INTRODUCTION

Biological interest

Nowadays, it is known that in mammals, Se is a trace element used by more than 30 Se-proteins in various metabolic pathways, in which it behaves as a cofactor. They incorporate the mineral co-transporting it as a cysteine residue, an amino acid that contains a sulphur atom which is replaced by Se to form Se-cysteine (Suttle, 2010; Schomburg, 2011; Combs, 2015).

Many Se-proteins have antioxidant functions. This is essential for mammals, since they protect the tissues from the attack of reactive oxygen species (ROS) which are secondary products of energy metabolism. An imbalance between the production of ROS and antioxidants can initiate oxidative chain reactions and peroxidation on lipids, proteins and DNA, causing serious cellular damage (Jaramillo et al., 2005; Chauhan et al., 2014).

In mammals, Se is incorporated into the organism primarily from food, but with differences in absorption efficiency (Se absorbed/Se in food). In non-ruminants and pre ruminants, efficiency ranges from 0.85 to 0.90, while in ruminants it is less, with values between 0.31-0.59, due to reduction of dietary Se to insoluble forms, such as elemental Se or selenide in the ruminal environment (Spears, 2003; Suttle, 2010).

Se absorption mainly occurs in the first portion of the small intestine, but for this to happen, it must be solubilised from food. In ruminants, Se release from food to the ruminal fluid is essential for its posterior absorption. So, It is crucial to clarify the release of the mineral from the forage to the ruminal liquor (Suttle, 2010; Combs, 2015).

Selenium quantification in ruminal fluid

The quantitative analysis of Se in biological samples has always been a challenge for researchers due to the low concentration present in this type of material and its high volatilization during the decomposition of the sample. There are a large number of methodologies for quantifying this mineral, such as: graphite furnace coupled to an atomic absorption spectrophotometer, spectrofluorometry, inductive coupling plasma to mass spectrometer, HG-AAS, among others. Spectrofluorometry and HG-EAA are the most frequently used techniques for the determination of Se in biological samples, due to their relatively low cost and minimal interference. The other techniques mentioned, although they allow the detection of Se at low concentrations and have more sensitivity, have more interference and are more expensive (Combs, 2015; Tinggi et al., 1992; Smrkolj et al., 2004).

The determination of Se by HG-EAA is a methodology that allows readings of low concentrations, of the order of 0.1 µg/l. To quantify a sample by this technique, it must be previously mineralized by wet digestion with concentrated acids and adequate heating can be carried out (Welz et al., 1985).

Performance verification

As mentioned above, it was important to develop a methodology to quantify Se in ruminal fluid. The methodology developed uses hydride generation atomic absorption spectroscopy (HG-AAS) to quantify Se.

To verify the reliability of the technique, the performance was evaluated. Such verification is extremely important. It is required on quality management systems, particularly the ISO 17025/IRAM 301 (IRAM, 2005), as a general condition for testing in calibration laboratories. This allows to know the functional characteristics of the method, providing a high level of confidence on it (AOAC, 2016).

To check the performance of the determination, Appendix F of the AOAC guide (AOAC, 2016) was used. It contains a detailed description of the minimum requirements of analytical performance to be analysed, according to the type of technique, in order to evaluate the performance. As Se quantification in ruminal bovine fluid by HG-AAS is categorized as a quantitative determination of a trace element, the analytical parameters necessary to evaluate are: applicable range, bias, precision by repeatability, intermediate precision, recovery, limit of detection (LOD) and relative uncertainty, all of which will be defined later.

Prior to the analysis of the above-mentioned parameters, acceptance criteria (AC) were established. These are the values that the method should fulfil in order to be reliably used for the required purposes. Some acceptance criteria were established by the AOAC guide (AOAC, 2016), while others were fixed by the professional in charge of the validation. Both are indicated below.

Objective

To develop and verify a methodological analytical performance of Se determination in ruminal fluid by HG-AAS.

MATERIALS AND METHODS

Sample preparation

The ruminal fluid used was extracted from a fistulous animal, which was fed with a stipulated and constant diet. This guaranteed approximately the same microbial population in the fluid over time. The ruminal fluid was wrung out and filtered through a cheesecloth to separate traces of forage. It was maintained in isothermal and anaerobic conditions (39 ºC) until its analysis to maintain the microbial environment.

Analytical technique

An HG-AAS was used where the reaction of acidified aqueous samples with a reducing agent, such as sodium borohydride (NaBH₄). This reaction generates volatile selenium hydride (SeH₂) which was transported to a quartz cell by means of an argon carrier gas. This cell was at 900ºC, which promotes the conversion to gaseous metal atoms (Se⁰). In the quartz cell, the generated analyte metal atoms are contained in the path of a light emitted by an EDL lamp (electrodeless discharge lamp). Thus, the atoms absorb part of the beam, and this is used by the computer to quantify the Se of the sample (PERKINELMER. 2000).

