全 文 ：热带亚热带植物学报 2016, 24(2): 128 ~ 142
Journal of Tropical and Subtropical Botany

Received: 2015–06–29 Accepted: 2015–09–25
This work was supported by the Natural Science Foundation of China (Grant No. 31070242), the Research Fund for the Doctoral Program, Ministry
of Education, China (Grant No. 20114407110006), the Leading Scientists Project of Pearl River Scholar Fund in Guangdong Province (Grant No.
2012).
GENG Yan, female, MD, interested in comparative transcriptomics and bioinformatics analysis of invasive plants. E-mail: gengyan_xx@163.com
* Corresponding author. E-mail: lishsh@scnu.edu.cn


de novo转录组学分析华南地区入侵植物五爪金龙
代谢特征

耿妍1, 陈玲玲1, 鲁焕1, 宁婵娟1, BJÖRN Lars Olof1,2, 李韶山1*
(1. 华南师范大学生命科学学院，广东省高校生态学与环境科学重点实验室，广州 510631; 2. 隆德大学生物学系，隆德 SE-22362, 瑞典)

摘要：为探究华南地区严重入侵植物五爪金龙(Ipomoea cairica)生物入侵的分子机制，对五爪金龙及其近缘种七爪金龙(I.
digitata)和裂叶牵牛(I. nil)进行 de novo转录组测序和组装，得到 56551 条 unigenes，其中 56522条得到注释，7815条 GO 注
释，15615条 COG注释，180201条 KEGG数据库注释。转录组分析结果表明，五爪金龙氮代谢通路关键酶基因的表达高于
对照。次生代谢关键酶(PAL、4CL、CAD、查耳酮合酶、苯基丙乙烯酮异构酶、槲皮黄 3-O-甲基转移酶等)基因在五爪金龙
与七爪金龙及裂叶牵牛中均得到协同性的差异表达，而这些代谢通路指导的产物合成对五爪金龙的抗逆境能力、生长、化感
作用等均起关键作用。关键基因的 RT-qPCR 验证结果与转录组结果具有一致性。因此，这从分子生物学层面上对解释五爪
金龙在华南地区的入侵机制提供了新的证据。
关键词：de novo；转录组；五爪金龙；生物入侵；代谢
doi: 10.11926/j.issn.1005-3395.2016.02.002

Metabolic Characteristics of Invasive Plant Ipomoea cairica in South China
by de novo Transcriptomics

GENG Yan
1
, CHEN Ling-ling
1
, LU Huan
1
, NING Chan-juan
1
, BJÖRN Lars Olof
1,2
, LI Shao-shan
1*

(1. School of Life Science, Key Laboratory of Ecology and Environmental Science in Guangdong Higher Education, South China Normal University,
Guangzhou 510631, China; 2. Department of Biology, Lund University, Sölvegatan 35, SE-223 62 Lund, Sweden)

Abstract: To explore the molecular mechanisms of Ipomoea cairica invasiveness in south China, the de novo
transcriptomes from I. cairica and two related species, I. digitata and I. nil, were sequenced and assembled. There
were 56551 all-unigenes obtained by assembling the reads, among them 56522 all-unigenes were annotated,
including 7815, 15615, and 180201 all-unigenes in GO, COG and KEGG databases, respectively. Moreover, the
activities of NR and GS in I. cairica, key enzyme in metabolic pathway for nitrogen, were greater than those in
related species. In addition, the transcriptome data showed that the genes of key enzymes in secondary
metabolism, such as pal, 4cl, cad, chs, and chi, had synergic differential expression in I. cairica, I. digitata and I.
nil. The production synthesis from metabolic pathway could play a key role in stress-resistant, growth and
allelopathy of I. cairica. The RT-qPCR verification results of key genes were similar to those from transcriptome.
Therefore, the result of the present research might explain partly the successful invasiveness of I. cairica in South
China at the level of molecular biology.
Key words: de novo; Transcriptome; RNA-seq; Ipomoea cairica; Biological invasion; Metabolism
第 2期 耿妍等: de novo转录组学分析华南地区入侵植物五爪金龙代谢特征 129


Ipomoea cairica (Convolvulaceae) is a perennial
herb with trumpet-shaped flowers and five-lobed leaf
blades
[1]
. It was planted as an ornamental plant at first,
but later became invasive. It originated in South or
Central America, naturalized in Hong Kong in 1921,
and is widespread in southern China
[2]
. The plant
grows in full sun with well-drained soil, such as flat
ground, road side and brush with sufficient sunlight.
Unfortunately, it grows not only along the ground but
also climbs and covers other plants. The covered plants
are damaged and even die because of reducing sun-
light
[3]
. Due to its rapid growth, strong environmental
adaptability, and high photosynthetic capacity, I.
cairica has rapidly spread in tropical and subtropical
regions and has become one of the most severe alien
invasive plants in south China. Thus, this invasive
species markedly reduces biodiversity and damages
the landscapes of gardens, becoming one of the most
detrimental weeds in southern China
[4]
, second only to
Mikania micrantha Kunth. Ipomoea digitata, an
annual weed, with flowers similar to I. cairica but
5-7-lobed leaf blades, is a non-invasive species
growing in south China. Ipomoea nil, an annual weed
with 3-lobed leaf blades, has become widespread in
southern China and listed as a mildly invasive plant.
There are several protein sequences for Ipomoea
available in GenBank, but as the total number of DNA
and mRNA sequences is limited, we investigated the
transcriptome of I. cairica, setting I. nil and I. digitata
as controls, using next-generation sequencing technology
with de novo transcriptome assembly. The assembled,
annotated transcriptome sequences will provide a
valuable genomic resource for further understanding
the molecular basis of plant invasion as well as for
conducting additional studies.
Many reports from Brazil, Japan
[5]
, Malaysia
[6]

and China (Hong Kong)
[7]
have mentioned the use of I.
cairica in folk medicine. Most studies on I. cairica
focused on biochemical
[1,8]
and ecological characteristics
[3]
,
but its genome and transcriptome has seldom been
reported. Shekhar

et al.
[9]
conducted a differential
transcript expression analysis, but only between two
cultivars of I. batatas. The allelopathy of I. cairica
inhibiting seed germination has been reported, but the
secondary metabolism of I. cairica that produces
chemical weapons and the molecular mechanisms of
this meta- bolism have not been studied. Two factors
limiting further study are the lack of physiological
comparisons between the original species in South
America and the native species in China, and not
understanding the reasons that I. cairica is invasive in
southern China.
Nitrogen, an essential element for plant growth
and development, is one of the important indicators of
resource capture capabilities in invasive plants.
Therefore, this study explored the nitrogen capture
capabilities of I. cairica by comparing this process in I.
cairica to that in related species, I. digitata (native to
southern China) and I. nil (mildly invasive), belonging
to Convolvulaceae.

