Appl Microbiol Biotechnol (2014) 98:695704

DOI 10.1007/s00253-013-4929-3

BIOTECHNOLOGICALLY RELEVANT ENZYMES AND PROTEINS

High-efficiency production of bioactive recombinant human

fibroblast growth factor 18 in Escherichia coli and its effects
on hair follicle growth
Lintao Song & Zhifeng Huang & Yu Chen & Haiyan Li &
Chao Jiang & Xiaokun Li

Received: 10 March 2013 / Revised: 10 April 2013 / Accepted: 13 April 2013 / Published online: 30 April 2013
# Springer-Verlag Berlin Heidelberg 2013

Abstract Using fusion tags, expression of recombinant human fibroblast growth factor 18 (rhFGF18) in mammalian cells
and Escherichia coli has been extensively used for fundamental
research and clinical applications, including chondrogenesis
and osteogenesis, hair growth, and neuroprotection. However,
high-level rhFGF18 expression is difficult and the products are
often not homogeneous. Furthermore, fusion-tagged protein
has higher immunogenicity and lower bioactivity, and the
removal of the fused tag is expensive. To overcome the limitations of fusion-tagged expression of protein and to prepare
soluble highly bioactive rhFGF18, we have developed a rapid
and efficient expression strategy. Optimized hFGF18 gene was
amplified by polymerase chain reaction and cloned into
pET22b and pET3c vectors, then transformed into E. coli
strains Origima (DE3) and BL21 (DE3)PlysS. The best combination of plasmid and host strain was selected, and only
Origima (DE3)/pET3c-rhFGF18 was screened for high-level
expressed rhFGF18. Under optimal conditions in a 30-L fermentor, the average bacterial yield and expression level of
rhFGF18 of three batches were more than 652 g and 30 %

Electronic supplementary material The online version of this article

(doi:10.1007/s00253-013-4929-3) contains supplementary material,
which is available to authorized users.
L. Song : Y. Chen : H. Li : C. Jiang (*) : X. Li (*)
Engineering Research Center of Bioreactor and Pharmaceutical
Development, Ministry of Education, Jilin Agricultural University,
Changchun 130118, China
e-mail: chaojiang10@hotmail.com
e-mail: proflxk@163.com
L. Song : Z. Huang : C. Jiang : X. Li
School of Pharmaceutical Science, Key Laboratory
of Biotechnology and Pharmaceutical Engineering
of Zhejiang Province, Wenzhou Medical College,
Wenzhou, Zhejiang 325035, China

respectively, after treatment with 1 mM isopropyl-thio-galactopyranoside for 10 h at 25 C. The target protein was
purified by CM Sepharose FF and heparin affinity chromatography. The purity of rhFGF18 was shown by HPLC to be
higher than 95 %, and the yield was 155 mg/L. In vitro MTT
assays demonstrated that the purified rhFGF18 could stimulate
significant proliferation of NIH3T3 cells, and animal experiments showed that rhFGF18 could effectively regulate hair
growth. In conclusion, this may be a better method of producing rhFGF18 to meet the increasing demand in its pharmacological application.
Keywords Human fibroblast growth factor 18 . Non-fusion
expression . Purification . Mitogenic activity . Hair follicle
growth

Introduction
Fibroblast growth factor 18 (FGF18) belongs to the heparinbinding growth factor family, and it was first identified in
1998 (Ohbayashi et al. 1998). Full-length human FGF18
consists of 207 amino acids containing two potential N-linked
glycosylation domains (Katoh 2005) and is structurally most
homologous to FGF8 and FGF17 among the FGF family
(Hu et al. 1998; Ohbayashi et al. 1998). FGF18 is expressed
predominantly in the adult lungs and kidneys (Ohbayashi et
al. 1998), as well as in several discrete regions during embryonic development (Dichmann et al. 2003; Ohuchi et al. 2000;
Sato et al. 2004; Usui et al. 2004). It plays critical roles in
multiple physiological functions, including skeletal growth
and development (Liu et al. 2002; Long and Ornitz 2013),
hair growth and skin maintenance (Greco et al. 2009; Kawano
et al. 2005; Kimura-Ueki et al. 2012; Plikus 2012), cortical
neuron activity (Hasegawa et al. 2004), and morphogenesis

696

and angiogenesis (Ishibe et al. 2005). It has been tested both

in experimental models of alopecia (Kimura-Ueki et al. 2012)
and clinically in patients for the treatment of osteoarthritis
(Beyer and Schett 2010; Vincent 2012). Intraperitoneal
administration of recombinant FGF18 induces significant
weight gain of the liver and intestine in mice (Hu et al. 1998).
Like other FGFs (Arakawa et al. 1989; Ghosh et al. 1996;
Xu et al. 1999; Zhan et al. 1992), FGF18 stimulated proliferation of NIH3T3 cells in a heparin sulfate-dependent
manner (Hu et al. 1998). In addition, FGF18 has been
extensively investigated as a factor of bone growth and
development (endochondral ossification) via the regulation of chondrogenesis and osteogenesis (Behr et al. 2011;
Ellsworth et al. 2002; Liu et al. 2002; Moore et al. 2005).
FGF18 also contributes to the control of mature chondrocytes
and their progenitors and is an important regulator of limb
development (Behr et al. 2011). Notably, Kimura-Ueki et al.
(2012) reported that FGF18 might be clinically applicable as a
potent therapeutic agent for alopecia.
It is well accepted that FGF18 has an important role in
the physiological regulation of bone development and the
hair cycle and has pharmacological significance in the pathogenesis of human disease. There is a remarkable demand in
the market for production of this protein. However, therapeutic applications of hFGF18 have been restricted by its
low level of expression and difficulty in producing highly
bioactive protein. Given these limiting factors, high priority
should be given to the development of strategies that could
enable significant preparation of sufficient, highly bioactive
hFGF18 for further mechanisms and pharmacological research. Hu et al. (1999) expressed FLAG-tagged FGF18 in
mammalian cells; however, several mammalian cell expression systems are more complicated and costly, making this
strategy difficult to use in the development of large-scale
fermentation processes. Moreover, the preparation of sufficient amounts of virus for large-scale expression is time
consuming. Recently, Jeon et al. (2012) expressed His6tagged FGF18 in Escherichia coli TOP 10 cells. However,
higher levels of expression were not achieved, and the
resulting hexahistidine-tagged FGF18 had higher immunogenicity and lower bioactivity, and it has more difficulty to
be an effective therapeutic drug. It has been previously
reported that therapeutic peptides/proteins, which express
non-fusion genes, exhibit clinical properties superior to
those of their corresponding tagged protein molecules
(Huang et al. 2008), and addition of fusion tags to the
recombinant protein is a key factor that affects the formation
of inclusion bodies (Zhu et al. 2013). Therefore, using the
structural properties of the target protein, such as disulfide
bond formation and heparin-binding ability, could be an intelligent choice for expression and purification. Accumulating
reports have shown the successful expression and purification
of bioactive human proteins such as FGF1 (Wu et al. 2005),