The method used is based on wet digestion of 5 ml of ruminal fluid with HClO₄, H₂SO₄ and HNO₃ concentrated (2:1:3). The sample was put in a thermostatic aluminium block programmed with ramps of temperatures ranging from 120 ºC to 180 ºC. The digest obtained must be transparent or slightly yellow (Alfthan. 1984; Raptis, et al., 1983).

Once digestion was completed, the digest was resuspended in 25 ml with doubly distilled water. In these conditions, all species of Se present in the fluid were in a state of oxidation +4 or +6. Since only the species +4 generate hydrides, it was necessary to pre-reduce the digest before reading it in the HG-AAS. For pre-reduction 10 ml of digest were put with 5 ml of HCl (c) in a water bath at 90º C for 20 minutes. Later, such mixture was taken to 25 ml with doubly distilled water.

Then, the generation of hydrides with a FIAS 100 generator took place, using HCl 10% (v/v) as a carrier solution, NaBH₄ 0.2% (w/v) in NaOH 0.05% (P/V) as reducing solution and Ar as a carrier gas at a flow rate of 80 ml/min (Perkinelmer, 2000).

Verification of the methodological performance

Acceptance criteria

They are boundary values established to compare with the results of the performance analysis and verify the applicability of the method. They were adjusted by the person responsible for the test, in a reliable and scientific manner (table 1).

Parameters analysed-Calculations

The AOAC, in its guide, prescribes the minimum requirements for analytical performance. These parameters are calculated as follows (AOAC. 2016):

Limit of detection (LOD): is the lowest concentration of analyte that can be quantified, with a 99% probability that it is not an equipment noise. It is considered that a quantification below LOD is no longer reliable, because the uncertainty related to the measurement would be larger than the measurement value itself.

LOD=Bco+3×Sbco

Bco: Average signal of 10 calibration blanks.

Sbco: Standard deviation of the mean of 10 targets

The AC established for this parameter was adjusted based on the lowest concentration found in the bibliography (Serra et al., 1994; Ruggieri et al., 2016).

Applicable range: is the Se concentration range that the method can measure without additional dilutions. It is characterized by the range that goes from the LOD until the last standard in the calibration curve, multiplied by 12.5, which is the dilution factor of the sample. The AC established had at least the lowest concentration (0.220 µg/l) and at most the highest concentration found in the bibliography (9.68 µg/l) (AOAC, 2016; Serra et al., 1994).

Bias (Bias %): is the analyte concentration that is recovered from certified standard when the sample is carried out throughout the whole method (or the maximum possible part). In our case, it was not possible to have a certified ruminal liquor standard. Then, as it is suggested by AOAC (2016), samples fortified with a certified standard solution of 980 ppm of Se in 1% HNO3 were used. We analyzed 5 independent replicates of 3 levels of fortification. The AC established corresponded to that determined by the Shah conference reports et al. (1992; 2000), widely accepted in bioanalytical analyses.The calculation was:

Bias(%)=[(Co/Ca)-1]×100

Co: Obtained concentration

Ca: Added concentration

Precision: is the degree of concordance between the obtained results from replicated measurements over the same object under specific conditions. It is a parameter only related to the random error. The specific conditions may be:

- Repeatability (RSDr): is the degree of concordance between the results of independent analyses done with the same method, over the same sample, in the same lab, by the same operator, on the same day. Seven independent replicates of three samples with different concentration levels were analysed. For this parameter, the AC is the one suggested by the AOAC in appendix F (2016); which is indicated according to the native concentration of the sample. The RSDr % was characterized for each level as:

Sm: Standard deviation of the mean of the replicates in each concentration.

Xm: Average of the replicates in each concentration.

- Intermediate Precision (RSDip): is the degree of concordance between the results of independent analyses done with the same method, over the same sample, in the same lab, by the same operator, on different days, with different calibration curves. The established AC is the one suggested by AOAC in appendix F (2016) for the precision by repeatability, assuming that the variation between days is not significant. Seven replicates of a sample were analysed on three different days, and the total RSDip was characterized as:

SM: Standard deviation of the mean of the replicates throughout the days.

XM: Mean of the replicates throughout the days.

Recovery (Recup.f %): is the fraction or percentage of the analyte that is quantified in a fortified sample. Seven replicates at three fortification levels (50, 25 and 10 µg/l) were analysed over the same sample of ruminal liquor.

The established AC is that suggested by AOAC in appendix F (2016), which is indicated according to the native concentration of the sample. The average percentage of recovery for each level of concentration was characterized as:

Lf: Concentration read in the fortified ruminal fluid.

