Access to this full-text is provided by Frontiers.
Content available from Frontiers in Plant Science
This content is subject to copyright.
Comparative metabolomics
provides novel insights into
correlation between dominant
habitat factors and constituents
of Stellaria Radix (Stellaria
dichotoma L. var. lanceolata Bge.)
Zhenkai Li
1
, Hong Wang
1
, Lu Feng
1
, Le Song
1
, Yongping Lu
2
,
Hongying Li
3
, Yanqing Li
1
, Gege Tian
1
, Yan Yang
1
, Haishan Li
1
,
Xiangui Mei
2
*and Li Peng
1
*
1
School of Life Sciences, Ningxia University, Yinchuan, China,
2
State Key Laboratory of Crop
Biology, College of Agronomy, Shandong Agricultural University, Taian, Shandong, China,
3
Ningxia Institute of Meteorological Sciences, Yinchuan, China
Stellaria dichotoma L. var. lanceolata Bge. (SDL) is the original plant of the
traditional Chinese medicine Yinchaihu (Stellaria Radix). It is mainly distributed
in the arid desert areas of northwest China, which is the genuine medicinal
material and characteristic cultivated crop in Ningxia. This study aims to analyze
the effects of different origins on SDL metabolites and quality, as well as to
screen the dominant habitat factors affecting SDL in different origins. In this
study, metabolites of SDL from nine different production areas were analyzed
by ultra-high performance liquid chromatography-quadrupole time-of-flight
mass spectrometry (UHPLC-Q-TOF MS) based metabolomics. And field
investigations were conducted to record thirteen habitat-related indicators.
Results showed that 1586 metabolites were identified in different origins, which
were classified as thirteen categories including lipids, organic acids and organic
heterocyclic compounds derivatives. Multivariate statistical analysis showed
that the metabonomic spectra of SDL from different origins had various
characteristics. What’s more, co-expression network correlation analysis
revealed that three metabolites modules (MEturquoise, MEbrown and
MEblue) were more closely with the habitat factors and 104 hub metabolites
were further screened out as the habitat-induced metabolite indicators.
Besides, soil texture, soil pH value and soil total salt content were found as
the dominant habitat factors which affect SDL metabolites. In conclusion, the
Frontiers in Plant Science frontiersin.org01
OPEN ACCESS
EDITED BY
Wenyan Han,
Tea Research Institute, Chinese
Academy of Agricultural Sciences,
China
REVIEWED BY
Shi Fei Li,
Shanxi University, China
Chengying Zhao,
Institute of Food Science and
Technology, Chinese Academy of
Agricultural Sciences, China
Yinshi Sun,
Institute of Special Animal and Plant
Sciences, Chinese Academy of
Agricultural Sciences, China
*CORRESPONDENCE
Li Peng
pengli1124@nxu.edu.cn
Xiangui Mei
meixiangui@163.com
SPECIALTY SECTION
This article was submitted to
Plant Metabolism and Chemodiversity,
a section of the journal
Frontiers in Plant Science
RECEIVED 03 September 2022
ACCEPTED 07 November 2022
PUBLISHED 25 November 2022
CITATION
Li Z, Wang H, Feng L, Song L, Lu Y,
Li H, Li Y, Tian G, Yang Y, Li H, Mei X
and Peng L (2022) Comparative
metabolomics provides novel insights
into correlation between dominant
habitat factors and constituents of
Stellaria Radix (Stellaria dichotoma L.
var. lanceolata Bge.).
Front. Plant Sci. 13:1035712.
doi: 10.3389/fpls.2022.1035712
COPYRIGHT
© 2022 Li, Wang, Feng, Song, Lu, Li, Li,
Tian, Yang, Li, Mei and Peng. This is an
open-access article distributed under
the terms of the Creative Commons
Attribution License (CC BY). The use,
distribution or reproduction in other
forums is permitted, provided the
original author(s) and the copyright
owner(s) are credited and that the
original publication in this journal is
cited, in accordance with accepted
academic practice. No use,
distribution or reproduction is
permitted which does not comply with
these terms.
TYPE Original Research
PUBLISHED 25 November 2022
DOI 10.3389/fpls.2022.1035712
study showed different habitat factors had various effects on SDL’s quality and
established relationship between them, which provide reference for revealing
SDL’s genuineness formation mechanism and guiding industrial crops practical
production by habitat factors selection.
KEYWORDS
Stellaria dichotoma L. var. lanceolata Bge., Metabolomics, Origin, Habitat factors,
Genuineness, Co-expression network analysis
1 Introduction
Yinchaihu (Stellaria Radix) is a kind of Chinese herbal
medicine, which is used to clear deficient heat and infantile
malnutrition with fever (Li et al., 2020;Chinese Pharmacopoeia
Commission, 2020). In modern medicine, it has been found to
have good medical prospects such as anti-inflammatory, anti-
allergic and anti-cancer and to be rich in active ingredients such
as sterols and flavonoids (Ba et al., 2018;Zhang et al., 2019a;
Zhang et al., 2019b;Li et al., 2020;Dong et al., 2021). Stellaria
dichotoma L. var. lanceolata Bge.(SDL) is the original plant of
Yinchaihu, and its dry roots are raw materials of Yinchaihu
(Chinese Pharmacopoeia Commission, 2020). SDL is mainly
distributed in semi-arid and arid areas in China and is
concentrated in Ningxia, Inner Mongolia, Shanxi. Over the
past few decades, with the scarcity of SDL wild resources,
people began to explore cultivation and production methods
of SDL. Ningxia took the lead in domesticating SDL successfully
in the 1980s. Since then, SDL has gradually been developed into
a medicinal crop and developed into a genuine medicinal
material of Ningxia. Besides, SDL is extended to areas with
harsh environments such as central arid areas due to its excellent
drought and barrenness tolerance. At present, the largest SDL
planting base in China has been built in Tongxin County,
Ningxia. The cultivation of SDL has eased the shortage of
resources of wild medicinal herbs, brought economic, social
and ecological benefits to the cultivation sites and become an
important source of economic income for local farmers.
The genuine medicinal materials have been recognized as
high-quality Chinese herbs with excellent efficacy which were
produced in a specific region since ancient times (Yuan and
Huang, 2020).Thegenuinenessistheuniqueattributeof
genuine medicinal herbs, and the habitat is an important
manifestation of the genuineness of medicinal herbs and an
important factor for the formation of their quality. The
secondary metabolites of medicinal plants are the material
basis for the therapeutic effects of Chinese herbal medicines.
And, different habitat factors affect the quality and therapeutic
effects of medicinal herbs by regulating the formation and
accumulation of secondary metabolites in medicinal plants
(Mudge et al., 2016;Jiang et al., 2020), which exerts influence
on medicinal herbs’genuineness. For SDL, changes in the
production methods have also led to the migration and change
of its origins. In addition to the central arid area, SDL has also
started to be cultivated and produced in the non-arid areas south
of the central arid area in Ningxia. However, it has not been
scientifically verified whether the migration of origin and the
change of habitat have impacts on the the secondary metabolites
of SDL and the quality of the medicinal herbs.
Metabolomics is a technique for qualitative and quantitative
analysis of all metabolites in living organisms, and has been
widely used in research fields such as quality evaluation of
traditional Chinese medicine, formation mechanism of
genuineness, screening biomarkers and new drug development
due to its advantages with high-throughput and high sensitivity
(Nicholson and Wilson, 2003;Nicholson and Lindon, 2008).
Besides, non-targeted metabolomics analysis is an important
mode which is based on a high-resolution mass spectrometer
and is capable of an unbiased, large-scale and systematic
detection for various metabolites in samples, reflecting the
changes of metabolic levels in organisms to the greatest extent
(Christ et al., 2018).
