Responses of Aroma Related Metabolic Attributes of Opisthopappus longilobus Flowers to Environmental Changes

Opisthopappus longilobus (Opisthopappus) and its descendant species, Opisthopappus taihangensis, commonly thrive on the Taihang Mountains of China. Being typical cliff plants, both O. longilobus and O. taihangensis release unique aromatics. To determine the potential differentiation and environmental response patterns, comparative metabolic analysis was performed on O. longilobus wild flower (CLW), O. longilobus transplant flower (CLT), and O. taihangensis wild flower (TH) groups. Significant differences in the metabolic profiles were found, not within O. longilobus, but between O. longilobus and O. taihangensis flowers. Within these metabolites, twenty-eight substances related to the scents were obtained (one alkene, two aldehydes, three esters, eight phenols, three acids, three ketones, three alcohols, and five flavonoids), of which eugenol and chlorogenic were the primary aromatic molecules and enriched in the phenylpropane pathway. Network analysis showed that close relationships occurred among identified aromatic substances. The variation coefficient (CV) of aromatic metabolites in O. longilobus was lower than O. taihangensis. The aromatic related compounds were significantly correlated with the lowest temperatures in October and in December of the sampled sites. The results indicated that phenylpropane, particularly eugenol and chlorogenic, played important roles in the responses of O. longilobus species to environmental changes.


Introduction
The metabolic processes of organisms are dynamic and complex systems that are regulated by multiple factors [1]. From the most upstream genomic DNA and mRNA, and from proteins to traits, organisms continually adjust their metabolic responses and metabolites to maintain a dynamic balance between their internal and external environments [2]. All reactions, from genes to traits, are the final comprehensive result of the co-regulation of genes and the external environment [3]. Thus, changes in the metabolites of organisms, the relationships between them, and their external manifestations are essential for elucidating their adaptations and evolutionary processes [4]. Conversely, metabolic data can more accurately reflect the physiological status and phenotypic presence of plants while clearly identifying minor changes in genetic and protein expression [2].
At different growth stages, differences in the metabolic activities of plants can facilitate the exploration and tracking of changing metabolite distribution patterns [5]. Based on the differences and interactions between metabolites, environmental influences on the metabolites of plants, as well as their adaptive pathways and metabolic networks, may be inferred [6].
Generally, continuous and dynamic changes in the environment typically result in adaptions by plants [7]. These adaptations involve the production and accumulation of a diverse set of metabolites, ranging from signaling hormones and primary metabolites to a wide array of specialized multifunctional metabolites [8,9]. When plants are subject to were the adaptive responses to the disruptions caused by different habitats?
To answer the above questions, we initially identified the types of metabolites i volved in the flowers of the two species (compared O. longilobus wild flowers, O. longilob transplant flowers, and O. taihangensis wild flowers). Subsequently, the different aroma metabolites and metabolic pathways were comparatively analyzed between the O. long lobus ancestor and O. taihangensis descendant. Finally, the potential mechanisms of O. lo gilobus in response to heterogeneous habitats were elucidated.