Ls: Concentration read in the not fortified ruminal fluid.

Cf: Fortification concentration.

Expanded relative uncertainty (UREL %): is a parameter associated to the result of a measurement which characterizes the dispersion of the values that might be reasonably attributed to the measure. The established AC is suggested by the United Nations Office on Drugs and Crime (UNODC) (2009) in its guide for methods validation (OAA, 2013; AOAC, 2013; Ruggieri et al., 2016).

The UREL % can be calculated through the “black box” model. In this model, proposed by Reyes and Cerezo (2009), only the Bias and the intermediate precision are considered for the calculation of the uncertainty, as can be seen in the following equation:

UREL %: expanded relative uncertainty.

RSDip%: coefficient of variation of intermediate precision.

n: number of days in which the intermediate precision was analysed.

RESULTS AND DISCUSSION

The results obtained by analysing the parameters required to evaluate the technique performance are shown in table 1:

The methodology was developed based on the available equipment and capabilities in the laboratory, following the guidelines established by Alfthan (1984); Tinggi et al. (1992) and Raptis et al. (1983).

As table 1 shows, all the parameters studied are within the limits of the AC. In the case of LOD, it was not possible to find references to discuss data for this parameter in ruminal fluid.

In the case of the applicable range, this exceeds what is required in the AC, which allows to loosely cover the range of possible Se concentrations in ruminal fluid (Serra et al., 1994; Koening et al., 1997).

For Bias, it is observed that the differences obtained among the three concentrations of fortification that were used and the theoretical value are both in agreement with the AC established. The values coincided with the ones obtained by Tinggi et al. (1992), who informed Bias % with fortified biological samples that fluctuated between 0-8.3%; but they were higher to the ones obtained by Navarro et al. (1995), which indicated up to 3.84%.

The precision by repeatability was within the terms of the AOAC (2016) guide, which takes into account the possible repeatability value based on the range of analyte concentrations. A limit of 21% is suggested for RSDr. Kadrabova et al. (1997) and Surai et al. (2008) obtained RSDr of 1.1-3.8% and 1-5%, respectively.

The intermediate precision obtained is below the established AC. Alfthan (1984) obtained RSDip between 3.6-5.7% for concentrations of 49.4; 129 and 1025 µg/l of Se, respectively. The RSDip mentioned are inferior to those obtained in this work, but the authors worked with higher Se concentrations (3.5 µg/l of Se).

The recoveries with fortified samples are included within the limits suggested by the AOAC (2016), according to the range of Se concentrations, of 60-115%. They were also similar to those obtained by other authors. Maher (1987) obtained recoveries of 97% in biological samples, Tinggi et al. (1992) of 104%, Alfthan (1984) of 101% and Kadravoba et al. (1997) from 93.9 to 101%, all of them in diverse concentrations and biological samples.

The expanded relative uncertainty was also below the AC for the purposes required. It was not possible to get references to compare this parameter for ruminal fluid.

CONCLUSIONS

The obtained values are within the minimum requirements indicated by the AC. In this way, it is concluded that the performance of the technique is suitable for the quantification of Se in ruminal fluid.

Acknowledgments

The authors thank Instituto Nacional de Tecnologia Agropecuaria (INTA) with headquarters in Balcarce, for having financed the research.

REFERENCES

ALFTHAN, G. 1984. A micromethod for the determination of selenium in tissues and biological fluids by single-test-tube fluorometry. Analytica Chimica Acta. 165, 187-194.

AOAC INTERNATIONAL. 2012. Appendix K: Guidelines for single-laboratory validation of chemical methods for dietary supplements and botanicals.1-32 (Available at: http://www.aoac.org/aoac_prod_imis/AOAC_Docs/StandardsDevelopment/eoma_appendix_k.pdf verified: March 2019).

AOAC INTERNATIONAL. 2016 Appendix F: Guidelines for standard methods performance requirements. 1-16 pp. (Available at: http://www.eoma.aoac.org/app_f.pdf verified: March 2019).

CHAUHAN, S.S.; CELI, P.; PONNAMPALAM, E.N.; LEURY, B.J., LIU, F.; DUNSHEA, F.R. 2014. Antioxidant dynamics in the live animal and implications for ruminant health and product (meat/milk) quality: role of vitamin E and selenium. Animal Production Science, 54(10), 1525-1536.

COMBS, F.G. 2015. Biomarkers of Selenium Status. Nutrients, 7(4), 2209-2236. doi:10.3390/nu7042209

IRAM. 301: 2000/ISO/IEC 17025. 2005. Requisitos generales para la competencia de los laboratorios de ensayo y de calibración. IRAM. Buenos Aires. 29 p.

