Determination of allantoin in blood by high-performance liquid chromatography with pre-column derivatization

A high-performance l iqu id chromatography method w i t h pre-column derivatization for separation and quantification o f al lantoin i n blood samples is described. Plasma after deproteinisation w i t h trichloroacetic acid was used for the derivatization procedure. The procedure was based on allantoin conversion to glyoxylic acid which forms a hydrazone w i th 2,4-dinitrophenylhydrazine. Al l an to in derivatives (synand anti-isomers) were separated on a reversed phase column (Nova-Pak C 1 8 , 4 fim) by gradient elution, and then moni tored at 360 nm. A l l components were completely resolved in about 46 min . The average recovery o f al lantoin added to plasma samples was 101.3 + 8.7% (n = 52). W i t h U V detector the smallest al lantoin concentration that gave reproducible integrations was 0.93 /mio l / l . The within-assay coefficient o f variat ion C V for derivatization and injection was 2 .7+ 1.2%, while C V for repeated injections was 0.68 + 0.33%. This H P L C method can also be used for determination o f allantoin in urine. K E Y W O R D S : allantoin, b lood, determination, H P L C


INTRODUCTION
In ruminants purines are metabolized in a series of reactions to form allantoin, uric acid, hypoxanthine and xanthine.Allantoin excreted in the urine is the main end product of purine metabolism (Balcells et al., 1991;Watts, 1980) and originated from three possible sourcess: purine bases of rumen microorganisms, endogenous purines from tissue turnover, and feed purines.In sheep and other ruminants allantoin appears to originate predominantly from nucleic acids synthesized by rumen microorganisms (Antoniewicz et al., 1980).In ruminants there may be a close relationship between the production of microbial protein in the rumen and the excretion of purine derivatives in urine and blood.Thus, measurements of the allantoin level in blood and urine provides an indice of the amount of microbial biomass supplied to ruminants (Rys et al., 1975;Antoniewicz et al., 1980Antoniewicz et al., , 1981;;Lindberg et al., 1989;Chen et al., 1992).Several authors have suggested that appropriate indicator of microbial protein synthesis may be the excretion of total purine derivatives (Chen et al., 1990;Puchala et al., 1991).
One major advantage is that this approach does not require the use of an invasive method to estimate microbial protein supply.Therefore it is necessary to resolve problem of determination of allantoin level in physiological fluids (Balcells et al., 1992;Terzuoli et al., 1994Terzuoli et al., , 1995)).However the method based on measurements of allantoin content in urine requires a total collection of urine (e.g.1-3 1/day for sheep) and good separation of urine from faeces is essential ( Lindberg et al., 1989;Gonda et al., 1994).Considering the above facts, it is essential to provide an accurate and selective method for the determination of allantoin in blood offering satisfactory estimation of the size of bacterial protein synthesis.
The aim of our work was to examine the suitability and accuracy of a pre-column derivatization (Young et al., 1942;Chen et al., 1993) and separation by HPLC for determination of allantoin in blood.

Reagents
HPLC-grade methanol, acetonitrile were purchased from Merck (Darmstadt, Germany), allantoin and 2, 4-dinitrophenylhydrazine (DNPH) from Sigma (St. Louis, MO, USA).All other chemicals were of analytical reagent grade and purchased from POCH (Gliwice, Poland).Water was distilled and then deionized prior to use.HPLC-grade water was prepared using a Milli-Q system (Millipore,Toronto, Canada).Mobile phases (solvents A and B) were filtered through a 0.2 /mi membrane filter (Millipore).A degassing of the solvents was made by 15 min ultrasonication prior to use.

HPLC configuration
A Waters 625 LC HPLC system (including a controller and pumps) was employed.The apparatus consisted of a turnable absorbance detector Waters Model 486, Waters 712 WISP autosampler, and computer data handling system (all equipment from Waters, Millipore, MA, USA).

Analytical solvents and gradient composition
The derivatizing solution was prepared by dissolving 100 mg of DNPH in 100 ml of 2 M HC1 and further the obtained solution was fdtered.Since the commercial DNPH contains 30% water, the actual concentration was 3.5 M. Thymol blue (pH indicator) was prepared at a concentration of 0.04% w/v.

Blood samples preparation
Blood samples from jugular vein of sheep were collected into tubes containing heparin and centrifuged at 1500 g for 20 min.The plasma was stored at 20°C.On the day of analysis, 0.5 ml of plasma was deproteinized with 0.5 ml of 10% (w/v) trichloroacetic acid and centrifuged at 10000-12000 rpm for 15 min.The supernatant was used for the derivatization procedure.