Total Metabolism of the Flowers
In the flower metabolites, a total of 95,655 features were detected in the positive io mode, and 36,478 features were detected in the negative ion mode. 5439 metabolites we obtained in the positive and negative ion mode, which accounted for 97.6% of those of t flowers. Based on quantitative analysis, 14,262 and 10,004 high-quality substances we obtained in the positive and negative ion modes, respectively.
Among the 3 groups, 14,713 substances were shared through the screened Venn di gram (Figure 1), in which there were 1954 metabolites between CLW and CLT, 1245 b tween CLW and TH, and 695 between CLT and TH. There were 638 metabolites that were unique to CLW, 748 to CLT, and 2248 to TH Significantly distinct metabolic expressions occurred between the CLW, CLT, and T groups. KEGG enrichment analysis revealed that the differential metabolites were enriche into 47 pathways in CLW/CLT, 59 metabolic pathways in TH/CLW, and 61 pathways TH/CLT (Table A1). Meanwhile, the significantly different metabolic pathways were a There were 638 metabolites that were unique to CLW, 748 to CLT, and 2248 to TH. Significantly distinct metabolic expressions occurred between the CLW, CLT, and TH groups. For the O. longilobus wild flowers and O. longilobus transplant flowers (CLW/CLT), the differential metabolites were upregulated by 919 and downregulated by 1189. The differential metabolites of wild flowers of O. taihangensis and those of O. longilobus (TH/CLW) were upregulated by 4020 and downregulated by 3722. Between TH/CLT, the differential metabolites were upregulated by 2560 and downregulated by 3135.
KEGG enrichment analysis revealed that the differential metabolites were enriched into 47 pathways in CLW/CLT, 59 metabolic pathways in TH/CLW, and 61 pathways in TH/CLT (Table A1). Meanwhile, the significantly different metabolic pathways were all involved in secondary plant metabolite biosynthesis, phenylpropanoid biosynthesis, and arginine biosynthesis between TH/CLT, TH/CLW, and CLW/CLT.
The most highly expressed metabolite for CLT was eugenol ( Figure 2 and Table 1), followed by caffeate, carvacrol, thymol, and chlorogenate. The least expressed metabolite was hesperetin.
The most highly expressed metabolite for CLT was eugenol ( Figure 2 and Table  followed by caffeate, carvacrol, thymol, and chlorogenate. The least expressed metabol was hesperetin.    For CLW, the most highly expressed metabolite was phenyl acetate, followed by chrysoeriol, hesperetin, dioctyl phthalate, and caffeate. The least expressed metabolite was p-Cresol. In TH, chlorogenate was the most highly expressed metabolite, followed by caffeate, scopoletin, dioctyl phthalate, and syringetin. The least expressed metabolite was p-Cresol. In addition, three groups could be distinguished from each other according to PCA ( Figure A1). The overall differences in aromatic substances were obvious among three groups, which indicated that the aromatic compounds in CLT, CLW, and TH were significantly different.

Coefficient of Variation of Aromatic Compounds
Based on the identified aromatic substances, the coefficients of variation (CV) ( Table A2) of fragrance metabolism were calculated. For CLT, the CV value ranged from 0.1 to 48.72, whereas for CLW, it ranged from 0.58 to 55.25, and, for TH, this index ranged from 2.17 to 80.90.
Furthermore, chlorogenic and eugenol had the relatively higher CV values among the identified fragrance metabolites.

Network Diagram among Fragrance Metabolites
A correlation metabolic network was developed for the odor metabolites ( Figure 3). Within the network diagram, closely linked metabolites were found likely involved in the same or related metabolic pathways. Most of the metabolites were relatively well connected, which implied that these aromatic substances were fundamental for the growth and survival of O. longilobus and O. taihangensis. same or related metabolic pathways. Most of the metabolites were relatively well connected, which implied that these aromatic substances were fundamental for the growth and survival of O. longilobus and O. taihangensis.

Environmental Correlation of Aromatic Substances
To explore the environmental correlation of aromatic substances, we firstly investigate the key environmental factors in each sampled site. Then, the correlation analysis was performed between the identified aromatic substances and environmental factors.
Using the PCA correlation matrix of environmental factors, fourteen variables were primary factors (Table A3). When simultaneously considering the VIF results (Table A3), it was found that two environmental factors, namely the lowest temperatures in October and December, were the most important factors for the sampled sites (Table 2). Thus, the correction analysis was carried out with the two factors.
According to correlation analysis (Tables A3 and 2), chlorogenic was significantly correlated with the minimum temperature in December, and the Pearson's correlation coefficient was 0.72105. Eugenol was significantly correlated with the minimum temperatures in October and December, and the Pearson's correlation coefficients were 0.88483 and 0.82039, respectively.

Environmental Correlation of Aromatic Substances
To explore the environmental correlation of aromatic substances, we firstly investigate the key environmental factors in each sampled site. Then, the correlation analysis was performed between the identified aromatic substances and environmental factors.
Using the PCA correlation matrix of environmental factors, fourteen variables were primary factors (Table A3). When simultaneously considering the VIF results (Table A3), it was found that two environmental factors, namely the lowest temperatures in October and December, were the most important factors for the sampled sites (Table 2). Thus, the correction analysis was carried out with the two factors.
According to correlation analysis (Tables 2 and A3), chlorogenic was significantly correlated with the minimum temperature in December, and the Pearson's correlation coefficient was 0.72105. Eugenol was significantly correlated with the minimum temperatures in October and December, and the Pearson's correlation coefficients were 0.88483 and 0.82039, respectively.

