The AP2/ERF transcription factor SmERF128 positively regulates diterpenoid biosynthesis in Salvia miltiorrhiza

Abstract

Key message
The novel AP2/ERF transcription factor SmERF128 positively regulates diterpenoid tanshinone biosynthesis by activating the expression of SmCPS1, SmKSL1, and SmCYP76AH1 in Salvia miltiorrhiza.

Abstract
Certain members of the APETALA2/ethylene-responsive factor (AP2/ERF) family regulate plant secondary metabolism. Although it is clearly documented that AP2/ERF transcription factors (TFs) are involved in sesquiterpenoid biosynthesis, the regulation of diterpenoid biosynthesis by AP2/ERF TFs remains elusive. Here, we report that the novel AP2/ERF TF SmERF128 positively regulates diterpenoid tanshinone biosynthesis in Salvia miltiorrhiza. Overexpression of SmERF128 increased the expression levels of copalyl diphosphate synthase 1 (SmCPS1), kaurene synthase-like 1 (SmKSL1) and cytochrome P450 monooxygenase 76AH1 (SmCYP76AH1), whereas their expression levels were decreased when SmERF128 was silenced. Accordingly, the content of tanshinone was reduced in SmERF128 RNA interference (RNAi) hairy roots and dramatically increased in SmERF128 overexpression hairy roots, as demonstrated through Ultra Performance Liquid Chromatography (UPLC) and Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS) analysis. Furthermore, SmERF128 activated the expression of SmCPS1, SmKSL1, and SmCYP76AH1 by binding to the GCC box, and to the CRTDREHVCBF2 (CBF2) and RAV1AAT (RAA) motifs within their promoters during in vivo and in vitro assays. Our findings not only reveal the molecular basis of how the AP2/ERF transcription factor SmERF128 regulates diterpenoid biosynthesis, but also provide useful information for improving tanshinone production through genetic engineering.

Full text

Vol.:(0123456789)
1 3
Plant Molecular Biology
https://doi.org/10.1007/s11103-019-00845-7
The AP2/ERF transcription factor SmERF128 positively regulates
diterpenoid biosynthesis inSalvia miltiorrhiza
YuZhang1,2· AijiaJi1,3· ZhichaoXu1· HongmeiLuo1,5· JingyuanSong1,4,5
Received: 15 March 2018 / Accepted: 18 February 2019
© Springer Nature B.V. 2019
Abstract
Key message The novel AP2/ERF transcription factor SmERF128 positively regulates diterpenoid tanshinone bio-
synthesis by activating the expression of SmCPS1, SmKSL1, and SmCYP76AH1 in Salvia miltiorrhiza.
Abstract Certain members of the APETALA2/ethylene-responsive factor (AP2/ERF) family regulate plant secondary
metabolism. Although it is clearly documented that AP2/ERF transcription factors (TFs) are involved in sesquiterpenoid
biosynthesis, the regulation of diterpenoid biosynthesis by AP2/ERF TFs remains elusive. Here, we report that the novel
AP2/ERF TF SmERF128 positively regulates diterpenoid tanshinone biosynthesis in Salvia miltiorrhiza. Overexpres-
sion of SmERF128 increased the expression levels of copalyl diphosphate synthase 1 (SmCPS1), kaurene synthase-like 1
(SmKSL1) and cytochrome P450 monooxygenase 76AH1 (SmCYP76AH1), whereas their expression levels were decreased
when SmERF128 was silenced. Accordingly, the content of tanshinone was reduced in SmERF128 RNA interference (RNAi)
hairy roots and dramatically increased in SmERF128 overexpression hairy roots, as demonstrated through Ultra Performance
Liquid Chromatography (UPLC) and Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS) analysis. Fur-
thermore, SmERF128 activated the expression of SmCPS1, SmKSL1, and SmCYP76AH1 by binding to the GCC box, and
to the CRTDREHVCBF2 (CBF2) and RAV1AAT (RAA) motifs within their promoters during invivo and invitro assays.
Our findings not only reveal the molecular basis of how the AP2/ERF transcription factor SmERF128 regulates diterpenoid
biosynthesis, but also provide useful information for improving tanshinone production through genetic engineering.
Keywords Salvia miltiorrhiza· Diterpenoid· Tanshinones· AP2/ERF transcription factor· Herbgenomics
Abbreviations
AP2/ERF APETALA2/ethylene-responsive factor
CMK 4-(Cytidine 5′-diphospho)-2-C-methyl-D-
erythritol kinase
CPS1 Copalyl diphosphate synthase 1
CYP76AH1 Cytochrome P450 monooxygenase 76AH1
DMAPP Dimethylallyl diphosphate
DXR 1-Deoxy-D-xylulose-5-phosphate
reductoisomerase
DXS2 1-Deoxy-D-xylulose-5-phosphate synthase
2
GFP Green fluorescent protein
GGPP Geranylgeranyl diphosphate
GGPPS Geranylgeranyl diphosphate synthase
GUS β-Glucuronidase
HDR1 4-Hydroxy-3-methylbut-2-enyl diphos-
phatereductase 1
HDS 4-Hydroxy-3-methylbut-2-enyl diphosphate
synthase
IPP Isopentenyl diphosphate
Yu Zhang and Aijia Ji have contributed equally to this work.
Electronic supplementary material The online version of this
article (https ://doi.org/10.1007/s1110 3-019-00845 -7) contains
supplementary material, which is available to authorized users.
* Jingyuan Song
jysong@implad.ac.cn
1 Key Lab ofChinese Medicine Resources Conservation,
State Administration ofTraditional Chinese Medicine
ofthePeople’s Republic ofChina, Institute ofMedicinal
Plant Development, Chinese Academy ofMedical Sciences,
Peking Union Medical College, Beijing100193, China
2 College ofChinese Materia Medica, Shanxi University
ofChinese Medicine, Jinzhong030619, China
3 School ofPharmaceutical Sciences, Guangzhou University
ofChinese Medicine, Guangzhou510006, China
4 Yunnan Branch, Institute ofMedicinal Plant Development,
Chinese Academy ofMedical Sciences, Peking Union
Medical College, Jinghong666100, China
5 Engineering Research Center ofChinese Medicine Resource,
Ministry ofEducation, Beijing100193, China