1 Materials and methods

1.1 Sample preparation and assembly
Samples of I. cairica, I. digitata, and I. nil were
collected from South China Normal University (Guang-
zhou, Guangdong, south China) in September, 2011 in
a phase of rapid growth. Samples had the same growth
vigour, young stem tips with a length from 10 cm to
15 cm were used in this study.
The total RNA extraction was conducted using a
polysaccharide and polyphenol-free RNA extraction
kit (BioTeke, Catalog number RP3201). The RNA was
enriched using magnetic beads with Oligo(dT) frag-
mented into short segments using fragmentation buffer.
Based on mRNA templates, the first-strand cDNA
synthesis was conducted with random hexamers, and
then buffer, dNTPs, RNase H and DNA polymerase I
was introduced for the synthesis of the second cDNA
strand. The products were purified using a QiaQuick
PCR kit and eluted with EB buffer. The ends were
repaired and the sequencing joints connected. The
fragments were detected using agarose gel electro-
phoresis followed by PCR. The cDNA library was
sequenced using an Illumina HiSeq 2000 instrument.
Bioinformatics analyses were applied to the
130 热带亚热带植物学报 第 24卷


sequencing results. The original image data were
transformed into sequence data using base calling.
Before assembly, adaptor sequences, meaning reads
containing more than 10% ‘N’ (where N represents
ambiguous bases in reads), and low quality sequences
(reads in which more than 50% of the bases had a
quality value<10) were removed. The 90 bp raw reads
were then filtered to obtain high-quality clean reads.
The short-sequencing reads were assembled using
SOAP de novo software
[10]
. SOAP de novo first connected
reads of significant overlap into longer fragments and
then obtained contigs without unknown sequences (N)
between two contigs. The reads were re-compared
with the contigs, ensuring different contigs from the
same transcripts and the distance of these contigs using
paired-end reads. SOAP de novo connected these contigs,
with the unknown sequences (N) in the middle, and
then obtained the scaffolds. The gaps in the scaffolds
were filled using paired-end reads, obtaining unigenes
with the least N that could not be extended at both ends.
The unigenes from different samples were further
assembled with redundancies eliminated to obtain the
longest possible unigenes with no redundancies.
Finally, BLASTx alignment (https://blast.ncbi.nlm.
nih.gov/Blast.cgi?PROGRAM=blastx&PAGE_TYPE=
BlastSearch&BLAST_SPEC=blast2seq&LINK_LOC=
blasttab) was conducted to compare the unigene
sequences to several protein databases, including non-
redundant (nr) NCBI, Swiss-Prot, Kyoto Encyclopedia
of Genes and Genomes (KEGG), and Cluster of Ortho-
logous Groups (COG). BLAST matches with E-value
scores less than 0.00001 were considered significant.
The resulting best score for the alignment was used to
determine the sequence direction. If contradictions
occurred in the results from the different databases,
then the orientation of the unigenes were determined
following the priority scheme of nr, followed by
Swiss-Prot, KEGG, and finally COG. If the unigenes
were not comparable with any of these protein data-
bases, then the coding region was predicted and the
orient of the sequence was determined using
ESTScan
[11–12]
. For unigenes with sequence direction,
we provided their sequences from the 5′ end to 3′ end,
for those without any direction, we provided their
sequences from assembly software.

1.2 Functional annotation
The unigenes were functionally annotated. The
BLAST search for the unigenes was run against the
COG database of proteins to predict the functional
categories and classification statistics, obtaining the
gene functional characteristics at the macro level.
The nr annotated unigenes were mapped against
the Gene Ontology (GO) database to retrieve functional
annotations through GO terms. A statistical analysis of
the GO terms offered the functional characteristics of
the genes at the macro level. After obtaining the GO
classifications, the unigenes were mapped to the KEGG
database to determine functional annotations and generate
corresponding pathway maps for the unigenes
[13]
,
providing relationship information for the unigenes, the
species and their interactions with the environment.

1.3 Predictive coding protein box
The sequence was translated into its amino acid
sequence using a standard code table, and the nucleotide
sequence was obtained. If none of the above databases
were successfully blasted against the unigene, then the
coding region was predicted to determine the orien-
tation of the nucleotide sequence (5′ to 3′) and amino
acid sequence using EST Scan
[11]
.

1.4 Differentially expressed genes
Differential gene expression analysis was conducted
according to the method of Audic, et al
[14]
. The distri-
bution of P(x) follows a Poisson distribution (Eq. 1), λ
is the number of real transcripts.
e
( )
!
x
P x
x

 (Eq. 1)
The probability that the expression of gene A is
equal in both samples is calculated using Eq. 2
[14]
.
!
2
( 1)
1 ! ! 2
1
( )
( | )
1
 
  
  
   
 
 
y
x y
N x y
P y X
N N
x y
N

0 0 0
2 ( | ) or 2(1 ( | )) if ( | ) 0.5
i y i y i y
i i i
p i x p i x p i x
  
  
    (Eq. 2)
The differences in the P-values for multiple hypo-
thesis testing were examined, and the thresholds of
第 2期 耿妍等: de novo转录组学分析华南地区入侵植物五爪金龙代谢特征 131


the P-values were decided using the false discovery rate
(FDR) method
[15]
. After this analysis was conducted, the
GO function and KEGG pathway were analyzed.
The expression of the unigene was calculated using
the RPKM method
[16]
(Eq. 3). It can determine the
effect of gene length and sequencing differences on
gene expression and can be used to directly compare
different samples.
6
3
10 G
RPKM
( ) /10NL
 (Eq. 3)
When RPKM(A) is assumed to be the expression
of unigene A, then C denotes the number of reads
blasted to only unigene A, N denotes the total number
of reads which could be blasted to all unigenes, and L
denotes the number of bases in unigene A.

1.5 GO and pathway analysis of differentially
expressed genes
The GO analysis
[17]
of differential gene expression
provides molecular function, biological process and
cellular component annotations, and can also determine
significantly enriched functions in the differentially
expressed genes. All of the differentially expressed
genes were first mapped onto the average terms in the
GO database (http://www.geneontology.org/). The
number of genes in every term was calculated and the
significantly enriched genes were determined by
comparing the genome with a super geometry inspection
according to Eq. 4.
1
0
m
i
M N M
i n i
P
N
n


  
  
  
 
 
 
 (Eq. 4)
N denotes the number of genes annotated by GO, n
denotes the number of differentially expressed
genes, and M denotes the number of genes annotated
with the special GO terms in all unigenes.
Thus, the main biological functions of the
differentially expressed genes were determined using
GO enrichment analysis
[17]
.

1.6 Confirmation using Real-time qPCR
Primers were designed using Primer 3.0. The primer
sequence and gene characteristics are summarized in
Table 1.
The RT-PCR reactions were performed using an
Applied Biosystems 7500 Real-time PCR system with
SYBR

Premix Taq

(TaKaRa) in 96 connected tubes
(TaKaRa). Each sample was analyzed in triplicate in a
20 μL reaction volume, containing 0.5 μL of each primer,
10 μL of SYBR Green I, 1 μL of diluted cDNA, and
8 μL of RNase free dH2O. For each experiment, the
endogenous control gene actin was analyzed in triplicate
(positive control) with a non-template reaction (negative
control) and a water only reaction (blank control). The
relative expression levels were calculated using the
2
–ΔΔCT method
[18]
.

Table 1 Nucleotide sequences of the primers used in SYBR Green I
RT-qPCR
Gene Species Primer sequences (5′~3′)
Actin I. cairica GCGGATAGAATGAGCAAGGAA
GGGCCGGACTCATCATACTC
I. digitata TTGTAGCACCACCTGAAAGGAAAT
CGGACTCATCATACTCTGCCTTG
I. nil TGTGACAATGGAACTGGAATGGT
TTGATTGAGCTTCATCTCCGACAT
Nitrate
reductase
I. cairica TCACTCGAGGTCGAGGTTC
CTCGAGGTGTCTTTCCTTC
I. digitata CCTTCATGAACACGGCTTCT
TTCTTCGCCTTATCGGAGTG
I. nil TCTTCATCTGCGCCGCCATTG
GGGAGTCGAGGTGTTGGGACAT
Glutamine
synthetase
I. cairica CTCCAGCAGGTGAGCCTATC
AGCCTTGTCTGCACCAATTC
I. digitata TGAGGAGCAAAGCCAGGACCAT
ACTGTCCTCTCCAGGAGCTTGG
I. nil CGGGAGAGGACAGTGAAGTC
ACCAAGTGGCCATTTCACAT

1.7 Determination of total nitrogen content of leaf (TN)
TN in leaf was determined by using sulfuric acid-
hydrogen peroxide digestion-distillation and titration
method
[19]
.