Co-expression network analysis (CNA) is a method of
systems biology for analyzing the correlation of gene, protein
or metabolite expression in multiple samples, which can classify
a large amount of biological information into different
information modules and conduct a correlation analysis with
phenotypes. The method takes full advantage of the overall
omics information, but also converts a large amount of
biological information into a module-phenotype association,
eliminating the need for multiple hypothesis testing correction
(Langfelder and Horvath, 2008;Qing et al., 2020). CNA analysis
was first applied to genomics analysis. Later, with the effective
practice of Matthew and others in correlating tomato
metabolomics data with genetic background analysis (DiLeo
et al., 2011), this method has begun to be widely used in
metabolomics analysis and become a powerful means of
helping metabolomics data explain more scientific problems.
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org02
In this study, SDL samples were collected from nine origins with
typical habitat characteristics. And their metabolites differences were
identified by ultra-high performance liquid chromatography-
tandem time-of-flight mass spectrometry (UHPLC-Q-TOF MS)
based metabonomics. Moreover, multivariate statistical analysis
and CNA analysis were adopted to analyze the correlation between
the metabolites and environmental habitats factors, so as to explore
the impact of different origins on SDL metabolites and screen the
main habitat factors.
2 Materials and methods
2.1 SDL collection and processing
The materials were collected from nine different producing areas
(Figure 1) in August 2020 and were all identified as roots of Stellaria
dichotoma L. var. Lanceolata Bge in this study. The specificsampling
location information is shown in Figure 1 Random sampling shall be
conducted at each sampling point for six repetitions. The roots of
SDL must be naturally dried to constant weight, crushed and sieved
through40meshesbeforestoringtheminadarkandrefrigerated
place for later use. The TX sample, growing for three years in a test
site, had no field cultivation such as fertilization, water after it was
sowed, the PY, YZ and HSP samples were collected from farmland
that had been abandoned for many years and the other samples were
from natural habitats.
2.2 Investigation on habitat factors in
different origins
2.2.1 Investigation on soil physical and
chemical properties
The soil samples were collected around the SDL. Soil
samples were collected by establishing a 30 × 30 cm sample
square with SDL as the center, each sample square was sampled
with a shovel at a depth of 500 g from 0-20 cm, 20-40 cm and 40-
60 cm respectively, and the soil from the three depths was mixed
to form one soil sample. Three biological replicates of soil
samples were taken from each origin. These soil samples were
naturally dried and stored in a refrigerated area away from light
for the subsequent testing. Soil particle size was determined by
soil sieve and laser particle size analyzer, respectively. The type of
soil texture was classified by referring to the National Standard
of the People’s Republic of China “Engineering Classification
Standard for Soil”(GB/T50145-2007). The total salt content of
soil suspensions was determined by means of the conductivity
method from “Soil Testing Part 16: Determination of Total
Water Soluble Salts in Soil”(NY/T1121.16-2006). The pH
value of soil suspensions was determined by using the pH
meter method from “Soil Testing Part 2: Determination of Soil
pH”(NY/T 1121.2-2006). And, the organic matter content of the
soil was determined by referring to “Soil Testing Part 6:
Determination of Soil Organic Matter”(NY/T1121.6-2006).
2.2.2 Meteorological factors collection
Meteorological factors were collected from meteorological
stations closest to the collection sites, corresponding to station
numbers 53811, Y3417, Y2810, Y3028, Y1710, Y1351, Y2534,
53730 and 53602, respectively. The data mainly includes the
annual average precipitation, the annual average temperature,
the highest temperature in July, the average temperature in July,
the lowest temperature in January and the average temperature
in January.
2.3 Determination of chemical
composition of SDL
2.3.1 Determination of the extract content
The cold leaching method with methanol was used to
determine the extract content of SDL as the general rule “2201
Determination of extract”in “Pharmacopoeia of the People’s
Republic of China (2020 edition)”. The specific method is: the
methanol solvent is used to extract the SDL sample, the
methanol in the resulting extract is evaporated, the methanol
extract is obtained, and the extract content of the sample is
calculated by weighing method.
2.3.2 Determination of total flavonoids content
The determination of total flavonoids content was based on
the previous reported methods (GuoH.etal.,2020)and
improved accordingly. The specific process was as follows:
2.00 g of the medicinal powder sample was weighed into a
centrifuge tube, 25 ml of 95% ethanol was added and the
supernatant was extracted by ultrasonication for 30 min and
then, the supernatant was separated. The residue was then added
with 25 ml of 95% ethanol to continue ultrasonic extraction for
15 min, and the supernatants of the two extractions were mixed
as the test sample solution. The total flavonoids content of the
herb was determined by measuring the absorbance value of the
sample at 496 nm by using rutin as the control.
2.3.3 Determination of total sterols content
The determination of total sterols content was based on the
previous reported methods (Zhang et al., 2012) and improved
accordingly. The specific process was as follows: 0.50 g medicinal
powder sample was weighted into a 25 ml volumetric flask and
then to add 20 ml of chloroform, which was extracted by
ultrasonication for 20 min. After 20 min of the ultrasonic
extraction, supernatant was acquired. The second step was to
dilute the supernatant with chloroform to the scale before
shaking it well and filtering it, thus obtaining the test sample
solution. The final step was to determine the total sterols content
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org03
of the herb by measuring the absorbance value at the wavelength
of 546 nm by using a- spinasterol as the control.
2.3.4 Metabolite detection
The metabolomic analysis of the herbs was carried out by
Shanghai Applied Protein Technology Co., Ltd. The method was
as follows: the first step was to grind herbal powder in liquid
nitrogen, after which 200 mg of the powder was weighted into a
2 ml centrifuge tube. Following this, 70% methanol aqueous
extraction solution was added to the centrifuge tube and then
vortexed it thoroughly. The second step was to extract the
solution, before drying extraction solution under vacuum to
get extract. Then, the extract must be stored at -80°C. The next
step was to dissolve the extract with 40% acetonitrile water
solution to get supernatant and then to analyze the metabolic
compositions in the supernatant. The separation was performed
with an Agilent 1290 Infinity LC HILIC column; column
temperature 25°C; flow rate 0.5 ml/min; injection volume 2 ml;
mobile phase composition A: water + 25 mM ammonium
acetate + 25 mM ammonia, B: acetonitrile; gradient elution
procedure as follows: 0~0.5 min, 95% B The gradient elution
procedure was as follows: 0 ~ 0.5 min, 95% B; 0.5~7 min, B
linearly varied from 95% to 65%; 7~8 min, B linearly varied from
65% to 40%; 8~9 min, B maintained at 40%; 9~9.1 min, B
linearly varied from 40% to 95%; 9.1~12 min, B maintained at
95%. During the whole process of analysis, the sample was
placedinanautosamplerat4°C.And,massspectrometric
analysis was performed with triple TOF 6600 mass
spectrometer, and the positive (pos) and negative (neg) ion
modes of electrospray spray ionization (ESI) were used for
detection. The metabolites were identified by matching the
retention time, molecular weight (error <25 ppm), secondary
fragmentation spectrum, collision spectra and other information
B
C
D
A
FIGURE 1
Specific information of different SDL sampling sites. (A–Care the information and pictures of the sampling sites. (D) is the soil of different
sampling sites, of which (1) is gravel soil, from DWK; (2) sandy soil, from LW, YC, EKTQQ and ALSZQ; (3) loam soil, from HSP; (4) clayey soil,
from TX, PY and YZ.).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org04
of metabolites by means of searching a local self-built standards
database established by Shanghai Applied Protein Technology.