Plant Molecular Biology
1 3
IPTG Isopropyl β-D-thiogalactoside
KSL1 Kaurene synthase-like 1
MCT 2-C-methyl-D-erythritol 4-phosphate
cytidylyltransferase
MDS 2-C-methyl-D-erythritol 2,4-cyclodiphos-
phate synthase
NADPH Nicotinamide adenine dinucleotide
phosphate
OX Overexpression
RNAi RNA interference
TIA Terpenoidindole alkaloid
TF Transcription factor
Introduction
The APETALA2/ethylene-responsive factor (AP2/ERF)
transcription factor (TF) family, one of the largest families
of TFs in the plant kingdom, is divided into five subfami-
lies: AP2 (APETALA2), ERF (ethylene-responsive fac-
tor), DREB (dehydration-responsive element binding pro-
teins), RAV (related to ABI3/VP1), and Soloist (Sakuma
etal. 2002). It has been reported that members of the AP2/
ERF family are involved in the biosynthesis of various
secondary metabolites, such as terpenoidindole alkaloid
(TIA) and nicotine. In Catharanthus roseus, octadecanoid-
responsive Catharanthus AP2/ERF-domain 3 (ORCA3) and
ORCA4 play key roles in TIA biosynthesis (van der Fits
and Memelink 2000, 2001; Paul etal. 2017). In tobacco,
ERF189, ERF221, ORC1, and NtERF32 are related to
nicotine biosynthesis (Shoji etal. 2010; De Boer etal.
2011; Sears etal. 2014). In addition, AP2/ERF TFs have
been reported to participate in sesquiterpenoid biosynthe-
sis. AaERF1, AaERF2, AaORA, and TRICHOME AND
ARTEMISININ (TAR1) play crucial roles in regulating
artemisinin biosynthesis in Artemisia annua (Yu etal. 2012;
Lu etal. 2013; Tan etal. 2015). Diterpenoids, an impor-
tant class of terpenoids, are widely distributed in nature
(Ma etal. 2012). Transcriptional expression of key enzyme
genes related to taxol biosynthesis in Taxus cuspidata was
upregulated by MeJA. The AP2/ERF TF TcAP2 was induc-
ible by MeJA, however the functional verification of TcAP2
has not been reported (Dai etal. 2009). To date, much less
is known about the AP2/ERF TF regulation of diterpenoid
biosynthesis.
Salvia miltiorrhiza Bunge (Danshen) is a model medici-
nal plant with great medicinal value, and its dried roots
are widely used to treat cardiovascular diseases (Cheng
2006). The main lipophilic bioactive components of S.
miltiorrhiza are diterpenoid tanshinones, including tan-
shinone I, tanshinone IIA, cryptotanshinone, and dihy-
drotanshinone I. More than 40 diterpenoid tanshinones
and structurally related compounds have been isolated and
characterized (Wang etal. 2007), many of which exhibit
strong anti-cancer, antioxidant, anti-atherosclerosis, and
anti-inflammatory activities (Dong etal. 2011; Zhang etal.
2012b; Tao etal. 2013; Chang etal. 2014; Liu etal. 2015).
Because of their unique structural characteristics and bio-
logical activities, tanshinones have attracted widespread
interest. Production and storage of tanshinones occur
in the root periderm of S. miltiorrhiza (Xu etal. 2015).
Many genes encoding key enzymes in tanshinone biosyn-
thesis have been cloned and analyzed, including copalyl
diphosphate synthase 1 (SmCPS1), kaurene synthase-like 1
(SmKSL1), and cytochrome P450 monooxygenase 76AH1
(SmCYP76AH1) (Xu etal. 2016b; Cui etal. 2015). Com-
pared with progress in tanshinone biosynthesis, regula-
tion of tanshinones remains poorly understood, especially
with regard to the regulation of AP2/ERF TF in tanshinone
biosynthesis.
According to the distribution of tanshinone, the expres-
sion patterns of AP2/ERF genes in different organs and
root tissues, and prediction of cis-regulatory elements in
the promoters, two genes (Sm128 and Sm152) related to the
biosynthesis of tanshinone were selected (Ji etal. 2016). In
this study, we cloned and characterized the Sm128 gene that
belongs to the ERF subfamily in S. miltiorrhiza and renamed
it SmERF128. We performed Agrobacterium tumefaciens-
mediated transformation, gene expression analysis, metabo-
lite analysis, subcellular localization, electrophoretic mobil-
ity shift assay (EMSA), microscale thermophoresis (MST),
and transient expression assay to investigate the function and
regulatory mechanism of SmERF128 in tanshinone biosyn-
thesis. Our results indicate that SmERF128 acts as a positive
regulator for diterpenoid tanshinone biosynthesis through
regulation of SmCPS1, SmKSL1, and SmCYP76AH1.
Results
Isolation andsubcellular localization ofSmERF128
The full-length cDNA of SmERF128 was isolated (Ji etal.
2016). SmERF128 (GenBank Accession No. MG897156)
contains an open reading frame (ORF) encoding a protein
of 210 amino acids with an AP2 domain (Supporting Infor-
mation Fig. S1A) that belongs to the ERF-B3 subgroup (Ji
etal. 2016).
To determine the subcellular localization of SmERF128,
a transient expression experiment was performed. We found
that SmERF128-GFP co-localized with the nuclear marker
SmMYC2a-mCherry (Fig.1a), whereas GFP alone localized
to the nucleus and cytoplasm (Fig.1b). The results suggested
that SmERF128 subcellularly localized to the nucleus, in
accordance with its putative role as a transcription factor.

