Effects of enzymatic browning reaction on the usability of tobacco leaves and identification of components of reaction products

The enzyme browning reaction results in grey speckles on tobacco leaves, which impairs the value and industrial usability of tobacco leaves. To demonstrate the influences of different browning degrees (BDs) of tobacco leaves on the usability of different cultivars and positions and identified structure of brown (grey) matter, we selected three flue-cured tobacco cultivars (K326, Yunyan87, and Honghuadajinyuan (Hongda)) and set four different BDs (<25%, 25% to 50%, 50% to 75%, and >75%). Indices related to: economic traits, chemical components, physical properties, and sensory quality of tobacco leaves with different cultivars were evaluated. Moreover, by utilising thin-layer chromatography and high-performance liquid chromatography, we analysed and identified the structure of the grey matter in terms of chemical composition. The experimental results show that the main component of grey speckles on tobacco leaves is 3-acetyl-6,7-dimethoxycoumarin (YC-ZJF). With the increase of BD, the amount of total sugar and reducing sugar, output value, the proportion of superior tobacco, shatter resistance index, and sensory evaluation score of the three cultivars significantly decrease, while the starch content increases significantly. The changes in protein, total nitrogen, and nicotine are insignificant with changing BD. In addition, other indices show different trends for different cultivars of flue-cured tobacco. After separation and identification of the components of grey speckled leaves, it is proved that the substance derived from grey speckles on tobacco leaves is YC-ZJF. The research is important to the study of browning mechanisms in tobacco leaves and provides corresponding targets for strategies to reduce browning thereof.

Tobacco with grey speckled or browning leaves, in undesirable in flue-cured tobacco (Fig. 1) and it not only affects the supply of tobacco leaves, but also inhibits sustainable development of tobacco production; however, the influences of grey speckles on usability of tobacco leaves in different cultivars and positions and the structure of grey matters from enzymatic browning reactions remains unclear. Based on this, we aim to study the influences of different BDs of tobacco leaves (referring to the degree of deterioration of tobacco leaves with grey speckles) on the usability of tobacco leaves in different cultivars and positions and identified components of grey matter, to explore the browning mechanisms of grey speckled tobacco leaves.
Flue-cured tobacco has an artificial flue-curing process that takes 5-7 days 1 . Many changes occur in chemical composition and appearance of such leaves during flue-curing 2 . According to a survey of tobacco production, the rate of occurrence of tobacco leaves with grey speckle during flue-curing would be about 20 to 30% 3 . Tobacco leaves with grey speckle are inferior tobacco in accordance with tobacco leaf grading standards and it either is cheap or not purchased. It reduces tobacco quality and even deprives industrial usability of tobacco, which incurs economic loss among tobacco growers. If the rate of occurrence of tobacco leaves with grey speckle can be reduced by one percentage point by using scientific means, $42 million (USD) of average annual benefits can be www.nature.com/scientificreports www.nature.com/scientificreports/ Field management. Hongda, K326, and Yunyan87 cultivars were produced by utilising high-quality, high-efficiency cultivation techniques to cultivate tobacco with balanced nutrition, normal growth, and fresh leaves yellowed and matured layer-by-layer. In August, after cultivating for 90 to 95 d and topping for 35 to 40 d, tobacco leaves became pale yellow and 80% of leaves were yellowed, showing white, bright main veins, white branch veins, and downward rolled leaf apices and leaf margins. When leaves were wrinkled, mature leaves were collected on five to seven occasions at different positions and flue-cured according to the requirements for experimental design. Other agronomic practices were carried out following guidelines recommended by the Integrated Technology Promotion Centre at the Yunnan Academy of Tobacco Agricultural Sciences 19 . Experimental design. Fresh tobacco leaves with the same quality were selected for flue-curing in the furnace and then different BDs were set by suddenly reducing the temperature in the flue-curing process. Four different treatments, namely, BDs lower than 25%, in the ranges of 25% to 50%, 50% to 75%, and higher than 75% were set. After flue-curing K326, Yunyan87, and Hongda cultivars, tobacco leaves with the four BDs were used as samples for the experiment. By using local graders for tobacco leaves to grade the samples according to the National Standard for Flue-cured Tobaccos (GB 2635(GB -1992, the appearance quality and economic traits of the three cultivars of treated tobacco samples were obtained. Some tobacco samples were sent to the laboratory of the Yunnan Academy of Tobacco Agricultural Sciences to determine physical indices and analyse chemical components, while the other samples were sent to the Technology Centre, China Tobacco Yunnan Industrial Co., Ltd for evaluation and rating. Meanwhile, after collecting grey matter from tobacco samples subjected to different treatments, the samples were sent to the College of Pharmacy, Huazhong University of Science and Technology (Key Laboratory of Natural Medicinal Chemistry and Resource Evaluation in Hubei Province) to identify their structure and components.