Appl Microbiol Biotechnol (2014) 98:695704

FGFR (Ryu et al. 2006), and KGF2 (Wu et al. 2009) utilizing
this strategy.
With the development of biotechnology, a variety of
alternative expression systems are now being used for expressing recombinant proteins for industrial production, as
well as in research for structural and biochemical studies
(Dong et al. 2008). However, the E. coli expression system
is the most frequently practice used for high-level production of recombinant protein (Ajikumar et al. 2010; Derynck
et al. 1984; Quick and Wright 2002; Verdon et al. 2012). It is
low cost and has high transformation efficiency, quick
growth, and suitability for large-scale manufacture, which
contribute to the selection of E. coli as an expression host
(Jana and Deb 2005).
Since hFGF18 is an important growth factor and its nonfusion expression in E. coli has not been reported, we
constructed an optimum recombinant human FGF18 expression vector (pET3c-rhFGF18), and used E. coli strain
Origima (DE3) cells to successfully express high levels of
rhFGF18. Because it could eliminate both active thioredoxin
and reductase expression, it may thus facilitate cytoplasmic
disulfide bond formation necessary for the correct folding of
the protein (Prinz et al. 1997). Furthermore, we undertook
purification of rhFGF18 protein using a CM Sepharose FF
and heparin affinity chromatography and obtained high
purity of the homogenous rhFGF18 protein. Importantly,
the produced rhFGF18 was found to have a significant
mitogenic effect on NIH3T3 cells and influenced hair follicle growth in vivo. This is the first report regarding the
expression and purification of non-tagged active rhFGF18
in E. coli and the resultant expressed recombinant protein
with relatively high activity for fundamental research and
therapeutic applications.

Materials and methods

Reagents and bacterial strain
TaKaRa Ex Taq, polymerase chain reaction (PCR) purification kit, gel extraction kit, bicinchoninic acid (BCA) kit, and
plasmid miniprep kit were purchased from TaKaRa Company
(Dalian, China). Restriction enzymes NdeI and BamHI were
purchased from NEB (Ipswich, MA, USA). Isopropyl--Dthiogalactoside (IPTG) and methylthiazol tetrazolium (MTT)
was obtained from Bio-Tech (Gold BioTechnology, St. Louis,
MO). Dulbeccos modified Eagle medium (DMEM) was purchased from Invitrogen (Carlsbad, CA). CM Sepharose FF,
Heparin-Sepharose column, and AKTA purifier were purchased from GE Healthcare (Piscataway, NJ, USA). All
primers were synthesized by Invitrogen (Shanghai, China).
Rabbit anti-FGF18 polyclonal antibody was purchased from
Abcam, Inc. (Abcam, Cambridge, MA, USA). The expression

Appl Microbiol Biotechnol (2014) 98:695704

vectors pET22b and pET3c, E. coli DH5, and Origami

(DE3) and BL21 (DE3)PlysS were obtained from the laboratory (Engineering Research Center of Bioreactor and
Pharmaceutical Development, Ministry of Education, Jilin
Agricultural University, China).
Construction of rhFGF18 expression vector
The coding sequence of hFGF18 (the amino acid sequence
of human FGF18, GenBank accession number NP003853)
was amplified from a T vector containing the sequence of
human FGF18 (GenBank accession number KC778400)
optimized according to the codon usage table of E. coli
(http://www.kazusa.or.jp/codon/) by PCR with the appropriate primers to recognize human FGF18, as follows: forward,
5-GGA ATT CCA TAT GTA CAG CGC ACC TAG CGC
ATG-3; reverse, 5-CGC GGA TCC TTA GGC CGG GTG
TGT CGG-3. The forward primer and reverse primers
contained NdeI and BamHI sites, respectively. PCR was
conducted with 50 L of reaction mixture containing
0.25 L TaKaRa Ex Taq (5 U/L), 5 L 10 Ex Taq buffer
(Mg2+ Plus), 4 L dNTPs (each 2.5 mM), and 1 L each of
the forward and reverse primers (each 20 M). The thermocycling parameters used for PCR were as follows: 0.5 min at
94 C for denaturation, 0.5 min at 55 C for annealing, and
1 min at 72 C for extension. After 30 cycles, the amplified
PCR products was digested with NdeI and BamHI and then
ligated into the previously digested pET22b and pET3c expression vectors to create the pET22b-rhFGF18 and pET3crhFGF18 constructs. The constructs were transformed into E.
coli DH5. The accurate insertion of the gene into the plasmid
was confirmed by automated DNA sequencing. The expression vectors (pET22b-rhFGF18 and pET3c-rhFGF18) were
further assessed by restriction enzymatic analysis and
transformed into competent cells of E. coli strains Origima
(DE3) and BL21 (DE3)PlysS.
Production and soluble screening of rhFGF18
For promoting the best expression of rhFGF18, the recombinant E. coli Origima (DE3) or BL21 (DE3)PlysS harboring
the accurate sequence of hFGF18 was shaker-cultured at
37 C and 180 rpm in 5 mL Luria-Bertani (LB) medium
containing 2 % glucose and 100 g/mL ampicillin until the
cell density reached an OD600 of 0.6. The cells were initiated
with a final concentration of 1 mM IPTG as inducer and
incubated at 37 C for 4 h with shaking at 180 rpm. The
expression of each culture was analyzed by Coomassie blue
staining of 12 % (v/v) sodium dodecyl sulfate polyacrylamide
gel electrophoresis (SDS-PAGE), and the expression level of
rhFGF18 was determined by densitometry. The highest expression of rhFGF18 transformant was used in subsequent
experiments.