Plant Molecular Biology
1 3
Expression ofcrucial pathway genes isaffected
inRNA interference andoverexpression transgenic
hairy roots
The expression profile of SmERF128 in different tissues in
S. miltiorrhiza was performed in our previous study (Ji etal.
2016). SmERF128 was highly expressed in the periderm and
co-expressed with terpenoid synthase genes (SmCPS1 and
SmDXS2). Therefore, we predicted that it could be related
to tanshinone biosynthesis. To further prove the transcrip-
tion regulatory function of SmERF128, the relative tran-
script levels of SmERF128 and enzyme genes involved in
tanshinone biosynthesis were determined by qRT-PCR in
three independent control lines and SmERF128 transgenic
S. miltiorrhiza hairy roots (Fig.2). All experiments were
performed with three biological replicates. Each data point
is the average of three independent biological repeats. We
observed that the expression levels of key tanshinone path-
way genes in three independent SmERF128 overexpres-
sion (SmERF128-OX) transgenic lines were significantly
increased compared with the control levels, including
SmCPS1, SmKSL1, and SmCYP76AH1 (Fig.2a). Coinci-
dently, their expression was significantly repressed in three
independent SmERF128 RNA interference (SmERF128-
RNAi) transgenic lines (Fig.2b). Our findings indicated that
SmERF128 probably regulated the expression of SmCPS1,
SmKSL1, and SmCYP76AH1, and then controlled the bio-
synthesis of tanshinone.
Overexpression ofSmERF128 increases tanshinone
yield andsilencing ofSmERF128 reduces tanshinone
production inS. miltiorrhiza
The SmERF128-OX hairy roots and their extracts appeared
much redder than the control lines (Fig.3a, b). In contrast,
the control lines and their extracts appeared much redder
than the SmERF128-RNAi hairy roots (Fig.4a, b). These
findings indicated that the content of tanshinone was reduced
in SmERF128-RNAi hairy roots and dramatically increased
in SmERF128-OX hairy roots. To test this hypothesis, the
contents of four major tanshinones (tanshinone I, tanshi-
none IIA, dihydrotanshinone I and cryptotanshinone) in
SmERF128 transgenic S. miltiorrhiza hairy roots were
measured by ultra-performance liquid chromatography
(UPLC) (Supporting Information Fig. S2). In comparison
with the control, the contents of tanshinone I, tanshinone
IIA, dihydrotanshinone I, and cryptotanshinone were mark-
edly increased in the SmERF128-OX lines (Fig.3c–e). Con-
versely, their levels were decreased in SmERF128-RNAi
lines (Fig.4c–e).
To further investigate the effect of SmERF128 on tan-
shinone production, an liquid chromatography–tandem
mass spectrometry (LC–MS) system was used to detect the
various metabolites. There were 370 metabolites in total,
23 of which were tanshinones. All of these metabolites
were identified by mass spectrometry. Overexpression of
SmERF128 increased the production of 23 tanshinones, and
20 of the metabolites were markedly increased (VIP > 1; fold
change ≥ 2; Supporting Information TableS1), including
tanshinone I, tanshinone IIA, dihydrotanshinone I, crypto-
tanshinone, miltirone, tanshinol B, methylenetanshinqui-
none, danshenxinkun A, tanshinone VI, 1R-hydroxymilti-
rone, hydroxytanshinone IIA, 1-Ketoisocryptotanshinone,
tanshindiolB, tanshindiol C, salvisyrianone, cryptojaponol,
methyltanshinonate, trijuganone C, tanshinone IIB, and
1,2-didehydrotanshinone V. In contrast, RNA interference
(RNAi) of SmERF128 in hairy roots decreased accumula-
tion of 21 tanshinones, and 10 tanshinones (sibiriquinone
A, danshenxinkun A, hydroxytanshinone IIA, tanshindiol
B, salvisyrianone, cryptojaponol, methyltanshinonate, tri-
juganone C, trijuganone B, and 1, 2-didehydrotanshinone
V) were dramatically decreased (VIP > 1; fold change ≤ 0.5;
Fig. 1 Subcellular localiza-
tion of SmERF128-GFP in
Arabidopsis protoplasts. a
SmERF128-GFP; b GFP.
Bright bright-field, GFP green
fluorescent protein, mCherry
fluorescence of nuclear marker
SmMYC2a-mCherry fusion
protein, Merged merged image
of GFP, mCherry, and bright.
Bars represent 10μm

Plant Molecular Biology
1 3
Supporting Information TableS2) compared with control
lines. An increase in tanshinone VI was observed in the
SmERF128 knockdown lines compared to the control lines.
These data revealed that SmERF128 contributed to the con-
tent of tanshinones in S. miltiorrhiza.
SmERF128 protein binds totheGCC, CBF2, andRAA
motifs
Some ERF subfamily proteins have been proven to spe-
cifically bind to the GCC box (AGC CGC C), and to the
CRTAREHVCBF2 (CBF2), and RAV1AAT (RAA)
motifs (Yu etal. 2012; Zhu etal. 2014; Tan etal. 2015).
To explore the regulatory mechanism of SmERF128 in tan-
shinone biosynthesis, the binding activities of SmERF128
to the SmCPS1, SmKSL1, and SmCYP76AH1 promoters
were examined. The GCC box and CBF2 and RAA motifs
were observed in SmCPS1, SmKSL1, and SmCYP76AH1
promoters (Supporting Information Fig. S3). The purified
His-SmERF128 protein (Supporting Information Fig. S4)
was mixed with the labeled probe or mutated probe in the
binding reaction (Fig.5a–d). The EMSA results showed that
the gel mobility shift was specific to the His-SmERF128
protein with the labeled 3 × GCC, 3 × CBF2 or 3 × RAA
probes, and no shifted bands were observed in the His-
SmERF128 protein with the labeled 3 × GCC, 3 × CBF2 or
3 × RAA mutation (Fig.5a–d). No complex was detected in
the labeled probes without His-SmERF128. These findings
strongly suggested that SmERF128 specifically bound to the
GCC box, and to the CBF2 and RAA motifs, which was
consistent with the expectation of AP2/ERF TFs.
To further examine the binding activities of the GCC box
and the CBF2 and RAA motifs, MST was performed. We
found that SmERF128 exhibited binding affinities to GCC
(KD = 13.89 ± 3.30µM), CBF2 (KD = 57.68 ± 0.22µM), and
RAA (KD = 199 ± 102.83µM) (Fig. 5e–g). In agreement
Fig. 2 Gene expression of the tanshinone biosynthetic pathway in
SmERF128 transgenic S. miltiorrhiza hairy roots. a qRT–PCR anal-
ysis of tanshinone biosynthetic genes in three independent PKOX
and three independent SmERF128 overexpression hairy root lines. b
qRT–PCR analysis of tanshinone biosynthetic genes in three inde-
pendent PKRNAi and three independent SmERF128-RNAi hairy
roots. SmActin was used as a control for normalization. Each data
point is the average of three independent biological repeats. Bars
indicate SD. Level of significance obtained with one-way analysis of
variance marked by the following: *P < 0.05 (one-way ANOVA)