1.8 Determination of activities of NR and GS
The activity of NR (NRA) was determined by
using colorimetric determination of sulfanilamide in
vivo
[20]
.
The activity of GS (GSA) was determined in the
following way: (1) From leaves, young stems and
roots of each species 0.2 g samples were collected; (2)
Three mL ice-cold extraction buffer [50 mmol L
−1

Tris-HCl (pH 7.5), 10 mmol L
−1
MgCl2, 1 mmol L
−1

EDTA, 10% (w/v) glycerol, 14 mmol L
−1
mercap-
132 热带亚热带植物学报 第 24卷


toethanol] and 1% (w/v) PVP] was added, then ground
on ice until homogenized; (3) The homogenate was
put into 1.5 mL microtubes, and centrifuged for
30 min at 4℃, 13000×g. Then the enzyme extracts
were transferred to new centrifuge tubes for standby.
For the next steps the GS testing kit directions of
Nanjing Jiancheng Bioengineering Institute were
followed.

2 Results and discussion

2.1 Sequencing and assembly
There were 26501994, 26901752 and 23139614
high-quality reads obtained from I. cairica, I. digitata
and I. nil, respectively, and assembled them using SOAP
de novo to obtain 384401 contigs with a mean length
of 164 nt in I. cairica, 352302 contigs with a mean
length of 171 bp in I. digitata, and 368222 contigs
with a mean length of 157 bp in I. nil (Table 2). With
paired-end read joining and gap-filling, the above
contigs were further assembled into 124007 scaffolds
with a mean size of 343 bp in I. cairica, 112882 scaffolds
with a mean size of 342 bp in I. digitata, and 98903
scaffolds with a mean size of 362 bp in I. nil. Using gap
filling and long sequence clustering, the scaffolds
were assembled into 90329 I. cairica unigenes with a
mean length of 423 bp, 86946 I. digitata unigenes
with a mean length of 441 bp, and 77255 I. nil unigenes
with a mean length of 428 bp. In addition, using the
same strategy, we obtained 56551 all-unigenes with a
mean length of 781 bp from the I. cairica, I. digitata
and I. nil reads combined (all-unigenes originated
from all of the I. cairica, I. digitata and I. nil reads,
and unigenes of each species from itself reads).

Table 2 Transcriptome qualities in Ipomoea cairica, I. digitata, and I. nil
I. cairica I. digitata I. nil Total
Total nucleotides (nt) 2385179460 2421157680 2082565260 –
Total number of reads 26501994 26901752 23139614 –
GC content (%) 47.75 47.79 48.66 –
Total number of contigs 384401 352302 368222 –
Mean length of contigs (bp) 164 171 157 –
Total number of scaffolds 124007 112882 98903 –
Mean length of scaffolds (bp) 343 369 362 –
The number of unigenes 90329 86946 77255 –
Mean length of unigenes 423 441 428 –
Number of all-unigenes – – – 56551
Mean length of all-unigenes – – – 781

For reads, the original data were transformed into
sequence data by base calling from raw data or raw
reads obtained as clean reads after removal of impu-
rities. For contigs, joint reads with overlap into longer
fragments without unknown sequences between each
two contigs (N) using SOAP de novo software. For
scaffolds, contigs joined using SOAP de novo soft-
ware. Unigenes represent sequences that cannot be
extended on either end and contain the least N.
The length and gap distributions of unigenes
from three species and all-unigene are shown in Fig. 1.
Overall, all-unigenes were better in quality than
unigenes from one of three species and obtained higher
scores when using BLASTx to search against the nr
database.
2.2 GO and COG classifications
We further classified all-unigenes using GO assign-
ments after functional annotation. Based on homologous
genes, 37815 all-unigenes were categorized into 44
GO terms under three domains: biological process,
cellular component and molecular function (Fig. 2). The
results indicated that the terms with the highest
percentage were cell, cell part, cell process and metabolic
process, whereas no genes were assigned to the terms
biology adhesion, cell killing, location, viral reproduction,
viron, viron part or electron activity. To further examine
the integrity of our transcriptome library and effec-
tiveness of the annotation process, we calculated the
unigene numbers with COG classifications. There were
15615 all-unigenes with a COG classification (Fig. 3).
第 2期 耿妍等: de novo转录组学分析华南地区入侵植物五爪金龙代谢特征 133


Of the 25 COG categories, the cluster predicting
general function had the highest number of unigenes,
followed by that for transcription, whereas the clusters
of nuclear structure and extracellular structure had the
fewest number of unigenes (Fig. 3).
GO categories, shown on the X-axis, are group
into three ontologies: biological process, cellular
component, and molecular function. The right Y-axis
indicates the number of genes in each category, and
the left Y-axis indicates the percentage of total genes
in that category. ‘All-unigene’ indicates that the unigenes
were assembled from reads of I. cairica, I. digitata
and I. nil samples.
15615 unigenes were assigned to 25 categories in
the COG functional classification. The Y-axis indicates
the number of genes in a specific functional cluster.
The legend to the right of the histogram lists the 25
functional categories.

2.3 Unigene networks and differentially expressed
gene pathway enrichment analysis
The biological pathways in I. cairica, I. digitata
and I. nil were analyzed. There were 25725 all-unigenes
mapped to Nr database, 30832 all-unigenes mapped to
Swiss-Prot database, 23659 all-unigenes mapped to
the KEGG database, 15615 mapped to COG database
and 37815 mapped to GO. In a more detailed analysis,
frequency of 30020 all-unigenes assigned to 119
different pathways (Data not show). To identify gene
expression without taking into account the effects of
different gene lengths or differences in the total
number of reads, we used the transformed RPKM value
with a false discovery rate (FDR)≤0.001 and a log2
ratio≥1 (representing at least a 2-fold change) to signify
differences in gene expression. We use these assumptions
to analyze the differences between I. cairica vs I.
digitata, I. cairica vs I. nil, and I. digitata vs I. nil.

2.4 Total leaf nitrogen content
The total nitrogen (TN) content represents an
important physiological indicator of nitrogen use
efficiency in plants. In present experiment, it is found
that TN in I. cairica was significantly higher than that
in the related species (Fig. 4).
2.5 Nitrate reductase and glutamine synthetase
activities
The activity of nitrate reductase (NR), involved
in the first and rate-limiting step for the NO3
−
assimi-
lation process, directly affects the efficiency of NO3
−

assimilation. The NR activity in leaf, stem, and root
was significantly higher in I. cairica than that in two
related species (Fig. 5: A).
Glutamine synthetase (GS) is a multifunctional
enzyme involved in the regulation a variety of
processes in nitrogen metabolism. GS activity can
reflect the capacity of nitrogen assimilation
[21]
. The
GS activity in stem was significantly higher in I.
cairica than that in two related species (Fig. 5: B).