2.4 Data analysis
Multivariate statistical analyses such as hierarchical cluster
analysis (HCA), principal component analysis (PCA), and K-
means cluster analysis were performed by R software (www.r-
project.org/). The significant changed metabolites(SCM) were
screened by Fold Change Analysis(FC analysis), Ttest and
Orthogonal Partial Least Squares Discriminant Analysis
(OPLS-DA) The specific screening criteria were VIP >1,FC >
1.5 or FC < 0.67 and p< 0.05. The KEGG (Kyoto Encyclopedia of
Genes and Genomes, https://www.kegg.jp/) database was used
for annotation and functional enrichment analysis of SCMs.
Based on the CNA method, metabolite co-expression networks
and modules were constructed. Based on the criteria of the
correlation coefficient R value being closer to ± 1 and the
correlation test pbeing less than 0.05, correlation analysis was
performed for co-expression modules and thirteen major habitat
factors. The expression of metabolites in the screened key
modules in all samples was subjected to cluster heat map
analysis to compare the distribution of eigenvalues of each
module in all samples. The hub metabolites were further
screened by analyzing the correlation(r) between metabolites
in key modules and modules, and the correlation between
metabolites in key modules and traits.
3 Results and analysis
3.1 Habitat characteristics of SDL in
different origins
3.1.1 Spatial distribution analysis
As shown in Figure 1 and Table 1, the nine SDL origins were
distributed at 35.00
°
N-40.00
°
N, 105.00
°
E-109.00
°
E, altitude 1
050.00-1 650.00 m. The naturally distributed samples grew at
38.00
°
N above and below 1 300 m above sea level; samples from
farmland or abandoned farmland grew at 38.00
°
N below and
above 1 300 m above sea level.
3.1.2 Meteorological factor analysis
The average values of the six meteorological factors collected
from 2015 to 2020 are shown in Table 1. The annual
precipitation for the nine producing areas ranged from 183 cm
to 583 cm. The average annual temperature ranged from 8.5°C to
11.1°C, the maximum temperature in July ranged from 27.7°C to
36.5°C and the average temperature in July ranged from 21.4°C
to 25.8°C. The minimum temperature in January ranged from
-23.2°C to -10.4°C and the average temperature in January
ranged from -8.5°C to -4.8°C. Besides, the significance analysis
showed that the six meteorological factors showed different
levels of variation due to different producing areas.
3.1.3 Analysis of the physical and chemical
properties of soils
As shown in Figure 1D, the soil from DWK (label refers to origin,
the same below) habitat showed a dark grey color and the rest were
yellowish brown. The soils from the DWK, ETKQQ, YC, LW,
ALSZQ and HSP habitats were relatively loose, while soils from
TX, PY and YZ were highly viscous and have soil agglomeration. The
physical and chemical properties of soils were further analyzed and
theresultsareshowninTable 1. In terms of soil texture, total salt
content, pH value and organic matter content, the habitat soils of SD L
from different origins were different in varying degrees. The results of
the soil texture analysis showed that DWK had the largest proportion
of large grained soils, which was classified as gravelly soils. Then, soils
from ETKQQ, YC, LW and ALSZQ were classified as sandy soils. TX,
PY and YZ had a higher proportion of finer grained soils, which was
classified as clayey soil. The soil texture of HSP is between sandy and
clayey soil, and is classified as loam soil. The total salt content of DWK
was 29.77 g/kg which was significantly higher than other producing
areas, while the pH value of DWK was 7.29 which was significantly
lower than that of others. And, for other producing areas, the total salt
content was less than 8.00 g/kg and pH values ranged from 8.31 to
9.17. In addition, the organic matter content of DWK, PY and TX was
significantly higher than that of other origins, which were 8.29 g/kg,
8.03 g/kg and 7.00 g/kg respectively, followed by YZ with 3.16 g/kg,
and the rest of the samples were around 1.00 g/kg.
3.2 Analysis of the characteristics of
medicinal materials, methanol extract,
total sterols and total flavonoids content
of SDL from different origins
3.2.1 Characteristics of medicinal materials
The “shayan”(hole-like or disk-like depression),
“zhenzhupan”(wart-like protruding buds, stems or rhizome
stumps) and yellow and white cross-section of the radial
texture are the most important characters of SDL. As shown
in Figures 2A,B, the SDL from the nine origins in this study all
had “shayan”,“zhenzhupan”and yellow and white cross-section
of the radial texture. The difference is that SDL collected from
wild natural habitats was darker in colour (DWK, LW, YC,
ETKQQ and ALSZQ) and was light brown or brown. SDL
collected from test or abandoned agricultural fields were
lighter in colour (TX, HSP, PY and YZ) and were pale yellow
or light brownish-yellow. In addition, SDL collected from clayey
soil (TX, PY and YZ) were mostly long-columnar in shape and
had no or few branches. SDL collected from loam, sandy or
gravelly soils (HSP, LW, YC, ETKQQ, ALSZQ and DWK) were
more variable in character and generally had multiple branches.
In summary, SDL from different origins has the basic
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org05
TABLE 1 Habitat analysis of SDL from different origins.
Label Latitude,
N
Longitude,
E
Elevation,
m
average
annual
precipitation,
mm
average
annual
temperature,
°C
highest
temperature
in July, °C
average
temperature
in July, °C
lowest
temperature
in January, °C
average
temperature
in January, °C
Soil
texture
Total salt
content,
g/kg
pH
value
Organic
matter
content,
g/kg
TX 36.76
°
106.36
°
1558 311.15 ± 65.95b 8.78 ± 1.08c 29.29 ± 1.35cd 22.31 ± 0.73de -15.33 ± 1.16b -8.52 ± 1.55c clayey
soil
3.31 ± 0.14
bc
8.71 ±
0.11 cd
7.00 ± 0.85 b
HSP 37.40
°
105.97
°
1335 148.72 ± 64.87c 10.81 ± 2.72ab 32.62 ± 1.09b 24.84 ± 1.41ab -11.29 ± 0.99a -7.09 ± 4.00abc loam soil 5.72 ± 0.37
bc
8.49 ±
0.15 de
1.01 ± 0.08 d
PY 35.75
°
106.80
°
1395 583.17 ± 145.53a 8.90 ± 0.47bc 29.10 ± 2.22cd 21.43 ± 1.25e -10.63 ± 1.63a -4.82 ± 1.58a clayey
soil
4.61 ± 0.32
bc
8.51 ±
0.05 de
8.03 ± 0.94 ab
YZ 36.07
°
106.30
°
1624 475.47 ± 182.34a 9.11 ± 1.23bc 27.67 ± 2.00d 21.37 ± 0.67e -10.42 ± 1.65a -5.03± 1.76ab clayey
soil
6.13 ± 0.51
bc
8.62 ±
0.09
cde
3.16 ± 0.37 c
DWK 39.18
°
106.38
°
1300 166.73 ± 56.52bc 11.07 ± 1.44a 32.48 ± 0.65b 25.80 ± 1.01a -11.96 ± 0.79a -6.34 ± 1.28abc gravelly
soil
29.77 ± 4.22
a
7.29 ±
0.47 e
8.29 ± 1.34 a
LW 38.05
°
106.59
°
1273 304.40 ± 212.85bc 9.22 ± 1.12abc 30.80 ± 0.57bc 23.51 ± 0.56cd -11.76 ± 0.75a -6.12 ± 1.10abc sandy
soil
2.43 ± 0.35
c
9.17 ±
0.03 a
0.72 ± 0.05 d
YC 37.88
°
107.56
°
1298 190.25 ± 73.14bc 8.46 ± 1.61c 31.50 ± 1.42b 24.01 ± 0.40bc -14.59 ± 1.27b -7.83 ± 1.72bc sandy
soil
2.90 ± 0.21
c
8.92 ±
0.07 bc
0.65 ± 0.01 d
ETKQQ 38.46
°
108.18
°
1290 266.10 ± 43.13bc 8.97 ± 0.18bc 36.48 ± 1.62a 24.07 ± 0.58bc -23.18 ± 1.28d -8.05 ± 1.62c sandy
soil
7.82 ± 0.17
b
8.67 ±
0.11 cd
1.08 ± 0.06 d
ALSZQ 39.39
°
106.73
°
1069 248.02 ± 47.00bc 9.50 ± 0.21abc 35.57 ± 1.56a 24.08 ± 0.50bc -17.73 ± 3.37c -7.00 ± 2.01abc sandy
soil
2.51 ± 0.08
c
8.31 ±
0.06 d
1.07 ± 0.08 d
The lowercase letters after the data indicate significant differences in habitat factors between different SDL origins at p<0.05 levels.