Plant Molecular Biology
1 3
Fig. 3 Contents of dihydrotanshinone I, tanshinone I, cryptotanshi-
none, and tanshinone IIA in SmERF128 overexpression and control
hairy roots of S. miltiorrhiza. a Phenotypes of SmERF128-OX lines
compared with empty vector control line. b Tanshinones extracted
from hairy roots of SmERF128-OX lines compared with empty vec-
tor control line. c–e UPLC analysis of dihydrotanshinone I, tanshi-
none I, cryptotanshinone, and tanshinone IIA of three independent
SmERF128-OX and control lines of S. miltiorrhiza. Orange arrow
dihydrotanshinone I, black arrow tanshinone I, red arrow cryptotan-
shinone, green arrow tanshinone IIA
Fig. 4 Contents of dihydrotanshinone I, tanshinone I, cryptotanshi-
none, and tanshinone IIA in SmERF128 silenced and control hairy
roots of S. miltiorrhiza. a The phenotypes of SmERF128-RNAi lines
compared with empty vector control lines. b Tanshinones extracted
from hairy roots of SmERF128-RNAi lines compared with empty vec-
tor control lines. c–e UPLC analysis of dihydrotanshinone I, tanshi-
none I, cryptotanshinone, and tanshinone IIA of three independent
SmERF128-RNAi and control lines of S. miltiorrhiza. Orange arrows
dihydrotanshinone I, black arrows tanshinone I, red arrows crypto-
tanshinone, green arrows tanshinone IIA

Plant Molecular Biology
1 3
with the results of EMSA, these observations strongly sug-
gested that SmERF128 could bind to the GCC box and to
the CBF2 and RAA motifs within SmCPS1, SmKSL1, and
SmCYP76AH1 promoters.
SmERF128 protein activates promoter activity
To test whether SmERF128 could activate promoters of
SmCPS1, SmKSL1, and SmCYP76AH1, each promoter-
GUS construct was co-transferred with 35S:SmERF128
into tobacco leaves. Each promoter-GUS was also co-
transformed with empty vector (pK7WG2D) into tobacco
leaves and used as a control. Compared to the control, the
tobacco leaves had obvious blue staining in 35S:SmERF128
transiently transformed N. benthamiana (Supporting Infor-
mation Fig. S5). The results of GUS staining showed that
SmERF128 activated the promoters of SmCPS1, SmKSL1,
and SmCYP76AH1 invivo.
To provide more evidence that SmERF128 regulates the
transcription of SmCPS1, SmKSL1, and SmCYP76AH1,
qRT-RCR analysis of the GUS gene expression in
SmERF128 transiently transformed N. benthamiana leaves
was performed. The transcript levels of GUS reporter were
significantly increased, ranging from 1.6- to 2.3-fold com-
pared with the control (Fig.6a–c). The transiently overex-
pressed SmERF128 showed a stronger ability to activate the
promoters of SmCPS1, SmKSL1, and SmCYP76AH1.
Discussion
With rapid advances in high throughput sequencing technol-
ogies and greatly reduced costs, herbgenomics has emerged
(Chen etal. 2015). Herbgenomics provides a foundation for
investigating the biosynthesis of plant secondary metabolites
and their regulation (Xin etal. 2018). Recently, genomic
and transcriptomic studies of S. miltiorrhiza have progressed
rapidly (Xu etal. 2015, 2016a). Based on genomic infor-
mation and transcriptomic data, the bHLH, AP2/ERF, and
bZIP gene families in S. miltiorrhiza were predicted (Zhang
etal. 2015, 2018; Ji etal. 2016). This progress has facilitated
study on regulation of diterpenoid tanshinone biosynthesis
in S. miltiorrhiza.
Fig. 5 Binding assay of SmERF128 to GCC-box, and to CBF2 and
RAA cis-elements. a–c EMSA binding analysis of SmERF128 pro-
tein to triple GCC boxes, and CBF2 and RAA using the Light Shift
Chemiluminescent EMSA Kit. Lane 1 negative control contain-
ing only biotin-labeled probe, lane 2 His-SmERF128 fusion protein
plus biotin-labeled probe, lane 3 His-SmERF128 fusion protein plus
mutated probe. Ten microgram of His-SmERF128 fusion protein was
incubated with 2µg of biotin-labeled probe or mutation at 25 °C for
30 min. Red arrow shifted band, blue arrow free probe. d Specific
binding of SmERF128 to triple GCC boxes, and CBF2 and RAA
using AidGreen Nucleic Acid Gel Stain. e–g MST measurements of
the binding affinity of SmERF128 for GCC box (e), and CBF2 (f)
and RAA motifs (g). Three independent repeats were performed. The
resulting binding curve from plotting the FNorm (percentage) versus
concentration was fit using a hyperbolic function to yield a KD of
13.89 ± 3.30µM for GCC box (e), 57.68 ± 0.22 µM for CBF2 motif
(f) and 199 ± 102.89 for RAA motif (g)