2.6 Differential gene expression in nitrogen metabolism
A comparison of the expression of genes involved
in the nitrogen pathway in I. cairica vs. I. digitata and
in I. cairica vs. I. nil demonstrated some DEGs
(FDR≤0.001 and log2 ratio≥1). A total of 84 DEGs
were detected between I. cairica and I. digitata, in
which 48 unigenes were up-regulated and 36 unigenes
were down-regulated. Similarly, a totalof 79 DEGs
were detected between I. cairica and I. nil, in which
52 unigenes were up-regulated and 27 unigenes were
down-regulated.
Before undertaking an analysis of the differences
in transcript abundance among the three species, we
used real-time quantitative PCR to confirm the
estimates of transcript abundance obtained with
RNA-seq. We chose to examine genes encoding NR
and GS. The results showed that the gene expression
profiles obtained with the two methods were similar
(Fig. 6). Therefore, the ratios of transcript abundance
obtained using RNA-seq are suitable for calling DEGs
between I. cairica and its two related species.

2.7 Functional annotation analysis of DEGs in
nitrogen metabolism
To further understand the function of co-DEGs,
we mapped them using the KEGG database for analysis
of nitrogen metabolism pathway (Table 4). DEGs code
for important enzymes in nitrogen metabolism, for
example, NR and GS. In addition, the levels of most
of co-DEGs were up-regulated in I. cairica compared
134 热带亚热带植物学报 第 24卷



Fig. 1 Length and gap distributions of unigenes after de novo assembly using the reads from Ipomoea cairica, I. digitata, I. nil and all samples.
第 2期 耿妍等: de novo转录组学分析华南地区入侵植物五爪金龙代谢特征 135



Fig. 2 Histogram of gene ontology classification


Fig. 3 Histogram of clusters of orthologous groups (COG) classification. A:
RNA processing and modification; B: Chromatin structure and dynamics; C:
Energy production and conversion; D: Cell cycle control, cell division,
chromosome partitioning; E: Amino acid transport and metabolism; F:
Nucleotide transport and metabolism; G: Carbohydrate transport and
metabolism; H: Coenzyme transport and metabolism; I: Lipd transport and
metabolism; J: Translation, ribosomal structure and biogensis, K: Transcrip-
tion, L: Replication, recombination and repair; M: Cell wall/membrane/
envelope biogenesis; N: Cell motility; O: Posttranslational modification,
protein turnover, chaperones; P: Inorganic ion transport and metabolism; Q:
Secondary metabolites biosynthesis, transport and catabolism; R: General
function prediction only; S: Function unkown; T: Signal transduction
mechanism; U: Intracellular trafficking, secretion, and vesicular transport; V:
Defense mechanism; W: Extracellular structures; Y: Nuclear structures; Z:
Cytoskeleton.

Fig. 4 Total nitrogen content in leaves

with those in other two species.

2.8 Transcripts related to secondary metabolism in
biosynthetic pathways
We also examine pathways related to the biosyn-
thesis of secondary metabolites, specifically the
biosynthesis of phenylpropanoid (KO00940), flavonoid
(KO00944), and flavones and flavonol (KO00941).
The results of the statistical analyses revealed that for
phenylpropanoid biosynthesis (KO00940) there were
333 differentially expressed unigenes between I. cairica
and I. digitata, as well as 348 between I. cairica and I.
nil; there were 220 unigenes differrentially expressed
for I. cairica compared with both I. digitata and I. nil.
For flavonoid biosynthesis (KO00944), there were
205, 208 and 139 differentially expressed unigenes,
and for flavones and flavonol biosynthesis (KO00941)
136 热带亚热带植物学报 第 24卷



Fig. 5 Nitrate reductase (NR) (A) glutamine synthetase (GS) (B) activity in leaves, stems and roots of Ipomoea cairica, I. digitata, and I. nil

Fig. 6 RT-PCR detection and RNA sequencing in Ipomoea cairica vs I. digitata (A) and I. cairica vs I. nil (B)

there were 59, 51 and 33 differentially expressed unigenes,
and for the biosynthesis of the secondary metabolites
with COGs there were 1398, 1446, and 926 between I.
cairica and I. digitata, between I. cairica and I. nil,
and for both I. digitata and I. nil, respectively.
Taken together, these data indicated the existence of
duplicate genes. Therefore, we removed the duplicates
and obtained 964 unigenes, with I. cairica different
from both I. digitata and I. nil (Data not show).
Among all those differences, between I. cairica and I.
digitata, I. cairica and I. nil, there were 493 unigenes
up-regulated and 303 unigenes down-regulated. These
results indicated that the differences were comparable
for both models, with 706 similarly expressed genes,
which was 82.57% of the total. Those differentially
expressed genes may explain the differences in the
biosynthesis of these secondary metabolites in I. cairica
compared with that in I. digitata and I. nil. It would be
interesting to determine whether those differences
were related to the invasiveness of I. cairica.
After further analysis of the annotations for those
differential genes, we found up-regulated expression
of some genes, Unigene32623_All, Unigene19708_ All
and Unigene32546_All code pal2, 33181_All code
pal4, related to a key enzyme in the biosynthesis of
secondary metabolites, phenylalanine ammonia lyase
(PAL), and the down-regulation of another PAL-
related gene, unigene4820_All code pal6, in I. cairica.

2.9 Analysis of physiological indexes
This study used high-throughput and de novo
transcriptome assembly methods to provide trans-
第 2期 耿妍等: de novo转录组学分析华南地区入侵植物五爪金龙代谢特征 137


criptome information for three species of Ipomoea,
three non-model plants. However, the invasion of I.
cairica is complex, with allelopathy and stress-
resistance both necessary and contributing to its
success. Previous studies have reported on the allelo-
chemicals and stress-resistance in I. cairica.
Our results indicated that the TN content in the
leaves of I. cairica was markedly higher than that in
the two related plants. However, this result did not
discriminate whether I. cairica had a greater capacity
for nitrogen absorption, reduction, or fixation. Nitrate
reductase is the first key enzyme in the nitrogen
metabolism pathway and is widely found in roots,
stems, leaves, and other tissues of higher plants. Evidence
indicates that the activity of NR reflects the status of
nitrogen nutrition and nitrogen metabolism in plants
as well as directly impacts the efficiency of inorganic
nitrogen use from soil and the synthesis rate of amino
acids in plants
[22]
. NO is an intracellular and extracellular
signaling molecule involved in various processes that
promote plant growth and development and alleviate
aging and stress damage
[23]
. Glutamine synthetase,
another key enzyme in the nitrogen metabolism path-
way, plays an important role in ammonia assimilation.
Data indicate that GS activity enhances ammonium
assimilation and nitrogen transformation
[24]
. In the
present study, NR activity was markedly higher in the
roots, stems and leaves and GS activity was signifi-
cantly higher in the stems of I. cairica compared with
those in the two related species (Fig. 5). Studies have
shown that increased nutrient availability favors the
invasion and success of invasive plant species in eco-
systems where resource availability is typically low
[25]
.
Nitrogen is the most important nutrient limiting plant
growth and development. Studying nitrogen utilization
benefits understanding of nutrient adaptation mecha-
nisms in invasive species
[26]
. The nitrogen assimilation
capacity of I. cairica is greater than that of the native
plant Paederia scandens
[27]
. Our results showed that
NR and GS activity as well as TN content in the leaves
of I. cairica were superior to those in the two related
species. Thus, the enahnced ability of I. cairica to
metabolize nitrogen and accelerate nitrogen transfor-
mation is likely an important mechanism in its successful
invasion.