Analyses. Economic traits. The raters graded the tobacco samples for the experiment according to National Standard (GB2635-92). The proportion of superior tobacco was estimated based on the national purchasing data and supervision and inspection data of grade quality during industry-commerce handover in that year. The flue-cured tobacco plants have distinct leaf numbers and weights due to different production areas and cultivars. For convenience, the number of leaves was estimated by using method proposed by Yan et al. 20 , thus allowing calculation of yield and output value of tobacco leaves.
Chemical components. The starch content was determined by spectrophotometry at 660 nm with HClO 4 extraction 21 . Total sugar and reducing sugar were determined by rapid colorimetric method with 3,5-dinitrosalicylic acid 22 . The content of protein was determined with continuous flow analytical method 10 . Total nitrogen determined was using an elemental analysis method 23 . Nicotine was determined using a spectrophotometric method 24 . The polyphenol content was determined with HPLC coupled with ESI-MS after solid-phase extraction 25 .
Industrial machinability of tobacco leaves. After removing green miscellaneous tobacco from flue-cured tobacco, 10 kg samples of C3F and C2F were selected and sent to the Technology Centre, China Tobacco Yunnan Industrial Co., Ltd for analysing the industrial machinability. After equilibration for 72 h at constant temperature (22 °C) and humidity (60%), the samples were analysed according to the method for detecting the shatter resistance index of tobacco leaves provided by Chen et al. 26 .
Sensory quality. The sensory quality of the tobacco samples was evaluated by seven certified experts from the Technology Centre, China Tobacco Yunnan Industrial Co., Ltd according to 11 indices: the original aroma, aroma volume, aroma quality, concentration, biting taste, physiological strength, offensive odour, cleanliness, wetness, and aftertaste. The evaluation results refer to Sensory Technical Requirements for Cigarettes (GB5606.  for evaluating the sensory quality of tobacco leaves and nine-point scale was used to score these 11 indices. The data shown in this study are the mean of seven reports.
Identification of main chemical components of grey matter in tobacco with grey speckles. Sample preparation for analysis using thin-layer chromatography. By separating the parts with grey speckles from parts without grey speckles on tobacco leaves, 50 g of the tobacco samples with grey speckles and 50 g without grey speckles were obtained. After crushing, they were extracted using ethanol at a concentration of 95%.
Extraction and isolation of the main chemical components of the tobacco with grey matter. After drying and crushing, the leaves of tobacco with grey matters were extracted with 95% ethanol four times (each for 3 d) at room temperature. The ethanol was removed under the reduced pressure to obtain 950 g residue. The extracts were separated and divided into three fractions (Fr.A, Fr.B, and Fr.C) by utilising silica gel (100 to 200 mesh) column chromatography (CC) eluted with a graduate methylene chloride/methanol solvent system. Fr.A (120 g) was further separated to three sub-fractions (Fr.Aa, Fr.Ab, and Fr.Ac) by silica gel CC eluted with methylene chloride/methanol (8:1). Fr.Ab (25 g) was separated using a reversed-phase C18 column eluted with a methanol/ water system to yield three sub-fractions (Fr.Ab1, Fr.Ab2, and Fr.Ab3). Fr.Ab1 (4 g) was separated and purified by using Spephadex LH-20 (MeOH) to obtain a fraction Fr.Ab1a. Fr.Abla (220 mg) was finally purified by a high-performance liquid chromatograph [YMC-pack ODS-A column (5 μm, 10 × 250 mm), 35% MeCN/H 2 O in a flow rate of 2 mL/min] to yield 9.7 mg YC-ZJF (t R = 20 min) (Fig. 2).