697

Thereafter, we analyzed IPTG concentration, temperature, and time after induction factors for rhFGF18 expression yield and production of soluble rhFGF18 in Origima
(DE3)/pET3c-rhFGF18. The bacteria were harvested by
centrifugation at 5,000g for 10 min at 4 C, and the cell
pellet was then resuspended in 20 mM TrisHCl (pH 6.5)
buffer containing 1 mM ethylene diamine tetraacetic acid,
1 mM phenylmethylsulfonyl fluoride, 0.05 % Tween80, and
5 mM -mercaptoethanol. Then, the cells were lysed by
sonication for 10 min in an ice bath. The cells lysates were
prepared by centrifugation for 10 min at 12,000g at 4 C,
and the supernatant was transferred to a fresh tube.
Purification and identification of rhFGF18
According to the isoelectric point of target protein, CM
Sepharose FF was chosen for the purification of rhFGF18.
First, the CM cation exchange column was equilibrated with
ten bed volumes of binding buffer (20 mM PB, pH 6.5,
25 mM NaCl) at a rate of 1 mL/min. Subsequently, the soluble
cell extract was applied to the column. After binding, the
column was further washing with binding buffer, and then
eluted with binding buffer containing different concentrations
of NaCl. Then, further purification was performed with a
heparin-sepharose column pre-equilibrated with binding buffer until the OD280 of the effluent reached base line. Finally,
rhFGF18 protein was collected from the column with stepwise
gradients of 0.3, 0.8, and 1.5 M NaCl. The elution fractions
were collected and determined by Coomassie blue staining of
12 % (v/v) SDS-PAGE. Fractions containing rhFGF18 were
concentrated with a Millipore filter, and the concentration was
evaluated with a BCA kit. Importantly, all purification procedures were carried out at 4 C and purified rhFGF18 fractions were stored at 80 C to retain biological activity.
The purity of rhFGF18 was further examined by HPLC
analysis, and the sample loaded onto a C18 column. The
HPLC was operated on Agilent 1260 Infinity equipped with
C18 column from Agilent. The elution was conducted using
a linear gradient of 090 % acetonitrile at a flow rate of
0.5 mL/min in the presence of 0.1 % trifluoroacetic acid.
The absorbance was measured continuously at 280 nm.
Western blotting was performed using a peroxidaseconjugated rabbit polyclonal anti-human FGF18 antibody
for further identification according to the manufacturers
protocol. The molecular sizes of the obtained protein were
verified by comparison with the migration of pre-stained
protein markers electrophoresed in parallel lanes.
Mitogenic activity of rhFGF18 assay
We used the NIH 3T3 cell line (American Type Culture
Collection, Rockville, MD) to determine the activity of the
rhFGF18 that was expressed in E. coli Origima (DE3). First,

698

the cells were grown in DMEM containing 10 % fetal

bovine serum (FBS), 100 U/mL ampicillin, and 100 U/mL
streptomycin until the culture reached the mid-logarithmic
time, then transferred to a 96-well plate (5103/well), and
incubated at 37 C for 24 h. The medium was then replaced
with DMEM supplemented with 0.4 % FBS, and the cells
were serum starved for 24 h. Next, the cells were treated
with different concentrations of rhFGF18 and human FGF1
protein for 48 h, and the number of viable cells was determined by adding 20 L MTT (5 mg/mL) per well for 4 h.
Finally, the medium was discarded and 150 L of DMSO
was added to each well. After incubation for 30 min at room
temperature, the absorbance was immediately measured at
600 nm using a microplate reader.
In vivo analysis of rhFGF18 activity
To characterize the effects of the purified rhFGF18 on hair
growth in vivo, we essentially followed the protocol developed
by Kawano et al. (2005). All procedures involving 56-day-old
C57BL/6 male mice (1822 g), and their cares were approved
by the Institutional Animals Care and Use Committee at Jilin
Agricultural University, China, and performed in accordance
with institutional guidelines for animal experiments. After
conditioning for 2 days, 16 mice (C57BL/6) were then anesthetized, and their dorsal hair was gently cut short with a
trimmer and photographed using a digital camera. Following
this, the mice were randomly divided into the treatment group
(FGF18, n=8) and negative control group (phosphate-buffered
saline (PBS), n=8). Then, eight mice were injected subcutaneously with rhFGF18 solution (1 mg/mL) into their backs at
a dose of 1 mg/kg body weight, while the remaining eight
received PBS at a corresponding volume. Thereafter, rhFGF18
and PBS were administered once daily for 14 days. The mice
were then maintained for selected numbers of days on a
standard laboratory diet and acidified water ad libitum. On
day 15, they were anesthetized, photographed, and killed,
and the full-thickness of the dorsal skin in the test area was
excised. The harvested samples were fixed and embedded in paraffin, and then, the embedded skin samples
were cut into 4 m-thick sections using standard procedures
(Paus et al. 1999). Finally, the sections were stained with
hematoxylin and eosin (H&E) and observed under a
microscope.

Results
Construction of hFGF18 expression vectors
To produce recombinant hFGF18 protein, expression vectors
bearing the optimized hFGF18 gene were constructed. The