2.10 Analysis of DEGs in the nitrogen metabolic
pathway
To better understand the biological functions of
DEGs in nitrogen metabolism, we used the KEGG
database to analyze the nitrogen metabolism pathway
(KO00910). In this experiment, 39 co-DEGs were iden-
tified and annotated for I. cairica and its two related
species (Table 3). Interestingly, most of the co-DEGs
were up-regulated, suggesting they may play an
important role in capturing nitrogen resources for I.
cairica.
It is generally believed that nitrate is the main
form of nitrogen absorbed by plants. Nitrate reductase
is the enzyme used in the first step for catalyzing
nitrate assimilation, and its synthesis is strictly modu-
lated at the transcriptional and post-transcriptional
levels
[28]
. It was reported that the double mutant
Arabidopsis NR (nia1, nia2) is etiolated, the content of
leaf amino acid is reduced, and it grows more slowly
compared with the wild type
[29]
. Nitrite reductase (NiR)
is involved in the second step for nitrate assimilation,
which catalyzes the reduction of nitrite to ammonium.
The higher activity of NiR prevents toxic effects by
accumulating NO2 on plants, and provides ammonia
for assimilation in next step
[30]
. The overexpression of
NR or NiR enhances mRNA levels, and improves
nitrogen assimilation in plants
[31]
. Transcriptomics
analysis in the present study showed differential regu-
lation for NR and NiR gene expression, with expression
up-regulated in I. cairica relative to the other two species,
indicating that the invasive I. cairica has a stronger
ability to assimilate nitrate.
Ammonia is assimilated in plants by the glutamine
synthetase/glutamate synthase (GS/GOGAT) cycle or
glutamate dehydrogenase (GDH) pathway
[32]
. The
GS/GOGAT cycle is the major pathway for ammonia
assimilation in higher plants, whereas the key enzyme
GS plays a central role in nitrogen metabolism of
138 热带亚热带植物学报 第 24卷


Table 3 Annotation of differentially expressed unigenes in nitrogen metabolism in Ipomoea cairica (IC) vs. I. digitata (ID) and in I. cairica vs. I. nil (IN)
Gene ID IC vs. ID IC vs. IN Annotation Gene name [species] Database
Unigene24221_All 13.7 1.4 Probable cytochrome b5 isoform 2 NR [Vf] KEGG
Unigene25910_All 12.6 12.6 Delta(5) fatty acid desaturase B NR [Nt] KEGG
Unigene10704_All 5.2 1.8 Ferredoxin-nitrite reductase ferredoxin-NiR [Nt] KEGG
Unigene27753_All 2.8 14.5 Laccase-12 NiR (NO-forming) [Gm] KEGG
Unigene38636_All 1.9 4.3 Laccase-17 NiR (NO-forming) [Pt] KEGG
Unigene14693_All –2.9 13.2 Glutamate synthase [NADH] GS [Vv] KEGG
Unigene31766_All 3.5 14.3 Glutamine synthetase leaf isozyme GS [Cl] KEGG
Unigene15163_All 5.0 4.7 Glutamine synthetase nodule isozyme GS [Hb] KEGG
Unigene31658_All 1.3 3.3 Glutamine synthetase root isozyme B GS [Eu] KEGG
Unigene18636_All 4.3 12.6 Glutamate dehydrogenase A GDH(NAD(P)+) [Np] KEGG
Unigene27731_All 5.4 4.2 Glutamate dehydrogenase A GDH(NAD(P)+) [Np] KEGG
Unigene9230_All 1.9 1.5 Glutamate dehydrogenase A GDH(NAD(P)+) [Np] KEGG
Unigene35570_All 1.6 1.4 Glutamate dehydrogenase A GDH(NAD(P)+) [Np] KEGG
Unigene31232_All 4.4 5.0 Probable glutamate dehydrogenase 3 GDH(NAD(P)+) [At] KEGG
Unigene49856_All 3.1 1.4 Glutamate dehydrogenase GDH(NAD(P)+) [Ac] KEGG
Unigene25819_All 3.7 8.4 Asparagine synthetase, nodule AS [Sp] KEGG
Unigene11157_All 2.5 4.3 Asparagine synthetase AS [Ha] KEGG
Unigene33181_All 17.4 17.4 Phenylalanine ammonia-lyase 1 PAL [Cc] KEGG
Unigene32546_All 3.7 17.4 Phenylalanine ammonia-lyase PAL [Me] KEGG
Unigene32623_All 1.9 2.1 Phenylalanine ammonia-lyase PAL [Ib] KEGG
Unigene19708_All 1.8 1.5 Phenylalanine ammonia-lyase PAL [Ib] KEGG
Unigene32744_All 2.2 2.8 Laccase-2 L-ascorbate oxidases [Pt] KEGG
Unigene31161_All 7.1 4.7 Laccase-3 L-ascorbate oxidase [At] KEGG
Unigene31213_All 12.4 12.4 Laccase-4 L-ascorbate oxidase [Pt] KEGG
Unigene26200_All 4.5 4.9 Laccase-6 L-ascorbate oxidase [At] KEGG
Unigene25388_All 8.1 13.8 Laccase-11 L-ascorbate oxidase [At] KEGG
Unigene26384_All 3.6 3.9 Laccase-12 L-ascorbate oxidase [At] KEGG
Unigene16438_All 1.1 1.0 Laccase-12 L-ascorbate oxidase [Pt] KEGG
Unigene19574_All 1.6 2.0 Laccase-17 L-ascorbate oxidase [Lt] KEGG
Unigene34362_All 2.3 10.2 L-ascorbate oxidase homolog L-ascorbate oxidase [Nt] KEGG
Unigene32435_All 5.4 4.0 L-ascorbate oxidase L-ascorbate oxidase [Nt] KEGG
Unigene10363_All –1.9 –1.5 L-ascorbate oxidase homolog L-ascorbate oxidase [Mt] KEGG
Unigene12570_All –1.0 –1.6 L-ascorbate oxidase homolog L-ascorbate oxidase [Nt] KEGG
Unigene147_All –2.3 –4.8 Laccase-4 L-ascorbate oxidase [Nt] KEGG
Unigene13298_All –4.6 –6.5 Phosphopantetheineadenylyltransferase L-ascorbate oxidase [At] KEGG
Unigene31031_All 13.2 4.1 Flavin-containing monooxygenase
FMOGS-OX5
Dimethylaniline monooxygenase [Alsl] KEGG
Unigene21078_All 1.1 2.4 Carbonic anhydrase Carbonic anhydrase [Alsl] KEGG
Unigene21626_All 1.4 1.6 Cyanatehydratase Cyanatelyase [Zm] KEGG
Unigene30713_All 6.5 14.3 Formamidase formamidase [Osjg] KEGG
Vv: Vitis vinifera; Cl: Canavalia lineata; Osjg: Oryza sativa Japonica; Hb: Hevea brasiliensis; Eu: Elaeagnus umbellata; Np: Nicotianaplum baginifolia;
At: Arabidopsis thaliana; Ac: Actinidia chinensis; Cc: Coffea canephora; Me: Manihot esculenta; Ib: Ipomoea batatas; Sp: Securigera parviflora; Ha:
Helianthus annuus; Pt: Populus trichocarpa; Gm: Glycine max; Lt: Liriodendron tulipifera; Mt: Medicago truncatula; Nt: Nicotiana tabacum; Alsl:
Arabidopsis lyrata ssp. lyrata; Vf: Vernicia fordii; Zm: Zea mays.