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org06
characteristics of medicinal herbs, but different habitats may
affect the overall color and morphology of SDL.
3.2.2 Methanol extract, total sterols and total
flavonoids content
The methanol extract is the content determination index for
evaluating SDL as stipulated in the Chinese Pharmacopoeia, and
total sterols and total flavonoids content are the most commonly
used indexes for evaluating the quality of SDL at present (Li
et al., 2020). In this study, the three indexes of SDL from
different origins were tested, and the determination results are
shown in Figure 2C. The content of SDL methanol extract, from
different origins reached the 20% index stipulated in the Chinese
pharmacopoeia. And, the lowest content of DWK was 24.77%
and the highest content of ETKQQ was 39.27%, while others did
not show significant differences (P < 0.05). The total sterol
content was the highest in ALSZQ and DWK with 5.85 g/kg
and 5.71 g/kg respectively, while the total sterols content of HSP,
ETKQQ, YZ and TX were significantly lower than the other
samples. The total flavonoids content of ETKQQ was 4.29 g/kg
which was significantly higher than the samples from other
origins. The sample with the lowest total flavonoid content was
PY, which was 1.93 g/kg and significantly lower than samples
from other origins. In summary, in addition to the differences in
the methanol extracts of SDL from different habitats, there were
greater differences in the contents of total flavonoids and total
sterols. It is speculated that there may be more different
substances in SDL from different origins.
3.3 SDL metabolites and quality control
The total ion chromatograms (TIC) of all QC samples were
compared by overlapping spectra, as shown in Figure 3A and
Supplementary Figure 1A, and the response intensities and
retention times of the peaks basically overlapped, indicating
good instrument precision throughout the experiment. The
proportion of QC samples with relative standard deviation
(RSD) less than 30% of the characteristic peaks exceeded 70%
(Figure 3B and Supplementary Figure 1B), indicating a good
stability of the instrument. PCA analysis was performed on all
samples and QC samples. As shown in Figure 3C and
Supplementary Figure 1C, the QC samples were closely
clustered together, indicating that the samples had good
repeatability. In addition, the results of PCA analysis also
showed that DWK had significantly different principal
component characteristics. ETKQQ, LW and YC had similar
principal component characteristics. And TX, YZ, PY, ALSZQ
and HSP had closer principal component characteristics.
The metabolites were identified by searching the local self-
built standards database. A total of 1586 substances
(Supplementary Table 1) were identified from nine origins of
SDL and divided into thirteen categories. Among them, 880
substances were identified in pos mode and 706 substances in
neg mode. And the categories mainly include lipids and lipid-
like molecules (331 species), organic acids and derivatives (327
species), organicheterocyclic compounds (201 species),
phenylpropane and polyketone compounds (170 species),
B
C
A
FIGURE 2
The characteristics of medicinal materials, methanol extract, total sterols and total flavonoids content of SDL in different origins. (Ashowed the
overall morphology of SDL from different origins. Bshowed the main medicinal characteristics of SDL, including ‘shayan’,‘zhenzhupan’and
yellow and white cross-section of the radial texture from left to right. Cis the content of methanol extract, total sterols and total flavonoids of
SDL from different origins,and the lowercase letters in the bar chart indicate significant differences in the content of components of SDL
between different origins at the p< 0.05 level. The same letter indicates no significant difference, while different letters indicate significant
difference.).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org07
organic oxygen compounds (169 species), benzenes (168
species), etc. (Figure 3D).
3.4 Analysis of SCMs in SDL of
different origins
A systematic clustering heat map analysis of the ionic
intensities of each metabolite was performed for SDL samples
of all origins, as shown in Figure 4. Result showed that DWK
clustered separately with samples from other origins, LW, YC
and ETKQQ clustered in a group, and TX and HSP clustered in a
group, which was generally consistent with the characteristics
showed by the PCA analysis (Figure 2C andSupplementary
Figure 1C). The total ion heat map of metabolites showed
significant expression differences for SDL metabolites of
different origins. According to the results of principal
component analysis and clustering, and habitat characteristics,
TX was selected for pairwise comparison with LW, PY, HSP,
ALSZQ and DWK samples. The SCMs between TX and other
samples were screened by FC analysis, Ttest and OPLS-DA,
(Figure 5,Supplementary Figures 2–4). The results showed that
there were significant differences in SCMs between different
comparison groups. The number of SCMs in different
comparison groups from high to low is: TX vs DWK (370
species), TX vs ALSZQ (336 species), TX vs PY (266species),
TX vs LW (234 species) and TX vs HSP(211 species)(Figure 6
and Supplementary Figure 5). These SCMs were reflected in a
variety of classifications such as lipid and lipid-like molecules,
organic acids and derivatives, phenylpropanoids and
polyketides, etc. Further Venn diagram analysis of all SCMs
showed that a total of 43 common SCMs were identified in the
five comparison groups (Figure 7). In addition, TX vs DWK has
65 unique SCMs; followed by TX vs PY and TX vs ALSZQ, 54
and 43 species, respectively, TX vs LW was the least, only seven
species. The KEGG database was further used to annotate and
enrich the scm of the five comparison groups. As shown in
Figure 8, SCMs in different comparison groups are enriched in
different signaling pathways, and the significance of the same
signaling pathway in different comparison groups is also
different. This further indicates that SDLs from different
origins have more differences in metabolites and
metabolic pathways.
B
C
D
A
FIGURE 3
Quality control and identification analysis of metabolite detection. (Ais TIC diagram in pos ion mode; Bis characteristic peak variation
coefficient in pos ion mode; Cis the PCA of metabolites detected in pos ion mode; Dis the category and pie chart of identified metabolites.).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org08
3.5 Correlation of SDL metabolites with
dominant environmental factors
The co-expression network of 1586 metabolites were
constructed by WGCNA, and the correlation analysis was
carried out with a total of thirteen habitat factors, including soil
physicochemical properties, meteorological factors, and spatial
distribution. As shown in Figure 9A, 1586 metabolites were
divided into six co-expression modules (MEyellow, MEred,
MEgreen, MEturquoise, MEblue and MEbrown) and one
module without obvious co-expression relationship (MEgrey).
And the heat map of metabolite expression and the distribution
map of module eigenvalues (Figure 9B) showed that MEyellow
from LW and YC had higher up-regulated expression in all
samples, MEred from HSP had higher up-regulated expression,
MEgreen from ALSZQ had higher upregulated expression,
MEturquoise from DWK had higher up-regulated expression,
MEblue from YZ had higher up-regulated expression, PY and TX
MEbrown had higher up-regulated expression. These specifically
expressed metabolite modules could be used as characteristic
groups of constituents of SDL in different origins. Further
correlation analysis of the seven metabolite modules with the
thirteen habitat factors showed (Figure 10)thatMEturquoisehad
asignificant positive correlation with total salt content of soil
(r=0.933, p<0.05) and average annual temperature(r=0.668,
p<0.05), while a negative correlation with soil pH value(r=-
0.886, p<0.05). MEbrown had a significant positive correlation
with average annual precipitation(r=0.749, p<0.05), and a
significant negative correlation with latitude (r=-0.690, p<0.05)
and soil texture(r=-0.668, p<0.05). MEblue had a significant
positive correlation with elevation(r=0.672, p<0.05). As key
modules, these highly correlated modules (MEturquoise
MEbrown and MEblue) were metabolite groups that were
largely influenced by habitat factors. Significantly, the total salt
content, pH value and soil texture, average annual temperature
and precipitation, elevation and latitude were the main habitat
factors resulting in SDL metabolite differences.