Appl Microbiol Biotechnol (2014) 98:695704

construction strategy and detailed procedure are described in

Materials and methods. The final product (full-length
hFGF18) was obtained by PCR (Fig. 1a), then digested with
NdeI and BamHI, and cloned into the expression vectors
(pET22b and pET3c) to create the recombinant plasmids
pET22b-rhFGF18 and pET3c-rhFGF18, which were confirmed by automated DNA sequencing and restriction enzymatic analysis (Fig. 1b). The results of the sequences of
hFGF18 (643 bp) were conformed to the desired sequence.
Expression screening and optimization of soluble rhFGF18
production
To elucidate the optimal production method of rhFGF18, we
utilized Origima (DE3) and BL21 (DE3)PlysS bacteria and
two kinds of recombinant plasmid for expression screening.
Here, we performed several small-scale expression experiments. The results showed that, except for the combination of
Origima (DE3)/pET3c-rhFGF18, all other combinations
resulted in target protein being greatly lower expressed
(Fig. 2).
The combination of Origima (DE3)/pET3c-rhFGF18 was
used for optimization of induction conditions for improved
production of soluble rhFGF18. The recombinant was shakercultured at 37 C and 180 rpm in LB medium until the culture
reached mid-logarithmic growth (OD600 =0.6). Next, E. coli
cells harboring hFGF18 were treated with 1 mM IPTG for
10 h at 25 C and then collected and lysed. The average
bacterial yield of three batches was more than 652 g in a 30L fermentor. SDS-PAGE analysis of the lysate supernatant
showed that the recombinant protein was expressed as a
soluble product with a molecular weight of about 23 kDa
corresponding to the target protein of rhFGF18 (Fig. 2b).
Densitometry scanning revealed that the amount of expressed
recombinant hFGF18 was more than 30 % of the total cellular
protein at 10 h after induction. Simultaneously, we found that
the rhFGF18 protein was expressed in inclusion bodies at
37 C (data not shown).
Purification and identification of rhFGF18
Next, the soluble product was purified as described in
Materials and methods. As demonstrated in Fig. 3a, the
fractions containing rhFGF18 were finally eluted by heparinaffinity chromatography using 20 mM PB containing 0.8 M
NaCl. The purified rhFGF18 protein yield was 155 mg/L. And
the recovered rhFGF18 was homogenous, and its purity was
over 90 %, as estimated by HPLC, with a retention time of
14.959 min (Fig. 4). Western blotting analysis with a specific
anti-human FGF18 antibody showed a specific reaction with
the target protein, and revealed part of the rhFGF18 protein
was degraded (Fig. 3b).

Appl Microbiol Biotechnol (2014) 98:695704

699

Fig. 1 Synthesis of FGF18 by PCR and identification by restriction

enzymatic analysis. The strategy for synthesizing FGF18 is described
in Material and methods. The molecular weight of the PCR fragment
containing FGF18 is shown in a. Lane 1 negative control (without
template vector), lane 2 FGF18 fragment (643bp). Identification of

recombinant plasmid by enzyme digestion (NdeI and BamHI) is shown

in b. Lanes 1 and 3 pET3c-rhFGF18 and pET22b-rhFGF18, respectively; lanes 2 and 4 restriction products of recombinant plasmid
pET3c-rhFGF18 and pET22b-rhFGF18, respectively

Mitogenic activity of recombinant hFGF18

To evaluate the effectiveness of recombinant hFGF18 bioactivity, the proliferative effect of rhFGF18 was determined
using a standard MTT assay on NIH3T3 cells and compared
with human FGF1 (positive control, Abcam, HongKong) because it has been well confirmed that FGF1 can strongly active
DNA synthesis in NIH3T3 cells (Gospodarowicz 1974). As

shown in Fig. 5, similar to FGF1 the purified rhFGF18 induced

a comparable mitogenic response in NIH3T3 cells, which is
consistent with the findings of previous studies (Hu et al.
1999). Furthermore, rhFGF18 was found to have a dosedependent effect on the viability of NIH3T3 cells, and the
negative control did not have any affirmative effect (Fig. 5).
Thus, we speculate that rhFGF18 produced with this method
had a remarkable mitogenic effect on NIH3T3 cells.

Fig. 2 SDS-PAGE analysis of rhFGF18 expression screening and optimization of induction conditions for production of soluble rhFGF18. Two
kinds of recombinant plasmid were transformed into E. coli strain
Origima (DE3) and BL21 (DE3)PlysS for expression screening. The
results showed that the Origima (DE3)/pET3c-rhFGF18 transformant
was effective for expression of target protein. a Lane M molecular weight
standards; lanes 1, 3, and 5 uninduced BL21 (DE3)PlysS/pET22b-

rhFGF18, BL21 (DE3)PlysS/pET3c-rhFGF18, and Origima (DE3)/

pET22b-rhFGF18, respectively; lanes 2, 4, and 6 induced BL21
(DE3)PlysS/pET22b-rhFGF18, BL21 (DE3)PlysS/pET3c-rhFGF18,
and Origima (DE3)/pET22b-rhFGF18, respectively. b Lane M molecular
weight standards, lanes 1 and 2 uninduced and induced Origima (DE3)/
pET3c-rhFGF18. The bacteria containing Origima (DE3)/pET3crhFGF18 were induced by 1 mM IPTG for 10 h at 25 C

700

Fig. 3 SDS-PAGE analysis of purification of rhFGF18 and its characterization by Western blotting. a SDS-PAGE analysis of the fraction
collected from CM Sepharose FF (lane 1) and heparin-affinity chromatography (lanes 24), respectively. Lane 1 fraction eluted with
0.6 M NaCl from CM Sepharose FF, lane 2 0.3 M NaCl-eluted from
heparin-sepharose chromatography, lane 3 0.8 M NaCl-eluted fraction,
lane 4 1.5 M NaCl-eluted fraction, lane M molecular weight standards.
b Western blotting analysis of the rhFGF18. Lane 1 bacterial lysate of
transformants uninduced by IPTG as negative control, lane 2 the
bacterial lysate of transformants induced by 1 mM IPTG

In vivo bioactivity of rhFGF18

To evaluate further the biological activity of rhFGF18, the
effect of the rhFGF18 on hair growth was analyzed by
subcutaneous administration of male C57BL/6 mice. In
our experiments, 14 days after treatment with rhFGF18
protein, the exterior surface of the dorsal skin of the mice

Fig. 4 HPLC analysis of the

purified rhFGF18. The purity of
rhFGF18 eluted from a heparinaffinity column was further
evaluated by HPLC analysis
using a C18 column. As seen
from the chromatogram, the
y-axis indicates the absorbance
(280 nm, mAu), while the
x-axis elution time (minutes).
The main peak was observed at
14.959 min. The purity of
purified rhFGF18 was more
than 90 %