Different Metabolites of the Flowers
In general, the regulation of primary metabolite allocation can affect fundamental processes such as the growth and development of an organism. In addition, other processes aimed at increasing plant fitness in specific environments are more associated with specialized metabolic processes, such as those involved with protection against herbivores and pathogens, reduction of oxidative stress, or competition with other plants [29]. During various developmental processes, plants need to continuously adjust the distribution of metabolites to maintain growth and survival under changing environments [30].
When O. longilobus individuals were transplanted from the wild site to the transplant garden, this species adjusted its metabolic pathways, such as phenylpropane biosynthesis, to respond the changed environment to ensure that its flowers bloomed normally.
For plants, arginine biosynthesis pathway was an important one of metabolites pathways. Within arginine biosynthesis pathways, arginine makes a key function in the regulation of abiotic and biotic stress in plants. When plants are exposed to various abiotic stresses, the expression of arginase genes is not only upregulated but also the enzyme activity is increased. When plants are subjected to biotic stresses, such as insect feeding and pathogen infestation, arginase also displays a significant effect. In addition, arginine promoted root growth and improved the salt tolerance of plants [31].
For O. longilobus and O. taihangensis, the significantly enriched arginine biosynthesis pathway might have affected the florescence development and the tolerance to the various stresses. From the evolutionary viewpoint, the O. taihangensis descendant would occupy different surroundings compared with its O. longilobus ancestor. Through adjusting its metabolites and metabolic pathways, O. taihangensis must ensure its development and survival when facing abiotic and biotic stresses from a novel environment.

Different Aromatic Substances of the Flowers
Plants produce a diversity of secondary metabolites, while generating and storing an extensive range of volatile organic compounds in their flowers. Typically, aromatic substances are important metabolites in plants for attracting insects, resisting diseases and pests, and adapting to altered habitats [32,33]. Further, the scents attract and repel various pollinators for different pollination processes [34,35].
In the CLW, CLT, and TH groups, the pathways related to flower fragrances were primarily phenylpropanoid, flavone, and flavonol biosynthesis, which are mainly composed of phenylpropanes (such as chlorogenic and eugenol) (Figure 4), which was consistent with the general pattern of the chemical composition of floral scents [36].
promoted root growth and improved the salt tolerance of plants [31].
For O. longilobus and O. taihangensis, the significantly enriched arginine biosynthesis pathway might have affected the florescence development and the tolerance to the various stresses. From the evolutionary viewpoint, the O. taihangensis descendant would occupy different surroundings compared with its O. longilobus ancestor. Through adjusting its metabolites and metabolic pathways, O. taihangensis must ensure its development and survival when facing abiotic and biotic stresses from a novel environment.