To further screen out the key metabolites with high
correlation from the key modules, we analyzed the scatter
distribution of metabolite significance (correlation between
metabolites and habitat factors) and module membership
(correlation between metabolites and modules) in each module
(Figure 11). And the key metabolite was filtered out based on the
metrics of metabolite significance and module membership
(Top50 and p<0.05). As shown in Supplementary Table 2, 104
species hub metabolites were screened, including 22 species
lipids and lipid-like molecules, 13 species benzenoids, 17
species organic oxygen compounds, 15 species organic acids
and derivatives, 14 species organoheterocyclic compounds and 9
species phenylpropanoids and polyketides, etc. Morever, the
largest number of hub metabolites significantly associated with
soil texture was 37 species, followed by soil total salt content (34
B
A
FIGURE 4
Heatmap of hierarchical clustering of SDL metabolites in habitats of different origin. (Ais the metabolite detected in pos mode and Bis the
metabolite detected in neg mode).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org09
species), soil pH value (34 species), average annual precipitation
(25 species), elevation(14 species), average annual temperature
(11 species) and latitude (11 species). The hub metabolites are
probably to be the main metabolites of SDL in response to
different habitat factors. On the contrary, the corresponding
habitat factors may be the dominant factors which affect
SDL’s quality.
4 Discussion
In this study, habitat surveys of nine origins of SDL revealed
that there were significant differences among them especially in
the soil physicochemical properties. In terms of medicinal
properties, further analysis revealed that the MeOH extract,
total sterols and total flavonoids content in SDLs from
different origins were also obvious different. Meanwhile,
metabolomics analysis showed that SDL from different origins
contained different metabolite characteristics and showed rich
diversity. Among which the metabolic characteristics of LW, YC
and ETKQQ were more similar, and the difference of DWK was
the largest. In addition, the results of SCMs screening showed
that the metabolites of TX were similar to those of LW and HSP.
At the same time, KEGG enrichment analysis also enriched the
SCMs of different comparison groups into different signaling
pathway. These results further indicate that different origins
have significant effects on the production, accumulation and
signaling pathways of SDL metabolites. This is consistent with
B
CD
EF
A
FIGURE 5
Volcano plot, OPLS-DA and permutation test analysis of differential metabolites of TXvsLW. (A, C, E are volcano plot, OPLS-DA and permutation
test analysis for detecting metabolites in pos mode, respectively; B, D, F are volcano plot, OPLS-DA and permutation test analysis for detecting
metabolites in neg mode, respectively.).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org10
the conclusion of metabolomics analysis of medicinal plants
such as Chrysanthemum (Zou et al., 2022) and American ginseng
(Si et al., 2021), from different origins. Furthermore, bioactive
substances are the material basis for medicinal plants (Li et al.,
2018) to exert their efficacy and are also important indicators for
evaluating the quality of medicinal materials. (Li et al., 2018).
However, SDL has not yet established systematic quality
evaluation system and its bioactive material basis research is
lagging behind. Herein, results showed that all of the methanol
extract content of SDL from different habitats reached the
requirements of Chinese Pharmacopoeia (20%), there were still
more significant differences in total sterols, total flavonoids and
other metabolites. These differential metabolites contain a
variety of bioactive ingredients, such as organic acids, phenol
propane, polyketide compounds, lipid and their derivatives,
which may lead to different efficacy effects and herb quality.
Therefore, the in-depth research on SDL quality biomarkers and
habitat selection should be paid more attention.
The bioactive substances are the products of long-term
adaptation to specific environments for medicinal plants, which
are mostly the secondary metabolites accumulated in response to
habitat abiotic and biotic stresses (Huang and Guo, 2007;Li et al.,
2019). Consequently, the same plant will also metabolize and
accumulate various secondary metabolites in different habitats.
The SDLs were collected from different ecological regions in this
study with complex and diverse habitat differences in the soil
environment, climate and spatial distribution. Therefore, the
characteristics of differential metabolites in SDL from different
habitats may be affected by the interaction of multiple habitat
factors. In this study, SDL metabolites were correlated with
thirteen habitat indicators by the WGCNA analysis. And, the
results showed that several metabolite modules (MEturquoise,
MEbrown and MEblue)were closely correlated with several
habitat indicators such as total soil salt content, pH value,
organic matter content, soil texture, annual precipitation,
average annual temperature and altitude.
B
A
FIGURE 6
FC analysis histogram of SCMs in TXvsLW. (Ais the metabolite detected in pos mode and Bis the metabolite detected in neg mode.).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org11
Further screening of the hub metabolites from the
metabolite module revealed that the hub metabolites were
most associated with soil texture, soil pH and soil salinity. Soil
is the material basis for the growth and development of
medicinal plants. The physical and chemical properties of soil
are important components of soil environment. SDL is a
psammophyte. Most of the wild SDLs collected in this study
grow in sandy soil, but non-wild SDLs collected from TX, PY,
YZ and HSP grow in clayey soil or loam soil. The change of soil
texture will not only change porosity and soil water retention
capacity, but also affect the soil available nutrients and soil
microorganism, and further affect the plant transpiration,
photosynthesis, respiration and other physiological and
biochemical effects and the accumulation of secondary
metabolites (Du et al., 2019;Haruna and Yahaya, 2021).
Therefore, different medicinal plants have preferences for soil
texture according to their physiological needs. For example,
Scrophularia ningpoensis is suitable for growing in limestone
heavy loam, and loess with deep soil layer is suitable for the
growth of Astragalus membranaceus, and Codonopsis pilosula
and Rehmannia glutinosa are suitable for growing on fertile
sandy soil (Chen and Tan, 2006;Tian et al., 2013). Therefore,
whether SDL is suitable to grow in non-sandy soil needs further
study. Salt stress is an important abiotic stress affecting
secondary metabolism of medicinal plants (Hemanta, 2017).
Zhang found that the yield of SDL herbs and the accumulation
of total flavonoids and total saponins reached the maximum
when the soil salt content was 0.3% (Zhang et al., 2017). Soil pH
is also an important environmental factor affecting the
metabolites of medicinal plants (Du et al., 2013), and different
medicinal plants and metabolites have different responses to soil
pH. For example, in the range of soil pH 4.5 ~ 9.5, the content of
active ingredients in P.multiflorum tubers decreased with the
increase of pH, and the contents of stilbene glycoside and bound
anthraquinone in P.multiflorum tubers reached the highest at
pH 4.5 (Leng et al., 2020). Zou found that the content of total
flavonoids, oleanolic acid and ursolic acid in G. longituba was
highest at pH 6.5, while the content of rosmarinic acid was
highest at pH 7.5 (Zou et al., 2019). In this study, a variety of
flavonoids, glycosides, alkaloids and organic acids were
significantly correlated with soil total salt content and pH such
as silybin (M453T51_1), catalposide (M465T412), poncirin
(M617T328_2), plantamajoside (M639T418), isorhamnetin 3-
galactoside (M501T353), harmane (M181T170), phenylalanine
betaine (M208T197), chicoric acid (M497T336) etc. These
secondary metabolites may be important substances for SDL
to respond to changes in soil pH and salt content. In conclusion,
soil physicochemical properties may be the most important
habitat factor affecting SDL metabolites.
