HPLC測定に基づいて近赤外分光法によるブルーベリーにおける成分含量の非破壊予測のためのモデルの構築

The overall goal of this procedure is to construct and validate models for nondestructive prediction of total sugar, total organic acid, and total anti-cyanine content in individual blueberries by near-infrared spectroscopy. This method will answer key problems in the field of evaluation of farm products, such as fruits and vegetables. To me, an advantage of this technique is that plenty of our products can be checked nondestructively by spectroscopy.

And the models constructed through the procedure shown here. Near-infrared spectroscopy is a powerful tool for the nondestructive evaluation of farm products. Prediction models are constructed by examining the correlation between the spectrum and the contents determined by chemical analysis of each target.

Demonstrating the procedure will be Yusuke Kimura, a graduated student from my laboratory and Jingai Che, from university and one of the authors of this protocol. Collect a sufficient number and various types of sample blueberries from the target cultivars. Collect preferably 100 blueberries for the construction of the calibration model.

And at least 10 for the validation of the constructed model. To construct robust models, collect samples of various types, such as with various colors, sizes, and at various ripening conditions. Weigh each blueberry.

The weights measured are used later for the calculation of content percent of ingredients of each blueberry. Next put a blueberry sample in the center of the window of the near-infrared spectrophotomoter for diffuse reflectance measurements. By using the sample single channel command, measure the spectra of each blueberry.

Preferably at several points in the fruit. Keep the blueberries in a freezer, below minus 30 degrees celsius if they are not to be analyzed just after the spectrum measurements. To begin the pre-treatment, cut a blueberry into several pieces so it can be easily homogenized.

Cutting the blueberry without defrosting when it is frozen. Put the pieces into a 50 milliliter beaker. Add approximately 10 milliliters of ultra pure water to the beaker.

Heat the cut blueberry in ultra pure water in a microwave oven for 20 seconds to deactivate the enzymes that might decompose sugars during the analyses. Add approximately 10 milliliters of ultra pure water to the beaker. Homogenize the mixture for five minutes at 12, 000 RPM with a homogenizer equipped with a standard shaft and generator.

Centrifuge the homogenized mixture for 10 minutes at 3, 000 RPM. Next collect the filtrate by vacuum filtration of the centrifuged sample using a five B paper filter. After repeating the homogenization steps twice, measure the pH of the filtrate and adjust it to seven with dilute hydrochloric acid and dilute aqueous solutions of sodium hydroxide.

Dilute the filtrate to 50 milliliters with ultra pure water. Divide the sample into two. One for the analysis of sugars and the other for the analysis of organic acids.

Pass the first sample solution through the columns connected in series to exclude pigments, cations, and anions. Throw away the first one milliliter of the sample solution from the columns. Then use the sample solution from the columns for the analysis of sugars by high performance liquid chromotography or HPLC.

Centrifuge the solution at 6, 600 RPM for 10 minutes in a microtube equipped with a 0.445 micron filter with a mini-centrifuge before the analysis by HPLC. Arrange the HPLC system using a gel permeation column in the column oven at 40 degrees celsius. Use de-gassed ultra pure water with a flow rate of 0.1 milliliters per minute as the eluent.

Also use a refractive index detector. Measure the HPLC spectrograms of standard solutions by injecting a 20 microliter aliquot for each measurement. Here packed solution software is used for the measurement.

Get the area intensity of the band of sugar on the chromatogram of each standard solution by clicking re-analysis with the right button of the mouse. Then plug the area intensities against the corresponding concentrations to get the working curve for each sugar by linear regression. From the curve, the equation representing the relationship between the area intensity and the concentration is obtained for each sugar.

After obtaining the standard curve, measure the HPLC spectrograms of sample solutions by injecting a 20 microliter aliquot for each measurement. Get the area intensities of the bands of sugars on the chromatogram of each sample solution as before. Refer to the text protocol for steps on performing HPLC measurements of organic acids.

Keep the blueberries in a freezer below minus 80 degrees celsius if they are not to be analyzed for anthocyanins just after the spectrum measurements. To perform pre-treatment for HPLC measurements of anthocyanins, dry each frozen fruit with a vacuum lyophalyzer for 12 hours. Extract anthocyanin from the dried blueberry in a one percent methanol solution of trifluruocetic acid by leaving the mixture in a refrigerator at four degrees celsius for 12 hours.

Centrifuge the extract for 15 minutes in a two milliliter microtube using an ultra-centrifuge at minus eight degrees celsius and 15, 000 RPM. Then filter the extract through 0.45 micron filter to get the sample for HPLC measurements. Arrange the HPLC system using a C18 reverse phase column in a column oven at 40 degrees celsius.

Apply the gradient method using a 0.1 percent aqueous trifluroacetic acid as LUNA. And 0.5 percent trifluroacetic acid in acetonitrile as LUNB. Use a flow rate of 0.1 milliliters per minute where the ratio of LUNB increases from eight percent to 15%during zero to 50 minutes after the injection.

And from 15%to 75%during 50 to 60 minutes after the injection. Use a photo diode array detector monitoring at 520 nanometers. Measure the HPLC spectrograms of standard solutions by injecting a 10 microliter aliquot through the auto-sampler for each measurement.

Use the LC solution software for the measurement. Finally measure the HPLC spectrograms of sample solutions by injecting a 10 microliter aliquot through the auto-sampler for each measurement. At this point, the construction of calibration models for prediction of ingredient contents, followed by the validation of these constructed models, is performed as described in the text protocol.

Prediction models are constructed using the observed spectra and the ingredient contents determined through the procedure shown here. An example of the result of cross validation of the model constructed for the prediction of total sugar content is shown here. The correlation between the values predicted by near-infrared spectroscopy and those determined by HPLC is satisfactory.

A similar result is observed for the cross validation of the model for total anthocyanin content. Again the correlation between the predicted values and the observed ones is satisfactory. After watching this video, you should have a good understanding of how to get previous values of content of blueberries with HPLC.

Different evaluation is possible only by using the models constructed properly based on the preset content measured by destructive chemical analysis. It means that previous chemical analysis of a target is mystery to construct high performance prediction models. Please refer to the text protocol for the procedure of the construction of prediction models that are not shown in this video.

Once constructed properly, the models for prediction are applicable for the prediction of sugar, organic acids, and any contents of blueberries. Near-infrared spectroscopy, as shown here, is expected to be applicable for the correlation of many kinds of farm products. Don't forget that it is necessary to get preset barriers of content through chemical analysis for the construction of a good model when applying near-infrared spectroscopy to evaluation of any farm product.