Appl Microbiol Biotechnol (2014) 98:695704

Fig. 5 Mitogenic activity analysis of different concentrations of

rhFGF18 and FGF1 on NIH3T3 cells in vitro. Cell cultures were
prepared as in Materials and methods. Increasing concentrations of
FGF18 (solid circle) or FGF1 (solid diamond) were added. PBS was
used as a negative control (solid triangles). Proliferation was quantified
by measuring the absorbance at 600 nm

exhibited vigorous hair growth throughout the entire test

area (Fig. 6a). Examination of the reverse side of the skin
revealed extensive growth of anagen hair follicles in the
mice with pigmentation or hair growth (black spots indicate
anagen hair follicles) (Fig. 6a). Furthermore, the skin from
one of the eight mice that received FGF18 showed no apparent changes, whereas none of the skin samples from the eight
control mice exhibited hair growth or strong pigmentation.
Examination of the reverse side of control skin revealed little

Appl Microbiol Biotechnol (2014) 98:695704

701

Fig. 6 Induction of hair growth by rhFGF18 protein in mice with

telogen stage hair follicles. Sixteen 56-day-old male C57BL/6 mice
were anesthetized and their dorsal hair was gently cut short, and
rhFGF18 was administered subcutaneously into the dorsal test area
(n=16). The squared area is the test area. a After 14 days, the mice that
received rhFGF18 showed vigorous hair growth. b As a control,
another group of mice received PBS. Representative mice for each

treatment are shown. On day 15, they were anesthetized and killed, and
the full-thickness of the dorsal skin in the test area was excised. A
representative result is shown. c Histology of the hair follicles from
mice administered with FGF18. d Histology of the hair follicles from
mice administered with PBS. The skin from the mice in a and b was
embedded in paraffin, sectioned, stained with H&E, and photographed.
Scale bars=50 m

or no anagen hair follicles, as is indicated by their white color

(Fig. 6b). Meanwhile, these results were further confirmed by
histological examination of skin sections (Fig. 6c, d). When
rhFGF18 was administered subcutaneously to mice in a uniform telogen state, anagen hair growth was compared between
the FGF18- and PBS-administered mice. Our findings suggest
that the rhFGF18 expressed using this method is important for
the regulation of hair growth and the maintenance of skin in
adult mice and may be an effective approach to stimulate hair
growth.

of FGF18 with high bioactivity for specific treatment of

FGF18-related diseases in the future. However, the low level
of expression and difficulty in producing highly bioactive
homogeneous protein has restricted its therapeutic applications. A widely used method to evade these limitations is the
fusion strategy expression of target protein, such as His6tagged, glutathione-S-transferase and thioredoxin, but these
methods have shortcomings in terms of efficient soluble
expression, bioactivity, and immunogenicity (Terpe 2003).
Specifically, to achieve release of target protein, the cleavage
reaction must be performed. Most importantly, the removal
of the fused tag is expensive and may impact on the bioactivity of the target protein. This may restrict development of
any therapeutic proteins including rhFGF18 because higher
biological activity, lower expenditure, and manufacturing
reproducibility are essential for regulatory approval. We
have reported a rapid and efficient strategy to express and
purify high levels of recombinant hFGF18.
In our experiments, we used the most commonly used E.
coli expression system for high-level expression of heterologous proteins due to its well-established, fast growth rate,
resulting high yields of protein expression. However, the
structural features of hFGF18 pose quite a challenge for its

Discussion
A wealth of pharmacological studies has previously shown
that FGF18, a member of the FGF family, has a high therapeutic potential for cartilage repair in preclinical and clinical
cartilage disorders (Moore et al. 2005). In addition, several in
vivo studies using different animal models have shown that
recombinant human FGF18 could regulate hair growth and
skin maintenance and have possibilities for alopecia therapies in humans (Greco et al. 2009; Kimura-Ueki et al. 2012).
It is necessary to develop methods for abundant production

702

production by genetic engineering in E. coli. The difficulties

primarily stem from the activity-dependent conformation
and potential toxicity of this peptide. To eliminate the adverse effect of expression levels of rhFGF18, preferential
codons of E. coli were introduced into the coding sequence
of rhFGF18(ESM Fig. S1). Origima (DE3) was chosen as
the expression host strain because it could eliminate both
active thioredoxin and reductase expression and thus may
facilitate cytoplasmic disulfide bond formation necessary
for the correct folding of the protein (Prinz et al. 1997).
Taken together, for efficient protein expression both were
indispensable. To further improve the production level of
the target protein, we successively optimized several pivotal
factors of expression conditions (IPTG concentration, temperature, and time after induction) (ESM Fig. S2). It has
been previously reported that stress conditions and high
rates of protein synthesis are the main causes of the formation of inclusion bodies (Gatti-Lafranconi et al. 2011).
Under optimal expression conditions, the high soluble expression level of rhFGF18 reached more than 30 % of total
protein after treatment with 1 mM IPTG for 10 h at 25 C in
Origima (DE3) host strain. Actually, the use of a lower
temperature (25 C) to increase soluble protein levels and
reduce the aggregation of recombinant proteins, thus reducing the inclusion bodies formed, is quite prevalent (de Groot
and Ventura 2006). In addition, a long induction period is
also necessary.
Because of the fact that the heparin binging ability of
hFGF18 could be specifically captured by heparin-sepharose,
the non-fusion rhFGF18 protein was efficiently separated from undesirable impurities of E. coli lysate supernatant by
a heparin-affinity chromatograph. In addition, heparinsepharose chromatography has been routinely used for the
separation and purification of some therapeutic proteins
(Berman et al. 1999; Huang et al. 2012; Kenig et al. 2008).
Therefore, heparin-sepharose may provide an environment
favorable for maintaining hFGF18 bioactivity during purification. Using the heparin-sepharose strategy, we successfully
purified the rhFGF18 and produced highly bioactive homogeneous rhFGF18. We utilized its heparin binging ability to
generate a non-fusion expression vector to produce rhFGF18
that is able to retain more bioactivity because of avoiding
cleavage of the fusion-tagged protein. Unfortunately, the
rhFGF18 was very sensitive to the environmental temperature
and degraded in the short term (Fig. 3b). Like other FGFs,
sensitivity to temperature and proteolytic enzymes significantly
impedes its clinical applications. Fortunately, poly(ethylene
glycol) (PEG) has been extensively used to improve protein
biostabilities and to increase the half-life time of protein in vivo
(Caliceti and Veronese 2003; Veronese and Pasut 2005).
Therefore, our research placed emphasis on the PEG modification of rhFGF18 to increase its in vivo bio-stability and
therapeutic potency.