Plant Molecular Biology
1 3
SmERF128 regulates theexpression ofditerpenoid
pathway genes
It has been reported that members of the AP2/ERF fam-
ily control the expression of sesquiterpenoid pathway genes
(Yu etal. 2012; Tan etal. 2015). In this study, we found an
AP2/ERF TF that participates in transcriptional regulation
of diterpenoid pathway genes. SmERF128 functions as a
positive regulator in diterpenoid tanshinone biosynthesis
in S. miltiorrhiza. The EMSA, MST, and transient expres-
sion assay indicated that SmERF128 directly bound to the
GCC box and to the CBF2 and RAA motifs within SmCPS1,
SmKSL1, and SmCYP76AH1 promoters and activated their
expression. Previous studies have found that AaERF1,
AaERF2, and TAR1 regulated sesquiterpenoid biosynthesis
in A. annua, and they were able to interact with the CBF2
and RAA motifs present in both Amorpha-4, 11-diene syn-
thase (ADS), and CYP71AV1 promoters, as demonstrated
through EMSA, yeast one-hybrid assay, and transient expres-
sion assay (Yu etal. 2012; Tan etal. 2015). However, much
less is known about the differences in binding activity of
AP2/ERF TFs to its binding sites. Interestingly, in our inves-
tigation, data from the MST assay showed that SmERF128
had higher affinity to the GCC box than to the CBF2 and
RAA motifs (Fig.5e–g). AP2/ERF TFs regulated the expres-
sion of diterpenoid biosynthetic genes and sesquiterpenoid
biosynthetic genes. In addition, ERF189, which belongs to
the AP2/ERF TF family, recognized a GCC-box element in
the promoter of genes involved in nicotine biosynthesis and
increased their expression in tobacco (Shoji etal. 2010).
All of the above results imply that many genes involved
in diverse secondary metabolic pathways contain binding
sites of AP2/ERF TF. Therefore, AP2/ERF TFs could play
an important role in the biosynthesis of various secondary
metabolites.
SmERF128 controls transcription ofthediterpene
synthases andterpenoid‑modifying enzyme genes
In this study, the function of SmERF128 in regulating the
tanshinone biosynthetic pathway was identified using trans-
genic approaches of overexpression and RNAi. Further-
more, we found that SmERF128 directly interacted with the
GCC box and the CBF2 and RAA motifs within SmCPS1,
SmKSL1, and SmCYP76AH1 promoters invitro, and acti-
vated their promoters invivo. SmKSL1 and SmCPS1, diter-
pene synthases, and SmCYP76AH1, a terpenoid-modifying
enzyme, are three key enzymes of the tanshinone biosyn-
thetic pathway (Supporting Information Fig. S6). Previ-
ous studies have described four TFs related to tanshinone
biosynthesis in S. miltiorrhiza by overexpression or RNAi
approaches, including SmMYB36, SmMYB9b, SmMYC2a,
and SmMYC2b (Zhou etal. 2016; Ding etal. 2017; Zhang
etal. 2017). Among these, SmMYB36 was identified to
bind to the promoter of geranylgeranyl diphosphate synthase
(GGPPS) based on an EMSA experiment, while the regula-
tory mechanism of the other TFs remains unclear. GGPPS is
a terpenoid synthase upstream of the terpenoid biosynthetic
pathway (Supporting Information Fig. S6), suggesting that
SmMYB36 is probably associated with the biosynthesis
of other diterpenoids by activating the GGPPS promoter.
Therefore, TFs could regulate tanshinone biosynthesis by
Fig. 6 GUS gene expression in transiently transformed N. bentha-
miana leaves. SmERF128 activated GUS gene expression in SmCP-
S1pro: GUS (a), and SmKSL1pro: GUS (b) and SmCYP76AH1pro:
GUS (c) reporters. Actin of N. benthamiana was used as the internal
standard. Each data point is the average of three independent biologi-
cal repeats. Error bars indicate SD. Student’s t test: *P < 0.05

Plant Molecular Biology
1 3
interacting with the diterpene synthases and terpenoid-mod-
ifying enzymes, as well as the terpenoid synthases involved
upstream of the terpenoid biosynthetic pathway. This study
provides the foundation for a better understanding of the
regulation of diterpenoid biosynthesis in plants.
Self‑regulation ofSmERF128
It was reported that the bHLH transcription factor mediated
secondary metabolism by binding to the promoter region of
AP2/ERF transcription. For instance, in C. roseus, CrMYC2
was capable of activating ORCA3 and co-regulating TIA
pathway genes concomitantly with ORCA3 (Zhang etal.
2011). Similarly, in tobacco, NtMYC2 positively regulated
nicotine biosynthesis not only by binding to the G-box in the
promoters of biosynthetic genes, but also by activating the
NIC2-locus ERF genes (Shoji and Hashimoto 2011; Zhang
etal. 2012a). Interestingly, W-box and TGACG motif were
both present in the promoter region of SmERF128 in this
study (Supporting Information Fig. S1B). Members of the
WRKY family and bZIP family were reported to interact
with the W-box and TGACG motif, respectively (Adachi
etal. 2015; IdrovoEspin etal. 2012). It will be interesting to
investigate whether SmERF128 is regulated by the WRKY
or bZIP transcription factor in tanshinone biosynthesis in
S. miltiorrhiza. Additionally, protein modifications play
an important role in the regulation of several TFs. Protein
phosphorylation is the best-studied example in plants. For
instance, the WRKY TFs phosphorylated by mitogen-
activated protein kinase (MAPK) regulate a plant immune
NADPH oxidase in N. benthamiana (Adachi etal. 2015).
More knowledge of protein interactions and modifications of
SmERF128 involved in tanshinone biosynthesis will further
our understanding of their regulation. Moreover, elucidating
the regulation of tanshinone biosynthesis will provide more
modules for synthetic biology (Naseri etal. 2017).
In conclusion, a novel AP2/ERF TF SmERF128 posi-
tively regulates tanshinone biosynthesis by activating the
diterpene synthases and terpenoid-modifying enzyme genes
involved in diterpenoid biosynthesis in S. miltiorrhiza. S.
miltiorrhiza provides a useful model system that has facili-
tated study of the regulation of diterpenoid biosynthesis in
plants.
Materials andmethods
Plant materials
Salvia miltiorrhiza (line 99–3) plants were obtained from the
Institute of Medicinal Plant Development (IMPLAD), Chi-
nese Academy of Medical Sciences (CAMS). For transient
expression assay, tobacco (Nicotiana benthamiana) seeds
were sown directly on the soil and grown in a greenhouse
with a 16-h light/8-h dark cycle. A 4-week-old Arabidopsis
plant was suitable for protoplast isolation. The optimal true
leaves (fifth, sixth, and seventh) were used as a mesophyll
protoplast source.
Cloning oftheSmERF128 gene
Total RNA was extracted from roots of S. miltiorrhiza using
RNAprep Pure Plant Kit (Tiangen Biotech, Beijing, China),
followed by reverse transcription using PrimeScript™
Reverse Transcriptase (Takara, Dalian, China). Com-
plete CDS of SmERF128 were amplified with the primers
SmERF128-PF and SmERF128-PR according to the manu-
facturer’s instructions for Takara Ex Taq DNA polymerase
(Ji etal. 2016). The PCR program was performed using the
following conditions: 95°C for 30s, followed by 30 cycles
of amplification (5s denaturation at 95°C, 34s annealing
at 58°C, 2min of extension at 72°C), and a final extension
at 72°C for 10min. The PCR product was ligated to the
Takara PMD 18-T vector and the sequence was confirmed
by Sanger sequencing. All primers are listed in Supporting
Information TableS3.
Construction ofSmERF128 RNAi andSmERF128
overexpression (OX) vector andplant
transformation
The 180-bp fragment of SmERF128 (loci: 89–268bp) was
amplified for RNAi constructs, and the full-length coding
region of SmERF128 was amplified for overexpression con-
structs (for primers, see Supporting Information TableS3).
The PCR products were inserted into the pDONR221 entry
vector using BP reaction and introduced into the destination
vector pK7GWIWG2D (II) or pK7WG2D using LR reaction
as described previously (Xu and Song 2017).
To generate transgenic hairy roots, pK7GWIWG2D (II)-
GFP-SmERF128i and pK7WG2D-GFP-SmERF128OX con-
structs were transformed into wild-type S. miltiorrhiza via
Agrobacterium rhizogenes strain ACCC10060 transforma-
tion. S. miltiorrhiza transformed with pK7GWIWG2D (II) or
pK7WG2D empty vector (PKRNAi or PKOX) was used as
a control, respectively. The transgenic hairy root lines were
cultured according to previously described methods (Xu and
Song 2017). The three independent lines were tested in these
experiments for gene expression and metabolite analyses.
Quantitative RT‑PCR expression analyses
The relative expression levels of general MEP pathway
enzyme genes were tested in control and transgenic
S. miltiorrhiza hairy root lines, including 1-Deoxy-
D-xylulose-5-phosphate synthase 2 (SmDXS2),