Annual precipitation and temperature are also important
habitat factors that are highly correlated with various pivot
metabolites in this study. Precipitation can reflect the water
environment of plant growth. Water is an important medium
linking the atmosphere, the soil to the plant and the metabolic
activities of the plant, which affects plant physiological and
biochemical processes such as photosynthesis, respiration,
oxidation and secondary metabolic activities in plants
(Bohnert et al., 1995;Jat and Gajbhiye, 2017). Lang found that
FIGURE 7
Venn diagram analysis of SCMs in Different Comparison Groups.
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org12
the root growth of SDL was inhibited to a certain extent during
moderate to severe drought stress (40% field water holding
capacity), which affected the yield; however, drought stress
significantly increased the content of total flavonoids, total
saponins and other secondary metabolites in SDL herbs (Lang
et al., 2014). This study found that flavonoids deoxyrhapontin
(M449T356) and saponins astragaloside ii (M871T33_3) showed
a significant correlation with annual precipitation, which further
verifying the previous results. Temperature can directly affect the
growth and development, physiological activity, harvest time
and the accumulation of compositions of medicinal plants. In
particular, during critical growth periods, it can alter the activity
of relevant enzymes in the medicinal plant, which directly affects
the level of secondary metabolism of plants (Eguchi et al., 2019).
Furthermore, elevation and latitude can indirectly affect the
growth and development of medicinal plants by influencing
habitat factors such as temperature and precipitation (Joshi et
al., 2021). In summary, the significant differences of SDL
metabolites from different origins may be due to the combined
and complex effects of habitat factors such as soil physical and
chemical properties, annual precipitation, annual mean
temperature, altitude and latitude.However,whatkindof
habitat is conducive to the growth of SDL and the formation
of medicinal quality, and how to promote the high-quality
production of SDL through scientific cultivation measures,
need deeper thinking and research.
“Simulative Habitat Cultivation”is the core model of
ecological cultivation of Chinese herbs proposed by
academician Luqi Huang and researcher Lanping Guo. Based
on the long-term adaptation of medicinal plants to specific
environmental stresses, it simulated various environmental
factors of wild medicinal plants, especially the original habitat
of genuine herbs, and then balance the growth and development
of Chinese herbs and secondary metabolism by utilizing
scientific design and clever human intervention, thus achieving
the optimal layout and high-quality development of genuine
Chinese herbs. Especially in the absence of more research bases
and special production purposes, “Simulative Habitat
B
CD
E
A
FIGURE 8
KEGG enrichment analysis of SCMs from different SDL comparison groups. (A–Eare the ones of TX vs LW, TX vs PY, TX vs HSP, TX vs DWK, and
TX vs ALSZQ, respectively.).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org13
Cultivation”of genuine medicinal materials can be used as a
basic model for high-yield and high-quality production of
Chinese herbs (Guo L. et al., 2020). SDL has a long history
and widely application. However, in modern research, basic
research on SDL is lagging behind or in the blank, such as the
research on efficacy mechanisms, material basis, quality markers
and habitat stress response mechanisms, etc. But, in recent
modern research, the basic research on its mechanism of
efficacy, therapeutic material basis, quality markers and the
response mechanism under environmental stress of SDL is at a
preliminary stage. This has led to a lack of effective evaluation
indicators in the production and quality evaluation of SDL,
which in turn has hindered the industrial development and
resource utilisation of SDL. Therefore, in the absence of a
research base, the “simulated biotope”model combined with
metabolomics technology offers a new idea for the more
scientific production of SDL.
During the period from 2017 to 2022, the author’s research
group conducted several surveys on SDL resources in and
around Ningxia (Li et al., 2022). The survey found that wild
B
A
FIGURE 9
Heat map of metabolite co-expression module division and their correlation with phenotypic traits. (Ais metabolite co-expression module; Bis
heat map of metabolite expression and the distribution map of module eigenvalues.).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org14
SDL was concentrated in Lingwu City and Yanchi County in
Ningxia and Etuokeqianqi County in Inner Mongolia, which all
belong to the wind-sand arid area and have similar habitat
characteristics such as soil physicochemical properties, climate
and spatial distribution. Metabolites of SDL collected in these
regions also have relatively similar characteristics. And the
content of various active substances in SDL collected from
Lingwu City (LW) was significantly higher than that in
SDLcollected from cultivated origin (TX), such as beta.-
sitosterol (M397T42), trigonelline (M138T291_2), betaine
(M118T277_2), fustin (M269T36), rotenone (M241T189),
arctiin (M557T165) and loganic acid (M399T284_2).
Therefore, the wind-sand arid area can be used as the
preferred ecological area for SDL “Simulative Habitat
Cultivation”production. In addition, the SDL of LW, YC and
ETKQQ are all distributed in the desert grassland in the region.
So the environmental characteristics of the desert grassland in
the arid area may be the main habitat characteristics in the
formation of SDL genuineness, such as less rain, alkaline sand
soil, etc.
The collection site of DWK is mainly located near the
Shitanjing coal mine, which is another area of concentrated
distribution of wild SDL. The results revealed that the habitat
characteristics of this collection site, especially the
physicochemical properties of the soil, were significantly
different from those of the other wild distribution areas, and
the characteristics of metabolites were also significantly different
from those of the others. At the meanwhile, the survey also
showed that SDL is the dominant species in these regions, which
is adapted to grow in the specific environment. A large number
of studies have confirmed that plants growing in a specific
environment for a long time are subject to the combined
influence of various ecological factors in the environment,
which may cause changes in genetic material such as
mutations in plant DNA and aberrations in chromosome
structure and number, and then change metabolic regulatory
enzymes, resulting in variation in the products of secondary
metabolism (Cao et al., 2021;Zhang et al., 2021). In the present
study, it is worthwhile to pay attention to and study in more
depth whether the distinct habitats of DWK have affected the
FIGURE 10
Correlation analysis between module and habitat factors. (The numbers in the colour block in the figure, the top one is the r-value and the
bottom one (in brackets) is the P-value).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org15
genetic material of SDL, thus causing metabolites of SDL to be
significantly different from other samples, or whether new
varieties have been produced.
As the largest concentrated cultivation area of SDL, Tongxin
County has a long history of cultivation and a good production
and processing base, and was awarded the “Tongxin Yinchai hu”
geographical indication certification for agricultural products in
2018, which is recognized as the genuine origin of cultivated
SDL. This research found some differences in characteristics of
metabolites between SDL in Tongxin County and samples
B
CD
EF
G
A
FIGURE 11
Scatter distribution of module membership vs metabolites significance. (A–Gare MEbrown-Latitude, MEbrown-Average annual precipitation,
MEbrown-Soil texture, MEturquoise-Average annual temperature, MEturquoise-Soil tital salt content, MEturquiose-Soil PH value and MEblue-
Elevation, respectively.).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org16
collected in the wild. However, key habitat indicators such as soil
total salt content, pH value, annual precipitation and average
annual temperature of the local natural habitat were very close to
those of the wild SDL habitat, with only some differences in soil
texture, which may be the main reason for the differences
between TX and wild SDL. The soil of the TX habitat is clayey
soil, rather than the sandy soil of the wild habitat. Therefore, it is
recommended that during the production and planting of SDL
in Tongxin County, measures such as deep ploughing and
applying soil cavitation amendments to increase soil porosity,
so as to achieve the purpose of “Simulative Habitat Cultivation”
and guarantee better quality of SDL.
5 Conclusion
In this study, a total of 1586 metabolites were identified in
SDLs from nine habitat by the UHPLC-Q-TOF MS based
metabolomics. Differential metabolites among nine origins
were analyzed through multivariate statistics and the
correlations between metabolites and the habitat factors were
also investigated and discussed. The results showed that SDLs
from different habitats had various metabolites, and the samples
with similar habitat factors also showed similar metabolite
characteristics. These differential metabolites are mainly some
lipids and lipid molecules, organic acids and their derivatives,
phenylpropane and polyketone compounds, etc. Further more,
1586 metabolites were clustered into seven co-expression
modules by the CNA. And the correlation analysis of seven
modules with thirteen habitat factors showed that three
metabolite modules(MEturquoise, MEbrown and MEblue)
showed significant correlations with different habitat factors
and 104 species hub metabolites were further screened out.