Appl Microbiol Biotechnol (2014) 98:695704

Certainly, the purified recombinant human FGF18 protein was biologically active in vitro. Like FGF1, FGF18 had
a dose-dependent effect on the proliferation of NIH3T3
cells, demonstrating that the rhFGF18 expressed from E.
coli Origima (DE3) is more favorable for retaining heparinbinding ability leaving an intact functional complex comprising FGF, FGFRs, and heparin. Thus, we speculate that FGF18
has a mitogenic effect on NIH3T3 cells, which is consistent
with the finding of previous studies (Hu et al. 1999).
Interestingly, rhFGF18 produced by the non-fusion method
had better biological activity than FLAG-tagged hFGF18 (Hu
et al. 1999). We speculate that the FLAG tag may lead to
lower bioactivity. Meanwhile, its specific activity on NIH3T3
cells was of a lesser magnitude than that of FGF1. This result
suggests that FGF18 may be less potent than FGF1 in fibroblasts. A possible explanation of this effect could be that
fibroblasts may not be the primary physiological target cell
type for FGF18. Interestingly, Shimoaka et al. (2002)
suggested that the mitogenic activity of FGF18 is more specific to bone and cartilage cells than FGF1 and FGF2.
Furthermore, FGF18 has been previously reported to activate
the proliferation of different cell types, such as human dermal
papilla cells, epidermal keratinocytes, vascular endothelial
cells, hepatic parenchymal cells, and pancreatic ductal epithelium (Hu et al. 1999; Kawano et al. 2005; Ohbayashi et al.
2002; Smith et al. 2007).
A variety of polypeptide growth factors, including various
members of the fibroblast growth factor family, are expressed
in skin. As reported previously, FGF18 mRNA is strongly
expressed in telogen hair follicles (Kawano et al. 2005).
When male C57BL/6 mice were subcutaneously administered
with FGF18 during telogen, hair growth occurred in the FGF18
treatment group earlier than the PBS control group. Reportedly,
FGF18 was delivered subcutaneously during anagen, and matrix cell proliferation was immediately inhibited, leading to
strong suppression of hair follicle growth (Kimura-Ueki et al.
2012). rhFGF18 protein has contrasting pharmacological effects on hair growth in mice, depending on the hair cycle stage
of the follicles. Furthermore, the anagen-promoting effects of
FGF7 and FGF10 are very different to the activity of FGF18
(Greco et al. 2009). It is likely that FGF7 and FGF10 activate
FGFR2b, an epithelial-specific FGF receptor subtype; in contrast, FGF18 activates FGFR3c and FGFR4 (Zhang et al.
2006). While signaling through FGFR2b results in epithelial
cell proliferation, signaling through FGFR3c and FGFR4 induces a complex phenotype that is sometimes inhibitory (Iwata
et al. 2000; Ornitz and Marie 2002). This probably explains
why the biological effects of FGF18 and FGF7/FGF10 differ
significantly. In our studies, we also observed significant hair
follicle growth after subcutaneously injecting rhFGF18 for
14 days during the telogen stages of the hair growth cycle,
compared to the negative control (PBS). However, KimuraUeki et al. (2012) reported that FGF18 might be performed the

Appl Microbiol Biotechnol (2014) 98:695704

inhibition of rhFGF18 on anagen progression by hair follicle.

The only certainty is that the effects of rhFGF18 on hair follicle
growth could be applied in further research and therapeutic
applications.
In summary, recombinant human FGF18 was successfully
expressed in the E. coli strain Origima (DE3) from optimized
non-fusion FGF18 gene. Purification was undertaken using
CM Sepharose FF and heparin affinity chromatography, the
identity of the purified protein was confirmed by Western
blotting, and its purity was determined by HPLC analysis.
The rhFGF18 could significantly promote proliferation of
NIH3T3 cells. In vivo, subcutaneous administration with
rhFGF18 during telogen of the hair growth cycle could remarkably regulate hair growth in C57BL/6 mice. This study
demonstrated that non-tagged expression of optimal rhFGF18
is simple, viable, and highly effective, making it convenient
for high level expression and purification of protein with high
bioactivity preserved.
Acknowledgments The project was supported by grants from the
National Natural Science Foundation of China (81102486), the
National 863 High Technology Research and Development Program (2011AA02A113), and National Natural Science Foundation
of Zhejiang Province of China (R2090550).

References
Ajikumar PK, Xiao WH, Tyo KE, Wang Y, Simeon F, Leonard E, Mucha
O, Phon TH, Pfeifer B, Stephanopoulos G (2010) Isoprenoid pathway optimization for Taxol precursor overproduction in Escherichia
coli. Science 330:7074. doi:10.1126/science.1191652330/6000/70
Arakawa T, Hsu YR, Schiffer SG, Tsai LB, Curless C, Fox GM (1989)
Characterization of a cysteine-free analog of recombinant human
basic fibroblast growth factor. Biochem Biophys Res Commun
161(1):335341
Behr B, Sorkin M, Manu A, Lehnhardt M, Longaker MT, Quarto N (2011)
Fgf-18 is required for osteogenesis but not angiogenesis during long
bone repair. Tissue Eng Part A 17:20612069. doi:10.1089/
ten.TEA.2010.0719
Berman B, Ostrovsky O, Shlissel M, Lang T, Regan D, Vlodavsky I,
Ishai-Michaeli R, Ron D (1999) Similarities and differences between the effects of heparin and glypican-1 on the bioactivity of
acidic fibroblast growth factor and the keratinocyte growth factor.
J Biol Chem 274:3613236138
Beyer C, Schett G (2010) Pharmacotherapy: concepts of pathogenesis
and emerging treatments. Novel targets in bone and cartilage. Best
Pract Res Clin Rheumatol 24:489496. doi:10.1016/j.berh.
2010.03.001S1521-6942(10)00028-8
Caliceti P, Veronese FM (2003) Pharmacokinetic and biodistribution
properties of poly(ethylene glycol)-protein conjugates. Adv Drug
Deliv Rev 55:12611277
de Groot NS, Ventura S (2006) Effect of temperature on protein quality
in bacterial inclusion bodies. FEBS Lett 580:64716476
Derynck R, Roberts AB, Winkler ME, Chen EY, Goeddel DV (1984)
Human transforming growth factor-alpha: precursor structure and
expression in E. coli. Cell 38:287297
Dichmann DS, Miller CP, Jensen J, Scott Heller R, Serup P (2003)
Expression and misexpression of members of the FGF and