Derivatization procedure
A 500 yul sample of deproteinized plasma or allantoin standards and 50 /d of the pH indicator were pipetted into a centrifuge tube.To plasma samples 280 /d of 0.6 M NaOH was added while to allantoin standards 100 /d of 0.6 M NaOH only.The mixture could be stored at about -10°C if necessary.After heating at 85°C for 60 min, 200 /d of the DNPH solution was added.Next, a sample was filtrered through 0.2 /mi fdter (Cole Parmers) into the autosampler vial.

RESULTS AND DISCUSSION
The major analytical problem in the present work was obtaining a suitable separation of allantoin-deriving products from the interfering compounds.The derivatization procedure, however is not specific due to interference by 2, 4-dinitrophenylhydrazone of other keto-acids in plasma samples.Moreover, plasma contain several components with a similar high polarity and UV absorption.To omit these problems the method was improved.
The effect of different mobile phases composition and applied gradient in the separation of allantoin derivatives (i.e.syn and and isomers: allantoin-A and allantoin-B peaks) is shown in Figure 1.As can be seen from the chromatograms in Figure la and lc, plasma samples contain peaks (I and II) of unidentified DMPH complexes that had areas comparable to those of allantoin derivative peaks (A, B).The unidentified substances are presumablly hydrazones of other keto-acids (Chen et al., 1993).As expected, the improved HPLC method enabled suitable separation of analytical peaks of allantoin A and B. Indeed, the peaks A and B were distinct from interfering substances in plasma, so, their presence do not affect the determination of allantoin.
Although no reduction in a area of peak A or B was observed, the increase in a peak area of a unidentified compound (I) (the retention time of 15.2 + 0.4 min; Figure la) was found (Table 1) when the processed plasma samples were stored for 3 days at room temperature.
The peaks A and B had a retention time of 18 + 2 and 43 ± 2 min (mean ± SD of 90 samples), respectively.The ratio of allantoin-B/A peak area was constant at 3.406 + 0.196 (calculation based on 60 samples), irrespective of the concentration of allantoin and type of sample.Obviously, peaks A and B were suitable for quantitative determination of the unknown, although the larger peak A was preferred since more precise and accurate results were achieved.Thus, all determinations of allantoin concentration were based on allantoin-A peak.
Accuracy of the method was assessed by examining the recovery of known quantities (27.0-229.7 /miol/1) of allantoin added to plasma samples and was on the average 101.2 + 8.7% (Table 2) .
The relationship between the concentration (y) and allantoin-A peak area (S N ) was linear over a wide range of allantoin content (6.7-401.6 /imol/1)  and from the data were calculated the following equations and correlation coefficient (R): y ijanol/1) = 1.922 x 10" 5 S N + 0.24 y 0anol/l) = 4.84 x 10" 13 S N 2 + 2.032 x 10" 5 S N -0.13 R = 0.9997 The lowest concentration of allantoin that gave a reproducible integration of allantoin-A peak was 0.93 /miol/1.
The within-assay CV based on three samples repeated 5-9 times (derivatization and injection) was 2.7 + 1.2% (n = 20).The CV for repeated injections was 0.68 + 0.33% based on three samples each with six injections.

CONCLUSION
The present method is highly selective for allantoin, and offers the necessary sensitivity to permit the determination of allantoin in blood samples.Obviously, this HPLC method can also be used for determination of allantoin in urine.The chromatographic separation uses a widely available a reversed-phase C 18 column.Due to the complete separation of allantoin derivatives, the proposed method has advantages over Chen's method (Chen et al., 1993).The allantoin derivatives are stable, however, increase of concentrations of some unidentified compounds was observed when the processed blood samples were stored for 3 days at room temperature.Thus, plasma samples should be subjected to chromatographic separation no later than after 3 days storage at room temperature following the derivatization procedure.
The run time of 46 min appears long, but information provided by plasma samples are worth the effort.Indeed, application of this HPLC method in the study of purine metabolism should provide further evidence of the influence of nutritional manipulation on microbial supply protein to the small intestine.

Figure 1 .
Figure1.Comparison of improved method with Chen's method(Chen et al., 1993).Chromatograms for plasma sample (a) and allantoin standards (b) by improved method.Chromatograms for plasma sample (c) and allantoin standards (d) by the Chen's method.Peaks A and B are allantoin A and B isomers of 2,4-dinitrophenylhydrazone of glyoxylic acid.Peaks I, II and 3 are two unidentified compounds and thymol blue (at 37+ lmin), respectively

TABLE 1
Effect of lenght of plasma storage at room temperature on levels of an unidentified compound (I) and allantoin derivatives