Development of a Hyperspectral Imaging System to Detect Nitrate Concentration of Spinach Callus

Abstract

The purpose of this study was to develop a hyperspectral imaging system that can detect gene recombinant calluses. Drug tolerance genes are commonly used as marker genes, although they may negatively affect the environment. Hence, we aimed to use the genes that code for nitrate reductase (NR) as marker genes. Once the NR genes are introduced into a cell, the recombined genes would show greater nitrate reduction compared with the wild-type genes, making it possible to use the nitrate concentration for discriminating gene recombinant calluses from non-recombinant ones. In the present study, a non-destruc-tive method of measuring the nitrate concentration in spinach ( Spinacia oleracea L. ‘Orai’) calluses was developed to allow for the rapid detection of the gene recombinant calluses. Absorption spectra of 349 calluses (concentration range: 126.0 mg L - 1 -2697 mg L - 1 ; standard deviation: 646.3 mg L - 1 ) were measured by a hyperspectral camera (wavelength range: 400 nm-1000 nm; resolution: 9 nm). The nitrate concentration of the calluses was measured by the RQ-Flex method. Mathematical models to estimate the nitrate concentration in the calluses from the absorption spectra were developed using the principal component regression (PCR) and partial least squares (PLS) methods, and the correlation coefficient between the measured and estimated nitrate concentration was found to be 0.7363. The standard errors of calibration ( SEC ) and prediction ( SEP ) of the model were 430.3 mg L - 1 and 438.7 mg L - 1 , respectively.