(YMC-pack ODS-A columns with a particle size of 5 μm and dimensions of 10 × 250 mm and 35% MeCN/ H 2 O solution at flow rate of 2 mL/min were used.) www.nature.com/scientificreports www.nature.com/scientificreports/ Statistical analysis. The data were analysed by using the general linear model (GLM) program of the SAS 9.3 computer package made by the SAS Institute Inc., Cary, NC. The statistical analysis of data in this study was based on the significance level of P < 0.05. Through Tukey's (HSD) test, the mean of data was divided into 95% confidence intervals. Sigma Plot 12.3 (Systat Software Inc., Chicago, IL, USA) was used for plotting.

Results
Effects of different BDs on economic traits of different cultivars of flue-cured tobacco. As shown in Table 1, BD and position exert significant effects (P < 0.05) on the average price and the proportion of superior tobacco. For the proportion of superior tobacco, BD has a synergistic effect with position. Figure 3 demonstrates that the average price with BD < 25% shows significant differences with BD > 75% in middle and upper leaves of the three flue-cured tobacco cultivars. Figure 4 demonstrates that the proportions of superior tobacco in different positions of the three flue-cured tobacco cultivars reduce with increasing BD. At the same BD, the proportions of superior tobacco in different positions of each cultivar have no significant difference.
As shown in Table 1, BD, cultivar and position significantly affect output value (P < 0.05) and the interaction between BD and position has significant influences on output value (P < 0.05). Moreover, BD shows no synergistic effects with cultivar and position. Figure 5 shows that output values of tobacco leaves in different positions of each cultivar decrease with increasing BD. Under the interaction between cultivar and BD, output values of lower and middle leaves of K326 and Hongda cultivars with BD < 25% are significantly different. The output values of three flue-cured tobacco cultivars with BD < 25% show significant differences to those with BD > 75% in each position.

Effects of different BDs on chemical characteristics of different cultivars. Effects of different BDs
on starch content. According to Table 2, BD, cultivar, position, and their interaction demonstrate significant influences on the starch content (P < 0.05). Figure 6 shows that, with increasing BD, the starch contents all rise in tobacco leaves at each position of the three cultivars. At the same BD, when BD is smaller than 25%, the starch contents in lower leaves of K326 and Hongda cultivars are significantly lower than those in middle and upper leaves. When BD ranges from 25% to 50%, starch contents in lower leaves of K326 and Yunyan87 cultivars are significantly lower than those in middle and upper leaves; at each position, the Hongda cultivar shows significant differences in starch contents. When the BD changes from 50% to 75%, starch contents in lower leaves of K326 cultivar are significantly lower than those in middle and upper leaves. Starch contents at each position of the  www.nature.com/scientificreports www.nature.com/scientificreports/ Hongda cultivar are significantly different, starch contents in lower leaves of the Yunyan87 cultivar are significantly higher than those in middle and upper leaves. When the BD exceeds 75%, starch contents in middle and lower leaves of Hongda cultivar are significantly lower than that in the upper leaves, while those in the middle and lower leaves of Yunyan87 cultivar are significant.