Different Aromatic Substances of the Flowers
Plants produce a diversity of secondary metabolites, while generating and storing an extensive range of volatile organic compounds in their flowers. Typically, aromatic substances are important metabolites in plants for attracting insects, resisting diseases and pests, and adapting to altered habitats [32,33]. Further, the scents attract and repel various pollinators for different pollination processes [34,35].
In the CLW, CLT, and TH groups, the pathways related to flower fragrances were primarily phenylpropanoid, flavone, and flavonol biosynthesis, which are mainly composed of phenylpropanes (such as chlorogenic and eugenol) (Figure 4), which was consistent with the general pattern of the chemical composition of floral scents [36]. Metabolism occurs between phenylpropane metabolism and other secondary metabolic pathways, which maintains a dynamic balance between phenylpropane metabolism and resisting environmental degradation [37]. When downregulated, chlorogenic based metabolites can be metabolized to storage forms such as glycosyl based derivatives or decomposed quinic acid and caffeic acid, and can be further metabolized to more complex molecules such as lignin [38]. These contribute to plant development and plant-environment interactions, and phenylpropanoid-based polymers (such as lignin) are required for Metabolism occurs between phenylpropane metabolism and other secondary metabolic pathways, which maintains a dynamic balance between phenylpropane metabolism and resisting environmental degradation [37]. When downregulated, chlorogenic based metabolites can be metabolized to storage forms such as glycosyl based derivatives or decomposed quinic acid and caffeic acid, and can be further metabolized to more complex molecules such as lignin [38]. These contribute to plant development and plant-environment interactions, and phenylpropanoid-based polymers (such as lignin) are required for mechanical support to facilitate growth and long-distance transport of water and nutrients [39].
Phenylpropanoid homeostasis between the different branches of phenylpropanoid metabolism exhibits extraordinary complexity and a high-level of plasticity during successive developmental stages in response to environmental stimuli and changes [39,40]. Lignin encapsulates carbohydrates composed of cellulose and hemicellulose to form a composite wood fiber barrier to resist the attack and destruction of plant tissues by microorganisms and the surrounding environment [41].
The metabolites of the phenylpropanoid metabolic pathway contribute to the upright growth of the plant, allowing it to better photosynthesize. At the same time, the involved metabolites also effectively protect plants from UV light, diseases, and insects for healthy plant growth. Accordingly, the phenylpropanoid pathway would make O. longilobus grow better with the changed habits, ensure its floral development, and maintain its survival and evolution.
Chlorogenic and eugenol are both involved in the phenylpropane pathway. For CLW, CLT, and TH (Figure 4), chlorogenic is upstream and eugenol is downstream. Moreover, the relative substances of synthetic chlorogenic are highly expressed in CLW and CLT ( Table 1).
The first step of the chlorogenic metabolic pathway involves the dissociation of ammonia from l-phenylalanine via phenylalanine ammonia lyase and the production of transcinnamic acid. This enzyme belongs to the aromatic amino acid lyase family, is the first key enzyme in the phenylpropane pathway [42], and is related to most bioactive metabolites, such as flavonoids and anthocyanins. Meanwhile, the caffeoyl-CoA is the final step in the synthesis of chlorogenic, where 4-coumaryl-CoA ligase is one of the crucial enzymes in the phenylpropane pathway toward the metabolism of other substances, which can catalyze cinnamic acid and its hydroxyl groups. Thus, 4-coumaryl-CoA ligase is of great value in the biosynthesis of phenylpropane compounds such as chlorogenic and flavonoids [43].
Intermediates of the phenylpropane metabolic pathway and their secondary metabolites can not only improve the disease resistance of plants, but also regulate and promote plant resistance to abiotic stresses such as low temperatures, high temperatures, and ultraviolet radiation. This might be an explanation for the difference in floral aromatic substances between two groups (CLW and CLT) of O. longilobus under different surroundings.
Being the main aromatic compound, chlorogenic plays a key role in the attraction of pollinators and against herbivores and pathogens. Further, chlorogenic is produced via the shikimic acid pathway during aerobic respiration, which exhibits anti-inflammatory, antibacterial, antiviral, and antidepressant pharmacological effects in vitro [44]. In this study, the content of chlorogenic in TH was higher than that of CLT and CLW ( Table 1). As the descendant of O. longilobus, additional chlorogenic might induce O. taihangensis to not only attract more and different pollinators to facilitate its reproduction, but also to respond to different environments compared with its ancestor and to better survive and develop. Certainly, higher contents of chlorogenic also may be responsible for the enhanced flower color and relatively different flower size of O. taihangensis in contrast to O. longilobus.
On the other hand, eugenol can attract biological pollinators, facilitate seed transmission, and protect against herbivores [45]. It is an important type of phenylpropane volatile that is commonly found in plant flowers and mature fruits, which possesses a variety of biological and therapeutic effects [13]. Eugenol had the highest content in CLT, followed by CLW (Table 1). When O. longilobus was transplanted into the transplant garden from the wild growth site, eugenol would accumulate to enhance survival under changing environmental conditions. Moreover, eugenol is a semi-volatile phenolic compound derived from plants that has a strong carnation musk odor [46] that creates the different fragrances for O. longilobus different from O. taihangensis.
To adapt to localized environmental conditions, the plants generate many specialized metabolites that contribute to their health and survival and play roles in their capacities [47]. The emission of fragrances from flowers is regulated by metabolic processes [48]. The biosynthesis and emission of floral fragrances are modulated by the stages of flower maturity, circadian rhythms, and by other environmental factors such as temperature [49,50]. Environmental conditions not only influence the vaporization of volatile compounds from flowers, but also their biosynthesis, particularly under different ambient air temperature regimes, which are known to play major roles in the biosynthesis and release of floral fragrances [50,51].
A significant correlation was found between chlorogenic and the minimum temperature in October, and between eugenol and the minimum temperature in October and December. These two environmental factors might directly affect the expression of chlorogenic and eugenol, or indirectly affect the expression of chlorogenic and eugenol synthase. It was reported that low temperatures can induce the expression of phenylalanine ammonia lyase (PAL) and other phenylpropane metabolism genes [52]. Flower buds could only be formed after low temperature treatment, which may be reasonable for O. longilobus and O. taihangensis species.