plants. It was reported that GS over-expression can
improve the efficiency of nitrogen use and promote
plant growth
[33]
. In addition, the GDH pathway supple-
ments the GS/GOGAT cycle for ammonia assimilation.
GDH has a dual function, as it can catalyze ammonia
assimilation and help glutamate oxidative deamination
[34]
.
Under conditions of excess ammonia, GDH deamination
is markedly increased in plants, providing a carbon
skeleton for the citric acid cycle
[35]
. Both GS and
GDH play an important role in maintaining balance
between carbon and nitrogen (C/N) in metabolic
processes
[36]
. The results of our transcriptomics analysis
showed that expression levels of GS and GDH were
higher in I. cairica than those in other two related
species, indicating that invasive I. cairica may have a
stronger ability to assimilate ammonia in plants.
第 2期 耿妍等: de novo转录组学分析华南地区入侵植物五爪金龙代谢特征 139


Asparagine synthetase (AS) catalyzes the synthesis
of asparagine from inorganic amine, which is the main
form of organic nitrogen for transport and storage. It is
involved in primary and secondary metabolism, signal
transduction pathways, metabolic stress, aging, and
other physiological and biochemical processes
[37]
.
Additionally, AS overexpression in plants increases
the level of free asparagine and promotes plant
growth
[38]
. We found that the invasive plant I. cairica
showed up-regulated expression of AS gene compared
with the related species, which would favor the trans-
portation and storage of organic nitrogen in I. cairica.
Phenylalanine ammonia-lyase (PAL) is a key
enzyme conducting the transition from primary to
secondary metabolism in plants. PAL catalyzes deami-
nation of L-phenylalanine into cinnamic acid. This
process is the central step in the secondary metabolic
pathway. PAL, the first enzyme in the phenyl acid
metabolic process, is deemed the first key enzyme
[39]
.
PAL plays an important role in pigment formation,
cell differentiation and lignification processes, resis-
tance to diseases, conditions of pest and adversity, and
other processes
[40]
. PAL catalyzes amino acids into
nitrogen compounds, which provides more nitrogen
resources for other metabolites in plants. Our trans-
criptomics analysis showed that the gene coding for
PAL had a higher expression level in I. cairica than in
the two related species, and this higher expression
may lead to the enhancement of nitrogen catalytic
efficiency in nitrogen metabolism.
L-ascorbate oxidase (AO) is a copper-containing
oxidase, mainly located in the cytoplasm or cell wall,
which can catalyze the oxidation of ascorbic acid. AO
regulates stress response, gene expression, growth and
development, and floral induction in plants
[41]
. Our
transcriptomics data suggested that some of the
unigenes modulating AO were up-regulated while
others were down-regulated in I. cairica compared
with those in the two related species. These levels may
adapt with the function of AO, which requires further
study.
The remaining genes related to plant growth and
development, resource utilization, and resistance
encode monooxygenases, carbonic anhydrase, cyanate
lyase and formamidase. Dimethylaniline monooxy-
genase is the most important monooxygenase system.
Carbonic anhydrase activity is positively correlated
with the net photosynthetic rate in Brassica campestris

and is very important for adapting to different nitrogen
environments
[42]
. Under conditions of nitrogen deprivation,
the gene that codes for formamidase will up-regulate
to adapt to the environment
[43]
.
Thus, overall, the DEGs involved in the nitrogen
metabolism pathway may provide some clues and
references from different aspects to understand nitrogen
utilization in I. cairica.

2.11 Secondary metabolism
PAL exists in all higher plants. It is encoded by
multigene families, and its genetic traits are conserved.
A previous study constructed a molecular evolutionary
tree using Clustalx by analyzing the bioinformatics of
PAL, and plants in the same families were clustered
[45]
.
Plants with a close kinship were also clustered
[44]
. The
Clustalx results offered valuable information about the
three plants used in the present study, which are in the
same family and genus. Disease- and stress-resistance
characteristics are hereditary, formed by mutual adap-
tation between plants and pathogenic microorganisms
during evolution. Plants may undergo a series of
physiological changes and defense reactions under
stress, including synthesis and accumulation of disease-
resistant substances such as lignin
[45]
. Phenylpropanoid
biosynthesis is also closely related to disease resis-
tance, and the activity of PAL is related to cold tole-
rance in banana plants
[46]
. I. cairica is less sensitive to
cold temperatures than the non-invasive or mildly
invasive plants I. triloba and I. nil, and this modifi-
cation likely contributes to its ecological adaptability
and invasiveness
[47]
. PAL plays an important role in
the biosynthesis of phytoalexin and lignin as well as
with other key enzymes, such as 4-coumarate, coenzyme
A Ligase (4CL) and cinnamyl-alcohol dehydrogenase
(CAD), all of which show highly coordinated expression.
Lignin is the main component in the plant secondary
cell wall
[48]
, increasing the mechanical strength and
140 热带亚热带植物学报 第 24卷


forming a hydrophobic network structure in the cell
wall to closely link the cellulose and prevent cell wall
dehydration
[49]
. Thus, lignin is advantageous for
transporting moisture through tissues and providing
resistance to adverse environments.
The expression of chalcone synthase (CHS), an
enzyme in flavonoid biosynthesis (KO00944), is
up-regulated in I. cairica compared with that in I.
digitata and I. nil. The expression of another key
enzyme in I. cairica, chalcone isomerase (CHI), is up-
regulated compared with that in I. digitata but down-
regulated compared with that in I. nil. The expression
of a key enzyme in the metabolic pathway of flavones
and flavonols, quercetin-3-O-trans-methylase, is up-
regulated in I. cairica compared with that in both I.
digitata and I. nil. Interestingly, 3-3′-5-trihydroxy-
4′-7-dimethoxyflavone and 3-3′-5-trihydroxy-4′-7-
dimethoxyflavone-3-O-sulfateare are found in I.
cairica are chemical weapons that may inhibit the
growth of other plants
[50]
and, intriguingly, belong to
flavonoids
[51−52]
.
These differences in the biosynthesis of secondary
metabolites among I. cairica and other plants in the
same genus may partly explain the invasiveness of I.
cairica and the non-invasiveness of the other plants.
These secondary metabolite differences may also
provide a foundation for additional studies examining
the stress resistance in and allelopathic effects of I.
cairica. In addition, the results from our bioinformatics
study examining the transcriptomics of pal, 4cl, cad,
chs, and chi indicate that further studies on these
enzymes are warranted in I. cairica.

3 Conclusion

As the results suggested that the transcriptome
quality obtained was good and the control samples
were effective. We analyzed the biological functions
and metabolic pathways associated with some of these
genes. An analysis of the biosynthesis of the secondary
metabolites suggested capabilities for resistance to
stress, efficient reproduction, rapid growth, and alle-
lopathy in I. cairica that may partly explain the
invasiveness of I. cairica in southern China, although
the complex processes of a plant invading a new
region likely includes multiple genes and pathways.
As previously mentioned, the effective compo-
nents of I. cairica have been used as folk medicine.
Whereas the extracted essential oil shows larvicidal
activity, arctigenin, trachelogenin, petunia-A, muri-
catin-B, and pinoresinol
[53]
provide antiviral, purgative,
anti-rheumatism, and antibacterial therapies. A crude
leaf extract of I. cairica was also shown to have
oviposition-deterring activity and oviciding potential
against dengue vectors
[54]
. Interestingly, these effec-
tive components are mainly products of secondary
metabolism, indicating that studies on the biosynthesis
of secondary metabolites may reveal an economic
value for I. cairica. Thus, even though I. cairica is an
invasive weed, it may also serve as a valuable resource,
providing effective medicinal compounds.