Effects of different BDs on the contents of total sugar. Based on Table 2, BD, cultivar, position and their interaction significantly affect the total sugar content (P < 0.05). Figure 7 shows that, as the BD increases, the total sugar contents in tobacco leaves at each position of different cultivars decrease. At the same BD, when the BD is smaller than 25%, lower and upper leaves of K326 exhibit significant differences in total sugars content: when the BD is between 25% and 75%, the total sugar content in lower leaves of K326 shows significant differences with those in www.nature.com/scientificreports www.nature.com/scientificreports/ middle and upper leaves. For the Hongda cultivar, the total sugar content in middle leaves is significantly higher than that in upper leaves. When the BD is higher than 75%, the total sugar contents in upper leaves of K326 and Yunyan87 cultivars are significantly higher than those in middle and lower leaves. Meanwhile, significant differences are found in contents of total sugar in leaves at different positions of Hongda cultivar. Table 2, BD, cultivar, position, and their interactions have significant influence on the reducing sugar content (P < 0.05). Figure 8 shows that the reducing sugar content in tobacco leaves in different positions of the three cultivars decreases with increasing BD. The reducing sugar contents in middle and upper leaves of Hongda and Yunyan87 cultivars are relatively high, while those in lower www.nature.com/scientificreports www.nature.com/scientificreports/ leaves of Hongda and K326 cultivars are high. When the BD is the same, namely BD < 25%, the lower leaves of the three cultivars show significant differences in reducing sugar content of their middle and upper leaves. When the BD ranges from 25% to 50%, the reducing sugar contents in upper leaves are significantly different from those in middle and lower leaves of the K326 cultivar, and the upper and middle leaves of Yunyan87 show significant differences therein to lower leaves. At a BD of 50% to 75%, the upper and middle leaves of the K326 cultivar demonstrate significant differences in the reducing sugar content with lower leaves, and middle and lower leaves of Hongda and Yunyan87 cultivars have significant differences therein. When the BD is higher than 75%, the reducing sugar contents in middle and lower leaves of the Hongda cultivar are significant.

Effects of different BDs on reducing sugar content. Based on
Effects of different BDs on protein contents. In accordance with Table 2, BD, cultivar, position, and their interaction have significant effects on protein content (P < 0.05). Figure 9 shows that, when the BD is smaller than 25%, protein contents in upper and lower leaves of the Hongda and K326 cultivars are significantly different, and the protein contents of Yunyan87 cultivar in lower leaves show significant differences with that in the middle leaves. When the BD ranges from 25% to 50%, protein contents in lower leaves of Hongda and Yunyan87 cultivars are significantly higher than those in middle and upper leaves. When BD is between 50% and 75%, the differences between protein contents in upper and middle leaves of the Hongda cultivar are significant. When the BD exceeds 75%, protein contents in middle leaves of the K326 cultivar are significantly higher than those in upper and lower leaves; middle leaves of the Hongda cultivar contain significantly lower protein contents than upper and lower leaves.
Effects of different BDs on the total nitrogen content. According to Table 2, BD, cultivar, position, and their interaction show significant effects on the total nitrogen content (P < 0.05). Figure 10 shows that, at the same BD of less than 25%, total nitrogen contents in lower leaves of the K326 cultivar are significantly lower than that in middle leaves. When BD varies from 25% to 50%, the total nitrogen contents in middle and upper leaves of the K326 cultivar are significantly higher than that in lower leaves. When the BD is in the range of 50% to 75%, significant differences in total nitrogen contents are found in each position of the K326 cultivar, while the total nitrogen contents in upper leaves of the Hongda cultivar are significantly higher than those in middle and lower leaves. When the BD is higher than 75%, the total nitrogen contents in upper leaves of the K326 cultivar are significantly lower than those in middle and lower leaves. The total nitrogen contents in middle leaves of the Hongda cultivar are significantly lower than those in upper and middle leaves.
Effects of different BDs on nicotine content. In accordance with Table 2, cultivar and position, as well as their interaction exert significant influences on nicotine content (P < 0.05) and interaction between BD and cultivar also significantly affects the nicotine content. Figure 11 shows that, at the same BD of less than 25%, the nicotine contents in lower leaves of the K326 cultivar are significantly lower than those in middle and upper leaves, while nicotine contents in middle and lower leaves of the Hongda and Yunyan87 cultivars are significantly lower than that in upper leaves. When BD ranges from 50% to 75%, nicotine contents in lower leaves of the K326 and Yunyan87 cultivars are significantly lower than those in middle and upper leaves. At other BDs, nicotine contents in middle and lower leaves of the Yunyan87 cultivar are significantly lower than that in upper leaves.