703
TGFbeta families of growth factors in the developing mouse pancreas. Dev Dyn 226:663674. doi:10.1002/dvdy.10270
Dong X, Tang B, Li J, Xu Q, Fang S, Hua Z (2008) Expression and
purification of intact and functional soybean (Glycine max) seed
ferritin complex in Escherichia coli. J Microbiol Biotechnol
18:299307
Ellsworth JL, Berry J, Bukowski T, Claus J, Feldhaus A, Holderman S,
Holdren MS, Lum KD, Moore EE, Raymond F, Ren H, Shea P,
Sprecher C, Storey H, Thompson DL, Waggie K, Yao L,
Fernandes RJ, Eyre DR, Hughes SD (2002) Fibroblast growth
factor-18 is a trophic factor for mature chondrocytes and their progenitors. Osteoarthr Cartil 10:308320. doi:10.1053/joca.2002.
0514S1063458402905145
Gatti-Lafranconi P, Natalello A, Ami D, Doglia SM, Lotti M (2011)
Concepts and tools to exploit the potential of bacterial inclusion
bodies in protein science and biotechnology. FEBS J 278:2408
2418. doi:10.1111/j.1742-4658.2011.08163.x
Ghosh AK, Shankar DB, Shackleford GM, Wu K, TAng A, Miller GJ,
Zheng J, Roy-Burman P (1996) Molecular cloning and characterization of human FGF8 alternative messenger RNA forms. Cell
Growth Differ 7(10):14251434
Gospodarowicz D (1974) Localisation of a fibroblast growth factor and
its effect alone and with hydrocortisone on 3T3 cell growth.
Nature 249:123127
Greco V, Chen T, Rendl M, Schober M, Pasolli HA, Stokes N, Dela
Cruz-Racelis J, Fuchs E (2009) A two-step mechanism for stem
cell activation during hair regeneration. Cell Stem Cell 4:155
169. doi:10.1016/j.stem.2008.12.009
Hasegawa H, Ashigaki S, Takamatsu M, Suzuki-Migishima R,
Ohbayashi N, Itoh N, Takada S, Tanabe Y (2004) Laminar patterning in the developing neocortex by temporally coordinated
fibroblast growth factor signaling. J Neurosci 24(40):87118719
Hu MC, Qiu WR, Wang YP, Hill D, Ring BD, Scully S, Bolon B,
DeRose M, Luethy R, Simonet WS, Arakawa T, Danilenko DM
(1998) FGF-18, a novel member of the fibroblast growth factor
family, stimulates hepatic and intestinal proliferation. Mol Cell
Biol 18:60636074
Hu MC, Wang YP, Qiu WR (1999) Human fibroblast growth factor-18
stimulates fibroblast cell proliferation and is mapped to chromosome
14p11. Oncogene 18:26352642. doi:10.1038/sj.onc.1202616
Huang Y, Rao Y, Feng C, Li Y, Wu X, Su Z, Xiao J, Xiao Y, Feng W, Li X
(2008) High-level expression and purification of Tat-haFGF19-154.
Appl Microbiol Biotechnol 77:10151022. doi:10.1007/s00253007-1249-5
Huang Z, Ye C, Liu Z, Wang X, Chen H, Liu Y, Tang L, Zhao H, Wang
J, Feng W, Li X (2012) Solid-phase N-terminus PEGylation of
recombinant human fibroblast growth factor 2 on heparinsepharose column. Bioconjug Chem 23:740750. doi:10.1021/
bc200550f
Ishibe T, Nakayama T, Okamoto T, Aoyama T, Nishijo K, Shibata KR,
Shima Y, Nagayama S, Katagiri T, Nakamura Y, Nakamura T,
Toguchida J (2005) Disruption of fibroblast growth factor signal
pathway inhibits the growth of synovial sarcomas: potential application of signal inhibitors to molecular target therapy. Clin
Cancer Res 11:27022712
Iwata T, Chen L, Li C, Ovchinnikov DA, Behringer RR, Francomano CA,
Deng CX (2000) A neonatal lethal mutation in FGFR3 uncouples
proliferation and differentiation of growth plate chondrocytes in
embryos. Hum Mol Genet 9:16031613
Jana S, Deb JK (2005) Strategies for efficient production of heterologous proteins in Escherichia coli. Appl Microbiol Biotechnol
67:289298. doi:10.1007/s00253-004-1814-0
Jeon E, Yun YR, Kang W, Lee S, Koh YH, Kim HW, Suh CK, Jang JH
(2012) Investigating the role of FGF18 in the cultivation and
osteogenic differentiation of mesenchymal stem cells. PLoS One
7:e43982. doi:10.1371/journal.pone.0043982PONE-D-12-00819

