 11 and temperatures of 3–5 °C.

Uricase oxidized UA to an allyl derivative that does not reduce ferricyanide. Change in the intensity of the color enabled calculation of UA content. The method is simple as only a single enzyme is involved. However, the reduction of ferricyanide can also be caused by other reducing substances present in biological samples. Thus, the results can be biased by the significant positive error. This especially applies for whole blood. UA can also get lost during the precipitation of proteins from the whole blood, this being another source of error.

The enzyme coupling method developed by Emerson [64] that calculated UA levels based on quantitative determination of hydrogen peroxide eliminated these deficiencies and improved the selectivity and sensitivity of UA detection. The concept of this enzymatic UA assay relied on oxidation of UA using uricase under formation of allantoin and the hydrogen peroxide according to the following reaction catalyzed by uricase:

Hydrogen peroxide was then used in the second enzymatic reactions of 4-aminoantipyrine (4-AAP) and 2-hydroxy-3,5-dichloro-benzenesulfonate (HDCBS) catalyzed by peroxidase to yield a chromogen [65]:

The absorbance of the chromogen product was then detected at a variety of wavelengths including 520 [24], [43], 548 [31], 546 [66], and 560 nm [67].

A different enzymatic uricase assay used measurement of fluorescence for detection of UA at excitation and emission wavelengths of 550 and 590 nm, respectively [25]. This method proved to be stable and reproducible and was greatly promoted for the early detection of UA.

Another approach for determination of UA comprised solution-based colorometric chemistries in the test-strip format. When the strip is exposed to a sample containing UA, cupric ion present is reduced to cuprous ion, which then forms a chelate with sodium bicinchoninate. The intensity of the resulting deep-violet precipitate is proportional to the amount of UA present in the sample. This test was used in the detection of UA in saliva of patients with the end-stage renal disease. The samples were collected from 19 patients immediately before and after dialysis and from 10 healthy controls. The test strips were imaged in a desktop digital scanner after immersing in the samples. The color intensities were determined by means of the histogram function in Adobe Photoshop and converted in concentrations by comparing the intensities with a calibration curve of digitally analyzed test strip color intensities produced by calibrator solutions of known concentration. Although the initial mean UA value 202 μmol/L for the patients with renal disease and the healthy controls (160 μmol/L) were not significantly different (p=0.12), the final mean concentrations observed for the patients and healthy controls were (p=0.0077 for UA). This saliva diagnostic test can reduce the number of blood tests in anemic and pediatric patients with renal disease [25]. Yet, it requires further study with comprehensive saliva samples cohorts to conclusively evaluate the clinical potential.

The Salimetrics^® (Carlsbad, USA) uric acid enzymatic assay kit with a LOD of 4.16 μmol/L is a colorimetric assay of UA in humans and some animals [24]. It was designed for research purposes only, not intended for diagnostic use and was not validated for detection in serum and plasma. This method requires only a short incubation period and utilizes enzymatic reaction that produces a red chromogen. The utilization is simple since the reaction is carried out and read in a 96-well microtiter plate with standard and controls provided. However, its sensitivity is low compared with HPLC methods and a large kit-to-kit variability was also observed [28]. Intra- and inter-laboratory assay variability depends on the platform used. The typical deviation range is 5–15% for the former and 15–30% for the latter. The kit was used in a study of relationship between baseline UA levels in saliva and brain function [39].

Although the Uric Acid Liquicolor Plus assay kit [26], the Uric Acid Plus [38] (both from Boehringer Mannheim GmbH, currently Roche Diagnostics GmbH, Mannheim, Germany), and the Infinite Uric Acid Liquid Kit (from Accurex Biomedical Pvt. Ltd., Mumbai, India) [27] are approved only for UA detection in serum and plasma, they were also used in clinical research with saliva [25], [67]. The selectivity of kits can be affected by matrix effects since immunoassays are generally not coupled with sample extraction or separation [28]. List of commercial kits using enzymatic methods for the determination of UA in saliva is presented in Table 4.

Vakh et al. published an innovative approach for the determination of salivary UA using an automatic chemiluminescence analysis of complex samples [56]. A multi-pumping flow system was successfully combined with a fluidized reactor and direct injection chemiluminescence detector. The fast luminol-N-bromosuccinimide reaction with UA in an alkaline medium was used to create the chemiluminescence. The method included online separation of UA from saliva using fluidized beds comprising anion exchange resin Dowex^® 2×8 followed by the elution and chemiluminescence determination in a direct injection detector. The stroke pulsations of the solenoid micro-pumps provided the floating of the anion exchange resin in the sample phase and then UA separation from the matrix in the sample pre-treatment block of the flow system. The LOD was 2 μmol/L. The determination linear logarithmic concentration range for UA was 6–1,000 μmol/L. The results were confirmed by reference HPLC-UV method. The F-test values ≤19 indicated insignificant difference in precision between both methods at a 95% confidence level. Thus, this method was the first automatic miniaturized procedure for the UA determination in saliva. The major advantages of this method were the need for only small quantities of sample (100 μL sample without dilution) and reagents, as well as generation of less waste (3 mL). However, the sample throughput and sensitivity was lower than in other flow and optical methods [68], [69]. Provided these limitations were eliminated, this approach may have a great potential in the determination of UA in clinical saliva samples.

Biosensors

Biosensors are small, fully integrated analytical instruments used for the detection and quantitation of a specific substance or analyte. The development of rapid, selective, sensitive, and cost-effective sensors has been a major issue in bioanalytical science over the last 20 years [70].