JARAMILLO, S.; VILLA, N.A.; PINEDA, A.F.; GALLEGO, Á.B. 2005. Actividad Sanguínea de Superóxido Dismutasa Y Glutatión Peroxidasa En Novillas a Pastoreo. 40(11), 1115-1121.

KADRABOVA, J.; MADARIC, A.; GINTER, E. 1997. The selenium content of selected food from the Slovak Republic. Food Chemistry. 58 (1-2): 29-32.

KOENIG, K.M.; BUCKLEY, W.T.; SHELFORD, J.A. 1990. True absorption of selenium in dairy cows: Stable isotope tracer methodology and effect of dietary copper. Canadian Journal of Animal. 71: 175-183.

MAHER, W.A. 1987. Decomposition of marine biological materials for the determination of selenium by fluorescence spectrometry. Microchemical Journal. 35(1), 125-129.

NAVARRO, M.; LÓPEZ, H.; RUIZ, M.L.; GONZÁLEZ, S.; PEREZ, V.; LÓPEZ, M.C. 1995. Determination of selenium in serum by hydride generation atomic absorption spectrometry for calculation of daily dietary intake. Science of the Total Environment. 175(3), 245-252.

ORGANISMO ARGENTINO DE ACREDITACIÓN (OAA). 2013. Guía para la validación de métodos de ensayo. GUI-LE-03. Versión: 1; 11.

PERKINELMER. 2000. Flow Injection Mercury/hydride Analysis: Recommended Analytical Conditions and General Information. USA. 1-16; 2-29 pp.

RAPTIS, S.E.; KAISER, G.; TÖLG, G. 1983. A survey of selenium in the environment and a critical review of its determination at trace levels. Fresenius’ Zeitschriftfüranalytische Chemie. 316(2), 105-123.

REYES, M.; CEREZO, M.J. 2009. Consideraciones sobre la evaluación de la incertidumbre. V Iberolab: 1-3 p.

RUGGIERI, F.; ALIMONTI, A.; BOCCA, B. 2016. Full validation and accreditation of a method to support human biomonitoring studies for trace and ultratrace elements. Trends in Analytical Chemistry. 80, 471-485. doi.org/10.1016/j.trac.2016.03.023.

SCHOMBURG, L. 2011. Selenium, selenoproteins and the thyroid gland: interactions in health and disease. NATURE REVIEWS ENDOCRINOLOGY, 8(3), 160-171. doi:10.1038/NRENDO.2011.174.

SERRA, A.B.; NAKAMURA, K.; MATSUI, T.; HARUMOTO, T.; FUJIHARA, T. 1994. Inorganic selenium for sheep I. Selenium balance and selenium levels in the different ruminal fluid fractions. Asian Journal of Animal Science 7, 83-89.

SHAH, V.P.; MIDHA, K.K.; DIGHE, S.; MCGILVERAY, I.J.; SKELLY, J.P.; YACOBI, A.; PITTMAN, K.A. 1992. Analytical methods validation: Bioavailability, bioequivalence and pharmacokinetic studies. Journal of Pharmaceutical Sciences. 81(3), 309-312.

SHAH, V.P.; MIDHA, K.K.; FINDLAY, J.W.; HILL, H.M.; HULSE, J.D.; MCGILVERAY, I.J.; TONELLI, A. 2000. Bioanalytical method validation-a revisit with a decade of progress. Pharmaceutical Research. 17(12), 1551-1557.

SMRKOLJ, P.; STIBILJ, V. 2004. Determination of selenium in vegetables by hydride generation atomic fluorescence spectrometry. Analytica Chimica Acta. 512(1), 11-17.

SPEARS, J.W. 2003. Trace mineral bioavailability in ruminants. Journal of Nutrition. 133(5), 1506-1509.

SURAI, P.F.; TAYLOR-PICKARD, J.A. 2008. Current advances in selenium research and applications. Wageningen Academic Publishers. (1), 351. doi: 10.3920/978-90-8686-642-7.

SUTTLE, N.F. 2010. Mineral nutrition of livestock. in: selenium. 4th. ed. CAB International. 377-425.

TINGGI, U.; REILLY, C.; PATTERSON, C.M. 1992. Determination of selenium in foodstuffs using spectrofluorometry and hydride generation atomic absorption spectrometry. Journal of Food Composition and Analysis. 5(4), 269-280.

UNITED NATIONS OFFICE ON DRUGS AND CRIME (UNODC). 2009. Guidance for the validation of analytical methodology and calibration of equipment used for testing of illicit drugs in seized materials and biological specimens. United Nations: New York. 67.

WELZ, B.; MELCHER, M. 1985. Decomposition of marine biological tissues for determination of arsenic, selenium, and mercury using hydride-generation and cold-vapor atomic absorption spectrometers. Analytical Chemistry. 57(2): 427-431.