Effects of different BDs on polyphenol content. As shown in Table 2, BD and position exert significant effects (P < 0.05) on the polyphenol content and BD has a synergistic effect with position. Figure 12 shows that the polyphenol content in different positions of the three flue-cured tobacco cultivars reduces with increasing BD.
When the BD is less than 25%, the polyphenol contents in upper leaves of the Hongda cultivar are significantly higher than those in middle and lower leaves. When BD ranges from 50% to 75%, the polyphenol contents in lower leaves of the K326 and Hongda cultivars are significantly lower than those in upper leaves. When the BD exceeds 75%, the polyphenol contents in middle and lower leaves of the K326 cultivar are significantly lower than that in upper leaves, and the polyphenol content in each position of the Hongda cultivar is significantly different.

Effects of different BDs on industrial usability of different flue-cured tobacco cultivars. Effects
of different BDs on shatter resistance index. As shown in Table 3, BD, variety, and position exert significant effects (P < 0.05) on the shatter resistance index. BD has a synergistic effect with position. Figure 13 shows that the shatter resistance index (screen aperture < 1 mm) of tobacco leaves in different positions of each cultivar increase with increasing BD, while the shatter resistance index (screen aperture ≥ 2 mm) of tobacco leaves in different positions of each cultivar decrease with increasing BD. When the BD is less than 25%, the shatter resistance index (screen aperture ≥ 2 mm) in different positions of each cultivar is significantly greater than that when the BD exceeds 75%; however, when the BD is less than 25%, the shatter resistance index (screen aperture < 1 mm) in different positions of each cultivar is significantly lower than that when the BD exceeds 75%.
Effects of different BDs on sensory evaluation score. Based on Table 3, BD, cultivar, and position show significant influences on sensory evaluation score (P < 0.05). Furthermore, interactions of BD and cultivar, cultivar and position, and the three combined also significantly affect the sensory evaluation score. Figure 14 shows that the sensory evaluation score in each position of different cultivars decreases with increasing BD. At the same BD, when is less than 25%, the sensory evaluation score of upper leaves of the K326 cultivar is significantly higher  www.nature.com/scientificreports www.nature.com/scientificreports/ than those of middle and lower leaves. When the BD ranges from 25% to 50%, middle and upper leaves of the Yunyan87 cultivar present significantly different sensory evaluation scores. For BDs between 50% and 70%, the sensory evaluation score of middle leaves of the K326 cultivar is significantly lower than those of upper and lower leaves, and the sensory evaluation score of middle leaves of the Hongda cultivar is significantly higher than those of upper and lower leaves. Furthermore, when the BD exceeds 75%, the sensory evaluation score of upper leaves of the K326 cultivar is significantly higher than those of middle and lower leaves.
