6. AOACSPIFANMethods-2018Awards

56

2 B runt et al .: J ournal of aoaC I nternatIonal V ol . 100, n o . 3, 2017 glucose, sucrose, maltodextrin, and/or starch. The subtraction of two large values in order to calculate much lower inulin/ FOS values generally results in less precise data with large SDs. Nevertheless, the method performs well when applied to products containing relatively low levels of interfering components.

of the fructans. The key difference is to avoid any borohydride treatment, thus avoiding the significant underestimation of Fm-type fructans. The method was codeveloped in two laboratories: Nestlé Research Centre (NRC) in Lausanne, Switzerland, and Carbohydrate Competence Centre of Eurofins (CCC) in Heerenveen, The Netherlands. Single-laboratory validations (SLVs) were performed independently in both centers. AOAC Official Method 2016.14 Fructans in Infant Formula and Adult Nutritionals HPAEC-PAD First Action 2016 Samples are reconstituted in water (if required) and further diluted until the concentration of fructan in solution is such that after hydrolysis, the fructose and glucose concentration is within the range covered by the standard curve. The diluted sample is treated with a mixture of sucrase and α-glucanases to hydrolyze sucrose and α-glucans, respectively, releasing their constituent monosaccharides. The sample is passed through an SPE cartridge packed with graphitized carbon. Salts and monosaccharides pass through and are washed away, while the fructans are retained. Fructans are released from the column using an acetonitrile solution. The released fructans are hydrolyzed with an inulinase mixture, and the released glucose and fructose are analyzed by high-performance anion-exchange chromatography (HPAEC) with pulsed amperometric detection (PAD). The fructan content is calculated by summing the glucose and 0.9× the fructose content measured. In some matrixes containing low amounts of fructans, a blank correction may be necessary and can be applied. In this study, blank corrections were required only at the lowest fructan concentration (0.03 g/100 g) and only for a few products. For unknown matrixes containing fructan concentrations below 0.1 g/100 g, it is advisable to check whether the blank correction is required. Samples from the SPIFAN II SLV Kit were provided by Covance (Madison, WI) and are listed in Table 2016.14A . In addition, two infant formulas from Nestlé (Vevey, Switzerland) were included in the SLV performed at NRC. Fructan ingredients used for spiking experiments were Orafti ® P95 and Orafti HP (both from Beneo, Tienen, Belgium) and NutraFlora ® P-95 (Ingredion, Inc., Westchester, IL). The samples were stored in the original package in a dry place and protected from light until the moment of use. According to the instructions for this SLV Kit and the AOAC Standard Method Performance Requirements (SMPRs ® ) defined in SMPR 2014.002 (5), all powder products except SRM 1849a were reconstituted by dissolving 25 g powder in 200 g water. The SRM 1849a was weighed directly or reconstituted by dissolving 10 g in 90 g water. In this validation study, three different standards of pure fructan ingredients were used: (1) Orafti HP, a long-chain inulin ingredient. (2) Orafti P95, a hydrolyzed inulin ingredient consisting of both GFn and Fm constituents. A. Principle B. Materials

The principle of Method 999.03 (4) differs from Method 997.08 (3), in that all monosaccharides present after combined α-glucanases and sucrase treatment are removed by converting them into alditols (via borohydride reduction). After enzymatic hydrolysis of the fructans, a colorimetric reducing sugar assay is then used to make the quantitative analysis. Fructans are thus accurately determined even in samples with high contents of monosaccharides, sucrose, maltodextrin, and/or starch, using relatively simple and inexpensive equipment. However, there is a drawback to this method: The reducing end groups of the fructan chains that do not terminate with a glucose (often referred to as Fm-type chains) are reduced into alditol end groups, which escape the analysis, resulting in low recoveries (e.g., the theoretical recoveries of fructobiose (F2) is only 50%, for F3, it is 67%; for F4, it is 75%, etc.). Fructan chains containing a terminal glucose (GFn-type chains) do not have a reducing end group, so these are recovered completely. FOS material prepared by depolymerization of inulin generally contains high amounts of F3 and F4, so total recovery can be below 80%. Although the method is not well suited to the analysis of samples containing FOS generated by partial hydrolysis of inulin, it is well suited to the analysis of long-chain fructans in a wide range of products and is quick and simple to apply. Neither of the methods is optimal for the determination of fructans in infant formula or adult nutritionals. Therefore, the Stakeholder Panel on Infant Formula and Adult Nutritionals (SPIFAN) issued a call for new methods to meet the AOAC Standard Method Performance Requirements (SMPRs ® ) defined in SMPR 2014.002 (5). After consideration by the Expert Review Panel, two methods, including the one described here and Method 2016.06 , were considered to have acceptable performance. Method 2016.06 (6, 7) is based on Method 999.03 (4) and on a method published by Cuany et al. (8), with a number of improvements. In Method 2016.06 (6, 7), a simplified sample preparation was introduced that reduces method turnaround time and, concomitantly, improves performance. The Cuany et al. method (8) required knowledge of the fructan type in the product to select appropriate correction factors (to correct for the “loss” of the terminal monomer of the fructan chains). Those correction factors are still required in Method 2016.06 (6, 7); however, a preanalysis step has been introduced to identify the fructan type and, thus, the appropriate correction factor, without the analyst needing (potentially confidential) recipe information. The method described here has been designed to determine the fructan content without the need for ingredient-specific correction factors, thus avoiding the need for preanalysis (or knowledge of the fructan type in the product), and without interference from other components such as sucrose and free sugars. As in Method 999.03 (4), an enzyme mixture is used to hydrolyze sucrose and α-glucans to their constituent monosaccharides. Next, following the strategy of Cuany et al. (8), a graphitized carbon SPE column is used to eliminate the released glucose and fructose before the enzymatic hydrolysis