Plant Molecular Biology
1 3
1-Deoxy-D-xylulose-5-phosphate reductoisomerase
(SmDXR), 2-C-methyl-D-erythritol 4-phosphate cytidy-
lyltransferase (SmMCT), 4-(Cytidine 5′-diphospho)-2-C-
methyl-D-erythritol kinase (SmCMK), 2-C-methyl-D-
erythritol 2,4-cyclodiphosphate synthase (SmMDS),
4-hydroxy-3-methylbut-2-enyl diphosphate synthase
(SmHDS), 4-hydroxy-3-methylbut-2-enyl diphosphate
reductase 1 (SmHDR1), as well as tanshinone pathway
genes SmCPS1, SmKSL1, SmCYP76AH1 (Zhang etal.
2018). First-strand cDNA was synthesized with 1µg total
RNAs from these samples. Quantitative real-time RT-PCR
(qRT-PCR) was performed with SYBR Premix Ex Taq™
(Takara, Dalian, China). The transcript level of SmACTIN
was used as an internal control. All qRT-PCR primers are
listed in Supporting Information TableS3. All experiments
were performed with three biological replicates.
Extraction ofcompounds andmetabolite analysis
Hairy roots from the 2-month-old S. miltiorrhiza were
dried at 45°C, and then 1g of powder was extracted
with 5mL methanol under ultrasound for 30min. The
extractions were centrifuged at 12,000rpm for 10min
and the supernatant was filtered through a 0.2µm organic
membrane for ultra-performance liquid chromatography
(UPLC) analysis. The UPLC analysis and LC–MS/MS
analysis were performed as previously described (Xu and
Song 2017). The differences of metabolite contents were
statistically significant (VIP > 1). All standards (tanshi-
none I, tanshinone IIA, dihydrotanshinone I, and cryp-
totanshinone) were purchased from Sigma-Aldrich (St.
Louis, MO, USA), and determined by mass spectrometry.
Subcellular localization ofSmERF128
For the subcellular localization assay, the open reading
frame (ORF) of SmERF128 was cloned into the tran-
sient expression vector pBI221. As a positive marker,
the cDNA of a previously characterized nuclear protein,
SmMYC2a, was fused to the mCherry gene to generate
35S: SmMYC2a-mCherry (Xu etal. 2014; Zhou et al.
2016). After cell walls were removed using cellulase and
macerozyme, mesophyll protoplasts were isolated from
leaves of Arabidopsis plants (Yoo etal. 2007). The two
plasmids were co-transformed into Arabidopsis protoplasts
and incubated in the dark at 28°C for 16h (Yoo etal.
2007). Arabidopsis protoplasts were co-transformed by
the empty pBI221 vector and 35S:SmMYC2a-mCherry,
used as a control. The fluorescence was observed using a
microscopy assay with a LSM710 laser scanning confocal
microscope (Zeiss).
Protein recombination andpurification
The full-length cDNA of SmERF128 was inserted into
pCold I vector (Takara, Dalian, China) using EcoRI and
XhoI restriction sites, and the recombinant plasmid pCold
I-SmERF128 was introduced into Escherichia coli strain
BL21. Transformants were cultured in LB medium to an
optical density at 600nm (OD 600) of 0.4–0.5 at 37°C.
Afterwards, target protein was expressed by adding 0.5mM
IPTG into LB medium (containing appropriate selection
antibiotics) for 24h at 15°C. The target protein was purified
using Ni–NTA agarose (Invitrogen), and quantified using the
Bradford assay (2-D Quant Kit; Amersham Biosciences, San
Francisco, CA, USA).
Electrophoretic mobility shift assay (EMSA)
To detect protein-nucleic acid interactions invitro, EMSA
was performed according to the manufacturer’s protocol.
The probe sequences, which contained triple tandem cop-
ies of GCC box, and CBF2, and RAA motifs, were synthe-
sized by Genewiz (Supporting Information TableS3). The
GCC box, CBF2, and RAA mutation were used as controls
(Supporting Information TableS3). The purified recombi-
nant protein His-SmERF128 and 3′ biotin-labeled probes or
mutated probes were incubated in binding buffer (10 × bind-
ing buffer, 50% glycerol, 1µg/µL poly (dI × dC), 100mM
MgCl2, 1M KCl, 20mM DTT, 1% NP-40 and 50% glycerol)
at room temperature for 20min, and then separated using
6% native polyacrylamide gels. Subsequently, the probe was
detected using AidGreen Nucleic Acid Gel Stain. Besides,
binding reactions were transferred to a Soak Amersham
HybondTM-N+ nylon membrane (GE Healthcare, USA), and
the fluorescence was detected using the Light Shift Chemilu-
minescent EMSA Kit (Pierce, Rockford, IL, USA).
Microscale thermophoresis (MST)
MST was performed to study protein–nucleic acid interac-
tion using purified His-SmERF128 protein and 3′ FAM-
labeled nucleic acid including triple tandem copies of the
GCC box, and the CBF2 and RAA motifs, which were syn-
thesized by Genewiz (Supporting Information TableS3). A
series of purified SmERF128 protein solutions with differ-
ent concentrations were prepared by dilutions in Phosphate
Buffer Saline (PBS). The labeled GCC box, and CBF2 and
RAA motifs were mixed with a series of His-SmERF128
solutions. Subsequently, the samples were loaded into silica
capillaries after incubation at room temperature. Binding
assays were performed with a Monolith NT.115 Micro-
scale Thermophoresis device using 40% LED power and