Acknowledgments We thank the Beijing Genomics
Institute at Shenzhen (BGI) for assisting in sequencing, and
associate professor Jing Li from Shanghai Jiao Tong University
for offering valuable suggestions regarding the bioinformatics
analyses.

Refferences
[1] GONG N, WANG Y T, BJÖRN L O, et al. DNA C-value of 20 invasive
alien species and 3 native species in South China [J]. Arch Biol Sci,
2014, 66(4): 1465–1472. doi: 10.2298/ABS1404465G.
[2] LI W H, TIAN X S, LUO J N, et al. Effects of simulated defoliation on
growth and photosynthetic characteristics of an invasive liana, Ipomoea
cairica (Convolvulaceae) [J]. Invas Plant Sci Managem, 2012, 5(2):
282– 289. doi: 10.1614/IPSM-D-11-00088.1.
[3] LI W H, LUO J N, TIAN X S, et al. Patterns of defoliation and their
effect on the plant growth and photosynthetic characteristics of Ipomoea
cairica [J]. Weed Biol Manage, 2012, 12(1): 40–46. doi: 10.1111/j.
1445-6664.2012.00432.x.
[4] WANG R L, ZENG R S, PENG S L, et al. Elevated temperature may
accelerate invasive expansion of the liana plant Ipomoea cairica [J].
Weed Res, 2011, 51(6): 574–580. doi: 10.1111/j.1365-3180.2011.
00884.x.
[5] THOMAS T G, RAO S, LAI S. Mosquito larvicidal properties of
essential oil of an indigenous plant, Ipomoea cairica Linn. [J]. Jpn J
Infect Dis, 2004, 57(4): 176–177.
[6] THIAGALETCHUMI M, ZUHARAH W F, RAMI R A, et al. Assess-
第 2期 耿妍等: de novo转录组学分析华南地区入侵植物五爪金龙代谢特征 141


ment of residual bio-efficacy and persistence of Ipomoea cairica plant
extract against Culex quinquefasciatus Say mosquito [J]. Trop Biomed,
2014, 31(3): 466–476.
[7] LIN R J, CHEN C Y, LO W L. Cytotoxic activity of Ipomoea cairica
[J]. Nat Prod Res, 2008, 22(9): 747–753. doi: 10.1080/14786410701
628739.
[8] LIN Z F, CHEN L H, ZHANG W Q. Peroxidase from Ipomoea cairica
(L.) SW. isolation, purification and some properties [J]. Proc Biochem,
1996, 31(5): 443–448. doi: 10.1016/0032-9592(95)00088-7.
[9] SHEKHAR S, MISHRA D, BURAGOHAIN A K, et al. Comparative
analysis of phytochemicals and nutrient availability in two contrasting
cultivars of sweet potato (Ipomoea batatas L.) [J]. Food Chem, 2015,
173: 957–965. doi: 10.1016/j.foodchem.2014.09.172.
[10] LI R, YU C, LI Y, et al. SOAP2: An improved ultrafast tool for short
read alignment [J]. Bioinformatics, 2009, 25(15): 1966–1967. doi: 10.
1093/bioinformatics/btp336.
[11] ISELI C, JONENEEL C V, BUCHER P. ESTScan: A program for
detecting, evaluating, and reconstructing potential coding regions in
EST sequences [C]// Proceedings of the 7th International Conference
on Intelligent Systems for Molecular Biology. California: AAAI Press,
1999: 138–158.
[12] PERTEA G, HUANG X Q, LIANG F, et al. TIGR gene indices
clustering tools (TGICL): A software system for fast clustering of large
EST datasets [J]. Bioinformatics, 2003, 19(5): 651–652. doi: 10.1093/
bioinformatics/btg034.
[13] KANEHISA M, ARAKI M, GOTO S, et al. KEGG for linking
genomes to life and the environment [J]. Nucl Acids Res, 2008, 36:
D480–D484. doi: 10.1093/nar/gkm882.，
[14] AUDIC S, CLAVERIE J M. The significance of digital gene
expression profiles [J]. Genome Res, 1997, 7(10): 986–995. doi: 10.
1101/gr.7.10.986.
[15] BENJAMINI Y, YEKUTIELI D. False discovery rate: Adjusted multiple
confidence intervals for selected parameters [J]. J Amer Statist Assoc,
2005, 100(469): 71–81. doi: 10.1198/016214504000001907.
[16] MORTAZAVI A, WILLIAMS B A, McCUE K, et al. Mapping and
quantifying mammalian transcriptomes by RNA-Seq [J]. Nat Methods,
2008, 5(7): 621–628. doi: 10.1038/nmeth.1226.
[17] CONESA A, GÖTZ S, GARCÍA-GÓMEZ J M, et al. Blast2GO: A
universal tool for annotation, visualization and analysis in functional
genomics research [J]. Bioinformatics, 2005, 21(18): 3674–3676. doi:
10. 093/bioinformatics/bti610.
[18] PFAFFL M W. A new mathematical model for relative quatification in
real-time RT-PCR [J]. Nucl Acids Res, 2001, 29(9): e45. doi: 10.1093/
nar/29.9.e45.
[19] LU R K. Chemical Analysis of Soil Agriculture [M]. Beijing: China
Agricultural Science and Technology Press, 2000: 308–311. (in Chinese)
[20] HAO J J, KANG Z L, YU Y. Plant Physiology Experimental
Techniques [M]. Beijing: Chemical Industry Press, 2007: 62–64. (in
Chinese)
[21] ZUO Y M, LIU Y X, ZHANG F S. Effects of improved iron nutrition
of peanut intercropped with maize on carbon and nitrogen metabolism
and nitrogen-fixing of peanut nodul [J]. Acta Ecol Sin, 2004, 24(11):
2584–2590. doi: 10.3321/j.issn:1000-0933.2004.11.034. (in Chinese)
[22] GUO Z L, SHEN A L, KOU C L, et al. The relationship between NRA
and nitrogen efficiency of different wheat varieties after flowering [J].
Chin Agri Sci Bull, 2008, 24(5): 219–223. (in Chinese)
[23] BESSON-BARD A, GRAVOT A, RICHAUD P, et al. Nitric oxide
contributes to cadmium toxicity in Arabidopsis thaliana by promoting
cadmium accumulation in roots and by up-regulating genes related to
iron uptake [J]. Plant Physiol, 2009, 149(3): 1302–1315. doi: 10.1104/
pp.108.133348.
[24] BECKER T W, CARRAYOL E, HIREL B. Glutamine synthetase and
glutamate dehydrogenase isoforms in maize leaves: Localization,
relative proportion and their role in ammonium assimilation or nitrogen
transport [J]. Planta, 2000, 211(6): 800–806. doi: 10.1007/s004250
000355.
[25] NORTON J B, MONACO T A, NORTON U. Mediterranean annual
grasses in western North America: Kids in a candy store [J]. Plant Soil,
2007, 298(1/2): 1–5. doi: 10.1007/s11104-007-9364-8.
[26] ELBERSE I A M, van DAMME J M M, van TIENDEREN P H.
Plasticity of growth characteristics in wild barley (Hordeum spontaneum)
in response to nutrient limitation [J]. J Ecol, 2003, 91(3): 371– 382. doi:
10.1046/j.1365-2745.2003.00776.x.
[27] LI W H, ZHANG C B, LIN J Y, et al. Characteristics of nitrogen
metabolism and soil nitrogen of invasive plants [J]. J Trop Subtrop Bot,
2008, 16(4): 321–327.
[28] MEYER C, STITT M. Nitrate reduction and signaling [M]// LEA P J,
MOROT-GAUDRY J-F. Plant Nitrogen. Berlin Heidelberg: Springer,
2001: 37–59. doi: 10.1007/978-3-662-04064-5_2.
[29] OLIVEIRA H C, JUSTINO G C, SODEK L, et al. Amino acid recovery
does not prevent susceptibility to Pseudomonas syringae in nitrate
reductase double-deficient Arabidopsis thaliana plants [J]. Plant Sci,
2009, 176(1): 105–111. doi: 10.1016/j.plantsci.2008.09.017.
[30] DIAO Z W, YU X, WANG Y B, et al. Study on nitrate reductase and
nitrite reductase coupled regulation of sugar beet [J]. J Nucl Agri Sci,
2014, 28(1): 138–145. doi: 10.11869/j.issn.1000-8551.2014. 01.0138.
142 热带亚热带植物学报 第 24卷