Sample Sites and Materials
The Shennong Mountains inhabit Henan Province, while the Xiangtang Mountains reside in Hebei Province. Both of these mountains occupy branches of the Taihang  This park is located in Fengfeng Mining District, with an annual dryness of 1.4 degrees. Maximum wind speed in the calendar year is 14 m/s. Spring precipitation accounts for 12.7% of the region's annual precipitation [53]. The environmental conditions of the park, such as annual average temperature, extreme maximum temperature, and annual average rainfall, are different with the Xiangtang Mountains, though both belong to the same climatic zone. From July to October of 2020, O. longilobus flowers were collected, including wild flowers and transplant flowers. Concurrently, O. taihangensis wild flowers were collected as comparative samples from the Shennong Mountains due to the phylogenetic relationship with O. longilobus [54] (Table A4).

Extraction of Total Metabolite
Due to the robustness, high selectivity and sensitivity, and a strong potential for both quantification and identification purposes, the liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) was performed in this study [55][56][57][58][59].
The collected samples were thawed on ice, 100 mg of tissue was weighed and grinded with liquid nitrogen, 120 µL of a precooled 50% methanol buffer was added, and the metabolites were extracted from 20 µL of each sample.
Subsequently, the metabolite mixture was vortexed for 1 min and incubated for 10 min at room temperature and stored at −20 • C overnight. The mixture was then centrifuged at 4000× g for 20 min, and the supernatant was transferred to 96-well plates and stored at −80 • C pending LC-MS analysis. Pooled quality control (QC) samples were also prepared by combining 10 µL of each extraction mixture.
All samples were analyzed using a TripleTOF 5600 Plus highresolution tandem mass spectrometer (SCIEX, Warrington, UK) with both positive and negative ion modes. Chromatographic separation was performed using an ultraperformance liquid chromatography (UPLC) system (SCIEX, UK). An ACQUITY UPLC T3 column (100 mm*2.1 mm, 1.8 µm, Waters, UK) was used for the reversed-phase separation. It was introduced for the separation of metabolites, and the mobile phase consisted of solvent A (water, 0.1% formic acid) and solvent B (Acetonitrile, 0.1% formic acid). The gradient elution conditions were as follows: with a flow rate of 0.4 mL/min: 5% solvent B for 0-0.5 min; 5-100% solvent B for 0.5-7 min; 100% solvent B for 7-8 min; 100-5% solvent B for 8-8.1 min; and 5% solvent B for 8.1-10 min. The column temperature was maintained at 35 • C.
The TripleTOF 5600 Plus system was used to detect metabolites eluted from the column. The curtain gas pressure was set at 30 PSI, the ion source gas1 and gas2 pressure was set at 60 PSI. The interface heater temperature was 650 • C. For the positive-ion mode, the ion spray floating voltage was set at 5 kV, and for the negative-ion mode, it was set at −4.5 kV. The MS data were acquired in the IDA mode. The TOF mass range was 60-1200 Da. Survey scans were acquired every 150 ms, and as many as 12 product ion scans were collected if the threshold of 100 counts/s was exceeded with a 1+ charge state. The total cycle time was fixed at 0.56 s. Four-time bins were summed for each scan at a pulse frequency of 11 kHz by monitoring the 40 GHz multichannel TDC detector with four-anode/channel detection. Dynamic exclusion was set for 4 s.