Soil texture, soil pH value and soil total salt content were
selected as the most dominant habitat factors affecting SDL
metabolites, and then followed by annual precipitation and
temperature, elevation and latitude. The research provides
theoretical and practical significance for guiding the
construction of genuine producing areas, the scientific
production and “Simulative Habitat Cultivation”for SDL.
Data availability statement
The original contributions presented in the study are
included in the article/Supplementary Material. Further
inquiries can be directed to the corresponding authors.
Author contributions
The manuscript was written through contributions of all authors.
ZKL, resource survey, Sample collection, Methodology, Sample
detection, Writing-review and editing, Data analysis, Visualization;
HW and LS, Resource survey, sample collection, sample detection; LF
and HSL, writing-review and editing; HYL, Meteorological data
collection; YPL, Metabonomic analysis; YQL, sample testing, data
analysis;YYandGGT,visualization;XGM,conceived,revisedand
supervised the manuscript; LP, conceived and designed the study,
methodology, supervision, funding acquisition. All authors
contributed to the article and approved the submitted version.
Funding
This study was supported by the Key Research and
Development Program of Ningxia (No.2021BEG02042) and
Ningxia Natural Science Foundation (No.2021AAC03103).
Acknowledgments
The authors acknowledge Shanghai Applied Protein
Technology Co., Ltd. for their support for metabolite testing,
and Ma Li (School of Foreign Chinese, Ningxia University) for
her help in language translation.
Conflict of interest
The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could
be construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the
authors and do not necessarily represent those of their affiliated
organizations, or those of the publisher, the editors and the
reviewers. Any product that may be evaluated in this article, or
claim that may be made by its manufacturer, is not guaranteed
or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found
online at: https://www.frontiersin.org/articles/10.3389/
fpls.2022.1035712/full#supplementary-material
Additional data relevant to this paper can be found in
the Annexes.
SUPPLEMENTARY FIGURE 1
Quality control analysis of metabolite detection. (Ais TIC diagram in neg
ion mode; Bis characteristic peak variation coefficient in neg ion mode; C
is the PCA of metabolites detected in neg ion mode.).
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org17
SUPPLEMENTARY FIGURE 2
Volcano plot analysis of different comparison groups (A, C, E, G were TX vs
PY, TX vs HSP, TX vs DWK, and TX vs ALSZQ in pos mode, respectively. B,
D, F, H were TX vs PY, TX vs HSP, TX vs DWK, and TX vs ALSZQ in pos
mode, respectively.).
SUPPLEMENTARY FIGURE 3
OPLS-DA and permutation test (pos). (A, C, E, G are OPLS-DA of TX vs PY,
TX vs HSP, TX vs DWK, and TX vs ALSZQ, respectively; B, D, F, H are
permutationtestofTXvsPY,TXvsHSP,TXvsDWK,andTXvs
ALSZQ, respectively.).
SUPPLEMENTARY FIGURE 4
OPLS-DA and permutation test (neg). (A, C, E, G are OPLS-DA of TX vs PY,
TX vs HSP, TX vs DWK, and TX vs ALSZQ, respectively; B, D, F, H are
permutationtestofTXvsPY,TXvsHSP,TXvsDWK,andTXvs
ALSZQ, respectively.).
SUPPLEMENTARY FIGURE 5
FC analysis histogram of SCMs in different comparison groups (A, C, E, G
are FC analysis of TX vs PY, TX vs HSP, TX vs DWK, and TX vs ALSZQ in pos
mode, respectively; B, D, F, H are FC analysis of TX vs PY, TX vs HSP, TX vs
DWK, and TX vs ALSZQ in neg mode, respectively.).
References
Ba,S.,Choi,J.,Park,B.,Lee,J.,Jo,E.,Lee,Y.,etal.(2018).Protectiveeffectsofa
traditionalherbalextractfromStellariadichoto mavar. lanceolata against mycobacterium
abscessus infections. PloS One 13 (11), e0207696. doi: 10.1371/journal.pone.0207696
Bohnert, H., Nelson, D., and Jensen, R. (1995). Adaptations to environmental
stresses. Plant Cell. 7 (7), 1099. doi: 10.1105/TPC.7.7.1099
Cao, Q., Liu, B., and Hu, F. (2021). Effects of hydrological connection and
human disturbance on genetic variation of submerged Vallisneria natans
populations in four lakes in China. J. Oceanol. Limnol. 39 (4), 1403–1416. doi:
10.1007/s00343-020-0277-1
Chen, W., and Tan , X. (2006). The relationship between Chinese herbal
medicine cultivation and soil ecological factors. Chin. J. Tradit. Chin. Med.
Inform. 12, 48–49. doi: 1005-5304(2006)12-0048-02
Chinese Pharmacopoeia Commission (2020). Pharmacopoeia of people’s republic
of China. Beijing: China Pharmaceutical Technology Press. 2020, 317.
Christ, B., Pluskal, T., Aubry, S., and Weng, J. (2018). Contribution of untargeted
metabolomics for future assessment of biotech crops. Trends Plant Sci. 23 (12),
1047–1056. doi: 10.1016/j.tplants.2018.09.011
DiLeo, M., Strahan, G., den Bakker, M., and Hoekenga, O. (2011). Weighted
correlation network analysis (WGCNA) applied to the tomato fruit metabolome.
PloS One 6 (10), e26683. doi: 10.1371/journal.pone.0026683
Dong, L., Zhou, X., Ma, J., Zhou, H., and Fu, X. (2021). Chemical constituents of
Stellaria dichotoma var. lanceolata and their anti-inflammatory effect on
lipopolysaccharide-stimulated RAW 264.7 cells. Chem. Natural Compounds 57
(1), 158–162. doi: 10.1007/S10600-021-03306-0
Du, Y., Fan, J., Li, S., Niu, Y., Yu, H., and Huang, J. (2019). Fractal dimension
characteristics of soil particle size distribution under different vegetation patches in
desert steppe and its relationship with soil nutrients. Chin. J. Appl. Ecol. 30 (11),
3716–3724. doi: 10.13287/j.1001-9332.201911.004
Du, L., Zhao, J., Farhat, A., and Liu, W. (2013). Higher nitrates, p and lower pH
in soils under medicinal plants versus crop plants. Environ. Chem. Lett. 11 (4), 385–
390. doi: 10.1007/s10311-013-0419-1
Eguchi, T., Tanaka, H., Yoshida, S., and Matsuoka, K. (2019). Temperature
effects on the yield and quality of the medicinal plant Pinellia ternata breit.Environ.
Control Biol. 57 (3), 83–85. doi: 10.2525/ecb.57.83
Guo, L., Zhou, L., Kang, C., Wang, H., Zhang, W., Wang, S., et al. (2020).