(in Chinese)
[31] KYAING M S, GU L J, CHENG H M. The role of nitrate reductase
and nitrite reductase in plant [J]. Curr Biotechn, 2011, 1(3): 159–164.
[32] MIFLIN B J, HABASH D Z. The role of glutamine synthetase and
glutamate dehydrogenase in nitrogen assimilation and possibilities for
improvement in the nitrogen utilization of crops [J]. J Exp Bot, 2002,
53(370): 979–987.
[33] CAI H M, ZHOU Y, XIAO J H, et al. Overexpressed glutamine
synthetase gene modifies nitrogen metabolism and abiotic stress
reponses in rice [J]. Plant Cell Rep, 2009, 28(3): 527–537. doi: 10.
1007/s00299-008-0665-z.
[34] GULATI A, JAIWAL P K. Effect of NaCl on nitrate reductitase,
glutamate dehydrogenase and glutamate in Vigna radiata calli [J]. Biol
Plant, 1996, 38: 177–183.
[35] ROBIN S A, SLADE A P, FORE G G, et al. The role of glutamate
dehydrogenase in plant nitrogen metabolism [J]. Plant Physiol, 1991,
95(2): 509–516.
[36] PACZEK V, DUBOIS F, SANGWAN R, et al. Cellular and subcellular
localisation of glutamine synthetase and glutamate dehydrogenase in
grapes gives new insights on the regulation of carbon and nitrogen
metabolism [J]. Planta, 2002, 216(2): 245–254. doi: 10.1007/s00425-
002-0854-x.
[37] GAUFIEHON L, REISDORF-CREN M, ROTHSTEIN S J, et al.
Biological functions of asparagine synthetase in plants [J]. Plant Sci,
2010, 179(3): 141–153. doi: 10.1016/j.plantsci.2010.04.010.
[38] BREARS T, LIU C, KNIGHT T J, et al. Ectopic overexpression of
asparagine synthetase in transgenic tobacco [J]. Plant Physiol, 1993,
103(4): 1285–1290.
[39] CAPELL T, CHRISTOU P. Progress in plant metabolic engineering [J].
Curr Opin Biotechn, 2004, 15(2): 148–154. doi: 10.1016/j.copbio.2004.
01.009.
[40] Zhang K C, Jin Q, Can Y P, et al. Research progress of PAL and its
control function of important secondary metabolites [J]. Chin Agri Sci
Bull, 2008, 24(12): 59–62. (in Chinese)
[41] SANMARTIN M, DROGOUDI P A, LYONS T, et al. Over-expression
of ascorbate oxidase in the apoplast of transgenic tobacco results in
altered ascorbate and glutathione redox states and increased sensitivity
to ozone [J]. Planta, 2003, 216(6): 918–928.
[42] XIA J R, HUANG J. Impacts of nitrogen and phosphorus on inorganic
carbon utilization and carbonic anhydrase activity in Nitzschia closterium



f. minutissima [J]. Acta Ecol Sin, 2010, 30(15): 4085–4092. (in Chinese)
[43] RATH M, SALAS J, PARHY B, et al. Identification of genes induced
in proteoid roots of white lupin under nitrogen and phosphorus depri-
vation, with functional characterization of a formamidase [J]. Plant Soil,
2010, 334(1): 137–150. doi: 10.1007/s11104-010-0373-7.
[44] KHALIK K A, OSMAN G, AL-AMOUDI W. Genetic diversity and
taxonomic relationships of some Ipomoea species based on analysis of
RAPD-PCR and SDS-PAGE of seed proteins [J]. Ausr J Crop Sci,
2012, 6(6): 1088–1093.
[45] YOON V, NODWELL J R. Activating secondary metabolism with
stress and chemicals [J]. J Ind Microbiol Biotechn, 2014, 41(2): 415–
424. doi: 10.1007/s10295-013-1387-y.
[46] WANG Y, XIE H, ZHANG Z Q, et al. Relationship between chilling
injury of banana fruit stored at low temperature and PAL activity and
soluble proteins [J]. J Fruit Sci, 2004, 21(2): 149–152. doi: 10.3969/
j.issn.1009-9980.2004.02.013. (in Chinese)
[47] PÁSKA C, INNOCENTI G, KUNVÁRI M, et al. Lignan production by
Ipomoea cairica callus cultures [J]. Phytochemistry, 1999, 52(5): 879–
883. doi: 10.1016/S0031-9422(99)00304-0.
[48] PICKETT J A. Genetic engineering of secondary metabolism for plant
protection [J]. New Biotechn, 2014, 31(Suppl): S55. doi: 10.1016/j.nbt.
2014.05.1737.
[49] THAKUR V K, THAKUR M K. Recent advances in green hydrogels
from lignin: A review [J]. Int J Biol Macromol, 2015, 72: 834–847. doi:
10.1016/j.ijbiomac.2014.09.044.
[50] MA R J, WANG N L, ZHU H, et al. Isolation and identification of
alleiochemicals from invasive plant Ipomoea cairica [J]. Allel J, 2009,
24(1): 77–84.
[51] ERREIRA A A, SILVEIRA D, ALVES R B, et al. Constituents of
Ipomoea cairica ethanolic extract [J]. Chem Nat Comp, 2005, 41(4):
465. doi: 10.1007/s10600-005-0178-8.
[52] FERREIRA A A, AMARAL F A, DUARTE I D G, et al. Antinoci-
ceptive effect from Ipomoea cairica extract [J]. J Ethnopharmac, 2006,
105(1/2): 148–153. doi: 10.1016/j.jep.2005.10.012.
[53] PÁSKA C, INNOCENTI G, FERLIN M, et al. Pinoresinol from
Ipomoea cairica callus cultures [J]. Nat Prod Lett, 2002, 16(5): 359–
363.
[54] AHBIRAMI R, ZUTMA F W, YAHAYA Z S, et al. Oviposition
deterring and oviciding potentials of Ipomoea cairica L. leaf extract
against dengue vectors [J]. Trop Biomed, 2014, 31(3): 456–465.