Extraction, isolation, and structure determination of the main chemical components of tobacco with grey speckles. Extracted solutions of the samples of leaves with and without grey speckles were analysed by using thin-layer chromatography. At a ratio of methylene chloride to methanol of 5:1 and a ratio www.nature.com/scientificreports www.nature.com/scientificreports/ of methylene chloride, methanol and formic acid of 5:1:0.1, ethanol extracts were separated and divided through a methylene chloride/methanol system and finally 220 mg of component Fr.Ab1a was separated and purified by utilising gel LH-20 (Fig. 15). Based on comparative analysis, it is known that parts with grey speckles and parts without grey speckles on tobacco leaves have different components (Fig. 16) and the components in the grey speckles are marked in red circles in Fig. 16. Sample Fr. Ab1a was separated using the Agilent1200 high-performance liquid chromatograph with an ultraviolet detector and YMC-pack ODS-A columns with a particle size of 5 μm and dimensions of 10 × 250 mm and passed through a system containing 35% MeCN/H 2 O solution at a flow rate of 2.0 mL/min. On this basis, 9.7 mg of the component 3-acetyl-6,7-dimethoxycoumarin (YC-ZJF) of grey speckles on leaves could be obtained (t R = 20 min) (Fig. 2). www.nature.com/scientificreports www.nature.com/scientificreports/ Figure 16 shows that the component from leaves with grey speckles has strong chromophores. After extracting extraction and separating tobacco with grey speckled leaves, the main chemical component, 9.7 mg of YC-ZJF was isolated from the tobacco with grey speckled leaves by repeated high-performance liquid chromatography (HPLC). The structure of the isolated compound (YC-ZJF) (Fig. 17) was determined to be 3-acetyl-6,7-dimethoxycoumarin as evinced by analysing its 1-d and 2-d NMR spectra (Figs. 18-24). Interestingly, 3-acetyl-6,7-dimethoxycoumarin was only reported as a synthetic compound, instead of a natural product (Takadate et al., 1995). Thus, 3-acetyl-6,7-dimethoxycoumarin should be an artifact during flue-curing of tobacco, therefore, 3-acetyl-6,7-dimethoxycoumarin should be one of the main components or related components in tobacco with grey speckled leaves.  www.nature.com/scientificreports www.nature.com/scientificreports/ mushroom preserved at room temperature [28][29][30] . There are many factors reflecting the quality of tobacco leaves, among which the proportion of superior tobacco directly reflects the economic value of tobacco leaves. In this experiment, the proportion of superior tobacco and output value of tobacco leaves in different positions of the three cultivars of flue-cured tobacco all decrease with increasing BD. The proportion of superior tobacco is an important index reflecting level of production technology used in the processing of tobacco leaves. According to the natural characteristics of tobacco plants, the theoretical maximum proportion of superior tobacco is 60% evaluated based on the number of leaves of a single plant 20 . Therefore, different BDs had different influences on the proportions of superior tobacco in different positions, however, these three cultivars all behaved such that the proportion of superior tobacco in the same position decreased with increasing BD. The output value is not only related to the quality of tobacco leaves, but is also correlated to the level of flue-curing technology used when www.nature.com/scientificreports www.nature.com/scientificreports/ processing tobacco leaves. Tobacco with grey speckled leaves has poor combustion characteristics, a low-quality aroma, and a small volume of aroma 31 . Moreover, it presents a strengthened biting taste and reduced quality. The higher the BD is, the lower the quality of tobacco leaves, so the output value also decreases.

Effects of different BDs on chemical characteristics of different cultivars. Effects of different
BDs on carbon metabolism conventional chemical components of different cultivars. Tobacco is an important economic crop. Its chemical composition is an important internal factor determining the quality of tobacco leaves and appropriate coordination of chemical components determines the value of tobacco 32 . The chemical www.nature.com/scientificreports www.nature.com/scientificreports/ components of tobacco leaves and their proportions decide the quality of tobacco leaves, thus exerting significant influence on the smoking quality of cigarettes 33 .