Plant Molecular Biology
1 3
red LED color (Zheng etal. 2015). Data analyses were
performed using the NT Analysis software (NanoTemper
Technologies).
Transient expression assay
With XbaI and SmaI sites, the 1021-bp fragment of CPS1
promoter, 481-bp fragment of KSL1 promoter, and 961-
bp fragment of CYP76AH1 promoter were fused with the
β-glucuronidase (GUS) reporter gene and cloned into the
binary vector pBI121. Each pro: GUS and SmERF128 over-
expression plasmid were co-introduced into tobacco (N.
benthamiana) leaves by Agrobacterium-mediated transfor-
mation. Each pro: GUS was co-transformed with pK7WG2D
empty vector into tobacco leaves, used as a control. The
transient transformation was performed as previously
described (Yu etal. 2012). The infiltrated plants were grown
in a greenhouse for an additional 3days, and histochemical
staining for GUS activity and qRT-PCR analysis of the GUS
gene expression in the infiltrated leaves were performed. The
experiments were repeated in triplicate.
Acknowledgements This work was supported by the National Nat-
ural Science Foundation of China (Grant No. 81573398), CAMS
Innovation Fund for Medical Sciences (CIFMS) (2017-I2M-1-
009) and Foundation of Educational Department of Guangdong
Province(E1-KFD015181K31).
Author contributions JS designed the study. YZ and AJ performed
experiments. YZ, AJ, ZX and HL analyzed the data. YZ, AJ, JS, and
ZX wrote the manuscript. All authors approved the final manuscript.
Compliance with ethical standards
Conflict of interest All the authors declare that they have no conflict
of interests.
References
Adachi H, Nakano T, Miyagawa N etal (2015) WRKY transcrip-
tion factors phosphorylated by MAPK regulate a plant immune
NADPH oxidase in Nicotiana benthamiana. Plant Cell
27:2645–2663
Chang CC, Chu CF, Wang CN etal (2014) The anti-atherosclerotic
effect of tanshinone IIA is associated with the inhibition of TNF-
α-induced VCAM-1, ICAM-1 and CX3CL1 expression. Phytomedi-
cine 21:207–216
Chen S, Song J, Sun C, Xu J, Zhu Y, Verpoorte R, Fan T (2015) Herbal
genomics: examining the biology of traditional medicines. Sci-
ence 347:S27–S29
Cheng TO (2006) Danshen: a popular Chinese cardiac herbal drug. J
Am Coll Cardiol 47:1498
Cui G, Duan L, Jin B etal (2015) Functional divergence of diterpene
syntheses in the medicinal plant Salvia miltiorrhiza. Plant Physiol
169:1607–1618
Dai Y, Qin Q, Dai D, Kong L, Li W, Zha X, Jin Y, Tang K (2009)
Isolation and characterization of a novel cDNA encoding methyl
jasmonate-responsive transcription factor TcAP2 from Taxus cus-
pidata. Biotechnol Lett 31:1801–1809
De Boer K, Tilleman S, Pauwels L etal (2011) APETALA2/ETHYL-
ENE RESPONSE FACTOR and basic helix-loop-helix tobacco
transcription factors cooperatively mediate jasmonate-elicited
nicotine biosynthesis. Plant J 66:1053–1065
Ding K, Pei T, Bai Z, Jia Y, Ma P, Liang Z (2017) SmMYB36, a Novel
R2R3-MYB transcription factor, enhances tanshinone accumula-
tion and decreases phenolic acid content in Salvia miltiorrhiza
hairy roots. Sci Rep 7:5104
Dong Y, Morris-Natschke SL, Lee KH (2011) Biosynthesis, total syn-
theses, and antitumor activity of tanshinones and their analogs as
potential therapeutic agents. Nat Prod Rep 28:529–542
IdrovoEspin FM, Peraza-Echeverria S, Fuentes G, Santamaria JM
(2012) In silico cloning and characterization of the TGA (TGACG
MOTIF-BINDING FACTOR) transcription factors subfamily in
Carica papaya. Plant Physiol Biochem 54:113–122
Ji A, Luo H, Xu Z etal (2016) Genome-wide identification of the AP2/
ERF gene family involved in active constituent biosynthesis in
Salvia miltiorrhiza. Plant Genome 9(2):1–11
Liu X, Guo CY, Ma XJ etal (2015) Anti-inflammatory effects of tan-
shinone IIA on atherosclerostic vessels of ovariectomized ApoE
mice are mediated by estrogen receptor activation and through
the ERK signaling pathway. Cell Physiol Biochem 35:1744–1755
Lu X, Zhang L, Zhang F etal (2013) AaORA, a trichome-specific
AP2/ERF transcription factor of Artemisia annua, is a positive
regulator in the artemisinin biosynthetic pathway and in disease
resistance to Botrytis cinerea. New Phytol 198:1191–1202
Ma Y, Yuan L, Wu B, Li X, Chen S, Lu S (2012) Genome-wide identi-
fication and characterization of novel genes involved in terpenoid
biosynthesis in Salvia miltiorrhiza. J Exp Bot 63:2809–2823
Naseri G, Balazadeh S, Machens F, Kamranfar I, Messerschmidt K,
Mueller-Roeber B (2017) Plant-derived transcription factors for
orthologous regulation of gene expression in the yeast saccharo-
myces cerevisiae. Acs Synth Biol 6:1742–1756
Paul P, Singh SK, Patra B, Sui X, Pattanaik S, Yuan L (2017) A dif-
ferentially regulated AP2/ERF transcription factor gene cluster
acts downstream of a MAP kinase cascade to modulate terpenoid
indole alkaloid biosynthesis in Catharanthus roseus. New Phytol
213:1107–1123
Sakuma Y, Liu Q, Dubouzet JG, Abe H, Shinozaki K, Yamaguchi-
Shinozaki K (2002) DNA-binding specificity of the ERF/AP2
domain of Arabidopsis DREBs, transcription factors involved in
dehydration- and cold-inducible gene expression. Biochem Bio-
phys Res Commun 290:998–1009
Sears MT, Zhang H, Rushton PJ etal (2014) NtERF32: a non-NIC2
locus AP2/ERF transcription factor required in jasmonate-induc-
ible nicotine biosynthesis in tobacco. Plant Mol Biol 84:49–66
Shoji T, Hashimoto T (2011) Tobacco MYC2 regulates jasmonate-
inducible nicotine biosynthesis genes directly and by way of the
NIC2-locus ERF genes. Plant Cell Physiol 52:1117–1130
Shoji T, Kajikawa M, Hashimoto T (2010) Clustered transcription fac-
tor genes regulate nicotine biosynthesis in tobacco. Plant Cell
22:3390–3409
Tan H, Xiao L, Gao S etal (2015) TRICHOME AND ARTEMISININ
REGULATOR 1 is required for trichome development and arte-
misinin biosynthesis in Artemisia annua. Mol Plant 8:1396–1411
Tao S, Zheng Y, Lau A etal (2013) Tanshinone I activates the Nrf2-
dependent antioxidant response and protects against as (III)-
induced lung inflammation invitro and invivo. Antioxid Redox
Signal 19:1647–1661
Van der Fits L, Memelink J (2000) ORCA3, a jasmonate-responsive
transcriptional regulator of plant primary and secondary metabo-
lism. Science 289:295–297
Van der Fits L, Memelink J (2001) The jasmonate-inducible AP2/ERF
domain transcription factor ORCA3 activates gene expression via