704
Katoh M (2005) Comparative genomics on FGF8, FGF17, and FGF18
orthologs. Int J Mol Med 16:493496
Kawano M, Komi-Kuramochi A, Asada M, Suzuki M, Oki J, Jiang J,
Imamura T (2005) Comprehensive analysis of FGF and FGFR
expression in skin: FGF18 is highly expressed in hair follicles and
capable of inducing anagen from telogen stage hair follicles. J Invest
Dermatol 124:877885. doi:10.1111/j.0022-202X.2005.23693.x
Kenig M, Gaberc-Porekar V, Fonda I, Menart V (2008) Identification of
the heparin-binding domain of TNF-alpha and its use for efficient
TNF-alpha purification by heparin-Sepharose affinity chromatography. J Chromatogr B Analyt Technol Biomed Life Sci 867:119125
Kimura-Ueki M, Oda Y, Oki J, Komi-Kuramochi A, Honda E, Asada M,
Suzuki M, Imamura T (2012) Hair cycle resting phase is regulated
by cyclic epithelial FGF18 signaling. J Invest Dermatol 132:1338
1345. doi:10.1038/jid.2011.490
Liu Z, Xu J, Colvin JS, Ornitz DM (2002) Coordination of chondrogenesis
and osteogenesis by fibroblast growth factor 18. Genes Dev 16:859
869. doi:10.1101/gad.965602
Long F, Ornitz DM (2013) Development of the endochondral
skeleton. Cold Spring Harb Perspect Biol 5(1) doi:10.1101/
cshperspect.a008334a008334
Moore EE, Bendele AM, Thompson DL, Littau A, Waggie KS,
Reardon B, Ellsworth JL (2005) Fibroblast growth factor-18
stimulates chondrogenesis and cartilage repair in a rat model of
injury-induced osteoarthritis. Osteoarthr Cartil 13:623631
Ohbayashi N, Hoshikawa M, Kimura S et al (1998) Structure and
expression of the mRNA encoding a novel fibroblast growth
factor, FGF-18. J Biol Chem 273:1816118164
Ohbayashi N, Shibayama M, Kurotaki Y, Imanishi M, Fujimori T, Itoh
N, Takada S (2002) FGF18 is required for normal cell proliferation and differentiation during osteogenesis and chondrogenesis.
Genes Dev 16:870879. doi:10.1101/gad.965702
Ohuchi H, Kimura S, Watamoto M, Itoh N (2000) Involvement of
fibroblast growth factor (FGF)18-FGF8 signaling in specification
of left-right asymmetry and brain and limb development of the
chick embryo. Mech Dev 95:5566
Ornitz DM, Marie PJ (2002) FGF signaling pathways in endochondral
and intramembranous bone development and human genetic disease. Genes Dev 16:14461465. doi:10.1101/gad.990702
Paus R, Muller-Rover S, Van Der Veen C, Maurer M, Eichmuller S,
Ling G, Hofmann U, Foitzik K, Mecklenburg L, Handjiski B
(1999) A comprehensive guide for the recognition and classification of distinct stages of hair follicle morphogenesis. J Invest
Dermatol 113:523532
Plikus MV (2012) New activators and inhibitors in the hair cycle clock:
targeting stem cells state of competence. J Invest Dermatol
132:13211324. doi:10.1038/jid.2012.38jid201238
Prinz WA, Aslund F, Holmgren A, Beckwith J (1997) The role of the
thioredoxin and glutaredoxin pathways in reducing protein disulfide bonds in the Escherichia coli cytoplasm. J Biol Chem
272:1566115667
Quick M, Wright EM (2002) Employing Escherichia coli to functionally
express, purify, and characterize a human transporter. Proc Natl Acad
Sci U S A 99:85978601. doi:10.1073/pnas.132266599132266599

Appl Microbiol Biotechnol (2014) 98:695704

Ryu EK, Cho KJ, Kim JK, Harmer NJ, Blundell TL, Kim KH (2006)
Expression and purification of recombinant human fibroblast
growth factor receptor in Escherichia coli. Protein Expr Purif
49:1522
Sato T, Joyner AL, Nakamura H (2004) How does Fgf signaling from
the isthmic organizer induce midbrain and cerebellum development? Dev Growth Differ 46:487494
Shimoaka T, Ogasawara T, Yonamine A, Chikazu D, Kawano H,
Nakamura K, Itoh N, Kawaguchi H (2002) Regulation of osteoblast,
chondrocyte, and osteoclast functions by fibroblast growth factor
(FGF)-18 in comparison with FGF-2 and FGF-10. J Biol Chem
277:74937500. doi:10.1074/jbc.M108653200M108653200
Smith SM, West LA, Hassell JR (2007) The core protein of growth
plate perlecan binds FGF-18 and alters its mitogenic effect on
chondrocytes. Arch Biochem Biophys 468:244251
Terpe K (2003) Overview of tag protein fusions: from molecular and
biochemical fundamentals to commercial systems. Appl Microbiol
Biotechnol 60:523533
Usui H, Shibayama M, Ohbayashi N, Konishi M, Takada S, Itoh N
(2004) Fgf18 is required for embryonic lung alveolar development. Biochem Biophys Res Commun 322:887892
Verdon J, Girardin N, Marchand A, Hechard Y, Berjeaud JM (2012)
Purification and antibacterial activity of recombinant warnericin
RK expressed in Escherichia coli. Appl Microbiol Biotechnol.
doi:10.1007/s00253-012-4417-1
Veronese FM, Pasut G (2005) PEGylation, successful approach to drug
delivery. Drug Discov Today 10:14511458
Vincent TL (2012) Explaining the fibroblast growth factor paradox in
osteoarthritis: lessons from conditional knockout mice. Arthritis
Rheum 64:38353838. doi:10.1002/art.34648
Wu X, Su Z, Li X, Zheng Q, Huang Y, Yuan H (2005) High-level
expression and purification of a nonmitogenic form of human
acidic fibroblast growth factor in Escherichia coli. Protein Expr
Purif 42:711
Wu X, Tian H, Huang Y, Wu S, Liu X, Wang C, Wang X, Huang Z,
Xiao J, Feng W, Li X (2009) Large-scale production of biologically active human keratinocyte growth factor-2. Appl Microbiol
Biotechnol 82:439444. doi:10.1007/s00253-008-1782-x
Xu J, Lawshe A, MacArthur CA, Ornitz DM (1999) Genomic structure, mapping, activity and expression of fibroblast growth factor
17. Mech Dev 83(12):165178
Zhan X, Hu X, Friedman S, Maciag T (1992) Analysis of endogenous
and exogenous nuclear translocation of fibroblast growth factor-1
in NIH 3T3 cells. Biochem Biophys Res Commun 188(3):982
991
Zhang X, Ibrahimi OA, Olsen SK, Umemori H, Mohammadi M,
Ornitz DM (2006) Receptor specificity of the fibroblast growth
factor family. The complete mammalian FGF family. J Biol Chem
281:1569415700
Zhu S, Gong C, Ren L, Li X, Song D, Zheng G (2013) A simple and
effective strategy for solving the problem of inclusion bodies in
recombinant protein technology: his-tag deletions enhance soluble
expression. Appl Microbiol Biotechnol 97:837845. doi:10.1007/
s00253-012-4630-y