Decomposition, transformation, and accumulation of starch determine the quality and appearance rating of tobacco leaves 34 . Leffingwell 35 believed that a high starch content in tobacco leaves during harvesting can cause imbalances in the chemical quality and the ratio of sugar to nicotine of flue-cured tobacco leaves and decrease their nicotine content. In addition, it can result in a smooth surface of cured tobacco leaves, raise proportions of tobacco leaves with miscellaneous colours and greenish tobacco leaves, and reduce industrial usability. In this experiment, with increasing BD, the starch contents in tobacco leaves in each position of the K326, Hongda, and Yunyan87 cultivars all significantly increase. On the whole, starch contents in upper and middle leaves of the three cultivars are significantly higher than those in lower leaves, however, for the Yunyan87 cultivar, when the BD ranges from 50% to 75%, the starch contents in lower leaves are significantly higher than those in middle www.nature.com/scientificreports www.nature.com/scientificreports/ and upper leaves. When the BD exceeds 75%, the starch contents in lower leaves of the Hongda cultivar are significantly higher than that in middle leaves. The BD affects the starch contents in leaves of different flue-cured tobacco cultivars, resulting in uncoordinated chemical compositions and low-quality flue-cured tobacco leaves.   Table 3. Analysis of variance for the effects of browning degree, variety, position, and their interactions on industrial usability. www.nature.com/scientificreports www.nature.com/scientificreports/ Sugar content is an important index used when evaluating the quality of tobacco and its content reflects the carbon-supply capacity 36 . When sugar content of tobacco leaves is too low, the biting taste strengthens. If the sugar content is too high, the smoke is acidic, which influences the acid-base equilibrium inherent to smoking, so that smoke becomes tasteless and the tar content of smoke increases 37 . Based on analysis of the results, the reducing sugar contents in different positions of the three cultivars all decrease with increasing BD. Under the four BDs, the reducing sugar contents in lower leaves of the K326 and Hongda cultivars are relatively high, while a high reducing sugar content is found in middle and upper leaves of the Yunyan87 cultivar. For lower tobacco leaves of the K326 and Hongda cultivars, with the increase of BD, the reducing sugar content decreases more in comparison with that in upper leaves: however, with increasing BD, the reducing sugar contents in different positions of   www.nature.com/scientificreports www.nature.com/scientificreports/ the Yunyan87 cultivar are reduced less compared with those in the K326 and Hongda cultivars, therefore, the BD exerts greater influences on the K326 and Hongda cultivars in comparison with that over the Yunyan87 cultivar and exerts greater influence on the reducing sugar content in lower leaves of the K326 and Hongda cultivars. The total sugar contents in different positions of the three cultivars all significantly decrease with increasing BD and, in particular, the decrease is greatest in middle leaves of the K326 cultivar. The reduction in total sugar content can affect the smoking quality of flue-cured tobacco, so that the biting taste increases and quality decreases 38 .

Effects of different BDs on nitrogen metabolism conventional chemical components of different cultivars.
In the production of flue-cured tobacco, nitrogen is the most important nutrient element influencing yield and quality of tobacco leaves and the total nitrogen and alkaloid contents in flue-cured tobacco leaves reflect nitrogen-supply capacity 39 . In this experiment, the total nitrogen contents in tobacco leaves in each position of the Hongda and Yunyan87 cultivars do not show significant differences with the increase of BD, however, tobacco leaves in each position of the K326 cultivar show that the total nitrogen content increases significantly at high BD, therefore, the BD more significantly influences the total nitrogen content in the K326 cultivar than in the Hongda and Yunyan87 cultivars. Clarke 40      are differences in cultivars and positions. For the K326 cultivar, nicotine contents in upper and middle leaves are significantly higher than that in lower leaves, while the Hongda and Yunyan87 cultivars demonstrate that nicotine contents in upper leaves are significantly higher than those in middle and lower leaves. On the whole, the three cultivars all contain the highest nicotine content in the upper leaves, with the lowest content in their lower leaves 41 .

Effects of different BDs on polyphenol components in different cultivars.
Polyphenol substances have important influences on the physiological and biochemical activities of tobacco, the colour and lustre of tobacco leaves, the aroma and taste of cigarettes, and their physiological strength 42 . In particular, as an important aroma precursor of flue-cured tobacco, it can be decomposed into a variety of aroma substances and be combined with proteins or be catalysed by polyphenol oxidase under the browning reaction 43 . This study showed that, with increased BD, the polyphenol content decreases. Chlorogenic acid and rutin are the main components of polyphenol substances. Sun et al. 44 demonstrated that the chlorogenic acid content is mainly affected by cultivar, followed by interaction between altitude and cultivar. Our result also showed that cultivar exerts significant effects (P < 0.05) on the polyphenol content and cultivar has a synergistic effect with BD and position.