Plant Molecular Biology
1 3
interaction with a jasmonate-responsive promoter element. Plant
J 25:43–53
Wang X, Morris-Natschke SL, Lee KH (2007) New developments in
the chemistry and biology of the bioactive constituents of Tan-
shen. Med Res Rev 27:133–148
Xin T, Zhang Y, Pu X, Gao R, Xu Z, Song J (2018) Trends in herb-
genomics. Sci China Life Sci. https ://doi.org/10.1007/s1142
7-018-9352-7
Xu Z, Song J (2017) The 2-oxoglutarate-dependent dioxygenase super-
family participates in tanshinone production in Salvia miltior-
rhiza. J Exp Bot 68:2299–2308
Xu Y, Wang Y, Long Q etal (2014) Overexpression of OsZHD1, a zinc
finger homeodomain class homeobox transcription factor, induces
abaxially curled and drooping leaf in rice. Planta 239:803–816
Xu Z, Peters RJ, Weirather J etal (2015) Full-length transcriptome
sequences and splice variants obtained by a combination of
sequencing platforms applied to different root tissues of Salvia
miltiorrhiza and tanshinone biosynthesis. Plant J 82:951–961
Xu H, Song J, Luo H etal (2016a) Analysis of the genome sequence
of the medicinal plant Salvia miltiorrhiza. Mol Plant 9:949–952
Xu Z, Ji A, Zhang X, Song J, Chen S (2016b) Biosynthesis and regula-
tion of active compounds in medicinal model plant Salvia miltior-
rhiza. Chin Herb Med 8:3–11
Yoo SD, Cho YH, Sheen J (2007) Arabidopsis mesophyll protoplasts:
a versatile cell system for transient gene expression analysis. Nat
Protoc 2:1565–1572
Yu ZX, Li JX, Yang CQ, Hu WL, Wang LJ, Chen XY (2012) The
jasmonate-responsive AP2/ERF transcription factors AaERF1 and
AaERF2 positively regulate artemisinin biosynthesis in Artemisia
annua L. Mol Plant 5:353–365
Zhang H, Hedhili S, Montiel G, Zhang Y, Chatel G, Pre M, Gantet P,
Memelink J (2011) The basichelix-loop-helix transcription fac-
tor CrMYC2 controls the jasmonate-responsive expression of the
ORCA genes that regulate alkaloid biosynthesis in Catharanthus
roseus. Plant J 67:61–71
Zhang HB, Bokowiec MT, Rushton PJ, Han SC, Timko MP (2012a)
Tobacco transcription factors NtMYC2a and NtMYC2b form
nuclear complexes with the NtJAZ1 repressor and regulate mul-
tiple jasmonate-inducible steps in nicotine biosynthesis. Mol Plant
5:73–84
Zhang Y, Jiang P, Ye M etal (2012b) Tanshinones: sources, pharma-
cokinetics and anti-cancer activities. In J Mol Sci 13:13621–13666
Zhang X, Luo H, Xu Z, Zhu Y, Ji A, Song J, Chen S (2015) Genome-
wide characterisation and analysis of bHLH transcription factors
related to tanshinone biosynthesis in Salvia miltiorrhiza. Sci Rep
5:11244
Zhang J, Zhou L, Zheng X, Zhang J, Yang L, Tan R, Zhao S (2017)
Overexpression of SmMYB9b enhances tanshinone concentration
in Salvia miltiorrhiza hairy roots. Plant Cell Rep 36:1297–1309
Zhang Y, Xu Z, Ji A, Luo H, Song J (2018) Genomic survey of bZIP
transcription factor genes related to tanshinone biosynthesis in
Salvia miltiorrhiza. Acta Pharm Sin B 8:295–305
Zheng J, Lang Y, Zhang Q etal (2015) Structure of human MDM2
complexed with RPL11 reveals the molecular basis of p53 activa-
tion. Gene Dev 29:1524–1534
Zhou Y, Sun W, Chen J etal (2016) SmMYC2a and SmMYC2b played
similar but irreplaceable roles in regulating the biosynthesis of
tanshinones and phenolic acids in Salvia miltiorrhiza. Sci Rep
6:22852
Zhu D, Wu Z, Cao G etal (2014) Translucent green, an ERF family
transcription factor, controls water balance in Arabidopsis by acti-
vating the expression of aquaporin genes. Mol Plant 7:601–615
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.