Determination of Differential Metabolites
The pretreatment of acquired LC-MS data was performed using XCMS software. Raw data files were converted to mzXML format and then processed using the XCMS, CAMERA, and metaX toolbox included in R software. The data was matched to in-house and public databases. The peak intensity data was further preprocessed using metaX, and metabolic substances with standard deviations of >30% were removed. The dataset groups were normalized prior to analysis using the probabilistic quotient normalization algorithm. The p value analyzed by Student's t-test was adjusted for multiple tests using an FDR (Benjamini-Hochberg), and then used for the selection of various metabolites. In conjunction with multivariate statistical analysis, the VIP value obtained by PLS-DA was analyzed to screen differential metabolic ions. The different ions met the following requirements: (1) Ratio ≥ 2 or Ratio ≤ 1/2; (2) Q value ≤ 0.05; (3) VIP ≥ 1 for the screening of differential metabolites. Assessment of nutritional value and quantitative analysis of bioactive phytochemicals was performed through the targeted LC-MS/MS method.
For annotation of differential metabolites, the metabolomic database was annotated using KEGG to retrieve all the pathways mapped by the differential metabolites. The pathways of the differential metabolites were further screened to identify key pathways that had the highest correlations with the different metabolites. Fold-change analysis and the t-test statistical test were employed to conduct BH corrections to obtain the Q-value.

Venn Diagram among Differential Metabolites
Subsequently, enrichment analysis of differential metabolites among the CLW, CLT, and TH groups was also performed. In addition, those with different expressions were screened by Origin to develop a Venn diagram.

Determination of Aromatic Substances
To evaluate the biochemical differences between O. longilobus and O. taihangensis flowers under various conditions, the metabolic scent profiles were compared. According to the original data, the differential metabolites of p < 0.05 and VIP > 1 were screened, and the similarity >700 was also screened. The non-volatile aromatic-related metabolites were then identified and classified, and a metabolic pathway map was illustrated using Visio software (Office Visio 2013, Microsoft USA, Redmond, WA, USA).

Variation Coefficient of Identified Aromatic Compounds
The variation coefficient (CV) is a statistic index used to quantify the degree of variation for each metabolite. For this study, the variation coefficient was calculated using the formula: CV = σ/µ, where σ is the standard deviation and µ is the average value. The variation coefficients of different metabolic aromatic compounds were subsequently plotted using SPSS (IBM SPSS Statistics 26, Chicago, IL, USA) software.

PCA (Principal Component Analysis) and Heat Map of Different Aromatic Compounds
A PCA diagram of different aromatic metabolites was plotted using Origin (Origin 2018) software. The relative content maps of differential metabolites were plotted with SigmaPlot version 10.0 (Systat Software Inc., San Jose, CA, USA) software. The metabolic pathway map was depicted using Visio software (Office Visio 2013, Microsoft USA).

Network Relationship among Different Aromatic Metabolites
To further explore the relationship among differential aromatic metabolites, the network analysis was displayed using Cytoscape 3.7.0 software.
A total of 103 climatic factors were obtained according to their longitude and latitude using R software. According to the variance inflation factor (VIF) and PCA of Origin software, significant environmental factors were obtained.
Finally, Pearson correlations between significant environmental factors and different aromatic substances were then calculated and plotted using Origin (Origin 2018) software [60]. It was generally believed that there was a significant relationship when the p value was < 0.05.  Data Availability Statement: All data generated or analyzed during this study are included in this published article.

Conclusions
Acknowledgments: Thanks to the support and help from fellow apprentices in the lab in carrying out the related work in general. Thanks to them for their enthusiastic help in the experimental process and writing process.

Conflicts of Interest:
The authors declare no conflict of interest.   The longer the projection, the greater the contribution of this metabolite to this principal component. Based on the feature vectors of metabolites, it can be clearly seen that the samples from different groups were far apart from each other, while samples from the same group were gathered together. In addition, the black, red, and green dots represent samples from the CLT, CLW, and TH groups, respectively. Figure A1. PCA of significant aromatic substances of CLT, CLW, TH groups. Note: The arrows indicate the feature vectors of the aromatic metabolites. The projection of the arrows to the coordinate axis represents the contribution of the original feature to each principal component in the new space. The longer the projection, the greater the contribution of this metabolite to this principal component. Based on the feature vectors of metabolites, it can be clearly seen that the samples from different groups were far apart from each other, while samples from the same group were gathered together. In addition, the black, red, and green dots represent samples from the CLT, CLW, and TH groups, respectively.