Precise high-throughput online near-infrared spectroscopy assay to determine key cell wall features associated with sugarcane bagasse digestibility

Abstract

Background: Sugarcane is one of the most crucial energy crops that produces high yields of sugar and lignocellulose. The cellulose crystallinity index (CrI) and lignin are the two kinds of key cell wall features that account for lignocellulose saccharification. Therefore, high-throughput screening of sugarcane germplasm with excellent cell wall features is considered a promising strategy to enhance bagasse digestibility. Recently, there has been research to explore near-infrared spectroscopy (NIRS) assays for the characterization of the corresponding wall features. However, due to the technical barriers of the offline strategy, it is difficult to apply for high-throughput real-time analyses. This study was therefore initiated to develop a high-throughput online NIRS assay to rapidly detect cellulose crystallinity, lignin content, and their related proportions in sugarcane, aiming to provide an efficient and feasible method for sugarcane cell wall feature evaluation.

Results: A total of 838 different sugarcane genotypes were collected at different growth stages during 2018 and 2019. A continuous variation distribution of the near-infrared spectrum was observed among these collections. Due to the very large diversity of CrI and lignin contents detected in the collected sugarcane samples, seven high-quality calibration models were developed through online NIRS calibration. All of the generated equations displayed coefficient of determination (R²) values greater than 0.8 and high ratio performance deviation (RPD) values of over 2.0 in calibration, internal cross-validation, and external validation. Remarkably, the equations for CrI and total lignin content exhibited RPD values as high as 2.56 and 2.55, respectively, indicating their excellent prediction capacity. An offline NIRS assay was also performed. Comparable calibration was observed between the offline and online NIRS analyses, suggesting that both strategies would be applicable to estimate cell wall characteristics. Nevertheless, as online NIRS assays offer tremendous advantages for large-scale real-time screening applications, it could be implied that they are a better option for high-throughput cell wall feature prediction.

Conclusions: This study, as an initial attempt, explored an online NIRS assay for the high-throughput assessment of key cell wall features in terms of CrI, lignin content, and their proportion in sugarcane. Consistent and precise calibration results were obtained with NIRS modeling, insinuating this strategy as a reliable approach for the large-scale screening of promising sugarcane germplasm for cell wall structure improvement and beyond.

Keywords: Biomass digestibility; Cell wall; Cellulose crystallinity; Lignin; NIRS; Sugarcane bagasse.

Fig. 1

Near-infrared spectral characterizations in the sugarcane population. A Original spectroscopy. B The first 13 principal components for near-infrared spectral characterization. C Sample variations in each principal component in the sugarcane samples. D 3D view of the collected sugarcane samples via PCA

Fig. 2

Diversity of cell wall features in the sugarcane population. A X-ray diffractograms. B Distribution and variations in cellulose crystallinity. C Variation in lignin content (% dry mass) in the collected sugarcane samples. D Variation in lignin proportion (% cell wall) in the collected sugarcane samples. ASL acid-soluble lignin; AIL acid-insoluble lignin

Fig. 3

Sample distribution in the calibration and validation sets for online NIRS modeling. A Cellulose crystallinity. B Lignin clean mass content in the dry biomass. C Lignin proportion in the cell wall. ASL acid-soluble lignin; AIL acid-insoluble lignin

Fig. 4

Equation performance during online NIRS calibration and external validation. A Cellulose crystallinity. B Lignin clean mass content in the dry biomass. C Lignin proportion in the cell wall. ASL acid-soluble lignin; AIL acid-insoluble lignin; RMSE root mean square error; RPD ratio performance deviation

Fig. 5

Equation prediction capacity of the offline NIRS models. A Original spectra of the dry sugarcane samples. B Principal component analysis of the samples in the calibration sets. C Calibration model of cellulose crystallinity, D lignin clean mass content in the dry biomass, and E lignin proportion in the cell wall. ASL acid-soluble lignin; AIL acid-insoluble lignin; RPD ratio performance deviation

Fig. 6

Comparison between the online and offline NIRS assays. A Proceedings of the online and offline NIRS analyses. B Near-infrared spectra collected during the online and offline procedures. C Statistical comparison of the model parameters between the offline and online NIRS assays. * and ** indicated statistically significant different at p  <  0.05 and 0.01, respectively