An optical or electrochemical transducer in the biosensor transforms target substances or their products in simple and quantifiable signals that are proportional to the concentration of the analyte. The signal is processed using associated electronics and integrated software systems [71]. Significant efforts are currently focused on new materials for the design of electrochemical biosensors featuring low limits of detection.

Salivary analysis using HPLC often requires relatively large volumes of sample and involves multiple step sample handling. These steps can include labeling, freezing, transportation, sample preparation, sorting, loading in an analyzer, analysis, and finally, reporting the results. This is a lengthy process in which each step has to be carried out carefully to avoid degradation of the salivary components. These issues have led to the requests for a rapid and reliable quantification of salivary biomarkers and resulted in the development of biosensors. Current biosensors allow the fast and accurate determination of UA [72]. The ability to immediately collect and analyze biomarkers in saliva with a single sensor offers numerous advantages in clinical applications [73]. Typical UA electrochemical sensors produced in recent years were summarized in the excellent review by Wang et al. [74].

The first UA sensors were developed already in the 1970s [75]. This approach included oxidation of UA using uricase with concomitant formation of allantoin and hydrogen peroxide. Uricase in the typical sensors is usually immobilized on an electrode, catalyzes reaction of UA, and the resulting product is detected. The uricase reaction in solution is controlled by temperature and pH. Its immobilization enhances its stability and activity. Electrodes derived from organic and inorganic materials collect electrons produced by the enzyme catalyzed reaction. UA determination using free and immobilized uricase was compared by Çete et al. [76]. The enzyme immobilized on platinum/polypyrrole-ferrocenium film enabled highly reproducible and stable results, maintaining 77.7% of the initial amperometric response at the end of the 30th measurement. Similarly, the amperometric response at the end of the 7th week of operation was 49% for the immobilized enzyme while 42% for its free counterpart. The kinetics of free uricase and immobilized in platinum/polypyrrole-ferrocenium film were characterized by K_m 0.39 and 0.44 mmol/L, and V_max 0.71 and 0.84 μmol/L/min under optimum conditions of pH 8.5 and 8.0 and temperature 35 and 55 °C, respectively.

Another approach to determination of UA developed by Moreas et al. relied on amperometric detection following the anodic oxidation of H₂O₂ [75]:

The major disadvantage of this technique was oxidation of UA itself on platinum, gold, and carbon electrodes. They prepared and characterized amperometric biosensors using layer-by-layer formed films comprising enzyme urate oxidase alternated with polyethylene imine or poly(diallyldimethylammonium chloride). The layer-by-layer method for immobilizing enzyme had advantages such as minimized protein denaturing and mild conditions for production of the films. Detection of H₂O₂ was carried out at 0.0 V (vs. SCE), thereby avoiding oxidation of UA. Biosensors had a limit of detection of 0.15 μA μmol/L cm² for a 10-bilayer film with polyethylene imine and 0.36 μA μmol/L cm² for poly(diallyldimethylammonium chloride). The linear response was between 0.1 and 0.6 μmol/L of UA. These values appeared as sufficient for use in clinical tests and this method is currently gaining on popularity [75]. However, no clinical results were presented yet.

Kim et al. designed biosensors for determination of salivary UA including a wireless amperometric circuit paired with a Bluetooth low energy communication system-on-chip for miniaturized and low-power operation [18]. The system was integrated in a salivary metabolite mouth guard sensor for a real time amperometric monitoring shown in Figure 3. This approach provided higher selectivity, sensitivity, and stability in comparison with previous study [77] in which O-phenylenediamine membrane and Prussian blue were used. The sensitivity was 2.32 μA/mmol/L in artificial saliva and 1.08 μA/mmol/L in real undiluted saliva. The mouthguard UA biosensor was also tested for practical use in monitoring hyperuricemia and the UA contents in saliva of healthy volunteers and hyperuricemia patients were compared. The UA concentration obtained as an average of 3 measurements per hour for 5 h was 178.5 ± 20.7 μmol/L for the healthy group and 822.6 ± 26.25 μmol/L for the hyperuricemic counterparts. Both groups produced consistent UA levels with small variations occurring due to circadian rhythm during the day and the food intake [18]. The effect of circadian rhytm on the salivary UA levels was described by Castagnola et al. [78]. All experiments were carried out using E_app = −0.3 V (vs. Ag/AgCl electrode) and a sampling time of 60 s. A great advantage of this biosensor was its ease of use. However, the toxicity and biocompatibility during the real-life application in humans remains to be studied. If nontoxic, this biosensor can enable production of a portable systems for monitoring UA in health and fitness environment.

Figure 3:

The mouthguard biosensor integrated with wireless amperometric circuit board (A). Reagent layer of the chemically modified printed Prussian-Blue carbon working electrode containing urease for salivary UA biosensor (B). Photograph of the wireless amperometric circuit board (C). front side (left) and back side (right).

Adapted with permission from [18]. Copyright (2015) Biosens. Bioelectr.

Several other groups also developed biosensors for quantification of UA in saliva. For example, Zhao et al. designed ZnS nanostructure-based electrochemical and photoelectrochemical biosensors [79]. Jindal et al. combined the n-type ZnO and p-type CuO to produce an array of p-n junction heterostructures and used these for low-powered biosensors [80]. Despite some potential for use in clinical practice, these methods were not used for the real-life analyses yet. Current UA sensors have high sensitivity and good specificity, but they are somewhat difficult to prepare and operate.