Strategies for medicinal plants adapting environmental stress and”simulative
habitat cultivation”of dao-di herbs. China J. Chin. Materia Med. 45 (9), 1969–
1974. doi: 10.19540/j.cnki.cjcmm.20200302.101
Guo, H., Zhu, H., Guo, J., Huang, Z., and Guo, P. (2020). Comparative study
on extraction and detection methods of total flavonoids of alfalfa at different
growth stages. Chin.J.Grassland42 (3), 141–146. doi: 10.16742/
j.zgcdxb.20190270
Haruna, A., and Yahaya, S. (2021). Recent advances in the chemistry of bioactive
compounds from plants and soil microbes: a review. Chem. Africa 4 (2), 231–248.
doi: 10.1007/s42250-020-00213-9
Hemanta, K. (2017). Effect of salt stress on medicinal plants and its amelioration
by plant growth promoting microbes. Int. J. Bio Resource Stress Manag. 8 (2), 316–
326. doi: 10.23910/IJBSM/2017.8.2.1772
Huang, L., and Guo, L. (2007). Secondary metabolites accumulating and
geoherbs formation under enviromental stress. China J. Chin. Materia Medica
04), 277–280. doi: 1001-5302(2007)04-0277-04
Jat, R., and Gajbhiye, N. (2017). Secondary metabolites production influenced
with soil fertility and irrigation in medicinal plant; mandukaparni (Centella asiatica
l.). Natl. Acad. Sci. Lett 40 (2), 87–90. doi: 10.1007/s40009-016-0531-4
Jiang, D., Wang, H., Kang, C., Jiang, J., Du, Y., Zhang, Y., et al. (2020). Influence
and mechanism of stress combination on medicinal plants secondary metabolism.
ChinaJ.Chin.MateriaMedica45 (9), 2009–2016. doi: 10.19540/
j.cnki.cjcmm.20200302.106
Joshi, B., Bhandari, N., Shrestha, S., Gyawali, R., and Thapa, P. (2021).
Comparative study of polyphenol, flavonoid, and antioxidant activity of various
medicinal plants collected from different altitudes. Asian J. Pharm. Clin. Res. 9 (14),
87–93. doi: 10.22159/ajpcr.2021.v14i9.41564
Lang, D., Cui, J., Da, Z., Li, Y., Zhou, L., and Zhang, X. (2014). Effect of drought
stress on growth and physiological-biochemical characteristics of Stellaria
dichotoma.China J. Chin. Materia Medica 39 (11), 1995–1999. doi: 10.4268/
cjcmm20141109
Langfelder, P., and Horvath, S. (2008). WGCNA: an r package for weighted
correlation network analysis. BMC Bioinf. 9 (1), 1–13. doi: 10.1186/1471-2105-9-
559
Leng, F., Yang, Z., Wu, Y., and He, D. (2020). Physiological and photosynthetic
characteristics and active component contents of Polygonum multiflorum thunb.
under different soil pH values. Acta Botanica Boreali-Occidentalia Sinica 40 (09),
1566–1573. doi: 10.7606/j.issn.1000-4025.2020.09.1566
Li, Z., Cui, Y., and Qin, X. (2018). Challenge of quality evaluation of traditional
Chinese medicinal materials and application progress on metabolomic approach
inl its quality valuation. Chin. Tradit. Herbal Drugs 49 (10), 2221–2229. doi: 0253-
2670(2018)10-2221-09
Li, M., Hou, X., Zhang, X., and Wang, Y. (2019). Construction of database of
growth environment of dao-di herbs. China J. Chin. Materia Medica 44 (14), 3010–
3014. doi: 10.19540/j.cnki.cjcmm.20190504.104
Li, Z., Song, L., Lei, Y., Liang, W., Wang, H., and Peng, L. (2020). Advances in
biology, chemical constituents and pharmacological activities of Stellaria
dichotoma var. lanceolata.J. Nanjing Univ. Tradit. Chin. Med. 36 (1), 136–140.
doi: 10.14148/j.issn.1672-0482.2020.0136
Li, Z., Song, L., Wang, H., Li, Y., Yang, Y., and Peng, L. (2022). Analysis of
inorganic elements characteristics of Stellariae radix from different producing areas
based on ICP-MS combined with multiple statistics. Chin. J. Inf. TCM 29 (6), 103–
109. doi: 10.19879/j.cnki.1005-5304.202110406
Mudge, E., Applequist, W., Finley, J., Lister, P., Townesmith, A., Walker, K., et al.
(2016). Variation of select flavonols and chlorogenic acid content of elderberry
collected throughout the Eastern united states. J. Food Composition Analysis 47,
52–59. doi: 10.1016/j.jfca.2015.12.003
Nicholson, J., and Lindon, J. (2008). Metabonomics. Nature 455 (7216), 1054–
1056. doi: 10.1038/4551054a
Nicholson, J., and Wilson, I. (2003). Understanding ‘global’systems biology:
metabonomics and the continuum of metabolism. Nat. Rev. Drug Discov. 2 (8),
668–676. doi: 10.1038/nrd1157
Qing, T., Sun, C., Bi, Z., Liang, W., Li, P., Zhang, J., et al. (2020). Identification of
drought-related co-expression modules and hub genes in potato roots based on
WGCNA. Acta Agronomica Sinica 46 (7), 1033–1051. doi: 10.3724/
SP.J.1006.2020.94130
Si, Y., Jiao, Y., Li, L., Lin, H., Wang, C., Zhou, B., et al. (2021). Comprehensive
investigation on metabolites of Panax quinquefolium l. in two main producing
areas of China based on ultra-high-performance liquid chromatography-
quadrupole time-of-flight mass spectrometry. J. Mass Spectro metry 56 (12),
e4791. doi: 10.1002/JMS.4791
Tian, F., Li, Y., Yu, K., Li, F., and Wang, Y. (2013). Cultivation characteristic and
suggestion of medicinal crops. Crop J. 3, 12–16. doi: 10.16035/j.issn.1001-
7283.2013.03.007
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org18
Yuan, Y., and Huang, L. (2020). Molecular pharmacognosy in daodi herbs (in
Chinese). Chin. Sci. Bull. 65, 1093–1102. doi: 10.1360/TB-2020-0184
Zhang, W., Cao, Z., Xie, Z., Lang, D., Zhou, L., Chu, Y., et al. (2017). Effect of
water stress on roots biomass and secondary metabolites in the medicinal plant
Stellaria dichotoma l. var. lanceolata bge. Scientia Horticult. 224, 280–285. doi:
10.1016/j.scienta.2017.06.030
Zhang, J., Li, T., Cai, Y., and Wang, Y. (2021). Genetic and environmental effects
on allometry of the medicinal plant Dendrobium officinale (Orchidaceae) from
yunnan, southwest China. Pak. J. Bot. 53 (5), 1675–1682. doi: 10.30848/
PJB2021-5(14)
Zhang, G., Yun, X., and Gao, Y. (2019a). Phytochemical constituents from
Stellaria dichotoma var. lanceolata.Chem. Natural Compounds 55 (1), 135–137.
doi: 10.1007/s10600-019-02635-5
Zhang,X.,Zhao,D.,Zhang,W.,andWang,Y.(2012).Methodologystudyon
qualification of total sterol in Stellaria dichotoma l. var. lanceolata bge by ultraviolet
spectrophotometry. Ningxia Med. J. 34 (2), 126–127. doi: 10.13621/j.1001-5949.2012.02.007
Zhang, G., Zheng, J., Yun, X., Lei, J., Bai, R., Shi, W., et al. (2019b). Flavonoids
from Stellaria dichotoma var. Lanceolata.Chem. Natural Compounds 55 (1), 110–
111. doi: 10.1007/s10600-019-02626-6
Zou, J., Guo, Q., Liu, L., Jin, L., Song, K., and liu, M. (2019). Effects of soil pH on
physiological characteristics and medicinal qualities of G. longituba.Soils 51 (1),
68–74. doi: 10.13758/j.cnki.tr.2019.01.010
Zou, Q., Guo, Q., Wang, T., Chen, J., Yang, F., and Yang, C. (2022). Comparison
of metabolome characteristics and screening of chemical markers in
Chrysanthemum indicum from different habitats. Physiol. Mol. Biol. Plants 28
(1), 65–76. doi: 10.1007/s12298-022-01137-z
Li et al. 10.3389/fpls.2022.1035712
Frontiers in Plant Science frontiersin.org19
Available via license: CC BY
Content may be subject to copyright.
