AOAC ISPAM Stakeholder Panel Meeting Book 9-24-17

214 215 216 217 218 219 220 221 222 223 224 225 226 227

Possible methods for estimating gluten in grain samples:

1. Method of Wieser et al. (4) using RP-HPLC after selective Osborne-like extraction. Katharina Scherf has been using this method to characterize grains. The method is well-

established, but requires HPLC.

2. We have tried water and NaCl extraction to remove soluble proteins (by definition non- gluten by Codex) followed by centrifugation, then Dumas nitrogen on the remaining pellet. Our results have been similar to the Wieser results. Gluten = Dumas N 2 *5.68 3. Similarly, we have tried extraction of non-gluten proteins with 0.4 M NaCl in PBS (pH=7.5) followed by Dumas nitrogen analysis of the resulting centrifuged pellet. 4. Potentially, an approximation could be made on a sample just by a Dumas nitrogen and use a factor for each grain (e.g., for wheat, multiply protein by 0.8, for rye and barley, 228 229 We assume the accuracy of such method should not be too critical, since our end criteria for 230 ELISA recoveries will be quite wide. Inaccuracy of this method will be dwarfed by the size of the 231 criteria. We could increase the width of the ELISA criteria to account for the uncertainty in the 232 “reference” method’s accuracy. 233 234 Example 235 Samples of contaminant wheat rye and barley were obtained from Katharina Scherf’s lab, with 236 gluten concentration as determined by the Wieser method. These samples were used to spike 237 the very low oat flour. For this series, we used oat groats that were made by mechanically 238 separating oats and them mechanically cleaning the groats again after dehulling. The groats 239 were ground in the Retsch mill to pass a 1.0 mm screen. The sample of oat flour was 240 homogenized in food processor for 5 minutes prior to spiking. A large quantity (ca. 5 kg) was 241 homogenized by splitting in 2 portions and stirring by Robot-Coupe for 5 minutes. After initial 242 mix, one half of each half was mixed together in the Robot coupe (ca. 2.5 kg). The final flour 243 was mixed in a large plastic bag. 244 245 The oat flour was partitioned into equal portions with a large riffle splitter to approximate target 246 weight for spiking. The exact weights of oat flour and spiked ground grains were recorded to 247 calculate spiked concentration. Spiked amounts of wheat, barley and rye were weighed in a 248 metal measuring cup prior to addition to the oat flour. Each portion of oat flour was put into a 249 separate Cuisinart DLC-X food processor bowl. The pre-weighed spiked ground grain was 250 sprinkled evenly over the top of the flour. The sample was then mixed in the Cuisinart for 5 251 minutes, after which the motor was stopped and the bowl scraped with a tongue depressor 252 then the sample was mixed for an additional 5 minutes. 253 254 Spiked data including masses are given in Table 1. In addition to the gluten concentration 255 estimates we obtained from Dr Scherf, we also analyzed the samples by liquid phase extraction 256 of non-gluten proteins and Dumas nitrogen analysis of the centrifuged pellet. We did this two 257 ways, the first by extraction in water (2x) followed by extraction in 0.5 M NaCl (2x). The second 258 method we tried was by 0.4 M NaCl in PBS (pH = 7.5) (3x). The second method was suggested 259 by Dr. Scherf as a more optimal extraction which will not solubilize certain omega gliadins which 260 can be lost in pure water extracts. (Personal correspondence with K. Scherf) 261 multiply protein by 0.55) based on Katharina Scherf’s empirical data.