Effects of different BDs on other industrial indices of different cultivars. Effects of different BDs on
shatter resistance index of different cultivars. The physical resistance to further processing of tobacco leaves is the focus of enterprises involved in re-drying tobacco and the better the physical resistance to further processing, the less tobacco is lost during defoliation 45 . The data show that the physical resistance to further processing, as represented by shatter resistance, decreases with increasing yellowing. The shatter resistance index (screen aperture <1 mm) of tobacco leaves in different positions of each cultivar increase with increasing BD, while the shatter resistance index (screen aperture ≥2 mm) of tobacco leaves in different positions of each cultivar decrease with increasing BD, which indicates that the capability of tobacco leaves to resist shattering decreases with increasing BD.  www.nature.com/scientificreports www.nature.com/scientificreports/ Effects of different BDs on sensory qualities of different cultivars. Tobacco is meant to be flavoursome and its quality is mainly evaluated through smoking; in this experiment, the sensory qualities of the three cultivars all decrease with increasing BD. Under each BD, the sensory quality of upper leaves of the K326 cultivar is higher than those in other positions. Moreover, the sensory qualities of middle leaves of the Hongda cultivar under each BD are higher than those of the other two cultivars. The sensory evaluation score is affected by chemical components, such as starch, total sugar, total nitrogen, and nicotine 46 . The influences of chemical components of flue-cured tobacco on sensory quality are generally explained through the theory of sugar-alkali equilibrium 47 . When tobacco leaves contain the ideal chemical components, the sensory quality of tobacco leaves is relatively high. For example, it is appropriate that the total sugar content is 20% to 28% and the difference between the contents of two sugars is less than 5% 48 . The experimental results show that the reducing sugar and total sugar contents in tobacco leaves in different positions of the three cultivars all decrease with increasing BD and the total sugar content is significantly positively correlated with sensory quality.
Discussion of structure identification of substances of grey speckles on leaves of flue-cured tobacco. Previous research into the causes of formation of tobacco with grey speckled leaves mainly focus on polyphenol oxidase and polyphenol substances 49 . Our previous study clarified that the identity of the bridging oxygen must be a water molecule in the tyrosinase which is a typical polyphenol oxidase 50 purified from a new polyphenol oxidase called PPO II found in tobacco and proved to accumulate in the injured parts of tobacco leaves. The poor quality of fresh tobacco leaves is one of causes of the appearance of tobacco with grey speckled leaves. If fresh tobacco leaves are infected by disease, much polyphenol oxidase accumulates therein and tobacco with grey speckled leaves become more likely during the curing process 51 . This experiment found that the structure of the component derived from the tobacco leaves with grey speckles is YC-ZJF, which was previously reported as a synthetic compound, instead of a natural product; however, due to the complexity of the formation of grey speckles, it is difficult to conclude that YC-ZJF is the major component of grey speckles. It is also possible that YC-ZJF might be an intermediate if grey speckles are largely composed of macromolecules. Nonetheless, the identification of the structure of YC-ZJF is still important to the further elucidation of the mechanisms of formation of tobacco with grey speckled leaves. Meanwhile, according to the identified structure of the substance, the corresponding components capable of reducing the presence of the substance forming grey speckles on tobacco leaves can be investigated, so as to reduce the incidence of tobacco with grey speckled leaves in the flue-curing process.

Conclusion
The effects of the browning reaction of flue-cured tobacco on tobacco value and industrial usability were elucidated along with the structural identification of grey matter thereon. With increasing BD, the proportion of superior tobacco and output value of tobacco leaves decrease regardless of cultivar or position. For carbon metabolism, the starch is not degraded with increased BD, and total sugar and reducing sugar contents decrease accordingly. For nitrogen metabolism, the protein, total nitrogen, and nicotine contents change albeit slightly. The industrial indices, including shatter resistance index and smoking quality decrease with increasing BD. More importantly, the chemical structure of the main components of tobacco with grey speckled leaves was identified and the main component of the separated parts with grey speckles was determined to be 3-acetyl-6,7-dimethoxycoumarin. These results demonstrated that the increase of BD reduces the economic value and industrial usability of tobacco. In the planting and production of tobacco, it is particularly important to reduce the incidence of tobacco with grey speckled leaves.

Data availability
All data generated or analysed during this study are included in this published article (and its Supplementary Information files).